High burden of recurrent cardiovascular events in heterozygous familial hypercholesterolemia: The French Familial Hypercholesterolemia Registry

: Background and aims: Cardiovascular risk is high in heterozygous familial hypercholesterolemia (HeFH). The objective of this study was to describe recurrent cardiovascular events in selected patients with HeFH attending lipid clinics in France. Methods: We included 781 patients with a clinical (Dutch Lipid Clinic Network score ≥ 6) or genetic diagnosis of HeFH who had experienced a first cardiovascular event (myocardial infarction, percutaneous coronary intervention or coronary bypass, unstable angina, stroke, peripheral arterial revascularization or cardiovascular death) and were enrolled in the French Familial Hypercholesterolemia Registry (November 2015 to March 2018). Results: The first cardiovascular event occurred at the mean age of 47 years (interquartile range 39-55) in a predominantly male population (72%); 48% of patients were on statin therapy. Overall, 37% of patients had at least 11 13 14 51 52 61 62 ABSTRACT Background and aims : Cardiovascular risk is high in heterozygous familial hypercholesterolemia (HeFH). The objective of this study was to describe recurrent cardiovascular events in selected patients with HeFH attending lipid clinics in France. Methods : We included 781 patients with a clinical (Dutch Lipid Clinic Network score ≥ 6) or genetic diagnosis of HeFH who had experienced a first cardiovascular event (myocardial infarction, percutaneous coronary intervention or coronary bypass, unstable angina, stroke, peripheral arterial revascularization or cardiovascular death) and were enrolled in the French Familial Hypercholesterolemia Registry (November 2015 to March 2018). Results : The first cardiovascular event occurred at the mean age of 47 years (interquartile range 39– 55) in a predominantly male population (72%); 48% of patients were on statin therapy. Overall, 37% of patients had at least one recurrent cardiovascular event (mean of 1.8 events per patient), of which 32% occurred in the 12 months after the index event; 55% of events occurred >3 years after the first event. Mean LDL-C at the last clinic visit was 144±75 mg/dL (132±69 mg/dL for patients on high-potency statin therapy and 223±85 mg/dL for untreated patients). Conclusions : The rate of recurrent cardiovascular events was high in French patients with HeFH in secondary prevention. The detection of FH in the childhood is crucial to prevent CV events at a young age by early initiating statin therapy. There is a clear urgent need to expand the actual very small target population which can be treated with PCSK9 inhibitor in France.

Abstract: Background and aims: Cardiovascular risk is high in heterozygous familial hypercholesterolemia (HeFH). The objective of this study was to describe recurrent cardiovascular events in selected patients with HeFH attending lipid clinics in France. Methods: We included 781 patients with a clinical (Dutch Lipid Clinic Network score ≥ 6) or genetic diagnosis of HeFH who had experienced a first cardiovascular event (myocardial infarction, percutaneous coronary intervention or coronary bypass, unstable angina, stroke, peripheral arterial revascularization or cardiovascular death) and were enrolled in the French Familial Hypercholesterolemia Registry (November 2015 to March 2018). Results: The first cardiovascular event occurred at the mean age of 47 years (interquartile range 39-55) in a predominantly male population (72%); 48% of patients were on statin therapy. Overall, 37% of patients had at least one recurrent cardiovascular event (mean of 1.8 events per patient), of which 32% occurred in the 12 months after the index event; 55% of events occurred >3 years after the first event. Mean LDL-C at the last clinic visit was 144±75 mg/dL (132±69 mg/dL for patients on highpotency statin therapy and 223±85 mg/dL for untreated patients). Conclusions: The rate of recurrent cardiovascular events was high in French patients with HeFH in secondary prevention. The detection of FH in the childhood is crucial to prevent CV events at a young age by early initiating statin therapy. There is a clear urgent need to expand the actual very small target population which can be treated with PCSK9 inhibitor in France.
One in two HeFH patients is not treated with statins at the time of the first CV even 37% of HeFH patients have CV recurrences after a first CV event Nearly half of the CV recurrences occurred more than 3 years after the first event

Introduction
Familial hypercholesterolemia (FH) is an autosomal dominant hereditary disease caused by mutations in genes involved in the catabolism of low-density lipoprotein cholesterol (LDL-C): LDLR, APOB and PCSK9 [1,2]. FH is one of the most common genetic diseases, with an estimated prevalence of 1 in 250 for heterozygous FH (HeFH) and 1 in 300,000 for homozygous FH [3][4][5].
People with FH have very high circulating levels of LDL-C from birth, leading to a lifelong exposure of the arteries to elevated levels of cholesterol and a high cardiovascular risk [6,7]. Young adults with FH have an up to 13-fold higher risk of myocardial infarction compared to similar young adults without the condition [3,4]. As placebo-controlled randomized trials with cardiovascular endpoints are unethical in FH, the reduction in cardiovascular risk with lipid-lowering treatment has to be extrapolated from trials in the general population or from observational studies [8][9][10].
The current burden of cardiovascular disease in patients with HeFH in France is unknown. We sought to evaluate the typical features of cardiovascular disease (mean age at the time of the event, type and number of events, use of lipid-lowering therapies) in HeFH patients in secondary prevention in the French Familial Hypercholesterolemia Registry.

Settings
In 2015, the New French Society of Atherosclerosis established a national multicenter registry to identify patients with FH in France, with the objective of assessing screening practices, treatments, and clinical and patient-reported outcomes. Eligible patients were those who visited a participating lipid clinic and were diagnosed, either clinically (Dutch Lipid Clinic Network score ≥6) or genetically, with homozygous or heterozygous FH. Clinical and laboratory data from the patients' medical records were obtained during routine clinic visits and were entered into the registry database by trained research staff. To obtain data from first contact, retrospective data extraction from the patients' medical records was done for patients who were already being treated at the site.
Sociodemographic, clinical and biological data (age, height, weight, blood pressure, smoking status, cardiovascular events, use of lipid-lowering drugs) were collected, along with family history (hypercholesterolemia or any cardiovascular disease), standard laboratory results (total cholesterol, LDL-C) and cardiovascular imaging data. Where available, genetic data (presence of known FHrelated mutations) were collected. Information on cardiovascular history included the date of diagnosis and the type of event. Cardiovascular events included coronary heart disease (acute coronary syndromes, encompassing myocardial infarction and unstable angina, defined according to the European Society of Cardiology/American College of Cardiology [11], percutaneous coronary intervention or coronary artery bypass graft), stroke or transient ischemic attack, peripheral artery disease (defined as carotid endarterectomy, carotid angioplasty, peripheral artery angioplasty or bypass), resuscitated sudden death and cardiovascular death.

Patients and Outcomes
The present analysis involved patients with either a Dutch Lipid Clinic Network score ≥6 or an FHcausing mutation, plus a history of at least one cardiovascular event. Cardiovascular event and recurrences of cardiovascular events were defined as: coronary heart disease (myocardial infarction, coronary angioplasty or bypass and unstable angina), stroke, transient ischemic attack, peripheral artery disease (carotid endarterectomy, carotid angioplasty, and peripheral arterial bypass), resuscitated sudden death or cardiovascular death. A recurrent cardiovascular event was defined as a second cardiovascular event, regardless of the type of the first or the second event. Coronary revascularization (i.e. percutaneous coronary intervention or coronary artery bypass graft) and myocardial infarction occurring <30 days after the index event were not considered as recurrences.
Overall follow-up duration was defined as the time from the first cardiovascular event to the last available visit at the lipid clinic. The length of time between the first and the last available visit was also 6 assessed. Exclusion criteria were a diagnosis of homozygous FH and the occurrence of minor cardiovascular events such as silent ischemia and stable angina when not followed by coronary revascularization.

Statistical analysis
Data are presented as frequency and percentage for categorical variables and as mean (standard deviation [SD]) or median with interquartile range (IQR) for continuous variables. A descriptive analysis was performed for cardiovascular event history. The characteristics of patients with and without at least one recurrence were compared using Student's t test or the Wilcoxon ranksum test for continuous variables, and the chi-square test or Fisher's exact test for categorical variables. P values were associated with a significance level of 0.05 for all tests. All analyses were performed with SAS version 9.4 software (SAS Institute Inc., Cary, NC, USA).  Table 1. Median follow-up was 7 (IQR: 2-17) years from the time of the first cardiovascular event until the last available clinic visit, totaling 5779 patient-years of follow-up. The population was predominantly male (72%) with a mean age of 60 years, and 64% had a family history of any cardiovascular event. Over half (55%) of the patients had a history of smoking, 36% had hypertension and 13% had type 2 diabetes. Data on maximum lifetime cholesterol and lipoprotein(a) were available for 535 (69%) and 414 (53%) patients, respectively; mean (SD) lifetime cholesterol was 420 (19) mg/dL and lipoprotein(a) was 55 (56) mg/dL. Overall, 75% patients underwent genetic testing, of which 437 (75%) had a mutation, with some patients having mutations in more than one gene. Median age at the first visit to a lipid clinic was 51 (IQR: 41-60) years .  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65   7

Index and recurrent cardiovascular events
Information on history of cardiovascular events, collected at the last clinic visit, is shown in Table   2. Median age at the first cardiovascular event was 47 (IQR 39-55) years. The most frequently observed first events were coronary revascularization procedure (64%) and myocardial infarction (27%).
Overall, 511 recurrences were observed, yielding an average incidence rate of 9 per 100 patientyears after the first cardiovascular event. During follow-up, 37% of patients had at least one recurrence, 14% had two or more recurrences and 7% had three or more recurrences. The most frequent recurrences were coronary revascularization (65%) and peripheral artery disease (15%).

Patients with recurrent cardiovascular events
Compared with patients who did not have a recurrent cardiovascular event, those with a recurrent event were more likely to be male (81% vs 67%, p < 0.0001) and to be older (62 vs 59 years, p < 0.0001) ( Table 3). No differences were found in terms of genetic mutations, lipoprotein(a) levels, and maximum lifetime total cholesterol.
In patients with a recurrent event, the first cardiovascular event had occurred 7 years earlier (p < 0.0001), at a median age of 42 years (IQR 37 -51) ( Table 3). The type of first cardiovascular event was more often myocardial infarction (28% vs 18%, respectively) and less often stroke or TIA (6% vs 13%, respectively) in patients with at least one recurrence (p = 0.0070).
At the time of the last clinic visit, 83% of patients were receiving statin therapy (alone or in combination with other lipid-lowering therapy); 58% were on a high-potency statin, as defined by Law et al. [12], and 57% were on statin plus ezetimibe (Table 4). No difference in treatments were found between patients with or without cardiovascular recurrences (p = 0.0690).

Discussion
The results from this contemporary French multicenter registry demonstrate the high burden of cardiovascular disease in HeFH. One in two patients was not receiving statin treatment at the time of their first cardiovascular event, which occurred at a mean age of 47 years, and only 58% were on a high-potency statin. Recurrent events occurred in 37% of patients, with almost two events (first and recurrent) per patient.

Statins in primary prevention
In FH, initiation of statin treatment early in life decreases long-term exposure of the arteries to high cholesterol levels and is highly beneficial in the prevention of cardiovascular events [13]. US and European guidelines recommend starting statins before the age of 10 years in FH [1,2,14]. French pediatric and atherosclerosis societies also promote the initiation of statins during childhood in FH [15,16]. In our study, approximately 40% of patients started statin therapy in primary prevention, suggesting that the hypercholesterolemic status of the remaining patients before the first event was either unknown or ignored by patients and/or their general practitioners. Furthermore, the mean age at which statin treatment was started was 41 years, by which time these patients have been exposed to elevated cholesterol levels for many years, placing them at high cardiovascular risk [1,15,16].

Recurrent events 9
Recurrent cardiovascular events are associated with poor quality of life as well as high costs due to hospitalization and physician visits [17]. In a Swedish national registry of patients post myocardial infarction, the probability of a recurrent event (defined as non-fatal myocardial infarction, non-fatal stroke or cardiovascular death) in patients aged <60 years within 12 months of the index event was 6.5%, and the rate remained relatively stable to 3 years [18]. In the UK cohort of the GRACE UK-Belgium registry, published in 2010, recurrent myocardial infarction occurred in 15.5% of patients with an acute coronary syndrome up to 5 years after the first cardiovascular event [19]. Rapsomaniki et al.
reported a 9.7% incidence of recurrent cardiovascular events (myocardial infarction or coronary heart disease death) at 5 years in a population with stable coronary artery disease [20]. The recurrence rates in these studies, albeit in different populations and with different endpoints, are markedly lower than in our registry of HeFH (9 per 100 patient-years).
Contemporary data on recurrent cardiovascular events in HeFH are scarce. Recently, a Swiss study reported a high risk of recurrent coronary events (odds ratio of 2-3.5 versus non-FH, depending on the criteria used for the clinical diagnosis of FH) in HeFH patients after an acute coronary syndrome [21]. In an analysis using data from two Danish registries, Rerup et al. demonstrated that patients with a possible diagnosis of FH (based on the Dutch Lipid Clinic Network criteria) had a higher risk of recurrent myocardial infarction than those without FH (16% vs 11%, respectively, over a period of 3.3 years) [22]. In 2018, Galema-Boers et al. reported a 30% rate of subsequent cardiovascular events in a Dutch cohort of 102 patients with HeFH (diagnosed using the same criteria as in our study), despite maximum-tolerated lipid-lowering therapy [23].
The design of our study may in part explain the high rate of recurrent cardiovascular events. First, data were collected over a period exceeding 7 years, versus 1 year in the Swiss study [21] and 3.3 years in the Danish study [22]. Second, 75% of the patients were diagnosed using genetic testing; among those diagnosed clinically, a Dutch Lipid Clinic Network score of ≥6 was used (indicating a probable or definite diagnosis). Third, our population may be biased towards patients with moresevere FH, as it was restricted to those referred to lipid clinics.

Cardiovascular events
The majority of first and subsequent cardiovascular events (approximately 80%) were coronary in origin, with peripheral artery disease and ischemic stroke occurring in roughly equal proportions 10 (approximately 10% each). In a Spanish study, Mata et al. described a higher proportion of coronary heart disease (93.4%) in HeFH patients with prior cardiovascular disease [24]. Few data are available on cardiovascular events other than coronary events in FH studies [3]. Most recurrences (65%) were new revascularization procedures, which is in line with other studies in distinct populations. For example, 58% of recurrent events were coronary revascularization in the IMPROVE-IT trial in patients after an acute coronary syndrome [25].

Cardiovascular risk factors
More than half of our patients were current or former smokers. In a previous study in 734 patients with HeFH (with or without cardiovascular events), 75% had at least one additional risk factor, of which 31% was tobacco smoking [26]. These results emphasize the importance of identifying ways in which to encourage smoking cessation in these high cardiovascular risk patients.
The results of our study confirm that patients with FH in secondary prevention are undertreated with lipid-lowering therapy and are not achieving the recommended LDL-C target of 70 mg/dL [7].
Despite the very high cardiovascular risk of these patients, the mean LDL-C value was 144±77 mg/dL. These findings are in concert with our previous work, in which LDL-C levels remained elevated despite maximal therapy with a high-potency statin plus a non-statin lipid-lowering drug (mainly ezetimibe) [27], and show that even when managed by lipid specialists, the majority of patients with HeFH have LDL-C levels that are above the recommended target. Whereas the high levels of LDL-C in HeFH reflect the disease itself, statin intoleranceprimarily manifesting as muscle symptomscan also lead to inability to take statins or to tolerate only low doses. In France, the PCSK9 inhibitors have been reimbursed since February 2018 for two conditions: alirocumab for HeFH requiring lipoprotein apheresis; and evolocumab for homozygous FH. In France, the criteria for apheresis are LDL-C >200 mg/dL on maximum-tolerated oral lipid-lowering therapy in secondary prevention, and LDL-C >300 mg/dL on maximum-tolerated oral lipid-lowering therapy in primary prevention.

Limitations
Whereas registries provide data from the spectrum of patients treated in everyday clinical practice, several limitations must be acknowledged. First, as we collected data from patients referred to lipid centers, our population may be biased towards more severe patients with HeFH. However, the 3 4 5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 11 prevalence of comorbidities are consistent with data from other FH registries. Second, we could not verify the reason for recurrent revascularization, whether ischemia-driven, urgent or due to the presence of silent ischemia. Third, it is important to note that data regarding the initiation of statin therapy before the first event were absent for almost half of the patients, so the results on exposure on statins may not be representative of the entire population.

Conclusions
The rate of recurrent cardiovascular events was high in this retrospective, contemporary French study of HeFH. A substantial proportion of these high-risk patients were undertreated with lipidlowering therapies and were not achieving the recommended LDL-C target. Detection of FH in the childhood is crucial to prevent cardiovascular events at a young age, through early initiation of statin therapy. There is also clear and urgent need to ensure that non-statin lipid-lowering therapies, such as the PCSK9 inhibitors, are considered in patients with HeFH who are unable to achieve the recommended LDL-C target with their current therapy alone.  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60  61  62  63  64  65 12 coordinated by the CHU of Nantes. Editorial support was provided by Sophie K. Rushton-Smith, PhD (MedLink Healthcare Communications, London), and was funded by French National Society of Atherosclerosis.

Fig. 1. Kaplan-Meier curve of freedom from recurrent cardiovascular event.
Coronary heart disease recurrences were censured when they were <30 days after a first coronary event. The shading shows 95% Hall-Wellner confidence interval. We again thank the reviewers for their overall positive comments and the pertinent points raised. Our responses to these comments are provided below, and the changes are marked in the manuscript.

Dear Professors von Eckardstein and Ray
We believe that the manuscript has benefited significantly from the reviewers' critiques, and hope that you will now find it acceptable for publication in your journal.

*Point-by-Point Response
It would be nice to have a cumulative event rate curve (Kaplan-Meier) illustrating your recurrent events.
Response: As suggested, we have included a Kaplan-Meier curve (Figure 1).

Figure 1: Kaplan-Meier curve of freedom from recurrent cardiovascular event.
Coronary heart disease recurrences were censured when they were <30 days after a first coronary event. The shading shows 95% Hall-Wellner confidence interval.
The English language must be carefully reviewed by a native English.
Response: As suggested, a native English speaker has reviewed the manuscript and corrected the language.

Reviewer #2:
The study is very well designed, analyzed and presented. However, the main finding is the recurrence of CV events in heFH patients, which has been reported before in several studies and populations. The observation of increased CV event recurrences in French heFH is very important for French clinicians in order to modify the therapeutic approaches for this disease (e.g. expand the group of patients that are eligible for PCSK9 prescription).

Response:
We thank the reviewer for his/her very positive and encouraging assessment of our manuscript.
The high CV risk of the HeFH population is indeed well known, and has been studied in many cohorts We observed that 68.3% of the recurrences occurred after the first year, emphazing the need for longterm cardiovascular follow-up in this population. Finally, as you acknowledge, we hope that our results will provide support to extend the criteria for reimbursement of PCSK9 monoclonal antibodies in patients with severe FH in France. We have also modified our definition of peripheral artery disease (PAD), so that it includes only PAD that required intervention, such as carotid endarterectomy, carotid angioplasty, and peripheral arterial bypass. Therefore, the number of PAD events decreased from 99 to 76. All these changes are in red in the manuscript (page 5).

-The inclusion of valvulopathy is very debatable and it should be defined what criteria was used. Any valve of any etiology?
Response: We agree with the reviewer: valvulopathy may not be due to atherosclerosis and should not be included in the analysis. We have therefore excluded the data from patients who had valve replacement or aortic intervention. Twenty events were excluded from the analysis, leading to 781 events instead of 801.
-40% of the events occur after the first year of follow-up, and a large part of them in the first month.

Response:
We have recalculated the number of events, excluding coronary events (recurrent MI, recurrent coronary revascularization) that occurred within 30 days after the index event. All these changes are in red in the manuscript (page 5).
This type of events is complicated to attribute to recurrence since they are possibly intimately related to the index event.
Response: We have censured coronary events that occurred less than 30 days after the index event, to exclude any that could be attributed to the index event.
-In relation to the previous comment, 61% of the recurrences are coronary revascularizations, when they were 32% of the first event. These differences should be explained carefully, and possibly discard all early revascularizations.

Response:
We believe that our high rate of recurrent coronary revascularization procedures is in line with other studies in this distinct population. As stated in the original version, "Most recurrences (65%) were new revascularization procedures, which is in line with other studies in distinct populations. For example, 58% of recurrent events were coronary revascularization in the IMPROVE-IT trial in patients after an acute coronary syndrome [24]." To address this issue, we have now excluded any coronary revascularization recurrences that occurred within 30 days of the index event.
-It would be better to clarify precisely the reasons for revascularization.
Response: Unfortunately we did not collect this information, which is included as a limitation of the study (page 10). We have also excluded any coronary revascularization recurrences that occurred within 30 days of the index event.
-The mortality data are not of interest because they refer only to patients who have survived.
While this information is prospective, only survivors can be studied, the rest of the information is retrospective.
Response: This is indeed another limitation of our retrospective study, but we believe that the number of deaths is interesting and we would prefer to retain it. We replaced the term "high intensity regimen" with the more obvious term "high-intensity statin". We have also added the data in Table 2. Coronary heart disease recurrences were censured when they were <30 days after a first coronary event. The shading shows 95% Hall-Wellner confidence interval.
-A very important limitation is that information about medication is absent in half of the patients at the time of the event. This means that the data in Table 4 must be analyzed with caution.
Response: The reviewer has raised an important point: the absence of information on the lipidlowering therapy at the time of the event for nearly half of patients (45%) is a limitation, and the data in the Table 4 must be interpreted with caution.
We have added the following sentence on pages 10−11: "Third, it is important to note that data regarding the initiation of statin therapy before the first event were absent for almost half of the patients, so the results on exposure on statins may not be representative of the entire population." -In the discussion, the comparison of recurrences with other cohorts is very biased since the definition of recurrences is very different. In fact, without removing the events of the first year, the recurrences of major events: IAM and stroke is very low, less than expected in other populations. All the discussion and the title itself is somewhat biased magnifying the risk.
Response: We agree that the definitions for events and recurrent events vary between studies. We have commented on this crucial point (page 9 in the discussion): "The recurrence rates in these studies, albeit in different populations and with different endpoints, are markedly lower than in our registry of HeFH (9 per 100 patient-years)." However, recurrences of CV events are frequent in the present cohort, even after the first year following the event. Indeed, 60% of HeFH patients had a CV recurrence >1 year later. We do not think that we are magnifying the risk. For example, in a Swedish registry, the prevalence of recurrent CV events in young patients (<60 years; nevertheless older that the HeFh population of 46 years) was very low compared with our cohort: respectively 10% vs 42% (Jernberg, Eur Heart J 2015; 36:1163-1170).
We have now added a Kaplan-Meier curve (Figure 1) to clarify the point on the time of recurrences.
-There are multiple typographical errors in the quotes within the text, and typos that should be thoroughly reviewed.