Abstract P6-14-03: Genome wide association study (GWAS) to identify variants conferring ramucirumab-associated hypertension in the ROSE/TRIO-012 breast cancer trial

Background: In a candidate single nucleotide polymorphism (SNP) association study, we previously identified VEGFR-1 and VEGFR-2 SNPs strongly associated with treatment emergent hypertension (HT) in the ramucirumab (RAM) and docetaxel (Doc) arm in the ROSE/TRIO-012 study, a double-blinded multinational phase III trial that randomized 1,144 patients with advanced breast cancer to receive first-line Doc in combination with RAM or placebo (Mackey et al, JCO Jan 10, 2015:141-148; Mackey et al, JCO, Volume 33, Issue 15_suppl, May 20, 2015: 547). However, candidate SNP studies limit the number of genes for interrogation and a more comprehensive genome wide search may identify critical variants associated with the phenotype of HT. Preliminary analysis indicated that patients experiencing HT with RAM showed better overall survival (Mackey JR, et al (2015). Reply to H. Lee, et al. JCO; in press). These observations provide the potential to identify those patients with genetic variants for predisposition to RAM-associated HT to inform therapeutic decisions.

Methods: Genotyping of samples is underway using Affymetrix SNP 6.0 arrays. Genotype data will be filtered for deviations from Hardy Weinberg Equilibrium and minor allele frequency of >0.05. Study subjects (n=792) provided ethics-committee approved prospective consent for this genetic study of whom 478 subjects were allocated RAM + Doc arm. Toxicity grades 0-1 (n= 394 controls; low toxicity) vs. grade >2 (n= 84 cases, high toxicity) is our binary outcome. Dominant genotypic model is assumed. Chi-square test, FDR and/or 10000 permutation tests will be employed (Golden Helix-SVS v8.3) and p<0.05 considered statistically significant. Population stratification will be identified (EIGENSTRAT) and association statistics will be corrected using Eigenvectors along with age as covariates. Fine mapping of loci showing significant associations will be attempted using imputation tools.

Results and conclusions: We expect up to 700,000 SNPs to be retained after filtering based on our previous breast cancer GWAS analyses (Damaraju et al. Cancer Research (suppl); Vol 70 (24), page 258s, 2010 and Sehrawat et al Hum Genet. 2011 Oct;130(4):529-37) and 30,000 SNPs to show significance at a nominal p-value (0.05); these will be analysed for regions of high linkage disequilibrium to narrow down potential loci showing association with HT to serve as candidate markers in further independent validation studies. Cumulative dose to adverse events will be considered in the analysis. Identified loci will be interrogated for potential genes in the flanking regions with biological relevance based on pathway analysis. Identified variants from candidate SNP and GWAS may allow developing predictive tools to enable stratification of patients for therapies. The analysis is expected to be completed by mid- November, 2015.

Citation Format: Mackey JR, Lipatov O, Martín M, Webster M, Hegg R, Verma S, Ramos-Vázquez M, Fresco R, Thireau F, Houé V, Press MF, Kumaran M, Damaraju S. Genome wide association study (GWAS) to identify variants conferring ramucirumab-associated hypertension in the ROSE/TRIO-012 breast cancer 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 P6-14-03.

Abstract P1-13-03: Genome wide association study (GWAS) of genetic variants associated with docetaxel toxicity in the ROSE/TRIO-012 trial

Background: Genetic predisposition to docetaxel (Doc) toxicity contributes to unacceptable toxicity and reduced dose intensity, and may influence disease outcomes. We previously reported variants associated with Doc toxicity in candidate gene single nucleotide polymorphism (SNP) associations in a breast cancer treatment setting (Damaraju et al. Eur J Cancer (suppl); Vol 8 (7), page 175, 2010) and the identified variants were confirmed in an independent validation study (Damaraju et al, J Clin Oncol Vol 33, Issue 15 suppl, 2015: 540). Others have reported candidate SNP (Breast Cancer Res Treat, 2011 SWOG 0221 study) and GWAS (Clin Cancer Res 2012 CALGB 40101 study) identified variants associated with paclitaxel mediated peripheral neuropathy. However, the overlap on the variants identified thus far between Doc and paclitaxel are limited, prompting a genome wide search to find variants contributing to Doc specific toxicity.

Methods: TRIO-012 is a double blinded, multinational trial that randomized 1,144 patients with advanced breast cancer to receive first-line Doc in combination with ramucirumab (RAM) or placebo (Mackey et al. J Clin Oncol Jan 10, 2015:141-148). Study subjects (n=719) in the Doc+RAM or Doc+placebo arm with available germline DNA are being genotyped; all subjects provided ethics-committee approved prospective consent for this genetic study. Genotyping are being performed with Affymetrix SNP 6.0 arrays. Genotype data will be filtered for deviations from Hardy Weinberg Equilibrium and minor allele frequency of >0.05. Doc-induced adverse events (AEs) are based on CTCAE (Common Terminology Criteria for Adverse Events v4.1) toxicity grades. Toxicities >2 scored for fatigue (n=96), myalgia (n=22), peripheral neuropathy (n=17) will be analysed as individual phenotypes in comparison with the no toxicity group (toxicity grades 0-1) and in a combined analysis of all Doc induced toxicities (0-1, n=599 vs. >2, n=120). Dominant genotypic model is assumed; Chi-square test, FDR and/or 10000 permutations were employed using SVS v8.3 and p<0.05 considered statistically significant. We will identify population stratification using EIGENSTRAT method and will correct the association statistics using the Eigenvectors along with age and BMI as covariates. Fine mapping of the identified loci will be attempted using imputation tools. We will interrogate the data for cumulative dose to toxicity and correlate SNPs identified with survival outcomes.

Results and conclusions: We expect to reconfirm the associations of loci reported in candidate SNP and previous GWAS studies; XKR4 (rs4737264) for peripheral neuropathy, CYP3A5*3 (rs776746) with fatigue, and FACND2 (rs7637888) with myalgia in addition to the potential novel variants distinct from paclitaxel AE GWAS studies. Fine mapping around these loci may help identify potential causal variants. Both candidate SNP and GWAS identified variants may aid in developing risk stratification models. The GWAS identified loci and the flanking genes will be interrogated using the ingenuity pathway analysis for insights in to the biological roles in the drug metabolism. We expect to complete the analysis by mid-November 2015.

Citation Format: Damaraju S, Gorbunova V, Gelmon K, García-Saenz J, Morales-Murillo S, AbiGerges D, Canon J-L, Kiselev I, Cohen GL, Jerusalem G, Thireau F, Fresco R, Houé V, Press MF, Kumaran M, Mackey JR. Genome wide association study (GWAS) of genetic variants associated with docetaxel toxicity in the ROSE/TRIO-012 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 P1-13-03.

Primary results of ROSE/TRIO-12, a randomized placebo-controlled phase III trial evaluating the addition of ramucirumab to first-line docetaxel chemotherapy in metastatic breast cancer

PURPOSE

Currently, antiangiogenic strategies in metastatic breast cancer have demonstrated modest improvements in progression-free survival (PFS) but not improved quality or duration of survival, warranting evaluation of new agents in a placebo-controlled setting. Ramucirumab is a human immunoglobulin G1 antibody that binds vascular endothelial growth factor receptor-2 and blocks ligand-stimulated activation. The ROSE/TRIO-012 trial evaluated ramucirumab with docetaxel in unresectable, locally recurrent, or metastatic breast cancer.

PATIENTS AND METHODS

In this double-blind, placebo-controlled, randomized, multinational phase III trial, 1,144 patients with human epidermal growth factor receptor 2 (HER2) -negative breast cancer who had not received cytotoxic chemotherapy in the advanced setting were randomly assigned at a two-to-one ratio to receive docetaxel 75 mg/m(2) plus ramucirumab 10 mg/kg or docetaxel 75 mg/m(2) plus placebo once every 3 weeks. Treatment continued until disease progression, unacceptable toxicity, or other withdrawal criteria. Patients were stratified by previous taxane therapy, visceral metastasis, hormone receptor status, and geographic region. An independent data monitoring committee oversaw the trial. The primary end point was investigator-assessed PFS.

RESULTS

Median PFS in patients treated with ramucirumab plus docetaxel was 9.5 months, compared with 8.2 months in patients who received placebo plus docetaxel (hazard ratio [HR], 0.88; P = .077). Median overall survival was 27.3 months in patients who received ramucirumab plus docetaxel, compared with 27.2 months in patients who received placebo plus docetaxel (HR, 1.01; P = .915). Toxicities seen at significantly higher rates in patients receiving ramucirumab included fatigue, hypertension, febrile neutropenia, palmar-plantar erythrodysesthesia syndrome, and stomatitis.

CONCLUSION

Addition of ramucirumab to docetaxel in HER2-negative advanced breast cancer did not meaningfully improve important clinical outcomes.