e18835 Background: Longitudinal data assessing seroconversion after different doses of SARS-CoV-2 vaccines and the role of personal protective measures to prevent COVID-19 infections remain scarce. This is the first study to evaluate the association between SARS-CoV-2 vaccination, tumor subgroups, cancer treatment, and personal protective measures with antibody titers as well as COVID-19 infections in a real-world cohort of cancer patients. Methods: This retrospective cohort study enrolled 633 adult patients all treated at a large oncology clinic in Hamburg (Germany) from January 2020 to July 2022. Data about demographics, cancer treatment, vaccinations, SARS-CoV-2-IgG antibody titers, personal protective measures, COVID-19 infections and infection-associated symptoms were reported in the patients’ health records. Primary outcomes were: (i) patients’ SARS-CoV-2-IgG antibody concentrations and (ii) COVID-19 infections. A multivariate regression model, reporting adjusted odds ratios (AOR), evaluated the association between collected variables and seroconversion rates after SARS-CoV-2 vaccination. A time-varying Kaplan-Meier survival analysis explored the association between the collected variables and the risk of infection. Results: Of 633 cancer patients, 52% had solid tumors and 48 % hematologic malignancies. Seroconversion was significantly associated with the number of vaccine doses, prior infections (AOR: 1.91, 95%CI: 1.01-3.60, p = .045), hematologic cancers (AOR: 0.58, 95%CI: 0.40-0.85, p = .005), and treatment with CD20/38 inhibitors or immune checkpoint inhibitors (ICI) (AOR: 0.58, 95%CI: 0.35-0.95, p = .030). 120 patients (19%) had a COVID-19 infection with 77% experiencing exhaustion, 67% fever, 62% rhinitis, 62% headache, 56% sore throat, and 48% a non-productive cough. Patients on anti-cancer treatments had a 63% (HR: 1.63, 95%CI: 1.04-2.57, p = .034) higher risk of infection. The risk of infection was 43% (AOR: 0.57, 95%CI: 0.38-0.85, p = .006) and 87% (AOR: 0.23, 95%CI: 0.11-0.47, p < .001) lower among patients with 2 and ≥3 vaccine doses, respectively. Wearing FFP2 masks (AOR: 0.74, 95%CI: 0.56-0.99, p = .046) and reducing social contacts (AOR: 0.81, 95%CI: 0.65-0.99, p = .042) were associated with a lower risk of infection, but this was not the case for wearing surgical masks or using a “Corona warning” app. Conclusions: In this study, SARS-CoV-2 vaccination was associated with lower infection rates in cancer patients. However, certain subgroups remained at an increased risk of infection despite vaccination, e.g. patients receiving CD20/38 or ICI anti-cancer therapy with hematologic cancers. More than two vaccine doses, FFP2 masks, and social distancing led to lower infection rates. These findings are instrumental in guiding recommendations for preventing COVID-19 infections in cancer patients.