Phenotypic Variability Among Patients With D4Z4 Reduced Allele Facioscapulohumeral Muscular Dystrophy

Key Points Question What are the phenotypes expressed among patients with facioscapulohumeral muscular dystrophy (FHSD) who are carriers of D4Z4 reduced allele with 7 to 8 repeat units? Findings In this cross-sectional study of 187 probands and 235 relatives who carry a D4Z4 reduced allele with 7 to 8 repeat units, 47.1% of probands did not have the classic FSHD phenotype, and 52.8% of the carrier relatives were nonpenetrant. In 106 families, 18.9% had a member with autosomal dominant FSHD, whereas in 34.9%, the proband was the only participant expressing a myopathic phenotype. Meaning The findings of this study suggest that knowledge of phenotypic variation in the expression of D4Z4 reduced allele with 7 to 8 repeat units in individuals with FSHD could be informative for clinical management and genetic counseling.


Introduction
Facioscapulohumeral muscular dystrophy (FSHD; OMIM 158900) is among the most common forms of hereditary myopathy. 1At present, 2 genetically distinct disease subtypes, FSHD1 and FSHD2, are described 2,3 on the basis of molecular features.In FSHD1, representative of 95% of patients, the molecular variation resides in a stretch of tandemly arrayed 3.3-kb repetitive elements named D4Z4.
Patients with FSHD1 carry D4Z4 alleles with 10 or fewer repeat units (RUs), with autosomal dominant inheritance. 4In FSHD2, individuals carry 2 D4Z4 arrays in the healthy range (ie, >10 RUs), but approximately 80% of these patients have a mutation in the SMCHD1 gene (OMIM 614982) with D4Z4 reduced CpG methylation and a permissive 4qA haplotype.7][8][9][10][11][12] The classic FSHD phenotype is characterized by onset in the first or second decade of life with progressive facial, shoulder girdle, and pectoral muscle weakness and atrophy, often asymmetric. 13sease progression may lead to the involvement of abdominal muscles and distal lower extremity weakness, causing a steppage gait before impairment of pelvic girdle muscles. 14tients with the smallest number of RUs display more severe phenotypes, including earlier wheelchair use and increased frequency of extramuscular manifestations. 15In contrast, patients with the largest number of residual D4Z4 fragments (ie, 7-10 RUs) have a milder disease and no affected relatives. 16wever, since its discovery, molecular analysis of the D4Z4 locus for FSHD diagnosis has revealed an unanticipated complexity, without a straightforward association of the clinical phenotype with molecular variations. 179][20][21][22][23] Several reports describe atypical phenotypes in carriers of D4Z4 reduced alleles (DRAs). 246][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41] Moreover, D4Z4 alleles in the range of FSHD1 (ie, 4-8 RUs) are carried by 3% of the healthy population. 5,19,20We also found that 1.3% of healthy people carry 1 DRA associated with the permissive haplotype 1614qA, and 2% carry 1 DRA with the 4qA allele. 5These observations argue for the role of modifying loci or epigenetic mechanisms influencing the clinical expression of disease. 42[45][46][47] Here we evaluate whether this genetic subgroup is different from those with classic FSHD.

Study Design and Participants
We performed a cross-sectional study of 187 probands (ie, the family member who first manifested symptoms and was the first individual analyzed) and 235 relatives from a consecutive group of 280 probands and 306 relatives, all carriers of a DRA with 7 to 8 RUs, accrued by the INRF between 2008 and 2016.All participants included in this study carry 1 DRA associated with the permissive haplotype 4qA.We did not analyze the short sequence-length polymorphism in all participants, given that several studies have shown that different haplotypes can be carried by patients with FSHD.

Procedures Clinical Investigation
We applied the CCEF, a recently published 48 novel clinical standardized clinical tool with interrater reliability. 48The CCEF consists of 4 sections.The first section, the evaluation form, investigates the patient's clinical history and disability and assesses muscle segmental involvement.The second section includes the FSHD evaluation scale to calculate the FSHD score (range, 0-15). 49The combination of the clinical features summarized in the clinical diagnostic form (section 3) assigns patients to different phenotypic categories (section 4

Statistical Analysis
Numerical variables were described using mean (SD) and compared between groups using 1-way analysis of variance, followed by Tukey honest significance post hoc test, in cases of more than 2 groups, or the t test for independent samples, in cases of 2-group comparisons.Linear models were used to adjust comparisons with respect to potential confounding factors.Categorical variables were synthetized using absolute frequencies with percentages, and differences in distribution were assessed using the χ 2 test.We obtained 95% CIs on proportions using the exact method for binomial and multinomial proportions.Missing values were not imputed.All reported P values were 2-sided, and statistical significance was set at a P < .05.All statistical analyses were performed with R version 3.5.0(R Project for Statistical Computing).

Distribution of Clinical Categories Among Probands
We grouped the probands in clinical categories on the basis of their clinical phenotype (Figure 1A

Distribution of Clinical Categories Among Relatives
Clinical evaluation of 235 relatives (Figure 1B) showed that 38 (

Distribution of Clinical Categories in Families
Facioscapulohumeral muscular dystrophy is characterized by great variability in clinical expression.
To investigate this aspect, we grouped families based on the proband's clinical category and subgrouped them based on the clinical patterns assessed in relatives (Figure 3; eFigure 2 in the Supplement).Among all families with the proband assessed as category A, we found that 30 relatives  while only 20 families (18.9%; 95% CI, 11.9%-27.0%)had a member with an autosomal dominant FSHD.The percentages of relatives with classic FSHD phenotype in families with the proband classified as category B and D were 6.1% (2 relatives) and 8.5% (5 relatives), respectively.

Age at Onset
The

Severity of Motor Impairment
We also established the degree of motor impairment among probands using the FSHD clinical score (Figure 4).The mean (SD) FSHD score was 5.8 (3.4).We did not detect a significant difference in FSHD score between men and women (5.7 [3.5]

Discussion
In FSHD genotype-phenotype correlation studies, the idea that there is an inverse correlation between the number of D4Z4 repeats and the severity of the disease has been favored. 43Alleles with DRA with 1 to 3 RUs were generally associated with a more severe form of disease, while DRA with 4 to 8 RUs was associated with the classic form of FSHD. 15,16However, it is now clear that many different phenotypes can be observed among individuals carrying a DRA, even DRAs of the same size, with critical consequences for clinical management. 24e recently published evidence-based guidelines for the diagnosis and management of FSHD 50 proposed that molecular testing, including the measurement of the size of D4Z4 alleles, the presence of 4qA polymorphism, and the D4Z4 methylation status, become a determinant aspect for diagnosis, whereas clinical features are not taken into account.In this study, we showed that among carriers with DRA with 7 to 8 RUs, the molecular test was not sufficient for diagnosis.
Considering the phenotypic variability of the probands and the high percentage of nonpenetrant individuals among relatives, finding a D4Z4 contraction might have little diagnostic and prognostic significance.We suggest applying the CCEF as a tool for the standard evaluation of the phenotype in conjunction with the molecular test.Our study showed that it is mandatory to extend the molecular test to the largest number of family members for proper genetic counselling.
Our analysis of 422 participants also provided elements for managing diagnosis, prognosis, and counseling among carriers of DRA with 7 to 8 RUs.We showed that this molecular group constitutes a very heterogeneous clinical group, including phenotypes different from the classic form of FSHD.
We observed that 52.9% of probands had the classic FSHD phenotype (ie, category A in the CCEF), whereas the rest (47.1%) displayed incomplete or atypical phenotypes.Among 187 probands, very few (4.2%) had severe facial involvement (category A1).This is a peculiar clinical aspect; in fact, we have shown previously that the percentage of patients with a classical phenotype in the carriers of smaller DRAs was close to 80% among carriers of DRA with 1 to 3 RUs. 15In addition, the percentage of relatives who are asymptomatic carriers was lower, ranging from 9.5% (for DRA with 1-3RUs) to 27.6% (for DRA with 4-6 RUs). 18Instead, the phenotypic expression of probands and relatives who

JAMA Network Open | Neurology
Phenotypic Variability Among Patients With D4Z4 Reduced Allele FSHD presented with a facial-sparing phenotype (ie, category B1).Thus, in patients with a DRA with 7 to 8 RUs, facial involvement was less frequent and less severe than previously reported among individuals carrying DRA with fewer RUs. 15 In our cohort, most symptomatic patients reported the first symptom when they were older than 20 years, without a statistically significant difference between the mean of age at onset for probands and relatives.Therefore, we can consider carriers of a DRA with 7 to 8 RUs as late-onset patients. 51is observation is also very interesting from another point of view.In our cohort, we reported several families with relatives in 3 generations, and the absence of statistically significant difference between the mean age at onset for probands and relatives suggests that no anticipation was detectable in our sample.
Among probands, 26.7% displayed atypical signs (ie, category D) and showed some distinctive features.First, patients in category D reported disease onset at older than 40 years.Therefore, we can consider them late-onset patients.Second, no probands assigned to category D displayed autosomal dominant inheritance; rather, they were sporadic cases.In this subgroup, most patients presented with some FSHD features as well as other uncommon characteristics suggestive of the possible copresence of an additional disease (ie, subcategory D1).In addition, 3.7% of these patients presented no signs that met the diagnostic criteria for FSHD (ie, subcategory D2).Considering that 3% of general population 5 carry a DRA with 4 to 8 RUs, some patients in category D2 have a different myopathy, and the association with the DRA with 7 to 8 RUs might be attributed to random occurrence.To our knowledge, this was the first study in which a large group of myopathic carriers of the molecular defect associated with FSHD1 was identified.These patients did not meet the clinical criteria, and they indicate alternative diagnoses.The next step is to perform muscle biopsies and exome sequencing to identify other causative genes in this subgroup of patients.
The evaluation of the FSHD clinical score confirms that myopathic carriers of 1 DRA with 7 to 8 RUs had a mild clinical impairment, 16,18,52 but at the same time, we observed large variability of clinical expression, particularly among probands.Family studies show that this variability does not depend exclusively on disease duration.We also found that most relatives (52.8%) carrying a DRA with 7 to 8 RUs had no muscle weakness.This percentage is much higher than the 25% to 30% reported by other studies. 18,23,52This difference is not because of the age at last clinical evaluation, given asymptomatic and nonpenetrant relatives in our cohort were older than the mean age at onset of symptomatic relatives.Thus, it is likely that they will never develop disease.
In 34.9% of families, all relatives were healthy, irrespective of the proband's clinical category.
This observation shows that disease penetrance varied among families and indicates that the genetic background or the presence of comorbidities might modulate disease onset and development.Our data point at the possibility that, in the heterozygous state, a D4Z4 reduction might produce a subclinical, sensitized condition that requires another contributing factor to cause overt myopathy.
6][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41] This possibility is also consistent with previous reports of expression changes of candidate proteins that were associated with FSHD in some families but were unchanged when other families were examined.It is also plausible that drugs or toxic agents might contribute to disease onset and clinical variability.In this respect, anamnestic records documented by the CCEF may provide useful information.Consequently, an extended evaluation of the family context is necessary to estimate prognosis for patients carrying or at risk of carrying DRA with 7 to 8 RUs.
Finally, by evaluating age at onset in combination with FSHD score and clinical category, we found that women had a later onset and frequently display atypical phenotypes.It is commonly reported that women have a milder phenotype, but the reasons are not well known. 53 anabolic factor promoting muscle protein synthesis and muscular regeneration) create a major sensitivity to the alterations caused by the FSHD pathogenic mechanism among men. 54Moreover, men and women may respond differently to catabolic conditions because of their hormonal profiles. 55,56

Limitations
This study has some limitations.The CCEF is an extensive clinical tool, which takes about 20 minutes to apply.Only a physician with expertise in neuromuscular disease can use the tool correctly.Thus, it is preferable they be properly trained.Second, a long follow-up period may be necessary to evaluate whether some symptomatic patients will be assigned to a different clinical category or if some nonpenetrant relatives will develop any sign of muscle impairment.
Third, most nonpenetrant relatives were older than 20 years, and the mean (SD) of age at last neurological examination (ie, 41.1 [15.3] years) was older than that of symptomatic relatives (ie, 33.4 [17.3] years).Thus, it is likely that they will never develop disease or that they might develop some symptoms at older age.In our cohort we had several patients with atypical phenotypes who developed the disease when older than 40 years.The clinical follow-up of nonpenetrant subjects will provide relevant clinical information on this matter.

Conclusions
The findings of this study indicate that in the case of probands who carry a DRA with 7 to 8 RUs and do not present the classic FSHD phenotype, it is necessary to consider alternative myopathies.In sporadic cases presenting with atypical phenotypes, the random association of a myopathic phenotype with a contracted D4Z4 allele has to be considered, given that there is a 1.7% frequency of DRA with 7 to 8 RUs in the general population.This study showed that the genetic background can influence the penetrance and phenotypic expression of disease in relatives carrying the same molecular signature.Based on the results of our study, the precise phenotypic characterization of patients and families should support molecular testing and could advance the management of diagnosis, genetic counseling, and selection procedures for randomized clinical trials.
tool developed to systematically describe clinical phenotypes in individuals with suspected FSHD, with the aim of obtaining additional information about the clinical significance of detecting a DRA with 7 to 8 RUs.These alleles have a 2 in 10 frequency (eFigure 1 in the Supplement) in the population accrued in the Italian National Registry for FSHD (INRF) and 1.7% in the general population 5 ; therefore, their detection has high clinical relevance but requires additional knowledge to establish their value for diagnosis and genetic counseling.At present, a standardized genotype-phenotype JAMA Network Open | Neurology Phenotypic Variability Among Patients With D4Z4 Reduced Allele FSHD JAMA Network Open.2020;3(5):e204040.doi:10.1001/jamanetworkopen.2020.4040(Reprinted) May 1, 2020 2/13 Downloaded From: https://jamanetwork.com/ by a University of Massachusetts User on 06/17/2020

Figure 1 .
Figure 1.Description of Clinical Phenotypes Observed Among Probands and Relatives

Figure 2 . 13 Downloaded
Figure 2. Distribution of Sex Across Clinical Category

Figure 3 .D
Figure 3. Distribution of Clinical Categories Of Relatives in Relationship With the Clinical Category of Probands

Figure 4 .
Figure 4. Severity of Muscle Impairment In Probands and Relatives [5][6][7][8][9][10][11][12]The carriers of DRA with 7 to 8 RUs represent 20% of all carriers accrued by INRF (eFigure 1 in the Supplement).We enrolled only patients for whom clinical evaluation was performed with the CCEF by a properly trained physician who belonged to the Italian Clinical Network for FSHD (ICNF).The ICNF is distributed across Italy and includes 1 diagnostic laboratory and 14 clinical centers.All clinical and molecular data were collected in the INRF database.Participant recruitment was approved by

Table .
Clinical Summary of Probands and Relatives a Calculated with 107 relatives (46 men; 61 women) with FSHD symptoms.
mean (SD) age at onset estimated among probands was 33.3 (17.9) years, and 124 (66.3%) limb girdle muscles significantly contributed to the whole FSHD score in those assessed as category D compared with those assessed as category A (mean [SD] contribution of impairment of limb girdle muscles: 29.4% [17.9%] vs 13.2% [14.9%];P < .001).This estimate was achieved evaluating the contribution of each subscore to the whole score.