BACKGROUND AND OBJECTIVES: Extended follow-up data from real-world cohorts of patients with multiple sclerosis treated with ocrelizumab (OCR) are becoming widely available. This monocentric retrospective study aimed to evaluate the long-term effectiveness of OCR and its impact on immunoglobulin (Ig) levels, lymphocyte subsets, and infections in a cohort of patients with relapsing remitting, primary progressive, and secondary progressive multiple sclerosis. METHODS: Patients followed at the Multiple Sclerosis Center of Bari with ≥ 2 years of OCR treatment were retrospectively recruited in 2024. Twelve-month confirmed disability worsening, improvement, and the annualized relapse rate before and after treatment start were estimated and follow-up magnetic resonance imaging scans were collected. Changes in IgG/IgM/IgA levels from baseline for up to 6 years of OCR treatment and serum levels of T CD4+, T CD8+, and natural killer lymphocytes were assessed. Infection occurrence, type, and outcomes were investigated. Multivariable Cox models examined the association of clinical and radiological baseline factors with the risk of confirmed disability worsening and the relationship of infections with clinical-laboratoristic risk factors. RESULTS: The final cohort retrieved 140 patients (80 relapsing remitting, 37 primary progressive, 23 secondary progressive) with a median (interquartile range) follow-up after treatment start of 4.62 (4.10-5.04) years. In the entire cohort, the mean annualized relapse rate decreased from 0.61 in the year before the start of OCR treatment to 0.02 in the second year, thereafter all patients in our cohort remained free of relapses and magnetic resonance imaging activity. The overall percentage of stable patients increased from the second to the fourth year, in parallel with a reduction in patients with confirmed disability worsening. A multifocal onset and the presence of at least two relapses before the start of OCR treatment were significant (p <
0.05) risk factors of confirmed disability worsening. During the follow-up, a reduction in the IgG serum level was observed, which further decreased, becoming stable after the third year. Immunoglobulin M levels slightly decreased over time, whereas IgA levels remained stable. No changes for T CD4+, natural killer cell absolute number, and a slight reduction in T CD8+ lymphocytes during follow-up were observed. Ocrelizumab did not determine a significant infection risk in the long term and no association was observed with Ig levels and severe infections. CONCLUSIONS: Ocrelizumab prevented disease activity over the long term and its effect on the immune system did not determine a significant infection risk in most patients.