1329 A Phase II Trial of Methotrexate, Cisplatin, 5-Fluorouracil, and Leucovorin in the Treatment of Invasive and Metastatic Urothelial Carcinoma William K. Oh, M.D.1 Judith Manola, M.S.2 Jerome P. Richie, M.D.3 Kevin R. Loughlin, M.D.3 Philip W. Kantoff, M.D.1 1 Lank Center for Genitourinary Oncology, Department of Adult Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts. 2 Department of Biostatistical Sciences, Dana-Farber Cancer Institute, Boston, Massachusetts. 3 Division of Urology, Brigham and Women’s Hospital, Boston, Massachusetts. BACKGROUND. The current treatment of advanced urothelial carcinoma generates high response rates but is associated with poor overall survival. The current study evaluated the efficacy and toxicity of a new combination of active drugs in the treatment of urothelial carcinoma. METHODS. Twenty-four patients with muscle invasive or metastatic urothelial carcinoma were enrolled. Fifteen patients (63%) had metastatic disease whereas 9 patients had T2–T4 disease. Three patients were unevaluable for response because of significant toxicity. Patients were treated every 28 days with methotrexate, 60 mg/m2, intravenously (i.v.) on Day 1; cisplatin, 25 mg/m2/day, by continuous i.v. infusion on Days 2– 6; 5-flurouracil (5-FU) 800 mg/m2/day by continuous i.v. infusion on Days 3– 6; and leucovorin, 500 mg/m2/day, by continuous i.v. infusion on Days 2– 6. Dosage in subsequent cycles was adjusted according to toxicity. RESULTS. The median follow-up was 81 months (range, 53–971 months). The overall response rate (complete response 1 partial response) for all 24 patients was 63% (95% confidence interval, 41– 81%). The median survival was 65 months in the patients with muscle invasive disease and 17 months in the patients with metastatic disease. The duration of response in patients with metastatic disease was 6 months (range, 4 –19 months). Toxicity was significant, with 82% of patients experiencing Common Toxicity Criteria Grade 3 or 4 neutropenia and 63% experiencing Grade 3 or 4 thrombocytopenia. However, only three patients developed febrile neutropenia and gastrointestinal and neurologic toxicity was moderate. CONCLUSIONS. The combination of methotrexate, cisplatin, 5-fluorouracil, and leucovorin represents an active regimen in the treatment of urothelial carcinoma with a moderate toxicity profile. As new drugs are found to treat urothelial carcinoma, further studies will be needed to evaluate the role of traditional agents such as 5-fluorouracil and methotrexate in new combination chemotherapeutic regimens. Cancer 1999;86:1329 –34. © 1999 American Cancer Society. KEYWORDS: bladder carcinoma, combination chemotherapy, metastatic, muscle invasive. A Address for reprints: William K. Oh, M.D., Lank Center for Genitourinary Oncology, Department of Adult Oncology, Dana-Farber Cancer Institute, 44 Binney Street, Boston, MA 02115. Received November 30, 1998; revision received April 12, 1999; accepted April 12, 1999. © 1999 American Cancer Society dvanced tumors of the urothelial tract continue to pose a difficult problem for the clinician. Although combination chemotherapeutic regimens have significant activity against the disease, overall survival rates for patients with advanced urothelial carcinoma remain poor.1 The current standard treatment of metastatic transitional cell carcinoma is combination chemotherapy with three- or four-drug regimens.2 Response rates to combination chemotherapy in metastatic disease are high, ranging from 40 –70%.1,3 The most commonly used treatment, M-VAC, was developed at the Memorial Sloan-Kettering Cancer Center and is comprised of methotrexate, vinblastine, 1330 CANCER October 1, 1999 / Volume 86 / Number 7 doxorubicin, and cisplatin. In a single institution series, 30% of patients with locoregional disease and 17% of patients with metastatic disease treated with M-VAC were alive and disease free 6 years after chemotherapy.3 However, an update of a large multicenter trial revealed that only 3.7% of patients treated with M-VAC were alive and disease free at 6 years.4 In randomized trials, M-VAC was demonstrated to be superior to single agent cisplatin5 and a combination of cisplatin, doxorubicin, and cyclophosphamide (CISCA).6 The combination of cisplatin, methotrexate, and vinblastine (CMV) also has shown significant activity in the treatment of transitional cell carcinoma.7,8 Although M-VAC and CMV demonstrate high initial response rates in patients with advanced urothelial carcinoma, the vast majority of patients develop disease recurrence and die.4,9 Furthermore, the toxicity of these regimens is considerable, with frequent need for dose reductions or early discontinuation of therapy.3,10 For this reason, investigators have been searching for drugs and combinations with comparable or greater activity and fewer side effects.11 The recent reported activity of paclitaxel plus platinum analogues12,13 and gemcitabine plus cisplatin14 has generated hope that treatments may improve.15 Nonetheless, the primary goal of chemotherapy for patients with advanced urothelial carcinoma should be the eradication of tumor and ultimate cure. To this end, the current study was designed to combine drugs with known or predicted activity in urothelial carcinoma to assess for activity and toxicity. Methotrexate16 and cisplatin5 as single agents each have response rates of approximately 30% in the treatment of transitional cell carcinoma. 5-fluorouracil (5-FU) as a single agent has modest activity against urothelial carcinoma.17 However, to our knowledge the combination of 5-FU with leucovorin potentiation has not been evaluated in urothelial carcinoma. Also of note is the embryologic relation between transitional and squamous epithelium.18 In fact, transitional cell tumors often display areas of squamous differentiation. Clinically, drugs that are active in squamous cell carcinoma (e.g., of the head and neck) also are active in transitional cell carcinoma. For these reasons, a Phase II study was undertaken to evaluate the activity and toxicity of M-PFL, a combination of methotrexate, cisplatin, 5-FU, and leucovorin. The combination of cisplatin, 5-FU, and leucovorin (PFL) has been administered safely to patients with carcinoma of the head and neck with high response rates.19 Because of methotrexate’s high single agent activity in transitional cell carcinoma, this drug was added to the PFL regimen. TABLE 1 Baseline Patient Characteristics No. enrolled Gender Age (yrs) Clinical stage Sites of metastases among patients with metastases Primary tumor Histology Performance status Creatinine clearance Evaluable Unevaluable Male Female Median (range) Metastatic Muscle invasive (T2–T4, N0M0) 21 3 15 9 61 (43–74) 15 9 Pelvic lymph nodes only Distant lymph nodes Lung Bone Liver Bladder Renal pelvis Transitional cell Squamous cell 0 1 2 Median (range) 1 10 4 3 1 23 1 23 1 14 9 1 84 (52–138) 87.5% 12.5% 63% 37% 63% 37% 7% 67% 27% 20% 7% 96% 4% 96% 4% 58% 38% 4% MATERIALS AND METHODS Twenty-four patients were enrolled in the study. Informed consent was obtained from all participants. Baseline patient characteristics are listed in Table 1. Patients were eligible if they had transitional, squamous, or mixed carcinoma of the urothelial tract. Patients with invasive tumors (T2–T4b based on the 1997 American Joint Committee on Cancer criteria) were eligible, as were patients with metastatic disease, including patients with lymph node or visceral involvement. However, measurable disease was necessary. Entry criteria included no prior chemotherapy for bladder carcinoma, a Cancer and Leukemia Group B performance status of 0 –2, age , 80 years, 24-hour creatinine clearance . 50 mL/minute, bilirubin and aspartate aminotransferase # 25% above normal, leukocyte count . 4000/mm3, platelet count . 100,000/ mm3, and signed informed consent. Twenty-three of 24 patients (96%) had transitional cell carcinoma, whereas 1 patient had squamous cell carcinoma. Nine patients (37%) had invasive bladder carcinoma without evidence of metastases. Any patient who was seen at the Dana-Farber Cancer Institute during the study period with a new diagnosis of muscle invasive or metastatic urothelial carcinoma and who met the criteria for study entry was considered for this protocol. The nine patients with muscle invasive disease clearly were derived from a selected cohort because they expressed a specific interest in a bladder-sparing approach after reviewing M-PFL in Advanced Urothelial Carcinoma/Oh et al. FIGURE 1. Schematic of the methotrexate, cisplatin, 5-fluorouracil, and leucovorin (M-PFL) regimen. MTX: methotrexate; 5-FU: 5-fluorouracil. 1331 cles of chemotherapy were administered, the extent of disease was assessed by physical examination, computed tomography (CT) scan of the abdomen and pelvis, bone scan, and chest X-ray. If indicated, cystoscopy was performed. Patients with metastatic disease were treated with up to six cycles of therapy if they continued to respond, until tumor progression or if toxicity became limiting. Patients with muscle invasive disease were treated with up to four cycles of therapy, provided they continued to respond. If any residual tumor was demonstrated by biopsy after four cycles, patients proceeded directly to radical cystectomy. If no evidence of tumor was observed, patients were candidates for pelvic radiotherapy in an attempt to preserve the bladder. Response evaluations were based on the site of initial measurable disease. A complete response (CR) was defined as absence of all clinically detectable tumor for at least 4 weeks. A partial response (PR) was defined as a . 50% reduction in the sum of the products of the greatest perpendicular dimensions of all measurable lesions lasting at least 4 weeks. Progressive disease was defined as a . 25% increase in tumor volume during the treatment period or the appearance of new lesions. Toxicity was graded according to the Common Toxicity Criteria. Duration of response and survival were measured from the start of treatment. Statistical Methods FIGURE 2. Overall survival of patients treated with the methotrexate, cisplatin, 5-fluorouracil, and leucovorin (M-PFL) regimen based on the presence of muscle invasive or metastatic disease at the time of presentation. their treatment options with an oncologist or urologist. Patients were treated every 28 days with a regimen comprised of methotrexate, 60 mg/m2, intravenously (i.v.) on Day 1; cisplatin, 25 mg/m2/day, by continuous i.v. infusion on Days 2– 6; 5-FU, 800 mg/m2/day, by continuous i.v. infusion on Days 3– 6; and leucovorin, 500 mg/m2/day, by continuous i.v. infusion on Days 2– 6. Figure 1 outlines the treatment in schematic form. Patients received at least 3 L of saline per day from the continuous infusion of chemotherapy, with further i.v. fluid provided only if patients became clinically dehydrated. Dose reductions were not allowed on subsequent cycles for methotrexate or cisplatin toxicity, although a treatment delay of up to 7 days was allowed if patients experienced renal toxicity. 5-FU dose reductions were allowed for myelotoxicity, diarrhea, stomatitis, or skin rash. A physician evaluation was performed at the beginning of each cycle of chemotherapy. After two cy- The primary endpoint of this study was to determine the objective response rate of advanced urothelial carcinoma to the M-PFL regimen. The study originally was designed to accrue 40 patients so that the lower 95% confidence limit was higher than the hypothesized “null” response rate of 30%, provided that the true response rate was $ 52%. This design provided 80% power. After 24 patients, the observed response rate was 63% and therefore the study was stopped early. The lower 95% confidence limit for the observed rate was 41%. The Kaplan–Meier method was used to provide graphs of survival and estimates of median survival.20 Confidence intervals were generated for the survival curve using the Greenwood method to estimate the variance.21 Exact binomial methods were used to generate confidence intervals for the response rate. RESULTS Twenty-four patients were enrolled on this trial. Three patients were unevaluable for response because of significant toxicity after the first cycle of therapy. One of these patients, who presented with a 7-cm muscle invasive tumor, developed intractable diarrhea and 1332 CANCER October 1, 1999 / Volume 86 / Number 7 TABLE 2 Characteristics of Responding Patients Patient no. Age (yrs)/Gender Extent of disease Sites of response Best response Duration of response (mos) Overall survival (mos) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 61/M 66/M 70/M 66/M 58/M 71/M 43/M 57/F 67/M 64/M 58/F 71/F 59/M 74/F 58/M INV INV INV INV INV INV MET MET MET MET MET MET MET MET MET Bladder Bladder Bladder Bladder Bladder Bladder, prostate Bladder, prostate, bone Pelvic, distant lymph nodes Lung, distant lymph nodes Lung Lung, pelvic mass Lung, distant lymph nodes Distant lymph nodes, bladder Distant lymph nodes Bone, pelvic mass CR pCR CR pCR CR PR PR CR PR CR PR PR CR CR PR 43 841 57 421 8 16 5 12 6 8 5 4 6 19 8 861 841 651 421 16 20 20 17 17 13 10 17 14 691 16 M: male; INV: invasive; CR: complete response; pCR: pathologic complete response; PR: partial response; MET: metastatic; F: female. abdominal pain. An exploratory laparotomy was performed, which was negative. On repeat cystoscopy, he was noted to have a CR after one cycle. Radiation therapy and radical cystectomy were both offered; the patient opted for surgery, at which time he was noted to have extensive carcinoma in situ but no residual invasive disease. The other two patients had metastatic disease. Both patients developed febrile neutropenia as well as Grade 4 mucositis and diarrhea, requiring hospitalization for intravenous fluids and antibiotics. Of the 21 evaluable patients, 13 had metastatic carcinoma whereas 8 had locally advanced or muscle invasive disease. The median follow-up was 81 months (range, 53–97 months). At last follow-up, five patients were alive: four patients with muscle invasive disease and one patient who presented with distant lymphadenopathy. One patient was lost to follow-up and the remaining patients had died of disease. A total of 75 cycles of chemotherapy was administered to 24 patients. Patients received a median of 3.5 cycles of M-PFL (range, 1–5 cycles), with 8 of the 24 patients (33%) requiring a dose reduction of chemotherapy. Of the 13 evaluable patients with metastatic disease, 4 had CRs and 5 had PRs to therapy for a response rate of 69%. Two patients developed progressive disease whereas two patients had stable disease. Of the 8 patients with muscle invasive disease, 5 had CRs and 1 had a PR for a response rate of 75%. Two patients demonstrated tumor progression. The overall response rate (CR 1 PR) for all 21 patients was 71% (95% confidence interval, 48 – 89%). If the three unevaluable patients were included in the analysis as non- responders, the response rate was 63% (95% confidence interval, 41– 81%). The median overall survival in the entire group was 17 months (range, 2– 861 months). Patients with muscle invasive disease had a median survival of 65 months (range, 5– 861 months), whereas patients with metastatic disease had a median survival of 17 months (range, 2– 691 months) (Fig. 2). The median duration of response in the 15 responding patients was 8 months (range, 4 – 84 months). Among the 9 patients with metastatic disease responding to treatment, the median response duration was 6 months (range, 4 –19 months). Characteristics of responding patients are outlined in Table 2. Among the 6 responding patients with muscle invasive carcinoma, 3 (50%) received postchemotherapy bladder irradiation. At last follow-up they had not developed muscle invasive recurrence of disease and maintained intact bladders. Patient 1 initially presented with a clinical T3 transitional cell carcinoma and developed a disease recurrence . 3 years later with invasive squamous cell carcinoma, which was treated effectively with radical cystectomy. Patient 5 developed a recurrent bladder mass 3 months after achieving a CR with chemotherapy and bladder irradiation. Salvage radical cystectomy was performed, revealing a pathologic T3b lesion; the patient subsequently developed a disease recurrence and died 6 months later. Patient 6 only achieved a PR after three cycles of chemotherapy and underwent a salvage cystectomy, which also demonstrated pathologic T3b disease. He subsequently developed a systemic recurrence and died 14 months later. M-PFL in Advanced Urothelial Carcinoma/Oh et al. TABLE 3 Toxicity of M-PFL 1333 DISCUSSION No. of patients with maximum grade of toxicity Toxicity grade 0 1 2 3 4 Neutropenia Anemia Thrombocytopenia Nausea/emesis Renal Mucositis Neurotoxicity Diarrhea Fever Allergy 2 4 6 13 17 14 15 14 19 23 0 6 2 6 4 1 4 1 1 0 3 14 1 2 3 4 3 4 4 1 8 0 8 3 0 2 2 2 0 0 11 0 7 0 0 3 0 3 0 0 M-PFL: methotrexate, cisplatin, 5-fluorouracil, and leucovorin. Toxicity was formidable (Table 3). Of the 24 patients enrolled, 3 discontinued treatment due to toxicity after 1 cycle as noted earlier. These patients have been included in Table 3. Eight of the remaining 21 patients (38%) required a dose reduction in chemotherapy, primarily by shortening the length of treatment by half a day. There were no treatmentrelated deaths. Significant myelosuppression was noted with this regimen. The majority of patients had Grade 3 or 4 neutropenia, although only three patients were admitted for four episodes of febrile neutropenia. All patients eventually recovered with empiric antibiotic coverage and negative blood cultures. Only mild anemia was noted, and no red blood cell transfusions were required. Platelet toxicity was more common, with 15 of 24 patients (63%) demonstrating Grade 3 or 4 thrombocytopenia. Six patients required at least 1 platelet transfusion, with 1 patient receiving a total of 12 infusions of pheresis platelets. However, no patient experienced an episode of bleeding. Gastrointestinal toxicity was moderate. Mucositis was mild or absent in the majority of patients, but three patients experienced severe mucositis requiring intravenous hydration. Diarrhea, nausea, and emesis generally were moderate and controllable. One patient developed Clostridium difficile diarrhea. Neurotoxicity was not a significant issue initially, but sensory paresthesias were noted in nine patients, and were severe enough in two to interfere with normal activities. Renal toxicity was mild and generally reversible. The majority of patients experienced mild or moderate fatigue while participating in the study, but no patients developed severe or disabling fatigue. Despite the high response rates to platinum-based combination chemotherapy for patients with advanced transitional cell carcinoma, overall survival remains poor.22 Efforts to develop regimens that may improve overall survival thus continue. M-PFL demonstrates significant activity in the treatment of invasive and advanced urothelial carcinoma. The overall response rate was 63% (95% confidence interval, 41– 81%), and was similar in both the muscle invasive and metastatic disease groups. This response rate is comparable to other platinum-based combination chemotherapy regimens that have been used in the treatment of transitional cell carcinoma, such as M-VAC3– 6 and CMV.7,8 Among those patients with T2–T4 tumors, three of six responding patients ultimately were able to avoid cystectomy with long term follow up. A fourth patient underwent cystectomy after 3 years for an invasive recurrence of a different histology from his initial tumor. Thus approximately 50% of the patients treated with M-PFL and subsequent irradiation after CR were likely to be cured of their initial invasive tumor. A high response rate was observed among patients with metastatic disease but the duration of response was limited to 6 months. Only 1 patient (6.7%) from this group was alive and disease free . 6 years later. The majority of the patients (67%) with metastases in this series had distant lymph node involvement, although visceral or bone involvement was noted in 8 of 15 patients (53%). Toxicity was significant but tolerable for the majority of the 24 patients treated. With a median of 3.5 cycles of treatment administered, Grade 3 or 4 neutropenia and thrombocytopenia was observed in 79% and 63% of patients, respectively. Nevertheless, only three patients required hospital admission for uncomplicated febrile neutropenia and only six patients required platelet transfusions, none for active bleeding episodes. Three patients were removed from the study after developing Grade 4 gastrointestinal toxicity, including mucositis and diarrhea. The major late toxicity of concern was peripheral neuropathy, which occurred in 37% of the group and was severe enough in 2 patients to be disabling. New combinations of chemotherapy currently are being developed that include paclitaxel12,13,23,24 and gemcitabine14,25,26. These newer drugs appear to have good activity against urothelial carcinoma and appear to have favorable side effect profiles. A large randomized trial comparing M-VAC with gemcitabine plus cisplatin in untreated patients with metastatic transitional cell carcinoma recently completed accrual. If 1334 CANCER October 1, 1999 / Volume 86 / Number 7 equivalent efficacy can be demonstrated from agents or combinations that have less toxicity, such regimens clearly will become the new standard of care. Nonetheless, new treatments that improve overall survival in patients with advanced urothelial carcinoma remain sorely needed. M-PFL represents an active regimen in the treatment of invasive and metastatic urothelial carcinoma. Response rates and the median survival of patients are comparable to those published with other platinumbased regimens. Toxicity was significant but could be managed effectively in the majority patients. Despite the availability of newer drugs with encouraging evidence of activity against urothelial carcinoma, older drugs with known activity should not be excluded from further study. Future studies will be needed to evaluate the role of traditional agents such as 5-FU and methotrexate in combination with newer drugs such as gemcitabine and paclitaxel. 10. 11. 12. 13. 14. 15. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. Kantoff PW, Scher HI. Chemotherapy for metastatic bladder cancer. Hematol Oncol Clin North Am 1992;6(1):195–203. Stadler WM, Kuzel TM, Raghavan D, Levine E, Vogelzang NJ, Roth B, et al. Metastatic bladder cancer: advances in treatment. Eur J Cancer 1997;33(Suppl 1):S23– 6. Sternberg CN, Yagoda A, Scher HI, Watson RC, Geller N, Herr HW, et al. Methotrexate, vinblastine, doxorubicin, and cisplatin for advanced transitional cell carcinoma of the urothelium. Efficacy and patterns of response and relapse. Cancer 1989;64(12):2448 –58. Saxman SB, Propert KJ, Einhorn LH, Crawford ED, Tannock I, Raghavan D, et al. Long-term follow-up of a phase III intergroup study of cisplatin alone or in combination with methotrexate, vinblastine, and doxorubicin in patients with metastatic urothelial carcinoma: a cooperative group study. J Clin Oncol 1997;15(7):2564 –9. Loehrer PJ Sr., Einhorn LH, Elson PJ, Crawford ED, Kuebler P, Tannock I, et al. A randomized comparison of cisplatin alone or in combination with methotrexate, vinblastine, and doxorubicin in patients with metastatic urothelial carcinoma: a cooperative group study [published erratum appears in J Clin Oncol 1993;11(2):384]. J Clin Oncol 1992;10(7): 1066 –73. Logothetis CJ, Dexeus FH, Finn L, Sella A, Amato RJ, Ayala AG, et al. A prospective randomized trial comparing MVAC and CISCA chemotherapy for patients with metastatic urothelial tumors. J Clin Oncol 1990;8(6):1050 –5. Harker WG, Meyers FJ, Freiha FS, Palmer JM, Shortliffe LD, Hannigan JF, et al. Cisplatin, methotrexate, and vinblastine (CMV): an effective chemotherapy regimen for metastatic transitional cell carcinoma of the urinary tract. A Northern California Oncology Group study. J Clin Oncol 1985;3(11): 1463–70. Pajk B, Cufer T, Cervek J, Zakotnik B. First-line chemotherapy with cisplatin, methotrexate and vinblastine in metastatic bladder cancer. Tumori 1996;82(5):453–5. Igawa M, Urakami S, Shiina H, Ishibe T, Kadena H, Usui T. Long-term results with M-VAC for advanced urothelial can- 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. cer: high relapse rate and low survival in patients with a complete response. Br J Urol 1995;76(3):321– 4. Witjes JA, Wullink M, Oosterhof GO, de Mulder P. Toxicity and results of MVAC (methotrexate, vinblastine, adriamycin and cisplatin) chemotherapy in advanced urothelial carcinoma. Eur Urol 1997;31(4):414 –9. Small EJ, Fippin LJ, Ernest ML, Carroll PR. A carboplatinbased regimen for the treatment of patients with advanced transitional cell carcinoma of the urothelium. Cancer 1996; 78(8):1775– 80. Zielinski CC, Schnack B, Grbovic M, Brodowicz T, Wiltschke C, Steger G, et al. Paclitaxel and carboplatin in patients with metastatic urothelial cancer: results of a phase II trial. Br J Cancer 1998;78(3):370 – 4. Redman BG, Smith DC, Flaherty L, Du W, Hussain M. Phase II trial of paclitaxel and carboplatin in the treatment of advanced urothelial carcinoma. J Clin Oncol 1998;16(5): 1844 – 8. Kaufman DW, Stadler W, Carducci M, Raghavan D, Murphy B, Aisner J, et al. Gemcitabine (GEM) plus cisplatin (CDDP) in metastatic transitional cell carcinoma: final results of a phase II study [abstract]. Proceedings Am Soc Clin Oncol 1998;17:320a. Fagbemi SO, Stadler WM. New chemotherapy regimens for advanced bladder cancer. Semin Urol Oncol 1998;16(1): 23–9. Yagoda A. Chemotherapy of urothelial tract tumors. Cancer 1987;60(3 Suppl):574 – 85. Rotman M, Macchia R, Silverstein M, Aziz H, Choi K, Rosenthal J, et al. Treatment of advanced bladder carcinoma with irradiation and concomitant 5-fluorouracil infusion. Cancer 1987;59(4):710 – 4. Kunze E. Histogenesis of nonurothelial carcinomas in the human and rat urinary bladder. Exp Toxicol Pathol 1998; 50(4 – 6):341–55. Clark JR, Busse PM, Norris CM Jr., Andersen JW, Dreyfuss AI, Rossi RM, et al. Induction chemotherapy with cisplatin, fluorouracil, and high-dose leucovorin for squamous cell carcinoma of the head and neck: long-term results. J Clin Oncol 1997;15(9):3100 –10. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 81. Collett D. Modelling survival data in medical research. London: Chapman and Hall, 1994. Brinkley WM, Torti FM. Systemic chemotherapy of transitional cell carcinoma of the urothelium. Semin Surg Oncol 1997;13(5):365–75. Roth BJ, Dreicer R, Einhorn LH, Neuberg D, Johnson DH, Smith JL, et al. Significant activity of paclitaxel in advanced transitional-cell carcinoma of the urothelium: a phase II trial of the Eastern Cooperative Oncology Group. J Clin Oncol 1994;12(11):2264 –70. Bajorin DF, McCaffrey JA, Hilton S, Mazumdar M, Kelly WK, Scher HI, et al. Treatment of patients with transitional-cell carcinoma of the urothelial tract with ifosfamide, paclitaxel, and cisplatin: a phase II trial. J Clin Oncol 1998;16(8):2722–7. Moore MJ, Tannock IF, Ernst DS, Huan S, Murray N. Gemcitabine: a promising new agent in the treatment of advanced urothelial cancer. J Clin Oncol 1997;15(12):3441–5. Stadler WM, Kuzel T, Roth B, Raghavan D, Dorr FA. Phase II study of single-agent gemcitabine in previously untreated patients with metastatic urothelial cancer. J Clin Oncol 1997; 15(11):3394 – 8.