Real-world evidence of ocrelizumab-treated relapsing multiple sclerosis cohort shows changes in progression independent of relapse activity mirroring phase 3 trials

Abstract

Ocrelizumab is a B cell-depleting drug widely used in relapsing-remitting multiple sclerosis (RRMS) and primary-progressive MS. In RRMS, it is becoming increasingly apparent that accumulation of disability not only manifests as relapse-associated worsening (RAW) but also as progression independent of relapse activity (PIRA) throughout the disease course. This study's objective was to investigate the role of PIRA in RRMS patients treated with ocrelizumab. We performed a single-center, retrospective, cross-sectional study of clinical data acquired at a German tertiary multiple sclerosis referral center from 2018 to 2022. All patients with RRMS treated with ocrelizumab for at least six months and complete datasets were analyzed. Confirmed disability accumulation (CDA) was defined as a ≥ 12-week confirmed increase from the previous expanded disability status scale (EDSS) score of ≥ 1.0 if the previous EDSS was ≤ 5.5 or a ≥ 0.5-point increase if the previous EDSS was > 5.5. PIRA was defined as CDA without relapse since the last EDSS measurement and at least for the preceding 12 weeks. RAW was defined as CDA in an interval of EDSS measurements with ≥ 1 relapses. Cox proportional hazard models were used to analyze the probability of developing PIRA depending on various factors, including disease duration, previous disease-modifying treatments (DMTs), and optical coherence tomography-assessed retinal degeneration parameters. 97 patients were included in the analysis. Mean follow-up time was 29 months (range 6 to 51 months). 23.5% developed CDA under ocrelizumab therapy (n = 23). Of those, the majority developed PIRA (87.0% of CDA, n = 20) rather than RAW (13.0% of CDA, n = 3). An exploratory investigation using Cox proportional hazards ratios revealed two possible factors associated with an increased probability of developing PIRA: a shorter disease duration prior to ocrelizumab (p = 0.02) and a lower number of previous DMTs prior to ocrelizumab (p = 0.04). Our data show that in ocrelizumab-treated RRMS patients, the main driver of disability accumulation is PIRA rather than RAW. Furthermore, these real-world data show remarkable consistency with data from phase 3 randomized controlled trials of ocrelizumab in RRMS, which may increase confidence in translating results from tightly controlled RCTs into the real-world clinical setting.

Figure 1

Overview of the cohort and patient selection according to inclusion and exclusion criteria. MS multiple sclerosis; OCR ocrelizumab; PPMS primary progressive multiple sclerosis; EDSS expanded disability status scale.

Figure 2

Disability change. In order to visualize EDSS changes, the total cohort (n = 97) was stratified into four groups: "stable" are patients without EDSS worsening (n = 71), "RAW" were patients with relapse-associated worsening in EDSS (n = 3), "PIRA -SIR" were patients with progression independent of relapses without superimposed relapses (n = 16), "PIRA + SIR" were PIRA patients with superimposed relapses (n = 4). There were no patients with a combination of PIRA and RAW in the cohort (i.e., no superimposed relapses altered the EDSS in any patient). n = 3 patients experienced relapses with complete remission, i.e., without detectable long-term EDSS worsening, and are depicted as blue dots within the stable group. EDSS worsening was defined as an increase of 1.0 EDSS points if the previous EDSS was ≤ 5.5 or 0.5 EDSS points if the previous EDSS was > 5.5 and if the new EDSS was confirmed at least 12 weeks thereafter. Significance was calculated using Kruskal–Wallis test and Dunn's multiple comparison test. *p < 0.05; **p < 0.01; ***p < 0.001. EDSS expanded disability status scale; RAW relapse-associated worsening; PIRA  progression independent of relapse activity; SIR superimposed relapses.

Figure 3

Kaplan-Meyer-estimates of CDA events in real-world ocrelizumab cohort. The figure depicts the probability of patients having experienced RAW or PIRA events in the observed cohort over time. CDA shows the probability of at least one of the two having occurred. CDA confirmed disability accumulation; RAW relapse-associated worsening; PIRA progression independent of relapse activity.

Figure 4

Factors associated with PIRA development probability. Kaplan-Meyer estimates of clinical, demographical, and OCT factors associated with PIRA event probabilities. The panels (A)–(I) depict the cumulative probabilities of developing PIRA according to different categories: (A) Disease duration prior to ocrelizumab stratified by quartiles (2.6 years, 9 years, 15.3 years). A shorter disease duration showed a significantly increased probability to developing PIRA (p = 0.02). (B) Number of DMTs used previously to ocrelizumab. Less previous DMTs was associated to a higher PIRA probability (p = 0.04). Other factors did not show significant alterations in PIRA probability: (C) Ocrelizumab as a first-line therapy vs. second-line, (D) ARR in the 2 years prior to ocrelizumab initiation, (E) EDSS at ocrelizumab initiation, (F) male vs. female and detectability of CD19⁺ cells in blood counts 6–12 months after OCR initiation (G). OCT based retinal measures were also not significantly associated with PIRA development probability with (H) showing peripapillary RNFL thickness, (I) showing macular RNFL volume and (J) showing macular GCIPL volume. OCT factors were stratified by quartiles with the worse of the two eyes taken into analysis. PIRA progression independent of relapse activity; OCR ocrelizumab; DD disease duration; DMTs disease-modifying therapies; ARR annualized relapse rate; EDSS expanded disability status scale; pRNFL peripapillary retinal nerve fiber layer; mRNFL macular retinal nerve fiber layer; mGCIPL macular ganglion cell and inner plexiform layer.