A Phase I trial of fixed dose rate gemcitabine and capecitabine in metastatic renal cell carcinoma

Abstract

BACKGROUND

Metastatic renal cell carcinoma (RCC) has modest response rates to chemotherapy with gemcitabine and 5-fluorouracil (5-FU). Fixed dose rate gemcitabine infusion leads to enhanced intracellular accumulation of drug and possible augmented clinical effect. To determine the toxicity of this combination therapy in metastatic RCC, a Phase I trial was conducted.

METHODS

Patients with metastatic RCC were enrolled in a Phase I dose escalation trial. Patients received fixed dose rate gemcitabine on Days 1, 8, and 15 in combination with capecitabine, an oral 5-FU analog, given on Days 1–21 of a 28-day cycle.

RESULTS

Nine patients were enrolled at one of two dose levels. The initial dose level produced dose-limiting toxicity (DLT), including prominent palmar-plantar erythrodysesthesia (hand-foot syndrome). A modified second dose level also resulted in DLT, precluding further study. No central nervous system (CNS) toxicity was observed in three patients with CNS metastases. Two patients demonstrated an objective partial response.

CONCLUSIONS

Fixed dose rate gemcitabine in combination with capecitabine produced unacceptable toxicity in patients with advanced RCC. Further development of this schedule in RCC cannot be recommended. Cancer 2005. © 2004 American Cancer Society.

Metastatic renal cell carcinoma (RCC) has only modest response proportions to cytokine therapy, including interleukin-2 (IL-2) and/or interferon-alpha (IFN-α) administered in varying doses and schedules.1 The majority of patients ultimately become refractory to this therapy with few effective alternatives. Chemotherapy is largely ineffective in RCC,2, 3 although the most encouraging results have been obtained with 5-fluorouracil (5-FU) and gemcitabine as single agents4–6 or in combination.7

5-FU is a pyrimidine analog that depletes intracellular deoxynucleotide triphosphates, including deoxythymidine triphosphate, by inhibiting thymidylate synthetase. 5-FU as a single agent has modest activity in metastatic RCC, producing almost exclusively partial responses (PR) in 5–10% of patients.2–4 Capecitabine is an orally administered, tumor-selective fluoropyrimidine that is converted to 5-FU in tissues by pyrimidine nucleoside phosphorylase.8 Capecitabine has demonstrated activity in treatment of patients with 5-FU–responsive tumors including breast and colorectal carcinoma.9, 10

Gemcitabine is a pyrimidine nucleoside antimetabolite that, once converted to difluorodeoxycytidine triphosphate (dFdCTP), inhibits DNA synthesis by inhibition of DNA polymerase and also by direct incorporation into DNA leading to premature termination of DNA elongation.11 The diphosphate intermediate of gemcitabine also inhibits ribonucleotide reductase and thereby depletes intracellular pools of deoxyuridine monophosphate. This results in enhanced binding of 5-fluorodeoxyuridine monophosphate, the active metabolite of 5-FU, to thymidylate synthase.12, 13 Gemcitabine as a single agent also has modest activity in RCC, producing a 6% and an 8% PR rate in 2 separate Phase II studies.5, 6

Based on single-agent activity and potentially complementary mechanisms of action, combination therapy with 5-FU and gemcitabine has been investigated in metastatic RCC. An initial Phase II study administered 5-FU by continuous infusion at a dose of 150 mg/m² per day on Days 1–21 and gemcitabine 600 mg/m² on Days 1, 8, and 15 of a 28-day cycle.7 Toxicities included bone marrow suppression, mucositis, nausea, and fatigue. PRs were observed in 7 of 39 evaluable patients (17%). Additional studies of combination gemcitabine/5-FU–based regimens have supported a 10–15% overall response rate.14–17

A Phase I study comprising 40 patients with refractory solid tumors investigated gemcitabine in combination with capecitabine.18 The capecitabine dose was fixed at 1660 mg/m² on Days 1–21 and gemcitabine was escalated to a maximum dose of 1000 mg/m², given as a standard 30-minute infusion on Days 1, 8, and 15 of the 28-day cycle. Dose-limiting toxicity (DLT) included myelosuppression and mucositis. Of note, only 1 patient experienced Grade 2 palmar-plantar erythrodysesthesia (hand-foot syndrome). The recommended Phase II doses were gemcitabine 1000 mg/m² and capecitabine 1660 mg/m².

To enhance the cytotoxic activity of gemcitabine, an alternative infusion regimen has been explored. As the active metabolite of gemcitabine, dFdCTP has a long intracellular half-life, a fixed dose rate infusion of 10 mg/m² per minute, and has been shown to lead to maximal intracellular accumulation.19, 20 The potential clinical effect of this prolonged infusion was investigated in a randomized Phase II trial of patients with pancreatic carcinoma.21 Patients were randomized to receive either 2200 mg/m² of gemcitabine over a standard 30-minute infusion or 1500 mg/m² of gemcitabine infused at a rate of 10 mg/m² per minute (fixed dose rate) on Days 1, 8, and 15 of a 28-day cycle. Significantly increased accumulation of dFdCTP in mononuclear cells was demonstrated in the fixed dose rate cohort. In addition, patients randomized to the fixed dose rate infusion demonstrated a superior overall survival (5.0 months vs. 8.0 months; P = 0.013), although the time to disease progression did not significantly differ between the cohorts. Grade 3 and 4 toxicity, predominantly hematologic, were observed with increased frequency in the fixed dose rate arm.

Based on the established modest response rate of gemcitabine and 5-FU in RCC and the potentially enhanced antitumor effect utilizing fixed dose rate infusion of gemcitabine, a Phase I study was conducted to determine the maximum tolerated dose (MTD) of fixed dose rate gemcitabine and capecitabine in patients with advanced RCC.

MATERIALS AND METHODS

Patient Selection

Eligible patients had histologically or cytologically confirmed Stage IV RCC of any histologic subtype. Patients had measurable disease defined by response evaluation criteria in solid tumors (RECIST) criteria.22 A life expectancy > 3 months and an Eastern Cooperative Oncology Group performance status ≤ 2 were required. Patients had normal organ and bone marrow function as defined by the following characteristics: leukocyte count ≥ 3000 cells per microliter, absolute neutrophil count ≥ 1500 cells per microliter, platelet count ≥ 100,000 cells per microliter, total bilirubin level ≤ 1.5 × the institutional upper limit of normal (ULN), aspartate aminotransferase/alanine aminotransferase levels ≤ 2.5 × the institutional ULN, and creatinine clearance ≥ 60 mL/1.73 m² per minute (calculated by the Cockcroft–Gault formula).

Previous gemcitabine or fluoropyrimidine (capecitabine, 5-FU, or fluorodeoxyuridine) for the treatment of metastatic RCC was not permitted. There was neither an exclusion of patients based on the number of previous therapies received nor were patients required to have received any previous systemic therapy for RCC. Patients who had received previous systemic therapy for RCC, surgery, or radiotherapy within 4 weeks before entering the study or those who had not recovered from adverse events due to previously administered agents were excluded. Concurrent antineoplastic therapy including investigational agents was prohibited. Other exclusion criteria were any history of significant gastric or small bowel resection, malabsorption syndrome or other lack of integrity of the upper gastrointestinal tract that might compromise absorption of capecitabine, previous severe reaction to fluoropyrimidine therapy, or known hypersensitivity to 5-FU. Informed consent approved by the University of California San Francisco Committee on Human Research was obtained from all patients.

Treatment

Oral capecitabine was administered twice daily at a dose of 1660 mg/m² per day (830 mg/m² twice per day) on Days 1–21 of a 28-day cycle. Gemcitabine was given as an intravenous infusion at a fixed dose rate of 10 mg/m² per minute on Days 1, 8, and 15 of a 28-day cycle (Fig. 1). The starting dose of gemcitabine was 600 mg/m², which was substantially lower than the dose in the previous Phase I trial with standard infusion rate gemcitabine.18 Treatment was administered on an outpatient basis. Gemcitabine doses were to be increased in increments of 200 mg/m² per week in a maximum of 6 cohorts of ≥ 3 new patients each until the MTD was established, not to exceed 1600 mg/m². No modification of the capecitabine dose was planned initially and no intrapatient dose escalation was permitted.

Figure 1. Study schema.

Determination of the Maximum Tolerated Dose

Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria, Version 2.0. DLTs were defined as any of the following events attributable to therapy: Grade 4 neutropenia lasting ≥ 7 days; Grade 3/4 neutropenia associated with fever ≥ 38.1 °C; Grade 4 thrombocytopenia lasting ≥ 7 days; Grade 3 or 4 nonhematologic toxicity except alopecia and hand-foot syndrome; Grade 3 or 4 nausea, emesis, or mucositis not reduced to Grade 1 within 2 weeks with maximal supportive therapy; Grade 3 or 4 diarrhea or a second occurrence of Grade 2 diarrhea not reduced to Grade 1 within 2 weeks with maximal supportive therapy; ≥ Grade 2 hand-foot syndrome not reduced to Grade 1 before the start of Cycle 2; the inability to administer 2 successive doses of gemcitabine within the first treatment cycle; or a delay of ≥ 14 days in initiating the second cycle of therapy due to persistent toxicity of ≥ Grade 2.

The MTD was defined by the occurrence of DLTs during the first cycle of therapy. If none of the three patients in a given gemcitabine dose cohort experienced a DLT, the next cohort would receive the next higher dose level of gemcitabine. If two or more of three patients experienced a DLT, then the MTD had been exceeded, dose escalation would stop, and the previous gemcitabine dose level would be considered to be the MTD. If one of three patients developed a DLT, then three additional patients would be enrolled and treated at the same dose level. If only one of the six total patients at this dose level experienced a DLT, then the dose escalation would continue as planned. If two or more of the six total patients at this dose level experienced a DLT, the MTD had been exceeded and the previous gemcitabine dose level would be considered to be the MTD.

Response Assessment

Computed tomography scans for disease assessment were performed at baseline and after every two cycles. A bone scan was performed at baseline for all patients and repeated after every two cycles only if indicative of metastases at baseline or if signs or symptoms of bone metastases developed. Disease response and progression was defined using the RECIST criteria.22 Patients continued to receive therapy until there was disease progression or unacceptable toxicity.

Statistical Considerations

The primary end point of this Phase I trial was to determine the MTD of prolonged infusion of gemcitabine in combination with capecitabine that would result in a < 33% DLT incidence in patients with metastatic RCC. Three to 6 patients per cohort were to be accrued, to a maximum of 36 patients.

RESULTS

Patient Characteristics

Nine patients were enrolled in this Phase I study. Patient characteristics are depicted in Table 1. This patient cohort was typical of an RCC population with a preponderance of lung and lymph node metastases. The majority of patients had three or more metastatic sites and had received two or more previous systemic treatment regimens for metastatic RCC.

Table 1. Patient Characteristics (n = 9)

Characteristics	No. of patients (%)
ECOG, Eastern Cooperative Oncology Group; IL-2: interleukin-2; IFN: interferon.
Median age (yrs) (range)	51 (26–73)
Gender
Male	6 (67)
Female	3 (33)
ECOG performance status
0	7 (78)
1	2 (22)
Site of metastases
Lung	9 (100)
Lymph node	8 (89)
Liver	2 (22)
Central nervous system	3 (33)
Bone	2 (22)
Other	4 (44)
No. of metastatic sites
1	0 (0)
2	2 (22)
≥ 3	7 (78)
Median time since diagnosis of metastatic disease to study entry (range)	40 (1–111 mos)
Previous therapy
Nephrectomy	7 (78)
Radiotherapy (any)	3 (33)
Immunotherapy (IL-2/IFN-based)	7 (78)
Investigational	5 (56)
No previous systemic treatment	2 (22)
One previous treatment	2 (22)
Two or more previous treatments	5 (56)
Histologic subtype
Clear cell	8 (89)
Papillary	1 (11)

Toxicity

The initial dose level (gemcitabine 600 mg/m² by fixed dose rate infusion on Days 1, 8, and 15 and capecitabine 1660 mg/m² on Days 1–21 of a 28-day cycle) resulted in DLT in 2 of 4 total patients. One of the initial three patients experienced a DLT, requiring cohort expansion to six patients. The fourth patient enrolled in this initial cohort also experienced a DLT. Thus, by protocol definition, the MTD had been exceeded. Both DLTs were defined by hand-foot syndrome (1 patient with Grade 3 and 1 patient with Grade 2 that did not resolve to ≤ Grade 1 by the planned start of Cycle 2). An additional patient in this cohort experienced Grade 3 hand-foot syndrome that resolved by the planned start of Cycle 2, and thus did not meet the protocol definition for a DLT. Pyridoxine use at doses of 50–100 mg per day by these patients after experiencing hand-foot syndrome did not ameliorate symptoms. The 2 patients who experienced DLTs continued off-protocol chemotherapy for an additional 11 and 15 months, respectively, with the same dose and schedule of fixed dose rate gemcitabine in combination with a capecitabine dose reduced by 50% to 830 mg/m² on Days 1–21 of a 28-day cycle. No further hand-foot syndrome or other notable toxicity was observed in these two patients, and each patient demonstrated an objective PR before subsequently experiencing disease progression. The other two patients in the initial dose cohort (including the patient who experienced the non-DLT hand-foot syndrome) demonstrated disease progression after one cycle of therapy, precluding further protocol treatment.

The protocol was thus modified to investigate a new starting dose level of capecitabine at 50% of the initial dose (830 mg/m²) administered on Days 1–21 and gemcitabine 600 mg/m² by fixed dose rate infusion on Days 1, 8, and 15 of a 28-day cycle. Dose escalation would then proceed for both gemcitabine and capecitabine according to the established DLT/MTD rules. At this modified dose level, two of five patients experienced DLTs. One of the initial 3 patients experienced a DLT defined by Grade 3 neutropenia with fever, and thus cohort expansion to 6 patients was required. The fifth patient in the second dose cohort experienced a DLT defined by a delay of ≥ 14 days in initiating the second cycle of therapy due to persistent toxicity (infection) of ≥ Grade 2. The MTD for this combination had again been exceeded. No hand-foot syndrome was observed in the second cohort. Given the inability to safely administer this combination chemotherapy, accrual to the protocol was halted. Table 2 lists the treatment-related toxicity for the entire study cohort for all cycles of therapy. In addition to the toxicity noted above, hematologic toxicity was prominent. Toxicity by chemotherapy dose level is listed in Table 3.

Table 2. Treatment-Related Toxicity

Toxicitya	No. of patients (n = 9) (%)
	Grade 2	Grade 3	Grade 4
AST: aspartate aminotransferase; ALT: alanine aminotransferase. a Toxicity was graded according to the second version of the Common Toxicity Criteria of the National Cancer Institute.
Hematologic
Anemia	5 (56)
Neutropenia	4 (44)	2 (22)
Leukopenia	6 (67)	1 (11)
Thrombocytopenia	1 (11)	1 (11)
Nonhematologic
Hand and foot syndrome	1 (11)	2 (22)
Diarrhea	3 (33)
Stomatitis	2 (22)
Fatigue	2 (22)	1 (11)
Cough	1 (11)
Pain	1 (11)
Infection	1 (11)	2 (22)
Fever	1 (11)
Elevated AST	1 (11)
Elevated ALT	1 (11)

Table 3. Chemotherapy Dose Levels and Outcome

Dose level	Gemcitabine Days 1, 8, and 15	Capecitabine Days 1–21	No. of patients	Result
DLT: dose-limiting toxicity.
Starting dose level	600 mg/m^2	1660 mg/m^2 per day	4	2 DLTs; 3 patients with hand-foot syndrome
Modified dose level	600 mg/m^2	830 mg/m^2 per day	5	2 DLTs: Grade 3 neutropenia with fever and persistent Grade 2 infection

Three patients with central nervous system (CNS) metastases were enrolled in the current study cohort. All had received previous therapy for CNS metastases. The first patient underwent surgical resection of a solitary parietal metastasis followed by prophylactic stereotactic radiosurgery (1800 centigrays [cGy] to 50% isodose) to the tumor bed. The second patient underwent whole-brain radiotherapy (3000 cGy) for multiple metastases. The third patient underwent surgical resection of 2 metastases followed by prophylactic whole-brain radiotherapy (5100 cGy). No CNS-related toxicity was observed in these or any patients.

Efficacy

Although toxicity assessment was the primary end point of the current Phase I study, patients were assessed radiographically at baseline and at every two cycles for disease response or progression. The best response observed included two PRs as defined by RECIST criteria.22 Both patients who experienced a PR demonstrated a DLT at the initial dose level (vide supra) but continued to receive chemotherapy off-protocol and demonstrated the objective response after being removed from the protocol. The best response of stable disease was observed in 5 patients (56%) and 2 patients (22%) demonstrated progressive disease prior to the end of Cycle 2.

DISCUSSION

RCC remains a treatment-resistant malignancy. Limited durable complete response rates observed with IL-2 and IFN-α result in the majority of cytokine-treated patients experiencing disease progression.1, 23 Chemotherapy has been largely unsuccessful in RCC, although modest antitumor activity has been observed with combination 5-FU and gemcitabine. In an attempt to improve these modest response rates, we investigated capecitabine in combination with gemcitabine, administered by a fixed dose rate infusion based on the potential augmented clinical effect. Our results demonstrate that this schedule of prolonged infusion of gemcitabine in combination with capecitabine in metastatic RCC produces unacceptable toxicity.

Our results are somewhat surprising, given previous experience with the combination of gemcitabine and fluoropyrimidines, as well as with fixed dose rate gemcitabine. Fixed dose rate gemcitabine as a single agent has been given in doses ≤ 1500 mg/m²,21 with the notable toxicity of myelosuppression. Fixed dose rate gemcitabine also has been given in doses ≤ 1000 mg/m² in combination with other chemotherapy (cisplatin) in nonsmall cell lung carcinoma with acceptable toxicity.24 Previous experience with standard rate infusion gemcitabine in combination with fluoropyrimidines demonstrated acceptable toxicity employing higher doses than used in the current study.18, 25 It is interesting to note that hand-foot syndrome was not a prominent toxicity observed in any of these studies. Thus, the fixed dose rate infusion of gemcitabine appeared to enhance the hand-foot syndrome toxicity of capecitabine. The mechanism of hand-foot syndrome is not fully understood. It is remains unknown why fixed dose rate gemcitabine should enhance this fluoropyrimidine toxicity. No pharmacokinetic parameters were measured in the current study to determine drug levels resulting from this combination therapy.

The current study has several limitations. Initial chemotherapy doses were chosen based on the demonstrated effects in metastatic RCC of 21-day 5-FU dosing7 and the Phase I study of capecitabine and standard infusion gemcitabine.18 Clearly, prolonged infusion of gemcitabine has a different toxicity profile, and starting doses in the current study resulted in unacceptable toxicity. Modification of the capecitabine dose produced different, but still unacceptable, toxicity. Further dose modification of one or both agents may have allowed for a treatment regimen with an acceptable safety profile, but this would have resulted in chemotherapy doses believed to be below those required for an antitumor effect, based on previous studies of these agents in RCC. No blood or tissue samples were obtained in the current study, precluding further characterization of the toxicity observed.

The further investigation of fixed dose rate gemcitabine in combination with capecitabine cannot be recommended for further development in RCC. However, given the modest but real response rates observed in previous studies, fluoropyrimidines in combination with standard infusion gemcitabine remain a viable therapeutic option for the treatment of metastatic RCC. Further, such therapy can be administered safely to RCC patients with treated CNS metastases.