The New Immuno-Oncology-Based Therapies and Their Perspectives in Hepatocellular Carcinoma

Simple Summary Hepatocellular carcinoma is a frequent and poor prognosis tumor, with most patients facing up, soon or later, to systemic therapies. So far, systemic therapies based on tyrosine kinase inhibitor monotherapies have been of modest benefit. The aim of this review article was to characterize the profile of efficacy and safety of immuno-oncology-based monotherapies that failed to demonstrate significant benefit, for comparison with the immuno-oncology-based combinational strategies. One of them has proven its drastic benefit in phase-3, whereas others have only shown promising data in phase-1/2, although the corresponding phase-3 results are pending. We showed that objective response rates and duration of response are important parameters for increased median overall survival and long survivals. We also pointed out that, being aware that there is an urgent unmet need for biomarkers, the pattern of safety and quality of life will guide the physician for the choice on the possible future combinations. Abstract Hepatocellular carcinoma is a poor prognosis tumor. Systemic therapies are frequently used due to frequent recurrences after surgical or radiologic treatments. Anti-angiogenic tyrosine kinase inhibitors have shown efficacy in monotherapy, but with very low rates of long survival and exceptional recovery. Immuno-oncology based on immune checkpoint inhibitors has revolutionized the systemic therapies since showing long survival rates without any tumor progression or recurrence for some patients in partial or complete response, and possibly for some patients in stable disease. However, the rate of responders under immuno-oncology monotherapy is too low to increase significantly the median overall survival of the treated patients. The immuno-oncology-based combinations with different types of immune checkpoint inhibitors (PD-1/PD-L1 and CTLA-4 inhibitors such as nivolumab, pembrolizumab, atezolizumab, durvalumab, ipilimumab, tremelimumab), or the association of immune checkpoint inhibitors plus anti-angiogenic agents (bevacizumab, lenvatinib, cabozantinib), have led to a breakthrough in the treatment of hepatocellular carcinoma. Indeed, the first phase-3 trial, combining atezolizumab with bevacizumab, has dramatically changed the outcome of patients. Data from several other types of combinations assessed in phase-3 trials are pending, and if positive, will drastically arm the physicians to efficiently treat the patients, and disrupt the current algorithm of hepatocellular carcinoma treatment.


Introduction
Hepatocellular carcinoma (HCC) is one of the most prevalent malignant tumors with an incidence reaching more than 840,000 new cases per year and ranks fourth in terms of cancer-related death (around 740,000 deaths) worldwide [1]. The main risk factors are B (HBV) and C (HCV) chronic viral hepatitis, alcohol abuse, non-alcoholic fatty liver disease, and generally speaking any kind of chronic liver injury and more especially of cirrhosis [2]. Unfortunately, HCC diagnosis is commonly late while the tumor has spread outside the liver parenchyma as portal vein invasion or distant metastasis. In the history of HCC patients, a substantial proportion of them will have to face up, early or later, to systemic therapies due to no longer compatibility with radical or loco-regional therapies. Cytotoxic chemotherapies and hormone therapies have never shown any significant benefit on overall survival (OS) [3,4]. The first systemic therapy having demonstrated a significant beneficial impact on HCC outcome is sorafenib, a tyrosine kinase inhibitor (TKI) harboring anti-angiogenic and anti-proliferative properties on HCC [5]. During almost ten years, all the systemic therapies tested in randomized controlled trials in first-line systemic therapy head-to-head vs. sorafenib, or in second line after failure of sorafenib, have not shown any significant benefit-i.e., brivanib, everolimus, sorafenib plus erlotinib, tivantinib, sunitinib, anti-glypican-3 antibodies [6][7][8][9][10][11][12]. In 2018, Kudo et al. have shown that lenvatinib, another TKI with anti-angiogenic properties, was at least equivalent to sorafenib in first line in the REFLECT non-inferiority study [13]. In subgroup analysis for OS, lenvatinib tended to differ from sorafenib, with better efficacy ( [13].
Other TKIs with anti-angiogenic and anti-cancerous properties on HCC cells have demonstrated efficacy in second line: regorafenib after progression under sorafenib in 2017 [14], and cabozantinib after progression or intolerance under sorafenib strategy in 2018 [15]. In subgroup analysis for OS, regorafenib tended to differ from placebo, with better efficacy ( [15].
Further, ramucirumab, a monoclonal antibody targeting vascular endothelial growth factor-2 (VEGFR-2), has shown significant benefit on a specific subpopulation of HCC patients, with high baseline AFP level ≥ 400 ng/mL [16]. In subgroup analysis for OS, ramucirumab tended to differ from placebo, with better efficacy ( [16].
However, all these options with TKI or ramucirumab monotherapies were strictly palliative with absence of long-term survivors and lack of potential recovery. Immunooncology approaches have completely revolutionized the paradigm of systemic therapies of HCC with nonetheless significant increase of median OS in combinational strategies, but also arising of possibility of long-term survivors and for some patients a hope of complete response and maybe definitive recovery. In this review article, we will focus on the advent of immuno-oncology currently assessed in phase-3 trials in the therapeutic strategies of HCC, the present standard of care of systemic therapy in first-line setting, and the potential strategies that will likely revolutionize the therapeutic algorithm of HCC therapy in a near future.
Quite similar data were obtained with pembrolizumab in the phase-1/2 Keynote-224 trial [18] with ORR of 17%, median DOR not reached (95% CI, 3.1-14.6+), DCR of 62%, median OS of 12.9 mo (95% CI, 9.7-15.5), and median PFS of 4.9 mo (95% CI, 3.4-7.2). However, very interestingly with nivolumab but not assessed so far with pembrolizumab, ORR seemed to be a reliable surrogate marker of outcome since median OS (95% CI) was: (i) strikingly high for complete and partial responders (CR/PR) since not reached; (ii) low and similar to placebo groups in clinical trials for progressive diseases (PD) at 8.9 mo (7.3-13.4); and (iii) intermediate at 16.7 mo (13.8-20.2) for stable diseases (SD) [19]. Of notice is the appearance of a late flat tail in the Kaplan-Meier OS curves with immuno-oncology, which represents the long-term survivors. This late flat tail will be further discussed later on in data from phase-3 trials.

The Same Immune Checkpoint Inhibitors Targeting PD-1 Were Disappointing in Phase-3 Trials
Nivolumab and pembrolizumab monotherapies have been assessed in phase-3 trials, either in first line for nivolumab or in second line for pembrolizumab. CheckMate-459 is a phase-3 prospective open trial that randomized in first-line nivolumab to sorafenib in the control arm [20]. The nivolumab arm showed ORR of 15% per Recist 1.1, median time to response (TTR) of 3.3 mo (range, 1.6-19.4), median DOR of 23.3 mo (95% CI, 3.1-34.5+), and DCR of 55%. However, outcomes were disappointing since they did not statistically differ from the sorafenib arm, regarding: (i) OS of 16.4 mo (95% CI, 13.9-18.4) vs. 14.7 mo (95% CI, 11.9-17.2), HR 0.85 (95% CI: 0.72-1.02), P = 0.0752; and (ii) PFS of 3.7 mo (95% CI, 3.1-3.9) vs. 3.8 mo (95% CI, 3.7-4.5), HR 0.93 (95% CI, 0.79-1.10). However, the late flat tail of the OS Kaplan-Meier curves tended to be better in the nivolumab than in the sorafenib arm, with median follow-up of 15.2 mo (range, 0.0-38.8) and 13.4 mo (range, 0.1-38.4), respectively, between arms. Illogically, a late flat tail was present in the sorafenib arm. One of the possible explanations could be that almost half of the patients who received a subsequent therapy after sorafenib withdrawal were treated by immuno-oncology. Treatment effect on OS in predefined subsets showed a trend for better benefit in: . Another remarkable data are the safety of nivolumab in comparison with sorafenib: fewer treatment-related adverse events (TRAE) of grade-3/4 (22% vs. 49% respectively), reason for discontinuation due to TRAE of 9% vs. 11%, and the clinically meaningful differences between treatment arms were observed for FACT Hep total in favor of nivolumab in terms of quality of life.
The same pattern of data was observed with pembrolizumab. Keynote-240 is a phase-3 prospective double-blinded trial that randomized in second-line pembrolizumab to placebo [21]. The pembrolizumab arm showed ORR of 18.3% per Recist 1.1, median TTR of 2.7 mo (range, 1.2-16.9), median DOR of 13.8 mo (95% CI, 1.5+-23.6+), and DCR of 62.2%. For nivolumab in CheckMate-459 [20], outcomes were disappointing, although with a strong trend for efficacy of pembrolizumab vs. placebo. However, the co-primary end-points did not reach the pre-specified P value required for statistical significance: (i) OS 13.9 mo (95% CI, 11.6-16.0) vs. for HBV and non-viral subsets. Identically to nivolumab, not much TRAE of grade-3/4 occurred with pembrolizumab (18.3%), and reason for discontinuation due to TRAE was 6.5%. Furthermore, pembrolizumab preserved quality of life during treatment following the European Organization for Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC QLQ-C30) and its HCC supplement (EORTC QLQ-HCC18) [22].

The Atezolizumab/Bevacizumab Combination Has Become the Gold-Standard in First-Line Systemic Therapy
Prior to IMbrave-150 [23], the GO30140 phase-1b study led to exciting data about the anti-cancer synergy of atezolizumab and bevacizumab combination [24]. Bevacizumab, in addition to its therapeutic anti-angiogenic properties by inhibiting VEGF, silences the immunosuppressive tumor microenvironment and thus enhances atezolizumab efficacy [25][26][27]. The GO30140 study included five cohorts, with groups A and F being dedicated to HCC. In group A, all patients received atezolizumab (1200 mg) and bevacizumab (15 mg/kg) every 3 weeks. In group F, patients were randomly assigned to receive atezolizumab (1200 mg) plus bevacizumab (15 mg/kg) every 3 weeks, or atezolizumab (1200 mg every 3 weeks) alone. In group A, atezolizumab/bevacizumab combination, in comparison to immuno-oncology monotherapies with nivolumab [20] or pembrolizumab [21], tended to show much higher levels of ORR (36% per Recist 1.1) with the same long DOR (median not reached (95% CI, 11.8 mo-NE), and higher DCR (71%) with a median follow-up of 12.4 mo (IQR, 8.0-16.2). Outcomes seemed better with atezolizumab/bevacizumab [24] than nivolumab [20]  IMbrave-150 has raised the atezolizumab/bevacizumab combination as the goldstandard of HCC in first line [23]. This is a phase-3 prospective open trial that randomized in first-line atezolizumab (1200 mg) + bevacizumab (15 mg/kg) every 3 weeks to sorafenib in the control arm. Such as in group A of the GO30140 phase-1b trial, anti-cancer properties of the atezolizumab/bevacizumab arm were exciting with ORR of 27% per Recist 1.1, median DOR not reached, and DCR of 73.6% with a median follow-up of 8.6 mo at the time of the primary analysis (Table 1) Another remarkable data is the safety of atezolizumab/bevacizumab in comparison with sorafenib: fewer TRAE of grade-3/4 in 36% vs. 46%, respectively, with different patterns of AE, thus explaining a substantial delay to deterioration of patient-reported quality of life, physical functioning, and role functioning using the EORTC QLQ-C30. However, reason for discontinuation due to AE of any cause was 15.5% for atezolizumab/bevacizumab vs. 10.3% for sorafenib, although the same data from TRAE are unavailable (Table 3).

Conclusions and Perspectives
The phase-1/2 of immuno-oncology-based combination strategies have shown exciting data in the systemic treatment of advanced HCC, letting arising hopes for very long survivals and maybe even recovery for some patients. So far, only the IMbrave-150 trial has proven the superiority of atezolizumab/bevacizumab compared to a TKI, sorafenib. If the ongoing phase-3 trials become positive (CheckMate-9DW, HIMALAYA, LEAP-002, and COSMIC-312), hepato-oncologists will dispose of a large diversity of strategies in first-line setting. It is likely that the choice of a specific combination will be driven by its pattern of efficacy and tolerance, fitting with the profile of each patient. Of evidence, reliable biomarkers (tumor genetics or epigenetics, genetic polymorphism, immunophenotyping of HCC, . . . ) predictive of efficacy of a specific immuno-oncology-based combination would be of huge help to make such a decision for personalized medicine, but none has been validated so far. At the moment, patients progressing under atezolizumab/bevacicumab combination are devoted to TKI monotherapy. However, clinical trials with immunooncology-based combinations will be soon carried out in second-line setting. Furthermore, immuno-oncology-based combinations might also change the therapeutic strategies of early HCCs in a neo-adjuvant and/or adjuvant setting, or improve the management of intermediate HCCs in association with, or in place of transarterial chemoembolization. Thus, the next coming years will be of great interest and might revolutionize the treatment of HCC and transform the poor prognosis of this disease.  treatment-related adverse event PD-1/PD-L1 programmed cell death protein 1 and programmed death-ligand 1 CTLA-4 cytotoxic T lymphocyte antigen 4