Background

Benefit of neoadjuvant chemotherapy in patients (pts) with luminal breast cancer (LBC) is limited. Palbociclib combined with endocrine treatment has shown impressive results in advanced LBC. We conducted a randomized parallel phase II study, assessing letrozole + palbociclib (LP) as neoadjuvant treatment in LBC.

Methods

Postmenopausal women were eligible if they had a stage II-III ER-positive HER2-negative BC, not candidate for breast conserving surgery (BCS), with either a PAM50 luminal B, or a PAM50 luminal A profile with proven lymph node involvement (N+). A parallel 1:1 randomization proposed 6 courses of 3^rd generation chemotherapy (FEC100 x 3 - docetaxel 100 x 3), or 19 weeks (wks) of L 2.5 mg/day plus P 125 mg/day, 3 wks/4. Surgery was performed at wk 20. Primary endpoint was locally assessed Residual Cancer Burden (RCB) rate. Main secondary endpoints included safety, response rate, positive and negative predictive values of PAM50 ROR (risk of recurrence)-defined status, centrally reviewed RCB, and BCS rates. The protocol planned that the trial should be stopped for futility if ≤ 5 local RCB 0-I events (16.7%) were observed in the first 30 pts in the LP arm.

Results

Out of 184 screened pts, 106 women with Stage II-IIIA, PAM50-ascertained LBC were randomized. Pts had T1-2 (73%) or T3 (27%); N + (26.5%); luminal B (89%) tumors. Median ROR score was 68 (22-93). At interim analysis, RCB 0-I was observed in 1 pt in the LP arm and inclusions were stopped. At final analysis, local RCB 0/I/II/III was observed in 3.8%/3.8%/52%/40.4% of pts in the LP arm, and in 5.9%/9.8%/37.3%/47.1% in the chemo arm, respectively. Central and local RCB results were identical. ROR score was not predictive of RCB 0/1. Clinical objective response rates were 74.5% and 76%, and BCS rates 69.2% and 68.6%, in the LP and chemo arms, respectively. Ki67 final median value was significantly lower in the LP arm [3% (range 1-40) vs 8% (2-15), p=.017). Of 19 serious adverse events, 2 occurred in the LP arm and 17 in the chemo arm (p < 0.001).

Conclusions

Neoadjuvant LP led to a slightly lower pCR/RCB 0-I rate than chemo, however clinical response and BCS rates were similar in both arms and LP had a much better safety profile. Extensive analyses are ongoing.

Clinical trial identification

NCT02400567

Legal entity responsible for the study

UNICANCER R&D

Funding

Pfizer - Nanostring

Disclosure

P. Cottu: Consulting Fees (e.g. advisory boards): Pfizer, Roche, Novartis. Travel: Pfizer, Roche, Novartis. F. Duhoux: Consulting Fees (e.g. advisory boards): Pfizer, Roche. Travel: Amgen, Pfizer, Roche, Teva. All other authors have declared no conflicts of interest.

Background

Taselisib is an oral, potent, and selective PI3-kinase (PI3K) inhibitor with enhanced activity against PIK3CA mutant (MUT) cancer cells. Confirmed partial responses were observed in pts with PIK3CA MUT metastatic breast cancer treated with taselisib as a single agent and combined with endocrine therapy (ET).

Methods

334 postmenopausal pts with ER+/HER2-, Stage I-III, operable EBC and evaluable tumor tissue for centralized PIK3CA genotyping were randomized (1:1) in 90 sites worldwide to receive LET with either taselisib (4mg 5 days on/2 days off) or PLA for 16 weeks, followed by surgery. Co-primary endpoints: objective response rate (ORR) by centrally assessed breast MRI and pathologic complete response (pCR, ypT0/is N0) rate at surgery, in all randomized pts and in pts with PIK3CA MUT tumors. The sample size was calculated to detect in the PIK3CA MUT subset an absolute increase of 24% in ORR (from 40% to 64%, minimal detectable difference [MDD] 15%; 2-sided α 16%, 80% power), and 18% in pCR (from 1% to 19%; MDD 13%; 2-sided α 4%, 80% power).

Results

The study met its primary endpoint: the addition of taselisib to LET increased the ORR from 38% to 56.2% in the PIK3CA MUT subset (N = 152; Odds ratio [OR] 2.03, 95%CI 1.06-3.88, p = 0.033) and in all randomized pts (from 39.3% to 50%, OR 1.55, 95%CI 1.00-2.38, p = 0.049). No significant difference was observed for pCR rate overall or in the PIK3CA MUT subset. Most common G3-4 adverse events in the taselisib arm: gastrointestinal disorders (7.8%), infections (4.8%), and skin/subcutaneous tissue disorders (4.8%). G3-4 hyperglycemia occurred in 1.2% of pts. One sudden death occurred in the taselisib arm, but was considered unrelated to study treatment. Taselisib discontinuation (10.8%) and dose reductions (11.4%) were infrequent.

Conclusions

LORELEI is the first randomized study to demonstrate a significant increase in ORR measured by MRI upon treatment with a PI3K selective inhibitor + ET in ER+/HER2- EBC pts. Toxicity was manageable. Ongoing comprehensive biomarker analyses will provide further insight into the antitumor responses observed with this combination.

Clinical trial identification

NCT02273973

Legal entity responsible for the study

Colead partnership between Genentech (Sponsor) - BIG (Governance & coordination) - ABCSG (data center) - SOLTI (biosamples)

Funding

Genentech Inc.

Disclosure

T.R. Wilson, Y. Shi: Employed by Genentech Inc., stocks in Roche. T.J. Stout: Employed by Genentech Inc. J.Y. Hsu: Employed by Genentech Inc., M. Piccart: Consultant (honoraria) from Roche-Genentech, Research grants to my Institute from Roche-Genentech. M. Gnant: Honoraria: Roche Research Funding: Roche. All other authors have declared no conflicts of interest.

Background

F is a selective estrogen receptor degrader for HR+ advanced breast cancer patients (pts). We designed this trial to compare A vs. A in combination with F (A+F) to address the hypothesis that a complete estrogen blockade can prevent resistance to aromatase inhibitors in the adjuvant setting.

Methods

This was a multicenter, open label, phase III study in which HR+/HER2- EBC postmenopausal pts were randomized 1:1 to A for 5 years (y) or A+F (A concurrently with F 250mg/4 weeks for 3y followed by 2y of A). Pts were stratified for prior chemotherapy (yes/no); number of positive lymph-nodes (0/1-3/≥4); HR status (both positive/one positive) and site. Primary objective was disease-free survival (DFS). To detect an absolute 5-y DFS increase of 3% (90% A, 93% A+F) a sample size of 2716 evaluable patients was required. On 2010, when the negative results of the FACT trial were made available, the financer decided to stop the study support after the inclusion of 872 pts.

Results

From January 2008 to June 2010, 437 pts were randomized to A and 435 to A+F. Pts characteristics were well balanced between arms; median age was 62 y (40-86), 46.9% were N0, 89.5% ER+/PgR+ and 68.2% had received prior chemotherapy. Treatment was completed as planned by 72.5% and 48.5% of A and A+F pts. Median relative dose intensity was 99% for A (both arms) and 81% for F. Most relevant G2-3 toxicities (>5% in either arm) with A vs. A+F were hypertension (13.2%; 9.9%), fatigue (5.2%; 11.8%), LDL-Cholesterol increase (9.4%; 5.3%), osteoporosis (5.5%; 6.9%) and musculoskeletal bone/joint pain (26.3%; 29.4%). After a median follow-up of 6.3y, the proportion of pts disease free at 5y was 90.99% (A 90.77%, A+F 91.25%, D = 0.48%, p = 0.357); 9.4% had BC relapse (A 10.5%; A+F 8.3%, D = 2.2%) and 4.3% had secondary tumors (A 3.9%; A+F 4.6%). Survival and breast cancer-specific survival were not reached.

Conclusions

The GEICAM/2006-10 study failed to demonstrate a statistically significant increase in DFS adding F to A as adjuvant therapy, though no formal conclusion can be extracted from this trial due to the F administered dose and the actual trial sample size.

Clinical trial identification

NCT00543127

Legal entity responsible for the study

GEICAM Spanish Breast Cancer Group

Funding

Astra Zeneca Astra Zeneca Astra Zeneca Astra Zeneca

Disclosure

M. Martin Jimenez: Speaker honoraria and AstraZeneca advisory boards participation. J.M. Baena-Canada: Consulting/relationship advice for AstraZeneca. M. Muñoz: Advisory Board from AstraZeneca. All other authors have declared no conflicts of interest.

Background

In the Herceptin Adjuvant (HERA) trial, 24% of patients (pts) who received T for 1 year (y) had a disease recurrence at 11 y follow-up. The primary analysis of the ExteNET trial, performed after 2 y follow-up, showed that a 1-y course of neratinib after T-based adjuvant therapy significantly improved invasive disease-free survival (iDFS) vs placebo in early-stage HER2+ BC (HR 0.67; 95% CI 0.50–0.91; p = 0.0091) [Chan et al. Lancet Oncol 2016]. We now report updated 5-y efficacy findings.

Methods

ExteNET is an international, multicenter, randomized, double-blind, placebo-controlled phase III trial. Pts received oral neratinib 240 mg/d or placebo for 1 y. After 2 y, randomized pts were asked to re-consent to collection of data concerning disease recurrences and survival from medical records for a further 3 y. The preplanned 5-y analysis was by intention-to-treat (ITT). Non-consenting pts were censored at their last physical examination. Primary endpoint: iDFS. HR (95% CI) were estimated using Cox proportional-hazards models. Data cut-off: March 2017. Clinicaltrials.gov: NCT00878709.

Results

ITT population comprised 2840 pts (neratinib, n = 1420; placebo, n = 1420); 53 pts died during the initial 2-y follow-up. Among 2787 available pts, 2117 (76%) re-consented to additional follow-up (neratinib, n = 1028; placebo, n = 1089). Updated results after a median follow-up of 5.2 y are shown below. Secondary efficacy endpoints were supportive of the primary analysis.Table:

149O

Conclusions

1 y of neratinib after T-based adjuvant therapy significantly improves iDFS at 5 y in pts with early-stage HER2+ BC, with a long-term sustained effect. A protocol-specified subgroup analysis suggested greater benefit in HR+ pts. Overall survival data are not yet mature.

Clinical trial identification

NCT00878709

Legal entity responsible for the study

Wyeth, Pfizer and Puma Biotechnology

Funding

Puma Biotechnology

Disclosure

B. Ejlertsen: Grants to institution from NanoString, Novartis, and Roche, outside the submitted work. Travel support for educational meetings from AstraZeneca and Celgene. S. Delaloge: Grants and personal fees from Roche and GSK. G. von Minckwitz: Research funding to the institution from Amgen, Roche, Novartis, Celgene, Teva, AstraZeneca, Myriad Genetics, AbbVie and Vifor Pharma. C. Barrios: Grants from Amgen, AstraZeneca, Boehringer, Novartis, Pfizer, Roche, Celgene, Sanofi, Lilly, Puma. Personal fees from GSK, Novartis, Pfizer, Roche, Eisai. M. Gnant: Grants from Sanofi-Aventis, Novartis, Roche, GSK, Pfizer, Smith Medical. Personal fees from Novartis, AstraZeneca, Accelsiors, Esai. S-B. Kim: Research funding to the institution from Novartis, Sanofi-Aventis, Kyowa-Kirin Inc, and Dongkook Pharma Co., Ltd. R.P. Bryce: Emplyee and stock options: Puma Biotechnology Inc. F. Xu: Emplyee: Puma Biotechnology Inc. M. Buyse: Emplyee and shareholder: IDDI. A. Chan: Personal fees for educational meetings from Pfizer, Amgen. All other authors have declared no conflicts of interest.

Background

Adjuvant systemic therapy for pT1abN0 breast cancer is controversial, as these tumors overall have a low relapse risk. The best tool to identify a subgroup that would benefit from adjuvant treatment is unknown.

Methods

This subgroup included patients with pT1abN0 tumors enrolled in MINDACT who had both their genomic risk G (per MammaPrint®) & clinical risk C (per a modified version Adjuvant! Online) assessed. Patients characterized as low-risk in both assessments were spared chemotherapy (CT), while in those characterized as C&G high CT was advised. Discordant cases were randomized to receive CT based on the C or the G result. Here, we report the 5-year rates of distant metastasis-free survival (DMFS), distant metastases-free interval (DMFI) & overall survival (OS) for pT1abN0 patients who received or not CT based on their G or C risk result.

Results

826/6693 (12.3%) patients with pT1abN0 tumors were enrolled in MINDACT. 310/826 (37.5%) were ≥ 60 years & 525/826 (63.6%) postmenopausal. 727/826 samples were reviewed by central pathology; 585/727 (80.5%) were invasive ductal, 662/727 (91.1%) ER positive, 46/727 HER2 positive (6.3%) & 81/727 (11.1%) were grade 3 tumors. IHC subtype classification identified 426/727 (58.6%) as Luminal A; 193/727 (26.5%) Luminal B; 38/727 (5.2%) Luminal B/HER2 positive; 8/727 (1.1%) HER2- positive; 37/727 (5.1%) triple-negative tumors. Almost all patients (820/826; 99.3%) were clinical low-risk (CL). Overall, 624/826 (75.5%) were CL/GL & 196/826 (23.7%) were CL/GH (5 patients were CH/GL, no cases were CH/GH, 1 missing). 5-year DMFS for patients with CL/GH pT1abN0 tumors who received CT was 97.3% (95% CI, 89.4-99.3) vs 91.4% (95% CI, 82.6-95.9) for those who did not. 5-years DMFI & OS for these patients treated with CT were 98.8% (95% CI, 91.9-99.8) & 98.5% (95% CI, 89.6- 99.8) vs 91.4% (95% CI, 82.6- 95.9) & 95.8% (89.1- 98.4%) respectively for those who did not receive CT.

Conclusions

Biological characteristics can be used as determinants of adjuvant CT administration for T1abN0 tumors. An important portion (23.7%) of these small tumors was CL but GH (MammaPrint^®) risk and derived a benefit from CT.

Clinical trial identification

The main MINDACT study: ClinicalTrials.gov number: NCT00433589, EudraCT number:2005-002625-31

Legal entity responsible for the study

European Organization for Research and Treatment of Cancer (EORTC)

Funding

None

Disclosure

L. Van \'t Veer: Personal fees and other support from Agendia NV outside the submitted work. Co- inventor on a patent related to MammaPrint: Diagnosis and prognosis of breast cancer patients (WO2002103320, licensed to Agendia, NV). M. Piccart: Consulting or advisory role for AstraZeneca, MSD, Novartis, Pfizer, Roche- Genentech, Periphagen, Huya, Debiopharm, PharmaMar, Radius. Research funding from most above mentioned companies. F. Cardoso: Consultancy/research grants from Astellas/Medivation, AstraZeneca, Celgene, Daiichi- Sankyo, Eisai, GE Oncology, Genentech, Glaxo Smith Kline (GSK), Macrogenics, Merck- Sharp, Merus BV, Novartis, Pfizer, Pierre- Fabre, Roche, Sanofi and Teva. All other authors have declared no conflicts of interest.