Molecular correlates differentiate response to atezolizumab (atezo) + bevacizumab (bev) vs sunitinib (sun): results from a Phase III study (IMmotio...

Date 20 October 2018
Event ESMO 2018 Congress
Session Proffered paper session - Genitourinary tumours, non prostate
Topics Renal Cell Cancer
Anticancer Agents
Presenter Brian Rini
Authors B.I. Rini1, M. Huseni2, M.B. Atkins3, D.F. McDermott4, T.B. Powles5, B. Escudier6, R. Banchereau7, L. Liu7, N. Leng8, J. Fan7, J. Doss7, S. Nalle7, S. Carroll7, S. Li8, C. Schiff7, M. Green7, R.J. Motzer9
  • 1Medical Oncology, Taussig Cancer Institute Cleveland Clinic, 44195 - Cleveland/US
  • 2Oncology Biomarkers Development, Genentech, Inc., South San Francisco/US
  • 3Oncology & Medicine, Lombardi Cancer Center Georgetown University, 20007 - Washington/US
  • 4Heme/onc, Beth Israel Deaconess Medical Center, 2215 - Boston/US
  • 5Queen Mary University Of London, Barts Cancer Institute and the Royal Free Hospital, EC1A 7BE - London/GB
  • 6Medical Oncology Department, Gustave Roussy, 94805 - Villejuif/FR
  • 7Oncology, Genentech, Inc., South San Francisco/US
  • 8Biostatistics, Genentech, Inc., South San Francisco/US
  • 9Medical Oncology, Memorial Sloan Kettering Cancer Center, 10065 - New York/US



Atezo + bev demonstrated improved PFS vs sun in pts with untreated mRCC expressing PD-L1 in Ph II (IMmotion150) and Ph III studies. Biomarker analyses in IMmotion150 suggested that T effector/IFNγ (Teff) and Angiogenesis (Angio) gene expression signatures (GEs) were associated with differential outcomes with atezo + bev and sun. We conducted genomic analyses to validate these GEs with clinical outcomes in IMmotion151 and evaluated their association with MSKCC risk groups and sarcomatoid histology.


Tumour GE analysis was performed by RNASeq in 823 pts from IMmotion151. Associations of Teff and Angio GEs with clinical outcome were evaluated at pre-specified expression level cutoffs identified in IMmotion150. PD-L1 status on immune cells was assessed with the SP142 IHC assay.


IMmotion151 met its co-primary endpoint, demonstrating improved PFS with atezo + bev vs sun in PD-L1+ pts (HR, 0.74 [95% CI: 0.57-0.96]; P = 0.02) across MSKCC groups. PFS was also improved in pts with sarcomatoid histology (HR, 0.56 [95% CI: 0.38-0.83]). High Teff GE was associated with PD-L1 expression by IHC and longer PFS in atezo + bev vs sun pts (HR, 0.76 [95% CI: 0.59-0.99]). High Angio GE was associated with improved PFS in the sun arm (HR, 0.59 [95% CI: 0.47-0.75]) but did not differentiate clinical activity between atezo + bev vs sun (HR, 0.95 [95% CI: 0.75-1.19]). Atezo + bev improved PFS vs sun in the low Angio subset (HR, 0.68 [95% CI: 0.52-0.89]). Angio GE was higher (P=4.28e-06) in favourable vs intermediate/poor MSKCC risk groups. PD-L1+ prevalence was higher (63% vs 39%) and Angio GE was lower (P=4.73e-16) in sarcomatoid vs non-sarcomatoid tumours.


These prospectively tested biomarker results validate molecular signatures that differentiate clinical outcomes with VEGF inhibition and immunotherapy in 1L mRCC. Moreover, these data identify tumour genomic profiles associated with prognostic risk groups and sarcomatoid histology. Findings from this study further our understanding of the biology of kidney cancer and inform future strategies to enable personalized therapy in mRCC pts.

Clinical trial identification


Editorial Acknowledgement

Medical writing assistance for this abstract was provided by Paige S. Davies, PhD, of Health Interactions.