Stem Cell Transplant Linked to Decreases in Multiple Sclerosis Biomarkers
The proportion of patients with concentrations of NfL and MBP above the upper limit of normal decreased from 67% and 63%, respectively, to 12% at 5 years.
A majority of patients with relapsing-remitting multiple sclerosis (RRMS) who received autologous hematopoietic stem cell transplantation (aHSCT) showed decreases in cerebrospinal fluid (CSF) concentrations of MS disease biomarkers neurofilament light (NfL) and myelin basic protein (MBP) to normal levels within 5 years of treatment, according to results from a single center study published in Multiple Sclerosis and Related Disorders.
Of the 43 patients included in the data, the proportion of patients with concentrations of NfL and MBP above the upper limit of normal at baseline decreased from 67% and 63%, respectively, to 12% at 5 years post-treatment with aHSCT. Mean concentrations had changed from 920 pg/mL to 270 pg/mL for NfL (P <.001) and from 1500 pg/mL to 680 pg/mL (P <.001) for MBP at the 5 year mark. However, mean concentrations of glial acidic fibrillary protein (GFAp), the third biomarker examined in the study, did not change significantly.
“NfL is now widely used as a biomarker for axonal injury in neurological disease, reflecting its utility,” corresponding author Joachim Burman MD, PhD, Department of Medical Sciences, Uppsala University, Sweden, and colleagues wrote. “All of the patients with gadolinium enhancing lesions had NfL values above the upper limit of normal, and more than half of those without. Patients with active disease usually have increased levels of NfL, and the high proportion of patients with increased concentrations of NfL likely reflects the fact that patients selected for aHSCT have active disease. Following aHSCT, CSF NfL concentrations were decreased successively at each follow-up, consistent with previous reports from the Canadian Bone Marrow Transplantation group as well as our own group.”
The investigators noted that MBP, in similarity to NfL, was initially at higher levels in all patients with gadolinium enhancing lesions and approximately half of patients without gadolinium enhancing lesions, and that its concentrations declined with each successive assessment. However, they pointed out that MBP is not used as an MS disease biomarker in prior studies as frequently as NfL, possibly due to its shorter half-life. In contrast to NfL and MBP, GFAp concentrations above the normal limit were only seen in 16% of the patients at baseline, and the concentrations did not meaningfully change after the patients were treated. The investigators attributed this to speculation that GFAp elevation indicates astrocytes reacting to damage via astrogliosis, rather than it being a direct result of axonal damage like NfL or demyelination like MBP. They highlighted the fact that while GFAp elevations are supported by previous literature as a biomarker for MS, no known previous study has ever demonstrated GFAp concentrations being meaningfully changed by MS therapies.
The study initially screened 63 patients who were administered aHSCT at Uppsala University Hospital during a period spanning from January 2012 to January 2019. The patients included in the final data (n=43) had a mean age of 31, a median baseline expanded disability status scale (EDSS) score of 3.5, and an annualized relapse rate of 1.6 in the year before aHSCT administration. The participants included 28 female patients and 15 male patients, 79% of whom who had received between 1 and 3 prior treatments with disease modifying drugs DMD (median, 2), and the rest of whom were treatment naïve (n=9). Follow-up data was available for all 43 patients at 1 year, 26 of the patients at 2 years, and 8 of the patients at 5 years (median, 3.9 years) after aHSCT treatment. A control group of 31 healthy volunteers was recruited and evaluated for concentrations of the 3 biomarkers at a single time point to establish a reference for normal levels.
In terms of outcomes after aHCST treatment, 15 patients had a stable EDSS during follow-up, 23 patients had an improved EDSS with confirmed disability improvement, and 5 patients had a worse EDSS with confirmed disability worsening. Evidence of disease activity was observed in 9 patients during follow-up, 4 having had a clinical relapse and 5 having new MRI lesions. It was noted that the patients with stable EDSS during follow-up were more likely to have lower concentrations of the biomarkers.
“All patients who were treated with aHSCT were offered to undergo lumbar puncture and participate in the study. About 1/3 declined,” Burman and colleagues noted, pointing out a limitation of the study. “We did not collect any data on the reason for declining, but it is possible that it may have led to a sampling bias towards patients with more active disease during follow-up. Only a minority of patients was followed for the entire 5 years. This means that the long-term effects of aHSCT must be interpreted with caution and further studies are needed to ascertain the results on the time scale 5–10 years after aHSCT. We relied on samples from healthy controls to assess normal levels of MBP and GFAp, but these were slightly mismatched in terms of age and sex. There is little indication of that sex has a large impact on the CSF concentrations of these proteins, but GFAp clearly increases with age. This means that our upper limit of normal may have been set too low and that a slightly larger percentage of patients should have been considered as having normal values.”
The investigators pointed out that at baseline all biomarkers were associated with one another and with EDSS, but the association ceased during follow-up, indicating that the disability-causing damage had been stopped. They ultimately concluded that aHCST is capable of arresting the process of axonal damage and demyelination that leads to injury in a majority of patients over time, though disease activity may persist.
