B-mode Ultrasound Related Combined Predictive Indicators for Benign Versus Malignant Subpleural Pulmonary Lesions.

Objective: The purpose of this study was to analyze the independent risk factors of malignant subpleural pulmonary lesions (SPLs) on B-mode ultrasound (US) images, to construct the combined predictive indicators, and to prospectively verify their predictive efficacy.

Methods: A total of 336 patients with SPLs were included in the prospective study, of whom the single-center included patients between September 2019 and December 2019 were the development cohort (DC) (n = 219); Patients who were concurrently enrolled in three centers between January and February 2020 were the validation cohort (VC) (n = 117). The clinical features and B-mode US parameters were collected. Based on the DC, a combined predictive indicators model was developed using binary logistic regression. Then the discrimination was verified externally in the VC. The reference criteria were from the comprehensive diagnosis of clinical-radiological-pathological made by two senior respiratory physicians.

Results: The combined predictive indicators model was finally constructed by five parameters: age, borderline, angle between the lesion border and thoracic wall, posterior echo of the lesion and invasion of the pleura. The fitting degree of the model was good (χ = 9.198, p = 0.326). The area under ROC curve of the model was 0.872 (DC) and 0.808 (VC), yielding a higher net benefit than individual risk factors.

Conclusion: The combined predictive indicators are useful in the assessment of malignant SPLs and are a useful adjunct diagnostic tool, especially in primary healthcare settings in developing countries.