A retrospective pharmacovigilance study of ocrelizumab in multiple sclerosis: Analysis of the FDA Adverse Event Reporting System (FAERS) database

Abstract

Ocrelizumab, a humanized anti-CD20 monoclonal antibody approved for multiple sclerosis in 2017, modulates B-cell activity to reduce disease progression. While clinical trials demonstrate its efficacy, real-world safety data remain critical to identify rare or long-term adverse events (AEs) not captured in controlled studies. This study analyzed post-marketing AEs of ocrelizumab using data from the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS) to assess safety under real-world conditions. A retrospective pharmacovigilance analysis was conducted on FAERS data from Q1 2017 to Q4 2023. Disproportionality analyses including reporting odds ratio, proportional reporting ratio, Bayesian Confidence Propagation Neural Network, and Multi-item Gamma Poisson Shrinker were used to detect safety signals. Subgroup, sensitivity, and time-to-onset analyses further characterized AE profiles. Among 50,967 reports, the analysis confirmed several AEs consistent with the ocrelizumab prescribing information, including urinary tract infection, depression, cough, herpes zoster, and breast cancer. Furthermore, potential new safety signals not previously documented in the drug’s label were identified, such as alopecia, insomnia, weight increased, and sepsis. This study confirms known AEs and identifies unexpected AEs, emphasizing the need for targeted monitoring, particularly during early therapy. Real-world FAERS data complement trial findings, aiding clinicians in optimizing multiple sclerosis treatment strategies.

1. Introduction

Multiple sclerosis (MS) is a chronic inflammatory demyelinating disorder of the central nervous system, predominantly affecting young adults, with a higher prevalence in females.^[1] The disease is marked by immune-mediated destruction of myelin sheaths, leading to disrupted nerve conduction and diverse neurological impairments, including visual deficits, motor dysfunction, balance abnormalities, and cognitive deterioration.^[2] While substantial progress has been made in understanding MS, its precise etiology remains incompletely defined.^[3] Proposed contributors encompass genetic predisposition,^[4] environmental triggers,^[5] aberrant immune responses targeting myelin,^[6] blood–brain barrier compromise,^[7] neuroinflammatory cascades,^[8] and gut microbiota dysregulation.^[9]

Ocrelizumab, a humanized anti-CD20 monoclonal antibody, selectively depletes CD20 + B cells through mechanisms involving antibody-dependent cellular cytotoxicity, complement-dependent cytotoxicity, and phagocytosis.^[10] It was granted approval for market launch by the U.S. Food and Drug Administration (FDA) in March 2017. Ocrelizumab suppresses the production of autoantibodies and the ability of B cells to present antigens and secrete proinflammatory cytokines that drive demyelination in MS.^[11] The OPERA I and OPERA II trials proved the efficacy of the drug in reducing the rate of annual relapses.^[12] It is noteworthy that ocrelizumab is the first agent approved for relapsing and primary progressive MS.^[13]

While randomized controlled trials are the gold standard for establishing drug efficacy, their strict inclusion criteria limit generalizability to the broader patient populations seen in clinical practice.^[14] The pivotal OPERA I/II and ORATORIO trials established ocrelizumab’s initial safety profile, identifying infusion-related reactions and infections (particularly of the upper respiratory and urinary tracts) as the most common adverse events (AEs), with a noted imbalance in malignancies.^[12,13] Since its approval, a growing body of evidence from long-term trial extensions and real-world studies has further characterized its safety profile over several years of use.^[15–19] However, these cohort studies and systematic reviews, while invaluable, are often designed to monitor for known risks or overall safety trends. Pharmacovigilance databases like U.S. Food and Drug Administration Adverse Event Reporting System (FAERS) offer a distinct advantage: their vast scale and spontaneous reporting nature allow for powerful, hypothesis-generating disproportionality analyses. This methodology is uniquely suited to detect rare, unexpected, or previously uncharacterized statistical signals that may not be apparent in smaller, structured cohorts. Therefore, a dedicated FAERS analysis is essential to complement existing knowledge.

The FAERS gives vital information on the safety of marketed drugs and it collects voluntary reports from healthcare providers, patients, and manufacturers.^[20] Previous FAERS analyses have provided valuable but focused insights into ocrelizumab’s safety. For instance, studies have investigated specific AEs such as neutropenia and immune-mediated colitis, or have performed comparative safety assessments against rituximab.^[21–23] While essential, these studies often analyze data from earlier post-marketing periods or concentrate on a narrow range of AEs. The rapid accumulation of real-world data necessitates a new, comprehensive evaluation. Our study builds upon this prior work by utilizing a significantly longer data period (through Q4 2023), strengthening the statistical power to confirm existing signals and detect new, potentially rarer AEs. Furthermore, we employ a more granular analytical approach, including subgroup, sensitivity, and time-to-onset (TTO) analyses, to better characterize the overall risk profile across the entire spectrum of reported AEs. By conducting the most current and extensive analysis to date, this study aims to track ocrelizumab’s evolving real-world safety profile, identify new potential safety signals, and offer valuable guidance for clinical decision-making.

2. Methods

2.1. Data source and extraction

A retrospective pharmacovigilance study was conducted using the FAERS database (https://open.fda.gov/data/faers/), including reports from the drug’s approval in Q1 2017 up to Q4 2023. To identify all relevant reports, the drug information file was queried using both the generic name (“ocrelizumab”) and the brand name (“Ocrevus”). The inclusion criteria of the study were AE records where ocrelizumab was listed as the “primary suspected” drug. Data management involved the removal of duplicates and standardization of AEs nomenclature.

2.2. Data management

Initial data processing was performed using R software (version 4.3.2, R Foundation for Statistical Computing, Vienna, Austria). Quarterly data files were merged, and duplicate reports were removed in strict accordance with the FDA’s recommended cross-referencing method. Briefly, for reports with the same case identifier (CASEID), the entry with the latest FDA receipt date (FDA_DT) was retained. In instances where both CASEID and FDA_DT values coincided, the report with the highest primary identifier (PRIMARYID) was selected. To ensure consistency for subsequent statistical analyses, all reported AE terms were standardized and coded using the Medical Dictionary for Regulatory Activities version 27.1. Figure 1 presents a comprehensive flow diagram of the study design, illustrating the sequential procedures of data extraction, processing, and analytical methodology.

Figure 1.

Process diagram depicting treatment-emergent AEs linked to ocrelizumab in the FAERS database. AEs = adverse events, FAERS = U.S. Food and Drug Administration Adverse Event Reporting System.

2.3. Weibull distribution analysis

TTO of AEs associated with ocrelizumab was determined by calculating the duration between the initiation of therapy (START_DT in the therapy information file) and the occurrence of the AE (EVENT_DT in the demographic information file). The Weibull distribution was utilized to forecast temporal variations in AE risk profiles over time.

2.4. Sensitivity analysis

To minimize potential confounding factors, this investigation excluded AE reports involving the concomitant use of ocrelizumab with medications commonly prescribed for MS symptom management, including baclofen, modafinil, pregabalin, oxybutynin, and methylprednisolone.

2.5. Statistical analysis

All data processing and statistical analyses were conducted using R software (version 4.3.2). A descriptive analysis was performed to summarize the baseline demographic and clinical characteristics of the reports. Frequencies and percentages were calculated for categorical variables (e.g., gender, reporter type, country), while the median and interquartile range were reported for continuous variables (e.g., age).

To investigate potential associations between ocrelizumab and specific AEs, 4 established disproportionality analysis methods were employed: Reporting odds ratio (ROR), proportional reporting ratio (PRR), Bayesian Confidence Propagation Neural Network, and Multi-item Gamma Poisson Shrinker. An AE was considered a potential safety signal if it met the threshold criteria in at least 1 of these 4 analytical methods. The computational logic and specific threshold criteria for each method are detailed in Tables S1 and S2, Supplemental Digital Content, https://links.lww.com/MD/Q212.

2.6. Ethical considerations

This study was based on publicly available, de-identified data from the FAERS database. Therefore, institutional review board approval and patient informed consent were not required.

3. Results

3.1. Descriptive analysis

A total of 50,967 AE reports associated with ocrelizumab were included in this comprehensive analysis. Females (32,469 cases, 63.7%) significantly outnumbered males (12,347 cases, 24.2%), with gender information missing in 6151 cases (12.1%). Regarding age distribution, the majority of cases (47.8%) fell within the 18 to 65 year age group. Geographically, the United States reported the highest number of cases (30,886, 60.6%), followed by Canada, Germany, Italy, and Australia. Healthcare professionals submitted 50.3% of the AE reports. The baseline characteristics of AE reports related to ocrelizumab administration are comprehensively outlined in Table 1.

Table 1.

Clinical characteristics of ocrelizumab adverse event reports in the FAERS database (Q1 2017–Q4 2024).

Characteristics	Numbers	Case proportion
Number of reports	50,967
Gender
Male	12,347	24.2%
Female	32,469	63.7%
Missing	6151	12.1%
Age	48 (38, 57)
<18	105	0.2%
18–64	24,345	47.8%
≥65	2573	5.0%
Missing	23,944	47.0%
Top 5 reported countries
United States	30,886	60.6%
Canada	7684	15.1%
Germany	4081	8.0%
Italy	1196	2.3%
Austrilia	815	1.6%
Reporter
Healthcare professional	21,148	41.5%
Non-healthcare professional	29,314	57.5%
Missing	505	1.0%
Reporting year
2017	882	1.7%
2018	3690	7.2%
2019	5107	10.0%
2020	4691	9.2%
2021	6331	12.4%
2022	9080	17.8%
2023	8961	17.6%
2024	12,225	24.00%

FAERS = U.S. Food and Drug Administration Adverse Event Reporting System, IQR = interquartile range.

3.2. AEs at the System Organ Class (SOC) levels

Table 2 presents the signal strength of ocrelizumab at the SOC level, while Figure 2 illustrates the distribution patterns of AEs across 27 SOC categories associated with this therapeutic agent. Positive signals were observed as follows: infections and infestations (n = 36935, ROR 4.93, PRR 4.06, empirical Bayesian geometric mean [EBGM] 3.96, information component [IC] 1.98), nervous system disorders (n = 22795, ROR 1.91, PRR 1.79, EBGM 1.78, IC 0.83), musculoskeletal and connective tissue disorders (n = 9526, ROR 1.12, PRR 1.12, EBGM 1.12, IC 0.16), respiratory, thoracic and mediastinal disorders (n = 9126, ROR 1.2, PRR 1.19, EBGM 1.19, IC 0.25), and immune system disorders (n = 2203, ROR 1.07, PRR 1.07, EBGM 1.07, IC 0.1).

Table 2.

Signal strength of ocrelizumab-related adverse events in system organ classes (SOC) in the FAERS database.

SOC	Numbers	ROR (95%CI)	PRR (χ2)	EBGM (EBGM05)	IC (IC025)
Infections and infestations*	36,935	4.93 (4.87–4.99)	4.06 (88382.05)	4 (3.96)	2 (1.98)
General disorders and administration site conditions	25,572	0.84 (0.83–0.85)	0.86 (665.45)	0.86 (0.85)	-0.21 (-0.23)
Nervous system disorders*	22,795	1.91 (1.88–1.94)	1.79 (8481.14)	1.78 (1.76)	0.83 (0.81)
Injury, poisoning and procedural complications	10,930	0.53 (0.52–0.54)	0.56 (4337.62)	0.56 (0.55)	-0.84 (-0.87)
Investigations*	10,611	1.1 (1.08–1.12)	1.1 (93.09)	1.09 (1.08)	0.13 (0.1)
Musculoskeletal and connective tissue disorders*	9526	1.12 (1.1–1.15)	1.12 (122.55)	1.12 (1.1)	0.16 (0.13)
Respiratory, thoracic and mediastinal disorders*	9126	1.2 (1.17–1.22)	1.19 (280.92)	1.19 (1.17)	0.25 (0.22)
Gastrointestinal disorders	7966	0.56 (0.55–0.57)	0.58 (2586.82)	0.58 (0.57)	-0.78 (-0.81)
Skin and subcutaneous tissue disorders	7319	0.74 (0.72–0.75)	0.75 (656.11)	0.75 (0.73)	-0.42 (-0.45)
Psychiatric disorders	5515	0.6 (0.59–0.62)	0.62 (1373.34)	0.62 (0.61)	-0.69 (-0.73)
Vascular disorders	2643	0.83 (0.8–0.86)	0.83 (93.23)	0.83 (0.8)	-0.27 (-0.32)
Eye disorders	2495	0.75 (0.72–0.78)	0.76 (198.45)	0.76 (0.73)	-0.4 (-0.46)
Renal and urinary disorders	2410	0.72 (0.69–0.75)	0.73 (254.67)	0.73 (0.7)	-0.46 (-0.52)
Immune system disorders*	2203	1.07 (1.03–1.12)	1.07 (10.31)	1.07 (1.03)	0.1 (0.04)
Neoplasms benign, malignant and unspecified (incl cysts and polyps)	2132	0.39 (0.38–0.41)	0.4 (1956.06)	0.4 (0.39)	-1.31 (-1.37)
Cardiac disorders	1661	0.48 (0.46–0.5)	0.49 (920.26)	0.49 (0.47)	-1.04 (-1.11)
Metabolism and nutrition disorders	1479	0.43 (0.41–0.46)	0.44 (1078.33)	0.44 (0.42)	-1.18 (-1.26)
Blood and lymphatic system disorders	1403	0.49 (0.47–0.52)	0.5 (731.22)	0.5 (0.48)	-1.01 (-1.09)
Ear and labyrinth disorders*	1202	1.68 (1.58–1.77)	1.67 (322.77)	1.67 (1.59)	0.74 (0.65)
Reproductive system and breast disorders	947	0.83 (0.78–0.88)	0.83 (32.99)	0.83 (0.79)	-0.27 (-0.36)
Hepatobiliary disorders	623	0.44 (0.41–0.48)	0.44 (441.55)	0.44 (0.41)	-1.17 (-1.29)
Social circumstances	494	0.64 (0.58–0.7)	0.64 (101.96)	0.64 (0.59)	-0.65 (-0.78)
Pregnancy, puerperium and perinatal conditions	420	0.66 (0.6–0.73)	0.66 (71.69)	0.66 (0.61)	-0.59 (-0.73)
Surgical and medical procedures	306	0.12 (0.11–0.14)	0.13 (1883.16)	0.13 (0.12)	-2.98 (-3.15)
Endocrine disorders	279	0.62 (0.55–0.7)	0.62 (63.75)	0.62 (0.57)	-0.68 (-0.85)
Congenital, familial and genetic disorders	137	0.3 (0.25–0.35)	0.3 (224.02)	0.3 (0.26)	-1.73 (-1.98)
Product issues	103	0.03 (0.03–0.04)	0.03 (2934.67)	0.03 (0.03)	-4.9 (-5.18)

AEs = adverse events, CI = confidence interval, EBGM = empirical Bayesian geometric mean, EBGM05 = the lower limit of the 95% CI of EBGM, IC = information component, FAERS = U.S. Food and Drug Administration Adverse Event Reporting System, IC025 = the lower limit of the 95% CI of the IC, PRR = proportional reporting ratio, ROR = reporting odds ratio.

^*Statistically significant signals in algorithm.

Figure 2.

Proportional representation of ocrelizumab-induced AEs stratified by system organ class (SOC). AEs = adverse events.

3.3. AEs at the preferred term (PT) levels

At the PT level, Table 3 systematically presents the frequency ranking and statistical significance analysis of the top 68 ocrelizumab-associated AEs. Among these AEs, some are consistent with the warnings listed on the drug label, such as urinary tract infection, depression, cough, herpes zoster, breast cancer, and so on. Furthermore, the investigation identified unexpected AEs not previously documented in the drug’s labeling information, including alopecia, insomnia, weight increased, and sepsis.

Table 3.

The top 70 adverse events associated with ocrelizumab at the preferred term (PT) level.

PT	Numbers	ROR (95% CI)	PRR (χ2)	EBGM (EBGM05)	IC (IC025)
Covid-19*	11,057	15.58 (15.28–15.9)	14.62 (131393.55)	13.69 (13.47)	3.78 (3.75)
Fatigue*	5084	2.35 (2.29–2.42)	2.31 (3799.32)	2.3 (2.25)	1.2 (1.16)
Urinary tract infection*	3422	7.64 (7.38–7.91)	7.5 (18643.47)	7.27 (7.06)	2.86 (2.81)
Headache*	2607	1.6 (1.54–1.67)	1.6 (579.89)	1.59 (1.54)	0.67 (0.61)
Multiple sclerosis*	2212	19.16 (18.34–20.02)	18.92 (34327.16)	17.37 (16.75)	4.12 (4.05)
Nasopharyngitis*	2182	4.36 (4.18–4.55)	4.31 (5454.64)	4.24 (4.1)	2.09 (2.02)
Asthenia*	2158	2.28 (2.18–2.38)	2.26 (1508.44)	2.25 (2.17)	1.17 (1.1)
Pneumonia*	2148	2.49 (2.38–2.6)	2.47 (1865.37)	2.45 (2.37)	1.29 (1.23)
Gait disturbance*	2035	4.18 (4–4.37)	4.14 (4758.87)	4.07 (3.93)	2.03 (1.96)
Infusion-related reaction*	2014	10.85 (10.37–11.35)	10.73 (16883.93)	10.23 (9.86)	3.36 (3.29)
Multiple sclerosis relapse*	1960	12.48 (11.92–13.07)	12.35 (19270.48)	11.69 (11.25)	3.55 (3.48)
Pain*	1959	1.1 (1.05–1.15)	1.1 (16.34)	1.09 (1.05)	0.13 (0.06)
Fall*	1883	2.21 (2.11–2.31)	2.19 (1217.61)	2.18 (2.1)	1.13 (1.06)
Influenza*	1728	5.54 (5.28–5.82)	5.5 (6200.49)	5.38 (5.17)	2.43 (2.36)
Cough*	1511	1.93 (1.83–2.03)	1.92 (660.81)	1.91 (1.83)	0.93 (0.86)
Pyrexia*	1492	1.69 (1.61–1.78)	1.69 (415.23)	1.68 (1.61)	0.75 (0.67)
Sars-cov-2 test positive*	1393	25.07 (23.71–26.51)	24.87 (28404.97)	22.24 (21.22)	4.47 (4.39)
Pruritus*	1392	1.35 (1.28–1.43)	1.35 (127.17)	1.35 (1.29)	0.43 (0.35)
Herpes zoster*	1325	8.58 (8.12–9.07)	8.52 (8445.29)	8.21 (7.84)	3.04 (2.96)
Muscular weakness*	1286	4.77 (4.51–5.04)	4.74 (3717.29)	4.66 (4.45)	2.22 (2.14)
Pain in extremity*	1276	1.72 (1.63–1.82)	1.72 (382.2)	1.71 (1.64)	0.78 (0.7)
Dizziness	1268	1.02 (0.97–1.08)	1.02 (0.66)	1.02 (0.98)	0.03 (-0.05)
Malaise*	1253	1.08 (1.02–1.14)	1.08 (6.86)	1.08 (1.03)	0.11 (0.02)
Hypoaesthesia*	1230	3.38 (3.2–3.58)	3.36 (2014.6)	3.33 (3.17)	1.73 (1.65)
Throat irritation*	1191	10.99 (10.37–11.65)	10.92 (10184.62)	10.41 (9.91)	3.38 (3.29)
Alopecia*	956	1.51 (1.42–1.61)	1.51 (164.92)	1.51 (1.43)	0.59 (0.5)
Infection*	912	2.23 (2.09–2.38)	2.23 (611.09)	2.21 (2.1)	1.15 (1.05)
Balance disorder*	901	4.25 (3.98–4.54)	4.24 (2183.98)	4.17 (3.94)	2.06 (1.96)
Memory impairment*	882	2.34 (2.19–2.5)	2.33 (666.8)	2.32 (2.19)	1.21 (1.12)
Oral herpes*	877	16.75 (15.63–17.95)	16.67 (11927.25)	15.46 (14.6)	3.95 (3.85)
Paraesthesia*	848	2.21 (2.06–2.36)	2.2 (549.62)	2.19 (2.07)	1.13 (1.03)
Oropharyngeal pain*	814	3.25 (3.03–3.48)	3.24 (1239.88)	3.2 (3.02)	1.68 (1.58)
Muscle spasms*	812	1.78 (1.66–1.9)	1.77 (271.66)	1.77 (1.67)	0.82 (0.72)
Back pain*	790	1.32 (1.23–1.42)	1.32 (61.97)	1.32 (1.25)	0.4 (0.3)
Depression*	784	1.53 (1.43–1.64)	1.53 (142.29)	1.52 (1.44)	0.61 (0.5)
Insomnia*	783	1.21 (1.12–1.29)	1.2 (27.17)	1.2 (1.13)	0.27 (0.16)
Sinusitis*	749	2.75 (2.56–2.96)	2.75 (822.1)	2.72 (2.56)	1.45 (1.34)
Feeling abnormal*	747	1.23 (1.15–1.32)	1.23 (32.02)	1.23 (1.16)	0.3 (0.19)
Weight increased*	692	1.2 (1.12–1.3)	1.2 (23.31)	1.2 (1.13)	0.26 (0.15)
Covid-19 pneumonia*	646	12.76 (11.78–13.82)	12.72 (6559.51)	12.02 (11.24)	3.59 (3.47)
Sepsis*	644	2.27 (2.1–2.46)	2.27 (451.67)	2.25 (2.11)	1.17 (1.06)
Hypersensitivity*	586	1.12 (1.04–1.22)	1.12 (7.94)	1.12 (1.05)	0.17 (0.05)
Cystitis*	583	7.22 (6.64–7.84)	7.19 (3003.42)	6.98 (6.51)	2.8 (2.68)
Tremor*	560	1.43 (1.31–1.55)	1.43 (70.66)	1.42 (1.33)	0.51 (0.39)
Bronchitis*	558	3.07 (2.82–3.33)	3.06 (762.3)	3.03 (2.82)	1.6 (1.47)
Urticaria*	550	1.27 (1.17–1.38)	1.27 (31.51)	1.27 (1.18)	0.34 (0.22)
Hypertension	549	1.02 (0.94–1.11)	1.02 (0.24)	1.02 (0.95)	0.03 (-0.09)
Upper respiratory tract infection*	549	4.58 (4.21–4.98)	4.57 (1496.2)	4.49 (4.18)	2.17 (2.04)
Migraine*	543	2.06 (1.9–2.25)	2.06 (293.82)	2.05 (1.91)	1.04 (0.91)
Illness*	510	1.38 (1.27–1.51)	1.38 (53.61)	1.38 (1.28)	0.46 (0.34)
Jc polyomavirus test positive*	505	178.75 (158.54–201.53)	178.21 (47109.27)	94.81 (85.75)	6.57 (6.41)
Flushing*	476	2.45 (2.24–2.68)	2.44 (401.6)	2.43 (2.25)	1.28 (1.15)
Visual impairment*	458	1.26 (1.15–1.39)	1.26 (25.03)	1.26 (1.17)	0.34 (0.2)
Muscle spasticity*	450	17.11 (15.54–18.85)	17.07 (6274.87)	15.81 (14.58)	3.98 (3.84)
Influenza like illness*	446	2.52 (2.3–2.77)	2.52 (403.96)	2.5 (2.31)	1.32 (1.18)
Mobility decreased*	441	2.26 (2.06–2.48)	2.26 (305.08)	2.24 (2.07)	1.16 (1.03)
Heart rate increased*	430	1.76 (1.6–1.94)	1.76 (139.45)	1.75 (1.62)	0.81 (0.67)
Chills*	428	1.46 (1.33–1.61)	1.46 (62.43)	1.46 (1.35)	0.55 (0.41)
Rhinorrhoea*	398	2.12 (1.92–2.34)	2.12 (233.76)	2.11 (1.94)	1.08 (0.93)
Musculoskeletal stiffness*	391	1.56 (1.42–1.73)	1.56 (78.77)	1.56 (1.43)	0.64 (0.49)
Vision blurred*	385	1.17 (1.06–1.29)	1.17 (9.11)	1.17 (1.07)	0.22 (0.07)
Fungal infection*	381	4.36 (3.94–4.82)	4.35 (962.86)	4.28 (3.93)	2.1 (1.95)
Cellulitis*	380	2.94 (2.65–3.25)	2.93 (477.45)	2.9 (2.67)	1.54 (1.39)
Breast cancer*	374	1.19 (1.08–1.32)	1.19 (11.57)	1.19 (1.09)	0.25 (0.1)
Stress*	365	1.88 (1.7–2.09)	1.88 (149.52)	1.87 (1.72)	0.91 (0.75)
B-lymphocyte count decreased*	365	123.44 (108.34–140.64)	123.17 (27396.55)	76.67 (68.74)	6.26 (6.08)
Cognitive disorder*	358	2.93 (2.64–3.25)	2.92 (446.55)	2.89 (2.65)	1.53 (1.38)
Feeling hot*	347	2.43 (2.19–2.7)	2.43 (288.19)	2.41 (2.21)	1.27 (1.11)
Immunodeficiency*	347	7.38 (6.63–8.22)	7.37 (1843.64)	7.15 (6.53)	2.84 (2.68)
Maternal exposure before pregnancy*	337	17.71 (15.84–19.8)	17.68 (4873.6)	16.33 (14.87)	4.03 (3.87)

CI = confidence interval, EBGM = empirical Bayesian geometric mean, EBGM05 = the lower limit of the 95% CI of EBGM, IC = information component, IC025 = the lower limit of the 95% CI of the IC, PRR = proportional reporting ratio, PT = preferred term, ROR = reporting odds ratio.

^*Statistically significant signals in algorithm.

3.4. Subgroup analysis

A subgroup analysis was subsequently conducted on the 100 most frequently reported AEs. Gender-specific subgroup analysis related to ocrelizumab revealed that certain AEs were exclusively observed in males, including seizure, nephrolithiasis, and amnesia (Tables S3, Supplemental Digital Content, https://links.lww.com/MD/Q212). Asthenia, maternal exposure before pregnancy, fungal infection, and stress were uniquely identified in females (Tables S4, Supplemental Digital Content, https://links.lww.com/MD/Q212).

The age subgroup analysis demonstrated a relatively consistent distribution of AEs across all age groups, with urinary tract infections, fatigue, headache, asthenia, and pain emerging as common positive signals across all cohorts. In the 18 to 65 age group, patients should be particularly attentive to the occurrence of nasopharyngitis, infusion-related reaction, throat irritation, and oropharyngeal pain. Further attention should be paid to sepsis, cystitis, and cellulitis in the over-65 age group. Comprehensive details supporting these results can be found in Tables S5 to S6, Supplemental Digital Content, https://links.lww.com/MD/Q212.

3.5. Sensitivity analysis

A sensitivity analysis was conducted to enhance the robustness of the findings in this study. To evaluate the independent AE profile of ocrelizumab, we have excluded 5 concomitant medications typically used in combination with ocrelizumab for the treatment of MS, including baclofen, modafinil, pregabalin, oxybutynin, and methylprednisolone. Subsequent disproportionality analysis was reconducted, and the following items remained as positive signals: urinary tract infection, depression, cough, herpes zoster, breast cancer, alopecia, insomnia, weight increased, and sepsis, as detailed in Table S7, Supplemental Digital Content, https://links.lww.com/MD/Q212.

3.6. TTO analysis

Figure 3 demonstrates that the majority of ocrelizumab-associated AEs were observed during the initial 30-day post-administration period (n = 3420, representing 15.0% of cases), with a progressive decline in event frequency thereafter. Notably, AE occurrences persisted beyond the first year of ocrelizumab therapy. The corresponding cumulative incidence pattern is displayed in Figure 4. The temporal characteristics of AE onset were available for a cohort of 22,808 treated patients. Statistical modeling using the Weibull distribution analysis demonstrated an initial failure pattern, confirming a temporal reduction in AE incidence. The complete statistical parameters characterizing this temporal pattern are enumerated in Table 4.

Figure 3.

Time-to-onset of AEs induced by ocrelizumab. AEs = adverse events.

Figure 4.

Ocrelizumab-related AEs: cumulative incidence over time. AEs = adverse events.

Table 4.

Time-to-onset of ocrelizumab-associated AEs and Weibull distribution analysis.

Drug	TTO (days)		Weibull distribution
Ocrelizumab	Case reports	Median (d) (IQR)	Scale parameter: α (95% CI)	Shape parameter: β (95% CI)	Type
	22,808	432 (146, 942)	549.34 (540.33, 558.35)	0.83 (0.81, 0.84)	Early failure

AEs = adverse events, CI = confidence interval, IQR = interquartile range, TTO = time-to-onset.

4. Discussion

This pharmacovigilance investigation employed 4 distinct disproportionality analysis methods to systematically evaluate AEs associated with ocrelizumab since its market approval in 2017, utilizing comprehensive FAERS database assessments. These real-world findings validate the majority of AEs listed in the drug’s prescribing information, containing urinary tract infection, depression, cough, herpes zoster, breast cancer, etc. The analysis additionally revealed unexpected AEs associated with the therapeutic agent, such as alopecia, insomnia, weight increased, and sepsis. It is important to note that the signal for “weight increased” corresponds to the specific Medical Dictionary for Regulatory Activities PT and should be distinguished from a formal diagnosis of obesity; however, its clinical significance warrants further observation. These newly identified safety signals underscore the critical need for enhanced pharmacovigilance throughout ocrelizumab treatment, thereby providing new safety information for clinicians managing MS with ocrelizumab.

It is important to interpret these findings within the context of pharmacovigilance. The FAERS database captures any adverse event reported after drug administration, and a statistical signal does not imply direct causation. For an immunosuppressive agent like ocrelizumab, strong signals for infections such as COVID-19 or herpes zoster are clinically significant, suggesting a potential increased susceptibility in treated patients. Therefore, such events are considered critical safety signals, even though the drug does not directly “cause” the virus. Our analysis presents these statistical associations as they appear in the data to provide a comprehensive and unbiased overview of the real-world safety profile.

This study confirmed several known AEs, including urinary tract infection, depression, cough, herpes zoster, and breast cancer. Herpes zoster, caused by reactivation of the varicella-zoster virus, can lead to significant physical discomfort and reduced quality of life. The hallmark presentation includes a painful unilateral dermatomal rash, with severe pain that may persist for months to years in some patients. Immunocompromised individuals may develop disseminated cutaneous or visceral infections, which can be life-threatening.^[24] According to the OPERA I and II trials, the patients that received ocrelizumab showed a higher percentage of developing herpesvirus infections than if they received interferon beta-1a (5.9% vs 3.4%).^[12] The ORATORIO study showed similar results, with ocrelizumab causing herpes virus infections significantly more than placebo (4.7% vs 3.3%), and with oral herpes basically only in ocrelizumab (2.3% vs 0.4%).^[13] Ocrelizumab achieves its therapeutic effect by depleting CD20 + B cells, which play a vital role in antiviral immunity through antibody production and immune regulation. This B-cell depletion may increase the risk of latent virus reactivation, such as varicella-zoster virus.^[25,26] In conclusion, given that herpes zoster-associated pain may compromise patient adherence to ocrelizumab, prompt antiviral therapy (e.g., acyclovir) should be initiated upon herpes zoster onset. Prior to ocrelizumab administration, clinicians should educate patients regarding the potential risk of herpes zoster and implement close monitoring.

In our study, breast cancer was also observed with ocrelizumab. Clinical trial data was also used to assess the risk associated with orizumab and cancer, such as: in the OPERA I/II trials, neoplasms occurred in 0.5% (4/825) of ocrelizumab-treated patients (breast cancer, renal carcinoma, melanoma) versus 0.2% (2/829) with interferon beta-1a during the 96-week period.^[12] In a case series, 2 instances of breast cancer were suspected to have emerged subsequent to ocrelizumab treatment in individuals with MS.^[27] This may be associated with the increased susceptibility to malignancies due to the inactivation of the protective functions of normal B cells, which are essential for preventing tumor formation and promoting tumor cell lysis. Emerging evidence from contemporary research indicates that the suppression of B cells can deteriorate the prognosis of breast cancer patients.^[28] OPERA I/II trials recommended age-related cancer screening guidelines be followed and that long-term studies be conducted to clarify the association between B-cell depletion and tumor risk in chronic autoimmune disease.

Notably, other unexpected AEs were also observed, including alopecia, insomnia, weight increased, and sepsis. Sepsis may present with septic shock (such as decreased blood pressure, pallor of the skin, and decreased urine output), and organ dysfunction resulting from insufficient organ perfusion (including acute respiratory distress syndrome, renal failure, changes in mental status, etc). In addition, there may also be coagulation disorders (such as petechiae, ecchymoses on the skin, epistaxis, gingival bleeding, etc). It has high mortality, especially in cancer patients.^[29,30] Serious infections occurred at rates of 2.01 per 100 patient-years and the incidence of this severe infection aligns with the range documented in epidemiological studies.^[31] In a case report, a young female patient with MS developed sepsis during treatment with ocrelizumab.^[32] The management measures for sepsis mainly include the timely use of effective antibacterial drugs for anti-infective treatment, and simultaneously providing supportive treatment to maintain the stability of the patient’s vital signs, enhance the body’s immunity, and address any possible complications. The observed studies underscore the need for heightened clinical surveillance for infectious complications during ocrelizumab therapy. Particular attention should be paid to patients with predisposing factors for infection, including those with prior recurrent infections or compromised immune function.

This investigation identified alopecia as another unexpected AE associated with ocrelizumab therapy. Moreover, the subgroup analysis demonstrated that this adverse event was exclusively observed in female patients. Current evidence demonstrates that alopecia in adult populations exerts profound psychosocial consequences extending beyond physical manifestations, including increased anxiety, anger, depressive symptoms, embarrassment, social isolation, and diminished self-esteem, all of which significantly compromise patients’ quality of life and occupational functioning.^[33,34] Given that alopecia represents an unlabeled potential adverse effect, proactive disclosure of this risk to MS patients initiating ocrelizumab treatment is clinically imperative. Systematic management of this and other treatment-emergent AEs is essential for optimizing therapeutic outcomes and preserving patients’ overall wellbeing.

Our subgroup analyses revealed distinct gender-specific AE profiles associated with ocrelizumab therapy. Male patients exhibited unique neurological and renal manifestations, including seizure, nephrolithiasis, and amnesia, whereas female patients were more prone to asthenia, maternal exposure-related event, fungal infection, and stress response. These findings suggest potential sex-based differences in drug metabolism or physiological susceptibility that warrant further investigation. Physicians should maintain heightened vigilance during therapeutic management of these patients. Age-stratified results demonstrated relative consistency in AE distribution across all cohorts, with urinary tract infections, fatigue, headache, asthenia, and pain emerging as common class effects. However, age-specific vigilance is advised: adults (18–65 years) showed higher susceptibility to nasopharyngitis, infusion-related reactions, throat irritation and oropharyngeal pain, while elderly patients (>65 years) faced elevated risks of sepsis, cystitis, and cellulitis. These age-related patterns may reflect immunological senescence or comorbid disease interactions in older populations. Current evidence advocates enhanced vigilance for population-specific adverse drug reactions, with monitoring intensity adjusted according to patient sex and age, to achieve superior safety profiles and therapeutic adherence.

It is noteworthy that our analysis did not highlight neutropenia or colitis among the top 68 most prominent signals, despite their identification in more focused pharmacovigilance studies. This difference likely stems from the differing scopes of analysis. Our study conducted a global assessment of all reported AEs to identify the most frequently reported signals across the entire safety profile. In contrast, the aforementioned studies were designed as specific investigations into these individual AEs of interest. Thus, while neutropenia and colitis may not be the most commonly reported AEs in the FAERS database for ocrelizumab, the evidence from targeted analyses confirms they remain important, specific risks that clinicians should continue to monitor. Our findings complement, rather than contradict, these focused investigations by providing a broader, ranked overview of the most common real-world safety signals.

The sensitivity analysis reinforced the robustness of our findings, as exclusion of concomitant MS medications (baclofen, modafinil, pregabalin, oxybutynin, and methylprednisolone) did not substantially alter the AE profile. The persistence of urinary tract infection, depression, cough, herpes zoster, breast cancer, alopecia, insomnia, weight increased, and sepsis as dominant signals underscores their likely direct association with ocrelizumab rather than polypharmacy effects. Notably, although these adverse reactions are predominantly non-life-threatening, their substantial impact on patients’ quality of life and treatment persistence necessitates prioritized integration into clinical management protocols.

This study has several limitations. First, the FAERS database relies on spontaneous reports from physicians, nurses, pharmacists, and patients, which may lead to underreporting, underestimation, and misreporting, thereby affecting the reliability of the results. Second, the FAERS database often lacks detailed information on drug dosage, treatment duration, comorbidities, and concomitant medications, limiting further exploration of potential confounding factors. In addition, 60.6% of the AE reports originated from the United States, which may affect the external validity of the findings; thus, future research incorporating data from broader geographic regions is warranted. Finally, disproportionality analysis can only establish statistical associations between drugs and AEs, rather than causal relationships, highlighting the need for further large-scale prospective studies to validate our findings.

5. Conclusion

This study analyzed AE reports related to ocrelizumab since its market approval using the FAERS database. We not only confirmed known AEs but also identified unexpected AEs, such as alopecia, insomnia, sepsis, and weight increased. Moreover, this study highlights the critical importance of early-phase monitoring during treatment. These findings provide valuable real-world safety information on ocrelizumab for clinicians and regulatory authorities, contributing to enhanced patient safety during therapy.

Author contributions

Conceptualization: Pei-Jie Ye.

Data curation: Hong-Yan Ma, Ya-Bin Ma.

Formal analysis: Hong-Yan Ma, Ya-Bin Ma.

Funding acquisition: Rong Bai.

Investigation: Hong-Yan Ma, Ya-Bin Ma.

Methodology: Pei-Jie Ye.

Project administration: Rong Bai.

Resources: Rong Bai.

Software: Pei-Jie Ye.

Visualization: Hong-Yan Ma, Ya-Bin Ma.

Writing – original draft: Hong-Yan Ma, Ya-Bin Ma.

Supplementary Material

medi-104-e44883-s001.docx ^{(65.7KB, docx)}