Nomogram to Predict Internal Mammary Lymph Nodes Metastasis in Patients With Breast Cancer

Abstract

Background: Numerous studies have showed that internal mammary lymph node (IMLN) metastasis is an important adverse prognostic factor in patients with breast cancer (BC), however, there are no available prediction model for the preoperative diagnosis of IMLN metastasis. Methods: Data from 102 breast cancer patients treated with IMLN operation were used to establish and calibrate a nomogram for IMLN status based on multivariate logistic regression. Prediction performance of this model was further validated with a second set of 50 patients with BC. Discrimination of the predict model was assessed by the C-index, and calibration assessed by calibration plots. Moreover, we conducted the decision curve analysis (DCA) to evaluate the clinical value of the nomogram. Finally, the survival status of patients in different risk groups based on nomogram were also compared. Results: The final multivariate regression model included tumor location, lymph vascular invasion (LVI), and pathological axillary lymph node stage (pALN stage). A nomogram was developed as a graphical representation of the model and had good calibration and discrimination in both sets (with C-index of 0.86 and 0.83 for the training and validation set, respectively). Moreover, the DCA showed the clinical usefulness of our constructed nomogram. False negative (FN) in low risk group classified by nomogram (FN-LR-nomogram) did not significantly impact adjuvant treatment decision making, and more importantly, patients with FN-LR-nomogram had recurrence-free survival equivalent to patients with pathologically ture negative in low risk group classified by nomogram (TN-LR-nomogram). Conclusions: As a non-invasive prediction tool, our nomogram shows favorable predictive accuracy for IMLN metastasis in patients with BC and can serve as a basis to integrate future molecular markers for its clinical application.

Keywords: IMLN metastasis; breast cancer; nomogram; pALN stage; recurrence-free survival.

Figure 1

Schematic results of constructed nomogram and final IMLN pathology.

Figure 2

Proposed nomogram to predict the probability of IMLN metastasis after mastectomy in patients with breast cancer. (A) Nomogram was build to predict IMLN status for BC patients, with the tumor location, LVI and pALN stage incorporated. Calibration plots for nomogram model in (B) training cohort and (C) validation cohort. The dashed line (the 45-degree line) represents a perfect prediction nomogram, and the black solid line represents the observed nomogram, of which a closer fit to the dashed line means a better prediction model. Plots (D) and (E) show the ROC curves of the constructed nomogram in the training and validation cohorts, respectively. UIQ, upper inner quadrant; LIQ, lower inner quadrant; UOQ, upper outer quadrant; LOQ, lower outer quadrant.

Figure 3

Decision curve analysis comparing the net-benefit of using the nomogram (black dashed line) depicted in (A) training cohort and (B) validation cohort. Black solid line: net benefit when all breast cancer patients are considered as not having the IMLN metastasis; gray solid line: net benefit when all breast cancer patients are considered as having the event. The ideal model is the model with the highest net benefit at any given threshold.

Figure 4

Recurrence-free survival for low risk patients with true negative and false negative groups and for high risk patients with true positive and false positive IMLN groups. Recurrence-free survival for low risk patients (A) with true negative and false negative groups and for high risk patients (B) with true positive and false positive IMLN groups.