Wazir, Muhammad Hisham Khan

Background CV mortality has been recognized as the preventable cause of death in most cancers. However, there is a scarcity of data exploring CV mortality in adults diagnosed with leukemia. Hence, understanding the timing, and quantifying the magnitude of CV mortality can facilitate strategies to reduce the burden of mortality in these patients. Methods Surveillance, Epidemiology, and End Results (SEER) database (2004-2018) was queried to obtain CV mortality (defined as deaths due to heart diseases, hypertension, cerebrovascular diseases, atherosclerosis, and aortic aneurysm and dissection) in adults (>18 years of age) diagnosed with leukemias including acute lymphocytic leukemia (ALL), chronic lymphocytic leukemia (CLL), acute myeloid leukemia (AML), and chronic myeloid leukemia (CML). Standardized mortality ratios (SMRs) were calculated. SMR was defined as the observed deaths from each category divided by the expected number of deaths in the age-matched US population for the same period. SMR >1 indicated increased while <1 indicated decreased mortality. SMR was presented by latency periods of <2, 2-5, and >5 years after the initial diagnosis of leukemias and stratified by age (<50 years, 50-70 years, >70 years), race (White, Black, Asian/Pacific Islander, Native American), ethnicity (Hispanic, Non-Hispanic), and marital status (married vs single/divorced/widowed). Results This analysis included 120,193 leukemia patients. Of which, 61,763 had lymphocytic leukemia and 58,430 were diagnosed with myeloid leukemia. In overall population, a significant increase in mortality was observed in adults diagnosed with leukemia (SMR: 1.34, 95% CI: 1.27-1.40). The risk of CV death was highest <2 years of initial diagnosis (1.94, 1.87-2.01) as compared to the population who died 2-5 years (1.17, 1.11-1.23) and >5 years (1.07, 1.01-1.14) after the initial diagnosis. In terms of different subgroups, patients aged <50 years observed the highest risk of CV mortality within < 2 years (9.35, 6.54-13.38) compared to patients aged 50-70 years (2.73, 2.45-3.04), and aged >70 years (1.80, 1.73-1.88). Similarly, regarding race, Asians/Pacific islanders had the highest risk of CV mortality (4.41, 3.70-5.26) followed by Blacks (2.58, 2.26-2.94), and Whites (1.85, 1.78-1.92) within <2 years of initial diagnosis. There was a significant increase in observed mortality in Hispanic (2.83, 2.47-3.23) as well as in non-Hispanic (1.90, 1.83-1.97) population within <2 years of initial diagnosis. However, the magnitude was relatively higher in Hispanic as compared to non-Hispanics. The risk of CV death was observed to be higher in single/divorced/widowed patients (2.30, 2.00-2.20) compared to those who were married (1.81, 1.71-1.91). The CV mortality risk remained consistent for all subtypes of leukemias (ALL, CLL, AML and CML) in overall patient population and among subgroups of interest over different latency periods of <2, 2-5, and > 5years after the initial diagnosis. Small sample size in Native Americans precluded any meaningful statistics. Conclusions: Risk of CV mortality among adults with leukemia is observed to be highest among younger (<50 years), Asians, Hispanics, and single/divorced/widowed patients, especially within 2 years of initial diagnosis. A risk-adapted approach may be warranted for cardiovascular care in these patients. Targeted cardio-oncological strategies should be aimed at improving CV mortality in patients who are at the highest risk of death.

6557 Background: Trials that inadequately include underrepresented minority (URM) populations widen inequities in cancer care. We assessed reporting patterns, representation, and trends of URM representation in breast cancer (BC) trials. Methods: We analyzed all phase 2/3 trials for age disparity and trials recruiting from the US for racial/ethnic disparity amongst BC RCTs (1990-2021). Enrollment-incidence ratios (EIR), which compare trial proportions (TP) from age subgroups against global estimates of incidence in the corresponding age group (from Global Burden of Disease) and racial/ethnic subgroups against US-population-based incidence in the corresponding racial/ethnic group (from SEER 22), were calculated. Individual-trial EIRs were pooled using random-effects meta-analyses. EIRs of <0.80 indicated meaningful underrepresentation, while EIRs >1.20 indicated meaningful overrepresentation. Meta-regression was used to explore associations between key trial characteristics and EIR, and trends in EIR over the last three decades. Results: Of 842 global trials with 490,657 participants, 232 (27.6%) reported the number of participants in specific age strata and 144 (17.1%) reported outcomes by age. Only 60 trials recruited exclusively from the US. Of these, 41 (68.3%) reported race and 24 (40.0%) reported ethnicity, and 2 (3.3%) analyzed outcomes by race. Older adults aged ≥65 years (TP: 24.3%; EIR: 0.53 [95% CI: 0.47-0.60]) and Hispanic women (TP: 5.3%; 0.54; 0.40-0.75) were underrepresented across global trials and US-only trials respectively. Black women were significantly underrepresented in trials that led to FDA approval (0.38; 0.25-0.58) but not in US-only trials. Stratified analysis by BC subtype (with subtype-specific population incidence) showed underrepresentation of Black women in US-only trials for HER2+ BC (0.66; 0.47-0.93), Hispanic women in ER+ BC (0.42; 0.20-0.90), Asians in triple-negative BC (0.24; 0.11-0.53), and older women in HER2+ BC (0.35; 0.31-0.40) and triple-negative BC (0.36; 0.29-0.45). From 1990-2021, the representation of older women declined significantly (meta-regression coefficient for EIR: -0.4 per 10 years), while Black and Hispanic representation showed no significant change. Black (p = 0.02) and Hispanic representation (p = 0.05) worsened significantly in larger trials. Conclusions: Published reports of BC trials do not consistently report the inclusion of URM. Older and Hispanic women were consistently underrepresented, whereas Black women were underrepresented in large, practice-changing trials. [Table: see text]