- Research article
- Open Access
- Open Peer Review
Response to gefitinib and erlotinib in Non-small cell lung cancer: a retrospective study
© Emery et al; licensee BioMed Central Ltd. 2009
- Received: 12 February 2009
- Accepted: 18 September 2009
- Published: 18 September 2009
In Non-small cell lung cancer (NSCLC), an overactive epidermal growth factor receptor (EGFR) pathway is a component of the malignant phenotype. Two tyrosine kinase inhibitors (TKIs) of EGFR, gefinitib and erlotinib, have been used with variable benefit.
We have analyzed outcome data of a population of NSCLC patients that received these TKIs to determine the benefit derived and to define the clinical and molecular parameters that correlate with response. Tumor tissue from a subgroup of these patients was analyzed by immunohistochemistry to measure the expression level of EGFR and four activated (phosphorylated) members of the pathway, pEGFR, pERK, pAKT, and pSTAT3.
Erlotinib was slightly superior to gefitinib in all measures of response, although the differences were not statistically significant. The most robust clinical predictors of time to progression (TTP) were best response and rash (p < 0.0001). A higher level of pEGFR was associated with longer TTP, while the total EGFR level was not associated with response. Higher levels of pAKT and pSTAT3 were also associated with longer TTP. In contrast, a higher level of pERK1/2 was associated with shorter TTP.
These observations suggest the hypothesis that tumor cells that have activated EGFR pathways, presumably being utilized for survival, are clinically relevant targets for pathway inhibition. An accurate molecular predictive model of TKI response should include activated members of the EGFR pathway. TKIs may be best reserved for tumors expressing pEGFR and pAKT or pSTAT, and little pERK. In the absence of molecular predictors of response, the appearance of a rash and a positive first scan are good clinical indicators of response.
- Epidermal Growth Factor Receptor
- Epidermal Growth Factor Receptor Expression
- Epidermal Growth Factor Receptor Pathway
Current chemotherapy combinations for advanced non-small cell lung cancer (NSCLC) have reached a plateau in overall response rate and are never curative. Response rates for first line and second line treatments are 35% and 8%, respectively, and time to progression averages four to six months [1–3]. New therapeutic agents and strategies to maximize the efficacy of current treatments are clearly needed.
In addition to classic cytotoxic agents, approved systemic therapies to treat NSCLC include inhibitors of the epidermal growth factor receptor (EGFR) pathway. The EGFR pathway is a principal transducer of growth and survival signals in lung cancer cells and therefore a logical target for therapy. Gefitinib and erlotinib are reversible inhibitors of the kinase domain of EGFR that compete with ATP for binding to the catalytic pocket. These small molecules inhibit EGFR autophosphorylation and, thus, they inhibit receptor dimerization, and the downstream signaling that would have otherwise stimulated proliferation (through the activation of Erk) and anti-apoptotic mechanisms (through activation of Akt and Stat).
TKIs were developed after accumulating evidence indicated that a large proportion of NSCLC over-expresses the EGF receptor and is likely to be dependent on this pathway for the malignant phenotype . However, TKIs in NSCLC therapy have not produced the gains in survival and time to progression that were anticipated when used alone or in combination with traditional cytotoxic agents [2, 3, 5]. Given this limitation, one way to increase the effectiveness of these agents is to rationally select patients based on tumor markers that predict response.
The identification of molecular predictors of response to anti-EGFR agents has been difficult. Unlike trastuzumab, whose clinical activity can be predicted by the evidence of over-expression of its target, the ErbB2 receptor, either by fluorescence in situ hybridization (FISH) or immunohistochemistry (IHC) , single markers have not produced accurate predictive models of response for anti-EGFR agents such as gefitinib and erlotinib. Early results on the relationship between EGFR protein overexpression or egfr gene amplification and response were contradictory and recent large studies indicate that there is no correlation between high EGFR expression by IHC or amplification by FISH and better response to TKIs [5, 7–9]. Specific mutations in the egfr gene have correlated with the dramatic responses that are sometimes seen in a small percentage of patients treated with these agents. However, these results are confounded by the prognostic value of the mutations, since it has been noted that these patients have a better prognosis, regardless of treatment used, than patients whose tumors do not bear egfr mutations . Moreover, the mutations do not identify the large number of patients that achieve more modest responses.
We evaluated data from our lung cancer patients to compare the benefits derived from erlotinib and gefitinib and to identify the clinical parameters that correlated with response. We subsequently analyzed tumor tissue from a subgroup of these patients for a pilot study to evaluate the utility of a combination of markers, downstream from the inactivated EGFR receptor, to predict response to TKIs and correlate outcomes. We based our design on the hypothesis that an accurate predictive model may need several markers, and that markers closer to the malignant phenotype (activated, phosphorylated proteins) may be better predictors than DNA/RNA markers or inactive (unphosphorylated) protein markers. In order to determine which tumor will be susceptible to anti-EGFR therapy, one may need to assess the level of phosphorylated EGFR as well as the level of phosphorylated downstream mediators, Akt, Erk1/2 and Stat3, the cytosolic end points to the three branches of the EGFR signaling pathway. Moreover, clinical data have identified these markers as relevant to the natural history of lung cancer [11–15].
At the Maine Center for Cancer Medicine, clinical data were compiled for 160 NSCLC patients that had received either gefitinib or erlotinib between 1996 and 2002 and organized into a database constructed in FileMaker Pro 8. Recorded variables included age, sex, smoking history, ECOG performance status score, date of diagnosis, stage at diagnosis, treatments, and toxicities. Outcome data included best response, time to progression, survival since start of TKI treatment, and overall survival. Best response to the new therapeutic agent was documented in the first computed tomography (CT) scans done 3-4 months after the start of TKI therapy. Time to progression was defined as the time interval from the initiation of TKI therapy to the time of disease progression as evidenced by CT scan or by a decline in performance status. Survival since TKI treatment was defined as the time interval from initiation of TKI therapy to death.
Paraffin-embedded tumor samples were obtained from a subgroup of patients (n = 32) that had received gefitinib or erlotinib and had adequate histological material. Twenty three of these were primary lung tumor resection samples, six were lymph node samples, and three were samples from other biopsy sites. All samples had been fixed in 10% buffered formalin. Clinical data were kept unavailable during the laboratory analysis until all data were evaluated. This study was approved by the Maine Medical Center Institutional Review Board.
Formalin-fixed, paraffin-embedded tumor samples were cut in 4 μm-thick sections. Consecutive sections were then treated with primary antibodies purchased from Cell Signaling Technologies (Beverly, MA, USA) to visualize phospho-EGFR (Tyr1173) (53A5 rabbit mAb), phospho-Akt (Ser473) (736E11 rabbit mAb), phospho-Stat3 (Tyr705) (D3A7 rabbit mAb) and phospho-Erk1/2 (Thr202/Tyr204) (20G11 rabbit mAb), according to the manufacturer's instructions. Primary antibodies were visualized with biotinylated secondary antibodies, streptavidin/HRP enzyme complex and DAB chromogen. The slides were counterstained with hematoxylin. Total EGFR levels were obtained from sections sent to NorDx Laboratories and stained on a Ventana BenchMark XT automated stainer using anti-EGFR mouse monoclonal antibody 3C6 and Basic DAB Detection kit (Ventana Medical Systems, Tucson, AZ, USA). Protocols were pre-tested using positive and negative control cell pellets for each anti-phospho protein antibody, available through Cell Signaling Technologies. Sections were scanned at low magnification and portions containing tumor tissue were delineated by the study pathologist (PA). Specimens were evaluated by light microscopy by two independent observers and scored based on a semiquantitative approach of percentage of positive tumor cells (0-100%), multiplied by staining intensity (0 = negative, 1 = weak, 2 = moderate, 3 = strong). In cases in which there were tumor cell populations within one sample with clear distinction in staining intensities, each region's intensity was multiplied by the appropriate percentage and added to produce the sample score. The final percentage of cells staining and score for each slide was calculated as the average between the percentage and scores of the two observers. For example, a sample that was scored by one observer as having 100% of cells staining, with 50% of them staining at intensity 1 and 50% at intensity 2, and was scored by the second observer as having 80% of cells staining, with 70% of them staining at intensity 1 and 10% at intensity 2, would receive a score of 150 [(50 × 1) + (50 × 2)] from observer one, and a score of 90 [(70 × 1) + (10 × 2)] from observer two. Accordingly, in this sample, the final percentage of cells staining would be 90% (mean between 100% and 80%), and the final score would be 120 (mean between 150 and 90). In this manner, a total score range of 0-300 was generated for each sample, where 0 was classified as no expression, 1-10 was classified as low expression, 11-100 was classified as moderate and 101-200 was classified as moderate-high expression level, and scores over 200 were classified as high expression level. The distribution of staining (membranous, cytoplasmic, or nuclear) was recorded. Positive expression of pERK1/2 in stromal fibroblasts and endothelial cells of tumor tissue was used as an internal standard of tissue quality and antigen preservation.
Univariate analysis between time to progression and survival according to best response, and appearance of rash were conducted using the Kaplan-Meier method. For the patients with tissue sample (n = 32), the associations between time to progression and survival according to level of expression of four phospho-proteins, and total EGFR were investigated using Rapid Insight Analytics statistics program software, and box plots analyses with Excel and XLSTAT program software.
Response to tyrosine kinase inhibitors
NSC not further specified
Avg. previous lines of chemo
As assessed by CT scans within 3-4 months after the initiation of therapy or by clinical examination when declining performance status prevented scanning, most patients in the gefitinib group had progressive disease (65%), but a quarter had stable disease (26%) and a small minority had a response (3% CR and 3% PR). In contrast almost half the patients in the erlotinib group experienced stable disease (47%) and 10% had a response (5% CR and 5% PR), while 41% showed progressive disease. The longest time to progression was 35 months for erlotinib and 72 months for gefitinib. The patient that experienced no progression for over 6 years was a Caucasian female with a 100 pk/year smoking history.
Instertitial Lung Disease
Clinical predictors of response
Association between phospho-protein level and response to gefitinib
The identification of patients that are likely to benefit from EGFR TKI therapy is a priority for patients with advanced non-small cell lung cancer. A recent study showed that TKI therapy-induced rash strongly correlated with better disease control and progression free survival, and concluded that the development of rash should be viewed as an indication of likelihood of clinical benefit . In this study, we also found that the appearance of rash correlated with longer TTP and survival with TKI treatment. In addition, we found that patients who achieved some degree of disease control at the first response evaluation had a significantly better outcome than patients who experienced disease progression.
Although early clinical indicators of response are useful, the identification of pre-treatment predictors of benefit would be even more desirable. We hypothesized that tumors that are actively using the EGFR pathway for survival and growth would be most susceptible to TKI therapy-mediated inhibition. We also hypothesized that the activation status of the EGFR pathway could be assessed by evaluating the phosphorylation level of key members of the pathway, rather than total protein levels. Accordingly, we determined the expression levels of the activated receptor (pEGFR), and the activated cytosolic end points to the three branches of the EGFR pathway (pAkt, pStat3 and pErk1/2) in pre-treatment samples of tumor tissue from patients undergoing TKI therapy and response assessment.
In general, the expression level of the phosphorylated proteins was significantly less than that of total EGFR. This reflects the fact that the phospho-proteins are more labile and more difficult to detect than total protein levels . In agreement with this, we found that the signal from phosphorylated antigens was more prominent in tumor cells located toward the periphery of the tissue blocks as opposed to cells in more internal regions. We presume that the time delay that occurs from the immersion of the block in fixative to its complete permeation results in the observed pattern of expression. Interestingly, when we compared the levels of the four phospho-proteins amongst each other, we found a significant correlation between the level of pEGFR and pStat3, in general agreement with the notion that the activation of the receptor gets transduced most directly to pStat3 in the signaling pathway.
Analysis of the association between outcome and phospho-protein expression suggested that the pattern of expression may be useful as a predictor of clinical benefit from TKI therapy. However, the small number of samples precluded obtaining statistically significant results. Higher levels of pEGFR, pAkt, and pStat3 were associated with longer TTP, while higher levels of pErk1/2 were associated with shorter TTP. These results make biologic sense given the presumed position of these proteins in the pathway. pAkt and pStat are thought to mediate cell survival, and it is possible that detecting these proteins indicates that a tumor is using the pathway for survival and that inhibition of the pathway provokes widespread apoptosis. On the other hand, pErk, mediates cell proliferation and is downstream of ras, a gene that is often mutated in these tumors. High levels of pErk may reflect activation of ras rather than EGFR, a condition where EGFR TKI would be ineffective.
In our experience, erlotinib was found slightly superior to gefitinib in objective response rate and disease control rate, but in terms of time to progression and survival after treatment, the two agents were statistically comparable.
In the absence of molecular predictors of response, the appearance of a rash and evidence of disease control at first follow-up exams are good indicators of response to TKIs. In combination, pEGFR, pAKT, pStat3 and pErk1/2 are promising predictive markers of response to EGFR TKI and deserve further analysis.
We would like to thank Lee Lucas, PhD for conducting the Kaplan-Meier analyses, Robert Ackroyd for procuring the tissue samples and Cindy Lester for conducting the immunohistochemical analysis for total EGFR. We would also like to acknowledge Thomas Ervin, MD and Stephen Toothhaker, MD for conceiving the idea of a study on patients that received gefitinib and initiating the assembly of their clinical data. Finally, we would like to thank Karen Rasmussen, PhD and Frederick Aronson, MD for their critical review of the research proposal and manuscript respectively. This work was funded by a grant from the Maine Cancer Foundation.
- National Comprehensive Cancer Network: Non-Small Cell Lung Cancer, Version 2. Clinical Practice Guidelines in Oncology. 2008, Accessed November 20, 2007, [http://www.nccn.org/professionals/physician_gls/f_guidelines.asp]Google Scholar
- Hanna N, Shepherd FA, Fossella FV, Pereira JR, De Marinis F, von Pawel J, Gatzemeier U, Tsao TC, Pless M, Muller T, Lim HL, Desch C, Szondy K, Gervais R, Shaharyar , Manegold C, Paul S, Paoletti P, Einhorn L, Bunn PA: Randomized phase III trial of pemetrexed versus docetaxel in patients with non-small cell lung cancer previously treated with chemotherapy. J Clin Oncol. 2004, 22: 1589-1597. 10.1056/NEJMoa050753.View ArticlePubMedGoogle Scholar
- Shepherd FA, Rodriguez Pereira J, 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, Santabárbara P, Seymour L: Erlotinib in previously treated non-small-cell lung cancer. N Engl J Med. 2005, 353: 123-132. 10.1016/S0169-5002(03)00137-5.View ArticlePubMedGoogle Scholar
- Hirsch F, Scagliotti GV, Langer CJ, Varella-Garcia M, Franklin WA: Epidermal growth factor family of receptors in preneoplasia and lung cancer: perspectives for targeted therapies. Lung Cancer. 2003, 41: S29-41. 10.1016/S0140-6736(08)61758-4.View ArticlePubMedGoogle Scholar
- Kim ES, Hirsh V, Mok T, Socinski MA, Gervais R, Wu YL, Li LY, Watkins CL, Sellers MV, Lowe ES, Sun Y, Liao ML, Osterlind K, Reck M, Armour AA, Shepherd FA, Lippman SM, Douillard JY: Gefitinib versus docetaxel in previously treated non-small-cell lung cancer (INTEREST): a randomised phase III trial. Lancet. 2008, 372 (9652): 1809-18. 10.1016/S0140-6736(08)61758-4.View ArticlePubMedGoogle Scholar
- Cobleigh MA, Vogel CL, Triphany D, Robert NJ, Scholl S, Fehrenbacher L, Wolter JM, Paton V, Shak S, Lieberman G, Slamon DJ: Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J Clin Oncol. 1999, 17: 2639-48. 10.1200/JCO.2005.05.1474.PubMedGoogle Scholar
- Gatzemeier U, Pluzanska A, Szczesna A, Kaukel E, Roubec J, De Rosa F, Milanowski J, Karnicka-Mlodkowski H, Pesek M, Serwatowski P, Ramlau R, Janaskova T, Vansteenkiste J, Strausz J, Manikhas GM, Von Pawel J: Phase III study of erlotinib in combination with cisplatin and gemcitabine in advanced non-small-cell lung cancer: the Tarceva Lung Cancer Investigation Trial. J Clin Oncol. 2007, 25 (12): 1545-1552. 10.1200/JCO.2005.02.840.View ArticlePubMedGoogle Scholar
- Herbst RS, Prager D, Hermann R, Fehrenbacher L, Johnson BE, Sandler A, Kris MG, Tran HT, Klein P, Li X, Ramies D, Johnson DH, Miller VA: TRIBUTE: A phase III trial of erlotinib hydrochloride combined with carboplatin and paclitaxel chemotherapy in advanced non-small-cell lung cancer. J Clin Oncol. 2005, 23 (25): 5892-5899. 10.1016/S0169-5002(03)91900-3.View ArticlePubMedGoogle Scholar
- Bailey R, Kris M, Wolf M, Kay A, Averbuch S, Askaa J, Janas M, Schmidt K, Fukuoka M: Gefitinib monotherapy for pretreated advanced non-small-cell lung cancer in IDEAL1 and 2: tumor response is not clinically predictable from tumor EGFR membrane staining alone. Lung Cancer. 2003, 41: S71-10.1016/S0169-5002(03)91900-3.View ArticleGoogle Scholar
- Eberhard DA, Johnson BE, Amler LC, Goddard AD, Heldens SL, Herbst RS, Ince WL, Jänne PA, Januario T, Johnson DH, Klein P, Miller VA, Ostland MA, Ramies DA, Sebisanovic D, Stinson JA, Zhang YR, Seshagiri S, Hillan KJ: 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 (25): 5900-9. 10.1136/jcp.2005.027615.View ArticlePubMedGoogle Scholar
- Sonnweber B, Dlaska M, Skvortsov S, Dirnhofer S, Schmid T, Hilbe W: High predictive value of epidermal growth factor receptor phosphorylation but not of EGFRvIII mutation in resected stage I non-small cell lung cancer (NSCLC). J Clin Pathol. 2006, 59 (3): 255-9. 10.1136/jcp.2005.027615.View ArticlePubMedPubMed CentralGoogle Scholar
- David O, Jett J, LeBeau H, Dy G, Hughes J, Friedman M, Brody AR: Phospho-Akt overexpression in non-small cell lung cancer confers significant stage-independent survival disadvantage. Clin Cancer Res. 2004, 10: 6865-71. 10.1158/1078-0432.CCR-04-0174.View ArticlePubMedGoogle Scholar
- Blackhall FH, Pintilie M, Michael M, Leighl N, Feld R, Tsao MS, Shepherd FA: Expression and prognostic significance of Kit, Protein Kinase B, and Mitogen-activated Protein Kinase in patients with small cell lung cancer. Clin Cancer Res. 2003, 9: 2241-2247.PubMedGoogle Scholar
- Haura EB, Zheng Z, Song L, Cantor A, Bepler G: Activated Epidermal Growth Factor Receptor- Stat3 signaling promotes tumor survival in vivo in non small cell lung cancer. Clin Cancer Res. 2005, 11 (23): 8288-8294. 10.1016/j.lungcan.2005.10.003.View ArticlePubMedGoogle Scholar
- Tang JM, He QY, Guo RX, Chang XJ: Phosphorylated Akt overexpression and loss of PTEN expression in non-small cell lung cancer confers poor prognosis. Lung Cancer. 2006, 51 (2): 181-91. 10.1158/1078-0432.CCR-06-2610.View ArticlePubMedGoogle Scholar
- Wacker B, Nagrani T, Weinberg J, Witt K, Clark G, Cagnoni PJ: Correlation between development of rash and efficacy in patients treated with the epidermal growth factor receptor tyrosine kinase inhibitor erlotinib in two large phase III studies. Clin Cancer Res. 2007, 13 (13): 3913-3921. 10.1158/1078-0432.CCR-06-2610.View ArticlePubMedGoogle Scholar
- Mandell JW: Phosphorylation state-specific antibodies: Applications in investigative and diagnostic pathology. Am J Pathol. 2003, 163: 1687-1698.View ArticlePubMedPubMed CentralGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2407/9/333/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.