Susan Jeffrey
April 03, 2016
CHICAGO, IL — Only a small fraction of those with very high LDL- cholesterol levels carry one of the known genes that cause familial hypercholesterolemia (FH), a new study shows[1]. However, for any two individuals with the same LDL level, those with these gene variants have a significantly increased risk for coronary artery disease.
On the basis of gene sequencing in more than 26,000 individuals with LDL cholesterols 190 mg/dL or higher, the researchers found that only 1.7% have a known FH mutation, a low diagnostic yield.
“However, with respect to clinical importance, for any given LDL, we demonstrate largely for the first time, that the risk of coronary disease is substantially higher in those who have an FH mutation, and we think this is likely related to cumulative exposure to LDL,” said Dr Amit V Khera (Massachusetts General Hospital, Boston).
“I think one of the most interesting results from my perspective of the study was we noted a substantial increase in coronary risk even at normal LDL cholesterol, less than 130 or the 130–160 range,” Khera commented during a press conference. “So I think this cumulative exposure issue is real, and it’s relevant, even at normal levels of observed LDL.”
The results were published online April 3, 2016 in the Journal of the American College of Cardiology to coincide with their presentation here at the American College of Cardiology 2016 Scientific Session.
Diagnostic Yield
It’s estimated that about 7% of US adults have severe hypercholesterolemia, defined as untreated LDL cholesterol of 190 mg/dL or higher. When levels are that high, Khera said, the question arises in the clinical setting of whether the condition is due to FH, caused by a single mutation in one of three known genes, whether it is “polygenic,” the net result of many more common variants that together confer risk, or environmental.
In this study, the researchers aimed to first assess the diagnostic yield of genetic testing for FH genes in this population of subjects selected only on the basis of severe hypercholesterolemia, and second to determine whether the CAD risk varies according to FH mutation status within a given stratum of LDL cholesterol.
Three genes causative for FH—LDL receptor (LDLR), apolipoprotein B (APOB), and proprotein convertase subtilisin/kexin type 9 (PCSK9)— were sequenced in 26,025 participants from seven CAD case-control studies (5540 CAD cases and 8577 controls free of CAD) and five prospective cohort studies including 11,908 participants.
Specific FH mutations included loss-of-function variants in LDLR, missense mutations in LDLR predicted to be damaging, and variants linked to FH in ClinVar, a clinical genetics database.
For the question on diagnostic yield, they combined the CAD-free controls with those in the population-based studies for a total of 20,485 subjects. “Of these, 1386, or about 7%, had severe hypercholesterolemia,” Khera reported. “However, when we then went on to look at their gene-sequencing data, an FH mutation was identified in only 24 of these individuals, about 1.7%.”
To explore the clinical importance of these mutations in CAD risk, they used participants with a normal LDL (<130 mg/dL) and no FH mutation as the reference group.
“Among those with a high LDL and no FH mutation, our study demonstrated a sixfold increase in risk for coronary disease,” Khera said. “However, if you had both a high LDL and an FH mutation, your risk was substantially higher, and in this population, we document an odds ratio for coronary disease of 22.3.”
CAD Risk by LDL Level and FH Mutation Status
| Group | Odds ratio (95% CI) |
| LDL <130 mg/dL, FH mutation negative | Reference |
| LDL >190 mg/dL, FH mutation negative | 6.0 (5.2–6.9) |
| LDL >190 mg/dL, FH mutation positive | 22.3 (10.7–53.2) |
Within any given category of LDL cholesterol, they found the risk for CAD was substantially higher for those with one of the mutations vs those without. They saw this pattern even among those with relatively normal LDL-cholesterol levels, he noted.
Among those with LDL cholesterol in the 190–320 mg/dL range, the odds ratio for CAD was 5.2 for those with no FH mutation but jumped to 17.0 among those positive for an FH mutation. “Importantly, these results and this discrepancy were observed despite almost identical levels of observed LDL cholesterol, with a mean of 203 vs 205 mg/dL in these groups,” Khera said.
These findings were “quite striking,” he noted, so they investigated further. “One potential mechanistic hypothesis is that FH mutation carriers have had the mutations from the time of birth, and as such, they may have been exposed to a much higher cumulative exposure to LDL cholesterol over their lifetime.”
Using data from the Atherosclerosis Risk in Communities (ARIC) study, they identified 18 carriers of an FH mutation with a high LDL and matched them to those with no FH mutation on the basis of age, gender, use of a statin, and an identical LDL at the time of ascertainment at the last study visit. By design, at time zero, the groups were similar in average LDL, about 195 mg/dL in both groups.
“However, we then took advantage of the fact that we had serial LDL levels available in these folks over time,” he said. They found that those with an FH mutation had significantly higher cumulative exposure to LDL in an area-under-the-curve analysis compared with those who didn’t. “Specifically, on average, the exposure to LDL cholesterol was 18-mg/dL higher in those who had an FH mutation,” he said.
Their findings raise a number of questions about what is the best and most cost-effective way forward, Khera said. For example, widespread gene sequencing and identification of carriers might make them adhere to preventive lifestyle changes, or it could make them fatalistic about not being able to avoid heart disease regardless of their efforts, he said.
However, they might also be in a position to start lipid therapy earlier, around 35 years of age, for example, and at least 50% of their first-degree relatives would have the same variant, “raising the possibility that cascade screening, meaning identifying the variant in these individuals’ relatives, may provide clinical gain for them as well.”
“Most important, if we’re able to perform gene sequencing for familial-hypercholesterolemia–related genes, could we reduce the morbidity and mortality of cardiovascular disease in the population, and if so, at what cost, and with what unintended consequences?” Khera concluded. “Genomic medicine as it relates to routine clinical medicine is in its early days, but I think these are the few critical questions that we and others are hoping to address in the coming years.”
Diagnostic Yield “Disappointing”
Commenting during a panel discussion after the presentation of these data here, Dr Kim Eagle (University of Michigan Health System, Ann Arbor) called this an important study.
“The diagnostic yield is disappointing, isn’t it? But potentially the impact of the finding, if you have it, is important,” he said. At this point, this information would not figure into individual risk calculation in current risk tools, Eagle said, “but as genetic testing becomes more economic and we deal with maybe the extremes of cholesterol findings, then perhaps it would not only be effective but cost-effective in moving across diagnostic thresholds where we would say, ‘Yes, this is enough that I would treat.’ ”
Khera agreed. “I think you hit the nail on the head. Traditionally, gene testing has been exorbitantly expensive, and it hasn’t necessarily been possible to apply this at scale,” he said. However, he added, “I think that if we do apply this at scale, there are a few key advantages. One is that we can really get a sense of the whole range of consequences of identifying those FH mutations. One important lesson of human genetics is that for many conditions, these mutations are incompletely penetrant.”
That is, many people with an FH mutation may have normal LDL cholesterol, he said. On average, carriers of FH mutations in this study had an LDL cholesterol of 190 mg/dL, “quite high,” but 25% had normal LDL levels of less than 130 mg/dL. Clinical criteria have traditionally been used to help predict who might have an FH mutation, but those criteria are also “imperfect” at identifying those individuals, he said.
In the future, it might be possible to make screening for FH genes more feasible by screening for other genetically driven conditions at the same time such as breast cancer or hypertrophic cardiomyopathy, he said. “So if we’re able to screen for multiple conditions that have a genetic cause, we might also increase our yield for clinically actionable results.”
Panelist Dr John JP Kastelein (University of Amsterdam, the Netherlands) said he was grateful to Khera and colleagues, “because you just gave me the ammunition to continue our $30 million grant in Holland for active identification of FH, because I think this is very, very important data.” While it has previously been suggested that these individuals had higher risk at a given LDL level, these new data confirm it, he said.
“For smaller countries, this is an enormous boost for active identification, cascade screening, an attempt to completely put all your FH patients in a database,” Kastelein noted, adding that this may be more difficult in the US from a geographical standpoint. He asked, though, whether by changing the LDL entry criterion that it might be possible to improve on the diagnostic yield of screening, “or do you think that genetic testing will become so massive and so much cheaper, that by having it low, you at least find them all?”
Khera replied that in 2016, it’s likely that screening would start at a high level of LDL, “but I don’t think that would necessarily be the same answer, say for example, in 2020.”
During a press conference here, Dr Frederick Masoudi (University of Colorado, Denver) also commented on these findings.
“I think it’s widely assumed in clinical practice that in patients with severe hypercholesterolemia, such as those that were the primary focus of the study, that there’s some sort of monogenic abnormality that’s leading to this problem, and this debunks that notion” in that only a very small minority, 1.7%, had a monogenic mutation, he said.
The study also used an “ingenious” approach to show those with FH were at particularly high risk because of the cumulative exposure to LDL over time, even among those with normal LDL levels.
“Ultimately the greater risk associated with these FH genes even at normal LDLs suggests that it could be an independent risk factor in future risk models,” Masoudi noted. “However, that remains to be seen. At this point, on the basis of this study, I would personally not be pursuing genetic testing in patients with severe hypercholesterolemia, because the effectiveness and cost-effectiveness of such a strategy on a population level remain unclear.”
The results, however, do support further study on this issue, he said. “In the meantime, adhering to recommended prevention guidelines, which generally advocate statin therapy in those patients with severe hypercholesterolemia, as well as assessing for concomitant conditions that may exacerbate hypercholesterolemia, rather than assuming that FH is in play, would be a very reasonable clinical approach.”
Khera is supported by an ACC/Merck Fellowship Award, and reported consulting fees from Merck and Amarin. Disclosures for the coauthors are listed in the article.