Predicting Lymph Node Metastases in Patients with Biopsy-Proven Ductal Carcinoma In Situ of the Breast: Development and Validation of the DCIS-met Model

Abstract

Abstract Purpose In patients with a biopsy-proven ductal carcinoma in situ (DCIS), axillary staging is frequently performed, but in hindsight often turns out to be superfluous. The aim of this observational study was to develop a prediction model for risk of lymph node metastasis in patients with a biopsy-proven DCIS. Methods Data were received from the Dutch Pathology Databank and the Netherlands Cancer Registry. The population-based cohort consisted of all biopsy-proven DCIS patients diagnosed in the Netherlands in 2011 and 2012. The prediction model was evaluated with the area under the curve (AUC) of the receiver operating characteristic, and a calibration plot and a decision curve analysis and was validated in a Dutch cohort of patients diagnosed in the period 2016–2019. Results Of 2892 biopsy-proven DCIS patients, 127 had metastasis (4.4%). Risk factors were younger age (OR = 0.97, 95% CI 0.95–0.99), DCIS not detected by screening (OR = 1.55, 95% CI 1.01–2.38), suspected invasive component at biopsy (OR = 1.86, 95% CI 1.01–3.41), palpable tumour (OR = 2.06, 95% CI 1.34–3.18), BI-RADS score 5 (OR = 2.41, 95% CI 1.53–3.78), intermediate-grade DCIS (OR = 3.01, 95% CI 1.27–7.15) and high-grade DCIS (OR = 3.20, 95% CI 1.36–7.54). For 24% (n = 708) of the patients, the predicted risk of lymph node metastasis was above 5%. Based on the decision curve analysis, the model had a net benefit for a predicted risk below 25%. The AUC was 0.745. Of the 2269 patients in the validation cohort, 53 (2.2%) had metastasis and the AUC was 0.741. Conclusions This DCIS-met model can support clinical decisions on axillary staging in patients with biopsy-proven DCIS.