Scan of 977 nonsynonymous SNPs in CLL4 trial patients for the identification of genetic variants influencing prognosis

January 26, 2008

Blood, 1 February 2008, Vol. 111, No. 3, pp. 1625-1633.

Prepublished online as a Blood First Edition Paper on November 15, 2007; DOI

10.1182/blood-2007-08-110130.

Scan of 977 nonsynonymous SNPs in CLL4 trial patients for the identification of

genetic variants influencing prognosis

le S. Sellick1, Wade2, s2, G. Oscier3,

Catovsky4, and S. Houlston1

1 Section of Cancer Genetics, Institute of Cancer Research, Sutton; 2 Clinical

Trial Service Unit, University of Oxford, Oxford; 3 Department of Haematology,

Royal Bournemouth Hospital, Bournemouth; 4 Section of Haemato-Oncology,

Institute of Cancer Research, Sutton, United Kingdom

To identify genetic variants associated with outcome from chronic lymphocytic

leukemia (CLL), we genotyped 977 nonsynonymous single nucleotide polymorphisms

(nsSNPs) in 755 genes with relevance to cancer biology in 425 patients

participating in a phase 3 trial comparing the efficacy of fludarabine,

chlorambucil, and fludarabine with cyclophosphamide as first-line treatment.

Selection of nsSNPs was biased toward those likely to be functionally

deleterious. SNP genotypes were linked to individual patient outcome data and

response to chemotherapy. The effect of genotype on progression-free survival

(PFS) and overall survival (OS) was assessed by regression analysis

adjusting for treatment and clinico-pathologic variables. A total of 78 SNPs (51

dominantly acting and a further 27 recessively acting) were associated with PFS

(9 also affecting OS) at the 5% level. These included SNPs mapping to the

immune-regulation genes IL16 P434S (P = .03), IL19 S213F (P = .001), LILRA4 P27L

(P = .004), KLRC4 S29I (P = .007), and CD5 V471A (P = .002); and DNA response

genes POLB P242R (P = .04) and TOPBP1 S730L (P = .02), which were all

independently prognostic of immunoglobulin heavy-chain variable region (IgVH)

mutational status. The variants identified warrant further evaluation as

promising prognostic markers of patient outcome. To facilitate the

identification of prognostic markers through pooled analyses, we have made all

data from our analysis publicly available.

Recommended Posts