TargetedTherapies in Combinationwith Chemotherapy in Non ^ Small Cell Lung Cancer
David H. Johnson
Abstract:With rare exceptions, attempts to combine so-called targeted agents with standard cytotoxic chemotherapy in advanced non ^ small cell lung cancer have yielded disappointing results. The reasons underlying these spectacular failures are not always fully understood, but certainly the lack of careful patient selection is a major contributing factor. In addition, recent preclinical and clinical studies indicate that antagonismmay exist between the epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors and chemotherapy primarily in tumor cells withwild-type EGFR. By contrast, tumor cells harboring somatic mutations in EGFR experience massive apoptosis when exposed to the EGFR tyrosine kinase inhibitors. Therefore, in theory, mutant tumor cells should exhibit enhanced cell kill when treatedwith concomitant chemotherapyand EGFR tyrosine kinase inhibitors akin to what is observed with chemotherapy and trastuzumab in breast cancer. Clinical data from the recently completedTRIBUTE trial support the latter possibility. Ideally, future studies of EGFR tyrosine kinase inhibitors and other targeted drugs will use careful patient selection criteria based on well-characterized and validated predictive markers. However, in the absence of suchbiomarkers, clinical judgment, common sense, and innovative clinical trial design are necessary to avoid undue delay in drug development. One only has to casually peruse the lay or scientific press to appreciate the high level of enthusiasm for the burgeoning field of so-called targeted therapy in the management of various malignancies, including non–small cell lung cancer (NSCLC; refs. 1, 2). However, in spite of some recent notable successes using targeted therapy (3), classic cytotoxic drugs remain the mainstay of treatment for locally advanced and metastatic NSCLC and more recently as an adjunct to surgery for patients with resected early-stage disease (4, 5). Thus, it is not surprising that investigators continue to use standard chemotherapy as the foundation on which newer treatment regimens are developed for NSCLC. Most of the earliest efforts to incorporate targeted agents into the treatment of NSCLC concentrated on patients with advanced, metastatic disease, which may not be the optimal setting in which to test these drugs. However, this approach is based in part on data derived from preclinical studies indicating that some targeted drugs yielded additive or even synergistic cytotoxic activity when combined with chemotherapy (6, 7). Unfortunately, with one notable exception (8), all of these efforts have failed. This review will assess some of the posited reasons for the failure of these early trials and will touch briefly on some promising new molecular targets other than the epidermal growth factor receptor (EGFR; to be discussed elsewhere in these conference proceedings). I will also outline some clinical trial strategies to avoid the spectacular phase III failures of the past. INTACT, TRIBUTE, and TALENT Trials By now, the results of the INTACT, TRIBUTE, and TALENT trials are well known (refs. 9–12; summarized in Table 1). Each of these studies was designed based on the knowledge that gefitinib and erlotinib have shown impressive antitumor activity in patients with refractory, advanced NSCLC with only modest toxicity (13, 14) and on the premise that these drugs could enhance the cytotoxic effects of standard chemotherapy agents as shown in preclinical studies (6, 7). Thus, it was widely anticipated that combining EGFR tyrosine kinase inhibitors (TKI) with standard chemotherapy would improve outcome in advanced NSCLC compared with chemotherapy alone, and yet all four trials failed to meet their primary end point of improved survival. Why? The principal criticism leveled at the INTACT and TRIBUTE trial centers on failure to select a proper study population likely to benefit from this class of targeted agents (15, 16). Of course, this particular criticism assumes that an appropriate predictive factor was already well characterized at the time these studies were undertaken. In fact, before initiating these studies, no such predictive marker was universally accepted, nor was there a clear correlation between expression of EGFR (the putative target) and the growth-inhibitory activity of gefitinib in Author’s Affiliation: Vanderbilt-Ingram Cancer Center Director and Division of Hematology and Oncology,Vanderbilt University School of Medicine Nashville, Tennessee Received1/15/06; revised 4/20/06; accepted 5/2/06. Grant support: NIH grants CA68485, CA49957, and P50 CA90949 (Lung Specialized Programs of Research Excellence) and an Unrestricted Institutional Grant for Cancer Research from the Bristol Myers Squibb Foundation. Presented at theThird Cambridge Conference on Novel Agents in theTreatment of Lung Cancer: Advances in EGFR-Targeted Agents, September 23-24, 2005, Cambridge, Massachusetts. Requests for reprints: David H. Johnson,Vanderbilt-Ingram Cancer Center, 777 Preston Research Building, Nashville, TN 37232-6307. Phone: 615-343-9454; Fax: 615-936-2236; E-mail: david.johnson@vanderbilt.edu. F2006 American Association for Cancer Research. doi:10.1158/1078-0432.CCR-06-0095 www.aacrjournals.org Clin Cancer Res 2006;12(14 Suppl) July15, 2006 4451s Research. on May 10, 2017. © 2006 American Association for Cancer clincancerres.aacrjournals.org Downloaded from preclinical or clinical studies (17, 18). In addition, these studies were undertaken before the discovery of the recently described somatic EGFR mutation in NSCLC that renders tumors more responsive to gefitinib and erlotinib (19–21). However, even today, the optimal molecular marker of gefitinib and erlotinib activity is a matter of controversy (20–25). Nonetheless, the likelihood of a positive outcome might have been increased by using specific clinical features to select patients for enrollment (26). Early on, even before the discovery of the activating EGFR mutations, certain clinical and histologic characteristics seemed to predict for a higher likelihood of response to EGFR TKIs (13, 26, 27). The predictive ability of some of these factors, most notably smoking history and histology, has held up in several subsequent reports (19, 27) and two prospective trials (3, 11). Whether such selection criteria would have made a difference in the outcome of these phase III trials is purely conjectural, of course, but in retrospect it seems as if it would have been a useful step, especially if one accepts the premise that an EGFR mutation is a necessary feature for a favorable outcome. This is because these clinical features seem to track fairly closely with the presence of EGFR mutations (28). Parenthetically, based on recent reports, mutational status alone might be insufficient to select all patients potentially benefited by EGFR TKIs (23, 24). In any case, had clinical features been used prospectively to select enrollees, these studies might have yielded very different outcomes. Another valid criticism of these studies, or at least of INTACT-2, is the fact that a contemporarily initiated phase II trial testing the activity of chemotherapy plus gefitinib in ‘‘unselected patients’’ yielded uninspiring results and should have interjected a precautionary note for these trials (29). Yet another criticism posited by some experts is that the concomitant use of chemotherapy and gefitinib or erlotinib results in antagonism akin to what has been reported with the concurrent use of tamoxifen and chemotherapy in breast cancer (30, 31). Like tamoxifen, initial preclinical studies indicated that gefitinib was a cytostatic agent (17). The antiproliferative effect of gefitinib is the result of p27-mediated G1 cell cycle arrest of EGFR-dependent tumor cells, which could render the cells less sensitive to cytotoxic agents. Thus, continuous administration of an EGFR TKI concurrent with chemotherapy might actually have a negative effect (32). Interestingly, a subset analysis of the TRIBUTE trial seems to support the possibility of antagonism between erlotinib and chemotherapy at least in patients with tumors harboring a wild-type EGFR (ref. 33; Table 2). To circumvent the theoretical problem of antagonism, some investigators have proposed sequencing or alternating use of chemotherapy and gefitinib based on preclinical models (32, 34, 35). However, the results of a recently halted Southwest Oncology Group trial showed that sequential use of gefitinib after combined modality therapy in locally advanced NSCLC yielded a numerically worse survival outcome (36). It is also now well established that gefitinib and erlotinib have proapoptotic effects as well as antiproliferative effects (32, 37–39). In theory, the proapoptotic effects of gefitinib or erlotinib should enhance the effectiveness of cytotoxic drugs. If true, combining the EGFR TKIs with chemotherapy should be a good strategy akin to combining trastuzumab with chemotherapy in breast cancer. Another look at the TRIBUTE data suggests that this may be the case specifically in patients with an EGFR mutation (ref. 33; Table 2). The TRIBUTE data Table1. Results of phase III trials combining chemotherapy and EGFR TKI Study Treatment Patient no. Odds ratio (%) Time to progression (mo) Mean survival time (mo) 1y (%) INTACT-1 P-Gem 363 47.2 6.0 10.9 44 P-Gem + #G 365 51.2 5.8 9.9 41 P-Gem + "G 365 50.3 5.5 9.9 43 INTACT-2 Cb-Pac 345 28.7 5.0 9.9 42 Cb-Pac + #G 345 30.4 5.3 9.8 41 Cb-Pac + "G 347 30.0 4.6 8.7 37 TALENT P-Gem 536 29.9 5.7 10.3 42 P-Gem + E 533 31.5 5.5 10.0 41 TRIBUTE Cb-Pac 533 19.3 4.9 10.5 44 Cb-Pac + E 526 21.5 5.1 10.6 47 NOTE: Data from Giaccone et al. (9), Herbst et al. (10,11), and Gatzemeier et al. (12). Abbreviations: P, cisplatin; Gem, gemcitabine; G, gefitinib; Cb, carboplatin; Pac, paclitaxel; E, erlotinib; #, 250 mg; ", 500 mg. Table 2. Overall response rates in TRIBUTE with or without EGFR mutation EGFR mutation (%) Wild-type EGFR (%)