Background

Defining the optimal dosing interval of bone-targeted agents (BTAs), such as denosumab (DN) and bisphosphonates, for patients (pts) with bone metastases remains an important clinical question. We performed a pragmatic randomised trial comparing the non-inferiority of 12- vs. 4-weekly BTAs in pts with breast and prostate cancer. We present data from the breast cohort.

Methods

Breast cancer pts who were either BTA-naïve, or already receiving, DN, zoledronate (ZA) or pamidronate (PAM) were eligible. They were randomised to their chosen BTA every 12- or 4-weeks for one year. The primary endpoint was Health Related Quality of Life (HRQL) (EORTC-QLQ-C30 Functional Domain-Physical Subdomain). Secondary endpoints included: pain (EORTC-QLQ-BM22-pain domain), Global Health Status (EORTC-QLQ-C30), symptomatic skeletal event (SSE) rates and SSE-free survival (composite of time to first SSE and time to death). Adverse events and toxicity were also compared.

Results

Of 160 breast cancer pts, 79 (49.4%) were randomised to 12- and 81 (50.6%) to 4-weekly therapy. 64 pts (40%) were BTA naïve. BTAs included; DN (n = 60, 37.5%), ZA (n = 48, 30%) or PAM (n = 52, 32.5%). Study-reported outcomes showed no significant difference in change from baseline in: HRQL-physical domain (median [range]: 0 [-87,20] vs. 0 [-66,53]), pain (median [range]: 0 [-80,33] vs. 0 [-27,20]), Global Health Status (median [range]: 0 [-67,50] vs. 0 [-50,42]) between the 12- and 4-weekly arms, respectively. Five (6%) and 7 (9%) pts had an SSE within 2 years and 1-year SSE-free survival was 65% (95% CI:52-75%) and 80% (69-88%) in the 12- and 4-weekly arms, respectively. Changes in dosing schedules were more common in the 4-weekly arm (17% vs 31%). Results were similar for subgroup analyses for BTA naïve and pre-treated pts, and for pts receiving DN, ZA or PAM.

Conclusions

The findings of this trial are consistent with those previously reported for de-escalating ZA. This trial also included pts receiving de-escalated DN and PAM. While the results of the REDUSE trial are awaited, the data presented would suggest that de-escalation of commonly used BTAs is a reasonable treatment option.

Clinical trial identification

NCT02721433.

Legal entity responsible for the study

The Ottawa Hospital Research Institute.

Funding

The Ottawa Hospital Research Institute and The Ottawa Hospital Foundation.

Disclosure

All authors have declared no conflicts of interest.

Background

Dynamic changes in circulating tumour DNA (ctDNA) levels may predict long-term outcome of therapy. We utilised samples from a phase I/II randomised trial (BEECH) to assess ctDNA dynamics as a surrogate for progression free survival (PFS) and early predictor of drug efficacy.

Methods

Patients with oestrogen receptor positive advanced metastatic breast cancer (ER+ mBC) in the BEECH study were randomised to paclitaxel plus placebo versus paclitaxel plus AKT inhibitor capivasertib. Plasma samples were collected for ctDNA analysis at baseline and at multiple timepoints in the development cohort (safety run-in, part A) and validation cohort (randomised, part B). Baseline sample ctDNA sequencing identified mutations for longitudinal analysis, and mutation specific digital droplet PCR (ddPCR) assays were utilised to assess change in ctDNA allele fraction between baseline and 872 on-treatment samples. Early suppression of ctDNA assessment was used to define criteria in the development cohort and independently evaluated in the validation cohort.

Results

In the development cohort, suppression of ctDNA was evident after 8 days of treatment (p = 0.014), with cycle 2 day 1 (4 weeks) identified as the optimal timepoint to predict PFS from early ctDNA dynamics. In the validation cohort, median PFS was 11.1 months in patients with suppressed ctDNA at 4 weeks and 6.4 months in patients with high ctDNA (HR = 0.20, 95% CI 0.083 – 0.50, p < 0.0001). No difference in the level of ctDNA suppression was observed between patients randomised to capivasertib or placebo overall (p = 0.904) nor in the PIK3CA mutant subpopulation (p = 0.071). Clonal haematopoiesis of indeterminate potential (CHIP) was evident in 30% (18/59) baseline samples and had no effect on tolerance of chemotherapy nor on PFS.

Conclusions

Early on-treatment ctDNA dynamics are a surrogate for PFS. Dynamic ctDNA assessment has the potential to substantially enhance early drug development through targeted therapy and biomarker evaluation.

Clinical trial identification

NCT01625286.

Legal entity responsible for the study

AstraZeneca.

Funding

AstraZeneca.

Disclosure

I. Kozarewa, J. Ratnayake, E.C. de Bruin, G. Schiavon: Employee: AstraZeneca. A.M. Antunes De Melo e Oliveira: Grant/research support (to the Institution): AstraZeneca. N. Turner: Advisory board honoraria, research funding: AstraZeneca. All other authors have declared no conflicts of interest.

Background

Predicting prognosis in metastatic HR+/HER2-negative disease might be of clinical value. Here, we developed and validated a prognostic biomarker in 765 patients (pts) recruited in 2 phase III trials evaluating endocrine-based therapies.

Methods

PAM50 and clinical data were available from 821 pts treated with letrozole+placebo/lapatinib in the first-line setting (EGF3008). Among them, 484 pts were selected based on HER2-negative disease and no prior endocrine therapy or relapse ≥6 months since tamoxifen discontinuation. To derive a prognostic model, the following variables were evaluated: 1) PAM50 subtypes, signatures and genes 2) ECOG, 4) visceral disease, 5) num. metastasis, 6) biopsy-type and 7) age. Using the variables, a progression-free survival (PFS) cox model was evaluated in 2/3 pts using Elastic Net (Monte Carlo CV). C-index of each model was estimated in 1/3 pts. The final model was tested in 261 pts treated with exemestane+placebo/everolimus (BOLERO2). PAM50 was performed in FFPE tumors (∼80% primary).

Results

In EGF3008, prognostic models that integrated PAM50 and clinical variables yielded superior C-index values compared to models with PAM50-only or clinical variables-only. The final model (PAM50MET) combined 21 variables, including 2 PAM50 subtypes, the Basal signature, 14 genes and 4 clinical variables. In EGF3008, the optimized cutpoint was associated with PFS (hazard ratio=0.41; p < 0.0001) and overall survival (OS; hazard ratio=0.41; p < 0.0001). In BOLERO2, PAM50MET score was associated with PFS (p = 0.004) and OS (p < 0.0001). Using the same cutpoint, PAM50MET-low was associated with better PFS (hazard ratio=0.72; p = 0.028) and OS (hazard ratio=0.51; p < 0.0001). The median PFS and OS in PAM50MET-low was 6.9 and 36.5 months compared to 5.2 and 23.4 months in PAM50MET-high.

Conclusions

PAM50MET could help identify pts with HR+/HER2-negative metastatic disease candidates, especially in the first-line setting, for endocrine therapy-only vs. endocrine therapy + CDK4/6 inhibitor vs. new treatment strategies. Further validation of PAM50MET in pivotal clinical trials that have evaluated endocrine-based therapies is justified.

Clinical trial identification

NCT00073528; NCT00863655.

Legal entity responsible for the study

Aleix Prat.

Funding

These trials was sponsored by Novartis Pharmaceuticals Corporation. This work was also supported by funds from the Spanish Society of Medical Oncology (to A.P.), Instituto de Salud Carlos III - PI13/01718 (to A.P.), Pas a Pas (to A.P.), Save the Mama (to A.P.), Instituto de Salud Carlos III - PI16/00904 (to A.P.), a Career Catalyst Grant (CCR13261208) from the Susan Komen Foundation (to A.P.), and Premio Jóven Investigador de la Fundación AstraZeneca (to A.P.).

Disclosure

A. Prat: Research funding: NanoString Technologies, GlaxoSmithKline, Susan G. Komen Foundation; Consulting/advisory relationship: NanoString Technologies, Novartis; Scientific advisory board: Novartis Pharmaceuticals Corporation, Pfizer, Roche. J.C. Brase: Employee: Novartis. S.R.D. Johnston: Consulting or advisory role: Eli Lilly, AstraZeneca, Puma Biotechnology; Speakers’ bureau: Pfizer, Novartis; Research funding: Pfizer (Inst). E.M. Ciruelos: Honoraria for consultancy and speaker: Lilly, Roche, Novartis, Pfizer. J.S. Parker: Co-inventor on patents related to the PAM50, licensed to NanoString Technologies, Inc. All other authors have declared no conflicts of interest.

Background

TNBC represents an unmet clinical need; the IMPassion 130 and KEYNOTE-086 trials have demonstrated efficacy of PD-1/PD-L1 targeted therapeutics in the metastatic/unresectable first line setting, with increased response in PD-L1 high populations, defined using different diagnostic assays and algorithms. This work addresses the need to establish PD-L1 assay concordance in TNBC.

Methods

196 TNBC cases were stained by IHC using the VENTANA PD-L1 (SP263), VENTANA PD-L1 (SP142), Dako PD-L1 IHC 28-8 pharmDx and Dako PD-L1 IHC 22C3 pharmDx assays. A single pathologist scored the proportions of membrane staining tumour cells (TC), staining immune cells (IC) and tumour occupied by immune cells. The proportion of tumour occupied by staining immune cells (IC_TA) and Combined Positive Score (CPS) were derived. Concordance between assays was evaluated using the Spearman coefficient ρ. Concordance in patient status was assessed using overall, positive and negative percent agreement (OPA/PPA/NPA) at matched algorithms and cut-offs.

Results

SP263, 22C3 and 28-8 assays showed good analytical correlation for TC staining (ρ = 0.84–0.89), but lower for SP142 (0.44-0.46), whereas all assays showed good correlation for IC_TA (0.77-0.96) and CPS (0.78-0.95). OPA ranged from 58%–97% at matched algorithms (Table). Prevalence for SP142 IC_TA≥1% and 22C3 CPS≥1 was in-line with IMPassion 130 and KEYNOTE-086, respectively.

Table: 10O

Prevalence (%) and OPA (%) of VENTANA SP263 with other assays at matched cut-offs

Conclusions

Analytical performance between SP263, 22C3 and 28-8 is very similar; SP142 stains fewer TCs and ICs. In contrast to other cancer types, in TNBC, TC scores are lower and IC scores are higher. Care should be taken when interchanging PD-L1 assays and interpreting efficacy data derived with different algorithms in TNBC; whilst SP263 would identify largely the same PD-L1 patient population as on KEYNOTE-086, it could potentially identify 22% additional cases as PD-L1 high than on IMPassion 130.

Legal entity responsible for the study

AstraZeneca.

Funding

AstraZeneca.

Disclosure

M. Scott, P. Scorer, C. Barker: Employee, shareholder: AstraZeneca. H. Al-Masri: Employee: Hematogenix.

Background

In patients (pts) with hormone receptor-positive (HR+), human epidermal growth factor receptor-2–negative (HER2–) advanced breast cancer (ABC), approximately 40% have tumors that exhibit PIK3CA mutations, resulting in phosphatidylinositol 3-kinase (PI3K) pathway hyperactivation. Use of the oral α-specific PI3K inhibitor alpelisib (ALP) + fulvestrant (FUL) significantly improved progression-free survival (PFS) in pts with ABC and a PIK3CA mutation (hazard ratio [HR]=0.65; P < 0.01) in SOLAR-1, a phase 3, randomized, double-blind study in men and postmenopausal women with HR+, HER2– ABC that progressed on or after aromatase inhibitor treatment. Here we report results by major region from SOLAR-1.

Methods

Pts (N = 572) received ALP 300 mg once daily (or placebo [PBO]) + FUL 500 mg every 28 days + Cycle 1, Day 15. Median PFS (mPFS) was estimated by Kaplan Meier in pts with a PIK3CA mutation (n = 341). Data for Japan will be presented separately and were excluded.

Results

For the PIK3CA-mutant (mut) cohort, pts were enrolled in Europe (EU; n = 173), North America (NA; n = 43), Asia (n = 34), and Latin America (LA; n = 31). Pts in the ALP arm (vs PBO) in EU had mPFS of 11.0 mo (vs 3.6 mo; HR = 0.56; 95% CI 0.39-0.81) and overall response rate (ORR) of 27.9% (vs 11.5%); in NA, mPFS was 15.2 mo (vs 3.6 mo; HR = 0.41; 95% CI 0.19-0.91), ORR 21.1% (vs 16.7%); in Asia, mPFS was 14.5 mo (vs 9.0 mo; HR = 0.55; 95% CI 0.20-1.51), ORR 46.7% (vs 10.5%); and in LA, mPFS was 9.4 mo (vs 12.9 mo; HR = 1.43; 95% CI 0.54-3.79), ORR 21.4% (vs 17.6%). Among all pts (mut and non-mut) in the ALP arm, median ALP exposure in EU, NA, Asia, LA, and overall was 5.5, 5.5, 7.6, 6.0, and 5.5 mo, respectively. Median average daily dose for ALP was 283.3, 294.0, 298.1, 260.2, and 286.1 mg/d. Most common all-grade AEs were hyperglycemia in EU, Asia, LA (63%, 75%, 65%) and nausea in NA (66%); followed by diarrhea in EU, NA, LA (61%, 66%, 53%) and decreased appetite (58%) in Asia.

Conclusions

Consistent with the overall SOLAR-1 population of the PIK3CA-mut cohort, PFS was generally improved across regions in the ALP vs PBO arm; however, low PFS events and pt numbers in some regions may limit conclusions.

Editorial acknowledgement

Healthcare Consultancy Group LLC.

Clinical trial identification

NCT02437318.

Legal entity responsible for the study

Novartis Pharmaceuticals Corporation.

Funding

Novartis Pharmaceuticals Corporation.

Disclosure

S. Loibl: Grant: Novartis, Pfizer, Celgene, Amgen. P.F. Conte: Speakers’ bureau: Roche/Genentech, Novartis, AstraZeneca; Research funding: Roche Relationship (Institution); Novartis (Institution); Merck Serono (Institution); Travel and accommodations: Novartis, Celgene, AstraZeneca. M. Campone: Grants: Pfizer, AstraZeneca, Sanofi, Pierre Fabre, Takeda; Personal fees: Novartis, Personal fees: Lilly, outside the submitted work. I. Mayer: Grants, clinical trial research funding: Novartis, Genentech, Pfizer; Ad board consulting: Novartis, Genentech, Lilly, Astra-Zeneca, Macrogenics, Immunomedics, Seattle Genetics, GSK. Y-S. Lu: Novartis (clinical trial study fee), during the conduct of the study: Novartis grants and personal fees outside the submitted work; Consulting fee: Pfizer, Boehringer Ingelheim; Contracted research: Roche, Merck Sharp & Dohmw, Pfizer, GlaxoSmithKlinem. N. Denduluri: Research funding: Lilly, Genentech, Novartis, Pfzier, Amgen; Consultant, research funding: Daiichi. C. Wilke: Employee, stockowner shares: Novarits. A. Ridolfi: Employee: Novartis Pharmaceuticals. F. André: Grants: Novartis, during the conduct of the study; Grants: Astra, Pfizer, Lilly, Roche, Daiichi, outside the submitted work.

Background

In EMBRACA, a randomised 2:1 phase 3 open-label study of pts with ABC and a gBRCA mutation, a statistically significant higher OR rate was observed with talazoparib (TALA) (n = 219) vs physician’s choice of chemotherapy (PCT; n = 114) (62.6% vs 27.2%; odds ratio, 5.0; 95% CI, 2.9, 8.8; P < 0.001). To evaluate the effects of OR on PRO, we performed post hoc analyses between pts who had OR vs those who did not, based on both treatments combined as well as for each treatment separately.

Methods

PRO were assessed on day 1 (baseline), at the start of each treatment cycle (every 3 wks), and at the end of treatment using the EORTC QLQ-C30 and breast cancer module QLQ-BR23. Higher scores indicated better global health status/quality of life (GHS/QoL). Pts were stratified by OR status (with OR vs without OR) within each treatment arm. Repeated measures mixed-effects analyses were performed to compare overall change from baseline scores while controlling for baseline. Time to definitive clinically meaningful deterioration (TDD) (decrease of ≥ 10 points) in GHS/QoL was compared between pts who had OR vs those who did not using stratified log-rank test and Cox proportional hazards model.

Results

With the 2 treatment arms combined, an overall change from baseline in GHS/QoL favoured pts with OR vs those without OR (3.9 [95% CI 0.7, 7.1]). Likewise, overall change from baseline favouring pts with OR vs those without OR was also seen in the TALA (2.1 [95% CI -1.6, 5.9]) and PCT arms (3.4 [95% CI -2.2, 9.1]) separately. Additionally, with the 2 treatment arms combined, a greater delay in TTD of GHS/QoL was observed in pts who had an OR vs those without OR (hazard ratio [HR]: 0.67; 95% CI 0.46, 0.98). A delay in TTD in those with OR vs no OR was also seen in the TALA (HR: 0.78; 95% CI 0.49, 1.24) and PCT arms (HR: 0.85; 95% CI 0.43, 1.71) separately.

Conclusions

Overall change from baseline and greater delay in TTD in GHS/QoL were observed favouring pts who experienced OR vs those who did not have OR. These results suggest that higher OR rates may lead to better overall improvement from baseline and greater delay in TTD in GHS/QoL in pts with ABC and a gBRCA mutation.

Legal entity responsible for the study

Pfizer, Inc.

Funding

Pfizer, Inc.

Disclosure

P.A. Fasching: Honoraria: Amgen, Novartis, Pfizer, Roche; Consulting or advisory role: Novartis, Pfizer, Roche; Research funding: Amgen, Novartis. R.G.W. Quek: Employee: Pfizer Inc.; Stock ownership: Pfizer Inc., Amgen Inc. H. Bhattacharyya: Employee, own stock: Pfizer. S.A. Hurvitz: Grants: Ambryx, Amgen, Bayer, BI Pharma, Biomarin, Cascadian, Daiichi Sankyo, Dignitana, Genentech, GSK, Lilly, Macrogenics, Medivation, Merrimack, Novartis, OBI Pharma, Pfizer, Pieris, Puma, Roche, Seattle Genetics; Travel: Lilly, Novartis, OBI Pharma. H.S. Rugo: Research funding to the University of California: Eisai, Genentech, GlaxoSmithKline, Lilly, Macrogenics, Merck, Novartis, OBI Pharma, Pfizer, Plexxikon; Travel expenses: Lilly, Mylan, Puma. J. Ettl: Consulting fees: Lilly, Novartis, Pfizer, Roche, Eisai; Contracted research: Celgene; Honoraria: Pfizer, Roche, Teva, Pierre Fabre; Travel support: Celgene, Novartis, Pfizer.

Background

The American Joint Committee on Cancer (AJCC) 8^th edition staging system introduced prognostic stage groups based on anatomic stage combined with biologic factors. We aimed to validate the AJCC prognostic classification in a large cohort of patients with HER2-positive breast cancer enrolled in the prospective ShortHER trial.

Methods

The ShortHER trial randomized 1253 HER2-positive patients to receive 9 weeks or 1 year of trastuzumab in combination with anthracycline and taxane chemotherapy. Patients were classified according to the classic AJCC anatomic groups and the AJCC prognostic groups (8^th edition). Distant disease-free survival (DDFS) was calculated as the time from randomization to relapse at a distant site or death. The Harrell’s C-index was used to compare the prognostic performance of the two staging systems.

Results

1244 patients had complete clinicopathological data for both AJCC anatomic and AJCC prognostic stage classifications. Compared with the anatomic AJCC, the prognostic AJCC moved 41.6% (n = 517) of the patients to a more favorable stage category: 100% of IB to IA (n = 40), 61.6% of IIA to IB or IA (n = 246), 63.0% of IIB to IB or IA (N = 94), 58.7% of IIIA to IIA or IB (n = 71) and 100% of IIIC to IIIB or IIIA (n = 66). Table shows the 5-years DDFS rates according to the two staging systems. The c-index was similar: 0.69209 for anatomic stage and 0.69249 for prognostic stage (P = 0.975).

Table: 88O

Conclusions

The AJCC prognostic classification reallocated 41.6% of HER2-positive patients to a more favorable stage category, while maintaining a similar prognostic performance as compared to the classic anatomic stage. With the AJCC prognostic staging, 59% of patients were classified as IA and showed an excellent prognosis after adjuvant treatment.

Clinical trial identification

EUDRACT number: 2007-004326-25 NCI; NCT00629278.

Legal entity responsible for the study

University of Modena and Reggio Emilia; University of Padua.

Funding

Agenzia Italiana del Farmaco (AIFA, grant FARM62MC97).

Disclosure

M.V. Dieci: Fees for consultancy role and participation on advisory boards: Eli Lilly; Fees for consultancy role: Genomic Health; Fees for participation on advisory boards: Celgene. A. Frassoldati: Advisory board: Roche, Novartis; Sponsored lectures: Pfizer, Novartis, Eli Lilly. O. Garrone: Fees for participation on advisory boards: Celgene, Eisai. V. Guarneri: Institutional research grant: Roche; Advisory boards: Eli Lilly, Roche, Novartis; Speaker’s bureau: Eli Lilly, Novartis. P.F. Conte: Fees and honoraria for participation on advisory boards: Eli Lilly, Novartis, Roche, AstraZeneca. All other authors have declared no conflicts of interest.