Introduction
Posterior reversible encephalopathy syndrome (PRES) is a clinico-radiographic syndrome
of varying causes and is associated with characteristic neuroimaging findings. Many
reports of PRES are associated with hypertension and the use of certain drugs. However,
the relationship between kidney disease and PRES, given that most renal patients have
the aforesaid risk factors, remains ill defined. We attempt here to share our experience
with PRES in 4 different circumstances.
Case Presentation
Case 1
Case 1, a 35-year-old woman, was found to have solitary functioning kidney and mild
renal insufficiency during routine evaluation. She was evaluated and continued on
conservative management. She underwent a right open simple nephrectomy in view of
recurrent pyelonephritis and was continued on hemodialysis. Renal biopsy showed chronic
pyelonephritis with marked glomerulosclerosis. She underwent deceased donor renal
transplantation. She was started on triple immunosuppression with tacrolimus, mycophenolate
mofetil, and prednisolone. On postoperative day 4, she developed generalized tonic
clonic seizures followed by respiratory failure. She was mechanically ventilated and
was started on parenteral antiepileptic drugs. Magnetic resonance imaging (MRI) of
the brain yielded findings suggestive of PRES with brain herniation. Her tacrolimus
level at this point was 6.4 ng/ml, and the drug was withheld. There was no significant
increase in antihypertensive requirement during the episode. She became symptomatically
better and was discharged on a mechanistic target of rapamycin (mTOR) inhibitor−based
triple immunosupression regimen (everolimus, mycophenolate mofetil, and prednisolone).
She had 2 episodes of biopsy-proven acute rejection within the first 6 months after
transplantation and was treated with i.v. methylprednisolone pulses, plasmapheresis,
and rituximab. There was significant chronic allograft nephropathy on the biopsy sample.
Graft function stabilized. The patient’s antihypertensive requirement increased at
this point. Six months after transplantation, she developed accelerated hypertension
and seizures, which once again necessitated mechanical ventilation. There were no
new lesions visible on brain MRI, and cerebrospinal fluid analysis results were normal.
Graft function slowly worsened, and she was re-initiated on hemodialysis in view of
graft failure within 1 year after transplantation. She has been on maintenance hemodialysis
since.
Case 2
Case 2, a 12-year-old girl, developed facial puffiness and pedal edema. Investigations
revealed nephrotic syndrome with mild renal failure (serum creatinine, 1.6 mg/dl).
Renal biopsy findings revealed focal segmental glomerulosclerosis. She was started
on oral corticosteroids. She developed accelerated hypertension and seizures within
1 month of starting treatment. Brain MRI revealed PRES. Cerebrospinal fluid analysis
results were normal. She was treated symptomatically with blood pressure (BP) control.
Her condition improved, and she was extubated. However, she needed 5 antihypertensive
agents for adequate BP control. She developed steroid resistance and was given 2 doses
of rituximab and a cyclophosphamide trial. At present, she is on steroids (40 mg once
daily) with the following basic parameters: serum creatinine, 1.85 mg/dl; serum albumin,
2.57 mg/dl; serum cholesterol, 529.3 mg%; and 3+ urine protein.
Case 3
Case 3, a 12-year-old boy, was diagnosed as having steroid-dependent nephrotic syndrome
at the age of 4 years. He developed severe renal failure and was initiated on hemodialysis.
He had seizures secondary to accelerated hypertension, which resolved with BP control
consisting of 2 antihypertensive agents. There was no evidence of PRES on MRI at this
point. He underwent living related renal allograft transplantation with a mildly positive
crossmatch. He was given plasmapheresis and started on triple immunosuppression with
tacrolimus, mycophenolate mofetil, and prednisolone. In the immediate posttransplantation
period, he had delayed graft function with biopsy-proven calcineurin toxicitiy and
2 acute rejections, for which he was treated with calcineurin inhibitor withdrawal,
plasmapheresis, i.v. Ig, rituximab, antithymocyte globulin, and i.v. corticosteroids.
He was then restarted on a tacrolimus-based regimen along with 3 antihypertensive
agents for BP control. Four months after surgery, he had severe headache, convulsions,
and altered sensorium with accelerated hypertension. Brain MRI showed evidence of
posterior reversible encephalopathy. The tacrolimus level was 4.2 ng/ml. BP was controlled
with 4 antihypertensive agents and the patient was switched to an mTOR inhibitor−based
regimen (everolimus, mycophenolate mofetil, and prednisolone). His elevated BPs were
controlled by optimizing his antihypertensive medications. He recovered fully with
these measures. In the seventh postoperative month, he was diagnosed with gastrointestinal
cytomegalovirus disease and started on treatment. Fourteen months after transplantation,
he developed headache and accelerated hypertension. MRI showed evidence of PRES again.
The patient improved with lowering of BP. Throughout the course of his illness, there
was progressive worsening of graft function. At the time of writing, serum creatinine
is 5.2 mg/dl.
Case 4
Case 4, a 34-year-old man, was found to have severe renal failure and was initiated
on twice-weekly hemodialysis. He was hypertensive at the time (requiring 3 drugs).
Fifteen months after initiation of hemodialysis, he developed headache, vomiting,
seizures, and accelerated hypertension. He had skipped his antihypertensive medication
prior to the episode. Brain MRI showed evidence of posterior reversible encephalopathy
syndrome. Antihypertensive and antiepileptic medications were continued, and with
this he showed symptomatic improvement. He is presently on maintenance hemodialysis
with 5 antihypertensive drugs.
Case 5
Mrs. E, a 25-year-old woman, had moderate renal failure, nephrotic-range proteinuria,
and microhematuria, along with joint pains and palpable purpuras. A renal biopsy showed
features of Henoch−Schönlein purpura nephritis with crescents and background interstitial
fibrosis and tubular atrophy (30%). The patient was treated with i.v. methylprednisolone
and was started on oral cyclophosphamide. However, 26 days after starting the drug,
she developed altered behavior, drowsiness, staring spells, and seizures. Blood pressure
was normal, and MRI brain showed evidence of PRES. She improved after withdrawal of
cyclophosphamide. The patient has had no episodes of repeated PRES. At the time of
writing, she has stage 5 chronic kidney disease and has not been initiated on renal
replacement therapy.
Discussion
Posterior reversible encephalopathy syndrome is a clinico-radiographic syndrome of
various etiologies, first described in a 1996 case series.
1
This described a clinical syndrome with varying presenting features such as headache,
confusion, decreased consciousness, visual disturbances, and seizures. The clinical
syndrome was associated with characteristic neuroimaging findings of posterior cerebral
white matter edema. The true incidence is not known; however, it has been described
in various case series. All age groups appear to be affected, with patients ranging
from 2 to 90 years.2, 3 Various case series suggest that PRES is more common in women,
even after excluding eclampsia.1, 4 In our series, 3 patients (60%) were female. The
pathogenesis is unclear, but it appears to be related to disordered cerebral autoregulation
and endothelial dysfunction.
1
A wide variety of medical conditions have been implicated as causes of PRES, including
the following: thrombotic thrombocytopenic purpura (TTP), drugs such as cyclosporine,
tacrolimus, sirolimus, cytarabine, gemcitabine, interferon-α, i.v. Ig, ipilimumab,
methotrexate, tyrosine kinase inhibitors (pazopanib), sorafenib, sunitinib, vincristine,
porphyria, hypercalcemia, hypomagnesmia, blood transfusion, contrast media exposure
(cerebral or coronary angiography), hypertensive encephalopathy, hemolytic uremic
syndrome, eclampsia, systemic lupus erythematosus, polyarteritis nodosa, cryoglobulinemia,
and Wegener granulomatosis.5, 6
Not much is known about PRES in kidney disease despite various independent risk factors
being present (sometimes simultaneously) in renal disease, such as hypertension, autoimmune
disease, and immunosuppression. PRES could be underrecognized in patients with end-stage
renal disease (ESRD). Marked hypertension is a cardinal risk factor in this population,
associated with extracellular fluid volume expansion. Neuroimaging findings can be
diverse, involving both anterior and posterior circulation territories. There have
been scattered reports of PRES in kidney disease, but systematic research is lacking.
In Hinchey’s original paper,
1
more than half of the patients had renal failure of varying degrees, and 80% had hypertension.
The findings in our case series are summarized in Table 1.
Table 1
Case patient characteristics
Characteristic
Case 1
Case 2
Case 3
Case 4
Case 5
Age, yr
35
12
12
34
25
Sex
F
F
M
M
F
Renal diagnosis at time of PRES
Renal allograft recipient POD 4
FSGS with mild renal failure, nephrotic syndrome
Renal allograft recipient, 4 months posttransplant, NKD: FSGS
CKD 5D, presumed CGN
HSP nephritis, moderate renal failure
Recent initiation of HD (<4 wk)
No
No
No
No
No
Recent immunosuppression
Yes; tacrolmus
Yes; steroids
Yes- Tacrolimus, Rituximab, steroids
No
Yes; steroids, cyclophosphamide
Clinical features
Seizures
Seizures, altered sensorium
Headache, seizures, altered sensorium
Headache, seizures, vomiting
Altered behavior, drowsiness, staring spells, seizures
BP (MAP)
Acceptable
High (130/90)
High
High
Acceptable
MRI picture
Symmetrical bilateral parieto occipital and frontal white matter hyperintensities
Diffuse gyral swelling and hyperintensity seen affecting the supratentorial brain
parenchyma. This is symmetrical bilaterally and affects all lobes with relative sparing
of watershed areas.
Multiple and symmetric gyral T2 FLAIR hyperintensities noted in the bilateral frontal,
parietal, and temporal cortices and bilateral cerebellar hemispheres
T2 and T2 FLAIR white matter hyperintensities seen in the bilateral high frontal,
centrum semiovale parieto occipital region and also involving the left hemi midbrain
with no asssociated diffusion restriction/blooming
T2 and FLAIR hyperintense signals noted in bilateral basal ganglia lateral and posterior
temporal cortex (left > right) and posterior parietal cortex
Outcome
Resolution after prolonged ventilation
Resolution
Initial resolution, then another episode of symptomatic PRES, which also resolved
completely
Resolution
Resolution prolonged
BP, blood pressure; CGN, chronic glomerulonephritis; CKD, chronic kidney disease;
F, female; FLAIR, fluid-attenuated inversion recovery; FSGS, focal segmental glomerulosclerosis;
HD, hemodialysis; HSP, Henoch−Schönlein purpura; M, male; MRI, magnetic resonance
imaging; NKD, native kidney disease; POD, postoperative day; PRES, posterior reversible
encephalopathy syndrome.
The incidence of PRES posttransplantation varies depending on the type of organ transplanted.
A retrospective single-center study of 4222 solid organ transplant recipients showed
an overall incidence of PRES of 0.49%,
7
with a 0.35% incidence in kidney transplants. This study showed that kidney transplant
recipients with PRES had higher mean arterial pressure at presentation than liver
transplant recipients but lower grades of vasogenic edema. The proposed causes were
many, with calcineurin inhibitor toxicity being the chief culprit. However, toxicity
has been difficult to prove, leading to the thought that it may not be a dose-related
effect. In the article by Wong et al.,
8
most patients with PRES had serum tacrolimus levels in the therapeutic range. Other
studies have shown that there is no correlation with neurotoxicity and the trough
levels of tacrolimus.9, 10 A few case reports have shown high tacrolimus levels in
cerebrospinal fluid, when corresponding serum levels have been within therapeutic
limits.
11
Observations of the onset of PRES with low drug levels is seemingly paradoxical. Both
transplant recipients in our series had PRES while on tacrolimus and had normal tacrolimus
levels, lending weight to the previously mentioned studies. In addition, 1 patient
while on tacrolimus did not have accelerated hypertension at the time of developing
PRES, thus making it difficult to propose an exact mechanism for calcineurin inhibitor−induced
PRES. The association of hypertension with PRES is well reported; however it may not
always be a component in patients on immunosuppressive agents.
11
The interval between the starting of tacrolimus and the development of PRES is also
variable; in our series, the intervals varied between 4 days and 16 weeks. The time
course of developing PRES from the onset of beginning immunosuppressant therapy is
particularly variable.
12
Cyclophosphamide use has been reported in several cases of PRES, but mostly in combination
with other cytotoxic agents for treatment of hematological malignancies and systemic
lupus erythematosus.13, 14, 15 PRES attributed to cyclophosphamide in these reports
was confounded by other contributory factors such as fluid overload, hypertension,
and/or renal failure. In contrast, our patient (20%) had only moderate renal failure
with a controlled BP.
Interestingly, reported recurrence of PRES appears to be infrequent in a reported
case series
7
despite the fact that the same precipitating factors may be present. In our case series,
case patient 3 had a recurrence of PRES, having accelerated hypertension at both times
despite multiple antihypertensive agents, reflecting the role of possible fluctuating
or intermittent hypertension as described earlier. In addition, the first episode
of PRES was associated with tacrolimus, which was then withdrawn. This patient also
had graft dysfunction, and therefore multiple factors may have contributed to the
second episode. Recurrent episodes of PRES have also been described in patients requiring
dialysis, and this group of patients needs close monitoring.16, 17
Neuroimaging is essential to the diagnosis of PRES. Typical findings are symmetrical
white matter edema in the posterior cerebral hemispheres, but variations do occur.
18
Lesions of the anterior cortex have been reported but tend to occur in more severe
cases, usually with edema being present in the posterior circulation territories.
19
Only 1 patient (20%, case patient 5) in our case series had lesions restricted to
the posterior cerebral areas, with the others having diffuse cerebral involvement
(Table 1).
Case series suggest that, in general, PRES has a benign prognosis, usually being reversible
within days to weeks.20, 21, 22 Radiologic improvement lags behind clinical recovery.
All our patients recovered completely with no neurological deficits, including case
patient 4, who had 2 documented episodes. Potentially grave consequences of this disorder
include progressive cerebral edema or intracerebral hemorrhage.
The hallmark of PRES is vasogenic edema. The role of hypertension, often cited as
the main cause of the vasogenic edema, is still controversial. Two theories have been
proposed.
23
In the most popular theory, severe hypertension overwhelms the autoregulation of the
cerebral vasculature, leading to hyperperfusion, arteriolar dilatation, and vasogenic
edema. A predilection of PRES for the posterior regions of the brain has been explained
by a relative lack of sympathetic innervation (as the sympathetic system raises the
threshold of autoregulation). However, BP is not always elevated in PRES, especially
in cases associated with immunosuppression and organ transplantation. In addition,
the degree of vasogenic edema does not always correlate with the severity of hypertension.
7
In the second theory, the principal problem is cerebral vasoconstriction resulting
in downstream hypoperfusion, ischemia, and vasogenic edema due to capillary leak.
Thus, hypertension may not be the sole driver of injury in all cases. Endothelial
cell dysfunction, seen in organ transplant recipients and autoimmune diseases, can
alter vascular tone and cause vasospasm and hypoperfusion. Cerebral vessels respond
to the ensuing hypoxia by secreting vascular endothelial growth factor (VEGF) and
increasing their permeability, thereby resulting in edema.
10
These theories may not be mutually exclusive. Although there is some evidence to support
either theory, retrospective study design, varying terminology, and different imaging
modalities all make it difficult to choose 1 theory over another. In addition, they
may not be distinct from one another. For instance, in our cohort, although hypertension
was a common feature among all 5 patients, the peak mean arterial pressure (MAP) was
below the recognized upper limit of autoregulation in 4 cases. Various BP cutoffs
have been proposed by different authors, a peak SBP usually being between 170 and
190 mm Hg. Other authors have proposed an upper MAP threshold of 116 mm Hg to define
clinically severe hypertension in the context of PRES.2, 12 Using this cut-off, all
5 of our case patients would have been classified as having severe hypertension. In
this context, it is important to recognize that PRES can occur with acute severe hypertension
or rapidly developing, fluctuating, and intermittent hypertension, as well as in patients
with normal BP. Therefore it is important to assess the percentage rise of BP from
the patient’s baseline. Renal failure appears to be prominent among patients with
hypertension who develop PRES, indicating that uremia, fluid overload, and electrolyte
imbalance may be contributing factors.
22
Two of our patients (case patients 1 and 4) had a coexistent systemic inflammatory
disorder with recent exposure to potent immunosuppressive agents. It is plausible
that another process could reduce the upper autoregulatory threshold by altering vascular
dynamics, thus increasing the potential for hypertension-mediated injury. PRES is
not uncommon in the pediatric population,
24
as shown in 2 of our patients (case patients 2 and 3). It is important to consider
this diagnosis in children presenting with seizure, visual disturbances, headache,
and altered mentation in an appropriate clinical setting.
The role of steroids in the treatment of PRES is debatable, with a possible role in
reducing the vasogenic edema. Case patients 2 and 3 were given short courses of i.v.
steroids in their management. It was not possible to be dogmatic about steroid efficacy
in these cases because whether the response was due to steroid effect or withdrawal
of the inciting factor (accelerated hypertension in both cases) was difficult to determine.
Literature evidence has shown multiple theories, with a possible role in systemic
lupus erythematosus (SLE)−related PRES,
25
but no controlled trials have been conducted. Other authors have suggested no role
for steroid therapy in PRES.
26
It remains a controversial subject.
PRES may occur after solid organ transplantation, but the exact incidence is unknown.
In a study published by Bartynski et al.,
27
PRES developed in around 0.5% (27) of 4222 patients who underwent solid organ transplantation
within the study period, thereby showing a low incidence in the transplantation group,
similar to the general population. Further categorization showed that PRES typically
developed in the first 2 months in patients who had liver transplants (9 of 10 patients)
and was associated with cytomegalovirus, mild rejection, or systemic bacterial infection.
PRES also typically developed after 1 year in patients who had a kidney transplant
(8 of 9 patients) and was associated with moderate rejection or bacterial infection.
The composite picture of PRES in the renal transplantation scenario is complex. As
mentioned previously in this article, the use of calcineurin inhibitors, particularly
tacrolimus, predisposes a transplant recipient to PRES even in the presence of apparently
normal drug levels. In this situation, conversion to an alternative regimen such as
cyclosporine or mTOR inhibitor−based therapy may help, but each is fraught with its
own risks. Sirolimus itself has been known to cause PRES, so the decision to substitute
tacrolimus with either cyclosporine or sirolimus should be done taking into consideration
the other side effects associated with each drug. Reduction in drug dosage or prompt
removal of the cytotoxic drug is usually recommended in cases of PRES associated with
cytotoxic agents. However, cases are reported in which symptoms resolve while the
medication is maintained.
28
Another possible strategy is to withdraw the presumed offending drug at the time of
PRES and then re-introduce it at a lower dose once PRES has resolved.
29
Opinions vary on the subject, and the lack of a definitive strategy reflects a paucity
of studies on the condition.
The relationship between PRES and acute rejections is difficult to define. In the
often-cited Bartynski et al. study,
27
PRES developed after 1 year in patients who had SOT of the kidney and was associated
with moderate rejection or bacterial infection. This association is not definitively
causal and may be multifactorial, with elevated BP due to steroid therapy, worsening
renal function, and increased dosage of calcineurin inhibitors being possible contributory
factors. The type of solid organ transplant seems to have little role in determining
the incidence of PRES.
In general, kidney transplant recipients developed PRES many years after transplantation
compared to early in the course in liver transplant recipients. Although the average
BP at the time of PRES development was higher in kidney transplant recipients (average
mean arterial pressure, 143 ± 20 mm Hg), brain edema was greater in liver transplant
patients. PRES was also seen in association with infections and narcotic or cocaine
use.
In conclusion, posterior reversible encephalopathy syndrome can be seen in association
with kidney disease in various scenarios, common ones being accelerated hypertension
and calcineurin inhibitor use. Patients on immunosuppressive regimens need not have
toxic levels of the drug or hypertension at the time of development of PRES. It is
a condition that requires early detection and prompt treatment in view of a generally
benign prognosis.
Disclosure
All the authors declared no conflict of interest.