A large body of evidence clearly demonstrates the protective effects of breastfeeding
and documents the transmission of specific infections to infants through breast milk.
The fear and anxiety that arise with the occurrence of any infectious disease are
even greater in the situation of the breastfeeding mother-infant dyad. Uncertainty
and lack of knowledge often lead to proscribing against breastfeeding out of fear,
which then deprives the infant of the potential protective, nutritional, and emotional
benefits of breastfeeding exactly at the time when they are most needed (see the discussion
of immunologic benefits of human milk in Chapter 5). Decisions concerning breastfeeding
in a mother with an infectious illness should balance the potential benefits of breastfeeding
versus the known or estimated risk for the infant acquiring a clinically significant
infection via breastfeeding and the potential severity of the infection.
Documenting transmission of infection from mother to infant by breastfeeding requires
not only the exclusion of other possible mechanisms of transmission but also the demonstration
of the infectious agent in the breast milk and a subsequent clinically significant
infection in an infant that was caused by a plausible infectious process. The first
step is to establish the occurrence of a specific infection (clinically or immunologically
evident) in a mother and demonstrate the persistence of the infectious agent such
that it could be transmitted to the infant. Isolation or identification of the infectious
agent from the colostrum, breast milk, or an infectious lesion of the breast is important
but not necessarily proof of transmission to an infant. Epidemiologic evidence of
transmission must be considered, including identifying characteristics of the organism
that relate an isolate from an infant to the maternal isolate. Infectious organisms
can reach the breast milk either by secretion in the fluid or cellular components
of breast milk or by contamination of the milk at the time of or after expression.
A reasonable mechanism of infection via breast milk should be evident and proved through
either animal or human studies. Demonstration of a subclinical or clinically evident
infection in an infant should follow these outlined steps.
Exclusion of other possible mechanisms of transmission (exposure to mother or other
persons/animals via airborne, droplet, arthropod, or vector modes of transmission
or through direct contact with other infectious fluids) would complete the confirmation
of transmission of infection via breastfeeding. It is essential to exclude prenatal
or perinatal transmission of infection to a fetus/infant, but doing this can often
be difficult.
Clinical case reports or studies confirming the isolation of an infectious agent from
the milk are important. To determine a reasonable estimate of the risk for infection
via breast milk, larger epidemiologic studies are needed that compare infection rates
in breastfed infants versus formula-fed infants, addressing the issues just identified.
Timing of breastfeeding is important relative to the timing of maternal infection
and to the presence of a pathogen in colostrum or breast milk. The duration of breastfeeding
is another important variable to consider in the estimate of risk because shedding
of a pathogen in breast milk may be intermittent.
These considerations are only some of the variables to be taken into account, in general,
to assess the risk for transmission of an infectious agent from mother to infant via
breast milk or breastfeeding. Efforts to prove transmission of infection in a particular
maternal-infant dyad can be just as difficult and must consider many of the same factors.
This chapter focuses on a discussion of specific, clinically relevant, infectious
agents and diseases, with reasonable estimates of the risk for infection to infants
from breastfeeding. The basic tenet concerning breastfeeding and infection is that
breastfeeding is rarely contraindicated in maternal infection.
243
The few exceptions relate to specific infectious agents with strong evidence of transmission
and to the association of an infant’s illness with significant morbidity and mortality.
The risk or benefit of breastfeeding relative to immunization of a mother or infant
is discussed for certain microorganisms. Appendix D addresses drugs in breast milk
and includes Table D-1, on antiinfective agents, and Chapter 5 reviews how breastfeeding
may protect against infection. Chapter 21 addresses specific concerns relating to
banked breast milk and includes standards developed by the human Milk Banking Association
of North America to guide the appropriate handling of banked human milk relative to
possible infectious agents.
Infection Control Considerations
Isolation precautions have undergone some revisions in terminology and conceptualization.
143
Understanding that the transmission of microorganisms can occur with a known infection
and with unrecognized sources of infection, recommendations have been made for standard
precautions to be applied to all patients to protect health care workers from potentially
infectious body fluids. Additionally, precautions based on the predominant modes of
transmission have been recommended to protect against infection through the airborne
route, direct contact, or contact with droplets. Although these precautions are intended
to be used in clinical situations to protect health care workers, they may be applied
in certain situations to the mother-infant dyad to prevent transmission of infectious
agents from one to the other or to other hospitalized mothers and infants. These precautions
are useful most often when a mother and infant are still hospitalized. The use of
such precautions within the home is not meant to limit breastfeeding. They are intended
to allow breastfeeding in the majority of cases and to facilitate the continuation
of breastfeeding with some additional safeguards in certain situations, after short
temporary periods of stopping breastfeeding, and when to safely use expressed breast
milk (see Appendix F).
Standard Precautions
Standard precautions include preventing contact with blood, all body fluids, secretions
and excretions, nonintact skin, and mucous membranes by (1) careful handwashing before
and after every patient contact; (2) use of gloves when touching body fluids, nonintact
skin, or mucous membranes or any items contaminated with body fluids (linens, equipment,
devices, etc.); (3) use of nonsterile gowns to prevent contact of clothing with body
fluids; (4) use of masks, eye protection, or face shields when splashing with body
fluids is possible; and (5) appropriate disposal of these materials. Standard precautions
should be applied to all patients regardless of actual or perceived risks. The Centers
for Disease Control and Prevention (CDC) does not consider breast milk a body fluid
with infectious risks and thus these policies do not apply to breast milk. (See section
on misadministration of breast milk later in this chapter as a possible exception
to this concept.)
In considering breastfeeding infant-mother dyads and standard precautions, body fluids
other than breast milk should be avoided, and only in specified situations should
breast milk also be avoided. In general, clothing or a gown for the mother and bandages,
if necessary, should prevent direct contact with nonintact skin or secretions. Avoiding
infant contact with maternal mucous membranes requires mothers to be aware of and
understand the risks and to make a conscious effort to avoid this type of contact.
The use of gloves, gowns, and masks on infants for protection is neither practical
nor appropriate. The recommendations concerning the appropriateness of breastfeeding
and breast milk are addressed for specific infectious agents throughout this chapter.
Human immunodeficiency virus (HIV) infection is an example of one infection that can
be prevented by the use of standard precautions, including avoiding breast milk and
breastfeeding. The recommendations concerning breastfeeding and HIV and the various
variables and considerations involved are discussed later.
Airborne Precautions
Airborne precautions are intended to prevent transmission via droplet nuclei (dried
respiratory particles smaller than 5 mcm that contain microorganisms and can remain
suspended in the air for long periods) or dust particles containing microorganisms.
Airborne precautions include the use of a private room with negative-air-pressure
ventilation and masks at all times. In the case of pulmonary tuberculosis (TB), respiratory
protective devices (requiring personal fitting and seal testing before use) should
be worn. Airborne precautions are recommended with measles, varicella or disseminated
zoster, and TB. Breastfeeding in the presence of these maternal infections is prohibited
for the infectious period. This is to protect against airborne transmission of the
infection from the mother and to allow the infant to be fed the mother’s expressed
breast milk by another individual. The exception to allowing breast milk would be
local involvement of the breast by varicella-zoster lesions or Mycobacterium tuberculosis,
such that the milk becomes contaminated by the infectious agent.
Droplet Precautions
Transmission via droplets occurs when an individual produces droplets that travel
only a short distance in the air and then contact a new host’s eyes, nose, mouth,
or skin. The common mechanisms for producing droplets include coughing, sneezing,
talking (singing or yelling), suctioning, intubation, nasogastric tube placement,
and bronchoscopy. In addition to standard precautions applied to all patients, droplet
precautions include the use of a private room (preferred) and a mask if within 3 feet
(0.9 m) of the patient. Droplet precautions are recommended for adenovirus, diphtheria,
respiratory infections, Haemophilus influenzae, Neisseria meningitidis or invasive
infection, influenza, mumps, mycoplasma, parvovirus, pertussis, plague (pneumonic),
rubella, and streptococcal pharyngitis, pneumonia, or scarlet fever. The institution
of droplet precautions with a breastfeeding mother who has these infections should
be specified for each particular infection. This may require some period of separation
for the infant and mother (for duration of the illness, for short-term or complete
treatment of the mother, for the infectious period) with use of expressed breast milk
for nutrition in the interim. Prophylactic treatment of the infant, maternal use of
a mask during breastfeeding or close contact combined with meticulous handwashing,
and the mother’s avoidance of touching her mucous membranes may be adequate and reasonable
for certain infections.
Contact Precautions
Contact precautions are meant to prevent transmission of infection via direct contact
(contact between the body surfaces of one individual with another) and indirect contact
(contact of a susceptible host with an object contaminated with microorganisms from
another individual). Contact precautions include cohorting or a private room, gloves
and gowns at all times, and handwashing after removal of gown and gloves. Contact
precautions are recommended for a long list of infections, such as diarrhea in diapered
or incontinent patients with Clostridium difficile infection, Escherichia coli O157:H7,
Shigella, rotavirus, hepatitis A, respiratory illness with parainfluenza virus or
respiratory syncytial virus (RSV), multidrug-resistant (MDR) bacteria (e.g., enterococci,
staphylococci, gram-negative organisms), enteroviral infections, cutaneous diphtheria,
impetigo, herpes simplex virus (HSV) infection, herpes zoster (disseminated or in
immunocompromised individuals), pediculosis, scabies, S. aureus skin infection, viral
hemorrhagic fevers (e.g., Ebola, Lassa), conjunctivitis and abscesses, cellulitis,
or decubitus that cannot be contained by dressings.
94
For a breastfeeding infant-mother dyad, implementation of precautions for each of
these infections in a mother requires meticulous attention to gowning and handwashing
by the mother and a specialized plan for each situation.
Each of these transmission-based precautions can be used together for organisms or
illnesses that can be transmitted by more than one route. They should always be used
in conjunction with standard precautions, which are recommended for all patients.
The Red Book: Report of the Committee on Infectious Diseases by the American Academy
of Pediatrics (AAP)
96
remains an excellent resource for infection control guidelines and recommendations
to prevent transmission in specific situations and infections.
Culturing Breast Milk
Routine culturing of breast milk or culturing breast milk to screen for infectious
agents is not recommended except when the milk is intended as donor milk to another
mother’s child directly or through human milk banks. See Chapter 21 for specific bacterial
count standards for raw donor milk and for pasteurization of donor milk. Breastfeeding
and the expression of or pumping of breast milk (referred to as expressed breast milk)
for later use are not sterile activities.
In general expressed breast milk should not contain large numbers of microorganisms
(less than 104 for raw milk and less than 106 for milk to be pasteurized), nor should
it contain potential pathogens such as Staphylococcus aureus, β-hemolytic streptococci,
Pseudomonas species, Proteus species, or Streptococcus faecalis or faecium. Few studies
have examined “routine” culturing of milk and the significance of specific bacterial
colony counts relative to illness in infants. The studies have been primarily concerned
with premature or low-birth-weight (LBW) infants who remain hospitalized and are commonly
fed via enteral tubes. A study from Canada tested 7610 samples of milk for use in
98 preterm infants.
242
The study did not identify any adverse events in the infants attributed to organisms
growing in the milk samples, and routine bacteriological testing of expressed breast
milk was not recommended. A study from Chicago examined gram-negative bacilli in the
milk used in premature infants.
48
Samples were tested before feeding and from the nasogastric tubes during feeding.
Milk samples from before feeding were less likely to contain gram-negative bacilli
(36%) than milk samples from the nasogastric tubing (60%). Feeding intolerance was
observed when there were more than 103 colony-forming units per milliliter (CFU/mL),
and episodes of sepsis were identified when the bacterial counts in the milk were
greater than or equal to 106 CFU/mL. This study recommended the routine bacteriologic
testing of expressed breast milk. Another study from Arkansas focused on contamination
of feeding tubes during administration of expressed breast milk or formula.
277
Ten infants in the neonatal intensive care unit (NICU) were exposed to greater than
105 gram-negative bacteria in their feeding tubes. The three infants who were fed
expressed breast milk with contamination at greater than 105 organisms remained well,
but the seven formula-fed infants with high levels of bacterial contamination in the
feeding tubes developed necrotizing enterocolitis. The gram-negative bacteria with
high level contamination in the feeding tubes were either Enterobacter or Klebsiella
in all cases. Many NICUs consider 105 to 106 CFU/mL as the significant bacterial count
for gram-negative bacilli in breast milk that places premature and LBW infants at
greater risk for infection.
Even less data are available concerning specific bacterial colony counts for gram-positive
organisms and the risk to the infant. Generally less than 103 gram-positive organisms
per mL of milk is considered acceptable, with only case reports and no controlled
trials to support this cutoff.
When the presence of an infectious illness in an infant and/or the breastfeeding mother’s
breast when breast milk is seriously considered as a possible mechanism of transmission
to the infant, culturing breast milk to identify the organism may be warranted and
useful. More important than hurrying to culture breast milk is the careful instruction
of mothers on the proper technique for collecting expressed breast milk, storing it,
and cleaning the collection unit. The reinforcement of proper technique from time
to time, especially when a question of contamination arises, is equally important.
Many small reports comment on the contamination of breast milk with different collection
methods. Relative comparisons suggest decreasing contamination of expressed breast
milk when collected by the following methods; drip milk, hand pumped milk, manual
expression, modern electric pumped milk. One group from Malaysia published results
showing no difference in contamination between milk collected by electric pump versus
manual expression when collected in the hospital. Expressed breast milk collected
at home by breast pump had higher rates of contamination with staphylococci and gram-negative
bacteria.
46
Discussion continues about the need to discard the first few milliliters of milk to
lower bacteria numbers in expressed breast milk without any evidence to suggest if
this is truly necessary.62., 337. No evidence shows that cleansing the breast with
anything other than tap water decreases the bacterial counts in cultured expressed
breast milk.
414
If an infant is directly breastfeeding, collecting milk for culture by manual expression
and trying to obtain a “midstream” sample (as is done with “midstream” urine collection
for culture) is appropriate. If an infant is being fed expressed breast milk, collecting
and culturing the milk at different points during collection (utilizing the same technique
the mother uses [manual expression, hand pump, or electric pump]) and administration
is appropriate. This might include a sample from immediately after collection, another
of stored expressed breast milk, and a sample of milk from the most recent infant
feeding at the time the decision to culture is made. Please see Box 13-1
for the basic steps in culturing expressed breast milk.
BOX 13-1
Culturing Breast Milk
1.
Wash hands as per routine.
2.
Wash breast with warm tap water and a clean washcloth.
3.
Manually express breast milk (“midstream” collection is not required) or attach breast
pump flange (previously cleaned as per routine) for collection and collect milk.
4.
Place a 3 to 5 mL sample of expressed breast milk in a sterile container with a nonleakable
top.
5.
Deliver to the labatory in less than 1 hour or refrigerate at 4° C until delivery.
Before sending samples to the viral lab or for nucleic acid/polymerase chain reaction
(PCR) testing, confirm that the laboratory will accept and process the sample as requested
and that the appropriate collection container and prelaboratory management of the
specimen are utilized.
6.
Processing of specimens:
a.
Direct examination by gram stain is not required.
b.
Culture on blood agar (BA) and MacConkey agar (MAC) media as per lab standards.
c.
Quantitate all isolates.
d.
Send separate samples for fungal culture, acid-fast bacilli, and viral culture as
indicated, based on the clinical situation.
e.
Perform routine sensitivity testing on all potential pathogens. (This will require
some discussion with the clinician and perhaps a pediatric infectious disease specialist.)
Interpretation of such culture results can be difficult and should involve a pediatric
infectious disease expert, a microbiologist, and hospital epidemiologist. Additional
organism identification is often required, utilizing antibiogram patterns or molecular
fingerprinting by various techniques to correlate a bacterial isolate from breast
milk with an isolate causing disease in infant or mother.
Misadministration of Breast Milk
Misadministration of breast milk, also known as misappropriation, breast milk exposure,
and accidental ingestion of breast milk, and other terms, is a medical-legal issue
when it occurs in a hospital. This scenario occurs when one infant receives breast
milk from another mother by mistake. This occurrence can be very distressing to the
families (recipient patient, recipient parent, and donor mother) and medical staff
involved. The actual risk for transmission of an infectious agent to an infant via
a single ingestion of expressed breast milk (the most common occurrence) from another
mother is exceedingly low. In this scenario, the CDC recommends treating this as an
accidental exposure to a body fluid, which could be infectious.
84
Bacterial, fungal, or parasitic infection from the one exposure is highly unlikely.
The concern is about viral pathogens, known to be blood-borne pathogens, which have
been identified in breast milk and include but are not limited to hepatitis B virus
(HBV), hepatitis C virus (HCV), cytomegalovirus (CMV), West Nile virus, human T-cell
lymphotropic virus (HTLV), and HIV.
Most hospitals have protocols for managing the situation from both the infection control/prevention
and the medical-legal perspectives. These protocols advise informing both families
about what occurred, discussing the theoretical risks of harm from the exposure, and
reviewing test results and/or recommending testing to determine the infectious status
of each mother relative to the above mentioned viruses. HCV is not a contraindication
to breastfeeding and West Nile virus infection in lactating women is rare.74., 177.
Neither infection has a documented effective form of prevention or acute treatment.
Testing either mother (donor or of recipient infant) for these agents is not warranted.
Prenatal testing for HIV is more commonplace throughout the world. The incidence of
HIV among women of childbearing age is low, although it varies significantly by geographic
location, and the hospital or locale-specific incidence would be important to know
to estimate risk. Most women and medical staff are aware that HIV can be transmitted
by breastfeeding; therefore breast milk from HIV-positive women is rarely if ever
stored in hospitals. The risk for transmission of HIV via breastfeeding is due to
the volume of feedings over months (estimated at 400 to 500 feedings in the first
2 months of life) compared with the small “dose of exposure” from one or two “accidental
feedings.” Transmission of HIV from a single breast milk exposure has never been documented.
Immunologic components in breast milk, along with time and cold of storage, inactivate
the HIV in expressed breast milk. For these reasons, the risk for transmission of
HIV via expressed breast milk consumed by another child is thought to be extremely
low. HTLV-I/II infection in childbearing women is uncommon except in certain geographic
regions (Japan, Africa, the Caribbean, and South America). Transmission of HTLV via
breast milk does occur and, like HIV, appears to be related to the volume and duration
of breastfeeding. Limiting the duration of breastfeeding is effective in decreasing
transmission.407., 409., 446. Freezing and thawing expressed breast milk decreases
the infectivity of HTLV-I.
11
In areas of low prevalence, a positive test in a mother should be suspected to be
a false positive test, and retesting with both antibody and polymerase chain reaction
(PCR) testing should be performed. For these reasons the transmission of HTLV-I/II
via accidental expressed breast milk exposure is thought to be extremely low. Although
the majority of women are CMV positive by childbearing age and CMV transmission occurs
via breastfeeding, the risk for CMV in a full-term infant is low. Premature or LBW
infants are at greater risk for developing disease with CMV infection. Freezing expressed
breast milk (at −20° C) for 3 to 5 days significantly decreases the infectivity of
CMV. Here again the risk for CMV transmission from a single accidental exposure to
CMV-positive expressed breast milk is extremely low.
With a discussion of theoretical risk should be a discussion of possible preventive
interventions, such as vaccination or antimicrobial postexposure prophylaxis. If donor
mothers are positive for HBV, it is appropriate to give recipient infants hepatitis
B virus immunoglobulin (HBIG) and HBV vaccines if they have not already received them.
If a donor mother is HIV or HTLV-I/II positive, the potential utility of postexposure
prophylaxis with antiretroviral medications should be considered on a case-by-case
basis. Clinicians participating in these decisions can refer to the AAP Red Book or
the updated United States Public Health Service Guidelines for the Management of Occupational
Exposures to HIV and Recommendations for Postexposure Prophylaxis (available at http://www.cdc.gov/mmwr/preview/mmwrhtml/rr5409al.htm).
324
It may also be appropriate to consult a pediatric infectious disease specialist.
Additional important components of the hospital-based protocols for managing accidental
expressed breast milk exposure include ongoing psychosocial support for the families
and staff, documentation of medical discussions with the families, investigative steps,
consents and interventions, and the demonstration of ongoing infection control efforts
to prevent additional events of misadministration of breast milk.
Clinical Syndromes and Conditions
Microorganisms produce a whole spectrum of clinical illnesses affecting mothers and
infants. Many situations carry the risk for transmission of the involved organism
from a mother to the infant, or vice versa; in general, however, infants are at greater
risk because of such factors as inoculum size and immature immune response. As always,
an infection must be accurately diagnosed in a timely manner. Empiric therapy and
initial infection control precautions should begin promptly based on the clinical
symptoms and the most likely etiologic agents. When dealing with a maternal infection,
clarifying the possible modes of transmission and estimating the relative risk for
transmission to the infant are essential first steps to decision-making about isolating
a mother from her infant and the appropriateness of continuing breastfeeding or providing
expressed breast milk. Breastfeeding infrequently is contraindicated in specific maternal
infections.
243
Often the question of isolation and interruption of breastfeeding arises when symptoms
of fever, pain, inflammation, or other manifestations of illness first develop in
a mother and the diagnosis is still in doubt. A clinical judgment must be made based
on the site of infection, probable organisms involved, possible or actual mechanisms
of transmission of these organisms to the infant, estimated virulence of the organism,
and likely susceptibility of the infant. Additionally, by the time the illness is
clearly recognized or diagnosed in a mother, the infant has already been exposed.
Given the dynamic nature of the immunologic benefits of breast milk, continuation
of breastfeeding at the time of diagnosis or illness in a mother can provide the infant
protection rather than continued exposure in most illnesses. Stopping breastfeeding
is rarely necessary. Many situations associated with maternal fever do not require
separation of mother and infant, such as engorgement of the breasts, atelectasis,
localized nonsuppurative phlebitis, or urinary tract infections.
Appendix F lists a number of clinical syndromes, conditions, and organisms that require
infection control precautions in hospitals. This appendix also includes short lists
of possible etiologic agents for these conditions and appropriate precautions and
recommendations concerning breastfeeding for different scenarios or organisms. This
chapter considers specific infectious agents that are common, clinically significant,
or of particular interest.
Bacterial Infections
Anthrax
Bacillus anthracis, a gram-positive, spore-forming rod, causes zoonotic disease worldwide.
Human infection typically occurs due to contact with animals or their products. Three
forms of human disease occur: cutaneous anthrax (the most common), inhalation anthrax,
and gastrointestinal (GI) disease (rare). Person-to-person transmission can occur
as a result of discharge from cutaneous lesions, but no evidence of human-to-human
transmission of inhalational anthrax is available. No evidence of transmission of
anthrax via breast milk exists. Standard contact isolation is appropriate for hospitalized
patients or patients with draining skin lesions.
The issue of anthrax as a biologic weapon has exaggerated its importance as a cause
of human disease. The primary concerns regarding anthrax and breastfeeding are antimicrobial
therapy or prophylaxis in breastfeeding mothers and the possibility that infant and
mother were exposed by intentional aerosolization of anthrax spores. The CDC published
recommendations for treatment and prophylaxis in infants, children, and breastfeeding
mothers.
72
The recommendations include the use of ciprofloxacin, doxycycline, amoxicillin, and
several other agents without discontinuing breastfeeding. Little available is information
on ciprofloxacin and doxycycline in breast milk for prolonged periods of therapy or
prophylaxis (60 days) and possible effects on infants’ teeth and bone/cartilage growth
during that time period. Depending on the clinical situation and sensitivity testing
of the identified anthrax strain, other agents can be substituted to complete the
60-day course. The CDC has approved the use of ciprofloxacin and doxycycline for breastfeeding
women for short courses of therapy (less than several weeks).
Simultaneous exposure of infant and mother could occur from primary aerosolization
or from spores “contaminating” the local environment. In either case decontamination
of the mother-infant dyad’s environment should be considered.
Breastfeeding can continue during a mother’s therapy for anthrax as long as she is
physically well. Open cutaneous lesions should be carefully covered and, depending
on the situation, simultaneous prophylaxis for the infant may be appropriate.
Botulism
Considerable justifiable concern has been expressed because of the reports of sudden
infant death from botulism. Infant botulism is distinguished from food-borne botulism
from improperly preserved food containing the toxin and from wound botulism from spores
entering the wound. Infant botulism occurs when the spores of Clostridium botulinum
germinate and multiply in the gut and produce the botulinal toxin in the GI tract.
17
The toxin binds presynaptically at the neuromuscular junction, preventing acetylcholine
release. The clinical picture is a descending, symmetric flaccid paralysis. Not every
individual who has C. botulinum identified in the stool experiences a clinical illness.
The age of infants seems to relate to their susceptibility to illness. The illness
is mainly in children younger than 12 months of age; the youngest patient described
in the literature was 6 days old.17 Most children become ill between 6 weeks and 6
months of age. The onset of illness seems to occur earlier in formula-fed infants
compared with breastfed infants. When a previously healthy infant younger than 6 months
of age develops constipation, then weakness and difficulty sucking, swallowing, crying,
or breathing, botulism is a likely diagnosis. The organisms should be looked for in
the stools, and electromyography may or may not be helpful.
In a group reviewed by Arnon et al,
19
33 of 50 patients hospitalized in California were still being nursed at onset of the
illness. A beneficial effect of human milk was observed in the difference in the mean
age at onset, with breastfed infants being twice as old as formula-fed infants with
the disease. The breastfed infants’ symptoms were milder. Breastfed infants receiving
iron supplements developed the disease earlier than those who were breastfed but unsupplemented.
Of the cases of sudden infant death from botulism, no infants were breastfed within
10 weeks of death. All were receiving iron-fortified formulas. In most cases, no specific
food source of C. botulinum can be identified, but honey is the food most often implicated,
and corn syrup has been implicated in infants older than 2 months of age. Honey may
contain botulism spores, which can germinate in the infant gut. However, botulin toxin
has not been identified in honey. It has been recommended that honey not be given
to infants younger than 12 months of age. This includes putting honey on a mother’s
nipples to initiate an infant’s interest in suckling.
Arnon
18
reviewed the first 10 years of infant botulism monitoring worldwide. The disease has
been reported from 41 of the 50 states in the United States and from eight countries
on four continents. The relationship to breastfeeding and human milk is unclear. In
general the acid stools (pH 5.1 to 5.4) of human milk fed infants encourage Bifidobacterium
species. Few facultative anaerobic bacteria, or clostridia, existing as spores, are
present in breastfed infants. In contrast, formula-fed infants have stool pHs ranging
from 5.9 to 8.0, with few bifidobacteria, primarily gram-negative bacteria, especially
coliforms and Bacteroides species. C. botulinum growth and toxin production decrease
with declining pH and usually stops below pH 4.6. Breast milk also contains additional
protective immunologic components, which purportedly have activity against botulinum
toxin.
269
The relationship between the introduction of solid foods or weaning in both formula-fed
and breastfed infants and the onset of botulism remains unclear. For a breastfed infant,
the introduction of solid food may cause a major change in the gut with a rapid rise
in the growth of enterobacteria and enterococci followed by progressive colonization
by Bacteroides species, clostridia, and anaerobic streptococci. Feeding solids to
formula-fed infants minimally changes the gut flora as these organisms already predominate.
Although more hospitalized infants have been breastfed, sudden-death victims are younger
and have been formula fed, which supports the concept of immunologic protection in
the gut of a breastfed infant.
Much work remains to understand this disease. Clinically, constipation, weakness,
and hypotonicity in a previously healthy child constitute botulism until ruled out,
especially with recent dietary changes. At this time, no reason exists to suspect
breastfeeding as a risk for infant botulism, and some evidence suggests a possible
protective effect from breastfeeding. Breastfeeding should continue if botulism is
suspected in mother or infant.
Brucellosis
Brucella melitensis has been isolated in the milk of animals. Foods and animals represent
the primary sources of infection in humans. Brucellosis demonstrates a broad spectrum
of illness in humans, from subclinical to subacute to chronic illness with nonspecific
signs of weakness, fever, malaise, body aches, fatigue, sweats, arthralgia, and lymphadenitis.In
areas where the disease is enzootic, childhood illness has been described more frequently.
The clinical manifestations in children are similar to those in adults.
259
Infection can occur during pregnancy, leading to abortion (infrequently), and can
produce transplacental spread, causing neonatal infection (rarely).
The transmission of B. melitensis through breast milk has been implicated in neonatal
infection.259., 260. There have been eight cases of brucellosis in infants that were
possibly associated with breastfeeding, but Brucella was not isolated from the breast
milk in any of those cases.∗
One case of brucellosis in an infant caused by breast milk transmission, with B. melitensis
isolated from the breast milk, before antibiotic treatment was given to the mother
has been documented.
415
Additionally, Brucella melitensis has been cultured from women with breast lumps and
abscesses.
295
Only one of six women described in this report was lactating at the time of diagnosis,
and no information about the infant was given. Brucellosis mastitis or abscess should
be considered in women presenting with appropriate symptoms and occupational exposure
to animals, contact with domestic animals in their environment, or exposure to animal
milk or milk products (especially unpasteurized products). The breast inflammation
tends to be granulomatous in nature (without caseation) and is often associated with
axillary adenopathy; occasionally systemic illness in the woman is evident. Treatment
of brucellosis mastitis or abscess should be treated with surgery or fine needle aspiration
as indicated and 4 to 6 weeks of combination antibiotic therapy with two or three
medications. Temporary interruption of breastfeeding with breast pumping and discarding
the milk to continue stimulation of milk production is appropriate. Breastfeeding
should then continue after an initial period of 48 to 96 hours of therapy in the mother.
Acceptable medications for treating the mother while continuing breastfeeding include
gentamicin, streptomycin, tetracycline, doxycycline, trimethoprim-sulfamethoxazole,
and rifampin (see Appendix D).
Chlamydial Infections
Chlamydial infection is the most frequent sexually transmitted disease (STD) in the
United States and is a frequent cause of conjunctivitis and pneumonitis in an infant
from perinatal infection. The major determinant of whether chlamydial infection occurs
in a newborn is the prevalence rate of chlamydial infection of the cervix.
364
Specific chlamydial immunoglobulin A (IgA) has been found in colostrum and breast
milk in a small number of postpartum women who were seropositive for Chlamydia. No
information is available on the role of milk antibodies in protection against infection
in infants.
389
It is not believed that Chlamydia is transmitted via breast milk. Use of erythromycin
or tetracycline to treat mothers and oral erythromycin and ophthalmic preparations
of tetracyclines, erythromycin, or sulfonamides to treat suspected infection in infants
are appropriate during continued breastfeeding. Separating infants from mothers with
chlamydial infections or stopping breastfeeding is not indicated. Simultaneous treatment
of mothers and infants may be appropriate in some situations.
Diphtheria
Corynebacterium diphtheriae causes several forms of clinical disease, including membranous
nasopharyngitis, obstructive laryngotracheitis, and cutaneous infection. Complications
can include airway obstruction from membrane formation and toxin-mediated central
nervous system (CNS) disease or myocarditis. The overall incidence of diphtheria has
declined even though immunization does not prevent infection but does prevent severe
disease from toxin production. Fewer than five cases are reported annually in the
United States.
Transmission occurs via droplets or direct contact with contaminated secretions from
the nose, throat, eye, or skin. Infection occurs in individuals whether they have
been immunized or not, but infection in those not immunized is more severe and prolonged.
As long as the skin of the breast is not involved, no risk for transmission exists
via breast milk. No toxin-mediated disease from toxin transmitted through breast milk
has been reported in an infant.
Breastfeeding, along with chemoprophylaxis and immunization of affected infants, is
appropriate in the absence of cutaneous breast involvement (see Appendix F).
Gonococcal Infections
Maternal infection with Neisseria gonorrhoeae can produce a large spectrum of illness
ranging from uncomplicated vulvovaginitis, proctitis, pharyngitis, conjunctivitis,
or more severe and invasive disease, including pelvic inflammatory disease, meningitis,
endocarditis, or disseminated gonococcal infection. The risk for transmission from
mother to infant occurs mainly during delivery in the passage through the infected
birth canal and occasionally from postpartum contact with the mother (or her partner).
Risk for transmission from breast milk is negligible, and N. gonorrhoeae does not
seem to cause local infection of the breasts. Infection in neonates is most often
ophthalmia neonatorum and less often a scalp abscess or disseminated infection. Mothers
with presumed or documented gonorrhea should be reevaluated for other STDs, especially
Chlamydia trachomatis and syphilis, because some therapies for gonorrhea are not adequate
for either of these infections.
With the definitive identification of gonorrhea in a mother, empiric therapy should
begin immediately, and the mother should be separated from the infant until completion
of 24 hours of adequate therapy. Treatment of the mother with ceftriaxone, cefixime,
penicillin, or erythromycin is without significant risk to the infant. Single-dose
treatment with spectinomycin, ciprofloxacin, ofloxacin, or azithromycin has not been
adequately studied but presumably would be safe for the infant given the 24-hour separation
and a delay in breastfeeding without giving the infant the expressed breast milk (pump
and discard). Doxycycline use in a nursing mother is not routinely recommended.
Careful preventive therapy for ophthalmia neonatorum should be provided, and close
observation of the infant should continue for 2 to 7 days, the usual incubation period.
Empiric or definitive therapy against N. gonorrhoeae may be necessary depending on
an infant’s clinical status and should be chosen on the basis of the maternal isolate’s
sensitivity pattern. The mother should not handle other infants until after 24 hours
of adequate therapy, and the infant should be separated from the rest of the nursery
population, with or without breastfeeding.
Haemophilus Influenzae
Haemophilus influenzae type B can cause severe invasive disease such as meningitis,
sinusitis, pneumonia, epiglottitis, septic arthritis, pericarditis, and bacteremia.
Shock can also occur. Because the increased utilization of the H. influenzae type
B conjugate vaccines, invasive disease caused by Haemophilus has decreased dramatically,
more than 95%, in the United States. Most invasive disease occurs in children 3 months
to 3 years of age. Older children and adults rarely experience severe disease but
do serve as sources of infection for young children. Children younger than 3 months
of age seem to be protected because of passively acquired antibodies from the mothers,
and some additional benefits may be received from breast milk.
Transmission occurs through contact with respiratory secretions, and droplet precautions
are protective. No evidence suggests transmission through breast milk or breastfeeding.
Evidence supports that breast milk limits the colonization of H. influenzae in the
throat.
185
In the rare case of maternal infection, an inadequately immunized infant in a household
is an indication to provide rifampin prophylaxis and close observation for all household
contacts, including the breastfeeding infant. Expressed breast milk can be given to
an infant during the 24-hour separation after the mother’s initiation of antimicrobial
therapy, or if the mother’s illness prevents breastfeeding, it can be reinitiated
when the mother is able (see Appendix F).
Leprosy
Although uncommon in the United States, leprosy occurs throughout the world. This
chronic disease presents with a spectrum of symptoms depending on the tissues involved
(typically the skin, peripheral nerves, and mucous membranes of the upper respiratory
tract) and the cellular immune response to the causative organism, Mycobacterium leprae.
Transmission occurs through long-term contact with individuals with untreated or multibacillary
(large numbers of organisms in the tissues) disease.
Leprosy is not a contraindication to breastfeeding, according to Jeliffe and Jeliffe.
202
The importance of breastfeeding and urgency of treatment are recognized by experts
who treat infants and mothers early and simultaneously. No mother-infant contact is
permitted except to breastfeed. Dapsone, rifampin, and clofazimine are typically and
safely used for infant and mother regardless of the method of feeding (see Appendix
D).
Listeriosis
Listeriosis is a relatively uncommon infection that can have a broad range of manifestations.
In immunocompetent individuals, including pregnant women, the infection can vary from
being asymptomatic to presenting as an influenza-like illness, occasionally with GI
symptoms or back pain. Severe disease occurs more frequently in immunodeficient individuals
or infants infected in the perinatal period (pneumonia, sepsis, meningitis, granulomatosis
infantisepticum).
Although listeriosis during pregnancy may manifest as mild disease in a mother and
is often difficult to recognize and diagnose, it is typically associated with stillbirth,
abortion, and premature delivery. It is thought that transmission occurs through the
transplacental hematogenous route, infecting the amniotic fluid, although ascending
infection from the genital tract may occur.
122
Early and effective treatment of a woman can prevent fetal infection and sequelae.206.,
257. Neonatal infection occurs as either early- or late-onset infection from transplacental
spread late in pregnancy, ascending infection during labor and delivery, infection
during passage through the birth canal, or, rarely, during postnatal exposure.
No evidence in the literature suggests that Listeria is transmitted through breast
milk. Treatment of the mother with ampicillin, penicillin, or trimethoprim-sulfamethoxazole
is not a contraindication to breastfeeding as long as the mother is well enough. Expressed
colostrum or breast milk also can be given if the infant is able to feed orally. The
management of lactation and feeding in neonatal listeriosis is conducted supportively,
as it is in any situation in which an infant is extremely ill, beginning feeding with
expressed breast milk or directly breastfeeding as soon as reasonable.
Meningococcal Infections
N. meningitidis most often causes severe invasive infections, including meningococcemia
or meningitis often associated with fever and a rash and progressing to purpura, disseminated
intravascular coagulation, shock, coma, and death.
Transmission occurs via respiratory droplets. Spread can occur from an infected, ill
individual or from an asymptomatic carrier. Droplet precautions are recommended until
24 hours after initiation of effective therapy. Despite the frequent occurrence of
bacteremia, no evidence indicates breast involvement or transmission through breast
milk.
The risk for maternal infection to an infant after birth is from droplet exposure
and exists whether the infant is breastfeeding or bottle feeding. In either case the
exposed infant should receive chemoprophylaxis with rifampin, 10 mg/kg/dose every
12 hours for 2 days (5 mg/kg/dose for infants younger than 1 month of age), or ceftriaxone,
125 mg intramuscularly (IM) once, for children younger than 15 years of age. Close
observation of the infant should continue for 7 days, and breastfeeding during and
after prophylaxis is appropriate. The severity of maternal illness may prevent breastfeeding,
but it can continue if the mother is able, after the mother and infant have been receiving
antibiotics for 24 hours. A period of separation from the index case for the first
24 hours of effective therapy is recommended; expressed breast milk can be given during
this period.
Pertussis
Respiratory illness caused by Bordetella pertussis evolves in three stages: catarrhal
(nasal discharge, congestion, increasing cough), paroxysmal (severe paroxysms of cough
sometimes ending in an inspiratory whoop, i.e., whooping cough), and convalescent
(gradual improvement in symptoms).
Transmission is via respiratory droplets. The greatest risk for transmission occurs
in the catarrhal phase, often before the diagnosis of pertussis. The nasopharyngeal
culture usually becomes negative after 5 days of antibiotic therapy. Chemoprophylaxis
for all household contacts is routinely recommended. No evidence indicates transmission
through breast milk, with similar risk to breastfed and bottle-fed infants.
In the case of maternal infection with pertussis, chemoprophylaxis for all household
contacts, regardless of age or immunization status, is indicated. In addition to chemoprophylaxis
of the infant, close observation and subsequent immunization (in infants older than
6 weeks of age) are appropriate. Despite chemoprophylaxis, droplet precautions and
separation of mother and infant during the first 5 days of effective maternal antibiotic
therapy are recommended. Expressed breast milk can be provided to the infant during
this period.
Staphylococcal Infections
Staphylococcal infection in neonates can be caused by either S. aureus or coagulase-negative
staphylococci (most often S. epidermidis) and can manifest in a wide range of illnesses.
Localized infection can be impetigo, pustulosis in neonates, cellulitis, or wound
infection, and invasive or suppurative disease includes sepsis, pneumonia, osteomyelitis,
arthritis, and endocarditis. S. aureus requires only a small inoculum (10 to 250 organisms)
to produce colonization in newborns, most often of the nasal mucosa and umbilicus.
193
By the fifth day of life, 40% to 90% of the infants in the nursery will be colonized
with S. aureus.
126
The organism is easily transmitted to others from mother, infant, family, or health
care personnel through direct contact.
Outbreaks in nurseries were common in the past. Mothers, infants, health care workers,
and even contaminated, unpasteurized, banked breast milk were sources of infection.298.,
326. Careful use of antibiotics, changes in nursery layout and procedures, standard
precautions, and cohorting as needed decreased the spread of S. aureus in nurseries.
Now the occurrence of methicillin-resistant S. aureus (MRSA) is again a common problem,
requiring cohorting, occasionally epidemiologic investigation, and careful infection
control intervention. There are numerous reports of MRSA outbreaks in NICUs.∗
The significance of colonization with Staphylococcus and the factors leading to development
of disease in individual patients are not clear. The morbidity and mortality related
to S. aureus infection in neonates is well described.192., 195., 219. Management of
such outbreaks has been reviewed.147., 250.
Little has been written about the role of breastfeeding in colonization with S. aureus
in NICUs, well-baby nurseries, or at home.
MRSA is an important pathogen worldwide. Community-acquired MRSA is different from
hospital-acquired MRSA. Community-acquired MRSA is usually defined as occurring in
an individual without the common predisposing variables associated with hospital-acquired
MRSA, lacking a MDR phenotype (common with hospital-acquired MRSA), frequently carrying
multiple exotoxin virulence factors (such as Panton-Valentine leukocidin toxin), as
well as carrying the smaller type IV staphylococcal cassette cartridge for the MecA
gene on a chromosome (hospital-acquired MRSA carries types I-III staphylococcal cassette
cartridge) and as being molecularly distinct from the common nosocomial strains of
hospital-acquired MRSA. Community-acquired MRSA is most commonly associated with skin
and soft tissue infections and necrotizing pneumonia and less frequently associated
with endocarditis, bacteremia, necrotizing fasciitis, myositis, osteomyelitis, or
parapneumonic effusions. Community-acquired MRSA is so common, it is now being observed
in hospital outbreaks.24., 144., 164., 358. Community-acquired MRSA transmission to
infants via breast milk has been reported.34., 144., 210., 253., 286. Premature or
small-for-gestational-age infants are more susceptible to and at increased risk for
significant morbidity and mortality due to MRSA due in part to prolonged hospitalization,
multiple courses of antibiotics, invasive procedures, and intravenous (IV) lines,
their relative immune deficiency due to prematurity and illness, and altered GI tract
due to different flora and decreased gastric acidity. Therefore colonization with
MRSA may pose a greater risk to infants in NICUs in the long run. Full-term infants
develop pustulosis, cellulitis, and soft tissue infections, but rarely has invasive
disease been reported.82., 132., 298. Fortunov et al
132
from Texas reported 126 infections in term or late-preterm previously well infants
including 43 with pustulosis, 68 with celluliltis or abscesses, and 15 invasive infections.
Family history of soft tissue skin infections and male sex were the only variables
associated with risk for infection; cesarean delivery, breastfeeding, and circumcision
were not.
132
Nguyen et al
298
reported MRSA infections in a well-infant nursery from California. The eleven cases
were all in full-term boys with pustular-vesicular lesions in the groin. The infections
were associated with longer length of stay, lidocaine injection use in infants, maternal
age older than 30 years, and circumcision. Breastfeeding was not an associated risk
factor for MRSA infection.
298
The question of the role of circumcision in MRSA outbreaks was addressed by Van Howe
and Robson.
426
They reported that circumcised boys are at greater risk for staphylococcal colonization
and infection.
426
Others report that S. aureus carriage in infants (and subsequent infection) is most
likely affected by multiple variables including infant factors (antibiotics, surgical
procedures [circumcision being the most common], duration of hospital stay as a newborn),
maternal factors (previous colonization, previous antibiotic usage, mode of delivery,
length of stay), and environmental factors (MRSA in the family or hospital, nursery
stay versus rooming-in, hand hygiene).∗
Gerber et al
147
from the Chicago area published a consensus statement for the management of MRSA outbreaks
in the NICU. The recommendations, which were strongly supported by experimental, clinical,
and epidemiologic data, included using a waterless, alcohol-based hand hygiene product,
monitoring and enforcing hand hygiene, placing MRSA-positive infants in contact precautions
with cohorting if possible, using gloves and gowns for direct contact and masks for
aerosol-generating procedures, cohorting nurses for care of MRSA-positive infants
when possible, periodic screening of infants for MRSA using nares or nasopharyngeal
cultures, clarifying the MRSA status of infants being transferred into the NICU, limiting
overcrowding, and maintaining ongoing instruction and monitoring of health care workers
in their compliance with infection control and hand hygiene procedures. Evaluation
of the outbreak could include screening of health care workers and environmental surfaces
to corroborate epidemiologic data and laboratory molecular analysis of the MRSA strains
if indicated epidemiologically. The use of mupirocin or other decolonizing procedures
should be determined on an individual basis for each NICU.
S. aureus is the most common cause of mastitis in lactating women.317., 394., 395.,
436. Recurrence or persistence of symptoms of mastitis is a well described occurrence
and an important issue in the management of mastitis. Community-acquired MRSA has
been associated with mastitis as well.342., 358., 395. (See Chapter 16 for a complete
discussion of mastitis.)
Two studies, one from France and one from Brazil, investigated the occurrence of MRSA
in expressed breast milk.26., 300. Barbe et al
26
cultured 9171 expressed breast milk samples from 378 women and tested 2351 samples
before pasteurization and 6820 samples after pasteurization. MRSA and methicillin-susceptible
S. aureus were identified respectively in eight samples (0.8%) from three mothers
and 281 samples (19.3%) from 73 mothers of the tested expressed breast milk before
pasteurization. After pasteurization, S. aureus was not detected in any of the 6820
samples of expressed breast milk. Colonization of one infant with MRSA was identified,
but no MRSA infections were identified in any of the hospitalized infants in the NICU
during the 18 months of the study.
26
Novak et al
300
identified MRSA in 57 of 500 samples (11%) of expressed fresh-frozen milk from 500
different donors from five Brazilian milk banks. Only 3 of the 57 samples were positive
with high-level bacterial counts of MRSA: greater than 10,000 CFU/mL. These were the
only samples that would not have been acceptable by bacteriological criteria according
to Brazilian or American criteria for raw milk use. They did not investigate other
epidemiological data to identify possible variables associated with low or high level
contamination of expressed breast milk with MRSA.
300
Management of an infant and/or mother with MRSA infection relative to breastfeeding
or use of breast milk should be based on the severity of disease and whether the infant
is premature, LBW, very-low-birth-weight (VLBW), previously ill, or full term.
Full-term infants who themselves or their mothers develop mild to moderate infections
(impetigo, pustulosis, cellulitis/abscess, mastitis/breast abscess, or soft tissue
infection) can continue breast feeding after a short period of interruption (24 to
48 hours). During this time, pumping to maintain the milk supply should be supported,
an initial evaluation for other evidence of infection should be done in the maternal-infant
dyad, the infected child and/or mother should be placed on “commonly” effective therapy
for the MRSA infection, and ongoing observation for clinical disease should continue.
The mother and infant can “room-in” together in the hospital, if necessary, with standard
and contact precautions. Culturing the breast milk is not necessary. Empiric therapy
for the infant may be chosen based on medical concerns for the infant and the known
sensitivity testing of the MRSA isolate. Appropriate antibiotic choices include short-term
use of azithromycin (erythromycin use during infancy [less than 6 weeks of age], or
breastfeeding associated with an increased risk for hypertrophic pyloric stenosis),
sulfamethoxazole-trimethoprim (in the absence of G6PD deficiency and older than 30
days of age), clindamycin, and perhaps linezolid for mild to moderate infections.
Infants in NICUs (premature, LBW, VLBW, and/or previously ill), who themselves or
their mothers have a MRSA infection, should have the breast milk cultured and suspend
breastfeeding or receiving breast milk from their mother until the breast milk is
shown to be culture negative for MRSA. The infant should be treated as indicated for
their infection or empirically treated if symptomatic (with pending culture results)
and closely observed for development of new signs or symptoms of infection. Pumping
to maintain the milk supply and the use of banked breast milk are appropriate. The
infant should be placed on contact precautions, in addition to the routine standard
precautions. The infant can be cohorted with other MRSA-positive infants with nursing
care cohorted as well. For the mother with MRSA infection, she should be instructed
concerning hand hygiene, the careful collection, handling, and storage of breast milk,
contact precautions to be used with her infant, and the avoidance of contact with
any other infants. The mother can receive several possible antibiotics for MRSA that
are compatible with breastfeeding when used for a short period. If the mother remains
clinically well, including without evidence of mastitis, but her breast milk is positive
for MRSA greater than 104 CFU/mL, empiric therapy to diminish or eradicate colonization
would be appropriate. Various regimens have been proposed to “eradicate” MRSA colonization,
but none have been proven to be highly efficacious. These regimens usually include
systemic antibiotics with one or two medications (rifampin added as the second medication),
nasal mupirocin to the nares twice daily for 1to 2 weeks with routine hygiene, with
or without the usage of hexachlorophene (or similar topical agent or cleanser) for
bathing during the 1 to 2 week treatment period. There is no clear information concerning
the efficacy of using similar colonization eradication regimens for other household
members or pets in preventing recolonization of the mother or infant. Before reintroducing
the use of the mother’s breast milk to the infant at least two to three negative breast
milk cultures should be obtained after completion of therapy.
Routine screening of breast milk provided by mothers for their infants in NICUs for
the presence of MRSA is not indicated in the absence of MRSA illness in the maternal-infant
dyad, an MRSA outbreak in NICUs, or a high frequency of MRSA infection in a specific
NICU.
Toxin-Mediated Staphylococcus Disease
One case of staphylococcal scalded skin syndrome was reported by Katzman and Wald
208
in an infant breastfed by a mother with a lesion on her areola that did not respond
to ampicillin therapy for 14 days. Subsequently the infant developed conjunctivitis
with S. aureus, which produced an exfoliative toxin, and a confluent erythematous
rash without mucous membrane involvement or Nikolsky sign. No attempt to identify
the exfoliative toxin in the breast milk was made, and the breast milk was not cultured
for S. aureus. The child responded to IV therapy with nafcillin. This emphasizes the
importance of evaluating mother and infant at the time of a suspected infection and
the need for continued observation of the infant for evidence of a pyogenic infection
or toxin-mediated disease, especially with maternal mastitis or breast lesions.
This case also raises the issue of when and how infants and their mothers become colonized
with S. aureus and what factors lead to infection and illness in each. The concern
is that Staphylococcus can be easily transmitted through skin to skin contact, colonization
readily occurs, and potentially serious illness can occur later, long after colonization.
In the case of staphylococcal scalded skin syndrome or toxic shock syndrome (TSS),
the primary site of infection can be insignificant (e.g., conjunctivitis, infection
of a circumcision, or simple pustulosis), but a clinically significant amount of toxin
can be produced and lead to serious disease.
Toxic shock syndrome can result from S. aureus or Streptococcus pyogenes infection
and probably from a variety of antigens produced by other organisms. TSS-1 has been
identified as a “superantigen” that affects the T lymphocytes and other components
of the immune response, producing an unregulated and excessive immune response and
resulting in an overwhelming systemic clinical response. TSS has been reported in
association with vaginal delivery, cesarean delivery, mastitis, and other local infections
in mothers. Mortality rate in the mother may be as high as 5%.
The case definition of staphylococcal TSS includes meeting all four major criteria:
fever greater than 38.9° C, rash (diffuse macular erythroderma), hypotension, and
desquamation (associated with subepidermal separation seen on skin biopsy). The definition
also includes involvement of three or more organ systems (GI, muscular, mucous membrane,
renal, hepatic, hematologic, or central nervous system); negative titers for Rocky
Mountain spotted fever, leptospirosis, and rubeola; and lack of isolation of S. pyogenes
from any source or S. aureus from the cerebrospinal fluid (CSF).
368
A similar case definition has been proposed for streptococcal TSS.
451
Aggressive empiric antibiotic therapy against staphylococci and streptococci and careful
supportive therapy are essential to decreasing illness and death. Oxacillin, nafcillin,
first-generation cephalosporins, clindamycin, erythromycin, and vancomycin are acceptable
antibiotics, even for a breastfeeding mother. The severity of illness in the mother
may preclude breastfeeding, but it can be reinitiated when the mother is improving
and wants to restart. Standard precautions, but allowing breastfeeding, are recommended.
Staphylococcal enterotoxin F has been identified in breast milk specimens collected
on days 5, 8, and 11 from a mother who developed TSS at 22 hours postpartum.
428
S. aureus that produced staphylococcal enterotoxin F was isolated from the mother’s
vagina but not from breast milk. Infant and mother lacked significant antibody against
staphylococcal enterotoxin F in their sera. The infant remained healthy after 60 days
of follow-up. Staphylococcal enterotoxin F is pepsin inactivated at pH 4.5 and therefore
is probably destroyed in the stomach environment, presenting little or no risk to
the breastfeeding infant.
35
Breastfeeding can continue if the mother is able.
Coagulase-Negative Staphylococcus
Coagulase-negative staphylococcal infection (Staphylococcus epidermidis is the predominant
isolate) produces minimal disease in healthy, full-term infants but is a significant
problem in hospitalized or premature infants. Factors associated with increased risk
for this infection include prematurity, high colonization rates in specific nurseries,
invasive therapies (e.g., IV lines, chest tubes, intubation), and antibiotic use.
Illness produced by coagulase-negative staphylococci can be invasive and severe in
high-risk neonates, but rarely in mothers. There are reports of necrotizing enterocolitis
associated with coagulase-negative Staphylococcus. At 2 weeks of age, for infants
still in the nursery, S. epidermidis is a frequent colonizing organism at multiple
sites, with colonization rates as high as 75% to 100%. Serious infections with coagulase-negative
staphylococci (e.g., abscesses, IV line infection, bacteremia/sepsis,endocarditis,
osteomyelitis) require effective IV therapy. Many strains are resistant to penicillin
and the semisynthetic penicillins, so sensitivity testing is essential. Empiric or
definitive therapy may require treatment with vancomycin, gentamicin, rifampin, teicoplanin,
linezolid, or combinations of these for synergistic activity. Transmission of infection
in association with breastfeeding appears to be no more common than with bottle feeding.
As with S. aureus infection control includes contact and standard precautions. Occasionally,
during presumed outbreaks, careful epidemiologic surveillance may be required, including
cohorting, limiting overcrowding and understaffing, surveillance cultures of infants
and nursery personnel, reemphasis of meticulous infection control techniques for all
individuals entering the nursery, and, rarely, removal of colonized personnel from
direct infant contact.
Staphylococcus epidermidis has been identified as part of fecal microbiota of breastfed
infants.
203
S. epidermidis has also been identified in the breast milk of women with clinical
evidence of mastitis.
107
Nevertheless, S. epidermidis is rarely associated with infection in full-term infants.
Conceivably breast milk for premature infants could be a source of S. epidermidis
colonization in the NICUs. The other factors associated with hospitalization in a
NICU noted previously presumably play a significant role in both colonization and
infection in premature infants. The benefits of early full human milk feeding potentially
outweigh the risk for colonization with S. epidermidis via breast milk.
348
Ongoing education and assistance should be provided to mothers about the careful collection,
storage, and delivery of human breast milk for their premature infants.
353
Streptococcal Infections
Group A
Streptococcus pyogenes (β-hemolytic group A Streptococcus [GAS]) is a common cause
of skin and throat infections in children, producing pharyngitis, cellulitis, and
impetigo. Illnesses produced by GAS can be classified in three categories: (1) impetigo,
cellulitis, or pharyngitis without invasion or complication; (2) severe invasive infection
with bacteremia, necrotizing fasciitis, myositis, or systemic illness (e.g., streptococcal
TSS); and (3) autoimmune-mediated phenomena, including acute rheumatic fever and acute
glomerulonephritis. GAS can also cause puerperal sepsis, endometritis, and neonatal
omphalitis. Significant morbidity and mortality rates are associated with invasive
GAS infection; mortality rate is 20% to 50%, with almost half the survivors requiring
extensive tissue débridement or amputation.
347
Infants are not at risk for the autoimmune sequelae of GAS (rheumatic fever or poststreptococcal
glomerulonephritis). Transmission is through direct contact (rarely indirect contact)
and droplet spread. Outbreaks of GAS in the nursery are rare, unlike with staphylococcal
infections. Either mother or infant can be initially colonized with GAS and transmit
it to the other.
In the situation of maternal illness (extensive cellulitis, necrotizing fasciitis,
myositis, pneumonia, TSS, mastitis), it is appropriate to separate mother and infant
until effective therapy (penicillin, ampicillin, cephalosporins, erythromycin) has
been given for at least 24 hours. Breastfeeding should also be suspended and may resume
after 24 hours of therapy for the mother.
Group B
Group B Streptococcus (GBS, Streptococcus agalactiae) is a significant cause of perinatal
bacterial infection. In parturient women, infection can lead to asymptomatic bacteriuria,
urinary tract infection (often associated with premature birth), endometritis, or
amnionitis. In infants, infection usually occurs between birth and 3 months of age
(1 to 4 cases per 1000 live births). It is routinely classified by the time of onset
of illness in the infant: early onset (0 to 7 days, majority less than 24 hours) and
late onset (7 to 90 days, generally less than 4 weeks). Infants may develop sepsis,
pneumonia, meningitis, osteomyelitis, arthritis, or cellulitis. Early-onset GBS disease
is often fulminant, presenting as sepsis or pneumonia with respiratory failure; three
quarters of neonatal disease is early onset. Type III is the most common serotype
causing disease.
Transmission is believed to occur in utero and during delivery. Colonization rates
of mothers and infants vary between 5% and 35%. Postpartum transmission is thought
to be uncommon, although it has been documented. Risk factors for early-onset GBS
disease include delivery before 37 weeks’ gestation, rupture of membranes for longer
than 18 hours before delivery, intrapartum fever, heavy maternal colonization with
GBS, or low concentrations of anti-GBS capsular antibody in maternal sera.
95
The common occurrence of severe GBS disease before 24 hours of age in neonates has
lead to prevention strategies. Revised guidelines developed by the AAP Committees
on Infectious Diseases and on the Fetus and Newborn
95
have tried to combine various variables for increased risk for GBS infection (prenatal
colonization with GBS, obstetric and neonatal risk factors for early-onset disease)
and provide intrapartum prophylaxis to those at high risk (Figure 13-1
)The utilization of these guidelines and intrapartum prophylaxis across the United
States has decreased the incidence of early-onset disease by approximately 80%. In
2005, the incidence of early-onset disease was 0.35 cases per 1000 live births.
95
Figure 13-1
Empiric management of neonate born to mother who received intrapartum antimicrobial
prophylaxis (IAP) for prevention of early-onset group B streptococcal (GBS) disease.
CSF, Cerebrospinal fluid; CBC, complete blood count. This algorithm is not an exclusive
course of management. Variations that incorporate individual circumstances or institutional
preferences may be appropriate.
(From Committee on Infectious Diseases, American Academy of Pediatrics: Red Book Report
of the Committee on Infectious Disease, ed 26, Elk Grove, Ill, 2003, American Academy
of Pediatrics, p 590.)
Late-onset GBS disease is thought to be the result of transmission during delivery
or in the postnatal period from maternal, hospital, or community sources. Dillon et
al
112
demonstrated that 10 of 21 infants with late-onset disease were colonized at birth,
but the source of colonization was unidentified in the others. Gardner et al
141
showed that only 4.3% of 46 children who were culture negative for GBS at discharge
from the hospital had acquired GBS by 2 months of age. Anthony et al
15
noted that many infants are colonized with GBS, but the actual attack rate for GBS
disease is low and difficult to predict.
Acquisition of GBS through breast milk or breastfeeding is uncommon. Cases of late-onset
GBS disease associated with GBS in the maternal milk have been reported.58., 214.,
313., 366., 438. Some of the mothers had bilateral mastitis, at least one had delayed
evidence of unilateral mastitis, and the others were asymptomatic. It was not clear
when colonization of the infants occurred or when infection or disease began in the
infants. The authors discussed the possibility that the infants were originally colonized
during delivery, subsequently colonized the mothers’ breasts during breastfeeding,
and then became reinfected at a later time. Butter and DeMoor
56
showed that infants initially colonized on their heads at birth had GBS cultured from
their throat, nose, or umbilicus 8 days later. Whenever they cultured GBS from the
nipples of mothers, the authors also found it in the nose or throat of the infants.
Byrne et al
58
presented a review of GBS disease associated with breastfeeding and made recommendations
to decrease the risk for transmission of GBS to infants via breastfeeding or breast
milk. Some of their recommendations included confirming appropriate collection and
processing procedures for GBS cultures
370
in medical facilities to decrease false-negative cultures, reviewing proper hygiene
for pumping, collection, and storage of expressed breast milk with mothers, reviewing
the signs and symptoms of mastitis with mothers, and utilizing banked human milk as
needed instead of mother’s milk. When a breastfed infant develops late-onset GBS disease,
it is appropriate to culture the milk. (See discussion of culturing breast milk earlier
in this chapter.) Consider treatment of the mother to prevent reinfection if the milk
is culture positive for GBS (greater than 104 CFU/mL), with or without clinical evidence
of mastitis in the mother. Withholding the mother’s milk until it is confirmed to
be culture negative for a pathogen is appropriate and should be accompanied by providing
ongoing support and instruction to the mother concerning pumping and maintaining her
milk supply. Serial culturing of expressed breast milk after treatment of the mother
for GBS disease or colonization would be appropriate to insure the ongoing absence
of a pathogen in the expressed breast milk. There are reports of reinfection of the
infant from breast milk.23., 225. Eradication of GBS mucosal colonization in the infant
or the mother may be difficult. Some authors have recommended using rifampin prophylactically
in both the mother and infant at the end of treatment to eradicate mucosal colonization.
23
(See Chapter 16 for management of mastitis in the mother.) A mother or infant colonized
or infected with GBS should be managed with standard precautions
94
while in the hospital. Ongoing close evaluation of the infant for infection or illness
and empiric therapy for GBS in the infant are appropriate until the child has remained
well and cultures are subsequently negative at 72 hours. Occasionally, epidemiologic
investigation in the hospital will utilize culturing medical staff and family members
to detect a source of late-onset GBS disease in the nursery. This can be useful when
more than one case of late-onset disease is detected with the same serotype. Cohorting
in such a situation may be appropriate. Selective prophylactic therapy for colonized
infants to eradicate colonization may be considered, but unlike GAS or Staphylococcus
infection, GBS infection in nurseries has not been reported to cause outbreaks. No
data support screening all breastfeeding mothers and their expressed breast milk for
GBS as a reasonable method for protecting against spread of GBS infection via expressed
breast milk. Selective culturing of expressed breast milk may be appropriate in certain
situations.
Tuberculosis
The face of tuberculosis (TB) is changing throughout the world. In the United States
the incidence of TB rose during 1986 through 1993 and has been declining since then.
60
Increased rates of TB were noted in adults between 25 and 45 years of age, and because
these are the primary childbearing years, the risk for transmission to children increased.
TB during pregnancy has always been a significant concern for patients and physicians
alike.
340
It is now clear that the course and prognosis of TB in pregnancy are less affected
by the pregnancy and more determined by the location and extent of disease, as defined
primarily by chest radiograph, and by the susceptibility of the individual patient.
Untreated TB in pregnancy is associated with maternal and infant mortality rates of
30% to 40%.
365
Effective therapy is crucial to the clinical outcome in both pregnant and nonpregnant
women. TB during pregnancy rarely results in congenital TB.
Any individual in a high-risk group for TB should be screened with a tuberculin skin
test (TST). No contraindication or altered responsiveness to the TST exists during
pregnancy or breastfeeding. Interpretation of the TST should follow the most recent
guidelines, using different sizes of induration in different-risk populations as cutoffs
for a positive test, as proposed by the CDC.
68
Figure 13-2
outlines the evaluation and treatment of a pregnant woman with a positive TST.
398
Figure 13-2
Evaluation and treatment of pregnant woman with positive tuberculin skin test.
(From Starke JR: Tuberculosis, an old disease but a new threat to mother, fetus, and
neonate, Clin Perinatol 24:107, 1997.)
Treatment of active TB should begin as soon as the diagnosis is made, regardless of
the fetus’ gestational age, because the risk for disease to mother and fetus clearly
outweighs the risks of treatment. Isoniazid, rifampin, and ethambutol have been used
safely in all three trimesters. Isoniazid and pyridoxine therapy during breastfeeding
is safe, although the risk for hepatotoxicity in the mother may be a concern during
the first 2 months postpartum.
391
Congenital TB is extremely rare if one considers that 7 to 8 million cases of TB occur
each year worldwide and that less than 300 cases of congenital TB have been reported
in the literature. As with other infectious diseases presenting in the perinatal period,
distinguishing congenital infection from perinatal or postnatal TB in infants can
be difficult.
Postnatal TB infection in infancy typically presents with severe disease and extrapulmonary
extension (meningitis, lymphadenopathy, and bone, liver, spleen involvement). Airborne
transmission of TB to infants is the major mode of postnatal infection because of
close and prolonged exposure in enclosed spaces, especially in their own household,
to any adult with infectious pulmonary TB. Potential infectious sources could be the
mother or any adult caregiver, such as babysitters, day care workers, relatives, friends,
neighbors, and even health care workers.
The suspicion of TB infection or disease in a household with possible exposure of
an infant is a highly anxiety-provoking situation (Figure 13-3
). Although protection of an infant from infection is foremost in everyone’s mind,
separation of the infant from the mother should be avoided when reasonable. Every
situation is unique, and the best approach will vary according to the specifics of
the case and accepted principles of TB management. The first step in caring for the
potentially exposed infant is to determine accurately the true TB status of the suspected
case (mother or household contact). This prompt evaluation should include a complete
history (previous TB infection or disease, previous or ongoing TB treatment, TST status,
symptoms suggestive of active TB, results of most recent chest radiograph, sputum
smears, or cultures), physical examination, a TST if indicated, a new chest radiograph,
and mycobacterial cultures and smears of any suspected sites of infection. All household
contacts should be evaluated promptly, including history and TST with further evaluation
as indicated.
68
Continued risk to the infant can occur from infectious household contacts who have
not been effectively evaluated and treated.
Figure 13-3
Management of newborn infant exposed to tuberculin-positive household contact. CXR,
Chest x-ray film; INH, isoniazid; MDR, multidry-resistant; TB, tuberculosis.
An infant should be separated temporarily from the suspected source if symptoms suggest
active disease or a recent TST documents conversion, and separation should continue
until the results of the chest radiograph are seen. Because of considerable variability
in the course of illness and the concomitant infectious period, debate continues without
adequate data about the appropriate period of separation.
278
This should be individualized given the specific situation. HIV testing and assessment
of the risk for MDR TB should be done in every case of active TB. Sensitivity testing
should be done on every Mycobacterium tuberculosis isolate. Table 13-1
summarizes the management of the newborn infant whose mother (or other household contact)
has TB.
TABLE 13-1
Management of Newborn Whose Mother (or Other Household Contact) Has Tuberculosis (TB)
Mother/Infant Status
Additional Workup Recommended1
Therapy For Mother/Contact
Therapy For Infant
Separation2
Breast Milk3
Breastfeeding3
1. TB infection, no disease4
None for mother/contact
Prophylactic5
None
No
Yes
Yes
2. TB infection: Abnormal CXR not suggestive of active disease
Decide active vs. inactive disease
a. Symptoms or physical findings suggestive of active TB
Aerosolized sputums (culture, smears)6
Active disease: empiric5
Isoniazid7
Yes
Yes
No8
Inactive disease: prophylactic5
None
No
Yes
Yes
b. No symptoms or physical findings suggestive of active TB
Aerosolized sputums in select cases
Prophylactic5
None
No
Yes
Yes
3. TB infection: Abnormal CXR suggestive of active disease
Aerosolized sputums (culture, smears)6
Empiric therapy5
Isoniazid7
Yes
Yes
No8
4. Active pulmonary TB: Suspected MDR TB
Aerosolized sputums (culture, smears)6
Consult TB specialist for best regimen9
Consult pediatric TB specialist9
Yes
Yes
No
Consider bacille Calmette-Guérin vaccine
5. TB disease: Suspected mastitis10
Aerosolized sputums (culture, smears)6
Empiric5
Isoniazid7
Yes
No11
No
6. TB infection: Status undertermined12
Perform/interpret CXR within 24 hours
Yes, until CXR interpreted (see a and b)
Yes
No
a. Abnormal CXR not suggestive of active disease
Proceed as in 2
As in 2
As in 2
As in 2
b. Abnormal CXR suggestive of active disease
Proceed as in 3
As in 3
As in 3
As in 3
Notes:
1
Further workup should always include evaluation of TB status of all other household
(or close) contacts by tuberculin skin testing (TST), review of symptoms, physical
examination, and chest x-ray (CXR). Sputum smears and cultures should be done as indicated.
2
Separation should occur until interpretation of CXR confirms absence of active disease,
or, with active disease, separation should continue until individual is no longer
considered infectious: three negative consecutive sputum smears, adequate ongoing
empiric therapy, and decreased fever, cough, and sputum production. Separation means
in a different house or location, not simply separate rooms in a household. Duration
of separation should be individualized for each case in consultation with TB specialist.
3
This assumes no evidence of breast involvement, suspected TB mastitis, or lesion (except
in status 5, when breast involvement is considered). Risk to infant is via aerosolized
bacteria in sputum from the lung. Expressed breast milk can be given even if separation
of mother and infant is advised.
4
TST positive, no symptoms or physical findings suggestive of TB, negative CXR.
5
Prophylactic therapy: isoniazid 10 mg/kg/day, maximum 300 mg for 6 months; pyridoxine
25 to 50 mg/day for 6 months. Empiric therapy: standard three- or four-drug regimens
for 2 months, and treatment should continue for total of 6 months with isoniazid and
rifampin when organism is shown to be sensitive. Suspected multidrug-resistant (MDR)
TB requires consultation with TB specialist to select optimum empiric regimen and
for ongoing monitoring of therapy and clinical response.
6
Sensitivity testing should be done on any positive culture.
7
Isoniazid 10 mg/kg/day for 3 to 9 months depending on mother’s or contact’s status;
repeat TST at 3 months and obtain normal CXR in infant before stopping isoniazid.
Before beginning therapy, workup of infant for congenital or active TB may be appropriate.
This workup should be determined by clinical status of infant and suspected potential
risk, and may include TST after 4 weeks of age, with CXR, complete blood count, and
erythrocyte sedimentation rate, liver function tests, cerebrospinal fluid analysis,
gastric aspirates, sonography/computed tomography of liver/spleen, and chest if congenital
TB is suspected.
8
Breastfeeding is proscribed when separation of mother and infant is indicated because
of risk for aerosolized transmission of bacteria. Expressed breast milk given to infant
via bottle is acceptable in absence of mastitis or breast lesions.
9
Consult with TB specialist about MDR TB. Empiric therapy will be chosen based on the
most recent culture sensitivities of index patient or perhaps suspected source case,
if known, as well as medication toxicities and other factors.
10
TB mastitis usually involves a single breast with associated axillary lymph node swelling
and, infrequently, a draining sinus tract. It can also present as a painless mass
or edema of breast.
11
With suspected mastitis or breast lesion caused by TB, even breast milk is contraindicated
until lesion or mastitis heals, usually 2 weeks or more.
12
Patient has a documented, recent TST conversion but has not been completely evaluated.
Evaluation should begin and CXR done and evaluated in less than 24 hours to minimize
separation of this person from infant. Further workup should proceed as indicated
by symptoms, physical findings, and CXR results.
Data from Committee on Infectious Diseases, American Academy of Pediatrics: Red Book:
Report of the Committee on Infectious Diseases, ed 26, Elk Grove Village, Ill. 2003,
American Academy of Pediatrics.
Initiation of prophylactic isoniazid therapy in the infant has been demonstrated to
be effective in preventing TB infection and disease in the infant. Therefore continued
separation of infant and mother is unnecessary after therapy in both mother and child
has begun.
114
The real risk to an infant requiring separation is from airborne transmission. Separation
of the infant from a mother with active pulmonary TB is appropriate, regardless of
the method of feeding. However, in many parts of the world, after therapy in the mother
and prophylaxis with isoniazid in the infant has begun, the infant and mother are
not separated. With or without separation, the mother and infant should continue to
be closely observed throughout the course of maternal therapy to ensure good compliance
with medication by both mother and infant and to identify, early on, any symptoms
in the infant suggestive of TB.
Tuberculous mastitis occurs rarely in the United States but does occur in other parts
of the world∗
and can lead to infection in infants, frequently involving the tonsils. A mother usually
has a single breast mass and associated axillary lymph node swelling and infrequently
develops a draining sinus. TB of the breast can also present as a painless mass or
edema. Involvement of the breast can occur with or without evidence of disease at
other sites. Evaluation of extent of disease is appropriate, including lesion cultures
by needle aspiration, biopsy, or wedge resection and milk cultures. Therapy should
be with multiple anti-TB medications, but surgery should supplement this, as needed,
to remove extensive necrotic tissue or a persistently draining sinus.
16
Neither breastfeeding nor breast milk feeding should be done until the lesion is healed,
usually 2 weeks or more. Continued anti-TB therapy for 6 months in the mother and
isoniazid for the infant for 3 to 6 months is indicated.
In the absence of tuberculous breast infection in the mother, transmission of TB through
breast milk has not been documented. Thus even though temporary separation of infant
and mother may occur pending complete evaluation and initiation of adequate therapy
in the mother and prophylactic isoniazid therapy (10 mg/kg/day as a single daily dose)
in the infant, breast milk can be expressed and given to the infant during the short
separation. Breastfeeding can safely continue whether the mother, infant, or both
are receiving anti-TB therapy. Anti-TB medications (isoniazid, rifampin, pyrazinamide,
aminoglycosides, ethambutol, ethionamide, p-aminosalicylic acid) have been safely
used in infancy, and therefore the presence of these medications in smaller amounts
in breast milk is not a contraindication to breastfeeding.
Although conflicting, reports indicate that breastfeeding by TST-positive mothers
does influence infants’ responses to bacille Calmette-Guérin vaccine, the TST, and
perhaps the M. tuberculosis bacillus. Despite efforts to identify either a soluble
substance or specific cell fractions (gamma/delta T cells) in colostrum and breast
milk that affect infants’ immune responsiveness, no unified theory explains the various
reported changes and no evidence has identified a consistent, clinically significant
effect.39., 213., 319., 367.
Viral Infections
Arboviruses
Arboviruses were originally a large collection of viruses grouped together because
of the common mode of transmission through arthropods. They have now been reclassified
into several different families: Bunyaviridae, Togaviridae, Flaviviridae, Reoviridae,
and others. They include more than 30 human pathogens.
These organisms primarily produce either CNS infections (encephalitis, meningoencephalitis)
or undifferentiated illnesses associated with fever and rash, severe hemorrhagic manifestations,
and involvement of other organs (hepatitis, myalgia, polyarthritis). Infection with
this array of viruses may also be asymptomatic and subclinical, although how often
this occurs is uncertain. Some of the notable human pathogens include Bunyaviridae
(California serogroup viruses), Hantavirus, Hantaan virus, Phlebovirus (Rift Valley
fever), Nairovirus (Crimean-Congo hemorrhagic fever), Alphavirus (western, eastern,
and Venezuelan equine encephalomyelitis viruses, chikungunya virus), Flavivirus (St.
Louis encephalitis virus, Japanese encephalitis virus, dengue viruses, yellow fever
virus, tick-borne encephalitis viruses), and Orbivirus (Colorado tick fever). Other
than for Crimean-Congo hemorrhagic fever and for reported cases of Colorado tick fever
associated with transfusion, direct person-to-person spread has rarely been described.
Recent outbreaks of chikungunya virus infection in Reunion Island and in India described
infection in young infants probably secondary to vertical spread from mother to infant
transplacentally.146., 339., 422. A few cases of early fetal deaths were associated
with infection in pregnant women. The cases of vertical transmission occurred with
near-term infection in the mothers, and the infants developed illness within 3 to
7 days of delivery.146., 339. No evidence for transmission via breast milk or breastfeeding
is available.
Little evidence indicates that these organisms can be transmitted through breast milk.
The exceptions to this include evidence of transmission of two Flaviviruses via breast
milk, West Nile virus, and yellow fever vaccine virus. Standard precautions are generally
sufficient. With any of these infections in a breastfeeding mother, the severity of
the illness may determine the mother’s ability to continue breastfeeding. Providing
the infant with expressed breast milk is acceptable. (See the discussion of West Nile
virus and yellow fever vaccine virus later in this chapter.)
In general, treatment for these illnesses is supportive. However, ribavirin appears
to decrease the severity of and mortality from Hantavirus pulmonary syndrome, hemorrhagic
fever with renal failure, and Crimean-Congo hemorrhagic fever. Ribavirin has been
described as teratogenic in various animal species and is contraindicated in pregnant
women. No information is available concerning ribavirin in breast milk, with little
information available on the use of IV or oral ribavirin in infants.
Arenaviruses
Arenaviruses are single-stranded ribonucleic acid (RNA) viruses that infect rodents
and are acquired by humans through the rodents. The six major human pathogens in this
group are (1) lymphocytic choriomeningitis virus, (2) Lassa fever virus, (3) Junin
virus (Argentine hemorrhagic fever), (4) Machupo virus (Bolivian hemorrhagic fever),
(5) Guanarito virus (Venezuelan hemorrhagic fever), and (6) Sabia virus. The geographic
distribution of these viruses and the illness they cause are determined by the living
range of the host rodent (reservoir). The exact mechanism of transmission to humans
is unknown and hotly debated.25., 69., 131. Direct contact and aerosolization of rodent
excretions and secretions are probable mechanisms.
Lymphocytic choriomeningitis virus is well recognized in Europe, the Americas, and
other areas. Perinatal maternal infection can lead to severe disease in the newborn,
but no evidence suggests transmission through breast milk.28., 224. Standard precautions
with breastfeeding are appropriate.
Lassa fever (West Africa) and Argentine hemorrhagic fever (Argentine pampas) are usually
more severe illnesses with dramatic bleeding and involvement of other organs, including
the brain. These fevers more frequently lead to shock and death than do the forms
of hemorrhagic fever caused by the other viruses in this group. Person-to-person spread
of Lassa fever is believed to be common, and transmission within households does occur.
212
This may relate to prolonged viremia and excretion of the virus in the urine of humans
for up to 30 days.
330
The possibility of persistent virus in human urine, semen, and blood after infection
exists for each of the arenaviruses. The possibility of airborne transmission is undecided.
Current recommendations by the CDC
69
are to use contact precautions for the duration of the illness in situations of suspected
viral hemorrhagic fever. No substantial information describes the infectivity of various
body fluids, including breast milk, for these different viral hemorrhagic fevers.
Considering the severity of the illness in mothers and the risk to the infants, it
is reasonable to avoid breastfeeding in these situations if alternative forms of infant
nutrition can be provided.
As more information becomes available, reassessment of these recommendations is advisable.A
vaccine is in clinical trials in endemic areas for Junin virus and Argentine hemorrhagic
fever. Preliminary studies suggest it is effective, but data are still being accumulated
concerning the vaccine’s use in children and pregnant or breastfeeding women.
Cytomegalovirus
Cytomegalovirus (CMV) is one of the human herpesviruses. Congenital infection of infants,
postnatal infection of premature infants, and infection of immunodeficient individuals
represent the most serious forms of this infection in children. The time at which
the virus infects the fetus or infant and the presence or absence of antibodies against
CMV from the mother are important determinants of the severity of infection and the
likelihood of significant sequelae (congenital infection syndrome, deafness, chorioretinitis,
abnormal neurodevelopment, learning disabilities).
234
About 1% of all infants are born excreting CMV at birth, and approximately 5% of these
congenitally infected infants will demonstrate evidence of infection at birth (approximately
five symptomatic cases per 10,000 live births). Approximately 15% of infants born
after primary infection in a pregnant woman will manifest at least one sequela of
prenatal infection.
96
Various studies have detected that 3% to 28% of pregnant women have CMV in cervical
cultures and that 4% to 5% of pregnant women have CMV in their urine.120., 172. Perinatal
infection certainly occurs through contact with virus in these fluids but usually
is not associated with clinical illness in full-term infants. The lack of illness
is thought to result from transplacental passive transfer of protective antibodies
from the mother.
Postnatal infection later in infancy occurs via breastfeeding or contact with infected
fluids (e.g., saliva, urine) but, again, rarely causes clinical illness in full-term
infants. Seroepidemiologic studies have documented transmission of infection in infancy,
with higher rates of transmission occurring in daycare centers, especially when the
prevalence of CMV in the urine and saliva is high. CMV has been identified in the
milk of CMV-seropositive women at varying rates (10% to 85%) using viral cultures
or CMV deoxyribonucleic acid (DNA) PCR.172., 301., 397., 430. CMV is more often identified
in the breast milk of seropositive mothers than in vaginal fluids, urine, and saliva.
The CMV isolation rate from colostrum is lower than that from mature milk.172., 396.
The reason for the large degree of variability in identification of CMV in breast
milk in these studies probably relates to the intermittent nature of reactivation
and excretion of the virus in addition to the variability, frequency, and duration
of sampling of breast milk in the different studies. Some authors have hypothesized
that the difference in isolation rates between breast milk and other fluids is caused
by viral reactivation in cells (leukocytes or monocytes) in the breast leading to
“selective” excretion in breast milk.
301
Vochem et al
430
reported that the rate of virolactia was greatest at 3 to 4 weeks postpartum, and
Yeager et al
455
reported significant virolactia between 2 and 12 weeks postpartum. Antibodies (e.g.,
secretory IgA) to CMV are present in breast milk, along with various cytokines and
other proteins (e.g., lactoferrin). These may influence virus binding to cells, but
they do not prevent transmission of infection.∗
Several studies have documented increased rates of postnatal CMV infection in breastfed
infants (50% to 69%) compared with bottle-fed infants (12% to 27%) observed through
the first year of life120., 281., 397., 430. In these same studies, full-term infants
who acquired CMV infection postnatally were only rarely mildly symptomatic at the
time of seroconversion or documented viral excretion. Also, no evidence of late sequelae
from CMV was found in these infants.
Postnatal exposure of susceptible infants to CMV, including premature infants without
passively acquired maternal antibodies against CMV, infants born to CMV-seronegative
mothers, and immunodeficient infants, can cause significant clinical illness (pneumonitis,
hepatitis, thrombocytopenia).∗
In one study of premature infants followed up to 12 months, Vochem et al
430
found CMV transmission in 17 of 29 infants (59%) exposed to CMV virolactia and breastfed
compared with no infants infected of 27 exposed to breast milk without CMV. No infant
was given CMV-seropositive donor milk or blood. Five of the 12 infants who developed
CMV infection after 2 months of age had mild signs of illness, including transient
neutropenia, and only one infant had a short increase in episodes of apnea and a period
of thrombocytopenia. Five other premature infants with CMV infection before 2 months
of age had acute illness, including sepsis-like symptoms, apnea with bradycardia,
hepatitis, leukopenia, and prolonged thrombocytopenia.
430
Vollmer et al
431
followed premature infants with early postnatal CMV infection acquired through breast
milk for 2 to 4.5 years to assess neurodevelopment and hearing function. None of the
children had sensorineural hearing loss. There was no difference between the 22 CMV-infected
children and 22 matched premature control CMV-negative infants in terms of neurologic,
speech and language, or motor development.
431
Neuberger et al
296
examined the symptoms and neonatal outcome of CMV infection transmitted via human
milk in premature infants in a case-control fashion; 40 CMV-infected premature infants
were compared with 40 CMV-negative matched premature infants. Neutropenia, thrombocytopenia,
and cholestasis were associated with CMV infection in these infants. No other serious
effects or illnesses were found directly associated with the infection including intraventricular
hemorrhage, periventricular leukomalacia, retinopathy of prematurity, necrotizing
enterocolitis, bronchopulmonary dysplasia, duration of mechanical ventilation or oxygen
therapy, duration of hospital stay or weight, gestational age, or head circumference
at the time of discharge.
Exposure of CMV-seronegative or premature infants to CMV-positive milk (donor or natural
mother’s) should be avoided.
379
Various methods of inactivating CMV in breast milk have been reported, including Holder
pasteurization, freezing (−20° C for 3 days), and brief high temperature (72° C for
10 seconds).120., 135., 155., 393., 455. One small, prospective study suggests that
freezing breast milk at −20°C for 72 hours protects premature infants from CMV infection
via breast milk. Sharland et al
379
reported on 18 premature infants (less than 32 weeks) who were uninfected at birth
and exposed to breast milk from their CMV seropositive mothers. Only one of 18 (5%)
infants became positive for CMV at 62 days of life, and this infant was clinically
asymptomatic. This transmission rate is considerably lower than others reported in
the literature. CMVseronegative and leukocyte-depleted blood products were used routinely.
Banked breast milk was pasteurized and stored at −20° C for various time periods and
maternal expressed breast milk was frozen at −20° C before use whenever possible.
The infants received breast milk for a median of 34 days (range 11 to 74 days) and
they were observed for a median of 67 days (range 30 to 192 days). Breast milk samples
pre- or postfreezing were not analyzed by PCR or culture for the presence of cytomegalovirus.
379
Buxmann et al
57
demonstrated no transmission of CMV in 23 premature infants receiving thawed frozen
breast milk until 33 weeks (gestational age + postnatal age) (less than or equal to
31 weeks gestational age) born to 19 mothers who were CMV-IgG negative. CMV infection
was found in five premature infants of 35 infants born to 29 mothers who were CMV-IgG
positive and who provided breast milk for their infants. Three of the five children
remained asymptomatic. One child development a respirator-dependent pneumonia and
the second developed an upper respiratory tract infection and thrombocytopenia in
association with their CMV infections.
57
Yasuda et al
454
reported on 43 preterm infants (median gestational age 31 weeks) demonstrating a peak
in CMV DNA copies, detected by a real-time PCR assay, in breast milk at 4 to 6 weeks
postpartum. Thirty of the 43 infants received CMV DNA-positive breast milk. Three
of the 30 had CMV DNA detected in their sera, but none of the three had symptoms suggestive
of CMV infection. Much of the breast milk had been stored at −20° C before feeding,
which the authors propose is the probable reason for less transmission in this cohort.
454
Lee et al
248
reported on the use of maternal milk frozen at −20°C for a minimum of 24 hours before
feeding to premature infants in a NICU; 23 infants had CMV-seropositive mothers and
39 infants had CMV-seronegative mothers. Two infants developed CMV infection, which
was symptomatic. They were both fed frozen thawed milk from CMV-seropositive mothers.
248
Others have reported individual cases of CMV infection in premature infants despite
freezing and thawing breast milk.268., 314. Simple freezing and thawing of breast
milk does not completely prevent transmission of CMV to premature infants. The efficacy
of freezing and thawing breast milk for varying lengths of time to prevent CMV infection
in premature infants has not been studied prospectively in a randomized controlled
trial. Eleven of 36 neonatal units in Sweden (27 of which have their own milk banks)
freeze maternal milk to reduce the risk for CMV transmission to premature infants.
314
A prominent group of neonatologists and pediatric infectious disease experts in California
who recognize the significant benefits of providing human milk to premature and LBW
infants recommend screening mothers of premature infants for CMV IgG at delivery and,
when an infant’s mother is CMV IgG positive at delivery, using either pasteurized
banked human milk or frozen then thawed maternal breast milk for premature infants
until they reach the age of 32 weeks.
445
In consideration of the low rates of CMV virolactia in colostrum169., 397. and the
predominant occurrence of virolactia between 2 and 12 weeks (peak at 3 to 4 weeks)
postpartum,430., 455. they reasonably propose beginning colostrum and breast milk
feedings for all infants until the maternal CMV serologic screening is complete. They
appropriately recommend close observation and follow-up of premature infants older
than 3 weeks of age for signs, symptoms, and laboratory changes of CMV infection until
discharge from the hospital.
445
CMV-seropositive mothers can safely breastfeed their full-term infants because, despite
a higher rate of CMV infection than in formula-fed infants observed through the first
year of life, infection in this situation is not associated with significant clinical
illness or sequelae.
Dengue Disease
Dengue viruses (serotypes dengue 1 to 4) are flaviviruses associated primarily with
febrile illnesses and rash; dengue fever, dengue hemorrhagic fever, and dengue shock
syndrome. The mosquito Aedes aegypti is the main vector of transmission of dengue
virus in countries lying between latitudes 35 degrees north and 35 degrees south.
More than 2.5 billion people live in areas where transmission occurs; dengue virus
infects over 100 millionindividuals a year and casuses approximately 24,000 deaths
a year.159., 163. Although dengue hemorrhagic fever and dengue shock syndrome occur
frequently in children younger than 1 year of age, they are infrequently described
in infants younger than 3 months of age.
167
There are also differences in the clinical and laboratory findings of dengue virus
infection in children compared with adults.
222
Boussemart et al
49
reported on two cases of perinatal/prenatal transmission of dengue and discussed eight
additional cases in neonates from the literature. Prenatal or intrapartum transmission
of the same type of dengue as the mother was confirmed by serology, culture, or PCR.
Phongsamart et al
333
described three additional cases of dengue virus infection late in pregnancy and apparent
transmission to two of the three infants with passive acquisition of antibody in the
third infant. Sirinavin et al
386
reported on 17 cases in the literature of vertical dengue infection, all presenting
at less than 2 weeks of age, but no observations or discussion of breast milk or breastfeeding
as a potential source of infection were published. Watanaveeradej et al
439
presented an additional three cases of dengue infection in infants documenting normal
growth and development at follow-up at 12 months of age.
It has been postulated that more severe disease associated with dengue disease occurs
when an individual has specific IgG against the same serotype as the infecting strain
in a set concentration, leading to antibody-dependent enhancement of infection. The
presence of preexisting dengue serotype specific IgG in an infant implies either previous
primary infection with the same serotype, passive acquisition of IgG from the mother
(who had a previous primary infection with the same serotype), or perhaps acquisition
of specific IgG from breast milk. Watanaveeradej et al
439
documented transplacentally transferred antibodies against all four serotypes of dengue
virus in 97% of 2000 cord sera at delivery. Follow-up of 100 infants documented the
loss of antibodies to dengue virus over time with losses of 3%, 19%, 72%, 99%, and
100% at 2, 4, 6, 9,and 12 months of age, respectively.
No evidence is available in the literature about more severe disease in breastfed
infants compared with formula-fed infants. No interhuman transmission of dengue virus
in the absence of the mosquito vector and no evidence of transmission via breast milk
are known. Only one report of a factor in the lipid portion of breast milk, which
inhibits the dengue virus, is available, and no evidence for antibody activity against
dengue virus in human breast milk is known.
127
Breastfeeding during maternal or infant dengue disease should continue as determined
by the mother’s or infant’s severity of illness.
Epstein-Barr Virus
Epstein-Barr virus (EBV) is a common infection in children, adolescents, and young
adults. It is usually asymptomatic but most notably causes infectious mononucleosis
and has been associated with chronic fatigue syndrome, Burkitt lymphoma, and nasopharyngeal
carcinoma. Because EBV is one of the human herpesviruses, concern has been raised
about lifelong latent infection and the potential risk for infection to a fetus and
neonate from the mother. Primary EBV infection during pregnancy is unusual because
few pregnant women are susceptible.149., 189. Although abortion, premature birth,
and congenital infection from EBV are suspected, no distinct group of anomalies is
linked to EBV infection in fetus or neonate. Also, no virologic evidence of EBV as
the cause of abnormalities was found in association with suspected EBV infection.
Culturing of EBV from various fluids or sites is difficult. The virus is detected
by its capacity to transform B lymphocytes into persistent lymphoblastoid cell lines.
PCR and DNA hybridization studies have detected EBV in the cervix and in breast milk.
One study, which identified EBV DNA in breast milk cells in more than 40% of women
donating milk to a breast milk bank, demonstrated that only 17% had antibody to EBV
(only IgG, no IgM).
204
Another study examining serologic specimens from breastfed and bottle-fed infants
showed similar seroprevalence of EBV at 12 to 23 months of age (36/66 [54.5%] and
24/43 [55.8%]) in the breastfed and bottle-fed children, respectively.236
The question of the timing of EBV infection and the subsequent immune response and
clinical disease produced requires continued study. Differences exist among the clinical
syndromes that manifest at different ages. Infants and young children are asymptomatic,
have illness not recognized as related to EBV, or have mild episodes of illness, including
fever, lymphadenopathy, rhinitis and cough, hepatosplenomegaly, or rash. Adolescents
or young adults who experience primary EBV infection more often demonstrate infectious
mononucleosis syndrome or are asymptomatic. Chronic fatigue syndrome is more common
in adolescents and young adults. Burkitt lymphoma, observed primarily in Africa, and
nasopharyngeal carcinoma, seen in southeast Asia, where primary EBV infection usually
occurs in young children, are tumors associated with early EBV infection.
246
These tumors are related to “chronic” EBV infection and tend to occur in individuals
with persistently high antibody titers to EBV viral capsid antigen and early antigen.
The questions of why these tumors occur with much greater frequency in these geographic
areas and what cofactors (including altered immune response to infection associated
with coinfections, immune escape by EBV leading to malignancy, or increased resistance
to apoptosis secondary to EBV gene mutations) may contribute to their development
remain unanswered.21., 289.
It also remains unknown to what degree breast milk could be a source of early EBV
infection compared with other sources of EBV infection in an infant’s environment.
Similar to the situation of postnatal transmission of CMV in immunocompetent infants,
clinically significant illness rarely is associated with primary EBV infection in
infants. More data concerning the pathogenesis of EBV-associated tumors should be
obtained before proscribing against breastfeeding is warranted, especially in areas
where these tumors are common but the protective benefits of breastfeeding are high.
In areas where Burkitt lymphoma and nasopharyngeal carcinoma are uncommon, EBV infection
in mother or infant is certainly not a contraindication to breastfeeding.
Filoviridae
Marburg and Ebola viruses cause severe and highly fatal hemorrhagic fevers. The illness
often presents with nonspecific symptoms (conjunctivitis, frontal headache, malaise,
myalgia, bradycardia) and progresses with worsening hemorrhage to shock and subsequent
death in 50% to 90% of patients. Person-to-person transmission through direct contact,
droplet spread, or airborne spread is the common mode of transmission. However, the
animal reservoir or source of these viruses in nature for human infection has not
been identified. Attack rates in families are 5% to 16%.
330
No postexposure interventions have proved useful in preventing spread, and no treatment
other than supportive is currently available.
A recent report documented the presence of Ebola virus in numerous body fluids including
in breast milk. One acute breast milk sample on day 7 after the onset of illness and
a “convalescent” breast milk sample on day 15 from the same woman were positive for
Ebola virus by both culture and PCR testing.
30
In the same study, saliva remained virus positive for a mean of 16 days after disease
onset, urine was positive for a mean of 28 days, and semen for a mean of 43 days after
the onset of disease.
No information is available concerning the risk for transmission of these viruses
in breast milk or additional risks or benefits from breastfeeding. Contact precautions
are recommended for Marburg virus infections and contact and airborne precautions
for Ebola virus infection. Given the high attack and mortality rates, these precautions
should be carefully instituted and breastfeeding not allowed. If any other suitable
source of nutrition can be found for an infant, expressed breast milk should also
be proscribed for the infant of a mother with either of these infections for at least
3 weeks postrecovery.
Hepatitis in The Mother
The diagnosis of hepatitis in a pregnant woman or nursing mother causes significant
anxiety. The first issue is determining the etiology of the hepatitis, which then
allows for an informed discussion of risk to the fetus/infant. The differential diagnosis
of acute hepatitis includes (1) common causes of hepatitis, such as hepatitis A, B,
C, and D; (2) uncommon causes of hepatitis, such as hepatitis E and G, CMV, echoviruses,
enteroviruses, EBV, HSV, rubella, varicella-zoster virus, yellow fever virus; (3)
rare causes of hepatitis, such as Ebola virus, Junin virus, and Machupo virus (cause
hemorrhagic fever), Lassa virus, and Marburg virus; and (4) nonviral causes, such
as hepatotoxic drugs, alcoholic hepatitis, toxoplasmosis, autoimmune hepatitis, bile
duct obstruction, ischemic liver damage, Wilson disease, α1-antitrypsin deficiency,
and metastatic liver disease. The following sections focus on hepatitis viruses A
to G. Other infectious agents that can cause hepatitis are considered individually
in other sections. Box 13-2
provides hepatitis terminology.
BOX 13-2
Terminology for Hepatitis
Hepatitis A Virus (HAV)
IgM anti-HAV
Immunoglobulin M (IgM) antibody against HAV
HAV RNA
HAV ribonucleic acid
Hepatitis B Virus
HBsAg
Hepatitis B surface antigen
HBeAg
Hepatitis Be antigen
HBcAg
Hepatitis B core antigen
Anti-HBe
Antibody against hepatitis Be antigen
IgM anti-HBcAg
IgM antibody against hepatitis B core antigen
HBV DNA
HBV deoxyribonucleic acid
HBIG
Hepatitis B immunoglobulin
Hepatitis C Virus (HCV)
Anti-HCV
Antibody against HCV
HCV RNA
HCV ribonucleic acid
Hepatitis D Virus (HDV)
Anti-HDV
Antibody against HDV
Hepatitis E Virus (HEV)
HEV RNA
HEV ribonucleic acid
Hepatitis G Virus (HGV)
HGV RNA
HGV ribonucleic acid
TT Virus (TTV)
TTV DNA
TT virus deoxyribonucleic acid
Other
NANBH
Non-A, non-B hepatitis
ISG
Immune serum globulin
Martin et al
266
outline a succinct diagnostic approach to a patient with acute viral hepatitis and
chronic viral hepatitis (Figure 13-4, Figure 13-5
). The approach involves using the four serologic markers (IgM anti-hepatitis A virus,
hepatitis B surface antigen [HBsAg], IgM anti-HBcAg, anti-HCV) as the initial diagnostic
tests. Simultaneous consideration of other etiologies of acute liver dysfunction is
appropriate depending on a patient’s history. If the initial diagnostic tests are
all negative, subsequent additional testing for anti-hepatitis D virus (HDV), HCV
RNA, hepatis G virus (HGV) RNA, anti-hepatis E virus (HEV), or HEV RNA may be necessary.
If initial testing reveals positive HBsAg, testing for anti-HDV, HBeAg, and HBV DNA
is appropriate. These additional tests are useful in defining the prognosis for a
mother and the risk for infection to an infant. During the diagnostic evaluation,
it is appropriate to discuss with the mother or parents the theoretic risk for transmitting
infectious agents that cause hepatitis via breastfeeding. The discussion should include
an evaluation of the positive and negative effects of suspending or continuing breastfeeding
until the exact etiologic diagnosis is determined. The relative risk for transmission
of infection to an infant can be estimated and specific preventive measures provided
for the infant (Table 13-2
).
Figure 13-4
Diagnostic approach to patient with acute viral hepatitis. See Box 13-2 for definitions
of abbreviations.
(From Martin P, Friedman L, Dienstag J: Diagnostic approach. In Zuckerman A, Thomas
H, editors: Viral Hepatitis: Scientific Basis and Clinical Management, Edinburgh,
1993, Churchill Livingstone.)
Figure 13-5
Diagnostic approach to patient with chronic viral hepatitis. See Box 13-2 for definitions
of abbreviations.
(From Martin P, Friedman L, Dienstag J: Diagnostic approach. In Zuckerman A, Thomas
H, editors: Viral Hepatitis: Scientific Basis and Clinical Management, Edinburgh,
1993, Churchill Livingstone.)
TABLE 13-2
Viral Hepatitis in Association With Breastfeeding∗
Hepatitis
Virus
Identified in Breast Milk
Factors for Perinatal/ Postnatal Transmission
Prevention
Breastfeeding†
A
Picornaviridae (RNA)
?
Vertical transmission uncertain or rare
ISG
Limited evidence of transmission via breastfeeding or of serious disease in infants
HAV in pregnancy associated with premature birth
HAV vaccine
Breastfeeding OK after ISG and vaccine
B
Hepadnaviridae (DNA)
HBsAg
Increased risk for vertical transmission with HBeAg+, in countries where HBV is endemic,
or early in maternal infection, before Ab production
HBIG
Low theoretic risk
HBV DNA
HBV vaccine
Virtually no risk after HBIG and HBV vaccineBreastfeeding OK after HBIG and vaccine
C
Flavivirus (RNA)
HCV RNA detected
Increased risk when mother HIV+ and HCV+ or with increased HCV RNA titers
None
Positive theoretic riskInadequate data on relative riskBreastfeeding OK after informed
discussion with parents
Vertical transmission uncommon
D
Delavirdine (RNA negative strand, circular)
?
Requires coinfection/superinfection with HBV
None (except to prevent HBV infection, give HBIG/HBV vaccine)
Prevent HBV infection with HBIG and vaccine
Vertical transmission rare
Breastfeeding OK after HBIG and vaccine
E
Caliciviridae (RNA)
+
Severe disease in pregnant women (20% mortality)
ISG and subunit vaccine being tested
Usually subclinical infection in childrenBreastfeeding OK
G
Related to calicivirus and flavivirus (RNA)
?
Vertical transmission occurs
None
Inadequate data
TT
TT virus (DNA, circular, single stranded)
TTV DNA detected
Vertical transmission occurs
None
Inadequate data
∗
See Box 13-2 for abbreviations. Ab, Antibody; HIV, human immunodeficiency virus.
†
With any type of infectious hepatitis, discussion of what is known and not known concerning
transmission should be related to the mother/parents, and together an informed decision
can be made concerning breastfeeding.
Data from Committee on Infectious Diseases, American Academy of Pediatrics: Red Book:
Report of the Committee on Infectious Diseases, ed 26, Elk Grove, IL, 2003, American
Academy of Pediatrics.
Hepatitis A
Hepatitis A virus (HAV) is usually an acute self-limited infection. The illness is
typically mild, and generally subclinical in infants. Occasionally, HAV infection
is prolonged or relapsing, extending 3 to 6 months, and rarely it is fulminant, but
HAV infection does not lead to chronic infection. The incidence of prematurity after
maternal HAV infection is increased, but no evidence to date indicates obvious birth
defects or a congenital syndrome.372., 464. HAV infection in premature infants may
lead to prolonged viral shedding.
349
Transmission is most often person to person (fecal-oral), and transmission in food-borne
or water-borne epidemics has been described. Transmission via blood products and vertical
transmission (mother to infant) are rare.
440
Transmission in daycare settings has been clearly described.
Infection with HAV in newborns is uncommon and does not seem to be a significant problem.
The usual period of viral shedding and presumed contagiousness lasts 1 to 3 weeks.
Acute maternal HAV infection in the last trimester or in the postpartum period could
lead to infection in an infant. Symptomatic infection can be prevented by immunoglobulin
(Ig) administration, and 80% to 90% of disease can be prevented by Ig administration
within 2 weeks of exposure. HAV vaccine can be administered simultaneously with Ig
without affecting the seroconversion rate to produce rapid and prolonged HAV serum
antibody levels.
Transmission of HAV via breast milk has been implicated in one case report, but no
data exist on the frequency of isolating HAV from breast milk.
440
Because HAV infection in infancy is rare and usually subclinical without chronic disease
and because exposure has already occurred by the time the etiologic diagnosis of hepatitis
in a mother is made, no reason exists to interrupt breastfeeding with maternal HAV
infection. The infant should receive Ig and HAV vaccine, administered simultaneously.
Hepatitis B
Hepatitis B virus (HBV) infection leads to a broad spectrum of illness, including
asymptomatic seroconversion, nonspecific symptoms (fever, malaise, fatigue), clinical
hepatitis with or without jaundice, extrahepatic manifestations (arthritis, rash,
renal involvement), fulminant hepatitis, and chronic HBV infection. Chronic HBV infection
occurs in up to 90% of infants infected via perinatal and vertical transmission and
in 30% of children infected between 1 to 5 years of age. Given the increased risk
for significant sequelae from chronic infection (chronic active hepatitis, chronic
persistent hepatitis, cirrhosis, primary hepatocellular carcinoma), prevention of
HBV infection in infancy is crucial. Transmission of HBV is usually through blood
or body fluids (stool, semen, saliva, urine, cervical secretions).
94
Vertical transmission either transplacentally or perinatally during delivery has been
well described throughout the world. Vertical transmission rates in areas where HBV
is endemic (Taiwan and Japan) are high, whereas transmission to infants from HBV carrier
mothers in other areas where HBV carrier rates are low is uncommon.
399
Transmission of HBV to infants occurs in up to 50% of infants when the mothers are
acutely infected immediately before, during, or soon after pregnancy.
462
HBsAg is found in breast milk, but transmission by this route is not well documented.
Beasley
31
and Beasley et al
32
demonstrated that although breast milk transmission is possible, seroconversion rates
are no different between breastfed and nonbreastfed infants in a long-term follow-up
study of 147 HBsAg-positive mothers. Hill et al
176
followed 101 breastfed infants and 268 formula-fed infants born to women who were
chronically HBsAg positive. All infants received hepatitis B immunoglobulin at birth
and a full series of hepatitis B vaccine. None of the breastfed infants and nine of
the formula-fed infants were positive for HBsAg after completion of the HBV vaccine
series. Breastfeeding had occurred for a mean of 4.9 months (range 2 weeks to 1 year).
Transmission, when it does happen, probably occurs during labor and delivery. Another
report from China followed 230 infants born to HBsAg-positive women. The infants received
appropriate dosing and timing of HBIG and HBV vaccine. At 1 year of age, anti-HBs
antibody was present in 90.9% of the breastfed infants and 90.3% of the bottle-fed
infants.
437
Risk factors associated with immunoprophylaxis failure against vertical transmission
of HBV include HBeAg-seropositive mothers and elevated HBV DNA “viral loads” in the
mothers.
392
In 2009 the AAP Committee on Infectious Diseases stated that “that breastfeeding of
the infant by a HBsAg-positive mother poses no additional risk for acquisition of
HBV infection by the infant with appropriate administration of hepatitis B vaccine
and HBIG.”
96
Screening of all pregnant women for HBV infection is an essential first step to preventing
vertical transmission. Universal HBV vaccination at birth and during infancy, with
administration of hepatitis B immunoglobulin (HBIg) immediately after birth to infants
of HBsAg-positive mothers, prevents HBV transmission in more than 95% of cases. Breastfeeding
by HBsAg-positive women is not contraindicated, but immediate administration of HBIG
and HBV vaccine should occur. Two subsequent doses of vaccine should be given at appropriate
intervals and dosages for the specific HBV vaccine product. This decreases the small
theoretic risk for HBV transmission from breastfeeding to almost zero.
When acute peripartum or postpartum hepatitis occurs in a mother and HBV infection
is a possibility, with its associated increased risk for transmission to the infant,
a discussion with the mother or parents should identify the potential risks and benefits
of continuing breastfeeding until the etiology of the hepatitis can be determined.
If an appropriate alternative source of nutrition is available for the infant, breast
milk should be withheld until the etiology of the hepatitis is identified. HBIG and
HBV vaccine can be administered to the infant who has not already been immunized or
has no documented immunity against HBV.
400
If acute HBV infection is documented in a mother, breastfeeding can continue after
immunization has begun.
Hepatitis C
Acute infection with HCV can be indistinguishable from hepatitis A or B infection;
however, it is typically asymptomatic or mild. HCV infection is the major cause of
blood-borne non-A, non-B hepatitis (NANBH). Chronic HCV infection is reported to occur
70% to 85% of the time regardless of age at time of infection. Sequelae of chronic
HCV infection are similar to those associated with chronic HBV infection. Bortolotti
et al
47
described two groups of children with HCV infection whom they observed for 12 to 48
months. The first group of 14 children, who acquired HCV infection early in life,
presumably from their mothers, demonstrated biochemical evidence of liver disease
in the first 12 months of life. Two of these children subsequently cleared the viremia
and had normal liver function, an additional three children developed normal liver
function despite persistent HCV viremia, and the remaining children had persistent
viremia and abnormal liver function. The second group of 16 children, with chronic
HCV infection, remained free of clinical symptoms of hepatitis, but 10 (62%) of them
had mild alanine aminotransferase elevations, and 7 of the 16 (44%) who had liver
biopsies had histologic evidence of mild to moderate hepatitis.
The two commonly identified mechanisms of transmission of HCV are transfusions of
blood or blood products and IV drug use. However, other routes of transmission exist
because HCV infection occurs even in the absence of obvious direct contact with significant
amounts of blood. Other body fluids contaminated with blood probably serve as sources
of infection. Transmission through sexual contact occurs infrequently and probably
requires additional contributing factors, such as coinfection with other sexually
transmitted agents or high viral loads in serum and other body fluids. Studies of
transmission in households without other risk factors have demonstrated either low
rates of transmission or no transmission.
The reported rates of vertical transmission vary widely. In mothers with unknown HIV
status or known HIV infection, the rates of vertical transmission were 4% to 100%,
whereas the rates varied between 0% and 42% in known HIV-negative mothers.
113
These same studies suggest that maternal coinfection with HIV, HCV genotype, active
maternal liver disease, and the serum titer of maternal HCV RNA may be associated
with increased rates of vertical transmission.263., 307., 461. The correlation between
HCV viremia, the HCV viral load in a mother, and vertical transmission of HCV is well
documented.288., 355., 406., 456. The clinical significance and risk for liver disease
after vertical transmission of HCV are still unknown. The timing of HCV infection
in vertical transmission is also unknown. In utero transmission has been suggested
by some studies,
125
whereas intrapartum or postpartum transmission was proposed by Ohto et al
308
when they documented the absence of HCV RNA in the cord blood of neonates who later
became HCV RNA positive at 1 to 2 months of age. More recently, Gibb et al
150
reported two pieces of data supporting the likelihood of intrapartum transmission
as the predominant time of vertical transmission: (a) low sensitivity of PCR for HCV
RNA testing in the first month of life with a marked increase in sensitivity after
that for diagnosing HCV infection in infants and (b) a lower transmission risk for
elective cesarean delivery (without prolonged rupture of membranes) compared with
vaginal or emergency cesarean delivery.
150
Another group, McMenamin et al,
275
analyzed vertical transmission in 559 mother-infant pairs. The overall vertical transmission
rate was 4.1% (18/441), with another 118 infants not tested or lost to follow-up.
Comparison of the vertical transmission rate was no different for vaginal delivery
or emergency cesarean in labor versus planned cesarean (4.2% vs. 3.0%). This held
true even when mothers had hepatitis C RNA detected antenatally (7.2% vs. 5.3%). The
authors did not support planned cesarean delivery to decrease vertical transmission
of hepatitis C infection. No prospective, controlled trials of cesarean versus vaginal
delivery and the occurrence of vertical hepatitis C transmission are available.
The risk for HCV transmission via breast milk is uncertain. Anti-HCV antibody and
HCV RNA has been demonstrated in colostrum and breast milk, although the levels of
HCV RNA in milk did not correlate with the titers of HCV RNA in serum.36., 162., 256.,
355. Nevertheless, transmission of HCV via breastfeeding (and not in utero, intrapartum,
or from other postpartum sources) has not been proven in the small number infants
studied. Transmission rates in breastfed and nonbreastfed infants appear to be similar,
but various important factors have not been controlled, such as HCV RNA titers in
mothers, examination of the milk for HCV RNA, exclusive breastfeeding versus exclusive
formula feeding versus partial breastfeeding, and duration of breastfeeding.∗
Zanetti et al
461
documented the absence of HCV transmission in 94 mother-infant pairs when the mother
had only HCV (no HIV) infection and no transmission in 71 mother-infant pairs who
breastfed, including 23 infants whose mothers were seropositive for HCV RNA. Eight
infants in that study were infected with HCV, their mothers had both HIV and HCV,
and three of these eight infants were infected with both HIV and HCV. The HCV RNA
levels were significantly higher in the mothers coinfected with HIV compared with
those mothers with HCV alone.
Overall, the risk for HCV infection via breastfeeding is low, the risk for HCV infection
appears to be more frequent in association with HIV infection and higher levels of
HCV RNA in maternal serum, no effective preventive therapies (Ig or vaccine) exist,
and the risk for chronic HCV infection and subsequent sequelae with any infection
is high. It is therefore appropriate to discuss the theoretic risk for breastfeeding
in HCV-positive mothers with the mother or parents and to consider proscribing breast
milk when appropriate alternative sources of nutrition are available for the infants.
HIV infection is a separate contraindication to breastfeeding. Additional study is
necessary to determine the exact role of breastfeeding in the transmission of HCV,
including the quantitative measurement of HCV RNA in colostrum and breast milk, the
relative risk for HCV transmission in exclusively or partially breastfed infants versus
the risk in formula-fed infants, and the effect of duration of breastfeeding on transmission.
The current position of the CDC is that no data indicate that HCV virus is transmitted
through breast milk.
83
Therefore breastfeeding by a HCV-positive, HIV-negative mother is not contraindicated.
Infants born to HCV RNA-positive mothers require follow-up through 18 to 24 months
of age to determine infants’ HCV status, regardless of the mode of infant feeding.
Infants should be tested for alanine aminotransferase and HCV RNA at 3 months and
12 to 15 months of age. Alanine aminotransferase and anti-HCV antibody should be tested
at 18 to 24 months of age to confirm an infant’s status: uninfected, ongoing hepatitis
C infection, or past HCV infection.
Hepatitis D
Hepatitis D virus (HDV) is a defective RNA virus that causes hepatitis only in persons
also infected with HBV. The infection occurs as either an acute coinfection of HBV
and HDV or a superinfection of HBV carriers. This “double” infection results in more
frequent fulminant hepatitis and chronic hepatitis, which can progress to cirrhosis.
The virus uses its own HBV RNA (circular, negative-strand RNA) with an antigen, HDAg,
surrounded by the surface antigen of HBV, HBsAg. HDV is transmitted in the same way
as HBV, especially through the exchange of blood and body fluids. HDV infection is
uncommon where the prevalence of HBV is low. In areas where HBV is endemic, the prevalence
of HDV is highly variable. HDV is common in tropical Africa and South America as well
as in Greece and Italy but is uncommon in the Far East and in Alaskan Inuit despite
the endemic occurrence of HBV in these areas.
390
Transmission of HDV has been reported to occur from household contacts and, rarely,
through vertical transmission. No data are available on transmission of HDV by breastfeeding.
HDV infection can be prevented by blocking infection with HBV; therefore HBIG and
HBV vaccine are the best protection. In addition to HBIG and HBV vaccine administration
to the infant of a mother infected with both HBV and HDV, discussion with the mother
or parents should include the theoretic risk for HBV and HDV transmission through
breastfeeding. As with HBV, once HBIg and HBV vaccine have been given to the infant,
the risk for HBV or HDV infection from breastfeeding is negligible. Therefore breastfeeding
after an informed discussion with the parents is acceptable.
Hepatitis E
Hepatitis E virus (HEV) is a cause of sporadic and epidemic, enterically transmitted
NANBH, which is typically self-limited and without chronic sequelae. HEV is notable
for causing high mortality rate in pregnant women. Transmission is primarily via the
fecal-oral route, commonly via contaminated water or food. High infection rates have
been reported in adolescents and young adults (ages 15 to 40 years). Tomar
416
reported that 70% of cases of HEV infections in the pediatric population in India
manifest as acute hepatitis. Maternal-neonatal transmission was documented when the
mother developed hepatitis E infection in the third trimester. Although HEV was demonstrated
in breast milk, no transmission via breast milk was confirmed in the report. Five
cases of transfusion-associated hepatitis E were reported.
416
Epidemics are usually related to contamination of water. Person-to-person spread is
minimal, even in households and day care settings. Although Ig may be protective,
no controlled trials have been done. Animal studies suggest that a recombinant subunit
vaccine may be feasible.
344
HEV infection in infancy is rare, and no data exist on transmission of HEV by breastfeeding.
No evidence of clinically significant postnatal HEV infection in infants or of chronic
sequelae in association with HEV infection and no documented HEV transmission through
breast milk is available. Currently no contraindication exists to breastfeeding with
maternal HEV infection. Ig has not been shown to be effective in preventing infection,
and no vaccine is available for HEV.
Hepatitis G
Hepatitis G virus (HGV) has recently been confirmed as a cause of NANBH distinct from
hepatitis viruses A through E. Several closely related genomes of HGV, currently named
GBV-A, -B, and -C, appear to be related to HCV, the pestiviruses, and the flaviviruses.
Epidemiologically, HGV is most often associated with transfusion of blood, although
studies have identified nontransfusion-related cases. HGV genomic RNA has been detected
in some patients with acute and chronic hepatitis and a small number of patients with
fulminant hepatitis. GBV-C/HGV has also been found in some patients with inflammatory
bile duct lesions, but the pathogenicity of this virus is unconfirmed. HGV RNA has
been detected in 1% to 3% of healthy blood donors in the United States.
8
Feucht et al
128
described maternal-to-infant transmission of HGV in three of nine children. Two of
the three mothers were coinfected with HIV and the third with HCV. None of these infants
developed signs of liver disease. Neither the timing nor the mode of transmission
was clarified. Lin et al
255
reported no HGV transmission in three mother-infant pairs after cesarean delivery
and discussed transplacental spread via blood as the most likely mode of HGV infection
in vertical transmission. Wejstal et al
442
reported on perinatal transmission of HGV to 12 of 16 infants born to HGV viremic
mothers, identified by PCR. HGV did not appear to cause hepatitis in the children.
442
Fischler et al
130
followed eight children born to HGV-positive mothers and found only one to be infected
with HGV. That child remained clinically well, while his twin, also born by cesarean
delivery and breastfed, remained HGV negative for 3 years of observation. Five of
the other six children were breastfed for variable periods without evidence of HGV
infection. Ohto et al
309
examined HGV mother-to-infant transmission. Of 2979 pregnant Japanese women who were
screened, 32 were identified as positive for GBV-C/HGV RNA by PCR; 26 of 34 infants
born to the 32 HGV positive women were shown to be HGV RNA positive. Reportedly, none
of the infants demonstrated a clinical picture of hepatitis, although two infants
had persistent mild elevations (less than two times normal) of alanine aminotransferase.
The viral load in mothers, who transmitted HGV to their infants, was significantly
higher than in nontransmitting mothers. Infants born by elective cesarean delivery
had a lower rate of infection (3 in 7) compared with infants born by emergency cesarean
delivery (2 of 2) or born vaginally (21 of 25). In this study, HGV infection in breastfed
infants was four times more common than in formula-fed infants, but this difference
was not statistically significant because only four infants were formula fed. The
authors report no correlation between infection rate and duration of breastfeeding
was seen. Testing of the infants was not done frequently and early enough routinely
through the first year of life to determine the timing of infection in these infants.
309
Schröter et al
371
reported transmission of HGV to 3 of 15 infants born to HGV RNA positive mothers at
1 week of age. None of 15 breast milk samples were positive for GBV-C/HGV RNA, and
all of the children who were initially negative for HGV RNA in serum remained negative
at follow-up between 1 to 28 months of age.
371
The foregoing data suggest that transmission is more likely to be vertical, before,
or at delivery rather than via breastfeeding. The pathogenicity and the possibility
of chronic disease due to HGV infection remain uncertain at this time. Insufficient
data are available to make a recommendation concerning breastfeeding by HGV-infected
mothers.
Herpes Simplex Virus
Herpes simplex virus types 1 and 2 (HSV-1, HSV-2) can cause prenatal, perinatal, and
postnatal infections in fetuses and infants. Prenatal infection can lead to abortion,
prematurity, or a recognized congenital syndrome. Perinatal infection is the most
common form of infection (1 in 2000 to 5000 live births, 700 to 1500 cases per year
in the United States) and is often fatal or severely debilitating. The factors that
facilitate intrapartum infection and predict the severity of disease have been extensively
investigated. Postnatal infection is uncommon but can occur from a variety of sources,
including oral or genital lesions and secretions in mothers or fathers, hospital workers
and home caregivers, and breast lesions in breastfeeding mothers. A number of case
reports have documented severe HSV-1 or HSV-2 infections in infants associated with
HSV-positive breast lesions in the mothers.116., 161., 338., 403. Cases of infants
with HSV gingivostomatitis inoculating the mothers’ breasts have also been reported.
In the absence of breast lesions breastfeeding in HSV-seropositive or culture-positive
women is reasonable when accompanied by careful handwashing, covering the lesions,
and avoiding fondling or kissing with oral lesions until all lesions are crusted.
Breastfeeding during maternal therapy with oral or IV acyclovir can continue safely
as well. Inadequate information exists concerning valacyclovir, famciclovir, ganciclovir,
and foscarnet in breast milk to make a recommendation at this time. Breastfeeding
by women with active herpetic lesions on their breasts should be proscribed until
the lesions are dried. Treatment of the mothers’ breast lesions with topical, oral,
and/or IV antiviral preparations may hasten recovery and decrease the length of viral
shedding.
Human Herpesvirus 6 and Human Herpesvirus 7
Human herpesvirus 6 (HHV-6) is a cause of exanthema subitum (roseola, roseola infantum)
and is associated with febrile seizures. HHV-6 appears to be most similar to CMV based
on genetic analysis. No obvious congenital syndrome of HHV-6 infection has been identified,
although prenatal infection has been reported.
118
Seroepidemiologic studies show that most adults have already been infected by HHV-6.
Therefore primary infection during pregnancy is unlikely, but reactivation of latent
HHV-6 infection may be more common. No case of symptomatic HHV-6 prenatal infection
has been reported. The significance of reactivation of HHV-6 in a pregnant woman and
the production of infection and disease in the fetus and infant remains to be determined.
Primary infection in children occurs most often between 6 and 12 months of age, when
maternally acquired passive antibodies against HHV-6 are waning. Febrile illnesses
in infants younger than 3 months of age have been described with HHV-6 infection,
but infection before 3 months or after 3 years is uncommon.
Various studies involving serology and restriction enzyme analysis of HHV-6 isolates
from mother/infant pairs support the idea that postnatal transmission and perhaps
perinatal transmission from the mothers are common sources of infection. One study
was unable to detect HHV-6 in breast milk by PCR analysis in 120 samples, although
positive control samples seeded with HHV-6-infected cells did test positive.
119
Given the limited occurrence of clinically significant disease and the absence of
sequelae of HHV-6 infection in infants and children, the almost universal acquisition
of infection in early childhood (with or without breastfeeding) and the absence of
evidence that breast milk is a source of HHV-6 infection, breastfeeding can continue
in women known to be seropositive for HHV-6.
Human herpesvirus 7 (HHV-7) is closely related to HHV-6 biologically. Primary infection
with HHV-7 occurs primarily in childhood, usually later in life than HHV-6 infection.
The median age of infection is 26 months, with 75% of children becoming HHV-7 positive
by 5 years of age.
63
Seroprevalence of HHV-7 antibody has been reported to be 80% to 98% in adults, and
passive antibody is present in almost all newborns.306., 408. Like HHV-6, HHV-7 infection
can be associated with acute febrile illness, febrile seizures, and irritability,
but in general it is a milder illness than with HHV-6 with fewer hospitalizations.
Virus excretion of HHV-7 occurs in saliva, and PCR testing of blood cells and saliva
are frequently positive in individuals with past infection.
463
Congenital infection of HHV-6 was detected via DNA PCR testing in 57 of 5638 of cord
blood samples (1%), but HHV-7 was not detected in any of 2129 cord blood specimens.
165
HHV-7 DNA was detected by PCR in 3 of 29 breast milk mononuclear cell samples from
24 women who were serum positive for HHV-7 antibody.
137
In the same study, small differences were seen in the HHV-7 seropositive rates between
breastfed infants and bottle-fed infants at 12 months of age (21.7% versus 20%), at
18 months of age (60% versus 48.1%), and at 24 months of age (77.3% versus 58.3%,
respectively,). None of these differences were statistically significant. Given that,
in general, HHV-6 infection occurs earlier than HHV-7 infection in most infants and
that HHV-6 is rarely found in breast milk, it seems unlikely that HHV-7 in breast
milk is a common source of infection in infants and children. The infrequent occurrence
of significant illness with HHV-7 infection, with the absence of sequelae except in
patients who had transplantation surgery at older ages and the common occurrence of
infection in childhood argue, that no reason to proscribe against breastfeeding for
HHV-7 positive women exists.
Human Papillomavirus
Human papillomavirus (HPV) is a DNA virus with at least 100 different types. These
viruses cause warts, genital dysplasia, cervical carcinoma (types 6 and 11), and laryngeal
papillomatosis. Transmission occurs through direct contact and sexual contact. Laryngeal
papillomas are thought to result from acquiring the virus in passage through the birth
canal. Infection in pregnant women or during pregnancy does not lead to an increase
in abortions or the risk for prematurity, and no evidence indicates intrauterine infection.
HPV is one of the most common viruses in adults and one of the most commonly sexually
transmitted infections.
Diagnosis is usually by histologic examination or DNA detection. Spontaneous resolution
does occur, but therapy for persistent lesions or growths in anatomically problematic
locations is appropriate. Therapy can be with podophyllum preparations, trichloroacetic
acid, cryotherapy, electrocautery, and laser surgery. Interferon is being tested in
the treatment of laryngeal papillomas, with mixed results.
109
Prevention against transmission means limiting direct or sexual contact, but this
may not be sufficient because lesions may not be evident and transmission may still
occur.
Rintala et al
346
examined the occurrence of HPV DNA in the oral and genital mucosa of infants during
the first 3 years of life. HPV DNA was identified in 12% to 21% of the oral scrape
samples and in 4% to 15% of the genital scrape samples by PCR. Oral HPV infection
was acquired by 42% of children, cleared by 11%, and persisted in 10% of children;
37% of the children were never infected. They did not report on breast milk or breastfeeding
in that study. The question of the source of the infection remains undetermined.
The breast is a rare site of involvement.
110
HPV types 16 and 18 can immortalize normal breast epithelium in vitro.
441
HPV DNA has been detected in breast milk in 10 of 223 (4.5%) of milk samples from
223 mothers, collected 3 days postpartum.
361
No attempt was made to correlate the presence of HPV DNA in breast milk with the HPV
status of an infant or to assess the “viral load” of HPV in breast milk or its presence
over the course of lactation. A second study found DNA of cutaneous and mucosal HPV
types in 2 of 25 human milk samples and 1 of 10 colostrum samples.
64
No reports of HPV lesions of the breast or nipple and documented transmission to an
infant secondary to breastfeeding are available.
No increased risk for acquiring HPV from breast milk is apparent, and breastfeeding
is acceptable. Even in the rare occurrence of an HPV lesion of the nipple or breast,
no data suggest that breastfeeding or the use of expressed breast milk is contraindicated.
Measles
Measles is another highly communicable childhood illness that can be more severe in
neonates and adults. Measles is an exanthematous febrile illness following a prodrome
of malaise, coryza, conjunctivitis, cough, and often Koplik spots in the mouth. The
rash usually appears 10 to 14 days after exposure. Complications can include pneumonitis,
encephalitis, and bacterial superinfection. With the availability of vaccination,
measles in pregnancy is rare (0.4 in 10,000 pregnancies),
148
although respiratory complications (primary viral pneumonitis, secondary bacterial
pneumonia), hepatitis, or other secondary bacterial infections often lead to more
severe disease in these situations.
Prenatal infection with measles may cause premature delivery without disrupting normal
uterine development. No specific group of congenital malformations have been described
in association with in utero measles infection, although teratogenic effects of measles
infection in pregnant women may rarely manifest in the infants.
Perinatal measles includes transplacental infection when measles occurs in an infant
in the first 10 days of life. Infection from extrauterine exposure usually develops
after 14 days of life. The severity of illness after suspected transplacental spread
of virus to an infant varies from mild to severe and does not seem to vary with the
antepartum or postpartum onset of rash in the mother. It is uncertain what role maternal
antibodies play in the severity of an infant’s disease. More severe disease seems
to be associated with severe respiratory illness and bacterial infection. Postnatal
exposure leading to measles after 14 days of life is generally mild, probably because
of passively acquired antibodies from the mother. Severe measles in children younger
than 1 year of age may occur because of declining passively acquired antibodies and
complications of respiratory illness and rare cases of encephalitis.
Measles virus has not been identified in breast milk, whereas measles-specific antibodies
have been documented.
1
Infants exposed to mothers with documented measles while breastfeeding should be given
immunoglobulin (Ig) and isolated from the mother until 72 hours after the onset of
rash, which is often only a short period after diagnosis of measles in the mother.
The breast milk can be pumped and given to the infant because secretory IgA begins
to be secreted in breast milk within 48 hours of onset of the exanthem in the mother.
Table 13-3
summarizes management of the hospitalized mother and infant with measles exposure
or infection.
148
TABLE 13-3
Guidelines for Preventive Measures After Exposure to Measles in Nursery or Maternity
Ward
Type of Exposure or Disease
MEASLES (PRODROME OR RASH) PRESENT∗
Disposition
Mother
Neonate
A. Siblings at home have measles∗ when neonate and mother are ready for discharge
from hospital.
No
No
1. Neonate: Protective isolation and immunoglobulin (IG) indicated unless mother has
unequivocal history of previous measles or measles vaccination.†
2. Mother: With history of previous measles or measles vaccination, she may either
remain with neonates or return to older children. Without previous history, she may
remain with neonate until older siblings are no longer infectious, or she may receive
IG prophylactically and return to older children.
B. Mother has no history of measles or measles vaccination exposure 6 to15 days antepartum.‡
No
No
1. Exposed mother and infant: Administer IG to each and send home at earliest date
unless siblings at home have communicable measles. Test mothers for susceptibility
if possible. If susceptible, administer live measles vaccine 8 weeks after IG.
2. Other mothers and infants: Same unless clear history of previous measles or measles
vaccination in the mother.
3. Hospital personnel: Unless clear history of previous measles or measles vaccination,
administer IG within 72 hours of exposure. Vaccinate 8 weeks or more later.
C. Onset of maternal measles occurs antepartum or postpartum.§
Yes
Yes
1. Infected mother and infant: Isolate together until clinically stable, then send
home.
2. Other mothers and infants: Same as B-3 except infants should be vaccinated at 15
months of age.
3. Hospital personnel: Same as B-3.
D. Onset of maternal measles occurs antepartum or postpartum.§
Yes
No
1. Infected mother: Isolate until no longer infectious.§
2. Infected mother’s infant: Isolate separately from mother. Administer IG immediately.
Send home when mother is no longer infectious. Alternatively, observe in isolation
for 18 days for modified measles,¶ especially if IG administration was delayed more
than 4 days.
3. Other mothers and infants: Same as C-2.
4. Hospital personnel: Same as B-3.
∗
Catarrhal stage or less than 72 hours after onset of exanthem.
†
Vaccination with live attenuated measles virus.
‡
With exposure less than 6 days antepartum, mother would not be potentially infectious
until at least 72 hours postpartum.
§
Considered infectious from onset of prodrome until 72 hours after onset of exanthem.
¶
Incubation period for modified measles may be prolonged beyond the usual 10 to 14
days.
From Gershon AA: Chickenpox, measles and mumps. In Remington JS, Klein JO, editors:
Infectious Diseases of the Fetus and newborn infant, ed 4, Philadelphia, 1995, WB
Saunders.
Mumps
Mumps is an acute transient benign illness with inflammation of the parotid gland
and other salivary glands and often involves the pancreas, testicles, and meninges.
Mumps occurs infrequently in pregnant women (1 to 10 cases in 10,000 pregnancies)
and is generally benign. Mumps virus has been isolated from saliva, respiratory secretions,
blood, testicular tissue, urine, CSF in cases of meningeal involvement, and breast
milk. The period of infectivity is believed to be between 7 days before and 9 days
after the onset of parotitis, with the usual incubation period being 14 to 18 days.
Prenatal infection with the mumps virus causes an increase in the number of abortions
when infection occurs in the first trimester. A small increase in the number of premature
births was noted in one prospective study of maternal mumps infection.383 No conclusive
evidence suggests congenital malformations are associated with prenatal infection,
not even with endocardial fibroelastosis, as originally reported in the 1960s.
Perinatal mumps (transplacentally or postnatally acquired) has rarely if ever been
documented. Natural mumps virus has been demonstrated to infect the placenta and infect
the fetus, and live attenuated vaccine virus has been isolated from the placenta but
not from fetal tissue in women vaccinated 10 days before induced abortion. Antibodies
to mumps do cross the placenta.
Postnatal mumps in the first year of life is typically benign. No epidemiologic data
suggest that mumps infection is more or less common or severe in breastfed infants
compared with formula-fed infants. Although mumps virus has been identified in breast
milk and mastitis is a rare complication of mumps in mature women, no evidence indicates
that breast involvement occurs more frequently in lactating women. If mumps occurs
in the mother breastfeeding can continue because exposure has already occurred throughout
the 7 days before the development of symptoms in the mother and secretory IgA in the
milk may help to mitigate the symptoms in the infant.
Parvovirus
Human parvovirus B19 causes a broad range of clinical manifestations, including asymptomatic
infection (most frequent manifestation in all ages), erythema infectiosum (fifth disease),
arthralgia and arthritis, red blood cell (RBC) aplasia (less often decreased white
blood cells or platelets), chronic infection in immunodeficient individuals, and rarely
myocarditis, vasculitis, or hemophagocytic syndrome.
Vertical transmission can lead to severe anemia and immune-mediated hydrops fetalis,
which can be treated, if accurately diagnosed, by intrauterine transfusion. Inflammation
of the liver or CNS can be seen in the infant, along with vasculitis. If the child
is clinically well at birth, hidden or persistent abnormalities are rarely identified.
No evidence indicates that parvovirus B19 causes an identified pattern of birth defects.
Postnatal transmission usually occurs person to person via contact with respiratory
secretions, saliva, and rarely blood or urine. Seroprevalence in children at 5 years
of age is less than 5%, with the peak age of infection occurring during the school-age
years (5% to 40% of children infected). The majority of infections are asymptomatic
or undiagnosed seroconversions.
417
Severe disease, such as prolonged aplastic anemia, occurs in individuals with hemoglobinopathies
or abnormal RBC maturation. Attack rates have been estimated to be 17% to 30% in casual
contacts but up to 50% among household contacts. In one study of 235 susceptible pregnant
women, the annual seroconversion rate was 1.4%.
223
No reports of transmission to an infant through breastfeeding are available. Excretion
in breast milk has not been studied because of limitations in culturing techniques.
Rat parvovirus has been demonstrated in rat milk. The very low seroconversion rate
in young children and the absence of chronic or frequent severe disease suggest that
the risk for parvovirus infection via breast milk is not significant. The possibility
of antibodies against parvovirus or other protective constituents in breast milk has
not been studied. Breastfeeding by a mother with parvovirus infection is acceptable.
Polioviruses
Poliovirus infections (types 1, 2, and 3) cause a range of illness, with 90% to 95%
subclinical, 4% to 8% abortive, and 1% to 2% manifest as paralytic poliomyelitis.
A review by Bates
29
from 1955 of 58 cases of poliomyelitis in infants younger than 1 month of age demonstrated
paralysis or death in more than 70% and only one child without evidence of even transient
paralysis. More than half the cases were ascribed to transmission from the mothers,
although no mention was made of breastfeeding. Breastfeeding rates at the time were
approximately 25%.
Prenatal infection with polioviruses does cause an increased incidence of abortion.
Prematurity and stillbirth apparently occur more frequently in mothers who developed
paralytic disease versus inapparent infection.
188
Although individual reports of congenital malformations in association with maternal
poliomyelitis exist, no epidemiologic data suggest that polioviruses are teratogenic.
Also, no evidence indicates that live attenuated vaccine poliovirus given during pregnancy
is associated with congenital malformations.89., 170.
Perinatal infection has been noted in several case reports of infants infected in
utero several days before birth who had severe disease manifesting with neurologic
manifestations (paralysis) but without fever, irritability, or vomiting. Additional
case reports of infection acquired postnatally demonstrate illness more consistent
with poliomyelitis of childhood. These cases were more severe and involved paralysis,
which may represent reporting bias.
89
No data are available concerning the presence of poliovirus in breast milk, although
antibodies to poliovirus types 1, 2, and 3 have been documented.
270
In this era of increasing worldwide poliovirus vaccination, the likelihood of prenatal
or perinatal poliovirus infection is decreasing. Maternal susceptibility to poliovirus
should be determined before conception and poliovirus vaccine offered to susceptible
women. An analysis of the last great epidemic in Italy in 1958 was done using a population-based
case-control study.
336
In 114,000 births, 942 infants were reported with paralytic poliomyelitis. A group
of matched control subjects was selected from infants admitted to the hospital at
the same time. Using the dichotomous variable of never breastfed and partially breastfed,
75 never-breastfed infants were among the cases and 88 among the control group. The
authors determined an odds ratio of 4.2, with 95% confidence interval of 1.4 to 14,
demonstrating that the risk for paralytic poliomyelitis was higher in infants never
breastfed and lowest among those exclusively breastfed. Because by the time the diagnosis
of poliomyelitis is made in a breastfeeding mother, the exposure of the infant to
poliovirus from maternal secretions has already occurred, and because the breast milk
already contains antibodies that may be protective, no reason exists to interrupt
breastfeeding. Breastfeeding also does not interfere with successful immunization
against poliomyelitis with oral or inactivated poliovirus vaccine.
71
Retroviruses
Human T-Cell Leukemia Virus Type I
The occurrence of human T-cell leukemia virus type I (HTLV-I) is endemic in parts
of southwestern Japan,66., 105., 207., 450. the Caribbean, South America,
156
and sub-Saharan Africa. HTLV-I is associated with adult T-cell leukemia/lymphoma and
a chronic condition with progressive neuropathy. The progressive neuropathy is called
HTLV-I associated myelopathy or tropical spastic paraparesis.
136
Other illnesses have been reported in association with HTLV-I infection including
dermatitis, uveitis, arthritis, Sjögren syndrome in adults, and infective dermatitis
and persistent lymphadenitis in children. Transmission of HTLV-I occurs most often
through sexual contact, via blood or blood products, and via breast milk. Infrequent
transmission does occur in utero or at delivery and with casual or household contact.
291
Seroprevalence generally increases with age and varies widely in different regions
and in populations of different backgrounds. In some areas of Japan, seropositivity
can be as high as 12% to 16%, but in South America, Africa, and some Caribbean countries
the rates are 2% to 6%. In Latin America seropositive rates can be as high as 10%
to 25% among female sex workers or attendees to STD clinics.
156
In blood donors in Europe, the seroprevalence of HTLV-I has been reported at 0.001%
to 0.03%. The seroprevalence in pregnant women in endemic areas of Japan is as high
as 4% to 5% and in nonendemic areas as low as 0.1% to 1.0%. HTLV-1 is not a major
disease in the United States. In studies from Europe the seroprevalence in pregnant
women has been noted to be up to 0.6%. These pregnant women were primarily of African
or Caribbean descent.
138
HTLV-I antigen has been identified in breast milk of HTLV-I positive mothers.
220
Another report shows that basal mammary epithelial cells can be infected with HTLV-I
and can transfer infection to peripheral blood monocytes.
254
Human milk from HTLV-I positive mothers caused infection in marmosets.221., 453. HTLV-I
infection clearly occurs via breastfeeding and a number of reports document an increased
rate of transmission of HTLV-I to breastfed infants compared with formula-fed infants.∗
Ando et al12., 13. in two separate reports demonstrated a parallel decline in antibodies
against HTLV-I in both formula-fed and breastfed infants to a nadir at approximately
1 year of age and a subsequent increase in antibodies from 1 to 2 years of age. The
percentage of children seropositive at 1 year of age in the breastfed and formula-fed
groups, respectively, was 3.0% and 0.6%, at 1.5 years of age it was 15.2% and 3.9%,
and at 2 years of age it was 41.9% and 4.6%. A smaller group of children followed
through 11 to 12 years of age demonstrated no newly infected children after 2 years
of age and no loss of antibody in any child who was seropositive at 2 years of age.12.,
13.
Transmission of HTLV-I infection via breastfeeding is also clearly associated with
the duration of breastfeeding.407., 409., 446., 447. It has been postulated that the
persistence of passively acquired antibodies against HTLV-I offers some protection
through 6 months of life (Table 13-4
).
TABLE 13-4
HTLV-I Transmission Related to Duration of Breastfeeding
Author (Reference)
Duration (mo)
Seroconversion Rate (%)
Number of Children∗
Takahashi
407
≤6
4.4
4/90
≥7
14.4
20/139
(bottle-fed)
5.7
9/158
Takezaki
409
≤6
3.9
2/51
>6
20.3
13/64
Wiktor
446
<12
9.0
8/86
≥12
32.%
19/60
HTVL, Human T-cell leukemia virus.
∗
Number of children positive for HTLV-I over the number of children examined.
Other factors relating to HTLV-I transmission via breast milk have been proposed.
Yoshinaga et al
460
presented data on the HTLV-I antigen producing capacity of peripheral blood and breast
milk cells and showed an increased mother-to-child transmission rate when the mother’s
blood and breast milk produced large numbers of antigen-producing cells in culture.
460
Hisada et al
183
reported on 150 mothers and infants in Jamaica, demonstrating that a higher maternal
provirus level and a higher HTLV-I antibody titer were independently associated with
HTLV-I transmission to the infant. Ureta-Vidal et al
421
reported an increased seropositivity rate in children of mothers with a high proviral
load and elevated maternal HTLV-I antibody titers.
Various interventions have been proposed to decrease HTLV-I transmission via breastfeeding.
Complete avoidance of breastfeeding was shown to be an effective intervention by Hino
et al180., 181. in large population of Japanese in Nagasaki. Avoiding breastfeeding
led to an 80% decrease in transmission. Breastfeeding for a shorter duration is another
effective alternative. Ando et al
11
showed that freezing and thawing breast milk decreased the infectivity of HTLV-I.
Sawada et al
363
demonstrated in a rabbit model that HTLV-I immunoglobulin protected against HTLV-I
transmission via milk. It is reasonable to postulate that any measure that would decrease
the maternal provirus load or increase the anti-HTLV-I antibodies available to infants
might decrease the risk for transmission. The overall prevalence of HTLV-I infection
during childhood is unknown because the majority of individuals do not manifest illness
until much later in life. The timing of HTLV-I infection in a breastfeeding population
has been difficult to assess because of passively acquired antibodies from the mother
and issues related to testing. Furnia et al
139
estimated the time of infection for a cohort of 16 breastfed infants in Jamaica. The
estimated median time of infection was 11.9 months as determined by PCR compared with
the estimated time of infection, based on whole virus Western blot, of 12.4 months.
In areas where the prevalence of HTLV-I infection (in the United States, Canada, or
Europe) is rare, the likelihood that a single test for antibody against HTLV-I would
be a false positive test is high compared with the number of true positive tests.
Repeat testing is warranted in many situations.
66
Quantification of the antibody titer and the proviral load is appropriate in a situation
when mother-to-child transmission is a concern. A greater risk for progression to
disease in later life has not been shown for HTLV-I infection through breast milk,
but early-life infections are associated with the greatest risk for adult T-cell leukemia.
402
The mother and family should be informed about all these issues. If the risk for lack
of breast milk is not too great and formula is readily available and culturally acceptable,
then the proscription of breastfeeding, or at least a recommendation to limit the
duration of breastfeeding to 6 months or less, is appropriate to limit the risk for
HTLV-I transmission to the infant. Freezing and thawing breast milk before giving
it to an infant might be another reasonable intervention to decrease the risk for
transmission, although no controlled trials document the efficacy of such an intervention.
Neither Ig nor antiviral agents against HTLV-I are available at this time.
Human T-Cell Leukemia Virus Type II
Human T-cell leukemia virus type II (HTLV-II) is endemic in specific geographic locations,
including Africa, the Americas, the Caribbean, and Japan. Transmission is primarily
through intravenous drug use, contaminated blood products, and breastfeeding. Sexual
transmission occurs but its overall contribution to the prevalence of HTLV-II in different
populations remains uncertain. Many studies have examined the presence of HTLV-I and
II in blood products. PCR testing and selective antibody tests suggest that about
half of the HTLV seropositivity in blood donors is caused by HTLV-II.
HTLV-II has been associated with two chronic neurologic disorders similar to those
caused by HTLV-I, tropical or spastic ataxia.
258
A connection between HTLV-II and glomerulonephritis, myelopathy, arthritis, T-hairy
cell leukemia, and large granulocytic leukemia has been reported.
Mother-to-child transmission has been demonstrated in both breastfed and formula-fed
infants. It appears that the rate of transmission is greater in breastfed infants.∗
HTLV-II has been detected in breast milk.
174
Nyambi et al
304
reported that HTLV-II transmission did correlate with the duration of breastfeeding.
The estimated rate of transmission was 20%. The time to seroconversion (after the
initial loss of passively acquired maternal antibodies) for infected infants seemed
to range between 1 and 3 years of age.
304
At this time avoidance of breastfeeding and limiting the duration of breastfeeding
are the only two possible interventions with evidence of effectiveness for preventing
HTLV-II mother-to-child transmission.207
With the current understanding of retroviruses, it is appropriate in cases of documented
HTLV-II maternal infection to recommend avoiding or limiting the duration of breastfeeding
and provide alternative nutrition when financially practical and culturally acceptable.
Mothers should have confirmatory testing for HTLV-II and measurement of the proviral
load. Infants should be serially tested for antibodies to HTLV-II and have confirmatory
testing if seropositive after 12 to 18 months of age. Further investigation into the
mechanisms of transmission via breast milk and possible interventions to prevent transmission
should occur as they have for HIV-1 and HTLV-I.
Human Immunodeficiency Virus Type 1
Human immunodeficiency virus type 1 (HIV-1) is transmitted through human milk. Refraining
from breastfeeding is a crucial aspect of preventing perinatal HIV infection in the
United States and many other countries. The dilemma is the use of replacement feeding
versus breastfeeding in countries where breastfeeding provides infants with significant
protection from illness and death due to malnutrition or other infections.
Breastfeeding and HIV Transmission
The question of the contribution of breastfeeding in mother-to-child HIV-1 transmission
is not a trivial one when one considers the following:
1.
The World Health Organization (WHO) has estimated that 33.2 million people (estimate
range 30.6 to 36.1) were living with HIV-1 in 2007.
419
2.
More than 90% of children younger than 13 years old infected with HIV-1 have been
infected by mother-to-child transmission. The number of children estimated to be living
with HIV increased to 2 million in 2007 (estimate range 1.9 to 2.3 million).
419
3.
The WHO estimates that 2.7 million people were newly infected with HIV-1 in 2007,
with children younger than 14 years old making up 370,000 of that 2.7 million. (This
number has declined due to increasing access to interventions to prevent mother-to-infant
transmission. Availability of antiretroviral therapy for prevention of mother-to-child
HIV transmission in developing countries in 2007 was estimated to reach 33% of the
mothers who needed it.)
419
4.
Breastfeeding contributes an estimated 10% to 20% increase in the overall mother-to-child
transmission rates, over and above intrauterine and intrapartum transmission, when
no specific interventions to prevent transmission via breastfeeding are utilized.
5.
Despite a dramatic increase in the number of people receiving antiretroviral therapy
in developing countries (3 million), this represented only 31% of the individuals
who needed treatment.
419
The evidence of HIV transmission via breastfeeding is irrefutable. Multiple publications
summarize the current evidence for HIV transmission via breastfeeding in the literature.232.,
341., 420. Since 1985, case reports have documented HIV transmission via breast milk
to children around the world.182., 198., 249., 465. Primary HIV infection in breastfeeding
mothers, with the concomitant high viral load, is associated with a particularly high
rate of HIV transmission via breast milk. Palasanthiran et al
322
estimated that risk at 27%.Large observational studies have demonstrated higher rates
of HIV transmission in breastfed infants of mothers with chronic HIV infection compared
with formula-fed infants.43., 108., 124. A systematic analysis of published reports
estimated the additional risk for perinatal HIV transmission due to breastfeeding
to be 14% (95% confidence interval 7% to 22%).
117
More recently published cohort studies similarly attributed additional risk for HIV
transmission due to breastfeeding at 4% to 22% over and above the risk from prenatal
and intrapartum transmission.38., 104., 121. Laboratory reports demonstrate the presence
of cell-free virus and cell-associated virus in breast milk as well as various immunologic
factors that could block or limit infection.∗
A dose-response relationship has been observed, correlating the HIV viral load in
human milk as well as a mother’s plasma viral load with an increased transmission
risk for the breastfed infant.335., 345., 351., 373.
Many of the potential risk factors associated with human milk transmission of HIV
have been described. The cumulative risk for HIV transmission is higher the longer
the duration of breastfeeding.108., 251., 282., 290., 424. Maternal characteristics
related to transmission of HIV via human milk include younger maternal age, higher
parity, lower CD4+ counts, higher plasma viral loads, and breast abnormalities (mastitis,
abscess, or nipple lesions). Characteristics of human milk that relate to a higher
risk for transmission include higher viral load in the milk, lower concentrations
of antiviral substances (lactoferrin, lysozyme), and lower concentrations of virus-specific
cytotoxic T-lymphocytes, levels of various interleukins (IL-7, IL-15),434., 435. secretory
IgA, and IgM. Mixed breastfeeding is also associated with a higher risk for HIV transmission
compared with exclusive breastfeeding.99., 100., 410. The measurable benefits of breast
milk versus the relative risk for HIV transmission to the infant due to exclusive
breastfeeding (with optimization of other factors to decrease HIV transmission) have
been reported in a couple of studies.97., 229. The measurable benefits of receiving
breast milk versus the relative increased risk for HIV transmission will need to be
determined in a prospective fashion in different locales.
247
Interventions to Prevent Breastfeeding-Related Transmission
A number of potential interventions to prevent breastfeeding transmission of HIV-1
can be utilized (Box 13-3
). The simplest and most effective is the compete avoidance of human milk. This is
a practical solution in places like the United States and other countries where replacement
feeding as well as other strictly medical interventions are feasible and reasonable,
and the risk for not providing breast milk to the infant is negligible. In resource-poor
situations, where the risk for other infections is high without the benefits of breast
milk, exclusive breastfeeding is appropriate, with any other reasonable and culturally
acceptable interventions to decrease HIV transmission via breast milk.
BOX 13-3
Recommendations on Breastfeeding and Transmission of Human Immunodeficiency Virus
(HIV)
•
Women and their health care providers need to be aware of the potential risk for transmission
of HIV infection to infants during pregnancy and in the peripartum period and through
breast milk.
•
Documented, routine HIV education and routine testing with the consent of women seeking
prenatal care are strongly recommended so that each woman knows her HIV status and
the methods available both to prevent the acquisition and transmission of HIV and
to determine whether breastfeeding is appropriate.
•
At delivery, education about HIV and testing with the consent of women whose HIV status
during pregnancy is unknown are strongly recommended. Knowledge of a woman’s HIV status
assists in counseling on breastfeeding and helps each woman understand the benefits
to herself and her infant of knowing her serostatus and the behaviors that would decrease
the likelihood of acquisition and transmission of HIV.
•
Women who are known to have HIV infections must be counseled not to breastfeed or
provide their milk for the nutrition of their own or other’s infants.
•
In general, women who are known to be HIV seronegative should be encouraged to breastfeed.
However, women who are HIV seronegative but at particularly high risk for seroconversion
(e.g., injection drug users and sexual partners of known HIV-positive persons or active
drug users) should be educated about HIV with an individualized recommendation concerning
the appropriateness of breastfeeding. In addition, during the perinatal period, information
should be provided on the potential risk for transmitting HIV through breast milk
and about methods to reduce the risk for acquiring HIV infection.
•
Each woman whose HIV status is unknown should be informed of the potential for HIV-infected
women to transmit HIV during the peripartum period and through breast milk and the
potential benefits to her and her infant of knowing her HIV status and how HIV is
acquired and transmitted. The health care provider needs to make an individualized
recommendation to assist the woman in deciding whether to breastfeed.
•
Neonatal intensive care units should develop policies that are consistent with these
recommendations for the use of expressed breast milk for neonates. Current standards
of the U.S. Occupational Safety and Health Administration (OSHA) do not require gloves
for the routine handling of expressed human milk. However, health care workers should
wear gloves in situations in which exposure to breast milk might be frequent or prolonged,
such as in milk banking.
•
Human milk banks should follow the guidelines developed by the U.S. Public Health
Service, which include screening all donors for HIV infection, assessing risk factors
that predispose to infection, pasteurizing all milk specimens.
From Lawrence RA: A review of the medical benefits and contraindications to breastfeeding
in the United States. In Maternal and Child Health Technical Information Bulletin,
Washington, DC, 1997, U.S. Health Resources and Services Administration.
Potentially effective interventions include exclusive breastfeeding, early weaning
versus breastfeeding for longer durations, education, and support to decrease the
likelihood of mastitis or nipple lesions.
191
Other possible interventions include treating a mother with antiretroviral therapy
for her own health (CD4 counts less than 350) or prophylactically to decrease the
human milk viral load, treating an infant prophylactically for a prolonged period
of time (6 weeks to 6 months) to protect against transmission via breastfeeding, treating
the milk itself to decrease the viral load (by pasteurization or other methods),316.,
318. treating acute conditions in mothers and infants (e.g., mastitis, breast lesions,
infant candidiasis), and enhancing an infant’s own defenses via vitamins, immunization,
or antiretroviral therapy. Some of these may not be feasible in certain settings such
as pasteurization or maternal antiretroviral therapy. Others may not be culturally
acceptable, such as treating expressed breast milk before giving it to an infant or
even exclusive breastfeeding.
Advantages of Breastfeeding
Significant data demonstrate the advantage of breastfeeding, even for HIV-infected
or HIV-exposed infants. The complete avoidance of breastfeeding in certain situations
may lead to increased risk for illness and death due to other reasons besides HIV
transmission.
106
A study from Kenya showed improved HIV-1-free survival rates in a formula-fed group
of children born to HIV-positive mothers, but the breastfed and formula groups had
similar mortality rates (24.4% versus 20.0%, respectively) and similar incidences
of diarrhea and pneumonia in the first 2 years of life.
272
No difference in the two groups was seen in the prevalence of malnutrition, but the
breastfed infants had better nutritional status in the first 6 months of life. Arpadi
et al
20
recommend additional nutritional interventions to complement breastfeeding in this
population after 6 months of age. Two reports from Zambia document the benefit of
exclusive breastfeeding for decreasing late HIV transmission and the lower mortality
at 12 months in infants who had continued breastfeeding rather than had discontinued
breastfeeding at 4 months of age.229., 385. In Malawi, HIV-infected and HIV-exposed
infants who were breastfed (exclusive breastfeeding for 2 months and mixed feeding
after that) had lower mortality at 24 months than those who were not breastfed.
405
A report from Botswana examined breastfeeding plus infant zidovudine prophylaxis for
6 months versus formula feeding plus infant zidovudine for 1 month; this study showed
a decreased risk for vertical transmission with formula feeding, but also increased
cumulative mortality for the HIV-infected infants at 7 months of age who were in the
formula-fed group.
411
A study from South Africa examining the use of vitamin A also demonstrated less morbidity
in HIV-infected children who were breastfed than not breastfed.
102
Other abstract reports have shown increased morbidity in HIV-infected children due
to diarrhea, gastroenteritis, and hospitalization after weaning from breastfeeding.205.,
226., 315., 413.
Exclusive Breastfeeding
Exclusive breastfeeding in most areas of the world is essential to infant health and
survival, even in the situation of maternal HIV infection.97., 99., 100., 229. The
duration of exclusive breastfeeding is crucial to decreasing the risk for HIV infection
in infants versus the risk for malnutrition and other infections with early weaning.
In the Mashi Study in Botswana, Thior et al
411
evaluated infants randomized to breastfeeding plus infant zidovudine for 6 months
or formula feeding plus 1 month of infant zidovudine. The cumulative infant mortality
was significantly higher at 7 months for the formula-fed group but at 18 months it
was similar between the two groups. The breastfed infants were more likely to become
HIV infected despite the 6 months of zidovudine prophylaxis.
411
Becquet et al
33
analyzed data from Cote d’Ivoire for 2001 to 2005; 47% of the HIV-exposed infants
were breastfed for a median of 4 months, and 53% were formula fed and observed for
2 years. No significant difference in the rate of HIV infection was seen in the two
groups, and no significant difference between the two groups was seen for morbid events
(diarrhea, acute respiratory infections or malnutrition) or hospitalization or death.
The authors attributed these good outcomes to effective nutritional counseling and
care, access to clean water, and the provision of a safe and continuous supply of
breast milk substitute.
33
Coovadia et al
97
studied exclusive breastfeeding in the first 6 months of life as an intervention in
South Africa. Of the exclusively breastfed infants, 14.1% at 6 weeks of age and 19.5%
at 6 months of age were HIV infected. Breastfed infants who also were fed solids or
formula milk were more likely to acquire infection than exclusively breastfed infants.
The cumulative mortality at 3 months of age was markedly lower for exclusively breastfed
infants (6.1%) versus 15.1% in the infants receiving mixed feedings.
Early Weaning
Kuhn et al
230
examined the effects of early, abrupt weaning on HIV-free survival of 958 children
in Zambia. Infants were randomly assigned to two different counseling programs that
advised either abrupt weaning at 4 months or prolonged breastfeeding. In the weaning
intervention group, 69% of mothers stopped breastfeeding by 5 months compared with
a median duration of breastfeeding of 16 months in the control group. The study found
no significant difference in HIV-free survival at 24 months in the two groups (83.9%
versus 80.7%). Children already infected by 4 months of age had a higher mortality
if they were assigned to the early weaning group (73.6% versus 54.8%). Additional
analysis showed that in mothers with less severe HIV disease early weaning was clearly
harmful to the infant.
231
Arpadi et al
20
studied the growth of HIV-exposed, uninfected children who were exclusively breastfed
for 4 months with rapid weaning to replacement foods or exclusively breastfed until
6 months and then continued breastfeeding with complementary foods. Weight-for-age
z scores dropped markedly in both groups from 4 to 15 months of age but less so in
the continued breastfeeding group. Length-for-age z score also dropped dramatically,
but was not influenced by continued breastfeeding. Even in this HIV-exposed, uninfected
group of children, additional nutritional interventions are essential to complement
breastfeeding beyond 6 months of age.
20
Antiretroviral Prophylaxis With Breastfeeding
In recent years the discussion around preventing HIV transmission via breastfeeding
and in the number of studies examining the important issues have increased.98., 233.,
283. The fact that intrapartum and perinatal transmission of HIV from mothers to infants
has decreased markedly due to the increased utilization of antiretroviral therapy
during pregnancy, delivery, and postnatally for prevention emphasizes the importance
of now working harder to decrease breast milk transmission of HIV. In considering
different possible interventions to decrease mother-infant HIV transmission, it is
crucial to reemphasize the goals of optimizing maternal health and survival and optimizing
infant health and survival at the same time.
A laboratory report shows that mothers receiving highly active antiretroviral therapy
(HAART) while breastfeeding do have decreased whole breast milk HIV-1 viral loads
(23/26 mothers had less than 50 copies/mL) compared with mothers who did not receive
HAART (9/25 with less than 50 copies/mL). However, the whole milk HIV-1 DNA load (measured
as “undetectable” at less than 10 copies/106 cells) was not significantly different
in the HAART (13 of 26 mothers)] and non-HAART (15 of 23) groups.
378
HIV-1 DNA is incorporated into cells found in breast milk. Another group showed significantly
lower HIV RNA levels in the breast milk of women treated with nevirapine, zidovudine,
and lamivudine compared with women not receiving antiretroviral therapy.
152
The use of maternal HAART seems to decrease HIV-1 transmission via breastfeeding.
One group working in Mozambique, Malawi, and Tanzania working with mother-infants
pairs receiving HAART as prevention during pregnancy compared one cohort (809 mother-infant
pairs) who received supplementary formula and water filters for the first 6 months
of life with a second cohort (251 mother-infant pairs) breastfeeding exclusively and
the mothers receiving HAART for the first 6 months. The cumulative incidence rate
of HIV infection at 6 months of age was 2.7% for the formula-fed infants and 2.2%
for breastfed infants. Through 6 months of age no apparent additional risk for late
postnatal transmission of HIV was observed.
323
The Petra study team working in Tanzania, South Africa, and Uganda examined the efficacy
of three short-course regimens of zidovudine and lamivudine in preventing early and
late HIV transmission in this predominantly breastfeeding population.
332
There were four regimens: A, zidovudine and lamivudine starting at 36 weeks’ gestation
plus intrapartum medication and 7–days’ postpartum treatment; B, same as A without
the prepartum component; C, intrapartum zidovudine and lamivudine only; D, placebo.
At week 6 the HIV transmission rates were 5.7% in group A, 8.9% in group B, 14.2%
in group C, and 15.3% in group D. At 18 months the HIV infection rates were 15% in
group A, 18% in group B, 20% in group C, and 22% in group D. Although a measurable
decrease in transmission at 6 weeks of age was observed, limited protection was seen
at 18 months of age. An observational study from Tanzania compared maternal HAART
for 6 months with exclusive breastfeeding and abrupt weaning at 5 to 6 months of age
with a historical control of the same feeding schedule without the postnatal maternal
HAART. In the treatment group the cumulative HIV transmission was 4.1% at 6 weeks,
5.0% at 6 months, and 6.0% at 18 months of age. The cumulative HIV infection or death
rate was 8.6% at 6 months and 13.6% at 18 months of age. The cumulative risk for HIV
transmission was 1.1% between 6 and 18 months. The HIV transmission in this treatment
group was half the transmission rate in the historical control group.
218
Another study in sub-Saharan Africa with 6 months of maternal HAART and exclusive
breastfeeding for 6 months demonstrated 94% HIV-free survival at 12 months of age;
the maternal and infant mortality rates for the treated mother-infant pairs were significantly
lower than the country’s maternal and infant mortality rates.
264
Antiretroviral therapy prophylaxis for infants is another investigated intervention
to decrease HIV transmission via breastfeeding. In a study from Cote d’Ivoire comparing
different groups over time, infants received zidovudine (ZDV) alone as ZDV prophylaxis,
a single dose of nevirapine (NVP) and 7 days of zidovudine (ZDV) as NVP/ZDV prophylaxis,
or a single dose of nevirapine plus zidovudine and lamivudine (3TC) for 7 days as
NVP/ZDV/3TC prophylaxis. Formula feeding (FF) was compared with exclusive shortened
breastfeeding (ESB) upto 4 months of age and prolonged breastfeeding (PB). The cumulative
transmission rates at 18 months were 22.3% in 238 infants in the ZDV + PB group, 15.9%
in 169 infants in the NVP/ZDV + ESB group, 9.4%, in the 195 infants in the NVP/ZDV
+FF group, 6.8% in the 198 infants in the NVP/ZDV/3TC + ESB group, and 5.6% in the
126 infants in the NVP/ZDV/3TC + FF group.
252
Kumwenda et al
235
working in Malawi demonstrated decreased HIV transmission with breastfeeding and two
different infant prophylaxis regimens. At 9 months of age, they observed a 10.6% occurrence
of HIV transmission for infants receiving a single dose of nevirapine plus 1 week
of zidovudine compared with 5.2% in the group receiving a single dose of nevirapine
plus 1 week of zidovudine plus 14 weeks of daily nevirapine, and 6.4% in the group
receiving a single dose of nevirapine plus 1 week of zidovudine plus 14 weeks of nevirapine
and zidovudine.
235
In the Mitra Study in Tanzania in which the median time of breastfeeding was 18 weeks,
the HIV transmission rate at 6 months in the infants who received zidovudine plus
3TC for 1 week plus 3TC alone for breastfeeding through 6 months of age was less than
50% of the transmission rate for those infants receiving only 1 week of zidovudine
plus 3TC.
217
A summary of three trials in Ethiopa, India, and Uganda compared a single dose of
nevirapine at birth for infants with 6 weeks of daily nevirapine in predominantly
breastfed infants whose mothers were counselled regarding feeding per the WHO/UNICEF
guidelines. At 6 months 87 of 986 infants in the single-dose group and 62 of 901 in
the extended-dose group were HIV infected, which was not statistically significant.
The authors suggested that a longer course of infant antiretroviral prophylaxis might
be more effective.
388
Human Immune Deficiency Virus in Maternal Health and Breastfeeding
The potential effect of breastfeeding on the HIV-positive mother needs to be adequately
assessed in relation to the mother’s health status. From Uganda and Zimbabwe Mbizvo
et al
271
reported no difference in the number of hospital admissions or mortality between HIV-positive
and HIV-negative women during pregnancy. In the 2 years after delivery the HIV-positive
women had higher hospital admission (approximately two times increased risk) and death
rates (relative risk greater than 10) than HIV-negative women.
271
Chilongozi et al
90
reported on 2292 HIV-positive mothers from four sub-Saharan sites followed for 112
months. Serious adverse events occurred in 166 women (7.2%); 42 deaths occurred in
the HIV-positive women, and no deaths occurred in 331 HIV-negative women.
90
Several studies have examined breastfeeding relative to mothers’ health and reported
conflicting results. The first study from Kenya demonstrated a significantly higher
mortality rate in breastfeeding mothers compared with a formula-feeding group in the
2 years after delivery. The hypothesized explanation offered by the authors for this
difference was increased metabolic demands, greater weight loss, and nutritional depletion.
294
A second study from South Africa showed an overall lower mortality rate in the two
groups with no significant difference in mortality rate in the 10 months of observation.
101
Kuhn et al
227
reported no difference in mortality at 12 months after delivery between 653 women
randomly assigned to a short breastfeeding group (326 women, median breastfeeding
duration 4 months, 21% still breastfeeding at 12 months) and a long breastfeeding
group (327 women, 90% breastfeeding at 5 months, 72% breastfeeding at 12 months, median
15 months). The HIV-related mortality rates were high (4.9%), but not associated with
prolonged lactation.
227
Walson et al
433
followed 535 HIV-positive women for 1 to 2 years in Kenya. The mortality risk was
1.9% at 1 year and 4.8% at 2 years of follow-up. Although less than 10% of women reported
a hospitalization during the 2 years, they experienced various common infections (pneumonia,
diarrhea, TB, malaria, STDs, urinary tract infections, mastitis). Breastfeeding was
a significant cofactor for diarrhea and mastitis but not for pneumonia, TB, or hospitalization.
433
In summary, breastfeeding of infants by HIV-positive mothers does lead to an increased
risk for HIV infection in the infants. Much remains to be understood about the mechanisms
of HIV transmission via breast milk and the action and efficacy of different interventions
to prevent such transmission. The complete avoidance of breastfeeding is a crucial
component for the prevention of perinatal HIV infection in the United States and many
other countries.
In resource-poor settings, where breastfeeding is the norm and where it provides vital
nutritional and infection protective benefits, the WHO, UNICEF, and the Joint United
Nations Programme on HIV/AIDS (UNAIDS) recommend education, counseling, and support
for HIV-infected mothers so they can make an informed choice concerning infant feeding.
Mothers choosing to breastfeed should receive additional education, support, and medical
care to minimize the risk for HIV transmission and to optimize their own health status
during and after breastfeeding. Mothers choosing to use replacement feedings should
receive parallel education, support, and medical care for themselves and their infants
to minimize the effect of the lack of breastfeeding.
Good evidence now shows that antiretroviral prophylactic regimens for mothers or infants
while continuing breastfeeding does decrease postnatal HIV transmission. Early weaning
is associated with increased morbidity and mortality. Further carefully controlled
research is indicated to adequately assess the risks and benefits to infants and mothers
of prolonged breastfeeding with antiretroviral prophylaxis for either or both mothers
and infants. Along with this, HIV testing rates must be improved at the same time
as increased availability and access to antenatal care, HIV prevention services, and
HIV medical care for everyone must be increased. The availability and free access
to antiretroviral medications must also improve.
The decision about infant feeding for HIV-positive mothers remains a difficult one,
but this is slowly changing with increasing options. The goals remain 100% HIV transmission
prevention, optimal maternal health and survival, and long-term infant health and
survival.
Human Immunodeficiency Virus Type 2
Human immunodeficiency virus type 2 (HIV-2) is an RNA virus in the nononcogenic, cytopathic
lentivirus genus of retroviruses. It is genetically closer to simian immunodeficiency
virus than to HIV-1. The clinical disease associated with HIV-2 has similar symptoms
to HIV-1 infection but progresses at a slower rate to severe immunosuppression.
HIV-2 is endemic in western Africa and parts of the Caribbean and found infrequently
in Europe and North and South America.190., 305. It is transmitted via sexual contact,
blood, or blood products and from mother to child.
Routine testing for HIV-2 is recommended in blood banks. Antibody tests used for HIV-1
are only 50% to 90% sensitive for detecting HIV-2.
65
Specific testing for HIV-2 is appropriate whenever clinically or epidemiologically
indicated.
Vertical transmission occurs infrequently. Ekpini et al
121
followed a large cohort of west African mothers and infants: 138 HIV-1 positive women,
132 HIV-2 positive women, 69 women seropositive for both HIV-1 and 2, and 274 HIV
seronegative women. A few cases of perinatal HIV-2 transmission occurred, but no case
of late postnatal transmission was observed.
121
It is probable that HIV-2 transmission via breast milk is less common than with HIV-1,
but insufficient data support that the risk for transmission is zero. Mothers who
test positive for HIV-2 should be tested for HIV-1, and guidelines for breastfeeding
should follow those for HIV-1 until additional information is available.
Rabies
Rabies virus produces a severe infection with progressive CNS symptoms (anxiety, seizures,
altered mental status) that ultimately proceeds to death; few reports of survival
exist. Rabies occurs worldwide except in Australia, Antarctica, and several island
groups. In 1992 more than 36,000 cases of rabies were reported to the WHO, a number
that is probably a marked underestimate of the actual cases.
67
Between 1990 and 2003, 37 cases of human rabies were reported in the United States.70.,
78. Postexposure prophylaxis is given to thousands of patients each year.
Rabies virus is endemic in various animal populations, including raccoons, skunks,
foxes, and bats. Because of aggressive immunization programs, rabies in domesticated
dogs and cats in the United States is uncommon. The virus is found in the saliva and
tears and nervous tissue of infected animals. Transmission occurs by bites, licking,
or simply contact of oral secretions with mucous membranes or nonintact skin. Many
cases of rabies in humans now lack a history of some obvious contact with a rabid
animal. This may be a result of the long incubation period (generally 4 to 6 weeks,
but can be up to 1 year, with reports of incubation periods of several years), a lack
of symptoms early in an infectious animal, or airborne transmission from bats in enclosed
environments (caves, laboratories, houses).
Person-to-person transmission via bites has not been documented, although it has occurred
in corneal transplants.
44
Rabies viremia has not been observed in the spread of the virus. No evidence exists
indicating transmission through breast milk.
In the case of maternal infection with rabies, many scenarios can occur before the
onset of progressive, severe CNS symptoms. The progression and severity of maternal
illness can preclude breastfeeding, but separation of an infant from the mother is
appropriate regardless of the mother’s status and method of infant feeding (especially
to avoid contact with saliva and tears). Breastfeeding should not continue when the
mother has symptoms of rabies, and the infant should receive postexposure immunization
and close observation. An infant may received expressed breast milk, but the expression
must occur without possible contamination with saliva or tears from the mother.
Depending on the scenario, the nature of a mother’s illness, the possible exposure
of an infant to the same source as the mother, and the exposure of a child to the
mother, postexposure immunization of an infant may be appropriate.
A more common scenario is a mother’s apparent exposure to rabies (without exposure
for the infant), necessitating postexposure immunization of the mother with rabies
vaccine. In the majority of cases, in the absence of maternal illness, breastfeeding
can reasonably continue during the mother’s five-dose immunization series in 28 days.
In a rare situation in which apparent exposure of mother and infant to rabies occurs
together, postexposure treatment of both mother and infant should be instituted, and
breastfeeding can continue.
Respiratory Syncytial Virus
Respiratory syncytial virus (RSV) is a common cause of respiratory illness in children
and is relatively common in adults, usually producing milder upper respiratory tract
infection in adults. No evidence indicates that RSV causes intrauterine infection,
adversely affects the fetus, or causes abortion or prematurity. RSV does produce infection
in neonates, causing asymptomatic infection, afebrile upper respiratory tract infection,
bronchiolitis, pneumonia, and apnea. Mortality rate can be high in neonates, especially
in premature infants and ill full-term infants, particularly those with preexisting
respiratory disease (hyaline membrane disease, bronchopulmonary dysplasia) or cardiac
disease associated with pulmonary hypertension.
RSV is believed to be transmitted via droplets or direct contact of the conjunctiva,
nasal mucosa, or oropharynx with infected respiratory secretions. Documentation of
RSV infection is rarely made in adults, and spread from a mother or other household
contacts probably occurs before a diagnosis can be made. Therefore risk for RSV transmission
from breast milk is probably insignificant compared with transmission via direct or
droplet contact in families. In nurseries, however, it is appropriate to make a timely
diagnosis of RSV infection in neonates to isolate infants from the others and prevent
spread in the nursery. Ribavirin is not recommended for routine use. It is infrequently
used in patients with potentially life-threatening RSV infection.
RSV infection should be suspected in any infant with rhinorrhea, nasal congestion,
or unexplained apnea, especially in October through March in temperate climates. Prophylaxis
against RSV with RSV-specific immunoglobulin IV (RSV-IGIV) during this season for
infants at highest risk for severe disease is appropriate.
Debate surrounds the effect of passively acquired antibodies (in infants from mothers
before birth) against RSV on the occurrence and severity of illness in neonates and
infants. It appears that a higher level of neutralizing antibody against RSV in neonates
decreases the risk for severe RSV disease.153., 239. Some controversy remains concerning
the measurable benefit of breastfeeding for preventing serious RSV disease.3., 54.,
115. Some studies have shown benefit and others no effect. Controlling for possible
confounding factors (e.g., smoking, crowded living conditions) in these studies has
been difficult. At this point, no reason exists to stop breastfeeding with maternal
RSV infection; a potential exists for benefit from nonspecific factors in breast milk
against the RSV. Infants with RSV infection should breastfeed unless their respiratory
status precludes it.
Rotaviruses
Rotavirus infections usually result in diarrhea, accompanied by emesis and low-grade
fever. In severe infections the clinical course can include dehydration, electrolyte
abnormalities, and acidosis and can contribute to malnutrition in developing countries.
Generally, every child will have at least one episode of rotavirus infection by 5
years of age.
157
In developed countries, rotavirus is often associated with diarrhea requiring hospitalization
in children younger than 2 years of age, but rarely associated with death. Worldwide
rotavirus is the leading cause of diarrhea-related deaths in children younger than
5 years old. Estimates suggest that in children younger than 5 years old rotavirus
infection leads to more than 100 million occurrences of diarrhea, 2 million hospital
admissions, and 500,000 deaths each year.
157
Fecal-oral transmission is the most common route, but fomites and respiratory spread
may also occur. Spread of infection occurs most often in homes with young children
or in daycare centers and institutions. In hospitalized infants or mothers with rotavirus
infection, contact precautions are indicated for the duration of the illness. No evidence
indicates prenatal infection from rotavirus, but perinatal or postnatal infection
from contact with the mother or others can occur.
No case of transmission of rotavirus via breast milk has been documented. Breast milk
does contain antibodies to rotavirus for up to 2 years. Human milk mucin has been
demonstrated to inhibit rotavirus replication and prevent experimental gastroenteritis.
457
The mechanisms of rotavirus immunity are not well understood. They are thought to
be multifactorial with cell-mediated immunity limiting severity and the course of
infection, while humoral immunity protects against subsequent infections. Innate and
adaptive responses at the level of the mucosa are probably the most important.
134
Exclusive breastfeeding may decrease the likelihood of severe rotavirus-related diarrhea
by as much as 90%.93., 377. Although breastfeeding does not prevent infection with
rotavirus, it seems to decrease the severity of rotavirus-induced illness in children
younger than 2 years old.93., 123., 184. At least one study suggested that this may
represent simply the postponement of severe rotavirus infection until an older age.
93
One study suggested that protection against rotavirus rapidly declines upon discontinuation
of breastfeeding.
356
This delay in rotavirus infection until the child is older may be beneficial in that
the older child may be able to tolerate the infection or illness with a lower likelihood
of becoming dehydrated or malnourished. Continuing breastfeeding during an episode
of rotavirus illness with or without vomiting is appropriate and often helpful to
the infant. No reason to suspend breastfeeding by a mother infected with rotavirus
is apparent.
Two rotavirus vaccines (RotaTeq and Rotarix) have been licensed for use in more than
90 countries, but less than 20 countries have routine immunization programs. Additional
types of rotavirus vaccines are undergoing study in various countries, specifically
examining the efficacy of the vaccines in low and medium income countries.
444
Some of the explanations for the slow implementation of an effective vaccine globally
include differences in protection with specific vaccines in high income countries
compared with low or medium income countries, the unfortunate association with intussusception
in the United States, the delayed recognition of the significant rotavirus-related
morbidity and mortality, and the cost of the new vaccines. The question of variable
efficacy of the specific rotavirus vaccines in developed and developing countries
remains an important one. Several trials are examining this issue and attempting to
address factors such as maternal transplacentally transferred antibodies, breastfeeding
practices (especially immediately before immunization with a live oral rotavirus vaccine),
stomach acid, micronutrient malnutrition, interfering gut flora, and differences in
the epidemiology of rotavirus in different locations.
327
Evidence indicates that maternal immunization with rotavirus vaccine can increase
both transplacental acquisition of antibodies and secretory IgA in breast milk.
334
Additionally, oral rotavirus vaccines have been able to stimulate a good serologic
response in both formula-fed and breastfed infants, although the antigen titers may
need to be modified to create an optimal response in all infants.
86
The actual protective effect of these vaccines in different situations and strategies
will require measurement in ongoing prospective studies.
Rubella Virus
Congenital rubella infection has been well described, and the contributing variables
to infection and severe disease have been elucidated. The primary intervention to
prevent congenital rubella has been to establish the existence of maternal immunity
to rubella before conception, including immunization with rubella vaccine and reimmunization
if indicated. Perinatal infection is not clinically significant. Postnatal infection
occurs infrequently in children younger than 1 year of age because of passively acquired
maternal antibodies. The predominant age of infection is 5 to 14 years old, and more
than half of those with infections are asymptomatic. Postnatal rubella is a self-limited,
mild viral infection associated with an evanescent rash, shotty adenopathy, and low-grade
transient fever. It most often occurs in the late winter and spring. Infants with
congenital infection shed the virus for prolonged periods from various sites and may
serve as a source of infection throughout the year. Contact isolation is appropriate
for suspected and proved congenital infection for at least 1 year, including exclusion
from day care and avoidance of pregnant women, whereas postnatal rubella infection
requires droplet precautions for 7 days after the onset of rash.
Rubella virus has been isolated from breast milk after natural infection (congenital
or postnatal) and after immunization with live attenuated vaccine virus. Both IgA
antibodies and immunoreactive cells against rubella have been identified in breast
milk. Breastfed infants can acquire vaccine virus infection via milk but are asymptomatic.
Because postpartum infection with this virus (natural or vaccine) is not associated
with clinically significant illness, no reason exists to prevent breastfeeding after
congenital infection, postpartum infection with this virus, or maternal immunization
with rubella vaccine.
Severe Acute Respiratory Syndrome
Severe acute respiratory syndrome (SARS) is a term that could be applied to any acute
serious respiratory illness caused by or associated with a variety of infections agents;
since 2003, however, it has been linked with SARS-associated coronavirus (SARS-CoV).
In the global outbreak of 2002 to 2003, more than 8400 probable cases of SARS and
more than 800 deaths occurred. More than the actual number of affected individuals
or its associated mortality rate (approximately 10% overall, and closer to 50% in
persons older than 65 years of age), it was what we did not know about this new unusual
illness, and the tremendous publicity surrounding it, that made SARS such a sensation.
We now know the cause of this illness, known as the SARS-CoV. SARS-CoV was shown not
to be closely related to the previously characterized coronavirus groups.265., 350.
Despite intense international collaboration to study the illness and the virus, many
things are not known, such as the degree of infectiousness, the actual period of transmissibility,
all the modes of transmission, how many people have an asymptomatic infection compared
with those with symptoms or severe illnesses, how to make a rapid diagnosis of confirmed
cases, and where it originated.
At least 21 cases of probable SARS in children have been described in the literature.42.,
187., 380., 387. In general, the illness in children is a mild, nonspecific respiratory
illness, but in adolescents and adults it is more likely to progress to severe respiratory
distress. It has been reported that children are less likely to transmit SARS than
adults.
187
The overall clinical course, the radiologic evolution, and the histologic findings
of these illness are consistent with the host’s immune response playing a significant
role in disease production.
Five infants were born to mothers with confirmed SARS. The infants were born prematurely
(26 to 37 weeks), presumably due to maternal illness. Although two of the five infants
had serious abdominal illnesses (other coronaviruses have been associated with reported
outbreaks of necrotizing enterocolitis), the presence of SARS-CoV could not be demonstrated
in any of these infants.
380
No evidence of vertical transmission of SARS is available.
The mode of feeding for any of the reported cases of young children with SARS or the
infants born to mothers with SARS was not mentioned. As with other respiratory viruses
predominantly transmitted by droplets, transmission via breast milk is an insignificant
mode of transmission, if it occurs at all. The benefits of breastfeeding being what
they are, mothers with SARS should continue breastfeeding if they are able, or expressed
breast milk can be given to an infant until the mother is able to breastfeed.
Smallpox
In this era of worry about biologic terrorism, smallpox is an important concern. The
concern for infants (breastfed or formula-fed) is direct contact with mothers or household
members with smallpox. Smallpox is highly contagious in the household setting due
to person-to-person spread via droplet nuclei or aerosolization from the oropharynx
and direct contact with the rash. Additional potential exposures for infants include
the release of a smallpox aerosol into the environment by terrorists, contact with
a smallpox-contaminated space or the clothes of household members exposed to an aerosol,
and infection via contact with a mother’s or a household member’s smallpox vaccination
site. These risks are the same for breastfed and formula-fed infants. No evidence
for transmission of the smallpox virus via breast milk exists.
A contact is defined as a person who has been in the same household or had face-to-face
contact with a patient with smallpox after the onset of fever. Patients do not transmit
infection until after progression from the fever stage to the development of the rash.
An exposed contact does not need to be isolated from others during the postcontact
observation period (usually 17 days) until the person develops fever. Temperature
should be monitored daily in the exposed contact. Personal contact and breastfeeding
between mother and infant can continue until the onset of fever, when immediate isolation
(at home) should begin. Providing expressed breast milk for the infant of a mother
with smallpox should be avoided because of the extensive nature of the smallpox rash
and the possibility of contamination (from the rash) of the milk during the expression
process. No literature documents transmission of the smallpox virus via expressed
breast milk.
The other issue for breastfeeding infants is the question of maternal vaccination
with smallpox in a preexposure event vaccination program. Children older than 1 year
of age can be safely and reasonably vaccinated with smallpox in the face of a probable
smallpox exposure. Smallpox vaccination of infants younger than 1 year of age is contraindicated.
Breastfeeding is listed as a contraindication to vaccination in the preevent vaccination
program. It is unknown whether vaccine virus or antibodies are present in breast milk.
The risk for infection due to contact or aerosolization of virus from a mother’s smallpox
vaccination site is the same for breastfed and formula-fed infants. The Advisory Committee
on Immunization Practices also does not recommend preevent smallpox vaccination of
children younger than 18 years old.
443
A report documents tertiary contact vaccinia in a breastfeeding infant.
140
A United States military person received a primary smallpox vaccination and developed
a local reaction at the inoculation site. Despite reportedly observing appropriate
precautions, the individual’s wife developed vesicles on both areolae (secondary contact
vaccinia). Subsequently, the breastfeeding infant developed lesions on her philtrum,
cheek, and tongue. Both the mother and infant remained well and the infections resolved
without therapy. Culture and PCR testing confirmed vaccinia in both the mother’s and
the infant’s lesions. The breast milk was not tested.
140
In a review from 1931 to 1981, Sepkowitz
375
reported on 27 cases of secondary vaccinia in households. The CDC reported 30 suspected
cases of secondary/tertiary vaccinia with 18 of those cases confirmed by culture or
PCR. The 30 cases were related to 578,286 vaccinated military personnel. This is an
incidence of 5.2 cases per 100,000 vaccinees and 7.4 cases per 100,000 primary vaccinees.
79
In a separate report on the civilian preevent smallpox vaccination program, 37,802
individuals were vaccinated between January and June 2003, and no cases of contact
vaccinia were reported.
77
The risk for contact vaccinia is low. The risk is from close or intimate contact.
In the above-mentioned case, the risk for the infant was contact with the mother’s
breasts, the inadvertent site of her contact vaccinia. Breastfed and formula-fed infants
are equally at risk from close contact in the household of a smallpox vaccinee or
a case of secondary vaccinia, and separation from the individual is appropriate in
both situations. If the breast of the nursing mother is not involved, expressed breast
milk can be given to the infant.
TT Virus
TT virus (TTV) is a recently identified virus found in a patient (TT) with posttransfusion
hepatitis not associated with the other hepatitis-related viruses A through G. TTV
has been described as an unenveloped, circular, single-stranded DNA virus.
311
This virus is prevalent in healthy individuals, including healthy blood donors, and
has been identified in patients with hepatitis. TTV DNA has been detected in infants
of TTV-positive and TTV-negative mothers. Ohto et al
310
reported no TTV DNA was detected in cord blood from 38 infants, and it was detected
in only 1 of 14 samples taken at 1 month of age. They noted an increasing prevalence
from 6 months (22%) to 2 years (33%), which they ascribed to acquisition via nonparenteral
routes. In comparisons of the TTV DNA in TTV-positive mothers and their TTV-positive
infants, 6 of 13 showed high level nucleotide sequence similarity, and 7 of 13 differed
by greater than 10%.
310
Schröter et al
371
reported on TTV DNA in breast milk examined retrospectively. Notably, TTV DNA was
detected in 22 of 23 serum samples of infants at 1 week of age, who were born to 22
women viremic for TTV DNA. Twenty-four women who were negative for TTV DNA gave birth
to 24 children who were initially negative for TTV DNA and remained negative throughout
the observation period (mean 7.5 months, range 1 to 28 months). TTV DNA was detected
in 77% of breast milk samples from TTV viremic women and in none of the breast milk
samples from TTV-negative women. No clinical or laboratory evidence of hepatitis was
found in the 22 children who were observed to be TTV DNA positive during the period
of the study.
371
Other authors have reported TTV in breast milk detected by PCR. They describe the
absence of TTV DNA in infants at 5 days and 3 months of age, and 4 of 10 infants were
positive for TTV DNA at 6 months of age, suggesting the late acquisition of infection
via breastfeeding.
197
TT virus is transmitted in utero and is found in breast milk. No evidence of clinical
hepatitis in infants related to TTV infection and no evidence for a late chronic hepatitis
exist. Given the current available information, no reason to proscribe breastfeeding
by TTV-positive mothers is compelling. Certainly more needs to be understood concerning
the chronic nature of this infection and the possible pathogenesis of liver disease.
Tumor Virus in Breast Milk
No documented evidence indicates that women with breast cancer have RNA of tumor virus
in their milk. No correlation between RNA-directed DNA polymerase activity has been
found in women with a family history of breast cancer. RNA-directed DNA polymerase
activity, a reserve transcriptase, is a normal feature of the lactating breast.91.,
129., 352.
Epidemiologic data conflict with the suggestion that the tumor agent is transmitted
through the breast milk. The incidence of breast cancer is low among groups who had
nursed their infants, including lower economic groups, foreign-born groups, and those
in sparsely populated areas.
262
The frequency of breast cancer in mothers and sisters of a woman with breast cancer
is two to three times that expected by chance. This could be genetic or environmental.
Cancer actually is equally common on both sides of the family of an affected woman.
If breast milk were the cause, it should be transmitted from mother to daughter. When
mother-daughter incidence of cancer was studied, no relationship was found to breastfeeding.
Sarkar et al
360
reported that human milk, when incubated with mouse mammary tumor virus, caused degradation
of the particular morphology and decreased infectivity and reverse transcriptase activity
of the virions. They suggest that the significance of this destructive effect of human
milk on mouse mammary tumor virus may account for the difficulty in isolating the
putative human mammary tumor agent. Sanner
359
showed that the inhibitory enzymes in milk can be removed by special sedimentation
technique. He ascribes the discrepancies in isolating virus particles in human milk
to these factors, which inhibit RNA-directed DNA polymerase.
The fear of cancer in breastfed female offspring of a woman with breast cancer does
not justify avoiding breastfeeding. Breastfed women have the same breast cancer experience
as nonbreastfed women, and no increase is seen in benign tumors. Daughters of breast
cancer patients have an increased risk for developing benign and malignant tumors
because of their heredity, not because of their breastfeeding history.280., 287.
Unilateral breastfeeding (limited to the right breast) is a custom of Tanka women
of the fishing villages of Hong Kong. Ing et al
194
investigated the question, “Does the unsuckled breast have an altered risk for cancer?”
They studied breast cancer data from 1958 to 1975. Breast cancer occurred equally
in the left and the right breasts. Comparison of patients who had nursed unilaterally
with nulliparous patients and with patients who had borne children but not breastfed
indicated a highly significantly increased risk for cancer in the unsuckled breast.
The authors conclude that in postmenopausal women who have breastfed unilaterally,
the risk for cancer is significantly higher in the unsuckled breast. They thought
that breastfeeding may help protect the suckled breast against cancer.
194
Others
274
have suggested that Tanka women are ethnically a separate people and that left-sided
breast cancer may be related to their genetic pool and not to their breastfeeding
habits. No mention has been made of other possible influences, such as the impact
of their role as “fishermen” or any inherent trauma to the left breast.
274
In 1926, Lane-Claypon
240
stated that a breast that had never lactated was more liable to become cancerous.
Nulliparity and absence of breastfeeding had been considered important risk factors
for breast cancer. MacMahon et al
262
reported in 1970 that age at first full-term pregnancy was the compelling factor,
and the younger the mother, the less the risk.
In a collective review of the etiologic factors in cancer of the breast in humans,
Papaioannou
325
concludes, “Genetic factors, viruses, hormones, psychogenic stress, diet, and other
possible factors, probably in that order of importance, contribute to some extent
to the development of cancer of the breast.”
325
Wing
449
concluded in her 1977 review on human milk and health that “in view of the complete
absence of any studies showing a relationship between breastfeeding and increased
risk of breast cancer, the presence of virus-like particles in breast milk should
not be a contraindication to breastfeeding.” Henderson et al
173
made a similar statement in 1974, whereas Vorherr
432
concluded in 1979 that the roles of pregnancy and lactation in the development and
prognosis of breast cancer had not been determined.
Gradually, studies have appeared challenging the dogma. Brinton et al,
52
McTiernan and Thomas,
276
and Layde et al
245
showed the clearly protective effects of breastfeeding. Another example is a study
conducted to clarify whether lactation has a protective role against breast cancer
in an Asian people, regardless of confounding effects of age at first pregnancy, parity,
and closely related factors.
458
In a hospital-based case-control study of 521 women without breast cancer, statistical
adjustment for potential confounders and a likelihood ratio test for linear trend
were done by unconditional logistic regression. Total months of lactation regardless
of parity was the discriminator. Regardless of age of first pregnancy and parity,
lactation had an independent protective effect against breast cancer in Japanese women.
458
Although breast cancer incidence is influenced by genetics, stress, hormones, and
pregnancy, breastfeeding clearly has a protective effect. “There is a reduction in
the risk of breast cancer among premenopausal women who have lactated. No reduction
in the risk of breast cancer occurred among postmenopausal women with a history of
lactation,” according to Newcombe et al,
299
reporting a multicenter study in 1993.
Varicella-Zoster Virus
Varicella-zoster virus infection (varicella/chickenpox, zoster/shingles) is one of
the most communicable diseases of humans, in a class with measles and smallpox. Transmission
is thought to occur via respiratory droplets and virus from vesicles. Varicella in
pregnancy is a rare event, although disease can be more severe with varicella pneumonia,
and can be fatal.
Congenital varicella-zoster virus infection occurs infrequently, causing abortion,
prematurity, and congenital malformations. A syndrome of malformations has been carefully
described with congenital varicella-zoster virus infection, typically involving limb
deformity, skin scarring, and nerve damage, including to the eye and brain.
148
Perinatal infection can lead to severe infection in infants if maternal rash develops
5 days or less before delivery and within 2 days after delivery. Illness in infants
usually develops before 10 days of age and is believed to be more severe because of
the lack of adequate transfer of antibody from the mother during this period and transplacental
spread of virus to the fetus and infant during viremia in the mother. Varicella in
a mother occurring before 5 days before delivery allows sufficient formation and transplacental
transfer of antibodies to the infant to ameliorate disease even if the infant is infected
with varicella-zoster virus. Mothers who develop varicella rash more than 2 days after
delivery are less likely to transfer virus to the infant transplacentally; they pose
a risk to the infants from postnatal exposure, which can be diminished by the administration
of varicella-zoster Ig to the infant. Postnatal transmission is believed to occur
through aerosolized virus from skin lesions or the respiratory tract entering the
susceptible infant’s respiratory tract. Airborne precautions are therefore appropriate
in the hospital setting. Infants infected with varicella-zoster virus in utero or
in the perinatal period (younger than 1 month of age) are more likely to develop zoster
(reactivation of latent varicella-zoster virus) during childhood or as young adults.
Table 13-5
summarizes management of varicella in the hospitalized mother or infant.
148
TABLE 13-5
Guidelines for Preventive Measures After Exposure to Chickenpox in Nursery or Maternity
Ward
Type of Exposure or Disease
CHICKENPOX LESIONS PRESENT
Disposition
Mother
Neonate
A. Siblings at home have active chickenpox when neonate and mother are ready for discharge
from hospital.
No
No
1. Mother: If she has a history of chickenpox, she may return home. Without a history,
she should be tested for varicella-zoster virus antibody titer.∗ If test is positive,
she may return home. If test is positive, she may return home. If test is negative,
varicella-zoster Ig† is administered and she is discharged home.
2. Neonate: May be discharged home with mother if mother has history of varicella
or is varicella-zoster virus-antibody positive. If mother is susceptible, administer
varicella-zoster Ig to infant and discharge home or place in protective isolation.
B. Mother has no history of chickenpox; exposed during period 6-20 days antepartum.‡
No
No
1. Exposed mother and infant: Send home at earliest date unless siblings at home have
communicable chickenpox.§ If so, may administer varicella-zoster Ig and discharge
home, as above.
2. Other mothers and infants: No special management indicated.
3. Hospital personnel: No precautions indicated if there is a history of previous
chickenpox or zoster. In absence of history, immediate serologic testing is indicated
to determine immune status.∗ Nonimmune personnel should be excluded from patient contact
until 21 days after an exposure.
4. If mother develops varicella 1 to 2 days postpartum, infant should be given varicella-zoster
Ig.
C. Onset of maternal chickenpox occurs antepartum‡ or postpartum.
Yes
No
1. Infected mother: Isolate until no longer clinically infectious. If seriously ill,
treat with acyclovir.¶
2. Infected mother’s infant: Administer varicella-zoster Ig† to neonates born to mothers
with onset of chickenpox less than 5 days before delivery and isolate separately from
mother. Send home with mother if no lesions develop by the time mother is noninfectious.
3. Other mothers and infants: Send home at earliest date. varicella-zoster Ig may
be given to exposed neonates.
4. Hospital personnel: Same as B-3.
D. Onset of maternal chickenpox occurs antepartum.§
1. Mother: Isolation unnecessary.
2. Infant: Isolate from other infants but not from mother.
3. Other mothers and infants: Same as C-3 (if exposed).
4. Hospital personnel: Same as B-3 (if exposed).
E. Congenital chickenpox
No
Yes
1. Infected infant and mother: Same as D-1 and D-2.
2. Other mothers and infants: Same as C-3.
3. Hospital personnel: Same as B-3.
ELISA, Enzyme-linked immunosorbent assay; FAMA, fluorescent antibody to membrane antigen;
LA, latex agglutination.
∗
Send serum to virus diagnostic laboratory for determination of antibodies to varicella-zoster
virus by a sensitive technique (e.g., FAMA, LA, ELISA). Personnel may continue to
work for 8 days after exposure pending serologic results because they are not potentially
infectious during this period. Antibodies to varicella-zoster virus greater than 1:4
probably are indicative of immunity.
†
Varicella-zoster Ig is available as VariZIG under an investigational new drug (IND)
application from the Food and Drug Administration. It is obtainable through FFF Enterprises
at 800-843-7477. The dose for a newborn is 1.25 mL (1 vial). The dose for a pregnant
woman is conventionally 6.25 mL (5 vials).
‡
If exposure occurred less than 6 days antepartum, mother would not be potentially
infectious until at least 72 hours postpartum.
§
Considered noninfectious when no new vesicles have appeared for 72 hours and all lesions
have crusted.
¶
Dosage of acyclovir for pregnant woman is 30 mg/kg/day; for seriously ill infant with
varicella, 750 to 1500 mg/m2/day.
From Gershon AA: Chickenpox, measles and mumps. In Remington JS, Klein JO, editors:
Infectious diseases of the fetus and newborn infant, ed 4, Philadelphia, 1995, WB
Saunders.
Postnatal varicella from nonmaternal exposure can occur but is generally mild when
it develops after 3 weeks of age or when a mother has passed on antibodies against
varicella-zoster virus via the placenta. Severe postnatal varicella does occur in
premature infants or infants of varicella-susceptible mothers. When a mother’s immune
status relative to varicella-zoster virus is uncertain and measurement of antibodies
to varicella-zoster virus in mother or infant cannot be performed promptly (less than
72 hours), administration of VZIG81 or IVIG to the infant exposed to varicella or
zoster in the postnatal period is indicated. Ideally a mother’s varicella status should
be known before pregnancy, when varicella virus vaccine could be given if indicated.
Varicella-zoster virus virus has not been cultured from milk, but varicella-zoster
virus DNA has been identified in breast milk.
459
Antibody against varicella-zoster virus has also been found in breast milk.
270
Breast milk from mothers who had received the varicella vaccine in the postpartum
period was tested for varicella-zoster virus DNA. Varicella DNA was not detected in
any of the 217 breast milk samples from the 12 women, all of whom seroconverted after
vaccination.
45
One case of suspected transfer of varicella-zoster virus to an infant via breastfeeding
has been reported, but virus may have been transmitted by respiratory droplet or exposure
to rash before the mother began antiviral therapy.
459
Isolation of an infant from the mother with varicella and interruption of breastfeeding
should occur only while the mother remains clinically infectious, regardless of the
method of feeding. As soon as the infant has received varicella-zoster Ig, expressed
breast milk can be given to an infant if no skin lesions involve the breasts. Persons
with varicella rash are considered noninfectious when no new vesicles have appeared
for 72 hours and all lesions have crusted, usually in 6 to 10 days. Immunocompetent
mothers who develop zoster can continue to breastfeed if the lesions do not involve
the breast and can be covered because antibodies against varicella-zoster virus are
provided to the infant via the placenta and breast milk and will diminish the severity
of disease, even if not preventing it. Conservative management in this scenario would
include giving an infant varicella-zoster Ig as well (see Table 13-5).
West Nile Virus
West Nile virus disease in the United States is one of the best examples of an emerging
infectious disease taking on new importance in public awareness about health issues.
In 2003 9136 human cases of West Nile infection were reported to the CDC (through
2/11/2004). Cases were reported from 45 states, including 6256 cases (68%) of West
Nile fever (milder cases), 2718 cases (30%) of West Nile meningoencephalitis, and
228 deaths related to West Nile disease.
80
West Nile virus is endemic in Israel and parts of Africa. Outbreaks have been reported
from Romania (1996), Russia (1999), Israel (2000), and Canada (2002) as well as the
United States (1999 to 2003).
331
It is estimated that 150 to 300 asymptomatic cases of West Nile infection occur for
every 20 febrile illnesses and for every one case of meningoencephalitis associated
with West Nile virus. West Nile fever is usually a mild illness of 3 to 6 days’ duration.
The symptoms are relatively nonspecific, including malaise, nausea, vomiting, headache,
myalgia, lymphadenopathy, and rash. West Nile disease is characterized by severe neurologic
symptoms (e.g., meningitis, encephalitis, or acute flaccid paralysis, and occasionally
optic neuritis, cranial nerve abnormalities, and seizures). Children are infrequently
sick with West Nile virus infection and infants younger than 1 year of age have rarely
been reported.
331
The case-fatality rate for 2003 in the United States was approximately 2.5%, but has
been reported as high as 4% to 18% in hospitalized patients. The case-fatality rate
for persons older than 70 years of age is considered to be higher, 15% to 29% among
hospitalized patients in outbreaks in Romania and Israel.
331
The primary mechanism of transmission is via a mosquito bite. Mosquitoes from the
genus Culex are primary vectors. The bird-mosquito-bird cycle serves to maintain and
amplify the virus in the environment. Humans and horses are incidental hosts. The
pathogenesis of the infection is believed to occur via replication of the virus in
the skin and lymph nodes, leading to a primary viremia that seeds secondary sites
before a second viremia causes the infection of the CNS and other affected organs.59.,
111. Transmission has been reported in rare instances during pregnancy7., 73. via
organ transplant
199
and percutaneously in laboratory workers.
75
A study of West Nile virus infection in pregnancy documented four miscarriages, two
elective abortions, and 72 live births. Cord blood samples were tested in 55 infants
and 54 of 55 were negative for anti-West Nile virus IgM. Three infants had West Nile
virus infection, which could have been acquired congenitally. Three of 7 infants who
had congenital malformations might have been caused by maternal West Nile virus infection
based on timing in pregnancy, but no evidence of West Nile virus etiology is conclusive.
312
West Nile virus transmission occurs via blood and blood product transfusion,
186
and the incidence has been estimated to be as high as 21 per 10,000 donations during
epidemics in specific cities.
40
No evidence of direct person-to-person transmission without the mosquito vector has
been found.
One case of possible West Nile virus transmission via breastfeeding has been documented.
74
The mother acquired the virus via packed RBC transfusions after delivery. The second
unit of blood she received was associated with other blood products from the same
donation causing West Nile infection in another transfusion recipient. Eight days
later the mother had a severe headache and was hospitalized with fever and a CSF pleocytosis
on day 12 after delivery. The mother’s CSF was positive for West Nile virus-specific
IgM antibody. The infant had been breastfed from birth through the second day of hospitalization
of the mother. Samples of breast milk were West Nile virus–specific IgG and IgM positive
on day 16 after delivery and West Nile virus-specific IgM positive on day 24. The
same milk was West Nile virus RNA positive by PCR testing on day 16, but not on day
24 after delivery. The infant tested positive for West Nile virus-specific IgM in
serum at day 25 of age, but remained well without fever. No clear-cut exposure to
mosquitoes for the infant were reported. The cord blood and placenta were not available
to be tested. IgM antibodies can be found in low concentrations in breast milk, but
this is not common or as efficient as the transfer of IgA, secretory IgA, or IgG into
breast milk.
A review of West Nile virus illness during the breastfeeding identified six occurrences
of breastfeeding during maternal West Nile virus illness.
177
Five of the six infants had no illness or detectable antibodies to West Nile virus
in their blood. One infant developed a rash and was otherwise well after maternal
West Nile virus illness, but was not tested for West Nile virus infection. Two infants
were identified who developed West Nile virus illness while breastfeeding, but no
preceding West Nile virus infection was demonstrated in their mothers. Two other breastfeeding
infants developed West Nile virus-specific antibodies after their mothers acquired
West Nile virus illness in the last week of pregnancy, but congenital infection could
not be ruled out. Live virus was not cultured from 45 samples of breast milk from
mothers infected with West Nile virus during pregnancy, but West Nile virus RNA was
detected in two samples and 14 samples had IgM antibodies to West Nile virus.
177
The above data suggest that West Nile virus infection through breastfeeding is rare.
To date evidence of significant disease due to West Nile virus infection in young
breastfeeding children is lacking. At this time, no reason exists to proscribe breastfeeding
in the case of maternal West Nile virus infection if a mother is well enough to breastfeed.
As with many other maternal viral illnesses, by the time the diagnosis is made in
a mother, the infant may have already been exposed during maternal viremia and possible
virolactia. The infant can and should continue to receive breast milk for the potential
specific and nonspecific antiviral immunologic benefits.
Yellow Fever Virus
Yellow fever virus is a flavivirus which is transmitted to humans by infected Aedes
and Haemogogus mosquitos in tropical areas of South America and Africa. Large outbreaks
occur when mosquitos in a populated area become infected from biting viremic humans
infected with yellow fever virus. Transmission from the mosquitos to other humans
occurs after an incubation period in the mosquito of 8 days. Direct person-to-person
spread has not been reported. Illness due to yellow fever virus usually begins after
an incubation period of 3 to 6 days, with acute onset of headache, fever, chills,
and myalgia. Photophobia, back pain, anorexia, vomiting, and restlessness are other
common symptoms. The individual is usually viremic for the first 4 days of illness
until the fever and other symptoms diminish. Liver dysfunction and even failure can
develop as can myocardial dysfunction. CNS infection is uncommon but symptoms can
include seizures and coma. Medical care should include intensive supportive care and
fluid management.
One case of congenital infection after immunization of a pregnant woman with the attenuated
vaccine strain has been reported. One of 41 infants whose mothers had inadvertently
received the yellow fever virus vaccine during pregnancy developed IgM and elevated
neutralizing antibodies against the yellow fever virus without any evidence of illness
or abnormalities.
418
A more recent study
404
from Brazil examined inadvertent yellow fever virus immunization during pregnancy
during a mass vaccination campaign in 2000; 480 pregnant women received the yellow
fever virus at a mean of 5.7 weeks’ gestation, the majority of whom did not know their
pregnancy status at the time. Seroconversion occurred in 98.2% of the women after
at least 6 weeks after vaccination. Mild postvaccination illness (headache, fever,
or myalgia) was reported by 19.6% of the 480 women. The frequency of malformations,
miscarriages, stillbirths, and premature deliveries was similar to that found in the
general population. At the 12-month follow-up point, 7% of the infants still demonstrated
neutralizing antibodies against yellow fever virus, but after 12 months only one child
was still seropositive.
404
Transmission of the yellow fever vaccine virus through breastfeeding was recently
reported from Brazil.
85
The mother was immunized during a yellow fever epidemic in a nonendemic area in Brazil;
15 days after delivering a healthy female infant (39 weeks’ gestational age) the mother
received the 17DD yellow fever vaccine, and 5 days later the mother reported headache,
malaise, and low-grade fever that persisted for 2 days. The mother continued breastfeeding
and did not seek medical care for herself. At 23 days of age the infant became irritable,
developed fever, and refused to nurse. The infant developed seizures and subsequent
evaluation of the infant demonstrated an abnormal CSF and CT of the brain showed bilateral
areas of diffuse low density suggestive of inflammation and consistent with encephalitis.
Yellow fever-specific IgM antibodies were identified in the infant’s serum and CSF.
Reverse-transcriptase polymerase chain reaction (RT-PCR) testing of the CSF also demonstrated
yellow fever virus RNA identical to the 17DD yellow fever vaccine virus. Breast milk
and maternal serum were not tested for yellow fever virus.
85
Yellow fever virus, wild or vaccine type, has not been identified in human breast
milk, although another flavivirus, West Nile virus, has been detected in milk from
a few lactating women with West Nile virus infection.
177
(See the section on West Nile virus.) Yellow fever vaccine–associated neurologic disease
occurs at different rates in different age-groups, including 0.5 to 4.0 cases per
1000 infants younger than 6 months of age.
285
The 17D-derived yellow fever vaccines are contraindicated in infants younger than
6 months of age. Since 2002, the Advisory Committee on Immunization Practices has
recommended, based on theoretical risk, that yellow fever vaccine be avoided in nursing
mothers, except when exposure in high-risk yellow fever endemic areas is likely to
occur.
76
No case of transmission of yellow fever virus from an infected mother to her infant
via breastfeeding or breast milk has been reported. Published information on the severity
of yellow fever virus infection in infants younger than 1 year of age, potential protection
from passively acquired antibodies, or protection from breast milk is limited. No
information on a differential risk in breastfed versus formula-fed infants is available.
Given the well documented method of transmission of yellow fever virus via mosquitos,
and the lack of evidence of transmission via breast milk, it makes more sense to protect
all infants against mosquito bites than to proscribe breastfeeding, even in the mother
infected with yellow fever virus. Continued breastfeeding or use of expressed breast
milk will depend on a mother’s health status and ability to maintain the milk supply
while acutely ill. If another source of feeding is readily available then temporarily
discarding expressed breast milk for at least 4 days of acute illness in the mother
is a reasonable precaution.
Spirochetes
Lyme disease
Lyme disease, as with other human illnesses caused by spirochetes, especially syphilis,
is characterized by a protean course and distinct phases (stages) of disease. Lyme
borreliosis was described in Europe in the early twentieth century. Since the 1970s,
tremendous recognition, description, and investigation of Lyme disease have occurred
in the United States and Europe. Public concern surrounding this illness is dramatic.
Lyme disease is a multisystem disease characterized by involvement of the skin, heart,
joints, and nervous system (peripheral and central). Stages of disease are identified
as early localized (erythema migrans, often accompanied by arthralgia, neck stiffness,
fever, malaise, and headache), early disseminated (multiple erythema migrans lesions,
cranial nerve palsies, meningitis, conjunctivitis, arthralgia, myalgia, headache,
fatigue, and, rarely, myocarditis), and late disease (recurrent arthritis, encephalopathy,
and neuropathy). The varied manifestations of disease may relate to the degree of
spirochetemia, the extent of dissemination to specific tissues, and the host’s immunologic
response.
The diagnosis of Lyme disease is often difficult in part because of the broad spectrum
of presentations, inapparent exposure to the tick, and the lack of adequately standardized
serologic tests. Culture of the spirochete, Borrelia burgdorferi, is not readily available.
Enzyme-linked immunosorbent assay (ELISA), immunofluorescent assay, and immunoblot
assay are the usual tests. PCR detection of spirochetal DNA requires additional testing
in clinical situations to clarify and standardize its utility.
Gardner
142
reviewed infection during pregnancy, summarizing a total of 46 adverse outcomes from
161 cases reported in the literature. The adverse outcomes included miscarriage and
stillbirth (11% of cases), perinatal death (3%), congenital anomalies (15%), and both
early- and late-onset progressive infection in the infants. Silver
384
reviewed 11 published reports and concluded that Lyme disease during pregnancy is
uncommon, even in endemic areas. Although the spirochete can be transmitted transplacentally,
a significant immune response in the fetus is often lacking, and the association of
Lyme infection with congenital abnormalities is weak.401., 448.
Little published information exists on whether B. burgdorferi can be transmitted via
breast milk. One report showed the detection of B. burgdorferi DNA by PCR in the breast
milk of two lactating women with untreated erythema migrans, but no evidence of Lyme
disease or transmission of the spirochete in the one infant followed for 1 year.
369
No attempt to culture the spirochete was made, so it is not possible to determine
if the detectable DNA was from viable spirochetes or noninfectious fragments. In that
same study of 56 women with untreated erythema migrans who had detectable B. burgdorferi
DNA in the urine, 32 still had detectable DNA in the urine 15 to 30 days after starting
treatment, but none had it 6 months after initiating therapy. Ziska et al
466
reported on the management of nine cases of Lyme disease in women associated with
pregnancy; seven of the nine women were symptomatic at conception and six received
antibiotics throughout pregnancy. Follow-up of the infants showed no transmission
of Lyme disease, even in the seven infants who had been breastfed.
466
The lack of adequate information on transmission of B. burgdorferi via breast milk
cannot be taken as proof that it is not occurring. If one extrapolates from data on
syphilis and the Treponema pallidum spirochete, it would be prudent to discuss the
lack of information on the transmission of B. burgdorferi via breast milk with the
mother or parents and to consider withholding breast milk at least until therapy for
Lyme disease has begun or been completed. If the infection occurred during pregnancy
and treatment has already been completed, an infant can breastfeed. If infection occurs
postpartum or the diagnosis is made postpartum, infant exposure may have already occurred.
Again, discussion with the mother or parents about withholding versus continuing breastfeeding
is appropriate.
After prenatal or postnatal exposure, an infant should be closely observed and empiric
therapy considered if the infant develops a rash or symptoms suggestive of Lyme borreliosis.
Treatment of mother and infant with ceftriaxone, penicillin, or amoxicillin is acceptable
during breastfeeding relative to the infant’s exposure to these medications. Doxycycline
should not be administered for more than 14 days while continuing breastfeeding because
of possible dental staining in the neonate. Continued surveillance for viable organisms
in breast milk and evidence of transmission through breastfeeding is recommended.
A large body of information is available on various “Lyme vaccines” used in dogs,
but these vaccines are only partially protective and must be repeated yearly. Preliminary
information suggests that a vaccine for use in humans safely produces good serologic
responses, but protective efficacy has not been demonstrated, and no information exists
on its use during pregnancy or breastfeeding.
Syphilis
Syphilis is the classic example of a spirochetal infection that causes multisystem
disease in various stages. Both acquired syphilis and congenital syphilis are well-described
entities. Acquired syphilis is almost always transmitted through direct sexual contact
with open lesions of the skin or mucous membranes of individuals infected with the
spirochete, Treponema pallidum. Congenital syphilis occurs by infection across the
placenta (placentitis) at any time during the pregnancy or by contact with the spirochete
during passage through the birth canal. Any stage of disease (primary, secondary,
tertiary) in a mother can lead to infection of the fetus, but transmission in association
with secondary syphilis approaches 100%. Infection with primary syphilis during pregnancy,
without treatment, leads to spontaneous abortion, stillbirth, or perinatal death in
40% of cases. Similar to acquired syphilis, congenital syphilis manifests with moist
lesions or secretions from rhinitis (snuffles), condylomata lata, or bullous lesions.
These lesions and secretions contain numerous spirochetes and are therefore highly
infectious.
Postnatal infection of an infant can occur through contact with open, moist lesions
of the skin or mucous membranes of the mother or other infected individuals. If the
mother or infant has potentially infectious lesions, isolation from each other and
from other infants and mothers is recommended. If lesions are on the breasts or nipples,
breastfeeding or using expressed milk is contraindicated until treatment is complete
and the lesions have cleared. Spirochetes are rarely identified in open lesions after
more than 24 hours of appropriate treatment. Penicillin remains the best therapy.
Evaluation of an infant with suspected syphilis should be based on the mother’s clinical
and serologic status, history of adequate therapy in the mother, and the infant’s
clinical status. Histologic examination of the placenta and umbilical cord, serologic
testing of the infant’s blood and CSF, complete analysis of the CSF, long bone and
chest radiographs, liver function tests, and a complete blood cell count are all appropriate
given the specific clinical situation. Treatment of the infant should follow recommended
protocols for suspected, probable, or proven syphilitic infection.
96
No evidence indicates transmission of syphilis via breast milk in the absence of a
breast or nipple lesion. When a mother has no suspicious breast lesions, breastfeeding
is acceptable as long as appropriate therapy for suspected or proven syphilis is begun
in the mother and infant.
Parasites
Giardia lamblia
Giardiasis is a localized infection limited to the intestinal tract, causing diarrhea
and malabsorption. Immunocompetent individuals show no evidence of invasive infection,
and no evidence exists that documents fetal infection from maternal infection during
pregnancy. Giardiasis is rare in children younger than 6 months of age, although neonatal
infection from fecal contamination at birth has been described.
22
Human milk has an in vivo protective effect against Giardia lamblia infection, as
documented by work from central Africa, where the end of breastfeeding heralds the
onset of Giardia infection.
145
This has been reaffirmed in undeveloped countries around the world.
The protective effect of breast milk has been identified in the milk of noninfected
donors.
151
The antiparasitic effect does not result from specific antibodies but rather from
lipase enzymatic activity. The lipase acts in the presence of bile salts to destroy
the trophozoites as they emerge from their cysts in the GI tract. Hernell et al
175
demonstrated that free fatty acids have a marked giardiacidal effect, which supports
the conclusion that lipase activity releasing fatty acids is responsible for killing
G. lamblia.
G. lamblia have also been reported to appear in the mother’s milk, and the parasite
has been transmitted to newborns via that route. The exact relationship of breastfeeding
to transmission of G. lamblia and the effect on infants continue to be studied, even
though symptomatic infection in breastfed infants is rare.
151
One report from the Middle East suggests that even partial breastfeeding is protective
against infection with intestinal parasites, including Cryptosporidium and Giardia
lamblia.
41
Breastfeeding by mothers with giardiasis is problematic mainly because of the medications
used for therapy. Metronidazole’s safety in infants has not been established, and
little information is available on quinacrine hydrochloride and furazolidone in breast
milk. Paromomycin, a nonabsorbable aminoglycoside, is a reasonable alternative recommended
for treatment of pregnant women. Breastfeeding by a mother with symptomatic giardiasis
is acceptable when consideration is given to the presence of the therapeutic agents
in the breast milk.
Hookworm Infection
Hookworm infection, most often caused by Ancylostoma duodenale and Necator americanus,
is common in children younger than the age of 4 years, and there is at least one report
on infantile hookworm disease from China.
374
This publication from the Chinese literature reports hundreds of cases of infantile
hookworm disease that include the common symptoms of bloody stools, melena, anorexia,
listlessness, and edema. Anemia, eosinophilia and even leukemoid reactions occur as
part of the clinical pictures in young children. They also note at least 20 cases
of hookworm diseases in newborn infants younger than 1 month of age. In the discussion
of infantile hookworm infection, they note four routes of infection: direct contact
with contaminated soil, “sand-stuffed” diapers, contaminated “washed/wet” diapers,
and vertical equal to transmammary transmission or transplacental transmission. They
postulated that infection of infants before 40 to 50 days of age would most likely
be due to transplacental transmission and infection before environmental contact would
most likely be due to transmammary transmission. Ample evidence is available in veterinary
medicine of transmammary spread of helminths.302., 382. At least two reports suggest
the possibility of transmammary transmission of hookworms in humans. Setasuban et
al
376
described the prevalence of Necator americanus in 128 nursing mothers as 61% and identified
N. americanus in breast milk in one case. Nwosu
303
documented positive stool samples for hookworms in 33 of 316 neonates (10%) at 4 to
5 weeks of age in southern Nigeria. The majority of neonatal infections were due to
Ancylostoma duodenale although Necator americanus is more prevalent in that area of
Nigeria. Examination of colostral milk did not demonstrate any hookworm larvae.
303
Additional epidemiologic work is necessary to determine the potential significance
of transmammary spread of helminths in humans, and more careful examination of breast
milk as a source of hookworm infection is required before reasonable recommendation
are possible.
Malaria
Malaria is recognized as a major health problem in many countries. The effect of malaria
infection on pregnant and lactating women and thus on the developing fetus, neonate,
and growing infant can be significant. The four species of malaria, Plasmodium vivax,
P. ovale, P. malariae, and P. falciparum, vary in the specific aspects of the disease
they produce. P. vivax exists throughout the world, but P. falciparum predominates
in the tropics and is most problematic in its chloroquine-resistant form. Malaria
in the United States is most often seen in individuals traveling from areas where
malaria is endemic. The parasite can exist in the blood for weeks, and infection with
P. vivax and P. malariae can lead to relapses years later. Transmission occurs through
the bite of the anopheline mosquito and can occur via transfusion of blood products
and transplacentally.
Congenital malaria is rare but seems to occur more often with P. vivax and P. falciparum.
It usually presents in the first 7 days of life (range 1 day to 2 months). It may
resemble neonatal sepsis, with fever, anemia, and splenomegaly occurring in the most
neonates and hyperbilirubinemia and hepatomegaly in less than half.
Malaria in infants younger than 3 months of age generally manifests with less severe
disease and death than in older children. Possible explanations include the effect
of less exposure to mosquitoes, passive antibody acquired from the mother, and the
high level of fetal hemoglobin in infants at this age.
22
The variations in the infection rates in children younger 3 months of age during the
wet and dry seasons support the idea that postnatal infection is more common than
congenital infection. No evidence indicates that malaria is transmitted through breast
milk. The greatest risk to infants is exposure to the anopheline mosquito infected
with malaria.
The main issues relative to malaria and breastfeeding are how to protect both mothers
and infants effectively from mosquitoes and what drugs for treating malaria in mothers
are appropriate during lactation. Protection from mosquito bites includes screened-in
living areas, mosquito nets while sleeping, protective clothing with or without repellents
on the clothes, and community efforts to eradicate the mosquitoes. Chloroquine, quinine,
and tetracycline are acceptable during breastfeeding. Sulfonamides should be avoided
in the first month of an infant’s life, but pyrimethamine-sulfadoxine (Fansidar) can
be used later.
Mefloquine is not approved for infants or pregnant women. However, the milk/plasma
ratio for mefloquine is less than 0.25, there is a large volume of distribution of
the drug, high protein binding of the drug limits its presence in breast milk, and
the relative importance of breastfeeding in areas where malaria is prevalent shifts
the risk/benefit ratio in favor of treatment with mefloquine. The single dose recommended
for treatment or the once-weekly dose for prevention allows for continued breastfeeding
with discarding of the milk for short periods after a dose (1 to 6 hours). Maternal
plasma levels of primaquine range from 53 to 107 ng/mL, but no information is available
on levels in human milk. Primaquine is used in children, and once daily dosing in
the mother would allow discarding milk with peak levels of drug. Therefore breastfeeding
during maternal malaria even with treatment is appropriate with specific medications.
Strongyloides
Strongyloides stercoralis is a nematode (roundworm). Most infections are asymptomatic,
but clinically significant infection in humans can include larval skin invasion, tissue
migration, intestinal invasion with abdominal pain and GI symptoms, and a Loeffler-like
syndrome due to migration to the lungs. Immune-compromised individuals can develop
dissemination of larvae systemically, causing various clinical symptoms. Humans are
the principal hosts, but other mammals can serve as reservoirs. Infection via the
skin by filariform larvae is the most common form of transmission; ingestion is an
uncommon occurrence. Transmammary transmission of Strongyloides species has been described
in dogs, ewes, and rats.211., 302., 382. Only one report of transmammary passage of
Strongyloides larvae in humans is available. In 76 infants younger than 200 days of
age, 34% demonstrated the presence of Strongyloides fuelleborni on stool examination.
The clinical significance of this was not elucidated. Strongyloides larvae was identified
in only one sample of milk from 25 nursing mothers.
53
In the absence of an understanding of the clinical significance of Strongyloides in
the stools of young infants, given the lack of exclusion of the most common mechanism
of transmission (through the skin) in the single report and the apparent infrequent
evidence of these larvae in human milk, it is difficult to make any recommendations
concerning breastfeeding and Strongyloides.
Toxoplasmosis
Toxoplasmosis is one of the most common infections of humans throughout the world.
The infective organism, Toxoplasma gondii, is ubiquitous in nature. The prevalence
of positive serologic test titers increases with age, indicating past exposure and
infection. The cat is the definitive host, although infection occurs in most species
of warm-blooded animals.
Postnatal infection with toxoplasmosis is usually asymptomatic. Symptomatic infection
typically manifests with nonspecific symptoms, including fever, malaise, myalgia,
sore throat, lymphadenopathy, rash, hepatosplenomegaly, and occasionally a mononucleosis-like
illness. The illness usually resolves without treatment or significant complications.
Congenital infection or infection in an immunodeficient individual can be persistent
and severe, causing significant morbidity and even death. Although most infants with
congenital infection are asymptomatic at birth, visual abnormalities, learning disabilities,
and mental retardation can occur months or years later. The syndrome of congenital
toxoplasmosis is clearly defined, with the most severe manifestations involving the
CNS, including hydrocephalus, cerebral calcifications, microcephaly, chorioretinitis,
seizures, or simply isolated ocular involvement. The risk for fetal infection is related
to the timing of primary maternal infection, although transmission can occur with
preexisting maternal toxoplasmosis.
241
In the last months of pregnancy the protozoan is more readily transmitted to the fetus,
but the infection is more likely to be subclinical. Early in pregnancy the transmission
to a fetus occurs less frequently but does result in severe disease. Treatment of
documented congenital infection is currently recommended, although duration and optimal
regimen have not been determined, and reversal of preexisting sequelae generally does
not occur.
343
Prevention of infection in susceptible pregnant women is possible by avoiding exposure
to cat feces or the organism in the soil. Pregnant or lactating women should not change
cat litter boxes, but if they must, it should be done daily and while wearing gloves.
The oocyst is not infective for the first 24 to 48 hours after passage. Mothers can
avoid ingestion of the organism by fully cooking meats and carefully washing fruits,
vegetables, and food preparation surfaces.
94
In various animal models, T. gondii has been transmitted through the milk to the suckling
young. The organism has been isolated from colostrum as well. The newborn animals
became asymptomatically infected when nursed by an infected mother whose colostrum
contained T. gondii. Only one report has identified T. gondii in human milk, and some
question surrounds the reliability of that report.
241
Transmission during breastfeeding in humans has not been demonstrated. Breast milk
may contain appropriate antibodies against T. gondii. Given the benign nature of postnatal
infection, the absence of documented transmission in human breast milk, and the potential
antibodies in breast milk, no reason exists to proscribe breastfeeding by a mother
known to be infected with toxoplasmosis.
Trichomonas vaginalis
Trichomonas vaginalis is a flagellated protozoan that can produce vaginitis (see Chapter
16 for a discussion of vaginitis) but frequently causes asymptomatic infection in
both men and women. The parasite is found in 10% to 25% of women in the childbearing
years. It is transmitted predominantly by sexual intercourse, but it can be transmitted
to the neonate by passage through the birth canal. This parasite often coexists with
other STDs, especially gonorrhea.
Infection during pregnancy or while taking oral contraceptives is more difficult to
treat. Some evidence suggests that infection with and growth of the parasite are enhanced
by estrogens or their effect on the vaginal epithelium. No evidence indicates adverse
effects on the fetus in association with maternal infection during pregnancy. Occasionally
female newborns have vaginal discharge during the first weeks of life caused by T.
vaginalis. This is thought to be influenced by the effect of maternal estrogen on
the infant’s vaginal epithelium and acquisition of the organism during passage through
the birth canal. The organism does not seem to cause significant disease in a healthy
infant. No documentation exists on transmission of T. vaginalis via breast milk.
The difficulty encountered with maternal infection during lactation stems from metronidazole
(Flagyl), the drug of choice, being contraindicated for infants. Case reports describe
treatment of neonates with metronidazole without adverse effect. Although topical
agents containing povidone-iodine (Betadine) or sodium lauryl sulfate (Trichotine)
can be effective when given as douches, creams, or suppositories, metronidazole remains
the treatment of choice. The AAP advises using metronidazole only with a physician’s
discretion and considers its effect on a nursing infant unknown but possibly a concern.
The potential concerns are metronidazole’s disulfiram-like effect in association with
alcohol, tumorigenicity in animal studies, and leukopenia and neurologic side effects
described in adults. On the other hand, metronidazole is given to children beyond
the neonatal period to treat serious infections with various other parasites, such
as Entamoeba histolytica.
The current recommendation for lactating women is to try local treatment first, and
if these fail, then to try metronidazole. A 2-g single-dose treatment produces peak
levels after 1 hour, and discarding expressed breast milk for the next 12 to 24 hours
is recommended. If this treatment also fails, a 1-g twice-daily regimen for 7 days
or a 2-g single daily dose for 3 to 5 days is recommended, with discarding of breast
milk close to the dose and timing of feedings distant from the dose. Infants who exclusively
breastfeed are presumed at greater risk from exposure to metronidazole than those
who are only partially breastfed.
Candida Infections
Candida consists of multiple species. The most common species affecting humans include
C. albicans as the dominant agent and C. tropicalis, C. krusei, and C. parapsilosis,
as well as many other uncommon species. In general, Candida exists as a commensal
organism colonizing the oropharynx, GI tract, vagina, and skin without causing disease
until some change disrupts the balance between the organism and the host. Mild mucocutaneous
infection is the most common illness, which can lead to vulvovaginitis, mastitis,
or, uncommonly, oral mucositis in a mother, and thrush (oral candidiasis) and candidal
diaper rash in an infant.
Invasive candidal infection occurs infrequently, usually when a person has other illness,
impaired resistance to infection (HIV, diabetes mellitus, neutropenia; decreased cell-mediated
immunity in premature infants or LBW or VLBW infants), or disrupted normal mucosal
and skin barriers and has received antibiotics or corticosteroids. Invasive disease
can occur through local spread, and may occur more often in the genitourinary tract
(urethra, bladder, ureters, kidneys), but usually develops in association with candidemia.
The bladder and kidney are more frequently involved, but when dissemination occurs
via candidemia, a careful search for other sites of infection should be made (e.g.,
retina, liver, spleen, lung, meninges).
279
Transmission usually occurs from healthy individuals colonized with Candida through
direct contact with them or through contact with their oral or vaginal secretions.
Intrauterine infection can occur through ascending infection through the birth canal
but is rare. No distinct syndrome of congenital candidal infection exists. Most often
an infant is infected in passing through the birth canal and remains colonized. Postnatal
transmission can occur through direct contact with caregivers.
The mother and infant serve as an immediate source of recolonization for each other,
especially during the direct contact of breastfeeding. For this reason, an infant
and breastfeeding mother should be treated simultaneously when treating thrush, vulvovaginitis,
diaper candidiasis, or mastitis. Colonization with this organism usually occurs in
the absence of any clinical evidence of infection. Simultaneous treatment should occur
even in the absence of any clinical evidence of Candida infection or colonization
in the apparently uninvolved individual of the breastfeeding dyad.
No well-controlled clinical trials define the most appropriate or most effective method(s)
of treatment for candidal infection in breastfeeding mother-infant pairs. The list
of possible treatment products is extensive and includes many anecdotal and empirical
regimens. In the face of this absence of data, Brent
51
conducted a survey of members of The Academy of Breastfeeding Medicine concerning
the respondents’ approach to diagnosis and treatment of thrush in the breastfeeding
dyad. Most of the respondents relied on the history and physical examination of the
infant, but only a third rated the examination of the mother as very important in
making a diagnosis. Only 7% reported using laboratory testing to make the diagnosis.
Twenty-one percent of the respondents reported using only oral nystatin for the infant
when the mother was asymptomatic. Almost half treated the infant and the mother with
topical nystatin, and 13% used oral nystatin for the infant and oral fluconazole for
the mother when the mother had breast pain. Less than 5% used oral fluconazole for
both infant and mother, and other therapies were used by about 15% of the respondents.
For recurrence of persistence of the thrush, more respondents reported treating the
mother or both the infant and mother with fluconazole, and almost a quarter reported
using other therapies.
Considerable discussion of mammary candidosis/candidiasis, the clinical diagnosis
of candidal involvement of the breast, the significance of pain with breastfeeding,
and the presence or absence of Candida albicans in milk samples is ongoing.14., 133.,
166. This topic will continue to be debated because additional prospective studies
are necessary to clarify specific issues. Data are inadequate to make specific recommendations
about various clinical situations regarding candida and breastfeeding. Clinical practice
will vary with experience, especially for the more problematic clinical situations.
Some general guidelines follow. (See Chapter 16 for a discussion of mastitis.)
The Treatment of mucocutaneous candidiasis should probably begin with a topical agent,
such as nystatin, clotrimazole, miconazole, econazole, butaconazole, terconazole,
or ciclopirox. Treatment should continue for at least 2 weeks, even with obvious improvement
in 1 or 2 days. Failures most often result from inadequate therapy involving the frequency
of application, careful washing and drying before application, or, in the case of
diaper candidiasis, decreasing the contact of the skin with moisture. Nystatin oral
suspension is less effective for the treatment of oral candidiasis in infants, now
compared with the past, supposedly due to increasing resistance.
154
Gentian violet (diluted to 0.25% to 1.0%) applied to the breast or painted onto an
infant’s mouth is being recommended more frequently. Other topical preparations have
been recommended for the mother’s breast including mupirocin, grapefruit seed extract,
or mixtures of mupirocin, betamethasone ointments, and miconazole powder. Controlled
clinical trials for efficacy and toxicity are not available.
When good adherence to the proposed regimen with topical agents fails, or when infant
or mother are severely affected by pain and decreased breastfeeding, systemic therapy
is appropriate. Fluconazole and ketoconazole are the most commonly used systemic agents
for oral or diaper candidiasis and vulvovaginitis or mastitis. Fluconazole has a better
side effect profile than ketoconazole, and more data are available concerning its
safe use in children younger than 6 months of age and even neonates and premature
infants.87., 154., 209. Fluconazole is not currently approved for use in infants younger
than 6 months of age. For severe invasive infections in infants, amphotericin B with
or without oral flucytosine, IV fluconazole, voriconazole or caspofungin are reasonable
choices in different situations. Use of itraconazole in infants has not been adequately
studied to date. Maternal use of fluconazole during breastfeeding is not contraindicated
because only a small amount of medicine compared with the usual infant dose reaches
the infant through breast milk. Amphotericin or caspofungin therapy in mothers is
also not contraindicated because these are both poorly absorbed from the GI tract.
Whenever a mother is treated for candidal mastitis or vulvovaginitis, the infant should
be treated simultaneously, at least with nystatin oral suspension as the first choice
of medication.
Any predisposing risk factors for candidal infection in mothers and infants should
be reduced or eliminated to improve the chance of rapid, successful treatment and
to decrease the likelihood of chronic or recurrent disease. For mothers, such interventions
might include decreasing sugar consumption, stopping antibiotic use as soon as possible,
and consuming some form of probiotic bacteria, such as acidophilus (in yogurt, milk,
or pill form), to reestablish a normal colonizing bacterial flora. For infants, breastfeeding
can enhance the growth of specific colonizing bacterial flora such as lactobacillus,
which can successfully limit fungal growth. Breastfeeding should continue with appropriate
support and problem-solving with a professional who is knowledgeable about breastfeeding.
Summary
HIV-1, HIV-2, HTLV-I, and HTLV-II are the only infectious diseases that are considered
absolute contraindications to breastfeeding in developed countries. When the primary
route of transmission is via direct contact or respiratory droplets/particles, temporary
separation of mother and infant may be appropriate (whether the infant is breastfed
or formula fed), but expressed breast milk should be given to the infant for the organism-specific
immunologic benefits in the mother’s milk. In most instances, by the time a specific
diagnosis of infection is made for a mother, the infant has already been exposed to
the organism and providing expressed breast milk to the infant should continue. (Refer
to Appendix F for specific exceptions, such as Lassa fever.) Regarding antimicrobial
therapy for mothers and continued breastfeeding, the majority of the medications commonly
used in adults can be used to treat the same infection in infants. The additional
amount of medication received by infants via breast milk is usually insignificant.
In almost all instances, an appropriate antimicrobial agent for treating mothers that
is also compatible with breastfeeding can be chosen.
Unless the risk to infants for transmission of an infectious agent via breast milk
that leads to a clinically significant illness in the infants is documented, breastfeeding
should continue.