Randomized phase II study of axitinib versus placebo plus best supportive care in second-line treatment of advanced hepatocellular carcinoma

Randomized phase II study of axitinib versus placebo plus best supportive care in second-line treatment of advanced hepatocellular carcinoma

Annals of Oncology original articles Annals of Oncology 26: 2457–2463, 2015 doi:10.1093/annonc/mdv388 Published online 18 September 2015 Randomized ...

211KB Sizes 2 Downloads 66 Views

Annals of Oncology

original articles Annals of Oncology 26: 2457–2463, 2015 doi:10.1093/annonc/mdv388 Published online 18 September 2015

Randomized phase II study of axitinib versus placebo plus best supportive care in second-line treatment of advanced hepatocellular carcinoma Y.-K. Kang1*, †, T. Yau2,†, J.-W. Park3, H. Y. Lim4, T.-Y. Lee5, S. Obi6, S. L. Chan7, SK. Qin8, R. D. Kim9, M. Casey10, C. Chen11, H. Bhattacharyya11, J. A. Williams12,‡, O. Valota13, D. Chakrabarti10 & M. Kudo14 1

Received 2 July 2015; revised 9 September 2015; accepted 10 September 2015

Background: The efficacy and safety of axitinib, a potent and selective vascular endothelial growth factor receptors 1–3 inhibitor, combined with best supportive care (BSC) was evaluated in a global, randomized, placebo-controlled phase II trial in patients with locally advanced or metastatic hepatocellular carcinoma (HCC). Patients and methods: Patients with HCC and Child–Pugh Class A who progressed on or were intolerant to one prior antiangiogenic therapy were stratified by tumour invasion ( presence/absence of extrahepatic spread and/or vascular invasion) and region (Asian/non-Asian) and randomized (2:1) to axitinib/BSC (starting dose 5 mg twice-daily) or placebo/ BSC. The primary end point was overall survival (OS). Results: The estimated hazard ratio for OS was 0.907 [95% confidence interval (CI) 0.646–1.274; one-sided stratified P = 0.287] for axitinib/BSC (n = 134) versus placebo/BSC (n = 68), with the median (95% CI) of 12.7 (10.2–14.9) versus 9.7 (5.9–11.8) months, respectively. Results of prespecified subgroup analyses in Asian versus non-Asian patients or presence versus absence of tumour invasion were consistent with the overall population. Improvements favouring axitinib/BSC (P < 0.01) were observed in secondary efficacy end point analyses [ progression-free survival (PFS), time to tumour progression (TTP), and clinical benefit rate (CBR)], and were retained among Asian patients in the prespecified subgroup analyses. Overall response rate did not differ significantly between treatments and patient-reported outcomes favoured placebo/BSC. Most common all-causality adverse events with axitinib/BSC were diarrhoea (54%), hypertension (54%), and decreased appetite (47%). Baseline serum analyses identified potential new prognostic (interleukin-6, E-selectin, interleukin-8, angiopoietin-2, migration inhibitory factor, and c-MET) or predictive (E-selectin and stromalderived factor-1) factors for survival. Conclusions: Axitinib/BSC did not improve OS over placebo/BSC in the overall population or in stratification subgroups. However, axitinib/BSC resulted in significantly longer PFS and TTP and higher CBR, with acceptable toxicity in patients with advanced HCC. Trial Registration: ClinicalTrials.gov, NCT01210495. Key words: antiangiogenic therapy, axitinib, best supportive care, hepatocellular carcinoma

*Correspondence to: Prof. Yoon-Koo Kang, Department of Oncology, Asan Medical Center, University of Ulsan, 88 Olympic-Ro 43-Gil, Songpa-Gu, Seoul 138-736, Republic of Korea. Tel: +82-2-3010-3230; E-mail: [email protected]

Equal contribution to the study. Employed by Pfizer at the time of the study.



© The Author 2015. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected].

Downloaded from http://annonc.oxfordjournals.org/ by guest on February 12, 2016

Department of Oncology, Asan Medical Center, University of Ulsan, Seoul, Republic of Korea; 2Department of Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong; 3National Cancer Center/Center for Liver Cancer, Goyang-si; 4Division of Hematology–Oncology, Samsung Medical Center, Sungkyunkwan University, Seoul, Republic of Korea; 5Division of Gastroenterology and Hepatology, Taichung Veterans General Hospital, Taichung, Taiwan; 6Department of Hepatology, Sasaki Foundation Kyoundo Hospital, Tokyo, Japan; 7State Key Laboratory in Oncology of South China, Department of Clinical Oncology, The Chinese University of Hong Kong, Hong Kong; 8Nanjing Bayi Hospital, Nanjing, China; 9H. Lee Moffitt Cancer Center, Tampa; 10Pfizer Inc, Collegeville; 11Pfizer Inc, New York; 12Pfizer Oncology, San Diego, USA; 13Pfizer srl, Milan, Italy; 14Department of Gastroenterology and Hepatology, Kinki University Hospital, Osaka, Japan

original articles introduction

methods study design This is a randomized, double-blind, global phase II study (ClinicalTrials.gov, NCT01210495). Patients were stratified by tumour invasion (presence versus absence of extrahepatic spread and/or vascular invasion) and geographic region (Asian versus non-Asian sites) and randomly assigned (2:1) to axitinib/BSC or placebo/BSC using a centralized interactive voice response system (IMPALA; supplementary Text, available at Annals of Oncolgy online). The primary end point was OS. Secondary end points included progression-free survival (PFS), time to tumour progression (TTP), objective response rate (ORR), clinical benefit rate (CBR), patient-reported outcomes (PROs), and safety. Prognostic or predictive potential of serum soluble proteins were also investigated. The study protocol was approved by institutional review boards or independent ethics committees and conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonization guidelines on Good Clinical Practice. Each patient provided informed consent.

patients and treatment Patients with histologically or cytologically confirmed locally advanced or metastatic HCC, who progressed on or were intolerant [discontinued treatment due to treatment-related grade 3/4 adverse event (AE) per the National Cancer Institute Common Terminology Criteria for Adverse Events (NCICTCAE) v3.0] to one prior antiangiogenic therapy, were eligible. Other key eligibility criteria included presence of ≥1 measurable lesion, Child–Pugh A

 | Kang et al.

liver function, and Eastern Cooperative Oncology Group performance status (ECOG PS) 0 or 1 (supplementary Text, available at Annals of Oncology online). Randomized patients received axitinib or placebo orally at a starting dose of 5 mg twice-daily (b.i.d.) in 4-week cycles. Doses could be modified per protocol-specified algorithm (supplementary Text, available at Annals of Oncology online). All patients additionally received BSC. The study team, investigators, and patients were blinded to treatment assignment.

assessments Baseline tumour assessments included CT or MRI scans, which were repeated every 8 weeks. Tumour response was assessed by investigators according to the Response Evaluation Criteria in Solid Tumours v1.1. Safety was monitored throughout the study and AEs graded per the NCI-CTCAE. Physical examinations and laboratory tests were carried out at baseline, Week 2 and/or 4, and every 4 weeks thereafter. Blood pressure (BP) was monitored at each clinic visit and by patients at home. PROs were assessed using the validated Functional Assessment of Cancer Therapy (FACT)– Hepatobiliary (Hep) questionnaire [11] including Hepatobiliary Symptom Index (FHSI-8) and EuroQol (EQ)-5D, and circulating serum levels of soluble proteins at baseline were measured (supplementary Text, available at Annals of Oncology online).

statistical analyses The study was designed to detect a hazard ratio (HR) of 0.60 for OS in favour of axitinib/BSC, assuming the median OS of 5.0 months with placebo/BSC and 8.3 months with axitinib/BSC. To detect a statistically significant difference between arms, 150 OS events were needed for a stratified one-sided log-rank test with a significance level of 0.025 and 80% power. Assuming a ∼13-month patient accrual period, 5% dropout rate per arm, and ∼11-month follow-up, 198 patients were required. One predefined interim analysis was carried out (supplementary Text, available at Annals of Oncology online). Time-to-event end points were analysed using the Kaplan–Meier method and median and two-sided 95% confidence intervals (CIs) were estimated. Comparison between treatment arms was made with stratified or unstratified one-sided log-rank test (α = 0.025), and HR and 95% CI were calculated. A Cox proportional hazards model was used to explore potential influences of baseline patient characteristics on OS. ORR ( proportion of patients with complete or partial response) or CBR ( proportion of patients with complete or partial response or ≥8 weeks stable disease) were compared using a stratified Cochran–Mantel–Haenszel test, and relative risk ratio and two-sided 95% CI were calculated. Statistical analysis methodology for PROs and biomarkers are provided in supplementary Text, available at Annals of Oncology online.

results patients and treatment Between December 2010 and July 2012, patients were assigned to axitinib/BSC (n = 134) and placebo/BSC (n = 68) (supplementary Figure S1, available at Annals of Oncology online). One patient assigned to the axitinib/BSC arm died due to disease progression and did not receive any study treatment and was excluded from safety analyses. The majority of patients were male (82%) and Asian (62%). Baseline patient characteristics were generally comparable between the axitinib/BSC and placebo/BSC arms (Table 1), although a ≥5% absolute difference was seen with age <65 years (60% versus 53%), hepatitis C (29% versus 16%),

Volume 26 | No. 12 | December 2015

Downloaded from http://annonc.oxfordjournals.org/ by guest on February 12, 2016

Liver cancer is the sixth most common cancer worldwide, with estimated 845 582 new cases and 806 873 deaths in 2015 [1]. Hepatocellular carcinoma (HCC), which accounts for the majority of liver cancers, is a highly vascularized tumour. Angiogenesis, via several pathways, including vascular endothelial growth factor (VEGF)/VEGF receptor (VEGFR) signalling, is thought to contribute to pathogenesis of HCC [2]. Multikinase inhibitor sorafenib, which blocks VEGFR kinases among others, has demonstrated improved overall survival (OS) in treatment-naïve patients with advanced HCC in randomized phase III studies [3, 4]. However, no other single-agent kinase inhibitor (e.g. sunitinib, linifanib, brivanib) [5–7] or combination (e.g. sorafenib/erlotinib) [8] has exceeded clinical benefits of sorafenib in first-line HCC, and sorafenib remains the only approved systemic therapy for advanced HCC. Yet, the efficacy of sorafenib may be short-lived due to toxicities, leading to disease progression following first-line sorafenib. There are currently no effective treatment options for patients who progress on or are intolerant to sorafenib. Axitinib is a potent and selective inhibitor of VEGFRs 1–3 [9], approved as second-line therapy for advanced renal cell carcinoma and has shown nonclinical activity in HCC animal models (Pfizer, data on file), providing evidence for its therapeutic potential. Furthermore, safety and pharmacokinetics of axitinib have been determined in patients with mild or moderate hepatic impairment [10]. This phase II trial evaluated the efficacy and safety of axitinib in combination with best supportive care (BSC) compared with placebo plus BSC in patients with advanced HCC previously treated with an antiangiogenic therapy. Additionally, serum proteins known to be involved in angiogenesis and tumour growth were explored as potential baseline biomarkers for axitinib efficacy.

Annals of Oncology

original articles

Annals of Oncology

Table 1. Demographics and baseline patient characteristics Axitinib + BSC (n = 134)

Placebo + BSC (n = 68)

a

Asian sites were China, Hong Kong, Japan, Korea, and Taiwan, whereas non-Asian sites were Belgium, France, Germany, the UK, Hungary, Italy, Slovakia, and the USA. b Unknown for one patient in the axitinib/BSC arm. c Include arterial chemotherapy, transarterial (chemo)embolization, percutaneous injection, radiofrequency ablation, microwave ablation, and other. BSC, best supportive care; ECOG PS, Eastern Cooperative Oncology Group performance status; BCLC, Barcelona Clinic Liver Cancer; HCC, hepatocellular carcinoma.

and liver cirrhosis (66% versus 60%). Additionally, more patients in the placebo/BSC than axitinib/BSC arm were intolerant to first-line antiangiogenic therapy (25% versus 10%). At data cut-off for the primary analysis (3 March 2014), 113 (84%) axitinib/BSC-treated and 54 (79%) placebo/BSC-treated

Volume 26 | No. 12 | December 2015

efficacy The difference in OS between the axitinib/BSC- and placebo/ BSC-treated patients did not reach statistical significance (HR 0.907, 95% CI 0.646–1.274; one-sided stratified P = 0.287). The median (95% CI) OS was 12.7 (10.2–14.9) months with axitinib/ BSC and 9.7 (5.9–11.8) months with placebo/BSC (Figure 1A). Prespecified subgroup analyses indicated no significant differences in OS between treatment arms regardless of geographic region or baseline tumour invasion (Figure 2). In an exploratory subgroup analysis, however, a positive trend for better OS with axitinib/BSC was observed in some subgroups, such as ECOG PS 1, excluding patients intolerant to prior antiangiogenic therapy, and baseline α-fetoprotein ≥400 ng/ml (Figure 2). Notably, when patients who were intolerant to prior therapy were excluded, the difference in OS showed an improvement for axitinib/BSC (n = 121) over placebo/BSC (n = 51) [HR 0.622; P = 0.013; median: 12.3 (95% CI 9.5–14.0) versus 9.2 (95% CI 4.3–10.0) months, respectively] (supplementary Figure S2A, available at Annals of Oncology online). In the Cox proportional hazards analysis, several baseline characteristics [ECOG PS, vascular invasion, prior therapy, Barcelona Clinic Liver Cancer staging, and trial outcome index (TOI) score (physical and functional well-being +HepCS-18)] were associated with OS; however, upon adjusting for these factors, treatment effect was consistent with the primary analysis (HR 0.811, P = 0.238). PFS was longer with axitinib/BSC than placebo/BSC [HR 0.618; P = 0.004; median: 3.6 (95% CI 2.3–4.6) versus 1.9 (95% CI 1.9–3.5) months; Figure 1B]. In prespecified subgroup analyses, PFS was significantly better with axitinib/BSC among patients at Asian, but not non-Asian, sites (supplementary Figure S2B/C, available at Annals of Oncology online). Similarly, a longer PFS with axitinib/BSC than placebo/BSC (P = 0.001) was observed among patients with, but not without, baseline tumour invasion. In the axitinib/BSC versus placebo/BSC arm, respectively, TTP was significantly longer (HR 0.621; P = 0.006; median 3.7 versus 1.9 months) and CBR was higher (31.3% versus 11.8%; P = 0.003), whereas the difference in ORR did not reach statistical significance (9.7% versus 2.9%; P = 0.091) (supplementary Table S1, available at Annals of Oncology online). The time to deterioration, based on the composite end point of death, tumour progression, or FHSI-8 mean score decrease ≥3 points, whichever occurred first, was 1.9 months in both arms (supplementary Text, available at Annals of Oncology online).

safety More than 90% of patients in each arm reported all-causality, allgrade AEs (Table 2). Common AEs with axitinib/BSC included

doi:10.1093/annonc/mdv388 | 

Downloaded from http://annonc.oxfordjournals.org/ by guest on February 12, 2016

Age, year Median (range) 61 (25–84) 63 (26–83) <65, n (%) 81 (60) 36 (53) Sex, n (%) Male 110 (82) 56 (82) Race, n (%) White 48 (36) 26 (38) Asian 84 (63) 42 (62) Black 2 (1) 0 ECOG PS, n (%) 0 78 (58) 39 (57) 1 56 (42) 29 (43) Geographic region, n (%) Asian sitesa 83 (62) 41 (60) Child–Pugh classification, n (%) A 134 (100) 68 (100) Tumour vascular invasion, n (%) Present 33 (25) 19 (28) Extrahepatic spread, n (%) Present 92 (69) 48 (71) Tumour vascular invasion and/or extrahepatic spread, n (%) Present 102 (76) 52 (76) BCLC stage,b n (%) A 5 (4) 3 (4) B 20 (15) 12 (18) C 108 (81) 53 (78) HCC aetiology, n (%) Hepatitis B 69 (51) 34 (50) Hepatitis C 39 (29) 11 (16) Liver cirrhosis, n (%) Present 89 (66) 41 (60) Time since histological diagnosis Median (range), months 22.9 (1.6, 125.6) 20.3 (2.4, 105.4) Prior therapy, n (%) Radiation therapy 33 (25) 12 (18) Surgery 76 (57) 39 (57) Local non-surgical therapyc 99 (74) 46 (68) Systemic therapy 134 (100) 68 (100) Sorafenib-containing regimen 124 (93%) 58 (85)

patients discontinued the study, primarily due to death. The main reasons for treatment discontinuation with axitinib/BSC (n = 127) and placebo/BSC (n = 67), respectively, were objective progression (52% versus 74%) and AEs (23% versus 12%). Patients in the axitinib/BSC arm received study treatment longer than those in the placebo/BSC arm (114.0 versus 57.5 days), but had more dose interruptions (80.5% versus 55.9%) or reductions (44.4% versus 5.9%) and fewer follow-up therapies (33.6% versus 54.4%). This study is ongoing with six patients in the axitinib/BSC arm still on treatment.

original articles A

Annals of Oncology

Survival distribution function

1.0

n Axitinib/BSC 134 Placebo/BSC 68

0.8

Events 101 52

Median OS, mo (95% CI) 12.7 (10.2, 14.9) 9.7 (5.9, 11.8)

HR = 0.907 (95% CI: 0.646, 1.274) P = 0.287 (one-sided stratified log-rank)

0.6

0.4

0.2

0.0 0

2

4

No. at risk: Axitinib/BSC 134 124 111 Placebo/BSC 68 62 48

8

10

12

14 16 18 20 Time (months)

22

24

26

28

30

32

34

93 41

81 38

74 29

68 25

51 23

27 13

17 9

8 7

4 6

2 3

1 1

0 0

48 18

43 17

33 15

Survival distribution function

1.0 n Axitinib/BSC 134 Placebo/BSC 68

0.8

Events Median PFS, mo (95% CI) 105 3.6 (2.3, 4.6) 54 1.9 (1.9, 3.5)

HR = 0.618 (95% CI: 0.438, 0.871) P = 0.004 (one-sided stratified log-rank)

0.6

0.4

0.2

0.0 0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

4 1

3 1

1 1

1 1

0 1

0 1

0 0

Time (months) No. at risk: Axitinib/BSC 134 72 Placebo/BSC 68 24

47 11

36 7

18 2

13 2

9 2

7 2

4 2

Figure 1. Kaplan–Meier estimates for (A) overall survival and (B) progression-free survival in all randomized patients. BSC, best supportive care; CI, confidence interval; HR, hazard ratio; OS, overall survival; PFS, progression-free survival.

diarrhoea, hypertension, and decreased appetite. A higher percentage of axitinib/BSC-treated than placebo/BSC-treated patients experienced grade ≥3 AEs, particularly hypertension, diarrhoea, and hand–foot skin reaction. Serious AEs were also higher with axitinib/BSC than placebo/BSC (47% versus 24%). Grade 3 laboratory abnormalities were generally more prevalent with axitinib/BSC than placebo/BSC, but the incidence of grade 4 laboratory abnormalities was low in both arms. Sixteen (12%) axitinib/BSC-treated patients died on study (within 28 days after the last dose of study medication) due to disease progression (n = 13) or AE ( pneumonia, liver failure, or septicaemia; n = 1 each). Eight (12%) placebo/BSC-treated patients died due to disease progression (n = 6) or AE (acute renal and liver failure or obstructive airway disorder; n = 1 each). AE-related treatment discontinuations (29% versus 13%), dose reductions (35% versus 7%), or temporary discontinuations (65% versus 25%) occurred more frequently with axitinib/BSC versus placebo/BSC. Axitinib/BSC treatment did not lead to increased incidence of haemorrhage (18% versus 16% with placebo/BSC) and arterial

 | Kang et al.

thromboembolic events, gastrointestinal perforation, and venous thromboembolic disorders were low (1% each).

serum soluble protein biomarkers Baseline specimens from 181 patients were included in the analysis (detailed results in supplementary text, available at Annals of Oncology online). Based on the unstratified log-rank test, low baseline serum level of E-selectin or stromal cell-derived factor (SDF)-1 were associated with significantly improved OS among patients treated with axitinib/BSC compared with placebo/BSC, suggesting their predictive potential. The Cox proportional hazards model identified lower than the median baseline levels of interleukin (IL)-6, E-selectin, IL-8, angiopoietin-2, migration inhibitory factor (MIF), or c-MET as potential positive prognostic factors for HCC (Punadjusted ≤ 0.009 or lower). In the axitinib/BSC arm, patients who achieved a clinical benefit had mean or median baseline serum levels of IL-6 and angiopoeitin-2 that were 50% and 35%, respectively, lower than

Volume 26 | No. 12 | December 2015

Downloaded from http://annonc.oxfordjournals.org/ by guest on February 12, 2016

B

6

original articles

Annals of Oncology

Favours axitinib

Favours placebo Sample Size (Event) Axitinib/BSC Placebo/BSC Hazard Ratio (95% CI) 134 (101)

68 (52)

0.881 (0.630, 1.233)

Baseline ECOG PS score 1

56 (46)

29 (26)

0.598 (0.368, 0.973)

Baseline ECOG PS score 0

78 (55)

39 (26)

1.084 (0.678, 1.732)

Age <65 y

81 (59)

36 (29)

0.723 (0.462, 1.131)

Age ≥65 y

53 (42)

32 (23)

1.094 (0.657, 1.821)

Treatment (axitinib vs placebo)

110 (85)

56 (43)

0.910 (0.630, 1.313)

24 (16)

12 (9)

0.834 (0.360, 1.930)

Asian sites

83 (64)

41 (31)

0.832 (0.539, 1.283)

Non-Asian sites

51 (37)

27 (21)

0.977 (0.572, 1.670)

Vascular invasion present

33 (26)

19 (15)

0.984 (0.520, 1.862)

Vascular invasion absent

101 (75)

49 (37)

0.869 (0.585, 1.290)

Excluding patients intolerant to prior therapy

121 (91)

51 (42)

0.662 (0.458, 0.956)

Baseline AFP <400 ng/mL

79 (58)

31 (20)

1.271 (0.763, 2.117)

Baseline AFP ≥400 ng/mL

52 (40)

37 (32)

0.671 (0.420, 1.071)

0.0

0.5 1.0 1.5 2.0 Hazard Ratio (95% CI)

2.5

Figure 2. Subgroup analysis of overall survival. The boxes indicate hazard ratios and whiskers represent the 95% confidence interval. AFP, α-fetoprotein; BSC, best supportive care; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status.

patients without achieving a clinical benefit (Punadjusted = 0.002 and 0.021, respectively).

discussion This study, which compared axitinib/BSC and placebo/BSC in second-line HCC, failed to meet the primary end point. Prespecified subgroup analyses did not reveal OS benefit of axitinib/BSC either, although an exploratory subgroup analysis suggested improved OS when patients intolerant to first-line antiangiogenic therapy were excluded. However, the study demonstrated improvements in PFS, TTP, and CBR with axitinib/BSC over placebo/BSC. Furthermore, the study showed the acceptable safety profile of axitinib/BSC in advanced HCC, consistent with the known safety profile of axitinib. PROs generally favoured placebo/BSC, which could be explained by more frequent treatment-related AEs experienced by axitinib/BSCtreated patients, and time to deterioration was the same in both arms. The absence of improvement in time to symptomatic progression, albeit OS benefit, has been reported for sorafenib in previous phase III trials [3, 4]. Because of a seemingly early separation in the OS curve, survival data from this study were further analysed, but the outcomes confirmed the primary results (supplementary Text, available at Annals of Oncology online). A failure to detect OS differences between treatment arms might indicate inadequate activity of axitinib in advanced HCC, unlike advanced renal cell carcinoma, but also could be explained by factors such as slight imbalance in baseline patient characteristics and/or post-study treatments (33.6% of axitinib/BSC-treated patients versus 54.4%

Volume 26 | No. 12 | December 2015

of placebo/BSC-treated patients received follow-up treatments). With regard to patient characteristics, other second-line treatment studies evaluating survival in patients with HCC following first-line sorafenib therapy reported that patients who discontinued sorafenib due to AEs or entered the second-line clinical trials had better prognosis, leading to a longer OS, than those who discontinued sorafenib due to tumour or liver disease progression or those who were ineligible, respectively, and that tumour progression pattern was associated with post first-line survival [12–14]. In our study, more patients (25%) in the placebo/BSC arm were intolerant to first-line therapy than in the axitinib/BSC arm (10%), which might explain improved OS with axitinib/BSC over placebo/BSC when intolerant patients were excluded from the analysis. Although patients were stratified by the presence or absence of extrahepatic spread and/or vascular invasion, it still remains possible that some imbalance existed between the two treatment arms with regard to the proportion of patients with different tumour progression pattern. It should also be noted that the median OS observed with the placebo/BSC arm (9.7 months) in this study was longer than the anticipated 5 months, as well as the 7.6 and 8.2 months reported for placebo/BSC in second-line HCC from phase III studies [15, 16]. Plausible reasons for longer than anticipated OS include frequent post-study treatments, selection of patients with more favourable prognosis (as explained above; also exclusion of patients with main portal vein invasion by HCC), and/or improved quality of BSC for patients with HCC in recent years. These findings point to the importance of selecting more appropriate patient eligibility criteria and stratification factor(s) in designing second-line clinical trials in advanced HCC.

doi:10.1093/annonc/mdv388 | 

Downloaded from http://annonc.oxfordjournals.org/ by guest on February 12, 2016

Male Female

original articles

Annals of Oncology

Table 2. Common treatment-emergent, all-causality adverse events Axitinib + BSC (n = 133a) All grades

Grade ≥ 3

Placebo + BSC (n = 68) All grades

Grade ≥ 3

b

131 (98) 72 (54) 72 (54) 62 (47) 46 (35) 45 (34) 45 (34) 36 (27) 35 (26) 33 (25) 33 (25) 27 (20) 27 (20) 26 (20) 23 (17) 21 (16) 21 (16) 14 (11) 11 (8)

109 (82) 27 (20) 34 (26) 16 (12) 10 (8) 9 (7) 20 (15) 5 (4) 6 (5) 0 0 12 (9) 7 (5) 4 (3) 2 (2) 0 1 (1) 2 (2) 3 (2)

63 (93) 8 (12) 9 (13) 14 (21) 18 (26) 14 (21) 4 (6) 2 (3) 7 (10) 0 0 3 (4) 1 (1) 7 (10) 2 (3) 8 (12) 3 (4) 10 (15) 11 (16)

26 (38) 0 1 (1) 4 (6) 7 (10) 1 (1) 0 0 0 0 0 0 0 0 1 (1) 1 (1) 1 (1) 0 2 (3)

125/133 (94) 114/133 (86) 101/133 (76) 99/133 (74) 92/133 (69) 84/132 (64) 75/125 (60) 76/130 (58) 73/128 (57) 67/133 (50) 63/133 (47) 44/105 (42) 49/126 (39) 51/133 (38) 39/127 (31) 37/131 (28)

20/133 (15) 39/133 (29) 11/133 (8) 13/133 (10) 5/133 (4) 1/132 (1) 7/125 (6) 19/130 (15) 1/128 (1) 16/133 (12) 16/133 (12) 2/105 (2) 3/126 (2) 1/133 (1) 6/127 (5) 12/131 (9)

54/67 (81) 55/67 (82) 48/67 (72) 34/67 (51) 40/67 (60) 28/67 (42) 29/63 (46) 36/66 (55) 42/64 (66) 27/67 (40) 21/67 (31) 21/48 (44) 17/63 (27) 14/67 (21) 11/62 (18) 20/66 (30)

8/67 (12) 23/67 (34) 5/67 (7) 3/67 (4) 3/67 (4) 1/67 (1) 0 6/66 (9) 4/64 (6) 2/67 (3) 5/67 (7) 0 0 0 0 5/66 (8)

a

One patient died before start of treatment. Reported by ≥15% of patients in either arm. c Reported by ≥30% of patients in either arm. d Denominator for each laboratory abnormality differed depending on the availability of baseline and at least one on-study test result. BSC, best supportive care; AST, aspartate aminotransferase; GGT, γ-glutamyl transferase; ALT, alanine aminotransferase. b

Patients in the axitinib/BSC arm achieved longer PFS (P = 0.004) and TTP (P = 0.006), and higher CBR (P = 0.003) compared with those in the placebo/BSC arm. Furthermore, these improvements were preferentially seen among patients from Asian, but not non-Asian, sites. It is unclear whether such regional disparities reflect differences in pathophysiology, genetic components, and/or in clinical practice standards in different regions. McNamara et al. [17] recently reported promising clinical activity of axitinib as second-line therapy for HCC in a single-arm, open-label phase II study. The study met its primary end point

 | Kang et al.

(>20% tumour control). Although the median PFS (3.6 months) was the same as in the current study, the median OS of 7.1 months was shorter than either treatment arm reported here, which might be due to more patients with advanced stages of disease enrolling in the McNamara study. The results of the current study are in line with outcomes reported in randomized phase III studies of other antiangiogenic agents for treatment of second-line HCC. The REACH trial reported lack of OS advantage (HR 0.866; P = 0.1391) for ramucirumab, a monoclonal antiVEGFR-2 antibody [15]. Ramucirumab/BSC, however, provided

Volume 26 | No. 12 | December 2015

Downloaded from http://annonc.oxfordjournals.org/ by guest on February 12, 2016

Adverse events , n (%) Any Diarrhoea Hypertension Decreased appetite Fatigue Abdominal pain Hand–foot skin reaction Weight decrease Nausea Dysphonia Hypothyroidism Asthenia Proteinuria Vomiting Rash Constipation Headache Oedema peripheral Back pain Laboratory abnormalitiesc,d, n/N (%) Elevated AST Elevated GGT Hyperglycaemia Elevated ALT Elevated alkaline phosphatase Hypoalbuminaemia Thrombocytopaenia Lymphopaenia Anaemia Hyperbilirubinaemia Hyponatraemia Prolonged prothrombin time Leukopaenia Hypocalcaemia Neutropaenia Hypophosphataemia

original articles

Annals of Oncology

acknowledgements Authors acknowledge earlier contributions from Evelyn Boucher, MD, formally at Department of d’Oncologie Medicale, Centre Eugene Marquis, Rennes Cedex, France, and Jie Tang, PhD, formally at Pfizer. Authors would like to thank Jane Q. Liang, ScD, of Pfizer and Yosuke Fujii, PhD, of Pfizer Japan for the statistical contribution, Yoshiko Umeyama, BPharm, of Pfizer Japan for the scientific contribution, and Liqiang Yang, PhD, of Pfizer for the critical review of the manuscript. Medical writing support was provided by Mariko Nagashima, PhD, of Engage Scientific Solutions (Southport, CT) and was funded by Pfizer.

funding This study was sponsored by Pfizer (no grant number).

disclosure Y-KK has been a consultant for Pfizer and Bayer and received research grant from Bayer and Sanofi; J-WP is a member of consultant/advisory board of Taiho Pharmaceutical, Bristol-Myers Squibb, and Roche, and received research funding from Bayer; SLC received research funding from Pfizer; MC, CC, HB, OV, and DC are employed by and own stock in Pfizer; JAW was employed by

Volume 26 | No. 12 | December 2015

Pfizer at the time of this study and owns stock in Pfizer; all remaining authors have declared no conflicts of interest.

references 1. Ferlay J, Soerjomataram I, Ervik M et al. GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon, France: International Agency for Research on Cancer, 2013; http://globocan.iarc.fr (17 April 2015, date last accessed). 2. Zhu AX, Duda DG, Sahani DV, Jain RK. HCC and angiogenesis: possible targets and future directions. Nat Rev Clin Oncol 2011; 8(5): 292–301. 3. Llovet JM, Ricci S, Mazzaferro V et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008; 359(4): 378–390. 4. Cheng AL, Kang YK, Chen Z 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(1): 25–34. 5. Cheng AL, Kang YK, Lin DY et al. Sunitinib versus sorafenib in advanced hepatocellular cancer: results of a randomized phase III trial. J Clin Oncol 2013; 31(32): 4067–4075. 6. Johnson PJ, Qin S, Park JW et al. Brivanib versus sorafenib as first-line therapy in patients with unresectable, advanced hepatocellular carcinoma: results from the randomized phase III BRISK-FL study. J Clin Oncol 2013; 31(28): 3517–3524. 7. Cainap C, Qin S, Huang WT et al. Linifanib versus Sorafenib in patients with advanced hepatocellular carcinoma: results of a randomized phase III trial. J Clin Oncol 2015; 33(2): 172–179. 8. Zhu AX, Rosmorduc O, Evans TR et al. SEARCH: a phase III, randomized, doubleblind, placebo-controlled trial of sorafenib plus erlotinib in patients with advanced hepatocellular carcinoma. J Clin Oncol 2015; 33(6): 559–566. 9. Hu-Lowe DD, Zou HY, Grazzini ML et al. Nonclinical antiangiogenesis and antitumor activities of axitinib (AG-013736), an oral, potent, and selective inhibitor of vascular endothelial growth factor receptor tyrosine kinases 1, 2, 3. Clin Cancer Res 2008; 14(22): 7272–7283. 10. Tortorici MA, Toh M, Rahavendran SV et al. Influence of mild and moderate hepatic impairment on axitinib pharmacokinetics. Invest New Drugs 2011; 29(6): 1370–1380. 11. Heffernan N, Cella D, Webster K et al. Measuring health-related quality of life in patients with hepatobiliary cancers: the functional assessment of cancer therapyhepatobiliary questionnaire. J Clin Oncol 2002; 20(9): 2229–2239. 12. Reig M, Rimola J, Torres F et al. Postprogression survival of patients with advanced hepatocellular carcinoma: rationale for second-line trial design. Hepatology 2013; 58(6): 2023–2031. 13. Shao YY, Wu CH, Lu LC et al. Prognosis of patients with advanced hepatocellular carcinoma who failed first-line systemic therapy. J Hepatol 2014; 60(2): 313–318. 14. Iavarone M, Cabibbo G, Biolato M et al. Predictors of survival of patients with advanced hepatocellular carcinoma who permanently discontinued sorafenib. Hepatology 2015; 62(3): 784–791. 15. Zhu AX, Park JO, Ryoo BY et al. Ramucirumab versus placebo as second-line treatment in patients with advanced hepatocellular carcinoma following first-line therapy with sorafenib (REACH): a randomised, double-blind, multicentre, phase 3 trial. Lancet Oncol 2015; 16(7): 859–870. 16. Llovet JM, Decaens T, Raoul JL 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(28): 3509–3516. 17. McNamara MG, Le LW, Horgan AM et al. A phase II trial of second-line axitinib following prior antiangiogenic therapy in advanced hepatocellular carcinoma. Cancer 2015; 121(10): 1620–1627. 18. Llovet JM, Pena CE, Lathia CD et al. Plasma biomarkers as predictors of outcome in patients with advanced hepatocellular carcinoma. Clin Cancer Res 2012; 18(8): 2290–2300.

doi:10.1093/annonc/mdv388 | 

Downloaded from http://annonc.oxfordjournals.org/ by guest on February 12, 2016

significantly improved PFS and ORR over placebo/BSC. Furthermore, the study revealed potential predictive value of baseline αfetoprotein, which was also observed here. In the BRISK-PS study, addition of brivanib, a dual inhibitor of VEGFR and fibroblast growth factor receptor, to BSC produced similarly disappointing OS results, although TTP and ORR were better in patients with sorafenib-refractory/intolerant HCC [16]. HCC is a complex disease of multifactorial causes, including viral infections, alcohol-related liver cirrhosis, and genetic alterations, and often occurs in the setting of chronic liver disease. Studies investigating biomarkers for tumour progression would facilitate understanding of molecular mechanisms involved in HCC pathogenesis. This study confirmed the previously reported prognostic association between lower baseline levels of circulating IL-6 or angiopoeitin-2 and longer OS in advanced HCC [18], and additionally identified proteins involved in immune cell mechanisms (E-selectin, IL-8, and MIF) and cell survival pathways (c-MET) as potential prognostic markers. The study also suggested possible predictive value of E-selectin and SDF-1. The association between low baseline serum SDF-1 and longer OS in axitinib/BSC-treated patients is especially compelling for metastatic potential via SDF-1/CXCR4 interaction. Low activation levels of SDF-1/CXCR4 pathway favouring clinical benefit in axitinib/BSC-treated patients may point towards strategies for patient selection. In conclusion, axitinib/BSC did not improve OS over placebo/ BSC, but resulted in longer PFS and TTP, and higher CBR, with an acceptable safety profile in patients with advanced HCC previously treated with antiangiogenic therapy. The potential prognostic and predictive biomarkers identified in this study may warrant further investigation.