Ribonucleotide reductase M2 and the consolidation tumor ratio combined with age at diagnosis for predicting invasive lung adenocarcinoma

Introduction: To investigate the expression of ribonucleotide reductase M2 (RRM2) in lung adenocarcinoma (LUAD) patients with different degrees of infiltration and to construct a neural network prediction model based on RRM2, the consolidation tumor ratio (CTR), and age at diagnosis (AAD) for predicting invasive LUAD. The test results of the model are compared with the pathological outcomes of patients to verify the accuracy of the prediction model. Methods: RRM2 expression was examined in 100 LUAD tissues collected from the First Affiliated Hospital of Jinzhou Medical University between January 2019 and December 2021 and confirmed by pathology after radical lung cancer surgery. This was achieved via immunohistochemistry. The integrated optical density (IOD) of RRM2 in cancer tissue was analyzed via ImageJ FIJI software. The expression of RRM2 in 100 cases of LUAD with varying degrees of infiltration was analyzed. The cases were divided into a training set of 60 cases and a validation set of 40 cases, in accordance with a 6:4 ratio. A neural network prediction model was constructed based on RRM2, CTR, and AAD to jointly predict invasive adenocarcinoma (IAC), and the accuracy and diagnostic efficacy of the prediction model were verified. Results: The differences in RRM2 IOD, CTR, and AAD between the MIA group and the IAC group were statistically significant ( P <0.05). In the training set, the receiver operating characteristic(ROC) curve demonstrated that the neural network model, which combined the three indicators of RRM2, CTR, and AAD, predicted IAC with an AUC of 0.93, an accuracy of 0.78, a sensitivity of 0.87, and a specificity of 0.86. This value was greater than the AUC of the dual-indicator combination and single indicators. In the validation set, the neural network prediction model, which combines the three indicators, had an AUC of 0.85, an accuracy of 0.78, a sensitivity of 0.94, and a specificity of 0.78. This value was greater than the AUC of the dual-indicator combination. Conclusion: Compared with models that use single or dual indicators, the neural network model established by combining RRM2, CTR, and AAD has a greater AUC for predicting IAC, demonstrating superior diagnostic efficacy and accuracy.