Fax: (313) 993-0559
Treatment of advanced colorectal carcinoma with oxaliplatin and capecitabine
A Phase II trial
Article first published online: 12 DEC 2003
Copyright © 2003 American Cancer Society
Volume 100, Issue 3, pages 531–537, 1 February 2004
How to Cite
Shields, A. F., Zalupski, M. M., Marshall, J. L. and Meropol, N. J. (2004), Treatment of advanced colorectal carcinoma with oxaliplatin and capecitabine. Cancer, 100: 531–537. doi: 10.1002/cncr.11925
- Issue published online: 20 JAN 2004
- Article first published online: 12 DEC 2003
- Manuscript Accepted: 21 OCT 2003
- Manuscript Revised: 20 OCT 2003
- Manuscript Received: 16 OCT 2003
- colon carcinoma;
- rectal carcinoma;
The current study was designed to evaluate the antitumor activity and toxicity of capecitabine and oxaliplatin in previously untreated patients with advanced colorectal carcinoma. The primary endpoint of the study was to determine the objective response rate, and a secondary endpoint was to measure the time to disease progression.
A 2-stage trial was planned with an accrual goal of 35 patients. The treatment included oxaliplatin given at a dose of 130 mg/m2 on Day 1 of each 3-week cycle. Initially, capecitabine at a dose of 2000 mg/m2/day in 2 divided doses was given on Days 1–14 of each cycle, but this was reduced to a dose of 1500 mg/m2/day because of toxicity. Patients were followed by computed tomography scans every two cycles to evaluate treatment response, and toxicity was monitored.
The first 13 patients on the trial received the higher dose of capecitabine. Although 5 responses (38.5%) were noted, 5 patients were hospitalized with diarrhea and dehydration. This toxicity led to a decrease in the dose of capecitabine to 1500 mg/m2/day and an additional 35 patients were treated. At the lower dose, the partial response rate was 37.1% (95% confidence interval [95% CI], 21.5–55.1%). The estimated median progression-free survival was 6.9 months (95% CI, 4.4–8.2 months). At the lower dose, four patients were hospitalized with diarrhea/dehydration (with one death reported), one with febrile neutropenia, and one with ventricular fibrillation. Overall, Grade (according to version 2.0 of the National Cancer Institute Common Toxicity Criteria) 3-4 diarrhea was reported to develop in 20% of those patients treated at the capecitabine dose of 1500 mg/m2/day compared with 62% of patients treated at the dose of 2000 mg/m2/day.
The combination of oxaliplatin and capecitabine is an active and convenient regimen for the treatment of patients with advanced colorectal carcinoma and should be compared with other front-line regimens as therapy for disease. Cancer 2004. © 2003 American Cancer Society.
Colorectal carcinoma remains a major cause of morbidity and mortality in the developed world. It is the second leading cause of cancer death in the U.S., with an estimated 57,100 deaths expected in 2003.1 Despite increased awareness and screening, many patients still present with or subsequently develop metastatic disease. Until recently 5-flourouracil (5-FU) was the only drug regularly used for the palliation of patients with metastatic colorectal carcinoma, and it has been reported to have a limited impact on survival.2 The introduction of irinotecan has led to an effective second-line treatment for those patients with disease that is refractory to 5-FU.3 The addition of irinotecan to 5-FU plus leucovorin (IFL) also has been reported to improve the survival of previously untreated patients with advanced disease.4, 5 Despite such improvements, the long-term prognosis of such patients remains poor. This has led to the search for alternative chemotherapeutic regimens for the treatment of patients with advanced colorectal carcinoma. The introduction of oxaliplatin has offered a promising new agent for the treatment of colorectal carcinoma. Although oxaliplatin is related to cisplatin, it has a different carrier ligand, which contributes to noncross-resistance. Unlike cisplatin, oxaliplatin has activity against colorectal carcinoma both in vitro and in vivo. As a single agent, it has demonstrable, but limited, activity.6 However, the combination of oxaliplatin plus 5-FU has shown significant benefit to patients as either first-line or second-line therapy.7 One of the more commonly used schedules developed by de Gramont et al. has combined oxaliplatin with 5-FU given by bolus and infusion over 2 days every 2 weeks (FOLFOX).8 In recently reported work, this regimen has been found to be more effective and less toxic than the weekly IFL regimen when given for first-line treatment of advanced colorectal carcinoma.7 In an effort to combine oxaliplatin with a fluoropyrimidine in a more convenient regimen, Phase I trials have combined oxaliplatin with capecitabine.9 This allows one to limit the number of intravenous treatments, the use of infusion pumps and central venous catheters, and outpatient clinic visits, making it potentially more attractive for patients. Based on this Phase I trial and the above considerations, we conducted a multicenter Phase II trial of capecitabine plus oxaliplatin in patients previously untreated for advanced colorectal carcinoma.
MATERIALS AND METHODS
Eligible patients had pathologically documented colorectal carcinoma that was not amenable to potentially curative resection. Patients could have had no prior chemotherapy for advanced disease, but prior adjuvant treatment was allowed if completed at least 6 months before treatment on the protocol. If patients had previously received a fluoropyrimidine as adjuvant treatment they could not have had severe toxicity. No prior therapy with irinotecan or oxaliplatin was allowed. Measurable disease (at least 2.0 cm × 1.0 cm) and an Eastern Cooperative Oncology Group (ECOG) performance status of 0–2 was required. Organ function requirements included creatinine ≤ 2.0 mg/dL, bilirubin ≤ 2.5 mg/dL, aspartate aminotransferase ≤ 4 times normal, an absolute neutrophil count ≥ 1500/mm3, and a platelet count ≥ 100,000/mm3. Patients could have no central nervous system metastases or other tumors within the previous 5 years, except for nonmelanoma skin cancer or in situ cervical carcinoma. Patients were excluded if they had a history of myocardial infarction within the prior 6 months or current evidence of congestive heart failure. Patients with an uncontrolled infection, those who were pregnant or lactating, or those with an organ allograft were excluded. The protocol was approved by the institutional review boards of the four participating centers and all patients were required to provide written informed consent prior to the initiation of treatment.
Patients received oxaliplatin at a dose of 130 mg/m2 as a 2-hour intravenous infusion on Day 1 of each 21-day cycle, 3–6 hours after the first dose of capecitabine. Capecitabine was given orally in 2 divided doses approximately 12 hours apart on Days 1–14. The trial began with a initial dose of capecitabine of 2000 mg/m2 on Days 1–14, but because of toxicity this was decreased to 1500 mg/m2/day. Capecitabine was to be taken within 30 minutes of eating. The total dose was rounded to the nearest 500 mg and divided into morning and evening doses. When an odd number of pills was to be used, the morning doses were higher.
Study Schedule and Evaluations
At baseline and before each treatment cycle, patients underwent a physical examination, complete blood count, and serum chemistries. Tumor measurements were obtained every other cycle. Patients could remain on treatment until disease progression or the development of unacceptable toxicity. The protocol was established as a 2-stage design with an interim analysis after enrolling the first 18 patients. We assumed we would not be interested in a response rate of ≤ 10% and would not wish to miss a response rate of 30%. We wanted a 90% power to detect the response rate of interest. To have a < 5% chance of continuing the trial if the response was unlikely to be > 10%, a cohort of 18 evaluable patients was to be treated and the trial would be terminated if there were ≤ 2 responses. If there were ≥ 3 partial responses, the trial was to continue to enroll an additional 17 patients for a total of 35 patients.
Tumor measurements were obtained using computed tomography scans in all patients and calculated as the sum of the products of the maximum perpendicular dimensions for measured lesions. A partial response required a 50% decrease in tumor measurements that persisted for at least 4 weeks. A complete response required the disappearance of all measurable lesions, and progressive disease required an increase of at least 25% in tumor measurements. Stable disease was defined as tumor measurements between those of a partial response and disease progression.
Toxicity and Dose Modifications
Toxicity was graded using the National Cancer Institute Common Toxicity Criteria, version 2.0 (available from URL: http://ctep.info.nih.gov/reporting/ctc.html), except for hand-foot syndrome and neurologic toxicity. Grade 1 hand-foot syndrome included numbness, dysesthesia, painless swelling, or erythema that did not disrupt normal activities. Grade 2 hand-foot syndrome was defined as painful erythema with swelling and discomfort that affected the activities of daily living. Grade 3 hand-foot syndrome included moist desquamation, ulceration, blistering, and severe pain, or symptoms that rendered the patient unable to work or perform activities of daily living. Grade 1 neurologic toxicity included paresthesias/dysesthesias of short duration. Grade 2 toxicity involved paresthesias/dysesthesias without functional impairment that persisted at the time of the next cycle, whereas Grade 3 toxicity included functional impairment.
Capecitabine dosing was not altered for Grade 1 toxicity. At the first appearance of Grade 2 toxicity, treatment was interrupted until it resolved to Grade 0–1. At the second or third appearance of Grade 2 toxicity, the capecitabine dose was reduced by 25% or 50%, respectively, from the starting dose. At the first or second appearance of Grade 3 toxicity, the dose was held until resolution and then decreased by 25% or 50%, respectively, from the starting dose. Once the dose had been decreased by 50% and a further Grade 2 or Grade 3 toxicity developed, that patient was removed from therapy permanently. At the first appearance of a Grade 4 toxicity, the patient was to discontinue therapy unless continuing was deemed by the physician to be in the patient's best interest, in which case the dose was reduced by 50%. Patients did not make up missed doses.
Oxaliplatin was reduced 25% for Grade 3 thrombocytopenia (platelet count ≥ 10,000–< 50,000/mm3) and 40% for Grade 4 thrombocytopenia. It also was reduced 25% for Grade 4 neutropenia (absolute neutrophil count < 500/mm3), Grade 4 mucositis, Grade 4 diarrhea, and Grade 4 emesis. The same reduction was employed for Grade 3 emesis that occurred despite optimal antiemetic prophylaxis. Oxaliplatin reductions were in addition to the reduction made in the capecitabine dose. The dose of oxaliplatin was decreased 25% for paresthesias persisting between cycles. Paresthesias with pain lasting > 7 days resulted in a 25% reduction during the first occurrence and a further 25% reduction during the second occurrence. The dose was reduced 50% for paresthesias with functional impairment. No dose reduction was made for cold-induced dysesthesia alone, even if it persisted between cycles.
A treatment cycle was delayed until the absolute neutrophil count was ≥ 1500/mm3 and the platelet count was ≥ 75,000/mm3. Other toxicities needed to return to baseline or Grade 1.
The median age of all 48 patients enrolled in the trial was 60 years (range, 28–79 years), with 28 men and 20 women (Table 1). Thirty-eight of the patients had colon carcinoma and 10 patients had rectal carcinoma; 45 of the patients had an ECOG performance status of 0 or 1. Eighteen patients had received prior adjuvant therapy. The liver was involved by metastatic disease in 33 patients and the lung was the site of metastatic disease in 15 patients.
|Patients included (n = 48)|
|2000 mg/m2 capecitabine||1500 mg/m2 capecitabine||Total|
|ECOG performance status|
|Prior adjuvant chemotherapy|
|Sites of metastasis|
|Time from advanced diagnosis to treatment (mos)|
Response to Therapy
Patients initially were treated with oxaliplatin at a dose of 130 mg/ m2 on Day 1 and capecitabine at a dose of 2000 mg/m2/day in 2 divided doses on Days 1–14 of each 3-week cycle. At this dose, only 13 patients were treated because of unacceptable toxicity. Of these 13 patients, 5 (38.5%) partial responses were reported. We subsequently treated 35 additional patients at a lower dose of capecitabine (1500 mg/m2/day). The partial response rate in this group was 37.1% (95% confidence interval [95% CI], 21.5–55.1%). The estimated median progression-free survival was 6.9 months (95% CI, 4.4–8.2 months) and the median survival had not been reached at the time of last follow-up (Fig. 1). Patients received a median of 6.5 cycles of treatment on the trial and were treated for a mean of 5.2 months (range, 0.7–15.9 months). The median duration of follow-up was 632 days and 484 days, respectively, for the high-dose and low-dose groups.
At the capecitabine dose of 2000 mg/m2/day, 7 of 13 patients were hospitalized a total of 10 times. Diarrhea/dehydration was the cause of hospitalization in five patients, one during the first cycle of treatment and four during the second cycle. The other hospitalizations were due to an elevated prothrombin time for a patient receiving warfarin, ureteral obstruction, deep vein thrombosis, and obstipation/constipation/hydronephrosis/partial small bowel obstruction (partial small bowel obstruction occurred twice in one patient). The hospitalizations were spread across all four medical centers. At the higher dose of capecitabine, a total of eight patients had Grade 3/4 diarrhea and four patients had Grade 3/4 dehydration (Table 2).
|Capecitabine, 2000 mg/m2/day (n = 13)|
|Prothrombin time prolonged||0||0||1||0||1|
|Capecitabine, 1500 mg/m2/day (n= 35)|
|Hand foot syndromeb||10||1||1||0||12|
|Prothrombin time prolonged||0||0||0||0||0|
At the lower capecitabine dose, 11 patients were hospitalized a total of 14 times for adverse events including 4 patients with diarrhea/dehydration, 3 patients with abdominal pain, 1 patient with febrile neutropenia, and 1 patient with ventricular fibrillation. One patient was hospitalized twice for small bowel obstruction attributed to the tumor and two patients were admitted for a pulmonary embolus. One patient was hospitalized for pneumonia and subsequently was one of the patients admitted for abdominal pain. One of the patients hospitalized with diarrhea later developed sepsis and died. Overall 7 patients (20%) developed Grade 3/4 diarrhea at the lower dose of capecitabine (4 of these patients were included in the hospitalized group). Two patients (5.7%) developed Grade 3/4 neutropenia, 1 of whom was hospitalized.
Of the patients without previous adjuvant therapy, 7 of 10 patients receiving the high dose and 3 of 20 patients receiving the low dose of capecitabine developed Grade 3/4 diarrhea. There was no clear difference with regard to the development of diarrhea between the men (8 of 28 patients) and women (7 of 20 patients) or those patients with rectal carcinoma (3 of 10 patients) and those with colon carcinoma (12 of 38 patients).
Compliance in taking the capecitabine was assessed by having the patients return with their bottles of pills and a diary at each visit. Although patients had to skip doses because of toxicity or other problems, we did not find unexpected differences between pill counts and the patient diaries. Overall, patients received 87% of their prescribed drug.
At the higher starting dose of capecitabine 7 of 13 patients had a total of 13 dose interruptions or reductions in capecitabine to ≤ 85% of the prescribed dose during the course of therapy with 6 interruptions or reductions occurring during Cycle 1 or beginning in Cycle 2. Overall, nine of the reductions were due to diarrhea and two were due to hand-foot syndrome. Three patients who were hospitalized with diarrhea during the second cycle did not receive further therapy and therefore were not categorized as having dose reductions. One patient with two previous dose reductions for diarrhea had recurrent diarrhea during the fourth treatment cycle and therapy subsequently was discontinued. At the lower dose of capecitabine, 13 of 35 patients required dose interruptions or reductions to ≤ 85% of the prescribed dose, with 3 interruptions or reductions occurring during Cycle 1 or at the start of Cycle 2. The reductions were due to diarrhea in five patients whereas four other patients with diarrhea were taken off the protocol.
Five patients and 16 patients, respectively, in the high-dose and low-dose treatment arms had reductions to ≤ 85% of the prescribed dose of oxaliplatin. This was due to neuropathy in all but six patients. Of the 48 patients enrolled, 18 developed Grade 2/3 neuropathy and 7 had treatment on protocol discontinued for this reason. Two patients underwent dose reductions for thrombocytopenia, whereas two additional patients had protocol treatment discontinued as a result of this toxicity.
Until a few years ago, 5-FU was the only standard agent used for the treatment of colorectal carcinoma. The most common regimens included 5-FU administered as a bolus or short infusion along with leucovorin, or as an infusion given for 24 hours once a week, or as a low-dose infusion given over many weeks.10, 11 All the regimens appeared to produce equivalent responses and survival rates.12 The infusion schedules did demonstrate improvements in toxicity, with less bone marrow suppression and mucositis. Conversely, palmar-plantar erythrodysesthesia (hand-foot syndrome) is commonly observed only with protracted infusion. Although this may be dose limiting, it rarely leads to hospitalization and readily resolves when therapy is discontinued or the dose is lowered.
The advent of capecitabine has offered patients a new oral therapeutic option. Capecitabine is a pro-drug that is converted to 5-FU through three enzymatic steps within the liver and/or tumor. It generally is administered twice a day for 2 of 3 weeks. Controlled trials have demonstrated that capecitabine produces survival rates equivalent to that obtained with the combination of 5-FU and leucovorin on the Mayo Clinic Regimen.13, 14 Although capecitabine was reported to produce a higher response rate than 5-FU, the lower toxicity of capecitabine appears to be of more importance. As is observed with protracted infusion of 5-FU, capecitabine also produces hand-foot syndrome. The convenience of an oral regimen is a significant advantage to patients.
The development of irinotecan initially provided a viable second-line option to patients failing treatment with a fluoropyrimidine.15 Irinotecan subsequently was combined with 5-FU and produced improved response rates and survival in previously untreated patients with advanced colorectal carcinoma.5 Oxaliplatin is another new agent that initially demonstrated responses in patients who did not respond to 5-FU. In combination with 5-FU, oxaliplatin (FOLFOX) has demonstrated response rates of 40–50%7, 8 in previously untreated patients. This regimen also has activity in patients failing the combination of irinotecan, 5-FU, and leucovorin. A recently reported trial demonstrated that FOLFOX had improved survival rates compared with bolus IFL and less toxicity.7 It is possible that this result was obtained because of different methods of 5-FU administration used in the study arms, as well as differences in crossover treatments after disease progression. Direct comparisons of irinotecan or oxaliplatin with infusional schedules of 5-FU appear to produce similar results.16
A recent trial reported administering oxaliplatin at a dose of 130 mg/m2 every 3 weeks with 5-FU given at a dose of 300 mg/m2/day and leucovorin at a dose of 20 mg/m2/day for 5 days.17 The partial response rate was 40% with a progression-free survival and overall survival of 5.9 months and 14 months, respectively. The results of this trial are similar to those of the current study, in which capecitabine replaced the 5-FU.
Although the FOLFOX regimen produces good response rates with acceptable toxicity, the schedule is demanding. Patients receive the oxaliplatin every 2 weeks followed by 5-FU, which is given in part as a 2-day infusion. This requires external infusion pumps, vascular access, additional medical care, and clinic visits. In an effort to provide a more simple regimen with a similar antitumor activity, investigators have combined oxaliplatin given every 3 weeks with a 2-week course of capecitabine.9 The initial Phase I trials demonstrated that oxaliplatin at a dose of 130 mg/m2 every 3 weeks and capecitabine at a dose of 2000 mg/m2 daily for 14 days could be delivered with acceptable toxicity.
To our knowledge, only 1 Phase II trial has been published to date using a treatment regimen of oxaliplatin at a dose of 130 mg/m2 every 3 weeks and capecitabine at a dose of 2500 mg/m2 daily for 14 days.18 This trial by Borner et al. included 43 previously untreated patients and 26 patients who had undergone previous fluoropyrimidine treatment. The response rate was 49% and 15%, respectively, in the previously untreated and treated patients. Grade 3/4 diarrhea was reported to occur in 35% and 50%, respectively, of these 2 cohorts. This result is comparable to our experience with Grade 3/4 diarrhea occurring in 20% of the patients treated at the lower dose of capecitabine (1500 mg/m2/day). More recently, a preliminary report of a European Phase II trial of oxaliplatin given at a dose of 130 mg/m2 every 3 weeks and capecitabine given at a dose of 2000 mg/m2 daily for 14 days was presented by Tabernero et al.19 Their study of 96 patients demonstrated an objective response rate of 55% (95% CI, 45–65%) and a median time to disease progression of 7.4 months (95% CI, 6.4–8.1 months). Grade 3/4 diarrhea was noted in only 15% of the patients.
Although the study of Tabernero et al.19 demonstrated a higher response rate than the current trial (55% vs. 37.1%), the time to disease progression was similar (7.4 months vs. 6.9 months). Despite the higher dose of capecitabine used in their European trial (2000 mg/m2/day vs. 1500 mg/m2/day), Tabernero et al. reported a slightly lower incidence of Grade 3/4 diarrhea (15% vs. 20%). The reason for the increased toxicity reported in the current study is unclear. There was one slight difference in the schedule: in the current study, capecitabine therapy was initiated in the morning 3–6 hours before the dose of oxaliplatin was given, whereas in the European trial capecitabine administration began on the evening after the oxaliplatin dose. We based our schedule on the desire to provide concurrent administration of both drugs, but it is unclear whether this difference in the schedule had an effect on toxicity. Analysis of differences in toxicity assessment or dose modification schemes between the studies must await final publication of the results of Tabernero et al.19 Given the limited number of patients involved in these trials, the reported toxicity rates also may have overlapping 95% CIs.
Neurotoxicity is a common side effect from oxaliplatin and in the current study of 48 patients, 14 patients and 4 patients, respectively, developed Grade 2 and Grade 3 neuropathy. Seven of the patients had treatment withdrawn because of this toxicity. Future studies may determine whether a lower starting dose of oxaliplatin is indicated or whether neuropathy can be mitigated by other techniques. For example, neuropathy might be decreased by slowing the oxaliplatin infusion or by adding infusions of calcium and/or magnesium. Such approaches may be of benefit if they do not impair the therapeutic response.
The results of the current study found the combination of capecitabine and oxaliplatin to be an efficacious, tolerable, and convenient treatment regimen. However, it should be compared with more established regimens such as FOLFOX, and expanded Phase III trials currently are being planned. The choice of the initial dose of capecitabine is problematic, given the differences between the current study and that of Tabernero et al.19 Based on the toxicity we observed in our initial cohort with capecitabine given at a dose of 2000 mg/m2/day, we suggest that patients treated with this regimen begin with a capecitabine dose of 1500 mg/m2/day for the first 2 cycles, with subsequent dose escalation for those patients with acceptable levels of toxicity. Real-time toxicity monitoring is encouraged in the next generation of large randomized trials.
The authors wish to thank Dr. Diane Harry for her assistance in data collection and biostatistics.
- 7N9741: oxaliplatin (oxal) or CPT-11 + 5-fluorouracil (5FU)/leucovorin (lv) or oxal + CPT-11 in advanced colorectal cancer. Updated efficacy and quality of life (qol) data from an intergroup study. Proc Am Soc Clin Oncol 2003; 22: 252., , , et al.
- 10Advanced Colorectal Cancer Meta-Analysis Project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: evidence in terms of response rate. J Clin Oncol. 1992; 10: 896–903.
- 16FOLFIRI followed by FOLFOX versus FOLFOX followed by FOLFIRI in metastatic colorectal cancer. Proc Am Soc Clin Oncol 2001; 20: 124a., , , et al.
- 19Capecitabine and oxaliplatin in combination as first line therapy for patients with metastatic colorectal cancer. Proc Am Soc Clin Oncol. 2002; 21: 133a., , , et al.