Clinicopathologic Reports, Case Reports, and Small Case Series
December 2004

Cortical Blindness Due to Reversible Posterior LeukoencephalopathySyndrome in a Patient With Thrombotic Thrombocytopenic Purpura and Preeclampsia

Author Affiliations



Copyright 2004 American Medical Association. All Rights Reserved.Applicable FARS/DFARS Restrictions Apply to Government Use.2004

Arch Ophthalmol. 2004;122(12):1885-1887. doi:10.1001/archopht.122.12.1885

Neurological involvement in thrombotic thrombocytopenic purpura (TTP)is frequent. In one series, magnetic resonance imaging revealed brain lesionsin 88% of the patients with TTP.1 The 2 mostcommon cerebral lesions associated with TTP are edema and infarction.1,2 Cerebral edema predominantly affectsthe white matter, but when it affects gray matter in the territory of theposterior cerebral circulation, it may resemble the radiological findingsof reversible posterior leukoencephalopathy syndrome (RPLS). We report a caseof reversible cortical blindness caused by RPLS in a patient with TTP exacerbationand preeclampsia. To our knowledge, there have been only 10 previously reportedcases of RPLS in the setting of TTP.1,2 Wealso discuss the role of diffusion-weighted imaging (DWI) in differentiatingreversible from irreversible ischemic lesions.

Case Report

A 19-year-old, 28-week pregnant white woman came to the hospital withgross hematuria and decreased urine output. Her medical history was significantfor recurrent episodes of TTP. Her pregnancy was complicated by 3 prior admissionsfor TTP exacerbation. Each time she responded to fresh frozen plasma transfusion.Her medications included iron supplements and prenatal vitamins.

At admission her blood pressure was 175/84 mm Hg, and she was afebrile.The remainder of the physical examination results were unremarkable. Laboratorytesting revealed a hematocrit level of 27.7%, platelet count of 13 × 103/μL and a creatinine level of 2.0 mg/dL (176.8 μmol/L). A peripheralblood test revealed schistocytes. Results of liver function tests and routinecoagulation studies were normal. We diagnosed TTP exacerbation, and the patientwas given daily fresh frozen plasma transfusions. Her blood pressure was controlledwith intravenous labetalol hydrochloride, ranging from 130/70 mm Hg to 140/80mm Hg. However, on day 3 of the hospital admission, her blood pressure reachedsystolic levels of 160 to 170 mm Hg and diastolic levels of 80 to 90 mm Hgdespite aggressive treatment with labetalol and the addition of hydralazinehydrochloride. Later the same day, she developed a severe headache, nausea,and vomiting, and ultrasonography revealed no fetal cardiac activity. Thefollowing day, she noted a sudden loss of vision in both eyes, which she describedas total blackness. Her blood pressure at the time of visual loss was 183/96mm Hg. Results of a cranial computed tomographic image without contrast wereinterpreted as normal. The visual acuity was hand motions OU. Both pupilswere briskly reactive to direct light with no relative afferent pupillarydefect. She did not generate nystagmus in either eye with the optokineticdrum. Anterior segment and dilated fundus examination findings were normal.Neurological examination revealed a lethargic, irritable patient with no focaldeficits. A cranial magnetic resonance imaging study showed increased signalintensity of the gray and subcortical white matter on fluid-attenuated inversionrecovery and T2-weighted images bilaterally within the posterior cerebralartery territories associated with mild mass effect, reflected as narrowingof the sulci in the involved areas. The DWI demonstrated an isointense signalin the same distribution as the T2-signal abnormalities, except for a smallincreased signal lesion in the right posterior temporal lobe (Figure 1). The following day, after an induced delivery of a stillbornfetus, the visual acuity improved to 20/200 OU, and the blood pressure normalized.At a follow-up examination 5 weeks later, the visual acuity improved to 20/20OD and 20/25 OS. Results of automated perimetry were normal in both eyes.Repeated magnetic resonance imaging demonstrated resolution of the previoussignal abnormalities in the temporal and occipital lobes (Figure 2).

Figure 1.
Image not available

Magnetic resonance imaging studyat the time of visual acuity examination measuring hand motions OU. A, Axialfluid-attenuated inversion recovery image through the level of the temporaland occipital lobes demonstrates bilateral asymmetric areas of increased signalintensity involving the gray and subcortical white matter of the posteriortemporal and occipital lobes. B, Diffusion-weighted image (DWI) shows allbut 1 lesion in the right posterior temporal lobe isointense to the normal-appearingbrain parenchyma. The lesion in the posterior temporal lobe (arrow) showsincreased signal intensity, suggestive of restricted diffusion. C, The apparentdiffusion coefficient (ADC) map does not demonstrate decreased signal intensityin the area of interest (arrow), as would be expected based on the DWI, butrather increased signal intensity suggestive of vasogenic edema. The contradictingfindings on the DWI and ADC map are due to the T2 “shine-through”phenomenon.

Figure 2.
Image not available

Follow-up magnetic resonance imagingstudy with visual acuity measuring 20/20 OD and 20/25 OS. Axial fluid-attenuatedinversion recovery (A) and diffusion-weighted (B) images obtained 4 weeksafter the initial magnetic resonance imaging study showed complete resolutionof the previous signal abnormalities.


Our patient’s ocular examination results, radiological findings,and clinical course were consistent with RPLS. Previously referred to as hypertensiveencephalopathy, RPLS has been associated with a variety of conditions includingpreeclampsia, eclampsia, an adverse effect of chemotherapeutic drugs, acuterenal insufficiency, and TTP.1,2

The exact mechanism of RPLS remains unclear. Most patients have elevatedsystemic blood pressure and develop reversible cerebral edema, with cerebralinfarction occurring rarely.1,2 Twomain theories have been proposed to explain the pathomechanism of the disease.The first suggests an acute rise in systemic arterial blood pressure inducingautoregulatory cerebral vasospasm, which results in ischemia and subsequentcytotoxic edema. When severe enough, the ischemia may progress to infarction.3 The second theory proposes that the acute hypertensionresults in dysautoregulation and passive vasodilation of the cerebral vasculature,resulting in hyperperfusion and extravasation of fluid into the interstitium(vasogenic edema). Recent radiological studies using DWI have demonstratedthat in most patients with RPLS, the cerebral edema is vasogenic in originand characterized by increased diffusion of water molecules, in contrast tocytotoxic edema, which shows restricted water diffusion.24 Furthermore,the reversibility of the lesions and their preference for the posterior circulationwith less vasomotor sympathetic innervation supports a vasogenic pathogenesis.

Before the advent of DWI, it was difficult to differentiate betweencytotoxic and vasogenic edema in patients with RPLS because they both producea bright signal on T2-weighted images. On DWI studies, vasogenic edema producesa dark or isointense signal (high diffusion) compared with cytotoxic edema,which produces a bright signal.2,4 However,a bright signal on DWI studies can sometimes be observed with vasogenic edemabecause of the prolongation of the T2 signal.2 Thisphenomenon, known as the T2 “shine-through” effect, occurs becausethe DWI is influenced by water diffusibility and the intrinsic T2 propertiesof the tissue.5 The difficulty of interpretingthis finding is eliminated by generating an apparent diffusion coefficientmap, which reflects only the diffusion characteristics of the lesions andis independent of the T2 shine-through effect.5 Onapparent diffusion coefficient maps, lesions with restricted water diffusion(ie, cytotoxic edema) are hypointense, and lesions with increased diffusion,as in our case of vasogenic edema, generate normal to increased signal intensity.The distinction between cytotoxic and vasogenic edema is important becausethe treatment and prognosis are different. Vasogenic edema is reversible withprompt treatment of the underlying cause. Cytotoxic edema, however, suggestsan acute infarction and may warrant invasive diagnostic studies and immediatethrombolytic therapy.

In our patient, the mechanism of the cerebral lesions was probably multifactorial,with TTP and preeclampsia as predisposing factors. Widespread endothelialinjury, hypertension, and renal failure most likely contributed to the developmentof vasogenic edema. Treatment with fresh frozen plasma interrupted plateletthrombi formation, preventing further endothelial damage and resulting inimprovement of renal function, blood pressure, and vasogenic cerebral edema.

Back to top
Article Information

Correspondence: Dr Bhatti, Department ofOphthalmology, University of Florida College of Medicine, Box 100284 JHMHSC,Gainesville, FL 32610-0284 (

Financial Disclosure: None.

Author Affiliations: Departments of Ophthalmology(Drs Soltes and Bhatti), Radiology (Dr Schmalfuss), Neurology (Dr Bhatti),and Neurosurgery (Dr Bhatti), University of Florida College of Medicine, Gainesville.

Bakshi  RShaikh  ZABates  VE  et al.  Thrombotic thrombocytopenic purpura: brain CT and MRI findings in 12patients. Neurology 1999;521285- 1289
Mukherjee  PMcKinstry  RC Reversible posterior leukoencephalopathy syndrome: evaluation withdiffusion-tensor MR imaging. Radiology 2001;219756- 765
Sundgren  PCEdvardsson  BHoltas  S Serial investigation of perfusion disturbances and vasogenic oedemain hypertensive encephalopathy by diffusion and perfusion weighted imaging. Neuroradiology 2002;44299- 304
Schaefer  PWBuonanno  FSGonzales  RG  et al.  Diffusion-weighted imaging discriminates between cytotoxic and vasogenicedema in a patient with eclampsia. Stroke 1997;281082- 1085
Provenzale  JMEngelter  STPetrella  JR  et al.  Use of MR exponential diffusion-weighted images to eradicate T2 “shine-through”effect. AJR Am J Roentgenol 1999;172537- 539