International Hepatology
Systemic therapy in HCC: Lessons from brivanib David Bolos, Richard S. Finn⇑ Department of Medicine, Division of Hematology/Oncology, Geffen School of Medicine at UCLA, USA
COMMENTARY ON: Brivanib versus sorafenib as first-line therapy in patients with unresectable, advanced hepatocellular carcinoma: results from the randomized phase III BRISK-FL study. Johnson PJ, Qin S, Park JW, Poon RT, Raoul JL, Philip PA, Hsu CH, Hu TH, Heo J, Xu J, Lu L, Chao Y, Boucher E, Han KH, Paik SW, Robles-Aviña J, Kudo M, Yan L, Sobhonslidsuk A, Komov D, Decaens T, Tak WY, Jeng LB, Liu D, Ezzeddine R, Walters I, Cheng AL. J Clin Oncol. 2013 Oct 1;31(28):3517–24. Reprinted with permission. Ó (2014) American Society of Clinical Oncology. All rights reserved. http://www.ncbi.nlm.nih.gov/pubmed/23980084 Abstract: PURPOSE: Brivanib is a dual inhibitor of vascular-endothelial growth factor and fibroblast growth factor receptors that are implicated in the pathogenesis of hepatocellular carcinoma (HCC). Our multinational, randomized, double-blind, phase III trial compared brivanib with sorafenib as first-line treatment for HCC. PATIENTS AND METHODS: Advanced HCC patients who had no prior systemic therapy were randomly assigned (ratio, 1:1) to receive sorafenib 400 mg twice daily orally (n = 578) or brivanib 800 mg once daily orally (n = 577). Primary end point was overall survival (OS). Secondary end points included time to progression (TTP), objective response rate (ORR), disease control rate (DCR) based on modified Response Evaluation Criteria in Solid Tumors (mRECIST), and safety. RESULTS: The primary end point of OS noninferiority for brivanib versus sorafenib in the per-protocol population (n = 1150) was not met (hazard ratio [HR], 1.06; 95.8% CI, 0.93 to 1.22), based on the prespecified margin (upper CI limit for HR 61.08). Median OS was 9.9 months for sorafenib and 9.5 months for brivanib. TTP, ORR, and DCR were similar between the study arms. Most frequent grade 3/4 adverse events for sorafenib and brivanib were hyponatremia (9% and 23%, respectively), AST elevation (17% and 14%), fatigue (7% and 15%), hand-foot-skin reaction (15% and 2%), and hypertension (5% and 13%). Discontinuation as a result of adverse events was 33% for sorafenib and 43% for brivanib; rates for dose reduction were 50% and 49%, respectively.
Keywords: Brivanib; Hepatocellular carcinoma; Sorafenib; Fibroblast Growth Factor; FGFR; Vascular Endothelial Growth Factor Receptor; VEGFR; Angiogenesis. Received 7 May 2014; received in revised form 11 June 2014; accepted 18 June 2014 ⇑ Corresponding author. Address: UCLA Translational Oncology, 2825 Santa Monica Blvd, Suite 200, Santa Monica, CA 90404, USA. Tel.: +1 310 586 2092; fax: +1 310 586 6830. E-mail address: Rfi[email protected] (R.S. Finn).
CONCLUSION: Our study did not meet its primary end point of OS noninferiority for brivanib versus sorafenib. However, both agents had similar antitumor activity, based on secondary efficacy end points. Brivanib had an acceptable safety profile, but was less welltolerated than sorafenib. AND Brivanib in patients with advanced hepatocellular carcinoma who were intolerant to sorafenib or for whom sorafenib failed: results from the randomized phase III BRISK-PS study. Llovet JM, Decaens T, Raoul JL, Boucher E, Kudo M, Chang C, Kang YK, Assenat E, Lim HY, Boige V, Mathurin P, Fartoux L, Lin DY, Bruix J, Poon RT, Sherman M, Blanc JF, Finn RS, Tak WY, Chao Y, Ezzeddine R, Liu D, Walters I, Park JW. J Clin Oncol. 2013 Oct 1;31(28):3509–16. Reprinted with permission. Ó (2014) American Society of Clinical Oncology. All rights reserved. http://www.ncbi.nlm.nih.gov/pubmed/23980090 Abstract: PURPOSE: Brivanib is a selective dual inhibitor of vascular endothelial growth factor and fibroblast growth factor receptors implicated in tumorigenesis and angiogenesis in hepatocellular carcinoma (HCC). An unmet medical need persists for patients with HCC whose tumors do not respond to sorafenib or who cannot tolerate it. This multicenter, double-blind, randomized, placebo-controlled trial assessed brivanib in patients with HCC who had been treated with sorafenib. PATIENTS AND METHODS: In all, 395 patients with advanced HCC who progressed on/after or were intolerant to sorafenib were randomly assigned (2:1) to receive brivanib 800 mg orally once per day plus best supportive care (BSC) or placebo plus BSC. The primary end point was overall survival (OS). Secondary end points included time to progression (TTP), objective response rate (ORR), and disease control rate based on modified Response Evaluation Criteria in Solid Tumors (mRECIST) and safety. RESULTS: Median OS was 9.4 months for brivanib and 8.2 months for placebo (hazard ratio [HR], 0.89; 95.8% CI, 0.69 to 1.15; P = .3307). Adjusting treatment effect for baseline prognostic factors yielded an OS HR of 0.81 (95% CI, 0.63 to 1.04; P = .1044). Exploratory analyses showed a median time to progression of 4.2 months for brivanib and 2.7 months for placebo (HR, 0.56; 95% CI, 0.42 to 0.76; P < .001), and an mRECIST ORR of 10% for brivanib and 2% for placebo (odds ratio, 5.72). Study discontinuation due to treatment-related adverse events (AEs) occurred in 61 brivanib patients (23%) and nine placebo patients (7%). The most frequent treatment-related grade 3 to 4 AEs for brivanib included hypertension (17%), fatigue (13%), hyponatremia (11%), and decreased appetite (10%).
Journal of Hepatology 2014 vol. 61 j 947–950
International Hepatology CONCLUSION: In patients with HCC who had been treated with sorafenib, brivanib did not significantly improve OS. The observed benefit in the secondary outcomes of TTP and ORR warrants further investigation. Ó 2014 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Hepatocellular carcinoma (HCC) remains a major health problem worldwide, and is the sixth most common cancer and third leading cause of cancer-related mortality globally [1]. The development of systemic agents in advanced hepatocellular carcinoma (HCC) has been a challenge. The results of the SHARP study and approval of sorafenib in 2007 was a giant step-forward in our management of the disease and how we design and conduct studies for this population [2]. However, since then we have been mired in numerous negative phase III studies [3]. One of the first agents to be evaluated in the post sorafenib era was brivanib, an oral, potent selective inhibitor of both the fibroblast growth factor (FGF) and vascular endothelial growth factor (VEGF) family of receptors (Table 1) [4]. Laboratory evidence suggests that both pathways play a role in the pathogenesis of HCC and, in context of the FGFR family, in the resistance to VEGF driven angiogenesis [5]. Despite encouraging phase II data, enpoints were not met in two randomised phase III registration studies (Table 2). To avoid similar disappointment in the future, we should look carefully at these studies to identify potential flaws in trial design and assumptions. Given that these are two different populations of patients, the front-line and second-line groups should be considered separately. Brivanib was initially evaluated in two single-arm open label phase II studies of patients with advanced HCC. The front-line study enrolled 55 patients with advanced HCC (BCLC B/C 11%/89%) which were treated in the study with the primary
Table 1. Brivanib targeted tyrosine kinases and inhibition profile.
Tyrosine kinase VEGF-R1 VEGF-R2 VEGF-R3 Flk-1 FGF-R1 FGF-R2 Tie-2
IC50 (nM) of brivanib 390 34 10 23 148 125 445
objective being 6-month progression-free survival (PFS) and progression free survival rate, and the secondary objectives of tumour response rate, median progression free and overall survival, and safety tolerability [6]. Using the modified World Health Organization Criteria (mWHO) the 6-month progression free survival (PFS) rate was 18.2% (95% CI, 9.1–30.9%), median PFS was 2.7 months (95% CI, 1.4–3.0) time to progression (TTP) was 2.8 months (95% CI, 1.4–3.5), and median overall survival (OS) was 10 months (95% CI, 6.8–15.2 months). Given the control arm from SHARP (placebo), these data would suggest that brivanib is an active drug in HCC, though if compared to the experimental arm in SHARP, it would seem comparable to sorafenib in efficacy. However, the brivanib study enrolled 64% Asian patients. In that context, if these efficacy data were compared to the Asia-Pacific sorafenib phase III study [7], a median OS of 10 months compares very favourably to the median OS of 6.5 months seen in the sorafenib arm in that study. With regard to toxicity, the adverse effects with brivanib were generally manageable with fatigue, hypertension, diarrhoea, nausea, and anorexia being most common and notably, the hand-foot skin reaction common with sorafenib, is uncommon with brivanib. After sorafenib’s approval for front-line therapy, the brivanib study was amended and evaluated brivanib in patients with progression on one line of prior therapy [8]. Again, this was an open label single-arm study. 46 patients were enrolled, 94% of them having received prior sorafenib, and again the majority being BCLC C (96%) and Asian (67%). Median TTP per mWHO criteria was 1.77 months (95% CI, 1.38–4.01), median PFS 2.73 months (95% CI, 1.41–3.91) and median OS was 9.79 months (95% CI, 5.52–13.57). For the first time, in a newly defined population of HCC patients, for those that had prior sorafenib, we were seeing a median OS of almost 10 months; a number that would not have been expected from interpreting the SHARP or Asia-Pacific datasets when comparing the differences between median OS and median TTP. One could have postulated that the median OS after stopping sorafenib was approximately 5 months from the SHARP study and just under 4 months in the Asia-Pacific study, and so the OS of 9.79 with brivanib was provocative in a population of patients with a great unmet need. In addition, both phase II studies included an assessment of response using the modified RECIST criteria for HCC in an attempt to better capture the clinical activity of brivanib [9]. With these data in hand, three global phase III randomised studies were performed with brivanib in HCC; a front-line study (BRISK-FL), a post-sorafenib study (BRISK-PS), and a third in combination with chemoembolisation (BRISK-TA). The first randomised second-line phase III study in the post sorafenib era to present was the BRISK-PS study [10]. This was
Table 2. Comparison of brivanib’s efficacy in phase II and phase III studies.
DCR (%) TTP (mo) PFS (mo) mOS (mo)
Phase II Front-line* 51 2.8 2.7 10
Phase III BRISK-FL^ 66 4.2 n.r. 9.5
Phase II Second-line* 45.7 1.77 2.0 9.79
⁄
Modified WHO criteria per independent radiology review. Modified RECIST criteria. n.r., not reported. DCR, disease control rate; TTP, time to progression; PFS, progression free survival; mOS, median overall survival.
^
948
Journal of Hepatology 2014 vol. 61 j 947–950
Phase III BRISK-PS^ 61 4.2 n.r. 9.4
JOURNAL OF HEPATOLOGY a global phase III, double-blind placebo control study of 395 patients randomised 2:1 that had progressed on or after sorafenib or were intolerant to sorafenib. The primary endpoint for the study was OS with a minimum hazard ratio of 0.78 being required for the study to be considered positive. Patients were stratified based on reason for sorafenib discontinuation and other strata commonly used in front-line HCC studies such as ECOGperformance status and the presence of extrahepatic spread, and/or vascular invasion. While the study was generally well balanced for these strata, there was an imbalance favouring the control arm in the number of patients that had vascular invasion, which was masked by the fact that the majority of patients had extrahepatic spread. In addition baseline a-fetoprotein (AFP), while not included in the stratification but a recognised negative prognostic factor in HCC, was increased in patients receiving brivanib. This study missed its primary endpoint; while median OS was better with brivanib, 8.2 months with placebo vs. 9.4 months with brivanib (HR, 0.89; 95.8% CI, 0.69–1.15; p = 0.3307) this was not statistically significant. Importantly, median OS in the phase III study was similar to that seen in the phase II for this population (9.4 months vs. 9.70). Time to progression (TTP) was significantly increased with brivanib as well as other secondary endpoints such objective response rate and disease control rate, all assessed using mRECIST. The side effect profile of brivanib was similar as in the phase II studies and was accompanied by a high rate of dose interruptions and reductions and a discontinuation rate of 23% secondary to AEs. Similarly, the BRISK-FL study did not meet its primary endpoint of non-inferiority in improving OS compared to sorafenib [11]. This study enrolled over 1100 patients and randomised them 1:1 in a double-blind fashion to brivanib or sorafenib and stratified them similarly as the SHARP study; by ECOG-PS, extrahepatic spread, and/or vascular invasion, and study site. Again, median OS with brivanib was similar as in the phase II study, 9.5 months vs. 10 months in phase III and II, respectively. However, this number did not differ significantly from the sorafenib survival of 9.9 months (HR, 1.06; 95.8% CI, 0.93–1.22) and exceeded the upper-limit HR 95% CI of 1.08 to indicate non-inferiority. Unlike the BRISK-PS study, secondary endpoints were similar using mRECIST in the BRISK-FL study, though the ORR with sorafenib of 9% with mRECIST was higher than that reported in the SHARP study using RECIST (2%). Again, the AE profile of brivanib was similar to the phase II studies. While the frequency of dose adjustments and interruptions was similar between the two arms, there were more discontinuations with brivanib due to adverse events, 43% with brivanib as compared to 33% with sorafenib. This fact highlights another relevant fact for drug development in HCC: systemic toxicity is not as well tolerated as in other malignancies. The failure of sunitinib in this population also echoes this observation [12]. So what can we learn from the development of brivanib? One, control arms are important, not only for phase III studies but for phase II as well. It could be argued that a randomised phase II study would have identified the equivalence of brivanib to sorafenib in the front-line setting making the BRISK-FL study less appealing. In addition, for the BRISK-PS study, a randomised phase II study would have given a better sense of the actual survival of the placebo group and possibly identify other important factors effecting survival in the post-sorafenib setting. Second, and along those lines, strata used in the front-line setting may not be applicable in the second-line setting. Perhaps, separating
extrahepatic-spread from vascular invasion is an important consideration in the second-line setting? Only recently, two publications identified prognosticators of outcome after sorafenib progression [13,14], one of which was just meeting the inclusion for a second-line study (i.e., selection bias). Third, is developing a better assessment tool for benefit in clinical trials with molecular targeted therapeutics? From the SHARP study and these prospective trials, it would appear that neither response by RECIST nor mRECIST correlates with outcome in HCC and further refinement of imaging endpoints need to be developed. In addition, perhaps it is time to move beyond anti-angiogenic agents in HCC. To date, no agents in this class, like brivanib, have been able to improve on sorafenib in the front-line or improve survival in the second-line setting. This is not to say these agents might not have done better than placebo in the front-line but as it stands, they need to do better than sorafenib or at least do as well with a more favourable toxicity profile. After sorafenib, we need to demand both a strong clinical rationale and biologic rationale to move agents with a similar mechanism of action into this space. Finally, drug development in HCC has suffered from the clinical heterogeneity of the disease and its molecular heterogeneity. More work on predictive markers of response needs to be done in HCC. While an ‘‘all-comers’’ approach worked for sorafenib, if we want to move forward in HCC with molecular therapeutics, we need not only better clinical trial tools and design but we also need to better select patients most likely to benefit from a given agent.
Conflict of interest RSF has served as a consultant to Bristol Myers Squibb, Onyx, Bayer, and Novartis. References [1] Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69–90. [2] Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008;359:378–390. [3] Llovet JM, Hernandez-Gea V. Hepatocellular carcinoma: reasons for phase III failure and novel perspectives on trial design. Clin Cancer Res 2014;20:2072–2079. [4] Bhide RS, Cai ZW, Zhang YZ, Qian L, Wei D, Barbosa S, et al. Discovery and preclinical studies of (R)-1-(4-(4-fluoro-2-methyl-1H-indol-5-yloxy)-5methylpyrrolo[2,1-f][1,2,4]triazin-6-yloxy)propan-2-ol (BMS-540215), an in vivo active potent VEGFR-2 inhibitor. J Med Chem 2006;49:2143–2146. [5] Chou T, Finn RS. Brivanib: a review of development. Future Oncol 2012;8:1083–1090. [6] Park JW, Finn RS, Kim JS, Karwal M, Li RK, Ismail F, et al. Phase II, open-label study of brivanib as first-line therapy in patients with advanced hepatocellular carcinoma. Clin Cancer Res 2011;17:1973–1983. [7] Cheng AL, Kang YK, Chen Z, Tsao CJ, Qin S, Kim JS, et al. Efficacy and safety of sorafenib in patients in the Asia-Pacific region with advanced hepatocellular carcinoma: a phase III randomised, double-blind, placebo-controlled trial. Lancet Oncol 2009;10:25–34. [8] Finn RS, Kang YK, Mulcahy M, Polite BN, Lim HY, Walters I, et al. Phase II, open-label study of brivanib as second-line therapy in patients with advanced hepatocellular carcinoma. Clin Cancer Res 2012;18:2090–2098. [9] Lencioni R, Llovet JM. Modified RECIST (mRECIST) assessment for hepatocellular carcinoma. Semin Liver Dis 2010;30:52–60. [10] Llovet JM, Decaens T, Raoul JL, Boucher E, Kudo M, Chang C, et al. Brivanib in patients with advanced hepatocellular carcinoma who were intolerant to sorafenib or for whom sorafenib failed: results from the randomized phase III BRISK-PS study. J Clin Oncol 2013;31:3509–3516. [11] Johnson PJ, Qin S, Park JW, Poon RT, Raoul JL, Philip PA, et al. Brivanib vs. sorafenib as first-line therapy in patients with unresectable, advanced
Journal of Hepatology 2014 vol. 61 j 947–950
949
International Hepatology hepatocellular carcinoma: results from the randomized phase III BRISK-FL study. J Clin Oncol 2013;31:3517–3524. [12] Cheng AL, Kang YK, Lin DY, Park JW, Kudo M, Qin S, et al. Sunitinib versus sorafenib in advanced hepatocellular cancer: results of a randomized phase III trial. J Clin Oncol 2013;31:4067–4075. [13] Reig M, Rimola J, Torres F, Darnell A, Rodriguez-Lope C, Forner A, et al. Postprogression survival of patients with advanced hepatocellular
950
carcinoma: rationale for second-line trial design. Hepatology 2013;58:2023–2031. [14] Shao YY, Wu CH, Lu LC, Chan SY, Ma YY, Yen FC, et al. Prognosis of patients with advanced hepatocellular carcinoma who failed first-line systemic therapy. J Hepatol 2014;60:313–318.
Journal of Hepatology 2014 vol. 61 j 947–950
Systemic therapy in HCC: Lessons from brivanib David Bolos, Richard S. Finn⇑ Department of Medicine, Division of Hematology/Oncology, Geffen School of Medicine at UCLA, USA
COMMENTARY ON: Brivanib versus sorafenib as first-line therapy in patients with unresectable, advanced hepatocellular carcinoma: results from the randomized phase III BRISK-FL study. Johnson PJ, Qin S, Park JW, Poon RT, Raoul JL, Philip PA, Hsu CH, Hu TH, Heo J, Xu J, Lu L, Chao Y, Boucher E, Han KH, Paik SW, Robles-Aviña J, Kudo M, Yan L, Sobhonslidsuk A, Komov D, Decaens T, Tak WY, Jeng LB, Liu D, Ezzeddine R, Walters I, Cheng AL. J Clin Oncol. 2013 Oct 1;31(28):3517–24. Reprinted with permission. Ó (2014) American Society of Clinical Oncology. All rights reserved. http://www.ncbi.nlm.nih.gov/pubmed/23980084 Abstract: PURPOSE: Brivanib is a dual inhibitor of vascular-endothelial growth factor and fibroblast growth factor receptors that are implicated in the pathogenesis of hepatocellular carcinoma (HCC). Our multinational, randomized, double-blind, phase III trial compared brivanib with sorafenib as first-line treatment for HCC. PATIENTS AND METHODS: Advanced HCC patients who had no prior systemic therapy were randomly assigned (ratio, 1:1) to receive sorafenib 400 mg twice daily orally (n = 578) or brivanib 800 mg once daily orally (n = 577). Primary end point was overall survival (OS). Secondary end points included time to progression (TTP), objective response rate (ORR), disease control rate (DCR) based on modified Response Evaluation Criteria in Solid Tumors (mRECIST), and safety. RESULTS: The primary end point of OS noninferiority for brivanib versus sorafenib in the per-protocol population (n = 1150) was not met (hazard ratio [HR], 1.06; 95.8% CI, 0.93 to 1.22), based on the prespecified margin (upper CI limit for HR 61.08). Median OS was 9.9 months for sorafenib and 9.5 months for brivanib. TTP, ORR, and DCR were similar between the study arms. Most frequent grade 3/4 adverse events for sorafenib and brivanib were hyponatremia (9% and 23%, respectively), AST elevation (17% and 14%), fatigue (7% and 15%), hand-foot-skin reaction (15% and 2%), and hypertension (5% and 13%). Discontinuation as a result of adverse events was 33% for sorafenib and 43% for brivanib; rates for dose reduction were 50% and 49%, respectively.
Keywords: Brivanib; Hepatocellular carcinoma; Sorafenib; Fibroblast Growth Factor; FGFR; Vascular Endothelial Growth Factor Receptor; VEGFR; Angiogenesis. Received 7 May 2014; received in revised form 11 June 2014; accepted 18 June 2014 ⇑ Corresponding author. Address: UCLA Translational Oncology, 2825 Santa Monica Blvd, Suite 200, Santa Monica, CA 90404, USA. Tel.: +1 310 586 2092; fax: +1 310 586 6830. E-mail address: Rfi[email protected] (R.S. Finn).
CONCLUSION: Our study did not meet its primary end point of OS noninferiority for brivanib versus sorafenib. However, both agents had similar antitumor activity, based on secondary efficacy end points. Brivanib had an acceptable safety profile, but was less welltolerated than sorafenib. AND Brivanib in patients with advanced hepatocellular carcinoma who were intolerant to sorafenib or for whom sorafenib failed: results from the randomized phase III BRISK-PS study. Llovet JM, Decaens T, Raoul JL, Boucher E, Kudo M, Chang C, Kang YK, Assenat E, Lim HY, Boige V, Mathurin P, Fartoux L, Lin DY, Bruix J, Poon RT, Sherman M, Blanc JF, Finn RS, Tak WY, Chao Y, Ezzeddine R, Liu D, Walters I, Park JW. J Clin Oncol. 2013 Oct 1;31(28):3509–16. Reprinted with permission. Ó (2014) American Society of Clinical Oncology. All rights reserved. http://www.ncbi.nlm.nih.gov/pubmed/23980090 Abstract: PURPOSE: Brivanib is a selective dual inhibitor of vascular endothelial growth factor and fibroblast growth factor receptors implicated in tumorigenesis and angiogenesis in hepatocellular carcinoma (HCC). An unmet medical need persists for patients with HCC whose tumors do not respond to sorafenib or who cannot tolerate it. This multicenter, double-blind, randomized, placebo-controlled trial assessed brivanib in patients with HCC who had been treated with sorafenib. PATIENTS AND METHODS: In all, 395 patients with advanced HCC who progressed on/after or were intolerant to sorafenib were randomly assigned (2:1) to receive brivanib 800 mg orally once per day plus best supportive care (BSC) or placebo plus BSC. The primary end point was overall survival (OS). Secondary end points included time to progression (TTP), objective response rate (ORR), and disease control rate based on modified Response Evaluation Criteria in Solid Tumors (mRECIST) and safety. RESULTS: Median OS was 9.4 months for brivanib and 8.2 months for placebo (hazard ratio [HR], 0.89; 95.8% CI, 0.69 to 1.15; P = .3307). Adjusting treatment effect for baseline prognostic factors yielded an OS HR of 0.81 (95% CI, 0.63 to 1.04; P = .1044). Exploratory analyses showed a median time to progression of 4.2 months for brivanib and 2.7 months for placebo (HR, 0.56; 95% CI, 0.42 to 0.76; P < .001), and an mRECIST ORR of 10% for brivanib and 2% for placebo (odds ratio, 5.72). Study discontinuation due to treatment-related adverse events (AEs) occurred in 61 brivanib patients (23%) and nine placebo patients (7%). The most frequent treatment-related grade 3 to 4 AEs for brivanib included hypertension (17%), fatigue (13%), hyponatremia (11%), and decreased appetite (10%).
Journal of Hepatology 2014 vol. 61 j 947–950
International Hepatology CONCLUSION: In patients with HCC who had been treated with sorafenib, brivanib did not significantly improve OS. The observed benefit in the secondary outcomes of TTP and ORR warrants further investigation. Ó 2014 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Hepatocellular carcinoma (HCC) remains a major health problem worldwide, and is the sixth most common cancer and third leading cause of cancer-related mortality globally [1]. The development of systemic agents in advanced hepatocellular carcinoma (HCC) has been a challenge. The results of the SHARP study and approval of sorafenib in 2007 was a giant step-forward in our management of the disease and how we design and conduct studies for this population [2]. However, since then we have been mired in numerous negative phase III studies [3]. One of the first agents to be evaluated in the post sorafenib era was brivanib, an oral, potent selective inhibitor of both the fibroblast growth factor (FGF) and vascular endothelial growth factor (VEGF) family of receptors (Table 1) [4]. Laboratory evidence suggests that both pathways play a role in the pathogenesis of HCC and, in context of the FGFR family, in the resistance to VEGF driven angiogenesis [5]. Despite encouraging phase II data, enpoints were not met in two randomised phase III registration studies (Table 2). To avoid similar disappointment in the future, we should look carefully at these studies to identify potential flaws in trial design and assumptions. Given that these are two different populations of patients, the front-line and second-line groups should be considered separately. Brivanib was initially evaluated in two single-arm open label phase II studies of patients with advanced HCC. The front-line study enrolled 55 patients with advanced HCC (BCLC B/C 11%/89%) which were treated in the study with the primary
Table 1. Brivanib targeted tyrosine kinases and inhibition profile.
Tyrosine kinase VEGF-R1 VEGF-R2 VEGF-R3 Flk-1 FGF-R1 FGF-R2 Tie-2
IC50 (nM) of brivanib 390 34 10 23 148 125 445
objective being 6-month progression-free survival (PFS) and progression free survival rate, and the secondary objectives of tumour response rate, median progression free and overall survival, and safety tolerability [6]. Using the modified World Health Organization Criteria (mWHO) the 6-month progression free survival (PFS) rate was 18.2% (95% CI, 9.1–30.9%), median PFS was 2.7 months (95% CI, 1.4–3.0) time to progression (TTP) was 2.8 months (95% CI, 1.4–3.5), and median overall survival (OS) was 10 months (95% CI, 6.8–15.2 months). Given the control arm from SHARP (placebo), these data would suggest that brivanib is an active drug in HCC, though if compared to the experimental arm in SHARP, it would seem comparable to sorafenib in efficacy. However, the brivanib study enrolled 64% Asian patients. In that context, if these efficacy data were compared to the Asia-Pacific sorafenib phase III study [7], a median OS of 10 months compares very favourably to the median OS of 6.5 months seen in the sorafenib arm in that study. With regard to toxicity, the adverse effects with brivanib were generally manageable with fatigue, hypertension, diarrhoea, nausea, and anorexia being most common and notably, the hand-foot skin reaction common with sorafenib, is uncommon with brivanib. After sorafenib’s approval for front-line therapy, the brivanib study was amended and evaluated brivanib in patients with progression on one line of prior therapy [8]. Again, this was an open label single-arm study. 46 patients were enrolled, 94% of them having received prior sorafenib, and again the majority being BCLC C (96%) and Asian (67%). Median TTP per mWHO criteria was 1.77 months (95% CI, 1.38–4.01), median PFS 2.73 months (95% CI, 1.41–3.91) and median OS was 9.79 months (95% CI, 5.52–13.57). For the first time, in a newly defined population of HCC patients, for those that had prior sorafenib, we were seeing a median OS of almost 10 months; a number that would not have been expected from interpreting the SHARP or Asia-Pacific datasets when comparing the differences between median OS and median TTP. One could have postulated that the median OS after stopping sorafenib was approximately 5 months from the SHARP study and just under 4 months in the Asia-Pacific study, and so the OS of 9.79 with brivanib was provocative in a population of patients with a great unmet need. In addition, both phase II studies included an assessment of response using the modified RECIST criteria for HCC in an attempt to better capture the clinical activity of brivanib [9]. With these data in hand, three global phase III randomised studies were performed with brivanib in HCC; a front-line study (BRISK-FL), a post-sorafenib study (BRISK-PS), and a third in combination with chemoembolisation (BRISK-TA). The first randomised second-line phase III study in the post sorafenib era to present was the BRISK-PS study [10]. This was
Table 2. Comparison of brivanib’s efficacy in phase II and phase III studies.
DCR (%) TTP (mo) PFS (mo) mOS (mo)
Phase II Front-line* 51 2.8 2.7 10
Phase III BRISK-FL^ 66 4.2 n.r. 9.5
Phase II Second-line* 45.7 1.77 2.0 9.79
⁄
Modified WHO criteria per independent radiology review. Modified RECIST criteria. n.r., not reported. DCR, disease control rate; TTP, time to progression; PFS, progression free survival; mOS, median overall survival.
^
948
Journal of Hepatology 2014 vol. 61 j 947–950
Phase III BRISK-PS^ 61 4.2 n.r. 9.4
JOURNAL OF HEPATOLOGY a global phase III, double-blind placebo control study of 395 patients randomised 2:1 that had progressed on or after sorafenib or were intolerant to sorafenib. The primary endpoint for the study was OS with a minimum hazard ratio of 0.78 being required for the study to be considered positive. Patients were stratified based on reason for sorafenib discontinuation and other strata commonly used in front-line HCC studies such as ECOGperformance status and the presence of extrahepatic spread, and/or vascular invasion. While the study was generally well balanced for these strata, there was an imbalance favouring the control arm in the number of patients that had vascular invasion, which was masked by the fact that the majority of patients had extrahepatic spread. In addition baseline a-fetoprotein (AFP), while not included in the stratification but a recognised negative prognostic factor in HCC, was increased in patients receiving brivanib. This study missed its primary endpoint; while median OS was better with brivanib, 8.2 months with placebo vs. 9.4 months with brivanib (HR, 0.89; 95.8% CI, 0.69–1.15; p = 0.3307) this was not statistically significant. Importantly, median OS in the phase III study was similar to that seen in the phase II for this population (9.4 months vs. 9.70). Time to progression (TTP) was significantly increased with brivanib as well as other secondary endpoints such objective response rate and disease control rate, all assessed using mRECIST. The side effect profile of brivanib was similar as in the phase II studies and was accompanied by a high rate of dose interruptions and reductions and a discontinuation rate of 23% secondary to AEs. Similarly, the BRISK-FL study did not meet its primary endpoint of non-inferiority in improving OS compared to sorafenib [11]. This study enrolled over 1100 patients and randomised them 1:1 in a double-blind fashion to brivanib or sorafenib and stratified them similarly as the SHARP study; by ECOG-PS, extrahepatic spread, and/or vascular invasion, and study site. Again, median OS with brivanib was similar as in the phase II study, 9.5 months vs. 10 months in phase III and II, respectively. However, this number did not differ significantly from the sorafenib survival of 9.9 months (HR, 1.06; 95.8% CI, 0.93–1.22) and exceeded the upper-limit HR 95% CI of 1.08 to indicate non-inferiority. Unlike the BRISK-PS study, secondary endpoints were similar using mRECIST in the BRISK-FL study, though the ORR with sorafenib of 9% with mRECIST was higher than that reported in the SHARP study using RECIST (2%). Again, the AE profile of brivanib was similar to the phase II studies. While the frequency of dose adjustments and interruptions was similar between the two arms, there were more discontinuations with brivanib due to adverse events, 43% with brivanib as compared to 33% with sorafenib. This fact highlights another relevant fact for drug development in HCC: systemic toxicity is not as well tolerated as in other malignancies. The failure of sunitinib in this population also echoes this observation [12]. So what can we learn from the development of brivanib? One, control arms are important, not only for phase III studies but for phase II as well. It could be argued that a randomised phase II study would have identified the equivalence of brivanib to sorafenib in the front-line setting making the BRISK-FL study less appealing. In addition, for the BRISK-PS study, a randomised phase II study would have given a better sense of the actual survival of the placebo group and possibly identify other important factors effecting survival in the post-sorafenib setting. Second, and along those lines, strata used in the front-line setting may not be applicable in the second-line setting. Perhaps, separating
extrahepatic-spread from vascular invasion is an important consideration in the second-line setting? Only recently, two publications identified prognosticators of outcome after sorafenib progression [13,14], one of which was just meeting the inclusion for a second-line study (i.e., selection bias). Third, is developing a better assessment tool for benefit in clinical trials with molecular targeted therapeutics? From the SHARP study and these prospective trials, it would appear that neither response by RECIST nor mRECIST correlates with outcome in HCC and further refinement of imaging endpoints need to be developed. In addition, perhaps it is time to move beyond anti-angiogenic agents in HCC. To date, no agents in this class, like brivanib, have been able to improve on sorafenib in the front-line or improve survival in the second-line setting. This is not to say these agents might not have done better than placebo in the front-line but as it stands, they need to do better than sorafenib or at least do as well with a more favourable toxicity profile. After sorafenib, we need to demand both a strong clinical rationale and biologic rationale to move agents with a similar mechanism of action into this space. Finally, drug development in HCC has suffered from the clinical heterogeneity of the disease and its molecular heterogeneity. More work on predictive markers of response needs to be done in HCC. While an ‘‘all-comers’’ approach worked for sorafenib, if we want to move forward in HCC with molecular therapeutics, we need not only better clinical trial tools and design but we also need to better select patients most likely to benefit from a given agent.
Conflict of interest RSF has served as a consultant to Bristol Myers Squibb, Onyx, Bayer, and Novartis. References [1] Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69–90. [2] Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008;359:378–390. [3] Llovet JM, Hernandez-Gea V. Hepatocellular carcinoma: reasons for phase III failure and novel perspectives on trial design. Clin Cancer Res 2014;20:2072–2079. [4] Bhide RS, Cai ZW, Zhang YZ, Qian L, Wei D, Barbosa S, et al. Discovery and preclinical studies of (R)-1-(4-(4-fluoro-2-methyl-1H-indol-5-yloxy)-5methylpyrrolo[2,1-f][1,2,4]triazin-6-yloxy)propan-2-ol (BMS-540215), an in vivo active potent VEGFR-2 inhibitor. J Med Chem 2006;49:2143–2146. [5] Chou T, Finn RS. Brivanib: a review of development. Future Oncol 2012;8:1083–1090. [6] Park JW, Finn RS, Kim JS, Karwal M, Li RK, Ismail F, et al. Phase II, open-label study of brivanib as first-line therapy in patients with advanced hepatocellular carcinoma. Clin Cancer Res 2011;17:1973–1983. [7] Cheng AL, Kang YK, Chen Z, Tsao CJ, Qin S, Kim JS, et al. Efficacy and safety of sorafenib in patients in the Asia-Pacific region with advanced hepatocellular carcinoma: a phase III randomised, double-blind, placebo-controlled trial. Lancet Oncol 2009;10:25–34. [8] Finn RS, Kang YK, Mulcahy M, Polite BN, Lim HY, Walters I, et al. Phase II, open-label study of brivanib as second-line therapy in patients with advanced hepatocellular carcinoma. Clin Cancer Res 2012;18:2090–2098. [9] Lencioni R, Llovet JM. Modified RECIST (mRECIST) assessment for hepatocellular carcinoma. Semin Liver Dis 2010;30:52–60. [10] Llovet JM, Decaens T, Raoul JL, Boucher E, Kudo M, Chang C, et al. Brivanib in patients with advanced hepatocellular carcinoma who were intolerant to sorafenib or for whom sorafenib failed: results from the randomized phase III BRISK-PS study. J Clin Oncol 2013;31:3509–3516. [11] Johnson PJ, Qin S, Park JW, Poon RT, Raoul JL, Philip PA, et al. Brivanib vs. sorafenib as first-line therapy in patients with unresectable, advanced
Journal of Hepatology 2014 vol. 61 j 947–950
949
International Hepatology hepatocellular carcinoma: results from the randomized phase III BRISK-FL study. J Clin Oncol 2013;31:3517–3524. [12] Cheng AL, Kang YK, Lin DY, Park JW, Kudo M, Qin S, et al. Sunitinib versus sorafenib in advanced hepatocellular cancer: results of a randomized phase III trial. J Clin Oncol 2013;31:4067–4075. [13] Reig M, Rimola J, Torres F, Darnell A, Rodriguez-Lope C, Forner A, et al. Postprogression survival of patients with advanced hepatocellular
950
carcinoma: rationale for second-line trial design. Hepatology 2013;58:2023–2031. [14] Shao YY, Wu CH, Lu LC, Chan SY, Ma YY, Yen FC, et al. Prognosis of patients with advanced hepatocellular carcinoma who failed first-line systemic therapy. J Hepatol 2014;60:313–318.
Journal of Hepatology 2014 vol. 61 j 947–950
