Abstract A143: Divergent Subtype Trajectories in Pancreatic Cancer: Age-Period-Cohort and Joinpoint Analysis of Surveillance, Epidemiology, and End Results Program (SEER) Data
Douglas Monroe, Ana A. Best, Philip Rosenberg, Hormuzd A. Katki, Rachael A. Stolzenberg-SolomonAbstract
Background:
Pancreatic cancer is clinically heterogeneous, and aggregate incidence summaries may obscure subtype-specific patterns. We evaluated whether SEER-based incidence trends differed for pancreatic ductal adenocarcinoma (PDAC), pancreatic neuroendocrine tumors (PNET), and poorly specified tumors.
Methods:
We analyzed Surveillance, Epidemiology, and End Results Program (SEER) pancreatic cancer incidence data from 1992 to 2022 by histologic subtype, sex, and race/ethnicity. Age-period-cohort models characterized age, period, and cohort curvature, and joinpoint period models estimated annual percent change (EAPC). Primary tabular results are reported for all race/ethnicity groups, both sexes, and all stages.
Results:
In the all-race/ethnicity, both-sex, all-stage joinpoint period model, PDAC changed from a nonsignificant decline in 1992.5-2000.5 (EAPC, -0.61%/year; 95% CI, -1.30 to 0.09) to a modest increase in 2000.5-2021.5 (0.36%/year; 95% CI, 0.21 to 0.50). PNET showed higher estimated period increases than PDAC in the longer period segments (3.37%/year during 1992.5-2009.5; 95% CI, 2.10 to 4.65; and 2.52%/year during 2012.5-2021.5; 95% CI, 0.97 to 4.09). Poorly specified tumors increased in 1992.5-2007.5 (2.53%/year; 95% CI, 1.52 to 3.55) and flattened in 2007.5-2021.5 (0.16%/year; 95% CI, -0.61 to 0.93). APC global period curvature was positive for PDAC (0.0147%/yr^2; 95% CI, 0.0051 to 0.0242) and PNET (0.0835%/yr^2; 95% CI, 0.0367 to 0.1303) but negative for poorly specified tumors (-0.0686%/yr^2; 95% CI, -0.0976 to -0.0396).
Conclusions:
The model output tables support divergent subtype-specific incidence trajectories rather than a single aggregate pancreatic cancer pattern. PDAC shows a modest recent increase, PNET shows larger estimated period increases, and poorly specified tumors flatten after 2007.5. These findings support subtype-specific surveillance and cautious interpretation of aggregate pancreatic cancer incidence trends.
Citation Format:
Douglas Monroe, Ana A. Best, Philip Rosenberg, Hormuzd A. Katki, Rachael A. Stolzenberg-Solomon. Divergent Subtype Trajectories in Pancreatic Cancer: Age-Period-Cohort and Joinpoint Analysis of Surveillance, Epidemiology, and End Results Program (SEER) Data [abstract]. In: Proceedings of the AACR Conference on Pancreatic Cancer: New Frontiers in Biology and Therapeutic Development; 2026 Sep 25-28; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(18_Suppl_2):Abstract nr A143.