Journal of Clinical Lipidology (2013) 7, 540–545
Clinical Lipidology Roundtable Discussion
JCL Roundtable: Diagnosis of severe familial hypercholesterolemia W. Virgil Brown, MD*, Daniel J. Rader, MD, John Kane, MD, PhD Emory University School of Medicine, 3208 Habersham Road NW, Atlanta, GA 30305, USA (Dr. Brown); University of Pennsylvania Health System, 3400 Civic Center Boulevard, Philadelphia, PA (Dr. Rader); and University of California, San Francisco School of Medicine, 555 Mission Bay Blvd South San Francisco, CA (Dr. Kane) KEYWORDS: Familial hypercholesterolemia; Homozygous; LDL receptor; LDL-C
Abstract: The diagnosis of familial hypercholesterolemia is usually straightforward. The severely elevated low-density lipoprotein cholesterol and the occurrence of high concentrations of low-density lipoprotein cholesterol in the parents provide the diagnosis. The presence of tendon xanthomata is confirmation but not necessary. However, this relatively simple picture becomes much more complicated when one attempts to define the genetic variants that actually produced this clinical syndrome. In this Roundtable discussion, I am joined by two experts in the identification of genetic abnormalities discovered in those with phenotypic familial hypercholesterolemia. Dr. John Kane from the University of California, San Francisco, and Dr. Daniel Rader from the University of Pennsylvania share their knowledge in and experience with this topic. Ó 2013 National Lipid Association. All rights reserved.
Financial disclosures This educational discussion was sponsored by Aegerion Pharmaceuticals, Inc. The two guest panelists received honoraria for their participation. The Editor-in-Chief, Dr. Brown, did not receive honoraria and maintained full editorial control over the content and editorial process. Dr. Rader has received consulting fees from Alnylam Pharmaceuticals, Inc, Bristol-Myers Squibb, Catabasis Pharmaceuticals, Inc, CSL, Eli Lilly and Company, Esperion Therapeutics, Inc, Johnson & Johnson, Inc, Merck & Co Inc, Novartis, Omthera Pharmaceuticals, Inc, Pfizer, Inc, Regeneron Pharmaceuticals, Inc, Roche, and Sanofi. Dr. Rader is a shareholder in Aegerion Pharmaceuticals, Inc, and VascularStrategies. Dr. Kane has received a research grant from Synageva BioPharma. * Corresponding author. E-mail address: [email protected] Submitted August 20, 2013. Accepted for publication August 22, 2013.
Dr. Brown: Let me start with a most difficult patient problem, a child younger than 5 years of age is referred because his pediatrician has discovered a low-density lipoprotein cholesterol (LDL-C) of 900 mg/dL. What should I do to refine the diagnosis? Dr. Kane: First, I would like to decide whether this is a primary disorder of cholesterol metabolism. The most likely possibility is the genetically determined absence or dysfunction of LDL receptors, but other possibilities exist. One of the entities to think about would be homozygous phytosterolemia. This can be evaluated quickly by gas chromatographic analysis of sterols. This is a disease in which the body cannot eliminate plant sterols (phytosterols) such as sitosterol and
1933-2874/$ - see front matter Ó 2013 National Lipid Association. All rights reserved. http://dx.doi.org/10.1016/j.jacl.2013.08.003
Dr. Brown
Dr. Rader
Brown et al
Diagnosis of severe FH
campesterol. These sterols may be identified as cholesterol in some analyses. In addition, phytosterol esters accumulate in circulating lipoproteins, especially LDL, along with cholesteryl esters, often leading to high serum cholesterol levels. These patients can have xanthomata that resemble those of homozygous familial hypercholesterolemia (FH). At least one other secondary dyslipidemia should be considered: cholestasis. We have seen patients with intrahepatic cholestasis with cholesterol levels .1000 mg/dL. They often have planar xanthomata of the skin. D. Rader: Although this case has all the characteristics of so-called classic homozygous FH, I think Dr. Kane is absolutely right that one needs to always think about potential alternative diagnoses, particularly because both of those he mentioned have clear implications for therapy. Dr. Brown: Family screening would be important in considering Dr. Kane these conditions because one would not expect both parents to have elevated LDL-C. Having ruled out liver disease and found one or both parents to have elevated LDL-C, would not one expect to find at least 1 and quite likely 2 defective LDL receptor alleles? Dr. Kane: Homozygosity for specific defective alleles at the LDL receptor locus is highly likely; however, a compound genetic state could be present with different defects at the LDL receptor on the 2 LDL receptor genes or a mixed heterozygous state with defects at other critical loci such as apolipoprotein B (apoB) or proprotein convertase subtilisin/kexin type 9 (PCSK9) accompanying a single defective LDL receptor gene. Dr. Brown: Do you have any statistics on the probability of true homozygosity vs mixed or compound heterozygosity? Is this a population-specific issue? Dr. Rader: First, I think we need to clarify whether we are talking about true homozygosity, meaning the exact mutation in both alleles of the same gene (such as the LDL receptor), or compound heterozygosity, which is basically 2 different mutations but both in the same gene (such as the LDL receptor). The situation of double heterozygosity is also possible in which a mutation is found in 1 allele of 1 gene (such as the LDL receptor) and then another mutation in 1 allele of a different gene (such as apoB). All of these situations have the potential to produce a phenotype that resembles ‘‘homozygous FH.’’ The most likely scenario is 2 different mutations in the LDL receptor. However, in regions where there is a founder effect with a specific LDL receptor mutation, or if there is consanguinity in the parents, the probability of having the same mutation in both alleles is considerably higher. Two mutations are not always found even in patients who appear to have phenotypic homozygous FH. Some mutations are missed because of the nature of the sequencing. Dr. Kane, I do not know if you agree with this, but I would say that in a patient who presents with phenotypic homozygous FH, up to
541 20% of the patients cannot be identified to have mutations in both alleles. Dr. Kane: That is true. So if all relevant defects are in the LDL receptor, we are really talking about 2 things here. The patient may have 2 defective but quite different alleles for the LDL receptor, one of which may be totally nonfunctional and the other a quantitatively defective receptor. In a mixed population such as the United States, this is the most frequent situation. However, the odds vary greatly when one looks at the heritage of each patient. If the patient is French Canadian or Lebanese or a South African ‘‘Afrikaner,’’ the chances are much greater of having true homozygosity for an ablative mutation. Dr. Brown: How reliable is history of consanguinity and how often do we find that? Dr. Rader: With a history of consanguinity in the parents, one should suspect the possibility of the same mutations in both alleles. However, consanguinity is not always reliably reported in the medical and family history. Dr. Brown: Practitioners should be aware that there are clusters of people, often in isolated environments, who have expanded in that setting from marriages between cousins and so forth. In this setting, a specific gene tends to be much more common, and the likelihood of it arriving twice in the same person is much higher. The French Canadians and Afrikaners in South Africa are classic examples. Then, of course, cultural issues can produce isolated populations such as the Lebanese Christians with resulting consanguinity. These are the settings in which true homozygous patients are found with some prevalence. However, there are reports in well-defined patients in whom severe hypercholesterolemia occurs with normal sequences in 1 or both LDL receptor genes. This seems to be unexplained in some patients. Have we made any progress in that area? What are the candidates for other alleles, other genes that might be participating? Dr. Kane: The PCSK9 gene product has a key role in the elimination of the LDL receptor protein. Gain-offunction mutations at this locus can result in elevated LDL levels. As yet unidentified loci are likely causes, because in some cases genotyping has failed to reveal LDL receptor mutations. Some of these may prove to be epigenetic. Dr. Brown: If a large number of people are screened for these genes related to elevated LDL-C, several gene variants are likely to be found, but often the variants do not provide an obvious molecular explanation for a disorder of LDL metabolism. Is that correct? Dr. Kane: That is right. Combined effects can happen at different loci. In our lipid clinic in San Francisco we have documented several people who have an LDL receptor allele plus dysbetalipoproteinemia. We have another family with adefective LDL receptor allele plus familial combined hyperlipidemia that resulted in extremely aggressive coronary disease. Dr. Brown: We should talk to the parents about a history of elevated LDL-C in the families. Would you not agree that examining the lipoproteins and the genes in both parents of
542
Journal of Clinical Lipidology, Vol 7, No 6, December 2013
the patient whose LDL is in the 800 or 900 range should be informative? Normal cholesterol in either parent is not consistent with classical homozygous FH. Dr. Rader: Yes, that is certainly true. To go back to what Dr. Kane said at the beginning about suspecting other diagnoses, if you did not see hypercholesterolemia in either parent, your suspicion for an autosomal recessive condition such as sitosterolemia would certainly be increased. Of course, there is also the autosomal recessive hypercholesterolemia due to homozygosity for mutations in ARH (lowdensity lipoprotein receptor adaptor protein 1; LDLRAP1) that has been described primarily in persons from Sardinia. Furthermore, certainly APOB is another gene in which mutations can interfere with the ability to bind to the LDL receptor and cause hypercholesterolemia. Generally, the panel of genes for molecular testing for diagnosis of severe hypercholesterolemia would be the LDL receptor, APOB (mutations in the receptor binding region), and PCSK9 (gain-of-function mutations). But I think your question, Dr. Brown, to paraphrase, is that if you took 1000 people who had a presentation of severe hypercholesterolemia and sequenced all 3 of those genes, what percentage of those people would you have a definitive molecular diagnosis for? I think you would be left with a reasonable number of persons, say 200 or so, on whom you were not able to make a definitive molecular diagnosis. I think that tells us, as Dr. Kane said, that other genes are in the pathway in which mutations cause severe hypercholesterolemia, and we just have not figured them out yet. Do you agree? Dr. Kane: I would agree with that. We have more research to do to elucidate all the significant loci. Dr. Brown: In the case of the APOB gene, I understand that it is quite rare to have a defect other than in the sequence near the amino acids at about 3500? Dr. Kane: Yes. Two major mutations in the ApoB-100 gene diminish the affinity of apoB-100 for the receptor, which, together, are about as abundant as the total of the receptor mutations, which can occur at many more sites in the sequence of that gene. Dr. Rader: Usually the exon that encodes the receptor binding domain specifically is sequenced, so those mutations are identified, as well as any other mutations, in the receptor binding domain. While we are talking about the apoB 3500 mutation that causes hypercholesterolemia, harkening back to the issue of isolated populations is the situation of the Lancaster County Amish population. Alan Shuldiner and his group have nicely shown that this apoB 3500 mutation, which we have always known came from Central Europe, is remarkably common in the Amish in Lancaster County because of a founder effect. Approximately 10% of the persons have that mutation, and these persons have higher LDL-C levels and increased risk of cardiovascular disease. Dr. Brown: How do the patients with abnormalities in apoB that interfere with its ligand function differ from patients with LDL receptor defects?
Dr. Rader: Those patients tend to have milder LDL-C levels and have less extensive xanthomas than patients with LDL receptor mutations. Dr. Brown: How does the rare but reported defect of PCSK9 hyperfunction fit into this picture? Dr. Kane: It is similar to severe heterozygous FH. The patients are not as severely affected as patients with homozygous FH, but they definitely convey risk. Dr. Brown: All right. Is there any way to link this genetic information to therapy? Does it help practitioners at all to know the gene sequence abnormalities in choosing how to treat these patients? Dr. Kane: Well, we certainly do not do this in practice with patients who seem to be obvious heterozygotes, for instance, with FH. Dr. Brown: In the case I gave as an example, if both parents have elevated LDL-C, how useful is it to do genetic typing? In patients who present like this, with a clear-cut genetic issue, with a high LDL after the evaluation of clinical conditions such as hypothyroidism and vascular disease is done, what guidance can gene sequencing provide? What should cause you to consider this level of diagnostic definition? Dr. Kane: If the 2 parents have what appears to be heterozygous FH, that information is probably enough for clinical purposes. If not, practitioners would want to penetrate this further. Today, whole exome studies are becoming sufficiently inexpensive. If a patient presented with a really abnormal phenotype, practitioners might move to whole exome studies, multiple defects have been found in unexpected loci, potentially revealing previously unrecognized mechanisms. Dr. Brown: Are you stating that for a patient who is in a large lipid specialty clinic with the markers of homozygous FH we have discussed, there is no mandate for gene sequencing? Is that correct? Dr. Rader: Yes, that is generally the case. I think, as Dr. Kane said, practitioners would not want to miss rare disorders that might have different implications for therapy, such as sitosterolemia, but the initial diagnostic evaluation for this disorder is biochemical. A second reason for consideration of sequencing might be related to treatment. Two new drugs were approved and are basically labeled for homozygous FH. So in some circumstances, particularly when the patient does not have classic homozygous FH, sequencing might help to clarify the situation and help with the choice of therapeutic agents. The third reason, perhaps one you want to expand on, Dr. Brown, is the whole issue of cascade screening and whether knowing the specific mutation(s) facilitates the ability to do cascade screening within extended families to a more effective extent than simply measuring LDL. It is obviously a topic of debate, but a case is to be made that if you identify mutation(s) and can tell the patient that he or she has a specific mutation, then you can screen that whole extended family effectively for that mutation. Certainly, in The Netherlands this approach has been used effectively. I predict that over time we are going to see more of a push to
Brown et al
Diagnosis of severe FH
define patients at the gene level as costs get even lower. Dr. Kane, I do not know if you would agree with that. Dr. Kane: I think we have arrived at that point. Already we find ourselves doing whole exome studies on patients who present with interesting phenotypes. Another entity we have been finding in our lipid clinic population is cholesteryl ester storage disease, a deficiency of the lysosomal acid lipase gene product, which causes accumulation of cholesteryl esters in the liver and other cells, and elevated LDL cholesterol levels, often with low levels of highdensity lipoprotein cholesterol. It is important to identify these patients who might otherwise die of aggressive cirrhosis, because a recombinant form of this lipase is now available for clinical use. Dr. Brown: Dr. Henry Ginsberg reported a study of a patient with cholesterol ester storage disease with isotopic labeling of LDL; he found reduced clearance of LDL from the blood. This patient responded to statin therapy. Dr. Kane: Although the blood levels of LDL are reduced by statins, the hepatic injury appears to progress in these patients. Dr. Brown: A molecular cousin of the LDL receptor, the very low-density lipoprotein (VLDL) receptor, exists. Is it worth looking for an abnormal VLDL receptor in this setting? Dr. Rader: Personally, I do not think so. I think the VLDL receptor is not really expressed in liver, and, to my knowledge, it has really no known role in regulating plasma levels of LDL. But I think once mutations in the standard genes are eliminated, then, as Dr. Kane is suggesting, that, if you really wanted to make a molecular diagnosis at this point, you would probably go to exome sequencing and look at the whole exome. Dr. Brown: One phenomenon that suggests genetic heterogeneity for the control of LDL-C concentrations is the markedly variable response to therapy in a given family. I have seen one sibling respond remarkably to a statin and another does not. Are you aware of elements that might control responsiveness? Dr. Kane: We, also, have observed great variation in the response to both ezetimibe and stains in different persons. This could involve differences in drug transporters and other indirectly involved mechanisms such as the affinity for a mutant hydroxymethylglutaryl coenzyme A reductase, and so forth. Dr. Brown: Are there other such signals during the care of the patient that might make you more interested in sequencing segments of the genome? Dr. Rader: I think that, if I had a patient with severe hypercholesterolemia who I knew was compliant with therapy and had a diminished response, I would look more carefully at the molecular etiology of the hypercholesterolemia, with the rationale that it might help me understand the drug response and help me care for the patient better. Dr. Brown: There are these interesting cases that probably should make us think about sequencing candidate genes. Dr. Kane: Now that whole exome, and even whole genome, capability is at hand, we can expect to identify
543 previously unsuspected participants in lipid metabolism that could also reveal new venues of treatment. Physicians who observe interesting phenotypes can contact academicians who have the capability of performing these studies. This is an opportunity for practitioners to be a critical part of the discovery process. Dr. Brown: If we remain curious and explore some of the potential pathways with selective sequencing, we may gain some fundamental knowledge. However, I agree that attempting to sequence genes in every patient that meets clinical criteria for FH is likely to have a low benefit-to-cost ratio. Dr. Rader: The estimate of 1 in 500 for the prevalence of heterozygous FH on the basis of classic phenotypic criteria needs careful examination in the modern era. Dr. Brown: Some investigators seem to have confirmed that prevalence with other mixed populations, but the method of defining LDL receptor defects has not been consistent, and methodology is changing as you both have stated. We should not forget that much higher frequencies occur in selected populations as we discussed before. Dr. Rader: Now that more people are being sequenced for the LDL receptor as well as other candidate genes and whole exomes, I think we are going to see a more reliable estimate of the prevalence of functional mutations in the LDL receptor in the general population and in people with premature coronary disease. Dr. Kane, would you have a comment on that? Dr. Kane: I would tend to agree. I think we are going to be made aware of the heterogeneity of mechanism here, and this is going to be important in the process. Gene products involved in the transport and function of receptors that we do not know about yet may be important in regulation. Epigenetic mechanisms are emerging that may be important and will require whole genome studies. Already, 4 species of regulatory RNA are recognized, pRNA, ceRNA, long intergenic RNA, and microRNA. More than 1000 microRNAs in the human genome are recognized now. In fact, there is a hierarchy for some microRNAs that regulate production of others, and so I think that, until we understand these epigenetic layers of regulation, we really are going to be missing important determinants of lipid metabolism. Such studies may help to explain the difference between the siblings who have the same mutations in an LDL receptor but who have quite different manifestations with respect to disease. Dr. Rader: So I will just say, to your last question, it has to do with family screening. Dr. Kane: Yes. Dr. Rader: I think we all recognize that there are still a lot of undiagnosed patients with FH and other forms of severe hypercholesterolemia. Estimates suggest that up to 90% of patients with FH in the United States are undiagnosed. The Familial Hypercholesterolemia Foundation is new, and one of its main efforts is to try to improve awareness of FH, increase the case finding, and then promote cascade screening, much like has been done in The Netherlands and a few other places around the world.
544
Journal of Clinical Lipidology, Vol 7, No 6, December 2013
This is an issue that lipidologists should be out in front on and that can really have a major public health effect. Dr. Brown: Many people in their 20s and 30s really are not aware of any genetic disorders or even early vascular disease in the family. These people and their children in such families will be totally missed if cholesterol is not checked in some systematic and routine fashion. The pediatricians are well aware of this, and there is a new movement in this country to make certain that every parent knows their child’s blood cholesterol by the child’s 11th birthday. Dr. Rader: Another potential reason to think about genotyping or sequencing would be if the mutation is identified in the parent, then practitioners screen the kids to identify the mutation. If it was known that the child has inherited a mutation and was going to be exposed to a lifelong hypercholesterolemia, practitioners might be inclined to start the statin sooner than if practitioners did not know the child had the mutation. Dr. Brown: I believe we all agree that this needs to be addressed as early as possible. The family diet and other risk factors as well as possible use of medications need to be addressed in children and affected parents with FH. Dr. Rader: We are really getting to an idea that others have advocated—that knowing the genotype gives one a sense of lifetime exposure that in some ways may actually be better than just measuring the LDL-C, which clearly can vary over time and, especially in kids, is not really telling the full story. So again, the concept of knowing genotype might be important clinically. Dr. Kane: I think the genomic data are likely to be of great value in determining lifetime risk. An example is loss-offunction mutations in PCSK9, resulting in low levels of plasma LDL which appear to confer longevity for the person. I think this may lead us to lifetime risk assessment rather than the current convention of estimating 10-year risk. I think medicine should refocus on lifetime risk so as to have an appropriate effect on risk in the later years of life. Dr. Brown: There is recognition of this problem and an effort to provide better estimates of lifetime risk related to causative risk factors. The autopsy data from the PDAY study certainly reminds us of the hidden nature of this disease for so many years before clinical events. From this study we have direct quantitative data on the extent of atherosclerotic lesions in 15- to 34-year-olds. This documents the relationship to elevated lipoproteins and other classical risk factors even in youth. The early treatment in selected persons could begin much earlier when there is a high probability of ongoing vascular lesions. Dr. Kane: Understanding the biology of lipid disorders and comprehending the appropriate treatment strategies will be expected to lead to significant effect on length and quality of life. Dr. Rader: If a 30-year-old with moderate hypercholesterolemia actually was genotyped and an LDL receptor mutation was found, would that in fact lead to a higher likelihood that the person would be placed on therapy by his or her physician? In addition, would the patient be
more compliant with the therapy because he or she was told to have a molecular diagnosis as opposed to just ‘‘diet-induced’’ high cholesterol? It is a hypothesis, but I think it needs to be tested. Dr. Kane: The lack of adherence informs us in this respect. I believe that the greater depth of understanding of the disease and its relationship to vascular events, the more likely will be the patient’s sustained compliance. Dr. Brown: I fully agree that when we define high-risk patients, such as those with FH, it is important to maintain therapy over the many years required for maximum effect. That means starting much earlier with therapy in such patients. Dr. Kane: That is another duty of ours, to explain clearly to patients in terms they can understand both the mechanisms of disease and the benefits of appropriate treatment. Dr. Brown: Pediatricians with a strong interest in prevention are concerned that their fellow physicians caring for children are not adequately impressed with the dangers of a high LDL-C and the need for drug therapy in children. Dr. Rader: Similarly, not all internists treat hypercholesterolemia in women as aggressively as they do in men. Dr. Brown: Interest is growing among practitioners in gynecology in making lipid management a part of their practice. Fortunately, women are more health conscious than men and are now being treated as frequently as men when they present with high cholesterol, according to data from the National Center for Health Statistics. Dr. Rader: If I might summarize my thoughts, severe inherited hypercholesterolemia is molecularly heterogeneous, with mutations in several different known genes having the potential to cause this phenotype. Patients currently labeled as ‘‘severe heterozygous FH’’ are likely to be eventually identified as having more than 1 mutation that leads to severe hypercholesterolemia. Ultimately, it seems likely that molecular diagnosis will be applied to patients with severe hypercholesterolemia to guide therapeutic choices, enhance adherence to treatment, and facilitate cascade screening of family members. Dr. Brown: I want to thank both of you for a rich discussion on this important matter of understanding the causative elements in genetically determined hypercholesterolemia. We have learned a great deal over the years from combining clinical studies with molecular biology. The fruits of this interaction are just beginning to be felt in the clinical and population arenas. Diagnosis of severe familial hypercholesterolemia: Recommended reading 1. Scientific Steering Committee on behalf of the Simon Broome Register Group. Risk of fatal coronary heart disease in familial hypercholesterolaemia. Br Med J. 1991;303:893–896. 2. Goldberg AC, Hopkins PN, Toth PP, et al. Familial hypercholesterolemia: screening, diagnosis and management of pediatric and adult patients: clinical guidance from the National Lipid Association Expert Panel on Familial Hypercholesterolemia. J Clin Lipidol. 2011;5: 133–140.
Brown et al
Diagnosis of severe FH
3. Huijgen R, Vissers MN, Kindt I, et al. Assessment of carotid atherosclerosis in normocholesterolemic individuals with proven mutations in the low-density lipoprotein receptor or apolipoprotein B genes. Circ Cardiovasc Genet. 2011;4:413–417. 4. Austin MA, Hutter CM, Zimmern RL, Humphries SE. Genetic causes of monogenic heterozygous familial hypercholesterolaemia: a HuGE prevalence review. Am J Epidemiol. 2004;160:407–420. 5. Lombardi MP, Redeker EJ, van Gent DH, et al. Molecular genetic testing for familial hypercholesterolaemia in the Netherlands: a stepwise screening strategy enhances the mutation detection rate. Genet Test. 2006;10:77–84. 6. Soutar AK, Naoumova RP. Mechanisms of disease: genetic causes of familial hypercholesterolemia. Nat Clin Pract Cardiovasc Med. 2007; 4:214–225. 7. Humphries SE, Norbury G, Leigh S, et al. What is the clinical utility of DNA testing in patients with familial hypercholesterolaemia? Curr Opin Lipidol. 2008;19:362–368. 8. Taylor A, Wang D, Patel K, et al. Mutation detection rate and spectrum in familial hypercholesterolaemia patients in the UK pilot cascade project. Clin Genet. 2010;77:572–580.
545 9. Hooper AJ, Nguyen LT, Burnett JR, et al. Genetic analysis of familial hypercholesterolaemia in Western Australia. Atherosclerosis. 2012; 224:430–434. 10. Usifo E, Leigh SE, Whittall RA, et al. Low-density lipoprotein receptor gene familial hypercholesterolemia variant database: update and pathological assessment. Ann Hum Genet. 2012;76:387–401. 11. Marteau T, Senior V, Humphries SE, et al. Psychological impact of genetic testing for familial hypercholesterolemia within a previously aware population: a randomized controlled trial. Am J Med Genet. 2004;128A:285–293. 12. Martin AC, Coakley J, Forbes DA, et al. Familial hypercholesterolaemia in children and adolescents: a new paediatric model of care. J Paediatr Child Health. 2013;49:E263–E272. 13. Lee M, Lu K, Patel SB. Genetic basis of sitosterolemia. Curr Opin Lipidol. 2001;12:141–149. 14. Garcia CK, Wilund K, Arca M, et al. Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet. 2003;34: 154–156. 15. Soutar AK, Naoumova RP. Autosomal recessive hypercholesterolemia. Semin Vasc Med. 2004;4:241–248.
Clinical Lipidology Roundtable Discussion
JCL Roundtable: Diagnosis of severe familial hypercholesterolemia W. Virgil Brown, MD*, Daniel J. Rader, MD, John Kane, MD, PhD Emory University School of Medicine, 3208 Habersham Road NW, Atlanta, GA 30305, USA (Dr. Brown); University of Pennsylvania Health System, 3400 Civic Center Boulevard, Philadelphia, PA (Dr. Rader); and University of California, San Francisco School of Medicine, 555 Mission Bay Blvd South San Francisco, CA (Dr. Kane) KEYWORDS: Familial hypercholesterolemia; Homozygous; LDL receptor; LDL-C
Abstract: The diagnosis of familial hypercholesterolemia is usually straightforward. The severely elevated low-density lipoprotein cholesterol and the occurrence of high concentrations of low-density lipoprotein cholesterol in the parents provide the diagnosis. The presence of tendon xanthomata is confirmation but not necessary. However, this relatively simple picture becomes much more complicated when one attempts to define the genetic variants that actually produced this clinical syndrome. In this Roundtable discussion, I am joined by two experts in the identification of genetic abnormalities discovered in those with phenotypic familial hypercholesterolemia. Dr. John Kane from the University of California, San Francisco, and Dr. Daniel Rader from the University of Pennsylvania share their knowledge in and experience with this topic. Ó 2013 National Lipid Association. All rights reserved.
Financial disclosures This educational discussion was sponsored by Aegerion Pharmaceuticals, Inc. The two guest panelists received honoraria for their participation. The Editor-in-Chief, Dr. Brown, did not receive honoraria and maintained full editorial control over the content and editorial process. Dr. Rader has received consulting fees from Alnylam Pharmaceuticals, Inc, Bristol-Myers Squibb, Catabasis Pharmaceuticals, Inc, CSL, Eli Lilly and Company, Esperion Therapeutics, Inc, Johnson & Johnson, Inc, Merck & Co Inc, Novartis, Omthera Pharmaceuticals, Inc, Pfizer, Inc, Regeneron Pharmaceuticals, Inc, Roche, and Sanofi. Dr. Rader is a shareholder in Aegerion Pharmaceuticals, Inc, and VascularStrategies. Dr. Kane has received a research grant from Synageva BioPharma. * Corresponding author. E-mail address: [email protected] Submitted August 20, 2013. Accepted for publication August 22, 2013.
Dr. Brown: Let me start with a most difficult patient problem, a child younger than 5 years of age is referred because his pediatrician has discovered a low-density lipoprotein cholesterol (LDL-C) of 900 mg/dL. What should I do to refine the diagnosis? Dr. Kane: First, I would like to decide whether this is a primary disorder of cholesterol metabolism. The most likely possibility is the genetically determined absence or dysfunction of LDL receptors, but other possibilities exist. One of the entities to think about would be homozygous phytosterolemia. This can be evaluated quickly by gas chromatographic analysis of sterols. This is a disease in which the body cannot eliminate plant sterols (phytosterols) such as sitosterol and
1933-2874/$ - see front matter Ó 2013 National Lipid Association. All rights reserved. http://dx.doi.org/10.1016/j.jacl.2013.08.003
Dr. Brown
Dr. Rader
Brown et al
Diagnosis of severe FH
campesterol. These sterols may be identified as cholesterol in some analyses. In addition, phytosterol esters accumulate in circulating lipoproteins, especially LDL, along with cholesteryl esters, often leading to high serum cholesterol levels. These patients can have xanthomata that resemble those of homozygous familial hypercholesterolemia (FH). At least one other secondary dyslipidemia should be considered: cholestasis. We have seen patients with intrahepatic cholestasis with cholesterol levels .1000 mg/dL. They often have planar xanthomata of the skin. D. Rader: Although this case has all the characteristics of so-called classic homozygous FH, I think Dr. Kane is absolutely right that one needs to always think about potential alternative diagnoses, particularly because both of those he mentioned have clear implications for therapy. Dr. Brown: Family screening would be important in considering Dr. Kane these conditions because one would not expect both parents to have elevated LDL-C. Having ruled out liver disease and found one or both parents to have elevated LDL-C, would not one expect to find at least 1 and quite likely 2 defective LDL receptor alleles? Dr. Kane: Homozygosity for specific defective alleles at the LDL receptor locus is highly likely; however, a compound genetic state could be present with different defects at the LDL receptor on the 2 LDL receptor genes or a mixed heterozygous state with defects at other critical loci such as apolipoprotein B (apoB) or proprotein convertase subtilisin/kexin type 9 (PCSK9) accompanying a single defective LDL receptor gene. Dr. Brown: Do you have any statistics on the probability of true homozygosity vs mixed or compound heterozygosity? Is this a population-specific issue? Dr. Rader: First, I think we need to clarify whether we are talking about true homozygosity, meaning the exact mutation in both alleles of the same gene (such as the LDL receptor), or compound heterozygosity, which is basically 2 different mutations but both in the same gene (such as the LDL receptor). The situation of double heterozygosity is also possible in which a mutation is found in 1 allele of 1 gene (such as the LDL receptor) and then another mutation in 1 allele of a different gene (such as apoB). All of these situations have the potential to produce a phenotype that resembles ‘‘homozygous FH.’’ The most likely scenario is 2 different mutations in the LDL receptor. However, in regions where there is a founder effect with a specific LDL receptor mutation, or if there is consanguinity in the parents, the probability of having the same mutation in both alleles is considerably higher. Two mutations are not always found even in patients who appear to have phenotypic homozygous FH. Some mutations are missed because of the nature of the sequencing. Dr. Kane, I do not know if you agree with this, but I would say that in a patient who presents with phenotypic homozygous FH, up to
541 20% of the patients cannot be identified to have mutations in both alleles. Dr. Kane: That is true. So if all relevant defects are in the LDL receptor, we are really talking about 2 things here. The patient may have 2 defective but quite different alleles for the LDL receptor, one of which may be totally nonfunctional and the other a quantitatively defective receptor. In a mixed population such as the United States, this is the most frequent situation. However, the odds vary greatly when one looks at the heritage of each patient. If the patient is French Canadian or Lebanese or a South African ‘‘Afrikaner,’’ the chances are much greater of having true homozygosity for an ablative mutation. Dr. Brown: How reliable is history of consanguinity and how often do we find that? Dr. Rader: With a history of consanguinity in the parents, one should suspect the possibility of the same mutations in both alleles. However, consanguinity is not always reliably reported in the medical and family history. Dr. Brown: Practitioners should be aware that there are clusters of people, often in isolated environments, who have expanded in that setting from marriages between cousins and so forth. In this setting, a specific gene tends to be much more common, and the likelihood of it arriving twice in the same person is much higher. The French Canadians and Afrikaners in South Africa are classic examples. Then, of course, cultural issues can produce isolated populations such as the Lebanese Christians with resulting consanguinity. These are the settings in which true homozygous patients are found with some prevalence. However, there are reports in well-defined patients in whom severe hypercholesterolemia occurs with normal sequences in 1 or both LDL receptor genes. This seems to be unexplained in some patients. Have we made any progress in that area? What are the candidates for other alleles, other genes that might be participating? Dr. Kane: The PCSK9 gene product has a key role in the elimination of the LDL receptor protein. Gain-offunction mutations at this locus can result in elevated LDL levels. As yet unidentified loci are likely causes, because in some cases genotyping has failed to reveal LDL receptor mutations. Some of these may prove to be epigenetic. Dr. Brown: If a large number of people are screened for these genes related to elevated LDL-C, several gene variants are likely to be found, but often the variants do not provide an obvious molecular explanation for a disorder of LDL metabolism. Is that correct? Dr. Kane: That is right. Combined effects can happen at different loci. In our lipid clinic in San Francisco we have documented several people who have an LDL receptor allele plus dysbetalipoproteinemia. We have another family with adefective LDL receptor allele plus familial combined hyperlipidemia that resulted in extremely aggressive coronary disease. Dr. Brown: We should talk to the parents about a history of elevated LDL-C in the families. Would you not agree that examining the lipoproteins and the genes in both parents of
542
Journal of Clinical Lipidology, Vol 7, No 6, December 2013
the patient whose LDL is in the 800 or 900 range should be informative? Normal cholesterol in either parent is not consistent with classical homozygous FH. Dr. Rader: Yes, that is certainly true. To go back to what Dr. Kane said at the beginning about suspecting other diagnoses, if you did not see hypercholesterolemia in either parent, your suspicion for an autosomal recessive condition such as sitosterolemia would certainly be increased. Of course, there is also the autosomal recessive hypercholesterolemia due to homozygosity for mutations in ARH (lowdensity lipoprotein receptor adaptor protein 1; LDLRAP1) that has been described primarily in persons from Sardinia. Furthermore, certainly APOB is another gene in which mutations can interfere with the ability to bind to the LDL receptor and cause hypercholesterolemia. Generally, the panel of genes for molecular testing for diagnosis of severe hypercholesterolemia would be the LDL receptor, APOB (mutations in the receptor binding region), and PCSK9 (gain-of-function mutations). But I think your question, Dr. Brown, to paraphrase, is that if you took 1000 people who had a presentation of severe hypercholesterolemia and sequenced all 3 of those genes, what percentage of those people would you have a definitive molecular diagnosis for? I think you would be left with a reasonable number of persons, say 200 or so, on whom you were not able to make a definitive molecular diagnosis. I think that tells us, as Dr. Kane said, that other genes are in the pathway in which mutations cause severe hypercholesterolemia, and we just have not figured them out yet. Do you agree? Dr. Kane: I would agree with that. We have more research to do to elucidate all the significant loci. Dr. Brown: In the case of the APOB gene, I understand that it is quite rare to have a defect other than in the sequence near the amino acids at about 3500? Dr. Kane: Yes. Two major mutations in the ApoB-100 gene diminish the affinity of apoB-100 for the receptor, which, together, are about as abundant as the total of the receptor mutations, which can occur at many more sites in the sequence of that gene. Dr. Rader: Usually the exon that encodes the receptor binding domain specifically is sequenced, so those mutations are identified, as well as any other mutations, in the receptor binding domain. While we are talking about the apoB 3500 mutation that causes hypercholesterolemia, harkening back to the issue of isolated populations is the situation of the Lancaster County Amish population. Alan Shuldiner and his group have nicely shown that this apoB 3500 mutation, which we have always known came from Central Europe, is remarkably common in the Amish in Lancaster County because of a founder effect. Approximately 10% of the persons have that mutation, and these persons have higher LDL-C levels and increased risk of cardiovascular disease. Dr. Brown: How do the patients with abnormalities in apoB that interfere with its ligand function differ from patients with LDL receptor defects?
Dr. Rader: Those patients tend to have milder LDL-C levels and have less extensive xanthomas than patients with LDL receptor mutations. Dr. Brown: How does the rare but reported defect of PCSK9 hyperfunction fit into this picture? Dr. Kane: It is similar to severe heterozygous FH. The patients are not as severely affected as patients with homozygous FH, but they definitely convey risk. Dr. Brown: All right. Is there any way to link this genetic information to therapy? Does it help practitioners at all to know the gene sequence abnormalities in choosing how to treat these patients? Dr. Kane: Well, we certainly do not do this in practice with patients who seem to be obvious heterozygotes, for instance, with FH. Dr. Brown: In the case I gave as an example, if both parents have elevated LDL-C, how useful is it to do genetic typing? In patients who present like this, with a clear-cut genetic issue, with a high LDL after the evaluation of clinical conditions such as hypothyroidism and vascular disease is done, what guidance can gene sequencing provide? What should cause you to consider this level of diagnostic definition? Dr. Kane: If the 2 parents have what appears to be heterozygous FH, that information is probably enough for clinical purposes. If not, practitioners would want to penetrate this further. Today, whole exome studies are becoming sufficiently inexpensive. If a patient presented with a really abnormal phenotype, practitioners might move to whole exome studies, multiple defects have been found in unexpected loci, potentially revealing previously unrecognized mechanisms. Dr. Brown: Are you stating that for a patient who is in a large lipid specialty clinic with the markers of homozygous FH we have discussed, there is no mandate for gene sequencing? Is that correct? Dr. Rader: Yes, that is generally the case. I think, as Dr. Kane said, practitioners would not want to miss rare disorders that might have different implications for therapy, such as sitosterolemia, but the initial diagnostic evaluation for this disorder is biochemical. A second reason for consideration of sequencing might be related to treatment. Two new drugs were approved and are basically labeled for homozygous FH. So in some circumstances, particularly when the patient does not have classic homozygous FH, sequencing might help to clarify the situation and help with the choice of therapeutic agents. The third reason, perhaps one you want to expand on, Dr. Brown, is the whole issue of cascade screening and whether knowing the specific mutation(s) facilitates the ability to do cascade screening within extended families to a more effective extent than simply measuring LDL. It is obviously a topic of debate, but a case is to be made that if you identify mutation(s) and can tell the patient that he or she has a specific mutation, then you can screen that whole extended family effectively for that mutation. Certainly, in The Netherlands this approach has been used effectively. I predict that over time we are going to see more of a push to
Brown et al
Diagnosis of severe FH
define patients at the gene level as costs get even lower. Dr. Kane, I do not know if you would agree with that. Dr. Kane: I think we have arrived at that point. Already we find ourselves doing whole exome studies on patients who present with interesting phenotypes. Another entity we have been finding in our lipid clinic population is cholesteryl ester storage disease, a deficiency of the lysosomal acid lipase gene product, which causes accumulation of cholesteryl esters in the liver and other cells, and elevated LDL cholesterol levels, often with low levels of highdensity lipoprotein cholesterol. It is important to identify these patients who might otherwise die of aggressive cirrhosis, because a recombinant form of this lipase is now available for clinical use. Dr. Brown: Dr. Henry Ginsberg reported a study of a patient with cholesterol ester storage disease with isotopic labeling of LDL; he found reduced clearance of LDL from the blood. This patient responded to statin therapy. Dr. Kane: Although the blood levels of LDL are reduced by statins, the hepatic injury appears to progress in these patients. Dr. Brown: A molecular cousin of the LDL receptor, the very low-density lipoprotein (VLDL) receptor, exists. Is it worth looking for an abnormal VLDL receptor in this setting? Dr. Rader: Personally, I do not think so. I think the VLDL receptor is not really expressed in liver, and, to my knowledge, it has really no known role in regulating plasma levels of LDL. But I think once mutations in the standard genes are eliminated, then, as Dr. Kane is suggesting, that, if you really wanted to make a molecular diagnosis at this point, you would probably go to exome sequencing and look at the whole exome. Dr. Brown: One phenomenon that suggests genetic heterogeneity for the control of LDL-C concentrations is the markedly variable response to therapy in a given family. I have seen one sibling respond remarkably to a statin and another does not. Are you aware of elements that might control responsiveness? Dr. Kane: We, also, have observed great variation in the response to both ezetimibe and stains in different persons. This could involve differences in drug transporters and other indirectly involved mechanisms such as the affinity for a mutant hydroxymethylglutaryl coenzyme A reductase, and so forth. Dr. Brown: Are there other such signals during the care of the patient that might make you more interested in sequencing segments of the genome? Dr. Rader: I think that, if I had a patient with severe hypercholesterolemia who I knew was compliant with therapy and had a diminished response, I would look more carefully at the molecular etiology of the hypercholesterolemia, with the rationale that it might help me understand the drug response and help me care for the patient better. Dr. Brown: There are these interesting cases that probably should make us think about sequencing candidate genes. Dr. Kane: Now that whole exome, and even whole genome, capability is at hand, we can expect to identify
543 previously unsuspected participants in lipid metabolism that could also reveal new venues of treatment. Physicians who observe interesting phenotypes can contact academicians who have the capability of performing these studies. This is an opportunity for practitioners to be a critical part of the discovery process. Dr. Brown: If we remain curious and explore some of the potential pathways with selective sequencing, we may gain some fundamental knowledge. However, I agree that attempting to sequence genes in every patient that meets clinical criteria for FH is likely to have a low benefit-to-cost ratio. Dr. Rader: The estimate of 1 in 500 for the prevalence of heterozygous FH on the basis of classic phenotypic criteria needs careful examination in the modern era. Dr. Brown: Some investigators seem to have confirmed that prevalence with other mixed populations, but the method of defining LDL receptor defects has not been consistent, and methodology is changing as you both have stated. We should not forget that much higher frequencies occur in selected populations as we discussed before. Dr. Rader: Now that more people are being sequenced for the LDL receptor as well as other candidate genes and whole exomes, I think we are going to see a more reliable estimate of the prevalence of functional mutations in the LDL receptor in the general population and in people with premature coronary disease. Dr. Kane, would you have a comment on that? Dr. Kane: I would tend to agree. I think we are going to be made aware of the heterogeneity of mechanism here, and this is going to be important in the process. Gene products involved in the transport and function of receptors that we do not know about yet may be important in regulation. Epigenetic mechanisms are emerging that may be important and will require whole genome studies. Already, 4 species of regulatory RNA are recognized, pRNA, ceRNA, long intergenic RNA, and microRNA. More than 1000 microRNAs in the human genome are recognized now. In fact, there is a hierarchy for some microRNAs that regulate production of others, and so I think that, until we understand these epigenetic layers of regulation, we really are going to be missing important determinants of lipid metabolism. Such studies may help to explain the difference between the siblings who have the same mutations in an LDL receptor but who have quite different manifestations with respect to disease. Dr. Rader: So I will just say, to your last question, it has to do with family screening. Dr. Kane: Yes. Dr. Rader: I think we all recognize that there are still a lot of undiagnosed patients with FH and other forms of severe hypercholesterolemia. Estimates suggest that up to 90% of patients with FH in the United States are undiagnosed. The Familial Hypercholesterolemia Foundation is new, and one of its main efforts is to try to improve awareness of FH, increase the case finding, and then promote cascade screening, much like has been done in The Netherlands and a few other places around the world.
544
Journal of Clinical Lipidology, Vol 7, No 6, December 2013
This is an issue that lipidologists should be out in front on and that can really have a major public health effect. Dr. Brown: Many people in their 20s and 30s really are not aware of any genetic disorders or even early vascular disease in the family. These people and their children in such families will be totally missed if cholesterol is not checked in some systematic and routine fashion. The pediatricians are well aware of this, and there is a new movement in this country to make certain that every parent knows their child’s blood cholesterol by the child’s 11th birthday. Dr. Rader: Another potential reason to think about genotyping or sequencing would be if the mutation is identified in the parent, then practitioners screen the kids to identify the mutation. If it was known that the child has inherited a mutation and was going to be exposed to a lifelong hypercholesterolemia, practitioners might be inclined to start the statin sooner than if practitioners did not know the child had the mutation. Dr. Brown: I believe we all agree that this needs to be addressed as early as possible. The family diet and other risk factors as well as possible use of medications need to be addressed in children and affected parents with FH. Dr. Rader: We are really getting to an idea that others have advocated—that knowing the genotype gives one a sense of lifetime exposure that in some ways may actually be better than just measuring the LDL-C, which clearly can vary over time and, especially in kids, is not really telling the full story. So again, the concept of knowing genotype might be important clinically. Dr. Kane: I think the genomic data are likely to be of great value in determining lifetime risk. An example is loss-offunction mutations in PCSK9, resulting in low levels of plasma LDL which appear to confer longevity for the person. I think this may lead us to lifetime risk assessment rather than the current convention of estimating 10-year risk. I think medicine should refocus on lifetime risk so as to have an appropriate effect on risk in the later years of life. Dr. Brown: There is recognition of this problem and an effort to provide better estimates of lifetime risk related to causative risk factors. The autopsy data from the PDAY study certainly reminds us of the hidden nature of this disease for so many years before clinical events. From this study we have direct quantitative data on the extent of atherosclerotic lesions in 15- to 34-year-olds. This documents the relationship to elevated lipoproteins and other classical risk factors even in youth. The early treatment in selected persons could begin much earlier when there is a high probability of ongoing vascular lesions. Dr. Kane: Understanding the biology of lipid disorders and comprehending the appropriate treatment strategies will be expected to lead to significant effect on length and quality of life. Dr. Rader: If a 30-year-old with moderate hypercholesterolemia actually was genotyped and an LDL receptor mutation was found, would that in fact lead to a higher likelihood that the person would be placed on therapy by his or her physician? In addition, would the patient be
more compliant with the therapy because he or she was told to have a molecular diagnosis as opposed to just ‘‘diet-induced’’ high cholesterol? It is a hypothesis, but I think it needs to be tested. Dr. Kane: The lack of adherence informs us in this respect. I believe that the greater depth of understanding of the disease and its relationship to vascular events, the more likely will be the patient’s sustained compliance. Dr. Brown: I fully agree that when we define high-risk patients, such as those with FH, it is important to maintain therapy over the many years required for maximum effect. That means starting much earlier with therapy in such patients. Dr. Kane: That is another duty of ours, to explain clearly to patients in terms they can understand both the mechanisms of disease and the benefits of appropriate treatment. Dr. Brown: Pediatricians with a strong interest in prevention are concerned that their fellow physicians caring for children are not adequately impressed with the dangers of a high LDL-C and the need for drug therapy in children. Dr. Rader: Similarly, not all internists treat hypercholesterolemia in women as aggressively as they do in men. Dr. Brown: Interest is growing among practitioners in gynecology in making lipid management a part of their practice. Fortunately, women are more health conscious than men and are now being treated as frequently as men when they present with high cholesterol, according to data from the National Center for Health Statistics. Dr. Rader: If I might summarize my thoughts, severe inherited hypercholesterolemia is molecularly heterogeneous, with mutations in several different known genes having the potential to cause this phenotype. Patients currently labeled as ‘‘severe heterozygous FH’’ are likely to be eventually identified as having more than 1 mutation that leads to severe hypercholesterolemia. Ultimately, it seems likely that molecular diagnosis will be applied to patients with severe hypercholesterolemia to guide therapeutic choices, enhance adherence to treatment, and facilitate cascade screening of family members. Dr. Brown: I want to thank both of you for a rich discussion on this important matter of understanding the causative elements in genetically determined hypercholesterolemia. We have learned a great deal over the years from combining clinical studies with molecular biology. The fruits of this interaction are just beginning to be felt in the clinical and population arenas. Diagnosis of severe familial hypercholesterolemia: Recommended reading 1. Scientific Steering Committee on behalf of the Simon Broome Register Group. Risk of fatal coronary heart disease in familial hypercholesterolaemia. Br Med J. 1991;303:893–896. 2. Goldberg AC, Hopkins PN, Toth PP, et al. Familial hypercholesterolemia: screening, diagnosis and management of pediatric and adult patients: clinical guidance from the National Lipid Association Expert Panel on Familial Hypercholesterolemia. J Clin Lipidol. 2011;5: 133–140.
Brown et al
Diagnosis of severe FH
3. Huijgen R, Vissers MN, Kindt I, et al. Assessment of carotid atherosclerosis in normocholesterolemic individuals with proven mutations in the low-density lipoprotein receptor or apolipoprotein B genes. Circ Cardiovasc Genet. 2011;4:413–417. 4. Austin MA, Hutter CM, Zimmern RL, Humphries SE. Genetic causes of monogenic heterozygous familial hypercholesterolaemia: a HuGE prevalence review. Am J Epidemiol. 2004;160:407–420. 5. Lombardi MP, Redeker EJ, van Gent DH, et al. Molecular genetic testing for familial hypercholesterolaemia in the Netherlands: a stepwise screening strategy enhances the mutation detection rate. Genet Test. 2006;10:77–84. 6. Soutar AK, Naoumova RP. Mechanisms of disease: genetic causes of familial hypercholesterolemia. Nat Clin Pract Cardiovasc Med. 2007; 4:214–225. 7. Humphries SE, Norbury G, Leigh S, et al. What is the clinical utility of DNA testing in patients with familial hypercholesterolaemia? Curr Opin Lipidol. 2008;19:362–368. 8. Taylor A, Wang D, Patel K, et al. Mutation detection rate and spectrum in familial hypercholesterolaemia patients in the UK pilot cascade project. Clin Genet. 2010;77:572–580.
545 9. Hooper AJ, Nguyen LT, Burnett JR, et al. Genetic analysis of familial hypercholesterolaemia in Western Australia. Atherosclerosis. 2012; 224:430–434. 10. Usifo E, Leigh SE, Whittall RA, et al. Low-density lipoprotein receptor gene familial hypercholesterolemia variant database: update and pathological assessment. Ann Hum Genet. 2012;76:387–401. 11. Marteau T, Senior V, Humphries SE, et al. Psychological impact of genetic testing for familial hypercholesterolemia within a previously aware population: a randomized controlled trial. Am J Med Genet. 2004;128A:285–293. 12. Martin AC, Coakley J, Forbes DA, et al. Familial hypercholesterolaemia in children and adolescents: a new paediatric model of care. J Paediatr Child Health. 2013;49:E263–E272. 13. Lee M, Lu K, Patel SB. Genetic basis of sitosterolemia. Curr Opin Lipidol. 2001;12:141–149. 14. Garcia CK, Wilund K, Arca M, et al. Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet. 2003;34: 154–156. 15. Soutar AK, Naoumova RP. Autosomal recessive hypercholesterolemia. Semin Vasc Med. 2004;4:241–248.
