Evaluating DNA damage in peripheral blood lymphocytes: a promising biomarker for diagnosis, prognosis, and treatment monitoring in colorectal cancer.

Colorectal cancer (CRC) remains a major health challenge due to its late-stage diagnosis and the variability in patient prognosis. This study explores the potential of DNA damage in peripheral blood lymphocytes (PBLs) as a biomarker for CRC, comparing it with standard clinical parameters. We assessed DNA strand breaks using the alkaline comet assay in 27 CRC patients at diagnosis and posttreatment, comparing these levels with 31 healthy controls. Patients received 5-fluorouracil (5-FU)-based chemotherapy (with irinotecan or oxaliplatin), radiotherapy, or combined chemoradiotherapy. At diagnosis (t0), DNA damage in PBLs was significantly higher in CRC compared to healthy controls (mean ± SD %DNA in tail: CRC 27.9 ± 14.0%; controls 6.5 ± 3.8%; P = .001), and independently of common confounding factors (sex, age, smoking, and alcohol consumption). Crucially, the prognostic signal came from baseline (t0): 6 of 27 patients relapsed/metastasized within 8-10 months, and high-DNA-damage basal levels were the only significant prognostic predictor (P = .0137), yielding an infinitely elevated odds ratio (95% CI: ≥2.203) and 100% sensitivity. In stark contrast, carcinoembryonic antigen (CEA) and cancer antigen 19-9 (CA19-9) showed limited performance. At t0, among patients with available serum data (n = 23), most values were below clinical cut-offs: CEA 3 ng/ml (14/23, 61%); CA19-9 37 U/ml (19/23, 83%). Prognostic sensitivities were 50.0% (CEA) and 16.7% (CA19-9). Posttreatment (t1) increases in DNA damage are pharmacodynamically expected with DNA-damaging therapy. t1 values were higher in patients who relapsed (P < .001), whereas the within-patient change (Δ = t1 - t0) was not associated with outcome (P = .148); these posttreatment findings are exploratory. Evaluating DNA damage in PBLs, therefore, offers a valuable noninvasive biomarker for early detection, treatment monitoring, and short-term risk stratification in CRC, warranting validation in larger, stage-balanced cohorts.