Proposal form for the evaluation of a genetic test for NHS Service Gene Dossier Test Disease Population Triad Disease name and description (please provide any alternative names Osteogenesis Imperfecta type 3 you wish listed) OMIM number for disease 259420 Gene name and description (please provide any alternative names you wish listed) OMIM number for Gene 120150, 120160 Mutational spectrum for which you test Technical Method (s) Validation Process Note please explain how this test has been validated for use in your laboratory) Are you providing this test already? If yes, how many reports have you produced? NB please give the number of mutation positive/negative samples you have reported For how long have you been providing this service? Is there specialised local clinical/research expertise for this disease? Are you testing for other genes/diseases closely allied to this one? Please give details Local Activity How many tests do you intend to provide annually in your laboratory? National Activity How many tests are being provided nationally? 30 June 2003 1 COL1A1, COL1A2 Type I collagen, alpha I and alpha 2 chain Covering all coding exons and 20-50bp of intronic sequence at intron/exon boundary of bothgenes, all missense, nonsense, splice site mutations and small deletions and insertions. The mutations are scattered throughout both genes though the majority are found in COL1A1. Direct DNA sequencing The process has been extensively validated within the laboratory against scanning techniques (SSCP, CSGE) and is the main route for mutation detection within the lab. We participate in EQA. Total number reported = 9 Mutation positive samples = 8 March 2001 Yes No Please provide details Professor N Bishop, Professor of Paediatric Bone Disease We are testing COL3A1 for vascular Ehlers Danlos Syndrome. We are introducing LRP5. We provide all tests which are required nationally each year for Osteogenesis imperfecta type I-IV. We have provided approximately 50 tests each year, of which approximately 15% are from individual with OI type 3. This obviously represents 100% of the national activity.. Analysis of the type I collagen genes is currently not available elsewhere in the UK, and has limited availability in the rest of Europe.
Based on experience how many tests will be required nationally? Based on previous activity, we expect to provide approximately 8 tests which are required nationally each year Epidemiology Estimated prevalence of disease in the general UK Estimated gene frequency (Carrier frequency or allele frequency) Estimated penetrance Target Population Please provide details on which the test is going to be used. For example a description of the particular patient group (see Note 1 below). 1 in 15,000-20,000 individuals (all types of OI) (Forlino and Marini 2000) 1 in 15,000-20,000 individuals (Forlino and Marini 2000) Type 3 has a very high penetrance (~100%). The target will be limited to patients with manifest disease and family members where appropriate. Gene mutations have been detected in individuals from different ethnic backgrounds and the testing will therefore not be limited to a particular ethnic group. Estimated prevalence of disease in the target 1 in 15,000-20,000 individuals Note 1- The characteristics of the for the test should be stated. Features such as ethnicity, age range, and affected families are all appropriate examples of characteristics. The concept of a test includes the in which it is to be applied, so, for example, a test for the Huntington gene evaluated for use in asymptomatic family members would receive a separate evaluation if it were additionally proposed that it should be used in the general. 2
Intended Use (Please use the questions in Annex A to inform your answers) Please tick the relevant clinical management criteria that this test effects. YES NO Diagnosis Treatment Prognosis & Management Presymptomatic testing Risk Assessment possibly Test Characteristics Analytical sensitivity and specificity This should be based on your own laboratory data for the specific test being applied for or the analytical sensitivity and specificity of the method/technique to be used in the case of a test yet to be set up. Clinical sensitivity and specificity of test in target The clinical sensitivity of a test is the probability of a positive test result when disease is known to be present; the clinical specificity is the probability of a negative test result when disease is known to be absent. The denominator in this case is the number with the disease (for sensitivity) or the number without disease (for specificity) Positive predictive value and penetrance are notionally equivalent for any single genetic allele the probability of developing disease given a positive test. The relationship is much more complex if more than one gene is responsible for the disease (locus heterogeneity), or if in any one gene there are multiple alleles (allelic heterogeneity), unless all the alleles are tested. In these cases, there are implications for the clinical sensitivity of the test and for its negative predictive value. For example, for a disease (such as APKD) that may be caused by either of two separate genes, even if each is 100 percent penetrant, the clinical sensitivity and the negative predictive value (and clinical validity) will both be reduced: clinical sensitivity since its maximum value can be no greater than the proportion of the disease that is caused by Sensitivity of DNA sequencing is over 95%. Since all mutations are checked in two separate amplicons, if possible by two independent methods, specificity is 100% where the mutation or type of mutation (eg Glycine substitution, nonsense) has been previously reported. Where the change is novel, particularly in type one families, it may be necessary to carry out family studies and it still may not be possible to reach a conclusion. The clinical sensitivity and specificity of COL1A1 and COL1A2 gene analysis is high in patients with clinically diagnosed Osteogenesis Imperfecta. Published literature suggests that approximately 90% of all patients with Osteogenesis Imperfecta carry COL1A1 or COL1A2 point mutations or small deletions which are detectable by direct DNA sequencing. The underlying genetic defect in the remainder is less certain, but at least some of them are suggestive of autosomal recessive inheritance patterns in as yet unidentified genes. Results from patients clinically diagnosed with Osteogenesis Imperfecta and referred to this laboratory for testing have shown a clinical sensitivity of 87% and a specificity of 100%, as we are testing both genes. 3
that particular gene, and negative predictive value since a negative test on Gene A will be no guarantee that the patient will not develop the phenotype, because the disease may be caused by Gene B. A similar form of analysis may be applied to genes with multiple alleles unless the test measures all the alleles Clinical validity (positive and negative predictive value in the target ) The clinical validity of a genetic test is a measure of how well the test predicts the presence or absence of the phenotype, clinical disease or predisposition. It is measured by its positive predictive value (the probability of getting the disease given a positive test) and negative predictive value (the probability of not getting the disease given a negative test). The denominator in this case is the number of people with a positive or a negative test respectively - not the number with or without the disease. The clinical validity may be calculated knowing the sensitivity and the specificity and the prevalence of the disease in the being studied. Positive and negative predictive values depend critically on the prevalence of the disease in the test According to the published literature, and personal clinical as well as scientific experience from this laboratory and the clinical collaborator, Prof N Bishop, this test has a 100% positive predictive value and a 93% negative predictive value in cases of clinically diagnosed Osteogenesis Imperfecta Positive predictive value and penetrance are notionally equivalent for any single genetic allele the probability of developing disease given a positive test. The relationship is much more complex if more than one gene is responsible for the disease (locus heterogeneity), or if in any one gene there are multiple alleles (allelic heterogeneity), unless all the alleles are tested. In these cases, there are implications for the clinical sensitivity of the test and for its negative predictive value. For example, for a disease (such as APKD) that may be caused by either of two separate genes, even if each is 100 percent penetrant, the clinical sensitivity and the negative predictive value (and clinical validity) will both be reduced: clinical sensitivity since its maximum value can be no greater than the proportion of the disease that is caused by that particular gene, and negative predictive value since a negative test on Gene A will be no guarantee that the patient will not develop the phenotype, because the disease may be caused by Gene B. A similar form of analysis may be applied to genes with multiple alleles unless the test measures all the alleles. 4
Clinical utility of test in target (Please refer to Appendix A) Please provide a full description of the clinical care pathway for those individuals undergoing testing. This is required to illustrate the clinical utility of the test (a template is provided on page 8). How will the test add to the management of the patient or alter clinical outcome? OI type 3 (severe) OI Type 3 generally but not exclusively occurs de novo as it is rarely compatible with reproduction (there are however a few individuals with type 3 who have been known to reproduce). Parents of a child with apparently de novo type 3 should be counselled. The risk of recurrence is associated with germline or somatic mosaicism in either parent. A priori recurrence risk is 7%. Where a mutation has been identified in the affected child, analysis of parental samples from as many tissues as available may identify somatic mosiacism. If present, the recurrence risk may be up to 50%. If negative the residual risk remains. If PND may possibly be considered in a future pregnancy, testing of the affected child should be carried out (this should be prior to conception due to length of time testing takes). If PND not acceptable no testing carried out - to be reviewed with parents if second affected pregnancy occurs. In the event that an individual with OI type 3 wishes to consider PND then the recurrence risk is 50%. Mutation analysis in the affected individual would be required prior to conception. Late prenatal testing is sometimes requested for delivery management. What impact will this test have on the NHS i.e. by removing the need for alternative management and/or investigations for this clinical Cost benefit of testing: Molecular diagnosis allows for confirmation of diagnosis, more precise recurrence risk, earlier PND with reduced parental anxiety/early termination possible. Reduction in birth of affected children. Is there an alternative means of diagnosis or prediction that does not involve molecular diagnosis? If so (and in particular if there is a biochemical test) please state the added advantage of the molecular test Diagnosis is possible by qualitative analysis of collagen in fibroblasts but this requires a skin or fibroblast sample from the affected individual which is not always available, would delay PND by up to 4 weeks and is not as reliable. PND for subsequent pregnancies is possible by scan at 16+ weeks. 5
Referral Pathway TARGET POPULATION All patients diagnosed with OI type 3 whose parents are seeking information regarding recurrence risk for future pregnancies. All patients of reproductive age diagnosed with OI type 3 where PND or predictive testing may be considered in future. Diagnosis of OI type 3 includes possible bone fractures at birth; pronounced short stature; scoliosis; greyish sclera; dentinogenesis imperfecta. However as diagnosis and subsequent termination is often in the second trimester not all characteristics may be present. WHAT TYPE AND LEVEL OF PROFESSIONAL OR REFERRER DO YOU ACCEPT SAMPLES FROM? [Foetal and neonatal samples may be collected from histopathologist carrying out a PM, or clinician on a Neonatal unit (paediatrican/obstetrician) or other. It is important that samples are stored from potential OI referrals prior to any decision being taken as to testing. In the absence of a sample no test is possible.] Parents then referred to Clinical Geneticist for discussion regarding recurrence risk for future pregnancies and options for testing including exclusion of somatic mosaicism and PND. Testing would be carried out under instruction from a Clinical Geneticist following the counselling of the parents. Patients of reproductive age referred to Clinical Geneticist for discussion regarding options for testing including PND and predictive testing. Samples obtained and testing would be carried out under instruction from a Clinical Geneticist following counselling. PLEASE PROVIDE DETAILS OF HOW REFERRALS WILL BE ASSESSED FOR APPROPRIATENESS? Patients will be assessed against the relevant test criteria HOW MANY TESTS DO YOU EXPECT TO PERFORM ANNUALLY? 6 8
Name of Disease/test: Osteogenesis Imperfecta type 3 / COL1A1, COL1A2 testing UKGTN Testing criteria OI type III Patient name: Patient postcode: Name of referrer: Title/Position: Referrals only will be accepted from one of the following: Referrer Clinical Geneticist Tick if this refers to you. Minimum required for testing to be appropriate is two or more of the following criteria as stated in the Gene Dossier: Criteria Patients diagnosed with OI type 3 whose parents are seeking information regarding recurrence risk or who may consider PND in future pregnancies. Clinical diagnosis should be made from: 1.Diagnosed as OI type III a) Fracture occurring during pregnancy, at birth or shortly thereafter and b) X ray evidence of OI type III supported by c) Typical facies with triangular face, small jaw, slight proptosis, brachycephaly. Supportive features: Multiple crush fractured vertebrae Progressive bowing deformity of long bones in absence of treatment Post-mortem report where undertaken Tick if this patient meets criteria If the sample does not fulfil these criteria and you still feel that testing should be performed please contact the laboratory to discuss testing of the sample. 7