Background:

The incidence of colorectal cancer (CRC) in younger patients (pts) is rising. This increase is most pronounced in tumors arising from the distal colon and rectum. Since tumor sidedness has emerged as an important prognostic and predictive biomarker in CRC, we aim to explore the impact of age on the tumor biology of left-sided colorectal cancer (LCC), defined as CRC from the splenic flexure to (and including) the rectum. Herein, we compare profiles of LCC from younger (≤ 45 years) and older pts (≥ 65 years).

Methods:

LCCs (n = 1,602) were examined by NextGen sequencing (NGS), protein expression, gene amplification, and microsatellite instability (MSI) fragment analyses. Tumor mutational load (TML) was calculated using only somatic non-synonymous missense mutations. Chi-square tests were used for comparisons. Besides comparing younger (≤ 45 years) and older (≥ 65 years) groups, a sub-analysis was also done comparing three rectal cancer groups based on age (≤ 30, 40-49, and ≥ 50 years).

Results:

LCCs from younger (median age 40, range 22-45 years, n = 229) and older (median age 71, range 65-89, n = 503) pts were studied. The most frequently mutated genes included APC, TP53, KRAS, PIK3CA, ARID1A, FBXW7, SMAD4, ATM, BRAF, and NRAS. Comparing younger v. older pts, there were no significant differences in the rates of APC (75.3% v. 82.9%, P = 0.139), TP53 (79.5% v. 73.1%, P = 0.261), KRAS (37.6% v. 43.0%, P = 0.403), PIK3CA (9.4% v. 14.6%), ARID1A (14.3% v. 13.2%, P = 0.884), FBXW7 (11.4% v. 10.5%, P = 0.830), SMAD4 (13.1% v. 7.4%, P = 0.129), BRAF (4.8% v. 5.7%, P = 0.762), or NRAS (3.5% v. 2.6%, P = 0.680) mutations. Additionally there were no significant differences in protein overexpression. However, there was a trend towards increased HER2 amplification in younger pts (5.7% v. 2.1%, P = 0.05). MSH6 (4.8% v. 0.5%, P = 0.015), MSH2 (2.4% v. 0%, P = 0.032), POLE (2.4% v. 0%, P = 0.032), and NF1 (7.9% v. 0%, P < 0.001) mutations were observed at higher rates in younger pts. High TML (≥ 17 mutations per megabase) was seen more frequently in younger pts (8.2% v. 2.6%, P = 0.02).

Conclusion:

The molecular differences between LCC in younger and older pts are mostly due to mutations in mismatch repair genes. Higher TML may predict a higher response rate to checkpoint inhibitors in younger pts with LCC. The differences in tumor biology observed here warrant further study and may eventually be used to tailor therapy.