Nomogram for predicting preoperative lymph node involvement in patients with invasive micropapillary carcinoma of breast: a SEER population-based study

Abstract

Background: Invasive micropapillary carcinoma (IMPC) is an unusual and distinct subtype of invasive breast tumor with high propensity for regional lymph node metastases. This study was to identify risk factors accounting for IMPC of the breast and to develop a nomogram to preoperatively predict the probability of lymph node involvement.

Methods: A retrospective review of the clinical and pathology records was performed in patients diagnosed with IMPC between 2003 and 2014 from Surveillance, Epidemiology, and End Results (SEER) database. The cohort was divided into training and validation sets. Training set comprised patients diagnosed between 2003 and 2009, while validation set included patients diagnosed thereafter. A logistic regression model was used to construct the nomogram in the training set and then varified in the validation set. Nomogram performance was quantified with respect to discrimination and calibration using R 3.4.1 software.

Results: Overall, 1407 patients diagnosed with IMPC were enrolled, of which 527 in training set and 880 in validation set. Logistic regression analysis indicated larger lesions, younger age at diagnosis, black ethnic and lack of hormone receptor expression were significantly related to regional nodes involvement. The AUC of the nomogram was 0.735 (95% confidential interval (CI) 0.692 to 0.777), demonstrating a good prediction performance. Calibration curve for the nomogram was plotted and the slope was close to 1, which demonstrated excellent calibration of the nomogram. The performance of the nomogram was further validated in the validation set, with AUC of 0.748 (95% CI 0.701 to 0.767).

Conclusions: The striking difference between IMPC and IDC remains the increased lymph node involvement in IMPC and therefore merits aggressive treatment. The nomogram based on the clinicalpathologic parameters was established, which could accurately preoperatively predict regional lymph node status. This nomogram would facilitate evaluating lymph node state preoperatively and thus treatment decision-making of individual patients.

Keywords: Breast cancer; IMPC; Lymph nodes involvement; Predict; Preoperative.

Fig. 1

Nomogram predicting lymph node involvement in patients with invasive micropapillary carcinoma of the breast. The top row shows the point assignment for each variable. Rows 2–8 indicate the variables included in the nomogram. For an individual patient, each variable is assigned a point value based on the tumor characteristics. The points assigned to each of the seven variables are summed, and the total points indicated in row 9. The bottom row shows the probability of the patient having lymph node metastasis. ER, estrogen receptor; PR, progesterone receptor

Fig. 2

Discrimination and calibration of the nomogram in primary and validation cohorts. a and c Receiver operating characteristic (ROC) curve for discrimination in the training and validation sets. The area under the curve (AUC) of the nomagram was 0.735 (95% CI 0.692 to 0.777) and 0.748 (95% CI 0.701 to 0.767) respectively, demonstrating very good prediction performance. b and d Calibration plot of actual (observed) and predicted probabilities for the nomogram in the training and validation sets. The x-axis represents the nomogram predicted probabilities as measured by logistic regression analysis, and the y-axis the actual probabilities. Vertical lines indicate the frequency distribution of predicted probabilities. The dotted line indicates the ideal reference line where predicted probabilities match the observed probabilities. The calibration curve indicated excellent calibration of the nomogram