Molecular Vision 2007; 13:1740-5 <http://www.molvis.org/molvis/v13/a194/> Received 14 December 2006 | Accepted 17 August 2007 | Published 19 September 2007 RB (OMIM 180200) is the most frequent intraocular tu- mor of childhood and is caused by mutations in the RB1 gene. The predisposition to develop RB is inherited as an autosomal dominant trait but mutations in both alleles are necessary to cause the disease [1]. Mutations of RB1 gene are highly het- erogeneous and spread in promoter and exons 1-25 [2-6]. Pre- vious reports described a wide range of detection rate, from 20 to 80%, according to the case selection and the screening technique [2,3,6]. While more than 900 RB1 mutations (see Retinoblastoma Genetics and [7]) the rate of mutation detec- tion remains relatively low. This is attributable to the large size of the RB1 gene, to the significant mutational heteroge- neity of the disease, and to limitations of currently available screening techniques. Furthermore, unusual mutation location can also impede on mutation detection. Single base substitu- tions represent the most frequent mutations and among them, nonsense mutations predominate [3,4,6]. Studies on genotype- phenotype correlation concluded there was an association be- tween nonsense or frameshift mutations and severity of the disease defined as bilateral multifocal RB [3-5]. Meanwhile, variable expression of RB is well-known with description of unaffected carriers, unilateral RB or benign retinoma [8,9]. Thus, low-penetrant phenotype was associated with the p.R661W mutation of the RB1 gene [10], with 4-kb deletion spanning exons 24 and 25 [11], alternative splicing mutation in exon 21 [12], p. L662P [13], promoter mutations [14], in- frame deletions affecting the N-terminal region of pRB [15], or alternative translation initiation associated with nonsense mutations in exon 1 [16] among other molecular changes. Genetic modifying factors or residual protein function due to either missense mutation or alternative translation initiation may influence phenotype expression, particularly in low-pen- etrant RB. Studies of patients from different parts of the world can help explain the spectrum of RB1 mutations and thus improve detection rate. Risk prediction is mandatory for current RB management [5,17] and justifies the continuous search for RB1 novel mutations and for phenotypic correlations. To charac- terize the spectrum of RB1 mutations and to analyze geno- type-phenotype correlation, we performed a phenotype and mutation analysis of 65 patients with isolated or familial RB who underwent treatment in our institution. Ten patients with sporadic and familial bilateral RB in whom novel mutations were detected during a non-systematic mutation screening program occurring between 1995 and 1998 are reported sepa- rately. METHODS Between March 2004 and January 2006, we performed a mutational screening of the RB1 gene in 65 consecutive ©2007 Molecular Vision Ten novel RB1 gene mutations in patients with retinoblastoma Hana Abouzeid, 1 Francis L. Munier, 1,2 Francine Thonney, 3 Daniel F. Schorderet 2,4,5 1 Jules-Gonin Eye Hospital, Lausanne, Switzerland, 2 Department of Ophthalmology, University of Lausanne, Lausanne, 3 Service of Medical Genetics, Centre Hospitalier Universitaire Vaudois, Lausanne, 4 IRO-Institut de Recherche en Ophtalmologie, Sion, 5 EPFL- Ecole polytechnique fédérale de Lausanne, Lausanne, Switzerland Purpose: To study phenotype-genotype correlations in 65 retinoblastoma patients, who were seen between March 2004 and January 2006 and to report undescribed retinoblastoma 1 (RB1) mutations identified in ten additional patients in whom mutations were detected before 2004. Methods: Complete ophthalmic examinations were performed in all patients and their parents. DNA was extracted from peripheral blood leukocytes, and the RB1 gene was screened by denaturing high-performance liquid chromatography and direct sequencing of the promoter and all the exons. Results: Seven cases were familial, 30 were sporadic bilateral, and 28 were sporadic unilateral. Screening of the RB1 gene resulted in the identification of four mutations in the familial cases (57%), 22 in the sporadic bilateral cases (73%), and three in the sporadic unilateral cases (10.7%). Twenty-two mutations, were single-base substitutions (76%). Of these mutations, 68% were of the nonsense type (15 cases). Ten patients with bilateral retinoblastoma in whom ten mutations were detected in a non-systematic approach between 1995 and 1998 were added to our recent series. In total, ten novel mutations were identified, including four single base substitutions, four small deletions and two small duplications. These are g.39445G>A, g.41924A>G, g.56851A>G, g.156795T>G, g.41983delT, g.44699_44706delAGCAGTTC, g.73788_73789delAA, g.78253delA, g.2157dupC, and g.2179_2183dupGGACC. Two patients had dysmorphic features associated with 13q14 large deletions. Conclusions: The detection rates of 73% in the sporadic bilateral cases and of 10.7% in the sporadic unilateral cases in our series are in accordance with recently published literature. Our pattern of mutations confirms the predominantly gene- inactivating mutations, i.e. single-base non-sense mutations and splice site mutations. Correspondence to: Francis Munier, Hôpital Ophtalmique Jules- Gonin, 15, Av de France, 1004 Lausanne, Switzerland; Phone: +41 21 626 81 11; FAX: +41 21 626 88 88; email: francis.munier@ophtal.vd.ch 1740