233P - Phase 2 clinical evaluation of preclinically defined biomarkers for vascular endothelial growth factor (VEGF) tyrosine kinase inhibitor (TKI) tivoz...

Date 28 September 2014
Event ESMO 2014
Session Poster Display session
Topics Cytotoxic agents
Renal Cell Cancer
Translational Research
Basic Principles in the Management and Treatment (of cancer)
Therapy
Biological therapy
Presenter Thomas Hutson
Citation Annals of Oncology (2014) 25 (suppl_4): iv58-iv84. 10.1093/annonc/mdu326
Authors T. Hutson1, W.K. Rathmell2, B. Feng3, M.O. Robinson3, J. Gyuris4, J. Lin3, T. Choueiri5
  • 1Gu Oncology Program, Baylor Sammons Cancer Center, 75246 - Dallas/US
  • 2Lineberger Comprehensive Cancer Center, North Carolina Cancer Hospital, University of North Carolina, 27599 - Chapel Hill/US
  • 3Clinical Research, AVEO Oncology, Cambridge/US
  • 4Clinical Research, AVEO Oncology, 02142 - Cambridge/US
  • 5Division Of Solid Tumor Oncology, Dana-Farber Cancer Institute, 02215 - Boston/US

Abstract

Aim

Despite significant effort, identifying predictive biomarkers for VEGF-targeted therapies remains a challenge. Using population-based tumor models, we identified a population of infiltrating myeloid cells associated with resistance to tivozanib, a selective VEGF receptor TKI. Myeloid cell biomarkers (immunohistochemistry [IHC] and RNA) from preclinical studies were evaluated in a phase 2 RCC clinical trial.

Methods

Prespecified biomarkers were evaluated in AV-951-10-202 (NCT01297244), a single-arm trial of tivozanib monotherapy in nephrectomized, targeted, therapy-naive RCC. RNA signatures were quantified using averaged qRT-PCR values on available tumor formalin-fixed paraffin-embedded archival tissues. CD68 (+)-infiltrating myeloid cells were quantified by IHC (Aperio Scanscope). RNA signatures were evaluated for prognostic impact in a dataset collected prior to the use of VEGF targeted therapies (Zhao et al. PLoS Med. 2006;3:e13).

Results

Patients (n = 105) were enrolled (90 clear cell [cc] histology): intent-to-treat (ITT) progression-free survival (PFS), 9.7 mo; ccRCC PFS, 9.7 mo). ccRCC samples that passed quality check (RNA, 63; IHC, 66) were analyzed. Low myeloid signature score was associated with significantly longer PFS based on median cutoff (PFS 14.7 vs 8.3 mo, hazard ratio [HR] 0.49, P = .035; 95% CI 0.25-0.96), and as a continuous variable (P = .03; N = 63). The CD68 IHC score exhibited a similar trend but was not significant (median cutoff PFS 13.3 vs 9.2, HR 0.55, P = .067; 95% CI 0.28–1.05; continuous P = .057, N = 66). This gene signature exhibited a prognostic effect in a historical data set (Zhao, 2006).

Conclusions

A preclinically derived myeloid signature identified a ccRCC population with longer PFS on tivozanib and further provides a candidate VEGF pathway resistance mechanism amenable to inhibition. These results warrant the consideration of combination therapies targeting both the VEGF pathway and myeloid cells.

Disclosure

T. Hutson: Advisory boards; Corporate sponsored research: Pfizer, Bayer, GSK, AVEO, Novartis; B. Feng, J. Gyuris and J. Lin have declared: AVEO Oncology employee o Stock o Salary M.O. Robinson: • Former AVEO Oncology employee; T. Choueiri: Advisory boards: AVEO, GSK, Bayer, Novartis, Pfizer Corporate sponsored research: Pfizer . All other authors have declared no conflicts of interest.