Early biomarker trajectories after stage-linked surgery for p16-negative oral squamous cell carcinoma
摘要
To describe early postoperative laboratory trajectories across two predefined stage-linked surgical pathways in p16-negative oral squamous cell carcinoma (OSCC). This single-center retrospective cohort study (2017–2024) included consecutive adults undergoing ablative OSCC surgery with neck dissection. Patients followed predefined pathways: T1/primary closure pathway (n = 35) or T3/T4a/anterolateral thigh (ALT) free-flap pathway (n = 36). For each analyte and patient, a single summary value was obtained at T0 (preoperative), T1 (postoperative day 1), T2 (days 2–6), and T3 (days 7–14). Trajectories were compared using random-intercept linear mixed-effects models. Because only 15 surgical site infections occurred, infection analyses were descriptive and unadjusted. No prediction model was developed. The T3/T4a/ALT pathway had higher preoperative C-reactive protein (CRP) levels (median 8.77 vs. 3.11 mg/L; p = 0.009) and a higher crude frequency of surgical site infection (33.3% vs. 8.6%). Albumin trajectories differed between pathways (time-by-pathway p < 0.001); CRP was higher overall in the ALT pathway (p < 0.001), whereas the CRP time-by-pathway interaction was not significant (p = 0.346). A patient-clustered covariance sensitivity analysis detected a CRP time-by-pathway interaction (p = 0.006), indicating that inference about CRP trajectory shape was sensitive to covariance specification. The pathways showed distinct albumin trajectories, different overall CRP levels, and crude morbidity profiles. These findings characterize laboratory patterns within selected institutional pathways and support multicenter studies of pathway-specific perioperative monitoring. Routine perioperative laboratory monitoring may help contextualize recovery in selected OSCC pathways, but formal risk stratification requires prospective validation.