Abstract Background: Diagnosing colorectal cancer in people below age 50 relies largely on the evaluation of symptoms despite recent recommendations to initiate screening at age 45. There is limited information on the positive predictive value (PPV) of symptoms as indicators of early-onset colorectal cancer. Methods: We identified patients aged 42 to 49 years in three community-based health systems between 2012 and 2020 whose first recorded colonoscopy had a diagnostic indication, with symptoms present in the two years before the colonoscopy. We computed the PPV for colorectal cancer by each individual symptom, combinations of the three most common symptoms, and combinations of the three symptoms with the highest PPVs and computed 95% confidence intervals (CI). We also evaluated the PPV of recent symptoms stratified according to the timing of onset and estimated the PPV of symptoms when followed by positive or negative fecal testing. Results: The study included 28,198 patients. The PPV for specific symptoms, with or without other symptoms, was 2.4% (95% CI, 2.2%–2.7%) for blood loss and 4.5% (95% CI, 3.3%–5.9%) for positive fecal testing. Pairwise and three-way combinations of blood loss, diarrhea, and abdominal mass had PPV point estimates for colorectal cancer above 3%. Conclusions: Our study suggests that specific symptoms and combinations may identify patients with a ≥3% prevalence of colorectal cancer in those aged 42 to 49 and that fecal testing results may further identify patients with a higher probability of colorectal cancer. Impact: Future research is needed to develop high PPV strategies for identifying early-onset colorectal cancer without compromising sensitivity. See related In the Spotlight, p. 1233

Supplementary Table S4 lists the 140 colorectal-cancer-associated loci and associations with colorectal cancer in European-ancestry population.

Supplementary Figure S6 shows the calibration on relative risk of PRS stratified by PRS with 7 bins in groups of different ancestry in the GERA cohort.

Supplementary Table S2 shows the descriptive statistics of GERA study participants by racial/ethnic groups and sex.

Supplementary Table S3 shows the comparison on characteristics between GECCO/CORECT study and GERA Europeans-ancestry participants

Supplementary Figure S7 shows the Sex-specific estimated baseline incidence rate of CRC based on SEER18 (2007-2015) CRC rate in European population.

Supplementary Table S7 shows the 10-year time-dependent AUC estimates of the PRS-enhanced model in the GERA European- ancestry participants.

Supplementary Table S1 shows the descriptive characteristics of study populations in GECCO and CORECT.

Background and Aims:Colorectal cancer (CRC) incidence in those under age 50 is increasing and minority populations are known to have worse CRC survival outcomes. Therefore, we evaluated 5-year CRC-specific survival by race and ethnicity among medically insured patients with early-onset CRC. Methods:This retrospective cohort study included Kaiser Permanente Northern California patients aged 18-49 years diagnosed with CRC between 2006 and 2019. Five-year CRC-specific survival by race and ethnicity was assessed using Kaplan-Meier survival analyses and unadjusted and adjusted Cox proportional hazards models. Results:Among 1620 patients, 50.3% were White, 21.6% Hispanic, 20.1% Asian or Pacific Islander, and 8.0% Black. Stage IV disease was found in 23.4% of White, 28.7% of Black, 30.1% of Asian or Pacific Islander, and 31.1% of Hispanic patients. Five-year CRC-specific survival probability estimates ranged from 74.4% in Hispanic patients to 79.9% in White patients with no statistically significant differences in unadjusted risk estimates. However, after adjusting for age, sex, comorbidities, and socioeconomic status measures, risk of death was higher in Hispanic vs White patients (hazard ratio: 1.46; 95% confidence interval: 1.08-1.97) but was attenuated (hazard ratio: 1.13; 95% confidence interval: 0.83-1.53) after further adjustment for stage at diagnosis. Conclusion:Among medically insured patients in a large integrated healthcare setting, Hispanic patients were more likely to be diagnosed with stage IV CRC and had the lowest 5-year CRC-specific survival probability compared to other race groups. Additional research is needed to identify factors contributing to the higher rate of late-stage disease diagnosis in Hispanic patients.

Supplementary methods show the details of study population, risk model development, calculation of the polygenic risk score, statistical analysis and funding of individual studies in GECCO and CORECT.