Imatinib mesylate for targeting the platelet-derived growth factor β receptor in combination with fluorouracil and leucovorin in patients with refractory pancreatic, bile duct, colorectal, or gastric cancer—A dose-escalation Phase I trial




In previous experimental models, because of its ability to inhibit the activity of platelet-derived growth factor β receptor, imatinib decreased the interstitial fluid pressure and improved the delivery and efficacy of anticancer drugs, including fluorouracil. The objective of this Phase I study was to define the dose-limiting toxicity (DLT) and maximum tolerated dose (MTD) of imatinib in combination with fluorouracil and leucovorin in patients with chemotherapy-refractory gastrointestinal cancer.


A 3-patient cohort dose-escalating study design was used. Patients received leucovorin 200 mg/m2 followed by fluorouracil 2000 mg/m2 as a 24-hour infusion on Days 1 and 2 combined with imatinib on Days −4, −3, −2, −1, 1, 2, 3, and 4. Cycles were repeated every 2 weeks, and the imatinib dose was escalated from 300 mg daily to 700 mg daily in 100-mg steps.


Thirty patients were enrolled at 5 dose levels. Frequent and dose-dependant National Cancer Institute Common Toxicity Criteria grade 1–4 adverse events with suspected relation to the treatment were anemia (43%), nausea (33%), fluid retention (27%), elevated serum γ-glutamyl-transpeptidase (20%), and diarrhea. DLTs were severe neutropenia, central fluid retention, and severe nausea observed in 1 patient each, resulting in an MTD for imatinib of 600 mg per day. There were no differences in imatinib pharmacokinetics before or during chemotherapy. A minor response was observed; and signs of clinical activity, including the resolution of ascites and improvement in performance status, were noted in some patients.


The combination of biweekly fluorouracil/leucovorin and imatinib 600 mg daily given in a week-on/week-off schedule was feasible and safe. Nausea and fluid retention represented the DLTs. Cancer 2007. © 2007 American Cancer Society.

Several mechanisms of resistance to conventional anticancer drugs have been described.1 One of these mechanisms is increased interstitial fluid pressure (IFP), which impairs the transvascular transport of molecules and reduces the tumor uptake of cytotoxic drugs.2 The etiology of increased IFP in tumors is poorly understood. However, growing evidence supports the hypothesis that the transmembrane platelet-derived growth factor β receptor (PDGFR-β) is involved in the regulation of IFP.3 Pietras and colleagues investigated the effect of imatinib mesylate, a potent PDGFR-β inhibitor, on the tumor IFP in animal models and provided experimental evidence that imatinib lowers tumor IFP and improves the delivery and efficacy of anticancer drugs.4–6 Those authors also demonstrated that imatinib alone had no effect on tumor cell growth but enhanced the efficacy of chemotherapeutic agents. One of these agents was fluorouracil (FU), which has been the mainstay of treatment for almost all types of adenocarcinoma of the gastrointestinal tract for 30 years.

Our clinical Phase I study was based on the rationale provided by these data. The objective of the study was to evaluate the safety and dose-limiting toxicity (DLT) of the combination of imatinib and FU modulated by leucovorin (LV) in patients with refractory pancreatic, bile duct, colorectal, and gastric cancer. An infusional biweekly FU/LV protocol was chosen, and imatinib was administered during the week of chemotherapy only.


Patient Eligibility

Patients with histologically confirmed, measurable metastatic gastric, pancreatic, colorectal, or bile duct cancer who were refractory to standard therapy (defined as having disease progression during or within 6 weeks of previous chemotherapy documented by 2 subsequent computed tomography [CT] scans using World Health Organization [WHO] criteria) or for whom no effective therapy existed were eligible for the study. Further criteria were age >18 years; an Eastern Cooperative Oncology Group (ECOG) performance status ≤3; sufficient renal, hepatic, and bone marrow function (defined as the following: total bilirubin <1.5 times the upper limit of normal [ULN], aspartate and alanine aminotransferase levels <2.5 times the ULN, creatinine <1.5 times the ULN, absolute neutrophil count >1.5 × 109/L, and platelets >100 × 109/L); complete resolution of all side effects caused by previous treatment; no concurrent, uncontrolled medical illness; and no other current or previous malignancy within the past 5 years (with the exception basal cell skin cancer or a cervical carcinoma in situ treated by surgery). Patients were excluded from the study if they had brain metastases, inflammatory bowel disease, coronary heart disease, cardiac insufficiency (New York Heart Association II-IV), known hypersensitivity to any of the study drugs, or if they were pregnant or breast-feeding. Women of child-bearing potential were advised to take adequate precautions to prevent pregnancy. Participants gave their written informed consent before they entered the study, which was approved by the responsible ethics committee. After a high drop-out rate (patients who did not complete 2 cycles of treatment for reasons other than toxicity) was observed in the first cohort, the protocol was amended to include only patients with an ECOG performance status ≤2.


Patients with pancreatic and bile duct cancer received LV 200 mg/m2 as a 2-hour intravenous infusion on Days 1 and 2 followed by FU 2000 mg/m2 as a 24-hour continuous infusion on Days 1 and 2 given at a constant infusion rate of 84 mg/m2 per hour, combined with imatinib given as a single dose (up to 600 mg) or as a divided dose on Days −4, −3, −2, −1, 1, 2, 3, and 4 of each cycle. Patients with colorectal or gastric cancer received a modified FOLFOX regimen7, 8: oxaliplatin 85 mg/m2 as a 2-hour intravenous infusion, followed by LV 200 mg/m2 as a 2-hour intravenous infusion, followed by FU 2600 g/m2 as a 24-hour continuous infusion on Day 1. Imatinib was administered on the same schedule that was used for the FU/LV regimen. Cycles were repeated every 2 weeks until patients developed disease progression or limiting toxicity. Prophylactic colony-stimulating factor or erythropoietin administration was not permitted. However, the therapeutic administration of hematopoietic growth factors was permitted at the discretion of the investigator.

Dose Escalation and Toxicity Assessment

For the assessment of toxicity, patients were interviewed using a standardized set of questions and were evaluated by physical examination and laboratory tests, including complete blood count, blood chemistry, and urine analysis, every week. In addition, electrocardiograms were obtained prior to each cycle of treatment. Toxicities were graded according to the National Cancer Institute Common Toxicity Criteria (NCI-CTC), version 2. DLT was defined as a hypersensitivity reaction grade ≥2, any grade ≥3 nonhematologic toxicity (except alopecia), grade 4 thrombocytopenia, or grade 4 neutropenia that persisted >5 days without the administration of hematopoietic growth factors. Adverse events were considered dose limiting when they were related at least possibly to the study treatment and when they occurred during the first 2 cycles of treatment. Abnormal laboratory values were considered dose limiting when, in addition, they were considered clinically significant by the investigator. A 3-patient cohort, dose-escalating study design was used. The initial dose of imatinib was 300 mg and was increased by 100 mg in the next cohort, provided that all patients in the previous cohort finished their first 2 cycles (4 weeks) of treatment without experiencing a DLT. In case a patient experienced a DLT, 3 additional patients were enrolled at the same dose level. If ≥2 of 6 or ≥1 of 3 patients at the same dose level experienced a DLT, then that dose level was considered a nontolerable toxic dose, and the dose level of the previous cohort was considered the maximum tolerated dose (MTD). Patients were considered evaluable for safety analysis when they had received at least 1 administration of any of the study drugs, and patients who did not complete 2 cycles for reasons other than toxicity were replaced. Intrapatient dose escalation or reduction was not permitted.

Patients were to be treated in mixed cohorts until the first DLT was observed in a patient who received the additional oxaliplatin. Thereafter, the groups were to be separated to define an MTD of imatinib for each group. However, a minority of trial patients (5 of 30) had colorectal or gastric cancer, and no DLTs actually were observed in these patients. Therefore, our work focuses on the combination of imatinib with FU/LV only. No conclusions can be drawn concerning the oxaliplatin combination.

Assessment of Response

Responses were classified according to WHO criteria.9 CT scans of target areas were obtained before the start of the treatment and were repeated every 6 weeks. Responses had to be confirmed by an independent radiologist. The time to progression (TTP) was measured from the start of treatment until disease progression or death from any cause.

Imatinib Pharmacokinetics

Imatinib pharmacokinetic studies were performed in patients who received imatinib doses ≥600 mg daily. Plasma samples were collected on Day −1 (before chemotherapy) and Day 2 (during chemotherapy) before the daily dose and after 0.5 hours, 1 hour, 2 hours, 3 hours, 5 hours, 6 hours, 7 hours, 10 hours, 12 hours, 24 hours, and 25 hours. Analysis of imatinib and N-desmethyl-imatinib, the main metabolite, was performed by using high-performance liquid chromatography.10 Dose-dependent parameters were calculated by linear interpolation of a 300-mg/m2, once-daily imatinib dose, assuming 100% imatinib absorption and 20% metabolization of N-desmethyl-imatinib.



Thirty patients were enrolled in 6 treatment cohorts at 5 dose levels. Patient characteristics are summarized in Table 1. Patients with colorectal or gastric cancer were included in Cohorts 1, 3, and 4. All patients except those with bile duct cancer had received at least 1 prior standard chemotherapy regimen in the palliative setting. All patients had documented progression of their disease prior to the study treatment, and all of those who were treated previously had chemotherapy-resistant disease, which was defined as having documented disease progression during or within 6 weeks of previous chemotherapy. All patients were assessable for the safety analysis, and 25 of 30 patients were assessable for efficacy. Four patients (3 patients in Cohort 1 and 1 patient in Cohort 3) did not complete 2 cycles for reasons other than toxicity and were replaced. Twenty-three of 30 patients (76.6%) discontinued the study treatment because of disease progression, 4 patients (13.3%) discontinued treatment because of toxicity (related to toxicity in 3 patients and unrelated to toxicity in 1 patient), and 3 additional patients (9.7%) withdrew consent for personal reasons.

Table 1. Patient Characteristics
CharacteristicNo. of patients (%)
Total, N = 30Pancreas, n = 17Bile duct, n = 8
  • ECOG indicates Eastern Cooperative Oncology Group.

  • *

    Other involved organs included the duodenum (1), bone (5), spleen (1), small intestine (1), and rectum (1).

 Women13 (43.3)11 (64.7)2 (25)
 Men17 (56.7)6 (35.3)6 (75)
Age, y
 18–505 (16.7)1 (5.9)2 (25)
 51–659 (30)6 (35.3)1 (12.5)
 66–9016 (53.3)10 (58.8)5 (62.5)
ECOG status
 03 (10)1 (5.9)1 (12.5)
 18 (26.7)6 (35.3)0 (0)
 212 (40)6 (35.3)4 (50)
 37 (23.3)4 (23.5)3 (37.5)
Type of disease
 Pancreas17 (56.6)
 Bile duct8 (26.7)
 Colorectal3 (10)
 Gastric2 (6.7)
 Chemotherapy27 (90)17 (100)5 (62.5)
 Radiation2 (6.7)1 (5.9)0 (0)
 Surgery0 (0)0 (0)0 (0)
Treatment-free interval before enrolment, mo
 <17 (23.3)7 (41.2)0 (0)
 1–211 (36.7)7 (41.2)1 (12.5)
 >28 (26.7)2 (11.8)4 (50)
No. of organs involved
 110 (33.3)7 (41.2)4 (50)
 213 (43.3)6 (35.3)4 (50)
 33 (10)1 (5.9)0 (0)
 43 (10)3 (17.6)0 (0)
 61 (3.3)0 (0)0 (0)
Organs involved (primary tumor excluded)
 Liver27 (90)16 (94.1)6 (75)
 Lymph nodes11 (36.7)5 (29.4)1 (12.5)
 Peritoneum6 (20)3 (17.6)1 (12.5)
 Lung7 (23.3)2 (11.8)3 (37.5)
 Other*9 (30)7 (41.2)1 (12.5)

Safety, Tolerability, and MTD

Patients received a total of 95 complete cycles at 6 dose levels. The mean and median number of cycles administered was 3.1 and 2, respectively (range, 1–11 cycles). Ten patients (30%) received ≥3 cycles of treatment. Three patients discontinued treatment because of related adverse events that were considered dose limiting and are discussed below. One additional patient discontinued treatment for a toxicity that was considered unrelated to imatinib (myocardial ischemia). No dose reductions, treatment interruptions, or delays for toxicity were required (with the exceptions of the 3 DLTs), and the majority of patients discontinued treatment for disease progression. Dose levels, number of cycles, DLTs, and toxic events are presented in Table 2, and the overall toxicities are presented in Table 3.

Table 2. Cohorts With Cycles Administered, Adverse Events, and Dose-limiting Toxicities
Cohort no.Dose level, mg/DayNo. of patientsNo. of cyclesDLTNo. of AEs*
  • DLT indicates dose-limiting toxicity; AEs, adverse events.

  • *

    Adverse events that at least possibly were related to imatinib and combined fluorouracil and leucovorin.

  • Three patients who received <2 cycles for reasons other than toxicity were replaced.

  • One patient who received <2 cycles for reasons other than toxicity was replaced.

Table 3. Toxicities According to the National Cancer Institute Common Toxicity Criteria, Version 2.0*
ToxicityNo. of patients (%)
Grade 1 or 2, n = 30Grade 3 or 4, n = 30
With suspected relationTotalWith suspected relationTotal
  • AST indicates aspartate aminotransferase; ALT, alanine aminotransferase; gGT, γ-glutamyl-transpeptidase.

  • *

    Data for events with suspected relation to the study drugs that occurred in ≥5% of patients.

 Neutropenia3 (10)6 (20)1 (3.3)1 (3.3)
 Leukopenia4 (13.3)6 (20)2 (6.7)2 (6.7)
 Thrombocytopenia3 (10)6 (20)1 (3.3)2 (6.7)
 Anemia12 (40)26 (86.7)1 (3.3)4 (13.3)
 Nausea9 (30)11 (36.7)1 (3.3)1 (3.3)
 Vomiting5 (16.7)6 (20)0 (0)0 (0)
 Diarrhea3 (10)4 (13.3)0 (0)0 (0)
 Constipation5 (16.7)5 (16.7)0 (0)1 (3.3)
 AST3 (10)18 (60)0 (0)2 (6.7)
 ALT2 (6.7)13 (43.3)0 (0)0 (0)
 gGT2 (6.7)14 (46.7)4 (13.3)11 (36.7)
 Bilirubin4 (13.3)8 (26.7)0 (0)2 (6.7)
 Edema/fluid retention7 (23.3)7 (23.3)1 (3.3)1 (3.3)
 Fever2 (6.7)7 (23.3)1 (3.3)1 (3.3)
 Alopecia2 (6.7)5 (16.7)
 Pain1 (3.3)14 (46.7)0 (0)5 (16.7)

Hematologic side effects were uncommon or mild in severity with the exception of anemia. NCI-CTC grade 1 or 2 and grade 3 or 4 anemia (related or not) were observed in 100% and 13.3% of patients, respectively. In contrast, severe (grade 3 or 4) neutropenia or neutropenic fever was observed in only 1 of 30 patients (3.3%). The nonhematologic toxicity profile generally was moderate, with NCI-CTC grade 1 through 4 nausea representing the leading toxicity observed in 40% of patients. The second most common toxicity was fluid retention followed by vomiting, which were observed in 26.6% and 20% of patients, respectively. The main laboratory abnormality observed in the study was the elevation of serum γ-glutamyl-transpeptidase (gGT), which (all grades; related or not) was observed in 25 of 30 patients (83.4%). NCI-CTC grade 3 or 4 elevation of serum gGT was observed in 11 patients (36.7%) and was considered related to the treatment in 4 patients (13.3%). A marked increase in the incidence at higher doses of imatinib was noted for gGT, nausea, diarrhea, edema, and anemia but not for neutropenia, as shown in Figure 1.

Figure 1.

Selected toxicities by cohort. Events that at least possibly were related to the treatment are shown in relation to the number of cycles completed within the cohort. Imatinib dose levels: cohort 1, 300 mg per day; cohort 2, 400 mg per day; cohort 3, 500 mg per day; cohorts 4 and 6, 600 mg per day; cohort 5, 700 mg per day. Cohorts that included patients who received the oxaliplatin-based treatment are indicated by an asterisk. gGT indicates γ-glutamyl-transpeptidase.

Three patients had adverse events that matched the protocol definition of DLT. One patient (pancreatic cancer) in cohort 3 (imatinib 500 mg) experienced a severe, unexpected hematologic toxicity during the first cycle, including NCI-CTC grade 4 leukopenia, granulocytopenia, thrombocytopenia, and neutropenic fever, and died 4 weeks after treatment from multiorgan failure despite maximal supportive therapy. Genotyping analysis for the most common mutation associated with dihydropyrimidine dehydrogenase (DPD) deficiency (DPYD*2A exon 14-skipping mutation) was performed in the patient and was negative. Cohort 3 was extended to include 3 additional patients. No further DLTs were observed in the cohort, and no comparable toxicities were observed in any other patients in the study. Two additional DLTs were observed in Cohort 5 (imatinib 700 mg). One patient (bile duct cancer) had grade 4 noncardiac fluid retention, and an additional patient (pancreatic cancer) had grade 3 nausea. Based on these dose-limiting events, the dose level of imatinib 700 mg daily (Cohort 5) was defined as toxic, and the dose level of imatinib 600 mg daily (Cohort 4) was defined as the MTD according to the protocol allegations. Additional patients (Cohort 6) were enrolled at the MTD level to a total number of 30 patients, as it was planed in the study protocol.

Five patients received the oxaliplatin-based regimen (1 patient received imatinib 300 mg per day, 2 patients received imatinib 500 mg per day, and 2 patients received imatinib 600 mg per day). All 5 patients tolerated the treatment well without DLTs or any NCI-CTC grade 3 or 4 toxicities that were considered related to the treatment. NCI-CTC grade 1 or 2 toxicities included anemia and leukopenia in 3 patients and 2 patients, respectively. Grade 1 or 2 fluid retention, neutropenia, nausea, and gGT elevation occurred in 1 patient each.


Five of 17 patients (29.4%) with pancreatic carcinoma, 4 of 8 patients (50%) with bile duct carcinoma, and 1 of 3 patients with colorectal carcinoma had stable disease as their best response (6-week assessment). However, most of those responses were not durable, and the majority of those patients had disease progression at their second tumor-evaluation visit (12-week assessment). A remarkable clinical development was observed in 2 of 17 patients with pancreatic carcinoma and in 2 of 8 patients with bile duct carcinoma, which may deserve further explanation: Three of those 4 patients had documented, rapid progression of disease under their previous treatment and, thus, clearly were resistant to chemotherapy, and 1 patient was previously untreated. One of the previously treated patients (pancreatic cancer) had a slight shrinkage of the primary tumor (20%; minor response), and 2 other patients (bile duct cancer and pancreatic cancer) had stable disease with a TTP in the study that was longer than the TTP documented with the previous treatment. The 1 previously untreated patient with bile duct cancer had a prolonged TTP that lasted for 12 months. In addition, the patients experienced significant relief of cancer-related symptoms, an improvement in ECOG performance status, and resolution of ascites (in 2 patients).


Three hundred thirty blood samples from 15 patients (all of whom received imatinib ≥600 mg) were analyzed. There were no significant differences in pharmacokinetics between samples that were obtained before the initiation of chemotherapy and samples that were obtained during treatment (Table 4), and no distinctive features were observed in the pharmacokinetics of patients who had a DLT.

Table 4. Pharmacokinetic Parameters of Imatinib and its Metabolite N-desmethyl-imatinib
Before chemotherapyDuring chemotherapyBefore chemotherapyDuring chemotherapy
  1. t ½ indicates half-life–time; AUC, area under the concentration-time curve; C max, peak concentration; t max, time to C max; Vss, volume of distribution.

t ½ Absorption, h1.41.4
t ½ Metabolism, h1.53.2
t ½, h26255253
AUC, μg × h/mL454412.012.0
C max, μg/mL1.
t max, h3.
Clearance total, mL/min/m21101108082
Vss, L/m2230250340360
Ratio AUC, % N-desmethyl-imatinib2727
Correlation coefficient0.980.990.990.99


For this Phase I study, we evaluated the feasibility and toxicity and determined the optimal dosing of imatinib as a PDGFR-β inhibitor combined with FU/LV in patients with different types of gastrointestinal cancer. We preferred the infusional use over the bolus use of FU in the study because of its more favorable hematologic toxicity profile. In addition, we chose a biweekly infusional schedule that previously demonstrated feasibility and safety in combination with additional cytotoxic drugs in patients with colorectal cancer11 to allow the intermittent administration of imatinib and to reduce the likelihood of cumulative toxicity. The intermittent use of imatinib also was supported, because no activity of imatinib as a single agent was anticipated.5

The combination generally was tolerated well, and most of the toxicities observed were related to the progression of disease and reflected the complications usually seen in patients with pancreatic or bile duct cancer at very advanced stages. Three NCI-CTC grade 3 or 4 events, 1 hematologic and 2 nonhematologic, were dose limiting during the trial. One patient (3.3%) in Cohort 3 had severe, unexpected myelotoxicity that was strongly suggestive of fluoropyrimidine intolerance. This pattern of toxicity was not observed in any other patient, nor was any other NCI-CTC grade 3 or 4 neutropenia observed during the study. FU was the agent considered responsible for the occurrence of this severely toxic case. However, the possibility cannot be excluded that imatinib may have intensified the effects of FU in that patient; especially because an exon-14-skipping mutation of the DPD gene, which accounts for approximately 25% of patients with fluoropyrimidine intolerance, was not identified. One patient (3.3%) in Cohort 5 developed severe fluid retention, including peripheral edema, ascites, bilateral pleural effusions, and anasarca, during the treatment. The patient had no history of cardiac disease or renal insufficiency, and no significant congestive heart failure or thromboembolic events were identified on clinical or radiologic examinations. No tumor cells could be detected in the patients' ascites or pleural effusions. On autopsy, no causal pathologic finding could be detected. Therefore, the event was considered as at least possibly related to the study treatment. One additional patient (3.3%) in Cohort 5 suffered from severe nausea that was related to the study treatment and discontinued the study for toxicity. Based on the last 2 DLTs observed in Cohort 5 (fluid retention and nausea), the imatinib 700-mg daily dose level was considered toxic, and the imatinib 600-mg daily dose level was considered the MTD.

Fluid retention is the most common dose dependent side effect of imatinib therapy and has been reported in approximately 50% of patients overall. Imatinib-related edema usually is mild and superficial. However, excessive or central fluid accumulation (eg, pleural effusion, pericardial effusion, pulmonary edema, ascites, anasarca) have been reported in approximately 1% to 3% of patients.12 Severe fluid retention also is a generally recognized, although rare, complication of conventional cytotoxic therapy,13 although it has not been described typically in patients receiving FU. A report of a Phase I trial with CDP860, a specific PDGFR-β inhibitor, in patients with cancer indicated that the trial was stopped early because 7 of 8 patients developed adverse events that were suggestive of fluid accumulation, and 3 of 8 patients developed dramatic ascites and/or pleural effusions. In that report, it was speculated that PDGFR-β inhibition may have been the mechanism of the fluid accumulation caused by the study drug CDP860, imatinib, or other multi-tyrosine-kinase inhibitors, such as SU11248. In a recent Phase I trial, 22 men with androgen-independent prostate cancer were treated with continuous imatinib 600 mg daily for 30 days plus escalating doses of docetaxel from 30 mg to 45 mg per week. A DLT was observed in 11 of the 22 patients (50%). The main toxicities were fatigue (35%) and nausea (20%), but 1 patient experienced a life-threatening, noncardiogenic pulmonary edema.

Two patients experienced the complete resolution of their malignant ascites during treatment. This seems noteworthy, because fluid retention was identified as a frequent adverse event and a DLT in the study. This may be explained by speculating that the antitumor effect of the combination on these patients' peritoneal disease may have overcome the adverse effect of inhibiting PDGFR-β in their normal tissues.

Liver transaminase elevations were a frequent adverse event in our study. The events, in the majority of patients, were related to the progression of hepatic metastases. However, in some patients, a relation to the combination of imatinib and FU/LV could not be excluded. NCI-CTC grade 3 or 4 elevations in serum gGT levels with suspected relation to the study medication was observed in 4 of 30 patients (13.3%), which clearly was higher than the rate reported in the official drug information for imatinib (<3%). The incidence of grade 3 or 4 abnormalities in liver function test results in our study also was higher than that reported for patients with gastrointestinal stromal tumors (approximately 3%14). It is worth noting that no clinical signs of hepatic dysfunction or failure were observed in our patients. However, this point should be considered with caution, because the treatment duration in the majority of the patients was comparatively short. Liver toxicity may be an issue with this combination and should be monitored closely in future studies.

A Phase I trial recently has shown that the use of continuous imatinib at a dose of 400 mg per day in combination with gemcitabine or doxorubicin in the treatment of patients with solid tumors was associated with unacceptable toxicities.15 In contrast, in our current study, we demonstrated that the combination of imatinib 600 mg daily given in a week-on/week-off schedule with biweekly FU/LV in patients with gastrointestinal cancer was feasible and safe, with nausea and fluid accumulation the DLTs.

Efficacy was a secondary endpoint of the trial. We observed a minor response; and, in 4 patients, there was a strong impression that the course of disease or disease-related symptoms was influenced favorably by the treatment. However, these few signs of efficacy are very difficult to interpret because of several aspects, including the small size and heterogeneity of the study population and the fact that most patients were not treated previously with FU/LV. Further studies are needed to investigate this issue.