Introduction

The overlapping of focal recurring inflammation, more widespread inflammatory processes, and neurodegeneration underlines the course of multiple sclerosis (MS) during all stages of the disease, challenging the classic described phenotypes.^{1, 2} There is a mutual contribution of progression independent of relapse activity (PIRA) and relapse-related worsening (RAW) events to progressive disability accrual since the disease onset, requiring an optimal timing of diagnosis and the appropriate disease-modifying therapy (DMT) to reduce as much as possible the disability burden.^3-6 Relapses remain a crucial determinant of disability accumulation, playing a role both in the short and long term,^{7, 8} and age is configured as the main risk factor associated with the occurrence of PIRA events.⁹

Exposure to new highly effective (HE) DMTs has been proven to have greater efficacy in reducing the accumulation of disability resulting from PIRA and RAW,^{2, 9} in comparison with moderate effective treatments. However, reliable comparative effectiveness studies between HE DMTs are scarce in the MS literature landscape¹⁰ and their effect on PIRA needs to be better clarified. Ocrelizumab (OCR) and natalizumab (NTZ) demonstrated a strong anti-inflammatory activity as HE DMTs, changing the prognosis of MS patients.¹¹ Studies based on large cohorts from disease registries demonstrated to be essential in exploring potential disease outcomes and DMT effectiveness over the medium and long term in real-life settings.¹²

In a large real life-cohort of naïve relapsing–remitting MS patients (RRMS) from the Italian MS and Related Disorders Register (I-MS&RD),¹³ we compared the risk of achieving the first 6 months confirmed PIRA and RAW events and irreversible Expanded Disability Status Scale (EDSS) 4.0 and 6.0 between patients treated with NTZ and those treated with OCR.

Materials and Methods

Results

Clinical data of 81,123 patients were available in the I-MS&RD at the time of data extraction. After applying the inclusion criteria, we retrieved a cohort of 770 RRMS patients, of which 568 were treated with NTZ and 212 with OCR. The PS matching procedure subsequently resulted in 195 pairs of patients. (Fig. 1) Baseline characteristics of the cohort before and after PS matching are reported in Table 1. Patients treated with NTZ, before the PS matching, were younger at DMT start (mean (SD): 30.19 (10.15)^10-15 years vs. 36.22 (11.23)), less disabled (mean (SD) EDSS score: 1.92 (1.31) vs. 2.51 (1.51)), more frequently females (65.49% vs. 58.02%), compared to patients treated with OCR in term of SMD values. After the matching procedure, considering the SMD values, the groups resulted balanced and comparable, with an age at DMT start of about 35 years, female gender in 60% of patients, monofocal onset in about 72% patients, OCBs presence in about 91% patients, EDSS score of about 2.5, number of EDSS evaluations of about 4.5, and at least one relapse in about 32% patients.

Median follow-up (IQR) after the treatment start was 1.63 (0.87–2.72) for NTZ group and 1.60 (0.80–2.68) years for OCR group.

A first PIRA event was reached by 23 (11.79%) NTZ-exposed patients and 25 (12.82%) OCR exposed patients. No RAW events were found in the NTZ group and only 1 RAW event was reported in the OCR group. An irreversible EDSS 4.0 was reached by 7 (4.14%) NTZ-treated patients and 10 (5.85%) OCR-treated patients, and an irreversible EDSS 6.0 was reached by 13 (6.67%) NTZ-treated patients and 15 (7.69%) OCR-treated patients. (Fig. 2).

KM curves for the probabilities of reaching the outcomes are shown in the Figure 3. KM for RAW events were not outlined because there were no events to display.

No differences between the two groups were found in the risk (HR, 95%CI) of reaching a first PIRA (Fig. 3A) (1.04, 0.59–1.84; p = 0.88) event, and irreversible EDSS 4.0 (Fig. 3B) (1.23, 0.57–2.66; p = 0.60) and EDSS 6.0 (0.93, 0.32–2.68; p = 0.89) (Fig. 3C).

Finally, 27 (13.85%) and 8 (4.10%) (p = 0.001) patients switched to another DMT, in NTZ- and OCR-treated groups, respectively. In the NTZ-treated switcher group, 9 patients (33.33%) switched to OCR.

In the OCZ-treated group, 3 patients switched to glatiramer acetate for pregnancy planning, 1 to NTZ for disease progression, and 4 to dimethyl fumarate due to a de-escalation strategy. In the NTZ group, we identified 20 patients switched for adverse events, with 2 cases of elevated levels of serum transaminases and 18 cases of suspension due to JCV antibody positivity, and within this group 8 patients switched to cladribine, 7 to OCR, 1 to siponimod, 1 to dimethyl fumarate, 1 to teriflunomide, 1 to interferon beta-1a, and 1 to glatiramer acetate; 4 patients suspended therapy due to allergic reactions, switching to cladribine, OCR, alemtuzumab and dimethyl fumarate; 1 patient switched to siponimod due to secondary progressive MS conversion, 1 to dimethyl fumarate for pregnancy planning and 1 to OCR for patient choice.

Discussion

To our knowledge, this is the first study aimed at comparing the effect of NTZ and OCR on the risk of achieving the first 6-month confirmed PIRA and RAW events and severe EDSS disability milestones in a large real-life cohort of naïve RRMS patients.

All the newly approved HE DMTs have demonstrated an excellent profile of effectiveness in reducing clinical and radiological disease activity, over the years.¹¹ Some recent meta-analyses have developed an effectiveness ranking for all available DMTs, placing NTZ and OCR at comparable levels.^{14, 15} NTZ proved to be highly effective in RRMS treatment both in clinical trials and real-world settings with a well-established safety profile, leading to a significant improvement in the control of disease activity.^{16, 17} Growing evidence suggests that OCR has a strong anti-inflammatory action, with an almost complete suppression of clinical and radiological disease activity, and a significant effect in slowing the progression of disability accrual in RR and progressive MS.^{18, 19}

A recent study²⁰ compared, in a real-world PS-matched cohort of RRMS patients, the effectiveness of NTZ and OCR using the NEDA-3 as outcome. After 30 months of follow-up, NEDA-3 was reached by 53.1% in the OCR group and by 36.1% in the NTZ group, the treatment with OCR was associated with a lower risk of relapse, while the reduction of serum NFL levels did not differ between the two treatment groups. Another study showed that, in RRMS who switched from NTZ to dimethyl fumarate or fingolimod or OCR, the switch to OCR was associated with the lowest ARR and discontinuation rates and the longest time to first relapse.²¹

To date, the assessment of DMT effect on PIRA events has not been sufficiently investigated.^{22, 23}

PIRA events, especially those that occurred early in the disease course, have been demonstrated to be predictors of unfavorable long-term disability prognosis.²⁴ Therefore, a timely and prolonged treatment with the appropriate DMT is also primarily aimed at reducing the occurrence of PIRA events.^{2, 9} The superiority of OCR over interferon beta 1a in reducing PIRA, has already been reported in OPERA 1 and 2 trials.⁴ OCR proved an all-rounder beneficial effect on cognitive function, fatigue, and quality of life of MS patients, considering composite clinical measures and magnetic resonance imaging (MRI) and fluid biomarkers to assess disease progression.¹⁹

Currently, the pathophysiological substrate of PIRA remains incompletely characterized.^{25, 26} The effect on PIRA of NTZ and OCR, with completely different mechanisms of action but a common strong anti-inflammatory effect, could be due to the suppression of inflammatory pathological processes in the “smoldering” MS lesions.^{2, 6} Despite predicted effects on inflammatory networks related to microglia in chronic active lesions, anti-CD20 therapies failed to fully resolve paramagnetic rim lesions after a 2-year MRI follow-up in a recent study.²⁷ These findings highlight the necessity of better identifying PIRA's pathological substrates to further evaluate appropriate treatment strategies.

Our results demonstrate that both OCR and NTZ strongly suppress RAW events and have a similar impact on PIRA in naïve RRMS patients, confirming that an early access to HE DMTs^28-30 may result in a beneficial effect on the accumulation of neurological damage, dependent and independent from relapses, early in the disease course.^{31, 32}

In OPERA I/II, in patients treated with OCR, PIRA events were the main contributors to both 12-week and 24-week CDA after 96 weeks (147 out of 167 [88%] and 115 out of 129 [89%]) and only a minority presented RAW events (29.7%).⁴ However, this pooled analysis of two randomized clinical trials (RCTs) used a composite CDA definition, including EDSS alongside with timed 25-feet walk and 9-hole peg test evaluations. Furthermore, considering the limitations of a comparison between our cohort and Kappos's, baseline data resulted similar in terms of age, sex, EDSS and number of relapses; however, in the pooled analysis only 73.3% of patients were treatment naïve.

Currently, to our knowledge there is no data about PIRA in RCTs populations of NTZ-treated patients.

A few limitations of our study should be taken into account. As with previous retrospective observational studies, incomplete or inaccurate data entered in the register must be acknowledged.

The lack of long-term follow-up has to be considered and the validation of these results over the long-term will constitutes a crucial point of future analysis. Although patients with less than 1 year of follow-up might not have greatly contributed to the 6-month confirmed PIRA evaluation, it is important to underline that we selected a cohort of closely monitored patients with at least three EDSS evaluations.

Moreover, as part of the data interpretation process, we also ran an analysis of the power calculation for equivalency study to assess limitations of this comparison of the two DMTs. Although the sample size calculation is not required in observational studies, we hypothesized several scenarios and calculated the necessary patient count to obtain the higher statistical power of 0.80. The number of patients to include per group would be 443 for a difference and 608 for an equivalence analysis, 439 for a superiority and 281 for a non-inferiority analysis. In our study, considering a cumulative incidence of 0.31 and 0.40 respectively for NTZ and OCR cohort, a power of 0.46 have been ensured with a sample size of 195 patients per group. With this cohort, the power is 0.38 for an equivalence study, 0.66 for a superiority, and 0.50 for a non-inferiority analysis. Therefore, a possible explanation of the absence of differences in PIRA outcomes between OCR and NTZ could be related to the sample size, too small to detect a difference.

In the definition of CDA events, we considered solely the EDSS score: although MRI features are considered a crucial prognostic factor and radiological activity is a cornerstone in defining the disease burden, we could not include MRI data because of the lack of a systematic acquisition of these data in the I-MS&RD. Furthermore, in comparison with RCTs, in this real-world study we are unable to account for potential confounders that could have affected results.

Despite all these considerations, our study used the large and validated database of the I-MS&RD to compare the impact of two of HE DMTs on PIRA and RAW events, so far not available in MS literature.

In conclusion, our results confirm that both OCR and NTZ effectively eliminate RAW events in RRMS patients. In the short term, the number and risk of PIRA, EDSS 4.0 irreversible and EDSS 6.0 events were not significantly different between the two treated groups. A longer follow-up and a larger cohort will be essential to confirm data on the effects of NTZ and OCR especially on disability progression outcomes.

Funding Information

The present work received no specific funding.

Conflict of Interest

The authors report no conflicts of interest with respect to the contents of the current study, but note that the patients in the study were treated with a number of disease-modifying drugs and that authors have received advisory board, membership, speakers honoraria, travel support, research grants, consulting fees, or clinical trial support from the manufacturers of those drugs, including Actelion, Allergan, Almirall, Alexion, Bayer Schering, Biogen, Celgene, Excemed, Genzyme, Forward Pharma, Ipsen, Medday, Merck, Mylan, Novartis, Sanofi, Roche, Teva, and their local affiliates.

Author Contributions

PI, TG, GL, and MT: conception and design of the study, acquisition and analysis of data, drafting the manuscript and figures; DP, EP, MI, MF, FP, FG, SR, PC, GDL, PG, PB, AG, SM, ADS, MV, RQ, DS, RC; VTR, EC, VBM, AGM, DVB, MF, and MPA: data acquisition and drafting a significant portion of the manuscript.