82804218.Pdf

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector ORIGINAL ARTICLE Single-Agent Pemetrexed or Sequential Pemetrexed/Gemcitabine as Front-Line Treatment of Advanced Non-small Cell Lung Cancer in Elderly Patients or Patients Ineligible for Platinum-Based Chemotherapy: A Multicenter, Randomized, Phase II Trial Cesare Gridelli, MD,* Eckhard Kaukel, MD,† Vanessa Gregorc, MD,‡ Maria Rita Migliorino, MD,§ Thomas R. Mu¨ller, MD,͉͉ Christian Manegold, MD,¶ Adolfo Favaretto, MD,# Andrea Martoni, MD,** Orazio Caffo, MD,†† Alexander Schmittel, MD,‡‡ Antonio Rossi, MD,* Francesca Russo, MD,§§ Patrick Peterson, PhD,͉͉͉͉ Marı´a Mun˜oz, PhD,¶¶ and Martin Reck, MD##

hematologic toxicity consisted of neutropenia (4.5% pemetrexed; 2.3% Introduction: This randomized phase II trial evaluated single-agent pemetrexed/gemcitabine), febrile neutropenia (4.5% pemetrexed; pemetrexed or sequential pemetrexed/gemcitabine in patients with non- 4.7% pemetrexed/gemcitabine), thrombocytopenia (4.5% pem- small cell lung cancer (NSCLC) who were elderly (Ն70 years) or etrexed; 7.0% pemetrexed/gemcitabine), and anemia (6.8% pem- younger than 70 years and ineligible for platinum-based chemotherapy. etrexed; 4.7% pemetrexed/gemcitabine). No grade 3/4 nonhema- Methods: Chemonaive patients with stage IIIB/IV NSCLC and an tologic toxicities exceeded 4.7% in either arm. Eastern Cooperative Oncology Group performance status of 0 to 2 received either 500 mg/m2 of pemetrexed (day 1, every 3 weeks) for Conclusions: Single-agent pemetrexed and sequential pemetrexed/ eight cycles, or the same dosage of pemetrexed for cycles 1 and 2 gemcitabine have shown moderate activity and are well tolerated as and then 1200 mg/m2 of gemcitabine (days 1 and 8, every 3 weeks) ﬁrst-line treatments for advanced NSCLC in elderly patients or for cycles 3 and 4 (repeated once for a total of eight cycles). All patients unsuitable for platinum-based combination chemotherapy.

patients were given vitamin B12 and folic acid supplementation. Key Words: Advanced NSCLC, Elderly patients, PS 2 patients, Results: From July 2003 to July 2004, 87 patients (44 pemetrexed; 43 Pemetrexed, Gemcitabine. pemetrexed/gemcitabine) received treatment. The median time to progression was 4.5 (95% confidence interval: 3.0–9.3) and 4.1 months (J Thorac Oncol. 2007;2: 221–229) (95% confidence interval: 1.7–5.8) for the pemetrexed and pemetrexed/ gemcitabine arms, respectively, and the median progression-free survival time was 3.3 months for both arms. Tumor response rates for the on-small cell lung cancer (NSCLC) represents approxi- pemetrexed and pemetrexed/gemcitabine arms were 4.5% and 11.6%, Nmately 80% to 87% of all lung cancers.1 Because the respectively. The median overall survival time was 4.7 months for the majority of patients present with metastatic NSCLC, pallia- pemetrexed arm and 5.4 months for the pemetrexed/gemcitabine arm, tive treatment is often the only therapeutic option. with respective 1-year survival rates of 28.5% and 28.1%. Grade 3/4 Platinum-based combination chemotherapy is currently the standard treatment for patients with advanced NSCLC. *Division of Medical Oncology, S.G. Moscati Hospital, Avellino, Italy; †Asklepios However, two categories of patients, the elderly and patients Klinik Harburg, Hamburg, Germany; ‡Division of Medical Oncology, San with poor performance status (PS; Յ2), are often not candi- Raffaele Hospital, Milan, Italy; §Pneumo-oncology Division, Forlanini Hospital, dates for platinum-based combinations because of poor gen- Rome, Italy; ͉͉Krankenhaus Hofheim am Taunus, Hofheim, Germany; ¶Heidel- eral conditions and comorbidities. berg University Medical Center, Mannheim, Germany; #Division of Medical Oncology, University Hospital, Padova, Italy; **Division of Oncology, Mal- More than 50% of advanced NSCLC cases are diag- pighi Hospital, Bologna, Italy; ††Division of Medical Oncology, Santa Chiara nosed in patients older than 65 years, and approximately 30% Hospital, Trento, Italy; ‡‡Charite´, Campus Benjamin Franklin, Department of to 40% of cases are diagnosed in patients older than 70 Hematology and Oncology, Berlin, Germany; §§Eli Lilly and Company, Flo- years,2,3 yet disproportionately low numbers of elderly par- rence, Italy; ͉͉͉͉Eli Lilly and Company, Indianapolis, IN; ¶¶Eli Lilly and Com- 4–7 pany, Madrid, Spain; and ##Hospital Grosshansdorf, Grosshansdorf, Germany ticipants are included in clinical trials. Three phase III Disclosure: Sponsored by Eli Lilly and Company, Indianapolis, Indiana. Carol J. trials constitute the published evidence-based literature for Shipley is the authorized representative signing on behalf of the Eli Lilly and the elderly.8–10 Chemotherapy treatment in the elderly is Company authors, only as work made for hire. Francesca Russo and Maria complicated by a number of age-related issues. Decreased Munoz are employees of Eli Lilly and Company. No other authors have financial disclosures to report. hepatic, renal, and bone marrow functions increase toxicity, Address for correspondence: Cesare Gridelli, MD, Division of Medical Oncol- particularly cisplatin-related toxicity.7,8 Furthermore, comor- ogy, S.G. Moscati Hospital, Contrada Amoretta, 83100 Avellino, Italy. bid diseases, which may significantly affect functional E-mail: [email protected] status, general health, and tumor symptoms, are frequently Copyright © 2007 by the International Association for the Study of Lung 11,12 Cancer present. Because of these complications, single-agent ISSN: 1556-0864/07/0203-0221 chemotherapy is often the recommended treatment, even by

Journal of Thoracic Oncology • Volume 2, Number 3, March 2007 221 Gridelli et al. Journal of Thoracic Oncology • Volume 2, Number 3, March 2007

American Society of Clinical Oncology guidelines, for el- Additionally, eligible patients met the following criteria: no derly patients with advanced NSCLC.6,8,10,13,14 The findings prior chemotherapy (prior surgery and radiotherapy were of the two phase III trials that compared single-agent versus permitted as long as relapse or disease progression had combination therapy in the elderly8,9 produced different con- occurred after the procedure or therapy); measurable disease clusions: the Multicenter Italian Lung Cancer in the Elderly according to Response Evaluation Criteria in Solid Tumors Study showed that the combination of gemcitabine and vi- criteria32; age Ն70 years or Ͻ70 years for patients who, in the norelbine was not more effective or tolerable than either investigator’s opinion, were ineligible for platinum-based agent alone, whereas the Southern Italy Cooperative Oncol- chemotherapy because of poor PS or comorbidities; an East- ogy Group found that the combination of gemcitabine and ern Cooperative Oncology Group PS of 0 to 2; an estimated vinorelbine was associated with significantly better overall life expectancy of at least 12 weeks; and adequate bone survival than vinorelbine alone. Nevertheless, the difference marrow, renal, and hepatic function. Patients with symptom- in sample sizes between the two trials (698 versus 120 atic brain metastases were ineligible for the study (whereas patients for the Multicenter Italian Lung Cancer in the Elderly patients with a history of brain metastases were permitted to Study and the Southern Italy Cooperative Oncology Group participate in the study), as were patients with an inability to trial, respectively) should be noted. interrupt nonsteroidal antiinflammatory drugs (2 days before, The proportion of patients with PS 2 enrolled in clinical the day of, and 2 days after receiving pemetrexed), uncon- trials is often less than 20% of the whole study population, trolled pleural effusions, or inability or unwillingness to take even though the proportion of PS 2 patients is consistently steroids or vitamin supplementation. The protocol was ap- higher in population-based studies.15 For these patients, no proved through institutional ethical review boards, and all treatment is widely accepted as standard. A subgroup analysis patients provided written informed consent before treatment. of several randomized trials16 suggests that several new- The study was conducted in accordance with the ethical generation cytotoxic drugs (gemcitabine, vinorelbine, and principles in the Declaration of Helsinki and good clinical taxanes) are superior to supportive care in this patient cate- practice. gory. Therefore, single-agent chemotherapy with these drugs may be the preferred option for palliative treatment of PS 2 Study Design and Sample Size patients.13,16–18 This open-label phase II study was designed to ran- Pemetrexed, a novel multitargeted antifolate, inhibits domly assign 90 patients to one of two regimens (single-agent three enzymes involved in pyrimidine and purine synthesis: pemetrexed or single-agent pemetrexed followed sequentially thymidylate synthase, dihydrofolate reductase, and glyci- by single-agent gemcitabine), and the trial was not considered namide ribonucleotide formyl transferase.19 Pemetrexed has complete until at least 73 patients progressed. This sample shown activity as a single agent in several phase II trials and size was chosen on the basis of a selection design using the one phase III trial in patients with NSCLC.20–23 Supplemen- primary endpoint of TtPD. The selection design assumed that tation with folic acid and vitamin B12, which is required to the primary goal of the trial was to select a regimen for a reduce pemetrexed toxicity,24 results in a mild toxicity profile larger phase III trial. Type I error, or alpha error, was not that allows for drug evaluation in the elderly. controlled under this design. The treatment that was observed Gemcitabine, a cytosine–arabinoside analogue, acts by to have a numerically longer TtPD and a better observed upsetting deoxynucleotide pools and interfering with DNA tolerability was to be selected as the best choice for subse- chain elongation.25 Gemcitabine has previously shown good quent studies. Assuming a hazard ratio of 0.82, there was at activity and tolerability in advanced NSCLC elderly patients least an 80% chance of correctly selecting the preferred in phase II trials26–30 and in a phase III study.8 regimen. Assuming a 5-month median TtPD in the inferior We conducted a multicenter phase II randomized trial arm, this design required a 9-month follow-up period after the to evaluate the time to progressive disease (TtPD) of single- last patient was enrolled. agent pemetrexed or sequential pemetrexed/gemcitabine as Randomization was controlled by a computerized voice first-line therapy in elderly patients or in patients with response unit at a central location for all study sites and was NSCLC who were ineligible for platinum-based chemother- stratified according to disease stage (stage IIIB versus stage apy. Sequential treatment with single-agent pemetrexed and IV) and PS (0–1 versus 2). then single-agent gemcitabine was chosen in an effort to maximize the activity of both drugs while producing a lower Treatment Plan toxicity profile than that seen with the combination.31 Sec- Eligible patients were randomly assigned either to re- ondary objectives included toxicity, response rate, and overall ceive 500 mg/m2 of pemetrexed as a 10-minute intravenous survival. infusion on day 1 every 3 weeks for a maximum of eight cycles, or pemetrexed at the same dosage for cycles 1 and 2 2 MATERIALS AND METHODS and then 1200 mg/m of gemcitabine as a 30-minute intravenous infusion on days 1 and 8 every 3 weeks for cycles 3 Eligibility Criteria and 4. This four-cycle schedule was repeated once, for a total Patients with histological or cytological confirmation of of eight cycles. stage IIIB NSCLC (with supraclavicular lymph node metas- All patients were instructed to take 350 to 1000 ␮gof tases or pleural effusion) or stage IV NSCLC not amenable to folic acid orally each day, beginning approximately 1 to 2 surgery or curative radiotherapy were eligible for the study. weeks before the first dose of pemetrexed and continuing

daily until 3 weeks after the last dose of pemetrexed. A a higher degree of comorbidity. During therapy, laboratory ␮ 1000- g vitamin B12 injection was administered intramuscu- assessments included hematology, blood chemistries, and larly approximately 1 to 2 weeks before the first dose of calculated creatinine clearance. After the second and fourth pemetrexed and was repeated approximately every 9 weeks cycles of chemotherapy, and then every two cycles thereafter, until 3 weeks after the last dose of pemetrexed. Patients radiologic tumor assessment was performed by the same treated with pemetrexed were instructed to take dexametha- method that had been used at baseline. Tumor response was sone (4 mg orally, twice daily, on the day before, the day of, assessed using the Response Evaluation Criteria in Solid and the day after pemetrexed) as a prophylactic measure Tumors criteria32 and required confirmation 3 to 4 weeks against skin rash. after initial response. Toxicity evaluations, performed at the Patients received full supportive care. Granulocyte col- end of each cycle, were based on the National Cancer Insti- ony–stimulating factor support was allowed for prolonged tute common toxicity criteria, version 2.34 myelosuppression and febrile neutropenia. Leucovorin was All patients who received at least one dose of a study allowed for common toxicity criteria grade 4 leukopenia drug were assessed for TtPD, progression-free survival lasting longer than 3 days, grade 4 neutropenia lasting longer (PFS), survival, response, and toxicity. than 3 days, grade 4 thrombocytopenia, bleeding associated with grade 3 thrombocytopenia, or grade 3 or 4 mucositis. Statistical Considerations The day 8 dose of gemcitabine was reduced by 25% for TtPD was defined as the time from the date of random- an absolute neutrophil count (ANC) of 0.5 to 1.0 ϫ 109/liter ization to the first date of documented disease progression or a platelet count of 50 to 100 ϫ 109/liter, and this dose was and was censored on the date of last tumor assessment for omitted for an ANC Ͻ0.5 ϫ 109/liter or a platelet count patients without documented progressive disease at the time Ͻ50 ϫ 109/liter. The day 8 dose of gemcitabine was reduced of analysis. Because many patients died without documented by 50% or was omitted for grade 3 nonhematologic toxicity progressive disease, there was a high rate of censoring for (except nausea and vomiting) and was omitted for grade 4 TtPD. Therefore, PFS was added retrospectively as a similar nonhematologic toxicity. Omitted doses were not admin- measure that would have less censoring. PFS was defined as istered later, and patients who required a dose reduction the time from the date of randomization to the first date of continued to receive the reduced dose for the remainder of documented disease progression or death from any cause and the study. was censored on the date of last follow-up for patients not Subsequent cycles were delayed because of toxicity known to have died at the time of the analysis. Overall (ANC Ͻ1.5 ϫ 109/liter, platelets Ͻ100 ϫ 109/liter, a calcu- survival was defined as the time from the date of randomiza- lated creatinine clearance Ͻ45 ml/min, or nonhematologic tion to the date of death from any cause and was censored on toxicity above grade 1 or baseline) for a maximum of 3 the date of last follow-up for patients not known to have died weeks, after which the patient was discontinued from the at the time of analysis. study. Doses of both drugs in subsequent cycles were reduced The Kaplan–Meier method35 was used to estimate pa- by 25% of the previous dose for platelet counts Ͻ50 ϫ rameters for TtPD, PFS, and overall survival (OS), and the 109/liter and/or ANC Ͻ0.5 ϫ 109/liter, or for diarrhea of log-rank test was used for comparison of these measures grade 3 or higher observed in the previous cycle. Pemetrexed between treatment arms. Interval estimates were calculated was reduced by 50% of the previous dose for grade 3 or 4 using 95% confidence intervals (CI). The tumor response mucositis. For other nonhematologic toxicities of at least rate, including a 95% CI, was reported.36 grade 3, doses were reduced to 75% of the previous dose if Cox regression analyses were used to examine any deemed appropriate by the treating physician. Patients were possible relationships between efficacy (TtPD, PFS, and OS) discontinued from the study if there was a recurrence of grade and the Charlson comorbidity index. Because of the small 3 or 4 thrombocytopenia, neutropenia, or mucositis after two sample size, these analyses were considered exploratory. dose reductions. RESULTS Baseline and Treatment Assessments No more than 4 weeks before enrollment, patients were Patient Characteristics assessed by chest and abdominal computed tomography From July 2003 through July 2004, 92 patients were scans, magnetic resonance imaging or computed tomography randomly assigned to receive either pemetrexed (47 patients) scan of the brain, bone scan, and chest x-ray (ultrasound was or sequential pemetrexed/gemcitabine therapy (45 patients). not permitted). No more than 1 week before enrollment, Of the 92 randomized patients, 87 patients received study patients were assessed by a history and physical examination, therapy, 44 in the pemetrexed arm and 43 in the sequential tumor measurement of palpable lesions, hematology, com- pemetrexed/gemcitabine arm. Of the 47 patients randomly plete blood chemistries, calculated creatinine clearance, elec- assigned to pemetrexed, three patients received no treatment: trocardiogram, evaluation of PS, and a comorbidities assess- one because of unmet protocol entry criteria, one because of ment using the Charlson comorbidity index,33 a 19-item worsening of PS, and one because of the patient’s decision. measure of health status by which the rater identifies the Of the 45 patients randomly assigned to sequential pem- presence or absence of each disease state, which has a etrexed/gemcitabine, one patient received no treatment be- corresponding numeric value; the values for all present dis- cause of unmet protocol entry criteria, and one patient was ease states are summed, with a higher score corresponding to discontinued immediately after receiving gemcitabine instead

Copyright © 2007 by the International Association for the Study of Lung Cancer 223 Gridelli et al. Journal of Thoracic Oncology • Volume 2, Number 3, March 2007 of pemetrexed in the first cycle (this patient was not consid- number of 2.5 cycles for pemetrexed and 3.0 cycles for ered to have received study therapy and was excluded from sequential pemetrexed/gemcitabine (range, 1–8 for both further analysis). arms). The actual median dose intensity was 164.8 mg/m2 per Baseline patient and disease characteristics for all week (range, 95.0–169.5) for the pemetrexed arm and 166.7 treated patients are listed in Table 1. The two arms were well mg/m2 per week (pemetrexed; range, 142.6–171.2) and 786.7 balanced for all demographic and stratification factors. All 87 mg/m2 per week (gemcitabine; range, 384.7–839.3) for the patients were evaluable for efficacy and toxicity, and clinical sequential pemetrexed/gemcitabine arm. Patients received data were collected up to March 14, 2005. 94.9% of the planned dose intensity in the pemetrexed arm and 98.3% (pemetrexed) and 90.8% (gemcitabine) of the Treatment Administration planned dose intensity in the sequential pemetrexed/gemcit- Eighty-seven patients received at least one cycle of abine arm. therapy, 40 patients (21 pemetrexed and 19 pemetrexed/ A total of 36 and 33 dose delays occurred in the gemcitabine) received at least four cycles, and 26 patients (13 pemetrexed and sequential pemetrexed/gemcitabine arms, re- per arm) received at least six cycles of therapy. The total spectively; most of these delays were caused by scheduling number of cycles administered was 163 for pemetrexed and conflicts (75% pemetrexed and 73% sequential pemetrexed/ 166 for sequential pemetrexed/gemcitabine, with a median gemcitabine). There were only seven dose reductions (five in the pemetrexed arm and two in the sequential pemetrexed/ gemcitabine arm). In the pemetrexed arm, reasons for dose TABLE 1. Baseline Patient and Disease Characteristics reductions were febrile neutropenia (two), neutropenia (two), No. of Patients (%) and thrombocytopenia (one).

Sequential Pem Pem/Gem Efficacy The median TtPD was similar in both arms: 4.5 months (43 ؍ n) (44 ؍ n) (95% CI: 3.0–9.3) and 4.1 months (95% CI: 1.7–5.8) for Sex pemetrexed and sequential pemetrexed/gemcitabine, respec- Male 35 (79.5) 29 (67.4) tively (Figure 1). A high rate of censoring occurred for TtPD Female 9 (20.5) 14 (32.6) (15 patients had no objective PD), mainly in the pemetrexed Age (yr) arm for 12 patients because of a lack of documented disease Median 73 73 progression. Of the 12 patients in the pemetrexed arm without Range 58–82 61–83 documented disease progression, four died either on study or Ն 70 35 (79.5) 34 (79.1) during the first follow-up visit (three from progression and Ͻ 70 9 (20.5) 9 (20.9) one from non–drug-related respiratory failure). Of the re- ECOG performance status maining eight patients, five discontinued early (during visit 1 0 4 (9.1) 3 (7.0) or 2), and no imaging assessments of disease status were done 1 26 (59.1) 23 (53.5) after discontinuation from the study drug. 2 14 (31.8) 17 (39.5) Because of the unexpectedly high rate of censoring for Stage of disease TtPD, PFS also was analyzed. Because PFS included death as IIIB 8 (18.2) 8 (18.6) an event, the rate of censoring for PFS was much lower than IV 36 (81.8) 35 (81.4) the rate of censoring for TtPD: 44 of 87 patients had censored Histology times for TtPD, whereas only 14 of 87 patients had censored Adenocarcinoma 18 (40.9) 24 (55.8) times for PFS. The median PFS time was 3.3 (months) for Squamous 11 (25.0) 15 (34.9) both the pemetrexed (95% CI: 2.0–4.4) and the sequential Large cell 7 (15.9) 3 (7.0) pemetrexed/gemcitabine (95% CI: 1.7–4.1) arms (Figure 2). Mixed 1 (2.3) 0 (0.0) Table 2 shows best overall response. Response could Undefined NSCLC 7 (15.9) 1 (2.3) not be assessed in 17 (38.6%) patients in the pemetrexed arm Number of disease sites and in four (9.3%) patients in the sequential pemetrexed/ 1–3 24 (54.6) 27 (62.8) gemcitabine arm. In the majority of the cases, this was due to 4–6 18 (40.9) 15 (34.9) a lack of follow-up tumor assessments attributable to early Ͼ 6 2 (4.5) 1 (2.3) discontinuations: 17 of 21 patients with unassessable re- Number of comorbidities sponses discontinued in visits 1 or 2, and of these 17, seven None 18 (40.8) 11 (25.6) died on study (four from progression and three from serious 1 14 (31.8) 20 (46.5) adverse events considered unrelated to the study drug). In the 2 5 (11.4) 7 (16.3) pemetrexed arm, early discontinuations of patients leading to 3 5 (11.4) 5 (11.6) nonassessable responses were attributable to lack of efficacy 4 1 (2.3) 0 (0) (in two patients), death (in three patients from progression Unknown 1 (2.3) 0 (0) and in two patients from serious adverse events considered Pem, pemetrexed; pem/gem, pemetrexed/gemcitabine; ECOG, Eastern Cooperative unrelated to the study drug), adverse events (unrelated to the Oncology Group; NSCLC, non-small cell lung cancer. study drug in four patients and related in two patients), and

FIGURE 1. Kaplan–Meier time to progressive disease.

FIGURE 2. Kaplan–Meier progression-free survival. other causes (in one patient); in the sequential pemetrexed/ in the sequential pemetrexed/gemcitabine arm (95% CI: 3.9– gemcitabine arm, early discontinuations were attributable to 25.1). The rates of stable disease were 34.1% and 23.3% for death (one from an adverse event considered unrelated to the pemetrexed and sequential pemetrexed/gemcitabine arms, re- study drug and one from progression) and adverse events spectively, resulting in overall disease control (response ϩ stable (related to the study drug in one patient). Objective tumor disease) in 17 (38.6%) and 15 (34.9%) patients. Additionally, responses were reported in two (4.5%) patients in the pem- there were four unconﬁrmed responses (restaging not completed etrexed arm (95% CI: 0.6–15.5) and in ﬁve (11.6%) patients after 1 month, per protocol), three in the pemetrexed arm and

TABLE 2. Best Overall Response TABLE 3. Worst Grade 3/4 NCI-CTC Hematologic Toxicity No. of Patients (%) No. of Patients (%)

Pem Sequential Pem/Gem Sequential Pem Pem/Gem (43 ؍ n) (44 ؍ Response (n (43؍ n) (44 ؍ n) Overall response 2 (4.5) 5 (11.6) (95% CI) (0.6–15.5) (3.9–25.1) Toxicity G3 G4 G3 G4 Complete response 0 (0) 0 (0) Neutropenia 1 (2.3) 1 (2.3) 1 (2.3) 0 (0) Partial response 2 (4.5) 5 (11.6) Febrile neutropenia 2 (4.5) 0 (0) 2 (4.7) 0 (0) Stable disease 15 (34.1) 10 (23.3) Anemia 2 (4.5) 1 (2.3) 2 (4.7) 0 (0) Progressive disease 10 (22.7) 24 (55.8) Thrombocytopenia 1 (2.3) 1 (2.3) 3 (7.0) 0 (0) Not assessable/unknown 17 (38.6) 4 (9.3) NCI-CTC, National Cancer Institute common toxicity criteria; Pem, pemetrexed; Pem, pemetrexed; pem/gem, pemetrexed/gemcitabine; CI, confidence interval. pem/gem, pemetrexed/gemcitabine; G, grade. one in the sequential pemetrexed/gemcitabine arm for response etrexed/gemcitabine arm had a PS of 2. The median PFS time rates of 11.4% and 14.0%, respectively. was 1.3 months for pemetrexed and 1.9 months for sequential The median overall survival time was 4.7 months (95% pemetrexed/gemcitabine, and the median OS time was 1.8 CI: 3.2–6.8) for pemetrexed and 5.4 months (95% CI: 3.8– months for pemetrexed and 3.9 months for sequential pem- 7.6) for sequential pemetrexed/gemcitabine (Figure 3). The etrexed/gemcitabine. There were no statistically significant 1-year survival rate was 28.5% (95% CI: 14.0–43.1) and differences between treatment arms within the PS 0/1 or PS 28.1% (95% CI: 13.2–42.9) for pemetrexed and sequential 2 subgroups. pemetrexed/gemcitabine, respectively. Per the Charlson Comorbidity Index, approximately As an exploratory analysis, we also examined efficacy 74% of patients had either no or one comorbidity at baseline, outcomes within the PS 0/1 and PS 2 subgroups. Thirty indicating a fairly low rate of comorbidity in this study. Of patients in the pemetrexed arm and 26 patients in the sequen- the 19 items, only the presence of diabetes and/or moderate to tial pemetrexed/gemcitabine arm had a PS of 0 or 1. The severe liver disease was suggested by the Cox models to be median PFS time was 4.4 months for pemetrexed and 3.7 associated with an increased risk of death. months for sequential pemetrexed/gemcitabine, and the median OS time was 5.2 months for pemetrexed and 7.6 months Toxicity for sequential pemetrexed/gemcitabine. Fourteen patients in Hematologic toxicity was mild (Table 3). Grade 3 and the pemetrexed arm and 17 patients in the sequential pem- 4 toxicities were neutropenia, febrile neutropenia, thrombo-

FIGURE 3. Kaplan–Meier overall survival.

Hematologic and nonhematologic toxicities were mild, with TABLE 4. Worst Grade 3/4 NCI-CTC Nonhematologic Toxicity no grade 4 hematologic toxicities seen in the sequential pemetrexed/gemcitabine arm and with no more than two No. of Patients (%) patients with grade 3/4 nonhematologic toxicities seen in Sequential both arms. Pem Pem/Gem Single-agent pemetrexed as both first-line and second- line treatment of advanced NSCLC has yielded encouraging (43 ؍ n) (44 ؍ n) response rates and survival and a good toxicity profile when Toxicity G3 G4 G3 G4 administered with adequate vitamin supplementation.20–24 In 2 Fatigue 0 (0) 0 (0) 0 (0) 1 (2.3) two phase II studies, 600 mg/m of pemetrexed was admin- 20,21 Rash 0 (0) 0 (0) 2 (4.7) 0 (0) istered to chemonaive patients with advanced NSCLC. Allergic reaction 1 (2.3) 0 (0) 0 (0) 0 (0) The objective response rates were 15.8% and 23.3%, respec- Pulmonary 2 (4.5) 0 (0) 0 (0) 0 (0) tively, and the median survival times were 7.2 and 9.2 20 Abdominal pain 0 (0) 0 (0) 0 (0) 1 (2.3) months, respectively. In the study by Clarke et al., the main Diarrhea 0 (0) 1 (2.3) 0 (0) 0 (0) grade 3/4 toxicities included neutropenia in 42% of patients, Mucositis 2 (4.5) 0 (0) 0 (0) 0 (0) cutaneous toxicity in 31% of patients, and asymptomatic Melena 1 (2.3) 0 (0) 0 (0) 0 (0) elevations in the hepatic biochemistry in 80% of patients. 2 Anorexia 0 (0) 1 (2.3) 0 (0) 0 (0) Although the study by Clarke et al.20 showed that 600 mg/m Edema 1 (2.3) 0 (0) 0 (0) 0 (0) could be safely administered, this conclusion was limited to ALT 1 (2.3) 0 (0) 0 (0) 0 (0) patients with a PS of 0 or 1. In the study by Rusthoven et al.,21 the main grade 3/4 toxicities included neutropenia in 39% of NCI-CTC, National Cancer Institute common toxicity criteria; Pem, pemetrexed; pem/gem, pemetrexed/gemcitabine; G, grade; ALT, alanine aminotransferase. patients, thrombocytopenia in 3% of patients, and skin rash in 39% of patients. On the basis of the toxicities seen during the study by Rusthoven et al.21 and those seen in the study by 37 cytopenia, and anemia. In the sequential pemetrexed/gemcit- Cripps et al., who used the same dosage and schedule in patients with colorectal cancer, pemetrexed was reduced to abine arm, no grade 4 toxicities occurred. 2 Nonhematologic toxicities were mild, with each grade 500 mg/m . It is important to note, nevertheless, that neither 3/4 toxicity occurring in no more than two patients in both of these studies included vitamin supplementation. arms (Table 4). Allergic reaction, diarrhea, melena, mucosi- Gemcitabine has been shown to be active and well tis, pulmonary, edema, increased alanine aminotransferase, tolerated in large randomized NSCLC studies.8,38 The com- and anorexia were reported in the pemetrexed arm, whereas bination of gemcitabine and pemetrexed has yielded encour- fatigue, rash, and abdominal pain were reported in the se- aging results in several phase II trials.31,39,40 When adminis- quential pemetrexed/gemcitabine arm. tered to 60 patients with advanced NSCLC, the combination Discontinuations attributable to treatment-related ad- resulted in a 15.5% response rate and a median overall verse events occurred in four patients in the pemetrexed arm survival time of 10.1 months. Reported toxicity incidences (because of pneumonia, maculopapular rash, exanthema, and were high, with grade 3 and 4 neutropenia in 61.7% (febrile neutropenia) and two patients in the sequential pemetrexed/ neutropenia in 15%), fatigue in 23.3%, and elevations of gemcitabine arm (because of abdominal pain and allergic aspartate aminotransferase and alanine aminotransferase in dermatitis). Twelve deaths occurred during the study, none of 15% and 20% of the patients, respectively; nevertheless, which were related to a study drug. Eight deaths were disease vitamin supplementation was added only after the enrollment related, and four deaths were attributable to serious adverse of the first 13 patients.31 The other two phase II studies also events. reported increased activity of the combination but with a poor toxicity profile despite vitamin supplementation.39,40 DISCUSSION We designed the present trial to administer pemetrexed In this phase II randomized trial of elderly patients (age as a single agent or to administer pemetrexed and then Ն70 years) and patients younger than 70 years who were gemcitabine as sequential treatment. Theoretically, sequential unsuitable for platinum-based chemotherapy (with a PS of 2), therapy allows the investigation of a single agent or a com- single-agent pemetrexed and sequentially applied pem- bination of drugs after effective chemotherapy with standard etrexed/gemcitabine were effective and well tolerated as regimens but before the emergence of clinical drug resis- first-line treatments of advanced NSCLC. For the pemetrexed tance. Additionally, sequential therapy may minimize the and sequential pemetrexed/gemcitabine arms, respectively, potential for toxicity seen with concurrently administered the median TtPD was 4.5 months and 4.1 months, and the two-drug regimens while permitting full doses of each indi- median overall survival time was 4.7 months and 5.4 months. vidual drug. This could result in a potentially more effective The median PFS time, which was analyzed because of an regimen with a lower toxicity profile than those seen in unexpectedly high rate of censoring for TtPD, was 3.3 combination treatment. Sequential administration has been months for each treatment arm. Objective tumor responses evaluated in both elderly and PS 2 patients using vinorelbine were reported in two patients treated with pemetrexed and in and docetaxel.41 In 75 elderly patients with a PS of 0 or 1, the five patients treated with sequential pemetrexed/gemcitabine. sequential approach yielded encouraging results—a response

Copyright © 2007 by the International Association for the Study of Lung Cancer 227 Gridelli et al. Journal of Thoracic Oncology • Volume 2, Number 3, March 2007 rate of 21% and a median survival time of 9 months—with with advanced non-small cell lung cancer. Int J Clin Pract 2006;60: good tolerability.41 340–343. 8. Gridelli C, Perrone F, Gallo C, et al. Chemotherapy for elderly patients One possible limitation to our study is the high rate of with advanced non-small-cell lung cancer: the Multicenter Italian Lung censoring for the TtPD analysis (15 patients with no objective cancer in the Elderly Study (MILES) phase III randomized trial. J Natl PD). Furthermore, the unexpectedly low TtPD and overall Cancer Inst 2003;95:362–372. survival results might be explained by the high percentage of 9. Frasci G, Lorusso V, Panza N, et al. Gemcitabine plus vinorelbine yields better survival outcome than vinorelbine alone in elderly patients with PS 2 patients in our study: 35.6%, 20.7% of whom were advanced non-small cell lung cancer. A Southern Italy Cooperative younger than 70 years. Finally, although other studies often Oncology Group (SICOG) phase III trial. Lung Cancer 2001;34(suppl group PS 2 and elderly patients together, which was typical at 4):S65–S69. the time our study was planned, elderly patients with good PS 10. The Elderly Lung Cancer Vinorelbine Italian Study (ELVIS) Group. tend to fare as well as younger patients with respect to Effects of vinorelbine on quality of life and survival of elderly patients with advanced non-small-cell lung cancer. J Natl Cancer Inst 1999;91: standard chemotherapy, whereas poor PS patients, regardless 66–72. of age, tend to have mostly palliative benefits.42–45 Because it 11. Gridelli C, Maione P, Rossi A, et al. Chemotherapy of stage I to III is well known that PS 0/1 and PS 2 patients constitute non-small cell lung cancer in elderly patients. American Society of different cohorts with different prognoses in advanced Clinical Oncology, 2005 Educational Book, 41st Annual Meeting, May 13–17, 2005, Orlando, FL, 2005. Pp. 611–617. NSCLC, we also examined efficacy by PS in our trial, and we 12. Sequist LV, Fidias P. Treatment of advanced non-small cell lung cancer found that PS 0/1 patients performed better than PS 2 pain the elderly. Semin Respir Crit Care Med 2005;26:289–297. tients. This finding confirms that PS 2 is an important pre- 13. Pfister DG, Johnson DH, Azzoli CG, et al. American Society of Clinical dictor of a poor outcome (shorter PFS and survival times). Oncology treatment of unresectable non-small-cell lung cancer guide- Thus, dedicated PS 2–specific studies are needed along with line: update 2003. J Clin Oncol 2004;22:330–353. 14. Gridelli C, Aapro M, Ardizzoni A, et al. Treatment of advanced a better understanding of the reasons for compromised PS non-small-cell lung cancer in the elderly: results of an international (comorbidity, disease burden, or both) and of how these expert panel. J Clin Oncol 2005;23:3125–3137. factors affect treatment outcome.6,44 15. Buccheri G, Ferrigno D, Tamburini M. Karnofsky and ECOG perfor- In conclusion, both single-agent pemetrexed (500 mg/m2) mance status scoring in lung cancer: a prospective, longitudinal study of and sequentially administered pemetrexed/gemcitabine (500 536 patients from a single institution. Eur J Cancer 1996;32A:1135– 2 2 1141. mg/m of pemetrexed and then 1200 mg/m of gemcitabine) 16. Gridelli C, Ardizzoni A, Le Chevalier T, et al. Treatment of advanced have shown moderate activity and are well-tolerated regi- non-small-cell lung cancer patients with ECOG performance status 2: mens as first-line treatments of advanced NSCLC in elderly results of a European Experts Panel. Ann Oncol 2004;15:419–426. patients (Ն70 years of age) or in patients younger than 70 17. Neubauer MA, Reynolds CH, Joppert MG, et al. Results of a phase II trial of gemcitabine in patients with non-small-cell lung cancer and a years who are unsuitable for platinum-based combination performance status of 2. Clin Lung Cancer 2005;6:245–249. chemotherapy (i.e., patients with a PS of 2). Although the 18. Blackhall FH, Bhosle J, Thatcher N. Chemotherapy for advanced non- study was designed to identify which of these two treatment small cell lung cancer patients with performance status 2. Curr Opin modalities should be selected for further study, the results Oncol 2005;17:135–139. 19. Shih C, Chen VJ, Gossett LS, et al. LY231514, a pyrrolo [2,3-d] suggest that either treatment is acceptable for further study in pyrimidine-based antifolate that inhibits multiple folate-requiring en- a more homogeneous patient population such as PS 0/1 zymes. Cancer Res 1997;57:1116–1123. elderly or PS 2 adult patients. 20. Clarke SJ, Abratt R, Goedhals L, et al. Phase II trial of pemetrexed disodium (ALIMTA, LY231514) in chemotherapy-naı¨ve patients with advanced non-small-cell lung cancer. Ann Oncol 2002;13:737–741. ACKNOWLEDGMENTS 21. Rusthoven JJ, Eisenhauer E, Butts C, et al. Multitargeted antifolate The authors would like to acknowledge Noelle Gasco LY231514 as first-line chemotherapy for patients with advanced non- small-cell lung cancer: a phase II study. National Cancer Institute of for her editorial assistance and Donna L. Miller for her Canada Clinical Trials Group. J Clin Oncol 1999;17:1194–1199. technical assistance. 22. Smit E, Mattson K, von Pawel J, et al. ALIMTA (pemetrexed disodium) as second-line treatment of non-small-cell lung cancer: a phase II study. Ann Oncol 2003;14:455–460. REFERENCES 23. Hanna N, Shepherd FA, Fossella FV, et al. Randomized phase III trial of 1. American Cancer Society. Cancer Facts & Figures 2006. Atlanta, GA: pemetrexed versus docetaxel in patients with non-small-cell lung cancer American Cancer Society, 2006. Available at: http://www.cancer.org. previously treated with chemotherapy. J Clin Oncol 2004;22:1589– Accessed June 19, 2006. 1597. 2. Ries LAG, Eisner MP, Kosary CL, et al. SEER Cancer Statistics Review 24. Niyikiza C, Baker SD, Seitz DE, et al. Homocysteine and methylmalonic 1975–2000. Bethesda, MD: National Cancer Institute, 2003. Available acid: markers to predict and avoid toxicity from pemetrexed therapy. at: http://seer.cancer.gov/csr/1975_2000/ Accessed June 19, 2006. Mol Cancer Ther 2002;1:545–552. 3. Gridelli C, Perrone F, Monfardini S. Lung cancer in the elderly. Eur J 25. Huang P, Chubb S, Hertel LW, et al. Action of 2=,2=-difluorodeoxycy- Cancer 1997;33:2313–2314. tidine on DNA synthesis. Cancer Res 1991;51:6110–6117. 4. Lewis JH, Kilgore ML, Goldman DP, et al. Participation of patients 65 26. Quoix E, Breton JL, Ducolone A, et al. First-line chemotherapy with years of age or older in cancer clinical trials. J Clin Oncol 2003;21: gemcitabine in advanced non-small cell lung cancer elderly patients: a 1383–1389. randomized phase II study of 3-week versus 4-week schedule. Lung 5. Hutchins LF, Unger JM, Crowley JJ, et al. Underrepresentation of Cancer 2005;47:405–412. patients 65 years of age or older in cancer-treatment trials. N Engl J Med 27. Bianco V, Rozzi A, Tonini G, et al. Gemcitabine as single-agent 1999;341:2061–2067. chemotherapy in elderly patients with stages III–IV non-small cell lung 6. Hennessy BT, Hanrahan EO, Breathnach OS. Chemotherapy options for cancer (NSCLC): a phase II study. Anticancer Res 2002;22:3053–3056. the elderly patient with advanced non-small cell lung cancer. Oncologist 28. Martoni A, Di Fabio F, Guaraldi M, et al. Prospective phase II study of 2003;8:270–277. single-agent gemcitabine in untreated elderly patients with stage IIIB/IV 7. Cheong KA, Chrystal K, Harper PG. Management of the elderly patient non-small-cell lung cancer. Am J Clin Oncol 2001;24:614–617.

29. Ricci S, Antonuzzo A, Galli L, et al. Gemcitabine monotherapy in regimens in advanced non-small-cell lung cancer: a meta-analysis of elderly patients with advanced non-small cell lung cancer: a multicentre survival outcomes. Lung Cancer 2005;49:69–80. phase II study. Lung Cancer 2000;27:75–80. 39. Ma CX, Nair S, Thomas S, et al. Randomized phase II trial of three 30. Altavilla G, Adamo V, Buemi B, et al. Gemcitabine as single agent in schedules of pemetrexed and gemcitabine as front-line therapy for the treatment of elderly patients with advanced non-small cell lung advanced non-small-cell lung cancer. J Clin Oncol 2005;23:5929–5937. cancer. Anticancer Res 2000;20:3675–3678. 40. McCleod M, Treat J, Christiansen NP, et al. Pemetrexed (P) plus 31. Monnerat C, Le Chevalier T, Kelly K, et al. Phase II study of pem- gemcitabine (G) as front-line chemotherapy for patients with locally etrexed-gemcitabine combination in patients with advanced-stage non- advanced or metastatic non-small cell lung cancer (NSCLC): a phase II small cell lung cancer. Clin Cancer Res 2004;10:5439–5446. clinical trial. J Clin Oncol 2005;23(16 suppl):650s. 32. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to 41. Hesketh PJ, Chansky K, Lau DH, et al. Sequential vinorelbine (V) and evaluate the response to treatment in solid tumors. J Natl Cancer Inst docetaxel (D) in advanced non-small cell lung cancer (NSCLC) patients 2000;92:205–216. age Ͼ70, or with performance status (PS) 2: a SWOG phase II trial 33. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of (S0027). J Clin Oncol 2004;22 (14 suppl):627. classifying prognostic comorbidity in longitudinal studies: development 42. Lilenbaum RC, Herndon 2nd, JE List MA, et al. Single-agent vs and validation. J Chronic Dis 1987;40:373–383. combination chemotherapy in advanced non-small cell-lung cancer: the 34. Cancer Therapy Evaluation Program. Common Terminology Criteria for Adverse Events, Version 2.0. DCTD, NCI, NIH, DHHS 1998. Available cancer and leukemia group B (study 9730). J Clin Oncol 2005;23:190– at: http://ctep.cancer.gov. Accessed June 19, 2006. 196. 35. Kaplan EL, Meier P. Nonparametric estimation from incomplete obser- 43. Crino L, Novello S, Migliorino MR, et al. A phase III randomized trial vation. J Am Stat Assoc 1958;53:457–481. comparing three platinum-based doublets in advanced non-small cell ϭ Ն Ͻ 36. Leemis LM, Trivedi KS. A comparison of approximate interval estima- lung cancer (NSCLC): impact of PS 2vs0or1andage 70 vs 70 tors for the Bernoulli parameter. Am Stat 1996;50:63–68. on chemotherapy outcome. Proc Am Soc Clin Oncol 2002;21:315a. 37. Cripps C, Burnell M, Jolivet J, et al. Phase II study of first-line 44. Billingham LJ, Cullen MH. The benefits of chemotherapy in patient LY231514 (multi-targeted antifolate) in patients with locally advanced subgroups with unresectable non-small-cell lung cancer. Ann Oncol or metastatic colorectal cancer: an NCIC Clinical Trials Group study. 2001;12:1671–1675. Ann Oncol 1999;10:1175–1179. 45. Langer CJ, Vangel M, Schiller J, et al. Age-specific subanalysis of 38. Le Chevalier T, Scagliotti G, Natale R, et al. Efficacy of gemcitabine ECOG 1594: fit elderly patients (70–80 yrs) with NSCLC do as well as plus platinum chemotherapy compared with other platinum containing younger pts (Ͻ70). Proc Am Soc Clin Oncol 2003;22:639.