Can RB1 Mutation Analysis be Used to Predict Disease Expression in Patients With Heritable Retinoblastoma?

Abstract

Abstract: : Purpose: Although some evidence suggests that heritable retinoblastoma (hRB) phenotype varies according to the nature of the predisposing retinoblastoma gene (RB1) mutation, population–based genotype–phenotype correlation data are lacking. We report the first genotype–phenotype analysis in a large consecutive cohort, which expands the total number of reported RB1 mutations in hRB patients by 30%. Methods: Germline RB1 mutations were characterized in 123 hRB patients enrolled at University of California at San Francisco and University of Toronto. Mutations were detected by protein truncation testing, quantitative multiplex polymerase chain reaction, fluorescent in situ hybridization, and karyotype analyses. Medical records were reviewed and disease expression was categorized as low, moderate or high severity according to disease stage and age at presentation and subsequent clinical course. There were sufficient clinical data available to include 88 of 123 patients in the study. Results: Sixty–six of 88 patients (75%) demonstrated truncating mutations. Twenty–seven of 39 nonsense mutations (69%) were C to T transitions involving eight CGA (arginine) codons located between exons 8 and 18, upstream of the highly conserved B box. Six distinct missense mutations were detected in 7 of 88 patients (8%). Mean predicted mutant protein length differed significantly between patients with high and low severity disease; mean predicted mutant protein length was longer in patients with low severity disease (p<0.035). Missense mutations correlated exclusively with low or moderate severity disease (p<0.0004), and were all predicted to result in residual protein function. Four of 6 missense mutations detected have not been previously described, including Trp377Ser, Val654Leu, Gln702Arg, and Phe755Val. Conclusions: RB1 mutation analysis could potentially provide a means to stratify hRB patients for disease risk. This study provides the first population–based support for the "weak allele" hypothesis, which proposes that RB1 mutations resulting in residual protein function confer reduced disease risk. Accumulation of further well–described clinical cases coupled with RB1 mutation analysis is likely to yield correlations which could someday guide diagnostic and therapeutic management of patients.

Keywords: retinoblastoma • genetics • oncology