AUTHOR=Chen Jindong , Cao Shuhui , Jin Yu , Rong Wenwen , Wang Hao , Xi Siqi , Gan Tian , He Ben , Zhong Hua , Zhao Liang TITLE=Construction and validation of a nomogram of risk factors for new-onset atrial fibrillation in advanced lung cancer patients after non-surgical therapy JOURNAL=Frontiers in Oncology VOLUME=13 YEAR=2023 URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2023.1125592 DOI=10.3389/fonc.2023.1125592 ISSN=2234-943X ABSTRACT=Objective

Risk factors of new-onset atrial fibrillation (NOAF) in advanced lung cancer patients are not well defined. We aim to construct and validate a nomogram model between NOAF and advanced lung cancer.

Methods

We retrospectively enrolled 19484 patients with Stage III-IV lung cancer undergoing first-line antitumor therapy in Shanghai Chest Hospital between January 2016 and December 2020 (15837 in training set, and 3647 in testing set). Patients with pre-existing AF, valvular heart disease, cardiomyopathy were excluded. Logistic regression analysis and propensity score matching (PSM) were performed to identify predictors of NOAF, and nomogram model was constructed and validated.

Results

A total of 1089 patients were included in this study (807 in the training set, and 282 in the testing set). Multivariate logistic regression analysis showed that age, c-reactive protein, centric pulmonary carcinoma, and pericardial effusion were independent risk factors, the last two of which were important independent risk factors as confirmed by PSM analysis. Nomogram included independent risk factors of age, c-reactive protein, centric pulmonary carcinoma, and pericardial effusion. The AUC was 0.716 (95% CI 0.661–0.770) and further evaluation of this model showed that the C-index was 0.716, while the bias-corrected C-index after internal validation was 0.748 in the training set. The calibration curves presented good concordance between the predicted and actual outcomes.

Conclusion

Centric pulmonary carcinoma and pericardial effusion were important independent risk factors for NOAF besides common ones in advanced lung cancer patients. Furthermore, the new nomogram model contributed to the prediction of NOAF.