Cancer treatment delays among patients with cancer living with HIV during the COVID-19 pandemic in the United States.

Abstract

131 Background: The COVID-19 pandemic led to care disruptions across the cancer continuum. It is unknown if immunosuppressed patients with cancer who may be at high risk for SARS-CoV-2 infection were disproportionately impacted. Our objective was to compare delays in cancer treatment initiation between people living with HIV (PLWH) and cancer, the general cancer population (GCP), and patients with cancer and a history of solid organ transplant (SOT), another immunosuppressed patient population. Methods: We used data from Flatiron Health electronic health record-derived de-identified database (2018-2021) comprised of US patients with cancer from 800 sites of care across the country. We included patients with lymphomas, myeloma, and melanoma, as well as cancers of the breast, lung, prostate, head & neck, and gynecologic or gastrointestinal systems. We calculated time to cancer treatment initiation (TTI) as the difference between the date of cancer diagnosis and the earliest date that one of the following treatments was recorded: cell based treatment, other local treatment, radiotherapy, surgery, systemic therapy, or transplant. 16.4%, 19.9%, and 13.3% of the GCP, PLWH, and SOT groups, respectively, did not have treatment information in the database. This may have been due to no receipt of treatment or lack of data in the medical record. All analyses reported were conducted among patients with available treatment data. PLWH and those with a history of SOT were selected using ICD 9/10 codes. After adjusting for age, sex, and metastatic disease (yes/no), we used a difference-in-difference analysis to compute TTI for each study group both prior to and during the COVID-19 pandemic. We then determined whether study group differences in cancer patient TTI (e.g., HIV-associated treatment delays) changed during the pandemic. Results: The sample included 181 PLWH, 65073 GCP patients, and 195 patients with a history of SOT. Overall, no significant delay in TTI was seen between the three groups pre-COVID. However, during the pandemic, the TTI was statistically significantly delayed for PLWH compared to the HIV-uninfected GCP (median TTI 41 days vs. 27 days, p = 0.0017). No significant difference was observed overall between the GCP and the SOT group (median TTI 27 days vs. 20 days, p = 0.376), evidence that COVID-related treatment delays were specific to PLWH. Difference in difference analysis revealed that, when compared to the GCP, there was a significant increase in TTI among PLWH over time, with delays increasing by almost one month during COVID [DID: 32.6 days (8.9-56.3); p = 0.007]. The increase in TTI for PLWH was consistent across treatment modalities, including surgery (DID: 55.1, 95% CI: 28.8-81.3, p < 0.001) and systemic therapy (DID: 30.4; 95% CI: 4.6-56.3, p = 0.021). Conclusions: PLWH and cancer experienced significant delays in cancer treatment initiation after diagnosis during the COVID-19 pandemic.