Atezolizumab plus bevacizumab in hepatocellular carcinoma (HCC): Safety and clinical activity results from a phase Ib study

Background: Patients (pts) with advanced HCC have limited treatment (tx) options; the multikinase inhibitor sorafenib is the global standard of care. Inhibition of either PD-L1/PD-1 or VEGF signalling has only limited activity in HCC. However, the additional immunomodulatory effects of bevacizumab (bev; anti-VEGF) may create a favourable tumour microenvironment that can enhance the efficacy of atezolizumab (atezo; anti–PD-L1), leading to an effective anti-tumour immune response. Therefore, atezo + bev may have synergistic effects in HCC. Methods: In a Phase Ib study cohort, pts with unresectable or metastatic 1L HCC received atezo 1200 mg + bev 15 mg/kg IV q3w until loss of clinical benefit. The primary endpoints were safety and ORR per investigator (INV) assessment. Secondary endpoints included PFS and DOR per RECIST v1.1, and OS. Results: 43 pts were evaluable for safety as of Jan 11, 2018. Tx-related all-grade (Gr) AEs occurred in 81% of pts; Gr 3-4 AEs were seen in 28%, most commonly hypertension (16%). 3 pts (7%) had tx-related SAEs (autoimmune encephalitis and mental status change in 1 pt, intra-abdominal haemorrhage, jugular vein thrombosis; all Gr 3). No tx-related Gr 5 AEs were seen. Immune-related AEs requiring systemic corticosteroid tx occurred in 5 pts (12%). 23 pts were evaluable for clinical activity. Response rates and 6-month PFS and OS rates are presented in the table. Medians for DOR, PFS and OS have not yet been reached. 12 of 14 responses were ongoing as of the data cutoff (≥ 6 mo in 10 pts). Updated data from an expanded pt cohort wilTable150O Efficacy (N = 23; minimum follow-up, 16 wk; median follow-up, 10.3 mo)Response Rates (confirmed per RECIST v1.1)INV AssessedIRF AssessedORR, n/N (%)14/23 (61%)15/23 (65%)CR01/23 (4%)PR14/23 (61%)14/23 (61%)SD5/23 (22%)7/23 (30%)PD4/23 (17%)1 (4%)DCR, n/N (%)CR + PR + SD19/23 (83%)22/23 (96%)CR + PR + SD ≥ 6 mo15/23 (65%)16/23 (70%)By region, n/N (%)Asia (excluding Japan)6/9 (67%)6/9 (67%)Japan/US8/14 (57%)9/14 (64%)By aetiology, n/N (%)HBV6/10 (60%)6/10 (60%)HCV7/9 (78%)7/9 (78%)Non-viral1/4 (25%)2/4 (50%)By AFP, n/Na (%)< 400 ng/mL8/13 (62%)8/13 (62%)≥ 400 ng/mL6/9 (67%)7/9 (78%)PFS Rate (RECIST v1.1)6-month PFS65%65%OS Rate6-month OS86%a Baseline AFP data from 1 patient is missing. AFP, alpha-fetoprotein; INV, investigator; IRF, independent review facility. Open table in a new tab l be shown. a Baseline AFP data from 1 patient is missing. AFP, alpha-fetoprotein; INV, investigator; IRF, independent review facility. Conclusions: Atezo + bev had a tolerable safety profile as well as high and durable objective responses. The ORR of 61% suggests that this combination has synergistic activity in HCC. Atezo + bev may offer a promising 1L tx option for advanced HCC and is being evaluated in the Phase III IMbrave150 study (NCT03434379). Editorial acknowledgement: Medical writing assistance for this abstract was provided by Steffen Biechele, PhD, of Health Interactions and funded by F. Hoffmann-La Roche, Ltd. Clinical trial identification: NCT02715531. Legal entity responsible for the study: F. Hoffmann-La Roche, Ltd. Funding: F. Hoffmann-La Roche, Ltd. Disclosure: C-H. Hsu: Membership on advisory board: BMS, Ono, MSD, Merck/Sorono, Novartis, Roche; Research funding: MSD. M.S. Lee: Consulting/advisory: Bayer; Research funding: EMD Sereno, Genentech/Roche. W. Verret, B. Liu, K. Iizuka: Employment: Genentech. A. Kwan: Employment & stock: Genentech/Roche. S. Stein: Consulting/advisory: Genentech/Roche. All other authors have declared no conflicts of interest.