Combined PD-1/VEGFR Blockade: A New Era of Treatment for Hepatocellular Cancer

In this issue of Clinical Cancer Research, Xu and colleagues (1) present the results of RESCUE, a phase II trial of the programmed cell death-1 (PD-1) inhibitor camrelizumab coadministered with apatinib, a VEGF receptor-2 (R2) antagonist for the treatment of advanced hepatocellular cancer (HCC).

The progression of HCC relies on T-cell exhaustion, with the PD-1 pathway acting as a master negative regulator of T-cell survival and proliferation within the tumor microenvironment (TME). Therapeutic modulation of this pathway leads to measurable responses in <20% of patients, defining HCC as a moderately immune-responsive malignancy (2). In recent years, PD-1 inhibitors failed to convincingly demonstrate overall survival (OS) benefit over sorafenib or placebo. Disease heterogeneity and postprogression sequencing of multiple lines of tyrosine kinase inhibitors (TKI) with immune checkpoint inhibitors (ICI) are likely to have influenced OS estimates in recent clinical trials (3).

Hypoxia and aberrant neovasculogenesis do not just underlie progression but are also diagnostic hallmarks of HCC, where progressive arterialization of blood inflow is the cornerstone of noninvasive diagnostic criteria utilized in clinical practice.

VEGF deprivation has long-standing evidence of clinical activity in HCC (4). However, the clinical development of the forerunner VEGF inhibitor bevacizumab was not pursued as a monotherapy in this indication, in view of the emergence of sorafenib as an orally available therapy with demonstrated efficacy against placebo.

The rapid expansion of cancer immunotherapy has reignited interest in VEGF pathway inhibition, given its potential to broaden the reach of efficacy of ICI in disease areas where anti-PD-1 monotherapy proved ineffective. The immunologic basis of synergy between VEGF and PD-1 pathway inhibition is being increasingly characterized. Normalization of vascular permeability leading to effective T-cell extravasation, inhibition of the suppressive myeloid and lymphoid compartment within the TME alongside a permissive role on dendritic cell maturation are among the complex elements of a cross-talk between the tumor and the host that is yet to be fully understood (5).

In the clinic, addition of bevacizumab to the programmed cell death ligand-1 (PD-L1) inhibitor atezolizumab synergistically extended progression-free survival (PFS) in the randomized arm F of GO30140 (5.6 vs. 3.4 months, P = 0.0108), the phase Ib study that paved the way to the subsequent approval this combination as a novel standard of care in first line on the basis of IMbrave150 (6). Independent evidence of synergistic efficacy from VEGF/PD-1 coinhibition comes from the pembrolizumab/lenvatinib combination, characterized by overall response rate (ORR) of 36% and a median PFS of 8.3 months by the RECIST (7).

The RESCUE study adds meaningful clinical evidence to the rapidly evolving portrait of combination immunotherapy in HCC. By testing the camrelizumab/apatinib combination in two independent open-label cohorts of treatment-naïve (n = 70) and treatment-experienced patients with HCC (n = 120), RESCUE provides important evidence of efficacy of PD-1/VEGFR2 blockade by demonstrating competitive ORRs of 34.3% and 22.5% across first and second line by RECIST and 45.7% and 25% by modified RECIST criteria, disease control rates in excess of 75% across cohorts and a median PFS of 5.7 in first and 5.5 months in second line. Objective responses in the study occurred 2 months from treatment initiation, lasted for over 14.8 months and were associated with the achievement of longer OS across cohorts (12 months OS rate 74.7% and 68.2%). Accounting for interstudy heterogeneity, these data compare favorably with first-line immunotherapy combinations such as atezolizumab/bevacizumab (ref. 8; ORR 27.3%, PFS 6.8 months) or pembrolizumab/lenvatinib (ref. 7; ORR 36%, PFS 8.3 months) and provide new evidence to suggest that PD-1/VEGFR2 blockade is active after prior sorafenib exposure. These findings are of particular translational importance given that many of the immunomodulatory properties of sorafenib, including reduction of regulatory T cells and myeloid-derived suppressor cells are dependent on its VEGF/VEGFR inhibitory capacity on the TME (9).

In addition to promising efficacy, the overall safety profile of the camrelizumab/apatininb combination appears acceptable in a patient population where well-controlled hepatitis B Virus infection represented the leading etiology of chronic liver disease. In RESCUE, severe treatment-emerging abnormalities in hepatic function, a key safety aspect in patients with hepatitis and underlying cirrhosis, were uncommon (10%) and outranked by hypertension (34%), a renowned class effect of VEGF blockade. Incidence of bleeding, in a patient group screened for evidence of gastrointestinal haemorrhage within 3 months from enrolment, was low (1.6%). With a low rate of permanent discontinuations due to toxicity (12%), the camrelizumab/apatinib regimen appears a clinically deliverable combination, although the need for dose interruptions/modifications in 72% of the patients suggests the need for individualized monitoring in the outpatient clinic.

An additional and important answer generated by the study by Xu and collaborators is confirmation of efficacy of PD-1/VEGF blockade in Chinese patients with HCC, a population where responses to PD-1 inhibitors in other indications such as malignant melanoma do not consistently replicate those seen in Western patients (10).

As treatment options continue to expand in advanced HCC a key question that remains unanswered relates to optimal patient selection for the most appropriate systemic therapy in the individual patient. Undeniably, adequate TKI sequencing remains a beneficial option to a subset of patients with advanced HCC where median OS can approach 26 months from initiation of therapy. On the other hand, PD-1 and CTL-associated protein 4 (CTLA-4) inhibitor combinations provide ORRs in the region of 27%–32% in treatment-experienced patients, much higher than the 17% observed for PD-1 monotherapies in second line. A number of PD-1/TKI combinations are on the horizon, including pembrolizumab/lenvatinib, atezolizumab/cabozantinib, pembrolizumab/regorafenib, and many others.

Despite showing evidence of activity, no combination therapy other than atezolizumab and bevacizumab is currently supported by evidence of OS improvement, demonstration of which is crucial for these therapies to become globally available. Achievement of an HR for PFS ≤0.6 reproducibly predicts for OS, and this threshold should guide phase II to III development (11). Another area of concern is the superadded toxicity associated with multitargeted TKIs compared with pure VEGF deprivation. The 67% rate of grade ≥3 treatment-related adverse events seen following pembrolizumab/lenvatinib therapy is an eloquent example of how toxicity is likely to inform future treatment decisions in an oncological indication where predictive biomarkers are lacking. In RESCUE, PD-L1 expression was not associated with response, further confirming the lack of reliability of this biomarker in HCC.

In the not so distant future, parallel reporting of a number of phase II/III studies comparing combination immunotherapy regimens against sorafenib or lenvatinib is likely to further expand treatment options without necessarily improving the long-standing practice of empirical prescribing in HCC. Qualification of predictive biomarkers to discriminate TKI from ICI responders (Fig. 1) remains a key unmet need, made now even more pressing by the expanding number of sequential treatment options in advanced HCC.

Lessons learnt from melanoma and renal carcinoma highlight clinical features (patient fitness and comorbid burden) to play a major role in determining candidacy to combination immunotherapies—a paradoxical shift from the promise of personalized medicine that we hoped immunotherapy would bring. Despite the number of unanswered questions, the RESCUE study represents a solid step forward in the qualification of a novel immunotherapy combination in advanced HCC. RESCUE highlights for the first time that PD-1/VEGFR targeting preserves efficacy post-sorafenib exposure, suggesting noncross resistance of the two therapeutic approaches. As phase III development programs are ongoing, combination immunotherapy receives, with this study, important confirmatory evidence of safety and efficacy in Chinese patients with advanced HCC.

D. J. Pinato reports personal fees from ViiV Healthcare, Bayer Healthcare, Roche, MiNa Therapeutics, EISAI, H3B, AstraZeneca, and DaVolterra, and grants and personal fees from Bristol-Myers Squibb and MSD outside the submitted work. A. Cortellini reports personal fees from AstraZeneca, Roche, MSD, Bristol-Myers Squibb, Astellas, and Novartis outside the submitted work. L. Rimassa reports personal fees for consulting/advisory role from Amgen, ArQule, AstraZeneca, Basilea, Bayer, Bristol-Myers Squibb, Celgene, Eisai, Exelixis, Genenta, Hengrui, Incyte, Ipsen, Lilly, MSD, Nerviano Medical Sciences, Roche, and Sanofi, personal fees for lectures from AbbVie, Amgen, Eisai, Gilead, Incyte, Ipsen, Lilly, Roche, and Sanofi, nonfinancial support (travel expenses) from Ipsen, and grants to institution from Agios, ARMO BioSciences, AstraZeneca, BeiGene, Eisai, Exelixis, Fibrogen, Incyte, Ipsen, Lilly, MSD, Nerviano Medical Sciences, Roche, and Zymeworks outside the submitted work. No disclosures were reported by the other author.

The authors would like to acknowledge the Imperial College National Institute for Health Research Biomedical Research Centre and the Imperial Experimental Cancer Medicine Centre. D.J. Pinato is supported by grant funding from the Wellcome Trust Strategic Fund (PS3416), by ASCO/Conquer Cancer Foundation Global Oncology Young Investigator Award 2019 (14704), and by the Cancer Research UK postdoctoral bursary grant (C57701/A26137).