Adjuvant chemotherapy in patients with invasive lobular carcinoma and use of the 21‐gene recurrence score: A National Cancer Database analysis

Abstract

BACKGROUND: Invasive lobular carcinoma (ILC) is traditionally considered less responsive to chemotherapy. Although the Oncotype recurrence score (RS) has been validated to identify high-risk patients who benefit from chemotherapy, some studies have questioned its relevance in patients with ILC. The objective of this study was to better characterize potential use of the RS in these patients. METHODS: The National Cancer Database was used to identify women with stage I through III, T1 through T3, N0 or N1, hormone receptor-positive, HER2-negative ILC or invasive ductal carcinoma (IDC) who had an available RS between 2010 and 2016. Multivariable Cox regression was used to model the effect of variables on 5-year overall survival (OS). The Kaplan-Meier method was used to estimate OS according to the RS, nodal status, and chemotherapy. RESULTS: In total, 15,763 patients with ILC and 100,070 with IDC were identified. The mean age of patients with ILC and IDC was 59.2 ± 9.1 and 57.2 ± 9.8, respectively. A lower percentage of patients with ILC versus those with IDC had a high RS, defined as >25 (6.6% vs 16.0%; P < .0001). ILC patients with a high RS who had N0 or N1 disease received approximately 10% less chemotherapy compared with similar patients who had IDC. The results indicated that the RS had statistically significant prognostic value for patients with ILC. In addition, an absolute OS advantage was correlated with the receipt of chemotherapy by patients with ILC who had a high RS with N0 or N1 disease. CONCLUSIONS: Patients with ILC who have a high RS are treated less often with chemotherapy compared with similar patients who have IDC. Nevertheless, the RS has a prognostic as well as a predictive value in ILC, with an association between OS benefit and chemotherapy receipt in patients who have ILC with a high RS, especially if they have N1 disease.

The use of national datasets to evaluate outcomes for invasive lobular carcinoma

“…It is unclear why the results differ. Perhaps the most glaring omission from Weiser et al 1 is that, in contrast to these previous studies, they did not stratify patients with high‐risk–RS ILC by comorbidity status. Patients with high‐risk–RS ILC who received adjuvant chemotherapy were likely younger and had fewer comorbidities than those patients who did not receive chemotherapy.…”

“…

The use of national datasets to evaluate outcomes for invasive lobular carcinoma Weiser et al 1 are to be congratulated for their work in identifying specific subgroups of patients with invasive lobular carcinoma (ILC) who might benefit from chemotherapy or targeted therapies. Using National Cancer Database (NCDB) data for patients with ILC, they reported that the recurrence score (RS) was prognostic for overall survival (OS) for patients with ILC: patients with a high-risk-RS had worse OS.

…”

“…However, Weiser et al 1 also reported that among patients with high‐risk–RS ILC, adjuvant chemotherapy was associated with improved OS. These findings are in contrast to other reports.…”

Reply to “The use of national datasets to evaluate outcomes for invasive lobular carcinoma”

“…We thank Marmor and colleagues for their interest in our publication and for their appreciation of our efforts 1 . Their letter discusses our finding that patients with invasive lobular carcinoma (ILC) and a high Oncotype recurrence score (RS; defined as >25), curated from the National Cancer Database (NCDB), have an association between chemotherapy (CT) use and improved overall survival (OS).…”

“…We thank Marmor and colleagues for their interest in our publication and for their appreciation of our efforts. 1 Their letter discusses our finding that patients with invasive lobular carcinoma (ILC) and a high Oncotype recurrence score (RS; defined as >25), curated from the National Cancer Database (NCDB), have an association between chemotherapy (CT) use and improved overall survival (OS). We found an absolute OS advantage of 4.6% in these patients and, when they were stratified by the N status, a small OS advantage of 4.2% in N0 patients versus a 12.2% advantage in N1 patients.…”