Clinomics and Its Role in Erlotinib Treatment of Non-Small-Cell Lung Cancer
A report on the BR.21 study published by Tsao et al. and Shepherd et al. in the July 14, 2005 issue of the New England Journal of Medicine*
Presented by Erika L. Artinger
MCB Graduate Program
Dartmouth Medical School
Genetics 144: Oncogenomics
Charles Brenner, Ph.D.
*An online subscription to the New England Journal of Medicine might be required to view figures.
Introduction
Clinomics is the use of genomic technologies to select individualized treatment therapies for patients (1). In cancer, for example, this could include testing a patient for a typical gene expression profile or known mutation prior to the selection of therapy. The results of testing would allow the caregiver to determine which available treatment option would be most likely to elicit a positive response in the patient. This would potentially reduce unnecessarily exposing a patient to ineffective drugs and their side effects.
In the BR.21 study published by Tsao, Shepherd and colleagues, tumors from cancer patients with non-small-cell lung cancer in a phase III clinical trial for the anti-cancer drug, erlotinib, were tested for a correlation with abnormalities in the epidermal growth factor receptor (EGFR). EGFR is an important receptor often found in non-small-cell lung cancer (NSCLC) because it is part of multiple signal transduction pathways that contribute to cancer-related cellular behaviors like proliferation and metastases (2, 7-8). The intended target of erlotinib is the intracellular tyrosine kinase (TK) domain of EGFR. Previous studies with a similar drug (gefitinib) that inhibits the same TK domain indicated that patients with mutations in the EGFR gene were more likely to respond to treatment (4,5). If a similar correlation to EGFR abnormalities is found with erlotinib treatment, clinicians would potentially be able to predict which NSCLC patients would most likely benefit from erlotinib treatment.
The Process and Expense of Clinical Trials
Clinical trials of drug therapies are conducted in three
phases:
Phase I: The objective of phase I is to determine the
safety of the compound being tested. It is the first phase in which humans are
the primary test organism and typically lasts several months. Small groups of
20 to 80 healthy participants are given the drug and monitored for toxicity and
acceptable dosage. Seventy percent of drugs pass and continue on to the second
phase of clinical trials, phase II (6).
Phase II: The purpose of phase II is to optimize
dosage and demonstrate effectiveness. It usually involves larger cohorts of
patients (100-300) and can last up to two years. Only 33% of drugs in phase II
pass and continue to phase III (6).
Phase III: Phase III is the final phase and involves
a large test population of 1000 to 3000 patients to demonstrate drug efficacy
and safety. It can last up to four years and only 25% to 30% of drugs that
make it to this phase pass and receive United States Food and Drug
Administration approval (6).
The process of developing a drug and seeing it through all
three phases of clinical trials is lengthy and expensive. US pharmaceutical
companies are investing increasing amounts of money to develop drugs. Over the
period of 10 years from 1993 to 2003, the amount of money spent by
pharmaceutical companies on drug research and development has consistently
grown at a faster rate than the National Institute of Health budget (7). During the same period, however, the number of drug and biological product
submissions to the FDA from pharmaceutical companies has substantially
decreased (7). Thus, while drug companies are investing increasing amounts of
money to find new treatments for disease, fewer drugs are making it through the
process of clinical trials to be approved by the FDA. The average investment
required for one successful drug launch has increased from $1.1 billion in
1995-2000 to $1.7 billion in 2000-2002, approximately half of which is spent on
clinical trials (Figure 1 below) (7). For comparison, the entire budget of the
NIH for the year 2006 was $29 billion (8).
Figure 1: Average Cost of a Drug Launch
Non-Small-Cell Lung Cancer
Lung carcinoma is the leading cause of cancer death in the United States (more than breast, prostate, and colon cancer combined) (10). Eighty percent of lung cancers are non-small-cell cancers. In 2004, there were 147,000 cases of NSCLC in the United States alone (11). There are four types of common lung cancers: adenocarcinoma, squamous cell carcinoma, large cell undifferentiated carcinoma, and bronchioaveolar carcinoma. The most common of these is adenocarcinoma, which comprises about 40% of NSCLCs (10). Patients diagnosed with adenocarcinoma are mostly non-smokers and have tumors near the edge of lung that can often spread to the chest lining and other body organs (10). About 30% of adenocarcinomas display increased EGFR expression (10). Adenocarcinomas comprised approximately half of the patients in the BR.21 study (Shepherd Table 1).
There are five stages of NSCLC development (10):
Stage 0 is defined as a single tumor contained within its location of origin in the lung.
Stage I limits cancer cells to the lung.
Stage II is defined by some spread to lymph nodes in the lung, chest wall or lining of the lungs.
Stage III includes spread to lymph nodes outside the lung and nearby locations like the neck and heart.
Stage IV is the final stage of development and is characterized by distant spread of cancer cells to other regions of the body.
All of the patients included in the BR.21 study were diagnosed with stage III or stage IV NSCLC (2). At these late stages of carcinogenesis, tumors are usually inoperable and the average life expectancy with chemotherapy and radiation treatments is 14 to 18 months (14).
Treatment options for patients with recurrent, late stage NSCLC include Docetaxel (Taxotere), Pemetrexad (Alimta), and the tyrosine kinase inhibitors Gefitinib (Iressa) and Erlotinib (Tarceva) (11). Erlotinib was approved by the FDA for treatment of NSCLC after the failure of at least one prior chemotherapy regimen on November 8, 2004 on the basis of the BR.21 study results.
Epidermal Growth Factor Receptor and NSCLC
The epidermal growth factor receptor (EGFR) is overexpressed in 40-80% of NSCLC (12) and correlates with decreased survival (16). EGFR functions by binding ligands (including epidermal growth factor (EGF) and transforming growth factor alpha (TGF-alpha)) and forming either homodimers or heterodimers with other members of the EGFR subfamily of receptors (like HER2/neu) (see Figure 2 below) (13). After dimerization, the cytoplasmic tyrosine kinase domains of EGFR autophosphorylate and activate signal transduction cascades that result in pro-growth behaviors (13). Increased EGFR expression results in increased invasion, metastasis, cell proliferation, maturation, angiogenesis, and inhibition of apoptosis (15).
Figure 2: Epidermal Growth Factor Receptor Function
Ten studies of EGFR mutations (located in exons 18-21
of the EGFR gene) in NSCLC have indicated that they are most common in patients
of Asian descent, females, patients with adenocarcinomas, and never-smokers (see
Figures 3-5 below) (17, 20).
Figure 3 (17)
Figure 5 (17)
Erlotinib
Erlotinib hydrochloride (clinical name Tarceva, produced by OSI Pharmaceuticals) is a small-molecule inhibitor of the Her1/EGFR tyrosine kinase (see Figure 6 below) (10). It is 6,7-bis(2-methoxy-ethoxy)-quinazolin-4-yl-(3-ethylnylphenyl)amine with chemical formula C22H23N3O4HCl (20). The intended mechanism of erlotinib is to inhibit autophosphorylation of the cytoplasmic tyrosine kinase domains of EGFR and thus prevent activation of the downstream signal transduction pathways, stopping growth (see Figure 2 above) (10).
Figure 6: Chemical Structure of Erlotinib
Earlier Erlotinib Clinical Trials (does not include combined clinical trials of erlotinib with other anti-cancer therapies):
Preclinical Studies:
In vitro studies with purified EGFR showed that Erlotinib inhibited EGFR at nanomolar concentrations and was highly specific (1000 fold higher inhibition of EGFR than other tyrosine kinases) (21). Erlotinib completely blocked EGF-induced EGFR autophosphorylation in in vivo studies with human HN5 cell xenografts in athymic mice (22). The xenografts also displayed substantial growth inhibition with erlotinib treatment (22).
Phase I Studies:
Healthy volunteers were given 10 to 1000 mg daily doses of erlotinib and displayed minimal side effects (23). Side effects were mostly observed at high doses and included headache, diarrhea, mild erythematous rash, and maculopapular dermatitis (23). Two other phase I studies were conducted on patients with treatment-refractory cancers and similar toxicity results were obtained (23, 24). Some of the patients in these studies displayed modest disease response to erlotinib (24). The minimal steady state dose to achieve inhibition was determined to be approximately 1.2 micrograms/ml with a half life of about a day (23). As a result, a dose of 150 mg erlotinib per day was recommended for future trials (20).
Phase II Studies:
Two studies on patients with NSCLC were performed with
encouraging results. The first study followed 57 patients with advanced or
recurrent NSCLC that had progressed after previous chemotherapy treatment (25). Patients were given 150 mg of erlotinib daily. A significant portion of
patients displayed response or disease stability improvement with treatment. The most positive responding patients all developed a rash. The second study
was restricted to 69 NSCLC patients with bronchioalveolar carcinoma given the
same size dose of erlotinib (26). A quarter of patients displayed a partial
response.
The phase II studies referenced above and other combined
erlotinib phase II and phase III trials demonstrated that erlotinib
responsiveness is most frequent in females, patients of Asian descent,
never-smokers, and patients with adenocarcinoma or BAC (20).
The BR.21 Clinical Study (Phase III Clinical Trial)
(NCIC CTG NCT00036647)
BR.21, the clinical trial investigated in the Tsao et al. and Shepherd et al. studies, was a phase III, randomized, placebo controlled, double blind trial conducted by the National Cancer Institute of Canada Clinical Trials Group (2,3). Seven-hundred and thirty-one stage IIIB or IV NSCLC patients who had previously received at least one or two regimens of chemotherapy and were not eligible to receive further chemotherapy treatment were given 150 mg of erlotinib per day or placebo in a ratio of 2:1 (2). Patients were from treatment centers all over the world (Figure 7 below). The BR.21 study was significant because it was the first randomized study of a drug targeting a tyrosine kinase domain to show substantial improvement in response and survival of NSCLC patients (9).
Figure 7: Map of Patient Origins (18)
Ethical Concerns for the BR.21 Study:
In a letter to the editor of the October 20, 2005 issue of the New England Journal of Medicine, Chadi Naban, M.D. and Jacob D. Bitran, M.D. raised concerns that it was unethical to give placebo to patients that had only undergone one prior chemotherapy regimen (19).The basis for this objection was a previous study reporting that docetaxel (clinical name Taxotere, Aventis Pharmaceuticals) treatment was superior to the best supportive care after first-line chemotherapy (27). Thus, patients in the placebo group (receiving no therapy) were not treated ethically because they were not given the best supportive care available.
In response, Shepherd et al. claim that all patients in the
BR.21 study that had only undergone one round of chemotherapy were evaluated by
their doctors as ineligible for second-line chemotherapy (19). Therefore, the decision
of the authors to include these patients in a placebo group was ethically
justified because no other course of treatment was deemed suitable for them.
Shepherd et al. Paper:
Clinical Results of the BR.21 Study
Erlotinib in Previously Treated Non-Small-Cell Lung Cancer
Shepherd FA, Pereira JR, Ciuleanu T, et al.
NEJM
2005 353:123-32
Purpose:
Shepherd et al. analyzed the BR.21 study to investigate
whether erlotinib prolonged survival or increased response in NSCLC patients
who had already failed at least one regimen of chemotherapy.
Baseline Characteristics of Clinical Group:
The study group of 731 patients was divided in a 2:1 ratio of patients treated with erlotinib (N=488) and placebo (N=243) (2). These two groups had comparable baseline characteristics (Shepherd Table 1) in every category compared except for the percent of patients above age 60 (however, the average age between the two groups only differed by 3 years) (2). The average age for the whole study was 61.4 years (2). Half of the patients had received two previous chemotherapy regimens, two-thirds were male, 12% were of Asian descent, half were diagnosed with adenocarcinoma, and three-quarters of the patients were current or previous smokers (2). Shepherd et al. also analyzed patient tissues for EGFR expression to correlate with the response data, but Tsao et al. more comprehensively studied that data and the results will be reserved for the Tsao et al. discussion below.
Statistical Analysis:
Baseline population characteristics like age, sex, and tumor
subtype in addition to EGFR expression were compared in a stratified log-rank
test and Cox regression analysis was used to identify factors related to
prolonged, progression-free survival (2).
A mini
guide to interpreting data on Kaplan-Meier curves:
P-value: The P-value is the probability of the null hypothesis (that the two curves are identical) is true. A P-value of 1 would indicate that there is a 100% probability that both the placebo curve and the test curve are the same. A P-value of 0.01 would indicate that there is only a 1% chance that the null hypothesis of both populations experiencing the same rate of survival is true.
Hazard: The hazard is calculated as the slope of the Kaplan-Meier curve and is equal to the rate of dying.
Hazard ratio: The hazard ratio is the ratio of the hazard of the test group to the hazard of the placebo group. It defines the relative risk of dying between the two groups. While the hazard values of the two curves may differ at different points on the graphs, it is assumed that their ratio will remain constant over time. A hazard ratio of 0.5 would indicate that the relative risk of dying of the test group is half of the risk of dying in the placebo group.
Censoring: A patient may be removed from continuing survival analysis for reasons other than NSCLC caused death (for example, treatment discontinued due to side effects or death from another unrelated cause). Cox regression statistically treats the removal of these patients by censoring them. When they leave the analysis, their absence does not affect the % survival. Thus, the number of patients may decrease at a specific time point without causing a dip in the Kaplan-Meier graph. In the BR.21 study, 5% of patients discontinued erlotinib treatment because of toxicity (2).
More information on Cox Regression
Univariate vs. Multivariate analysis:
Tsao et al. and Shepherd et al. used both univariate and multivariate analysis to analyze response and survival in the BR.21 study (2, 3). Univariate analysis is statistical analysis that involves looking for correlations between data with only one variable at a time (for example: survival by gender). Multivariate analysis, on the other hand, looks for interrelationships between multiple variables at a time.
Results:
Response (as scored by RECIST (28)):
Shepherd et al. found that erlotinib-treated patients demonstrated a measurable response 8% more often than placebo-treated patients (P<0.001) (Shepherd Table 2). Notably, females, adenocarcinoma patients, Asians, and non-smokers had the best response rates. This confirms previous studies mentioned above that found similar erlotinib responsiveness in these patient subgroups.
Survival:
The hazard ratio for overall survival of erlotinib treatment compared with placebo was 0.7 (P<0.001) (Shepherd Figure 1A) (2). Overall survival was increased from 4.7 months on placebo to 6.7 months with erlotinib treatment. When the data is constricted to include only progression-free survival, the relative risk of dying of the Erlotinib group decreases further to 0.61 (P<0.001) (Shepherd Figure 1B).
In univariate analysis, Shepherd et al. confirmed that erlotinib was correlated to longer overall survival (P=0.002) (Shepherd Table 3) and found that adenocarcinoma, never smoking, and Asian descent were also correlated to survival (Hazard ratios 0.8, 0.8, and 0.7, respectively). In multivariate analysis, only smoking status and erlotinib treatment had enough statistical significance to be predictive of increased survival.
Quality of Life:
Patients receiving erlotinib had improved quality of life based on survey questions that included cough, pain, dyspnea, and overall physical function (Shepherd Supplemental Appendix).
Conclusions:
The 8.9% response rate to erlotinib treatment and increased overall survival justify the use of erlotinib as second or third-line treatment for NSCLC.
Tsao et al. Paper: The
Molecular Results of the BR.21 Study
Erlotinib in lung cancer - molecular and clinical predictors of outcome
Tsao M-S, Sakurada A, Cutz JC, et al.
Purpose:
Tsao et al. set out to answer the following questions (18, 3):
1. Will mutation analysis improve the ability of the clinician to predict response in patients treated with erlotinib?
2. Does mutation status predict for a differential effect on survival from erlotinib treatment?
3. Do patients without mutations benefit from erlotinib therapy?
Facts to remember:
Previous studies demonstrated that erlotinib responsiveness is most frequent in females, patients of Asian descent, never-smokers, and patients with adenocarcinoma or BAC (20).
Previous studies of EGFR mutations (in exons 18-20) in NSCLC
have indicated that they are most common in patients of Asian descent, females,
patients diagnosed with adenocarcinomas, and never-smokers (see Figures 4-6 above)
(17, 20).
Strategy:
Tsao et al. set out to analyze EGFR expression and mutational status in the BR.21 patients and determine if EGFR abnormalities correlated with response to erlotinib or increased survival with erlotinib treatment (3).
Methods:
Immunohistochemical (IHC) analysis:
To detect the expression of EGFR in tumor cells, paraffin blocks or unstained slides of tissue sections from 325 patients were stained with antibodies to EGFR with Dako EGFR PharmDx kits. Sections were considered positive when greater than 10% of the tumor cells showed partial or complete staining of EGFR. Tsao Figures 2A and 2B show two examples of positive staining (EGFR antibody is stained brown with DAB).
The baseline characteristics of patients in the subgroup of patients that were tested with IHC were compared with those of the entire study and found to be mostly comparable (Tsao Table 1). Significant differences were visible, however, in the proportion of patients with Asian descent (P<0.001), performance status (P=0.04), number of prior regimens (P=0.001), and time from diagnosis to randomization (P=0.01) in the subgroup.
Fluorescence In Situ Hybridization (FISH) analysis:
To detect polysomies (abnormal chromosome copy number) and amplifications (increased copies of a gene) of the EGFR locus in tumor cells, FISH was performed on tissue sections from 125 patients with probes to EGFR (Spectrum Orange) and CEP7 (Spectrum Green), a marker for the centromere of chromosome seven. EGFR is located on chromosome 7 so comparing the signals of CEP7 to EGFR will reveal the ratio of EGFR genes per chromosome (amplifications) and comparing the number of CEP7/EGFR signals to the number of cells will reveal number of chromosomes per cell (polysomies). Tsao Figure 2C is an example of FISH results in a diploid tumor cell that has a 1 EGFR : 1 CEP7 signal ratio in the right proportion to the number of cells. Tsao Figure 2D is an example of an amplification in tumor cells because there is an excess of EGFR signals. High polysomy was defined as greater than or equal to 4 copies of EGFR in at least 40% of cells (3). Amplification was defined as a gene:chromosome ratio of at least 2 or at least 15 gene copies per cell in greater than or equal to 10% of cells (3).
The baseline characteristics of patients in the subgroup of patients that were tested with FISH were compared with those of the entire study and found to be mostly comparable (Tsao Table 1). Differences were visible, however, in the proportion of patients with Asian descent (P=0.03), smoking history (P=0.02), number of prior regimens (P<0.001), and best response to prior chemotherapy (P=0.01).
Mutational analysis:
To determine the mutational status of the EGFR gene in exons 18-21 (regions previously shown to harbor significant mutations for TK inhibitors) (4,5), the authors used tumor samples from 177 patients, digested them with proteinase K, isolated DNA with phenol-chlorophorm, and PCR (polymerase chain reaction) amplified the significant exons with internal and external primers previously used in the gefitinib study by Paez et al. (5). To get the most homogenous tumor sample possible for amplification, the authors occasionally utilized microdissection and laser capture microdissection techniques. The PCR products were purified and sequenced. Sequence variations found in greater than 15% of specimens were considered significant and included in subsequent statistical analysis.
The baseline characteristics of patients in the subgroup of patients that were tested for EGFR mutations were compared with those of the entire study and found to be mostly comparable (Tsao Table 1). Differences were visible, however, in the proportion of patients with Asian descent (P=0.01), number of prior regimens (P<0.001), best response to prior chemotherapy (P=0.003), and decrease in body weight (P= 0.03).
More Information about:
Results:
EGFR tests as a group:
The subpopulation of patients that were analyzed with at least one EGFR test showed a similar benefit with erlotinib treatment to the entire BR.21 study as a whole (hazard ratio of EGFR tested patients was 0.76, P=0.03 compared to 0.70, P<0.001) (Tsao Figure 1A and 1B).
IHC:
After immunohistochemical analysis, the authors found that 57% of tumors displayed positive EGFR expression (Tsao Table 2). Multivariate analysis showed an increase in responsiveness (scored by RECIST (28)) in EGFR positive tumors (11% versus 4% in negative patients) but it was not statistically significant (P=0.1) (Tsao Table 3). Similarly, survival increased in positively staining patients (hazard ratio in positive patients was 0.68 versus 0.70 in all patients, and 0.93 in EGFR negative patients) but was not statistically significant (Tsao Table 3, Tsao Figures 1A, 1C and 1D).
FISH:
After FISH analysis, the authors found that 34% of tumors analyzed had high polysomy and 11% displayed EGFR gene amplification (Tsao Table 2). EGFR-amplified tumors responded in 20% of cases (compared to 2% in non-amplified tumors) which was statistically significant (P=0.03), unlike the IHC results (Tsao Table 3). Longer survival was also correlated to amplification and statistically significant. The hazard ratio for EGFR amplified patients was 0.44, compared to 0.7 for all patients and 0.85 for non-amplified patients (P values = 0.008, <0.001, and 0.59, respectively) (Tsao Table 3, Tsao Figures 1A, 1E and 1F).
Mutation Analysis:
The authors found 45 distinct EGFR mutations in the tumors of 40 patients (23% of patients tested) (Tsao Table 2). The demographics of the patients with mutations were consistent with those previously described (most prevalent in women, never-smokers, Asians*, and adenocarcinomas*). Twenty-nine percent of the mutations were deletions in exon 19 and 18% of mutations were identified as the L858R mutation in exon 21 previously identified in responses to gefitinib (Tsao Figure 3A) (3, 4).
* Statistically significant
Unlike the Paez and Lynch studies of gefitinib, Tsao et al. did not find a statistically significant correlation of EGFR mutational status to responsiveness or survival with erlotinib treatment. Mutants responded in 16% of patients analyzed, compared to a 7% response in wild type tumors but had a non-significant P value of 0.37 (Tsao Table 3). The hazard ratios for mutants, wild type tumors, and all patients were 0.77, 0.73, and 0.70, respectively, with individual P values of 0.13, 0.45, <0.001 and an overall P value for the mutational status interaction with erlotinib of 0.97 (Tsao Table 3 and Figures 1A, 1G and 1H).
Conclusions:
While Tsao et al. found increased responsiveness in all the tumors that tested positive for EGFR abnormalities with erlotinib, the only statistically significant correlation was within the high polysomy or amplification subgroup identified by FISH analysis. The authors also found prolonged survival in the EGFR positive tumors identified by IHC and the high polysomy and amplification subgroup identified by FISH. Only the FISH tumors were statistically significant, however. Surprisingly, the authors found no survival benefit for the tumors found to contain mutations in the EGFR locus. On the basis of this study, the authors cannot recommend mutation analysis for patients considering erlotinib treatment because, unlike the gefitinib studies, it does not have significant predictive power for positive response to treatment.
Responses to Shepherd et al. and Tsao et al.
Several letters to the editor regarding the BR.21 papers have been published in the New England Journal of Medicine since the original publication of the papers in July of 2005. A brief summary of some of the letters and the author replies is below:
Pao W, Ladanyi M, Miller V. October 2005 (19):
Pao and colleagues contest that it is inappropriate to draw conclusions from the mutations categorized by Tsao et al. for several reasons: Other studies suggest a European and North American EGFR mutation rate of 10%, which is lower than the 23% reported by Tsao et al. (29, 30, 3). Also, Tsao et al. reported exon 19 deletions and L858R substitutions in only 47% of mutations, which is substantially lower than the 90% reported elsewhere (31). Ninety-two percent of the mutations that were not among the previously identified exon 19 deletions and L858R substitutions were transitions, which could indicate potential PCR artifacts (32). If only 47% of the Tsao et al. reported mutations are real and associated with erlotinib responsiveness, statistical analysis would have to be restricted to these mutations to show a significant correlation.
Tsao M-S, Kamel-Reid S, and Shepherd FA Response (19):
The rate of mutation for just the exon 19 deletions and
L858R substitutions was 11%, in accordance with the above published 10% EGFR
mutation rate. Responsiveness in these classic mutations was only 25% in the
BR.21 study (2 of 8 tumors) and did not show a greater survival benefit (hazard
ratio for classic mutations was 0.67, compared to 0.65 for novel mutations and
0.73 for wild type EGFR).
In previous studies, PCR artifacts were found to increase
with decreases in the amount of cells used for analysis (33). However, Tsao et
al. found the opposite relationship with their novel mutations.
Marchetti
A, Felicioni L, Buttitta F. February 2006 (34):
Marchetti and colleagues examined PCR artifacts more extensively. They repeatedly (10x) PCR amplified 70 samples of DNA from a single paraffin-embedded lung tumor section in exons 18 Ò 21 of the EGFR locus. They found all 22 transitions in the ÏnovelÓ mutations reported by Tsao et al. and identified them as artifacts. They were also able to find the same mutations in DNA samples from normal tissue (Marchetti Figures 1A and 1B) and propose that postmortem deamination of cytosine or adenine could be the source of the artifacts.
Tsao M-S, Kamel-Reid S, and Shepherd FA Response (34):
The authors successfully added 25 additional BR.21 samples to their mutational assay and reanalyzed all data only in regards to the exon 19 deletions and L858R substitutions. The results of this new analysis were consistent with all their previous analyses and still showed no statistically significant correlation (Tsao Response Figures 1A and 1B).
Discussion
The absence of a correlation between mutations in the EGFR locus and response to erlotinib treatment reported by Tsao et al. is not entirely convincing, especially in the light of potential PCR artifacts. In addition to the PCR concerns raised by Pao and Marchetti, the sample size of tissues actually suitable for mutation analysis in the Tsao et al. paper was small enough that the conclusiveness of the results can be questioned. However, even if a more conclusive correlation between mutation status and drug response in erlotinib were established, it still may not be beneficial to recommend mutational analysis of patients prior to the selection of second and third line treatment. The clinical demographics of gender, ethnicity, smoking status, and histological subtype may be strong enough predictors of erlotinib response without having to order a mutational analysis test. The clinical demographics also have the added benefit of being inexpensive to evaluate and immediately available to the clinician for treatment decisions. If the average life expectancy of a patient in stage III NSCLC that has already failed previous regimens of chemotherapy is only 6.7 to 4.7 months (with and without erlotinib treatment) and quality of life improves on erlotinib (2), saving a week of waiting for test results prior to beginning treatment may outweigh the benefits of mutational analysis.
References
1. Brenner, Charles, ed., and
David Duggan, ed. Oncogenomics:
Molecular Approaches to Cancer. Hoboken, New Jersey: John Wiley & Sons,
Inc., 2004. [Chapter 15 Clinomics: Postgenomic Cancer Care, pg. 341-358].
2. Shepherd FA, Pereira JR, Ciuleanu T, Tan EH, Hirsh V, Thongprasert S, Campos D, Maoleekoonpiroj s, Smylie M, Martins R, van Kooten M, Dediu M, Findlay B, Tu D, Johnston D, Bezjak A, Clark G, Santabarbara P, Seymour L. Erlotinib in Previously Treated Non-Small-Cell Lung Cancer. N Engl J Med, 2005; 353: 123-32.
3. Tsao M-S, Sakurada A, Cutz J-C, Zhu C-Q, Kamel Reid S, Squire J, Lorimer I, Zhang T, Liu N, Daneshmand M, Marrano P, da Cunha Santos G, Lagarde A, Richardson F, Seymour L, Whitehead M, Ding K, Pater J, Shepherd FA. Erlotinib in lung cancer molecular and clinical predictors of outcome. N Engl J Med 2005; 353:133-44.
4. Lynch, T.J., Bell, D.W., Sordella, R., Gurubhagavatula, S., Okimoto, R., Brannigan, B.W., Harris, P.L., Haserlat, S.M., Supko, J.G., Haluska, F.G., Louis, D.N., Christiani, D.C., Settleman, J., Haber, D.A., 2004. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N. Engl. J. Med., 350(21): 2129-2139.
5. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004; 101:13306-11.
6. Mayo Clinic Staff. Clinical Trials: A Chance to Try Evolving Therapies. July 8, 2005. http://www.mayoclinic.com/health/clinical-trials/DI00033
7. Innovation or Stagnation? Challenge and opportunity on the critical path to new medical products. Food and Drug Administration, United States Department of Health and Human Services. March 2004. Http://www.geospiza.com/media-v4/white-papers/ fda-innovation-whitepaper.pdf
8. Summary of the FY 2006 PresidentÌs Budget. February 7, 2005. http://www.nih.gov/news/budget/FY2006presbudget.pdf
9. Doroshow, Targeting EGFR in Non-Small-Cell Lung Cancer. Editorial. N Engl J Med 2005. 353;2 200-1.
10. Learning about Lung Cancer, Considering Tarceva. Accessed February 2006. http://www.tarceva.com
11. Frankly speaking about lung cancer. Accessed February 2006. http://www.thewellnesscommunity.org/programs/frankly/lung/lung_cancer_home.asp.
12. Arteaga CL. ErbB-targeted therapeutic approaches in human cancer. Exp Cell Res 2003; 284:122-30
13. Jorissen RN, Walker F, Pouliot N, Garrett TP, Ward CW, Burgess AW. Epidermal growth factor receptor: mechanisms of activation and signaling. Exp Cell Res 2003; 284:31-53
14. Machtay M. Progression of Non-Small-Cell Lung Cancer. Abramson Cancer Center of the University of Pennsylvania. Accessed February 2006. http://www.oncolink.org/experts/article.cfm?c=3&s=19&ss=50&id=1708
15. Dowell J, Minna JD, Kirkpatrick P. Fresh from the Pipeline: Erlotinib hydrochloride. Nature Reviews Drug Discovery. Published Online: 21 December 2004; doi:10.1038/nrd1612.
16. Nicholson RI, Gee JMW, Harper ME. EGFR and cancer prognosis. Eur J Cancer 2001; 37: Suppl 4:S9-S15
17. Slides from Giuseppe GiacconeÌs presentation at the 2005 ASCO meeting
Giaccone, Giuseppe. Clinical
Trials and the Biology of EGFR Blocking Agents. 2005 ASCO Annual Meeting.
18. Questions are a direct
quotation from Dr. TsaoÌs presentation to ASCO, slide is directly from his
presentation. Tsao M-S. Molecular analysis
of the epidermal growth factor receptor (EGFR) gene and protein expression in
patients treated with erlotinib in National Cancer Institute of Canada Clinical
Trials Group (NCIC CTG) trial BR.21. 2005 ASCO meeting. 19. Nabhan, C., Bitran,
J.D., Takano, T., Ohe, Y., Pao, W., Ladanyi, M., Miller, V.A., Shepherd, F.A.,
Seymour, L., Tsao, M.S., Kamel-Reid, S., Shepherd, F.A. Correspondence:
Erlotinib in Lung Cancer. N Engl J Med 353; 16. 20. Brown, ER, Shepherd, FA. Erlotinib in the treatment of non-small-cell lung cancer. Expert Rev Anticancer
Ther. 2005 Oct;5(5):767-75. 21. Moyer JD, Barbacci EG,
Iwata KK et al. Induction of apoptosis and cell cycle arrest by CP-358,774, an
inhibitor of epidermal growth factor tyrosine kinase. Cancer Res. 57(21),
4838-4848 (1997). 22. Pollack V, Savage D,
Baker DA et al. Inhibition of epidermal growth factor receptor-associated
tyrosine phosphorylation in human carcinomas with CP-358,774: dynamics of
receptor inhibition in situ and antitumor effect in athymic mice. J. Pharmacol. Exp. Ther. 291, 739-748 (1999). 23. Hidalgo M, Siu LL,
Nemunaitis J et al. Phase I and pharmacological study of OSI-774, an epidermal
growth factor receptor tyrosine kinase inhibitor, in patients with advanced
solid malignancies. J. Clin. Oncol. 19 (13), 3267-3279 (2001). 24. Karp D, Silberman R,
Csudae R et al. Phase I dose escalation study of epidermal growth factor
receptor (EGFR) tyrosine kinase (TK) inhibitor CP-358,774 in patients with
advanced solid tumors. Proc. Am. Soc. Clin. Oncol. 18, 388a (1999) (Abstract
1499). 25. Perez-Soler R, Chachoua
A, Hammond L et al. Determinants of tumor response and survival with erlotinib
in patients with non-small-cell lung cancer. J. Clin. Oncol. 22(16), 3238-3247
(2004). 26. Patel JD, Miller VA,
Kris MG et al. Encouraging activity and durable responses demonstrated by the
EGFR tyrosine kinase inhibitor erlotinib (Tarceva TM, OSI-774) in patients with
advanced bronchioalveolar (BAC) cell carcinoma. Lung Cancer 41, S56 (2003). 27. Shepherd FA, Dancey J,
Ramlau R, et al. Prospective randomized trial of docetaxel versus best
supportive care in patients with non-small-cell lung cancer previously treated
with platinum-based chemotherapy. J.
Clin. Oncol. 2000; 18:2095-103. 28. Therasse P,
Arbuck SG, Eisenhauer EA, et al. New
guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer
Inst 2000;92:205-16. 29. Eberhard DA, Johnson BE,
Amler LC, et al. Mutations in the epidermal growth factor receptor and in KRAS
are predictive and prognostic indicators in patients with non-small-cell lung
cancer treated with chemotherapy alone and in combination with erlotinib. J
Clin Oncol 2005; 23:5900-9. 30. 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 USA 2004: 101:13306-11. 31. Pao W, Miller VA.
Epidermal growth factor receptor mutations, small molecule kinase inhibitors,
and non-small-cell lung cancer: current knowledge and future directions. J Clin
Oncol 2005; 23:2556-68. 32. Wong C, DiCioccio RA,
Allen HJ, Werness BA, Pivers MS. Mutations in BRCA1 from fixed,
paraffin-embedded tissue can be artifacts of preservation. Cancer Genet
Cytogenet 1998; 107: 21-7. 33. Cappuzzo F, Hirsh FR,
Rossi E, et al. Epidermal growth factor receptor gene and protein and gefitinib
sensitivity in non-small-cell lung cancer. J Natl Cancer Inst 2005;97:643-55. 34. Marchetti A, Felicioni
L, Buttitta F, Tsao M-S, Kamel-Reid S, Shepherd FA. Correspondence: Assessing
EGFR Mutations. N
Engl J Med 2006. 354;5 526-8.