Recurrence Score Testing Does not Appear to Benefit Patients With Grade 1, Progesterone Receptor-Positive Breast Cancers: An Opportunity to Eliminate Overtreatment and Decrease Testing Costs

Background: We previously described a risk prediction model (Anne Arundel Medical Center [AAMC] model) based on pathology which may eliminate the need for recurrence score (RS) testing in select early-stage breast cancers. There is a concern that patients in discordant risk prediction groups (AAMC vs. RS) may be overtreated or undertreated if RS testing were omitted. Methods: We queried the Surveillance, Epidemiology, and End Results (SEER) database for all breast cancer patients between 2004 and 2015. AAMC low-risk was defined as Grade 1 and progesterone receptor-positive (PR + ) tumors, while AAMC high-risk was defined as Grade 3 or estrogen-negative tumors. RS low-risk group was defined as RS < 16 and age ≤ 50 years, or RS ≤ 25 and age > 50 years. RS high-risk group was defined as RS > 25. Results: A total of 71,212 cases were analyzed. Of these, 590 were AAMC low-risk/RS high-risk discordant, while 5,596 were AAMC high-risk/RS low-risk discordant. For AAMC low-risk/RS high-risk discordant, 10-year breast cancerspecific survival (BCSS) did not differ for patients who received adjuvant chemotherapy versus those who did not (93% chemotherapy vs. 99% unknown/no chemotherapy, p = .12). Overall survival (OS) was also comparable (92% chemotherapy vs. 91% unknown/no chemotherapy, p = .42). In the AAMC high-risk/RS low-risk discordant group, 10-year BCSS (92% chemotherapy vs. 96% unknown/no chemotherapy, p = .06) and OS (87% chemotherapy vs. 90% unknown/ no chemotherapy, p = .52) did not differ between adjuvant chemotherapy and unknown/no chemotherapy groups. Conclusions: Adjuvant chemotherapy in the AAMC low-risk/RS high-risk and AAMC high-risk/RS low-risk discordant groups did not improve survival. This supports consideration of omission of RS testing in Grade 1, PR + tumors. Patients with Grade 3 tumors do benefit from RS testing.