Response to Fixed-Dose Capecitabine in Advanced Colorectal Cancer Is Comparable to Body Surface Area Calculated Dosing

SYDNEY, Australia-Fixed-dosecapecitabine (Xeloda) in advanced colorectalcancer patients allows ease of dosingwith an acceptable toxicity profile anda response rate comparable to body surfacearea dosing. Furthermore, an assessmentof serum thymidine and folate concentrationsas predictors of toxicitydetermined that a high pretreatment concentrationof serum folate is predictive ofgreater toxicity according to Rohini Sharma,MD, of Sydney Cancer Centre, Australia(ASCO abstract 573).The recommended dose of capecitabine(1,250 mg/m² bid for 14 days of a 21day cycle) produces an overall responserate of 25.7% in metastatic colorectal cancer,compared to a response rate of 16.7%for fluorouracil 425 mg/m² plus leucovorin20 mg/m² days 1 to 5 every 4 weeks.As Dr. Sharma pointed out, however,"There is a lack of evidence for dose selectionof capecitabine according to bodysurface area. Therefore, we studied theactivity and toxicity profile of capecitabinegiven as a fixed dose of 2,000 mg bid.We also assessed potential predictors oftoxicity."Pretreatment folate status correlatesto the development of toxicity in patientstreated with antifolates, but Dr. Sharmasaid that it is unknown whether this relationshipalso holds for capecitabine. Theinvestigators evaluated the correlationbetween folate status and cytotoxicity andalso assessed homocysteine and methylmalonicacid (MMA), sensitive surrogatemarkers for folate and B12 status.Lower Median DoseEligibility criteria for the study includedlocally advanced or metastatic colorectalcancer with measurable or evaluabledisease, Karnofsky performance status 0to 2, fewer than two prior chemotherapyregimens, and a life expectancy greaterthan 12 weeks.Patients were treated with capecitabine2,000 mg bid for 14 days followed bya 1-week break. Dose reduction by 1,000mg per day was planned for any grade 2 or3 toxicities.At the time of analysis, patients hadcompleted a median number of five cyclesof capecitabine for a median durationof therapy of 80.4 days. The median bodysurface area calculated dose was 63,000mg. Therefore, patients received a median dose that was 11% lower than the bodysurface area calculated dose.Of the 51 patients enrolled in the study,13 were excluded because they had beendosed in a standard fashion. An objectivetumor response was seen in seven of the33 patients (21%) assessable for response.Fixed-Dose Well ToleratedDr. Sharma said, "Fixed-dose capecitabinewas tolerated well with no grade3/4 nonhematologic or hematologic toxicitiesrecorded." A single dose reductionbecause of toxicity was required in 10patients (26%). Multiple reductions wererequired in one patient (3%). Toxicityresulted in treatment delays in 13 patients(34%). Two patients (5%) ceased treatmentbecause of toxicity.In terms of the assessment of possiblepredictors of toxicity, a high pretreatmentserum folate was associated with increasedtoxicity (P = .004). No significant associationwas found between serum homocysteineor thymidine levels and toxicity.'Particularly Attractive'"Fixed dosing is particularly attractivebecause it is difficult to split tablets. Inaddition, a BSA-calculated dose is difficultto precisely determine, and there islittle evidence supporting such an approach,"Dr. Sharma summarized. "Ourstudy confirmed that fixed dosing has acomparable response rate to BSA dosingin patients with metastatic colorectal cancerand, importantly, this type of dosingwas well tolerated. Fixed-dose capecitabineallows ease of dosing with an acceptabletoxicity profile and a response ratesimilar to that found in phase III trials,"she continued."The second part of our study consideredthe role of serum markers as predictorsfor toxicity. A nonsustained fall inthymidine was noted during treatment,indicative of thymidylate synthase (TS)inhibition. A high pretreatment concentrationof serum folate is predictive ofgreater toxicity. This finding has not beenpreviously reported and we can only hypothesizethat the basis for this may beincreased suppression of TS and thereforeincreased toxicity," she added."Finally," Dr. Sharma said, "our studyaddresses the role of TS genotype ina Caucasian and Chinese population andstudies whether differences in toxicitycan be predicted by differences in TSgenotype. The results of this question arestill pending and are awaited with greatinterest."