Risk profiles and a concise prediction model for lymph node metastasis in patients with lung adenocarcinoma

Abstract

Abstract Background Lung cancer is the second most commonly diagnosed cancer and ranks the first in mortality. Pathological lymph node status(pN) of lung cancer affects the treatment strategy after surgery while systematic lymph node dissection(SLND) is always unsatisfied. Methods We reviewed the clinicopathological features of 2,696 patients with LUAD and one single lesion ≤ 5 cm who underwent SLND in addition to lung resection at the Sun Yat-Sen University Cancer Center. The relationship between the pN status and all other clinicopathological features was assessed. All participants were stochastically divided into development and validation cohorts; the former was used to establish a logistic regression model based on selected factors from stepwise backward algorithm to predict pN status. C-statistics, accuracy, sensitivity, and specificity were calculated for both cohorts to test the model performance. Results Nerve tract infiltration (NTI), visceral pleural infiltration (PI), lymphovascular infiltration (LVI), right upper lobe (RUL), low differentiated component, tumor size, micropapillary component, lepidic component, and micropapillary predominance were included in the final model. Model performance in the development and validation cohorts was as follows: 0.861 (95% CI: 0.842–0.883) and 0.840 (95% CI: 0.804–0.876) for the C-statistics and 0.803 (95% CI: 0.784–0.821) and 0.785 (95% CI: 0.755–0.814) for accuracy, and 0.754 (95% CI: 0.706–0.798) and 0.686 (95% CI: 0.607–0.757) for sensitivity and 0.814 (95% CI: 0.794–0.833) and 0.811 (95% CI: 0.778–0.841) for specificity, respectively. Conclusion Our study showed an easy and credible tool with good performance in predicting pN in patients with LUAD with a single tumor ≤ 5.0 cm without SLND and it is valuable to adjust the treatment strategy.

Keywords

• Concise prediction model
• Lymph node metastasis
• Lung adenocarcinoma
• Specimen features