Application of the CSF JCV antibody index to early natalizumab-associated progressive multifocal leukoencephalopathy.
Warnke C1,2, Wijburg MT3,4, Hartung HP1, Killestein J4, Adams O5, Wattjes MP3.
Progressive multifocal leukoencephalopathy (PML) is a severe adverse event in patients with multiple sclerosis (MS) treated with natalizumab. Brain MRI can detect small PML lesions at an early disease stage.1 Therefore, MRI screening is recommended for patients at high risk of developing PML.2 Diagnosis of PML is confirmed by brain biopsy or detection of JC virus (JCV) DNA from cerebrospinal fluid (CSF).3 However, in spite of using ultrasensitive quantitative PCR (qPCR) protocols, patients with PML can be JCV DNA negative in CSF, hampering the diagnosis. Recently, we demonstrated that the detection of intrathecally produced anti-JCV IgG antibodies is highly specific for diagnosis of PML, making the so-called CSF JCV antibody index (AIJCV) a candidate tool for earlier diagnosing of PML in a proportion of natalizumab-associated PML patients.
The aim of this study was to assess the value of the AIJCV in a case series of natalizumab-treated patients with MS undergoing enhanced MRI pharmacovigilance for PML.
I spent better half of two years complaining to anyone who'd listen that the JCV antibody index as it stands was not fit for purpose. Firstly, as a diagnostic test it doesn't have good specificity (the proportion of subjects without the target condition, in this case PML-Progressive Multifocal Leucoencephalopathy, in whom the test is negative) or sensitivity (the proportion of subjects with the target condition in whom the test is positive) required by the regulatory agencies, including the FDA. This is because the target and the control group (those who didn't develop PML) JCV raw values overlap. Secondly, more of a technical issue, as a biochemist when I talk about an index, I specifically mean a ratio between the CSF and serum taking into account any blood brain barrier (BBB) dysfunction (i.e. transference from the blood into the CSF of proteins because of a leaky BBB), which is an indicator of the relative amounts of CSF IgG (antibody production) compared to the serum. A raised index in the latter scenario is an indicator of IgG/antibody production in the CNS. Therefore, if I was designing a JCV IgG index to study antibody production against the JCV in the brain, I would do this. Anyway, I dwell in possibility.
Although, it would appear that I didn't have to wait too long...Warnke et al. have developed a CSF JCV antibody index (that detects CSF produced anti-JCV IgG antibodies) for the diagnosis of PML. They recruited subjects with an MRI suspicious of PML for this study, and collected paired serum and CSF to calculate the JCV antibody index in the CSF. The laboratory methods are not available in this paper, it's like following the white rabbit down a rabbit hole, if you click on the links for long enough, a whole Wonderland of adventure begins!
Their findings were, at the time of PML suspicion on MRI, JCV DNA (indicative of virus replication in the CSF) was undetectable in 4 out of 8 cases (i.e. 50%). Only 25% of cases had a positive CSF JCV antibody index at the time of first PML suspicion; although one of the cases was PCR negative but index positive.
The AAN (American Academy of Neurology) defines definite PML diagnosis as symptomatic disease (i.e. has clinical signs and symptoms consistent with PML) confirmed by MRI and the detection of JCV DNA in the CSF. As it stands the CSF JCV antibody index needs work, and wouldn't replace the JCV DNA PCR as confirmation in cases of PML suspicion, although it may provide added value. Yearly MRI, is currently the most sensitive way of screening for PML. As far as the index is concerned the steps being taken now at least make more logical sense. The mammoth task of now testing more cases and controls awaits the group.