187PHigh dose neo-adjuvant chemotherapy in triple-negative breast cancer with evidence of homologous recombination deficiency (HRD)

Abstract Background Subgroup analyses of randomized, controlled trials (RCTs) using high dose chemotherapy (HD) with autologous stem cell support (autoSCT) reveal a potential benefit of HD in patients with triple negative breast cancer (TNBC) and in tumors harboring features of HRD (Vollebergh et al, 2011). The phase III part of the neo-TN RCT, compared neo-adjuvant HD with conventional treatment (CONV) in BRCA1-like TNBC. Methods Patients with cT2-3N0-3M0 TNBC with HRD were randomized between HD or CONV after 3 courses of 2-weekly doxorubicin/cyclophosphamide (ddA60C600). CONV treatment consisted of another 3xddAC, or in case of an unfavorable response on MRI a switch to 3 cycles capecitabine800BID14-7/docetaxel75; after an amendment all patients switched to 3xcarboplatinAUC6/paclitaxel80weekly. The HD regimen consisted of a 4th course ddAC and 2 courses of cyclophosphamide3000d1/carboplatin400d1,2/ thiotepa250d2 both followed by autoSCT. Primary outcome was the Neo-adjuvant Response Index (NRI), secondary outcomes included overall (OS) and recurrence free survival (RFS). Results From 2010 to 2016, 122 patients were randomized (intention-to-treat). Median follow-up is 45 months. There was no significant difference in NRI between HD and CONV (mean NRI 0.78 versus 0.72, range 0-1, p = 0.41 Wilcoxon test). An NRI of > 0.7 was strongly associated with good prognosis (RFS of 97% [95%CI 93%-100%] versus 67% [95%CI 55%-82%] at 4 years). See Table for 4-years OS and RFS comparing HD with CONV. There were no treatment related deaths. However, 7 out of 55 patients who actually received HD did not complete HD mainly because of infections or allergic reactions. Table . 187P HD % (95%CI) CONV % (95%CI) HR (95%CI) p-value All (n = 122) 4-yrs OS 4-yrs RFS 92 (85-99) 92 (85-99) 79 (69-91) 78 (68-89) 0.43 (0.15-1.23) 0.41 (0.15-1.07) 0.12 0.07 Stage III only (n = 35) 4-yrs OS 93 (81-100) 60 (39-93) 0.14 (0.02-1.14) 0.07 Conclusions No significant efficacy differences were found between HD and CONV. The NRI is of prognostic value. Whether the HD regimen is promising, especially in very high risk BRCA1-like TNBC [HR = 0.14], requires additional data including a comparison with platinum treatment in all control arm patients. Clinical trial identification NCT01057069. Legal entity responsible for the study The authors. Funding The Dutch Cancer Foundation (KWF) and by the Schumacher Kramer Foundation. Disclosure S.C. Linn: Research grant / Funding (self), research support for patients fees in D-Care study: Amgen; Advisory / Consultancy, Research grant / Funding (self), Research grant / Funding (institution): AstraZeneca; Advisory / Consultancy: Cergentis; Research grant / Funding (self): Genentech; Advisory / Consultancy: IBM; Research grant / Funding (self): Novartis; Advisory / Consultancy, Research grant / Funding (institution), patients fees in TEAM 2b study: Pfizer; Advisory / Consultancy, Research grant / Funding (self), Research grant / Funding (institution): Roche; Research grant / Funding (institution): Tesaro; Speaker Bureau / Expert testimony, Fee for teaching, paid to institution: Bayer. All other authors have declared no conflicts of interest.