Context
Progressive multifocal leukoencephalopathy (PML) is a demyelinating disease of the central nervous system (CNS) caused by the JC virus. This virus is ubiquitous in the population but may become pathological in immunocompromised individuals, such as patients with multiple sclerosis (MS) having im- munosuppressive treatments, AIDS, or in the context hematological malignancies. This infection causes a progressive destruction of infected oligodendrocytes and astrocytes and of the white matter in the CNS. The clinical signs of this life-threatening condition may include cognitive alterations, sensitive troubles disorders, paresis, hemianopsia, cerebellar symptoms or altered mental status.
The treatment strategy of PML aims to foster the immune system reconstitution. However, the patient may encounter an excessive reaction of the immune system, called IRIS, leading to a paradoxical neuro- logical degradation and MRI signs not associated with clinical PML.
Usually, the diagnostic of PML is established through a combination of compatible clinical state, imaging features and a positive PCR for the JC virus in the CSF.
Biomarkers have been proposed to gain insight about the disease’s activity in the CNS. For example, they can help to estimate neuronal loss in the patient’s brain. For this purpose, neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP), beeing proteins expressed in specific CNS cells, are a neuronal biomarker candidate. Indeed, NfL or GFAP levels increase proportionally to the neuronal damage degree in various neurological disorders. More specifically, NfL, either measured in serum or CSF, showed some interest for estimating MS activity and for its follow up. Furthermore, serum NfL levels usually increase among pre-PML patients. It could also indicate the development of an IRIS when NfL levels increase even further.
Ob jectives
Hence, the NfL and glial fibrillary acidic protein (GFAP) can be used as a biomarker for both the amount of neuronal loss estimation and the PML occurrence monitoring. The correlation between serum and CSF levels is known but we aim to confirm it here. However, whether they can be used to estimate PML prognosis is not actually known. This interrogation constitutes the main subject of this study. Secondarily, we are interested in investigating into the longitudinal correlation between these two factors.
Methods
This clinical research study is based on the Cohort for research on PML and immune reconstitution syndrome (CORPUS) database. This study includes all patients over 18 years old having a PML from any causes. Some patients suffered with various predisposing conditions including, but not limited, to MS patients treated with immunosupressors, HIV patients or patients suffering from an hematological malignancies. For this work, we would focus only to patients from Lausanne University Hospital. The disease evolution and outcome is quantified thanks to a scale called Modified Rankin Scale (mRS), commonly used to measure neurological disability. The statistical analysis is done using python code and use mainly linear regression to study correlation and Mann-Whithney test to explore difference amoung groups.
Results
In this study, we found correlation between CSF and serum biomarkers levels for both NfL and GFAP. Also, biomarkers concentration measures at the time of PML diagnostic but also biomarkers concentration evolution seem to have an impact on the prognosis and thus may be used in a clinical context.
Conclusion
The correlation between serum and CSF levels allow to monitor biomarker levels rather in the blood, which is safer and simpler for the patient. The biomarkers concentration and evolution seem to be have indication about the prognosis. However, it is not significant in all cases. This may be tackled by exploring with other statistical tool but also with more data, thus including other centers.