- © 2009 by American Society of Clinical Oncology
Phase III Trial of Cisplatin Plus Gemcitabine With Either Placebo or Bevacizumab As First-Line Therapy for Nonsquamous Non–Small-Cell Lung Cancer: AVAiL
- Martin Reck,
- Joachim von Pawel,
- Petr Zatloukal,
- Rodryg Ramlau,
- Vera Gorbounova,
- Vera Hirsh,
- Natasha Leighl,
- Jörg Mezger,
- Venice Archer,
- Nicola Moore and
- Christian Manegold
- From the Krankenhaus Grosshansdorf, Grosshansdorf; Asklepios Fachkliniken München-Gauting, Muenchen-Gauting; St. Vincentius-Kliniken, Karlsruhe; Heidelberg University Medical Center, Mannheim, Germany; Third Faculty of Medicine, Charles University, Postgraduate Medical School, Prague, Czech Republic; Wielkopolskie Centrum Chorob Pluc i Gruzlicy, Poznan, Poland; Cancer Research Center, Moscow, Russia; McGill University Health Centre–Royal Victoria Hospital, Montreal, Quebec; Princess Margaret Hospital, Toronto, Ontario, Canada; and F. Hoffmann-La Roche Ltd, Basel, Switzerland.
- Corresponding author: Martin Reck, MD, Department of Thoracic Oncology, Hospital Grosshansdorf, Wohrendamm 80, 22927 Grosshansdorf, Germany; e-mail: dr.martin.reck{at}web.de.
-
Presented in part at the 43rd Annual Meeting of the American Society of Clinical Oncology, June 1-5, 2007, Chicago, IL, and at the 12th World Conference on Lung Cancer, September 2-6, 2007, Seoul, Korea.
Abstract
Purpose Bevacizumab, a monoclonal antibody targeting vascular endothelial growth factor, improves survival when combined with carboplatin/paclitaxel for advanced nonsquamous non–small-cell lung cancer (NSCLC). This randomized phase III trial investigated the efficacy and safety of cisplatin/gemcitabine (CG) plus bevacizumab in this setting.
Patients and Methods Patients were randomly assigned to receive cisplatin 80 mg/m2 and gemcitabine 1,250 mg/m2 for up to six cycles plus low-dose bevacizumab (7.5 mg/kg), high-dose bevacizumab (15 mg/kg), or placebo every 3 weeks until disease progression. The trial was not powered to compare the two doses directly. The primary end point was amended from overall survival (OS) to progression-free survival (PFS). Between February 2005 and August 2006, 1,043 patients were randomly assigned (placebo, n = 347; low dose, n = 345; high dose, n = 351).
Results PFS was significantly prolonged; the hazard ratios for PFS were 0.75 (median PFS, 6.7 v 6.1 months for placebo; P = .003) in the low-dose group and 0.82 (median PFS, 6.5 v 6.1 months for placebo; P = .03) in the high-dose group compared with placebo. Objective response rates were 20.1%, 34.1%, and 30.4% for placebo, low-dose bevacizumab, and high-dose bevacizumab plus CG, respectively. Duration of follow-up was not sufficient for OS analysis. Incidence of grade 3 or greater adverse events was similar across arms. Grade ≥ 3 pulmonary hemorrhage rates were ≤ 1.5% for all arms despite 9% of patients receiving therapeutic anticoagulation.
Conclusion Combining bevacizumab (7.5 or 15 mg/kg) with CG significantly improved PFS and objective response rate. Bevacizumab plus platinum-based chemotherapy offers clinical benefit for bevacizumab-eligible patients with advanced NSCLC.
INTRODUCTION
Vascular endothelial growth factor (VEGF) promotes tumor angiogenesis, which is critical for tumor progression.1–4 In non–small-cell lung cancer (NSCLC), increased VEGF expression is common and associated with adverse clinical outcomes. Bevacizumab, a humanized monoclonal anti-VEGF antibody,5 has demonstrated significant clinical benefit in first- and second-line colorectal cancer and in first-line treatment of nonsquamous NSCLC, metastatic breast cancer, and renal cell cancer.6–10
In a randomized, three-arm, phase II trial in advanced NSCLC comparing carboplatin and paclitaxel (CP) alone versus CP in combination with bevacizumab (7.5 or 15 mg/kg every 3 weeks), improved time to progression was observed in the bevacizumab 15 mg/kg arm,11 leading to the selection of the 15 mg/kg dose by the Eastern Cooperative Oncology Group (ECOG) for the phase III NSCLC trial E4599.8 In the phase II study,11 pulmonary hemorrhage was seen in four of 13 patients with predominantly squamous cell histology receiving bevacizumab, prompting the exclusion of squamous cell histology from most NSCLC trials with bevacizumab. E4599 demonstrated a substantial clinical benefit for nonsquamous NSCLC patients treated with bevacizumab plus CP, with a hazard ratio (HR) of 0.66 for progression-free survival (PFS) and 0.79 for overall survival (OS) versus chemotherapy alone.8 Safety was acceptable, and severe (grade 3 to 5) pulmonary hemorrhage occurred in eight patients (1.9%) receiving bevacizumab.
This trial evaluates bevacizumab in combination with cisplatin/gemcitabine (CG), a commonly used and efficacious regimen in NSCLC. No large trials to date have assessed safety and efficacy of bevacizumab in combination with CG. Therefore, it was decided to re-evaluate both doses of bevacizumab versus placebo, in combination with this chemotherapy doublet.
PATIENTS AND METHODS
Patients
Eligible patients had histologically or cytologically documented, advanced (stage IIIB, with supraclavicular lymph node metastasis or malignant pleural or pericardial effusion, or stage IV) or recurrent nonsquamous NSCLC. Other inclusion criteria included age ≥ 18 years; ECOG performance status of 0 to 1; and adequate hematologic, hepatic, and renal function.
Exclusion criteria included mixed non–small-cell and small-cell tumors or mixed adenosquamous carcinomas with a predominant squamous component; history of grade ≥ 2 hemoptysis (≥ one-half teaspoon of bright red blood per event); CNS metastases; history of thrombotic or hemorrhagic disorders; therapeutic anticoagulation at enrollment (full-dose therapeutic anticoagulation was allowed on study for treatment of venous thrombosis); current and recent use of aspirin (> 325 mg/day); clinically significant cardiovascular disease; medically uncontrolled hypertension; nonhealing wound, ulcer, or bone fracture; palliative radiotherapy for bone lesions outside the thorax within 2 weeks before treatment; surgery within 4 weeks before treatment; and pregnancy or lactation. Patients with tumors invading or abutting major blood vessels (based on radiologist assessment) were also excluded. Although guidance was provided in the trial protocol, tumor location determination relied on the expertise of the investigators and radiologists when interpreting scans. The protocol was approved by local independent ethics committees and conducted in accordance with the Declaration of Helsinki and Guideline for Good Clinical Practice, an International Conference on Harmonization Tripartite Guideline.
Study Design
Eligible patients were assigned by a centralized stratified random assignment procedure to CG with bevacizumab (7.5 or 15 mg/kg) or CG with high- or low-dose placebo (Fig 1). Because only one size of bevacizumab vial was available, to maintain the blind to trial treatment, the placebo arm was split into high dose and low dose; centers were blinded to bevacizumab versus placebo but not to dose assignment. Patients assigned to high- and low-dose placebo were pooled into one placebo group for all analyses.
Study schema. CG, cisplatin plus gemcitabine; NSCLC, non–small-cell lung cancer; PD, progressive disease.
Stratification factors included region, stage (IIIB, IV, or recurrent), ECOG performance status (0 or 1), and sex. Cisplatin was administered intravenously at 80 mg/m2 on day 1, and gemcitabine was administered intravenously at 1,250 mg/m2 on days 1 and 8. Chemotherapy was repeated every 3 weeks for up to six cycles unless there was evidence of disease progression or unacceptable toxicity. Blinded placebo or bevacizumab was administered intravenously and concurrently with chemotherapy every 3 weeks on day 1.
The blind to treatment allocation was maintained for all patients who withdrew from trial treatment during chemotherapy plus bevacizumab/placebo as a result of progressive disease. For patients who completed six cycles of chemotherapy plus bevacizumab/placebo, maintenance of the study blind varied because of a protocol amendment. Before the amendment, patients assigned to the bevacizumab groups were permitted to continue with single-agent bevacizumab at the randomized dose level. Patients assigned to placebo were observed with no further treatment until progression. The protocol amendment eliminated the optional unblinding after completion of cycle 6 of treatment. Patients enrolled after this amendment continued to receive blinded single-agent bevacizumab or placebo every 3 weeks until evidence of disease progression or unacceptable toxicity (Fig 2). Cross over from placebo to the bevacizumab arms was not permitted at any time. Bevacizumab and placebo were supplied by F. Hoffmann-La Roche Ltd (Basel, Switzerland).
Summary of patient disposition (as randomly assigned). Trial treatment (TT) included bevacizumab or placebo plus cisplatin and gemcitabine (CG). (*) Completed all components of TT as planned in the protocol. (†) Patients who prematurely withdrew from chemotherapy could enter the single-agent bevacizumab or placebo/observation phase of the study. (‡) Placebo, n = 19; observation, n = 109.
Safety and Efficacy Assessments
After baseline evaluation, tumor status was assessed every three cycles (approximately every 9 weeks) during trial treatment. If patients withdrew from trial treatment for reasons other than progressive disease, follow-up tumor assessments were repeated every 2 months until disease progression. Tumor responses were assessed by the investigator according to Response Evaluation Criteria in Solid Tumors.12
An independent Data and Safety Monitoring Board was responsible for ongoing review of unblinded safety data. Adverse events (AEs) were graded using the National Cancer Institute Common Toxicity Criteria for Adverse Events version 3.0 and coded according to the Medical Dictionary for Regulatory Activities. Laboratory data were also obtained and provided to the Data and Safety Monitoring Board.
Statistical Analysis
The primary end point was unstratified PFS, which was defined as time from random assignment to first documented disease progression (by Response Evaluation Criteria in Solid Tumors) or death, whichever occurred first. Secondary end points included objective response rate, duration of response, and OS. Patients who discontinued study therapy for reasons other than progression or death were still assessed for progression. Patients without documented progression or death were censored at the last tumor assessment. The primary analysis population for efficacy is the intent-to-treat population, which was defined as all patients who were randomly assigned.
The study was originally initiated with a two-stage adaptive design and a primary end point of OS. After stage 1, one of the bevacizumab groups was to be dropped based on interim analysis of safety and efficacy (PFS), resulting in a two-arm study comparing CG plus placebo with the selected bevacizumab dose plus CG. After the positive OS results of E4599,13 the study design was amended to change the primary end point from OS to PFS. This decision was made to ensure completion of the trial, to accelerate reporting of the efficacy data and thus expedite the availability of a potentially active treatment option for patients, and to avoid the risk of the OS end point being confounded by the increasing use of second-line therapies (including nonprotocol cross over of patients in the placebo arm to bevacizumab). In addition, a single-stage design was adopted, comparing CG plus two doses of bevacizumab with CG plus placebo, although the trial was not powered to allow a direct comparison of the two bevacizumab-containing arms. An interim analysis was performed to assess safety only.
On the basis of the PFS benefit observed in E4599 (HR = 0.7; median PFS, 4.5 months with placebo v 6.4 months with bevacizumab), the final PFS analysis was planned after 430 events or when median follow-up time for all patients was at least 7 months, whichever occurred latest. All calculations were performed using the log-rank test and involved two-sided P values with an α = .05 and 80% power. All analyses reported here were performed from a data cutoff on October 7, 2006, at which time 665 PFS events and 356 deaths had been observed. On the basis of the survival benefit observed in E4599 (HR = 0.8; median OS, 10.2 months with placebo v 12.8 months with bevacizumab),13 an OS analysis is planned to be performed after 709 deaths, which is sufficient to yield 80% power for a two-sided log-rank test at an α = .05.
Event-time distributions were estimated using the Kaplan-Meier method. Multiplicity was adjusted for using the closed test procedure performing a log-rank test of equality over strata before performing the pair-wise PFS comparisons. Other planned analyses included PFS censoring for other antineoplastic therapy administered before progression, response rates, and duration of response. For the comparison of objective response rates, a χ2 test with Schouten correction was performed.
RESULTS
Patient Characteristics
Between February 2005 and August 2006, 1,043 patients were randomly assigned at 150 centers in 20 countries across Europe, Eastern Asia, Australia, Central and South America, and Canada. A total of 347 patients were assigned to placebo plus CG, 345 were assigned to low-dose bevacizumab plus CG, and 351 were assigned to high-dose bevacizumab plus CG. Of the randomly assigned patients, 57 received no study therapy as a result of eligibility violations (n = 27), consent withdrawal (n = 16), AEs (n = 8), and other reasons (n = 6). Table 1 lists selected demographic and baseline characteristics of all randomly assigned patients, which were balanced between treatment arms.
Summary of Key Demographic Characteristics (intent-to-treat population)
Treatment
The median number of cycles of chemotherapy was five for the placebo and high-dose bevacizumab arms and six for the low-dose bevacizumab arm (Table 2). The median number of cycles of bevacizumab or placebo was five for the placebo and high-dose bevacizumab arms and six for the low-dose bevacizumab arm. Slightly higher withdrawal rates as a result of AEs were observed in the bevacizumab-containing arms (26% and 30% of patients in the low-dose and high-dose bevacizumab plus CG arms, respectively) compared with the placebo arm (23%).
Bevacizumab and Chemotherapy Exposure by Treatment Arm (safety population), NPT Exposure, and Patient Status at Clinical Cutoff (October 7, 2006; ITT population)
Figure 2 shows that 45% to 49% of patients completed six cycles of trial treatment, with 37% to 42% continuing into the single-agent or observation periods. Ninety-four percent of patients eligible to receive single-agent bevacizumab received bevacizumab maintenance therapy at cycle 7. There was increased initiation of nonprotocol antineoplastic therapy (NPT) before progression in the high-dose bevacizumab group (9%) versus the low-dose bevacizumab and placebo groups (6%).
Efficacy Analysis
PFS, the primary end point of the study, was significantly longer in both bevacizumab treatment arms compared with the placebo arm. The HR for PFS was 0.75 (P = .003) for the low-dose bevacizumab group versus placebo (median PFS, 6.7 v 6.1 months, respectively) and 0.82 (P = .03) for the high-dose bevacizumab group versus placebo (median PFS, 6.5 v 6.1 months, respectively; Fig 3). Although the treatment effect of both bevacizumab groups relative to placebo seemed to be similar, the trial was not powered to directly compare the two doses. Moreover, bevacizumab demonstrated activity across many patient subgroups (Fig 4).
Plot of Kaplan-Meier estimates for progression-free survival (intent-to-treat population) for the (A) 7.5 mg/kg bevacizumab (Bev) arm and (B) 15 mg/kg bevacizumab arm compared with placebo. CG, cisplatin plus gemcitabine; HR, hazard ratio.
Forest plots of hazard ratios for progression-free survival by subgroup for each bevacizumab arm compared with placebo. CG, cisplatin plus gemcitabine; ECOG, Eastern Cooperative Oncology Group.
The objective response rate was also significantly higher in both bevacizumab arms compared with the placebo arm; rates were 20.1%, 34.1%, and 30.4% for the placebo, low-dose bevacizumab (P < .0001), and high-dose bevacizumab (P = .0023) arms, respectively. Median durations of response for the placebo, low-dose bevacizumab, and high-dose bevacizumab arms were 4.7, 6.1, and 6.1 months, respectively.
In the Avastin in Lung (AVAiL) primary PFS analysis, patients without disease progression or death at the time of the data cutoff were censored at the time of their last tumor assessment, even if patients had received subsequent lines of therapy. A preplanned sensitivity analysis was performed to compensate for the possible confounding effects of NPT administered before disease progression. Patients without an event before the start of NPT were censored at the last tumor assessment before the start of NPT. In this sensitivity analysis, the HR was 0.68 (95% CI, 0.56 to 0.83; P = .0001) for the low-dose bevacizumab group and 0.74 (95% CI, 0.60 to 0.90; P = .002) for the high-dose bevacizumab group, each compared with placebo.
The secondary end point of OS is currently immature because of limited follow-up. Only half (n = 356) of the required 709 deaths had been observed at the time of the final PFS analysis.
Safety
All patients who were randomly assigned and had received at least one dose of trial treatment (safety population) were included in the safety analysis. Any patient receiving bevacizumab in error on one or more occasions was assigned to the respective bevacizumab arm.
The majority of AEs reported during the study were grade 1 or 2. Table 3 lists severe (≥ grade 3) AEs. The overall incidence of AEs ≥ grade 3 was similar across all arms (Table 3). The overall incidence of serious AEs was higher in the high-dose bevacizumab plus CG arm (44%) compared with the placebo plus CG and low-dose bevacizumab plus CG arms (35% in each arm). The rates of ≥ grade 3 hypertension, vomiting, neutropenia, bleeding, and proteinuria were modestly higher in the bevacizumab arms than in the placebo arm. Febrile neutropenia occurred in 2% of patients in each of the bevacizumab arms compared with 1% in the placebo arm. Although the rates of severe hypertension were higher in the high-dose arm, clinically significant events were infrequent (there were two hypertensive crises in the low-dose bevacizumab arm and one in the high-dose bevacizumab arm).
Summary of Selected Severe (grade ≥ 3) AEs, Severe AEs of Special Interest, and Pulmonary Hemorrhage Events (safety population)
Grade 3 or greater pulmonary hemorrhage events were observed in two patients in the placebo plus CG arm, five patients in the low-dose bevacizumab plus CG arm, and three patients in the high-dose bevacizumab plus CG arm (Table 3). Four of these 10 events occurred in patients with tumors classified as central in location, which was consistent with the proportion of patients with centrally located tumors in the overall study population (38%). Moreover, although the overall rate of grade 3 to 4 thrombocytopenia was 23% to 27%, only one of 10 pulmonary hemorrhage events was associated with any grade of thrombocytopenia. Nine percent of the study population received therapeutic anticoagulation with either warfarin or low molecular weight heparin after baseline; no pulmonary hemorrhage was observed in these patients.
The rate of treatment discontinuations as a result of AEs was slightly higher in the bevacizumab-containing arms (26% and 30% of patients in the low-dose and high-dose bevacizumab arms, respectively) compared with the placebo arm (23%). Rates of death as a result of AEs, regardless of causality, were 4%, 4%, and 5% of patients in the placebo, low-dose bevacizumab, and high-dose bevacizumab arms, respectively.
The most common cause of death was progressive disease. The most common causes of non–progression-related deaths were cardiopulmonary failure, respiratory failure, hemoptysis, and pulmonary embolism. One death, in the high-dose bevacizumab arm, was associated with febrile neutropenia.
DISCUSSION
The results of this large, international, randomized phase III study augment a growing body of evidence that combining bevacizumab with standard platinum-based chemotherapy provides important clinical benefits for patients with advanced nonsquamous NSCLC. This study combined bevacizumab with CG and demonstrated significant improvements in the primary end point of PFS and secondary end points of response rate and duration of response. These results are consistent with two prior randomized clinical studies evaluating bevacizumab plus CP in patients with previously untreated advanced-stage NSCLC.8,11 The results were consistent across most baseline characteristics. In this trial, bevacizumab was administered as a single agent until disease progression after completion of combined treatment with CG. Given that VEGF is a key proangiogenic mediator, continued inhibition of VEGF is essential to prevent tumor revascularization and/or neovascularization, which are required for further growth.14
Patients enrolled onto this trial had a number of favorable prognostic features when compared with other randomized trials of bevacizumab in NSCLC. The overall population was younger, with a median age of 57 to 60 years versus 63 years in E4599, and also included approximately 8% dry stage IIIB patients versus none in E4599. In addition, the eligibility criteria excluded tumors abutting or invading major blood vessels, which may have selected a more favorable population. These differences may explain the relatively high 6.1-month median PFS result in the control group. It should be noted that CG is a standard regimen in Europe based on its favorable efficacy and tolerability profile.15,16 In a phase III trial comparing four standard chemotherapy doublets, the longest time to progression was observed with CG,17 although no significant difference in OS was observed. Nevertheless, the median PFS observed in the placebo plus CG arm in this trial is in line with the results of other large randomized trials; in a recent phase III trial of erlotinib plus CG, the median PFS observed in the placebo plus CG arm was 5.4 months.18
Safety in this trial seemed similar to that observed in prior clinical studies of bevacizumab in NSCLC.8,11 AEs occurring more commonly in the bevacizumab groups included neutropenia (including one fatal event), asthenia, vomiting, hypertension, proteinuria, and bleeding. There was no observed increase in the incidence of arterial or venous thromboembolic events or GI perforation. Severe pulmonary hemorrhage was increased in the bevacizumab groups and included seven fatal events. However, the overall rate of severe pulmonary bleeding was lower than that observed in the previous phase II trial of bevacizumab plus CP11 and similar to that observed in the phase III E4599 trial8 despite the use of full-dose therapeutic anticoagulation in approximately 9% of study patients. The only bevacizumab-associated AEs that seemed to be dose related were hypertension and proteinuria. No new safety signals were observed.
The planned discontinuation of one of the bevacizumab arms, according to the original design, did not take place because of the amended design after the disclosure of the E4599 results. This provided an opportunity to evaluate the efficacy and safety of two different doses of bevacizumab. The results indicate similar efficacy (in terms of PFS and response rate) and a similar safety profile for the 7.5 and 15 mg/kg bevacizumab arms, although the trial was not powered to draw any conclusions from direct comparison of the two bevacizumab arms. When the possible confounding effects of NPT administered before documented disease progression were taken into account by censoring patients who received NPT at the last tumor assessment before starting NPT, an analysis identical to the E4599 primary PFS analysis, the HRs for the bevacizumab arms in the AVAiL trial (7.5 mg/kg: HR = 0.68; 95% CI, 0.56 to 0.83; and 15 mg/kg: HR = 0.74; 95% CI, 0.60 to 0.90) were comparable to the HR observed for the intent-to-treat population in the E4599 trial (HR = 0.66; 95% CI, 0.57 to 0.77).8
In summary, AVAiL is the second large randomized phase III trial to demonstrate that combining bevacizumab with standard of care chemotherapy significantly improves clinical outcomes for patients with advanced nonsquamous NSCLC. The positive results of the AVAiL and E4599 trials provide further evidence that bevacizumab can be effectively combined with different chemotherapy regimens. These results suggest that bevacizumab in combination with platinum-based chemotherapy offers tangible clinical benefit for bevacizumab-eligible patients with advanced nonsquamous NSCLC.
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: Venice Archer, F. Hoffmann-La Roche (C); Nicola Moore, F. Hoffmann-La Roche (C) Consultant or Advisory Role: Martin Reck, F. Hoffmann-La Roche (C), Boehringer Ingelheim (C); Rodryg Ramlau, F. Hoffmann-La Roche (C); Vera Hirsh, F. Hoffmann-La Roche (C), AstraZeneca (C), Eli Lilly & Co (C), Amgen Inc (C), Abraxis (C) Stock Ownership: Nicola Moore, F. Hoffmann-La Roche Honoraria: Martin Reck, F. Hoffmann-La Roche, Eli Lilly & Co, Boehringer Ingelheim, Bayer Healthcare; Petr Zatloukal, F. Hoffmann-La Roche, Eli Lilly & Co; Rodryg Ramlau, F. Hoffmann-La Roche; Natasha Leighl, F. Hoffmann-La Roche; Christian Manegold, F. Hoffmann-La Roche, Eli Lilly & Co Research Funding: None Expert Testimony: None Other Remuneration: None
AUTHOR CONTRIBUTIONS
Conception and design: Christian Manegold
Provision of study materials or patients: Martin Reck, Joachim von Pawel, Petr Zatloukal, Rodryg Ramlau, Vera Gorbounova, Vera Hirsh, Natasha Leighl, Jörg Mezger, Christian Manegold
Collection and assembly of data: Martin Reck, Joachim von Pawel, Petr Zatloukal, Rodryg Ramlau, Vera Hirsh, Natasha Leighl, Jörg Mezger, Venice Archer, Christian Manegold
Data analysis and interpretation: Martin Reck, Rodryg Ramlau, Jörg Mezger, Venice Archer, Nicola Moore
Manuscript writing: Martin Reck, Natasha Leighl, Jörg Mezger, Venice Archer, Nicola Moore, Christian Manegold
Final approval of manuscript: Martin Reck, Joachim von Pawel, Petr Zatloukal, Rodryg Ramlau, Vera Gorbounova, Vera Hirsh, Natasha Leighl, Jörg Mezger, Venice Archer, Nicola Moore, Christian Manegold
Acknowledgment
We thank Yfke Hager of Gardiner-Caldwell Communications for medical writing support.
Footnotes
-
Supported by F. Hoffmann-La Roche Ltd.
-
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
- Received September 20, 2007.
- Accepted October 20, 2008.
