Although guidelines exist for the use of several clinical prediction models based on logistic regression to help estimate the risk of lung cancer before treatment decision-making, they almost all focus on solitary pulmonary nodules, and have not shown accuracy in predicting malignancy in multiple pulmonary nodules. A study by Chen et al that sought to develop a Web-based machine learning model to predict the risk of malignancy among patients presenting with multiple pulmonary nodules has found that the model performed better than radiologists, surgeons, validated mathematical models, and a previously established artificial intelligence tool. According to the study authors, the prediction model can be used as a convenient tool for precision diagnosis of multiple pulmonary nodules before surgical treatment, thereby decreasing unnecessary invasive procedures. The study is published in Clinical Cancer Research.
The researchers developed an ensemble algorithm (XGBoost) to predict malignancy using the clinicoradiologic variables of 1,739 nodules from a cohort of 520 patients with multiple pulmonary nodules who were treated at Peking University People’s Hospital in China between 2007 and 2018. The model, called PKU-M, was trained using 10-fold cross-validation in which hyperparameters were selected and fine-tuned. The model was then validated and compared with solitary pulmonary nodule models, clinicians, and a computer-aided diagnosis (CADx) system in an independent transnational cohort and prospective multicentric cohort.
According to the study findings, the PKU-M model showed excellent discrimination (area under the curve [AUC] = 0.909, 95% confidence interval [CI] = 0.854–0.946) and calibration (Brier score = 0.122) in the development cohort. External validation (583 nodules) revealed that the AUC of the PKU-M model was 0.890 (95% CI = 0.859–0.916), higher than those of the Brock model (AUC = 0.806, 95% CI = 0.771–0.838), PKU model (AUC = 0.780, 95% CI = 0.743–0.817), Mayo model (AUC = 0.739, 95% CI = 0.697–0.776), and VA model (AUC = 0.682, 95% CI = 0.640–0.722).
Prospective comparison (200 nodules) showed that the AUC of the PKU-M model (AUC = 0.871, 95% CI = 0.815–0.915) was higher than that of surgeons (AUC = 0.790, 95% CI = 0.711–0.852; AUC = 0.741, 95% CI = 0.662–0.804; and AUC = 0.727 95% CI = 0650–0.788), radiologist (AUC = 0.748, 95% CI = 0.671–0.814), and CADx system (AUC = 0.757, 95% CI = (0.682–0.818). In addition, the model outperformed the clinicians with an increase of 14.3% in sensitivity and 7.8% in specificity.
“After its development using machine learning algorithms, validation using transnational multicentric cohorts, and prospective comparison with clinicians and the CADx system, this novel prediction model for multiple pulmonary nodules presented solid performance as a convenient reference to help decision-making,” concluded the study authors.
“The increasing detection rate of multiple pulmonary nodules has led to an emerging problem for lung cancer diagnosis,” said Young Tae Kim, MD, PhD, Professor in the Department of Thoracic and Cardiovascular Surgery at Seoul National University Hospital and the Seoul National University College of Medicine in the Republic of Korea, and a coauthor of this study, in a statement. “Because many nodules are found to be benign either after long-term follow-up or surgery, it is important to carefully evaluate these nodules prior to invasive procedures. Our prediction model, which was exclusively established for patients with multiple nodules, can help not only mitigate unnecessary surgery, but also facilitate the diagnosis and treatment of lung cancer.”
Jun Wang, MD, Peking University People’s Hospital, Beijing, is the corresponding author of this study.
Disclosures: Funding for this study was provided by the National Natural Science Foundation of China and Peking University People’s Hospital. For full disclosures of the study authors, visit clincancerres.aacrjournals.org.The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.