Fax: (615) 342-1745
Combination chemotherapy with gemcitabine and irinotecan in patients with previously treated carcinoma of an unknown primary site†
A Minnie Pearl Cancer Research Network Phase II trial
Article first published online: 29 AUG 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 9, pages 1992–1997, 1 November 2005
How to Cite
Hainsworth, J. D., Spigel, D. R., Raefsky, E. L., Kuzur, M. E., Yost, K., Kommor, M., Litchy, S. and Greco, F. A. (2005), Combination chemotherapy with gemcitabine and irinotecan in patients with previously treated carcinoma of an unknown primary site. Cancer, 104: 1992–1997. doi: 10.1002/cncr.21416
The following institutions are Minnie Pearl Cancer Research Network Participating Sites: Tennessee Oncology, PLLC (Nashville, TN); Wellstar Cancer Research (Marietta, GA); Consultants in Blood Disorders & Cancer (Louisville, KY); Graves-Gilbert Clinic (Bowling Green, KY); Grand Rapids CCOP (Grand Rapids, MI); Upstate Carolina CCOP (Spartanburg, SC); Northeast Georgia Medical Center (Gainesville, GA); King's Daughters' Medical Center (Ashland, KY); Pheobe Cancer Center (Albany, GA); The Reading Hospital and Medical Center (Reading, PA); and Montana Cancer Specialists (Missoula, MT).
- Issue published online: 17 OCT 2005
- Article first published online: 29 AUG 2005
- Manuscript Accepted: 3 JUN 2005
- Manuscript Revised: 20 MAY 2005
- Manuscript Received: 2 MAR 2005
- Eli Lilly, Inc.
- Minnie Pearl Cancer Foundation
- combination chemotherapy;
- unknown primary site;
- Phase II trial
The current study was performed to evaluate the activity of combination chemotherapy with gemcitabine and irinotecan in patients with previously treated carcinoma of an unknown primary site.
Patients with carcinoma of an unknown primary site who had received one previous chemotherapy regimen were eligible for this study. All patients received gemcitabine at a dose of 1000mg/m2 intravenously (i.v.) and irinotecan at a dose of 100 mg/m2 i.v. on Days 1 and 8; treatment courses were repeated every 21 days. Patients were evaluated for response after completing two courses of treatment; responders/stable patients continued treatment for a recommended six courses.
Forty patients entered this multicenter, community-based Phase II trial between September 2000 and July 2003. Four of these 40 patients (10%) achieved objective responses (a partial response in 3 patients and a complete response in 1 patient). An additional 17 patients (43%) had stable disease/minor response at first reevaluation; 7 of these patients (18%) remained stable for longer than 6 months. The median survival for the entire group was 4.5 months, with 1-year and 2-year survival rates of 25% and 13%, respectively. The treatment was well tolerated by most patients. Neutropenia was the most common Grade 3/4 toxicity (according to the National Cancer Institute Common Toxicity Criteria, version 3.0) (occurring in 36% of patients). Myelosuppression-related complications were uncommon, as were severe nonhematologic toxicities.
The combination of gemcitabine and irinotecan has modest activity and is well tolerated in patients with recurrent/refractory carcinoma of an unknown primary site. Treatment-related toxicity, particularly myelosuppression, appears to be less severe than toxicity produced by the taxane and platinum regimens frequently used in the first-line therapy of these patients. Evaluation of the gemcitabine and irinotecan combination as first-line therapy is indicated. Cancer 2005. © 2005 American Cancer Society.
Carcinoma of an unknown primary site is a common clinical entity, accounting for approximately 3% of all cancer diagnoses. Within this heterogeneous group, a number of treatable subsets of patients have been recognized based on specific clinical and pathologic features. Specific therapies have been defined for patients in these subsets, and excellent treatment results are reportedly achieved in some of these patients. Unfortunately, the majority of patients do not fit into any identified treatable subset, and require empiric treatment with combination chemotherapy.
The empiric first-line therapy of patients with carcinoma of an unknown primary site has improved with the introduction of the taxanes and other new chemotherapeutic agents. Platinum agents, in combination with taxanes, gemcitabine, or irinotecan, have produced response rates of 25–45%, with 1-year and 2-year survival rates of approximately 45% and 20%, respectively.1–5 However, complete responses with first-line regimens are uncommon, and the large majority of patients eventually require additional therapy.
To our knowledge, effective second-line therapy for patients with carcinoma of an unknown primary site has not been identified to date. We previously evaluated single-agent gemcitabine in this setting.6 Although the objective response rate was low (8%), an additional 25% of patients achieved stable disease with improved symptoms. Several of the newer chemotherapeutic agents, such as irinotecan and vinorelbine, have a broad spectrum of activity but have not been evaluated specifically in patients with carcinoma of an unknown primary site.
The combination of gemcitabine and irinotecan has proven to be efficacious and well tolerated in patients with advanced pancreatic carcinoma.7 In this difficult group of patients, objective responses were observed in 20% of patients, and approximately a third of patients had decreases in their serum CA 19-9 levels of greater than 50%. Recently, we evaluated this combination as a component of first-line treatment for patients with carcinoma of an unknown primary site.8 Patients initially received two courses of combination paclitaxel, carboplatin and etoposide, and then received two courses of combination gemcitabine and irinotecan. In this study, the gemcitabine and irinotecan regimen demonstrated independent activity, producing objective responses in 15% of patients who did not respond to the combination of paclitaxel, carboplatin, and etoposide. In the Phase II trial reported in the current study, we evaluated the feasibility, toxicity, and efficacy of the combination of gemcitabine and irinotecan in patients with previously treated carcinoma of an unknown primary site.
MATERIALS AND METHODS
This nonrandomized Phase II trial was initiated in September 2000 and was performed in the Minnie Pearl Cancer Research Network, a multicenter, community-based clinical trials group.
Patients who were eligible for this trial had histologically confirmed carcinoma of an unknown primary site. For the purposes of the current study, patients were considered to have carcinoma of an unknown primary site if no primary site was evident after an evaluation including history, physical examination, chemistry profile, computed tomography (CT) scan of the chest and abdomen, and directed radiologic workup of symptomatic areas. The following light microscopic histologies were eligible: adenocarcinoma, poorly differentiated adenocarcinoma, poorly differentiated carcinoma, poorly differentiated neuroendocrine carcinoma, and poorly differentiated squamous carcinoma. Patients were required to have received one previous chemotherapy regimen for carcinoma of an unknown primary site. Previous therapy could not have included treatment with either gemcitabine or irinotecan, unless the patient had previously achieved either an objective response or stable disease to a regimen containing 1 of these agents, and had been off therapy for an interval of greater than 3 months. Additional eligibility criteria included measurable disease, a leukocyte count of ≥ 4000/μL and a platelet count of ≥ 100,000/μL, and normal liver and kidney function. Patients with parenchymal brain metastases were eligible only if brain metastases had been treated previously with surgical resection and/or radiation therapy and were inactive at the time of study enrollment. Patients with meningeal involvement were ineligible. All patients were required to provide written informed consent prior to entering this clinical trial. This trial was approved by the Institutional Review Board of Centennial Medical Center (Nashville, Tennessee) and by the institutional review boards of all participating Network sites.
Prior to entering this clinical trial, all patients were required to have a thorough history and physical examination, complete blood counts, serum chemistry profile, and CT scans of the chest and abdomen. In addition, specific radiologic or endoscopic evaluation of symptomatic areas was recommended. All patients with poorly differentiated carcinoma had specialized immunoperoxidase staining performed to rule out other poorly differentiated neoplasms with specific treatment requirements (e.g., lymphoma). Men with metastatic adenocarcinoma underwent measurement of the serum prostate-specific antigen (PSA) level to rule out metastatic prostate carcinoma. Young men with poorly differentiated carcinoma had serum determinations of human chorionic gonadotropin and α-fetoprotein.
All patients received treatment with gemcitabine at a dose of 1000 mg/m2 intravenously (i.v.) and irinotecan at a dose of 100 mg/m2 i.v., both administered on Days 1 and 8. Treatment courses were repeated at 21-day intervals. All patients received prophylactic antiemetic therapy with a 5HT3 receptor antagonist plus dexamethasone at a dose of 20 mg either orally or i.v.
Patients were reevaluated for response after completing two courses of therapy. Patients who had evidence of objective response or stable disease with improved symptoms continued therapy at 21-day intervals. For responding patients, the recommended number of treatment courses was six. However, patients whose disease remained in remission and who tolerated treatment were allowed to continue treatment beyond six courses, at the discretion of the treating physician.
Dose modifications of both chemotherapeutic agents were based on hematologic and nonhematologic toxicities. Dose reductions for myelosuppression were based on blood counts measured on the day of scheduled treatment. Nadir blood counts were not used to determine dose reductions. Patients who had a leukocyte count of ≥ 3000/μL and a platelet count of ≥ 100,000/μL received full doses of both agents. If the leukocyte count was 2000–3000/μL or the platelet count was 75,000–100,000/μL, 75% doses of both agents were administered. For a leukocyte count of < 2000/μL or a platelet count of < 75,000/μL, treatment was withheld and blood counts were repeated in 1 week. Treatment was resumed at 75% doses when the blood counts had risen to a leukocyte count of of ≥ 3000/μL and a platelet count of ≥ 100,000/μL. If the Day 8 dose of any treatment course was delayed because of low counts, blood counts were repeated on Day 15; if they had recovered sufficiently, the Day 8 treatment was given on Day 15 (at a 75% dose). If this treatment modification occurred, the following treatment course was not delayed, and began 1 week later.
Patients who were hospitalized for the treatment of fever associated with neutropenia received 75% doses of both drugs for all subsequent courses. Cytokines (granulocyte–colony-stimulating factor or granulocyte-macrophage–colony-stimulating factor) were not used during the first course of treatment. However, cytokines could be used at the discretion of the treating physician and according to standard guidelines during subsequent cycles. Treatment with cytokines could not substitute for the dose reductions outlined earlier.
Patients who experienced irinotecan-related Grade 1 or Grade 2 diarrhea were treated using standard recommendations. Irinotecan doses were not reduced, but the patients immediately began prophylactic treatment with loperamide (at a dose of 4 mg at the first onset of diarrhea, then 2 mg every 2 hours around the clock until the patient was free of diarrhea for at least 12 hours). If patients developed Grade 3 or 4 diarrhea, loperamide prophylaxis was initiated immediately and further doses of irinotecan were held until diarrhea improved to Grade 1 or less. Subsequently, irinotecan was administered at a 75% dose level.
Patients who experienced other reversible Grade 3 or Grade 4 nonhematologic toxicity (with the exception of nausea, emesis, or alopecia) had doses of the offending agent or agents withheld until the toxicity improved to less than Grade 2. Subsequent doses of the offending agent were administered at a 75% dose level. Patients who developed irreversible nonhematologic toxicity or toxicity that required more than 3 weeks to resolve were removed from study treatment.
Definition of Response
All patients were assigned a response category based on World Health Organization response criteria. Complete response required the total disappearance of clinically and radiologically detectable disease for at least 4 weeks. A partial response required a reduction by at least 50% in the size of all measurable lesions, as determined by the product of the greatest dimensions, with no new lesions appearing for at least 4 weeks. Patients with stable disease had a reduction of less than 50%, or an increase of greater than 25% in the size of measurable lesions, as measured by the products of the greatest dimensions, with no new lesions appearing. Patients with progressive disease had increase of at least 25% in the size of existing lesions or the appearance of new lesions.
This nonrandomized, Phase II study was performed for the purposes of determining the feasibility, toxicity, and efficacy of the gemcitabine and irinotecan regimen in patients with recurrent/refractory carcinoma of an unknown primary site. To our knowledge, no “standard” treatment has been defined in this patient population to date,. In a previous Phase II trial, single-agent gemcitabine produced an 8% response rate with temporary symptom improvement in an additional 25% of patients.6 Therefore, evidence of at least this level of activity was expected in this Phase II trial to justify further development of the combination of gemcitabine and irinotecan. To achieve an accurate assessment of the response rate to this treatment regimen, 40 patients were enrolled in this clinical trial.
All patients who received at least one dose of treatment were assessed for treatment-related toxicity. Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria (version 3.0). Progression-free survival was calculated from the date of the first treatment with this regimen to the date of carcinoma progression. Overall survival was calculated from the date of the first treatment with this regimen to the date of death. Actuarial survival curves were constructed using the method of Kaplan and Meier.9
Between September 2000 and July 2003, 40 patients were enrolled on this clinical trial and were treated with gemcitabine and irinotecan. The patient characteristics are summarized in Table 1. The median age was 54 years (range, 24–77 yrs); 90% of patients had an Eastern Cooperative Oncology Group performance status of 1. Thirty-four patients (85%) had received first-line treatment with regimens containing a taxane and a platinum agent. Only 9 patients (23%) had previous major responses to treatment.
|Characteristic||No. of patients (%)|
|Median age in yrs (range)||54 (24–77)|
|Poorly differentiated adenocarcinoma||10 (25)|
|Poorly differentiated carcinoma||7 (17)|
|Poorly differentiated neuroendocrine carcinoma||6 (15)|
|Poorly differentiated squamous carcinoma||2 (5)|
|ECOG performance status|
|Paclitaxel, carboplatin, and etoposide||14 (35)|
|Paclitaxel and carboplatin||11 (27.5)|
|Paclitaxel, carboplatin, and gemcitabine||8 (20)|
|Docetaxel, cisplatin, and etoposide||1 (2.5)|
|Carboplatin and etoposide||2 (5)|
|Cisplatin, etoposide, and topotecan||1 (2.5)|
|Cisplatin, bleomycin, and etoposide||1 (2.5)|
|Irinotecan, 5-FU, and leucovorin||1 (2.5)|
|Best response to previous therapy|
|Stable disease/progression||31 (77)|
|Site of treatment|
|Sarah Cannon Research Institute||25 (63)|
|Network sites||15 (37)|
The median number of courses of the gemcitabine and irinotecan regimen received by patients on this trial was four (range, one to six courses). Thirty-six patients (90%) received at least 2 courses of treatment and were evaluable for response. Four patients were removed from treatment prior to completing two courses for the following reasons: rapid tumor progression in two patients, treatment-related toxicity in one patient, and primary tumor site identified (in the prostate) in one patient. A total of 30 patients were removed from treatment before completing 6 courses of therapy. Twenty-six of these 30 patients were removed from treatment because of carcinoma progression, 1 patient was removed because of toxicity, and 3 patients were removed for other reasons (poor compliance in 1 patient, intercurrent illness in 1 patient, and by patient request in 1 patient).
During the first 2 courses of therapy, 14 patients received 100% of the scheduled doses of both chemotherapeutic agents. Reduction in the Day 8 dose due to myelosuppression was the most common reason for dose reduction. For the 36 patients who completed 2 courses of treatment, the percentage of planned chemotherapy administered was 86% for irinotecan and 86% for gemcitabine.
Four of 40 patients entering this trial (10%) achieved objective responses to treatment with the regimen of gemcitabine and irinotecan (a partial response in 3 patients and a complete response in 1 patient). These four patients had diverse clinical presentations, with varying histologies (adenocarcinoma in one patient, poorly differentiated adenocarcinoma in two patients, and poorly differentiated carcinoma in one patient) and one or more sites of visceral metastases (the liver in two patients, the lung in one patient, bones in one patient, and the peritoneum in one patient). All four of these patients had received previous treatment with the combination of paclitaxel, carboplatin, and gemcitabine. Two of these patients had also achieved a partial response with first-line chemotherapy, whereas 2 patients had had stable disease for more than 6 months. The 4 patients with objective responses had response durations of 3 months, 8 months, 11 months, and 48 months, respectively.
Seventeen additional patients (43%) had either stable disease or a minor response (i.e., a measurable decrease in the size of indicator lesions, but not meeting the criteria for a partial response) at the time of the first reevaluation; in 7 of these patients (18%), stable disease persisted for at least 6 months while they were receiving therapy. Seventeen patients (43%) had progressive disease at the time of the first reevaluation (or before) and were removed from treatment.
Figure 1 shows the actuarial overall survival for the entire group. The median survival, from the time of entry into this trial, was 4.5 months (95% confidence interval [95% CI], 4.2–7.0 mos), with 1-year and 2-year survival rates of 25% and 13%, respectively.
The treatment-related toxicity experienced by patients in this trial was similar to that previously reported with this gemcitabine and irinotecan regimen,7 and is detailed in Table 2. Myelosuppression was the most common toxicity, with Grade 3/4 neutropenia experienced by 13 patients (33%). Grade 3/4 thrombocytopenia occurred in 13 patients (33%). However, only three patients required hospitalization for the treatment of neutropenia and fever, and three patients required platelet transfusion. There were no episodes of bleeding. Ten patients required 1 or more red blood cell transfusions during the course of treatment.
|Toxicity||No. of patients (%)|
|Grade 3||Grade 4|
|Neutropenia||12 (30)||1 (3)|
|Hospitalization, neutropenia/fever||1 (3)|
|Platelet transfusion||3 (8)|
|Fatigue||4 (10)||1 (3)|
|Infection (non-neutropenic)||3 (8)||0|
Severe nonhematologic toxicity was uncommon in this clinical trial. Three patients (8%) experienced Grade 3/4 diarrhea associated with irinotecan, 2 of whom required hospitalization for management of this problem. There was one treatment-related death reported during this study (severe diarrhea followed by sepsis).
To the best of our knowledge, standard treatment has not been defined for patients with carcinoma of an unknown primary site who have recurrent or refractory disease after first-line empiric chemotherapy. To our knowledge, few prospective clinical trials have been reported to date in this patient population, and response rates have always been less than 15% with either single agents or combination regimens.6, 10 Because gemcitabine had previously demonstrated some single-agent activity in patients with recurrent/refractory carcinoma of an unknown primary site, and irinotecan has activity against adenocarcinomas from a variety of primary tumor sites, we evaluated the combination of gemcitabine and irinotecan in this setting.
In this multicenter Phase II trial, we demonstrated an objective response rate of 10% in patients with recurrent/refractory carcinoma of an unknown primary site who were treated with the combination of gemcitabine and irinotecan. Although this objective response rate was somewhat disappointing, an additional 45% of patients had stable disease or a minor response at the time of the first reevaluation. Seven of these patients (18%) with stable disease/minor response remained free of disease progression for longer than 6 months. Therefore, a total of 28% of patients derived substantial benefit from this treatment. In addition, the treatment-related toxicity with the gemcitabine and irinotecan regimen was reported to be mild to moderate in the majority of patients. Although dose reductions were relatively frequent, usually for the Day 8 dose, the myelosuppression was manageable and predictable in most patients; we therefore recommend starting with the doses used in this regimen, and reducing doses if necessary. There were only a few hospitalizations for treatment-related toxicity reported, and only 3 patients (8%) had treatment discontinued due to toxicity. In particular, Grade 3/4 irinotecan-related diarrhea was uncommon in this trial, most likely because of the dose and schedule used in this regimen.
Because of the heterogeneous nature of these patients, retrospective comparisons of Phase II trials are difficult. However, this combination regimen appeared to be more active than single-agent gemcitabine in terms of the percentage of patients with clinical improvement (55% vs. 33%) and overall survival (median 4.5 mo vs. 3 mos).6 Although the Grade 3 myelosuppression was found to be increased with the combination regimen, Grade 4 toxicity was uncommon with either regimen. It is our clinical impression, based on these considerations, that the combination of gemcitabine and irinotecan is more active in this setting than gemcitabine alone. However, we believe that a definitive answer to this question is not possible based on existing data.
Results in the current Phase II trial with the combination of gemcitabine and irinotecan are consistent with the results we previously obtained in a first-line trial in patients with carcinoma of an unknown primary site. In that trial, patients received sequential regimens, beginning with two courses of the paclitaxel, carboplatin, and etoposide combination, followed by two courses of the gemcitabine and irinotecan regimen.8 Approximately 10% of the patients who did not respond to two courses of the paclitaxel, carboplatin, and etoposide regimen achieved objective responses to the combination of gemcitabine and irinotecan. In addition, the gemcitabine and irinotecan regimen was associated with substantially less toxicity than the paclitaxel, carboplatin, and etoposide combination. Specific toxicities experienced less often with the gemcitabine and irinotecan regimen included myelosuppression and its complications, nausea/emesis, alopecia, and peripheral neuropathy. Based on these observations, we have initiated a randomized, prospective Phase III trial comparing the efficacy and toxicity of the paclitaxel, carboplatin, and etoposide regimen versus the combination of gemcitabine and irinotecan in the first-line treatment of patients with carcinoma of an unknown primary site.
The combination of gemcitabine and irinotecan has modest activity and appears to be well tolerated in patients with recurrent/refractory carcinoma of an unknown primary site. When patients with prolonged stable disease are considered along with the major responders, the results with this combination compare favorably with our previous experience with single-agent gemcitabine and with other single agents or combinations previously evaluated in the second-line treatment of this patient group. However, more active treatments are necessary for these patients, and ongoing clinical trials should focus on the evaluation of novel agents.
- 4A safety and efficacy trial of docetaxel and cisplatin in patients with cancer of unknown primary [abstract 2597]. Proc Am Soc Clin Oncol. 2003; 22: 646., , , et al.
- 5Cisplatin in combination with either gemcitabine or irinotecan in carcinomas of unknown primary site: results of a randomized phase II study—trial for the French Study Group on Carcinomas of Unknown Primary (GEFCAPI 01). J Clin Oncol. 2003; 21: 3479–3482., , , et al.