Beyond the usual regimens based on streptozocin and doxorubicin or 5-fluorouracil, no second-line therapy of metastatic neuroendocrine tumor has gained wide acceptance. Gemcitabine and oxaliplatin are generally well tolerated and have shown activity against a wide range of malignancies. The authors assessed the efficacy of gemcitabine-oxaliplatin combination (GEMOX) in the treatment of patients with metastatic neuroendocrine tumors.
Twenty consecutive patients with progressive disease were treated with GEMOX, in most cases after failure of other chemotherapy regimens (median = 2). Patients were followed for evidence of toxicity, response, and survival. Two patients were chemotherapy-naive at treatment initiation and were excluded from the efficacy analysis.
Toxicity was manageable overall; however, 6 (30%) patients had to discontinue treatment because of oxaliplatin-induced neurotoxicity (grade 2). Three (17%) of 18 patients had a partial response, median progression-free survival was 7.0 months, and median overall survival was 23.4 months.
Neuroendocrine tumors (NETs) comprise a heterogeneous group of rare malignancies. These tumors can be classified according to their organ of origin as well as according to their degree of differentiation.1 Well-differentiated endocrine carcinomas are considered low-grade malignancies, but 20% to 50% of patients present with metastases at the time of diagnosis.2, 3 Once metastases are present, patients are usually not curable, but may live for long periods of time, up to several years. Several treatment options are available for patients with metastatic disease, ranging from watchful waiting to chemotherapy, transcatheter arterial chemoembolization, and liver transplantation for liver metastases. The precise role and the best sequence for these different therapies are currently not known for several reasons: rarity and heterogeneity of the disease, lack of randomized trials, and lack of consensus between experts. The European Neuroendocrine Tumor Society has recently released guidelines for pathological diagnosis and staging of endocrine tumors,4, 5 which derive from consensus conferences held in 2006 and 2007. These guidelines may help homogenize the management of these rare tumors in the future.
Most reports have addressed the efficacy of chemotherapy in the first-line setting, with response rates of 20% to 35%.6 Combinations including streptozocin, 5-fluorouracil, or doxorubicin have yielded only modest response rates, and are associated with significant toxicity. These combinations are, however, considered to be the most efficient regimens available for patients with carcinoid and pancreatic islet cell tumors.7-9 Currently few data exist regarding chemotherapy for metastatic endocrine carcinoma beyond first line. A prospective study from the Eastern Cooperative Oncology Group investigated the activity of single-agent dacarbazine in patients failing streptozocin-based regimens and reported only modest activity in this setting.10 Although there is currently no standard therapy after failure of streptozocin combination chemotherapy, several treatment options are available aside from chemotherapy: promising results with peptide receptor radionuclide therapy were recently reported by Kwekkeboom et al using (177Lu-DOTA0, Tyr3)octreotate11; several targeted therapies are also currently under active investigation, and preliminary reports of promising activity have been recently reported for everolimus and sunitinib.12, 13 However, chemotherapy regimens with less associated toxicity than streptozocin-based regimens are needed.
Gemcitabine is a nucleoside analog with structural similarities to cytarabine, and was widely used in the treatment of patients with advanced pancreatic adenocarcinoma.14 Kulke et al reported its use as a single agent in 18 patients with metastatic NETs. Although treatment was well tolerated, no radiological responses were observed, 65% of patients experienced disease stabilization, and the median overall survival was <1 year.15
Oxaliplatin is a platinum analogue with a favorable safety profile in comparison with other platinum derivatives and has significant activity in various gastrointestinal cancers.16-19 Tetzlaff et al first reported the response to an oxaliplatin-based regimen in a patient with metastatic carcinoid tumor in 2005.20 Bajetta et al recently reported the interesting results of a phase 2 trial of capecitabine and oxaliplatin combination in patients with advanced untreated neuroendocrine tumor. The response rate in the 27 patients with low-grade neuroendocrine tumor was 30% (8 of 27).21 Several phase 2 trials are currently ongoing with oxaliplatin-based regimens in patients with well-differentiated NETs.21-23
The favorable toxicity profile of gemcitabine-oxaliplatin combination (GEMOX), together with the significant antitumor activity in several gastrointestinal malignancies,17, 19 led us to evaluate its efficacy and tolerability for patients with metastatic NETs who failed 1 or several previous chemotherapy regimens and/or biotherapy. To our knowledge, the current study is the first to evaluate GEMOX in this setting.
MATERIALS AND METHODS
This was a retrospective study aiming to assess the efficacy of GEMOX in patients with advanced NETs. The study was approved by the review board of the gastroenterology-oncology department.
The study population consisted of patients with histologically confirmed, progressive, unresectable, or metastatic NETs. Previous treatments with chemotherapy, chemoembolization, interferon, or octreotide were allowed. Between June 2004 and April 2008, 20 consecutive patients received GEMOX chemotherapy for metastatic or locally advanced endocrine carcinoma with palliative intent in 2 centers of a single institution. Patients' names were retrieved from an institutional database of patients with endocrine tumors. Patients' files were then reviewed for efficacy outcomes (eg, response rate and progression-free survival), toxicities, and overall survival. All patients had a confirmed diagnosis of (neuro)endocrine carcinoma24, 25 by an expert pathologist in the field (J.-Y. S.).
The GEMOX regimen comprised gemcitabine 1000 mg/m2 as a 30-minute intravenous infusion followed by oxaliplatin 100 mg/m2 as a 2-hour intravenous infusion. Cycles were repeated every 2 weeks. Complete blood count, serum biochemistry, and liver function tests were repeated 24 to 48 hours before Day 1 of each treatment cycle. Chemotherapy was administered if neutrophils were >1.5 billion/L and platelets >100 billion/L, except for 1 patient with documented hypersplenism in whom treatment was administered if platelets were >75 billion/L. Chemotherapy doses were reduced if needed according to the following guidelines: for patients experiencing excessive (grade 4 hematological side effects or grade ≥3 nonhematological side effects) “general” toxicities (eg, asthenia, hematological toxicity, nausea and vomiting), doses of both drugs were reduced to 75% and to 50% if a subsequent dose reduction was needed. For oxaliplatin-specific toxicities (paresthesia and neuropathy), only the doses of oxaliplatin were reduced, first to 75 mg/m2 and then to 50 mg/m2.
All patients were assessed every 4 cycles using clinical examination and computed tomography scan and/or magnetic resonance imaging when appropriate. Radiological response was classified according to Response Evaluation Criteria in Solid Tumors criteria.26 Complete response required disappearance of all target lesions lasting for at least 4 weeks. Partial response (PR) required a decrease of >30% in the sum of the largest perpendicular diameters of all measurable lesions, persisting for at least 4 weeks, without progression of any nonmeasurable sites and without the appearance of new lesions. Progressive disease (PD) included an increase of 20% or more in the sum of the longest diameters of target lesions, taking as references the smallest longest diameter recorded since the treatment started, or the appearance of 1 or more new lesions. Stable disease was defined as having neither sufficient shrinkage to qualify as a PR, nor sufficient increase to qualify as PD. After chemotherapy withdrawal (because of either progression or toxicity, or for patients on chemotherapy “holiday”) patients were followed-up every 3 month with clinical examination and imaging studies.
Overall survival was calculated from the date of Cycle 1-Day 1 of GEMOX chemotherapy to the date of death. Progression-free survival (PFS) was calculated from the date of Cycle 1-Day 1 of GEMOX chemotherapy to the date of disease progression or death from any cause, whichever occurred first. Survival distributions were estimated using the Kaplan-Meier method.
The baseline characteristics of the 20 patients are summarized in Table 1. Briefly, median age was 56 years (range, 32-73 years), and most patients had digestive endocrine carcinomas (the primary site was ileal in 5 cases, pancreatic in 5, and bronchial in 4; the 6 remaining patients had primary tumors of other origin). Thirteen patients had a World Health Organization (WHO) performance scale (PS) of 0 or 1, and 7 patients had a PS of 2. The median number of previous chemotherapy regimen was 2 (range, 0-4), and all patients but 2 had progressed on at least 1 previous chemotherapy regimen; only 1 patient had previously received a cisplatin-containing regimen (etoposide-cisplatin). The median time from original diagnosis to start of GEMOX chemotherapy was 4.7 years (range, 0.6-26.9 years), and the median time from diagnosis of metastatic disease was 3.6 years (range, 0.6-21.3 years). All patients but 1 had their tumor assigned to Group 2 of the WHO classification for endocrine tumors24 (ie, well-differentiated endocrine carcinoma); 1 patient was diagnosed as having poorly differentiated endocrine carcinoma, but nonetheless experienced overall prolonged survival, including after metastatic relapse (>6 years). Proliferation index using Ki67 was available for only 4 patients; 3 patients had pancreatic endocrine carcinoma with 5%, 20%, and 60% of cells staining positive for Ki67, and 1 patient with endocrine carcinoma of the ovary had 20% of cells staining positive. Patients with ≥20% cells positive for KI67 were still considered as Group 2 (grade 3) of the WHO classification based on morphology.
Table 1. Patient Characteristics (N=20)
WHO, World Health Organization; FED, fluorouracil, epirubicin, dacarbazine; FOLFIRI, folinic acid, fluorouracil, and irinotecan.
Other primary sites include: ovary (n=1), rectum (n = 1), gall bladder (n = 1), and unknown primary site (n=1).
Including 4 small intestine carcinoid tumors, 1 bronchial carcinoid tumor, and 1 pancreatic gastrinoma (with secondary ectopic adrenocorticotropic dhormone production).
Other chemotherapy regimens include etoposide-cisplatin (n=1), single-agent gemcitabine (n=1), and single-agent temozolomide (n=1).
Ten (50%) patients had previously received α-interferon. Two patients had relapsed 21 and 42 months after liver transplantation for liver metastases of NETs (1 pancreatic endocrine carcinoma, 1 bronchial carcinoid, respectively). All patients had received GEMOX for documented disease progression within the last 3 months: 14 patients had Response Evaluation Criteria in Solid Tumors-defined progression (ie, >20% increase in size or appearance of new target lesions), whereas 6 patients had some kind of radiological progression (median 11% increase in size; range, 6%-15%) with clinical deterioration or symptom worsening.
The 2 patients who received GEMOX as first-line therapy were 1 patient with a carcinoid tumor of the ovary with metastases to the liver, lung, and peritoneum, and 1 patient with a bronchial carcinoid tumor with liver metastases and peritoneal carcinomatosis. They received 4 and 10 cycles of full-dose GEMOX, respectively. The patient with the carcinoid tumor of the ovary progressed after 4 cycles but is still alive with >7 months of follow-up. The patient with the bronchial carcinoid tumor had a PR and went on a chemotherapy holiday after 10 cycles. Disease progression was documented 9.4 months after initiation of GEMOX. These 2 patients were excluded from further analysis.
Safety and Duration of Treatment
GEMOX was given at half dose (gemcitabine 500 mg/m2 and oxaliplatin 50 mg/m2) in 5 patients who were considered too fragile to tolerate full-dose combination; 1 patient had received 3 prior chemotherapy regimens and had grade 1 thrombocytopenia because of hypersplenism, 1 patient had just undergone bilateral adrenalectomy for ectopic adrenocorticotropic hormone (ACTH) secretion inducing Cushing syndrome, 1 patient had poor hematological tolerance to chemotherapy after liver transplantation, and the other 2 patients had only received 1 prior chemotherapy regimen but with poor overall tolerance. The median number of GEMOX cycles administered was 7 (range, 1-16). Ten patients discontinued GEMOX because of toxicity, in 6 cases because of oxaliplatin-induced grade 2 neuropathy, whereas 5 patients discontinued because of disease progression (Table 2); the 3 remaining patients are still receiving GEMOX (n = 1), are still on chemotherapy holiday (n = 1), or were lost to follow-up (n = 1).
Table 2. Reasons for Early Treatment Discontinuation (N = 18)
Reason for Early Study Discontinuation
Grade ≥2 neurotoxicity
Grade ≥3 emesis/vomiting
Grade ≥3 infection
Grade ≥3 fatigue
Overall, 122 cycles of GEMOX were administered to 18 patients; 77 cycles were administered at full dose, 2 at 75% dose, and 43 at 50% dose (because of the finding that 5 patients started treatment at 50% dose). Thirteen (11%) cycles were delayed in 6 patients; 6 cycles were delayed because of thrombocytopenia (3 patients), 4 cycles because of asthenia (2 patients), 2 because of poor digestive tolerance (2 patients), and 1 because of infection. Two patients required dose reductions because of recurrent thrombocytopenia. No clinically significant thrombocytopenia occurred (eg, bleeding because of grade 3 or 4 thrombocytopenia). Two patients received primary prophylaxis with granulocyte colony-stimulating factor, and no patient required rehospitalization for febrile neutropenia. Two patients required erythrocyte transfusion for grade 3 anemia.
Response Rate, PFS, and Overall Survival
Partial response was observed in 3 of 18 patients (17%; 95% confidence interval, 3.6-41.4), 12 (67%) patients had stable disease for a median of 5.5 months (range, 3.4-20.3 months), and 3 (17%) patients had disease progression at first assessment. Median PFS and overall survivals were 7.0 months and 23.4 months, respectively (Figs. 1, 2, respectively). Four patients had disease stabilization for ≥12 months, 1 of which is still receiving intermittent cycles of GEMOX chemotherapy >20 months after the first cycle.
Patients with a PR included 1 patient with thymic carcinoid (time to progression, 14.6 months) and 2 patients with pancreatic endocrine carcinomas (time to progression, 12 and 7 months, respectively).
Baseline chromogranin A (within 1 month of Cycle 1-Day 1 of GEMOX) was available for 6 of 18 patients, and follow-up values during GEMOX chemotherapy were available for 3 of them: in 1 patient with a radiological PR, chromogranin A decreased by 17%; in 2 patients with radiologically stable disease, chromogranin A rose by 64% and 67%, respectively. One patient with ectopic ACTH production from an advanced gastrinoma had a 10% decrease in ACTH during GEMOX chemotherapy.
We report here our experience with GEMOX chemotherapy for the treatment of patients with well-differentiated advanced progressive NETs. In our hands, this combination was feasible, with acceptable toxicity in the setting of heavily pretreated patients, albeit doses were adapted for 5 patients. Six patients discontinued chemotherapy because of oxaliplatin-related neurotoxicity, no grade 4 toxicity was observed, and no treatment-related death was recorded. The overall response rate was 17%, 67% of patients had stable disease, and the median PFS was 7 months.
In the setting of first-line therapy for advanced disease, streptozocin-containing regimens produce objective response in 30% to 35% of patients with islet-cell carcinoma, with PFS of 18-24 months. However, PFS is usually shorter when patients receive these combinations as second-line therapy.27, 28 In patients with well-differentiated gastrointestinal endocrine carcinomas, streptozocin-based combinations (usually with 5-fluorouracil) show response rates of 10% to 20% with PFS of 5 to 7 months.8, 10 Other combinations such as 5-fluorouracil-epirubicin-dacarbazine seem to have similar activity.29 Finally, patients with metastatic pulmonary carcinoids are usually treated with platinum-based chemotherapy, but results are disappointing overall.30, 31
After failure with these “standard” first-line regimens, no single chemotherapy regimen has emerged as a standard for second-line therapy. In a randomized phase 2/3 study from the Eastern Cooperative Oncology Group, 67 patients with carcinoid tumors received dacarbazine in a crossover arm after failure of streptozotocin and doxorubicin or streptozotocin and fluorouracil combination. In this study, 5 of 61 (8.2%) evaluable patients had a response and 7 (11.3%) had stable disease; the PFS was not reported, but the overall survival was 11.9 months.10 In another arm of this study, 21 patients received single-agent dacarbazine either because they had previously received doxorubicin, fluorouracil, or streptozocin, or because they had renal or heart disease counterindicating streptozotocin and doxorubicin or streptozotocin and fluorouracil combination. In this latter arm, the response rate (PRs) was 9.5% (2 of 21), and the median PFS and overall survival were 4.1 and 8.8 months, respectively.10 In contrast, our results show that gemcitabine-oxaliplatin combination has significant activity in this setting, with a 17% response rate and a 7.0-month median PFS in patients with neuroendocrine tumors previously treated with chemotherapy (in most cases 2 or more previous regimens). Overall toxicity was mild; however, half of the patients discontinued treatment because of toxicity, in most cases because of oxaliplatin-related neurotoxicity.
Newer agents, such as paclitaxel, docetaxel, topotecan, and gemcitabine, have shown little activity as single agents.15, 32-34 More recently, temozolomide has been shown to have interesting single-agent activity with a favorable toxicity profile in previously treated patients.35 Several phase 2 trials have shown promising results with temozolomide-based combinations using antiangiogenic drugs such as bevacizumab or thalidomide,36, 37 and capecitabine-temozolomide was reported to produce responses in 59% to 71% of patients with advanced pancreatic endocrine tumor.38, 39
Current research focuses on the incorporation of antiangiogenic drugs and/or drugs targeting the mTOR pathway in the treatment of NETs either alone or in combination with conventional chemotherapy or interferon. Several phase 2 trials have been reported with agents such as temsirolimus,40 thalidomide,36 sorafenib,41 everolimus,12 sunitinib,13 and bevacizumab,22, 23, 42 and phase 3 trials are currently ongoing with everolimus, sunitinib, and bevacizumab. Some published results are promising, but others are either disappointing from the efficacy standpoint (temsirolimus)40 or raise concern regarding the safety of such agents in this setting (sorafenib).41
Preliminary results of 2 phase 2 trials of bevacizumab and oxaliplatin-based combinations seem to support the activity of oxaliplatin in the treatment of advanced NETs.22, 23 These combinations also include bevacizumab, which was previously shown to have activity in NETs37, 42; however, the reported response rate suggests increased activity with oxaliplatin-base chemotherapy. In a phase 2 trial, Bajetta et al reported a 27.5% response rate in 40 patients with neuroendocrine tumors treated with oxaliplatin and capecitabine.21 In this study, 27 patients had well-differentiated endocrine carcinomas; in this subgroup, the response rate was 30%, 48% of patients had stable disease as their best response, the median time to progression was 20 months, and median overall survival was 40 months. Toxicity was mild overall, and only 12.5% of patients experienced oxaliplatin-related paresthesia, which is likely because of the predetermined number of cycles of chemotherapy in this study.21
Although no definitive conclusions can be drawn from our series of patients because of the heterogeneity of the underlying disease, GEMOX combination seems active, given the heavily pretreated population. Interestingly, whereas gemcitabine as a single agent was shown to have little activity in this disease,15 the combination with oxaliplatin yielded an interesting response rate (17%), with a PFS of 7.0 months, which compares favorably with other agents in this setting, including targeted agents.41, 43, 44
Overall, our experience with GEMOX chemotherapy shows that this combination is feasible and has promising activity in patients with previously treated NETs.