This is the first study to identify and investigate the prevalence of pathogenic TTR variants in the population of Saudi Arabia, a population that is not represented in current databases. The investigation of TTR variants in a previously unstudied population would be expected to provide insights into the prevalence of known amyloidogenic variants and potentially identify novel variants that might be associated with systemic amyloidosis. Of the variants known to be associated with amyloidosis, c.424G > A(p.Val142Ile) was the most frequently identified in the Saudi population with an allele frequency of 0.001, similar to the frequency reported in a multinational database . This variant is most often detected in individuals of African descent, with approximately 3% of African-Americans carrying at least one copy of the gene (allele frequency 0.0173) . It is known to be associated with late onset CM, more commonly in men (Table 2) [8, 17–20]. Another known amyloidogenic variant, c.238A > G(p.Thr80Ala), was found at a frequency of 0.00004 in our database, suggesting that it is rare in the Saudi population. This variant is relatively common in Ireland and the UK with 1.1% of the northwest Irish population being carriers [15, 16]. It has also been found in other regions associated with Irish immigration, as well as in populations with no known Irish or UK ancestry . This variant is associated with cardiac amyloidosis as well as autonomic and peripheral neuropathy (Table 2) [21, 22].
Phenotypes of known amyloidogenic TTR variants with alleles identified in the Saudi population
c.424G > A(p.Val142Ile) [8, 13, 18–20]
• Late-onset cardiomyopathy (over the age of 65), more commonly in men
• Low amyloid load, remaining subclinical in many carriers
c.238A > G(p.Thr80Ala) [8, 15, 16, 21, 22]
• Age of onset in seventh decade of life
• Cardiac amyloidosis and autonomic and peripheral neuropathy
• Gastrointestinal disorders are common. Carpal tunnel syndrome precedes other symptoms in nearly three quarters of patients
c.239C > T(p.Thr80Ile) 
• Likely pathogenic, but there are no reports of the clinical manifestation
• It is similar to p.Thr80Ala, as both substitute a native uncharged polar residue with a nonpolar residue
Our study also identified two individuals with another known amyloidogenic variant, c.239C > T(p.Thr80Ile), with an allele frequency of 0.00007. There are few published reports of this variant, suggesting that it is rare. The phenotype associated with this variant has not been previously reported in the literature. It is absent in major databases (1000 Genomes Project, Exome Variant Server, Genome Aggregation Database), consistent with it being ethnically restricted. However, it is recorded in the Mutations in Hereditary Amyloidosis database as being amyloidogenic  and its similarity to the p.Thr80Ala variant suggests that its clinical implications might be predicted (Table 2). Interestingly, some of the most common variants identified in other populations are not observed in the Saudi database of 13,906 individuals. These include: c.148G > A(p.Val50Met), the most common variant in Western Europe ; p.Leu131Met, a variant with cardiac manifestations predominantly found in Denmark [8, 23]; p.Ile88Leu, which is seen predominantly in the Italian population [8, 24] and has primarily cardiac manifestations ; p.Val50Ala, p.Ala117Ser and p.Gly103Arg, the most common amyloidogenic variants in the Chinese population ; and p.Ser70Arg, the most common variant in Mexico . The absence of these variants from our database does not imply that they are absent in the Saudi population. However, if present, they are likely to have a very rare prevalence. Our findings further indicate that there is significant population variation in the prevalence of TTR variants, including some variants that are relatively restricted to certain ethnic groups. The fact that some of the TTR variants identified in the Saudi database have not been previously reported suggests that they might be restricted to the Saudi and possibly neighboring populations. Two of the three novel variants in our study were predicted by computational structural analysis to result in a reduced capacity of the TTR protein to associate with thyroid hormone or RBP4. These associations are thought to aid in the stabilization of the TTR tetramer and prevent it from dissociating into fibril-forming monomer aggregates [1–3]. Thus, it is possible that patients carrying these variations are at risk of developing future clinical amyloidosis. However, this in silico analysis may be considered as only “predictive” and further investigations are required to confirm the clinical implications of these novel variants.
The study is limited by sample size (13,906) which, while relatively large, will not identify very rare or de novo ATTR alleles within the population. The use of a cohort of 13,906 unrelated individuals, primarily sequenced to investigate rare inherited diseases globally, is not expected to introduce any bias in the ascertainment of allele frequencies. Individuals represented in our database originate from a large geographic area encompassing different regions within Saudi Arabia, thus allowing adequate estimation of TTR variants that are not vanishingly rare in the country [27, 28]. Another limitation of our study is that the interpretation of the clinical implication of novel pathogenic variants is restricted, as there is little clinical information and follow-up linked to the exome data. While one may predict that at least two of the three novel variants identified in our database are potentially amyloidogenic using structural analysis, without family histories and segregation, one cannot assume that such variants will ultimately result in clinical ATTR amyloidosis. This limitation also applies to variants of known function. How these data link to clinical manifestations in the population of Saudi Arabia remains unknown and, therefore, no comparison of penetrance or manifestations of identified variants between Saudi and other studied populations can be made. Such information could provide a valuable insight on how genetic background may influence the penetrance and manifestations of TTR variants. Furthermore, the prevalence of identified variants cannot be compared to the number of patients with a known diagnosis, precluding any attempts to determine if there is significant underdiagnosis in Saudi Arabia. Studies in other populations have noted the potential for extensive underdiagnosis [29, 30].
The data described in this manuscript suggest that there are TTR variants potentially associated with amyloidosis in Saudi Arabia and highlight the need for further clinical data regarding this patient population. Reports from other populations suggest that a concerted effort is required to identify, monitor and manage individuals with pathogenic TTR variants. This approach would likely allow therapeutic intervention before considerable deposition of amyloid fibrils induces symptoms and advanced organ damage [12, 31]. In Saudi Arabia, such efforts could focus on: the routine genetic testing of patients with phenotypes that raise suspicion of amyloidosis irrespective of age; the development of genetic testing programs for relatives of patients with known ATTR amyloidosis and the utilization of existing newborn and pre-marital genetic screening programs to identify carriers of pathogenic TTR variants; the development of a consensus on optimal monitoring and management of patients with pathogenic TTR variants, including appropriate genetic counseling for family members; and finally, establishing a registry for patients with amyloidosis. In individuals with a phenotype suspicious for amyloidosis such as hypertrophic cardiomyopathy, the presence of a TTR mutation has diagnostic implications and points away from a myocardial sarcomeric disease and towards amyloidosis. The identification of potentially pathogenic TTR mutations has important clinical implications for the classification, diagnosis, and treatment of amyloidosis. Within the disease context, the detection of a mutation allows one to classify TTR amyloidosis as hereditary rather than wild-type and should prompt consideration for genetic screening of siblings. Genetic screening conducted on siblings of patients with the hereditary form allows for the detection of mutation carriers who are at risk of developing future clinical amyloidosis. Saudi Arabia benefits from extensive existing screening programs, meaning that there is infrastructure already in place to facilitate screening for TTR variants in people known to be at risk and in the wider population.
The aim of these screening programs is to identify patients before they become symptomatic. A recent article by Conceição and colleagues  on early diagnosis and follow-up cautioned that it is important to minimize any anxiety that could be caused through over-medicalization from the knowledge of carrier status and repeated follow-up. In order to determine how these patients should be monitored in Saudi Arabia, it is therefore important to carefully consider and gain consensus on when patients should begin to be monitored and at what time intervals; which clinicians and other healthcare professionals should be involved; what should be done for patients with TTR variants of currently uncertain effect; when treatment should be initiated; and which tests should be performed to monitor patients. In this context, biomarkers such as serum TTR levels or urine RBP4  have potential for more widespread use in the clinic. Furthermore, current imaging tools can allow the detection of organ involvement at a subclinical stage, before overt morphological abnormalities become evident. For example, cardiac imaging using Tc-labelled phosphate compounds with high affinity for TTR allows early diagnosis of amyloid myocardial deposition, not only at a pre-symptomatic stage, but also before an increase in wall thickness is detected by echocardiography and prior to the development of electrocardiographic abnormalities . A recently published Japanese expert opinion recommended periodic clinical assessment as well as various investigations in monitoring asymptomatic gene mutation carriers . With the advent of drugs that help stabilize the TTR tetramer, reduce tissue deposition and possibly slow disease progression,  an early diagnosis guided by genetic screening should be strongly pursued. Such drugs may well prove to be more effective if administered before the development of overt tissue involvement and organ damage. Figure 3 illustrates a framework for the initial assessment and follow-up of individuals who are asymptomatic carriers of TTR mutations.