Abstract GS5-05: Primary analysis of NRG-BR005, a phase II trial assessing accuracy of tumor bed biopsies in predicting pathologic complete response (pCR) in patients with clinical/radiological complete response after neoadjuvant chemotherapy (NCT) to explore the feasibility of breast-conserving treatment without surgery

Background: Increasing use of NCT regimens and improved selection of appropriate candidates has increased rates of pCR and has raised the question of whether surgical removal of the primary tumor is required for these patients (pts). To avoid surgery, the ability to predict pCR after NCT must be very accurate. Breast imaging alone (even when it includes mammogram, ultrasound, and breast MRI), is inadequate to achieve this. We hypothesized that performing core needle biopsy (bx) of the tumor bed in addition to tri-modality imaging in pts who have clinical complete response (cCR) to NCT will increase the ability to predict pCR. Methods: NRG-BR005 was planned as a two-stage Phase II trial, designed to assess the accuracy of post-NCT stereotactic tumor bed bx for pCR (absence of both invasive and DCIS residual disease), in pts with cCR and radiologic complete response (rCR)/near rCR by tri-modality imaging. Pts with operable (T1-T3, stage I/IIIA) invasive ductal carcinoma were eligible. Pts must have completed NCT and achieved a cCR as well as rCR/near rCR by mammography (mass ≤1 cm and no malignant microcalcifications), ultrasound (mass ≤2cm), and MRI (no mass with rapid rise or washout kinetics). Pts underwent marker-directed stereotactic multiple-core needle bx of the tumor bed with marker placement to facilitate BCS. The primary endpoint was the negative predictive value (NPV) of the bx, defined as the number of pts with a negative bx and confirmed pCR at BCS divided by the total number of pts with a negative bx. A bx detection NPV of >90% was required to support the feasibility of foregoing BCS. We also calculated the sensitivity of bx, defined as the number of pts with positive bx who had residual tumor at BCS divided by the total pts with residual tumor at BCS. The point estimate and 95% confidence intervals (CI) for these parameters were calculated using the exact method. Planned accrual was 175 pts in order to obtain 35 pts who had residual tumor at surgery. Results: From 8/17 to 6/19, 105 pts were accrued. As of 7/31/19, 98 pts were evaluable for analysis. Pts were 62.2% white, 87.7% non-Hispanic or Latino, 44.9% estrogen receptor-positive, and 44.9% HER2 positive. Among evaluable pts, 36 had residual disease at surgery. Stereotactic bx was positive in 18/36 pts (50%). Stereotactic bx was negative for all 62 pts who achieved pCR. NPV of the bx was 77.5% (95% CI: 66.8% to 86.1%) indicating that a negative bx provided incorrect diagnosis in 22.5% of pts. Sensitivity of the bx was 50% (95% CI: 32.9% to 67.1%) indicating that when disease was present, the bx detected it only 50% of the time. Complications from bx were reported in 7/98 evaluable pts (7.1%) (6 post-procedure hematomas, 1 breast pain). Conclusions: In the current findings of NRG-BR005, bx did not achieve an NPV of >90% and identified only 50% of the pts who had residual disease at surgery following NCT. The findings do not support breast-conserving treatment without surgery based on the study criteria for cCR and rCR/near rCR and negative tumor bed bx. Further analyses including central review of the tri-modality imaging and assessment of the imaging algorithm with and without the addition of bx are underway. Once these analyses are combined with information on biologic subtypes, a new prediction model may be defined. Support: U10CA180868, U10CA180822. Citation Format: Mark Basik, Reena S Cecchini, Jennifer F De Los Santos, Heidi R Umphrey, Thomas B Julian, Eleftherios P Mamounas, Julia White, Peter C Lucas, Christa Balanoff, Antoinette R Tan, Joseph J. Weber, David A Edmonson, Ursa A. Brown-Glaberman, Emilia J. Diego, Mediget Teshome, Cindy B Matsen, Samantha A Seaward, Irene L. Wapnir, Jamie L Wagner, Judy A Tjoe, Alastair M Thompson, Norman Wolmark. Primary analysis of NRG-BR005, a phase II trial assessing accuracy of tumor bed biopsies in predicting pathologic complete response (pCR) in patients with clinical/radiological complete response after neoadjuvant chemotherapy (NCT) to explore the feasibility of breast-conserving treatment without surgery [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr GS5-05.