Phase II Study of Pemetrexed-Gemcitabine Combination in Patients with Advanced-Stage Non-Small Cell Lung Cancer

Vol. 10, 5439–5446, August 15, 2004 Clinical Cancer Research 5439

Christian Monnerat,1 Thierry Le Chevalier,1 deaths attributed to treatment. Common Toxicity Criteria Karen Kelly,2 Coleman K. Obasaju,3 grade 3/4 toxicities were neutropenia (61.7%), febrile neu- Julie Brahmer,4 Silvia Novello,1 tropenia (16.7%), fatigue (23.3%), and elevations of aspar- 3 3 tate aminotransferase (15.0%) and alanine aminotrans- Takashi Nakamura, Astra M. Liepa, ferase (20.0%). 5 2 Laurence Bozec, Paul A. Bunn, Jr, and Conclusions: This combination had good tolerance and David S. Ettinger4 achieved promising overall survival with extended 1- and 1Institut Gustave Roussy, Villejuif, France; 2The University of 2-year survival rates. This cisplatin-free regimen warrants Colorado Cancer Center, Denver, Colorado; 3Eli Lilly and Company, further evaluation in randomized trials. Indianapolis, Indiana; 4The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore, Maryland; and 5Eli Lilly and Company, St. Cloud, France INTRODUCTION Pemetrexed (LY231514, Alimta; Eli Lilly and Company, ABSTRACT Indianapolis, IN) is a novel antifolate that is metabolized intra- Purpose: Cisplatin is one of the most active agents for cellularly to a pentaglutamate form. Pemetrexed inhibits three the treatment of non-small cell lung cancer (NSCLC). It is enzymes (thymidylate synthase, dihydrofolate reductase, and also known for significant toxicity, which makes it unsuit- glycinamide ribonucleotide transferase) involved in folate me- able for certain patients. Our purpose was to evaluate the tabolism and DNA synthesis (1–3). The cytotoxicity of pem- efficacy and toxicity of a promising cisplatin-free combina- etrexed is caused by inhibition of both the pyrimidine and purine tion, gemcitabine plus pemetrexed, in NSCLC. pathways. The maximum tolerated dose established in Phase I 2 Experimental Design: Chemo-naive patients with inop- studies of pemetrexed was 600 mg/m every 3 weeks (4, 5). This 2 erable NSCLC were eligible for this study. Gemcitabine dose was later reduced to 500 mg/m every 3 weeks in several (1250 mg/m2) was given intravenously on days 1 and 8, Phase II trials, particularly when used in combination with other followed by intravenous pemetrexed (500 mg/m2) on day 8. cytotoxic agents (6, 7). Folic acid and vitamin B12 supplementation became a requirement for pemetrexed-based therapy early After inclusion of 13 patients, folic acid and vitamin B12 supplementation was added to lower pemetrexed-induced in this study, when analysis of safety data from multiple pem- toxicity. Quality of life was assessed with the Lung Cancer etrexed trials suggested that supplementation may decrease tox- Symptom Scale. icity (8). Single-agent pemetrexed was tested in two Phase II Results: Sixty patients enrolled; 58 were evaluable for non-small cell lung cancer (NSCLC) trials enrolling 92 chemo- response. All patients had a World Health Organization naive patients (without vitamin supplementation) by Clarke performance status of 0 or 1. Eighty-seven percent had stage et al. (9) and Rusthoven et al. (7). Response rates of 15.8% and IV disease. Nine patients had a confirmed partial response 23.3% and median survivals of 7.2 and 9.2 months were ob- [overall response rate, 15.5%; 95% confidence interval (CI), served, respectively. In these two studies, the main toxicities 7.3–27.4%]. Twenty-nine (50.0%) patients had stable dis- reported were grade 3/4 neutropenia (42% and 39%) and grade ease. Median overall survival was 10.1 months (95% CI, 3/4 skin rash (31% and 39%). 7.9–13.0 months), with a 1- and 2-year overall survival of Gemcitabine (Gemzar; Eli Lilly and Company) is a 42.6% (95% CI, 30.0–55.3%) and 18.5% (95% CI, 7.9– pyrimidine antimetabolite that is intracellularly transformed to 29.1%). Median progression-free survival was 5.0 months. difluorodeoxycytidine triphosphate. Its incorporation into DNA Median response duration was 3.3 months. There were no results in chain termination. Gemcitabine also inhibits ribonucleotide reductase, an enzyme required for deoxynucleoside formation and DNA synthesis (10) Gemcitabine has been widely explored in NSCLC, as a single-agent and in combination. The United States Food and Drug Administration-approved Received 2/4/04; revised 4/6/04; accepted 4/13/04. dose is 1000 mg/m2 weekly for 3 weeks every 4 weeks (11, 12) Grant support: Supported by Eli Lilly and Company (Indianapolis, or 1250 mg/m2 weekly for 2 weeks every 3 weeks (13). In a IN). T. Le Chevalier, P. Bunn, and D. Ettinger have acted as consultants Phase II study of gemcitabine (1250 mg/m2) weekly for 3 of 4 of Eli Lilly and Company. The costs of publication of this article were defrayed in part by the weeks enrolling 161 patients with advanced NSCLC, the re- payment of page charges. This article must therefore be hereby marked sponse rate was 21.8%, and median survival was 11.5 months advertisement in accordance with 18 U.S.C. Section 1734 solely to (13). Main toxicities were grade 3 neutropenia, nausea, and indicate this fact. elevations of hepatic transaminases. Requests for reprints: Thierry Le Chevalier, Department of Medicine, Institut Gustave Roussy, 39 rue Camille-Desmoulins, 94805 Villejuif Because pemetrexed and gemcitabine have both shown Cedex, France. Phone: 33-142-114322; Fax: 33-142-115219; E-mail: tle- single-agent activity against a wide range of solid tumors in- [email protected]. cluding NSCLC (7, 9, 11, 12, 14, 15), the combination of these

two agents was evaluated in vitro, and cytotoxic synergy was to 100 ml of normal saline for intravenous infusion. Gemcitab- found when gemcitabine exposure preceded that of pemetrexed ine was supplied as a lyophilized powder in 200- or 1000-mg (16). Whereas preclinical studies suggested pemetrexed on day vials. The appropriate dose was prepared with normal saline to 1 with a 90-min interval between administrations, a Phase I make a solution containing approximately 10 mg/ml for intra- study led to administration of pemetrexed on day 8 due to better venous infusion. Chemotherapy was given on a 3-week schedule hematological tolerance (16). The recommended dose was gem- and consisted of 1250 mg/m2 gemcitabine over 30 min on days citabine (1250 mg/m2) on days 1 and 8, followed by pemetrexed 1 and 8 and 500 mg/m2 pemetrexed over 10 min on day 8, (500 mg/m2) on day 8 of a 3-week cycle. The most common approximately 90 min after the gemcitabine infusion. Dose dose-limiting toxicity was neutropenia. Other toxicities included adjustments within a cycle were as follows: in case of Common nausea, fatigue, rash, and elevated hepatic transaminases (16). Toxicity Criteria Version 2.0 (18) grade 3/4 neutropenia, throm- The objective of the present Phase II study was to deter- bocytopenia, or nonhematological toxicity (except grade 3 trans- mine the antitumor activity of the pemetrexed-gemcitabine com- aminase elevation) after day 1, the day 8 doses were held until bination, a cisplatin-free outpatient regimen, in chemo-naive resolution of the toxicity, and pemetrexed and gemcitabine were patients with locally advanced or metastatic NSCLC. Secondary given at full dose in case of a grade 3 toxicity and at 50% end points were to assess toxicity of treatment, evaluate overall reduced dose in case of a grade 4 toxicity. No dose escalation and progression-free survival, and assess the impact on symp- was allowed after a dose reduction for the remainder of the toms and quality of life. study. The next cycle was not initiated until 14 days had elapsed from the administration of a delayed day 8 dose and until all toxicities resolved. Treatment could be delayed up to 35 days PATIENTS AND METHODS from the day 1 dose to allow a patient sufficient time to recover Patient Selection. Eligible patients had histological or from all toxicities. If a further cycle could not be given by day cytologic diagnosis of stage IIIB or IV NSCLC, with evidence 35, the patient was discontinued from study. For subsequent of bidimensionally measurable disease. Any clinically signifi- cycles, pemetrexed and gemcitabine doses were adjusted ac- cant effusion had to be drained and pleurodesis had to be cording to neutrophil and platelet nadir counts or maximum performed at least 2 weeks before baseline imaging. Previous nonhematological toxicity from the preceding cycle. Treatment systemic chemotherapy, as well as prior radiation therapy to was interrupted for a persistent drop of creatinine clearance target lesions, was not permitted. Other eligibility criteria in- under 45 ml/min or the appearance of a clinically significant cluded the following: age Ն 18 years; World Health Organiza- effusion. Beginning on December 10, 1999, folic acid (350– ␮ ␮ tion performance status of 0 or 1; calculated creatinine clear- 1000 g/day orally) and vitamin B12 (1000 g) i.m. every 9 ance Ն 45 ml/min (based on modified Cockcroft and Gault weeks became mandatory, starting 1 to 2 weeks before initiation method; Ref. 17); adequate hepatic function (i.e., serum biliru- of chemotherapy. Dexamethasone (4 mg or equivalent) was bin of Յ1.5 times the upper normal limit; alkaline phosphatase, given orally twice daily the day before, the day of, and the day aspartate aminotransferase, and alanine aminotransferase of after pemetrexed infusion to prevent rash. Antiemetic treatment Յ Յ 3.0 times the upper normal limit or 5.0 times the upper with a 5HT3 antagonist and dexamethasone was recommended normal limit in case of liver metastases); adequate bone marrow before each chemotherapy infusion. Dexamethasone was rec- function (absolute neutrophil count Ն 1.5 ϫ 109/liter, platelet ommended for up to 2 days after the infusion. Salicylates or count Ն 150 ϫ 109/liter, and hemoglobin Ն 9 g/dl); albumin Ն other nonsteroidal anti-inflammatory agents were interrupted 2 3.0 g/dl; absence of Ͼ10% weight loss in the 6 weeks before days before (5 days for long- acting agents), the day of, and 2 inclusion; absence of serious concomitant illness that would days after pemetrexed infusion to avoid any interaction with compromise chemotherapy delivery; absence of second primary renal function. malignancy (except cervical carcinoma in situ and nonmelano- Measurement of Study End Points. This multicenter matous skin cancer that was adequately treated); absence of study had response rate as the primary end point, and the Simon clinical evidence of brain metastases; effective contraceptive optimum sequential two-stage design was used (19). In stage 1, methods in patients with reproductive potential and negative at least 7 of 30 enrolled patients had to have an objective tumor pregnancy test for women before study entry; and written in- response to proceed to stage 2, where 30 additional patients formed consent. were to be treated, for a total of 60 patients. After this trial began, evidence of excessive pemetrexed- All patients who received one dose of chemotherapy were related toxicity in some patients treated in other studies was assessable for toxicity. A clinical and laboratory assessment was

linked to folic acid and vitamin B12 deficiency, as measured by performed weekly, and toxicities were graded according to an elevated plasma homocysteine level (8). Thus, after Decem- Common Toxicity Criteria. Within 3 weeks before the first ber 10, 1999, all pemetrexed-based trials were amended, and cycle, tumor measurements were performed by computed to-

systematic folic acid and vitamin B12 supplementation was mography scan or magnetic resonance imaging. The same provided to all patients. method of measurement was performed before every other cycle This study was conducted according to the Helsinki Dec- of therapy. Confirmation of response was performed 3–4 weeks laration, and the protocol was approved by each local institu- after the first evidence of response. Patients were evaluable for tional review board and local ethical review committee. response if they had a baseline exam and at least one follow-up Drug Administration. Pemetrexed was supplied as a exam and had received at least one cycle of chemotherapy. All 40-mg/ml aqueous solution in vials of 200 or 1000 mg of the sites with measurable lesions were followed for response. Meas- base compound. The appropriate dose of pemetrexed was added urements of unidimensional (i.e., single largest dimension) and

bidimensional lesions (i.e., the product of the largest diameter Table 1 Patient and tumor characteristics and its largest perpendicular) were summed at each assessment, Variables No. of patients (%) and the best response on study was recorded. A complete N 60 response required the disappearance of all clinical and radiolog- Age (yrs) ical evidence of tumor for at least 3–4 weeks. A partial response Median (range) 58 (34–73) required a Ն50% decrease in the sum of the products of the Sex diameters of all measurable lesions for at least 3–4 weeks. Female 22 (36.7) Ն Male 38 (63.3) Appearance of any new lesion or a 25% increase in the sum of WHO performance status the products of the diameters of measurable lesions compared 0 20 (66.7) with the smallest sum constituted progressive disease. Stable 1 40 (33.3) disease was a tumor decrease less than a partial response and a Histology Adenocarcinoma 31 (51.7) tumor increase less than progressive disease. Overall survival Squamous cell carcinoma 10 (16.7) was defined as the time from study entry to the date of death or Large cell carcinoma 7 (11.7) last follow-up. Progression-free survival was defined as the time Bronchioloalveolar carcinoma 5 (8.2) from study entry until the date of disease progression or death NSCLC, not specified 7 (11.7) from any cause or the date of last follow-up. Response duration Stage at study entry IIIB 8 (13.3) was defined as the time of a complete response or partial IV 52 (86.7) response until disease progression. All survival estimates were Sites of disease calculated by the Kaplan-Meier method (20). Lung 49 (84.5) Before each cycle, symptoms and quality of life data were Lymph node 35 (60.3) Pleural effusion 12 (20.7) collected with the Lung Cancer Symptom Scale (LCSS), as Adrenal gland 11 (19.0) reported by patients and investigators (21). The patient LCSS Liver 11 (19.0) items are assessed as 100-mm visual analog scales, and scores Bone 11 (19.0) were reported as 0 to 100, with 0 representing the best score. No. of organ sites involved Clinically meaningful changes were defined as Ն10-mm 1 6 (10.0) Ն1 54 (90.0) changes from baseline that were maintained for at least two cycles. Patients were considered symptomatic if the baseline Abbreviation: WHO, World Health Organization. assessment was Ն10 mm. Observer LCSS scores were reported on 5-point categorical scales, and clinically meaningful changes were defined as Ն1-point changes from baseline that were maintained for at least two cycles. Patients with baseline assess- occurring at cycles 6, 9, 11, and 12. Reasons for dose reduction ment and at least one on-study assessment were evaluable for included neutropenia (82.5%), febrile neutropenia (7.5%), leu- each of the LCSS analyses. Symptoms were classified as im- kopenia (5.0%), and fatigue (5.0%). A dose delay was required proved, worsened, or stable, based on the pattern of change; in 28 patients (46.7%) and in 45 cycles (19.7%) . For pem- 2 symptoms for patients without a confirmatory assessment were etrexed, the theoretical dose intensity was 166.7 mg/m /week, 2 classified as unknown. and the actual mean dose intensity received was 145.4 mg/m / week, representing a relative dose intensity of 87.2%. For gemcitabine, the theoretical dose intensity was 833.3 mg/m2/week, RESULTS and the actual mean dose intensity received was 644.0 mg/m2/ Patients. From August 1999 to May 2001, 72 patients week, representing a relative dose intensity of 77.3%. were considered for inclusion and signed informed consent. Antitumor Activity. Fifty-eight patients were evaluable After fulfillment of all baseline exams, 11 patients did not meet for tumor response (seven had baseline computed tomography the inclusion criteria (7 had inadequate organ function, 1 did not scans performed within 28 days of initiation of chemotherapy, have histological/cytologic confirmation, 1 had inadequate per- instead of 21 days as defined by the protocol). Two of the formance status, 1 had excess weight loss, and 1 had both enrolled patients were not evaluable: one patient had a baseline inadequate organ function and performance status), and 1 pa- computed tomography scan done 42 days before treatment; and tient withdrew consent. The demographic and disease charac- the other had no baseline computed tomography scan available. teristics of the 60 eligible patients are listed in Table 1. Among No patient had a complete response, but 9 patients (7 of them the 60 eligible patients, 47 received vitamin supplementation from the 30 patients in stage I of the trial) had a confirmed throughout their course of therapy, 10 received supplementation partial response, for an overall response rate of 15.5% [95% during part of their therapy, and 3 completed the study before confidence interval (CI), 7.3–27.5%]. Twenty-nine patients vitamin supplementation was initiated. (50.0%) had stable disease, 12 patients (20.7%) had progressive Drug Administration. A total of 229 cycles were ad- disease, and 8 patients (13.8%) were not evaluated (6 patients ministered (median, 4; range, 1–12). Thirty-two patients went off study before tumor evaluation, and 2 patients did not (53.3%) received full doses of chemotherapy, 23 patients have a subsequent measure of same lesions recorded at base- (38.3%) had dose reductions, and 5 patients (8.3%) had one line). Therefore, 38 of the evaluable patients (65.5%) had either omitted dose. Dose reduction was required in a total of 40 cycles a partial response or stable disease. Responses were seen in (17.5%), 20 (50.0%) occurring at cycle 2, 6 occurring at cycle 3, lung, lymph nodes, liver, soft tissue, and pleura. Thirty-three 7 occurring at cycle 4, 3 occurring at cycle 5, and 1 each patients received post-study chemotherapy, primarily platinum-

based combination therapy. Median overall survival of all 60 patients was 10.1 months (95% CI, 7.9–13.0 months); median overall survival was 13.8 months (95% CI, 6.7–21.7 months) for partial response patients, 13.6 months (95% CI, 10.1–23.7 months) for stable disease patients, and 4.9 months (95% CI, 3.4–8.6 months) for progressive disease patients. The 1- and 2-year overall survival rates were 42.6% and 18.5%, respectively (see Fig. 1). Median progression-free survival was 5.0 months (95% CI, 3.5–6.3 months; see Fig. 2). Median duration of response was 3.3 months (95% CI, 2.7–7.1 months). Lung Cancer Symptom Scale. Fifty-nine patients completed the patient LCSS at baseline. The percentage of patients considered symptomatic were as follows: anorexia, 63.7%; fatigue, 75.9%; cough, 67.2%; dyspnea, 83.1%; hemoptysis, 13.8%; and pain, 57.6%. Rates reported on the observer LCSS were similar. Fifty-five patients were included in at least one of the patient LCSS analyses. The results of the symptom items are displayed in Fig. 3. Most patients reported their symptoms as stable or improved. The highest rates of improvement were seen with anorexia, cough, and pain. The majority of patients reported no hemoptysis at baseline, so they had no opportunity to improve. The highest rate of symptoms worsening was seen Fig. 2 Kaplan-Meier estimates for progression-free survival of all 60 with fatigue. Similar results were seen with the observer LCSS. patients. For the patient LCSS summary items (symptom distress, inter- ference with activity level, and global quality of life), most patients reported these as stable or improved. Patients with a partial response were more likely to be classified as improved patients (5.0%) had grade 3/4 thrombocytopenia. There was no than stable for each of the LCSS items. evidence of any cumulative hematological toxicity. Toxicity. All 60 patients were assessable for toxicity. A summary of nonhematological and biochemical tox- Hematological toxicities by worst grade for all patients are listed icity is listed in Table 3. The regimen was generally well in Table 2. Anemia was generally mild, with 9 patients (15.0%) tolerated, with no grade 3/4 nausea, vomiting, or sensory requiring a blood transfusion. Grade 3 and 4 neutropenia was neuropathy. Pemetrexed-related skin rash was frequent, oc- observed in 28.3% and 33.3% of the patients, respectively, and curring in 23 patients (38.3%), but only 2 cases (3.3%) were 9 patients (15.0%) experienced febrile neutropenia. Only 3 grade 3. Fatigue was the most frequent nonhematological toxicity. Grade 3 fatigue occurred in 14 patients (23.3%) after a median of 3 cycles and did not appear to be related to either age or gender. One patient on long-term corticosteroids for severe lung dyspnea was diagnosed with pneumocyctis cari- nii pneumonia at cycle 3, and two additional patients experienced documented pneumonitis of noninfectious origin. One patient experienced grade 4 diarrhea requiring hospital- ization. Two patients developed pulmonary emboli, which were judged nonrelated to study medication; one resulted in death at cycle 1, and the other developed during a hospital- ization for febrile neutropenia. There were no deaths attributed to chemotherapy treatment. Seven patients experienced alopecia; all but one were grade 1. Other toxicities were rare and generally mild. One patient with a history of past and present alcoholism had a grade 4 hyperbilirubinemia due to an underlying liver disorder and went off therapy at cycle 3. Vitamin Supplementation. Because vitamin supplementation was complete for 47 patients, partial for 10 patients, and absent for 3 patients, we examined whether vitamin supplementation had an effect on the clinical outcome and rate of toxicities. No numerical difference (data not shown) was noticed for the response rate, overall survival, LCSS scores, median number of chemotherapy cycles admin- Fig. 1 Kaplan-Meier estimates for overall survival of all 60 patients. istered, or toxicity rates.

Fig. 3 Summary of the changes in the patient LCSS. To be included in the analysis, patients must have had baseline and at least one post-baseline on-study assessment; patients not pre- sented did not meet criteria for evaluation or were classified as “unknown.”

Table 2 Hematological toxicity (N ϭ 60) Phase II trials in which survival has been considerably more Toxicity grade than twice the TTP, subsequent Phase III trials have failed to confirm patient benefit. In this trial, the median TTP was 5.0 1234months, which was approximately half of the overall median Hemoglobin 2 (3.3) 7 (11.7) 7 (11.7) 0 (0.0) survival. This median TTP compares favorably with other dou- Leukocytes 8 (13.3) 14 (23.3) 21 (35.0) 4 (6.7) blet combinations in the literature. Neutrophils 4 (6.7) 3 (5.0) 17 (28.3) 20 (33.3) Febrile neutropenia NA NA 8 (13.3) 1 (1.7) Response rates in advanced NSCLC have not always cor- Thrombocytes 24 (40.0) 6 (10.0) 3 (5.0) 0 (0.0) related well with patient survival and are thus not accepted as definitive evidence of patient benefit by United States regula- NOTE. National Cancer Institute Common Toxicity Criteria Ver- sion 2.0 grades for hematological values are reported per patient. Data tory agencies. In this study, the objective response rate was are expressed as n (%). 15.5%, and the disease stabilization rate was 50%. This response Abbreviation: NA, not applicable. rate is somewhat lower than that reported in many trials of other

DISCUSSION Table 3 Nonhematological and biochemical toxicity (N ϭ 60) Our study was designed to confirm preclinical data sug- Toxicity grade gesting synergistic activity of pemetrexed and gemcitabine and Toxicity 1234 to evaluate both clinical activity and tolerance of this regimen in the treatment of advanced NSCLC. The most accepted end point Nonhematological Skin rash 13 (21.7) 8 (13.3) 2 (3.3) 0 (0) of patient benefit in advanced lung cancer trials is overall Fatigue 13 (21.7) 14 (23.3) 14 (23.3) 0 (0) survival. Although there is heterogeneity in the patients enrolled Fever 14 (23.3) 4 (6.7) 1 (1.7) 0 (0) in different clinical trials, the heterogeneity tends to be least in Anorexia 12 (20.0) 3 (5.0) 1 (1.7) 0 (0) trials of chemo-naive patients and in multicenter trials. Most Nausea 19 (31.7) 8 (13.3) 0 (0) 0 (0) Vomiting 6 (10.0) 4 (6.7) 0 (0) 0 (0) Phase II and III trials of standard doublet chemotherapy com- Diarrhea 8 (13.3) 4 (6.7) 2 (3.3) 1 (1.7) binations in advanced, chemo-naive NSCLC patients show me- Stomatitis 9 (15.0) 1 (1.7) 1 (1.7) 0 (0) dian survival times of 8–10 months. In this trial, the median Edema 4 (6.7) 1 (1.7) 0 (0) 0 (0) survival was 10.1 months. The 1-year survival rate was 43%, the Pneumonitis 0 (0) 0 (0) 2 (3.3) 1 (1.7) 2-year survival rate was 19%, and the 3-year survival rate was Alopecia 6 (10.0) 1 (1.7) NA NA Biochemical 11%. These results compare favorably with studies with the AST 20 (33.3) 9 (15.0) 8 (13.3) 1 (1.7) most active doublet combinations in the literature. ALT 17 (28.3) 10 (16.7) 9 (15.0) 3 (5.0) The time to progression (TTP) or progression-free survival Alkaline phosphatase 0 (0.0) 1 (1.7) 0 (0.0) 0 (0.0) is another well-accepted end point in advanced lung cancer Bilirubin 0 (0.0) 0 (0.0) 0 (0.0) 1 (1.7) Creatinine 2 (3.3) 0 (0.0) 0 (0.0) 0 (0.0) trials. The TTP is not influenced by subsequent therapies given after progression. In most multicenter trials of double chemo- NOTE. National Cancer Institute Common Toxicity Criteria Ver- sion 2.0 grades for nonhematological toxicities are reported per patient. therapy combinations in advanced, chemo-naive NSCLC pa- Data are expressed as n (%). tients, the median TTP is 4–5 months. Generally, patient sur- Abbreviations: NA, not applicable; AST, aspartate aminotrans- vival is about twice the duration of the median TTP. In some ferase; ALT, alanine aminotransferase.

active doublet combinations (20–30%). The 21% rate of disease severe nausea and vomiting, renal toxicity often requiring hos- progression compares more favorably with results from other pitalization for intravenous hydration, and long-lasting ototox- active doublets. Otherwise, there is not a clear explanation for icity and neuropathy, are all detrimental to the patient’s quality the lower than anticipated response rate. of life. Furthermore, cisplatin resistance has been found to be Recently, a new end point of clinical benefit has emerged associated with elevated excission repair cross-complement in the management of advanced NSCLC. As in our study, it has group 1 (ERCC1) mRNA levels in tumor tissue (37, 38). High been seen that chemotherapy regimens with modest response ERCC1 level has also been associated with shorter survival in profiles may be associated with prolonged survival. Carboplatin, NSCLC patients treated with gemcitabine plus cisplatin (39). for instance, was quoted with a low 9% response rate but was Theoretically, NSCLC patients with high levels of ERCC1 associated with longer survival than multidrug regimens (22). could benefit more from effective non-platinum-based cytotoxic As demonstrated for single-agent topotecan, patients who therapy. achieved a stabilization of their disease have an overall survival Other Phase II studies of cisplatin-free regimens in NSCLC identical to that of patients with a partial response (23). In our have recently been reported. The combination of gemcitabine study, two-thirds of patients achieved either partial response or and vinorelbine has been investigated in previously treated and stable disease, and their median overall survival was identical untreated patients with stage IIIb and IV disease (40). In the (13.8 months for partial response patients and 13.6 months for untreated group, 16 of 42 patients (36%) experienced a partial stable disease patients). Most of these patients also experienced response, median overall survival was 10.1 months, and 1- and symptom improvement or stabilization. 2-year survival was 42.9% and 31.6%. A randomized Phase II Although second-line therapies can produce a small im- study in untreated advanced NSCLC of gemcitabine plus irino- provement in survival, the influence of second-line chemother- tecan or docetaxel has been conducted and reported by the apy on survival in this study is not possible to evaluate (24). Cancer and Leukemia Group B (41). In this study, gemcitabine Assessment of clinical benefit brought by new chemotherapy and irinitoecan produced an overall response rate of 12.8%, and should not be restricted to an evaluation of response rate and gemcitabine and docetaxel produced an overall response rate of duration of survival but should also be balanced by measures of 23.1%. Median overall survival was 7.95 months and 12.8 treatment-related toxicities, disease-related symptoms, and qual- months, respectively. Estimated 1-year survival was 23% and ity of life (25). Recently, the results of a randomized study 51%, respectively. comparing docetaxel with pemetrexed in the second-line setting To move beyond the apparent plateau in NSCLC, cytotoxic were published (26). Pemetrexed treated patients had equivalent agents are being used in combination with agents having novel efficacy, as evidenced by response rates, time to progression, mechanisms of action. An example of this is the study of and survival, but with significantly reduced rates of hematolog- cisplatin, vinorelbine, and bexarotene, a novel retinoid X recep- ical and nonhematological toxicity. tor agent, in advanced NSCLC. This regimen produced an Assessment of lung cancer-related symptoms revealed im- overall response rate of 25%, a median survival time of 14 provements in pain, cough, and anorexia and worsening of months, and 1-year, 2-year, and projected 3-year survival rates fatigue. Evaluation of these end points is limited by the small of 61%, 32%, and 30%, respectively. This apparent advance is number of patients symptomatic at baseline and by the lack of a encouraging, but due to the presence of cisplatin, this regimen comparison in the setting of a Phase II trial. would not be a treatment option for all patients. An effective, The pemetrexed-gemcitabine regimen was well tolerated in platinum-free cytotoxic regimen to be combined with such a our experience. The only significant grade 3/4 clinical toxicities novel agent is still needed. encountered in Ͼ10% of the patients were fatigue and febrile Pemetrexed has now clearly demonstrated antitumor activ- neutropenia. Fifty percent of the dose reductions occurred at ity against NSCLC, either as a single agent or in association cycle two. Cumulative myelosuppression was not observed. with cisplatin or gemcitabine (7, 9, 16, 26, 42). Evidence for There were no severe cases of anorexia, edema, fever, or sto- non-cross-resistance of pemetrexed in NSCLC has been sug- matitis. The only case of severe diarrhea occurred in a patient gested by its activity in cisplatin-pretreated patients (42). Com- with an elevated homocysteine level. The number of patients bination studies associating pemetrexed with other active who completed or began treatment without vitamin supplemen- agents, such as paclitaxel, docetaxel, irinotecan, vinorelbine, tation is too small in this study for a meaningful comparison of and carboplatin, are being pursued (5). Antitumor activity of supplemented versus nonsupplemented patients. However, folic these new regimens is awaited to further confirm the role of

acid and vitamin B12 supplementation has been shown to dra- pemetrexed in the treatment of NSCLC. matically reduce the incidence of grade 3/4 hematological and Selection of day 8 pemetrexed administration was based on nonhematological toxicities in other studies (27–29). Because Phase I data of Adjei et al. (16). However, those data were nearly all patients received vitamin supplementation, changes in produced without vitamin supplementation. A randomized, dose or schedule would be required to reduce the rates of three-arm study of pemetrexed and gemcitabine is currently neutropenia. being conducted that should further define the optimal schedule The utility of cisplatin-free combinations in the treatment for this combination regimen. Another Phase I/II study in of advanced NSCLC is still a matter of debate. There is no NSCLC is testing the administration of these two agents on a question that cisplatin is currently the drug of choice for meta- biweekly schedule. In addition, results from a Phase I and static NSCLC (30, 31). Nevertheless, some plateau in the overall pharmacokinetic study testing the necessity for the 90-min delay survival has been reached with the use of cisplatin-based com- between administrations should be available in the near future. binations (32–36). The many toxicities of cisplatin, including Despite the disappointing response rate, this cisplatin-free

regimen is characterized by promising overall survival, progres- 17. Shargel L, Yu AB. Dosage adjustment in renal disease. In: Shargel sion-free survival, and time to progressive disease, which are L, Yu AB, editors. Applied biopharmaceutics and pharmacokinetics, better indicators of efficacy than response. In addition, this 2nd ed. Norwalk, CT: Appleton-Century-Crofts; 1985. p. 305–30. regimen offers good tolerance and convenient outpatient admin- 18. National Cancer Institute: common toxicity criteria (version 2). Bethesda, MD: Division of Cancer Treatment and Diagnosis, National istration. Our study showed the activity of the pemetrexed- Cancer Institute; 1999. gemcitabine combination in advanced NSCLC. Its 10.1-month 19. Simon R. Optimal two-stage designs for phase II clinical trials. median survival warrants further evaluation in a randomized Controlled Clinical Trials 1989;10:1–10. fashion against the current platinum-based standards, once the 20. Kaplan EL, Meier P. Non-parametric estimation of incomplete optimal administration schedule is established. observations. J Am Stat Assoc 1958;53:457–81. 21. Hollen PJ, Gralla RJ, Kris MG, et al. Measurement of quality of life ACKNOWLEDGMENTS in patients with lung cancer in multicenter trials of new therapies. Psychometric assessment of the Lung Cancer Symptom Scale. Cancer We thank Patrick McAndrews and Donna L. Miller for assistance (Phila) 1994;73:2087–98. in preparing the manuscript. 22. Bonomi PD, Finkelstein DM, Ruckdeschel JC, et al. Combination chemotherapy versus single agents followed by combination chemother- REFERENCES apy in stage IV non-small-cell lung cancer: a study of the Eastern 1. Chen VJ, Bewley JR, Gossett L, et al. Activity of LY231514 against Cooperative Oncology Group. J Clin Oncol 1989;7:1602–13. several enzymes in the folate-dependent pathways. Proc Am Assoc 23. Cesano A, Lane SR, Ross GA, Fields SZ. Stabilization of disease as Cancer Res 1996;37:381. an indicator of clinical benefit associated with chemotherapy in non- 2. Grindey GB, Shih C, Barnett CJ, et al. LY231514, a novel pyrrolo- small cell lung cancer patients. Int J Oncol 2000;17:587–90. pyrimidine antifolate that inhibits thymidylate synthase (TS). Proc Am 24. Huisman C, Smit EF, Giaccone G, Postmus PE. Second-line chem- Assoc Cancer Res 1992;33:411. otherapy in relapsing or refractory non-small-cell lung cancer: a review. 3. Shih C, Chen VJ, Gossett LS, et al. LY231514, a pyrrolo[2,3-D]- J Clin Oncol 2000;18:3722–30. pyrimidine-based antifolate that inhibits multiple folate-requiring en- 25. Gridelli C, Perrone F, Nelli F, Ramponi S, De Marinis F. Quality of zymes. Cancer Res 1997;57:1116–23. life in lung cancer patients. Ann Oncol 2001;12(Suppl 3):S21–5. 4. Rinaldi DA, Kuhn JG, Burris HA, et al. A phase I evaluation of 26. Hanna N, Shepherd FA, Fossella FV, et al. Randomized phase III multitargeted antifolate (MTA, LY231514), administered every 21 days, trial of pemetrexed versus docetaxel in patients with non-small cell lung utilizing the modified continual reassessment method for dose escala- cancer previously treated with chemotherapy. J Clin Oncol 2004;22: tion. Cancer Chemother Pharmacol 1999;44:372–80. 1589–97. 5. Adjei AA. Pemetrexed (Alimta®): a novel multitargeted antifolate 27. Bunn PA Jr, Paoletti P, Niyikiza C, et al. Vitamin B12 and folate agent. Expert Rev Anticancer Ther 2003;3:145–56. reduce toxicity of Alimta™ (pemetrexed disodium, LY231514, MTA), 6. Cripps C, Burnell M, Jolivet J, et al. Phase II study of first-line a novel antifolate/antimetabolite. Proc Am Soc Clin Oncol 2001;20:76a. LY231514 (multi-targeted antifolate) in patients with locally advanced 28. Scagliotti GV, Shin DM, Kindler HL, et al. Phase II study of or metastatic colorectal cancer: an NCIC Clinical Trials Group study. pemetrexed with and without folic acid and vitamin B12 as front-line Ann Oncol 1999;10:1175–9. therapy in malignant pleural mesothelioma. J Clin Oncol 2003;21: 7. Rusthoven JJ, Eisenhauer E, Butts C, et al. Multitargeted antifolate 1556–61. LY231514 as first-line chemotherapy for patients with advanced non- 29. Vogelzang NJ, Rusthoven JJ, Symanowski J, et al. A phase III study small-cell lung cancer: a phase II study. National Cancer Institute of of pemetrexed in combination with cisplatin versus cisplatin alone in Canada Clinical Trials Group. J Clin Oncol 1999;17:1194. patients with malignant pleural mesothelioma. J Clin Oncol 2003;21: 8. Niyikiza C, Baker SD, Seitz DE, et al. Homocysteine and methyl- 2636–44. malonic acid: markers to predict and avoid toxicity from pemetrexed 30. Johnson DH. Evolution of cisplatin-based chemotherapy in non- therapy. Mol Cancer Ther 2002;1:545–52. small cell lung cancer: a historical perspective and the Eastern Coop- 9. Clarke SJ, Abratt R, Goedhals L, et al. Phase II trial of pemetrexed erative Oncology Group experience. Chest 2000;117:133S–7S. disodium (ALIMTA, LY231514) in chemotherapy-naive patients with 31. Chemotherapy in non-small cell lung cancer: a meta-analysis using advanced non-small-cell lung cancer. Ann Oncol 2002;13:737–41. updated data on individual patients from 52 randomised clinical trials. 10. Plunkett W, Huang P, Gandhi V. Preclinical characteristics of Non-Small Cell Lung Cancer Collaborative Group. Br Med J 1995;311: gemcitabine. Anticancer Drugs 1995;6(Suppl 6):7–13. 899–909. 11. Gatzemeier U, Shepherd FA, LeChevalier T, et al. Activity of 32. Schiller JH, Harrington D, Belani CP, et al. Comparison of four gemcitabine in patients with non-small cell lung cancer: a multicentre, chemotherapy regimens for advanced non-small-cell lung cancer. extended phase II study. Eur J Cancer 1996;32A:243–8. N Engl J Med 2002;346:92–8. 12. Manegold C, Bergman B, Chemaissani A, et al. Single-agent gem- 33. Kelly K, Crowley J, Bunn PA Jr, et al. Randomized phase III trial citabine versus cisplatin-etoposide: early results of a randomised phase of paclitaxel plus carboplatin versus vinorelbine plus cisplatin in the II study in locally advanced or metastatic non-small-cell lung cancer. treatment of patients with advanced non-small-cell lung cancer: a South- Ann Oncol 1997;8:525–9. west Oncology Group trial. J Clin Oncol 2001;19:3210–8. 13. Cardenal F, Lopez-Cabrerizo MP, Anton A, et al. Randomized 34. Scagliotti G, De Marinis F, Rinaldi M, et al. Phase III randomized phase III study of gemcitabine-cisplatin versus etoposide-cisplatin in the trial comparing three platinum-based doublets in advanced non-small- treatment of locally advanced or metastatic non-small-cell lung cancer. cell lung cancer. J Clin Oncol 2002;20:4285–91. J Clin Oncol 1999;17:12–8. 35. Fossella F, Pereira JR, von Pawel J, et al. Randomized, multina- 14. Guchelaar HJ, Richel DJ, van Knapen A. Clinical, toxicological and tional, phase III study of docetaxel plus platinum combinations versus pharmacological aspects of gemcitabine. Cancer Treat Rev 1996;22: vinorelbine plus cisplatin for advanced non-small-cell lung cancer: the 15–31. TAX 326 study group. J Clin Oncol 2003;21:3016–24. 15. Adjei AA. Pemetrexed: a multitargeted antifolate agent with prom- 36. Smit EF, van Meerbeeck JP, Lianes P, et al. Three-arm randomized ising activity in solid tumors. Ann Oncol 2000;11:1335–41. study of two cisplatin-based regimens and paclitaxel plus gemcitabine in 16. Adjei AA, Erlichman C, Sloan JA, et al. Phase I and pharmacologic advanced non-small-cell lung cancer: a phase III trial of the European study of sequences of gemcitabine and the multitargeted antifolate agent Organization for Research and Treatment of Cancer Lung Cancer Group– in patients with advanced solid tumors. J Clin Oncol 2000;18:1748–57. EORTC 08975. J Clin Oncol 2003;21:3909–17.

37. Dabholkar M, Vionnet J, Bostick-Bruton F, Yu JJ, Reed E. Mes- 40. Herbst RS, Khuri FR, Lu C, et al. The novel and effective nonplati- senger RNA levels of XPAC and ERCC1 in ovarian cancer tissue num, nontaxane combination of gemcitabine and vinorelbine in ad- correlate with response to platinum-based chemotherapy. J Clin Investig vanced nonsmall cell lung carcinoma: potential for decreased toxicity 1994;94:703–8. and combination with biological therapy. Cancer (Phila) 2002;95: 38. Metzger R, Leichman CG, Danenberg KD, et al. ERCC1 mRNA 340–53 [erratum in Cancer (Phila) 2003;97:1367]. levels complement thymidylate synthase mRNA levels in predicting 41. Rocha Lima CM, Rizvi NA, Zhang C, et al. Randomized phase II response and survival for gastric cancer patients receiving combination trial of gemcitabine plus irinotecan or docetaxel in stage IIIB or stage IV cisplatin and fluorouracil chemotherapy. J Clin Oncol 1998;16:309–16. NSCLC. Ann Oncol 2004;15:410–8. 39. Lord RV, Brabender J, Gandara D, et al. Low ERCC1 expression 42. Smit EF, Mattson K, von Pawel J, et al. ALIMTA (pemetrexed correlates with prolonged survival after cisplatin plus gemcitabine chemo- disodium) as second-line treatment of non-small-cell lung cancer: a therapy in non-small cell lung cancer. Clin Cancer Res 2002;I:2286–91. phase II study. Ann Oncol 2003;14:455–60.

Downloaded from clincancerres.aacrjournals.org on September 29, 2021. © 2004 American Association for Cancer Research. Phase II Study of Pemetrexed-Gemcitabine Combination in Patients with Advanced-Stage Non-Small Cell Lung Cancer

Christian Monnerat, Thierry Le Chevalier, Karen Kelly, et al.

Clin Cancer Res 2004;10:5439-5446.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/10/16/5439

Cited articles This article cites 34 articles, 16 of which you can access for free at: http://clincancerres.aacrjournals.org/content/10/16/5439.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/10/16/5439.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://clincancerres.aacrjournals.org/content/10/16/5439. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.