Continuous variables are expressed as mean±standard deviation and were compared using either the t test or Mann-Whitney U test according to their distribution. Noncontinuous variables are expressed as absolute and relative frequencies and were compared using the chi-square or Fisher exact test, as appropriate. P < .05 was considered significant. The analysis was performed using R software, version 3.4.3 (R Foundation for Statistical Computing).

The following composite end points were analyzed: a) major arrhythmic events (MAEs), if appropriate ICD discharge or SCD was recorded, and b) heart failure death (HFD), if a heart transplant (HTx) was performed or death occurred due to heart failure.

The cumulative probability of MAEs and HFD was calculated using the Kaplan-Meier method. Factors were compared using the Mantel-Cox log-rank test. The initial date of diagnosis was taken as the start of follow-up. Cox regression analysis was used to identify independent predictors of events. Risk ratios and 95% confidence intervals (95%CIs) for end points were calculated with Cox proportional hazards models, which included as covariables sex, type of genetic variant (nonmissense vs missense), LVEF < 45%, first-degree AVB, complete AVB, or any degree of AVB with NSVT.

In the REDLAMINA registry cohort, 11 new pathogenic variants were identified (4 missense and 7 nonmissense), as well as 21 known variants (16 missense and 5 nonmissense) (table 1 of the supplementary data). The most frequent variant was p.Arg190Trp, described in 16 carriers from 5 different families.

The clinical characteristics of the cohort are summarized in table 1, and more detail can be found in table 3 of the supplementary data. At first medical contact, no significant differences were found between men and women in age, symptoms, New York Heart Association (NYHA) functional class, clinical myopathy, or coronary risk factors. Left ventricular dilatation and systolic dysfunction were more common in men. There was a high incidence of conduction disorders, with 42.9% of carriers having some degree of AVB and 17.9% having third-degree AVB, which was nonsignificantly more frequent in men (P = .057).

ICDs were implanted in 62 carriers, with a higher frequency in men (41 of 71; 58%) than in women (21 of 69; 30%) (P = .003). Of the total devices implanted, 90% (56 carriers: 20 women and 36 men) were in primary prevention and 10% (6 carriers: 5 men and 1 woman) were in secondary prevention.

At the end of follow-up, 7 SCDs were documented, 4 in probands and 3 in relatives (detected during the family study). Five SCDs occurred in the 78 carriers that did not undergo device implantation (table 3 of the supplementary data).

There was only 1 death due to advanced heart failure (1 proband with contraindications for HTx). More HTxs were performed in probands (17 of 52; 32.7%) than in family members (12 of 88; 13.6%) (P = .007). In the subgroup of carriers that received an ICD, the proportion of appropriate discharges was similar in probands and family members (11 of 36 [27.9%] vs 6 of 26 [22.2%]; P = .6).

There was no significant difference in the frequency of events (embolic events, SCD, ICD discharges, or HTx) between men and women (table 1).

Two low-risk carriers who had experienced SCD were detected (both carriers of the p.Arg190Trp variant). The guidelines and Wahbi score were used to identify the remaining carriers who experienced SCD or who had an appropriate ICD discharge6–8 (table 3 of the supplementary data).

Of the 16 carriers described, 7 had HTxs and 2 experienced SCD. An asymptomatic 28-year-old man with slight left ventricular dilatation and normal LVEF experienced SCD while sleeping. He had normal Holter and stress test findings. His mother was also a p.Arg190Trp carrier and was diagnosed with DCM and 40% LVEF at 43 years of age; her atrial fibrillation had reverted to sinus rhythm and she had no NSVT on Holter monitoring. She died suddenly while in NYHA class II and under medical treatment.

American and European guidelines recommend an ICD in patients with 2 or more of the following risk factors: LVEF < 45%, NSVT, nonmissense variants, or male sex. In their new scale, Wahbi et al.6–8 added first-degree or higher AVB and LVEF as a linear variable. These recommendations are mainly based on a 2012 article by van Rijsingen et al.10 that included a European cohort of 6 countries comprising 109 families and a total of 269 carriers. Wahbi et al.8 also used data from this cohort.

According to current recommendations, the presence of a single LMNA nonmissense variant is associated with a worse prognosis than a missense variant.6–8 However, this difference is not confirmed by our study.

One possible reason is the pathogenicity attributed to some variants included in the original study by van Rijsingen et al.10 With pathogenicity as a criterion, the variants collected in that cohort were absent in at least 150 ethnically compatible controls. However, of the 37 missense genetic variants included in the original article, at least 6 (p.Lys117Arg, p.Arg397Cys, p.Arg545His, p .Ser573Leu, p.Gly638Arg, and p.Arg644Cys) can currently be classified as probably nonpathogenic or of unknown significance (table 4 of the supplementary data). All of these variants have low but significant frequencies in public databases, such as gnomAD,11 and are classified as either probably nonpathogenic or of unknown significance in ClinVar.12 With the information available in 2012, it was impossible to perform this evaluation. These variants were found in 12 index cases and 19 family members that should have been excluded from the analysis. The inclusion of nonpathogenic LMNA variants and, as a result, the inclusion of DCM of other etiologies that may have had a more benign clinical course, could explain the different prognosis observed in missense genetic variant carriers.

In the Wahbi et al. score, the authors do not mention the genetic variants included in the study. It is therefore impossible to confirm their pathogenicity. Their work also included patients with phenotypes other that of cardiac laminopathy (eg, 72 patients with lipodystrophy, 65 with Emery-Dreifuss), which may have influenced the results.8,13

The REDLAMINA registry included only pathogenic or genetic variants with a high probability of being pathogenic (according to present criteria),9 excluding variants not clearly related to the cardiac laminopathy phenotype, as defined in the Methods section (table 1 of the supplementary data and table 2 of the supplementary data).

The other possible explanation for the discrepancy in the results is that not all missense pathogenic variants included in these studies have the same prognosis. The missense genetic variants included in the REDLAMINA registry cohort might be associated with a worse prognosis than those included in both the European and Wahbi et al.8 cohorts.

In the present registry, the risk of MAEs and DCM was similar in men and women. However, the number of ICDs implanted was higher in men than in women (table 1). Although this could be considered to be due to the strict application of the guidelines (given that male sex is a risk factor), every center in the REDLAMINA registry applied its own criteria when implanting an ICD (table 3 of the supplementary data). This approach is not uncommon for patients with rare diseases associated with SCD.

In our cohort, men had a lower LVEF at first examination and, even though the number of HTxs was higher in men than in women, there were no significant differences between the sexes in the incidence of this event (SCD) during follow-up (figure 3). Neither were there differences between men and women in final composite events (MAEs and DCM) or in embolic phenomena such as SCD, AVB, pacemaker implantation, or ICD discharges (table 1).

The results of previous cohorts could have been influenced by the inclusion of nonpathogenic missense variants in previous studies and the different prognoses of some missense variants.10,14

“Low-risk” missense variants with the founder effect have been described in the literature. As well as the p.Arg331Gln (58 carriers) and p.Arg216Cys (36 carriers) variants, they have been associated with a delayed presentation and a good prognosis vs other pathogenic LMNA variants.15,16 Captur et al.17,18 carried out a study of all LMNA genetic variants published and their relationship with the phenotype described and found that malignant ventricular arrhythmias occur with greater frequency in nonmissense variant carriers. Interestingly, they observed that not all missense variants confer the same prognosis.

In contrast, some LMNA missense variants are associated with poor prognosis. A clear example is the most frequent variant reported in the REDLAMINA registry: p.Arg190Trp. The REDLAMINA registry describes 2 carriers of this missense variant. According to the established risk criteria (the guidelines and the Wahbi et al. score8), these patients did not require ICD implantation but nevertheless experienced SCD. Both carriers had only 1 risk factor (the male patient because of his sex and the female patient because her LVEF was reduced). Application of the Wahbi score would also not have indicated the need for ICD implantation (< 7%). However, a review of the literature concerning this variant would probably have led to a “high-risk” classification. At least 19 articles provide clinical information on 23 affected families, with a high incidence of HTx, SCD, and heart failure. This variant was mentioned for the first time in 2002 by Arbustini et al.19 in an Italian family with DCM and SCD. It was subsequently described on various occasions in other European, Asian, and American families and always associated with unfavorable prognoses (table 5 of the supplementary data).

Our data indicate that the current risk criteria in cardiac laminopathies are not appropriate for missense variant carriers because not all of them have the same prognosis. American and European guidelines, as well as the Wahbi et al. score, underestimate the risk of these missense variants.6–8 The work of van Rijsingen et al.10,14 and Wahbi et al.,8 apart from the already mentioned limitations, do not provide clinical data that allow us to determine which events are related to which particular variant.13 It is essential to develop international registries that share clinical information on LMNA missense variants that would allow the deduction of prognostic information and in turn help to stratify SCD risk.

The current recommendations concerning ICD implantations in cardiac laminopathies represent a good strategy but can underestimate the risk of SCD in women with certain missense variants. To manage these aspects, more studies overcoming the above limitations are needed.

The REDLAMINA registry has all of the limitations inherent to a retrospective multicenter study. Because most participating hospitals are referral centers with HTx programs, there could be a selection bias regarding the severity of the patients included in the study. The REDLAMINA registry compiled data from various centers, which may have introduced an unforeseen bias. Our sample size is smaller than that of previously published registries, although it has the advantage of being from a single country, in contrast to others that might be biased by the use of different patient treatments. The most frequent variant was p.Arg190Trp (already associated with an unfavorable prognosis), which might have influenced the results.

In our cohort, few events occurred in the low-risk groups, and no definitive conclusions can be drawn, which could be a limitation of our study. In fact, only 17 MAEs were recorded, lower than of the other mentioned series. Nevertheless, it must be considered that these series also included events related to missense variants with very doubtful pathogenicity (such as in the series of van Rinjensen et al.10,14) or simply did not provide data on which missense variants were related to events, as in the Wahbi et al. series.8 In addition, in the REDLAMINA registry, when “high-risk” patients (with at least 2 risk factors) were considered, there were no significant differences between men and women, as previously demonstrated.

Not all patients included in our study underwent complete screening (through massive ultrasequencing) for genes related to DCM because some of the older studies were performed using the Sanger technique. This could signify a bias, given that the patients whose LMNA gene was exclusively studied could have a pathogenic variant in another gene that may have influenced the prognosis.

Another limitation could also be the low number of patients that underwent cardiac magnetic resonance imaging (52 patients out of 140 included).