Article Text


Dramatic improvement in non-AIDS related progressive multifocal leucoencephalopathy
  1. K Auré1,
  2. A Béhin1,
  3. F Louillet1,
  4. C Lafitte1,
  5. M Sanson1,
  6. J P Vernant2
  1. 1Service de Neurologie Mazarin, Groupe Hospitalier Pitié-Salpêtrière, Assistance Publique-Hôpitaux de Paris, Paris, France
  2. 2Service d’Hématologie, Groupe Hospitalier Pitié-Salpêtrière
  1. Correspondence to:
 Dr Anthony Béhin
 Federation de Neurologie-Mazarin, Groupe hospitalier Pitié-Salpêtrière, 47-83 Boulevard de l’hôpital, 75013 Paris Cedex, France;

Statistics from

Progressive multifocal leucoencephalopathy (PML) is a rare disorder occurring when a strain of papovavirus (JC virus) infects the central nervous system. It results in a generally quick and fatal outcome. It is associated with cell mediated immune deficient diseases but some few cases were reported in immunocompetent hosts. Since 1981, it has been commonly associated with AIDS. In AIDS related PML, long term survival without real neurological improvement has been reported in patients treated with highly active antiretroviral therapy (HAART).1 Few cases of improvement with cidofovir or cytosine arabinoside have been described in AIDS related or non-AIDS-related PML,2,3 but in larger trials in AIDS related PML, no clinical benefit was found.4 As a whole, the treatment of this progressive demyelinating disease remains controversial, in particular in the rare cases of non-AIDS-related PML. We describe a patient with an underlying haematological disease, without clear cut immune cell deficiency, who developed rapidly progressive PML. The patient showed clinical, virological, and imaging improvement when treated with an association of intravenous and intrathecal cytosine arabinoside combined with intravenous cidofovir.

A 48 year old man presented with progressive multiple lymphadenopathies, hepatosplenomegaly, weight loss, and blood cell count abnormalities. Fine needle aspiration cytology of lymphadenopathy diagnosed marginal zone B cell lymphoma. There was also bone marrow and blood proliferation. A few weeks after the diagnosis, the patient noticed rotatory vertigo and visual problems suggestive of a right homonymous hemianopia. Because of dissemination and the large tumour mass, chemotherapy (CHOP: cyclophosphamide, doxorubicin, vincristine and prednisone) was started one month after the onset of neurological symptoms. Following the first course of chemotherapy, his neurological symptoms worsened, and language disorders appeared. No real immunodeficiency was shown—the absolute CD4+ count was 549/mm3 (normal 858±260), and there was discrete hypogammaglobulinaemia (4.6 g/l; normal 5–12). The patient was HIV and HTLV1 seronegative. Cerebral magnetic resonance imaging showed a non-contrast-enhancing lesion in the left occipital white matter (fig 1A).

Figure 1

 evolution of cerebral magnetic resonance imaging (MRI) before and during treatment. (A) The first cerebral MRI, two months after the first neurological symptoms. Flair weighted axial images (TR 9000 ms; TE 114 ms) showing high signal intensity lesion in the left occipital white matter with respect to cortical grey matter. (B) Brain MRI one week after treatment onset. Flair weighted axial images (TR 10 002 ms; TE 160 ms) showing extension of the lesions. At that time, extension of the lesions was greater, with bilateralisation of the lesions in both occipital white matter and extension through temporal white matter. (C) Brain MRI three months after treatment onset. Flair weighted axial images (TR 10 002 ms; TE 160 ms) showing dramatic improvement under treatment. Hyperintense signal has diminished in both occipital lobes, but cerebral atrophy had appeared in the left occipital and temporal cortex. TE, time of echo; TR, time of repetition.

Cerebrospinal fluid (CSF) analysis was normal apart from a moderate increase in CSF protein (0.7 g/l). Suspicion of PML was confirmed by a positive polymerase chain reaction (PCR) for JC virus DNA. Chemotherapy was discontinued (after one cycle) but neurological symptoms worsened rapidly and the patient developed a right hemiplegia, global aphasia, alexia and agraphia, apraxia, and cortical blindness concurrently with MRI deterioration (fig 1B).

Three months after his first symptoms, treatment was started with intravenous aracytine 2 mg/kg/d for five days every three weeks, combined with intrathecal aracytine (30 mg) weekly and intravenous cidofovir 5 mg/kg/d once every two weeks. The main adverse effect of this treatment was grade IV bone marrow toxicity, inducing spacing in the rhythm of treatment administration. One week after treatment onset, the patient stabilised and after one month began improving. After three months, he had recovered completely from his hemiplegia, and had significant improvement in his aphasia and cortical blindness. Right hemianopia and minor alexia without agraphia persisted. This dramatic improvement was confirmed by cerebral imaging, by the absence of JC virus DNA detection in CSF, and by a specific response of CD4+ T cells against JC virus. We decided to continue subcutaneous aracytine 2 mg/kg/d for five days monthly and intravenous cidofovir twice weekly. Fifteen months after treatment onset, the patient was ambulatory and cerebral MRI continued to improve (fig 1C). Lymphoma tumour burden did not clearly change during the treatment.


We report the favourable outcome of a patient with non-AIDS-related PML treated with a combination of intravenous and intrathecal aracytine and cidofovir. As the patient was not immunocompromised and had not received immunosuppressive treatment at PML onset, risk factors for the occurrence of PML are unclear. Treatment led to a rapid clinical and radiological improvement which was long lasting despite treatment delay and the patient’s worrying clinical condition at treatment onset. Dose and administration schedules of cytarabine and cidofovir were derived from reports suggesting individual beneficial effects of these drugs.2,3 To our knowledge, these drugs have not been used in combination before. In comparison with previous reports, the present case suggests a more rapid and prolonged effect of this therapeutic combination than with either aracytine or cidofovir treatment alone. This efficacy may be explained by a synergy between the drugs and by their different routes of administration. We thought that the improvement in our patient was related to the treatment because there was a temporal link with treatment onset and because of the radiological features (the rare cases of PML stabilising without specific treatment have usually been associated with an inflammatory response to the virus, indicated by contrast enhancement on imaging).

The main limiting factor of this treatment was bone marrow toxicity. During the periods of immune deficiency, the patient’s neurological condition did not deteriorate, suggesting that PML occurrence in this patient was linked to a qualitative defect of CD4 cells rather than to their absolute count. Immunological studies have shown that JCV specific CD4-T cell responses play a major role in the control of PML development; for example, in one study5 no specific T cell response was demonstrated in a series of 14 patients before treatment, whereas nine of 10 survivors recovered specific immunity. Our patient thus probably had no specific T cell response against JCV before treatment, but recovered a moderate but significant response while on treatment, possibly explaining PML regression. Such a restoration of T cell response can be achieved by HAART in AIDS. This observation suggests that the combination of aracytine and cidofovir could have had a similar action in restoring a specific T cell response against JCV in our patient. A direct effect of these nucleosidic analogues against JC virus DNA may also explain the rapid clinical and radiological improvement in our patient.

Despite its substantial bone marrow toxicity, this observation suggests that the new association of intrathecal and intravenous aracytine with intravenous cidofovir could be useful in patients with PML, particularly those with an underlying haematological disease. It is worth noting that bone marrow toxicity did not lead to deterioration of the neurological status of the patient, supporting the view that a specific defect in CD4 function is more important than the absolute CD4 count. The dramatic improvement observed in our patient warrants further prospective studies testing this drug combination.


We thank Dr Yacine Taoufik, Department of Immunology, Kremlin-Bicêtre Hospital, for his help with the immunological data and critical discussion.


View Abstract


  • Competing interests: none declared

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.