Atezolizumab (anti-programmed death-ligand 1 (PD-L1)) and bevacizumab (anti-vascular endothelial growth factor (VEGF)) combination therapy has become the new standard of care in patients with unresectable hepatocellular carcinoma. However, potential predictive biomarkers and mechanisms of response and resistance remain less well understood. We report integrated molecular analyses of tumor samples from 358 patients with hepatocellular carcinoma (HCC) enrolled in the GO30140 phase 1b or IMbrave150 phase 3 trial and treated with atezolizumab combined with bevacizumab, atezolizumab alone or sorafenib (multikinase inhibitor). Pre-existing immunity (high expression of CD274, T-effector signature and intratumoral CD8+ T cell density) was associated with better clinical outcomes with the combination. Reduced clinical benefit was associated with high regulatory T cell (Treg) to effector T cell (Teff) ratio and expression of oncofetal genes (GPC3, AFP). Improved outcomes from the combination versus atezolizumab alone were associated with high expression of VEGF Receptor 2 (KDR), Tregs and myeloid inflammation signatures. These findings were further validated by analyses of paired pre- and post-treatment biopsies, in situ analyses and in vivo mouse models. Our study identified key molecular correlates of the combination therapy and highlighted that anti-VEGF might synergize with anti-PD-L1 by targeting angiogenesis, Treg proliferation and myeloid cell inflammation.
Your institute does not have access to this article
Subscribe to Nature+
Get immediate online access to the entire Nature family of 50+ journals
Subscribe to Journal
Get full journal access for 1 year
only $4.92 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
All clinical, raw RNA-seq and WES data for the GO30140 and IMbrave150 trials are deposited in the European Genome-Phenome Archive under accession no. EGAS00001005503. Qualified researchers may request access to individual patient-level data through the clinical study data request platform (https://vivli.org/). Further details on Roche’s criteria for eligible studies are available at https://vivli.org/members/ourmembers. For further details on Roche’s Global Policy on the Sharing of Clinical Information and how to request access to related clinical study documents, see https://www.roche.com/research_and_development/who_we_are_how_we_work/clinical_trials/our_commitment_to_data_sharing.htm. MSigDB v.7.5.1 was used for GSEA. GENCODE Human release 40 was used for sequencing pipeline gene modeling. Source Data are provided with this paper.
Sung, H. et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 71, 209–249 (2021).
Lee, M. S. et al. GO30140 Investigators. Atezolizumab with or without bevacizumab in unresectable hepatocellular carcinoma (GO30140): an open-label, multicentre, phase 1b study. Lancet Oncol. 21, 808–820 (2020).
Finn, R. S. et al. IMbrave150 Investigators. Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma. N. Engl. J. Med. 382, 1894–1905 (2020).
Socinski, M. A. et al. Atezolizumab for first-line treatment of metastatic nonsquamous NSCLC. N. Engl. J. Med. 378, 2288–2301 (2018).
McDermott, D. F. et al. Clinical activity and molecular correlates of response to atezolizumab alone or in combination with bevacizumab versus sunitinib in renal cell carcinoma. Nat. Med. 24, 749–757 (2018).
Hack, S. P., Zhu, A. X. & Wang, Y. Augmenting anticancer immunity through combined targeting of angiogenic and PD-1/PD-L1 pathways: challenges and opportunities. Front. Immunol. 11, 598877 (2020).
Aran, D., Hu, Z. & Butte, A. J. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol. 18, 220 (2017).
Lei, Y., Li, X., Huang, Q., Zheng, X. & Liu, M. Progress and challenges of predictive biomarkers for immune checkpoint blockade. Front. Oncol. 11, 617335 (2021).
Zhang, W. et al. Fully automated 5-plex fluorescent immunohistochemistry with tyramide signal amplification and same species antibodies. Lab. Invest. 97, 873–885 (2017).
Li, J. L. et al. DLL4-notch signaling mediates tumor resistance to anti-VEGF therapy in vivo. Cancer Res. 71, 6073–6083 (2011).
Zhu, A. X. et al. REACH-2 study investigators. Ramucirumab after sorafenib in patients with advanced hepatocellular carcinoma and increased α-fetoprotein concentrations (REACH-2): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol. 20, 282–296 (2019).
Rosenberg, J. E. et al. Atezolizumab in patients with locally advanced and metastatic urothelial carcinoma who have progressed following treatment with platinum-based chemotherapy: a single-arm, multicentre, phase 2 trial. Lancet 387, 1909–1920 (2016).
Rizvi, N. A. et al. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. Science 348, 124–128 (2015).
Carbone, D. P. et al. First-line nivolumab in stage IV or recurrent non-small-cell lung cancer. N. Engl. J. Med. 376, 2415–2426 (2017).
Kim, Y. J. et al. A phase II study of pembrolizumab and paclitaxel in patients with relapsed or refractory small-cell lung cancer. Lung Cancer 136, 122–128 (2019).
de Galarreta, R. M. et al. β-catenin activation promotes immune escape and resistance to anti-PD-1 therapy in hepatocellular carcinoma. Cancer Discov. 9, 1124–1141 (2019).
Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation. Cell 144, 646–674 (2011).
Khan, K. A. & Kerbel, R. S. Improving immunotherapy outcomes with anti-angiogenic treatments and vice versa. Nat. Rev. Clin. Oncol. 15, 310–324 (2018).
Lee, H. G., Cho, M. Z. & Choi, J. M. Bystander CD4+ T cells: crossroads between innate and adaptive immunity. Exp. Mol. Med. 52, 1255–1263 (2020).
Rini, B. I. et al. Pembrolizumab plus axitinib versus sunitinib for advanced renal-cell carcinoma. N. Engl. J. Med. 380, 1116–1127 (2019).
Motzer, R. J. et al. Avelumab plus axitinib versus sunitinib for advanced renal-cell carcinoma. N. Engl. J. Med. 380, 1103–1115 (2019).
Makker, V. et al. Lenvatinib plus pembrolizumab in patients with advanced endometrial cancer: an interim analysis of a multicentre, open-label, single-arm, phase 2 trial. Lancet Oncol. 20, 711–718 (2019).
Topalian, S. L. et al. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N. Engl. J. Med. 366, 2443–2454 (2012).
Herbst, R. S. et al. Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature 515, 563–567 (2014).
Borghaei, H. et al. Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer. N. Engl. J. Med. 373, 1627 (2015).
Fehrenbacher, L. et al. Atezolizumab versus docetaxel for patients with previously treated non-small-cell lung cancer (POPLAR): a multicentre, open-label, phase 2 randomised controlled trial. Lancet 387, 1837–1846 (2016).
Herbst, R. S. et al. Pembrolizumab versus docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung cancer (KEYNOTE-010): a randomized controlled trial. Lancet 387, 1540–1550 (2016).
Sangro, B. et al. Association of inflammatory biomarkers with clinical outcomes in nivolumab-treated patients with advanced hepatocellular carcinoma. J. Hepatol. 73, 1460–1469 (2020).
Pinato, D. J. et al. Clinical implications of heterogeneity in PD-L1 immunohistochemical detection in hepatocellular carcinoma: the Blueprint-HCC study. Br. J. Cancer 120, 1033–1036 (2019).
Peng, L. S. et al. Increased tumor-infiltrating CD8(+)Foxp3(+) T lymphocytes are associated with tumor progression in human gastric cancer. Cancer Immunol. Immunother. 61, 2183–2192 (2012).
Yoon, H. H. et al. Prognostic impact of FoxP3+ regulatory T cells in relation to CD8+ T lymphocyte density in human colon carcinomas. PLoS ONE 7, e42274 (2012).
deLeeuw, R. J., Kost, S. E., Kakal, J. A. & Nelson, B. H. The prognostic value of FoxP3+ tumor-infiltrating lymphocytes in cancer: a critical review of the literature. Clin. Cancer Res. 18, 3022–3029 (2012).
Kalathil, S. G., Hutson, A., Barbi, J., Iyer, R. & Thanavala, Y. Augmentation of IFN-γ+ CD8+ T cell responses correlates with survival of HCC patients on sorafenib therapy. JCI Insight 4, e130116 (2019).
Sangro, B., Sarobe, P., Hervás-Stubbs, S. & Melero, I. Advances in immunotherapy for hepatocellular carcinoma. Nat. Rev. Gastroenterol. Hepatol. 18, 525–543 (2021).
Ormandy, L. A. et al. Increased populations of regulatory T cells in peripheral blood of patients with hepatocellular carcinoma. Cancer Res. 65, 2457–2464 (2005).
Shetty, S., Lalor, P. F. & Adams, D. H. Liver sinusoidal endothelial cells—gatekeepers of hepatic immunity. Nat. Rev. Gastroenterol. Hepatol. 15, 555–567 (2018).
Dunham, R. M. et al. Hepatic stellate cells preferentially induce Foxp3+ regulatory T cells by production of retinoic acid. J. Immunol. 190, 2009–2016 (2013).
Sharma, A. et al. Onco-fetal reprogramming of endothelial cells drives immunosuppressive macrophages in hepatocellular carcinoma. Cell 183, 377–394 (2020).
Hellmann, M. D. et al. Nivolumab plus ipilimumab in lung cancer with a high tumor mutational burden. N. Engl. J. Med. 378, 2093–2104 (2018).
Hellmann, M. D. et al. Tumor mutational burden and efficacy of nivolumab monotherapy and in combination with ipilimumab in small-cell lung cancer. Cancer Cell 33, 853–861 (2018).
Hainsworth, J. D. et al. Targeted therapy for advanced solid tumors on the basis of molecular profiles: results from MyPathway, an open-label, Phase IIa multiple basket study. J. Clin. Oncol. 36, 536–542 (2018).
Ang, C. et al. Prevalence of established and emerging biomarkers of immune checkpoint inhibitor response in advanced hepatocellular carcinoma. Oncotarget 10, 4018–4025 (2019).
Wong, C. N. et al. Qualification of tumour mutational burden by targeted next-generation sequencing as a biomarker in hepatocellular carcinoma. Liver Int. 41, 192–203 (2021).
Lachenmayer, A. et al. Wnt-pathway activation in two molecular classes of hepatocellular carcinoma and experimental modulation by sorafenib. Clin. Cancer Res. 18, 4997–5007 (2012).
Sohn, B. H., Park, I. Y., Shin, J. H., Yim, S. Y. & Lee, J. S. Glutamine synthetase mediates sorafenib sensitivity in β-catenin-active hepatocellular carcinoma cells. Exp. Mol. Med. 50, e421 (2018).
Berraondo, P., Ochoa, M. C., Olivera, I. & Melero, I. Immune desertic landscapes in hepatocellular carcinoma shaped by β-catenin activation. Cancer Discov. 9, 1003–1005 (2019).
Pinyol, R., Sia, D. & Llovet, J. M. Immune exclusion-Wnt/CTNNB1 class predicts resistance to immunotherapies in HCC. Clin. Cancer Res. 25, 2021–2023 (2019).
Harding, J. J. et al. Prospective genotyping of hepatocellular carcinoma: clinical implications of next-generation sequencing for matching patients to targeted and immune therapies. Clin. Cancer Res. 25, 2116–2126 (2019).
Trujillo, J. A. et al. Secondary resistance to immunotherapy associated with β-catenin pathway activation or PTEN loss in metastatic melanoma. J. Immunother. Cancer 7, 295 (2019).
Pfister, D. et al. NASH limits anti-tumour surveillance in immunotherapy-treated HCC. Nature 592, 450–456 (2021).
Terme, M. et al. VEGFA-VEGFR pathway blockade inhibits tumor-induced regulatory T-cell proliferation in colorectal cancer. Cancer Res. 73, 539–549 (2013).
Lee, W. S., Yang, H., Chon, H. J. & Kim, C. Combination of anti-angiogenic therapy and immune checkpoint blockade normalizes vascular-immune crosstalk to potentiate cancer immunity. Exp. Mol. Med. 52, 1475–1485 (2020).
Fessas, P. et al. Phenotypic characteristics of the tumour microenvironment in primary and secondary hepatocellular carcinoma. Cancers 13, 2137 (2021).
Lawrence, M., Degenhardt, J. & Gentleman, R. Package VariantTools. Bioconductor https://www.bioconductor.org/packages/devel/bioc/manuals/VariantTools/man/VariantTools.pdf (2022).
Karosiene, E., Lundegaard, C., Lund, O. & Nielsen, M. NetMHCcons: a consensus method for the major histocompatibility complex class I predictions. Immunogenetics 64, 177–186 (2012).
Shukla, S. A. et al. Comprehensive analysis of cancer-associated somatic mutations in class I HLA genes. Nat. Biotechnol. 33, 1152–1158 (2015).
Canzler, S. & Hackermüller, J. multiGSEA: a GSEA-based pathway enrichment analysis for multi-omics data. BMC Bioinformatics 21, 561 (2020).
We are grateful for the participation and commitment of patients, families and doctors in biomarker studies of the GO30140 and IMbrave150 trials. Without their contribution, this study would not have been possible. We thank the following foundations for their financial support. M.R.G. was supported by Fundación Alfonso Martín Escudero Fellowship and a Damon Runyon-Rachleff Innovation Award (no. DR52-18). A.L. was supported by a Damon Runyon-Rachleff Innovation Award (no. DR52-18), an R37 Merit Award (no. R37CA230636) and Icahn School of Medicine at Mount Sinai. The Tisch Cancer Institute and related research facilities are supported by P30 CA196521. We thank J. Munnoz-Rodriguez, X. Wang and X. Wang from Roche Tissue Diagnostics for performing and developing the digital pathology algorithm, and for data analysis of multiplex immunofluorescence assays. We also thank A. Pochiraju, N. Yang, S. Nalle and L. Ma for coordinating sample collection, assay implementation and data transfer. We thank C. V. Wong and J. Cheung for managing and coordinating preclinical studies at Mount Sinai. Figure 1 was created with BioRender.com.
A.X.Z. received consulting fees from Bayer, Eisai, Eli Lilly, Exelixis, F. Hoffmann–La Roche, Merck, Gilead, Sanofi Aventis and Sirtex. M.R.G. has received grant support from Genentech. A.R.A., Y.G., S.L., H.K., J.S., W.V. and Y.W. are employees of Genentech and hold stock or other ownership interests in F. Hoffmann–La Roche. W.Z. is an employee of Roche Tissue Diagnostics and holds stock and other ownership interests in Roche. C.-H.H. received consulting fees from AstraZeneca, Eli Lilly and Roche, and received honoraria from Bristol-Meyers Squibb, Eisai, Merck Sharp & Dohme, Ono Pharmaceutical and Roche. A.R.H. has received research funding from Genentech and Merck, and has been on a speakers’ bureau for Eisai, BMS and Exelixis. B.-Y.R. has nothing to declare. T.Y. has received honoraria and consulting fees from Bristol-Myers Squibb. A.O.K. has received honoraria from Bayer Health, Bristol-Myers Squibb, Eisai, Exelixis, Genentech/Roche and Merck; has received consulting fees from Bayer Health, Bristol-Myers Squibb, Eisai, Exelixis, Genentech/Roche and Merck; has received institutional research funding from Adaptimmune, Bayer/Onyx, Bristol-Myers Squibb, Genentech, Hengrui Pharmaceutical and Merck; and travel, accommodation and other expenses support from Bayer/Onyx, Bristol-Myers Squibb, Exelixis and Merck. A.M.B. has received consulting fees from Deciphera, Exelixis and Genentech. F.D. has received consulting fees from Genentech/Roche, Array BioPharma, Exelixis, Eisai, QED Therapeutics and Signatera; has been on a speakers’ bureau for Genentech/Roche, Amgen, Eisai, Ipsen, Exelixis, Sirtex Medical, Deciphera, Natera and Servier; has received institutional research funding from Bristol-Myers Squibb, AstraZeneca, Merck, Genentech, Taiho Pharmaceutical, Exelixis and Ipsen; and has an immediate family member who is an employee of Roche Diagnostics. R.S.F. has received consulting fees from AstraZeneca, Bayer, Bristol-Myers Squibb, Eisai, Exelixis, F. Hoffmann–La Roche/Genentech, Lilly, Merck, Novartis and Pfizer; has received institutional research funding from Bayer, Bristol-Myers Squibb, Eisai, Lilly, Merck, Novartis, Pfizer and F. Hoffmann–La Roche/Genentech; and has provided expert testimony for Novartis. H.C.T. has received honoraria from Roche, MSD Merck, Ipsen and AstraZeneca. A.L. has received consulting fees from AstraZeneca and research funding from Pfizer.
Peer review information
Nature Medicine thanks David Pinato, Xin Wang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Saheli Sadanand, in collaboration with the Nature Medicine team.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended Data Fig. 1 Clinical outcomes in the biomarker-evaluable population (BEP) are representative of those in the intention-to-treat (ITT) population in GO30140 and IMbrave150 clinical trials.
Kaplan-Meier estimates of progression-free survival (PFS) and overall survival (OS) in patients in the ITT and BEP of a, GO30140 group A & F (A n = 90, F n = 91 biologically independent subjects) or b, IMbrave150 (n = 119 biologically independent subjects). Stratified hazard ratios and two-sided log-likelihood P values for progression or death are reported. 95% confidence intervals are indicated in parentheses. A table comparing independent review facility–assessed objective response rate in the BEP and ITT population in GO30140 group A is included in (a). The variables used for stratification in the Cox model for IMBrave 150 (b) were geographic region (Asia [excluding Japan] vs the rest of the world), α-fetoprotein categorical level at baseline (<400 ng per milliliter vs ≥400 ng per milliliter), and the presence of either macrovascular invasion or extrahepatic spread (vs the d of both). Tick marks indicate censored data. Numbers in parentheses indicate CIs. Atezo, atezolizumab; Bev, bevacizumab; CR, complete response; HR, hazard ratio; PD, progressive disease; PR, partial response; SD, stable disease.
Deconvolution scores for various cell types and biological programs yielded from performing xCell analysis on expression data from the GO30140 study (n = 90 biologically independent subjects), separated by responder/non-responder status. Responders were defined as patients with complete or partial response as assessed by an independent review facility according to Response Evaluation Criteria in Solid Tumors, version 1.1. Boxes represent median and first/third quartile values. Box whiskers represent the most extreme values within 150% of the interquartile range. P-values are two-sided Student t tests P value comparing scores between responders and non-responders. aDC, activated dendritic cells; DC, dendritic cells; cDC, conventional dendritic cells.
Extended Data Fig. 3 Key biomarker expression by RECIST response and PFS/OS association within treatment groups of IMbrave150.
a, Expression of ABRS, CD274 and Teff signature score by RECIST response in IMbrave150 Atezo-Bev treatment group (n = 119 biologically independent subjects). Box plots are of log2 gene expression scores for response categories, with two-sided P values from Student t tests. Boxes represent median (middle line) and first/third quartile (bottom and upper lines) values. Box whiskers represent the most extreme values within 150% of the interquartile range. b, KM plots of PFS stratified by ABRS, CD274 or Teff signature high vs low expression (median split) within the Aatezo-Bbev (n = 119 biologically independent subjects) or sorafenib (n = 58 biologically independent subjects) treatment group in IMbrave150. c, KM plots of OS stratified by ABRS, CD274 or Teff signature high vs low expression (median split) within the Aatezo-Bbev (n = 119 biologically independent subjects) or sorafenib (n = 58 biologically independent subjects) treatment group in IMbrave150. 95% confidence intervals are indicated in parentheses.
a, Images of PD-L1 staining on tumor cells and immune cells by PD-L1 immunohistochemistry (SP263) (scale bars in each main image and its magnified inset are 90 µm and 20 µm, respectively) are representing. PD-L1 IHC done with the same assay conditions in 180 GO30140 group A and F samples and 199 IMbrave150 tissue samples. b, Prevalence and proportions of PD-L1 staining in either tumor cells (TC) or immune cells (IC). c, Discrete PD-L1 staining at canonical thresholds are tested with a Cox proportional hazards model for association with either overall survival (OS) or progression-free survival (PFS). Points and whiskers in the plot are hazard ratio and confidence interval, respectively. Atezo, atezolizumab; Bev, bevacizumab.
Forest plots of progression-free survival (PFS) (a) or overall survival (OS) (b) in IMbrave150 patients (n = 130) either possessing (Mut) or not possessing (WT) mutations in the indicated gene, for the six genes observed to be mutated in > 10% of patients in the study. Mutations are defined as somatic short variants that are either recurrent in cancer (known) or disrupt tumor suppressor genes or are in known hotspot regions (likely) by Foundation Medicine criteria. Statistics are calculated by a two-sided Cox model stratified with variables as stated elsewhere. Points and whiskers in the plot are hazard ratio and confidence interval, respectively. Atezo, atezolizumab; Bev, bevacizumab; NA, not achieved.
Extended Data Fig. 6 Efficacy and mechanism of action study of atezolizumab-bevacizumab in an immunogenic hepatocellular carcinoma mouse model.
a, Survival curves in C57BL/6 WT females. Number of mice per group is shown as well as median survival in days. Log-rank Mantel-Cox test. Treatment groups are compared with the control group (IgG1 + IgG2) and two-sided P value is adjusted. b-i, Quantification of (b) CD8 T cells, (c) proliferating CD8 T cells, (d) proliferating SIINFELKL-specific CD8 T cells, (e) granzyme B-expressing SIINFELKL-specific CD8 T cells, (f) proliferating Tregs, (g) ratio of proliferating Tregs, (h) conventional dendritic cells (cDCs), and (i) monocyte-derived macrophages (MoMa) in the livers of MYC-lucOS;CTNNB1 2 weeks after starting treatments (n = 5 biologically independent animals). Results of an analysis of variance test with multiple comparisons is shown. Mean and standard deviation (SD) are shown. j, Representative pictures of the stainings for CD8 (pink) and endomucin (brown) in tumor area quantified in (k). k & l, Number of CD8 + T cells and endomucin-positive vessels per mm2 of (k) tumor and (l) peritumor areas in MYC-lucOS; CTNNB1 mice treated with the corresponding treatments for 2 weeks (n = 2-5 biologically independent animals). An analysis of variance test with multiple comparisons is shown. Mean and standard deviation (SD) are shown. C, control (IgG1 + IgG2), P, anti–PD-L1; PV, anti–PD-L1 + anti-VEGF; Undef, undefined; V, anti-VEGF.
About this article
Cite this article
Zhu, A.X., Abbas, A.R., de Galarreta, M.R. et al. Molecular correlates of clinical response and resistance to atezolizumab in combination with bevacizumab in advanced hepatocellular carcinoma. Nat Med (2022). https://doi.org/10.1038/s41591-022-01868-2