TRIBUTE: A Phase III Trial of Erlotinib Hydrochloride (OSI-774) Combined With Carboplatin and Paclitaxel Chemotherapy in Advanced Non–Small-
http://www.100md.com
《临床肿瘤学》
The University of Texas M.D. Anderson Cancer Center, Houston, TX
University of California Los Angeles, Los Angeles
Kaiser Permanente, Oakland
Genentech Inc, South San Francisco, CA
Georgia Oncology Partners, Marietta, GA
Dana-Farber Cancer Institute, Boston, MA
Vanderbilt-Ingram Cancer Center, Nashville, TN
Memorial Sloan-Kettering Cancer Center, New York, NY, for the TRIBUTE Investigator Group
ABSTRACT
PURPOSE: Erlotinib is a potent reversible HER1/epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor with single-agent activity in patients with non–small-cell lung cancer (NSCLC). Erlotinib was combined with chemotherapy to determine if it could improve the outcome of patients with NSCLC.
PATIENTS AND METHODS: TRIBUTE randomly assigned patients with good performance status and previously untreated advanced (stage IIIB/IV) NSCLC to erlotinib 150 mg/d or placebo combined with up to six cycles of carboplatin and paclitaxel, followed by maintenance monotherapy with erlotinib. Random assignment was stratified by stage, weight loss in the previous 6 months, measurable disease, and treatment center. The primary end point was overall survival (OS). Secondary end points included time to progression (TTP), objective response (OR), and duration of response.
RESULTS: There were 1,059 assessable patients (526 erlotinib; 533 placebo). Median survival for patients treated with erlotinib was 10.6 v 10.5 months for placebo (hazard ratio, 0.99; 95% CI, 0.86 to 1.16; P = .95). There was no difference in OR or median TTP. Patients who reported never smoking (72 erlotinib; 44 placebo) experienced improved OS in the erlotinib arm (22.5 v 10.1 months for placebo), though no other prespecified factors showed an advantage in OS with erlotinib. Erlotinib and placebo arms were equivalent in adverse events (except rash and diarrhea).
CONCLUSION: Erlotinib with concurrent carboplatin and paclitaxel did not confer a survival advantage over carboplatin and paclitaxel alone in patients with previously untreated advanced NSCLC. Never smokers treated with erlotinib and chemotherapy seemed to experience an improvement in survival and will undergo further investigation in future randomized trials.
INTRODUCTION
Lung cancer is the most common cause of cancer deaths worldwide, and the development of more effective therapy remains challenging.1 The development of agents that target the epidermal growth factor receptor (EGFR) signal transduction pathways have provided a class of novel targeted therapeutic agents.2
Erlotinib (Tarceva; Genentech Inc, San Francisco, CA) is a potent reversible HER1/EGFR tyrosine kinase inhibitor (EGFR-TKI). This is a report of a phase III trial that evaluated whether concurrent administration of erlotinib with standard chemotherapy could enhance survival in chemotherapy-na?ve patients with advanced or metastatic non–small-cell lung cancer (NSCLC).
Erlotinib has been evaluated in a number of phase I and II studies as a single agent.3,4 In a phase II trial, 57 patients with advanced or recurrent NSCLC who had previously received platinum-based chemotherapy received erlotinib 150 mg/d as a single agent.5 In this study, patients had measurable tumors that expressed a low level of EGFR by immunohistochemical (IHC) analysis. The objective response rate of 12.3% (95% CI, 5.1% to 23.7%) and median survival of 8.5 months were comparable to results noted for docetaxel, pemetrexed, or gefitinib in refractory NSCLC.6,7 The most common adverse events with erlotinib consisted of mild/moderate rash and diarrhea.
Preclinical evaluations of erlotinib in athymic nude mice bearing tumor xenografts in HN5 head and neck and A431 epidermoid carcinoma cell lines established the effective dose for 50% inhibition of growth. Further evaluation in H460a and A549 NSCLC, and HN5 tumor models at the 50% inhibition of growth in combination with chemotherapy were conducted. No antagonism was detected with any of the chemotherapy agents evaluated, including paclitaxel and cisplatin. Additive effects were observed with some cytotoxic combinations of erlotinib, including paclitaxel and cisplatin.8,9
On the basis of phase II results and the therapeutic advantage demonstrated preclinically with erlotinib with platinum-based and taxane chemotherapy, a phase III trial was conducted using erlotinib in combination with paclitaxel and carboplatin.
This trial, designated TRIBUTE (Tarceva responses in conjunction with paclitaxel and carboplatin), was a multicenter (United States), randomized, placebo-controlled trial of paclitaxel and carboplatin with or without erlotinib in chemotherapy-na?ve patients with advanced NSCLC. The primary objective was to determine if the concurrent addition of erlotinib to paclitaxel plus carboplatin could prolong overall survival. Secondary objectives included improvement in time to progression (TTP), objective response rate, and safety.
PATIENTS AND METHODS
Eligibility Criteria
Inclusion required histologically documented stage IIIB or stage IV NSCLC; age 18 years; and Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1. Exclusion criteria included prior systemic chemotherapy for NSCLC; symptomatic or untreated brain metastases, prior exposure to agents directed at the HER axis; unstable systemic disease that would potentially preclude chemotherapy treatment with or without erlotinib; inadequate hematologic (granulocytes 1.5/mL), renal (creatinine > 1.5x upper limit of normal [ULN]), or hepatic function (liver function tests > 2.5x ULN). EGFR protein expression determination (by IHC methodology) was not an a priori condition for enrollment.
All patients provided informed consent, and approval was obtained from the ethics committee at each center. The study followed the Declaration of Helsinki and good clinical practice guidelines.10
Trial Design
Patients were randomly assigned to receive either daily oral erlotinib or placebo concurrently with chemotherapy (Fig 1). Randomization was stratified by disease stage (IIIB; IV), weight loss during the previous 6 months ( 5%; > 5%), tumor measurability (measurable; nonmeasurable), and treatment center. Patients received a maximum of six cycles of paclitaxel (200 mg/m2) followed by carboplatin (to achieve an area under concentration/time curve of 6 mg/mL · min) every 3 weeks in the absence of disease progression.
To evaluate the safety of erlotinib plus carboplatin/paclitaxel and to establish the starting erlotinib dose in this setting, the first 50 patients started therapy at 100 mg in the week before and during the first cycle of concurrent chemotherapy. If the study drug was tolerated, dose was escalated to 150 mg daily on day 15 of the first cycle.
Based on an acceptable safety profile, an independent Dose Selection Committee determined that a starting dose of 150 mg was tolerable. Following dose selection, approximately 1,000 patients were enrolled using this starting dose. Recommendations were provided for management of severe or intolerable rash or diarrhea (including guidelines for dose-reduction or interruption of study drug).
Due to the fact that there was no prior experience with the combination of erlotinib with carboplatin and paclitaxel in NSCLC, the first 400 patients were intensely monitored for safety to further assess for possible interaction between the chemotherapy specified for the study and concurrent administration of erlotinib. All adverse events were collected regardless of causality or relationship to study drugs. For the remaining 650 patients, collection of safety information was limited to serious adverse events, and all adverse events leading to study-drug or chemotherapy-dose interruption, reduction, or discontinuation. All adverse events were coded using the Medical Dictionary for Regulatory Activities (MedDRA), version 6.0. Severity was determined using the National Cancer Institute Common Toxicity Criteria (NCI-CTC), version 2.0.
Twenty-four patients (12 in each treatment arm) enrolled at the M.D. Anderson Cancer Center underwent pharmacokinetic sampling for erlotinib, paclitaxel, and carboplatin. These patients were enrolled after the final erlotinib dose had been selected. Sampling for OSI-774 (and its major metabolite OSI-420) occurred within 2 days of cycle 2, at 3 and 4 hours post-erlotinib. The sample schedule for day 1 of cycle 2 for erlotinib, paclitaxel, and carboplatin is presented in Table 1. Blood samples for OSI-774 and carboplatin were drawn into 2-mL heparinized tubes, while blood for paclitaxel assay was drawn into a 2-mL K-EDTA tube. All samples were centrifuged within 30 minutes of sample collection, for 10 minutes at 4°C at 3,000 x g. Plasma was extracted, frozen, and stored at –80°C.
Statistical Analysis
The primary end point was survival for the intent-to-treat population. For all patients alive at the time of the statistical analysis, survival was censored at the date of last contact. Secondary end points included TTP and response rate.
The study was designed to have 80% power for a two-sided hypothesis test that erlotinib with carboplatin/paclitaxel increased median survival by 25% relative to survival reported for carboplatin/paclitaxel alone within a randomized trial of four chemotherapy regimens. A median survival of approximately 8 months was reported in ECOG and Southwest Oncology Group trials with carboplatin/paclitaxel in a similar population.11,12 A 25% improvement in survival associated with the combination of erlotinib with this regimen would result in a median survival of 10.13 months (a hazard ratio of 0.80). Based on these assumptions, 642 events were required corresponding to enrollment of 1,050 patients. This sample size was also sufficient to detect a 33% improvement in 1-year survival, with similar statistical power.
Survival, TTP, and objective response rate were summarized by predefined categories including age, sex, ECOG performance status, weight loss during the previous 6 months, disease stage, smoking history, and selected markers on the EGFR pathway. Smoking history was collected prospectively as a data element within the screening case report forms. Patients were asked whether they had a prior smoking history, and, if so, the number of prior pack years. No minimum prior exposure (eg, 100 cigarettes) was used to define never smokers for this study.
Based on emerging results regarding efficacy in never smokers treated with EGFR tyrosine kinase inhibitors, the statistical analytic plan for the study was revised before unblinding to include such an analysis in this trial.
The Cox proportional hazards model was used to estimate the effect of risk factors on survival and TTP and to evaluate any modifications of treatment effect. To evaluate the effects of baseline characteristics on objective response rates and to assess modifications of treatment effect, a logistic regression model was applied.
An independent Data Monitoring Committee (DMC) reviewed safety on a quarterly basis and performed an interim analysis of efficacy and safety (321 deaths), in order to potentially terminate the study if the DMC had safety concerns or if robust efficacy was demonstrated.
Assessments
Tumor response was determined by Response Evaluation Criteria in Solid Tumors (RECIST).13 Tumor assessments were performed on day 21 of chemotherapy cycles 2, 4, and 6, and every 8 weeks following completion of chemotherapy. Time to symptomatic deterioration was assessed via the Lung Cancer Symptom Scale at these intervals. EGFR expression was determined retrospectively via IHC staining analysis on unstained slides, if available.14 Positive EGFR expression was indicated by weak to strong membranous staining in 10% or more tumor cells. Safety was evaluated by analyzing adverse events and laboratory tests.
RESULTS
Patients
Between July 18, 2001, and August 19, 2002, 1,079 patients were randomly assigned: 539 patients to erlotinib and 540 patients to placebo. Demographic and disease characteristics were balanced between the two treatment arms (Table 2). The majority of patients in both treatment arms had stage IV disease, with 5% weight loss in the previous 6 months.
Efficacy
Median survival was 10.6 months in the erlotinib arm versus 10.5 months in the placebo arm (hazard ratio, 0.995; 95% CI, 0.86 to 1.16; P = .95; Fig 2A). TTP was 5.1 months for erlotinib and 4.9 months for placebo (P = .36; Fig 2B).
The objective response rate (complete and partial) was similar between the erlotinib and placebo arms (21.5% v 19.3%, respectively; P = .36) and were partial responses in both arms.
Subset analyses failed to demonstrate any significant improvement in survival by age, sex, race, cancer stage, ECOG status, prior weight loss, histology, previous cancer-related surgery, or EGFR expression. There was no correlation between the level of EGFR expression (as assessed by IHC staining analysis using the DakoCytomation PharmDx test kit; DakoCytomation, Carpinteria, CA) and clinical outcome. Unstained slides were available for 445 of the 1,079 patients enrolled on the study. Of these, 344 were assessable for analysis (180 erlotinib, 164 placebo). The hazard ratio for EGFR-positive histology (93 erlotinib, 74 placebo) was 1.00 (95% CI, 0.69 to 1.45) versus a hazard ratio for EGFR-negative histology (87 erlotinib, 90 placebo) of 1.02 (95% CI, 0.71 to 1.46).
Analysis for the newly identified somatic mutations in the EGFR kinase domain was performed on 228 assessable samples of diagnostic tissue and will be reported in detail elsewhere.15,16
One prespecified subset had a difference in outcome. Patients who reported that they had never smoked experienced substantial prolongation in survival with erlotinib treatment: 22.5 months (n = 72) versus 10.1 months in the placebo arm (n = 44; hazard ratio of 0.49; 95% CI, 0.28 to 0.85; Fig 3), which was independent of tumor histology. Median overall survival of never smokers treated with chemotherapy alone was similar to that of former or current smokers in the same treatment arm (10 months). These results for never smokers contrasted sharply with the outcome for current (8.4 months erlotinib; 9.1 months placebo) and previous smokers (10.0 months erlotinib; 10.9 months placebo).
The median TTP in never smokers was 6.0 months in the erlotinib arm versus 4.3 months for placebo (hazard ratio, 0.50; 95% CI, 0.31 to 0.80). The response rate of never smokers receiving erlotinib plus carboplatin/paclitaxel was significantly higher than that of those receiving carboplatin/paclitaxel alone (21 [30%] of 69; 95% CI, 20% to 43% v five [11%] of 44; 95% CI, 4% to 25%; P = .02). The never-smoker subset tended to be younger (58 v 64 years), female (60% v 37%), and to have adenocarcinoma (82% v 58%) in comparison to prior/current smokers.
An exploratory analysis of patients who survived beyond chemotherapy and received study drug as monotherapy thereafter showed an improvement in survival in the erlotinib arm that suggested a maintenance effect.
In the 861 patients who survived beyond 4 months (erlotinib 408; placebo 453), median survival with erlotinib was 13.6 v 12.2 months (P = .04). In the 740 patients who survived beyond 6 months (erlotinib 348; placebo 392), median survival with erlotinib was 15.4 v 13.8 months for placebo (P = .007).
Duration of Therapy, Dose Adherence, Dose Intensity, and Exposure
The median duration on study drug was 4.6 months with erlotinib and 5.3 months with placebo. There were more dose interruptions (because of compliance or toxicity) and reductions with erlotinib versus placebo (250 v 190 interruptions and 135 v 17 reductions, respectively). Mean total dose of study drug per patient was lower in the erlotinib arm versus in the placebo arm (24.0 g v 26.5 g, respectively; P = .035); this was also true for the mean daily dose (137.8 mg/d v 146.8 mg/d, respectively; P < .0001). Overall, 14 patients (10 erlotinib, four placebo) were unblinded early because a potential investigational new drug safety report was filed.
Exposure to chemotherapy was similar between the two treatment arms. The median number of cycles of carboplatin and paclitaxel was five for both arms (P = .246), whereas respective mean doses per cycle for carboplatin (619.9 mg for erlotinib 635.0 mg for placebo; P = .1) and paclitaxel (358.4 for erlotinib; 362.8 mg for placebo; P = .2) were also similar.
Carboplatin and paclitaxel sampling for pharmacokinetic analysis was performed in 21 patients. Overall, exposure to carboplatin and paclitaxel was similar between erlotinib and placebo-treated patients. Erlotinib exposure was similar to historical data from phase I and II trials.3,5,17 Detailed results will be reported elsewhere.
Results of assessment for time to symptomatic progression as determined by assessment of lung cancer symptom scores will be presented elsewhere.
Safety and Tolerability
Adverse events in the erlotinib arm were similar in incidence and severity to those of the placebo arm, with the exception of an increased incidence of rash and diarrhea, which are known to be associated with EGFR tyrosine kinase inhibition (Table 3).
The erlotinib arm had a higher incidence of study drug–related serious adverse events relative to the placebo arm (8.6% erlotinib; 2.4% placebo). The most common events included diarrhea (3.8% erlotinib; 1.1% placebo) and rash (dermatitis acneiform; 0.8% erlotinib; 0 placebo). The rate of interstitial lung disease (ILD) was estimated by inclusion of a set of 24 preferred coding terms that have been associated with such events, regardless of causality or possible confounders. There were five severe ILD-like events in the erlotinib arm (1.0%), versus one event in the placebo arm (0.2%). All ILD-like events were fatal.
A total of 662 treated patients died before the data cutoff: 322 (61.2%) in the erlotinib arm and 340 (63.8%) in the placebo arm. The majority of the deaths in both arms were attributed to disease progression by investigators (275 erlotinib; 318 placebo).
Forty-eight deaths were attributed to an adverse event: 33 (10.2%) of 322 deaths in the erlotinib arm and 15 (4.4%) of 340 deaths in the placebo arm. The majority of the imbalance was due to infection (seven erlotinib; one placebo) and GI events (four erlotinib; one placebo).
With respect to infectious events leading to death, pneumonias accounted for six events (five erlotinib, one placebo). All such events occurred during concomitant administration of chemotherapy and study drug. None of the pneumonias leading to death were assessed as related to study drug by the respective investigators (events were related to "disease under study"/protocol-specified chemotherapy). The two remaining infectious events leading to death in the erlotinib arm were neutropenic sepsis (occurring after the first cycle of chemotherapy; assessed as related to erlotinib and protocol-specified chemotherapy) and septic shock (occurring on the fifth day of study; assessed by the investigator as related to the disease under study).
All of the GI events leading to death occurred during the phase of concomitant administration of chemotherapy and study drug. Of the GI events leading to death in the erlotinib arm, there was one case of GI hemorrhage assessed as related to erlotinib and concomitant medication (ie, warfarin sodium). Another erlotinib patient experienced a GI perforation on study day 17 (after one cycle of chemotherapy) and declined surgical intervention. Diarrhea led to death in a patient with a history of diabetes mellitus who had experienced nausea, vomiting, and diarrhea since the first cycle of chemotherapy, on study day 8 (assessed as related to concurrent illness and medication). Finally, intestinal ischemia was the cause of death on study day 11 and was assessed as related to the single cycle of protocol-specified chemotherapy by the treating physician. The single GI event leading to death in the placebo cohort was perforation of a gastric ulcer after five cycles of chemotherapy (assessed as related to disease under study, and concurrent medication and illness).
Twenty-one deaths were attributed to other causes: 14 (4.3%) of 322 deaths in the erlotinib arm and seven (2.1%) of 340 in the placebo arm. Nine deaths in this category were due to unknown causes (eight erlotinib; one placebo), while eight deaths were attributed to cardiac/pulmonary arrest (five erlotinib; three placebo).
There were more early deaths (ie, during concomitant administration of chemotherapy and study drug) in the erlotinib arm. At study day 120 (the point of maximum difference between treatment arms), there were 105 deaths in erlotinib arm versus 70 in placebo. The majority of early deaths in TRIBUTE were attributed to progression of underlying lung cancer (78% erlotinib, 87% placebo). This effect was also evident in the phase III trial of gefitinib with concurrent administration of the same chemotherapy regimen (INTACT-2).18
DISCUSSION
This randomized, placebo-controlled trial examined the efficacy and safety of erlotinib in combination with paclitaxel and carboplatin for first-line therapy of advanced NSCLC in 1,079 unselected patients. There was no improvement in survival, TTP, or response rate compared with chemotherapy given alone. These clinical results contrast with preclinical studies, suggesting additive effects of combining erlotinib and paclitaxel and a platinum.
These results are consistent with previous reports in which paclitaxel and carboplatin were combined with an EGFR-TKI.18 In contrast to these combination studies, erlotinib monotherapy effected a statistically significant survival advantage over placebo in 731 patients with recurrent, previously treated NSCLC. Erlotinib was associated with a 37% improvement in overall survival (log-rank hazard ratio, 0.73; 95% CI, 0.6 to 0.87; P = .001) and progression-free survival (hazard ratio, 0.61; 95% CI, 0.51 to 0.73; P < .001).19
Although a post hoc analysis of patients who had lived beyond concomitant administration of study drug with chemotherapy suggested a survival benefit associated with erlotinib, such results must be viewed with obvious caution. One has to consider that such patients who demonstrated a benefit in this analysis had completed chemotherapy with erlotinib. As such, use of erlotinib in a maintenance setting after first-line chemotherapy has not been proven and will need to be evaluated prospectively in a randomized trial setting to demonstrate its potential therapeutic benefit.
In light of the positive preclinical data, the lack of benefit in the first-line trial when erlotinib was combined with chemotherapy may be based on an as yet unidentified negative interaction when an EGFR-TKI is given concurrently with a platinum-based regimen and a taxane. However, the efficacy of erlotinib to carboplatin and paclitaxel in the never smoker population argues against antagonism. In addition, there was no apparent pharmacokinetic interaction between erlotinib and carboplatin and paclitaxel in this trial.
The patients enrolled in TRIBUTE were not selected based on known prognostic factors. Betensky et al presented a model for the potential effect of molecular heterogeneity in trial design.20 If such heterogeneity were to confer a treatment benefit in a subset of patients, the overall trial could be underpowered to detect an effective therapy. Thus, one possible explanation for the negative results may be the lack of patient selection for such a factor, at least in the first-line setting. For example, the discovery of mutations in the EGFR tyrosine kinase domain in some patients may be one example of the potential to appropriately select for a subset of patients who may benefit from EGFR inhibition in the first-line setting.
Sensitivity to erlotinib or gefitinib has been strongly associated with mutations, most commonly, deletions of four to six amino acids in exon 19 or a point mutation (L858R) in exon 21. The tissue blocks from the patients participating in TRIBUTE have been studied for EGFR mutations, and this analysis will be reported in a separate publication.
The fact that never smokers who received chemotherapy plus erlotinib achieved a 120% increase in median survival as compared with the patients who received chemotherapy alone supports this hypothesis, and argues against a universally unfavorable negative interaction with chemotherapy. This finding, however, needs to be validated in an appropriately designed clinical trial.
While the observation in never smokers might be spurious, the similar demographics and balanced baseline characteristics between the two arms of this study argues against this. In addition, similar results were also observed in the recently completed trial of single-agent erlotinib in patients with refractory advanced NSCLC conducted by the National Cancer Institute of Canada (NCIC), as well as in INTACT-2 with gefitinib.18,19 However, in the NCIC trial, a survival benefit was also observed for patients who were current/prior smokers.
Recent work by several groups has suggested that never smokers are more likely to harbor mutations than smokers in the tyrosine kinase domain of EGFR. For example, Pao et al reported such mutations in seven of 15 unselected early-stage never smokers with adenocarcinoma.21
The low response rate of never smokers treated with chemotherapy and placebo in this trial (11%) raises the question of whether patients who may have an increased likelihood of these mutations are less sensitive to cytotoxic chemotherapy. Unpublished, preclinical studies in which cell lines that harbor these mutations are treated with both EGFR-TKIs and cytotoxic chemotherapy are ongoing.21A However, the long survival observed in these patients, which is far greater than that seen in any reported phase II or III trial to date, might simultaneously suggest a favorable interaction between chemotherapy and erlotinib. In never smokers, it would be important to investigate prospectively the value of combining erlotinib with chemotherapy versus erlotinib alone.
Studies have now shown that tumors of never smokers are molecularly less complex and may confer a better prognosis for patients.22 In some countries within the Pacific Rim, lung cancer, particularly in females, is far more common in never smokers than smokers. Thus, racial background and smoking history should be considered when designing and interpreting clinical trials in NSCLC, particularly with EGFR-TKIs.
The side effects of rash and diarrhea associated with erlotinib seemed to be characteristic of this class of drugs. In addition, the incidence of life-threatening toxicities such as ILD, which have rarely been reported with gefitinib, was likewise rarely observed with erlotinib in this trial (1.0% erlotinib; 0.2% placebo).18 There was however, a higher rate of deaths (regardless of whether related to progression of NSCLC or toxicity) in the erlotinib arm relative to placebo associated with concomitant administration of chemotherapy (163 erlotinib, 125 placebo). This effect was not noted in the European counterpart trial with erlotinib, which specified a different chemotherapy regimen (gemcitabine and cisplatin). As noted previously, this is suggestive of a potential negative interaction between the carboplatin and paclitaxel regimen specified in this study and erlotinib.
Additional preclinical and clinical investigations are ongoing to identify whether alternative schedules and or dosing regimens might allow erlotinib to be combined with chemotherapy in patients with first-line stage IIIB/IV disease. In addition, the identity of clinical or molecular surrogates that may predict for improved outcome (ie, never smokers or somatic mutation in the kinase domain of EGFR) are still preliminary and will require further investigation in prospective, randomized trial settings.
Editor’s Note
A related article on this subject will be published in the November 1, 2005, issue titled, Epidermal Growth Factor Receptor Mutations and Gene Amplification in Non–Small-Cell Lung Cancer: Molecular Analysis of the IDEAL/INTACT Gefitinib Trials; by Daphne W. Bell, Thomas J. Lynch, Sara M. Haserlat, Patricia L. Harris, Ross A. Okimoto, Brian W. Brannigan, Dennis C. Sgroi, Beth Muir, Markus J. Riemenschneider, Renee Bailey Iacona, Annetta D. Krebs, David H. Johnson, Giuseppe S. Giaccone, Roy S. Herbst, Christian Manegold, Masahiro Fukuoka, Mark G. Kris, Jose Baselga, Judith S. Ochs, and Daniel A. Haber.
Authors' Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following authors or their immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. 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 Informaton for Contributors.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Parkin DM, Bray FI, Devesa SS: Cancer burden in the year 2000: The global picture. Eur J Cancer 37:S4-S66, 2001 (suppl 8)
Lynch TJ, Adjei AA, Bunn PA Jr, et al: Novel agents in the treatment of lung cancer: Conference summary statement. Clin Cancer Res 10:4199s-4204s, 2004 (12 Pt 2)
Hidalgo M, Siu LL, Nemunaitis J, et al: Phase I and pharmacologic study of OSI-774, an epidermal growth factor receptor tyrosine kinase inhibitor, in patients with advanced solid malignancies. J Clin Oncol 19:3267-3279, 2001
Soulieres D, Senzer NN, Vokes EE, et al: Multicenter phase II study of erlotinib, an oral epidermal growth factor receptor tyrosine kinase inhibitor, in patients with recurrent or metastatic squamous cell cancer of the head and neck. J Clin Oncol 22:77-85, 2004
Perez-Soler R, Chachoua A, Hammond LA, et al: Determinants of tumor response and survival with erlotinib in patients with non-small-cell lung cancer. J Clin Oncol 22:3238-3247, 2004
Hanna N, Shepherd FA, Fossella FV, et al: Randomized phase III trial of pemetrexed versus docetaxel in patients with non-small-cell lung cancer previously treated with chemotherapy. J Clin Oncol 22:1589-1597, 2004
Kris MG, Natale RB, Herbst RS, et al: Efficacy of gefitinib, an inhibitor of the epidermal growth factor receptor tyrosine kinase, in symptomatic patients with non-small cell lung cancer: A randomized trial. JAMA 290:2149-2158, 2003
Data on file with OSI Pharmaceuticals
Higgins B, Kolinsky K, Smith M, et al: Antitumor activity of erlotinib (OSI-774, Tarceva) alone or in combination in human non-small cell lung cancer tumor xenograft models. Anticancer Drugs 15:503-512, 2004
Declaration of Helsinki. Adopted by the 18th World Medical Assembly, Helsinki, Finland, June 1964, amended by the 29th World Medical Assembly, Tokyo, Japan, October 1975, and the 35th World Medical Assembly, Venice, Italy, October, 1983
Schiller JH, Harrington D, Belani CP, et al: Comparison of four chemotherapy regimens for advanced non–small-cell lung cancer. N Engl J Med 346:92-98, 2002
Kelly K, Crowley J, Bunn PA Jr, et al: Randomized phase III trial of paclitaxel plus carboplatin versus vinorelbine plus cisplatin in the treatment of patients with advanced non–small-cell lung cancer: A Southwest Oncology Group trial. J Clin Oncol 19:3210-3218, 2001
Therasse P, Arbuck SG, Eisenhauser EA, et al: New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst 92:205-216, 2000
DakoCytomation PharmDxTM kit. Carpenteria, CA, DakoCytomation Inc, 2004
Paez G, J?nne J, Lee JC, et al: EGFR mutations in lung cancer: Correlation with clinical response to gefitinib therapy. Science 304:1497-1500, 2004
Lynch T, Bell D, Sordella R, et al: Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 350:2129-2139, 2004
Tran HT, Zinner R, Blumenschein GR, et al: Pharmacokinetic study of the phase III, randomized, double-blind, multicenter trial of paclitaxel (pac) and carboplatin (c) combined with erlotinib (e) or placebo in patients with advanced non-small cell lung cancer (NSCLC). J Clin Oncol 22:139, 2004 (abstr 2050)
Herbst RS, Giaccone G, Schiller JH, et al: Gefitinib in combination with paclitaxel and carboplatin in advanced non-small-cell lung cancer: A phase III trial—INTACT 2. J Clin Oncol 22:785-794, 2004
Shepherd FA, Pereira J, Ciuleanu TE, et al: A randomized placebo-controlled trial of erlotinib in patients with advanced non-small cell lung cancer (NSCLC) following failure of 1st line or 2nd line chemotherapy: A National Cancer Institute of Canada Clinical Trials Group (NCIC CTG) trial. J Clin Oncol 22:18, 2004 (abstr 7022)
Betensky R, Louis DN, Cairncross JG: Influence of unrecognized molecular heterogeneity on randomized clinical trials. J Clin Oncol 20:2495-2499, 2002
Pao W, Miller V, Zakowski M, et al: EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 101:13306-13311, 2004
Tracy S, Mukohara T, Hansen M, et al: Gefitinib induces apoptosis in the EGFRL858R non-small-cell lung cancer cell line H3255. Cancer Res 68:7241-7244, 2004
Sanchez-Cespedes M, Ahrendt SA, Piantadosi S, et al: Chromosomal alterationsin lung adenocarcinoma from smokers and nonsmokers. Cancer Res 61:1309-1313, 2001(Roy S. Herbst, Diane Prag)
University of California Los Angeles, Los Angeles
Kaiser Permanente, Oakland
Genentech Inc, South San Francisco, CA
Georgia Oncology Partners, Marietta, GA
Dana-Farber Cancer Institute, Boston, MA
Vanderbilt-Ingram Cancer Center, Nashville, TN
Memorial Sloan-Kettering Cancer Center, New York, NY, for the TRIBUTE Investigator Group
ABSTRACT
PURPOSE: Erlotinib is a potent reversible HER1/epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor with single-agent activity in patients with non–small-cell lung cancer (NSCLC). Erlotinib was combined with chemotherapy to determine if it could improve the outcome of patients with NSCLC.
PATIENTS AND METHODS: TRIBUTE randomly assigned patients with good performance status and previously untreated advanced (stage IIIB/IV) NSCLC to erlotinib 150 mg/d or placebo combined with up to six cycles of carboplatin and paclitaxel, followed by maintenance monotherapy with erlotinib. Random assignment was stratified by stage, weight loss in the previous 6 months, measurable disease, and treatment center. The primary end point was overall survival (OS). Secondary end points included time to progression (TTP), objective response (OR), and duration of response.
RESULTS: There were 1,059 assessable patients (526 erlotinib; 533 placebo). Median survival for patients treated with erlotinib was 10.6 v 10.5 months for placebo (hazard ratio, 0.99; 95% CI, 0.86 to 1.16; P = .95). There was no difference in OR or median TTP. Patients who reported never smoking (72 erlotinib; 44 placebo) experienced improved OS in the erlotinib arm (22.5 v 10.1 months for placebo), though no other prespecified factors showed an advantage in OS with erlotinib. Erlotinib and placebo arms were equivalent in adverse events (except rash and diarrhea).
CONCLUSION: Erlotinib with concurrent carboplatin and paclitaxel did not confer a survival advantage over carboplatin and paclitaxel alone in patients with previously untreated advanced NSCLC. Never smokers treated with erlotinib and chemotherapy seemed to experience an improvement in survival and will undergo further investigation in future randomized trials.
INTRODUCTION
Lung cancer is the most common cause of cancer deaths worldwide, and the development of more effective therapy remains challenging.1 The development of agents that target the epidermal growth factor receptor (EGFR) signal transduction pathways have provided a class of novel targeted therapeutic agents.2
Erlotinib (Tarceva; Genentech Inc, San Francisco, CA) is a potent reversible HER1/EGFR tyrosine kinase inhibitor (EGFR-TKI). This is a report of a phase III trial that evaluated whether concurrent administration of erlotinib with standard chemotherapy could enhance survival in chemotherapy-na?ve patients with advanced or metastatic non–small-cell lung cancer (NSCLC).
Erlotinib has been evaluated in a number of phase I and II studies as a single agent.3,4 In a phase II trial, 57 patients with advanced or recurrent NSCLC who had previously received platinum-based chemotherapy received erlotinib 150 mg/d as a single agent.5 In this study, patients had measurable tumors that expressed a low level of EGFR by immunohistochemical (IHC) analysis. The objective response rate of 12.3% (95% CI, 5.1% to 23.7%) and median survival of 8.5 months were comparable to results noted for docetaxel, pemetrexed, or gefitinib in refractory NSCLC.6,7 The most common adverse events with erlotinib consisted of mild/moderate rash and diarrhea.
Preclinical evaluations of erlotinib in athymic nude mice bearing tumor xenografts in HN5 head and neck and A431 epidermoid carcinoma cell lines established the effective dose for 50% inhibition of growth. Further evaluation in H460a and A549 NSCLC, and HN5 tumor models at the 50% inhibition of growth in combination with chemotherapy were conducted. No antagonism was detected with any of the chemotherapy agents evaluated, including paclitaxel and cisplatin. Additive effects were observed with some cytotoxic combinations of erlotinib, including paclitaxel and cisplatin.8,9
On the basis of phase II results and the therapeutic advantage demonstrated preclinically with erlotinib with platinum-based and taxane chemotherapy, a phase III trial was conducted using erlotinib in combination with paclitaxel and carboplatin.
This trial, designated TRIBUTE (Tarceva responses in conjunction with paclitaxel and carboplatin), was a multicenter (United States), randomized, placebo-controlled trial of paclitaxel and carboplatin with or without erlotinib in chemotherapy-na?ve patients with advanced NSCLC. The primary objective was to determine if the concurrent addition of erlotinib to paclitaxel plus carboplatin could prolong overall survival. Secondary objectives included improvement in time to progression (TTP), objective response rate, and safety.
PATIENTS AND METHODS
Eligibility Criteria
Inclusion required histologically documented stage IIIB or stage IV NSCLC; age 18 years; and Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1. Exclusion criteria included prior systemic chemotherapy for NSCLC; symptomatic or untreated brain metastases, prior exposure to agents directed at the HER axis; unstable systemic disease that would potentially preclude chemotherapy treatment with or without erlotinib; inadequate hematologic (granulocytes 1.5/mL), renal (creatinine > 1.5x upper limit of normal [ULN]), or hepatic function (liver function tests > 2.5x ULN). EGFR protein expression determination (by IHC methodology) was not an a priori condition for enrollment.
All patients provided informed consent, and approval was obtained from the ethics committee at each center. The study followed the Declaration of Helsinki and good clinical practice guidelines.10
Trial Design
Patients were randomly assigned to receive either daily oral erlotinib or placebo concurrently with chemotherapy (Fig 1). Randomization was stratified by disease stage (IIIB; IV), weight loss during the previous 6 months ( 5%; > 5%), tumor measurability (measurable; nonmeasurable), and treatment center. Patients received a maximum of six cycles of paclitaxel (200 mg/m2) followed by carboplatin (to achieve an area under concentration/time curve of 6 mg/mL · min) every 3 weeks in the absence of disease progression.
To evaluate the safety of erlotinib plus carboplatin/paclitaxel and to establish the starting erlotinib dose in this setting, the first 50 patients started therapy at 100 mg in the week before and during the first cycle of concurrent chemotherapy. If the study drug was tolerated, dose was escalated to 150 mg daily on day 15 of the first cycle.
Based on an acceptable safety profile, an independent Dose Selection Committee determined that a starting dose of 150 mg was tolerable. Following dose selection, approximately 1,000 patients were enrolled using this starting dose. Recommendations were provided for management of severe or intolerable rash or diarrhea (including guidelines for dose-reduction or interruption of study drug).
Due to the fact that there was no prior experience with the combination of erlotinib with carboplatin and paclitaxel in NSCLC, the first 400 patients were intensely monitored for safety to further assess for possible interaction between the chemotherapy specified for the study and concurrent administration of erlotinib. All adverse events were collected regardless of causality or relationship to study drugs. For the remaining 650 patients, collection of safety information was limited to serious adverse events, and all adverse events leading to study-drug or chemotherapy-dose interruption, reduction, or discontinuation. All adverse events were coded using the Medical Dictionary for Regulatory Activities (MedDRA), version 6.0. Severity was determined using the National Cancer Institute Common Toxicity Criteria (NCI-CTC), version 2.0.
Twenty-four patients (12 in each treatment arm) enrolled at the M.D. Anderson Cancer Center underwent pharmacokinetic sampling for erlotinib, paclitaxel, and carboplatin. These patients were enrolled after the final erlotinib dose had been selected. Sampling for OSI-774 (and its major metabolite OSI-420) occurred within 2 days of cycle 2, at 3 and 4 hours post-erlotinib. The sample schedule for day 1 of cycle 2 for erlotinib, paclitaxel, and carboplatin is presented in Table 1. Blood samples for OSI-774 and carboplatin were drawn into 2-mL heparinized tubes, while blood for paclitaxel assay was drawn into a 2-mL K-EDTA tube. All samples were centrifuged within 30 minutes of sample collection, for 10 minutes at 4°C at 3,000 x g. Plasma was extracted, frozen, and stored at –80°C.
Statistical Analysis
The primary end point was survival for the intent-to-treat population. For all patients alive at the time of the statistical analysis, survival was censored at the date of last contact. Secondary end points included TTP and response rate.
The study was designed to have 80% power for a two-sided hypothesis test that erlotinib with carboplatin/paclitaxel increased median survival by 25% relative to survival reported for carboplatin/paclitaxel alone within a randomized trial of four chemotherapy regimens. A median survival of approximately 8 months was reported in ECOG and Southwest Oncology Group trials with carboplatin/paclitaxel in a similar population.11,12 A 25% improvement in survival associated with the combination of erlotinib with this regimen would result in a median survival of 10.13 months (a hazard ratio of 0.80). Based on these assumptions, 642 events were required corresponding to enrollment of 1,050 patients. This sample size was also sufficient to detect a 33% improvement in 1-year survival, with similar statistical power.
Survival, TTP, and objective response rate were summarized by predefined categories including age, sex, ECOG performance status, weight loss during the previous 6 months, disease stage, smoking history, and selected markers on the EGFR pathway. Smoking history was collected prospectively as a data element within the screening case report forms. Patients were asked whether they had a prior smoking history, and, if so, the number of prior pack years. No minimum prior exposure (eg, 100 cigarettes) was used to define never smokers for this study.
Based on emerging results regarding efficacy in never smokers treated with EGFR tyrosine kinase inhibitors, the statistical analytic plan for the study was revised before unblinding to include such an analysis in this trial.
The Cox proportional hazards model was used to estimate the effect of risk factors on survival and TTP and to evaluate any modifications of treatment effect. To evaluate the effects of baseline characteristics on objective response rates and to assess modifications of treatment effect, a logistic regression model was applied.
An independent Data Monitoring Committee (DMC) reviewed safety on a quarterly basis and performed an interim analysis of efficacy and safety (321 deaths), in order to potentially terminate the study if the DMC had safety concerns or if robust efficacy was demonstrated.
Assessments
Tumor response was determined by Response Evaluation Criteria in Solid Tumors (RECIST).13 Tumor assessments were performed on day 21 of chemotherapy cycles 2, 4, and 6, and every 8 weeks following completion of chemotherapy. Time to symptomatic deterioration was assessed via the Lung Cancer Symptom Scale at these intervals. EGFR expression was determined retrospectively via IHC staining analysis on unstained slides, if available.14 Positive EGFR expression was indicated by weak to strong membranous staining in 10% or more tumor cells. Safety was evaluated by analyzing adverse events and laboratory tests.
RESULTS
Patients
Between July 18, 2001, and August 19, 2002, 1,079 patients were randomly assigned: 539 patients to erlotinib and 540 patients to placebo. Demographic and disease characteristics were balanced between the two treatment arms (Table 2). The majority of patients in both treatment arms had stage IV disease, with 5% weight loss in the previous 6 months.
Efficacy
Median survival was 10.6 months in the erlotinib arm versus 10.5 months in the placebo arm (hazard ratio, 0.995; 95% CI, 0.86 to 1.16; P = .95; Fig 2A). TTP was 5.1 months for erlotinib and 4.9 months for placebo (P = .36; Fig 2B).
The objective response rate (complete and partial) was similar between the erlotinib and placebo arms (21.5% v 19.3%, respectively; P = .36) and were partial responses in both arms.
Subset analyses failed to demonstrate any significant improvement in survival by age, sex, race, cancer stage, ECOG status, prior weight loss, histology, previous cancer-related surgery, or EGFR expression. There was no correlation between the level of EGFR expression (as assessed by IHC staining analysis using the DakoCytomation PharmDx test kit; DakoCytomation, Carpinteria, CA) and clinical outcome. Unstained slides were available for 445 of the 1,079 patients enrolled on the study. Of these, 344 were assessable for analysis (180 erlotinib, 164 placebo). The hazard ratio for EGFR-positive histology (93 erlotinib, 74 placebo) was 1.00 (95% CI, 0.69 to 1.45) versus a hazard ratio for EGFR-negative histology (87 erlotinib, 90 placebo) of 1.02 (95% CI, 0.71 to 1.46).
Analysis for the newly identified somatic mutations in the EGFR kinase domain was performed on 228 assessable samples of diagnostic tissue and will be reported in detail elsewhere.15,16
One prespecified subset had a difference in outcome. Patients who reported that they had never smoked experienced substantial prolongation in survival with erlotinib treatment: 22.5 months (n = 72) versus 10.1 months in the placebo arm (n = 44; hazard ratio of 0.49; 95% CI, 0.28 to 0.85; Fig 3), which was independent of tumor histology. Median overall survival of never smokers treated with chemotherapy alone was similar to that of former or current smokers in the same treatment arm (10 months). These results for never smokers contrasted sharply with the outcome for current (8.4 months erlotinib; 9.1 months placebo) and previous smokers (10.0 months erlotinib; 10.9 months placebo).
The median TTP in never smokers was 6.0 months in the erlotinib arm versus 4.3 months for placebo (hazard ratio, 0.50; 95% CI, 0.31 to 0.80). The response rate of never smokers receiving erlotinib plus carboplatin/paclitaxel was significantly higher than that of those receiving carboplatin/paclitaxel alone (21 [30%] of 69; 95% CI, 20% to 43% v five [11%] of 44; 95% CI, 4% to 25%; P = .02). The never-smoker subset tended to be younger (58 v 64 years), female (60% v 37%), and to have adenocarcinoma (82% v 58%) in comparison to prior/current smokers.
An exploratory analysis of patients who survived beyond chemotherapy and received study drug as monotherapy thereafter showed an improvement in survival in the erlotinib arm that suggested a maintenance effect.
In the 861 patients who survived beyond 4 months (erlotinib 408; placebo 453), median survival with erlotinib was 13.6 v 12.2 months (P = .04). In the 740 patients who survived beyond 6 months (erlotinib 348; placebo 392), median survival with erlotinib was 15.4 v 13.8 months for placebo (P = .007).
Duration of Therapy, Dose Adherence, Dose Intensity, and Exposure
The median duration on study drug was 4.6 months with erlotinib and 5.3 months with placebo. There were more dose interruptions (because of compliance or toxicity) and reductions with erlotinib versus placebo (250 v 190 interruptions and 135 v 17 reductions, respectively). Mean total dose of study drug per patient was lower in the erlotinib arm versus in the placebo arm (24.0 g v 26.5 g, respectively; P = .035); this was also true for the mean daily dose (137.8 mg/d v 146.8 mg/d, respectively; P < .0001). Overall, 14 patients (10 erlotinib, four placebo) were unblinded early because a potential investigational new drug safety report was filed.
Exposure to chemotherapy was similar between the two treatment arms. The median number of cycles of carboplatin and paclitaxel was five for both arms (P = .246), whereas respective mean doses per cycle for carboplatin (619.9 mg for erlotinib 635.0 mg for placebo; P = .1) and paclitaxel (358.4 for erlotinib; 362.8 mg for placebo; P = .2) were also similar.
Carboplatin and paclitaxel sampling for pharmacokinetic analysis was performed in 21 patients. Overall, exposure to carboplatin and paclitaxel was similar between erlotinib and placebo-treated patients. Erlotinib exposure was similar to historical data from phase I and II trials.3,5,17 Detailed results will be reported elsewhere.
Results of assessment for time to symptomatic progression as determined by assessment of lung cancer symptom scores will be presented elsewhere.
Safety and Tolerability
Adverse events in the erlotinib arm were similar in incidence and severity to those of the placebo arm, with the exception of an increased incidence of rash and diarrhea, which are known to be associated with EGFR tyrosine kinase inhibition (Table 3).
The erlotinib arm had a higher incidence of study drug–related serious adverse events relative to the placebo arm (8.6% erlotinib; 2.4% placebo). The most common events included diarrhea (3.8% erlotinib; 1.1% placebo) and rash (dermatitis acneiform; 0.8% erlotinib; 0 placebo). The rate of interstitial lung disease (ILD) was estimated by inclusion of a set of 24 preferred coding terms that have been associated with such events, regardless of causality or possible confounders. There were five severe ILD-like events in the erlotinib arm (1.0%), versus one event in the placebo arm (0.2%). All ILD-like events were fatal.
A total of 662 treated patients died before the data cutoff: 322 (61.2%) in the erlotinib arm and 340 (63.8%) in the placebo arm. The majority of the deaths in both arms were attributed to disease progression by investigators (275 erlotinib; 318 placebo).
Forty-eight deaths were attributed to an adverse event: 33 (10.2%) of 322 deaths in the erlotinib arm and 15 (4.4%) of 340 deaths in the placebo arm. The majority of the imbalance was due to infection (seven erlotinib; one placebo) and GI events (four erlotinib; one placebo).
With respect to infectious events leading to death, pneumonias accounted for six events (five erlotinib, one placebo). All such events occurred during concomitant administration of chemotherapy and study drug. None of the pneumonias leading to death were assessed as related to study drug by the respective investigators (events were related to "disease under study"/protocol-specified chemotherapy). The two remaining infectious events leading to death in the erlotinib arm were neutropenic sepsis (occurring after the first cycle of chemotherapy; assessed as related to erlotinib and protocol-specified chemotherapy) and septic shock (occurring on the fifth day of study; assessed by the investigator as related to the disease under study).
All of the GI events leading to death occurred during the phase of concomitant administration of chemotherapy and study drug. Of the GI events leading to death in the erlotinib arm, there was one case of GI hemorrhage assessed as related to erlotinib and concomitant medication (ie, warfarin sodium). Another erlotinib patient experienced a GI perforation on study day 17 (after one cycle of chemotherapy) and declined surgical intervention. Diarrhea led to death in a patient with a history of diabetes mellitus who had experienced nausea, vomiting, and diarrhea since the first cycle of chemotherapy, on study day 8 (assessed as related to concurrent illness and medication). Finally, intestinal ischemia was the cause of death on study day 11 and was assessed as related to the single cycle of protocol-specified chemotherapy by the treating physician. The single GI event leading to death in the placebo cohort was perforation of a gastric ulcer after five cycles of chemotherapy (assessed as related to disease under study, and concurrent medication and illness).
Twenty-one deaths were attributed to other causes: 14 (4.3%) of 322 deaths in the erlotinib arm and seven (2.1%) of 340 in the placebo arm. Nine deaths in this category were due to unknown causes (eight erlotinib; one placebo), while eight deaths were attributed to cardiac/pulmonary arrest (five erlotinib; three placebo).
There were more early deaths (ie, during concomitant administration of chemotherapy and study drug) in the erlotinib arm. At study day 120 (the point of maximum difference between treatment arms), there were 105 deaths in erlotinib arm versus 70 in placebo. The majority of early deaths in TRIBUTE were attributed to progression of underlying lung cancer (78% erlotinib, 87% placebo). This effect was also evident in the phase III trial of gefitinib with concurrent administration of the same chemotherapy regimen (INTACT-2).18
DISCUSSION
This randomized, placebo-controlled trial examined the efficacy and safety of erlotinib in combination with paclitaxel and carboplatin for first-line therapy of advanced NSCLC in 1,079 unselected patients. There was no improvement in survival, TTP, or response rate compared with chemotherapy given alone. These clinical results contrast with preclinical studies, suggesting additive effects of combining erlotinib and paclitaxel and a platinum.
These results are consistent with previous reports in which paclitaxel and carboplatin were combined with an EGFR-TKI.18 In contrast to these combination studies, erlotinib monotherapy effected a statistically significant survival advantage over placebo in 731 patients with recurrent, previously treated NSCLC. Erlotinib was associated with a 37% improvement in overall survival (log-rank hazard ratio, 0.73; 95% CI, 0.6 to 0.87; P = .001) and progression-free survival (hazard ratio, 0.61; 95% CI, 0.51 to 0.73; P < .001).19
Although a post hoc analysis of patients who had lived beyond concomitant administration of study drug with chemotherapy suggested a survival benefit associated with erlotinib, such results must be viewed with obvious caution. One has to consider that such patients who demonstrated a benefit in this analysis had completed chemotherapy with erlotinib. As such, use of erlotinib in a maintenance setting after first-line chemotherapy has not been proven and will need to be evaluated prospectively in a randomized trial setting to demonstrate its potential therapeutic benefit.
In light of the positive preclinical data, the lack of benefit in the first-line trial when erlotinib was combined with chemotherapy may be based on an as yet unidentified negative interaction when an EGFR-TKI is given concurrently with a platinum-based regimen and a taxane. However, the efficacy of erlotinib to carboplatin and paclitaxel in the never smoker population argues against antagonism. In addition, there was no apparent pharmacokinetic interaction between erlotinib and carboplatin and paclitaxel in this trial.
The patients enrolled in TRIBUTE were not selected based on known prognostic factors. Betensky et al presented a model for the potential effect of molecular heterogeneity in trial design.20 If such heterogeneity were to confer a treatment benefit in a subset of patients, the overall trial could be underpowered to detect an effective therapy. Thus, one possible explanation for the negative results may be the lack of patient selection for such a factor, at least in the first-line setting. For example, the discovery of mutations in the EGFR tyrosine kinase domain in some patients may be one example of the potential to appropriately select for a subset of patients who may benefit from EGFR inhibition in the first-line setting.
Sensitivity to erlotinib or gefitinib has been strongly associated with mutations, most commonly, deletions of four to six amino acids in exon 19 or a point mutation (L858R) in exon 21. The tissue blocks from the patients participating in TRIBUTE have been studied for EGFR mutations, and this analysis will be reported in a separate publication.
The fact that never smokers who received chemotherapy plus erlotinib achieved a 120% increase in median survival as compared with the patients who received chemotherapy alone supports this hypothesis, and argues against a universally unfavorable negative interaction with chemotherapy. This finding, however, needs to be validated in an appropriately designed clinical trial.
While the observation in never smokers might be spurious, the similar demographics and balanced baseline characteristics between the two arms of this study argues against this. In addition, similar results were also observed in the recently completed trial of single-agent erlotinib in patients with refractory advanced NSCLC conducted by the National Cancer Institute of Canada (NCIC), as well as in INTACT-2 with gefitinib.18,19 However, in the NCIC trial, a survival benefit was also observed for patients who were current/prior smokers.
Recent work by several groups has suggested that never smokers are more likely to harbor mutations than smokers in the tyrosine kinase domain of EGFR. For example, Pao et al reported such mutations in seven of 15 unselected early-stage never smokers with adenocarcinoma.21
The low response rate of never smokers treated with chemotherapy and placebo in this trial (11%) raises the question of whether patients who may have an increased likelihood of these mutations are less sensitive to cytotoxic chemotherapy. Unpublished, preclinical studies in which cell lines that harbor these mutations are treated with both EGFR-TKIs and cytotoxic chemotherapy are ongoing.21A However, the long survival observed in these patients, which is far greater than that seen in any reported phase II or III trial to date, might simultaneously suggest a favorable interaction between chemotherapy and erlotinib. In never smokers, it would be important to investigate prospectively the value of combining erlotinib with chemotherapy versus erlotinib alone.
Studies have now shown that tumors of never smokers are molecularly less complex and may confer a better prognosis for patients.22 In some countries within the Pacific Rim, lung cancer, particularly in females, is far more common in never smokers than smokers. Thus, racial background and smoking history should be considered when designing and interpreting clinical trials in NSCLC, particularly with EGFR-TKIs.
The side effects of rash and diarrhea associated with erlotinib seemed to be characteristic of this class of drugs. In addition, the incidence of life-threatening toxicities such as ILD, which have rarely been reported with gefitinib, was likewise rarely observed with erlotinib in this trial (1.0% erlotinib; 0.2% placebo).18 There was however, a higher rate of deaths (regardless of whether related to progression of NSCLC or toxicity) in the erlotinib arm relative to placebo associated with concomitant administration of chemotherapy (163 erlotinib, 125 placebo). This effect was not noted in the European counterpart trial with erlotinib, which specified a different chemotherapy regimen (gemcitabine and cisplatin). As noted previously, this is suggestive of a potential negative interaction between the carboplatin and paclitaxel regimen specified in this study and erlotinib.
Additional preclinical and clinical investigations are ongoing to identify whether alternative schedules and or dosing regimens might allow erlotinib to be combined with chemotherapy in patients with first-line stage IIIB/IV disease. In addition, the identity of clinical or molecular surrogates that may predict for improved outcome (ie, never smokers or somatic mutation in the kinase domain of EGFR) are still preliminary and will require further investigation in prospective, randomized trial settings.
Editor’s Note
A related article on this subject will be published in the November 1, 2005, issue titled, Epidermal Growth Factor Receptor Mutations and Gene Amplification in Non–Small-Cell Lung Cancer: Molecular Analysis of the IDEAL/INTACT Gefitinib Trials; by Daphne W. Bell, Thomas J. Lynch, Sara M. Haserlat, Patricia L. Harris, Ross A. Okimoto, Brian W. Brannigan, Dennis C. Sgroi, Beth Muir, Markus J. Riemenschneider, Renee Bailey Iacona, Annetta D. Krebs, David H. Johnson, Giuseppe S. Giaccone, Roy S. Herbst, Christian Manegold, Masahiro Fukuoka, Mark G. Kris, Jose Baselga, Judith S. Ochs, and Daniel A. Haber.
Authors' Disclosures of Potential Conflicts of Interest
Although all authors completed the disclosure declaration, the following authors or their immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. 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 Informaton for Contributors.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Parkin DM, Bray FI, Devesa SS: Cancer burden in the year 2000: The global picture. Eur J Cancer 37:S4-S66, 2001 (suppl 8)
Lynch TJ, Adjei AA, Bunn PA Jr, et al: Novel agents in the treatment of lung cancer: Conference summary statement. Clin Cancer Res 10:4199s-4204s, 2004 (12 Pt 2)
Hidalgo M, Siu LL, Nemunaitis J, et al: Phase I and pharmacologic study of OSI-774, an epidermal growth factor receptor tyrosine kinase inhibitor, in patients with advanced solid malignancies. J Clin Oncol 19:3267-3279, 2001
Soulieres D, Senzer NN, Vokes EE, et al: Multicenter phase II study of erlotinib, an oral epidermal growth factor receptor tyrosine kinase inhibitor, in patients with recurrent or metastatic squamous cell cancer of the head and neck. J Clin Oncol 22:77-85, 2004
Perez-Soler R, Chachoua A, Hammond LA, et al: Determinants of tumor response and survival with erlotinib in patients with non-small-cell lung cancer. J Clin Oncol 22:3238-3247, 2004
Hanna N, Shepherd FA, Fossella FV, et al: Randomized phase III trial of pemetrexed versus docetaxel in patients with non-small-cell lung cancer previously treated with chemotherapy. J Clin Oncol 22:1589-1597, 2004
Kris MG, Natale RB, Herbst RS, et al: Efficacy of gefitinib, an inhibitor of the epidermal growth factor receptor tyrosine kinase, in symptomatic patients with non-small cell lung cancer: A randomized trial. JAMA 290:2149-2158, 2003
Data on file with OSI Pharmaceuticals
Higgins B, Kolinsky K, Smith M, et al: Antitumor activity of erlotinib (OSI-774, Tarceva) alone or in combination in human non-small cell lung cancer tumor xenograft models. Anticancer Drugs 15:503-512, 2004
Declaration of Helsinki. Adopted by the 18th World Medical Assembly, Helsinki, Finland, June 1964, amended by the 29th World Medical Assembly, Tokyo, Japan, October 1975, and the 35th World Medical Assembly, Venice, Italy, October, 1983
Schiller JH, Harrington D, Belani CP, et al: Comparison of four chemotherapy regimens for advanced non–small-cell lung cancer. N Engl J Med 346:92-98, 2002
Kelly K, Crowley J, Bunn PA Jr, et al: Randomized phase III trial of paclitaxel plus carboplatin versus vinorelbine plus cisplatin in the treatment of patients with advanced non–small-cell lung cancer: A Southwest Oncology Group trial. J Clin Oncol 19:3210-3218, 2001
Therasse P, Arbuck SG, Eisenhauser EA, et al: New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst 92:205-216, 2000
DakoCytomation PharmDxTM kit. Carpenteria, CA, DakoCytomation Inc, 2004
Paez G, J?nne J, Lee JC, et al: EGFR mutations in lung cancer: Correlation with clinical response to gefitinib therapy. Science 304:1497-1500, 2004
Lynch T, Bell D, Sordella R, et al: Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 350:2129-2139, 2004
Tran HT, Zinner R, Blumenschein GR, et al: Pharmacokinetic study of the phase III, randomized, double-blind, multicenter trial of paclitaxel (pac) and carboplatin (c) combined with erlotinib (e) or placebo in patients with advanced non-small cell lung cancer (NSCLC). J Clin Oncol 22:139, 2004 (abstr 2050)
Herbst RS, Giaccone G, Schiller JH, et al: Gefitinib in combination with paclitaxel and carboplatin in advanced non-small-cell lung cancer: A phase III trial—INTACT 2. J Clin Oncol 22:785-794, 2004
Shepherd FA, Pereira J, Ciuleanu TE, et al: A randomized placebo-controlled trial of erlotinib in patients with advanced non-small cell lung cancer (NSCLC) following failure of 1st line or 2nd line chemotherapy: A National Cancer Institute of Canada Clinical Trials Group (NCIC CTG) trial. J Clin Oncol 22:18, 2004 (abstr 7022)
Betensky R, Louis DN, Cairncross JG: Influence of unrecognized molecular heterogeneity on randomized clinical trials. J Clin Oncol 20:2495-2499, 2002
Pao W, Miller V, Zakowski M, et al: EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 101:13306-13311, 2004
Tracy S, Mukohara T, Hansen M, et al: Gefitinib induces apoptosis in the EGFRL858R non-small-cell lung cancer cell line H3255. Cancer Res 68:7241-7244, 2004
Sanchez-Cespedes M, Ahrendt SA, Piantadosi S, et al: Chromosomal alterationsin lung adenocarcinoma from smokers and nonsmokers. Cancer Res 61:1309-1313, 2001(Roy S. Herbst, Diane Prag)