Non-random patterns of multiple primary malignancies in the era of cancer survival.

Abstract

10584 Background: Improved cancer therapies have increased survival, placing survivors at risk for subsequent primary cancers, yet the epidemiology of multiple primary cancers (MPC) is poorly defined. Methods: Using the VA Cancer Registry (2003–2023), we quantified MPC frequency, classified cancers by SEER ICD-O-3 codes and histology, and analyzed incidence and expected versus observed occurrence. Results: Among 822,564 veterans with cancer, 11.1% (n=91,279) developed ≥2 primary cancers. The frequencies of patients with 2, 3, 4, or ≥5 MPC were 81715 (9.9%), 8655 (1.1%), 800 (0.1%), and 109 (0.01%), with similar relative risks across successive cancers. Interestingly, the median age at first cancer was similar across groups (67.3 years for 1 cancer; 66.6, 65.8, and 65 years for 2, 3 and 4 cancers). In contrast, intervals between successive cancers progressively shortened (2.7 years – 2 cancers; 2.5 and 1.8 years – 3 cancers; 2.2, 1.9, and 1.4 years – 4 cancers), suggesting prior cancer influences the risk of subsequent cancer. The 7 most common first and subsequent cancers were prostate, bladder, lung, colon, melanoma, kidney, and oral cavity cancers. We identified three etiologic patterns underlying MPC. First, exposure-related cancers, clustered as expected, lung cancer was the most common 2 nd and 3 rd cancer after bladder (n=1850 and 231) and oral cavity (n=1581 and 148) cancers, suggesting a relationship to smoking. Second, genetically predisposed cancers did not cluster, likely reflecting the older population. Third, therapy-related cancers increased, with acute myeloid leukemia (AML)/acute lymphoblastic leukemia (ALL) standard rates rising from 0.02/0.003 (1 st ) to 39.01/3.19 (AML) and 5.04/0.03 (ALL) as 2 nd and 3 rd primary cancers. Interestingly, the frequency of prostate cancer and chronic lymphocytic leukemia (CLL) declined with increasing cancer order, consistent with earlier detection through established screening practices, and in contrast, lung cancer and multiple myeloma (MM) frequency increased as subsequent cancers, suggesting more complex etiologies, including aging-related risk. Conclusions: In this large VA registry analysis, over 11% of cancer survivors developed MPC with consistent patterns of occurrence, shortening inter-cancer intervals, and non-random clustering by exposure-, therapy-, and disease-specific factors. These findings underscore the need for risk-adapted surveillance and long-term survivorship strategies that account for shared etiologies and treatment-related risks rather than assuming stochastic occurrence. Frequency of a cancer as MPC. 1 st primary 2 nd primary 3 rd primary 4 th primary Pts with 2 cancers Prostate 29.4% 14.1% CLL 2.3% 1.1% Lung 7.8% 20% MM 1.1% 1.7% Pts with 3 cancers Prostate 28.1% 15.6% 9.3% CLL 2.8% 2.2% 1% Lung 5.4% 11.7% 20.6% MM 0.8% 1.5% 2% Pts with 4 cancers Prostate 26% 14.8% 12.1% 7.3% CLL 2.4% 1.8% 1.9% 1.4% Lung 5.4% 8.4% 13.9% 22.9% MM 1% 1.1% 1.8% 2.1%