Competing-Risk Nomograms for Predicting the Prognosis of Patients With Infiltrating Lobular Carcinoma of the Breast

Abstract

BackgroundInfiltrating lobular carcinoma (ILC) is the second most common histologic subtype of breast cancer. We assessed the rates of cause-specific death in ILC patients with the aim of establishing competing-risk nomograms for predicting their prognosis. Patients and MethodsData on ILC patients were extracted from the Surveillance, Epidemiology, and End Results (SEER) database. The cumulative incidence function was used to calculate the cumulative incidence rates of cause-specific death, and Gray's test was applied to test the differences in cumulative incidence rates among groups. We then identified independent prognostic factors by applying the Fine–Gray proportional subdistribution hazard analysis method and established nomograms based on the results. Calibration curves and the concordance index were employed to validate the nomograms. ResultsThe study enrolled 11,361 patients. The 3-, 5-, and 10-year overall cumulative incidence rates for those who died of ILC were 3.1%, 6.2%, and 12.2%, respectively, whereas the rates for those who died from other causes were 3.2%, 5.8%, and 14.1%. Age, marriage, grade, size, regional node positivity, American Joint Committee on Cancer M stage, progesterone receptor, and surgery were independent prognostic factors for dying of ILC, whereas the independent prognostic factors for dying of other causes were age, race, marriage, size, radiation, and chemotherapy. The nomograms were well calibrated and had good discrimination ability. ConclusionWe applied competing-risk analysis to ILC patients based on the SEER database and established nomograms that perform well in predicting the cause-specific death rates at 3, 5, and 10 years after the diagnosis.