Identification of a Subset of Stage I Colorectal Cancer Patients With High Recurrence Risk

• Published in: JNCI : Journal of the National Cancer Institute 2022-05, Vol.114 (5), p.732-739
• Main Authors: Lee, Lik Hang, Davis, Lindy, Ylagan, Lourdes, Omilian, Angela R, Attwood, Kristopher, Firat, Canan, Shia, Jinru, Paty, Philip B, Cance, William G
• Format: Article
• Language: English
• Subjects: Biomarkers; Cancer; Colon; Colorectal cancer; Colorectal carcinoma; Colorectal Neoplasms - pathology; Confidence intervals; Focal adhesion kinase; Health hazards; Humans; Immunohistochemistry; Kinases; Metastases; Neoplasm Staging; Prognosis; Proportional Hazards Models; Protein-tyrosine kinase; Statistical analysis; Statistical significance; Statistical tests; Survival; Survival analysis; Tyrosine
• ISSN: 0027-8874; 1460-2105
• DOI: 10.1093/jnci/djac023

Abstract Background A challenge in early-stage colorectal cancer (CRC) is identifying biomarkers that predict an increased risk for recurrence. A potential clinically adaptable biomarker is focal adhesion kinase (FAK), a tyrosine kinase that promotes invasion and metastasis. Methods An initial, single-institution, 298-patient cohort with all stages of CRC and long-term follow-up was assessed for FAK with tissue microarrays using immunohistochemistry. FAK expression was scored and dichotomized into high and low. Subsequently, a validation cohort of 517 early-stage CRCs from a separate institution was evaluated. All statistical tests were 2-sided. Results FAK overexpression did not correlate with any known histologic feature and was an early event in CRC, increasing from normal colon to stage I, and stage I to II, but not different at higher stages. High FAK was associated with decreased 10-year recurrence-free survival (RFS) among stage I patients (70.2% for high FAK vs 94.1% for low, P = .02), but not among higher stages in the initial cohort. The same finding was seen in the validation cohort (73.1% for high FAK vs 93.1% for low, P = .004). Multivariable survival analysis for stage I patients showed only two statistically significant factors predicting RFS: FAK (hazard ratio = 5.27, 95% confidence interval = 1.81 to 15.33, P = .002) and perineural invasion (hazard ratio = 7.38, 95% confidence interval = 1.01 to 53.96, P = .049). FAK was the only statistically significant factor in multivariable analysis across RFS, overall, and disease-specific survivals. Conclusions High FAK expression identified a subset of stage I CRC patients with high incidence of recurrence and reduced survival, suggesting that FAK has important prognostic value. These patients would immediately benefit from more rigorous surveillance protocols for recurrent disease.