Article Text

Download PDFPDF

Punctate lesion pattern suggestive of perivascular inflammation in acute natalizumab-associated progressive multifocal leukoencephalopathy: productive JC virus infection or preclinical PML-IRIS manifestation?
  1. Mike P Wattjes1,
  2. Lonneke Verhoeff2,
  3. Willemijn Zentjens2,
  4. Joep Killestein3,
  5. Erik T van Munster2,
  6. Frederik Barkhof1,
  7. Jeroen J J van Eijk2
  1. 1Department of Radiology, Nuclear Medicine & PET research, MS Center Amsterdam, VU University Medical Center, Amsterdam, The Netherlands
  2. 2Department of Neurology, Jeroen Bosch Ziekenhuis, Den Bosch, The Netherlands
  3. 3Department of Neurology, MS Center Amsterdam, VU University Medical Center, Amsterdam, The Netherlands
  1. Correspondence to Dr Mike P Wattjes, Department of Radiology, Nuclear Medicine & PET research, MS Center Amsterdam, VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands; m.wattjes{at}

Statistics from


Natalizumab (Tysabri, Biogen-Idec Inc, Cambridge, Massachusetts, USA) is an effective drug for the treatment of multiple sclerosis (MS). Patients treated with natalizumab are at higher risk of developing progressive multifocal leukoencephalopathy (PML). MRI has become crucial in the drug surveillance in terms of an early (preferably presymptomatic) detection and monitoring of natalizumab-associated PML with special regard to the occurrence of an immune reconstitution inflammatory syndrome (IRIS).1

Natalizumab-associated PML is characterised by rather heterogeneous imaging findings particularly in the early disease course. Recently, a rare PML imaging pattern with punctiform enhancing lesions has been described in a patient with an asymptomatic course of natalizumab-associated PML.2 It remains unclear what this inflammation pattern in acute PML represents regarding the underlying histopathology and possible consequences of the disease progression or PML-IRIS development. In this case report, we describe a patient with acute natalizumab-associated PML presenting with punctate enhancing lesions suggestive of perivascular inflammation and discuss the importance of correctly interpreting this imaging finding in terms of differentiating productive JC virus infection from early presymptomatic PML-IRIS manifestation.

Case report

A 45-year-old female with relapsing-remitting MS (5 years disease duration, EDSS 2.0) complained of subacute slurred speech and mild ataxia of her right arm after completing 51 infusions of natalizumab. Natalizumab treatment was immediately discontinued. She had tested positive for serum antibodies against JC virus 6 months earlier. MRI showed multifocal classical PML lesions in the cerebellum and supratentorial brain. In addition, a group of small punctate T2-/FLAIR-hyperintense and contrast-enhancing lesions was identified in the deep white matter of the left parietal lobe, suggesting a perivascular distribution pattern with contrast enhancement (figure 1). Quantitative PCR of the cerebrospinal fluid (CSF) revealed 5301 copies/ml of the JC-virus DNA. The diagnosis of inflammatory PML was established (day 1). Therapy with mirtazepine (60 mg/day) and mefloquine (250 mg/day) was initiated.

Figure 1

Summary of the disease course of our patient demonstrating the diagnostic and therapeutic procedures as well as the corresponding MR images. Bottom row: axial fluid-attenuated inversion recovery (FLAIR). Top row: contrast-enhanced T1-weighted images, except the first (baseline) MRI scan performed on 2 January 2012, which was performed according to the scheduled follow-up scheme without any signs of PML clinically and on MRI. Please note the perivascular inflammation at the time of PML diagnosis and during follow-up (white closed-head arrow) and the subcortical PML lesion (white open-head arrow). During follow-up the perivascular lesion showed progression in combination with an increase of JC virus load in the CSF. At later stages (3 October 2012, 14 November 2012) when the JC virus became undetectable, the perivascular inflammation regressed and disappeared, whereas imaging findings suggestive of IRIS were visible with new areas of contrast enhancement (black circles) in combination with progressive T2-lesions showing subtle mass effect (white closed-head arrows). The timeline presented at the top represents the duration (weeks) of natalizumab treatment. C, copies/ml; CSF, cerebrospinal fluid; Dx, diagnosis; JCV, JC virus; NTZ, natalizumab; PLEX, plasma exchange; PML, progressive multifocal leukoencephalopathy.

Because of subtle clinical deterioration (without radiological progression) during the second week, a course of five plasma exchange treatments was started (natalizumab concentration decreased from 15 to 0.2 μg/ml). At week 4, intravenous methylprednisolone (IVMP) treatment was initiated because of suspected early PML-IRIS based on progressive clinical symptoms and PML lesion progression, particularly of the lesions with perivascular enhancement on MRI, resulting in clinical stabilisation for 1 week. At week 6, clinical decline followed, with anarthria, visual loss, a right-sided hemiparesis and severe ataxia. A second course of IVMP was given and a lumbar puncture was performed. The number of JC-virus DNA copies in the CSF increased to 14 823 copies/ml, suggesting PML progression. Follow-up MRI showed further increase of the contrast-enhancing lesions suggestive of perivascular inflammation. At week 13, MRI showed multifocal new areas of pathological contrast enhancement as well as progressive T2-lesions, indicating PML-IRIS. Simultaneously, the areas of perivascular contrast enhancement regressed and subsequently resolved. At the same time, no JC-virus DNA could be observed in the CSF anymore. After the completion of a new 3-day course of IVMP, she developed pneumonia, which was successfully treated with antibiotics. She clinically improved slightly but still showed signs of PML-IRIS on MRI. She slowly improved upon a new IVMP treatment (week 17) that was tapered over weeks. After the fifth IVMP course at week 21 with tapering over months, her clinical situation improved with residual neurological symptoms including minimal right-sided paresis, moderate ataxia of the extremities, moderate to severe gait ataxia and motor aphasia.


The role of MRI has become crucial in the early and preferably preclinical diagnosis of PML. However, early MRI findings in natalizumab-associated PML are rather heterogeneous. Recently, Phan-Ba and colleagues described a rare type of inflammation pattern with punctate gadolinium-enhancing lesions in the posterior fossa in a patient with an asymptomatic course of natalizumab-associated PML.2

Our case study stresses that punctate enhancing lesions suggestive of perivascular inflammation in the deep white matter of the supratentorial brain reflecting inflammatory PML can be one of the earliest and leading imaging findings in natalizumab-associated PML patients. The clinical relevance as well as the underlying histopathology of this inflammation pattern in PML is not completely understood. Several histopathological studies have demonstrated that perivascular distribution of PML can be observed. However, perivascular inflammation pattern is more prominent in PML-IRIS demonstrated by CD8-positive T-cells in the perivascular spaces.3 From the imaging point of view, it is challenging to determine the correct meaning of this inflammation pattern since even in early stages PML and PML-IRIS can occur simultaneously. It is crucial to have a correct imaging-based diagnosis because corticotherapy for suspected IRIS at the stage of productive JC virus infection might very well lead to inferior clinical outcomes.4

This case report describes this clinical dilemma in detail. Initially, we made the diagnosis of inflammatory PML, but based on the increased perivascular enhancement after natalizumab discontinuation we changed our diagnosis to PML-IRIS, leading to early corticosteroid therapy. In retrospect, it is most likely that the initial perivascular inflammation and enhancement was due to productive JC virus infection, which is also reflected by the substantial increase of virus load in the CSF and the corresponding increase of PML manifestations on MRI with stronger perivascular enhancement. Another argument is the regression of the enhancement suggestive of perivascular inflammation after the virus became undetectable in the CSF and more classical IRIS imaging findings were observed. Therefore, longitudinal CSF analysis in combination with follow-up MRI may substantially aid in this difficult diagnostic process.

A similar type of inflammation, tiny punctate T2-lesions adjacent to the PML white matter manifestations, has been observed in 72% of natalizumab-associated PML patients.5 Since these lesions were associated with gadolinium enhancement in PML-IRIS stages, it has been suggested that these lesions might represent an early immune response in the perivascular spaces which will become more prominent in PML-IRIS stages.5

In conclusion, the presented punctate enhancing lesion type pattern suggestive of perivascular inflammation in early and/or preclinical natalizumab-associated PML emphasises the broad spectrum of imaging findings we have to be aware of. This pattern particularly at early stages does not exclusively represent PML-IRIS but can be based on productive JC virus infection with consequences for the correct treatment strategy and clinical outcome.


The authors are thankful to this PML patient for agreeing to present her clinical information and MRI material and to Iris D Kilsdonk for the help in preparing the figure.



  • Contributors MPW and JJE contributed to the conception, design, analysis and interpretation of the data and drafting the article. LV, WZ, JK, EvM and FB contributed to the interpretation of the data and drafting the article. All authors approved the final version of the manuscript.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Since this was an observational case study with permission from the patient, a formal review by the Institutional Review Board at our institution was not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.