CARWL score as a predictor of radiation-induced periodontitis in locally advanced head and neck cancer undergoing concurrent chemoradiotherapy

Résumé fourni par la source

Although concurrent chemoradiotherapy (CCRT) has improved outcomes in locally advanced head and neck cancer (LA-HNC), radiation-induced periodontitis (RIP) remains an under-recognized oral toxicity with significant consequences, including tooth loss and osteoradionecrosis. This study evaluates the utility of the novel CARWL score-a combined index of the C-reactive protein-to-albumin ratio (CAR) and significant weight loss (SWL)-for stratifying the risk of RIP in LA-HNC patients without baseline periodontitis undergoing CCRT. We conducted a retrospective analysis of 67 LA-HNC patients who underwent CCRT and received detailed oral examinations before and after treatment; none had periodontitis at the initiation of CCRT. Receiver operating characteristic (ROC) curve analysis identified an optimal pretreatment CAR cutoff of 3.07, with SWL defined as greater than 5% body weight loss in the preceding six months. Based on CAR (≥3.07 vs. <3.07) and SWL (present vs. absent), patients were categorized into three CARWL groups. The primary endpoint was the association between the baseline CARWL group and the rates of RIP following CCRT. RIP was diagnosed in 17 patients (25.4%) during follow-up, with incidences increasing progressively across CARWL-0, CARWL-1, and CARWL-2 groups (11.8% vs. 20.8% vs. 38.5%; p = 0.007). In multivariable Cox proportional-hazards analysis, a higher CARWL score emerged as an independent predictor of increased RIP risk (adjusted HR = 3.64; 95% CI 1.41-9.37; p = 0.007), and supplementary logistic regression sensitivity analysis corroborated these findings (adjusted OR = 3.58; 95% CI 1.35-9.45). These findings demonstrate that the pretreatment CARWL score serves as a straightforward and readily available biomarker that effectively stratifies the risk of radiation-induced periodontitis in LA-HNC patients treated with CCRT.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.