- Email: [email protected]
Multikinase Inhibitors in Metastatic Renal Cell Carcinoma: Indirect Comparison Meta-Analysis
Clinical Therapeutics/Volume 33, Number 6, 2011
Multikinase Inhibitors in Metastatic Renal Cell Carcinoma: Indirect Comparison Meta-Analysis Henry W.C. Leung, MD, MBA1; and Agnes L.F. Chan, PhD, MAMM, BscPharm2 1 2
Department of Radiation Oncology, Wan Fang Hospital, Taipei Medical University, Taipei, Taiwan; and Department of Pharmacy, Chi Mei Medical Center, Tainan, Taiwan
ABSTRACT Background: Randomized controlled trials (RCTs) of multikinase inhibitors sunitinib, sorafenib, and pazopanib have reported efficacy compared with results from placebo and interferon-␣ (INF-␣). To date, these drugs have not been compared in head-to-head trials. Objective: To review systematically the evidence of clinical effectiveness of multikinase inhibitors in the treatment of metastatic renal cell carcinoma (mRCC) and, via an indirect meta-analysis, to determine an optimal treatment among these agents. Methods: A systematic literature search of MEDLINE, EMBASE, CANCERLIT, and Cochrane controlled trials register databases was performed. All RCTs of multikinase inhibitors (sorafenib, sunitinib, and pazopanib) used to treat mRCC were included. The study selection, data extraction, and quality assessment were performed independently by 2 reviewers, with all disagreements being resolved by consensus. The effects of multikinase inhibitors on progression-free survival (PFS) were compared using an indirect treatment comparison method with INF-␣ or placebo as a comparator. Results: Four studies were included. Two studies examined sunitinib or sorafenib versus IFN-␣, and the other 2 studies investigated sorafenib or pazopanib versus placebo. Compared with placebo, 2 interventions reported improvement for PFS (sorafenib: hazard ratio [HR] ⫽ 0.44, P ⫽ 0.01; pazopanib: HR ⫽ 0.46, P ⫽ 0.0001), whereas only sunitinib improved PFS over IFN-␣ (HR ⫽ 0.539, P ⫽ 0.001). An indirect comparison suggests that sunitinib is likely to demonstrate greater clinical benefit than sorafenib in terms of PFS (HR ⫽ 0.47; 95% CI, 0.316 – 0.713; P ⬍ 0.001), using IFN-␣ as the comparator. Sorafenib was not statistically different from pazopanib using placebo as the comparator in the indirect comparison (HR ⫽ 0.957; 95% CI, 0.657–1.39; P ⫽ 0.24). Conclusion: Some multikinase inhibitors have a favorably reported PFS for patients with mRCC compared
708
with results using IFN-␣ or placebo. Our findings suggest that sunitinib might offer some clinical benefit over sorafenib in terms of PFS. No statistical difference was found between sorafenib and pazopanib treatments. However, these conclusions are based on 2 indirect comparisons of single RCTs. More RCTs are required to confirm these findings and investigate the clinical effectiveness of multikinase inhibitors in the treatment of mRCC. (Clin Ther. 2011;33:708 –716) © 2011 Elsevier HS Journals, Inc. All rights reserved. Key words: metastatic renal cell carcinoma, multikinase inhibitors, progression-free survival.
INTRODUCTION In 2009 in the United States, approximately 57,760 people (35,430 men and 22,330 women) were diagnosed with cancer of the kidney and renal pelvis, and approximately 12,980 people died of the disease. Furthermore, more than 63,000 new cases of renal cell carcinoma (RCC) and 26,000 deaths were reported in 2006 in Europe. The incidence rates by race in Asian and Pacific Island countries were approximately 9.6 and 4.8 per 100,000 men and women, respectively. For blacks and whites, the incidence rates were 21.3 per 100,000 men versus 10.3 per 100,000 women and 19.2 per 100,000 men versus 9.9 per 100,000 women, respectively.1 The overall 5-year survival rate for RCC is now nearly 60%; however, actual survival depends on the stage of the disease at diagnosis and the grade of the tumor cells.2 In England and Wales in 2004, the 5-year survival rate for metastatic RCC (mRCC) was approximately 10%.3 In Taiwan, the incidence rate for mRCC in 2007 was approximately 1.05% and 522 in the same year.4 Accepted for publication May 3, 2011. doi:10.1016/j.clinthera.2011.05.003 0149-2918/$ - see front matter © 2011 Elsevier HS Journals, Inc. All rights reserved.
Volume 33 Number 6
H.W.C. Leung and A.L.F. Chan As our understanding of the molecular mechanisms involved in the pathogenesis of mRCC develops, more advanced treatment options are emerging, such as multikinase inhibitors (eg, sunitinib, sorafenib, pazopanib), antivascular endothelial growth factor agents (eg, bevacizumab), and mammalian targets of rapamycin inhibitors (eg, temsirolimus and everolimus). According to the National Institute for Health and Clinical Excellence (NICE) technology appraisal guidance 178, the assessment group concluded that for patients eligible for immunotherapy, bevacizumab plus interferon-␣ (IFN-␣) appears to be superior to IFN-␣ alone as a first-line treatment, as measured by progressionfree survival (PFS) and tumor response. For patients with poor prognosis, temsirolimus appears to be superior to IFN-␣ in terms of overall survival (OS), PFS, and tumor response rate. However, for patients who do not respond to immunotherapy, sorafenib and sunitinib are likely to demonstrate clinically and statistically significant benefits compared with best supportive care in terms of PFS and tumor response rate.5 New multikinase inhibitors have now been compared with IFN-␣ and placebo6-9 as treatments for mRCC, although they have not been evaluated directly in headto-head trials (Table I). Sunitinib, but not sorafenib, has been reported to have a longer PFS when compared with IFN-␣ in Phase III trials (Table II).7,8 In addition, phase III trials of pazopanib have reported significant improvements in PFS and tumor response compared with results using placebo (Table II).9 The limited amount of information on the clinical efficacy, cost, patient experience, and adverse events (AEs) associated with these agents will likely play an important role both in decisions related to reimbursement by the Taiwan national health insurance program and in the availability of these new drugs to hospitals. To determine the optimal treatment for a rapidly developed targeted therapy, we conducted a meta-analysis of all randomized controlled trials (RCTs) of multikinase inhibitors used to treat mRCC by an indirect comparison method.10
METHODS Inclusion Criteria Any randomized clinical trial evaluating the therapeutic efficacy of the multikinase inhibitors sunitinib, sorafenib, or pazopanib for the treatment of mRCC was included. All included trials evaluated 1 of the study drugs with a control intervention as the sole treatment. Conference abstracts were included if there
June 2011
were sufficient data to assess the quality of the studies or if detailed results were provided. Nonrandomized trials, pharmacokinetic studies, animal studies, and laboratory studies were excluded from this study. Only studies with institutional review board approval and appropriate consent were included.
Search Strategy A systematic literature search of MEDLINE, EMBASE, CANCERLIT and Cochrane databases was performed. The strategy consisted of searching for publications using key terms related to the target drugs (eg, trade names) and various terms used to describe renal cancer and neoplasms. The initial searches combined these terms with the heading neoplasm from the cancer Medical Subject Headings [MeSH] and looked for publications from January 2008 to March 2009. A subsequent search was run in March 2010 to identify any relevant publications added to the databases since March 2009. The searches were limited to original publications in English that described RCTs with human subjects. Bibliographies of included studies were searched manually to identify additional relevant studies. Individual conference abstracts from the American Society of Clinical Oncology (ASCO) were also searched. Specifically, we searched the ASCO website for abstracts from the 2007 and 2009 Annual Meeting and the 2007 and 2009 Genitourinary Cancers Symposium.
Data Extraction Two reviewers independently examined the titles and abstracts of the identified studies and the full texts of all potentially relevant studies. Data extraction was conducted using a standardized form. Two reviewers initially screened the abstracts to exclude obviously ineligible studies and reviewed the full texts of all remaining studies. The following data were collected into the standardized form: patient characteristics, study design, interventions, and direct outcomes. The quality and content of each study were assessed independently and appraised according to the inclusion and exclusion criteria of the Jadad scale, an RCT appraisal tool.11,12 Data were extracted independently by 2 reviewers with all disagreements resolved by consensus. Details of the data on primary end point PFS, OS, median duration of treatment, duration of treatment response, and AEs were recorded, if available, on a standardized form.
709
Study
Intervention and Comparator
Patients’ Ages (Median)/Gender (%)
ECOG
PS
0
1
MSKCC Prognostic Risk (%) Low/Intermediate*
Undergone Previous Nephrectomy (%)
Dosage Regimen Severe AEs (ⱖGrade 3 or Case Death) (%)
Dose (mg/kg), Duration
Sorafenib: placebo ⫽ 34:24 (hospitalization or death) Sor: INF ⫽ 16.5:14.4
Sor: 400 mg BID 6-wk cycle
Design
Diagnosis
TARGET Escudier et al 20076 Escudier et al 20097
Phase III RCT 903 patients
Metastatic clear-cell RCC Metastatic clear-cell RCC
Sor vs Pl
58 vs 59 Male: 70 vs 75
52%/48%
Sor: 49 Pl: 46
Sor: 49 Pl: 52
Sor:Pl ⫽ 94: 93
Sor vs 1FN-␣-2a
62 vs 62.5 Male/female 67 vs 33/56.5 vs 43.5
Sor: 53.6/45.4 1FN: 51.1/47.8
Sor: 57.7 INF: 53.3
Sor: 42.3 INF: 46.7
Sor:INF ⫽ 97.9:90.2
SUTENT Motzer et al 20098
Phase III RCT 750 patients
Metastatic clear-cell RCC
Sun vs 1FN-␣
62 vs 59 Male/female 71 vs 29/72 vs 28
Favorable/intermediate/ poor sun group: 38%/ 56%/6% IFN group: 32%/51%/7%
NA
NA
NA
Sor: IFN ⫽ 3:1
VEG105192 Sternberg et al 20109
Phase III RCT 435 patients
Metastatic clear-cell RCC
Pa vs Pl
59 vs 60 Male: 68 vs 75
Favorable/intermediate/ poor Pa: 39/55/3 P1: 39/55/3
Pa: 42 P1: 41
Pa: 58 P1: 59
Pa: 89 P1: 88
Pa/IFN death ⫽ 4:3 grade 3 ⫽ 33:14 grade 4 ⫽ 7:6
Phase II RCT 189 patients
Sor: 400 mg BID (period 1) 600 mg BID (period 2) IFN:9 Mu 3 times weekly Sun: 50 mg once daily on a 4 wk on, 2 wk off IFN:SC thrice weekly, 3 Mu/dose the first week; 6 Mu the second week; 9 Mu thereafter Pa: 800 mg once daily
Volume 33 Number 6
AE ⫽ adverse event; ECOG ⫽ Eastern Cooperation Oncology Group; IFN-␣-2a ⫽ interferon alpha-2a; MSKCC ⫽ Memorial Sloan-Kettering Cancer Center; Mu ⫽ million unit; NA ⫽ not applicable; Pa ⫽ pazopanib; Pl ⫽ placebo; PS ⫽ performance status; RCC ⫽ renal cell carcinoma; RCT ⫽ randomized controlled trial; SC ⫽ subcutaneous; Sor ⫽ sorafenib; Sun ⫽ sunitinib. *Unless noted otherwise.
Clinical Therapeutics
710 Table I. Characteristics of patients and included studies.
H.W.C. Leung and A.L.F. Chan
2010, USA
2009, USA
2009, USA
TARGET Escudier et al6 17. Escudier et al7 SUTENT Motzer et al8 VEG105192 Sternberg et al9
HR ⫽ hazard ratio; INF-␣ ⫽ interferon-␣; NS ⫽ not significant; OS ⫽ overall survival; PFS ⫽ progression-free survival; RCC ⫽ renal cell carcinoma; RCT ⫽ randomized controlled trial.
435
–
9.2 vs 4.2
–
0.46
–
0.34–0.62
OS: NS PFS: 0.001 0.0001 0.451–0.643 0.637–1.001 0.539 0.821 11 vs 5 26.4 vs 21.8 750
– 5.7 vs 5.6 189
Not reached
0.88
–
0.61–1.27
OS: 0.02 PFS: 0.01 0.504 0.35–0.55 0.54–0.94 0.44 –0.72 5.5 vs 2.8 903
Sorafenib vs placebo Sorafenib vs IFN-␣ Sunitinib vs IFN-␣ Pazopanib vs placebo 2007, USA
19.3 vs 15.9
PFS OS PFS OS N
June 2011
Study/Authors
Year of Intervention Publication, and Country Comparator
Table II. Direct comparisons of included studies.
Outcome Measured (median months)
OS
HR
PFS
95% CI for HR
P
Data Analysis The hazard ratios (HRs) and appropriate 95% CIs of survival data (OS and PFS) were extracted from the original studies. The HR describes the relative risk complications based on a comparison of event rates; it has also been used to describe the outcome of therapeutic trials as a measure of the extent to which a treatment shortens the duration of the illness. It is a constant under the Cox proportional hazard model. The P value is used to reject the null hypothesis that HR ⫽ 1, that is, that treatment is not beneficial. Therefore, an HR value of ⬍1 favors the drug being evaluated.13 Studies were pooled if more than 1 trial of the studied drug existed. Studies for each drug were pooled using a random effects model.14 Owing to the lack of head-to-head comparison that RCTs study between the interventions, we performed indirect comparisons of the interventions with the same comparators across similar patient groups using an indirect treatment comparison method, as described by Bucher et al.10 In this method, the randomization of each trial was maintained and the summary estimates of pooled interventions were compared with 95% CIs. SAS (version 9.0, Cary, North Carolina) was used for all data analysis.
RESULTS Quantity of Studies The primary literature search identified 101 abstracts. We excluded 80 papers because they were not RCTs or original studies or did not examine at least 1 of our target drugs. A total of 21 full-length articles were reviewed. After reviewing the full text, 6 articles matched our inclusion criteria.6-9,15 Two of 6 relevant articles were excluded because their results were based on the same patients and the same trial.8,15 Eight abstracts from relevant conferences and annual meetings were reviewed, but no new abstracts met our inclusion criteria. The flow diagram of the 4 included studies is depicted in Figure 1.
Characteristics of Patients All patients enrolled in the included studies were ⱖ18 years, were diagnosed with metastatic clear-cell RCC, had a life expectancy ⱖ12 weeks, and had an Eastern Cooperation Oncology Group (ECOG) performance status of 0 or 1. The majority of patients had undergone a nephrectomy or had received a cytokinebased treatment. Among those in the included studies, patients treated with sorafenib, sunitinib, or pazo-
711
Clinical Therapeutics
101 abstracts 80 were excluded: • Not an RCT • Not an original article • Did not examine targeted drugs 21 full-length articles 15 were excluded: • Not an RCT • Not an original study • Did not examine targeted drugs
2 were excluded: • Same population, same trial
4 full-length articles included
Figure 1. The flow diagram of included studies. RCT ⫽ randomized controlled trial.
panib were stratified as having either intermediate-risk or low-risk status and in other studies as having favorable-, intermediate-, or poor-risk status, according to the Memorial Sloan-Kettering Cancer Center prognostic score.17-19 From the AEs reported in these trials, the frequency of serious AEs leading to hospitalization or death may be higher in the sorafenib group than in the placebo (34% vs 24%) or IFN-␣ (16.5% vs 14.4%) groups.6,7 In the sunitinib trial, the proportion of grade 3 or 4 AEs with sunitinib ranged from 1% to 16% for all categories. The most commonly reported sunitinibrelated grade 3 AEs included hypertension (12%), fatigue (11%), diarrhea (9%), and hand-foot syndrome (9%). No grade 4 AEs were reported during treatment.8 During treatment with pazopanib, AEs resulting in death were reported in 4% of patients in the pazopanib arm and in 3% of patients in the placebo arm. The proportions of patients experiencing an AE with grade 3 or 4 were 33% and 7%, respectively, in the pazopanib arm and 14% and 6%, respectively, in the placebo arm. The most common grade 3 or 4 AEs in the pazopanib arm were hypertension (4%) and diarrhea (4%) (Table I).9 Patients with brain metastases, previous malignancy, or a history of severe cardiac disease were excluded from all studies. All studies explained the inclusion and exclusion criteria that were utilized.
712
Characteristics of Included Studies All included studies were RCTs; 3 studies were Phase III, international, multicenter, randomized clinical trials; and 1 study was a Phase II trial. Two studies compared the effects of sorafenib (400 mg– 600 mg twice daily) or pazopanib (80 mg once daily) on PFS with the placebo.6,9 Another 2 studies compared the effects of sorafenib or sunitinib (50 mg once daily taken orally in 6-week cycles with 4 weeks on and 2 weeks off) with those of IFN-alfa-2a (median daily dose of 8.59 million U) on PFS.7,8 The characteristics of each study are provided in Table I.
Meta-analysis of Progression-free Survival Four trials were analyzed that compared the effect of at least 1 new multikinase inhibitor (sorafenib, sunitinib, or pazopanib) with IFN-␣ or placebo on PFS. One of the trials compared the effects of pazopanib and placebo on PFS in 2 patient subpopulations that were either treatment-naïve or cytokine-pretreated. We pooled 2 trials that compared sorafenib or sunitinib with IFN-␣ and found a pooled HR of 0.47 (95% CI, 0.316 – 0.713; P ⬍ 0.001). Two trials that compared sorafenib and pazopanib with placebo were also pooled, and a pooled HR of 0.957 was found (95% CI, 0.657– 1.393; P ⫽ 0.24). Figure 2 and Figure 3 illustrate the
Volume 33 Number 6
H.W.C. Leung and A.L.F. Chan
Hazard Ratio
SE
Weight
Hazard Ratio IV, Random, 95% CI
Escudier 20097
0.88
0.51
20.2%
0.88 [–0.12, 1.88]
Motzer 200716
0.42
0.64
12.8%
0.42 [–0.83, 1.67]
0.54
0.28
67.0%
0.54 [–0.01, 1.09]
100.0%
0.59 [0.14, 1.04]
Study or Subgroup
Motzer
20098
Total (95% CI)
Heterogeneity: τ² = 0.00; X² = 0.43, df = 2 (P = 0.81); I² = 0% Test for overall effect: Z = 2.59 (P = 0.010)
Hazard Ratio IV, Random, 95% CI
–4 –2 Favors control
0 2 4 Favors experimental drug
Figure 2. Forest plot of 3 studies comparing sorafenib to interferon-␣ (control). SE ⫽ standard error.
intervention compared with IFN-␣ or placebo on PFS outcomes. We did not pool study outcomes for OS because completed data were not available in all studies.
Indirect Comparisons of Multikinase Inhibitors Versus IFN-␣ and Placebo On Progression-free Survival We performed an indirect comparison of the effects of multikinase inhibitors on PFS. The results of the indirect comparison of sunitinib versus sorafenib on PFS suggested that sunitinib may have significant benefits compared with sorafenib (HR ⫽ 0.47; 95% CI, 0.316 – 0.713; P ⬍ 0.001) using IFN-␣ as the comparator. We did not observe a significant difference between sorafenib and pazopanib in terms of PFS (HR ⫽ 0.957; 95% CI, 0.657– 1.393; P ⫽ 0.24) using placebo as a comparator (Table III).6,9
Study Quality The results of the quality appraisal were analyzed using the Jadad scale.11,12 Two studies scored a 5 because
Study or Subgroup Escudier 20076 Escudier
20097
Sternberg 20109 Total (95% CI)
Hazard Ratio
SE
0.44
0.28
0
0
0.46
0.32
Weight
the description of randomization and technique was adequate.6,9 By contrast, the other 2 studies scored a 3 on the Jadad scale because the description of double-blind or the method of blinding was inappropriate.7,8 The populations in all studies were similar and predominantly included patients with clear-cell mRCC (Table IV).
DISCUSSION The results of the meta-analysis of 3 multikinase inhibitors (sorafenib, sunitinib, and pazopanib) suggest that treatment with sorafenib or sunitinib is likely to have clinical advantages over treatment with IFN-␣ alone. The results also suggest that treatment with sorafenib or pazopanib is likely to have a greater treatment effect than placebo when the effect on PFS is considered. In trials using IFN-␣ as a comparator, the median PFS was significantly prolonged following treatment with sunitinib rather than sorafenib, (5–11 months vs 5.6 – 5.7 months, respectively). When placebo was used as a comparator, the median PFS with pazopanib was slightly longer than that with sorafenib (9.2 vs 4.2
Hazard Ratio IV, Random, 95% CI
56.6%
0.44 [–0.11, 0.99]
43.4%
0.46 [–0.17, 1.09]
100.0%
0.45 [0.04, 0.86]
Hazard Ratio IV, Random, 95% CI
Not estimable
Heterogeneity: τ² = 0.00; X² = 0.00, df = 1 (P = 0.96); I² = 0% Test for overall effect: Z = 2.13 (P = 0.03)
–4 –2 Favors control
0 2 4 Favors experimental drug
Figure 3. Forest plot of 3 studies comparing sorafenib and pazopanib to placebo (control). SE ⫽ standard error.
June 2011
713
Clinical Therapeutics
Table III. Indirect comparison of studies. Study
Intervention
TARGET vs VEG105192 Escudier et al vs SUTENT
Sorafenib vs pazopanib Sunitinib vs sorafenib
Direct HR (PFS)
Indirect HR (PFS)
95% CI for HR
0.44 vs 0.46
0.35–0.55 vs 0.34–0.62
0.539 vs 0.88
0.451–0.643 vs 0.61–1.27
95% CI for HR
0.957
0.657–1.393
0.47
0.316–0.713
CI ⫽ confidence interval; HR ⫽ hazard ratio; PFS ⫽ progression-free survival.
months and 5.5 vs 2.8 months, respectively). The reason for the slightly lower value of PFS for sorafenib in the trials was not clear. The discrepancy could be the result of small differences in patient characteristics or differences in the PFS assessment. We did not perform an indirect comparison of the effect of interventions on OS data because there was a lack of final OS data reported in the studies analyzed and because of uncertainty regarding poststudy medication usage. We sought all RCTs of targeted multikinase inhibitors for mRCC. A variety of factors reportedly influence prognosis of mRCC, including the following: patient access (whether patients will be able to receive several lines of therapy), high level of lactate dehydrogenase (⬎1.5 times upper limit of normal), low serum
hemoglobin (⬍ lower limit of normal), high corrected serum calcium (⬎10 mg/dL), alkaline phosphatase, time from initial diagnosis, tumor stage (nodes and sites of metastasis), number of organs involved, absence of prior nephrectomy, and scores of performance, such as ECOG status or Karnofsky score.17,20,21 In this review, analysis of the populations’ baseline characteristics in the included studies revealed that the age, sex, and ECOG performance status were similar among all patients. However, according to Memorial Sloan-Kettering Cancer Center criteria for mRCC prognosis, patient stratification differed among studies. Patients were classified as either favorable/low or intermediate risk in some studies, whereas other studies stratified patients as having favorable/low, intermediate, or poor risk. In addition, refrac-
Table IV. Quality assessments for each study with Jadad Scale. Questions
Answer
Score
Escudier et al6
Escudier et al7
Motzer et al8
Sternburg et al9
Randomized Study? Randomization technique described and adequate Randomization technique described and inadequate
Yes/No
⫹1/0
Yes
Yes
Yes
Yes
Yes
⫹1
Yes
Yes
Yes
Yes
Yes
–1
No
No
No
No
Yes/No
⫹1/0
Yes
No
No
Yes
Yes
⫹1
Yes
No
No
Yes
Yes
–1
No
No
No
No
Yes/No
⫹1/0
Yes 5
Yes 3
Yes 3
Yes 5
Double blinded study? Technique of blinding described and adequate Technique of blinding described and inadequate Description of withdrawals and dropouts? Result of Jaded scale
714
Volume 33 Number 6
H.W.C. Leung and A.L.F. Chan tory patients were identified as low or intermediate risk. Because we hoped to estimate the overall treatment effect of targeted drugs, we included all groups of treatment classifications in our meta-analysis. This might have resulted in a selection bias that could have altered the outcome. An additional source of bias was the subjective and intermittent assessment of progression in the included studies. Two trials used an independent blinded central review to evaluate the primary PFS in the first assessment but did not in the second assessment.7,8 Moreover, differing opinion on diagnoses may exist between investigators and independent central reviewers. Finally, 1 trial did not clearly and appropriately mention whether it was double-blinded. In 2 trials comparing sorafenib or pazopanib with placebo, fatal AEs were reported in 10% of patients in the sorafenib arm versus 6% in the placebo arm and in 4% of patients in the pazopanib arm versus 3% in the placebo. From these data, the safety profile of sorafenib is likely to be worse than that of pazopanib and sunitinib, which did not report any fatal AEs in their trial. The relative efficacy of different multikinase inhibitors has not been established because no head-to-head comparisons have been reported. We therefore performed an indirect comparison analysis to evaluate the effectiveness of these agents, which suggested that sunitinib may have comparable clinical benefits to sorafenib as a first-line treatment for mRCC patients at preventing progression. Findings here are consistent with a recently published systematic review, which included an adjusted comparison of the effects of treatment with sunitinib and treatment with sorafenib on PFS for mRCC.22 This study reported sunitinib to be more effective than sorafenib (HR ⫽ 0.58; 95% CI, 0.38 – 0.86; P ⱕ 0.001). By contrast, our results were unable to demonstrate significant differences between sorafenib and pazopanib on PFS (HR⫽ 0.957; 95% CI, 0.657–1.393; P ⫽ 0.24). A limitation of this review is that our study included only RCTs; therefore, it is possible that potentially useful trials of cohort or case-controlled design were not identified. Our findings should be interpreted with caution, as only 4 studies were analyzed, and, among these, only 2 indirect comparisons were made. In addition, this small selection of studies was not adequate for the exploration of patient subgroup analyses.
June 2011
CONCLUSION Compared with IFN-␣ or placebo, some multikinase inhibitors have reported favorable PFS for patients with mRCC. Our findings suggest that sunitinib might offer some clinical benefit over sorafenib in terms of PFS. No statistical difference was found between sorafenib and pazopanib treatments. However, these conclusions are based on 2 indirect comparisons of single randomized clinical trials. More such trials are required to confirm these findings and investigate the clinical effectiveness of multikinase inhibitors in the treatment of mRCC.
ACKNOWLEDGMENT This was an unfunded study. The authors have indicated that they have no conflicts of interest regarding the content of this article. The authors would like to thank the referees and editors for their useful comments. All authors contributed equally to the conduct of the study and creation of the manuscript.
REFERENCES 1. National Cancer Institute. SEER Stat Fact Sheets: Kidney and Renal Pelvis. http://seer.cancer.gov/statfacts/html/ kidrp.html. Accessed August 2010. 2. National Institute for Health and Clinical Excellence. Sunitinib for the first-line treatment of advanced and/or metastatic renal cell carcinoma. NICE technology appraisal guidance TA 169. March 2009. http://guidance.nice. org.uk/TA169. Accessed May 9, 2011. 3. Ljungberg B, Hanbury DC, Kuczyk MA, et al. Renal cell carcinoma guideline. Eur Urol. 2007;51:1502–1510. 4. Taiwan Cancer Registry. http://www.mc.ntu.edu.tw/CRS/ main.php?Page⫽A5. Accessed May 9, 2011. 5. National Institute for Health and Clinical Excellence. Bevacizumab (first-line), sorafenib (first- and second-line), sunitinib (second-line) and temsirolimus (first-line) for the treatment of advanced and/or metastatic renal cell carcinoma. NICE technology appraisal guidance 178. August 2009. http://www.nice.org. uk/nicemedia/pdf/TA178Guidance1.pdf.AccessedMay6,2011. 6. Escudier B, Eisen T, Stadler WM, et al, for the TARGET Study Group. Sorafenib in advanced clear-cell renal-cell carcinoma. N Engl J Med. 2007;356:125–134. 7. Escudier B, Szczylik C, Hutson TE, et al. Randomized phase II trial of first-line treatment with sorafenib versus interferon Alfa-2a in patients with metastatic renal cell carcinoma. J Clin Oncol. 2009;27:1280 –1289. 8. Motzer RJ, Hutson TE, Tomczak P, et al. Overall survival and updated results for sunitinib compared with interferon alfa in patients with metastatic renal cell carcinoma. J Clin Oncol. 2009;27:3584 –3590.
715
Clinical Therapeutics 9. Sternberg CN, Davis ID, Mardiak J, et al. Pazopanib in locally advanced or metastatic renal cell carcinoma: results of a randomized phase III trial. J Clin Oncol. 2010;28:1061–1068. 10. Bucher HC, Guyatt GH, Griffith LE, Walter SD. The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. J Clin Epidemiol. 1997; 50:683– 691. 11. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17:1–12. 12. Schwenk W, Haase O, Günther N, Neudecker J. Methodological quality of randomised controlled trials comparing short-term results of laparoscopic and conventional colorectal resection. Int J Colorectal Dis. 2007;22:1369 –1376. 13. Spruance SL, Reid JE, Grace M, Samore M. Hazard ratio in clinical trials. Antimicrob Agents Chemother. 2004;48:2787–2792. 14. Fleiss JL. The statistical basis of meta-analysis. Stats Methods Med Res. 1993;2:121–145. 15. Escudier B, Eisen T, Stadler WM, et al. Sorafenib for treatment of renal cell carcinoma: final efficacy and safety results of the phase III treatment approaches in renal cancer global evaluation trial. J Clin Oncol. 2009;27:3312–3318. 16. Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa in metastatic renal-cell carcinoma. N Engl J Med. 2007;356:115–124. 17. Motzer RJ, Mazumdar M, Bacik J, et al. Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol. 1999;17:2530 –2540. 18. Motzer RJ, Bacik J, Murphy BA, et al. Interferon-alfa as a comparative treatment for clinical trials of new therapies against advanced renal cell carcinoma. J Clin Oncol. 2002;20: 289 –296.
716
19. Memorial Sloan-Kettering Cancer Center. Renal cell carcinoma nomogram. http://www.mskcc.org/mskcc/ html/6156.cfm. Accessed May 9, 2011. 20. Mills EJ, Rachlis B, O’Regan C, et al. Metastatic renal cell cancer treatments: an indirect comparison meta-analysis. BMC Cancer. 2009;9:34.
21. Escudier B, Albiges L, Blesius A, et al. How to select targeted therapy in renal cell cancer. Ann Oncol. 2010;21(Suppl 7):vii59 –vii62. 22. Leibovich BC, Cheville JC, Lohse CM, et al. A scoring algorithm to predict survival for patients with metastatic clear cell renal cell carcinoma: a stratification tool for prospective clinical trials. J Urol. 2005;174:1759–1763.
Address correspondence to: Agnes L.F. Chan, PhD, Director of Pharmacy, Chi Mei Medical Center, 901, Cheung Hwa Road, Tainan, Taiwan. E-mail: [email protected]
Volume 33 Number 6
Multikinase Inhibitors in Metastatic Renal Cell Carcinoma: Indirect Comparison Meta-Analysis Henry W.C. Leung, MD, MBA1; and Agnes L.F. Chan, PhD, MAMM, BscPharm2 1 2
Department of Radiation Oncology, Wan Fang Hospital, Taipei Medical University, Taipei, Taiwan; and Department of Pharmacy, Chi Mei Medical Center, Tainan, Taiwan
ABSTRACT Background: Randomized controlled trials (RCTs) of multikinase inhibitors sunitinib, sorafenib, and pazopanib have reported efficacy compared with results from placebo and interferon-␣ (INF-␣). To date, these drugs have not been compared in head-to-head trials. Objective: To review systematically the evidence of clinical effectiveness of multikinase inhibitors in the treatment of metastatic renal cell carcinoma (mRCC) and, via an indirect meta-analysis, to determine an optimal treatment among these agents. Methods: A systematic literature search of MEDLINE, EMBASE, CANCERLIT, and Cochrane controlled trials register databases was performed. All RCTs of multikinase inhibitors (sorafenib, sunitinib, and pazopanib) used to treat mRCC were included. The study selection, data extraction, and quality assessment were performed independently by 2 reviewers, with all disagreements being resolved by consensus. The effects of multikinase inhibitors on progression-free survival (PFS) were compared using an indirect treatment comparison method with INF-␣ or placebo as a comparator. Results: Four studies were included. Two studies examined sunitinib or sorafenib versus IFN-␣, and the other 2 studies investigated sorafenib or pazopanib versus placebo. Compared with placebo, 2 interventions reported improvement for PFS (sorafenib: hazard ratio [HR] ⫽ 0.44, P ⫽ 0.01; pazopanib: HR ⫽ 0.46, P ⫽ 0.0001), whereas only sunitinib improved PFS over IFN-␣ (HR ⫽ 0.539, P ⫽ 0.001). An indirect comparison suggests that sunitinib is likely to demonstrate greater clinical benefit than sorafenib in terms of PFS (HR ⫽ 0.47; 95% CI, 0.316 – 0.713; P ⬍ 0.001), using IFN-␣ as the comparator. Sorafenib was not statistically different from pazopanib using placebo as the comparator in the indirect comparison (HR ⫽ 0.957; 95% CI, 0.657–1.39; P ⫽ 0.24). Conclusion: Some multikinase inhibitors have a favorably reported PFS for patients with mRCC compared
708
with results using IFN-␣ or placebo. Our findings suggest that sunitinib might offer some clinical benefit over sorafenib in terms of PFS. No statistical difference was found between sorafenib and pazopanib treatments. However, these conclusions are based on 2 indirect comparisons of single RCTs. More RCTs are required to confirm these findings and investigate the clinical effectiveness of multikinase inhibitors in the treatment of mRCC. (Clin Ther. 2011;33:708 –716) © 2011 Elsevier HS Journals, Inc. All rights reserved. Key words: metastatic renal cell carcinoma, multikinase inhibitors, progression-free survival.
INTRODUCTION In 2009 in the United States, approximately 57,760 people (35,430 men and 22,330 women) were diagnosed with cancer of the kidney and renal pelvis, and approximately 12,980 people died of the disease. Furthermore, more than 63,000 new cases of renal cell carcinoma (RCC) and 26,000 deaths were reported in 2006 in Europe. The incidence rates by race in Asian and Pacific Island countries were approximately 9.6 and 4.8 per 100,000 men and women, respectively. For blacks and whites, the incidence rates were 21.3 per 100,000 men versus 10.3 per 100,000 women and 19.2 per 100,000 men versus 9.9 per 100,000 women, respectively.1 The overall 5-year survival rate for RCC is now nearly 60%; however, actual survival depends on the stage of the disease at diagnosis and the grade of the tumor cells.2 In England and Wales in 2004, the 5-year survival rate for metastatic RCC (mRCC) was approximately 10%.3 In Taiwan, the incidence rate for mRCC in 2007 was approximately 1.05% and 522 in the same year.4 Accepted for publication May 3, 2011. doi:10.1016/j.clinthera.2011.05.003 0149-2918/$ - see front matter © 2011 Elsevier HS Journals, Inc. All rights reserved.
Volume 33 Number 6
H.W.C. Leung and A.L.F. Chan As our understanding of the molecular mechanisms involved in the pathogenesis of mRCC develops, more advanced treatment options are emerging, such as multikinase inhibitors (eg, sunitinib, sorafenib, pazopanib), antivascular endothelial growth factor agents (eg, bevacizumab), and mammalian targets of rapamycin inhibitors (eg, temsirolimus and everolimus). According to the National Institute for Health and Clinical Excellence (NICE) technology appraisal guidance 178, the assessment group concluded that for patients eligible for immunotherapy, bevacizumab plus interferon-␣ (IFN-␣) appears to be superior to IFN-␣ alone as a first-line treatment, as measured by progressionfree survival (PFS) and tumor response. For patients with poor prognosis, temsirolimus appears to be superior to IFN-␣ in terms of overall survival (OS), PFS, and tumor response rate. However, for patients who do not respond to immunotherapy, sorafenib and sunitinib are likely to demonstrate clinically and statistically significant benefits compared with best supportive care in terms of PFS and tumor response rate.5 New multikinase inhibitors have now been compared with IFN-␣ and placebo6-9 as treatments for mRCC, although they have not been evaluated directly in headto-head trials (Table I). Sunitinib, but not sorafenib, has been reported to have a longer PFS when compared with IFN-␣ in Phase III trials (Table II).7,8 In addition, phase III trials of pazopanib have reported significant improvements in PFS and tumor response compared with results using placebo (Table II).9 The limited amount of information on the clinical efficacy, cost, patient experience, and adverse events (AEs) associated with these agents will likely play an important role both in decisions related to reimbursement by the Taiwan national health insurance program and in the availability of these new drugs to hospitals. To determine the optimal treatment for a rapidly developed targeted therapy, we conducted a meta-analysis of all randomized controlled trials (RCTs) of multikinase inhibitors used to treat mRCC by an indirect comparison method.10
METHODS Inclusion Criteria Any randomized clinical trial evaluating the therapeutic efficacy of the multikinase inhibitors sunitinib, sorafenib, or pazopanib for the treatment of mRCC was included. All included trials evaluated 1 of the study drugs with a control intervention as the sole treatment. Conference abstracts were included if there
June 2011
were sufficient data to assess the quality of the studies or if detailed results were provided. Nonrandomized trials, pharmacokinetic studies, animal studies, and laboratory studies were excluded from this study. Only studies with institutional review board approval and appropriate consent were included.
Search Strategy A systematic literature search of MEDLINE, EMBASE, CANCERLIT and Cochrane databases was performed. The strategy consisted of searching for publications using key terms related to the target drugs (eg, trade names) and various terms used to describe renal cancer and neoplasms. The initial searches combined these terms with the heading neoplasm from the cancer Medical Subject Headings [MeSH] and looked for publications from January 2008 to March 2009. A subsequent search was run in March 2010 to identify any relevant publications added to the databases since March 2009. The searches were limited to original publications in English that described RCTs with human subjects. Bibliographies of included studies were searched manually to identify additional relevant studies. Individual conference abstracts from the American Society of Clinical Oncology (ASCO) were also searched. Specifically, we searched the ASCO website for abstracts from the 2007 and 2009 Annual Meeting and the 2007 and 2009 Genitourinary Cancers Symposium.
Data Extraction Two reviewers independently examined the titles and abstracts of the identified studies and the full texts of all potentially relevant studies. Data extraction was conducted using a standardized form. Two reviewers initially screened the abstracts to exclude obviously ineligible studies and reviewed the full texts of all remaining studies. The following data were collected into the standardized form: patient characteristics, study design, interventions, and direct outcomes. The quality and content of each study were assessed independently and appraised according to the inclusion and exclusion criteria of the Jadad scale, an RCT appraisal tool.11,12 Data were extracted independently by 2 reviewers with all disagreements resolved by consensus. Details of the data on primary end point PFS, OS, median duration of treatment, duration of treatment response, and AEs were recorded, if available, on a standardized form.
709
Study
Intervention and Comparator
Patients’ Ages (Median)/Gender (%)
ECOG
PS
0
1
MSKCC Prognostic Risk (%) Low/Intermediate*
Undergone Previous Nephrectomy (%)
Dosage Regimen Severe AEs (ⱖGrade 3 or Case Death) (%)
Dose (mg/kg), Duration
Sorafenib: placebo ⫽ 34:24 (hospitalization or death) Sor: INF ⫽ 16.5:14.4
Sor: 400 mg BID 6-wk cycle
Design
Diagnosis
TARGET Escudier et al 20076 Escudier et al 20097
Phase III RCT 903 patients
Metastatic clear-cell RCC Metastatic clear-cell RCC
Sor vs Pl
58 vs 59 Male: 70 vs 75
52%/48%
Sor: 49 Pl: 46
Sor: 49 Pl: 52
Sor:Pl ⫽ 94: 93
Sor vs 1FN-␣-2a
62 vs 62.5 Male/female 67 vs 33/56.5 vs 43.5
Sor: 53.6/45.4 1FN: 51.1/47.8
Sor: 57.7 INF: 53.3
Sor: 42.3 INF: 46.7
Sor:INF ⫽ 97.9:90.2
SUTENT Motzer et al 20098
Phase III RCT 750 patients
Metastatic clear-cell RCC
Sun vs 1FN-␣
62 vs 59 Male/female 71 vs 29/72 vs 28
Favorable/intermediate/ poor sun group: 38%/ 56%/6% IFN group: 32%/51%/7%
NA
NA
NA
Sor: IFN ⫽ 3:1
VEG105192 Sternberg et al 20109
Phase III RCT 435 patients
Metastatic clear-cell RCC
Pa vs Pl
59 vs 60 Male: 68 vs 75
Favorable/intermediate/ poor Pa: 39/55/3 P1: 39/55/3
Pa: 42 P1: 41
Pa: 58 P1: 59
Pa: 89 P1: 88
Pa/IFN death ⫽ 4:3 grade 3 ⫽ 33:14 grade 4 ⫽ 7:6
Phase II RCT 189 patients
Sor: 400 mg BID (period 1) 600 mg BID (period 2) IFN:9 Mu 3 times weekly Sun: 50 mg once daily on a 4 wk on, 2 wk off IFN:SC thrice weekly, 3 Mu/dose the first week; 6 Mu the second week; 9 Mu thereafter Pa: 800 mg once daily
Volume 33 Number 6
AE ⫽ adverse event; ECOG ⫽ Eastern Cooperation Oncology Group; IFN-␣-2a ⫽ interferon alpha-2a; MSKCC ⫽ Memorial Sloan-Kettering Cancer Center; Mu ⫽ million unit; NA ⫽ not applicable; Pa ⫽ pazopanib; Pl ⫽ placebo; PS ⫽ performance status; RCC ⫽ renal cell carcinoma; RCT ⫽ randomized controlled trial; SC ⫽ subcutaneous; Sor ⫽ sorafenib; Sun ⫽ sunitinib. *Unless noted otherwise.
Clinical Therapeutics
710 Table I. Characteristics of patients and included studies.
H.W.C. Leung and A.L.F. Chan
2010, USA
2009, USA
2009, USA
TARGET Escudier et al6 17. Escudier et al7 SUTENT Motzer et al8 VEG105192 Sternberg et al9
HR ⫽ hazard ratio; INF-␣ ⫽ interferon-␣; NS ⫽ not significant; OS ⫽ overall survival; PFS ⫽ progression-free survival; RCC ⫽ renal cell carcinoma; RCT ⫽ randomized controlled trial.
435
–
9.2 vs 4.2
–
0.46
–
0.34–0.62
OS: NS PFS: 0.001 0.0001 0.451–0.643 0.637–1.001 0.539 0.821 11 vs 5 26.4 vs 21.8 750
– 5.7 vs 5.6 189
Not reached
0.88
–
0.61–1.27
OS: 0.02 PFS: 0.01 0.504 0.35–0.55 0.54–0.94 0.44 –0.72 5.5 vs 2.8 903
Sorafenib vs placebo Sorafenib vs IFN-␣ Sunitinib vs IFN-␣ Pazopanib vs placebo 2007, USA
19.3 vs 15.9
PFS OS PFS OS N
June 2011
Study/Authors
Year of Intervention Publication, and Country Comparator
Table II. Direct comparisons of included studies.
Outcome Measured (median months)
OS
HR
PFS
95% CI for HR
P
Data Analysis The hazard ratios (HRs) and appropriate 95% CIs of survival data (OS and PFS) were extracted from the original studies. The HR describes the relative risk complications based on a comparison of event rates; it has also been used to describe the outcome of therapeutic trials as a measure of the extent to which a treatment shortens the duration of the illness. It is a constant under the Cox proportional hazard model. The P value is used to reject the null hypothesis that HR ⫽ 1, that is, that treatment is not beneficial. Therefore, an HR value of ⬍1 favors the drug being evaluated.13 Studies were pooled if more than 1 trial of the studied drug existed. Studies for each drug were pooled using a random effects model.14 Owing to the lack of head-to-head comparison that RCTs study between the interventions, we performed indirect comparisons of the interventions with the same comparators across similar patient groups using an indirect treatment comparison method, as described by Bucher et al.10 In this method, the randomization of each trial was maintained and the summary estimates of pooled interventions were compared with 95% CIs. SAS (version 9.0, Cary, North Carolina) was used for all data analysis.
RESULTS Quantity of Studies The primary literature search identified 101 abstracts. We excluded 80 papers because they were not RCTs or original studies or did not examine at least 1 of our target drugs. A total of 21 full-length articles were reviewed. After reviewing the full text, 6 articles matched our inclusion criteria.6-9,15 Two of 6 relevant articles were excluded because their results were based on the same patients and the same trial.8,15 Eight abstracts from relevant conferences and annual meetings were reviewed, but no new abstracts met our inclusion criteria. The flow diagram of the 4 included studies is depicted in Figure 1.
Characteristics of Patients All patients enrolled in the included studies were ⱖ18 years, were diagnosed with metastatic clear-cell RCC, had a life expectancy ⱖ12 weeks, and had an Eastern Cooperation Oncology Group (ECOG) performance status of 0 or 1. The majority of patients had undergone a nephrectomy or had received a cytokinebased treatment. Among those in the included studies, patients treated with sorafenib, sunitinib, or pazo-
711
Clinical Therapeutics
101 abstracts 80 were excluded: • Not an RCT • Not an original article • Did not examine targeted drugs 21 full-length articles 15 were excluded: • Not an RCT • Not an original study • Did not examine targeted drugs
2 were excluded: • Same population, same trial
4 full-length articles included
Figure 1. The flow diagram of included studies. RCT ⫽ randomized controlled trial.
panib were stratified as having either intermediate-risk or low-risk status and in other studies as having favorable-, intermediate-, or poor-risk status, according to the Memorial Sloan-Kettering Cancer Center prognostic score.17-19 From the AEs reported in these trials, the frequency of serious AEs leading to hospitalization or death may be higher in the sorafenib group than in the placebo (34% vs 24%) or IFN-␣ (16.5% vs 14.4%) groups.6,7 In the sunitinib trial, the proportion of grade 3 or 4 AEs with sunitinib ranged from 1% to 16% for all categories. The most commonly reported sunitinibrelated grade 3 AEs included hypertension (12%), fatigue (11%), diarrhea (9%), and hand-foot syndrome (9%). No grade 4 AEs were reported during treatment.8 During treatment with pazopanib, AEs resulting in death were reported in 4% of patients in the pazopanib arm and in 3% of patients in the placebo arm. The proportions of patients experiencing an AE with grade 3 or 4 were 33% and 7%, respectively, in the pazopanib arm and 14% and 6%, respectively, in the placebo arm. The most common grade 3 or 4 AEs in the pazopanib arm were hypertension (4%) and diarrhea (4%) (Table I).9 Patients with brain metastases, previous malignancy, or a history of severe cardiac disease were excluded from all studies. All studies explained the inclusion and exclusion criteria that were utilized.
712
Characteristics of Included Studies All included studies were RCTs; 3 studies were Phase III, international, multicenter, randomized clinical trials; and 1 study was a Phase II trial. Two studies compared the effects of sorafenib (400 mg– 600 mg twice daily) or pazopanib (80 mg once daily) on PFS with the placebo.6,9 Another 2 studies compared the effects of sorafenib or sunitinib (50 mg once daily taken orally in 6-week cycles with 4 weeks on and 2 weeks off) with those of IFN-alfa-2a (median daily dose of 8.59 million U) on PFS.7,8 The characteristics of each study are provided in Table I.
Meta-analysis of Progression-free Survival Four trials were analyzed that compared the effect of at least 1 new multikinase inhibitor (sorafenib, sunitinib, or pazopanib) with IFN-␣ or placebo on PFS. One of the trials compared the effects of pazopanib and placebo on PFS in 2 patient subpopulations that were either treatment-naïve or cytokine-pretreated. We pooled 2 trials that compared sorafenib or sunitinib with IFN-␣ and found a pooled HR of 0.47 (95% CI, 0.316 – 0.713; P ⬍ 0.001). Two trials that compared sorafenib and pazopanib with placebo were also pooled, and a pooled HR of 0.957 was found (95% CI, 0.657– 1.393; P ⫽ 0.24). Figure 2 and Figure 3 illustrate the
Volume 33 Number 6
H.W.C. Leung and A.L.F. Chan
Hazard Ratio
SE
Weight
Hazard Ratio IV, Random, 95% CI
Escudier 20097
0.88
0.51
20.2%
0.88 [–0.12, 1.88]
Motzer 200716
0.42
0.64
12.8%
0.42 [–0.83, 1.67]
0.54
0.28
67.0%
0.54 [–0.01, 1.09]
100.0%
0.59 [0.14, 1.04]
Study or Subgroup
Motzer
20098
Total (95% CI)
Heterogeneity: τ² = 0.00; X² = 0.43, df = 2 (P = 0.81); I² = 0% Test for overall effect: Z = 2.59 (P = 0.010)
Hazard Ratio IV, Random, 95% CI
–4 –2 Favors control
0 2 4 Favors experimental drug
Figure 2. Forest plot of 3 studies comparing sorafenib to interferon-␣ (control). SE ⫽ standard error.
intervention compared with IFN-␣ or placebo on PFS outcomes. We did not pool study outcomes for OS because completed data were not available in all studies.
Indirect Comparisons of Multikinase Inhibitors Versus IFN-␣ and Placebo On Progression-free Survival We performed an indirect comparison of the effects of multikinase inhibitors on PFS. The results of the indirect comparison of sunitinib versus sorafenib on PFS suggested that sunitinib may have significant benefits compared with sorafenib (HR ⫽ 0.47; 95% CI, 0.316 – 0.713; P ⬍ 0.001) using IFN-␣ as the comparator. We did not observe a significant difference between sorafenib and pazopanib in terms of PFS (HR ⫽ 0.957; 95% CI, 0.657– 1.393; P ⫽ 0.24) using placebo as a comparator (Table III).6,9
Study Quality The results of the quality appraisal were analyzed using the Jadad scale.11,12 Two studies scored a 5 because
Study or Subgroup Escudier 20076 Escudier
20097
Sternberg 20109 Total (95% CI)
Hazard Ratio
SE
0.44
0.28
0
0
0.46
0.32
Weight
the description of randomization and technique was adequate.6,9 By contrast, the other 2 studies scored a 3 on the Jadad scale because the description of double-blind or the method of blinding was inappropriate.7,8 The populations in all studies were similar and predominantly included patients with clear-cell mRCC (Table IV).
DISCUSSION The results of the meta-analysis of 3 multikinase inhibitors (sorafenib, sunitinib, and pazopanib) suggest that treatment with sorafenib or sunitinib is likely to have clinical advantages over treatment with IFN-␣ alone. The results also suggest that treatment with sorafenib or pazopanib is likely to have a greater treatment effect than placebo when the effect on PFS is considered. In trials using IFN-␣ as a comparator, the median PFS was significantly prolonged following treatment with sunitinib rather than sorafenib, (5–11 months vs 5.6 – 5.7 months, respectively). When placebo was used as a comparator, the median PFS with pazopanib was slightly longer than that with sorafenib (9.2 vs 4.2
Hazard Ratio IV, Random, 95% CI
56.6%
0.44 [–0.11, 0.99]
43.4%
0.46 [–0.17, 1.09]
100.0%
0.45 [0.04, 0.86]
Hazard Ratio IV, Random, 95% CI
Not estimable
Heterogeneity: τ² = 0.00; X² = 0.00, df = 1 (P = 0.96); I² = 0% Test for overall effect: Z = 2.13 (P = 0.03)
–4 –2 Favors control
0 2 4 Favors experimental drug
Figure 3. Forest plot of 3 studies comparing sorafenib and pazopanib to placebo (control). SE ⫽ standard error.
June 2011
713
Clinical Therapeutics
Table III. Indirect comparison of studies. Study
Intervention
TARGET vs VEG105192 Escudier et al vs SUTENT
Sorafenib vs pazopanib Sunitinib vs sorafenib
Direct HR (PFS)
Indirect HR (PFS)
95% CI for HR
0.44 vs 0.46
0.35–0.55 vs 0.34–0.62
0.539 vs 0.88
0.451–0.643 vs 0.61–1.27
95% CI for HR
0.957
0.657–1.393
0.47
0.316–0.713
CI ⫽ confidence interval; HR ⫽ hazard ratio; PFS ⫽ progression-free survival.
months and 5.5 vs 2.8 months, respectively). The reason for the slightly lower value of PFS for sorafenib in the trials was not clear. The discrepancy could be the result of small differences in patient characteristics or differences in the PFS assessment. We did not perform an indirect comparison of the effect of interventions on OS data because there was a lack of final OS data reported in the studies analyzed and because of uncertainty regarding poststudy medication usage. We sought all RCTs of targeted multikinase inhibitors for mRCC. A variety of factors reportedly influence prognosis of mRCC, including the following: patient access (whether patients will be able to receive several lines of therapy), high level of lactate dehydrogenase (⬎1.5 times upper limit of normal), low serum
hemoglobin (⬍ lower limit of normal), high corrected serum calcium (⬎10 mg/dL), alkaline phosphatase, time from initial diagnosis, tumor stage (nodes and sites of metastasis), number of organs involved, absence of prior nephrectomy, and scores of performance, such as ECOG status or Karnofsky score.17,20,21 In this review, analysis of the populations’ baseline characteristics in the included studies revealed that the age, sex, and ECOG performance status were similar among all patients. However, according to Memorial Sloan-Kettering Cancer Center criteria for mRCC prognosis, patient stratification differed among studies. Patients were classified as either favorable/low or intermediate risk in some studies, whereas other studies stratified patients as having favorable/low, intermediate, or poor risk. In addition, refrac-
Table IV. Quality assessments for each study with Jadad Scale. Questions
Answer
Score
Escudier et al6
Escudier et al7
Motzer et al8
Sternburg et al9
Randomized Study? Randomization technique described and adequate Randomization technique described and inadequate
Yes/No
⫹1/0
Yes
Yes
Yes
Yes
Yes
⫹1
Yes
Yes
Yes
Yes
Yes
–1
No
No
No
No
Yes/No
⫹1/0
Yes
No
No
Yes
Yes
⫹1
Yes
No
No
Yes
Yes
–1
No
No
No
No
Yes/No
⫹1/0
Yes 5
Yes 3
Yes 3
Yes 5
Double blinded study? Technique of blinding described and adequate Technique of blinding described and inadequate Description of withdrawals and dropouts? Result of Jaded scale
714
Volume 33 Number 6
H.W.C. Leung and A.L.F. Chan tory patients were identified as low or intermediate risk. Because we hoped to estimate the overall treatment effect of targeted drugs, we included all groups of treatment classifications in our meta-analysis. This might have resulted in a selection bias that could have altered the outcome. An additional source of bias was the subjective and intermittent assessment of progression in the included studies. Two trials used an independent blinded central review to evaluate the primary PFS in the first assessment but did not in the second assessment.7,8 Moreover, differing opinion on diagnoses may exist between investigators and independent central reviewers. Finally, 1 trial did not clearly and appropriately mention whether it was double-blinded. In 2 trials comparing sorafenib or pazopanib with placebo, fatal AEs were reported in 10% of patients in the sorafenib arm versus 6% in the placebo arm and in 4% of patients in the pazopanib arm versus 3% in the placebo. From these data, the safety profile of sorafenib is likely to be worse than that of pazopanib and sunitinib, which did not report any fatal AEs in their trial. The relative efficacy of different multikinase inhibitors has not been established because no head-to-head comparisons have been reported. We therefore performed an indirect comparison analysis to evaluate the effectiveness of these agents, which suggested that sunitinib may have comparable clinical benefits to sorafenib as a first-line treatment for mRCC patients at preventing progression. Findings here are consistent with a recently published systematic review, which included an adjusted comparison of the effects of treatment with sunitinib and treatment with sorafenib on PFS for mRCC.22 This study reported sunitinib to be more effective than sorafenib (HR ⫽ 0.58; 95% CI, 0.38 – 0.86; P ⱕ 0.001). By contrast, our results were unable to demonstrate significant differences between sorafenib and pazopanib on PFS (HR⫽ 0.957; 95% CI, 0.657–1.393; P ⫽ 0.24). A limitation of this review is that our study included only RCTs; therefore, it is possible that potentially useful trials of cohort or case-controlled design were not identified. Our findings should be interpreted with caution, as only 4 studies were analyzed, and, among these, only 2 indirect comparisons were made. In addition, this small selection of studies was not adequate for the exploration of patient subgroup analyses.
June 2011
CONCLUSION Compared with IFN-␣ or placebo, some multikinase inhibitors have reported favorable PFS for patients with mRCC. Our findings suggest that sunitinib might offer some clinical benefit over sorafenib in terms of PFS. No statistical difference was found between sorafenib and pazopanib treatments. However, these conclusions are based on 2 indirect comparisons of single randomized clinical trials. More such trials are required to confirm these findings and investigate the clinical effectiveness of multikinase inhibitors in the treatment of mRCC.
ACKNOWLEDGMENT This was an unfunded study. The authors have indicated that they have no conflicts of interest regarding the content of this article. The authors would like to thank the referees and editors for their useful comments. All authors contributed equally to the conduct of the study and creation of the manuscript.
REFERENCES 1. National Cancer Institute. SEER Stat Fact Sheets: Kidney and Renal Pelvis. http://seer.cancer.gov/statfacts/html/ kidrp.html. Accessed August 2010. 2. National Institute for Health and Clinical Excellence. Sunitinib for the first-line treatment of advanced and/or metastatic renal cell carcinoma. NICE technology appraisal guidance TA 169. March 2009. http://guidance.nice. org.uk/TA169. Accessed May 9, 2011. 3. Ljungberg B, Hanbury DC, Kuczyk MA, et al. Renal cell carcinoma guideline. Eur Urol. 2007;51:1502–1510. 4. Taiwan Cancer Registry. http://www.mc.ntu.edu.tw/CRS/ main.php?Page⫽A5. Accessed May 9, 2011. 5. National Institute for Health and Clinical Excellence. Bevacizumab (first-line), sorafenib (first- and second-line), sunitinib (second-line) and temsirolimus (first-line) for the treatment of advanced and/or metastatic renal cell carcinoma. NICE technology appraisal guidance 178. August 2009. http://www.nice.org. uk/nicemedia/pdf/TA178Guidance1.pdf.AccessedMay6,2011. 6. Escudier B, Eisen T, Stadler WM, et al, for the TARGET Study Group. Sorafenib in advanced clear-cell renal-cell carcinoma. N Engl J Med. 2007;356:125–134. 7. Escudier B, Szczylik C, Hutson TE, et al. Randomized phase II trial of first-line treatment with sorafenib versus interferon Alfa-2a in patients with metastatic renal cell carcinoma. J Clin Oncol. 2009;27:1280 –1289. 8. Motzer RJ, Hutson TE, Tomczak P, et al. Overall survival and updated results for sunitinib compared with interferon alfa in patients with metastatic renal cell carcinoma. J Clin Oncol. 2009;27:3584 –3590.
715
Clinical Therapeutics 9. Sternberg CN, Davis ID, Mardiak J, et al. Pazopanib in locally advanced or metastatic renal cell carcinoma: results of a randomized phase III trial. J Clin Oncol. 2010;28:1061–1068. 10. Bucher HC, Guyatt GH, Griffith LE, Walter SD. The results of direct and indirect treatment comparisons in meta-analysis of randomized controlled trials. J Clin Epidemiol. 1997; 50:683– 691. 11. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials. 1996;17:1–12. 12. Schwenk W, Haase O, Günther N, Neudecker J. Methodological quality of randomised controlled trials comparing short-term results of laparoscopic and conventional colorectal resection. Int J Colorectal Dis. 2007;22:1369 –1376. 13. Spruance SL, Reid JE, Grace M, Samore M. Hazard ratio in clinical trials. Antimicrob Agents Chemother. 2004;48:2787–2792. 14. Fleiss JL. The statistical basis of meta-analysis. Stats Methods Med Res. 1993;2:121–145. 15. Escudier B, Eisen T, Stadler WM, et al. Sorafenib for treatment of renal cell carcinoma: final efficacy and safety results of the phase III treatment approaches in renal cancer global evaluation trial. J Clin Oncol. 2009;27:3312–3318. 16. Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa in metastatic renal-cell carcinoma. N Engl J Med. 2007;356:115–124. 17. Motzer RJ, Mazumdar M, Bacik J, et al. Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol. 1999;17:2530 –2540. 18. Motzer RJ, Bacik J, Murphy BA, et al. Interferon-alfa as a comparative treatment for clinical trials of new therapies against advanced renal cell carcinoma. J Clin Oncol. 2002;20: 289 –296.
716
19. Memorial Sloan-Kettering Cancer Center. Renal cell carcinoma nomogram. http://www.mskcc.org/mskcc/ html/6156.cfm. Accessed May 9, 2011. 20. Mills EJ, Rachlis B, O’Regan C, et al. Metastatic renal cell cancer treatments: an indirect comparison meta-analysis. BMC Cancer. 2009;9:34.
21. Escudier B, Albiges L, Blesius A, et al. How to select targeted therapy in renal cell cancer. Ann Oncol. 2010;21(Suppl 7):vii59 –vii62. 22. Leibovich BC, Cheville JC, Lohse CM, et al. A scoring algorithm to predict survival for patients with metastatic clear cell renal cell carcinoma: a stratification tool for prospective clinical trials. J Urol. 2005;174:1759–1763.
Address correspondence to: Agnes L.F. Chan, PhD, Director of Pharmacy, Chi Mei Medical Center, 901, Cheung Hwa Road, Tainan, Taiwan. E-mail: [email protected]
Volume 33 Number 6