Frontiers in Oncology (Nov 2022)

Using inflammatory indexes and clinical parameters to predict radiation esophagitis in patients with small-cell lung cancer undergoing chemoradiotherapy

  • Jianjian Qiu,
  • Dongmei Ke,
  • Hancui Lin,
  • Yilin Yu,
  • Qunhao Zheng,
  • Hui Li,
  • Hongying Zheng,
  • Lingyun Liu,
  • Jiancheng Li

DOI
https://doi.org/10.3389/fonc.2022.898653
Journal volume & issue
Vol. 12

Abstract

Read online

ObjectiveRadiation esophagitis (RE) is a common adverse effect in small cell lung cancer (SCLC) patients undergoing thoracic radiotherapy. We aim to develop a novel nomogram to predict the acute severe RE (grade≥2) receiving chemoradiation in SCLC patients.Materials and methodsthe risk factors were analyzed by logistic regression, and a nomogram was constructed based on multivariate analysis results. The clinical value of the model was evaluated using the area under the receiver operating curve (ROC) curve (AUC), calibration curves, and decision curve analysis (DCA). The correlations of inflammation indexes were assessed using Spearman correlation analysis.ResultsEighty-four of 187 patients (44.9%) developed grade ≥2 RE. Univariate analysis indicated that concurrent chemoradiotherapy (CCRT, p < 0.001), chemotherapy cycle (p = 0.097), system inflammation response index (SIRI, p = 0.048), prognostic-nutrition index (PNI, p = 0.073), platelets-lymphocyte radio (PLR, p = 0.026), platelets-albumin ratio (PAR, p = 0.029) were potential predictors of RE. In multivariate analysis, CCRT [p < 0.001; OR, 3.380; 95% CI, 1.767-6.465], SIRI (p = 0.047; OR, 0.436; 95% CI, 0.192-0.989), and PAR (p = 0.036; OR, 2.907; 95% CI, 1.071-7.891) were independent predictors of grade ≥2 RE. The AUC of nomogram was 0.702 (95% CI, 0.626-0.778), which was greater than each independent predictor (CCRT: 0.645; SIRI: 0.558; PAR: 0.559). Calibration curves showed high coherence between the predicted and actual observation RE, and DCA displayed satisfactory clinical utility.ConclusionIn this study, CCRT, SIRI, and PAR were independent predictors for RE (grade ≥2) in patients with SCLC receiving chemoradiotherapy. We developed and validated a predictive model through these factors. The developed nomogram with superior prediction ability can be used as a quantitative model to predict RE.

Keywords