SARS-CoV-2 infection risk and COVID-19 prevalence in cancer patients during the first wave of COVID-19 pandemic in a northernItaly virus epicenter area.

e22560 Background: The COVID-19 pandemic significantly impacted cancer patients, who are more vulnerable to severe outcomes. This study examines the risk of SARS-CoV-2 infection, COVID-19 diagnosis, and mortality in cancer patients, exploring variations based on tumor subtypes and stage during the first wave of the pandemic in Northern Italy. Methods: We collected data from the Province of Parma Cancer Registry, COVID-19 hospital medical records, and the local surveillance system of all cases positive for SARS-CoV-2 from the beginning of the outbreak (24 th February) to 11 th June 2020. We defined patients as either “users” or “non-users” based on the evidence of any referral to health services during the study period. We classified cancer patients diagnosed with COVID-19 as either "active" or "non-active" based on whether they were currently receiving cancer treatment. Cumulative incidence of COVID-19 diagnosis and all-causes death were analyzed by Cox regression and depicted using Aalen-Johansen curves. For interpretative purposes, multiple correspondence analysis (MCA) and a Hierarchical Clustering on Principal Components (HCPC) were performed to identify clusters of patients. Results: 40,148 cancer patients (mean age 68 years; 58% females; 45% “users”) were analyzed. The cumulative risk of SARS-CoV-2 infection was significantly higher for cancer patients compared to non-cancer patients (11% vs. 7%; p < 0.0001). Among cancer patients, “users” had a significantly higher incidence of COVID-19 diagnosis than “non-users” (HR 1.18 [95% CI: 1.04-1.34]; p = 0.011). Breast cancer (BC) and lung cancer (LC) patients had respectively the lowest and highest cumulative incidence of COVID-19 (HR 0.86 and HR 4.31). “Active” patients resulted at a higher risk of all-causes death compared to “non-active” subjects (HR 1.83 [95% CI: 1.32-2.53; p < 0.001]). BC patients were at lower risk of all-causes death compared to patients affected by the other cancer subtypes (HR 0.53 [95% CI: 0.31-0.89]; p = 0.017). MCA explained 31.5% of the overall variability by the two first dimensions, the first one mainly described by disease stage and activity status and the second one by subtypes and age. Based on this map, the HCPC analysis identified three clusters. Cluster 1 includes young women with BC, no comorbidities, and recovery from COVID-19. Cluster 2 comprises patients with other cancers, mainly male, Stage I-III, and COVID-19 non-recovery. Cluster 3 consists of “active” stage IV LC patients with poor radiological responses and death. Conclusions: Our analysis found that cancer patients were more vulnerable to SARS-CoV-2 infection, with varied outcomes based on tumor subtypes. These findings highlight the need for robust safety protocols, prioritizing vaccination, and balancing cancer treatments with the risk of viral exposure, especially during any future pandemics.