Note that only those deaths that occurred during study treatment (ie, were the direct cause of study treatment discontinuation) are shown.
Kaplan-Meier estimates of overall survival (A) and time to progression (B). The curves were truncated when the percentage of patients remaining at risk was approximately 10%. Vertical ticks on the curves indicate censoring events.
Hazard ratios (HRs) and 95% CIs were calculated using a stratified Cox proportional hazards model for all patients and an unstratified Cox proportional hazards model for subgroup analyses. BCLC indicates Barcelona Clinical Liver Cancer; ECOG, Eastern Cooperative Oncology Group; HCC, hepatocellular carcinoma.
eTable 1. Guidelines for Treatment and Monitoring Based on Results of Hepatitis B Screening at Baseline
eTable 2. Time to Definitive Deterioration of EORTC QLQ-C30 Scores in the Full Analysis Set
eTable 3. Clinically Notable Adverse Events of Any Cause in the Safety Population
Customize your JAMA Network experience by selecting one or more topics from the list below.
Zhu AX, Kudo M, Assenat E, et al. Effect of Everolimus on Survival in Advanced Hepatocellular Carcinoma After Failure of Sorafenib: The EVOLVE-1 Randomized Clinical Trial. JAMA. 2014;312(1):57–67. doi:10.1001/jama.2014.7189
Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Aside from the multikinase inhibitor sorafenib, there are no effective systemic therapies for the treatment of advanced hepatocellular carcinoma.
To assess the efficacy of everolimus in patients with advanced hepatocellular carcinoma for whom sorafenib treatment failed.
Design, Setting, and Participants
EVOLVE-1 was a randomized, double-blind, phase 3 study conducted among 546 adults with Barcelona Clinic Liver Cancer stage B or C hepatocellular carcinoma and Child-Pugh A liver function whose disease progressed during or after sorafenib or who were intolerant of sorafenib. Patients were enrolled from 17 countries between May 2010 and March 2012. Randomization was stratified by region (Asia vs rest of world) and macrovascular invasion (present vs absent).
Everolimus, 7.5 mg/d, or matching placebo, both given in combination with best supportive care and continued until disease progression or intolerable toxicity. Per the 2:1 randomization scheme, 362 patients were randomized to the everolimus group and 184 patients to the placebo group.
Main Outcomes and Measures
The primary end point was overall survival. Secondary end points included time to progression and the disease control rate (the percentage of patients with a best overall response of complete or partial response or stable disease).
No significant difference in overall survival was seen between treatment groups, with 303 deaths (83.7%) in the everolimus group and 151 deaths (82.1%) in the placebo group (hazard ratio [HR], 1.05; 95% CI, 0.86-1.27; P = .68; median overall survival, 7.6 months with everolimus, 7.3 months with placebo). Median time to progression with everolimus and placebo was 3.0 months and 2.6 months, respectively (HR, 0.93; 95% CI, 0.75-1.15), and disease control rate was 56.1% and 45.1%, respectively (P = .01). The most common grade 3/4 adverse events for everolimus vs placebo were anemia (7.8% vs 3.3%, respectively), asthenia (7.8% vs 5.5%, respectively), and decreased appetite (6.1% vs 0.5%, respectively). No patients experienced hepatitis C viral flare. Based on central laboratory results, hepatitis B viral reactivation was experienced by 39 patients (29 everolimus, 10 placebo); all cases were asymptomatic, but 3 everolimus recipients discontinued therapy.
Conclusions and Relevance
Everolimus did not improve overall survival in patients with advanced hepatocellular carcinoma whose disease progressed during or after receiving sorafenib or who were intolerant of sorafenib.
clinicaltrials.gov Identifier: NCT01035229
Patients with advanced hepatocellular carcinoma (HCC) have a median overall survival of less than 1 year, largely because of the absence of effective therapies.1 The multikinase inhibitor sorafenib is the only systemic therapy shown to significantly improve overall survival in advanced HCC.2,3 However, sorafenib benefits are mostly transient and modest, and disease eventually progresses. As many as 28.9% of patients with Child-Pugh A liver cirrhosis discontinue sorafenib because of adverse events (AEs),4 and there is no approved alternative in this setting. Effective therapies are needed for patients who experience progression during or after receiving sorafenib or who have sorafenib intolerance.
The phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin (mTOR) pathway, a key regulator of cellular growth, proliferation, angiogenesis, and survival,5 is a novel therapeutic target for HCC. The mTOR pathway is implicated in hepatocarcinogenesis,6,7 with activation occurring in up to 45% of HCC6 and associated with less differentiated tumors, early recurrence, and poor prognosis.7,8 In preclinical models, mTOR inhibition with everolimus and other rapamycin analogues prevented tumor progression and improved survival, and mTOR activation was associated with HCC recurrence.7,9,10 In transgenic mice, mTOR activation caused by liver-specific Tsc1 knockout was sufficient for HCC development,11 and in Pten-deficient mice with constitutive mTOR activation, steatohepatitis and HCC development were induced.12 In early-phase clinical studies, everolimus demonstrated manageable safety and clinical activity.13,14 In a phase 1 study conducted in Taiwan, the maximum tolerated dose was 7.5 mg/d with a 71% disease control rate across daily doses.13 In a phase 1/2 study conducted in the United States, the disease control rate was 44% at 10 mg/d, the maximum dose tested.14
The first Everolimus for Liver Cancer Evaluation (EVOLVE-1) study compared the efficacy and safety of everolimus, 7.5 mg/d, with matching placebo, both given with best supportive care, in patients with advanced HCC who were refractory to or intolerant of sorafenib.
EVOLVE-1 was an international, double-blind, placebo-controlled, phase 3 study. Adults aged 18 years and older with Barcelona Clinic Liver Cancer stage B or C advanced HCC15 and radiologically confirmed disease progression during or after sorafenib therapy or sorafenib intolerance were eligible. Prior local and hormonal therapy and 0 or 1 systemic chemotherapy regimen before sorafenib were allowed. Additional inclusion criteria included current cirrhotic status of Child-Pugh class A (5-6 points) without encephalopathy (calculated based on clinical findings and laboratory results during the screening period), Eastern Cooperative Oncology Group (ECOG) performance status 2 or lower, adequate organ function, and alcohol intake less than 80 g/d. Previous organ transplantation requiring immunosuppression, long-term immunosuppressive therapy, known history of HIV seropositivity, clinically significant third-space fluid accumulation, acute or chronic infectious disorders except for chronic hepatitis B (HBV) and hepatitis C (HCV) virus infection, or other severe or uncontrolled medical condition resulted in exclusion from the study.
The protocol was approved by the appropriate ethics body at each participating center. The study was performed per the protocol, Good Clinical Practice guidelines, local regulations, and the ethical principles of the Declaration of Helsinki. All patients provided written informed consent.
Patients were assigned unique identification numbers at enrollment and registered using an interactive voice- or web-response system that randomly allocated patients 2:1 to receive everolimus or placebo. Randomization was stratified by region (Asia [China, South Korea, Taiwan, and Thailand] vs rest of the world [Australia, Austria, Belgium, Canada, France, Germany, Greece, Hungary, Israel, Italy, Japan, Spain, and the United States]) and macrovascular invasion (yes vs no). Based on possible differences in treatment patterns and disease etiology, Australia and Japan were grouped with rest of the world. A validated system randomly assigned patient numbers to randomization numbers linked to the treatment groups and medication numbers. The medication randomization list was produced by Novartis Drug Supply Management using a validated system that automated the random assignment of medication numbers to study drug packs. If necessary, emergency treatment unblinding was performed using the interactive response system.
Patients received everolimus, 7.5 mg, once daily or matching placebo, both with best supportive care (anything believed to be in the patient’s best interest, excluding other antineoplastic treatments, delivered per normal local practice). Treatment continued until progression, intolerable toxicity, or another reason. For patients unable to tolerate the protocol-specified dosing scheme, dose adjustments and interruptions per a protocol-specified algorithm were permitted. The chosen everolimus dose was based on the results of the phase 1 study of everolimus for HCC conducted in Taiwan.13 Results of a pharmacokinetic study suggest everolimus, 7.5 mg, in patients with Child-Pugh class A liver function provides similar exposure to 10 mg given to patients with normal liver function.16
Per US Food and Drug Administration guidance on the collection of race and ethnicity data in clinical trials17 and their potential prognostic value, race/ethnicity information was recorded on the baseline clinical report form. The clinical report form included categories of Caucasian, Black, Asian, Native American, Pacific Islander, and other for race and Hispanic/Latino, Chinese, Indian (Indian subcontinent), Japanese, mixed ethnicity (specify), and other. Study sites were not given specific instruction regarding the collection of this information. Therefore, it is possible that for some patients, race/ethnicity information was determined by the site without input from the patient.
During screening, all patients were evaluated for HBV and HCV infection (eMethods in the Supplement). To reduce viral reactivation risk, at-risk patients received prophylactic antiviral therapy (eTable 1 in the Supplement). Definitions of HBV reactivation and HCV flare are provided in the eMethods in the Supplement.
Primary end point was overall survival, defined as time from randomization to date of death of any cause. The key secondary end point was time to progression, defined as time from randomization to date of radiologically confirmed disease progression. Other secondary end points included disease control rate, defined as percentage of patients who achieved complete or partial response or stable disease per RECIST (Response Evaluation Criteria in Solid Tumors) version 1.0, time to definitive 5% deterioration in the global quality-of-life and physical functioning scales of the European Organisation for the Research and Treatment of Cancer 30-item Quality of Life Questionnaire (EORTC QLQ-C30), and safety.
Tumors were measured by computed tomography or magnetic resonance imaging with contrast of the chest, abdomen, and pelvis at baseline, every 6 weeks until disease progression, and at study end or early discontinuation. Complete or partial responses required confirmation at least 4 weeks after initial response observation. After disease progression or start of new anticancer therapy, patients were followed up for survival every 4 weeks until the number of deaths required for final analysis was observed. Safety assessments included incidences of AEs and serious AEs and changes in vital signs and laboratory results. Adverse events were assessed according to the Common Terminology Criteria for Adverse Events version 3.0 and coded using the Medical Dictionary for Regulatory Activities.
Sample size estimation was based on the ability to detect a 28.6% reduction in risk of death (hazard ratio [HR] for overall survival, 0.714), corresponding to a 40% prolongation in median overall survival from 5 months with placebo to 7 months with everolimus. In the absence of definitive data, the anticipated overall survival for placebo was estimated based on the difference between time to progression and overall survival observed in phase 3 studies of sorafenib for advanced HCC.2,3 An overall survival of 7 months for everolimus was considered reasonable based on preliminary survival observed in early-phase studies of everolimus for HCC (final data published as Shiah et al13 and Zhu et al14). Given the 2:1 randomization scheme and 2 interim analyses planned after occurrence of the first 135 and 270 deaths, it was estimated that 449 deaths would provide 90% power to detect a clinically meaningful overall survival improvement with the use of a group sequential log-rank test with a 2.5% cumulative 1-sided significance level. Assuming uniform patient accrual, 531 patients were needed to observe the required number of deaths. When the Lan-DeMets procedure18 with an O’Brien-Fleming stopping boundary was used to account for the 2 interim analyses in light of the 454 overall survival events at final analysis, the significance boundary for final analysis was P ≤ .023573.
All randomly assigned patients who received 1 or more doses of study drug and had 1 or more postbaseline assessments were evaluated for safety. All enrolled patients were assessed for efficacy. Overall survival, time to progression, and deterioration in quality of life were analyzed by Kaplan-Meier methods. Study groups were compared using a stratified log-rank test (1-sided 2.5% significance level). Hazard ratios and 95% confidence intervals were estimated using a stratified Cox proportional hazards model. For analysis of overall survival, patients without a known date of death at the time of analysis were censored at the date of last contact. For analysis of time to progression, patients without radiologically confirmed disease progression at the time of analysis or receipt of further antineoplastic therapy were censored at the time of the last adequate tumor assessment before the analysis cutoff date or start date of further antineoplastic therapy. For analysis of time to definitive deterioration in quality of life, patients receiving further antineoplastic therapy before definitive worsening were censored at the date of last assessment before start of therapy; patients without worsening were censored at the date of last assessment before cutoff. Effects of potential prognostic factors on overall survival were investigated using a stratified Cox regression model including ECOG performance status, extrahepatic spread, age, sorafenib status, and baseline alpha-fetoprotein level. Time to progression was compared between treatment groups only if the overall survival difference was statistically significant. Disease control rate was compared between groups using the Cochran-Mantel-Haenszel test, with adjustment for stratification factors. All statistical analyses were performed using SAS version 9.2 (SAS Institute).
Between May 27, 2010, and March 21, 2012, 546 patients from 111 centers in 17 countries with advanced HCC whose disease progressed during or after sorafenib or who were sorafenib intolerant were randomly assigned to receive treatment with everolimus (n = 362) or placebo (n = 184) and were included in the full analysis population (Figure 1). One patient in the everolimus group and 2 in the placebo group did not receive study treatment and were excluded from the safety population.
At the time of analysis (June 14, 2013), 303 deaths (83.7%) in the everolimus group and 151 (82.1%) in the placebo group had occurred, and treatment was ongoing for 0.8% of everolimus and 3.3% of placebo recipients (Figure 1). The most common reasons for treatment discontinuation in the everolimus and placebo groups were disease progression (72.1% and 81.0%, respectively), AEs (16.9% and 7.6%, respectively), and consent withdrawal (5.5% and 3.8%, respectively).
Baseline demographics and disease characteristics were well balanced between treatment groups (Table 1). In the total population, most patients were enrolled from the rest of the world (83.3%); did not have macrovascular invasion (67.2%); and had ECOG performance status 0 to 1 (95.4%), Barcelona Clinic Liver Cancer stage C disease (86.3%), and extrahepatic disease (74.0%). The most common disease etiologies were HBV (26.2%), HCV (25.1%), and alcohol abuse (20.0%). Before sorafenib, 15.9% of patients received another antineoplastic medication, most commonly chemotherapy (10.3%). Previous sorafenib was discontinued for progression in 80.8% and for intolerance in 19.0%; 1 patient in the everolimus group discontinued sorafenib for planned surgery for inguinal hernia.
Median follow-up duration from randomization to the cutoff date was 24.6 months (range, 14.8-36.6). Median duration of exposure to everolimus and placebo was 9.43 weeks (range, 0.1-120.0) and 8.93 weeks (range, 0.4-136.3), respectively. Median relative dose intensity was 100% (range, 27%-100%) and 100% (range, 58%-105%), respectively. After study treatment discontinuation, 38.1% of everolimus and 41.3% of placebo recipients received further antineoplastic therapy, most commonly chemotherapy (19.6% and 18.5%, respectively).
No difference in the risk of death was noted between everolimus and placebo (HR, 1.05; 95% CI, 0.86-1.27; P = .68) (Figure 2A). Median overall survival was 7.6 months (95% CI, 6.7-8.7) and 7.3 months (95% CI, 6.3-8.7), respectively. Overall, 91.4% of deaths in the everolimus group and 94.7% of deaths in the placebo group were the result of disease progression. Estimated overall survival rates in the everolimus and placebo groups were 67.0% and 65.6%, respectively, at 5 months, and 53.4% and 51.4%, respectively, at 7 months. Adjustment for prognostic factors did not affect the risk of death between treatment groups (HR, 1.11; 95% CI, 0.91-1.36). Prespecified subgroup analyses revealed similar results in most subgroups (Figure 3). The exception was HCC etiology; patients with HBV appeared to have prolonged overall survival with everolimus (HR, 0.64; 95% CI, 0.45-0.93), whereas patients with an etiology other than HBV or HCV appeared to have a shorter overall survival with everolimus (HR, 1.35; 95% CI, 1.01-1.80).
There was no apparent difference between treatment groups in time to progression (HR, 0.93; 95% CI, 0.75-1.15) (Figure 2B). Per the prespecified analysis plan, the statistical significance of this difference was not formally tested. At the time of analysis, 252 progression events (69.6%) in the everolimus group and 133 (72.3%) in the placebo group had occurred. Median time to progression was 3.0 months (95% CI, 2.8-4.0) and 2.6 months (95% CI, 1.5-2.8), respectively.
The disease control rate was 56.1% (95% CI, 50.8%-61.3%) and 45.1% (95% CI, 37.8%-52.6%) in the everolimus and placebo groups, respectively (P = .01) (Table 2). There were no complete responses. A greater proportion of placebo recipients experienced progressive disease. No significant difference in time to definitive deterioration in the EORTC QLQ-C30 global quality-of-life scale was observed (P = .47); time to definitive deterioration in physical functioning was significantly shorter with everolimus (P = .01) (eTable 2 in the Supplement).
Adverse events of any grade, regardless of relationship with study drug, were experienced by 99.2% of everolimus and 94.0% of placebo recipients (Table 3). Grade 3/4 AEs and serious AEs were more frequent with everolimus (70.9% vs 52.2% and 47.4% vs 35.2%, respectively). Adverse events led to treatment discontinuation in 16.6% of everolimus and 7.7% of placebo recipients and dose interruptions or reductions in 55.7% and 29.1%, respectively. Grade 3/4 AEs with incidence 5% or greater in the everolimus group (everolimus vs placebo) were asthenia (7.8% vs 5.5%, respectively), anemia (7.8% vs 3.3%, respectively), decreased appetite (6.1% vs 0.5%, respectively), HBV (as reported on AE forms) (6.1% vs 4.4%, respectively), ascites (5.6% vs 8.7%, respectively), and thrombocytopenia (5.6% vs 0.5%, respectively) (Table 3).
On baseline screening, 73 patients (20.2%) in the everolimus and 44 (24.2%) in the placebo group were positive for HBV-DNA or surface antigen (HBsAg). Despite antiviral prophylaxis, HBV reactivation based on central laboratory results was experienced by 27 (37.0%) everolimus and 10 (22.7%) placebo recipients. An additional 69 (19.1%) and 29 (15.9%), respectively, were negative for HBV-DNA or HBsAg but positive for hepatitis B surface or core antibody (HBsAb or HBcAb) and, per protocol, did not receive antiviral prophylaxis. Two of these patients, both in the everolimus group, experienced HBV reactivation. All reactivation cases were asymptomatic, although 3 everolimus recipients discontinued therapy because of reactivation. None of the 79 patients in the everolimus group or 39 in the placebo group with detectable HCV-RNA or known history of HCV infection without detectable HCV-RNA experienced HCV flare. Incidence of other AEs of clinical interest for everolimus is shown in eTable 3 in the Supplement.
Seventy-five patients (13.8%) died during treatment or within the 28 days after discontinuation, with 47 deaths (13.0%) in the everolimus and 28 (15.4%) in the placebo group. Progressive disease was the primary cause of death in 36 (10.0%) everolimus and 24 (13.2%) placebo recipients. Adverse events led to death in 11 (3.0%) everolimus recipients (2 cases each of death of unknown cause and renal failure and 1 case each of cerebrovascular accident, gastrointestinal hemorrhage, interstitial lung disease, peritonitis, pneumonitis, respiratory failure, and upper gastrointestinal hemorrhage) and 4 (2.2%) placebo recipients (1 case each of cardiac arrest, gastrointestinal hemorrhage, hepatic cirrhosis, and multiorgan failure).
No systemic therapies have been established for patients with advanced HCC for whom sorafenib fails or who cannot tolerate sorafenib, highlighting the unmet need in this setting. Since sorafenib was approved for HCC, interest in developing molecularly targeted agents in this disease has been renewed. Although most phase 3 efforts have focused on vascular endothelial growth factor–targeted agents, everolimus inhibits mTOR, another critical target implicated in hepatocarcinogenesis.
Despite the strong scientific rationale6-8,11,12,19-22 and preclinical data,7,9,10,23 everolimus plus best supportive care failed to improve survival over placebo plus best supportive care in patients with advanced HCC that progressed during or after receiving sorafenib or who were sorafenib intolerant in the EVOLVE-1 study. The known prognostic factors, including ECOG performance status and the presence of macrovascular invasion or extrahepatic disease, were well balanced between treatment groups, and adjustment for these factors did not improve the risk of death between groups. A comprehensive, prespecified subgroup analysis revealed similar outcomes in most groups. Patients with HBV appeared to have prolonged overall survival with everolimus. Whether this finding truly represents a benefit in this subgroup or rather reflects an imbalance of prognostic factors between populations requires further investigation. In EVOLVE-1, improvements in secondary outcomes, including time to progression and overall response rate, were not observed. Although the disease control rate favored everolimus, this finding should be considered exploratory given the primary end point was not met. Furthermore, any benefit could be counterbalanced by the higher AE rate observed with everolimus. Despite the negative results from the EVOLVE-1 trial, given the immunosuppressive and antitumor effects of mTOR inhibitors, the potential benefits of this class of agents in the adjuvant setting are being assessed in a phase 3 trial of sirolimus for patients with HCC after liver transplantation (NCT00355862).
The AE profile observed for everolimus in this study was mostly consistent with the known safety profile of everolimus in other cancers. Hepatitis B virus reactivation, a known complication of immunosuppressive therapy, has been reported in other everolimus trials13,24 and in published case reports.25-27 In the phase 1 trial of everolimus for advanced HCC, routine antiviral prophylaxis was not given, and HBV reactivation with hepatitis flare was reported in 4 of 27 patients (15%) who were HBsAg-positive.13 In all 4 patients, reactivation resolved within 4 to 8 weeks of everolimus withdrawal and initiation of antiviral therapy. Despite preemptive antiviral therapy in EVOLVE-1, HBV reactivation based on central laboratory findings occurred in 37.0% of everolimus and 22.7% of placebo recipients who were HBV-DNA or HBsAg-positive (or both) at baseline. An additional 2.9% of everolimus recipients who were HBsAg-negative but HBsAb- or HBcAb-positive experienced HBV reactivation. Of note, all cases were asymptomatic. Our study represents the largest prospective experience with HBV reactivation in patients receiving everolimus with antiviral prophylaxis and, to the best of our knowledge, is the first report describing events in patients with past HBV infection receiving everolimus. These results suggest antiviral prophylaxis and close monitoring may be important for everolimus recipients with underlying HBV infection. The mechanism and clinical significance of subclinical HBV reactivation in HBsAg-positive placebo recipients require elucidation.
The results from EVOLVE-1 extend the list of failed phase 3 studies in advanced HCC, highlighting the challenge of developing effective therapies for this cancer. In the first-line setting, 3 multikinase inhibitors—sunitinib, brivanib, and linifanib—have been compared with sorafenib, and all failed to achieve their primary end point.28-30 Another first-line trial failed to demonstrate a survival benefit for erlotinib plus sorafenib over sorafenib alone.31 In the phase 3 BRISK-PS study of patients with advanced HCC who progressed during or after sorafenib or who were sorafenib intolerant, brivanib failed to improve overall survival compared with placebo, despite improvements in time to progression and objective response rate.32 Similar to BRISK-PS, median overall survival in the EVOLVE-1 control group was longer than expected, likely reflecting patient selection and improved supportive management of this population. The shorter median overall survival in the placebo group of EVOLVE-1 compared with that of BRISK-PS (7.3 vs 8.2 months) could reflect enrollment of a higher percentage of patients with poor prognostic factors in EVOLVE-1 (eg, macrovascular invasion, 32.8% vs 18%). The natural history of patients for whom sorafenib treatment failed remains to be explored.
EVOLVE-1 and the other failed phase 3 studies have provided several important lessons. First, it is difficult to assess efficacy signals from phase 2 trials. For everolimus, the initial modest efficacy signal was obtained from 2 phase 1/2 studies.13,14 As has been recommended,15 assessing efficacy in a randomized phase 2 study should be considered before proceeding to phase 3 testing. Second, surrogate end points such as time to progression, progression-free survival, and response rate inconsistently predict overall survival in phase 3 trials. Third, clinical and biologic heterogeneity likely affects the performance of targeted therapies in HCC. In the absence of well-characterized and validated predictive biomarkers, targeted agents will likely continue to have a high risk of failure if phase 3 trials are conducted in unselected populations. Relevant biomarkers that may predict clinical outcome in patients receiving everolimus are being assessed in the EVOLVE-1 population. Future studies of targeted agents for HCC should aim to enrich patient populations based on molecular classification and predictive biomarkers. A limitation of EVOLVE-1 was that it was not designed to identify a molecularly or clinically selected population that may have benefited from everolimus.
Everolimus did not improve overall survival in patients with advanced hepatocellular carcinoma whose disease progressed during or after taking sorafenib or who were intolerant of sorafenib.
Corresponding Author: Andrew X. Zhu, MD, PhD, Massachusetts General Hospital Cancer Center, Harvard Medical School, PO Box 232, 55 Fruit St, Boston, MA 02114 (firstname.lastname@example.org).
Author Contributions: Dr Zhu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Zhu, Poon, C.-L. Chen, Furuse, Anak, L.-T. Chen.
Acquisition, analysis, or interpretation of data: Zhu, Kudo, Assenat, Cattan, Kang, Lim, Poon, Blanc, Vogel, Dorval, Peck-Radosavljevic, Santoro, Daniele, Furuse, Jappe, Perraud, Anak, Sellami, L.-T. Chen.
Drafting of the manuscript: Zhu, Lim, C.-L. Chen, Jappe, Perraud, Anak, Sellami.
Critical revision of the manuscript for important intellectual content: Zhu, Kudo, Assenat, Cattan, Kang, Poon, Blanc, Vogel, Dorval, Peck-Radosavljevic, Santoro, Daniele, Furuse, Perraud, Anak, Sellami, L.-T. Chen.
Statistical analysis: Jappe, Perraud, Anak.
Administrative, technical, or material support: Zhu, Kang, Vogel, C.-L. Chen, Santoro, Furuse, Jappe.
Study supervision: Zhu, Kudo, Lim, Peck-Radosavljevic, Anak, Sellami, L.-T. Chen.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Zhu reports consultancy for sanofi-aventis, Exelixis, Eisai, and Daiichi Sankyo. Dr Kudo reports having received a grant, consulting fee or honorarium, and support for travel to meetings for the study or other purposes and consultancy for Novartis. Dr Cattan reports receiving support for travel to meetings for the study or other purposes from Novartis and consultancy for Novartis. Dr Kang reports board membership, consultancy, and grants or grants pending for Novartis. Dr Vogel reports consultancy, payment for lectures, and payment for manuscript preparation from Bayer, Celgene, and Roche; consultancy and payment for lectures from Amgen; and payment for lectures from Dr Falk Pharma. Dr Peck-Radosavljevic reports consulting fee or honorarium and support for travel to meetings for the study or other purposes from Novartis; board membership for BioAlliance Pharma; consultancy for Bayer, Gilead, Merck Sharpe Dohme, Jansen, Roche, AbbVie, Bristol-Myers Squibb, and Boehringer Ingelheim; grants or grants pending for Bayer, Abbott, and Gilead; payments for lectures including service on speakers’ bureaus for Bayer, Merck Sharpe Dohme, Roche, Eli Lilly, Bristol-Myers Squibb, and Jansen; payment for manuscript preparation for Bayer; and payment for development of educational presentation for Bayer and Gilead. Dr Daniele reports board membership for Bayer and Daiichi Sankyo, consultancy for Bayer, and payments for lectures including service on speakers’ bureaus for Bayer, Daiichi Sankyo, and Novartis. Dr Furuse reports consulting fee or honorarium from Novartis; fees for participation in review activities such as data monitoring boards, statistical analysis, end point committees, and the like from Novartis; board membership for Chugai Pharmaceutical, Taiho Pharmaceutical, Bayer Pharmaceutical, Zeria Pharmaceutical, GlaxoSmithKline, and Yakult; consultancy for GlaxoSmithKline, Zeria Pharmaceutical, Ono Pharmaceutical, and Boehringer Ingelheim; grants or grants pending from Taiho Pharmaceutical, Chugai Pharmaceutical, Pfizer, Bayer Pharmaceutical, GlaxoSmithKline, Takeda BioDevelopment Center Limited, Eli Lilly Japan, Yakult, Ono Pharmaceutical, Onco Therapy Science, and Takeda Pharmaceutical; payment for lectures including service on speakers’ bureaus for Chugai Pharmaceutical, Taiho Pharmaceutical, Bayer Pharmaceutical, Eli Lilly Japan, and Pfizer. Ms Jappe reports employment and stock/stock options for Novartis Pharma AG. Mr Perraud reports employment for Novartis Pharma AG. Dr Anak reports employment and stock/stock options for Novartis Pharma AG. Dr Sellami reports employment and stock/stock options for Novartis Pharmaceuticals Corporation. Dr L.-T. Chen reports support for travel to meetings for the study or other purposes, consultancy, grants/grants pending, and payment for lectures including service on speakers’ bureaus for Novartis. No other disclosures were reported.
Funding/Support: This study was sponsored by Novartis Pharmaceuticals. Medical writing and editorial support in the preparation of the manuscript was provided by Melanie Leiby, PhD, and Dolores Matthews, MEd, of ApotheCom (Yardley, Pennsylvania) and was supported by Novartis Pharmaceuticals.
Role of the Sponsor: Representatives of the study sponsor contributed to the design of the trial. Data were collected through the sponsor’s data management system and were analyzed by their statistical team. An independent data monitoring committee reviewed interim efficacy analyses and oversaw safety data. As authors of the manuscript, representatives of the sponsor contributed to data interpretation and writing, reviewing, and amending the manuscript and participated in the decision to submit the manuscript for publication.
Previous Presentation: These data were presented in part at the 2014 Gastrointestinal Cancers Symposium; January 16-18, 2014; San Francisco, California.
Study Steering Committee: Co-chairs: Zhu, L.-T. Chen (senior author). Members: Kang, Lim, Poon, Vogel, Daniele, Furuse, Perraud, Anak, Sellami.
Additional Contributions: We thank Ghassan Abou-Alfa, MD (Memorial Sloan-Kettering Cancer Center, New York, New York), Hironobu Minami, MD (Kobe University Hospital, Kobe, Japan), Richard S. Finn, MD (University of California, Los Angeles), and Lee-Jen Wei, PhD (Harvard School of Public Health, Boston, Massachusetts) for serving on the independent data monitoring committee (chaired by Dr Abou-Alfa) that reviewed interim efficacy analyses and oversaw safety data. These individuals were compensated. We thank the following investigators for enrolling patients in the EVOLVE-1 study: Australia: Jonathan Cebon (Ludwig Institute for Cancer Research, Heidelburg), Jacob George (Westmead Hospital, Westmead), Simone Strasser (Royal Prince Alfred Hospital, Camperdown), Amany Zekry (St George Hospital Clinical School, University of New South Wales, Sydney). Austria: Rudolf Stauber (LKH Graz Med. Universität Klinik, Graz), Wolfgang Vogel (Univ Klinik fur Innere Medizin Innsbruck, Innsbruck). Belgium: Ivan Borbath (Cliniques Universitaires Saint-Luc, Brussels), Marc Peeters (UZ Antwerpen, Edegem), Jean-Luc Van Laethem (CHU Erasme, Brussels), Chris Verslype (Gasthuisberg University Hospital, Leuven). Canada: Yoo-Joung Ko (Sunnybrook Health Sciences Center, Toronto), Walter Kocha (London Regional Cancer Cantre, London), Richard Létourneau (CHUM–Campus St Luc, Montreal), Peter Metrakos (MUHC–Royal Victoria Hospital, Montreal). China: Minshan Chen (Sun Yat-sen University Cancer Center, Guangzhou), Guohong Han (Xijing Hospital, Fourth Military Medical University, Xi’an), Hongming Pan (Sir Run Run Shaw Hospital and Zhejiang University Medical College, Zhejiang), Shukui Qin (PLA No. 81 Hospital, Nanjing), Jianming Xu (307 Hospital of the PLA, Beijing), Lvnan Yan (West China Hospital of Sichuan and University Chengdu, Sichuan). France: Sandrine Faivre (Hôpital Beaujon, Clichy), Armand Abergel (CHU de Clermont-Ferrand–Hôpital d’Estaing, Clermont Ferrand), Michel Doffoel (HUS–Nouvel Hôpital Civil, Strasbourg), Cyrille Féray (Hôtel-Dieu Hospital, Nantes), René Gerolami Santandrea (AP-HP Hôpital de la Conception, Marseille), Alexandra Heurgué-Berlot (Hôpital Robert Debré, Reims), Come Lepage (CHU–Hôpital du Bocage, Dijon), Phillipe Merle (Hôpital de la Croix Rousse, Lyon), Pierre Michel (CHU de Rouen, Rouen), Laurent Mineur (Institut Sainte Catherine, Avignon), Eric Nguyen-Khac (CHU d’Amiens–Hôpital Nord, Amiens), Isabelle Ollivier (CHU d’Amiens–Hôpital Côte de Nacre, Caen), Jean Marc Phelip (CHU de Saint Etienne–Hôpital Nord, Saint Priest en Jarez), Albert Tran (Hôpital de l’Archet 2, Nice). Germany: Christoph Antoni (University Medical Centre Mannheim, University of Heidelberg, Mannheim), Thomas Berg (Univ. Klinikum Leipzig, Leipzig), Michael Geissler (Städt Kliniken Esslingen, Esslingen), Frank Kolligs (Klinikum der Univ. München–Grosshadern, Munich), Michael Scheurlen (Klinik der Universität Würzburg, Würzburg), Eckart Schott (Charité Berlin Campus Vircho-Klinikum, Berlin), Joerg Schlaak (University Hospital Essen, Essen), Cameron Silke (Universitätskliniken Göttingen, Göttingen), Jörg Trojan (Universitätsklinikum Frankfurt, Frankfurt), Henning Wege (Universitätsklinikum Hamburg-Eppendorf, Hamburg). Greece: Christos Papandreou (University General Hospital of Larisa, Larisa), Amanda Psyrri (Attikon University General Hospital of Athens, Athens), Nikolaos Touroutoglou (Interbalkan Medical Center, Thessaloniki). Hungary: András Csejtei (Markusovszky Korhaz, Szombathely), István Láng (Orszagos Onkologiai Intezet, Budapest), János Szántó (Institute of Oncology, Medical and Health Science Center, University of Debrecen, Debrecen), László Thurzó (SZOTE Onkoterapias Klinika Szeged). Israel: Dan Aderka (Chaim Sheba Medical Center, Ramat Gan), Salomon Stemmer (Rabin Medical Center, Petah-Tiqva). Italy: Raffaele Addeo (Osp. Presidio Ospedaliero S. Giovanni di Dio, Frattamaggiore), Carlo Barone (Policlinico Univ A Gemelli–Univ Cattolica del Sacro Cuore, Rome), Antonio Bernardo (Fondzione Salvatore Maugeri IRCCS–1st Scientifico Pavia, Pavia), Luigi Bolondi (Az Osp. di Bologna Policl. S. Orsola-Malpighi, Bologna), Ignazio Carrea (Az Osp. Univ. Policlinico P. Giaccone Università Studi Palermo, Palermo), Francesco Cognetti (Istituti Fisioterapici Ospitalieri–Polo Onco. Regine Elena, Rome), Massimo Colombo (Fond. IRCCS Cà Granda Osp. Maggiore Policlinico, Milan), Camillo Porta (Fodazione IRCCS Policlinico S. Matteo, Univ. degli Studi Pavia, Pavia), Fabio Farinati (Azienda Ospedaliera–Università di Padova, Univer. Degli Studi, Padova), Sergio Frustaci (Centro di Riferimento Oncologico IRCCS, Aviano), Rosario Vincenzo Iaffaioli (IRCCS Fondazione G Pascale, Napoli), Gabriele Luppi (A.O. Univ Policl de Modena, Univ. Studi Modena e R. Emilia, Modena), Sante Romito (AO Universitaria OO.RR Foggia–Presidio Osp Riuniti Foggia, Foggia). Japan: Masayuki Furukawa (National Kyushu Cancer Center, Fukuoka), Masafumi Ikeda (National Cancer Center Hospital East, Kashiwa), Yoshitaka Inaba (Aichi Cancer Center Hospital, Aichi), Takao Iwasaki (Tohoku University Hospital, Sendai), Akihide Masumoto (Iizuka Hospital, Iizuka), Seijin Nadano (National Hospital Organization, Shikoku Cancer Center, Matsuyama), Yoichi Nishigaki (Gifu Municipal Hospital, Gifu), Kazushi Numata (Yokohama City University Medical Center, Yokohama), Takuji Okusaka (National Cancer Center Hospital Chuoku, Tokyo), Hideki Saitsu (NHO Kyushu Medical Center Fukuoka, Fukuoka), Yutaka Sasaki (Kumamoto University Hospital, Kumamoto), Takuma Teratani (Kanto Medical Center NTT EC, Shinagawa-ku), Hidenori Toyoda (Ogaki Municipal Hospital, Ogaki), Tatsuya Yamashita (Kanazawa University Hospital, Kanazawa), Osamu Yokosuka (Chiba University Hospital, Chiba). Korea: Hyun Cheol Chung (Severance Hospital, Seoul), Tae-You Kim (Seoul National University Hospital, Seoul). Spain: Jordi Bruix (Hospital Clinic I Provincial de Barcelona, Barcelona), Meritxell Casas Rodrigo (Consorci Hospitalari Parc Tauli, Barcelona), Lluis Castells Fuste (Hospital Vall D’Hebron, Barcelona), Enrique Grande Pulido (Hospital Ramon y Cajal Madrid, Madrid), Jose Montero Alvarez (Hospital Universitario Reina Sofia, Córdoba), Bruno Sangro Gomez Acebo (Clinica Universitaria de Navarra, Pamplona). Taiwan: Yee Chao (Taipei Veterans General Hospital, Taipei), Long-Bin Jeng (China Medical University Hospital, Taichung), Cheng-Yao Lin (Chi Mei Foundation Hospital–Liouying, Liouying Township), Deng-Yn Lin (Chang-Gung Memorial Hospital-Lino, Lin-Ko), Ying-Chun Shen (National Taiwan University Hospital, Taipei). Thailand: Touch Ativitavas (Ramathibodi Hospital, Bangkok), Vichien Srimuninnimit (Siriraj Hospital, Bangkok). United States: Jared Acoba (Queen’s Medical Center, Honolulu, Hawaii), Sanjiv Agarwala (St Luke’s Hospital and Health Network, Bethlehem, Pennsylvania), Ari Baron (California Pacific Medical Center, San Francisco), J. Thaddeus Beck (Highlands Oncology Group, Fayetteville, Arkansas), Allen Cohn (Rocky Mountain Cancer Centers, Denver, Colorado), David Cosgrove (Sidney Kimmel Cancer Center at Johns Hopkins Hospital, Baltimore, Maryland), Bradley Freilich (Midwest Cancer Care Physicians, Kansas City, Missouri), William Harris (Fred Hutchinson Cancer Research Center/Seattle Cancer Care Alliance, Seattle, Washington), Aram Hezel (University of Rochester Medical Center, Rochester, New York), Haresh Jhangiani (Compassionate Cancer Care Medical Group, Fountain Valley, Fountain Valley, California), Roy MacKintosh (VA Sierra Nevada Health Care System, Reno), Anthony Reid (University of California, San Diego, La Jolla), Spencer Shao (Northwest Cancer Specialists, Portland, Oregon), Carlos Taboada (Methodist Charlton Cancer Center, Dallas, Texas), Udit Verma (University of Texas Southwestern Medical Center, Dallas).
Create a personal account or sign in to: