Baseline Ki-67 Research Articles

218P - Prognostic estimates of Ki-67 percentage drop after neoadjuvant chemotherapy (NAC) in luminal B (lumB) and triple negative breast cancer (TNBC)

Background: Pathologic complete response (pCR) and residual cancer burden (RCB) after NAC are validated prognostic markers in BC. We assessed the impact of adding Ki-67% drop (baseline biopsy - surgery) to distant metastasis relapse-free survival (dRFS) models containing CP factors plus post-treatment stage (MDACC CPS score), estrogen receptor status and tumor grade (MDACC CPS+EG score). Methods: Records from 341 patients (pts) with lumB/HER2 neg (baseline Ki67 >20%) or TNBC who received NAC from 2008 to 2015 at our hospital were reviewed. Uni- and multivariate Cox models were constructed and concordance-index (c-index) calculated. Results: Pts: median age 47 years (24-83), 60% lumB and 40% TNBC, 62% stage 2, 38% stage 3. pCR: 12% lumB, 32% TNBC (p < 0.01). Median Ki-67% drop: 24% lumB, 5% TNBC (p < 0.01), without differences by NAC regimen. dRFS at 5 year-median follow-up was 75% in lumB (CI95% 67-83) vs 62% in TNBC (CI95% 53-74, p < 0.01); 90% in RCB 0/1 (CI95% 81-97) vs 74% in RCB 2/3 (CI95% 65-83, p < 0.01). As compared to pts with RCB 0/1, those with RCB 2/3 plus Ki-67% drop > = 20% (best cut-off in univariate model) had similar dRFS at 5 years (90%, CI95% 81-100, p = 0.48), irrespective of molecular group. Enrichment for Ki-67 > =20% drop in lumB (60%) vs TNBC (30%, p < 0.01) was observed. Both CPS and CPS+EG scores were validated as independent prognostic factors in univariate dRFS models (c-index of 0.70 and 0.78, respectively). The addition of Ki-67% drop (< 20% vs > = 20%) to CPS and CPS+EG scores in multivariable models significantly improved their performance (c-index of 0.74 and 0.81, respectively). Ki-67 > = 20% drop associates with 70-80% reduction in distant relapse risk (HR 0.27 and 0.16 in CPS and CPS+EG models, respectively, p < 0.05). Conclusions: Our data support the addition of Ki-67% drop after NAC in lumB and TNBC to existing dRFS online outcome calculators. In the context of RCB 2/3, Ki-67 > = 20% drop is mainly seen in lumB/HER neg tumors. Importantly, Ki-67 < 20% drop identifies a high-risk population that may be eligible to clinical trials with novel therapeutic interventions in the adjuvant setting. Legal entity responsible for the study: Vall d'Hebron University Hospital Funding: Vall d'Hebron Institute of Oncology Disclosure: M. Oliveira: Consulting or Advisory Role in Roche-Genentech and Puma Biotechnology. Research Funding in AstraZeneca and Genentech/Roche. R. Dienstmann: Consulting or Advisory Role in Roche-Genentech and in Astellas Pharma. Research Funding in Merck. All other authors have declared no conflicts of interest.

Abstract CT045: Ribociclib + letrozole for first-line treatment of hormone receptor-positive (HR+), human epidermal growth factor receptor 2-negative (HER2-) advanced breast cancer (ABC): efficacy by baseline tumor markers

Abstract Background: Cyclin D-cyclin-dependent kinase (CDK) 4/6 complexes promote cell proliferation through phosphorylation of retinoblastoma protein (Rb). In breast cancer, cyclin D-CDK4/6 activity can be increased through cyclin D gene (CCND1) amplification or loss of the CDK4/6 negative regulator p16. Here we present efficacy data from the Phase III MONALEESA-2 study of ribociclib (CDK4/6 inhibitor) + letrozole vs. placebo + letrozole for first-line treatment of HR+, HER2- ABC, assessed in baseline tumors by protein levels of Rb, p16, the cell proliferation marker Ki67, and by gene expression levels of CDKN2A (p16) and CCND1. Methods: Postmenopausal women with HR+, HER2- ABC with no prior systemic therapy for advanced disease were randomized 1:1 to receive ribociclib or placebo (600 mg/day 3-weeks-on/1-week-off) + letrozole (2.5 mg/day continuous). The primary endpoint was investigator-assessed progression-free survival (PFS). Provision of a representative baseline tumor biopsy or archival tissue at screening was mandatory if available. Baseline tumor tissue was evaluated for protein biomarkers (immunohistochemistry) and gene expression (NanoString nCounter® Human Cancer Reference panel). Results: Of 668 patients randomized, 479 were evaluable for total Rb, and 416 (87%) displayed high levels (H-score ≥100). p16 protein levels were evaluable in 405 patients; 165 (41%) had low (H-score &amp;lt;50), 182 (45%) medium (H-score ≥50-149), and 58 (14%) had high (H-score ≥150) p16 levels. Ki67 was detected in ≤14% of tumor cells in 216 (47%) patients and in &amp;gt;14% of cells in 247 (53%) patients. The median messenger RNA expression level was used as the cut-off to define patients with low or high baseline CDKN2A and CCND1 gene expression. An improved PFS was observed by the addition of ribociclib to letrozole in all the above patient subgroups, with hazard ratios ranging from 0.40 (high p16 by H-score; 95% confidence interval [CI] 0.16-1.0; p=0.06) to 0.64 (≤14% Ki67-positive cells; 95% CI 0.39-1.0; p=0.07). Patients with less or greater than 14% Ki67-positive cells, lower or higher p16 levels, Rb levels, or CDKN2A or CCND1 gene expression benefitted from the addition of ribociclib to letrozole to a similar extent. Conclusions: A consistent benefit from ribociclib + letrozole vs. placebo + letrozole was observed irrespective of baseline Rb, p16, and Ki67 levels or CDKN2A and CCND1 gene expression levels. Hormone receptor positivity remains the only established biomarker of response to CDK4/6 inhibitors. Citation Format: Fabrice Andre, Salomon M. Stemmer, Mario Campone, Katarina Petrakova, Shani Paluch-Shimon, Yoon-Sim Yap, Norbert Marschner, Arlene Chan, Cristian Villanueva, Lowell L. Hart, Carlos L. Arteaga, Gabe S. Sonke, Eva-Maria Grischke, Emilio Alba, Arnd Nusch, Denise A. Yardley, Erik Jakobsen, Sibel Blau, Sara M. Tolaney, Faye Su, Wei He, Caroline Germa, Gabriel N. Hortobagyi. Ribociclib + letrozole for first-line treatment of hormone receptor-positive (HR+), human epidermal growth factor receptor 2-negative (HER2-) advanced breast cancer (ABC): efficacy by baseline tumor markers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr CT045. doi:10.1158/1538-7445.AM2017-CT045

Characterization of potential predictive biomarkers of response to nivolumab in CheckMate 141 in patients with squamous cell carcinoma of the head and neck (SCCHN).

6050 Background: Nivolumab, an anti-programmed death-1 (PD-1) monoclonal antibody, demonstrated longer median overall survival (7.5 vs 5.1 months) and improved response (13.3% vs 5.8%) versus investigator choice chemotherapy (ICC) in patients with recurrent SCCHN after platinum failure in CheckMate 141 (NCT02105636), a randomized, open-label Phase 3 trial. We screened peripheral blood lymphocytes (PBL) to identify biomarkers which may predict response to nivolumab. Methods: Paired baseline (day 1) and on treatment (day 43) PBL samples (n = 36; 24 nivolumab; 12 ICC) were analyzed using multicolor flow cytometry and a non-competing anti-PD-1 antibody. Results were correlated with clinical outcome: responders (complete/partial response) and non-responders (stable or progressive disease). Results: Levels of CD8+ T cells at baseline and on treatment were higher in nivolumab responders compared to non-responders (23% vs 13%; P&lt; 0.05). Interestingly, PD-1+ CD8+ and PD-1+ CTLA-4+ CD8+ effector T cells (likely exhausted T cells) decreased about 2-fold following nivolumab in both responders and non-responders ( P&lt; 0.05), whereas, the decrease in CTLA-4+ CD8+ effector T cells following nivolumab was significant in responders only (8% vs 5%; P&lt; 0.05). Levels of PD-1+ TIM-3+ CD8+ effector cells decreased following nivolumab in non-responders only (11% vs 7%; P&lt; 0.05), a similar non-significant reduction was observed in responders. Levels of PD-1+ Tregs were lower in responders than non-responders at baseline (19% vs 33%; P&lt; 0.01), and following nivolumab (12% vs 20%; P&lt; 0.001). As in T-effector cell populations, PD-1+ Tregs decreased about 1.6-fold after nivolumab in both responders and non-responders ( P&lt; 0.01). Interestingly, baseline Ki67+ Treg levels were lower in non-responders (28% vs 17%; P&lt; 0.05). Conclusions: Response to nivolumab may be associated with higher levels of CD8+ T cells and CTLA-4+ CD8+ effector T cells, and lower PD-1+ CD8+ effector T cells and PD-1+ Tregs at baseline. Targeting both PD-1 and CTLA-4 axes is warranted in SCCHN to overcome suppressive signals in CD8+ effector T cells and in Treg cells expressing both checkpoint receptors. Clinical trial information: NCT02105636.

The dynamics of the oesophageal squamous epithelium 'normalisation' process in patients with gastro-oesophageal reflux disease treated with long-term acid suppression or anti-reflux surgery.

Proton pump inhibitors and laparoscopic anti-reflux surgery (LARS) offer long-term symptom control to patients with gastro-oesophageal reflux disease (GERD). To evaluate the process of 'normalisation' of the squamous epithelium morphology of the distal oesophagus on these therapies. In the LOTUS trial, 554 patients with chronic GERD were randomised to receive either esomeprazole (20-40 mg daily) or LARS. After 5 years, 372 patients remained in the study (esomeprazole, 192; LARS, 180). Biopsies were taken at the Z-line and 2 cm above, at baseline, 1, 3 and 5 years. A severity score was calculated based on: papillae elongation, basal cell hyperplasia, intercellular space dilatations and eosinophilic infiltration. The epithelial proliferative activity was assessed by Ki-67 immunohistochemistry. A gradual improvement in all variables over 5 years was noted in both groups, at both the Z-line and 2 cm above. The severity score decreased from baseline at each subsequent time point in both groups (P < 0.001, all comparisons), attaining a normal level by 5 years. Corresponding decreases in Ki-67 expression were observed (P < 0.001, all comparisons). No significant differences were found between esomeprazole treatment and LARS. Neither baseline severity score nor Ki-67 expression predicted the risk of treatment failure. Five years of treatment is generally required before squamous epithelial cell morphology and proliferation are 'normalised' in patients with chronic GERD, despite endoscopic and symptomatic disease control. Control of the acid component of the refluxate seems to play the predominant role in restoring tissue morphology.

Characterization of potential predictive biomarkers of response to nivolumab in CheckMate-141 in patients with squamous cell carcinoma of the head and neck (SCCHN).

5 Background: Nivolumab, an anti-programmed death-1 (PD-1) monoclonal antibody, demonstrated longer median overall survival (7.5 vs. 5.1 months) and improved response (13.3% vs. 5.8%) versus investigator choice chemotherapy (ICC) in patients with recurrent SCCHN after platinum failure in CheckMate 141 (NCT02105636), a randomized, open-label Phase 3 trial. We screened peripheral blood lymphocytes (PBL) to identify biomarkers which may predict response to nivolumab. Methods: Paired baseline (day 1) and on treatment (day 43) PBL samples (n=36; 24 nivolumab; 12 ICC) were analyzed using multicolor flow cytometry and a non-competing anti-PD-1 antibody. Results were correlated with clinical outcome: responders (complete/partial response) and non-responders (stable or progressive disease). Results: Levels of CD8+ T cells at baseline and on treatment were higher in nivolumab responders compared to non-responders (23% vs. 13%; p&lt;0.05). Interestingly, PD-1+ CD8+ and PD-1+ CTLA-4+ CD8+ effector T cells (likely exhausted T cells) decreased about 2-fold following nivolumab in both responders and non-responders (p&lt;0.05), whereas, the decrease in CTLA-4+ CD8+ effector T cells following nivolumab was significant in responders only (8% vs. 5%; p&lt;0.05). Levels of PD-1+ TIM-3+ CD8+ effector cells decreased following nivolumab in non-responders only (11% vs. 7%; p&lt;0.05), a similar non-significant reduction was observed in responders. Levels of PD-1+Tregs were lower in responders than non-responders at baseline (19% vs. 33%; p&lt;0.01), and following nivolumab (12% vs. 20%; p&lt;0.001). As in T effector cell populations, PD-1+ Tregs decreased about 1.6-fold after nivolumab in both responders and non-responders (p&lt;0.01). Interestingly, baseline Ki67+ Treg levels were lower in non-responders (28% vs. 17%; p&lt;0.05). Conclusions: Response to nivolumab may be associated with higher levels of CD8+ T cells and CTLA-4+ CD8+ effector T cells, and lower PD-1+ CD8+ effector T cells and PD-1+ Tregs at baseline. Targeting both PD-1 and CTLA-4 axes is warranted in SCCHN to overcome suppressive signals in CD8+ effector T cells and in Treg cells expressing both checkpoint receptors. Clinical trial information: NCT02105636.

Abstract P2-10-02: The impact of intrinsic subtypes and molecular features on aromatase inhibitor induced reduction of proliferation marker of Ki67 in primary ER+ breast cancer: A POETIC study (CRUK/07/015)

Abstract Background Neoadjuvant endocrine therapy (NAE) is often a good option for postmenopausal (PM) women with estrogen receptor positive (ER+) breast cancers(BC). Fall in Ki67 is widely accepted as valid for predicting favorable tumor response to NAE and improved outcome. We report our planned correlative study to investigate if intrinsic subtype impacts on Ki67 changes (ΔKi67) as measured by immunohistochemistry. We also explored the correlation of several ER+ BC relevant molecular features at baseline(B) with ΔKi67. Patients and methods POETIC is a phase III, randomized 2:1 study for 4486 PM patients with ER+ BC to determine whether peri-operative aromatase inhibitor (AI) followed by standard adjuvant therapy improves outcome compared with standard adjuvant therapy alone. The proliferation rate was estimated as percentage (%) of cancer cells staining for Ki67. Primary biological endpoint was defined as two-week (2wk) change in Ki67 (2wkΔKi67): ln[(2wk Ki67+0.1)/(B Ki67+0.1)]. Secondary endpoint: “responders”, was % change of Ki67 defined as (2wk Ki67 – B Ki67) *100/B Ki67. “Responder” was defined as follows: reduction &amp;lt;50% as poor (PR), 50-75% moderate and &amp;gt;75% as good responder (GR). Human whole genome expression(GE) Illumina BeadChips were performed. Data was obtained from 137 paired samples from the treatment group(T) and 49 pairs from the control(C) group with GE data passing quality check and baseline Ki67≥5% to minimise the impact of extreme values based on proportional ΔKi67. Intrinsic subtype and risk of recurrence(ROR) groups were calculated using PAM50. GE scores from Oncotype Dx, MammaPrint, p53 mutation/wildtype(Troester 2006), ER+ early response (ERE)(Hatzis 2011), estrogen-regulated genes subtypes (Oh 2006) and markers for 23 different immune cell types(Bindea 2013) were calculated. Associations of GE scores to endpoints of response were determined by Spearman correlation and chi-square tests. Bonferroni correction was used to control error rate with p&amp;lt;0.0005 deemed significant. Results At B of the 137 paired T, 64% were Luminal A (LumA), 22% Luminal B (LumB), 9% as HER-2 enriched (HER2-E), 2% as Basal-like (BLBC) and 3% as Normal-like. Subtypes at B were associated with response, with LumA showing the biggest reduction of Ki67 (p=0.0001) and GR. All GE, except ERE, correlated significantly with 2wkΔKi67 and response: higher risk groups associated with lowest reduction rate. None of immune cell types correlated with 2wkΔKi67, except that tumors enriched with T-helper 1 cell type were associated with PR (p &amp;lt; 0.000001). Comparing subtypes between time-points, 85% of LumB and 42% of HER2-E were assigned instead as LumA at 2wk regardless of response. Of the 15 ROR defined high-risk group, only 33% were assigned instead as low-risk at 2wk. Conclusion Both LumA and LumB are endocrine sensitive. A fall of Ki67 was observed in majority of cases. Most tumors estimated as high-risk by molecular profiling showed less response and most remained moderate or high risk of recurrence on endocrine therapy. Whether molecular profiling at 2wk after starting AI predicts for long-term outcome in PM women with ER+ better than at diagnosis will need to be determined. Citation Format: Cheang MCU, Morden J, Gao Q, Parker J, López-Knowles E, Detre S, Hills M, Zabaglo L, Tomiczek M, Mallon E, Robertson J, Smith I, Bliss J, Dowsett M, On Behalf of the POETIC Trialists. The impact of intrinsic subtypes and molecular features on aromatase inhibitor induced reduction of proliferation marker of Ki67 in primary ER+ breast cancer: A POETIC study (CRUK/07/015) [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P2-10-02.

Abstract S6-07: Comparison of 12 weeks neoadjuvant Nab-paclitaxel combined with carboplatinum vs. gemcitabine in triple- negative breast cancer: WSG-ADAPT TN randomized phase II trial

Abstract Background: Pathological complete response (pCR) is associated with improved prognosis in TNBC, but optimal chemotherapy remains unclear. Use of weekly nab- paclitaxel (Nab-Pac) vs. conventional paclitaxel and also addition of carboplatinum(Carbo) to anthracycline-taxane(A/T) containing chemotherapy results in significantly higher pCR rates in TNBC with unclear impact on survival and increased toxicity. The ADAPT study seeks to compare Carbo vs. gemcitabine(Gem) added to nab- paclitaxel as a short 12-week A-free regimen. It also assesses efficacy in early responders vs. non-responders by 3-week proliferation and/or imaging response. Methods: ADAPT TN compares 12-week neoadjuvant regimens: Carbo vs. Gem combined with Nab-Pac and aims to identify early-response markers for pCR (yPN0 and ypT0/is). TNBC patients (centrally confirmed ER/PR &amp;lt;1%, HER2 neg.), cT1c- cT4c, cN0/+ were randomized to arm A (Nab-Pac 125/Gem 1000 d1,8 q3w) vs. B (Nab-Pac 125/Carbo AUC2 d1,8 q3w). Randomization was stratified by center and nodal status. The trial is powered for pCR comparison by therapy arm and by presence vs. absence of early response markers. Pre-planned interim analysis aimed to identify a dynamic biomarker, e.g. drop of 3-week Ki-67, and to validate trial assumptions. Results: 336 patients were enrolled from 47 centers between 06/13-02/15 (n=182 ArmA: Nab-Pac/Gem and n=154 ArmB: Nab-Pac/Carbo). 90% and 95% completed therapy according to protocol respectively (n.s.). Median age was 50y. At baseline: A/B: 73% and 74%% had G3 tumors, median Ki-67 of 70% and 75%; 62.6% and 62.9%% had cT2-4c tumors, pN0 status prior to chemotherapy was confirmed in 50.5% and 50%, respectively. pCR (ypT0/is/ypN0) was A: 28.7% and B: 45.9% (p&amp;lt;0.001). Total pCR (ypT0/ypN0) was A: 25.8% and B: 45.2% respectively (p &amp;lt;0.001). Nab/Gem arm was associated with significantly higher frequency of dose reductions (20.6% vs. 11.9% (p=0.03), treatment related SAE's (13% vs. 5%, p=0.02), grade 3-4 infections (6.1% vs. 1.3%, p=0.04) and ALAT elevations (11.7 vs. 3.3%, p=0.01) compared to the Nab-Carbo arm. Within the planned interim analysis (n=130: A/B: 69/61), baseline Ki-67 (Nab- Pac/Carbo arm), age&amp;gt;50 years, and low cellularity (&amp;lt;500 tumor cells and/or Ki-67≤10% in the 3-week biopsy) (Nab-Pac/Gem arm) were positively associated with pCR by logistic regression analysis (separately by therapy arm). In all patients, therapy arm itself was significant for pCR. Validation of responder definitions for the whole study will be presented at the meeting. Conclusions: This is the first large randomized study comparing two short 12-week anthracycline- free regimens in unselected TNBC. Our results suggest superior efficacy and excellent toxicity of Nab-Pac/Carbo vs. Gem. Longer A/T-Carbo containing regimens render quite comparable pCR rates, thus overtreatment by 4xEC in unselected TNBC may be present in some patients. Early response criteria seem to differ according to regimen; their assessment may be impaired by substantial tumor necrosis already after the first therapy cycle. Citation Format: Gluz O, Nitz U, Liedtke C, Christgen M, Sotlar K, Grischke EM, Forstbauer H, Braun M, Warm M, Hackmann J, Uleer C, Aktas B, Schumacher C, Bangemann N, Lindner C, Kuemmel S, Clemens M, Potenberg J, Staib P, Kohls A, Pelz E, Kates RE, Wuerstlein R, Kreipe HH, Harbeck N. Comparison of 12 weeks neoadjuvant Nab-paclitaxel combined with carboplatinum vs. gemcitabine in triple- negative breast cancer: WSG-ADAPT TN randomized phase II trial. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr S6-07.

Abstract P1-13-01: Comparison of 12 weeks neoadjuvant Nab-paclitaxel combined with carboplatinum vs. gemcitabine in triple- negative breast cancer: WSG-ADAPT TN randomized phase II trial

Abstract Background: Pathological complete response (pCR) is associated with improved prognosis in TNBC, but optimal chemotherapy remains unclear. Use of weekly nab- paclitaxel (Nab-Pac) vs. conventional paclitaxel and also addition of carboplatinum(Carbo) to anthracycline-taxane(A/T) containing chemotherapy results in significantly higher pCR rates in TNBC with unclear impact on survival and increased toxicity. The ADAPT study seeks to compare Carbo vs. gemcitabine(Gem) added to nab- paclitaxel as a short 12-week A-free regimen. It also assesses efficacy in early responders vs. non-responders by 3-week proliferation and/or imaging response. Methods: ADAPT TN compares 12-week neoadjuvant regimens: Carbo vs. Gem combined with Nab-Pac and aims to identify early-response markers for pCR (yPN0 and ypT0/is). TNBC patients (centrally confirmed ER/PR &amp;lt;1%, HER2 neg.), cT1c- cT4c, cN0/+ were randomized to arm A (Nab-Pac 125/Gem 1000 d1,8 q3w) vs. B (Nab-Pac 125/Carbo AUC2 d1,8 q3w). Randomization was stratified by center and nodal status. The trial is powered for pCR comparison by therapy arm and by presence vs. absence of early response markers. Pre-planned interim analysis aimed to identify a dynamic biomarker, e.g. drop of 3-week Ki-67, and to validate trial assumptions. Results: 336 patients were enrolled from 47 centers between 06/13-02/15 (n=182 ArmA: Nab-Pac/Gem and n=154 ArmB: Nab-Pac/Carbo). 90% and 95% completed therapy according to protocol respectively (n.s.). Median age was 50y. At baseline: A/B: 73% and 74%% had G3 tumors, median Ki-67 of 70% and 75%; 62.6% and 62.9%% had cT2-4c tumors, pN0 status prior to chemotherapy was confirmed in 50.5% and 50%, respectively. pCR (ypT0/is/ypN0) was A: 28.7% and B: 45.9% (p&amp;lt;0.001). Total pCR (ypT0/ypN0) was A: 25.8% and B: 45.2% respectively (p &amp;lt;0.001). Nab/Gem arm was associated with significantly higher frequency of dose reductions (20.6% vs. 11.9% (p=0.03), treatment related SAE's (13% vs. 5%, p=0.02), grade 3-4 infections (6.1% vs. 1.3%, p=0.04) and ALAT elevations (11.7 vs. 3.3%, p=0.01) compared to the Nab-Carbo arm. Within the planned interim analysis (n=130: A/B: 69/61), baseline Ki-67 (Nab- Pac/Carbo arm), age&amp;gt;50 years, and low cellularity (&amp;lt;500 tumor cells and/or Ki-67≤10% in the 3-week biopsy) (Nab-Pac/Gem arm) were positively associated with pCR by logistic regression analysis (separately by therapy arm). In all patients, therapy arm itself was significant for pCR. Validation of responder definitions for the whole study will be presented at the meeting. Conclusions: This is the first large randomized study comparing two short 12-week anthracycline- free regimens in unselected TNBC. Our results suggest superior efficacy and excellent toxicity of Nab-Pac/Carbo vs. Gem. Longer A/T-Carbo containing regimens render quite comparable pCR rates, thus overtreatment by 4xEC in unselected TNBC may be present in some patients. Early response criteria seem to differ according to regimen; their assessment may be impaired by substantial tumor necrosis already after the first therapy cycle. Citation Format: Gluz O, Nitz U, Liedtke C, Christgen M, Sotlar K, Grischke EM, Forstbauer H, Braun M, Warm M, Hackmann J, Uleer C, Aktas B, Schumacher C, Bangemann N, Lindner C, Kuemmel S, Clemens M, Potenberg J, Staib P, Kohls A, Pelz E, Kates RE, Wuerstlein R, Kreipe HH, Harbeck N. Comparison of 12 weeks neoadjuvant Nab-paclitaxel combined with carboplatinum vs. gemcitabine in triple- negative breast cancer: WSG-ADAPT TN randomized phase II trial. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P1-13-01.

Abstract S5-03: Final analysis of WSG-ADAPT HER2+/HR+ phase II trial: Efficacy, safety, and predictive markers for 12-weeks of neoadjuvant TDM1 with or without endocrine therapy versus trastuzumab+endocrine therapy in HER2-positive hormone-receptor-positive early breast cancer

Abstract Background: In HER2+ early breast cancer (eBC) pCR rates after standard neoadjuvant chemo- + anti-HER2 therapy differ according to hormone-receptor (HR) status. Molecular analysis reveals HER2+/HR+ BC as a distinct entity within HER2+ BC. The ADAPT HER2+/HR+ phase II trial aims to identify early responders to endocrine + anti-HER2 therapy. Methods: The trial completed recruitment in January 2015 (n=376). Patients (pts.) were randomized to 12 weeks of neoadjuvant therapy: A:T-DM1 (3.6 mg/kg q3w) vs. B:T-DM1 with endocrine therapy (ET) (pre-: tamoxifen; postmenopausal: aromatase inhibitor) vs C:trastuzumab q3w+ET. After surgery, standard chemotherapy at investigators' discretion and completion of 1y trastuzumab were recommended. Trial tests pCR (yPN0 and ypT0/is) in after T-DM1 or T-DM1+ET compared to T+ET. Biomarkers are measured at baseline and after 3 weeks. Results: Pre-planned interim analysis (n=130) aimed to identify an early-response biomarker (e.g. Ki-67 drop) and to validate trial assumptions. Median age was 49 years; 55% were pre-menopausal; 40% had cT1 tumors, 51% cT2; 68% had cN0, 27% cN1; 75% had G3. Median baseline Ki67 was 30%. In all arms, more than 95% received all 4 therapy cycles. 16 SAEs were reported in 13 pts (A:7; B:6; C:3); all CTC grade 1 (1), 2 (11) or 3 (4); all pts completely recovered without sequelae. Overall pCR rate was 30.8%: T-DM1: 40.5%, T-DM1+ET: 45.8%, T+ET: 6.7%. The difference between either arm T-DM1 arm vs. T+ET was significant (p&amp;lt;0.001). Exploratory analysis suggests benefit of adding ET to T DM1 in pre- (pCR: 27.3% for T-DM1 single agent vs. 45.5% with ET) but not in postmenopausal pts (pCR: 60% vs. 46.2%). Substantial early therapy response did not permit Ki-67 quantification in the 3-week biopsy in 43.1% due to low cellularity (&amp;lt;500 tumor cells). PIK3CA mutation analysis (n=114) revealed a mutation rate of 15.8% (n=18). Overall pCR rate was 35.4% (n=96) for wildtype and 17.6% (n=18) for tumors with PIK3CA mutation. Ongoing biomarker analyses include further mutation analysis and intrinsic subtypes in the total trial collective. Conclusions: The WSG-ADAPT HER2+/HR+ phase II trial is internationally the first large prospective randomized phase II trial specifically conducted within this distinct subtype. Interim analysis demonstrated for the first time clinically meaningful pCR rates (&amp;gt;40%) after short therapy (12 weeks) of T-DM1 ± ET without systemic chemotherapy in HER2+/HR+ eBC. Final efficacy and safety data will be presented at the meeting together with results of the correlative science program. Citation Format: Harbeck N, Gluz O, Christgen M, Braun M, Kuemmel S, Schumacher C, Potenberg J, Kraemer S, Kleine-Tebbe A, Augustin D, Aktas B, Forstbauer H, Tio J, Liedtke C, Kates RE, Wuerstlein R, de Haas SL, Kiermaier A, Kreipe HH, Nitz U. Final analysis of WSG-ADAPT HER2+/HR+ phase II trial: Efficacy, safety, and predictive markers for 12-weeks of neoadjuvant TDM1 with or without endocrine therapy versus trastuzumab+endocrine therapy in HER2-positive hormone-receptor-positive early breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr S5-03.

Abstract P5-13-03: Fulvestrant plus anastrozole as neoadjuvant therapy in postmenopausal women with hormone receptor positive early breast cancer

Abstract Background: Aromatase inhibitors (AIs) are effective in reducing the risk of recurrence from breast cancer (BC) but 20% of patients (pts) with early BC still recur despite adjuvant AIs. Thus more effective endocrine therapies (HTs) are needed. In metastatic BC (MBC), combination of lower dose fulvestrant plus anastrozole improves survival compared to anastrozole alone. The 21-gene Recurrence Score® (RS; Oncotype DX®) has been validated to predict benefit from adding chemotherapy (CT) to HT where pts with a low score have little benefit from CT and derive a large benefit from HT. Ki-67 response to neo-adjuvant HT may predict adjuvant outcomes to HT. Postoperative Endocrine Prognostic Index (PEPI) and modified PEPI may further identify a subset of HT sensitive cancers that do not require adjuvant CT (PEPI 0 category). We conducted a single arm phase II trial to assess the efficacy of fulvestrant plus anastrozole as neoadjuvant HT in pts with operable BC. Methods: Postmenopausal pts with stage II and III, ER/PR+, HER2 (-) BC with a RS&amp;lt;25 (performed on initial core bx) were included. Duration of neo-adjuvant HT was 4 months. Pts received anastrozole 1mg (PO) daily continuously from day 1 until surgery + fulvestrant (IM) 500mg on day 1, 14 and 28 of cycle 1, and on the last day of three subsequent 28 day cycles (total 6 doses of fulvestrant). At week 4, an optional core bx was repeated to assess change in Ki-67. Response assessments were made clinically every 4 wks. All pts had breast/axillary surgery after the 6th dose of fulvestrant. Ki-67, histologic grade, ER/PR status, and RS were assessed at baseline, core bx at 4 wks, and at definitive surgery. Primary end points were pathologic complete response (pCR) rate and change in Ki-67. Adjuvant CT was left to the discretion of treating physician. Results: 42 pts were enrolled 7/2009 to 11/2014. Median age was 62. 32 (76%) patients had stage IIA, 7 (17%) had stage IIB and 3 (7%) had stage III disease. 14% had clinically node positive disease. The median RS was 12 (0-24). Median tumor size was 3.5cm. 21%, 74%, and 5% had grade 1, 2 and 3 tumors respectively. Mean ER expression was 95%. 16 (38%) pts had a clinical complete response (cCR), 13 (31%) had a clinical partial response (cPR) and 12 (29%) had stable disease. One pt had progression on therapy. There were no pCRs. Median baseline Ki-67 was 5% (1-36%). 94% of pts had decrease in Ki-67 from baseline to 4-week bx and 97% of pts had decrease in Ki-67 from baseline to surgery. Modified PEPI score at surgery was 0 in 53% of patients. 78% of pts did not receive adjuvant CT. At median follow up of 38 mos only 1 pt had a recurrence with 98% free of a recurrence. There were no grade 3 or grade 4 toxicities. Conclusions: The neoadjuvant combination of anastrozole and fulvestrant in pts with RS&amp;lt;25 markedly improves Ki-67 response with more than half of pts achieving a modified PEPI score of 0 at surgery. At a relatively short median follow up, recurrence rate is very low. Given the efficacy and tolerability of anastrozole plus fulvestrant in MBC and now in the neo-adjuvant setting, an adjuvant trial of this combination is warranted in pts with ER+ BC. Citation Format: Khan QJ, Barr JA, Britt AS, Kimler BF, Connor CS, McGinness M, Mammen JMV, Wagner JL, Amin A, Springer M, Baccaray S, Fabian CJ, Sing AP, Sharma P. Fulvestrant plus anastrozole as neoadjuvant therapy in postmenopausal women with hormone receptor positive early breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P5-13-03.

Gene transcript analysis blood values correlate with ⁶⁸Ga-DOTA-somatostatin analog (SSA) PET/CT imaging in neuroendocrine tumors and can define disease status.

Precise determination of neuroendocrine tumor (NET) disease status and response to therapy remains a rate-limiting concern for disease management. This reflects limitations in biomarker specificity and resolution capacity of imaging. In order to evaluate biomarker precision and identify if combinatorial blood molecular markers and imaging could provide added diagnostic value, we assessed the concordance between (68)Ga-somatostatin analog (SSA) positron emission tomography (PET), circulating NET gene transcripts (NETest), chromogranin A (CgA), and Ki-67 in NETs. We utilized two independent patient groups with positive (68)Ga-SSA PET: data set 1 ((68)Ga-SSA PETs undertaken for peptide receptor radionuclide therapy (PRRT), as primary or salvage treatment, n = 27) and data set 2 ((68)Ga-SSA PETs performed in patients referred for initial disease staging or restaging after various therapies, n = 22). We examined the maximum standardized uptake value (SUVmax), circulating gene transcripts, CgA levels, and baseline Ki-67. Regression analyses, generalized linear modeling, and receiver-operating characteristic (ROC) analyses were undertaken to determine the strength of the relationships. SUVmax measured in two centers were mathematically evaluated (regression modeling) and determined to be comparable. Of 49 patients, 47 (96 %) exhibited a positive NETest. Twenty-six (54 %) had elevated CgA (χ(2) = 20.1, p < 2.5×10(-6)). The majority (78 %) had Ki-67 < 20 %. Gene transcript scores were predictive of imaging with >95 % concordance and significantly correlated with SUVmax (R (2) = 0.31, root-mean-square error = 9.4). The genes MORF4L2 and somatostatin receptors SSTR1, 3, and 5 exhibited the highest correlation with SUVmax. Progressive disease was identified by elevated levels of a quotient of MORF4L2 expression and SUVmax [ROC-derived AUC (R (2) = 0.7, p < 0.05)]. No statistical relationship was identified between CgA and Ki-67 and no relationship with imaging parameters was evident. (68)Ga-SSA PET imaging parameters (SUVmax) correlated with a circulating NET transcript signature. Disease status could be predicted by an elevated quotient of gene expression (MORF4L2) and SUVmax. These observations provide the basis for further exploration of strategies that combine imaging parameters and disease-specific molecular data for the improvement of NET management.

Abstract S2-03: Preoperative window of opportunity study of the PI3K inhibitor pictilisib (GDC-0941) plus anastrozole vs anastrozole alone in patients with ER+, HER2-negative operable breast cancer (OPPORTUNE study)

Abstract Background: Preclinical and clinical data support a key role for the PI3K pathway in resistance to endocrine therapy in patients with ER+ breast cancer and suggest that combining PI3K inhibitors with endocrine therapy may overcome resistance. Short-term preoperative window of opportunity (WOO) studies are a validated strategy for novel treatments to provide proof-of-concept and define the most appropriate patient population by directly assessing treatment effects in tumour tissue before and after treatment. This is the first WOO study with the PI3K inhibitor pictilisib (GDC-0941) in combination with anastrozole (ANA). Methods: 73 postmenopausal patients (pts) have been randomized (2:1 in favour of the combination) to receive 2-week preoperative treatment with ANA plus pictilisib (n=50, "ANA+PIC" arm) or ANA alone (n=23, "ANA" arm). Pts had newly diagnosed, operable, ER+, HER2-negative breast cancer of ≥1 cm size. Pts receiving HRT were excluded. Treatment effects and correlative studies were assessed using FFPE and frozen tumour biopsies taken before and after 14 days of study treatment. The primary endpoint was inhibition of tumour-cell proliferation, as measured by change in Ki67 expression, determined centrally by 2 investigators. Secondary endpoints include induction of apoptosis (Caspase3) and safety. Comprehensive biomarkers analyses include targeted NGS of a comprehensive cancer panel of &amp;gt;400 genes, copy number analyses, and pre- and post-treatment reverse-phase protein arrays (RPPA) and RNA profiling. Results: Baseline (BL) disease characteristics were similar between both study arms. PAM50 analysis showed that 53% and 47% of tumors were Luminal (Lum) A and B, respectively. 65% of tumors had &amp;gt;14% Ki67-positive cells. Observed treatment-emergent AEs were consistent with those previously described for single-agent pictilisib and anastrozole. Mean post-treatment percentage reduction of Ki67 was 84% (95% CI, 75%-89%) for ANA+PIC and 72% (54%-87%) for ANA. Ki67-response (≥50% drop in % of Ki67+ cells) was 86% for ANA+PIC and 60% for ANA. By using the definition that pts with a natural logarithm of %Ki67+ cells of ≤1 or 1-2 have a day 15 anti-proliferative response, 93% [ln(ki67): &amp;lt;1, 46%; 1-2, 46%] of ANA+PIC were responders compared with 60% [&amp;lt;1, 47%; 1-2, 13%] of ANA-treated pts (p = 0.01). Preplanned subgroup analyses showed a significant interaction of response to ANA+PIC with molecular subtype and Ki67 levels. Patients with LumB tumors or high BL Ki67 (&amp;gt;14%) had a higher Ki67 response with ANA+PIC compared to ANA (LumB, 83% vs 38%; Ki67&amp;gt;14%, 94% vs 55%), whereas Ki67 response was similar for both treatments for LumA tumors (ANA, 75%; ANA+PIC, 73%) or tumors with low BL Ki67. Mean post-treatment % reduction of Ki67 in LumB tumors was 87% (95% CI, 49%-96%) for ANA+PIC and 56% (16%-77%) for ANA (p=0.03). Additional data on apoptosis and comprehensive pre- and post-treatment biomarkers analyses will be presented. Conclusions: This first report of a preoperative WOO study evaluating a PI3K inhibitor in early breast cancer demonstrated addition of pictilisib to ANA was associated with increased anti-proliferative response over single-agent ANA. Citation Format: Peter Schmid, Sarah E Pinder, Duncan Wheatley, Jane Macaskill, Charles Zammit, Jennifer Hu, Robert Price, Nigel Bundred, Sirwan Hadad, Alice Shia, Louise Lim, Shah-Jalal Sarker, Patrycja Gazinska, Natalie Woodman, Darren Korbie, Matt Trau, Paul Mainwaring, Peter Parker, Arnie Purushotham, Alastair M Thompson. Preoperative window of opportunity study of the PI3K inhibitor pictilisib (GDC-0941) plus anastrozole vs anastrozole alone in patients with ER+, HER2-negative operable breast cancer (OPPORTUNE study) [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr S2-03.

Abstract P4-15-01: Distinct early proliferation response to neoadjuvant anti-HER2 antibody drug conjugate +/- endocrine therapy in early breast cancer in the WSG ADAPT HER2+/HR+ trial

Abstract Background: HER2+, hormone receptor-positive (HR+) breast cancer is a distinct subtype associated with good prognosis but poor response to standard chemotherapy + anti-HER2 (single or dual blockade). Substantial overtreatment by poly-chemotherapy combined with anti-HER2 therapy is suspected in this subtype. Yet, the efficacy of combining endocrine therapy (ET) with anti-HER2 therapy or novel antibody drug conjugates like T DM1 without systemic chemotherapy remains unclear. Methods: ADAPT HER2+/HR+ is a phase II, randomized, neoadjuvant, 3-arm trial (12 weeks) in patients with cT1c-cT3, cN0/+ HER2+, ER+ and/or PR+ early BC. Arm A: T-DM1 (3.6 mg/kg) alone; Arm B: T DM1 + ET (premenopausal: tamoxifen, postmenopausal: AI); Arm C: trastuzumab + ET. Postoperative chemotherapy is recommended together with completion of one year of trastuzumab. Initial and serial core biopsies were obtained prior to therapy and after 3 weeks. First translational analysis of the trial run-in phase (n=130) focuses on dynamics of HER2, Ki67, ER and PR. Results: 162 tumors, HR+ and HER2+ by local pathology, were screened; n=130 were HR+ and HER2+ by central pathology and randomized at 40 trial sites in Germany between 11/2012 and 03/2014 (Arm A/B/C: 37/49/44). Median age was 49 years; 60% were cT2-3, 32% cN+, 75% central G3; median baseline Ki67 was 30%; 49 patients were treated by TAM and 44 postmenopausal patients by AI in ET containing arms. Three-week core biopsies were available in 117 patients (arm A/B/C: n=33/43/41), n=99 with invasive tumor tissue (61/76 (80%) in T-DM1 containing arms and 38/41(93%) in trastuzumab + ET arm). Three-week Ki67 could only be analyzed in n=73 (53% of patients in T-DM1-containing arms, 81% in the T+ET arm) due to a lacking amount of cells for counting (&amp;lt;500). Median fractional decrease in proliferation (Ki67) after 3 weeks of therapy was 40% in the T-DM1 + ET arm (B) as compared to 14% and 25% in the T-DM1 (A) and T+ET (C) arms, respectively. Among postmenopausal patients, the contrast (52% (B) vs. 0% (A) and 28% (B)) was significant. Mean PR expression change (absolute, measured in %) was -15 in the B Arm vs. 7 and -7 in the A and C arms, respectively (p=0.04) – particularly in postmenopausal women ( 25 vs. +13 and -10, respectively, p=0.04). Baseline ER expression was positively associated with early proliferation response (fractional Ki67 decrease), e.g., by logistic regression using 30% decrease as response criterion. Data on the impact of gene mutations and further molecular markers on proliferation response will be available for presentation at the meeting. Conclusions: In the unique neoadjuvant ADAPT HER2+/HR+ trial, the combination of T-DM1 with ET (particularly AI) seems to be associated with a strong early proliferation response and PR expression drop in anti-HER2 antibody (drug conjugate) +/- ET, without systemic pre-operative chemotherapy. Interim analysis is scheduled after 130 completely treated patients. The high percentage of non-invasive tissue biopsied after 3 weeks in the T-DM1 arms is intriguing, as it would be consistent with higher pathological complete response rates. Citation Format: Oleg Gluz, Ulrike Nitz, Kuemmel Sherko, Kraemer Stefan, Michael Braun, Claudia Schumacher, Bahriye Aktas, Helmut Forstbauer, Toralf Reimer, Peter Fasching, Jochem Potenberg, Daniel Hofmann, Ronald E Kates, Rachel Wuerstlein, Matthias Christgen, Hans H Kreipe, Nadia Harbeck. Distinct early proliferation response to neoadjuvant anti-HER2 antibody drug conjugate +/- endocrine therapy in early breast cancer in the WSG ADAPT HER2+/HR+ trial [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P4-15-01.

Abstract P6-11-04: OOTR-N007: A phase II neoadjuvant study of letrozole plus palbociclib in postmenopausal patients with ER positive, HER2 negative breast cancer

Abstract Background : CDK4-associated kinase activity is required to maintain breast tumorigenesis. Virtually all ER-positive cell lines harbour loss of p16ink4a. Low expression of CDK inhibitors p21 and p27 and high expression level of cyclin E and D1 have all been associated with resistance to anti-estrogen therapy. In preclinical models, Rb deficiency is associated with resistance to antiestrogen therapy. Palbociclib (PD 0332991, Pfizer Inc.) is an oral, potent, highly selective reversible inhibitor of CDK4/6 that prevents cellular DNA synthesis by prohibiting progression from G1 into the S phase. Palbociclib in combination with letrozole has shown promising activity in metastatic setting. Methods : In an open-label, multi-center, single arm pilot study efficacy and safety of neo-adjuvant palbociclib 100 mg QD for 3 weeks plus letrozole 2.5 mg QD in 4 week cycles for 4 months was studied. Postmenopausal patients of mostly Chinese ancestry, with histologically confirmed ER+, HER2- invasive breast cancer and tumor greater than 2 cm were registered. Patients with T3N1, T4 or any N2, N3 were excluded. Aim : The primary endpoint is Objective Response Rate (ORR). The secondary end-point include Pathologic Response Rate (PRR), Disease-Free Survival (DFS), safety. Exploratory analysis of gene and protein expression in tumors and serial whole blood is planned. Results : As of June 2014, 11 patients were recruited. 9 patients completed treatment and 2 are still on study. Of the 9 patients that completed study at time of this abstract, 1 patient had a complete pathological response (pCR) and 7 had a partial response (ORR of 89%). Baseline Ki67 levels were low (median 18%), consistent with luminal type A disease. On average, patients underwent surgery 24 days after completion of the study. Ki67 was measured at baseline, cycle 1 day 15, and at the time of surgery. Median Ki67 at time of surgery was 10%. There were no significant changes in Ki67 that could be related to treatment outcome based on this preliminary data analysis in surgical samples only. In the first 8 patients, whom all started at a dose of 125mg palbociclib, 4 patients developed grade 3/4 neutropenia in the absence of fever. Neutropenia was well manageable with G-CSF and/or dose modification. Protocol amendment allowed starting dose of 100 mg with dose titration to 125mg after the first cycle. 3 patients started at 100mg and grade 4 neutropenia without fever occurred in 1 subject. Another common side effect was low grade mucositis (n=5). Conclusion : Based on these initial results, the addition of palbociclib to neo-adjuvant letrozole appears to be safe and effective. Historically, ORR in this patient population with letrozole alone is below 55%. The addition of palbociclib increased the preliminary ORR to 89%. One patient had a pCR, uncommon for patients on neo-adjuvant endocrine therapy. Complete safety and efficacy data will be included for at least 11 patients, including initial biomarker results. The study continues enrolment up to 45 patients. Citation Format: Louis WC Chow, Chi-Kei Lam, Wings TY Loo. OOTR-N007: A phase II neoadjuvant study of letrozole plus palbociclib in postmenopausal patients with ER positive, HER2 negative breast cancer [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P6-11-04.

A presurgical window-of-opportunity study of metformin in obesity-driven endometrial cancer

BackgroundMetformin use is associated with reduced cancer risk in several observational studies of patients with type 2 diabetes. Results from preclinical studies in endometrial cancer show that metformin reduces cellular proliferation by inhibition of the PI3K-AKT-mTOR pathway. We tested the hypothesis that metformin would reduce cellular proliferation in vivo in atypical endometrial hyperplasia and endometrial endometrioid adenocarcinoma. MethodsWe recruited women attending gynaecological oncology clinics in Manchester, UK, with atypical endometrial hyperplasia or endometrial endometrioid adenocarcinoma. Women received metformin (850 mg twice daily) or no drug (control) during the 1–4 week presurgical window between cancer diagnosis and hysterectomy according to patient preference. Paired blood and tumour samples were obtained at recruitment and hysterectomy. Cellular proliferation was assessed by Ki-67 proliferation index. Automated scoring on two separate occasions provided consistent replicate scores (SD <10%). This study is registered with the ISRCTN register, number ISRCTN81570194. FindingsSamples from 40 women have been analysed (28 metformin-treated [median age 64 years, IQR 58–69]; 12 control [70, 64–70]). 24 of the patients (60%) were obese. 22 patients (55%) had either undiagnosed diabetes (fasting glucose >7·0 mmol/L, n=4) or insulin resistance (homoeostatic model assessment of insulin resistance >2·8, n=18). Metformin was taken for a median of 20 days (IQR 17–24), and mild gastrointestinal side-effects were reported by 22 metformin-treated patients. In the metformin-treated group, Ki-67 was 12·9% lower at hysterectomy than at recruitment (95% CI 3·7–22·1, p=0·008) after adjustment for baseline Ki-67, Ki-67 change in controls, age, and body-mass index. No significant changes in phosphorylation of AKT or markers of insulin resistance after adjustment for treatment arm were seen. InterpretationUndiagnosed insulin resistance or diabetes were common in our study population. Short-term presurgical metformin was associated with a reduction in Ki-67 proliferation index. We are now exploring the hypothesis that metformin reduces Ki-67 expression by inducing phosphorylation of AMP-activated kinase and subsequent mTOR proproliferative inhibition, independent of insulin and insulin-like growth factor receptor activation. FundingWellbeing of Women, Wellcome Trust.

Clinical Activity of Duvelisib (IPI-145), a Phosphoinositide-3-Kinase-δ,γ Inhibitor, in Patients Previously Treated with Ibrutinib

Introduction: Duvelisib (IPI-145) belongs to an emerging class of therapeutic small molecule kinase inhibitors that target B-cell receptor (BCR) signaling pathways important in various lymphoproliferative disorders. Duvelisib, a novel oral inhibitor of PI3K-δ,γ, has shown clinical activity across a range of hematologic malignancies in an ongoing Phase 1 study.Some patients (pts) treated with ibrutinib (IBR), a BCR inhibitor that targets Bruton’s tyrosine kinase (BTK), have had disease progression and various mechanisms of IBR resistance have been characterized. Since the mechanism of action of duvelisib differs from IBR, duvelisib was evaluated in a subset of pts previously treated with IBR enrolled in an ongoing Phase I study.Methods: The study was designed to evaluate the safety, maximum tolerated dose, pharmacokinetics, and activity of orally administered duvelisib twice daily in 28-day cycles. Pts with relapsed/refractory (R/R) hematologic malignancies were enrolled, including pts previously treated with IBR with R/R chronic lymphocytic leukemia (CLL) and aggressive B-cell NHL (aNHL including DLBCL and Richter’s transformation [RT]). Pharmacodynamic studies in CLL pts with measurable disease in the peripheral blood (PB) included flow cytometry to evaluate whether duvelisib inhibits the phosphorylation of AKT (pAKT) at S473 and the CLL cell proliferation index via measurement of Ki67. DNA sequencing on malignant PB cells was performed to determine the mutation status of BTK and other genes involved in hematologic malignancies. Clinical responses were based on IWCLL (2008) criteria and revised IWG (2007) criteria as applicable.Results: Twelve pts previously treated with IBR were enrolled (R/R CLL, n=6; aNHL, n=6 [DLBCL=2; RT=4]) and received IPI-145 at 25 mg BID (n=2) or 75 mg BID (n=10). The median age in R/R CLL pts was 58 years (y) (42-76), 67% were male, 100% had ≥3 prior systemic therapies, the median time from prior therapy to first dose of duvelisib was 0.34 months (mo) (0.0-1.6), and 67% started duvelisib within 2 wks of last dose of IBR. The median age in R/R aNHL pts was 62 y (36-81), 67% were male, 50% had ≥3 prior systemic therapies, a median time from last prior therapy to first dose of duvelisib of 0.74 months (0.2-3.7), and 50% started duvelisib within 3 weeks of last dose of IBR.With a median of 4.1 treatment cycles (range 3.0-9.2) in the R/R CLL pts and 2.5 treatment cycles (range 1.8-5.4) in R/R aNHL pts, the safety profile in these pts previously treated with IBR appears consistent with all pts (n=206) with advanced hematologic malignancies receiving duvelisib. The best response observed in R/R CLL pts includes 1 partial response (PR) and 5 stable disease (SD). Of these 6 pts, 2 pts (1 PR, 1 SD) remain on duvelisib for 8 and 9 months, respectively, and 4 pts have discontinued treatment (progressive disease [PD], n=3; physician decision, n=1). The best response observed in R/R aNHL pts included 2 PRs, 1 SD, 2 PD, and 1 not evaluated [NE]. Three pts remain on duvelisib (1 PR, 1 SD, 1 NE) for 3-5 mo and 3 pts have discontinued treatment (progressive disease, n=2; AE, n=1).The median baseline proliferative index (Ki67) in R/R CLL pts previously treated with IBR (n=5) was substantially higher (22.2% vs 3.8%) than in non-IBR exposed R/R CLL pts (n=23). Across both R/R CLL populations, Ki67 was reduced with duvelisib treatment by Cycle 2 Day 1 (at steady state); however, the effect was more pronounced in non-IBR treated pts. Despite this apparent difference in baseline proliferative index, the IBR previously treated pts with measureable disease in PB (R/R CLL=5), including 3 pts with a known IBR-resistance mutation in BTK (2 with C481S, 1 with C481F) demonstrated reductions of pAKT that were observed within 1 hour of duvelisib treatment and sustained through 24 hours postdose. This pAKT pharmacodynamic response was consistent with results in R/R CLL pts not previously treated with IBR (n=35).Conclusions: Pharmacodynamic data demonstrates duvelisib inhibits pAKT in R/R CLL pts, including those with an IBR-resistance mutation in BTK, and is consistent with duvelisib having a non-overlapping MOA with IBR. Baseline Ki67 data suggest a more aggressive clinical phenotype in R/R CLL pts who progress on IBR compared to those without previous IBR treatment. Early evidence of clinical activity and a tolerable safety profile suggest additional studies of duvelisib in pts who have progressed on IBR are warranted. DisclosuresPorcu:Actelion (e), Cutaneous Lymphoma Foundation (h), United States Cutaneous Lymphoma Consortium (h), Infinity (d), Celgene (d), : Consultancy, Employment, Honoraria, Membership on an entity’s Board of Directors or advisory committees, Research Funding. Off Label Use: IPI-145. Flinn:Infinity Pharmaceuticals: Consultancy. Kahl:Infinity Pharmaceuticals: Consultancy, Research Funding. Horwitz:Research: Celgene, Millennium, Infinity, Kiowa-Kirin, Seattle Genetics, Spectrum-Consulting: Amgen, Bristol-Myers Squibb, Celgene, Jannsen, Millennium, seattle genetics: Consultancy, Honoraria, Research Funding. Oki:Infinity Pharmaceuticals: Research Funding. Byrd:Pharmacyclics, Genentech: Research Funding. Sweeney:Infinity Pharmaceuticals: Employment. Allen:Infinity Pharmaceuticals: Employment. Faia:Infinity Pharmaceuticals, Inc.: Employment. Ni:Infinity Pharmaceuticals: Employment. Stern:Infinity Pharmaceuticals, Inc.: Employment. Kelly:Infinity Pharmaceuticals: Employment.

Prognostic significance of Ki-67 index value at the primary breast tumor in recurrent breast cancer.

The Ki-67 index value is a prognostic factor in primary breast cancer and is a proliferation marker that also distinguishes between luminal type A and type B breast cancer. Moreover, a change in Ki-67 index values due to treatment and recurrence is considered to be important in treating breast cancer. In this study, we investigated whether the baseline Ki-67 value in the primary tumor is useful as a prognostic factor following disease recurrence. Immunohistochemical analysis of the Ki-67 index was performed on 4,701 patients with primary breast cancer from 1987 until March, 2013. Among these patients, there were 666 consecutive cases exhibiting recurrence after primary surgery. The fraction of proliferating cells was based on a count of at least 500 tumor cells in the area including the hot spot. The Ki-67 values were divided into 3 groups, namely <20, ≥20 and ≥50%. The investigated items included estrogen receptor (ER), progesterone receptor (PgR), human epidermal growth factor receptor 2 (HER2), tumor size, nodal status for the primary tumor, recurrence site (soft tissue, bone and viscera) and disease-free interval (DFI). The Cox's proportional hazard model was used to perform univariate and multivariate analyses of the factors associated with overall survival (OS) following recurrence. The median follow-up period was 65.9 months in the surviving group. The median Ki-67 value at baseline was 20% in all the cases and 27% in the recurrent cases. The Ki-67 values were low (24%) in patients with bone metastasis and significantly higher in patients with liver or brain metastasis (38 and 55%, respectively). Moreover, DFI was found to be inversely correlated with the Ki-67 values. Univariate analysis was performed to identify the prognostic factors for OS after recurrence. The significant factors included tumor size, lymph node status, ER, PgR, DFI, recurrence site and Ki-67 index value. Among these factors, a multivariate analysis identified the Ki-67 index value in the primary tumor as an independent significant factor, particularly in luminal type tumors. The Ki-67 index value in the primary tumor was a significant prognostic factor for OS after disease recurrence. It is, therefore, important to take the Ki-67 index value into consideration for the treatment and follow-up of breast cancer patients.

Abstract P4-01-07: Evaluation of FLT PET-CT as an imaging biomarker of proliferation in primary breast cancer

Abstract Background [18F]flurothymidine (FLT) is proposed as an positron emission tomography (PET) imaging biomarker of proliferation for breast cancer. The aim of this prospective study was to assess the feasibility of FLT PET-CT as a technique for predicting response to neoadjuvant chemotherapy (NAC) in operable breast cancer and to compare baseline FLT to Ki-67. Methods 20 patients with primary breast cancer were recruited into this study and 19 received NAC with FEC (n = 6) or FEC-T (n = 13). A baseline FLT PET-CT scan was performed and repeated before the second cycle of chemotherapy. Expression of Ki-67 in the diagnostic biopsy was quantified after being stained by Dako Ki67 MIB-1 antibody and% positive cells scored. From the FLT PET-CT scans standardized uptake value maximum (SUVmax) were calculated. Results All 20 patients completed the baseline scan and 17 completed the second scan. Baseline Ki-67 results were available for 19 patients. Median age was 52 years (range 32 to 67), 12 patients were ER/ PR +ve, 6 Her-2 +ve and 6 triple negative. Mean baseline SUVmax was 7.3 (range 2.92 to 13.87) and 4.62 (range 1.79 to 14.15) post 1 cycle of NAC (range 7 to 14 days), representing a drop of 2.68 (36.3%). Mean baseline Ki-67 was 32.23 (range 2.3 to 68.4). Pearsons correlation showed a significant correlation between pre-chemotherapy Ki-67 and SUVmax of 0.604 (p = 0.006). 7 out of 17 (41%) patients achieved near pCR or pCR (pathological responders) after NAC and all of these had a reduction in SUVmax with a mean percentage value of –36.5% (range -0.5% to -62.5%). 10 out of 17 (59%) patients were non-responders and 8 of these had a reduction in SUVmax with a mean percentage value of -36.2% (range 11.3% to -70.7). Conclusions Baseline SUVmax measurements of FLT PET-CT were significantly related to Ki-67 suggesting that it is a proliferation biomarker. However, in this series neither the baseline value or the change in SUVmax after one cycle of NAC were able to predict response as most patients had a sizeable SUVmax reduction. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-01-07.

Abstract P2-16-04: A phase 2 study of ridaforolimus (RIDA) and dalotuzumab (DALO) in estrogen receptor positive (ER+) breast cancer

Abstract Background: Preclinical studies indicate that the dual inhibition of IGFR and mTOR may be additive or synergistic and abrogates the feedback activation of AKT due to rapamycin analog mTOR inhibitors. A phase 1 study of the mTOR inhibitor RIDA and the anti-IGFR antibody DALO demonstrated that the combination was feasible and well-tolerated at doses that were nearly those used for the two single agents. The dose limiting toxicity was stomatitis, similar to RIDA monotherapy. Preliminary signals of anti-tumor activity, including partial responses and prolonged progression free survival (PFS), were observed in ER+ advanced breast cancer (ABC), especially in high proliferation tumors (Ki67 ≥15%). Methods: The trial was a multi-center, international randomized study with PFS as the primary endpoint. Key eligibility included ABC with prior treatment with a non-steroidal aromatase inhibitor. The original phase 2 study design was a two-part, adaptive design intended to first test the combination of RIDA (30 mg by mouth daily for 5 out of every 7 days), -DALO (10 mg/kg IV weekly) against a standard agent, exemestane in Part A. Patients were stratified into high and low proliferation strata based on baseline Ki67. Following a demonstration of PFS benefit of the combination in Part A, Part B was intended to show the PFS benefit of the combination over each single agents by comparing RIDA-DALO to RIDA and DALO. Results: The study was initiated in October 2011. Accrual was suspended after the first 66 patients were randomized due to a higher than expected rate of stomatitis in the RIDA-DALO arm. Preliminary data indicated an overall incidence of any grade stomatitis was 68% (22/33 pts), and of grade 3 stomatitis was 35% (11/33 pts). In an effort to identify a more tolerable regimen, the study was amended to eliminate Part B and to evaluate two sequential reduced dose RIDA-DALO cohorts in a non-randomized design: 20mg and 10mg for 5 out of every 7 days. The dose of DALO was unchanged. Preliminary safety results of overall and grade 3 stomatitis in the 20 mg were 81.5% (22/27 pts) and 37% (10/27 pts), respectively. Although the incidence of overall stomatitis in the 10 mg cohort remained high, 88% (22/25 pts), grade 3 stomatitis was dramatically reduced to 8% (2/25 pts). Conclusion: Preliminary evaluation of safety from this phase 2 study demonstrates that the previously recommended phase 2 dose of RIDA-DALO was not tolerable. However lower doses of RIDA (10 mg) in combination with DALO appeared to be tolerable with markedly reduced rates of grade 3 stomatitis. Final results of efficacy, safety and RNA profiling analysis from the two RIDA-DALO dose cohorts as well as from the randomized portion of the study will be available at the time of the meeting. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P2-16-04.

Abstract PD3-7: Disease-free survival and Ki67 analysis of a randomized controlled trial comparing zoledronic acid plus chemotherapy with chemotherapy alone as a neoadjuvant treatment in patients with HER2-negative primary breast cancer (JONIE-1 study)

Abstract Background: Zoledronic acid (ZOL) has been found to have a synergistic anti-proliferative effect when used in combination with antitumor drugs. We suggested that the addition of ZOL to neoadjuvant chemotherapy (CT) has potential anti-cancer benefit in postmenopausal patients with triple-negative breast cancer in JONIE-1 Trial (50% pCR rate in ZOL+CT: CTZ versus 0% in CT, p = 0.029). We analyzed the disease-free survival (DFS) as a secondary endpoint and baseline Ki67 levels between two groups. Methods: Women with Stage IIA-IIIB HER-2-negative breast cancer were randomly assigned 1:1 to CTZ group or CT group, CT was FEC100 q3w × 4 cycles followed by weekly paclitaxel for 12 cycles and ZOL 4mg was administered every 3-4 weeks. Among 188 patients recruited between March 2010 and April 2012 excluding 10 from the primary assessment, 178 patients were assessed. The aims of this study were to compare the relative efficacy or CTZ with the efficacy of CT in prolonging DFS in all patients and also to compare the pCR rates between baseline Ki67 high (20% and &amp;gt;20%) with Ki67 low (&amp;lt;20%) in ER positive cohort. Results: During a mean follow-up period of 34.4 months, breast cancer specific events (recurrence and death) occurred 17 participants, 7 in CTZ group and 10 in CT group. The 1-year, 2-year, and 3-year DFS rates were 97.7%, 88.4%, and 88.4% in CTZ versus 100%, 84.8%, and 81.1% in CT, respectively. In the ER positive cohort studied for Ki67 consist of 109 patients, 40 were in Ki67 low group and 69 were in Ki67 high group. Among Ki67 low group, number of pCR was none out of 18 in CTZ and 1 out of 22 in CT (p = .550). Among Ki high group, that was 6 out of 38 in CTZ and 3 out of 31 (p = .352). Conclusion: We could not find a modest improvement in disease-free survival compared the addition of ZOL to neoadjuvant CT with CT alone. Ki67 study for central analysis has been under investigation. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr PD3-7.