Similar overall safety profiles were observed between the masitinib and placebo groups, although there was a higher incidence of severe and serious AEs associated with masitinib treatment. The most frequent masitinib-associated AEs were consistent with the known safety profile of tyrosine kinase inhibitors, notably rash, nausea, edema, and diarrhea, which are generally considered manageable with symptomatic treatments when of non severe intensity. The majority of AEs leading to permanent discontinuation in the present study were of non severe intensity, suggesting therefore a fairly cautious investigator approach to AEs or difficulties experienced in their management. As rash was the leading cause of discontinuation in this and other non-oncology masitinib trials (data not shown), future studies might consider consulting a dermatologist on matters of rash management and possible treatment interruption or dose adjustment prior to any decision on discontinuation.
Although efficacy data did not produce statistically significance differences between treatment groups, it does suggest a positive effect of masitinib on MS-related impairment and potential retardation of disease progression for both PPMS and rfSPMS patients. For example, in patients treated with masitinib we observed an improvement in MSFC scores relative to baseline, compared with a worsening MSFC score in patients receiving placebo. These changes were mainly driven by the T25FW and 9-HPT subscores, with the clinical implications being that masitinib might slow down the degeneration of lower limb function (as evidenced by a milder deterioration of T25FW) and improve upper limb function (as evidenced by improvement in 9-HPT). However, no adjustments were made for learning effects associated with some of the MSFC component measures, which may therefore have influenced these findings . Also, for progressive diseases such as PPMS, the use of LOCF analysis is inclined to underestimate functional deterioration. Conversely however, considering the number of positive MSFC clinical responses achieved by masitinib patients (32%) compared with placebo patients (0%), it is unlikely that such effects had a major impact on the overall results.
Initially, 35 patients were planned for a treatment period of 36 months; however, this was amended to at least 20 patients who had completed at least 12 months of treatment. This protocol amendment, which effectively unblinded the study early, was implemented in part because even under blinded conditions it was probable that some masitinib-treated MS patients were among those showing positive response. In view of the pressing medical need for an effective treatment in progressive forms of MS, if this were the case then the primary objective to demonstrate acceptable safety and possible therapeutic response, i.e. establish proof-of-concept, would have been sufficiently accomplished, thereby enabling progression to the next development stage (i.e. phase 2b/3). One negative consequence of this reduced study population however, given the final dataset, was that it precluded any demonstration of statistical significance between the masitinib and placebo treatment. A second factor in the decision to amend the study population size was due to design factors and minor protocol deviations that would have complicated any definitive interpretation of efficacy, even if statistical significance had been demonstrated. This included a study amendment to close the 3.0 mg/kg/day treatment arm because of lack of response, effectively pooling all patients into the 6.0 mg/kg/day treatment arm. Also, it became apparent that there was a minor protocol deviation in the timed 25-foot walk (T25FW) test measuring leg function and ambulation, which forms part of the MSFC composite score. This test was misunderstood by two test centers representing 7 and 10 patients of the mITT population, respectively. One conducted the test on 25 steps and the other on 25 meters instead of 25 feet. The resultant disparity between centers was statistically compensated for by individually calculating each subpopulation’s T25FW z-score (i.e. with respect to units of steps, meters or feet) with reference to its overall patient average, and then taking the average of these z-scores for the overall T25FW z-score. This protocol deviation is expected to have had little or no effect on the interpretation of the MSFC score because the z-score (or standard score, a dimensionless quantity indicating how many standard deviations an observation is above or below the mean) allows direct comparison of observations from different units of measure.
The possible mechanisms of action by which masitinib may be capable of inducing the observed positive therapeutic response in patients with progressive MS are multifaceted. Although a topic of debate, there is growing evidence that the different courses of MS, i.e. relapsing as opposed to relapse-free, are due to distinct pathophysiologic processes. That is, RRMS and SPMS are probably different stages of the same disease while PPMS may imply different processes. Relapses are considered the clinical expression of acute inflammatory focal lesions whereas progression is considered to reflect the occurrence of demyelination, axonal loss and gliosis . This distinction in MS types appears to be reflected by the unsuccessful treatment of PPMS with powerful disease modifying drugs. In turn, this may relate to the dominant cause of progression of disability in PPMS being more strongly related to nerve cell death, in addition to inflammation-induced neuronal damage (swelling) commonly attributed to relapsing forms of MS. As mentioned previously, there is good evidence in support of mast cells being actively involved in the pathogenesis of MS [12, 13]. For example, sites of inflammatory demyelination contain cellular infiltrates with mast cell accumulation in the brain and spinal cord,  and the percentage of degranulated mast cells in the central nervous system correlates with the clinical onset of disease symptoms in acute EAE . The contribution of mast cells to the pathological cascade of MS is in part because they release large amounts of proinflammatory mediators and therefore play a prominent role in sustaining the inflammatory network of the central nervous system . The involvement of inflammation in the development of brain injury in MS is well-established, neurodegeneration being provoked in part by soluble inflammatory mediators, with a significant correlation existing between inflammation and acute axonal injury . Moreover, perivascular mast cells secrete pro-inflammatory and vasoactive molecules that can regulate the BBB’s permeability, a defective BBB being a common finding that precedes clinical or pathological signs of MS [14, 24, 25]. Additionally, it has been shown in vitro that mast cell activation can lead to neuronal damage by inducing astroglia to produce neurotoxic quantities of nitric oxide (NO) ; NO being a molecule implicated in the pathogenesis of MS, especially for those patients in progression [27, 28]. It has also been reported that mast cells can be a source of NO derivatives, which they synthesize spontaneously or following activation, depending on their subtype . This evidence supports the notion that mast cells, which can be found in close vicinity to neurons, could influence the survival and functions of NO-sensitive cells and through this mechanism participate in the pathophysiology of chronic neurodegenerative diseases of the nervous system. Additionally, it is plausible that masitinib’s inhibitory action also effects the activation of dendritic cells, which are integral to the differentiation of T helper cells and regulate T cell responses, through inhibition of c-Kit and Lyn [30, 31]. This hypothesis may be of significance as recent genetic findings particularly implicate T helper cell differentiation in the pathogenesis of MS .