Phase II/III Study of Doxorubicin With Fluorouracil Compared With Streptozocin With Fluorouracil or Dacarbazine in the Treatment of Advanced
http://www.100md.com
《临床肿瘤学》
the University of Pennsylvania, Philadelphia, PA
Dana Farber Cancer Institute, Boston, MA
Creighton University, Omaha, NE
ABSTRACT
PURPOSE: Optimal treatments for metastatic carcinoid tumor remain undefined, and the role of chemotherapy for symptomatic patients with progressive disease is uncertain.
PATIENTS AND METHODS: Two hundred forty-nine patients with advanced carcinoid tumors were randomized to either doxorubicin with fluorouracil (FU/DOX) or streptozocin with fluorouracil (FU/STZ). Patients crossed over to the dacarbazine (DTIC) treatment after disease progression following first-line treatment (either FU/DOX or FU/STZ), and 73 patients were assigned to one of these three treatments based on their previous treatment or on abnormal baseline cardiac or renal function.
RESULTS: In the randomized group, there was no difference between FU/DOX and FU/STZ in response rates (15.9% v 16%) and progression-free survival (4.5 v 5.3 months). FU/STZ (24.3 months) was superior to FU/DOX (15.7 months; P = .0267) in median survival. The response rate of crossover DTIC treatment was 8.2%, with a median survival of 11.9 months. Hematologic toxicities were the major treatment-related toxicities for both FU/DOX and FU/STZ, and mild to moderate renal toxicity was reported in 40 (34.8%) of 115 patients in the FU/STZ arm.
CONCLUSION: Response to all three treatment regimens were modest. FU/STZ improved survival compared with the doxorubicin-based regimen, suggesting that the combination should be considered to be an active regimen of therapy when chemotherapy is judged to be an option for selected patients with carcinoid tumors.
INTRODUCTION
Carcinoid tumors are the most common neuroendocrine tumors of the gastrointestinal tract, accounting for up to one third of all small-intestinal malignancies.1,2 Because of the relatively indolent nature of the disease, many patients present with metastatic disease at the time of diagnosis, and two thirds of these patients develop carcinoid syndrome from the secretion of bioactive substances.2 Although the introduction of somatostatin analogs has improved the hormonally related carcinoid syndrome, the prognosis of patients with progressive unresectable metastases is fairly dismal, with significant symptoms or disability resulting from their malignant disease and poor overall outcome, and palliation of metastatic carcinoid tumor has continued to be a challenge to oncologists.
This Eastern Cooperative Oncology Group (ECOG) study was designed to evaluate and confirm the toxicities, response rate, progression-free survival (PFS), and overall survival benefit of fluorouracil with streptozocin (FU/STZ), compared with fluorouracil with doxorubicin (FU/DOX) in the treatment of selected patients with metastatic carcinoid tumor. The protocol was based the results of two ECOG trials (EST 3272 and EST 5375), which demonstrated objective response rates of 22% and 33%, respectively, in patients treated with FU/STZ and 21% in patients with doxorubicin alone.3,4 However, there was uncertainty in the palliative or survival benefit from these two trials, and gastrointestinal and hematologic toxicities plus STZ-associated renal toxicity were of concern. Dacarbazine (DTIC) has also been reported to have some effectiveness in metastatic carcinoid tumor.5,6 The secondary goal of this study, therefore, was to assess the activity of DTIC.
PATIENTS AND METHODS
All patients had histologically proven, unresectable advanced carcinoid tumors that were measurable or assessable by biochemical or radiographic measures. Because of the characteristically indolent nature of carcinoid tumors, patients were chosen for treatment only if they had significant symptoms or disability (unable to carry out usual work activities) resulting from their malignant disease by radiologic or biologic measures. Written informed consent conforming to federal and institutional guidelines was obtained.
Patients were ineligible for the study if they had major surgical procedures with resection or anastomosis within 3 weeks, or exploration or biopsy within 2 weeks. Patients who had prior radiotherapy to the primary area of measurable disease were excluded. Patients who had any chemotherapy or radiation therapy within the previous month were ineligible. Patients with any of the following conditions were excluded from the study: any active infection process or active and significant coexistent disease that in the judgment of the investigator would make the risk of chemotherapy prohibitive; severe malnutrition, severe nausea, or frequent vomiting; a concomitant neoplasm in other sites, except for nonmetastatic carcinoma of the skin or carcinoma-in-situ of the cervix; total disability (ECOG performance status 4). Estimated food intake should have been 30 calories/kg of body weight or greater each day. Patients with leucopenia (WBC < 4000/mm3) or thrombocytopenia (platelets < 150,000/mm3) at entry were ineligible. Patients were requested to start protocol therapy within 1 week of random assignment.
Patients were stratified by ECOG performance status (0, 1 v 2, 3), objective response indicators (measurable tumors versus biochemical marker, ie, a laboratory assay of functioning carcinoid), prior chemotherapy with study agents (none, DOX only, STZ only, FU, or both DOX and STZ), active heart disease (absent, present), and active renal disease (absent, present).
Patients previously untreated with chemotherapy and patients previously treated with somatostatin analogs only were randomly assigned to either FU/STZ or FU/DOX. Patients who had previous treatment with doxorubicin or who had active heart disease were directly assigned to treatment with FU/STZ if they had not been previously treated with FU or STZ and had no active renal disease. Patients who had previous treatment with STZ or who had active renal disease (creatinine > 1.5 mg, blood urea nitrogen (BUN) > 30, or persistent proteinuria) were directly assigned to treatment with FU/DOX if they had not been previously treated with FU or DOX and had no heart disease. Patients who had previous treatment with FU or both DOX and STZ, or who had active heart and renal disease, or who had STZ and active heart disease or DOX and renal disease were directly assigned to DTIC. Patients could be also directly assigned to treatment with FU/DOX if they had not been previously treated with FU or STZ or DOX, and there was no heart or renal disease present.
The FU/DOX therapy consisted of 40 mg/m2 of DOX (25 mg/m2 if jaundiced) on day 1 by rapid intravenous (IV) push every 5 weeks, and 400 mg/m2 of FU per day on days 1 to 5 by rapid IV push every 5 weeks. The FU/STZ therapy consisted of 500 mg/m2 of STZ on days 1 to 5 by rapid IV push every 10 weeks, and 400 mg/m2 of FU per day on days 1 to 5 and days 36 to 40 by rapid IV push every 10 weeks. The DTIC was given 250 mg/m2 per day on days 1 to 5 by rapid IV push every 4 weeks. The initial course of these treatments was reduced by 50% if the 24-hour urine 5-HIAA was 150 mg to prevent carcinoid crisis.
Patients were re-evaluated every 5 weeks for the combination and every 4 weeks for DTIC. If objective response or disease stabilization was observed, then the treatment was continued. DOX was discontinued at a total dose of 500 mg/m2 and FU was continued alone. If disease progression was observed, patients with initial combination therapy were switched to DTIC, and patients treated with DTIC initially were taken off study.
The study was designed mainly to evaluate tumor response and survival. The responses are defined as: complete response (CR), complete disappearance of all measurable and assessable disease lasting for at least 4 weeks with no new lesions; partial response (PR), 50% decrease of the sum of the products of perpendicular diameters of all measurable lesions with no new lesions. PFS was computed from random assignment to progression or death, whichever came first, or was censored at the time last known alive without progression. Survival was computed from random assignment to death or censored at the time last known alive.
The sample size calculations for those on the randomly assigned arms were based on the ability to accrue approximately 40 to 50 cases annually. Based on this, 4 years of accrual and 1 year of follow-up were planned. Regarding objective tumor response, a 100% or greater increase in response rates (from 20% to 40%) could be detected with at least 80% power under the assumption of a two-side binomial test with a .05 level of significance. Regarding survival, an exponential distribution was assumed. The study was initially designed to detect a 50% or greater increase in median survival (from 50 weeks to 75 weeks) with at least 80% power with a two-sided log-rank test with a .05 level of significance and 1 year of follow-up. However, the exact calculation was in error, with detection of at least 60% power when the study was designed initially. There was no provision for interim analyses.
The distribution of patient characteristics between treatment arms was evaluated by Fisher’s exact test for balance. The exact Wilcoxon mid-rank test for ordered categorical outcome was used to compare the distribution of maximum toxicity grade between the two arms and also to evaluate the distribution of the four age categories in the patient characteristics section. Fisher’s exact test was also used to analyze contingency tables of response. Survival curves were estimated by the method of Kaplan and Meier, with differences assessed by the log-rank test. Logistic regression of response and proportional hazards regression of survival to identify simultaneously significant prognostic covariates were based on the likelihood ratio test.
RESULTS
The study was activated in December 1981, and was closed on February 21, 1990, with 32 participating institutions. In all, 249 patients were enrolled (Fig 1); 176 patients were accrued to the randomized arms, with 88 patients on each arm. Among them, 163 patients (78 on the FU/STZ arm and 85 on the FU/DOX arm) were analyzable. Of 13 patients removed from study evaluation, four were cancelled before receiving the assigned treatment and nine were ineligible for the study.
In the direct assignment group, accrual was slow in the DTIC and FU/DOX arms and their accrual goals were not met, with 21 patients entering the DTIC arm, 25 patients the FU/DOX arm and 27 patients the FU/STZ arm. There were 70 patients who were crossed over to DTIC treatment after their disease progressed after FU/STZ or FU/DOX treatment, and 62 of these cases were analyzable. Of those nonassessable patients, four were cancelled before receiving DTIC, and four were ineligible for the study.
Patients Characteristics
The distribution of patient characteristics at the study were balanced for age, sex, performance status, presence of carcinoid syndrome, and liver metastasis in both the randomly assigned group and the directly assigned groups, except for more prior surgery in the patients treated with FU and DOX (P = .036; Tables 1, 2, and 3).
Toxicity
The treatment-related toxicities were evaluated based on treatment regimens (FU/DOX, FU/STZ, and DTIC) and with combining random and direct assignment arms (Tables 4, 5, 6, and 7). The degree of toxicity was classified as mild, moderate, severe, life-threatening, or lethal. Overall, there were four lethal toxicities; three on the FU/DOX arm and one on the FU/STZ arm. Renal toxicity was reported in 40 (34.8%) of 115 patients in the FU/STZ arm, most being mild to moderate in degree. Two patients developed life-threatening renal failure. The distribution of maximum grade toxicities for each arm was compared, and there were no significant differences in maximal overall toxicities between FU/DOX and FU/STZ, with DTIC exhibiting overall less adverse maximum grade toxicity (P = .0782).
Response
Overall, there were two CRs and 21 PRs in the randomly assigned group, with two CRs (2.4%) and nine PRs (13.5%) on the FU/DOX arm and 12 PRs on the FU/STZ arm (16%). These differences were not statistically significant (P = .82). There were 13 cases (15.3%) of disease stabilization (SD) in the FU/DOX arm and 12 (15.4%) SDs in the FU/STZ arm. In the direct assignment group, there were no CRs and 12 PRs (seven PRs for FU/DOX [28%], three PRs for FU/STZ [11.5%], and two PRs [9.5%] for DTIC). There were two CRs and three PRs in the 61 patients who received second-line DTIC (8.2%). There were seven patients who crossed over to DTIC and had stable disease (11.3%) after failing their initial treatment.
PFS
Overall median PFS on the randomized and direct groups was 4.7 and 4.8 months, respectively. In the randomized group, there was no statistical significant difference between FU/DOX and FU/STZ arms (4.5 v 5.3 months; P = .17). However, the substratified analysis based on performance status instead of treatment regimens showed the PFS significantly improved in those patients with relatively better performance status (P = .0013). The overall median PFS on the direct assignment arms was 4.8 months (7.3 months in the FU/DOX arm, 4.7 months in the FU/STZ arm, and 4.1 months in the DTIC arm).
Survival
Overall median survival for the randomized population was 18.4 months, with FU/STZ superior to FU/DOX (24.3 v 15.7 months; P = .0267; Fig 2). Based on the analysis of single covariates, the following factors may relate survival with treatment: performance status (0, 1 v 2, 3; P = .0001), age ( 60 v > 60; P = .05); objective indicator (biochemical indicator versus measurable disease; P = .0127).
The final model or a step-down proportional hazards regression of survival (keeping all variable with P < .1) showed performance status (P = .0001) and treatment with FU/STZ (P = .0099) as significant predictors for survival. Subsequent analysis with combining these two factors showed that those individual with performance of 0 or 1 and treated with FU/STZ had the longest survival (Fig 3).
Median survival for the direct assignment group was 13.2 months (18.4 months in the FU/DOX arm, 10.9 months in the FU/STZ arm, and 8.8 months in the DTIC arm). The median survival for second-line DTIC was 11.9 months.
DISCUSSION
Because of the heterogeneity of carcinoid tumors and the wide variations in natural history, therapeutic choices for patients with advanced diseases are greatly influenced by tumor burden, distribution, aggressiveness, and presence and severity of hormone-associated syndromes. The introduction of somatostatin analogs has improved the management of carcinoid syndrome. However, there is little evidence for tumor reduction, though there are anecdotal reports of cytostatic impact on tumor growth and possible survival benefit, and many physicians now use these agents in patients without syndrome but with advancing disease, in the hope of slowing or stabilizing the tumor.7-9 Although typical carcinoid tumors have been considered to be insensitive to classic cytotoxic chemotherapy, systemic chemotherapy has not been well or systematically studied and may be considered for selected patients, such as those with more rapidly growing tumors that have progressed on less toxic treatments.
A phase II study conducted by the Southwest Oncology Group (SWOG) evaluated the combination of the combination of FU, DOX, STZ, and cyclophosphamide in patients with metastatic carcinoid tumor and showed a 31% response rate with no survival benefit and high toxicity.10 The current study confirmed the activity of combinations of FU with DOX or STZ in patients with advanced carcinoid tumor, with no difference in response rate and PFS between the two combinations. On the basis of these measures, the efficacy of these regimens is modest, consistent with previous studies.3,4 In this study, performed before current supportive measures such as newer antiemetics and growth factors, toxicity was significant. However, despite a lack of observable difference in cytotoxicity between the two regimens in the randomized portion of the trial, a significant survival advantage emerged favoring the STZ-containing regimen. This supports the activity of this drug in this patient population, and suggests that surrogate markers of efficacy such as tumor response and PFS may not be adequate determinants of activity in certain gut neuroendocrine tumors, such as the carcinoid patients eligible for this trial. Stratification for variables such as degree of urine 5-HIAA elevation, serum chromogranin A elevation, and previous therapies including surgical debulking or ablation and hormonal treatment should be included in future trials for patients with advanced carcinoid tumors for precisely evaluation of the efficacy of new treatment regimens, especially novel agents.
DTIC as a single agent has some activity in those patients previously treated with FU/DOX or FU/STZ. A later phase II study showed that outcomes from the combination of FU, DTIC, and epirubicin (FED) were similar to those of FU/STZ and FU/DOX in previously untreated metastatic carcinoid tumor patients with response a rate of 30% and median response duration of 10 months.11 However, another trial conducted by the Italian Trials in Medical Oncology (ITMO) group with a similar regimen showed no more activity than any one of the single agents given alone.12
Recent trials with other newer cytotoxic agents have been disappointing. Only minimal activity has been shown with mitoxantrone and paclitaxel.13,14 Interferon (IFN) may have activity in the treatment of metastatic carcinoid tumor via antiproliferative effects, with decreasing 5-HIAA levels in the urine in the majority of patients treated with IFN-, but with tumor regression in only 15% to 20%.15,16 Adding IFN- to somatostatin analogs also appears effective in controlling the carcinoid syndrome refractory to somatostatin analogs alone.17 A recent phase II study result suggested that the treatment with IFN- in patients with metastatic carcinoid tumor was well tolerated, but without significant antitumor effects.18 Systemic chemotherapy, using regimens typically considered active in small cell carcinoma of the lung, may benefit selected patients with more aggressive, atypical histologic variants of carcinoid tumors. It has been reported that a response rate of 67% was achieved with a combination of etoposide and cisplatin in patients with anaplastic neuroendocrine carcinomas, while only 7% responded with well-differentiated carcinoid tumors.19 Combination of STZ with infusional FU has been shown to have a somewhat better toxicity profile in the treatment of neuroendocrine tumors.20
In summary, there is still no standard chemotherapy treatment for patients with metastatic carcinoid tumors of typical histologic grade. Selected patients, however, may both require systemic treatment for tumor bulk or uncontrolled symptoms, even when modern techniques such as somatostatin analogs, liver resection, embolization, and chemoembolization are utilized. New molecular targeting agents also need to be investigated for the treatment of advanced carcinoid tumor. For selected patients, the use of a STZ and fluoropyrimidine regimen would seem an appropriate choice for systemic chemotherapy, with the use of current supportive care measures to reduce the level of toxicity seen in the patients treated on this trial.
Authors’ Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by Public Health Service grants CA23318, CA66636, CA21115, CA15488, CA13650, and from the National Cancer Institute, National Institutes of Health, and the Department of Health and Human Services.
This articles’ contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Godwin JD II: Carcinoid tumors: An analysis of 2837 cases. Cancer 36: 560-569, 1975
Modlin IM, Lye KD, Kidd M: A 5-decade analysis of 13,715 carcinoid tumors. Cancer 97: 934-959, 2003
Moertel CG, Hanley JA: Combination chemotherapy trials in metastatic carcinoid tumor and the malignant carcinoid syndrome. Cancer Clin Trials 2: 327-334, 1979
Engstrom PF, Lavin PT, Moertel CG, et al: Streptozocin plus fluorouracil versus doxorubicin therapy for metastatic carcinoid tumor. J Clin Oncol 2: 1255-1259, 1984
Kessinger A, Foley JF, Lemon HM: Use of DTIC in the malignant carcinoid syndrome. Cancer Treat Rep 61: 101-102, 1977
van Hazel GA, Rubin J, Moertel CG: Treatment of metastatic carcinoid tumor with dactinomycin or dacarbazine. 67: 583-585, 1983
Di Bartolomeo M, Bajetta E, Buzzoni R, et al: Clinical efficacy of octreotide in the treatment of metastatic neuroendocrine tumors: A study by the Italian Trials in Medical Oncology group. Cancer 77: 402-408, 1996
Bousquet C, Puente E, Buscail L, et al: Antiproliferative effect of somatostatin and analogs. Chemotherapy 47: 30-39, 2001 (suppl 2)
Garland J, Buscombe JR, Bouvier C, et al: Sandostatin LAR (long-acting octreotide acetate) for malignant carcinoid syndrome: A 3-year experience. Aliment Pharmacol Ther 17: 437-444, 2003
Bukowski RM, Johnson KG, Peterson RF, et al: A phase II trial of combination chemotherapy in patients with metastatic carcinoid tumors: A Southwest Oncology Group study. Cancer 60: 2891-2895, 1987
Bajetta E, Rimassa L, Carnaghi C, et al: 5-fluororuacil, dacarbazine, and epirubicin in the treatment of patients with neuroendocrine tumors. Cancer 83: 372-378, 1998
Di Bartolomeo M, Bajetta E, Bochicchio AM, et al: A phase II trial of dacarbazine, fluorouracil and epirubicin in patients with neuroendocrine tumours: A study by the Italian Trials in Medical Oncology (I.T.M.O.) Group. Ann Oncol 6: 77-79, 1995
Neijt JP, Lacave AJ, Splinter TAW, et al: Mitoxantrone in metastatic apudomas: A phase II study of the EORTC gastro-Intestinal Cancer Cooperative Group. Br J Cancer 71: 106-108, 1995
Ansell SM, Pitot HC, Burch PA, et al: A phase II study of high-dose paclitaxel in patients with advanced neuroendocrine tumors. Cancer 91: 1543-1548, 2001
Oberg K, Eriksson B: The role of interferons in the management of carcinoid tumors. Acta Oncol 30: 519-522, 1991
Oberg K: Chemotherapy and biotherapy in the treatment of neuroencocrine tumors. Ann Oncol 12: S111-S114, 2001 (suppl 2)
Janson ET, Oberg K: Long-term management of the carcinoid syndrome: Treatment with octreotide alone and in combination with alpha-interferon. Acta Oncol 32: 225-229, 1993
Stuart K, Levy DE, Anderson T, et al: Phase II study of interferon gamma in malignant carcinoid tumor (E9292): A trial of the Eastern Cooperative Oncology Group. Invest New Drugs 22: 75-81, 2004
Moertel CG, Kvols LK, O’Connell MJ, et al: Treatment of neuroendocrine carcinomas with combined etoposide and cisplatin: Evidence of therapeutic activity in the anaplastic variants of these neoplasms. Cancer 68: 227-232, 1991
Gonzalez MA, Biswas S, Clifton L, et al: Treatment of neuroendocrine tumours with infusional 5-fluorouracil, folinic acid and streptozocin. Br J Cancer 89: 455-456, 2003(Weijing Sun, Stuart Lipsi)
Dana Farber Cancer Institute, Boston, MA
Creighton University, Omaha, NE
ABSTRACT
PURPOSE: Optimal treatments for metastatic carcinoid tumor remain undefined, and the role of chemotherapy for symptomatic patients with progressive disease is uncertain.
PATIENTS AND METHODS: Two hundred forty-nine patients with advanced carcinoid tumors were randomized to either doxorubicin with fluorouracil (FU/DOX) or streptozocin with fluorouracil (FU/STZ). Patients crossed over to the dacarbazine (DTIC) treatment after disease progression following first-line treatment (either FU/DOX or FU/STZ), and 73 patients were assigned to one of these three treatments based on their previous treatment or on abnormal baseline cardiac or renal function.
RESULTS: In the randomized group, there was no difference between FU/DOX and FU/STZ in response rates (15.9% v 16%) and progression-free survival (4.5 v 5.3 months). FU/STZ (24.3 months) was superior to FU/DOX (15.7 months; P = .0267) in median survival. The response rate of crossover DTIC treatment was 8.2%, with a median survival of 11.9 months. Hematologic toxicities were the major treatment-related toxicities for both FU/DOX and FU/STZ, and mild to moderate renal toxicity was reported in 40 (34.8%) of 115 patients in the FU/STZ arm.
CONCLUSION: Response to all three treatment regimens were modest. FU/STZ improved survival compared with the doxorubicin-based regimen, suggesting that the combination should be considered to be an active regimen of therapy when chemotherapy is judged to be an option for selected patients with carcinoid tumors.
INTRODUCTION
Carcinoid tumors are the most common neuroendocrine tumors of the gastrointestinal tract, accounting for up to one third of all small-intestinal malignancies.1,2 Because of the relatively indolent nature of the disease, many patients present with metastatic disease at the time of diagnosis, and two thirds of these patients develop carcinoid syndrome from the secretion of bioactive substances.2 Although the introduction of somatostatin analogs has improved the hormonally related carcinoid syndrome, the prognosis of patients with progressive unresectable metastases is fairly dismal, with significant symptoms or disability resulting from their malignant disease and poor overall outcome, and palliation of metastatic carcinoid tumor has continued to be a challenge to oncologists.
This Eastern Cooperative Oncology Group (ECOG) study was designed to evaluate and confirm the toxicities, response rate, progression-free survival (PFS), and overall survival benefit of fluorouracil with streptozocin (FU/STZ), compared with fluorouracil with doxorubicin (FU/DOX) in the treatment of selected patients with metastatic carcinoid tumor. The protocol was based the results of two ECOG trials (EST 3272 and EST 5375), which demonstrated objective response rates of 22% and 33%, respectively, in patients treated with FU/STZ and 21% in patients with doxorubicin alone.3,4 However, there was uncertainty in the palliative or survival benefit from these two trials, and gastrointestinal and hematologic toxicities plus STZ-associated renal toxicity were of concern. Dacarbazine (DTIC) has also been reported to have some effectiveness in metastatic carcinoid tumor.5,6 The secondary goal of this study, therefore, was to assess the activity of DTIC.
PATIENTS AND METHODS
All patients had histologically proven, unresectable advanced carcinoid tumors that were measurable or assessable by biochemical or radiographic measures. Because of the characteristically indolent nature of carcinoid tumors, patients were chosen for treatment only if they had significant symptoms or disability (unable to carry out usual work activities) resulting from their malignant disease by radiologic or biologic measures. Written informed consent conforming to federal and institutional guidelines was obtained.
Patients were ineligible for the study if they had major surgical procedures with resection or anastomosis within 3 weeks, or exploration or biopsy within 2 weeks. Patients who had prior radiotherapy to the primary area of measurable disease were excluded. Patients who had any chemotherapy or radiation therapy within the previous month were ineligible. Patients with any of the following conditions were excluded from the study: any active infection process or active and significant coexistent disease that in the judgment of the investigator would make the risk of chemotherapy prohibitive; severe malnutrition, severe nausea, or frequent vomiting; a concomitant neoplasm in other sites, except for nonmetastatic carcinoma of the skin or carcinoma-in-situ of the cervix; total disability (ECOG performance status 4). Estimated food intake should have been 30 calories/kg of body weight or greater each day. Patients with leucopenia (WBC < 4000/mm3) or thrombocytopenia (platelets < 150,000/mm3) at entry were ineligible. Patients were requested to start protocol therapy within 1 week of random assignment.
Patients were stratified by ECOG performance status (0, 1 v 2, 3), objective response indicators (measurable tumors versus biochemical marker, ie, a laboratory assay of functioning carcinoid), prior chemotherapy with study agents (none, DOX only, STZ only, FU, or both DOX and STZ), active heart disease (absent, present), and active renal disease (absent, present).
Patients previously untreated with chemotherapy and patients previously treated with somatostatin analogs only were randomly assigned to either FU/STZ or FU/DOX. Patients who had previous treatment with doxorubicin or who had active heart disease were directly assigned to treatment with FU/STZ if they had not been previously treated with FU or STZ and had no active renal disease. Patients who had previous treatment with STZ or who had active renal disease (creatinine > 1.5 mg, blood urea nitrogen (BUN) > 30, or persistent proteinuria) were directly assigned to treatment with FU/DOX if they had not been previously treated with FU or DOX and had no heart disease. Patients who had previous treatment with FU or both DOX and STZ, or who had active heart and renal disease, or who had STZ and active heart disease or DOX and renal disease were directly assigned to DTIC. Patients could be also directly assigned to treatment with FU/DOX if they had not been previously treated with FU or STZ or DOX, and there was no heart or renal disease present.
The FU/DOX therapy consisted of 40 mg/m2 of DOX (25 mg/m2 if jaundiced) on day 1 by rapid intravenous (IV) push every 5 weeks, and 400 mg/m2 of FU per day on days 1 to 5 by rapid IV push every 5 weeks. The FU/STZ therapy consisted of 500 mg/m2 of STZ on days 1 to 5 by rapid IV push every 10 weeks, and 400 mg/m2 of FU per day on days 1 to 5 and days 36 to 40 by rapid IV push every 10 weeks. The DTIC was given 250 mg/m2 per day on days 1 to 5 by rapid IV push every 4 weeks. The initial course of these treatments was reduced by 50% if the 24-hour urine 5-HIAA was 150 mg to prevent carcinoid crisis.
Patients were re-evaluated every 5 weeks for the combination and every 4 weeks for DTIC. If objective response or disease stabilization was observed, then the treatment was continued. DOX was discontinued at a total dose of 500 mg/m2 and FU was continued alone. If disease progression was observed, patients with initial combination therapy were switched to DTIC, and patients treated with DTIC initially were taken off study.
The study was designed mainly to evaluate tumor response and survival. The responses are defined as: complete response (CR), complete disappearance of all measurable and assessable disease lasting for at least 4 weeks with no new lesions; partial response (PR), 50% decrease of the sum of the products of perpendicular diameters of all measurable lesions with no new lesions. PFS was computed from random assignment to progression or death, whichever came first, or was censored at the time last known alive without progression. Survival was computed from random assignment to death or censored at the time last known alive.
The sample size calculations for those on the randomly assigned arms were based on the ability to accrue approximately 40 to 50 cases annually. Based on this, 4 years of accrual and 1 year of follow-up were planned. Regarding objective tumor response, a 100% or greater increase in response rates (from 20% to 40%) could be detected with at least 80% power under the assumption of a two-side binomial test with a .05 level of significance. Regarding survival, an exponential distribution was assumed. The study was initially designed to detect a 50% or greater increase in median survival (from 50 weeks to 75 weeks) with at least 80% power with a two-sided log-rank test with a .05 level of significance and 1 year of follow-up. However, the exact calculation was in error, with detection of at least 60% power when the study was designed initially. There was no provision for interim analyses.
The distribution of patient characteristics between treatment arms was evaluated by Fisher’s exact test for balance. The exact Wilcoxon mid-rank test for ordered categorical outcome was used to compare the distribution of maximum toxicity grade between the two arms and also to evaluate the distribution of the four age categories in the patient characteristics section. Fisher’s exact test was also used to analyze contingency tables of response. Survival curves were estimated by the method of Kaplan and Meier, with differences assessed by the log-rank test. Logistic regression of response and proportional hazards regression of survival to identify simultaneously significant prognostic covariates were based on the likelihood ratio test.
RESULTS
The study was activated in December 1981, and was closed on February 21, 1990, with 32 participating institutions. In all, 249 patients were enrolled (Fig 1); 176 patients were accrued to the randomized arms, with 88 patients on each arm. Among them, 163 patients (78 on the FU/STZ arm and 85 on the FU/DOX arm) were analyzable. Of 13 patients removed from study evaluation, four were cancelled before receiving the assigned treatment and nine were ineligible for the study.
In the direct assignment group, accrual was slow in the DTIC and FU/DOX arms and their accrual goals were not met, with 21 patients entering the DTIC arm, 25 patients the FU/DOX arm and 27 patients the FU/STZ arm. There were 70 patients who were crossed over to DTIC treatment after their disease progressed after FU/STZ or FU/DOX treatment, and 62 of these cases were analyzable. Of those nonassessable patients, four were cancelled before receiving DTIC, and four were ineligible for the study.
Patients Characteristics
The distribution of patient characteristics at the study were balanced for age, sex, performance status, presence of carcinoid syndrome, and liver metastasis in both the randomly assigned group and the directly assigned groups, except for more prior surgery in the patients treated with FU and DOX (P = .036; Tables 1, 2, and 3).
Toxicity
The treatment-related toxicities were evaluated based on treatment regimens (FU/DOX, FU/STZ, and DTIC) and with combining random and direct assignment arms (Tables 4, 5, 6, and 7). The degree of toxicity was classified as mild, moderate, severe, life-threatening, or lethal. Overall, there were four lethal toxicities; three on the FU/DOX arm and one on the FU/STZ arm. Renal toxicity was reported in 40 (34.8%) of 115 patients in the FU/STZ arm, most being mild to moderate in degree. Two patients developed life-threatening renal failure. The distribution of maximum grade toxicities for each arm was compared, and there were no significant differences in maximal overall toxicities between FU/DOX and FU/STZ, with DTIC exhibiting overall less adverse maximum grade toxicity (P = .0782).
Response
Overall, there were two CRs and 21 PRs in the randomly assigned group, with two CRs (2.4%) and nine PRs (13.5%) on the FU/DOX arm and 12 PRs on the FU/STZ arm (16%). These differences were not statistically significant (P = .82). There were 13 cases (15.3%) of disease stabilization (SD) in the FU/DOX arm and 12 (15.4%) SDs in the FU/STZ arm. In the direct assignment group, there were no CRs and 12 PRs (seven PRs for FU/DOX [28%], three PRs for FU/STZ [11.5%], and two PRs [9.5%] for DTIC). There were two CRs and three PRs in the 61 patients who received second-line DTIC (8.2%). There were seven patients who crossed over to DTIC and had stable disease (11.3%) after failing their initial treatment.
PFS
Overall median PFS on the randomized and direct groups was 4.7 and 4.8 months, respectively. In the randomized group, there was no statistical significant difference between FU/DOX and FU/STZ arms (4.5 v 5.3 months; P = .17). However, the substratified analysis based on performance status instead of treatment regimens showed the PFS significantly improved in those patients with relatively better performance status (P = .0013). The overall median PFS on the direct assignment arms was 4.8 months (7.3 months in the FU/DOX arm, 4.7 months in the FU/STZ arm, and 4.1 months in the DTIC arm).
Survival
Overall median survival for the randomized population was 18.4 months, with FU/STZ superior to FU/DOX (24.3 v 15.7 months; P = .0267; Fig 2). Based on the analysis of single covariates, the following factors may relate survival with treatment: performance status (0, 1 v 2, 3; P = .0001), age ( 60 v > 60; P = .05); objective indicator (biochemical indicator versus measurable disease; P = .0127).
The final model or a step-down proportional hazards regression of survival (keeping all variable with P < .1) showed performance status (P = .0001) and treatment with FU/STZ (P = .0099) as significant predictors for survival. Subsequent analysis with combining these two factors showed that those individual with performance of 0 or 1 and treated with FU/STZ had the longest survival (Fig 3).
Median survival for the direct assignment group was 13.2 months (18.4 months in the FU/DOX arm, 10.9 months in the FU/STZ arm, and 8.8 months in the DTIC arm). The median survival for second-line DTIC was 11.9 months.
DISCUSSION
Because of the heterogeneity of carcinoid tumors and the wide variations in natural history, therapeutic choices for patients with advanced diseases are greatly influenced by tumor burden, distribution, aggressiveness, and presence and severity of hormone-associated syndromes. The introduction of somatostatin analogs has improved the management of carcinoid syndrome. However, there is little evidence for tumor reduction, though there are anecdotal reports of cytostatic impact on tumor growth and possible survival benefit, and many physicians now use these agents in patients without syndrome but with advancing disease, in the hope of slowing or stabilizing the tumor.7-9 Although typical carcinoid tumors have been considered to be insensitive to classic cytotoxic chemotherapy, systemic chemotherapy has not been well or systematically studied and may be considered for selected patients, such as those with more rapidly growing tumors that have progressed on less toxic treatments.
A phase II study conducted by the Southwest Oncology Group (SWOG) evaluated the combination of the combination of FU, DOX, STZ, and cyclophosphamide in patients with metastatic carcinoid tumor and showed a 31% response rate with no survival benefit and high toxicity.10 The current study confirmed the activity of combinations of FU with DOX or STZ in patients with advanced carcinoid tumor, with no difference in response rate and PFS between the two combinations. On the basis of these measures, the efficacy of these regimens is modest, consistent with previous studies.3,4 In this study, performed before current supportive measures such as newer antiemetics and growth factors, toxicity was significant. However, despite a lack of observable difference in cytotoxicity between the two regimens in the randomized portion of the trial, a significant survival advantage emerged favoring the STZ-containing regimen. This supports the activity of this drug in this patient population, and suggests that surrogate markers of efficacy such as tumor response and PFS may not be adequate determinants of activity in certain gut neuroendocrine tumors, such as the carcinoid patients eligible for this trial. Stratification for variables such as degree of urine 5-HIAA elevation, serum chromogranin A elevation, and previous therapies including surgical debulking or ablation and hormonal treatment should be included in future trials for patients with advanced carcinoid tumors for precisely evaluation of the efficacy of new treatment regimens, especially novel agents.
DTIC as a single agent has some activity in those patients previously treated with FU/DOX or FU/STZ. A later phase II study showed that outcomes from the combination of FU, DTIC, and epirubicin (FED) were similar to those of FU/STZ and FU/DOX in previously untreated metastatic carcinoid tumor patients with response a rate of 30% and median response duration of 10 months.11 However, another trial conducted by the Italian Trials in Medical Oncology (ITMO) group with a similar regimen showed no more activity than any one of the single agents given alone.12
Recent trials with other newer cytotoxic agents have been disappointing. Only minimal activity has been shown with mitoxantrone and paclitaxel.13,14 Interferon (IFN) may have activity in the treatment of metastatic carcinoid tumor via antiproliferative effects, with decreasing 5-HIAA levels in the urine in the majority of patients treated with IFN-, but with tumor regression in only 15% to 20%.15,16 Adding IFN- to somatostatin analogs also appears effective in controlling the carcinoid syndrome refractory to somatostatin analogs alone.17 A recent phase II study result suggested that the treatment with IFN- in patients with metastatic carcinoid tumor was well tolerated, but without significant antitumor effects.18 Systemic chemotherapy, using regimens typically considered active in small cell carcinoma of the lung, may benefit selected patients with more aggressive, atypical histologic variants of carcinoid tumors. It has been reported that a response rate of 67% was achieved with a combination of etoposide and cisplatin in patients with anaplastic neuroendocrine carcinomas, while only 7% responded with well-differentiated carcinoid tumors.19 Combination of STZ with infusional FU has been shown to have a somewhat better toxicity profile in the treatment of neuroendocrine tumors.20
In summary, there is still no standard chemotherapy treatment for patients with metastatic carcinoid tumors of typical histologic grade. Selected patients, however, may both require systemic treatment for tumor bulk or uncontrolled symptoms, even when modern techniques such as somatostatin analogs, liver resection, embolization, and chemoembolization are utilized. New molecular targeting agents also need to be investigated for the treatment of advanced carcinoid tumor. For selected patients, the use of a STZ and fluoropyrimidine regimen would seem an appropriate choice for systemic chemotherapy, with the use of current supportive care measures to reduce the level of toxicity seen in the patients treated on this trial.
Authors’ Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by Public Health Service grants CA23318, CA66636, CA21115, CA15488, CA13650, and from the National Cancer Institute, National Institutes of Health, and the Department of Health and Human Services.
This articles’ contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Godwin JD II: Carcinoid tumors: An analysis of 2837 cases. Cancer 36: 560-569, 1975
Modlin IM, Lye KD, Kidd M: A 5-decade analysis of 13,715 carcinoid tumors. Cancer 97: 934-959, 2003
Moertel CG, Hanley JA: Combination chemotherapy trials in metastatic carcinoid tumor and the malignant carcinoid syndrome. Cancer Clin Trials 2: 327-334, 1979
Engstrom PF, Lavin PT, Moertel CG, et al: Streptozocin plus fluorouracil versus doxorubicin therapy for metastatic carcinoid tumor. J Clin Oncol 2: 1255-1259, 1984
Kessinger A, Foley JF, Lemon HM: Use of DTIC in the malignant carcinoid syndrome. Cancer Treat Rep 61: 101-102, 1977
van Hazel GA, Rubin J, Moertel CG: Treatment of metastatic carcinoid tumor with dactinomycin or dacarbazine. 67: 583-585, 1983
Di Bartolomeo M, Bajetta E, Buzzoni R, et al: Clinical efficacy of octreotide in the treatment of metastatic neuroendocrine tumors: A study by the Italian Trials in Medical Oncology group. Cancer 77: 402-408, 1996
Bousquet C, Puente E, Buscail L, et al: Antiproliferative effect of somatostatin and analogs. Chemotherapy 47: 30-39, 2001 (suppl 2)
Garland J, Buscombe JR, Bouvier C, et al: Sandostatin LAR (long-acting octreotide acetate) for malignant carcinoid syndrome: A 3-year experience. Aliment Pharmacol Ther 17: 437-444, 2003
Bukowski RM, Johnson KG, Peterson RF, et al: A phase II trial of combination chemotherapy in patients with metastatic carcinoid tumors: A Southwest Oncology Group study. Cancer 60: 2891-2895, 1987
Bajetta E, Rimassa L, Carnaghi C, et al: 5-fluororuacil, dacarbazine, and epirubicin in the treatment of patients with neuroendocrine tumors. Cancer 83: 372-378, 1998
Di Bartolomeo M, Bajetta E, Bochicchio AM, et al: A phase II trial of dacarbazine, fluorouracil and epirubicin in patients with neuroendocrine tumours: A study by the Italian Trials in Medical Oncology (I.T.M.O.) Group. Ann Oncol 6: 77-79, 1995
Neijt JP, Lacave AJ, Splinter TAW, et al: Mitoxantrone in metastatic apudomas: A phase II study of the EORTC gastro-Intestinal Cancer Cooperative Group. Br J Cancer 71: 106-108, 1995
Ansell SM, Pitot HC, Burch PA, et al: A phase II study of high-dose paclitaxel in patients with advanced neuroendocrine tumors. Cancer 91: 1543-1548, 2001
Oberg K, Eriksson B: The role of interferons in the management of carcinoid tumors. Acta Oncol 30: 519-522, 1991
Oberg K: Chemotherapy and biotherapy in the treatment of neuroencocrine tumors. Ann Oncol 12: S111-S114, 2001 (suppl 2)
Janson ET, Oberg K: Long-term management of the carcinoid syndrome: Treatment with octreotide alone and in combination with alpha-interferon. Acta Oncol 32: 225-229, 1993
Stuart K, Levy DE, Anderson T, et al: Phase II study of interferon gamma in malignant carcinoid tumor (E9292): A trial of the Eastern Cooperative Oncology Group. Invest New Drugs 22: 75-81, 2004
Moertel CG, Kvols LK, O’Connell MJ, et al: Treatment of neuroendocrine carcinomas with combined etoposide and cisplatin: Evidence of therapeutic activity in the anaplastic variants of these neoplasms. Cancer 68: 227-232, 1991
Gonzalez MA, Biswas S, Clifton L, et al: Treatment of neuroendocrine tumours with infusional 5-fluorouracil, folinic acid and streptozocin. Br J Cancer 89: 455-456, 2003(Weijing Sun, Stuart Lipsi)