Investigation of hand-foot syndrome (HFS) observed in pazopanib (P)-treated patients (pts) with renal cell carcinoma (RCC).

Abstract

3061 Background: HFS is an adverse-effect of several agents used to treat cancer, particularly the VEGF TKIs. HFS can be debilitating for patients, affecting their physical, mental, and social well-being. This study sought to explore demographic, pharmacokinetic (PK) and pharmacogenetic (PGx) factors which may contribute to pazopanib-related HFS in pts with RCC. Methods: Week (wk) 4 plasma P trough concentrations (Cmin) were determined for 204 out of 225 pts with advanced RCC from a phase II study (VEG102616). Logistic and ordinal regression analyses were used to evaluate the effect of demographic (gender, race, baseline weight and height) and wk 4 Cmin on occurrence and severity of HFS, respectively. For the PGx analyses, DNA samples were collected from consenting pts with advanced RCC from three P clinical studies; VEG102616 (phase II), VEG105192 (phase III), and VEG107769. Pts were defined as HFS cases (NCI-CTCv3 grade ≥ 1, n=33) or controls (grade = 0 with treatment duration >28 days, n=321). A total of 3012 SNPs in 89 candidate genes were evaluated. The penalized maximum likelihood logistic regression was used to evaluate the effect of genotypes on the HFS case-control status. Results: While the overall incidence was low (7.7%), a high proportion of east and south-east Asians exhibited HFS (21/71, 30%) compared to Caucasians (22/489, 4%). Wk4 Cmin (P = 0.0005) and race (P = 0.03) were significantly associated with occurrence of HFS by logistic regression. A similar correlation was observed between maximal HFS grades (CTC v3) and wk4 Cmin (P < 0.0001) and race (P = 0.0045). None of the genetic markers were significantly associated with HFS after correction for multiple comparisions. Five SNPs in ABCG2, PDGFRA, GZMB and TPCN2 genes showed suggestive association with HFS at a nominal significance level P<0.01 and with a plausible genetic model. Conclusions: Both race (Asian vs. Caucasian) and wk4 P Cmin concentrations are significant predictors of HFS. Candidate gene analyses revealed that some genetic markers may be associated with P- related HFS. However, as none of the markers passed the significance level of multiple testing corrections, further evaluation in independent data sets is required to confirm or refute these findings.