572 Phase II Trial of Trimetrexate for Patients with Advanced Gastric Carcinoma An Eastern Cooperative Oncology Group Study (E1287) Ramesh K. Ramanathan, M.D.1 Stuart Lipsitz, Sc.D.2 Robert F. Asbury, M.D.3 Raman Qazi, M.D.4 Bernard R. Greenberg, M.D.5 Daniel G. Haller, M.D.6 1 University of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania. 2 Division of Biostatistics, Dana-Farber Cancer Institute, Boston, Massachusetts. 3 Interlakes Oncology and Hematology P.C., New York. 4 Highland Hospital, Rochester, New York. 5 University of Connecticut Health Center, Farmington, Connecticut. 6 University of Pennsylvania Cancer Center, Philadelphia, Pennsylvania. BACKGROUND. A Phase II study was conducted to evaluate the response, duration of response, and duration of survival of patients with measurable gastric carcinoma treated with trimetrexate (TMTX) who had not had prior chemotherapy. METHODS. Thirty-three patients with unresectable or metastatic gastric adenocarcinoma who had not received previous chemotherapy were treated with intravenous TMTX 12 mg/m2 daily for 5 days. The dosage of TMTX was reduced to 8 mg/m2 daily for 5 days for those who had received prior radiotherapy. The cycle was repeated every 3 weeks until disease progression or unacceptable toxicity occurred. RESULTS. Thirty-three patients could be analyzed with follow-up data. There was one Grade 5 (lethal) toxicity and four Grade 4 toxicities. Hematologic toxicity was the most common. The overall response rate was 21%, the overall median progression free survival was 2.7 months, and the overall median survival was 5.9 months for the entire cohort. No patients were alive at last follow-up. CONCLUSIONS. Though TMTX as a single agent has activity in gastric carcinoma with manageable toxicity, it cannot be recommended for routine use as a single agent due to the brief duration of response and median survival. Cancer 1999;86: 572– 6. © 1999 American Cancer Society. KEYWORDS: antifolates, gastric carcinoma, Phase II study, trimetrexate. D Supported in part by Public Health Service grants CA18653, CA23318, CA07190, CA15488, CA66636, and CA21115 from the National Cancer Institute, National Institutes of Health and the Department of Health and Human Services. This study was conducted by the Eastern Cooperative Oncology Group (Robert L. Comis, M.D., Chair). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute. Address for reprints: Ramesh K. Ramanathan, M.D., University of Pittsburgh Cancer Institute, N755 MUH, 200 Lothrop Street, Pittsburgh, PA 15213. Received November 11, 1998; revision received March 10, 1999; accepted March 10, 1999. © 1999 American Cancer Society espite its decreasing incidence, gastric carcinoma remains a significant cause of cancer death in the United States. In 1997, the estimated incidence of gastric carcinoma was 22,400, with 14,000 deaths1. In patients who undergo surgical resection, there is no known effective adjuvant therapy. Many chemotherapy agents have been evaluated in this disease, with maximal response rates of '20% reported for 5-fluorouracil (5-FU), adriamycin, mitomycin-C, methylCCNU, dacarbazine, carmustine, and hydroxyurea.2,3 In general, single agent therapy has been associated with short duration of response and minimal impact on survival. Though numerous combination regimens have been studied in advanced gastric carcinoma, most regimens are associated with substantial toxicity, and their impact on survival has not been significant. Therefore, there is a need to develop new active agents in this disease.4 – 8 Trimetrexate (NSC-249008) (TMTX) was one in a series of nonclassical (i.e., without a glutamate residue) folate antagonists that were developed originally as antimalarials but were found later to be potent growth inhibitors of cultured tumor cells and more highly concentrated in leukemia cells than methotrexate (MTX).9,10 Unlike MTX, TMTX is more lipophilic and does not require activation by the enzyme folylpolyglutamyl synthetase. It also differs from MTX in that it does not require the reduced folate transport system to enter cancer cells, thus avoiding competition for uptake when administered with TMTX in Gastric Carcinoma/Ramanathan et al. leucovorin.11,12 TMTX has demonstrated activity in murine systems, including B16 melanoma, L1210 leukemia, P388 leukemia, and CD8F mammary carcinoma. The spectrum of this activity was greater than that of MTX.13 In Phase I studies, the toxicities of TMTX included myelosuppresion, nausea, mucositis, cutaneous toxicity, and renal dysfunction.14 The starting dose of TMTX in solid tumors is 8 –12 mg/m2 given daily for 5 days, with cycles repeated every 3 weeks. The higher dose of 12 mg/m2 is used in patients who were untreated previously.14 MATERIALS AND METHODS Patients Patients with surgically unresectable primary gastric adenocarcinomas were eligible for this study. Patients were required to have no previous chemotherapy. Previous radiation therapy to the indicator lesion was allowed if there was documented growth of the lesion on two consecutive evaluations resulting in a 25% increase in the indicator lesion. Histologic proof of residual, recurrent, or metastatic carcinoma was required, except in the case of pulmonary metastases. Patients were required to have an area of measurable, known malignant disease to serve as an objective indicator of response to therapy. Patients had a minimum of 3 weeks of recovery from any major surgical procedure involving bypass, 2 weeks from exploration and biopsy only, and at least 4 weeks from any previous radiation. An Eastern Cooperative Oncology Group (ECOG) performance status of 0, 1, or 2 was required. Eligible patients had normal bone marrow reserve (white blood cells [WBC] . 4000/mm3 and platelets . 150,000/mm3), renal function (blood urea nitrogen , 30 mg/dL or creatinine , 1.5 mg/dL), and normal liver function (serum albumin . 3.0 gm/dL and total bilirubin , 1.5mg/dL). All patients provided written, informed consent conforming to federal and institutional guidelines. Patients with a prior history of malignant neoplasms other than curatively treated basal cell or squamous cell carcinoma of the skin or surgically cured carcinoma of the cervix in situ were ineligible for this study. Patients with an active infectious process or malignant effusions were ineligible. Treatment Plan The treatment schedule for TMTX consisted of 12 mg/m2 intravenously (IV) daily for 5 days. The doses were diluted in D5W 100 ml and were infused over 10 minutes each day. The cycle was repeated every 3 weeks and was administered until there was disease progression, unacceptable toxicity, or patient discontinuation. When the dosage was calculated, 573 the lesser of the ideal weight or the actual weight was used to determine surface area. Patients who had received prior radiation started therapy at 8 mg/m2 IV daily for 5 days. Patients were monitored for adverse events, and therapy was repeated every 3 weeks if laboratory tests of bone marrow, kidneys, and liver had normalized. If treatment was held for more than 3 weeks, then patients were taken off of the study. Dose modifications allowed for escalation as well as reduction based on nadir counts. If WBC was .3000/mm3 and platelets . 90,000/mm3, then 125% of the prior dose of TMTX was given. If nadir counts of WBC were 2000 –2999/mm3and platelets were 50,000 – 89999/mm3, then the dose was unchanged. If WBC , 2000/mm3 or platelets were ,50,000/mm3, then 50% of the prior dose of TMTX was given. If mucositis was present on the day of scheduled new course, then treatment was held until the mucositis had resolved. For Grade 3 mucositis at any time, treatment was reinsituted at 75% of the prior dose when mucositis had resolved. If serum creatinine was ,1.5 mg/dL at the time of scheduled therapy, then the dose of TMTX was unchanged; however, if serum creatinine was in the range of 1.6 –2.0 mg/dL, then 50% of the prior dose was given. If serum creatinine was .2.0 mg/dL, then therapy was not given. Patient Registration The Central Randomization Desk randomized patients at the ECOG Operations Office over the telephone. Patients were stratified by performance status (0 –1 vs. 2) and radiotherapy (prior vs. no prior). Written informed consent was required from all patients prior to treatment assignment. Accessibility Criteria Patients were considered assessable for response if they received at least one cycle of therapy. Tumor response was classified as complete, partial, no change, or progression. Complete response was defined as the complete disappearance of all measurable or evaluable disease for at least 1 month. Partial response was defined as a minimum decrease of 50% in the sum of the products of the greatest perpendicular tumor dimensions and a definite improvement in evaluable disease estimated .50% lasting at least 1 month. No change was defined as a decrease of ,50% or an increase of ,25% in the sum of the products of the greatest perpendicular tumor dimensions and no change in evaluable disease lasting at least month. Progression was defined as a minimum increase of 25% in the sum the of products of the greatest perpendicular tumor dimensions, an increase in the area of evaluable disease estimated .25%, or the develop- 574 CANCER August 15, 1999 / Volume 86 / Number 4 ment of new lesions. Patients who relapsed had the appearance of new measurable or evaluable disease, an increase .25% over the minimum sum of the products of the greatest perpendicular tumor dimensions during response, or an estimated increase of .25% over the minimum area of evaluable disease during response. Statistical Analysis Fisher’s exact test was used to analyze contingency tables of response. Survival curves were estimated by using the method of Kaplan and Meier,15 with differences assessed by using the log-rank test.16 Logistic regression17 of response and proportional hazards regression18 of survival to identify simultaneously significant prognostic covariates were based on the likelihood ratio test. To calculate confidence intervals for a binomial parameter for a two-stage, Phase II clinical trial, we used the method of Atkinson and Brown.19 RESULTS Administrative Information The protocol was activated in September 1987, with an accrual goal of 32 patients (30 evaluable) with gastric carcinoma to be entered by the 23 participating institutions. In May 1988, the study was redesigned as a two-step study and also added the site of small bowel carcinoma as another stratum. The first stage of the new design had separate accrual goals of 15 evaluable patients in the 2 disease sites. After accruing 15 evaluable patients in a given site, if 3 or more responses were seen, then the accrual was to be extended for another 15 evaluable patients. There were 3 responses in the first 15 patients with gastric carcinoma, and accrual continued until September 1989. In October 1989, after having accrued only 2 small bowel patients, the study was closed to small bowel patients because of poor accrual. The ineligibility rate was low, with the 1 ineligible patient having small bowel carcinoma and nonmeasurable disease. TABLE 1 Patient Characteristics on Study Characteristic Age (yrs) Median (range) Performance status ECOG 0–1 ECOG 2 Gender Female Male Prior Radiotherapy Yes No Total (%) 62 (46–83) — 29 4 88 12 7 26 21 79 1 32 3 97 ECOG: Eastern Cooperative Oncology Group. patient died 26 days after treatment was stopped. The adverse drug report states “microangiopathic hemolytic anemia probably associated with diffuse bone marrow disease, and probably not drug related.” There have been four Grade 4 hematologic toxicities, all leucopenia in nature. Table 2 summarizes the incidence of toxicities related to therapy, which ranged from none to lethal in nature. Hematologic toxicity, leucopenia in particular, was the most common toxicity. Response A total of 7 patients had complete or partial responses. Objective response data are summarized in Table 3. Three patients had a complete response with sites of disease in the liver and the lymph nodes and a soft tissue mass. The overall response rate was 21% (7 of 33 evaluable patients). A 90% confidence interval for the probability of response in the gastric patients, using the method described by Atkinson and Brown,19 was 11.3–39.4%. The median response duration for the 7 patients with responses was 6 months, with durations ranging from 1 month to 11.5 months. Patient Characteristics Progression Free Survival Thirty-three patients with gastric adenocarcinoma were accrued. None had received prior chemotherapy. One patient had received previous radiotherapy. The median age was 62 years. Table 1 summarizes the distribution of patient characteristics from the study, including the stratification factors. Progression free survival time was computed from randomization to progression or death, which ever came first, or was censored at the time last known alive without progression. Of the 33 patients who could be analyzed, all have died. The overall median progression free survival was 2.7 months. The analysis of single covariates shows that disease free survival was related significantly only to performance status (P 5 0.05). Toxicity Toxicity was evaluated on all 33 patients. One patient had a lethal toxicity. In this patient, treatment was stopped after one cycle (Day 8) due to severe anemia (hemoglobin 5 3.1 gm/dL). The patient received transfusions, but blood counts remained low, and the Survival Survival time is computed from randomization to death or is censored at the time last known alive. The TMTX in Gastric Carcinoma/Ramanathan et al. 575 TABLE 2 Treatment-Related Toxicities According to Eastern Cooperative Oncology Group Toxicity Criteria Grade Toxicity 1 2 3 4 5 Gastrointestinal Vomiting Diarrhea Infection Hemorrhage Skin, mucous membrane Neurologic Genitourinary Hematologic Liver Othera 7 6 7 — — 7 6 7 4 1 7 4 8 — 4 — 5 5 — — 1 2 1 4 — — 6 — 2 — — — — — — — — 4 — — — — — — — — — — 1 — — a 11 2 5 The Grade 3 others were cardiac. TABLE 3 Objective Response Response Total no. % Complete response Partial response No change Progression Unevaluablea Cases analyzed 3 4 2 22 2 33 9 12 6 65 9 100 a Unevaluable for response (1 patient had inadequate follow-up, and 1 patient was taken off the study due to toxicity after 1 cycle). overall median survival was 5.9 months. All patients have died. DISCUSSION Patients with advanced gastric carcinoma have a poor prognosis. The median survival in this group of patients is in the range of 5–10 months, and few untreated patients survive more than 1 year after initial diagnosis. 5-FU is the single most extensively studied agent in this disease, with a response rate of 21%.20 Among the newer agents, docetaxel has shown single agent activity of 14% and currently is being studied in combination with platinum and other agents.21 However, complete responses with single agent therapy are rare, and response duration generally is brief, without any appreciable improvement in overall survival. In this study, single agent TMTX showed a response rate of 21% in patients with gastric adenocarcinoma. The response rate in patients with small bowel adenocarcinoma could not be assessed due to poor accrual and early closure of the study to this site. The activity of TMTX in this study is comparable to other single agents in this disease. There is only one other study evaluating TMTX in advanced gastric carcinoma. The Southwest Oncology Group used a similar dose schedule of TMTX (8 –12 mg/m2/day for 5 days). A response rate of 4% was reported (95% confidence interval, 0 –22%), and the median survival was 6 months. These results are comparable to our study, in which the response rate was 21% (90% confidence interval, 11.3– 39.4%), and the median survival was 5.9 months. Hematologic toxicity, similar to our report, also was predominant, with Grade 3 toxicity seen in 10 out of 23 patients. In general, as a single agent, the activity of TMTX in various solid tumors has been modest, with response rates generally '10%.14 It appears that the combination of TMTX and 5-FU may be more active than either drug alone. Studies suggest that TMTX may modulate the activity of 5-FU and enhance its cytotoxicity.23,24 The sequence of 5-FU and TMTX administration appears to be important, because synergistic cytotoxicity was seen when TMTX was added before 5-FU. When the sequence was reversed and 5-FU was given before TMTX, antagonism was observed.25 The mechanism of synergy is thought to be due to increased intracellular accumulation of 5-phosphoribosyl-1-pyroposphate caused by antifolate inhibition of purine synthesis.25 In cell lines, 5-FU and leucovorin showed increased cytotoxicity in the presence of TMTX, which was greater than the combination of 5-FU and TMTX alone.26 Based on these observations, trials of the combination TMTX, 5-FU, and leucovorin have been conducted.27–29. In colorectal carcinoma, this regimen has shown activity in both untreated and previously treated patients.27–29 Gastrointestinal toxicity, in particular diarrhea, appears to be the most significant toxicity of this regimen, and this 576 CANCER August 15, 1999 / Volume 86 / Number 4 may be ameliorated by a high dose loperamide regimen. We do not recommend the use of single agent TMTX in gastric carcinoma due to the brief duration of response and median survival. However, further evaluation of TMTX with 5-FU and leucovorin is warranted, because this regimen has sound preclinical rationale as well as encouraging activity in colorectal carcinoma. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Parker SL, Tong T, Bolden S, Wingo PA: Cancer statistics, 1997. CA Cancer J Clin 1997;47:5–27. Comis RL, Carter SK. Integration of chemotherapy into combined modality therapy of solid tumors. Cancer Treat Rev 1974;1:221. Moertel CG, Mittelman AJ, Bakemeier RF. Sequential and combination chemotherapy of advanced gastric cancer. Cancer 1976;38:678. Moertel CG, Rubin J, O’Connell MJ, Schutt AJ, Wienand HS. A Phase II study of combined 5-fluorouracil, doxorubicin and cisplatin in the treatment of advanced upper gastrointestinal adenocarcinomas. J Clin Oncol 1986;4:1053–7. Rougier P, Ducreux M, Mahjoubi M, Pignon JP, Bellefqih S, Olivera J, et al. Efficacy of combined 5-fluorouracil and cisplatinum in advanced gastric carcinomas. A phase II trial with prognostic factor analysis. Eur J Cancer 1994;30A: 1263–9. Kelsen D, Atiq OT, Saltz L, Niedzwiecki D, Ginn D, Chapman D, et al. FAMTX versus etoposide, doxorubicin, and cisplatin: a random assignment trial in gastric cancer. J Clin Oncol 1992;10:541– 8. Wilke H, Wils JA, Rougier P, Lacave A, Van Cutsem E, Vanhoeferi U, et al. Preliminary analysis of a randomized Phase III trial of FAMTX versus ELF versus cisplatin/ 5FU in advanced gastric cancer. A trial of the EORTC gastrointestinal tract cooperative group and the AIO. Proc Am Soc Clin Oncol 1995;14:206. Preusser P, Wilke H, Achterrath W, Fink U, Lenaz L, Heinicke A, et al. Phase II study with the combination etoposide, doxorubicin, and cisplatin in advanced measurable gastric cancer. J Clin Oncol 1989;7:1310 –7. Elslager EF. In: Castle RN, Townsend LB, editors. Lectures in heterocyclic chemistry, vol 11. Orem, UT: Hetero Corp., 1974;97. Kamen, BA, Eibl B, Cashmore A, Bertino J. Uptake and efficacy and transport of trimetrexate (TMQ, 2,4-diamin-5methyl-6-[(3,4,5-trimethoxyanilino) methyl] quinazoline in methotrexate resistant leukemia cells in vitro. Biochem Pharmacol 1984;33:1697–9. Jackson RC, Fry DW, Boritzki TJ, Besserer JA, Leopold WR, Sloan BJ, et al. Biochemical pharmacology of the lipophilic antifolate, trimetrexate. Adv Enyz Regul 1984;22:187–206. Weir EC, Cashmore AR, Dreyer RN, Graham ML, Hsiao N, Moroson BA, et al. Pharmacology and toxicity of a potent “nonclassical” 2,4-diamino quinazoline folate antagonist, trimetrexate, in normal dogs. Cancer Res 1982;42:1696 –702. Ohnoshi T, Ohnuma T, Takahashi I, Scanlon K, Kamen BA, Holland JF. Establishment of methotrexate resistant human 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. acute lymphoblastic leukemia cell in culture and effects of folate antagonists. Cancer Res 1982;42:1665–70. Haller DG. Trimetrexate: experience with solid tumors. Semin Oncol 1997;24:S18 –76. Kaplan EL, Meier P. Nonparametric estimation of incomplete observations. J Am Stat Assoc 1958;53:457– 81. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163–70. Cox DR. Analysis of binary data. London: Chapman and Hall, 1970. Cox DR. Regression models and life tables. J R Stat Soc B 1972;34:187–202. Atkinson EN, Brown SW. Confidence limits for probability of response in multistage Phase II clinical trials. Biometrics 1985;41:741– 4. Cocconi G, DeLisi V, Di Blasio B. Randomized comparison of 5-FU alone or combined with mitomycin and cytarabine (MFC) in the treatment of advanced gastric cancer. Cancer Treat Rep 1982;66:1263– 6. Einzig AI, Neuberg D, Remick SC, Karp DD, O’Dwyer PJ, Stewart JA, et al. Phase II trial of docetaxel (taxotere) in patients with adenocarcinoma of the upper gastrointestinal tract previously untreated with cytotoxic chemotherapy: the Eastern Cooperative Oncology Group (ECOG) results of protocol E1293. Med Oncol 1996;13:87–93. Hantel A, Tangen CM, Macdonald JS, Richman SP, Pugh RP, Pollock T, et al. Phase II trial of trimetrexate in untreated advanced gastric carcinoma. A Southwest Oncology Group study. Invest N Drugs 1994;12:155–7. Cadman E, Heimer R, Davis L. Enhanced 5-fluorouracil nucleotide formation after methotrexate administration: explanation for drug synergism. Science 1979;205:1135–7. Hudes GR, LaCreta F, DeLap RJ, Grillo-Lopez AJ, Catalano R, Comis RL. Phase I clinical and pharmacologic trial of trimetrexate in combination with 5-fluorouracil. Cancer Chemother Pharmacol 1989;24:117–2. Sobrero A, Romanini A, Russello O, Nicolin A, Rosso R, Bertino JR. Sequence-dependent enhancement of HCT-8 cell kill by trimetrexate and fluoropyrimidines: implications for the mechanism of this interaction. Eur J Cancer Clin Oncol 1989;25:977– 82. Romanini A, Li WW, Colofiore JR, Bertino JR. Leucovorin enhances cytotoxicity of trimetrexate/fluorouracil, but not methotrexate/fluorouracil, in CCRF-CEM cells. J Natl Cancer Inst 1992;84:1033– 8. Conti JA, Kemeny NE, Tong W, Andre M, Bertino JR. Final results of a phase I trial of sequential trimetrexate (TMTX), fluorouracil (FU) and high-dose leucovorin (LV) in previously treated patients (pts) with gastrointestinal (GI) carcinoma (CA). Proc Am Soc Clin Oncol 1997;16:A1014. Blanke CD, Kasimis B, Schein P, Capizzi R, Kurman M. Phase II study of trimetrexate, fluorouracil, and leucovorin for advanced colorectal cancer. J Clin Oncol 1997;15: 915–20. Conti JA, Kemeny N, Seiter K, Goker E, Tong W, Andre M, et al. Trial of sequential trimetrexate, fluorouracil, and high-dose leucovorin in previously treated patients with gastrointestinal carcinoma. J Clin Oncol 1994;12:695– 700.