Abstract Introduction: SARS-CoV2 (S-2) infection duration (S-2-D) and its impact on patients with cancer and mild to moderate COVID-19 undergoing cancer-directed therapy (CDT), especially in the underserved population, is not well described. We conducted a retrospective study to analyze S-2 positive (+) patients on CDT to describe the S-2-D and its impact on CDT. Methods: 299 patients with cancer were tested with nasopharyngeal (NP) S-2 PCR assay at Columbia University Medical Irving Center (CUIMC), a Minority-NCI Community Oncology site, of whom 77(26%) tested positive. We retrospectively analyzed 25 S-2 (+) patients with mild to moderate COVID-19 receiving CDT. NP PCR were repeated every one to two weeks until two successive negative (-) PCRs were obtained prior to restarting CDT. Time to two (-) PCR and serology results were recorded. Cycling thresholds (Ct) were obtained for S-2 specific targets and represented an indirect measure of viral load. Results: Demographics of N=25 patients included Hispanic (N=17, 68%), Black (N=1, 4%), White (N=6, 24%), and undeclared (N=1, 4%). Among the tumor histologies represented, gastrointestinal (N = 9, 36%), breast (N = 4, 16%), sarcoma (N = 3, 12%), and myeloma (N=2, 8%) were most common. Median time in days (d) to two (-) PCR was 24 (6-69). Nine (36%) patients were persistently (+) for more than six weeks. 20 patients had CDT interruption with a median time of 55 days off CDT. Two (10%) patients experienced disease progression during CDT interruption. 13 patients had >1 sequential (+) PCR. The mean time between 1st and 2nd NP S-2 PCR was 16d. There was a non-statistically significant trend towards a higher S-2 specific Ct on the 2nd NP S-2 PCR as compared to the 1st (31.21 vs 22.31, p=0.10). One patient was noted to be PCR (+) despite having developed S-2-specific IgG. Conclusion: Patients with cancer receiving CDT appear to have prolonged detectable S-2 by PCR, which can lead to interruption of CDT and POD in patients. Patients from underserved communities may be at greater risk given health care disparities. Further data are needed to help select patients with mild to moderate COVID-19 who can safely continue CDT while balancing the risk of worsening infection and risk of spread. The increased Ct over time in our patients suggest a decrease in viral load and infectious probability. Our lack of statistical significance is likely due to small sample size. Ct along with serology can be tools to help guide when and which patients can restart CDT. However, these tests need further investigation for applicability and validity. Citation Format: Winston Wong, Claire Brevia, Michael May, Susan Whittier, Eldad Hod, Ilenia Pellicciotta, Gary K. Schwartz, Maura Abbott, Gulam A. Manji. A double-edged sword: Prolonged detection of SARS-COV-2 in patients receiving cancer-directed therapy [abstract]. In: Proceedings of the AACR Virtual Meeting: COVID-19 and Cancer; 2020 Jul 20-22. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(18_Suppl):Abstract nr PO-004.