Cetuximab Preclinical Antitumor Activity

104

Cetuximab preclinical antitumor activity (monotherapy and combination based) is not predicted by relative total or activated epidermal growth factor receptor tumor expression levels

Robert Wild, Krista Fager, Christine Flefleh, the cetuximab resistant/weakly responsive HT29, A549, David Kan, Ivan Inigo, Stephen Castaneda, and WiDr models. We conclude that preclinical activity Feng (Roger) Luo, Amy Camuso, associated with cetuximab monotherapy does not corre- Kelly McGlinchey, and William C. Rose late directly with relative basal levels of total or activated (pY1068) EGFR in a tumor. Moreover, robust single-agent Bristol-Myers Squibb Co., Pharmaceutical Research Institute, activity by cetuximab may be the best predictor for this Oncology Drug Discovery, Princeton, New Jersey agent to potentiate chemotherapy-mediated antitumor activities. [Mol Cancer Ther 2006;5(1):104–13] Abstract Although Erbitux (cetuximab) has proven therapeutic Introduction benefit in the clinical setting, the molecular determinants Erbitux (cetuximab) is a chimeric mouse/human monoclo- predicting responsiveness to this agent are still not very nal antibody of the IgG1 subclass that targets the human well understood. Here, we assessed the relationship epidermal growth factor receptor (EGFR; ref. 1). EGFR is between basal total and activated (pY1068) epidermal overexpressed in about one third of all human cancers and growth factor receptor (EGFR) levels in a tumor and the has been directly implicated in tumor growth and responsiveness to cetuximab monotherapy or combina- progression (1). In particular, elevated expression levels tion-based treatment using human xenograft models. of EGFR in colon carcinomas have been linked to more Cetuximab treatment alone (0.25–1 mg/mouse/injection, aggressive disease and poor prognosis (2). Hence, strate- q3d, i.p.) effectively delayed the growth of GEO and gies aimed at disrupting the function of the EGFR have L2987 tumors by a minimum of 10 days corresponding to been investigated as exciting new approaches for the log cell kill values of z1.0. Borderline activity was seen in treatment of cancer and include small molecule inhibitors the A549 and WiDr xenografts. However, cetuximab as well as various biologics approaches (3). In the case of failed to show any significant antitumor activity in the the monoclonal antibody cetuximab, clinical studies have HT29, HCT116, LOVO, Colo205, LX-1, HCC70, and N87 shown that this agent is capable of significantly inhibiting models. All of the studied tumors had detectable yet tumor growth (4). In particular, about 11% to 23% of variable levels of EGFR. For combination regimens, irinotecan (CPT-11) refractory, EGFR-positive colon cancer cetuximab (1 mg/mouse/injection, q3dx5, i.p.) and cis- patients dramatically benefited from this antibody therapy platin (4.5 mg/kg/injection, q3dx5, i.v.) proved to be when administered as a single agent or in combination significantly more efficacious than individual monothera- with irinotecan (5). Thus, in February 2004, cetuximab was pies in the cisplatin-refractory yet cetuximab-responsive approved in the United States for the treatment of GEO tumor model (P < 0.001). However, no therapeutic advanced colon cancer in combination with irinotecan. enhancement was observed in the cisplatin and cetuximab Nevertheless, several questions remain unanswered about weakly responsive A549 xenograft. Similarly, combina- this therapeutic. Specifically, it is still not very well tions of CPT-11 (48 mg/kg/injection, q3dx5, i.v.) with understood why only a fraction of patients with EGFR- cetuximab (1 mg/mouse/injection, q3dx5, i.p.) failed to positive tumors respond to cetuximab treatment (i.e., which show any improvements over individual monotherapies in exact patient population will most likely benefit from this antibody therapy). Preclinically, many studies have shown that treatment with cetuximab results in inhibition of tumor growth in nude mice bearing xenografts of human cancer cell lines Received 7/19/05; revised 10/11/05; accepted 11/15/05. (6–14). Moreover, treatment with the antibody in combi- The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked nation with chemotherapeutic drugs or radiation has been advertisement in accordance with 18 U.S.C. Section 1734 solely to reported to be more efficacious than individual mono- indicate this fact. therapies in the same or similar xenograft model systems Requests for reprints: Robert Wild, Bristol-Myers Squibb Co., Oncology (6–11, 13–20). Although published preclinical reports Drug Discovery, P.O. Box 4000, Mailstop K23-03, Princeton, NJ 08543- 4000. Phone: 609-252-5808; Fax: 609-252-7340. suggest that EGFR expression may be needed for cetux- E-mail: [email protected] imab activity (21, 22), recent clinical evidence indicates that Copyright C 2006 American Association for Cancer Research. response to cetuximab treatment is independent of the doi:10.1158/1535-7163.MCT-05-0259 relative degree of EGFR expression levels in tumors (23).

Mol Cancer Ther 2006;5(1). January 2006

The objective of the following studies was to further Tu m o r L in e s expand on our knowledge of cetuximab monotherapy All human carcinoma xenografts were propagated in activities in a broad panel of EGFR-positive human nude mice as s.c. transplants using tumor fragments carcinoma xenografts. In particular, we wanted to assess obtained from donor mice. The models employed included whether any correlation existed between basal total EGFR the colon carcinomas: GEO, HT29, HCT116, LOVO, WiDr, or basal activated (pY1068) EGFR levels in a tumor and the and Colo205; the lung carcinomas: L2987, A549, and LX-1; responsiveness to cetuximab treatment. To this effect, we the breast carcinoma, HCC70; and the gastric carcinoma, employed tumor models from several different histologies N87. Tumor passage occurred approximately every 2 to 4 (colon, lung, breast, and gastric cancer), all known to weeks depending on the model. All tumor implants for frequently overexpress EGFR (24). Finally, we also evalu- efficacy testing were s.c. in nude mice and employed tumor ated the utility of combining cetuximab with the chemo- fragments from tumors passaged at least twice in vivo. therapeutic agents irinotecan (CPT-11) or cisplatin in Quantitation of Total EGFR and Activated (pY1068) chemotherapy-refractory or poorly responding tumor EGFR Levels in HumanTumor Xenograft Samples models because positive clinical results have been reported Human tumor xenografts were established in athymic in the same setting (5, 25). Based on our collective mice and surgically removed when they reached an preclinical data, attempts were then made to identify the average tumor size of 150 to 300 mg. Resected tumors determinants that predict synergistic antitumor activity of were immediately snap-frozen in liquid nitrogen and cetuximab with conventional anticancer agents. stored at 80jC until further use. Frozen tumor tissues were then lysed in ice-cold lysis buffer that contained 10% Triton X-100, 5% glycerol, 20 mmol/L Tris-HCl (pH 7.7), Materials and Methods 0.5 mmol/L EDTA, 0.1 mol/L NaCl, 1 mmol/L sodium Compounds and Their Administrations orthovanadate, 1% Phosphatase Inhibitor Cocktail 2 Clinical-grade cetuximab was supplied by ImClone (Sigma), and 2% Complete protease inhibitors (Roche Systems (New York, NY) at a concentration of 2 mg/mL Molecular Biochemicals, Indianapolis, IN). The sample in a buffer consisting of 10 mmol/L sodium phosphate and was subsequently homogenized using a Wharton tissue 145 mmol/L sodium chloride at pH 7.2. For all efficacy homogenizer. Cell lysates were clarified by centrifugation studies requiring lower concentrations of cetuximab, (10 minutes at 10,000 g), and protein concentrations the stock solution was diluted with sterile PBS (pH 7.4). were determined using the MicroBCA method (Pierce). Cetuximab was administered i.p. at a constant volume of Tumor lysates from all of the samples were adjusted 0.5 mL/mouse. Clinical-grade cisplatin (Platinol) was to a concentration of 3 mg/mL protein. Equal amounts of obtained from Bristol-Myers Squibb (Princeton, NJ), and total protein (30 Ag) from each tumor sample were then irinotecan (CPT-11) was purchased commercially; they used to analyze the relative total human EGFR and were both dissolved in sterile 0.9% NaCl (saline) and activated EGFR (pY1068) levels in the different xenograft administered i.v. in a volume of 0.01 mL/g of mouse body samples as determined by ELISA assay according to the weight. CPT-11 and cisplatin were administered up to their protocol recommended by the manufacturer (Biosource maximum tolerated dose (MTD) as previously reported International). Expected values for different tumor cell (26, 27). lines, as provided by the assay manufacturer, were used Chemicals and Reagents as a guideline to classify samples as high, medium, or low Complete protease inhibitor tablets were from Roche EGFR expressers. Furthermore, we considered samples Diagnostics (Indianapolis, IN). MicroBCA reagents were with EGFR levels higher than the calculated mean for all from Pierce (Rockford, IL). Human-specific EGFR and evaluated tumor samples (f20 ng/mL) as high receptor pY1068 EGFR ELISA kits were purchased from Biosource expressers. International, Inc. (Camarillo, CA). All other chemicals and In vivo Antitumor Testing reagents were from Sigma (St. Louis, MO). Unless For each experiment, athymic mice were implanted otherwise specified, sterile buffers and solutions were s.c. with f1mm3 tumor fragments obtained from donor obtained from Life Technologies (Carlsbad, CA). Sterile animals. Tumors were allowed to grow to a size between tissue culture ware was obtained from Fisher Scientific Co. 100 to 200 mg, and animals were then distributed to various (Hanover Park, IL). treatment and control groups (n = 8 per group). Treatment Animals of each animal was based on individual body weight on a Female athymic, nude mice (5–6 weeks of age) were mg per kg (mg/kg) basis for all drugs except cetuximab, obtained from Harlan Sprague-Dawley Co. (Indianapolis, which was administered on a mg per mouse (mg/mouse) IN) and maintained in an ammonia-free environment basis delivered in 0.5 mL/mouse. Treated animals were in a defined and pathogen-free colony. Animals were checked daily for treatment related toxicity/mortality. quarantined for f1 week before their use for tumor In addition, average body weights for each group were propagation and drug efficacy testing. They were fed food determined before the initiation of treatment (Wt1) and and water ad libitum. All studies were done in accordance following the last treatment dose (Wt2). The difference with Bristol-Myers Squibb and the American Association of body weight (Wt2 Wt1) was then used to estimate for Accreditation of Laboratory Animal Care guidelines. the degree of treatment-related toxicity. Tumors were

Mol Cancer Ther 2006;5(1). January 2006

measured with a caliper twice a week, until they reached a Results predetermined ‘‘target’’ size (0.5–1 g depending on tumor EGFR Expression Levels of Tumor Xenografts model used). Tumor weights (mg) were estimated from the Cetuximab was evaluated for its antitumor activity in a following formula: panel of 11 different human tumor xenografts with various EGFR expression levels. The employed tumor models were 2 Tumor weight ðmgÞ¼flength ðmmÞwidth ðmmÞ g=2 maintained in vivo for a minimum of two passages before use in an experiment and included six colon carcinomas Efficacy of antitumor agents was determined in two (LOVO, WiDr, GEO, HCT-116, Colo-205, and HT-29), three ways: lung carcinomas (L2987, A549, and LX-1), one breast (a) Calculating the relative % tumor growth inhibition carcinoma (HCC70), and one gastric carcinoma (N87). (%TGI) of treated animals at different time points using the EGFR expression levels were determined in tumor lysates formula: via ELISA (specific for human EGFR) and ranged from high (>20 ng/mL) in L2987, HCC70, and LOVO; to medium %TGI ¼fðCt TtÞ=ðCt C0Þg 100 (10–20 ng/mL) in WiDr, GEO, A549, LX-1, and HCT116; to low (<10 ng/mL) in Colo205, N87, and HT29 (Fig. 1A). All where Ct = the median tumor weight (mg) of control (C) samples were above the limit of EGFR detection. mice at time t (days after tumor implant); Tt = median Cetuximab MonotherapyAntitumor Activity tumor weight (mg) of treated (T) mice at time t; C0 = Cetuximab administered i.p. at 1 mg/mouse/injection on median tumor weight (mg) of control mice at time 0. an intermittent dosing schedule has been successfully used (b) Calculating the tumor growth delay (T-C value), in a number of xenograft models (6–8, 11, 13, 19, 20, 30, 31). defined as the difference in time (days) required for the However, a systematic evaluation of the optimal dose and treated tumors (T) to reach a predetermined target size schedule of this antibody in preclinical models has not been compared with those of the control group (C). Antitumor described. Moreover, a comprehensive and simultaneous efficacy could then be compared between different treat- analysis of the antitumor activity of this antibody in a large ment groups by estimating the overall gross log cell kill panel of tumor models has not been reported. (LCK). The following formula was employed to calculate Schedule comparison of 1 mg/mouse/injection in the the LCK: EGFR moderately expressing A549 tumor model showed that there was no appreciable difference in efficacy when the LCK ¼ T C=ð3:32 TVDTÞ antibody was administered either on a daily (qdx13), every other day (q2dx7), or every 3-day (q3dx5) schedule (data not where, tumor volume doubling time (TVDT) = median time shown). Moreover, a dose response experiment using a (days) for control tumors to reach target size median time q3dx5 schedule in the GEO colon carcinoma model showed (days) for control tumors to reach half the target size. that cetuximab was equally effective at dose levels between Antitumor activity was defined as a statistically 0.25 and 1 mg/mouse/injection (data not shown). Hence, significant LCK of z1.0 (19). Borderline activity, partic- cetuximab administrations of 0.25 to 1 mg/mouse/injection, ularly for monotherapy experiments employing the i.p., every 3 days (q3d) for a minimum of five injections cytostatic agent cetuximab, was defined as a continuous represented a likely optimal treatment regimen, which was TGI z50% for at least 1 TVDT any time after the start of then applied to profile the antitumor activity of the antibody treatment, which was accompanied by a statistically in our large panel of EGFR-positive xenograft models. significant tumor growth delay (T-C value). Statistical Using this preclinically defined optimal treatment regi- significance was determined by the Gehan’s generalized men, cetuximab showed marked and reproducible anti- Wilcoxon test and required P < 0.05 (28). For combina- tumor activity in the L2987 (human lung carcinoma) and tion drug treatment regimens, a combination treatment GEO (human colon carcinoma) models, achieving contin- was termed therapeutically synergistic if it was statisti- uous tumor growth inhibitions in excess of 50% for at least cally superior to the best achievable monotherapy tested 1 TVDT. This was accompanied by a significant tumor (19, 29). growth delay corresponding to >1 LCK in the majority of The dose of a compound that yielded the maximum experiments. Responses synonymous with borderline therapeutic effect was termed the optimal dose. Groups of activity based on tumor growth inhibition (%TGI z 50% mice with more than one animal death attributable to drug for 1 TVDT, accompanied with a statistically significant toxicity were considered to have had excessively toxic tumor growth delay) were occasionally seen in the A549 treatments, and their data were not used in the evaluation (human lung carcinoma) and WiDr (human colon carcino- of a compound’s (or combination’s) antitumor activity. A ma) xenograft models. However, the antibody was com- MTD was defined as one whose toxicity approached but pletely inactive in the HT29, HCT116, LOVO, LX-1, HCC70, did not attain the degree of lethality just described as being N87, and Colo205 human carcinoma models, where neither excessive. significant tumor growth inhibitions nor considerable tumor A tumor was considered refractory (resistant) to a growth delays were observed. In addition, longer treatment particular drug treatment if the agent was inactive at its durations did not seem to affect the antitumor responses in optimal dose or MTD, corresponding to a LCK of <1.0. the cetuximab-resistant and cetuximab-borderline sensitive

Mol Cancer Ther 2006;5(1). January 2006

xenografts, whereas extended dosing periods prolonged Conversely, GEO reproducibly showed significant sensi- tumor growth inhibitions in the GEO and L2987 cetuximab- tivity to cetuximab treatment (Table 1), yet this tumor sensitive tumor models. These results are summarized in model expressed a relatively low level of total basal EGFR Table 1. (10.88 ng/mL; Fig. 1A). Lack of Correlation between Basal Levels of Total EGFR or Relative Activated EGFR (pY1068) and Antitu- mor Response to Cetuximab Treatment We also assessed the relationship between the basal levels of EGFR expression in the 11 human carcinoma xenografts and their responses to cetuximab monotherapy, as measured by the maximum LCK achieved on a q3dx5 schedule (or a q3dx10 schedule for the N87 and HCC70 tumor models because other schedules were not evaluated). In the cetuximab-responsive GEO and L2987 tumors, maximum LCK values of 1.4 and 1.3, respectively, were achieved. In addition, tumor growth inhibitions in excess of 50% for at least 1 TVDT were attained in 100% of the experiments involving these two tumor models (Table 1). In the borderline cetuximab-sensitive A549 and WiDr xenografts, maximum LCK values of only 0.7 and 0.3, respectively, were obtained. Moreover, mean TGIs in excess of 50% for at least one TVDT were achieved in only two of four experiments for the WiDr xenograft and four of six experiments for the A549 tumor model (Table 1). Finally, in the cetuximab-resistant models, maximum LCK values rarely went higher than 0.1, and %TGI values were always below levels sufficient to be defined as borderline active (Table 1). ELISA analysis of human EGFR expression showed a >10-fold difference in the levels of EGFR protein among these tumors (Fig. 1A). Upon examination, the levels of EGFR expression did not correlate significantly with tumor response to cetuximab treatment (r = 0.38238, P = 0.2458; Fig. 1B). Although L2987 had the highest levels of EGFR expression among the tested xenograft samples (61.23 ng/mL) and showed one of the most sensitive phenotypes to cetuximab treatment (maximum 1.3 LCK), other high expressers of EGFR, such as HCC70 and LOVO (40.55 and 27.77 ng/mL, respectively), failed to respond effectively to cetuximab therapy (maximum LCK values achieved were 0.1 and 0.2 LCK, respectively).

Figure 1. A, basal total EGFR expression levels in various xenograft tumor samples. An equal amount of total protein (30 Ag) from each tumor sample was used to analyze the relative total EGFR levels in the different xenograft samples by ELISA. B, basal total EGFR levels versus antitumor efficacy of cetuximab (maximum LCK achieved at an optimal dose using the q3dx5 schedule, or q3dx10 for the N87 and HCC70 models since other schedules were not tested). Basal total EGFR levels were plotted versus the maximum antitumor efficacy obtained in each individual xenograft model. A linear regression analysis was done to determine degree of correlation between these two data sets (solid line). Dotted lines, 95% confidence interval of the linear regression line. C, basal relative activated EGFR (pY1068/Total EGFR) versus cetuximab responsiveness. Basal activated EGFR (pY1068) levels were normalized to basal total EGFR levels to determine basal ‘‘relative’’ activated EGFR levels. A one-way ANOVA showed that there was no significant difference in basal relative activated EGFR levels between cetuximab-sensitive (GEO, L2987, A549, and WiDr) and cetuximab-resistant (HCT116, HT29, LOVO, Colo205, N87, HCC70, and LX-1) tumor models (P = 0.19662).

Mol Cancer Ther 2006;5(1). January 2006

Table 1. Optimal antitumor activity of cetuximab monotherapy in carcinoma xenografts

Tumor (type) EGFR Experiment Dose Route, LCK at tumor target size z50 %TGI for status* no. (mg/mouse/injection) schedule 1 TVDT 0.5 g 1.0 g

c c GEO (colon) + 1 1 i.p., q3dx10,14 1.8 2.7 Yes c c 2 1 i.p., q3dx5,14 1.2 0.8 Yes c c 3 0.25 i.p., q3dx5,11 1.1 1.4 Yes c c 4 0.25 i.p., q3dx6,10 1.3 1.3 Yes c c 5 0.25 i.p., q3dx4,10 1.0 0.8 Yes c c 6 1 i.p., q3dx6,11 1.1 0.7 Yes WiDr (colon) ++ 1 1 i.p., q3dx10,14 0.2 0.6 Yes 2 1 i.p., q3dx10,15 0.0 0.1 No b 3 0.25 i.p., q3dx5,14 0.2 0.3 No b b 4 1 i.p., q3dx6,19 0.2 0.3 Yes HCT116 (colon) ++ 1 1 i.p., q3dx5,14 0.0 0.0 No HT29 (colon) + 1 1 i.p., q3dx5,13 0.1 0.1 No 2 1 i.p., q3dx10,14 0.0 0.0 No 3 1 i.p., q3dx5,10 0.0 ND No LOVO (colon) ++ 1 1 i.p., q3dx5,14 0.0 0.2 No Colo205 (colon) + 1 1 i.p., q3dx5,14 0.1 0.0 No N87 (gastric) + 1 1 i.p., q3dx10,16 0.1 0.0 No HCC70 (breast) +++ 1 1 i.p., q3dx10,28 0.1 0.1 No c c L2987 (lung) +++ 1 1 i.p., q3dx5,16 1.0 1.3 Yes b 2 0.5 i.p., q3dx4,22 0.9 0.8 Yes 3 1 i.p., q3dx6,20 1.0x 1.3x Yes c c 4 1 i.p., q3dx6,23 2.4 2.4 Yes LX-1 (lung) ++ 1 1 i.p., q3dx5,14 0.3 ND No b A549 (lung) ++ 1 1 i.p., q3dx10,18 0.4 0.6 Yes b 2 1 i.p., q3dx5,26 0.5 0.7 Yes 3 1 i.p., q3dx5,21 0.1 0.3 No 4 1 i.p., q3dx5,22 0.1 0.3 No 5 1 i.p., q3dx5,21 0.7 ND Yes 6 1 i.p., q3dx6,21 1.0x ND Yes

NOTE: LCK numbers in bold illustrate maximum antitumor activity obtained in each tumor model using a q3dx5 schedule (q3dx10 for N87 and HCC70). LCK values were then plotted versus total EGFR levels to correlate these two variables (see Fig. 1B). Abbreviation: ND, not determined. *EGFR receptor levels: +++ (>20 ng/mL), ++ (10-20 ng/mL), + (<10 ng/mL). cStatistically significant at P < 0.001 versus untreated control. bStatistically significant at P < 0.05 versus untreated control. xStatistically significant at P < 0.01 versus untreated control.

We also quantitated the basal levels of activated EGFR poorly responding human tumor models. Specifically, we (pY1068) in the tumor xenografts by ELISA. Considerable evaluated cetuximab in combination with cisplatin in the levels were detected in all but two models (i.e., A549 and GEO and A549 cisplatin-refractory tumor models and HT29 tumor xenografts). Upon analysis, there was no cetuximab combined with CPT-11 in the HT29, A549, and significant difference in relative activated EGFR levels WiDr CPT-11 poorly responding/refractory xenografts. In (pY1068/total EGFR) among the four cetuximab-sensitive/ these combination studies, all test agents were adminis- borderline-sensitive tumor models (L2987, GEO, A549, tered using an every 3 days 5 schedule (q3dx5), and a and WiDr) and the seven cetuximab-resistant xenografts tumor end point of 1 g was used unless otherwise specified. (Colo205, LX-1, HT29, HCT116, N87, HCC70, and LOVO) In addition, cetuximab treatment always followed, within as determined by one-way ANOVA (F = 1.9447, P = 15 minutes, the cytotoxic drug administration. All combi- 0.19662; Fig. 1C). Responsiveness to cetuximab was also nation treatments were well tolerated without significant independent of absolute values of activated (pY1068) EGFR toxicity. as well as TVDT (data not shown). As previously described, cetuximab monotherapy repro- Cetuximab In vivo Antitumor Activity (Combination ducibly showed antitumor activity in the GEO human Therapy) tumor xenograft (Table 1). In contrast, cisplatin mono- Antitumor activity of cetuximab in combination with the therapy was ineffective in this colon carcinoma model at all cytotoxic agents cisplatin or CPT-11 was assessed in a series tested dose levels, including its near MTD of 4.5 mg/kg/ of chemotherapy refractory (i.e., resistant to treatment) or injection (Table 2). Interestingly, combination regimens

Mol Cancer Ther 2006;5(1). January 2006

of cetuximab and cisplatin clearly showed improved antitumor activities over their respective monotherapies. For instance, the combination of 1 mg/mouse/injection of cetuximab with 4.5 mg/kg/injection cisplatin resulted in a LCK of 1.6 versus a maximum LCK of 0.8 for cetuximab monotherapy (P < 0.001 versus untreated control and cetuximab monotherapy; Fig. 2; Table 2). Moreover, higher doses of cisplatin were associated with better antitumor response rates in the cetuximab plus cisplatin combination regimens, although cisplatin treatment by itself did not show any activity (Table 2). We also attempted to extend the observation of therapeutic synergy obtained in the GEO model to another cisplatin-refractory xenograft tumor, the A549 human lung carcinoma. As described previously, cetuximab monotherapy was at best borderline active in some of the experiments employing this tumor model (Table 1). In this particular combination study with cisplatin, cetuximab monotherapy at 1 mg/mouse/injection (q3dx5;21, i.p.) failed to achieve mean tumor growth inhibitions in excess of 50% for at least one continuous TVDT resulting in a 0.3 Figure 2. Preclinical antitumor activity of cetuximab plus cisplatin LCK. Cisplatin at an MTD of 6 mg/kg/injection was combination chemotherapy versus s.c. GEO human colon carcinoma. Horizontal line, 1 LCK at tumor target size of 1.0 g. Arrows, days of drug likewise ineffective (0.3 LCK). Moreover, when cetuximab treatment. therapy was added to the cisplatin regimen, there was no significant increase in antitumor activity over the monotherapy regimens. Hence, combinations of these two agents Combinations of cetuximab with CPT-11 were tested did not improve the overall efficacy profile in the A549 initially in the HT29 human colon carcinoma model. As human lung carcinoma model. These results are summa- described previously (Table 1), cetuximab monotherapy rized in Table 2. was inactive in this tumor model. CPT-11 monotherapy

Table 2. Preclinical antitumor activity of cetuximab plus cisplatin combination therapy

Tumor (type) Cetuximab dose Cisplatin dose Schedule LCK at tumor (mg/mouse/injection), i.p. (mg/kg/injection), i.v. target size of 1 g

GEO (colon) 1 — q3dx5,14 0.8* — 4.5 q3dx5,14 0.0 c 1 4.5 q3dx5,14 1.6*, b 1 3 q3dx5,14 1.6*, 1 2 q3dx5,14 1.4* 0.5 4.5 q3dx5,14 2.0*,x b 0.5 3 q3dx5,14 1.5*, 0.5 2 q3dx5,14 1.5*,x b 0.25 4.5 q3dx5,14 1.6*, b 0.25 3 q3dx5,14 1.5k, 0.25 2 q3dx5,14 1.1* A549 (lung) 1 — q3dx5,21 0.3 — 6 q3dx5,21 0.3 1 6 q3dx5,21 0.5{ 1 4.5 q3dx5,21 0.1 1 3 q3dx5,21 0.4 0.25 6 q3dx5,21 0.3 0.25 4.5 q3dx5,21 0.2 0.25 3 q3dx5,21 0.2

*Statistically significant at P < 0.001 versus untreated control. cStatistically significant at P < 0.001 versus the most effective monotherapy regimen tested (cetuximab for GEO; cisplatin for A549). bStatistically significant at P < 0.01 versus the most effective monotherapy regimen tested (cetuximab for GEO; cisplatin for A549). xStatistically significant at P < 0.05 versus the most effective monotherapy regimen tested (cetuximab for GEO; cisplatin for A549). kStatistically significant at P < 0.01 versus untreated control. {Statistically significant at P < 0.05 versus untreated control.

Mol Cancer Ther 2006;5(1). January 2006

inhibited HT29 tumor growth to a borderline active rating. CPT-11 was active, achieving 1.1 LCK at its MTD of 48 For example, a LCK of 0.7 (P < 0.01 versus untreated mg/kg/injection (P < 0.001 versus untreated control). control) was achieved for a tumor target size of 1.0 g at the Adding cetuximab to CPT-11 failed to significantly MTD of 48 mg/kg/injection. However, adding cetuximab increase antitumor activity over CPT-11 monotherapy therapy to CPT-11 failed to significantly increase antitumor (Table 3). activity over CPT-11 monotherapy, and no synergy was attained in this tumor model (Table 3). We also tested CPT-11 and cetuximab combinations in Discussion the A549 human lung carcinoma model. As described Cetuximab monotherapy showed unequivocal activity in 2 previously, mean tumor growth inhibitions in excess of of 11 evaluated human xenograft tumor models (L2987, 50% for at least one continuous TVDT period were human lung carcinoma; Geo, human colon carcinoma). achieved for cetuximab treatment in only four of six Modest levels of tumor growth suppression were seen experiments, one of which was this particular combination on occasion in two additional models, WiDr (human colon therapy experiment. Cetuximab single-agent administra- carcinoma) and A549 (human lung carcinoma). In contrast, tion was borderline active and achieved a tumor growth cetuximab failed completely to show any antitumor delay corresponding to 0.7 LCK (P < 0.05 versus untreated activity in the majority of evaluated EGFR-positive tumors, control). CPT-11 monotherapy was inactive at its MTD of including Colo205 (human colon carcinoma), HCC70 48 mg/kg/injection (0.3 LCK). However, adding CPT-11 (human breast carcinoma), HCT116 (human colon carcino- to cetuximab therapy failed to significantly increase anti- ma), LOVO (human colon carcinoma), N87 (human tumor activity over the cetuximab monotherapy regimen gastric carcinoma), LX-1 (human lung carcinoma), and (Table 3). HT29 (human colon carcinoma). Although tumor EGFR- Lastly, cetuximab and CPT-11 combination regimens positivity may be a prerequisite for achieving antitumor were evaluated in the WiDr human colon carcinoma activity with cetuximab, our data suggests that the relative tumor model. As described previously (Table 1), cetux- tumor receptor levels do not predict responsiveness to the imab has weak and only occasional borderline activity in antibody and do not correlate with cetuximab efficacy. this xenograft model. In this particular combination Moreover, basal levels of relative activated EGFR (pY1068 therapy experiment, cetuximab was inactive and failed EGFR/total EGFR) do not directly correlate with the status to achieve mean tumor growth inhibitions in excess of of sensitivity to cetuximab therapy. These results draw 50% for at least one continuous TVDT period. In contrast, multiple parallels to the clinical experience with cetuximab,

Table 3. Preclinical antitumor activity of cetuximab plus CPT-11 combination therapy

Tumor (type) Cetuximab dose CPT-11 dose Schedule LCK at tumor (mg/mouse/injection), i.p. (mg/kg/injection), i.v. target size of 1 g

HT29 (colon) 1 — q3dx5,13 0.1 — 48 q3dx5,13 0.7* 1 48 q3dx5,13 0.8* 1 36 q3dx5,13 0.6* 0.5 48 q3dx5,13 1.0* 0.5 36 q3dx5,13 0.7* 0.25 48 q3dx5,13 0.8* 0.25 36 q3dx5,13 0.7* c A549 (lung) 1 — q3dx5,26 0.7 — 48 q3dx5,26 0.3* — 36 q3dx5,26 0.4* b 1 48 q3dx5,26 0.9 1 36 q3dx5,26 0.6* 0.25 48 q3dx5,26 0.7* 0.25 36 q3dx5,26 0.7* WiDr (colon) 1 — q3dx5,14 0.0 b — 48 q3dx5,14 1.1 b 1 48 q3dx5,14 1.1 b 1 36 q3dx5,14 0.9 b 0.25 48 q3dx5,14 1.0 b 0.25 36 q3dx5,14 0.9

*Statistically significant at P < 0.01 versus untreated control. cStatistically significant at P < 0.05 versus untreated control. bStatistically significant at P < 0.001 versus untreated control.

Mol Cancer Ther 2006;5(1). January 2006

where monotherapy activity with this antibody has been erlotinib sensitivity cluster around the active site of the likewise low, showing f11% overall response rates in kinase, presumably stabilizing the interaction between the EGFR-positive, irinotecan-refractory metastatic colorectal EGFR-tyrosine kinase domain and its competitive inhibitor cancer patients (5). Although increased gene copy number (e.g., gefitinib and erlotinib; ref. 44). Because cetuximab is for EGFR may correlate with clinical response to cetuximab interacting with the extracellular domain of the receptor, it (32), others have reported that lower mRNA levels of EGFR is likely that the antibody will act independently of these may provide a better predictive marker of positive clinical mutations. Thus, antibody-based EGFR inhibitors may outcome in colorectal cancer (33). Furthermore, clinical have different overall activity profiles compared with observations showed that response to cetuximab was not gefitinib and erlotinib, although they all target the same related to the relative strength of EGFR protein expression molecular entity/pathway. Consequently, cetuximab, and of the tumor specimen (23). In fact, one report even showed perhaps other related antibodies targeting the extracellular that cetuximab had activity in colorectal cancer patients domain of the EGFR, may have different molecular and with tumors that do not express the EGFR protein at all as genetic determinants that predict sensitivity compared determined by current immunohistochemistry techniques with the small molecule inhibitors. Clearly, further studies (34). Therefore, it is likely that molecular events other are needed to fully address this question. than basal EGFR levels (total and/or activated) are the true Previous reports have shown that cetuximab can poten- determinants of the tumor responsiveness to cetuximab tiate antitumor effects of many cytotoxic agents, including treatment. For example, the dynamics of the EGFR cisplatin and CPT-11 (16, 45). However, few reports have activation status in tumors (i.e., changes in EGFR activation tried to rationalize the determinants that predict therapeu- and its associated downstream pathway components upon tic synergy. Here, we report that cetuximab and cisplatin antibody treatment) may have a much better correlation combination regimens proved to be significantly more with efficacy and sensitivity to cetuximab treatment, which efficacious than either monotherapy alone in the cisplatin warrants further investigation. nonresponsive GEO human colon carcinoma tumor model. Interestingly, clinical data suggest that response to GEO tumors were also one of the most sensitive models the small molecule EGFR tyrosine kinase inhibitors, tested for cetuximab monotherapy regimens. In contrast, gefitinib (Iressa) and erlotinib (Tarceva), may be dependent the same cetuximab and cisplatin combination was without on the number of EGFR gene copies and protein expression benefit in the cisplatin poorly responsive A549 human lung levels (35, 36). Moreover, recent reports show that somatic carcinoma model. In this particular experiment, cetuximab mutations in the tyrosine kinase domain of the EGFR gene monotherapy was likewise inactive. Similarly, combina- significantly correlated with the clinical response to tions of CPT-11 with cetuximab failed to show any clear gefitinib in patients with chemotherapy-refractory non– improvements over the best monotherapy used in the small cell lung cancer (37, 38). Specifically, nearly all CPT-11 weakly responsive HT29, A549, and WiDr tumor gefitinib-responsive patients harbored mutations in the models. Both HT29 and WiDr experiments showed clear EGFR kinase domain, whereas no mutations have been resistance to cetuximab monotherapy. Interestingly, in the identified in nonresponsive, refractory cases. The main A549 combination study, cetuximab alone approached EGFR mutations identified in these studies included single-agent activity, and the best antitumor effect occurred missense mutations (L858R) and in-frame deletions (L747- using a combination of cetuximab with CPT-11. Although P753) in the kinase domain, which corresponded with this effect was not statistically different from the best enhanced EGFR and downstream pathway activation monotherapy response, it may indicate a general trend. and were associated with increased sensitivity to inhibition Collectively, our data suggest that robust cetuximab single- by gefitinib in vitro compared with wild-type receptor agent efficacy in a particular tumor may be a prerequisite to (37–39). A similar scenario has been reported for the clearly potentiate chemotherapy-mediated antitumor activ- related small molecule ATP-competitive EGFR tyrosine ities, even in chemotherapy nonresponsive tumor models. kinase inhibitor, erlotinib (Tarceva), where five of seven This hypothesis is further strengthened by recently tumors sensitive to erlotinib had somatic mutations in published data, indicating therapeutic synergy of cetux- EGFR analogous to gefitinib; however, all of 10 erlotinib- imab with BMS-275183, an oral taxane, in the cetuximab- refractory tumors were mutation free (40, 41). Although, sensitive L2987 and GEO tumor models (19). Moreover, Pao et al. showed some differences in EGFR activation others have reported synergy of cetuximab with drugs, among the mutants compared with other published such as gemcitabine, paclitaxel, docetaxel, topotecan, data, the relative changes in sensitivity to the two small 5-fluorouracil, doxorubicin, and radiation (7, 9, 14, 15, 17, molecule inhibitors were similar (i.e., changes in sensitivity 18, 20, 31, 46, 47). The tumor models used in these studies to erlotinib tracked with changes in sensitivity to gefitinib). were all at least borderline responsive to cetuximab Preliminary evidence suggests that the discussed EGFR monotherapy. Prewett et al. (16) previously showed mutations associated with sensitivity to the small molecule enhanced antitumor activity of cetuximab in combination inhibitors are not required for response to cetuximab nor with CPT-11 in the HT29 human colon carcinoma xeno- do they seem predictive of response to cetuximab in the graft, which stands in contrast to our observation in the clinical setting (42, 43). This may be due in part to the fact same tumor model. However, in their hands, HT29 tumor that the sequence alterations published for gefitinib and growth was inhibited significantly more effectively by

Mol Cancer Ther 2006;5(1). January 2006

cetuximab monotherapy than we report, achieving tumor 7. Prewett M, Rockwell P, Rockwell RF, et al. The biologic effects of C225, a chimeric monoclonal antibody to the EGFR, on human prostate growth inhibitions in excess of 50% for at least one carcinoma. J Immunother Emphasis Tumor Immunol 1996;19:419 – 27. continuous TVDT. As such, HT29 would be classified at 8. Prewett M, Rothman M, Waksal H, Feldman M, Bander NH, Hicklin DJ. the minimum as a borderline-sensitive tumor model, Mouse-human chimeric anti-epidermal growth factor receptor antibody C225 inhibits the growth of human renal cell carcinoma xenografts in nude thereby perhaps explaining the ability to potentiate anti- mice. Clin Cancer Res 1998;4:2957 – 66. tumor activity of the chemotherapeutic agent. 9. Ciardiello F, Bianco R, Damiano V, et al. Antitumor activity of In summary, cetuximab showed strikingly similar activ- sequential treatment with topotecan and anti-epidermal growth factor ity profiles in preclinical models of EGFR-positive tumors receptor monoclonal antibody C225. Clin Cancer Res 1999;5:909 – 16. compared with the clinical setting. The identification of 10. Saleh MN, Raisch KP, Stackhouse MA, et al. Combined modality therapy of A431 human epidermoid cancer using anti-EGFr antibody C225 accurate biomarkers predicting the response to cetuximab and radiation. Cancer Biother Radiopharm 1999;14:451 – 63. should be aggressively pursued as they may not only 11. Bianco C, Bianco R, Tortora G, et al. Antitumor activity of combined significantly improve cetuximab monotherapy response treatment of human cancer cells with ionizing radiation and anti-epidermal growth factor receptor monoclonal antibody C225 plus type I protein rates but could also positively affect the success rate in kinase A antisense oligonucleotide. Clin Cancer Res 2000;6:4343 – 50. treating patients in combination with various cytotoxic 12. Ciardiello F, Bianco R, Damiano V, et al. Antiangiogenic and antitumor agents and/or radiation. Pharmacogenomic and proteomic activity of anti-epidermal growth factor receptor C225 monoclonal approaches may be suitable to identify a set of genes or antibody in combination with vascular endothelial growth factor antisense oligonucleotide in human GEO colon cancer cells. Clin Cancer Res proteins, which will accurately reflect active and intact 2000;6:3739 – 47. signaling through the EGFR and correlate with responsive- 13. Milas L, Mason K, Hunter N, et al. In vivo enhancement of tumor ness to cetuximab treatment. Conversely, some markers radioresponse by C225 antiepidermal growth factor receptor antibody. may be discovered that predict resistance to the antibody Clin Cancer Res 2000;6:701 – 8. therapy. Most recently, attempts were made to address the 14. Overholser JP, Prewett MC, Hooper AT, Waksal HW, Hicklin DJ. Epidermal growth factor receptor blockade by antibody IMC-C225 inhibits latter alternative. More specifically, Skvortsov et al. showed growth of a human pancreatic carcinoma xenograft in nude mice. Cancer that they were able to identify proteins potentially involved 2000;89:74 – 82. in resistance mechanisms against cetuximab therapy by 15. Buchsbaum DJ, Bonner JA, Grizzle WE, et al. Treatment of pancreatic cancer xenografts with Erbitux (IMC-C225) anti-EGFR antibody, gemcita- selecting two colorectal cancer cell lines with a high bine, and radiation. Int J Radiat Oncol Biol Phys 2002;54:1180 – 93. expression of EGFR but a different sensitivity toward 16. Prewett MC, Hooper AT, Bassi R, Ellis LM, Waksal HW, Hicklin DJ. treatment with the antibody (48). By comparing the Enhanced antitumor activity of anti-epidermal growth factor receptor proteomic profile of these cell lines, significant differences monoclonal antibody IMC-C225 in combination with irinotecan (CPT-11) against human colorectal tumor xenografts. Clin Cancer Res 2002;8: in the expression of 14 proteins were detected, which 994 – 1003. may serve as markers predicting resistance to this antibody 17. Nakata E, Hunter N, Mason K, Fan Z, Ang KK, Milas L. C225 therapeutic. Likewise, attempts were recently made to antiepidermal growth factor receptor antibody enhances the efficacy of identify biomarkers predictive of response to cetuximab by docetaxel chemoradiotherapy. Int J Radiat Oncol Biol Phys 2004;59: 1163 – 73. profiling cetuximab-sensitive and cetuximab-resistant 18. Nasu S, Ang KK, Fan Z, Milas L. C225 antiepidermal growth factor tumor cell lines as well as in vivo tumor xenografts receptor antibody enhances tumor radiocurability. Int J Radiat Oncol Biol (49, 50). Using pharmacogenomic and proteomic Phys 2001;51:474 – 7. approaches, several candidate genes and proteins were 19. Rose WC, Wild R. Therapeutic synergy of oral taxane BMS-275183 and cetuximab versus human tumor xenografts. Clin Cancer Res 2004;10: identified, which are currently being assessed for their 7413 – 7. predictive value in the clinical setting. 20. Inoue K, Slaton JW, Perrotte P, et al. Paclitaxel enhances the effects of the anti-epidermal growth factor receptor monoclonal antibody ImClone Acknowledgments C225 in mice with metastatic human bladder transitional cell carcinoma. Clin Cancer Res 2000;6:4874 – 84. We thank Drs. Robert Kramer and Jack Hunt for their support and critical 21. Sato JD, Kawamoto T, Le AD, Mendelsohn J, Polikoff J, Sato GH. comments on the article. Biological effects in vitro of monoclonal antibodies to human epidermal growth factor receptors. Mol Biol Med 1983;1:511 – 29. References 22. Goldenberg A, Masui H, Divgi C, Kamrath H, Pentlow K, Mendelsohn J. Imaging of human tumor xenografts with an indium-111-labeled anti- 1. Mendelsohn J. The epidermal growth factor receptor as a target for epidermal growth factor receptor monoclonal antibody. J Natl Cancer Inst cancer therapy. Endocr Relat Cancer 2001;8:3 – 9. 1989;81:1616 – 25. O’Dwyer PJ, Benson AB, III. Epidermal growth factor receptor-targeted 2. 23. Saltz LB, Meropol NJ, Loehrer PJ, Sr., Needle MN, Kopit J, Mayer RJ. therapy in colorectal cancer. Semin Oncol 2002;29:10 – 7. Phase II trial of cetuximab in patients with refractory colorectal cancer that 3. Rowinsky EK. The erbB family: targets for therapeutic development expresses the epidermal growth factor receptor. J Clin Oncol 2004;22: against cancer and therapeutic strategies using monoclonal antibodies and 1201 – 8. tyrosine kinase inhibitors. Annu Rev Med 2004;55:433 – 57. 24. Nicholson RI, Gee JM, Harper ME. EGFR and cancer prognosis. Eur J 4. Mendelsohn J. Antibody-mediated EGF receptor blockade as an Cancer 2001;37 Suppl 4:S9 – 15. anticancer therapy: from the laboratory to the clinic. Cancer Immunol 25. Baselga J, Trigo JM, Bourhis J, et al. Phase II multicenter study of the Immunother 2003;52:342 – 6. antiepidermal growth factor receptor monoclonal antibody cetuximab in 5. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and combination with platinum-based chemotherapy in patients with platinum- cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal refractory metastatic and/or recurrent squamous cell carcinoma of the cancer. N Engl J Med 2004;351:337 – 45. head and neck. J Clin Oncol 2005;23:5568 – 77. 6. Goldstein NI, Prewett M, Zuklys K, Rockwell P, Mendelsohn J. 26. Saulnier MG, Balasubramanian BN, Long BH, et al. Discovery of a Biological efficacy of a chimeric antibody to the epidermal growth factor fluoroindolo[2,3-a]carbazole clinical candidate with broad spectrum anti- receptor in a human tumor xenograft model. Clin Cancer Res tumor activity in preclinical tumor models superior to the marketed 1995;1:1311 – 8. oncology drug, CPT-11. J Med Chem 2005;48:2258 – 61.

Mol Cancer Ther 2006;5(1). January 2006

27. Rose WC, Crosswell AR, Schurig JE, Casazza AM. Preclinical 40. Pao W, Miller VA. Epidermal growth factor receptor mutations, small- antitumor activity of orally administered platinum (IV) complexes. Cancer molecule kinase inhibitors, and non-small-cell lung cancer: current Chemother Pharmacol 1993;32:197 – 203. knowledge and future directions. J Clin Oncol 2005;23:2556 – 68. Epub 28. Gehan EA. A Generalized Wilcoxon test for comparing arbitrarily 2005 Mar 14. singly-censored samples. Biometrika 1965;52:203 – 23. 41. Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations are 29. Lavelle F, Bissery MC, Combeau C, Riou JF, Vrignaud P, Andre S. common in lung cancers from ‘‘never smokers’’ and are associated with Preclinical evaluation of docetaxel (Taxotere). Semin Oncol 1995;22: sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 3 – 16. 2004;101:13306 – 11. 30. Shaheen RM, Ahmad SA, Liu W, et al. Inhibited growth of colon 42. Lenz HJ, Mayer RJ, Gold PJ, et al. Activity of cetuximab in patients cancer carcinomatosis by antibodies to vascular endothelial and epidermal with colorectal cancer refractory to both irinotecan and oxaliplatin. Journal growth factor receptors. Br J Cancer 2001;85:584 – 9. of Clinical Oncology (2004 ASCO Annual Meeting Proceedings) 2004; 22:3510. 31. Karashima T, Sweeney P, Slaton JW, et al. Inhibition of angiogenesis by the antiepidermal growth factor receptor antibody ImClone C225 in 43. Mukohara T, Engelman JA, Hanna NH, et al. Differential effects androgen-independent prostate cancer growing orthotopically in nude of gefitinib and cetuximab on non-small-cell lung cancers bearing mice. Clin Cancer Res 2002;8:1253 – 64. epidermal growth factor receptor mutations. J Natl Cancer Inst 2005; 97:1185 – 94. 32. Moroni M, Veronese S, Benvenuti S, et al. Gene copy number for Herbst RS, Fukuoka M, Baselga J. Gefitinib: a novel targeted epidermal growth factor receptor (EGFR) and clinical response to antiEGFR 44. treatment in colorectal cancer: a cohort study. Lancet Oncol 2005;6: approach to treating cancer. Nat Rev Cancer 2004;4:956 – 65. 279 – 86. 45. Fan Z, Baselga J, Masui H, Mendelsohn J. Antitumor effect of anti-epidermal growth factor receptor monoclonal antibodies plus cis- 33. Vallbohmer D, Zhang W, Gordon M, et al. Molecular determinants of diamminedichloroplatinum on well established A431 cell xenografts. cetuximab efficacy. J Clin Oncol 2005;23:3536 – 44. Cancer Res 1993;53:4637 – 42. 34. Chung KY, Shia J, Kemeny NE, et al. Cetuximab shows activity in 46. Bruns CJ, Harbison MT, Davis DW, et al. Epidermal growth factor colorectal cancer patients with tumors that do not express the epidermal receptor blockade with C225 plus gemcitabine results in regression of growth factor receptor by immunohistochemistry. J Clin Oncol 2005;23: human pancreatic carcinoma growing orthotopically in nude mice by 1803 – 10. antiangiogenic mechanisms. Clin Cancer Res 2000;6:1936 – 48. 35. Cappuzzo F, Hirsch FR, Rossi E, et al. Epidermal growth factor 47. Baselga J, Norton L, Masui H, et al. Antitumor effects of doxorubicin receptor gene and protein and gefitinib sensitivity in non-small-cell lung in combination with anti-epidermal growth factor receptor monoclonal cancer. J Natl Cancer Inst 2005;97:643 – 55. antibodies. J Natl Cancer Inst 1993;85:1327 – 33. 36. Tsao MS, Sakurada A, Cutz JC, et al. Erlotinib in lung cancer: 48. Skvortsov S, Sarg B, Loeffler-Ragg J, et al. Different proteome molecular and clinical predictors of outcome. N Engl J Med 2005;353: pattern of epidermal growth factor receptor-positive colorectal cancer cell 133 – 44. lines that are responsive and nonresponsive to C225 antibody treatment. 37. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the Mol Cancer Ther 2004;3:1551 – 8. epidermal growth factor receptor underlying responsiveness of non-small- 49. Khambata Ford S, Perkins N-A, Yoganathan S, et al. Pharmacoge- cell lung cancer to gefitinib. N Engl J Med 2004;350:2129 – 39. nomic approaches for identifying markers predictive of tumor response to 38. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: Cetuximab (Erbitux). In: Proc Amer Assoc Cancer Res 2004, Orlando FL, correlation with clinical response to gefitinib therapy. Science 2004;304: March 2004. 1497 – 500. 50. Gao J, Wild R, Carney R, et al. Protein profiling approaches to identify 39. Sordella R, Bell DW, Haber DA, Settleman J. Gefitinib-sensitizing predictive markers of tumor response to Cetuximab (Erbitux) using EGFR mutations in lung cancer activate anti-apoptotic pathways. Science xenograft models. In: Proc Amer Assoc Cancer Res 2005, Anaheim CA, 2004;305:1163 – 7. April 2005.

Mol Cancer Ther 2006;5(1). January 2006

Downloaded from mct.aacrjournals.org on September 27, 2021. © 2006 American Association for Cancer Research. Cetuximab preclinical antitumor activity (monotherapy and combination based) is not predicted by relative total or activated epidermal growth factor receptor tumor expression levels

Robert Wild, Krista Fager, Christine Flefleh, et al.

Mol Cancer Ther 2006;5:104-113.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/5/1/104

Cited articles This article cites 48 articles, 22 of which you can access for free at: http://mct.aacrjournals.org/content/5/1/104.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/5/1/104.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/5/1/104. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.