Hazard Ratios (95% CI) of Overall Mortality by Comparing Early-onset and Later-onset CRC

AgeUnadjustedAdjusted for Stage OnlyFully Adjusted except StageFully Adjusted
<50 1.04 (1.02-1.05) 0.89 (0.88-0.90) 1.02 (1.00-1.03) 0.95 (0.93-0.96) 
51 - 55 1.00 1.00 1.00 1.00 
<20 0.83 (0.71-0.97) 1.08 (0.93-1.26) 1.05 (0.90-1.23) 1.04 (0.88-1.22) 
20 - 24 1.04 (0.95-1.14) 1.08 (0.99-1.18) 1.05 (0.96-1.16) 1.07 (0.97-1.17) 
25 - 29 1.07 (1.01-1.14) 0.93 (0.88-0.99) 1.01 (0.94-1.08) 0.94 (0.88-1.00) 
30 - 34 1.05 (1.00-1.09) 0.85 (0.82-0.89) 1.00 (0.95-1.06) 0.90 (0.85-0.95) 
35 - 39 0.99 (0.96-1.03) 0.83 (0.80-0.85) 0.97 (0.93-1.02) 0.88 (0.84-0.92) 
40 - 44 1.02 (1.00-1.05) 0.88 (0.86-0.90) 1.00 (0.98-1.03) 0.94 (0.92-0.96) 
45 - 49 1.05 (1.03-1.07) 0.91 (0.90-0.93) 1.02 (1.00-1.04) 0.96 (0.94-0.98) 
51 - 55 1.00 1.00 1.00 1.00 
AgeUnadjustedAdjusted for Stage OnlyFully Adjusted except StageFully Adjusted
<50 1.04 (1.02-1.05) 0.89 (0.88-0.90) 1.02 (1.00-1.03) 0.95 (0.93-0.96) 
51 - 55 1.00 1.00 1.00 1.00 
<20 0.83 (0.71-0.97) 1.08 (0.93-1.26) 1.05 (0.90-1.23) 1.04 (0.88-1.22) 
20 - 24 1.04 (0.95-1.14) 1.08 (0.99-1.18) 1.05 (0.96-1.16) 1.07 (0.97-1.17) 
25 - 29 1.07 (1.01-1.14) 0.93 (0.88-0.99) 1.01 (0.94-1.08) 0.94 (0.88-1.00) 
30 - 34 1.05 (1.00-1.09) 0.85 (0.82-0.89) 1.00 (0.95-1.06) 0.90 (0.85-0.95) 
35 - 39 0.99 (0.96-1.03) 0.83 (0.80-0.85) 0.97 (0.93-1.02) 0.88 (0.84-0.92) 
40 - 44 1.02 (1.00-1.05) 0.88 (0.86-0.90) 1.00 (0.98-1.03) 0.94 (0.92-0.96) 
45 - 49 1.05 (1.03-1.07) 0.91 (0.90-0.93) 1.02 (1.00-1.04) 0.96 (0.94-0.98) 
51 - 55 1.00 1.00 1.00 1.00 

Despite significantly reduced colorectal cancer (CRC) incidence in Americans aged 50 and older since 2000, the incidence of CRC among those <50 has been steadily rising in the U.S. Understanding survival of early-onset CRC (EOCRC; <50 years), relative to those over 50 years, is fundamental to informing treatment approaches and understanding unique biological distinctiveness within EOCRC. To inform these questions, we established a large, nationwide, registry-based CRC cohort (N=769,871) from the National Cancer Database between January 1, 2004 through December 31, 2015. After initial analyses, to avoid possible screening detection bias at age 50 in our population (0-90 years), which was marked with a remarkable proportion of earlier stage and the lowest hazard ratio (HR) in all ages, patients diagnosed at age 51-55 were selected as the comparison group. Overall survival was assessed by the Kaplan-Meier method and Cox proportional hazards regression. Of 769,871 CRC patients (49% women; 83% white), 46% died over a median follow-up of 2.9 years, and 14% were EOCRC. Compared with CRC diagnosed at ages 51-55 years, EOCRC patients had a modestly lower 10-year survival rate (53.6% [95% CI, 53.2%-54.0%] vs. 54.3% [95% CI, 53.8%-54.8%]) in unadjusted analysis. However, after adjusting for stage at diagnosis, EOCRC patients had better survival compared to subjects diagnosed at ages 51-55 (adjusted HR for mortality: 0.95 [95% CI, 0.93-0.96]). The survival advantage appeared greatest for patients diagnosed at stages I-II and ages 35-39. In conclusion, after adjusting for stage at diagnosis, EOCRC is associated with a modestly improved survival when compared to subjects who presented between ages 51-55 years. This reinforces the importance of early CRC detection in the younger population. Further study is needed to understand the underlying heterogeneity of survival within early-onset patients by stage and age.

Citation Format: En Cheng, Holly N. Blackburn, Kimmie Ng, Donna Spiegelman, Melinda L. Irwin, Xiaomei Ma, Cary P. Gross, Fred K. Tabung, Edward L. Giovannucci, Pamela L. Kunz, Xavier Llor, Kevin Billingsley, Jeffrey A. Meyerhardt, Nita Ahuja, Charles S. Fuchs. Survival for patients with early-onset colorectal cancer - An overall survival analysis from the National Cancer Database, 2004-2015 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 898.