Genetic and familial predisposition to rotator cuff disease: a systematic review.

The initial literature search produced 251 articles which were assessed for relevance by their title and abstract. Of these, 234 were excluded due to lack of relevance to our topic. After the remaining 17 full texts were reviewed, seven studies were found to meet the inclusion criteria and were thus included in our final analysis ( Figure 1 ). Four studies investigated whether there is a familial predisposition to rotator cuff disease. One of these studies (n = 129) demonstrated that siblings of an individual with a rotator cuff tear were twice as likely to develop a full-thickness tear and nearly five times more likely to suffer symptoms, when compared to spouses of these individuals ( Table I ) 7 . A five-year follow-up (n = 62) showed that the relative risk for the siblings to have a full-thickness tear was 2.85 (95% CI 1.75–4.64) compared to the control population, the relative risk for a tear to progress in size was 2.08 (95% CI 1.58–2.7), and the relative risk of having a symptomatic tear was 1.44 (95% CI 2.04–8.28) ( Table I ) 6 . Tashjian et al’s 2009 study (n = 3,091) used the Genealogical Index of Familiality to demonstrate a significant excess relatedness when all generations were used but not when only looking at more distant relationships ( Table I ) 29 . When only individuals diagnosed before age 40 (n = 652) were studied, significant excess relatedness was found when both all generations and only more distant relationships were used ( Table I ). Close relationships were defined as those between first- and second-degree relatives, while distant relationships were those with a genetic path length of at least three. Excess relatedness was used interchangeably with excess familial clustering or heritable predisposition. Tashjian et al’s 2014 study (n = 92) demonstrated that a significantly higher number of individuals with rotator cuff tears (32.3%) also had family members with a history of rotator cuff tears or surgery when compared to individuals without rotator cuff tears (18.3%) ( Table I ) 31 . Three studies investigated the genes associated with rotator cuff disease by using association analyses to map genes to rotator cuff disease. One of these studies (n = 203) found a significant association of certain haplotypes in DEFB1 , FGFR1 , FGFR3 , and ESRRB with rotator cuff disease ( Table II ) 17 . After adjusting for ethnic group and sex, another association was found for FGF10 ( Table II ). A second study (n = 175) discovered two haplotypes in ESRRB that significantly increased the risk of tears ( Table II ) 32 . The third study (n = 311) found two single nucleotide polymorphisms (SNPs) within genes SAP30BP and SASH1 associated with rotator cuff tears ( Table II ) 30 . The specific SNPs and haplotypes associated with rotator cuff disease from these three prior studies are presented in Table III . The accompanying forest plots are in Figures 2 and 3 . The results of bias assessment according to the MINORS criteria are in Table IV 25 . All of the studies had a clearly stated aim, prospective collection of data, end points appropriate to the aim of the study, loss to follow-up <5%, and contemporary groups. Six studies included consecutive patients 6 , 7 , 17 , 29 , 30 , 32 , two stated they had unbiased assessments of study end points 6 , 7 , one had a prospective calculation of the study size 17 , three had adequate control groups 6 , 7 , 29 , five had baseline equivalence of groups 6 , 7 , 17 , 29 , 31 , and five had adequate statistical analyses 7 , 17 , 29 , 30 , 32 . The lowest score was 16 31 , and the highest was 22 7 .

Rotator cuff disease is a common tendon disorder that is associated with shoulder pain and functional disability. The pathogenesis of rotator cuff disease is not completely understood. Identifying a possible genetic association could help our understanding of the disease process that leads to rotator cuff pathology. This systematic review summarized studies on familial and genetic predisposition to rotator cuff disease. Although there were only a limited number of studies on this topic, they do in general constitute a consensus that rotator cuff disease is a heritable trait. Rotator cuff disease is a generic term that can be used to describe impingement syndrome, subacromial/subdeltoid bursal pathology, rotator cuff tendinopathy, and rotator cuff tear. An issue with prior literature is the absence of a uniform definition and diagnostic criteria for rotator cuff disease. Even in studies limited to rotator cuff tears, the case definition is variable. One study used clinical diagnosis 29 by a physician as their criterion, whereas other studies used ultrasound 6 , 7 and MRI 17 , 30 – 32 for diagnosis. One of the studies used a criterion of whether the patient underwent a surgical rotator cuff repair 29 . Rotator cuff pathology is a clinical syndrome since structural defects found on imaging have been demonstrated in asymptomatic individuals. 16 , 24 Hence the case definition of rotator cuff pathology needs to account for both clinical presentation and structural deficit. Harvie et al compared the rates of symptomatic and asymptomatic tears in siblings and spouses of individuals with rotator cuff tears, and determined that both were higher in siblings 7 . A follow-up study demonstrated that rotator cuff tears in siblings also had a higher risk of progressing 6 . Another study reported differing results when analyzing all of their subjects versus analyzing only those diagnosed before the age of 40 29 . The entire cohort did not demonstrate excess relatedness when only distant relationships were studied, implying that perhaps environmental factors were playing a confounding role. In contrast, the subgroup of younger patients showed excess relatedness in both close and distant relationships. The genetic association studies observed associations between rotator cuff disease and SNPs in seven candidate genes ( Table III , Figure 2 , Figure 3 ) 17 , 30 , 32 . DEFB1 (Defensin, Beta 1) encodes the protein antimicrobial peptide defensin β-1, which aids in preventing epithelial surfaces from being colonized by microbes. The rs1800972 C>G variant was significantly more frequent in individuals with rotator cuff disease 17 . This base change is thought to lead to a decreased production of defensin β-1 production and higher expression levels. 20 , 21 The G allele is also more common in individuals with severe acute pancreatitis and less predominant in individuals with diabetes and S. aureus nasal colonization 8 , 20 , 21 . ESRRB (estrogen-related receptor beta) encodes a protein similar to the estrogen receptor and is believed to have an inhibitory effect on estrogen signaling 27 . Mutations in this gene have also been associated with hearing impairment and dental decay 35 . In addition, upregulation of ESRRB has been linked with the progression of endometriosis 3 . In vitro studies have demonstrated a correlation between estrogen deficiency and poor tendon healing 34 , implying a possible role ESRRB may have in rotator cuff disease. FGF3 (fibroblast growth factor 3) and FGF10 (fibroblast growth factor 10) encode fibroblast growth factor proteins and are involved in a number of processes such as cell growth and tissue repair, including tendons, and could thus be associated with the pathogenesis of rotator cuff disease. Mutations in FGF3 have been linked with improper embryonic development of the inner ear 33 . Mutations in FGF10 can lead to aplasia of lacrimal and salivary glands 5 . FGFR1 encodes one of the receptors also associated with fibroblast growth factor; however, this gene is more specific to limb development. Mutations have been associated with cleft lip and cleft palate, Pfeiffer syndrome, and osteoglophonic dysplasia 18 , 22 , 36 . SAP30BP is implicated in cell death. SASH1 is a tumor suppressor gene implicated in a number of cancers 23 , 37 . Thus, many of the SNPs associated with rotator cuff disease have a potential biologic mechanism for their association with rotator cuff disease, but further research is needed in this area. A few studies on familial and genetic predisposition to rotator cuff disease have used controls from the general population, and have used cases and controls genotyped on different platforms in different experiments 30 – 32 . This is problematic since these studies assume that the prevalence of rotator cuff disease in the general population is low. This can also cause bias if the cases and the controls vary in other characteristics related to risk of rotator cuff disease, or if there are systematic differences in genotyping error between platforms. However, asymptomatic rotator cuff tears are demonstrated in 40% of persons over the age of 50 years, 54% in those over 60 years, and 65% in persons over 70 16 , 24 . Many of these patients will progress to a symptomatic rotator cuff tear in the future 15 . Hence, true controls are individuals from the general population that are asymptomatic and do not have structural evidence for a rotator cuff tear. The findings of previous studies provide evidence that there may be an important relationship between genes and rotator cuff disease. However, data on this issue is still limited. To our knowledge, only one study performed a genome wide association analysis (GWAS), which found two SNPs within genes SAP30BP and SASH1 associated with rotator cuff tears 30 . Both gene products, SAM and SH3 domain-containing protein 1 (SASH1) and SAP30-binding protein (SAP30BP), were reported in the process of cell apoptosis 30 . The effort to identify susceptibility genes of common multifactorial traits could lead to some insight into the pathogenesis mechanism that would, in turn, facilitate the development of better therapeutic and prophylactic approaches. However, many significant GWAS signals associated with multifactorial traits, e.g. type 2 diabetes, were mapped outside the coding gene sequence, which imposes a barrier to understanding their identities and function as well as limits their usefulness in experimental studies 2 , 10 . It is equally important to assess the reproducibility of reported genotype-phenotype association, as more of them failed to replicate. This was attributed to several issues, including inappropriate reliance on standard significant thresholds, small samples, and genotype and phenotype heterogeneity 11 , 12 . At the advent of GWAS, it appeared a promising proposal to estimate disease-risk by capturing the profile of common genetic variants. Nevertheless, the majority of common variants identified by GWAS only possess a very moderate effect size, and even the sum of these genetic effects only accounts for a minor portion of estimated trait heritability 26 . To address this issue, the effort has been made to identify the variants with lower frequency but higher penetrance by using imputed variants as well as to explore gene-gene and gene-environment interactions 26 . Further such studies would be beneficial, with the awareness that GWAS studies are designed to find common SNPs associated with complex traits and thus may not reveal conclusive information. Even if multiple SNPs are identified, they may still not fully explain the relationship between genes and rotator cuff disease. GWAS studies that identify SNPs would also need to be replicable and use a consistent definition of the clinical phenotype of rotator cuff tears.

There is data suggesting a genetic predisposition to rotator cuff disease. A large GWAS study with adequate controls could discover SNPs associated with symptomatic rotator cuff tears. The results from such a study could assist with early detection of individuals at risk of developing non-traumatic tears or suggest mechanisms of idiopathic rotator cuff disease. This may lead to medical treatments or prophylactic rehabilitation therapies to avoid development of symptomatic rotator cuff tears.

The term rotator cuff disease is used loosely in the literature. This term can encompass disorders ranging from impingement to tendinopathy to rotator cuff tearing. The transition from rotator cuff tendinopathy to rotator cuff tear was described as a continuum by Neer 19 . Hence, in our study we used the umbrella term rotator cuff disease and included studies on impingement syndrome and rotator cuff tendinopathy/tear. A systematic literature search on familial or genetic predisposition to rotator cuff disease of PubMed and EMBASE databases was performed from their years of inception through March 2016. The database search was performed with the help of a trained librarian, and the keywords used included “rotator cuff disease,” “genetics,” “polymorphism,” and “family.” The full search criteria can be found in the appendix (Appendix A). Initially, 251 citations were identified, and two of the authors (C.G. and N.B.J.) independently reviewed the titles and abstracts for relevance. The full texts of 17 of the citations were then reviewed, and 10 studies were found not to be relevant to our topic. Bibliographies of full text articles that met our inclusion criteria were also reviewed for additional articles. No additional articles were gained from the bibliography search. The studies included in this review were assessed with the Methodological Index for Non-Randomized Studies (MINORS) and were scored accordingly 25 . The maximum possible score was 24. When required, authors of included articles were contacted for additional information. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology for reporting our manuscript 13 .