Final survival and safety results from a multicenter, open-label, phase 3b trial of erlotinib in patients with advanced nonsmall cell lung cancer

Authors


  • Genentech, Inc. provided writing assistance for this article.

Abstract

BACKGROUND.

Erlotinib is an orally available, reversible inhibitor of epidermal growth factor receptor (EGFR) with a proven survival advantage for patients with locally advanced or metastatic nonsmall cell lung cancer (NSCLC) who have failed a prior chemotherapy. This phase 3b, multicenter, open-label trial of erlotinib in patients with advanced NSCLC who had progressed after standard chemotherapy treatment was conducted to examine the efficacy and safety of erlotinib monotherapy in patients with advanced NSCLC who had developed disease progression after previous chemotherapy and to characterize the duration of survival and the response rate of erlotinib-treated patients in subpopulations defined by other patient characteristics before U.S. Food and Drug Administration approval.

METHODS.

A total of 229 patients were enrolled and treated with the standard dose of erlotinib (150 mg once daily). The coprimary objectives were to characterize the overall response rate (ORR) and overall survival (OS) associated with erlotinib therapy in this group and in patient subsets defined by tobacco history. Secondary objectives were to assess safety, to characterize OS and ORR in patient subpopulations, and to determine duration of time on treatment. Patients could remain on study up to 9 months after approval.

RESULTS.

The ORR was 8.3% (95% confidence interval [95% CI], 5.2–2.4%). The ORR in never-smokers, previous smokers, and current smokers was 28.6% (95% CI, 13.2–50.6%), 6.0% (95% CI, 3.0–10.4%), and 7.3% (95% CI, 2.0–19.0%), respectively. The median OS for all patients was 6.3 months (95% CI, 4.7–8.0 months). In previous and current smokers, the median survival was 5.2 months (95% CI, 4.2–7.3 months) and 6.3 months (95% CI, 3.6–9.2 months), respectively, and was not reached in never-smokers. The median duration of treatment was 10.6 weeks. One (0.4%) interstitial lung disease-like event was reported.

CONCLUSIONS.

No new safety signals were noted. The observed ORR and survival data are consistent with results from the pivotal trial BR.21. Cancer 2008. © 2008 American Cancer Society.

Lung cancer is the leading cause of cancer-related death in North America, and is expected to account for approximately 160,000 deaths in the U.S.1 It is estimated that in the same period >174,000 new cases will be diagnosed. The overall 5-year survival rate for patients who develop lung cancer in the U.S. is 14%, having only slightly improved over the past 20 years.2 Approximately 85% of patients with lung cancer have nonsmall cell lung cancer (NSCLC),3 and of those, approximately 40% present with incurable metastatic disease.4 Cytotoxic chemotherapy, which has historically been the mainstay of treatment for advanced NSCLC, appears to have reached a plateau of therapeutic efficacy.5, 6

Therapies that target proteins implicated in the development and progression of cancer have demonstrated efficacy. One of the most intensively investigated targeted therapies includes those directed against members of the HER family of receptor tyrosine kinases, which includes HER-1/epidermal growth factor receptor (EGFR) and HER-2. The HER family has been implicated in the development and progression of multiple cancers, including breast cancer, head and neck cancer, and NSCLC.7–9 Preclinical studies have demonstrated that, in malignant cells, EGFR downstream signaling pathways are often dysregulated.10 The strongest evidence for a direct role of HER signaling in tumorigenesis is based on findings that EGFR is more highly expressed in tumors than in normal tissue.11 Although EGFR is expressed at low levels in normal lung tissue, in NSCLC patients, EGFR is expressed at higher levels in 13% to 80% of tumors.12 Consequently, EGFR was considered a good candidate for targeted NSCLC therapy.

Erlotinib (Tarceva; Genentech, South San Francisco, Calif) is an orally available, small molecule inhibitor of EGFR tyrosine kinase activity. It is currently approved by the U.S. Food and Drug Administration (FDA) as monotherapy for treating locally advanced or metastatic NSCLC in patients who have failed at least 1 prior chemotherapy regimen, and in combination with gemcitabine for treating locally advanced, unresectable, or metastatic pancreatic cancer.13 In the randomized, double-blind pivotal trial (BR.21) that included 731 patients with recurrent NSCLC, patients treated with erlotinib demonstrated prolonged progression-free survival (PFS) and overall survival (OS) compared with placebo-treated patients.14 In the BR.21 study, the median survival for the overall population improved by 42.5%, from 4.7 months to 6.7 months (hazards ratio of 0.71; P < .0001). As assessed by both response and survival, the BR.21 study suggested that patients with no prior tobacco exposure and patients of Asian descent respond well to erlotinib. Although patients in the BR.21 study who had adenocarcinoma appeared to have a higher response rate than patients with squamous cell carcinoma (13.9% vs 3.8%), the survival benefit was similar for the 2 subpopulations.

Based on the BR.21 study, the current open-label, single-arm, advanced access, phase 3b trial was initiated before the regulatory approval of erlotinib. It was designed to examine the efficacy and safety of erlotinib monotherapy in patients with advanced NSCLC who had developed disease progression after previous chemotherapy and to characterize the duration of survival and response rate of erlotinib-treated patients in subpopulations defined by other patient characteristics, including histology, sex, race/ethnicity, and EGFR mutation status.

MATERIALS AND METHODS

Study Design

The current study was a phase 3b, multicenter, open-label, expanded access clinical study of daily oral erlotinib in patients with advanced (inoperable stage IIIb or IV) NSCLC who had developed disease recurrence after standard chemotherapy (determined according to the American Joint Committee on Cancer TNM Staging System). This trial was designed to provide patients access to erlotinib before FDA approval and to examine the benefit in several subsets of patients.

Eligible patients were age ≥18 years with inoperable and incurable, locally advanced, recurrent, or metastatic NSCLC; who had developed disease recurrence after standard chemotherapy (combination chemotherapy such as 2-drug combination chemotherapy or a single-agent chemotherapy agent for patients who were elderly or had poor performance status); who were able to receive oral medications; who had recovered from the toxic effects of prior therapy; who had an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0 to 3; and who had adequate cardiac, hematologic, hepatic, and renal function (granulocyte count ≥1.0 × 109/L, platelet count ≥75 × 109/L, serum bilirubin <1.5 × upper limit of normal [ULN], aspartate aminotransferase [AST] <2 × ULN [unless elevation was clearly because of liver metastases; then AST had to be <5 × ULN], and serum creatinine ≤1.5 mg/dL).

Patients were ineligible if they had unstable systemic disease (including active infection; unstable angina; congestive heart failure; myocardial infarction within the last 6 months; or hepatic, renal, or metabolic disease); prior therapy with any systemic HER-1/EGFR small molecule or other investigational agents in this class; a history of another malignancy within the past 2 years unless the malignancy had been adequately treated and was associated with a 5-year anticipated survival of ≥90%; known central nervous system metastases that had not yet been definitively treated with surgery and/or radiotherapy or that were symptomatic or unstable; or were pregnant or nursing.

During the treatment period, patients received single-agent erlotinib until disease progression at the approved standard recommended dosage for patients with NSCLC (150 mg/day). Patients were removed from treatment if they experienced unacceptable toxicity or were unwilling or unable to comply with study requirements. Dose reduction or interruption of erlotinib for grade 3 to 4 diarrhea or rash or for any grade of pulmonary events could occur at any time during the study (in 50-mg decrements).

Disease was assessed within 30 days of the initiation of study therapy and was reassessed at intervals determined by the investigator and based on standard practices. Overall best response and date of disease progression were determined by study investigators.

Study Objectives

The primary objectives of the current study were to characterize the best overall response rate (ORR) and the duration of survival in all patients and in subsets of NSCLC patients defined by their previous tobacco exposure. The secondary objectives were to evaluate safety in the patient population through the assessment of serious adverse events (SAEs) and adverse events (AEs) leading to treatment discontinuation and to characterize the duration of survival and response rate of erlotinib in subpopulations of patients defined by other patient characteristics, including histology, sex, and race/ethnicity; and to characterize the duration of time on treatment.

Safety was assessed by analyzing SAEs and all AEs leading to discontinuation from the study. SAEs and AEs were graded according to National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE; version 3.0).15 Investigators assessed whether there was a reasonable possibility that erlotinib caused or contributed to the AE.

Statistical Methods

The original sample size of 5000 patients was chosen so that relatively uncommon safety events could be assessed, and so that patient subgroups could be analyzed for the primary endpoints. Because the FDA approved erlotinib sooner than anticipated (November 2004), the study was terminated after 233 patients had enrolled. Therefore, because of insufficient data, it was not possible to conduct all the intended analyses.

The duration of survival was defined as the time from the initiation of erlotinib treatment until death from any cause. For patients who were discontinued from the study, survival was censored at the date of last contact. Kaplan-Meier methodology was used to estimate the survival rate, median survival time, and 95% confidence intervals (95% CIs). Overall best responses as determined by investigators were summarized and the 95% Blyth-Still-Casella exact CIs were calculated.

RESULTS

Patients and Demographics

Between August 2004 and November 2004, 233 patients were enrolled at 94 sites in the U.S. A total of 229 patients (98.3%) received the study drug (Table 1). Four patients were not treated because they refused medication, were noncompliant, or were lost to follow-up before receiving study drug. At baseline, the median patient age was 68 years. The majority of patients were white (85.6%) and were former or current smokers (90.8%). Of those patients with a history of smoking, they had smoked for a mean of 39.4 years, and >80% had accumulated >20 pack-years of smoking. The most common cancer histology was adenocarcinoma (45%), followed by squamous cell carcinoma (33.6%). A total of 46.7% of patients had received ≥2 lines of therapy; the majority (83.4%) had received prior platinum, and 38% had received both platinum and a taxane. Nearly 31% of patients had an ECOG PS of 2 or 3.

Table 1. Patient Characteristics
DemographicTreated patients N = 229
  1. SD indicates standard deviation; ECOG, Eastern Cooperative Oncology Group; BAC, bronchioalveolar carcinoma.

Median age (range), y68 (27–90)
Sex no. (%)
 Male139 (60.7)
 Female90 (39.3)
Race/ethnicity, no. (%)
 White196 (85.6)
 Black18 (7.9)
 Hispanic11 (4.8)
 Asian or Pacific Islander2 (0.9)
 American Indian or Alaskan Native1 (0.4)
 Other1 (0.4)
Smoking status, no. (%)
 Previous smoker167 (72.9)
 Current smoker41 (17.9)
 Never smoked21 (9.2)
Y of being a regular smoker (smokers only)
 Mean ± SD39.4 ± 13.5
 Range4–64
Pack-y of smoking (smokers only), no. (%)
 ≤57 (3.4)
 >5 but ≤109 (4.3)
 >10 but ≤2025 (12.0)
 >20167 (80.3)
ECOG performance status, no. (%)
 058 (25.3)
 1101 (44.1)
 253 (23.1)
 317 (7.4)
Type of current cancer, no. (%)
 Adenocarcinoma103 (45.0)
 Squamous cell carcinoma77 (33.6)
 Other or unknown31 (13.5)
 Large cell carcinoma18 (7.9)
 Pure BAC or BAC-like4 (1.7)
Time since initial lung cancer diagnosis, no. (%)
 <6 mo39 (17.0)
 6–12 mo72 (31.4)
 >12 mo118 (51.5)
Prior chemotherapy, no. (%)
 Platinum191 (83.4)
 Taxane105 (45.9)
 Taxane plus platinum87 (38.0)
No. of prior chemotherapy regimens, no. (%)
 1122 (53.3)
 >1107 (46.7)

Efficacy

Of the 229 patients with advanced, recurrent NSCLC who received erlotinib monotherapy, 3 patients achieved a complete response and 16 achieved a partial response (Table 2). At the time of study termination, 137 patients (59.8%) had died (Table 3). The median duration of survival was 6.3 months (95% CI, 4.7–8.0 months), and the best ORR was 8.3% (95% CI, 5.2–12.4%) (Table 2). In those patients who had never smoked, the ORR was 28.6% versus 6.0% in previous smokers and 7.3% in current smokers (Table 2). For previous smokers, the median duration of survival was 5.2 months (95% CI, 4.2–7.3 months) and for current smokers it was 6.3 months (95% CI, 3.6–9.2 months) (Table 3). For patients who had never smoked, the median survival was not reached. Kaplan-Meier estimates of survival are shown in Figure 1.

Figure 1.

Kaplan-Meier estimates of survival (A) All patients. (B) All patients stratified by smoking history.

Table 2. Tumor Response by Smoking Exposure*
 All treated patients n = 229Never smoked n = 21Previous smokers n = 167Current smokers n = 41
  • CR indicates complete response; PR, partial response; SD, stable disease; PR, progressive disease; ORR, overall response rate; 95% CI, 95% confidence interval.

  • *

    Never smoked, patients who have smoked a total of ≤100 cigarettes in patient lifetime; previous smoker, patients who smoked >100 cigarettes but quit smoking; current smoker, patients currently smoking cigarettes.

Best tumor response over the course of treatment, no. (%)
 CR3 (1.3)0 (0.0)3 (1.8)0 (0.0)
 PR16 (7.0)6 (28.6)7 (4.2)3 (7.3)
 SD53 (23.1)6 (28.6)38 (22.8)9 (22.0)
 PD124 (54.1)6 (28.6)90 (53.9)28 (68.3)
 Unevaluable33 (14.4)3 (14.3)29 (17.4)1 (2.4)
ORR (CR or PR)19 (8.3)6 (28.6)10 (6.0)3 (7.3)
95% CI of ORR (%)5.2–12.413.2–50.63.0–10.42.0–19.0
Table 3. Overall Survival by Smoking Exposure*
 Treated patients n = 229Never smoker n = 21Previous smokers n = 167Current smokers n = 41
  • NR indicates not reached; 95% CI, 95% confidence interval.

  • *

    Never smoked, patients who have smoked a total of ≤100 cigarettes in patient lifetime; previous smoker, patients who smoked >100 cigarettes but quit smoking; current smoker, patients currently smoking cigarettes.

  • Estimated using the Kaplan-Meier method.

  • 95% Blyth-Still Casella exact confidence interval.

  • §

    Censored value.

  • Number could be larger because the median survival was not reached.

Median duration of survival, mo6.3NR5.26.3
95% CI for the median4.7–8.0NR4.2–7.33.6–9.2
25th–75th percentile2.6-NR7.2-NR2.2-NR3.3–9.7
No. of patients who died (%)137 (59.8)6 (28.6)104 (62.3)27 (65.9)
Duration of survival, mo
 Median (95% CI)6.3 (4.7–8.0)NR5.2 (4.2–7.3)6.3 (3.6–9.2)
 25th–75th percentile§2.67.2-2.2-3.3–9.7
 Minimum-maximum0.2–12.10.8–12.10.2–12.10.6–10.6

Table 4 summarizes the median duration of survival and best ORR among different subgroups, as defined by age group, sex, histology, number of prior chemotherapy regimens, baseline ECOG PS, and pack-years of smoking. The median survival of patients with an ECOG PS of 0 or 1 was approximately 6 months longer than those with an ECOG PS of 2 or 3 (9.2 months vs 3.0 months). Trends toward longer median survival and higher response rates were observed in patients age ≥65 years, female patients, patients with adenocarcinoma or squamous cell carcinoma, and patients who had smoked ≤20 pack years (Table 4).

Table 4. Median Survival and Response Rate Among Subgroups
SubgroupNo.Overall survivalBest overall response
Median, months95% CIResponserate, %95% CI
  1. 95% CI, 95% confidence interval; NR, not reached; ECOG, Eastern Cooperative Oncology Group.

Age group
 18–64 y924.6(3.3–6.3)6.5(2.9–13.1)
 ≥65 y1377.8(5.8–9.7)9.5(5.2–15.3)
Sex
 Male1394.7(4.0–7.2)7.9(4.2–13.3)
 Female908.0(5.6–NR)8.9(3.9–16.1)
Histologic type
 Squamous cell carcinoma778.1(4.7–NR)9.1(4.3–17.7)
 Large cell carcinoma182.5(1.2–3.0)5.6(0.3–26.0)
 Adenocarcinoma1037.9(5.2–9.7)8.7(4.5–15.5)
 Other/unknown314.4(3.3–6.7)6.5(1.2–20.0)
No. of prior chemotherapy regimens
 11225.6(3.6–9.2)6.6(2.9–12.3)
 >11076.7(4.6–8.6)10.3(5.5–17.3)
ECOG performance status
 0–11599.2(6.7-NR)9.4(5.4–14.7)
 2–3703.0(2.2–3.5)5.7(2.0–13.7)
Pack-y of smoking (smokers only)
 ≤20417.0(4.2–9.7)17.1(7.6–30.7)
 >201675.0(4.0–7.2)3.6(1.6–7.5)

The median duration of treatment, a secondary efficacy outcome measure of this study, was 10.6 weeks; the 25% and 75% percentiles were 5.4 weeks and 21.0 weeks, respectively.

Safety and Tolerability

Overall, erlotinib was well tolerated. Eight patients (3.5%) died during the study because of an AE. One of these AEs, interstitial lung disease, was determined by the investigator to be related to therapy with erlotinib. The total incidence of nonfatal SAEs was 39.3% (Table 5). The total incidence of NCI-CTCAE grade 3 or 4 SAEs was 33.7%. The most common SAE was pneumonia (7.9%) (Table 5). Six SAEs (2.6%) were considered by investigators to be related to the study drug. Of these 6 patients, 3 (1.3%) had diarrhea, 1 (0.4%) had pneumonia, 1 (0.4%) had uveitis, 1 (0.4%) had a drug interaction, and 1 (0.4%) had dehydration.

Table 5. Incidence of Serious Adverse Events (≥1%, Excluding Fatal Adverse Events; n = 229)*
MedDRA AE system organ class and preferred termNo. of treated patients (%)Highest NCI-CTCAE grade, no. (%)
1234
  • NCI-CTCAE indicates National Cancer Institute Common Terminology Criteria for Adverse Events; MedDRA, Medical Dictionary for Regulatory Activities; AE, adverse events; SAE, serious adverse events; CNS, central nervous system; COPD, chronic obstructive pulmonary disorder.

  • *

    All data expressed as number (%).

  • Benign, malignant, and unspecified (including cysts and polyps).

Any SAEs-overall90 (39.3)3 (1.3)10 (4.4)51 (22.3)26 (11.4)
Gastrointestinal disorders
 Overall13 (5.7)1 (0.4)012 (5.2)0
 Diarrhea4 (1.7)1 (0.4)03 (1.3)0
General disorders and administration site conditions
 Overall7 (3.1)005 (2.2)2 (0.9)
Infections and infestations
 Overall28 (12.2)1 (0.4)3 (1.3)19 (8.3)5 (2.2)
 Pneumonia18 (7.9)1 (0.4)2 (0.9)10 (4.4)5 (2.2)
 Sepsis4 (1.7)01 (0.4)3 (1.3)0
Blood and lymphatic disorders
 Overall4 (1.7)003 (1.3)1 (0.4)
 Anemia3 (1.3)002 (0.9)1 (0.4)
Cardiac disorders
 Overall7 (3.1)01 (0.4)5 (2.2)1 (0.4)
Injury, poisoning, and procedural complications
 Overall4 (1.7)01 (0.4)3 (1.3)0
Metabolism and nutrition disorders
 Overall5 (2.2)02 (0.9)2 (0.9)1 (0.4)
 Dehydration3 (1.3)01 (0.4)1 (0.4)1 (0.4)
Neoplasms
 Overall8 (3.5)002 (0.9)6 (2.6)
 Metastasis to bone3 (1.3)0003 (1.3)
 Metastasis to CNS3 (1.3)001 (0.4)3 (0.9)
Nervous system disorders
 Overall9 (3.9)01 (0.4)5 (2.2)3 (1.3)
Psychiatric disorders
 Overall3 (1.3)01 (0.4)2 (0.9)0
Respiratory, thoracic, and mediastinal disorders
 Overall23 (10.0)2 (0.9)4 (1.7)9 (3.9)8 (3.5)
 Pleural effusion6 (2.6)0 (0.0)2 (0.9)4 (1.7)0 (0.0)
 Pulmonary embolism6 (2.6)0 (0.0)0 (0.0)1 (0.4)5 (2.2)
 COPD4 (1.7)1 (0.4)0 (0.0)3 (1.3)0 (0.0)
 Dyspnea3 (1.3)0 (0.0)1 (0.4)2 (0.9)0 (0.0)
 Hemoptysis3 (1.3)1 (0.4)1 (0.4)1 (0.4)0 (0.0)

Of the 229 treated patients, 209 patients (91.3%) discontinued taking erlotinib before the study ended. Disease progression accounted for nearly 60% of those discontinuations (Table 6). Approximately 19% of patients discontinued the study drug because of AEs and SAEs (Table 6). Two of these events (0.9%) (1 case of pneumonia and 1 case of dehydration) were suspected by the investigator to be related to erlotinib treatment. Diarrhea and rash were the most common treatment-related AEs leading to discontinuation (5.3%).

Table 6. Reasons for Discontinuation of Erlotinib (n = 229)
Reason for discontinuationNo. (%)
Disease progression136 (59.3)
Serious adverse events28 (12.2)
Patient withdrawals20 (8.7)
Sponsor's decision to terminate study20 (8.7)
Adverse events (nonserious)15 (6.6)
 Rash7 (3.1)
 Diarrhea5 (2.2)
 Other3 (1.3)
Physician's decision5 (2.2)
Clinical need for concomitant or ancillary therapy4 (1.7)
Lost to follow-up1 (0.4)

Of the 229 treated patients, 45 (19.7%) required an erlotinib dose reduction; 32 patients (14.0%) had a dose reduction to 100 mg/day, 11 patients (4.8%) had a dose reduction to 50 mg/day, and 1 patient (0.4%) had a dose reduction to 25 mg/day. For 1 patient, the lowest dose was not recorded.

DISCUSSION

This multicenter, open-label, phase 3b trial of erlotinib in patients with recurrent NSCLC was designed to examine the efficacy and safety of erlotinib monotherapy before its regulatory approval. Because of the approval of erlotinib in November of 2004, enrollment in the trial was prematurely terminated. Consequently, instead of the expected accrual of 5000 patients, only 233 were enrolled. At the time of this final analysis, 60% of patients had died and therefore the data are mature.

Overall, baseline characteristics were similar to those reported in the pivotal BR.21 trial, a placebo-controlled study of erlotinib in patients with advanced or metastatic NSCLC who had failed prior chemotherapy.14 Although compared with BR.21 this study had fewer younger patients, fewer patients who had never smoked, and fewer patients of Asian ethnicity, the final efficacy results from this trial are comparable to those of the BR.21 trial. In that study, the ORR was 8.9%, the median duration of survival was 6.7 months, and the median duration of treatment in the erlotinib arm was 9.6 weeks. Thus, it would appear that in this trial, which mimics a real-world clinical setting, erlotinib has an efficacy similar to that which has been observed in the randomized phase 3 clinical trial.

Although only 21 patients who had never smoked were enrolled in this trial, which is consistent with previous findings, erlotinib-treated patients who had never smoked survived longer and had a better response rate than current smokers or those who had smoked previously (see Table 2). The median duration of survival for those who had never smoked was not reached. The duration of survival was found to be similar between current and former smokers. These results are consistent with the findings from the controlled BR.21 study. Because of small patient numbers, we could not analyze subgroups defined by race and EGFR mutation status.

Exploratory analyses demonstrated a trend toward a longer median survival and a higher response rate in females, patients age ≥65 years, patients with adenocarcinoma or squamous cell carcinoma, and patients who smoked <20 pack-years. A similar trend for age groups and sex was observed on a smaller magnitude in the BR.21 study. Also consistent with the BR.21 study, the OS was found to be similar between patients with adenocarcinoma and squamous cell carcinoma. However, because of the small sample sizes, these analyses are descriptive in nature, and some of the 95% CIs are wide.

The results of the current study also demonstrated that erlotinib was well tolerated in this patient population. No new safety signals were noted in this trial. The incidence of erlotinib-related SAEs was low: 3.1% to 0.4% for fatal SAEs and 2.6% for nonfatal SAEs. The incidence of interstitial lung disease-like events (n = 1 patient; 0.4%) was consistent with that reported in other studies. The incidence of rash (3.1%) and diarrhea (2.2%) leading to drug discontinuation were also similar to those observed in other trials. Although the percentage of patients who discontinued erlotinib treatment because of SAEs or AEs was relatively high (18.8%), only approximately one-third of these events (6.6%) were determined to be related to erlotinib treatment.

An ongoing expanded access study of erlotinib conducted by Roche will enroll many more patients than the current study (approximately 5000 patients from outside the U.S. have been enrolled to date). That study, for which the authors are currently collecting tissue samples, will provide additional molecular data.

Ancillary