Abstract PS13-19: Achieving a pathologically negative axilla after neoadjuvant chemotherapy for breast cancer is associated with presenting tumor size and subtype

Abstract

Abstract Introduction: Axillary pathologic complete response (pCR) after neoadjuvant chemotherapy (NAC) confers higher overall and recurrence-free survival, compared to residual axillary disease. Recent findings suggest that pCR in the breast (ypT0) post-NAC is associated with a pathologically negative axilla (ypN0) in patients (pts) presenting with lower stage HER2+ and triple negative breast cancer (TNBC). Additional studies are needed to understand how clinical T (cT) and N (cN) staging are associated with ypN0 in other tumor subtypes, including hormone-receptor (HR) positive tumors. The ability to reliably predict axillary pathologic response post-NAC may allow identification of a subset of pts for whom axillary staging may be safely omitted. We hypothesize that tumor subtype and lower clinical stage at presentation are associated with ypN0. Methods: A single institution cancer registry was retrospectively reviewed for pts receiving NAC followed by surgery from January 2010-June 2018. Fisher’s exact tests were used to compared proportion of breast and axillary pCR by tumor subtype (TNBC, HR+/HER2-, HR+/HER2+ and HR-/HER2+). Univariable logistic regression determined factors associated with ypN0. Multivariable logistic regression determined the association between ypN0 and tumor subtype adjusting for factors that retained significance on univariable analysis. Sensitivity analyses determined how cN status affected ypN status by tumor subtype. Results: Of the 1348 pts who received NAC followed by surgery, median age was 54 (IQR 44-63); 59% (n=738) were postmenopausal. Proportion of tumor subtypes were: 15% (n=197) TNBC, 12% (n=155) HR+/HER2-, 48% (n=653) HR+/HER2+, and 25% (n=343) HR-/HER2+. Tumor size at diagnosis was: 1% (n=18) T0, 20% (n=272) T1, 53% (n=713) T2, 17% (n=230) T3 and 9% (n=111) T4. Clinical nodal staging at diagnosis was: 52% cN0 (n=695), 41% cN1 (n=550), 5% cN2 (n=61), and 3% cN3 (n=43). TNBC and HER2+ subtypes were associated with the highest rate of breast pCR and ypN0. On univariable analyses of the cN positive pts, younger age at diagnosis, non-postmenopausal status, oral contraceptive use, alcohol consumption, cT stage, cN stage and tumor subtype were significantly associated with ypN0 (Table1A). In the adjusted model, postmenopausal status, cT, and tumor subtype were associated with ypN0. Lower cT and HR- subtypes had significantly higher odds of ypN0 (Table 1B). In sensitivity analyses, cN2/cN3 was associated with lower odds of ypN0 compared to cN0/cN1 disease in TNBC (OR0.11 95%CI 0.03,0.40, p=0.001), HR-/HER2+ disease (OR0.42, 95%CI 0.22,0.77, p=0.005), and HR+/HER2+ (OR0.26 95%CI 0.11,0.61 p=0.002), but not in HR+/HER2- disease (OR1.17, 95%CI 0.25,5.57, p=0.838). Conclusion: HR- and low cT stage at diagnosis are associated with ypN0 in this large cohort. Younger age, pre-menopausal status and cN stage may be important considerations in future investigations aimed at defining the subset of patients most likely to achieve ypN0 and ultimately to be considered for de-escalation of axillary staging post NAC. A.Univariable logistic regression analysisVariableOdds Ratio95% Confidence Intervalp valueAge at diagnosis0.990.98, 0.99<0.001RaceWhite1.060.70, 1.600.777Black0.600.17, 2.080.421Asian1.200.36, 4.000.768Postmenopausal0.760.60, 0.960.023BMI1.000.99. 1.000.424Oral contraceptive use1.250.95, 1.630.114Alcohol consumption1.601.05, 2.440.027 Tobacco use1.130.99, 1.290.064Family history of cancer1.250.92, 1.680.148Grade12.170.75, 6.250.15320.950.45, 2.010.89131.700.81, 3.590.163HistologyIDC1.451.10, 1.890.007ILC0.370.23, 0.60<0.001Mixed0.950.68, 1.320.764Other0.440.18, 1.080.073Clinical T stage11.370.47, 3.980.56920.760.27, 2.160.60530.350.12, 1.020.05540.150.05, 0.450.001Clinical N stage10.140.11, 0.18<0.00120.130.07, 0.23<0.00130.110.06, 0.21<0.001Tumor subtypeTNBC1.591.00, 2.530.049HR+/ HER2 -0.550.35, 0.870.010HR+/ HER2+0.690.47, 1.000.049HR-/ HER2 +1.641.09, 2.460.019B.Multivariable logistic regression analysisAge at diagnosis0.980.97, 0.99<0.001Alcohol use1.190.74, 1.930.476HistologyIDC1.150.82, 1.610.414ILC0.570.32, 1.010.053Mixed1.080.71, 1.630.721Other0.790.28, 2.220.660Clinical T stage11.190.39, 3.580.76120.740.25, 2.160.58530.510.17, 1.540.23540.290.09, 0.910.034Clinical N stage10.130.10, 0.18<0.00120.140.08, 0.26<0.00130.100.05, 0.20<0.001Tumor subtypeTNBC1.440.84, 2.470.181HR+/ HER2 -0.540.31, 0.940.028HR+/ HER2+0.600.39, 0.930.024HR-/ HER2 +1.701.05, 2.730.030 Citation Format: Sara P Myers, Gillian M Ahrendt, Joanna S Lee, Jennifer G Steiman, Atilla Soran, Ronald R Johnson, Priscilla F McAuliffe, Emilia J Diego. Achieving a pathologically negative axilla after neoadjuvant chemotherapy for breast cancer is associated with presenting tumor size and subtype [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS13-19.