A dose escalation study of docetaxel plus capecitabine in combination with oxaliplatin in patients with advanced solid tumors

Abstract

Objectives. Capecitabine (CAP), Oxaliplatin (OX) and Docetaxel (DOC) have shown considerable activity in a wide range of solid tumors. A phase I study was conducted in order to determine the maximum-tolerated dose (MTD) and dose-limiting toxicities (DLTs) of their combination in patients with advanced solid tumors. Patients and methods. Twenty-one patients were enrolled. The patient's median age was 68 years, 15 were male, and 12 were chemo-naïve. DOC was administered on day 1 as an 1-hour (iv) infusion at a standard dose of 50 mg/m². OX was administered on day 1 as a 2-hour (iv) infusion at escalating doses ranging from 70–80 mg/m². CAP was administered orally on days 1 to 7 at escalating doses ranging from 2 000–2 750 mg/m² given as two daily divided doses. Treatment was repeated every two weeks. Results. Six different dose-levels were examined. At dose-level VI, two of three enrolled patients presented DLTs (one patient diarrhea and asthenia grade 3 and another grade 3 diarrhea), and thus, the recommended MTD for future phase II studies is CAP 2 750 mg/m² , DOC 50 mg/m² and OX 75 mg/m². A total of 121 treatment cycles were administered. Grade 3 neutropenia was observed in six (5%) treatment cycles and grade 3 thrombocytopenia in one (1%). There was no febrile episode. Grade 3 asthenia was observed in three (14%) patients, grade 3 diarrhea in four (19%), grade 3 neuropathy in one (5%), and grade 1/2 hand-foot syndrome in three (14%). Other toxicities were uncommon. There was no treatment related death. Four (29%) PRs and seven (50%) SD were observed among 14 evaluable patients. Responses were seen in patients with renal (n = 1), gastric (n = 2) and pancreatic (n = 1) cancer. Conclusions. These results demonstrate that CAP, DOC and OX can be safely combined at clinically relevant doses and this regimen merits further evaluation.

Docetaxel (DOC, Taxotere™, Sanofi-Aventis, France), oxaliplatin (Eloxatin™, Sanofi-Aventis, France), and capecitabine (Xeloda™, Hoffmann-La Roche, Nutley, NJ) are three widely used chemotherapeutic agents with different mechanisms of action. Capecitabine (CAP), a fluoropyrimidine carbamate, was designed as an orally active drug that would deliver fluorouracil (5-FU) selectively to the tumor cells [1]. CAP has demonstrated a high level of activity in breast and colorectal cancer [2–4]. Furthermore, CAP has a favorable safety profile that is typical of infused fluoropyrimidines. Dose limiting toxicities (DLTs) when given as monotherapy consist of diarrhea and infrequent leucopenia at higher dose levels. Palmarplantar erythrodysesthesia (PPE) after prolonged treatment with CAP is not uncommon [5].

Oxaliplatin (OX), a platinum derivative, is characterized by a 1, 2-diaminocycloexane carrier ligand that offers important advantages in terms of activity and toxicity compared with cisplatin and carboplatin. OX lacks the nephrotoxicity of cisplatin and the myelosupression of carboplatin and may be active in tumors with intrinsic or acquired resistance to cisplatin [6]. In clinical practice, OX used alone or in combination with 5-fluorouracil and leucovorin has produced high response rates and long progression-free survival in patients with advanced colorectal cancer [7–9]. The principal toxicity of OX is transient peripheral neuropathy, which is manifested as paresthesia and dysesthesia in the extremities and is triggered or enhanced by exposure to cold [10]. Hematologic toxicity consists of dose-related grade 1 or 2 neutropenia or thrombocytopenia [11]. Other toxicities are mild and uncommon.

DOC and CAP appear to be a good combination regimen since these two active agents overlaps little in terms of their key side effects [12]. In addition, DOC further upregulates thymidine phosphorylase in tumor tissues acting synergistically with CAP in breast cancer xenograft models [13]. The addition of CAP to DOC has resulted in improved response rates in comparison with either agent alone in gastric, lung and breast cancer [14–16]. Similarly, the combination of OX with CAP seems to be particularly attractive. This seems to be true not only because of the experimentally and clinically confirmed supra-additive effect of OX with CAP [17], but also because of the simplicity of concomitant administration without the need of indwelling vascular devices or pumps, and the different relatively mild toxicity profile of each drug. Preliminary results of a number of clinical phase I/II trials evaluating this combination in both pretreated and chemo-naive patients with advanced colorectal cancer have demonstrated feasibility and encouraging antitumor activity [18,19].

Based on the encouraging efficacy of various doublets between CAP, OX, and DOC we investigated the feasibility of combining these three agents in a phase I trial. Since intermittent rather than continuous administration of this drug may results in a toxicity reduction (by incorporation of drug-free intervals), we have decided to use a weekly administration schedule of CAP, instead of the more widely used 14-days schedule, in combination with a bimonthly OX plus DOC administration in patients with advanced neoplasms.

Patients and methods

Eligibility criteria

Patients with histologically or cytologically confirmed advanced solid tumors who had failed to respond to standard therapy or for which no standard or effective treatment was available were eligible for the trial. Eligibility criteria included age < 75 years, a life expectancy of at least three months, adequate P.S (WHO scale 0-2), adequate renal (creatinine clearance > 60 ml/min) and hepatic function (serum bilirubin < 1.5 the upper limit of the reference range; transaminases < 5 the upper limit of the reference range), adequate hematological function (hemoglobin > 10 gr/dl; neutrophil count > 1.5 × 10⁹/l; PLT > 100 × 10⁹/l) and adequate cardiac function (LVEF > 50% and no history of myocardial infarction in the past six months). Patients with brain metastases were eligible if they had been irradiated, the brain lesions were radiographically stable and clinical improvement was evident. Patients should be off previous anticancer therapy for at least four weeks (six weeks for nitrosureas or mitomycin C). The study excluded patients unable to receive oral medications, those with a history of allergic reaction to any of the study agents or peripheral neuropathy, or patients receiving other investigational agents. Patients were also excluded if they demonstrated evidence of any other medical problems severe enough to affect their compliance to the protocol. The study was approved by the Ethical and Scientific Committee of our Institution and written informed consent was obtained from all patients prior to patient screening and enrollment.

Treatment schedule

An open-label, single-center, non-randomized, dose-escalating phase I study was conducted. This was a triplet combination and, for safety reasons, we selected lower starting doses to those recommended for each one of the constituent agents in the literature based from previous experience in doublets. DOC was administered on day 1 as a 1-hour i.v. infusion at the standard dose of 50 mg/m². OX was given on day 1 diluted in 250 ml of D/W 5% as a 2-hour i.v. infusion, at escalating doses ranging from 70 to 80 mg/m². CAP was administered orally on days 1 to 7 at escalating doses ranging from 2 000 to 2 750 mg/m² given as two daily divided doses taken with 150–200 ml water and within 30 min after a meal (ideally 12 h apart). The latter drug was supplied as film-coated tablets in two dose strengths, 150 and 500 mg, which were not to be split. Compliance with the oral medication regimen was assessed by tablet counts at each clinical visit. To prevent PPE patients were advised to avoid tight clothing and shoes, steam and hot baths/showers, direct sunlight, friction-causing activities and excessive exercise [20]. All patients developing PPE were treated with pyridoxine 100 mg/d [21]. Patients were medicated with dexamethasone 8 mg oral q 12 hours × 3 doses, starting the night before the scheduled docetaxel dose. All patients were premedicated with ondansetron 16 mg or an equivalent 5-hydroxytryptamine-3 inhibitor. Cycles were repeated every two weeks if the absolute granulocyte count was > 1.5 × 10⁹/l and PLT > 100 × 10⁹/l and non-hematologic toxicity was resolved. DLT was defined on the first chemotherapy cycle by the occurrence of one of the following; grade 4 neutropenia or thrombocytopenia lasting more than five days; grade 3/4 febrile neutropenia (fever > 38.5 persisting for > 48 h), any grade ≥ 3 non-hematologic toxicity except from nausea/vomiting and alopecia; any treatment delay lasting more than one week at day 1 of the 2^nd treatment cycle because of unresolved toxicity. An initial cohort of three patients had to be treated at each dose-level, at a minimum of four completed cycles before including new patients at the next dose-level. This mode was followed for the early assessment of possible cumulative toxicities at the examined dose-level. Dose escalation to the next dose-level was allowed when no dose-limiting toxicity (DLT) for three patients or no more than one DLT for six patients was registered. If two or more patients developed DLTs, further dose-escalation was interrupted, and the previous dose-level was defined as the recommended dose for phase II testing. Patients who did not complete seven days of CAP administration during the first cycle of chemotherapy for reasons other than toxicity were replaced to determine DLTs and MTD. All patients presenting grade 3 neutropenia received rhGCSF. Cumulative toxicity was also recorded for all subsequent chemotherapy cycles at all dose-levels in all patients.

Dose modifications

Dose modifications were performed on the basis of toxicity. Administration of the agents was delayed until adequate hematologic recovery up to a maximum of two weeks. In case of grade 4 hematologic toxicity in individual patients, chemotherapeutic drug doses were reduced by 25% for subsequent courses. More specifically, the OX dose was reduced by 25% for subsequent courses in case of persistent (> 14 days) or temporary (7 to 14 days) painful paresthesia or functional impairment. In case of persistent painful paresthesia or functional impairment, or if a patient experienced any other severe neurotoxicity despite a 25% dose attenuation, OX was omitted in subsequent cycles. In general, the doses of both DOC and OX were reduced by 25% for all subsequent chemotherapy cycles in patients who experienced a second occurrence of a given grade 2 toxicity or any grade 3 and 4 neurologic toxicity. CAP administration was interrupted if patients developed > grade 2 diarrhea, stomatitis, nausea/vomiting until resolution of the toxicity to grade 1. Doses of CAP omitted for toxicity were not replaced or restored on resolution of the toxicity to grade 1 and subsequent treatment cycles were given at the appropriate dose adjustment. No dose reductions or interruptions were performed for anemia, or for any grade 1 toxicity. As a general rule, the daily dose of CAP was reduced by 25% at the first occurrence of grade 3 toxicity or the second occurrence of grade 2 toxicity. The daily dose was reduced by 50% at the first occurrence of grade 4 toxicity, the second occurrence of grade 3 toxicity and the third occurrence of grade 2 toxicity. CAP was discontinued at the second occurrence of grade 4 toxicity.

Patients’ evaluation

Baseline evaluations included the following: patient history, physical examination, chest x-rays, complete blood count (CBC) with differential and platelet counts, blood chemistry, electrocardiograph (ECG), computed tomography (CT scans) of the chest, abdomen and pelvis, while whole brain CT scans were performed when clinically indicated. CBCs with differential and platelet counts were performed twice weekly or daily in patients with severe myelosuppression; whole blood chemistry and physical examination were performed every two weeks. All adverse events were documented. WHO performance status was assessed every two weeks.

Patients were considered evaluable for response if they had measurable disease and completed at least four cycles of chemotherapy [22]. In the case of a complete response, patients received four additional cycles of chemotherapy after the criteria of complete response were first met. Patients with partial response and stable disease were treated for a maximum of 12 cycles. Patients were withdrawn from the study at any evidence of progressive disease. Tumor responses had to be confirmed four to six weeks after their initial documentation.

Results

From January 2004 to April 2006, 21 patients suffering from refractory tumors or tumors for which no established therapy existed were enrolled into the study. Their median age was 68 years and the majority (80%) of them had a good PS (WHO: 0). Fifteen of them were male and 12 had received no prior chemotherapy regimens. Patients’ characteristics at baseline are summarized in Table I.

Table I. Patients' characteristics.

Patients enrolled	21
Median age (range)	68 (44–75)
Gender
Male/female	15/6
Performance status (WHO)
0	17 (80%)
1	4 (20%)
Tumor types
Pancreatic cancer	4
Renal cancer	4
Gastric cancer	9
Soft tissue sarcoma	1
Colorectal cancer	1
Unknown primary site	1
Hepatocellularcancer	1
Previous chemotherapy regimens
None	12
1	6
2 or more	3

DLT's and MTD

Six different dose-levels were examined (Table II).

Table II. Dose levels examined (mg/m²).

Level	No Patients	Docetaxel mg/m^2	Capecitabine mg/m^2	Oxaliplatin mg/m^2
I	3	50	2 000	70
II	3	50	2 250	70
III	3	50	2 250	75
IV	3	50	2 500	75
V	6	50	2 750	75
VI	3	50	2 750	80

None of the 18 patients participating in dose-levels I to V experienced any DLT during the first cycle. At dose level I, one patient experienced grade 3 neutropenia on days 10 to 11, accompanied by fever lasting for six hours during the 1^st treatment cycle. However, this patient presented frequent febrile episodes before treatment initiation. In addition, the patient had liver involvement, the febrile episodes were easily managed with NSAIDS, there were no positive cultures, and therefore, this patient was not regarded as having a DLT according to the protocol requiring persistence of fever for > 48 h. At dose-level VI, two of three enrolled patients experienced DLTs; one patient developed diarrhea and asthenia of grade 3 and another patient grade 3 diarrhea. The former patient required hospitalization for hydration and after resolution of the toxicity refused further treatment. The other patient did not required hospitalization, recovered well within five days with standard loperamide treatment, and continued his therapy after appropriate dose adjustments. Therefore, the MTD had been reached and no additional patients entered this dose level. Three additional patients were enrolled at dose level V and since none of them developed a DLT by the first treatment cycle, this level (DOC 50 mg/m², OX 75 mg/m² and CAP 2 750 mg/m²) represents the recommended dosing for future phase II testing.

Compliance to the protocol

A total of 121 chemotherapy cycles were administered with a median of six cycles per patient and toxicity data are available from all patients and all cycles. Treatment was discontinued after the first cycle of treatment in only one patient because of grade 3 diarrhea and asthenia and patient's refusal to continue treatment. Twenty patients stopped the treatment after at least two cycles for the following reasons: completion of treatment (n = 7), PD (n = 7), non-hematologic toxicity (n = 4) and patient's decision (n = 2). Dose reductions in all three drugs due to neutropenia were required in seven (6%) cycles. The dose of DOC was reduced in six (5%) additional cycles due to grade 3 asthenia and that of CAP in nine (7%) cycles due to grade 3 diarrhea. OX dose was reduced in four (3%) cycles due to painful acral dysesthaesia. Nineteen (16%) cycles were delayed because of unresolved hematologic toxicity (n = 6), non-hematologic toxicity (n = 1), and other reasons not related to disease or treatment (n = 12). At the recommended dose-level V, one patient presented neutropenia grade 3 at treatment cycles 2 and 8 and another patient diarrhea grade 3 at cycle 4 and asthenia grade 3 at cycle 6. Both patients continued their treatment after appropriate dose-adjustments. The other four enrolled patients at dose-level V completed all the planned treatment cycles without requirement for dose-adjustments or treatment delays. Overall, the median interval between cycles was 15 days (range 14–29). The planned and administered doses per patient and dose level are given in Table III.

Table III. Planned and administered doses (mg/m²/week) per patient and dose level.

Patient	Level	Taxotere (mg/m^2/week)		Oxaliplatin (mg/m^2/week)		Capecitabine (mg/m^2/week)		DLT
		Planned	Admin.	Planned	Admin.	Planned	Admin.
1	I	25	25	35	35	1000	1000	-
2	I	25	21	35	29	1000	840	-
3	I	25	21	35	30	1000	850	-
4	II	25	21	35	33	1125	1057	-
5	II	25	25	35	32	1125	1060	-
6	II	25	19	35	33	1125	1068	-
7	III	25	25	37.5	37.5	1125	1125	-
8	III	25	23	37.5	35	1125	1046	-
9	III	25	25	37.5	37.5	1125	1050	-
10	IV	25	22	37.5	37.5	1250	1250	-
11	IV	25	25	37.5	31	1250	1250	-
12	IV	25	25	37.5	37.5	1250	1250	-
13	V	25	20	37.5	30	1375	1100	-
14	V	25	22	37.5	33	1375	1196	-
15	V	25	17	37.5	25	1375	921	-
16	V	25	21	37.5	32	1375	1168	-
17	V	25	20	37.5	34.5	1375	1265	-
18	V	25	25	37.5	33	1375	1375	-
19	VI	25	20	40	37	1375	1265	Gr 3 diarrhea and asthenia (1pt)
20	VI	25	25	40	40	1375	1375	-
21	VI	25	25	40	40	1375	1375	Gr 3 diarrhea (1pt)

Hematologic toxicity

Hematologic toxicity was generally mild. Grade 3 neutropenia occurred in 6 (5%) cycles. The granulocyte nadir usually occurred between days 9 and 12 and the median time for the recovery of neutrophils was five days (range 3–6). There was no significant difference when the median neutrophil nadir counts were compared at cycles 1 and 5 of treatment. Grade 2 anemia was observed in only one (1%) cycle. Grade 3 thrombocytopenia was recorded in one cycle (1%) and one patient lasting for eight days. Hematologic toxicity is summarized in Table IV.

Table IV. Hematologic toxicity per cycle and dose-level (n = 121).

Level	No pts.	Cycles	Neutropenia		Anemia		Thrombocytopenia
			Gr 1/2	Gr 3	Gr 1/2	Gr 3	Gr 1/2	Gr 3
I	3	30	1 (3%)	3 (10%)	-	-	-	-
II	3	19	6 (31%)	1 (5%)	-	-	-	-
III	3	15	-	1 (7%)	2 (13%)	-	-	1 (7%)
IV	3	13	-	-	-	-	1 (8%)	-
V	6	35	7 (20%)	1 (3%)	4 (11%)	-	-	-
VI	3	9	3 (33%)	-	-	-	-	-
Total	21	121	17 (14%)	6 (5%)	6 (5%)	-	1 (1%)	1 (1%)

Non-hematologic toxicity

During the entire treatment period (all cycles) a total of eight patients experienced grade 3 non-hematologic toxicities. These included four patients with grade 3 diarrhea [at cycles 1 (n = 2), 3 and 4], three patients with asthenia grade 3 (cycles 1, 4 and 6) and one patient with grade 3 neuropathy (cycle 9). Five additional patients developed neuropathy grade 2. In all cases this toxicity presented as acral paresthesias. Grade 2 nail disorders were recorded in three patients after prolonged DOC treatment. Mild infusion-related reactions, such as sudden onset of flushing, headache and chills were observed in two patients receiving DOC during the first treatment cycle. In the subsequent cycles such episodes were avoided by slowing the infusion rate. Non-hematologic toxicity is summarized in Table V.

Table V. Non-hematologic toxicity per patient and dose-level (n = 21).

Level	Pts	Nausea		Diarrhea		Asthenia		Mucositis		PPE		Neuropathy
		Gr 1/2	Gr 3	Gr 1/2	Gr 3	Gr 1/2	Gr 3	Gr 1/2	Gr 3	Gr 1/2	Gr 3	Gr 1/2	Gr 3
I	3	1 (33%)	-	-	-	2 (66%)	-	1 (33%)	-	3 (100%)	-	1 (33%)	1 (33%)
II	3	-	-	-	-	-	-	1 (33%)	-	-	-	2 (66%)	-
III	3	-	-	1 (33%)		1 (33%)	1 (33%)	1 (33%)	-	-	-	2 (66%)	-
IV	3	1 (33%)	-	-	1 (33%)	-	-	1 (33%)	-	-	-		-
V	6	4 (67%)	-	2 (33%)	1 (17%)	1 (17%)	1 (17%)		-	-	-	2 (33%)	-
VI	3	-	-	-	2 (67%)	-	1 (33%)		-	-	-		-
Total	21	6 (29%)	-	3 (14%)	4 (19%)	4 (19%)	3 (14%)	4 (19%)	-	3 (14%)	-	7 (33%)	1 (5%)

Response to treatment

At the time of analysis, 14 patients had measurable disease, had completed at least four treatment cycles and were evaluable for response. Partial response was seen in four (29%) patients, stable disease in seven (50%), and PD in three (21%). Partial responders were suffering from gastric (n = 2), renal (n = 1), and pancreatic (n = 1) cancer. Both patients with gastric cancer who responded had failed previous 5-FU based 1^st-line chemotherapy, and received the regimen as second-line treatment. The median duration of response was 4.2 months. The median survival for the entire group was 8.5 months.

Discussion

Combining chemotherapeutic agents with known antitumor activity and non-overlapping mechanisms of action may result in improved efficacy in the treatment of solid tumors. The combination of DOC with cisplatin and 5-FU has been studied as treatment in patients with head and neck and gastric cancer [23–25]. However, despite the reported high activity, this efficacy is counterbalanced by significant toxicity, mainly febrile neutropenia affecting 20–29% of patients [23–25]. More specifically, the incidence of febrile neutropenia in the large phase III V-325 study, in patients with gastroesophageal adenocarcinoma was 29% along with 82% grade 3–4 neutropenia [25]. The combination of OX, CAP, and DOC may be particularly interesting as these agents have a broad spectrum of activity in a variety of cancers. A number of Phase I and II trials have now been completed that evaluated various ways of combining bimonthly schedules of doublets between DOC, OX, and CAP in several solid tumors [18,19,26]. Based on preclinical data of an additive and/or synergistic effect between these agents , we sought to develop a combination of all three drugs.

In the present phase I study, we determined that the recommended dosing of the combination is CAP 2 750 mg/m² D1-7 given orally as two daily divided doses, OX 75 mg/m² D1, and DOC 50 mg/m² D1, administered every two weeks, without growth factor support. The regimen was generally well tolerated, with the main hematologic toxicity being grade 2/3 neutropenia which complicated only 16 (13%) treatment cycles. With the exception of one episode of grade 3 thrombocytopenia other hematologic toxicity was remarkably low. This low incidence of hematologic toxicity in the present study, compared to other studies using similar drug combinations, could be interpreted by the replacement of cisplatin with oxaliplatin and the weekly instead of the more widely used 14-days administration schedule of capecitabine. Additionally, the every two weeks mode of administration permits an early evaluation of toxicity before proceeding to the next cycle. Similarly, the non-hematologic side effect profile of the combination was generally manageable and consistent with the known toxicities of the individual agents. Grade 2 PPE was seen in only one patient. This often painful, paraesthetic erythema of the palmar and plantar skin is reported to occur in up to 52% (all grades) of patients treated with single-agent CAP (1 250 mg/m² bid) [14]. This low incidence of clinically significant PPE observed could be attributed to the 7-days instead of the usual 14-days mode of CAP administration. Nausea was very mild and it never impaired oral intake of CAP. Twelve patients received more than six treatment cycles and no episode of cumulative toxicity was recorded. Grade 2/3 asthenia occurred in four (19%) patients. Neurotoxicity was also mild with only one patient experiencing grade 3 acral paresthesia after nine cycles of OX treatment.

Although response was not a primary end-point of the present study, the observed response rate of 29% with an additional 50% of patients experiencing disease stabilization is of interest for this poor-prognosis patient population with chemoresistant neoplasms. This finding is further supported by the study of Scheitauer et al. who tested two different schedules of an OX plus CAP combination in a randomized phase II trial and concluded that a weekly CAP plus bimonthly OX regimen is of similar toxicity and possibly more active to standard three-weekly regimen [18]. The responses were seen in patients with gastric (two PRs), renal (one PR) and pancreatic cancer (one PR). This activity encountered, may be due to the good performance status for the majority of our patients or to the preclinical observation of synergy, since significantly lower than single-agent doses of these agents still achieved a promising preliminary antitumor effect. Disease stabilization was also observed in seven (50%) patients, who, therefore, received multiple cycles of treatment. A Phase II study is now ongoing in patients with pancreatic and gastric cancer to further define the antitumor activity of this regimen.

In conclusion, the present study confirms the feasibility of combining DOC, OX and CAP in a bi-weekly intermittent schedule. The regimen was associated with acceptable toxicity and considerable activity. The recommended doses for future phase II studies consist of DOC 50 mg/m² day 1, OX 75 mg/m² day 1, and CAP 2 750 mg/m² orally bid days 1 to 7. Patients suffering from pancreatic and gastric cancer may be potential candidates for the regimen.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.