Departments/Institutions to which this work is attributable: Departments of Paediatric Nephrology, Infection and Paediatric Neurology, Guy's and St Thomas' NHS Foundation Trust, and the Virus Reference Department, Centre for Infections, Health Protection Agency.
Two children, boys of 8 and 13 years, presented with measles-associated encephalopathy several years after kidney transplantation for congenital nephrotic syndrome. In the absence of prior clinical measles, the neurological symptoms initially eluded diagnosis, but retrospective analysis of stored samples facilitated the diagnosis of measles-associated encephalopathy without recourse to biopsy of deep cerebral lesions. Each had received a single dose of measles mumps and rubella vaccine before 12 months of age. Prior vaccination, reduction of immunosuppression and treatment with intravenous immunoglobulin and ribavirin may have contributed to their survival. Persistent measles virus RNA shedding, present in one child, was not controlled by treatment with i.v. ribavirin. Two years later, both patients continue to have functioning allografts with only minimal immunosuppression. These cases illustrate the difficulty in diagnosing measles-associated encephalopathy in the immunocompromised host, even in the era of molecular diagnostics, and highlight the renewed threat of neurological disease in communities with incomplete herd immunity.
Measles vaccination was successfully implemented in developed countries prior to the widespread expansion of transplant programs. Vaccination is contra-indicated following organ transplantation, and where possible children are vaccinated pre-transplant. Neither prior vaccination nor detectable measles immunoglobulin G (IgG), however, ensures protection in the immunocompromised and administration of human normal immunoglobulin (HNIG) is recommended following known exposure to measles (1,2). Organ recipients are afforded protection through herd immunity and the resultant lack of exposure to infection. Measles inclusion body encephalitis (MIBE) has therefore rarely been implicated as a cause of encephalitis in the transplant setting. In an apparent reemergence of this complication, we now describe two temporally related cases of measles-associated encephalopathy, occurring many years following transplantation.
Case 1, an 8-year-old male renal transplant recipient, presented with headache and focal neurological signs (day 0). He had developed nephrotic syndrome in association with diffuse mesangial sclerosis at 7 months and at 18 months received a live-related renal transplant. Post-transplant induction therapy included anti-lymphocyte globulin followed by maintenance cyclosporine (CsA), azathioprine and prednisolone. Seventy-two days (day −72) prior to neurological presentation, MMF was substituted for azathioprine to facilitate CsA withdrawal because of chronic allograft nephropathy.
On day −44, he was treated with amoxicillin and clavulanic acid for pharyngitis. Five days later he was more unwell with fever, increasing drowsiness, diarrhea, bilateral conjunctivitis and a fleeting diffuse, maculopapular rash over his trunk, chest and arms. Chest X-ray showed bilateral streaky shadowing for which he was treated with i.v. antibiotics.
Six weeks later (day −2), he presented with a week's history of drowsiness, worsening right frontal headaches and unsteadiness. He was afebrile with a normal neurological examination. MMF was discontinued. On day 0, he was admitted with weakness of his left leg with associated paresthesia and myoclonic jerks. An initial computed tomography (CT) brain scan was normal and he was commenced empirically on cefotaxime, clarithromycin and acyclovir.
By day 3, he had developed dysarthria, intermittent ophthalmoplegia and an evolving left hemiparesis with associated myoclonus. He suffered a generalized tonic-clonic seizure and was commenced on phenytoin and sodium valproate. Electroencephalogram (EEG) showed rhythmic slow waves with intermittent focal sharp waves and magnetic resonance imaging (MRI) revealed altered signal in the right basal ganglia, predominantly in the lentiform nucleus (Figure 1A). Laboratory indices of cerebrospinal fluid (CSF) were unremarkable. Microscopy, bacterial culture, cryptococcal ag and CSF polymerase chain reaction (PCR) for herpes simplex virus, varicella zoster virus, enteroviruses, cytomegalovirus, Epstein-Bull virus, JC virus, human herpes virus 6 and toxoplasma gondii were negative. By day 4, he required assisted ventilation for fluctuating conscious level and bulbar palsy. Repeat MRI showed marked progression in the abnormality of the right lentiform nucleus with new changes in the subthalamic nuclei, the right cerebral peduncle, the left basal ganglia and the medulla oblongata.
Encephalitis, vasculitis or pediatric autoimmune neuropsychiatric disorders associated with streptococcal infection (PANDAS) were considered. On day +10, a single dose of intravenous immunoglobulin (IVIG) was commenced in view of slight elevation of markers of streptococcal infection. Clinical improvement coincided with this. He showed no further neurological deterioration, but required a tracheostomy on day +12 for continuing bulbar palsy.
On day +17 serum from day +1 tested measles immunoglobulin (IgM) positive and IgG equivocal (Figure 2). Sequential samples (collected prior to IVIG) confirmed recent seroconversion. There was evidence of ongoing infection with measles RNA detected in multiple samples, including serum from day −44, later genotyped as wild-type measles virus strain D8. No CSF was available for measles PCR testing. CsA was discontinued. Intravenous ribavirin (33 mg/kg loading dose then 16 mg/kg 6 hourly for 4 days followed by 8 mg/kg 8 hourly) was administered for 3 weeks. Despite this, measles RNA remained detectable in multiple clinical samples. Retrospective testing showed him to be seronegative for measles when transplanted despite a single dose of measles mumps and rubella (MMR) vaccine at 10 months of age.
Over the next 12 months, he steadily improved and is walking with a wide-based gait with residual mild swallowing difficulty and dysarthria following tracheostomy reversal. His current immunosuppression is azathioprine 50 mg once daily (od) and prednisolone 10 mg on alternate days. His creatinine remains stable at around 80 μmol/L.
Case 2, a 13-year-old male renal transplant recipient, presented 18 days after Case 1, with a generalized tonic-clonic seizure and bilateral leg pain (day 0). He had been well but for another episode of recurrent conjunctivitis 2 weeks previously. On admission, he was afebrile and alert with painful cramps in his legs, and myoclonus and weakness of his left leg. The lower limb reflexes were bilaterally brisk.
In the past, Finnish-type congenital nephrotic syndrome was diagnosed at 6 weeks of age. He underwent bilateral nephrectomy and dialysis at 15 months and, at 28 months, received a cadaveric renal transplant. Induction therapy and maintenance were as for Case 1. At 13 years of age, 3 months prior to presentation, MMF was likewise substituted for azathioprine.
An EEG on day 1 revealed abnormal parasagittal discharges and MRI showed an area of signal change in the right thalamus (Figure 1B). Lumbar puncture revealed a raised CSF protein of 800 mg/L (normal < 400 mg/L) with other indices in the normal range. Gram stain, culture and CSF PCR for viruses, including measles, and toxoplasma gondii were negative. CSF measles IgG was negative.
Treatment with cefotaxime, clarithromycin and acyclovir was commenced. By day 3, he had an evolving left hemiparesis, mild left upper motor neurone lesions, facial weakness and epilepsia partialis continua (EPC) of his left foot and arm. MMF was stopped, CsA reduced and prednisolone increased. On day 4, he developed bilateral optic neuritis and central retinitis. His seizures were difficult to control and by day +6 he required intubation and ventilation for 9 days for fluctuating conscious level and reduced gag reflex.
On day 7 a single dose of IVIG 2 g/kg was commenced. On day 14, a serum sample from day 7 was shown to be measles IgG and IgM positive with seroconversion confirmed by comparison with a previous sample (Figure 2). Measles RNA, however, was not detected in any sample. With a presumptive diagnosis of measles-associated encephalitis, he received i.v. ribavirin as for Case 1. CsA was stopped and he was maintained on daily prednisolone alone. He had received a single dose of MMR vaccine at 8 months of age but, by retrospective testing, was seronegative for measles when transplanted.
He has made slow progress and 2 years later is wheelchair bound, registered blind and with limited communication. His current immunosuppression consists of azathioprine 75 mg od and prednisolone 10 mg on alternate days and his creatinine remains stable at around 75 μmol/L.
The Centers for Disease Control clinical case definition for measles specifies (a) a generalized maculopapular rash for 3 days or more, (b) fever of 38.3°C, and (c) cough, coryza or conjunctivitis (3). Although both had been non-specifically unwell, neither of these two children was diagnosed with measles prior to presentation with encephalopathy. Even in retrospect there were few clues. Case 2 remained at home with isolated conjunctivitis—a recurrent problem—while Case 1 had presented with symptoms suggestive of a viral infection—fever, conjunctivitis, coryza, chest X-ray changes and an evanescent rash. In both cases, these symptoms were only later attributed to measles. The lack of a prior measles diagnosis or known measles exposure is well recognized in individuals with deficient cellular immunity who develop measles complications: the typical rash and/or other characteristic features of measles are frequently absent. Thus, while immunocompromised individuals suffer specific complications—giant cell pneumonia and MIBE—these often arise unheralded. Where there is a known exposure to measles, symptoms of giant cell pneumonia first appear after 2–3 weeks, whereas the median interval to presentation with MIBE is 4 months (range 1–7 months) (4). In a review of measles in 68 immunocompromised patients, including a single organ transplant recipient, 27–40% cases had no rash (5). Eighty per cent suffered severe complications with fatality rates of 70% in patients with malignant tumors and 40% in those with HIV infection. The data reviewed included that from early leukemia trials in the United Kingdom where 15 of 51 (30%) deaths in children in first remission were due to measles or its complications.
Measles was considered as a possible cause in Case 1 only after extensive molecular analysis and serological testing had failed to identify a cause for his neurological condition. Due to the relative inaccessibility of the lesions, brain biopsy was not performed in either case. There is, however, compelling evidence for the diagnosis of measles-associated encephalopathy: measles IgM was detected in serum, and retrospective analysis of stored samples showed seroconversion to measles coincident with neurological presentation in both cases (Figure 2). Evidence for active measles infection was sought and measles RNA, but not culturable virus, was detected in samples of saliva, plasma, nasopharyngeal aspirate and urine of Case 1 dating back to his presentation on day −44 and persisting for several weeks after his neurological condition had stabilized. In contrast, Case 2 was not associated with evidence of peripheral viral replication (Figure 2). The absence of measles RNA in CSF does not exclude measles as a cause of encephalitis. On the contrary, MIBE is characterized by the absence of both infectious measles virus and measles-specific Ab in the CSF. In the absence of brain biopsy or autopsy evidence of inclusion bodies, probable cases are more correctly designated ‘sub-acute measles encephalitis’ (SME).
The two cases may be directly linked. Case 1 lived in South East London and was infected with measles virus strain D8, the wild-type virus strain implicated in local outbreaks (6). Both had attended the same pediatric transplant outpatients clinic when Case 1 presented with symptoms of a viral illness (Case 1, day −44; Case 2, day −62), now known to be measles (Figure 2). As samples from Case 2 yielded neither measles virus nor RNA, it was not possible to confirm or rule out this possibility. This does seem to be the most likely explanation however, as no measles outbreaks were reported from Case 2's more rural locality to account for his infection.
Of note, both patients were afebrile when they presented neurologically. Fever is not a feature of SME, noted in only 1 of 33 cases. Both had altered levels of consciousness and developed seizures, features almost invariably present in SME (4). Both also developed EPC, a rare disorder manifest by spontaneous, continuous, regular or irregular muscle twitching, which is focal and may persist for prolonged periods—hours, days or weeks (7). The commonest cause of EPC is thought to be Rasmussen's encephalitis but it is a known feature of SME and may occur in as many as a third of cases (4).
Just three cases of measles-associated encephalitis have been reported following organ transplantation, all of them in kidney transplant recipients. The main features of these are summarized, together with the present cases, in Table 1 (8–10). The clinical features in all five were similar. Common presenting features including seizures, and in particular EPC, hemiplegia and altered level of consciousness. The course of the disease is usually rapidly progressive, and ultimately was fatal in the three prior cases.
Table 1. Sub-acute measles encephalopathy in organ transplant recipients
Underlying renal diagnosis
Measles vaccination pre-transplant
Time since transplant (years)
Immunosuppression at presentation
Preceding clinical measles? (interval to CNS disease)
Coincident seroconversion. Measles RNA detected from multiple sites
Survived Residual dysarthria and wide-based gait
Congenital nephrotic syndrome
Yes. Single dose MMR at 8 months
MMF Prednisolone CsA
Myoclonus Generalized seizures EPC Hemiparesis Bilateral optic neuritis and retinitis Altered level of consciousness
MRI – right thalamic changes
Abnormal parasagittal discharges
Survived Registered blind. Limited mobility
While the diagnosis of MIBE can only be confirmed by brain biopsy, the following combination of clinical findings are said to be highly suggestive: (a) history of measles or measles exposure within 7 months of presentation in an unimmunized, immunocompromised patient; (b) refractory focal seizures with or without subsequent generalized activity; (c) absence of fever; and (d) normal CSF analysis. These criteria apply to these cases except perhaps in one respect—prior immunization. Each child had received a single dose of MMR vaccine, but in effect they were unimmunized—at least from the perspective of humoral immunity. Both were seronegative for measles at the time of transplantation in retrospective testing of stored sera. Several factors might account for the lack of serological response. Just 80% dialysis patients seroconvert following a single dose of MMR vaccine, compared with 90% healthy 15-month olds (11). Vaccine was administered at 10 and 8 months, respectively—below the recommended age for MMR (12–15 months in the United Kingdom)—and neither received a second dose. Nephrotic syndrome, the pre-transplant diagnosis in both patients, is accompanied by immune deficits. Nephrotic patients have lower levels of IgG, both during relapse and in remission (12). Thus they may respond less well to vaccination than other children. Interestingly, two of the three previously described cases of measles-associated encephalopathy in transplant patients also had nephrotic syndrome as their pre-transplant diagnosis and one of these had received a single dose of vaccine in infancy (Table 1).
Why did both these children survive while all three previously reported transplant recipients with measles encephalopathy died? Did prior vaccination, albeit suboptimal—-as noted above—ultimately tip the balance in their favor? Net immunosuppression had been increased in both with the switch from azathioprine to MMF shortly prior to presentation. Immune reconstitution through immunosuppression dose reduction, not a therapeutic option in the majority of reported cases, was perhaps the most important therapeutic maneuver in stalling disease progression. IVIG, given to treat possible PANDAS, may have been beneficial. Any advantage mediated by ribavirin seems unlikely to represent inhibition of viral replication given the continued detection of measles RNA (4). That both retained their transplants may reflect the immunosuppressive effects of measles. Nearly two years later, both patients' transplants are still functioning well on minimal immunosuppression.
These two cases arose in the course of a measles outbreak of 169 confirmed cases in London (13). Measles continues to be a threat. With MMR uptake as low as 60% in some areas, the risk of endemic measles in London has prompted the recent MMR ‘Catch-up Campaign’ in London schools (14,15). Nor is this problem confined to the United Kingdom. Insufficient MMR immunization in other developed countries—Japan, Germany, Ireland, for example—put the growing numbers of immunocompromised patients at risk. Even in the United States, where MMR uptake is high, susceptible individuals may unknowingly be exposed to sporadic cases. This was illustrated only recently in Chicago, when MIBE was the surprise diagnosis made following brain biopsy of a pediatric bone marrow transplant recipient, with no history of either exposure or foreign travel (16).
What can be done? Worldwide eradication of measles is a realizable goal. Indigenous transmission of measles in the Americas has all but ceased but there is still much to do in the developing countries of the third world. The new World Health Organization target for elimination in Europe is 2010 (17). In the meantime all organ transplant recipients should receive two doses of MMR if possible and have detectable measles-specific antibodies when listed for transplantation. Some advocate annual IgG testing, with revaccination and suspension from the transplant list as required (18). Despite these precautions, however, neither prior vaccination, nor seropositivity for measles IgG, ensures protection in the immunocompromised, and administration of HNIG is recommended following known exposure to measles (1,2). The risk of household exposure can be minimized by ensuring that all healthy close contacts of immunocompromised transplant patients have received two doses of MMR. Indeed, at a community level, all efforts should be made to maximize the herd immunity through high uptake of MMR immunization.
Staff in highly specialized units need to be aware of measles outbreaks both locally and further afield. Even in the absence of known cases, measles warrants consideration in the differential diagnosis of unexplained clinical findings. Measles will remain a special threat to transplant recipients and other immunocompromised individuals until it has been successfully eradicated.
BC and LJ were responsible for the measles serology, PCR and sequencing. AT, DJ, FH, MS, EH, MK, SR and EMM were actively involved in management of the patients and contributed to the manuscript.
EMM has received sponsorship from Aventis Pasteur MSD Ltd and SmithKline Beecham toward conference attendance in the last 5 years. The authors would like to thank Miss Alice Gem for secretarial help.