Hand-foot syndrome (HFS) in patients treated with capecitabine (CAP) and the role of thymidine phosphorylase (TP) and dihydropyrimidine dehydrogenase (DPD)

Abstract

8615 Background: HFS is the most common toxicity of CAP. Preclinical studies have shown that radiation (XRT) up-regulates TP, which may in turn increase efficacy of CAP. CAP is degraded by DPD, and a deficiency in this enzyme may increase toxicity of CAP. However, effect of XRT on frequency of HFS and association with TP and DPD has not been fully characterized. Methods: Toxicity data were collected from pts with LA pancreatic cancer enrolled in 3 clinical trials conducted at UAB between Apr 2001 and Jul 2005. Overall results of these trials have been reported elsewhere. Pts received XRT (50.4 Gy) with CAP (1,200–1,600 mg/m2 BID M-F) followed CAP (2,000 mg/m2 BID x 14 days). Pts were classified into 2 groups to evaluate HFS: CAP-XRT and CAP. Roche grading was used to assess HFS. Dose modifications were according to drug insert. Pts received prophylactic udder cream and pyridoxine. Tumor specimens were procured in 36 pts by EUS-FNA 1 wk pre- and 2 wks post XRT to evaluate TP and DPD. Age, race, sex, and PS were evaluated as prognostic factors. Results: Median duration of CAP was 6 wks (range: 3–6) for CAP-XRT and 2.5 cycles (range: 0–17) for CAP. Among 58 pts, 14 developed HFS (24%). CAP group had a higher incidence of HFS than CAP-XRT (17.2 % vs. 10.3 %; P = 0.12). Grade 2/3 HFS was observed in 15.5 % of CAP and 1.7 % of CAP-XRT (P = 0.0078). Median cumulative dose of CAP for first development of HFS was 235,000 mg/m2 in CAP-XRT group and 3,185,000 mg/m2 in CAP, with relative frequency of an event occurring in CAP-XRT vs. CAP of 0.59. HFS occurred at a median of 5 wks in CAP-XRT and 6 wks in CAP. Log-rank test showed neither age, sex, ECOG PS, or race was associated with development of HFS. There was no difference in tumor responses of pts with vs. without HFS. Mean tumor TP was higher among pts with vs. without HFS (275.77 vs. 215.29; P = 0.32). Mean tumor DPD was lower among pts with vs. without HFS (55.18 vs. 63.58; P = 0.49). Mean TP:DPD ratio was higher among pts with vs. without HFS (10.29 vs. 3.04; P = 0.31). Conclusions: This study suggests that incidence, severity, and time to occurrence of HFS with CAP-XRT < CAP, indicating no effect of XRT. No significant association of HFS with higher tumor TP or lower tumor DPD was found. Pharmacological basis for HFS with CAP needs to be explored. [Table: see text]