Долгосрочная оценка эффективности и иммунологических последствий Ocrelizumab в реальном мире.

Key Points

Ocrelizumab stabilizes disability progression and prevents clinical and radiological disease activity over the long term. Ocrelizumab affects both humoral and cellular immune responses from the first infusion but does not determine a significant infection risk in the long term. No association was observed with decreased immunoglobulin levels and severe infections.

Introduction

Growing evidence in immunological and neuropathology studies over the past few years strongly supports the crucial involvement of B cells in the pathogenesis of multiple sclerosis (MS). B cells play a fundamental role in the disease not only by producing antibodies but also through the abnormal inflammatory cytokine and chemokine production and their antigen-presenting function that activates T cells, exacerbating the chronic non-resolving neuroinflammation. Furthermore, a B-cell infiltration in tertiary lymphoid-like structures in the meninges has been proven in pathological studies [ 1 – 3 ].

Highly effective disease-modifying therapies (DMTs), including B cell-depleting therapies, either specifically targeted to B cells or not, have significantly contributed to modifying the natural history of MS [ 4 ].

Ocrelizumab (OCR) is an anti-CD20 monoclonal antibody approved for the treatment of relapsing-remitting MS (RRMS) and primary progressive MS (PPMS). The OPERA I and II and ORATORIO trials demonstrated a significant benefit in relapse recurrence, disability progression, and magnetic resonance image (MRI) activity [ 5 , 6 ]. Data on the efficacy and safety of OCR were also confirmed in the open-label extension phases of the pivotal trials [ 7 , 8 ].

Strong real-world evidence, also based on disease registries, demonstrated that OCR reduces clinical and radiological activity over time, with a favorable safety profile maintained over years of treatment [ 9 , 10 ]. Ocrelizumab has a wide range of effects on both humoral and cellular immune responses from the first infusion: a precise monitoring of immunological consequences and the related risk of infections throughout time needs to be addressed [ 11 , 12 ].

In this scenario, real-world observational studies represent a great source to explore the long-term clinical and immunological impact of a therapy in real clinical settings [ 13 ]. The study aimed to evaluate the effectiveness and safety of OCR and assess its long-term effect on immunoglobulin (Ig) levels, lymphocyte subsets, and infection-related events for patients with RRMS, including patients with active secondary progressive MS (SPMS) and PPMS, who were followed at the Multiple Sclerosis Center of the University Hospital Policlinico of Bari.

Methods

Data Collection and Study Population

Patients with ≥ 2 years of treatment with OCR were retrospectively recruited at the Multiple Sclerosis Center of Bari (Department of Translational Biomedicine and Neuroscience DiBraiN, University Hospital Policlinico of Bari). All the data on MS history, demographics, treatments, and regular neurological evaluations of patients were collected, according to the Italian MS Register (RISM). Data extraction was performed by applying manually sequential filters to the data set extracted in January 2024. A minimum data set was retrieved including the following variables: date of birth, sex, date of disease onset and confirmed diagnosis, onset symptoms, dates of relapses, dates of Expanded Disability Status Scale (EDSS) evaluations with complete information regarding functional scores, start and end dates of all the administered DMTs, presence and number of cerebrospinal fluid oligoclonal IgG bands, MRI data (presence/absence of new or enlarged T2 lesions and gadolinium enhancement), and data about OCR infusions. Changes in serum IgG/IgM/IgA levels and serum values of T CD4+, T CD8+, and natural killer (NK) lymphocytes from baseline for up to 6 years of OCR treatment were also retrieved from clinical records and analyzed. Data about infection occurrence, type, and outcomes were also collected from RISM-App and clinical records.

Radiological activity was defined as the occurrence of gadolinium-enhancing lesions or new/enlarged T2-hyperintense lesions. An MRI assessment was performed approximately once a year: both brain MRI and MRI of the spinal cord were included in the follow-up assessment, and gadolinium administration was performed at all scans collected.

RISM was approved by the Ethics Committee of the “Azienda Ospedaliera/Universitaria-Policlinico of Bari” (Study REGISTRO SM001, approved on 08/07/2016). Patients signed an informed consent that allows the use of clinical data for research purposes [ 14 ].

Statistical Analysis

Summaries of continuous variables have been calculated as median with interquartile ranges or mean and standard deviation (SD), categorical variables have been presented as frequencies (proportions). The Wilcoxon paired test was used to evaluate the EDSS changes over time. Twelve-month confirmed disability worsening (CDW), and confirmed disability improvement were assessed. Confirmed disability worsening was calculated as an increase in EDSS score of 1.5 points for patients with a baseline EDSS score of 0, or 1 point for a baseline EDSS score between 1 and 5.5, or 0.5 points for a baseline EDSS above 5.5, confirmed after 12 months. Confirmed disability improvement was calculated as a decrease in the EDSS score of 1 point for patients with a baseline EDSS score between 1 and 5.5, or 0.5 points for patients with a baseline EDSS score above 5.5, confirmed after 12 months. The annualized relapse rate (ARR) before and after the start of OCR treatment was estimated considering a crude ARR (person time method; 95% confidence interval [CI] by fitting Poisson distributions) [ 15 ].

Factors associated with CDW were assessed using a multivariable Cox proportional hazards model. The exposure time was censored when the outcome was reached or at the last neurological assessment. The results of the proportional hazards regression model were expressed as hazard ratios (HRs) and 95% CIs. The model was adjusted for sex (female sex as reference), age, disease duration, symptoms at onset (monofocal/multifocal), number of relapses pre-initiation of OCR, and the presence of oligoclonal bands in the cerebrospinal fluid.

The evolution over time of the serum levels of the laboratory parameters examined in the study was estimated by calculating the mean value of Igs for each follow-up timepoint, in 6 months corresponding to the infusion period; therefore, we plotted graphically these values. A linear mixed-effects regression model with a random intercept for individuals was used to compare measurements of Igs at different timepoints after initiation of OCR. The IgG, IgM, and IgA levels were included as the dependent variable, while time categorized at different timepoints and sex and age were included as independent variables.

Using the threshold values from earlier research [ 9 , 16 ], hypogammaglobulinemia was defined as IgG < 600 mg/dL, IgM < 40 mg/dL, and IgA < 70 mg/dL. Considering the safety assessment, the associations of clinical, laboratory, and demographic risk factors with infection occurrence were estimated using a Cox proportional hazards model. The exposure time was censored when the outcome was reached or at the last neurological assessment; the model was adjusted including sex, age, disease duration, Ig levels, and EDSS score.

All statistical analyses were performed in Windows SPSS Statistics Version 25 (IBM Corporation, Armonk, NY, USA). A threshold of 0.05 was used for statistical significance.

Results

The final cohort retrieved 140 patients, of whom 80 were diagnosed with RRMS, 37 with PPMS, and 23 with SPMS. The median (interquartile range) follow-up after treatment started was 4.62 (4.10–5.04) years.

The baseline demographics and clinical characteristics of the cohort are shown in Table 1 . With 70 subjects recorded for each sex, the entire cohort had the same percentage of male and female subjects, while considering the subgroup, we reported 44 (55%), 18 (48.6%), and 8 (34.8%) female subjects for the RRMS, PPMS, and SPMS cohorts, respectively. The mean (SD) age at OCR initiation was 45.38 (9.95) years for the overall cohort, 43.26 (9.8) for patients with RRMS, 49.97 (9.37) years for patients with PPMS, and 45.37 (9.15) years for patients with SPMS. The mean (SD) disease duration was 17.13 (9.22) years and was longer for patients with SPMS, 22.59 (7.84) years, compared with patients with RRMS and patients with PPMS [17.41 (9.36) and 13.13 (7.92), respectively] ( p < 0.001).

Table 1 Baseline clinical and demographic characteristics Variable Overall population ( n  = 140) RR ( n  = 80) PP ( n  = 37) SP ( n  = 23) Female sex, n (%) 70 (50%) 44 (55%) 18 (48.64%) 8 (34.78%) Age at onset (mean ± SD), years 31.57 ± 11.70 29.05 ± 10.78 40.4 ± 10.6 26.09 ± 8.81 Time from onset to diagnosis (mean ± SD), years 3.95 ± 4.61 3.72 ± 5.10 4.52 ± 3.98 3.85 ± 3.72 Time from diagnosis to ocrelizumab initiation (mean ± SD), years 9.85 ± 8.08 10.49 ± 8.32 5.02 ± 5.57 15.43 ± 6.33 Follow-up (mean± SD), years 4.62 ± 0.72 4.52 ± 0.63 4.88 ± 0.93 4.60 ± 0.60 Age at ocrelizumab initiation (mean ± SD), years 45.38 ± 9.95 43.26 ± 9.82 49.97 ± 9.37 45.37 ± 9.15 EDSS at ocrelizumab initiation, median (IQR) 5.25 (1.5–8.0) 4.5 (1.5–8.0) 5.5 (4.0–8.0) 6.5 (4.0–7.5) Presence of CSF oligoclonal IgG bands, n (%) 111 (79.3) 64 (80) 28 (75.7) 19 (82.6) Number of CSF oligoclonal IgG bands, median (IQR) 10.5 (1–28) 11 (1–28) 8.5 (1–25) 16 (1–27) Disease duration (mean ± SD), years 17.13 ± 9.22 17.41 ± 9.36 13.13 ± 7.92 22.59 ± 7.84 Previous DMTs exposure, n (%)  Naïve 26 (18.57) 12 (15) 14 (37.83) /  Fingolimod 34 (24.28) 22 (27.50) 4 (10.81) 8 (34.78)  Natalizumab 6 (4.28) 5 (6.25) 1 (2.70) /  Teriflunomide 5 (3.57) 1 (1.25) 1 (2.70) 3 (13)  Dimethyl-fumarate 23 (16.42) 17 (21.25) 3 (8.10) 3 (13)  Interferon-β 21 (15) 12 (15) 4 (10.81) 5 (21.73)  Glatiramer acetate 12 (8.57) 7 (8.75) 4 (10.81) 1 (4.34)  Azathioprine 8 (5.71) 1 (1.25) 6 (16.21) 1 (4.34)  Alemtuzumab 5 (3.57) 3 (3.75) / 2 (8.69) CSF, cerebrospinal fluid; DMTs, Disease Modifying Therapies; EDSS, Expanded Disability Status Scale; IQR, interquartile range; SD, standard deviation

Baseline clinical and demographic characteristics

CSF, cerebrospinal fluid; DMTs, Disease Modifying Therapies; EDSS, Expanded Disability Status Scale; IQR, interquartile range; SD, standard deviation

Patients with SPMS presented with significantly higher ( p < 0.001) EDSS scores at OCR initiation, with a median (interquartile range) EDSS score of 6.5 (4.0–7.5) compared with patients with RRMS, 4.5 (1.5–8.0) and patients with PPMS, 5.5 (4.0–8.0). In the entire cohort, the mean ARR decreased from 0.61 (95% CI 0.47–0.74) in the year before the initiation of OCR to 0.04 (95% CI 0.01–0.09) in the first year and 0.02 (95% CI 0.004–0.062) in the second year. Subsequently all patients in our cohort remained free of relapses (Fig. 1 ). Considering the subgroups, comparable reductions in the mean ARR were seen before and after the start of OCR treatment in the RRMS group, 0.64 versus 0.03, and in the SPMS group, 0.58 versus 0.07. All patients also remained free of MRI activity up to the end of the follow-up.

Annualized relapse rate (ARR) in the year before ocrelizumab initiation and in the follow-up years

The overall percentage of stable patients increased from the second year ( n = 90, 64.28%) to the fourth year ( n = 121, 86.42%), in parallel with a decrease in patients with CDW ( n = 37, 26.42% vs n = 11, 7.85%) and confirmed a disability improvement ( n = 13, 9.28% vs n = 8, 5.71%). The EDSS score after the start of OCR treatment significantly increased compared with the baseline values only in the SPMS group ( p < 0.01), but it remained stable in the PPMS and RRMS groups. The multivariable Cox regression models revealed that the presence of two or more relapses before the start of OCR treatment (HR 1.44, 95% CI 1.13–1.88) and a multifocal onset (HR, 1.53, 95% CI 1.51–1.55) were significant ( p < 0.05) risk factors for CDW (Table 2 ).

Table 2 Multivariable Cox model performed on the entire cohort to estimate the predictors of CDW Variables HR (95% CI) p value Sex, female as reference 1.17 (0.82–1.70) 0.571 Age at disease onset 0.93 (0.61–1.41) 0.321 Disease duration 1.11 (0.76–1.64) 0.655 Type of clinical onset, monofocal reference group  Multifocal 1.53 (1.51–1.55) < 0.001  Number of relapses prior to first DMT start  > 1 vs 0 1.01 (0.98–1.05) 0.578  ≥ 2 vs 0 1.44 (1.13–1.88) < 0.001 Presence of CSF oligoclonal IgG bands 1.28 (0.86–1.89) 0.492 CSF, cerebrospinal fluid; DMTs, Disease Modifying Therapies

Multivariable Cox model performed on the entire cohort to estimate the predictors of CDW

CSF, cerebrospinal fluid; DMTs, Disease Modifying Therapies

As a thorough and meticulous immunological assessment is required at our clinic before the start of each DMT, in line with Italian MS guidelines, the laboratory data were available at baseline for every patient. Before each OCR infusion, blood tests were performed at intervals of 2–21 days [mean (SD): 8.4 (6.7)]. The mean (SD) baseline Ig values were 1043 (232) for IgG, 107 (36) for IgM, and 202 (72) for IgA (mg/dL). Table 3 presents the mean values of IgG, IgM, and IgA at the different timepoints of observation, which corresponds to the OCR infusions. The graphical distribution of these values is reported in Fig. 2 : after the first infusion, a reduction in the IgG serum level was observed, which further decreased during the follow-up, therefore becoming stable after the third year. In parallel, IgM levels decreased over time, whereas IgA levels remained stable.

Table 3 Mean serum immunoglobulins (IgG, IgA, IgM) values throughout the observation period Time point (months) Number of patients IgG mean value (mg/dL) IgM mean value (mg/dL) IgA mean value (mg/dL) 0 140 1043 107 222 6 140 1010 82 220 12 140 995 76 217 18 140 1023 67 225 24 140 974 73 212 30 140 982 60 222 36 133 972 70 212 42 121 964 56 211 48 110 924 51 208 54 97 910 51 216 60 77 920 51 187 66 60 930 60 207 72 43 928 57 217

Mean serum immunoglobulins (IgG, IgA, IgM) values throughout the observation period

Serum immunoglobulin G (IgG), immunoglobulin M (IgM), and immunoglobulin A (IgA) mean levels (mg/dL) over time

The mixed-effects regression model assessed this trend and revealed that IgG levels decreased with time since the initiation of OCR: IgG levels decreased by 6.61% at 24 months after initiation and by 11.40% (95% CI − 18.6 to − 6.5,  p  < 0.001) at 48 months and persisted at similar levels throughout the follow-up. Total IgM levels decreased significantly by 31.77% (95% CI − 29.78 to − 33.75, p < 0.0001), 52.33% (95% CI − 50.31 to − 54.28, p < 0.0001), and 46.72% (95% CI − 44.73 to − 48.70, p < 0.0001) at the 24-, 48-, and 72-month visits, compared with baseline. Immunoglobulin A levels did not change significantly during the follow-up.

We did not find significant differences in the total NK-cell absolute number at 12, 36, and 48 months. No changes for T CD4+ and a slight reduction in T CD8+ lymphocytes during the follow-up were observed (Fig. 3 ).

Laboratory changes over time: profile plots of T CD4+ ( A ) and CD8+ ( B ) lymphocytes

Considering the assessment of infections, we reported three cases of localized herpes zoster, one sepsis, two complicated coronavirus disease 2019 infections, nine recurrent urinary tract infections, and one otitis. We also reported one case of anti CD20-related-organizing pneumonia. We collected these infusion-related reactions: itchy throat, 15 patients (10.7%); low-grade fever, 7 patients (5%); headache, 11 patients (7.8%). Notably, all selected patients exhibited normal IgG levels and IgA levels. Conversely, in four patients with urinary tract infections, we observed IgM values < 40 mg/dL.

The Cox model (Table 4 ) revealed that only age ≥ 55 years (HR 1.67, 95% CI 1.53–1.83) and an EDSS score ≥ 5.0 (HR 1.37, 95% CI 1.16–1.59) were associated with an increased risk of infection occurrence. No association was observed with decreased Ig levels.

Table 4 Multivariable Cox model performed on the entire cohort to estimate the predictors of the risk of infections. Variables HR (95% CI) p value Sex, female as reference 1.38 (0.69–2.75) 0.345 Age ≥ 55 years 1.67 (1.53–1.83) 0.005 Disease duration 1.05 (0.88–1.23) 0.482 Decreased IgG level 1.16 (0.99–1.34) 0.670 Decreased IgA level 1.03 (0.80–1.28) 0.674 Decreased IgM level 1.11 (0.84–1.42) 0.577 EDSS score ≥ 5.0 1.37 (1.16–1.59) 0.001 EDSS, Expanded Disability Status Scale

Multivariable Cox model performed on the entire cohort to estimate the predictors of the risk of infections.

EDSS, Expanded Disability Status Scale

Discussion

Real-world data of this cohort indicate that OCR stabilized disability progression and prevented clinical and radiological disease activity over the long term. The effect of OCR on the humoral and cellular immune responses did not determine a significant infection risk in the long term, with no association between decreased Ig levels and severe infections.

The amount of data currently available on the effectiveness and safety of B-cell-depleting therapies led to an increasing consideration of these as a viable strategic option for MS treatment [ 1 – 4 , 17 ]. A systematic literature review of real-world data for OCR highlighted a trend of effectiveness similar to the results reported in the OPERA I/II and ORATORIO clinical trials, considering diverse patient populations across different geographical locations [ 6 ].

Results from our study proved a marked reduction in the ARR with OCR use. Similar outcomes were observed in a study from the Danish MS Register, where the mean ARR for the entire cohort decreased from 0.58 (95% CI 0.53–0.62) in the year prior to OCR to 0.09 (95% CI 0.07–0.11) after treatment. Pontieri et al. also found a 24-week CDW and a 24-week confirmed disability improvement in 8.7% and 16.7% of the cohort, respectively, and, after the first year of treatment, an almost complete suppression of MRI activity [ 18 ]. Strong relapse suppression is also observed in a Spanish study, in which the ARR decreased from 1.11 (95% CI 0.95–1.23) before treatment to 0.09 (95% CI 0.05–0.16) after treatment initiation [ 19 ].

Lorefice et al. evaluated NEDA-3 status achievement in a cohort of OCR-treated patients, which was observed in 163/192 (84.9%) patients with RRMS and 9/16 (56%) patients with PPMS, with younger age and ARR in the year before OCR identified as significant determinants of disease stability [ 20 ].

The percentage of patients who experienced clinical relapses after starting OCR was modest across studies, primarily occurring between 3 and 30 months, and was linked to washout dynamics of prior DMTs, according to a recent review by Montalban et al. [ 9 ]. A recent study from RISM proved that OCR, compared with natalizumab, strongly suppresses relapse-associated worsening events and has a similar impact also on progression independent of relapse activity in naïve patients with RRMS [ 21 ].

Similar percentages of CDW observed in our study in the second year are reported in a multicentric study of patients with PPMS, where 50 (28.4%) subjects experienced CDW during the follow-up, and a baseline EDSS score of ≥ 5.0 resulted in a statistically significant positive predictor for the development of CDW and/or stopping treatment [ 22 ]. In another study, at the second year of follow-up, 90.5% of patients with RRMS, 64.7% of patients with PPMS, and 68.8% of patients with SPMS treated with OCR had no disability worsening, also significantly associated with lower baseline EDSS scores [ 23 ].

This study represents a long-term real-world cohort of patients treated with OCR: this is highly relevant not only for the efficacy assessment but also for safety-related questions, given the increasing concerns with a longer treatment duration and the need for longer term data. Several trials and real-world studies have described Ig levels during the treatment of patients with MS with anti-CD20 monoclonal antibodies [ 12 ]: in this recent review, OCR trials reported decreasing IgG levels, while real-world studies found IgG levels to be stable or decrease only slightly. Capasso et al. disclosed that the significant impact of OCR on both humoral and cellular immunity starts from its first infusion, with the reduction in CD19 and CD20 lymphocyte counts and serum IgG levels, followed by a stabilized trend of B cells and a further decline of IgG and CD8 lymphocyte levels [ 24 ].

The VELOCE trial suggested a relative stability of mean IgG levels, with some fluctuations and a minimal reduction after 24 weeks [ 25 ]. Moreover, in line with our results, several real-world evidence studies [ 24 ] showed an overall decrease in IgG levels during treatment with OCR in subgroups of patients. In the same trials and real-world evidence studies that reported an IgG decrease, IgM levels were found to be reduced with OCR treatment [ 12 ].

OPERA I/II and ORATORIO data revealed a correlation between reduced IgG, IgM, and IgA levels and an increased risk of severe infection, with the association being most significant for IgG rather than IgM or IgA [ 12 , 24 , 25 ]. Conversely, a retrospective cohort study suggested that a higher serum IgA level was associated with a reduced odds of infection and older age, higher serum IgA and IgG were associated with a reduced odds of infection, and a higher EDSS score was associated with an increased odds of antimicrobial use [ 26 ].

In the multivariable analysis, age ≥ 55 years and an EDSS score ≥ 5.0 were associated with an increased odds of infection. A higher EDSS score was linked to a higher frequency of urological complications and respiratory dysfunction [ 26 , 27 ]. Our results are also in line with those of a retrospective study of patients treated with anti-CD20 antibodies [ 28 ], where the risk of severe infection was associated with comorbidities, higher age, a longer duration of treatment, and higher EDSS scores. Our investigation did not reveal a significative correlation between infection susceptibility and Ig levels. In parallel with our observations, Oksbjerg at al. also highlighted that serum levels of IgG and IgM decreased over time but were not associated with a risk of severe infection.

A reduction in Ig levels is relevant because of the significant role of IgG in protective immunity but CD20-negative plasma cells are not fully targeted by anti-CD20 therapies and OCR proved to have multiple roles in the development and maintenance of regulatory plasma cells [ 29 , 30 ]. Values of NK and T CD4+ and T CD8+ lymphocytes over time are in line with other real-world studies [ 11 ].

Several limitations of this study deserve discussion. As with previous retrospective observational studies, incomplete or inaccurate data collection must be acknowledged. The collection of clinical and laboratory parameters was retrospectively performed considering both the extraction of data from RISM-App and paper medical records, thus lacking the power of a prospective collection and laboratory analysis ad hoc performed on the cohort. Unfortunately, we were unable to provide more thoughtful insights into MRI parameters because of the lack of standardized radiological protocols in our center.

Conclusions

Data from this cohort suggest that, over time, OCR reduced clinical and radiological disease activity and stabilized the progression of disability. Lower Ig levels were not associated with serious infections, and the impact of OCR on humoral and cellular immune responses did not determine a long-term infection risk.

Declarations

Funding

No funding was received for the preparation of this article.

Conflicts of interest/competing interests

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 the authors have received advisory board, membership, and 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, and Teva, and their local affiliates.

Ethics approval

The Italian MS Register (RISM) was approved by the Ethics Committee of the “Azienda Ospedaliera/Universitaria-Policlinico of Bari” (Study REGISTRO SM001, approved on 08/07/2016). The study was performed in accordance with the standards of ethics outlined in the Declaration of Helsinki.

Consent to participate

Patients signed an informed consent, in line with the Italian MS Register (RISM) rules, that allows the use of clinical data for research purposes.

Consent for publication

Not applicable.

Availability of data and material

Anonymized data will be shared on reasonable request from a qualified investigator.

Code availability

Not applicable.

Authors’ contributions

PI, TG, CF: conception and design of the study, acquisition and analysis of data, drafting the manuscript and figures; DP, AB: data acquisition and drafting a significant portion of the manuscript. All authors read and approved the final version.

References