ABC | Volume 111, Nº4, Octuber 2018

Original Article Silva et al Predictors of family enrollment Arq Bras Cardiol. 2018; 111(4):578-584 Currently, cascade genetic screening is themost cost‑effective method for FH diagnosis. 9–12 The screening begins with the clinical and genetic diagnosis of an IC, of which all first‑degree relatives are screened for the same mutation. After the identification of all affected relatives, the cascade gives sequence to all 2 nd -degree relatives, and then successively. Most importantly, the higher the number of screened relatives, the more cost-effective the cascade becomes. 13,14 Interestingly, despite being recognized as the most cost- effective strategy for population-wide FH identification, little is known about the best strategies for maximizing the enrollment of at-risk individuals in a cascade screening program. One can argue that this information is evenmore important than devising ways to identify index-cases to be tested from the overall population.. Therefore, the aim of this study is to identify the main predictors of family enrollment into cascade screening using the IC as a starting point. Methods The Brazilian FH screening program (HipercolBrasil) is performed by the Laboratory of Genetics and Molecular Cardiology at the Heart Institute (InCor) of the University of São Paulo Medical School Hospital. This study was approved by the institutional ethics committee (CAPPesq 3757/12/013). All participants read and signed an informed consent form authorizing the study. Participants included in our analysis were previously registered in the HipercolBrasil Program and were referred to the programby institutional physicians or by other collaborators. Individuals that spontaneously contacted the programby phone or website were also included. Once the inclusion criteria were met, participants were referred to molecular genetic testing. Study population and inclusion criteria The inclusion criterion for program enrollment was the presence of a baseline LDL-C value ≥ 210 mg/dL. However, some individuals with LDL-C < 210 mg/dL were also enrolled when suggestive signs of FH were detected by the physicians. All genetic positive ICs (individuals in which a pathogenic or likely pathogenic mutations were identified) who authorized the screening of relatives from January 2011 to July 2015 were included in the present study. Whole blood was collected after a physical exam was performed and a standardized questionnaire was applied by trained personnel from the HipercolBrasil team. In case of a positive genetic result, the IC was contacted and informed about the importance of the genetic results and the possibility of free family screening. After a comprehensive explanation of the disease risks and early diagnosis benefits, the IC was asked to provide information on all at-risk first-degree relatives. These were then contacted by phone and invited to join the cascade program by trained specialized health professionals. The screening in relatives is restricted to the same mutation found in the IC, despite the presence or not of FH clinical features. All relatives also signed the informed consent form and were submitted to the same standardized questionnaire application. Study variables Possible predictor variables from ICs were obtained before the genetic test results were available . The standardized questionnaire consisted of socioeconomic, clinical and biochemical variables. Information regarding employment status consisted of three categories: employed (working age, individual currently working); unemployed (working age, individual not currently working) and inactive (students, elderly and/or retired individuals and those with special needs unable to work). Educational level was defined as: illiterate, elementary education, high-school education, and college/university. IC origin was defined according to whom or from where the patient was referred to the program. ICs could have been referred by physicians from the Lipid Clinic of the Heart Institute (the lipids referral center closely associated with the HipercolBrasil program); from partner centers located at other tertiary care institutions; from private physicians; by the patient itself through the program website (www. hipercolesterolemia.com.br) ; or by a primary health care unit. Enrollment criteria were the same for all ICs regardless of origin. The participation of other partner centers in the study was approved by the institutional ethics committee (CAAE 00594212.0.0000.0068 /nº:1.213.994). Required clinical information was: occurrence of atherosclerotic or familial history of early atherosclerotic cardiovascular disease and/or altered lipid levels; clinical stigmata such as corneal arcus, xanthelasmas or xanthomas. Biochemical exams were obtained from medical records or from previous exams brought by the patient. The following values were recorded: total cholesterol (TC), LDL-C, HDL-C, triglycerides (TG) and fasting glucose. The Dutch Lipid Clinic Network (DLNC) score and Simon Broome criteria were calculated using the available information at the baseline visit. Whenever possible, the baseline value of LDL-C was used. In case of a patient receiving lipid-lowering treatment with unavailable baseline LDL-C values, the current value was used to calculate the score. Those clinical scores were applied only with the intention of collecting and storing data and were not used as criteria for program enrollment. Genetic testing IC samples were sequenced for six FH-related genes: LDLR , APOB , PCSK9 , LDLRAP1 , LIPA and APOE . Target regions were considered as coding exons plus 10bp of introns up and downstreamand captured using a specially designed enrichment reagent. Templates were prepared on Ion One Touch System and sequenced in Ion Torrent PGM®platform, with 32 samples per run in a 316v2 Ion Chip. Bioinformatics analyses were performed in CLC Genomics Workbench 9.5 (QIAGEN) in a custom pipeline. Minimum quality requirements for variant call were: Base quality of PhredQ ≥ 20; Target‑region coverage ≥ 10x; Frequency of variant allele ≥ 20% and bidirectional presence of variant allele. After filtering for a MAF ≤ 0.002 with control populations (NHLBI-ESP6500, AbraOM, ExAC and 1000Genomes), all potential mutations were consulted for previous description 579