Spectrum of MEFV variants in 8,252 patients from Eastern and Southeastern Anatolia: a large single-center cohort study (2019–2026)

Abstract

Background

Familial Mediterranean Fever (FMF) is a monogenic autoinflammatory disorder caused by variants in MEFV. Although Türkiye carries a considerable FMF burden, large contemporary single-center genetic datasets from Eastern and Southeastern Anatolia are still limited. Because the frequency and distribution of MEFV variants differ across populations and regions, locally derived reference data remain important for clinical interpretation.

Objective

To characterize the MEFV variant spectrum and genotype categories in a large single-center cohort representing Eastern and Southeastern Anatolia.

Methods

We performed a retrospective review of 8,252 consecutive referrals tested for recurrent MEFV common variants at a tertiary genetics laboratory between 2019 and 2026. Demographic variables and genotype results were retrieved from the laboratory information system. Results were grouped as wild-type (WT), monoallelic, or multi-allelic, and multi-allelic findings were further classified as homozygous, compound heterozygous, or complex multi-site configurations.

Results

The median age was 17 years (mean 22.0 ± 16.9 years; range 0–84), and 53.8% of individuals were female. Genotype categories were distributed as follows: WT 60.4% (4,985/8,252), monoallelic 28.2% (2,327/8,252), and multi-allelic 11.4% (940/8,252), including homozygous 3.8% (311/8,252), compound heterozygous 6.9% (569/8,252), and complex multi-site 0.7% (60/8,252). The most frequent variants were M694V 14.5% (1,199/8,252), E148Q 14.0% (1,158/8,252), V726A 5.2% (426/8,252), and M680I (G > C) 4.1% (337/8,252). Exon 10 variants accounted for a substantial proportion of variant-positive findings, while E148Q was common but predominantly heterozygous. Variant rankings were broadly similar between females and males.

Conclusions

This study provides a large region-specific reference for MEFV variation in Eastern and Southeastern Anatolia. The predominance of M694V and other exon 10 variants, together with the high background frequency of E148Q, has direct relevance for pre-test counseling, targeted testing strategies, and interpretation of routine FMF results. Further work incorporating sequencing-based assays and standardized clinical follow-up will be necessary to refine genotype-phenotype relationships and improve variant interpretation in this high-burden setting.

Introduction

Familial Mediterranean fever (FMF) is the most common monogenic autoinflammatory disease and is caused by variants in the MEFV gene, which encodes pyrin, a key regulator of innate immune signaling and inflammasome activity [1–2]. Clinically, FMF is characterized by recurrent episodes of fever, serositis, pleuritis, arthritis, and inflammatory pain syndromes, although disease expression is highly variable and may range from classical childhood-onset attacks to milder or atypical adult presentations [1, 3–4]. This variability reflects not only clinical heterogeneity, but also differences in the underlying genetic background and in the way specific variants are interpreted across populations.

Türkiye remains one of the countries with the highest FMF burden, and MEFV testing is part of routine diagnostic work in both pediatric and adult practice [3, 5–9]. Across Turkish cohorts, M694V has consistently been reported as the dominant or one of the dominant variants, while M680I, V726A, E148Q, and several lower-frequency alleles contribute to the broader mutational spectrum [5–18]. At the same time, studies from different parts of the country have shown that mutation frequencies are not uniform. Reports from East Anatolia, southeast Turkey, the Aegean region, Central Anatolia, the Black Sea region, Van, Adıyaman, and Hatay all suggest measurable regional variation in both allele ranking and genotype distribution [5–7, 11–18]. These differences are clinically relevant because they influence which variants are most likely to be encountered in daily practice and how strongly a given result supports the diagnosis in a local referral population.

The interpretation of MEFV results is further complicated by the coexistence of clearly pathogenic exon 10 variants with more context-dependent alleles. Variants such as M694V, M680I, and V726A have long been associated with typical FMF phenotypes and, in some settings, with more severe disease or increased complication risk, including amyloidosis [3, 5, 7, 19–20]. In contrast, the significance of variants such as E148Q remains less straightforward, particularly when detected in the heterozygous state and outside a strong clinical context [1, 4, 20]. For that reason, population-specific reference data remain important not only for epidemiologic description but also for the practical interpretation of positive laboratory findings.

Although national and multicenter Turkish studies have substantially improved understanding of the country-wide MEFV spectrum [8–9, 12], large recent single-center datasets focused specifically on Eastern and Southeastern Anatolia are still limited. This matters because pooled national estimates may obscure local gradients created by referral structure, regional ancestry patterns, and assay-specific practice. The present study was therefore designed to characterize the MEFV variant spectrum and genotype categories in a large contemporary single-center cohort representing Eastern and Southeastern Anatolia. By defining the relative frequencies of common variants and the distribution of wild-type, monoallelic, and multi-allelic genotypes, we aimed to provide a practical regional reference for FMF testing in this high-burden setting.

Materials and methods

This retrospective laboratory-based cohort study included 8,252 consecutive unrelated individuals referred for MEFV analysis to a tertiary medical genetics laboratory serving Eastern and Southeastern Anatolia between 2019 and 2026. The study population reflected routine clinical referrals for suspected Familial Mediterranean Fever or related indications, and no additional selection criteria were applied beyond the availability of complete laboratory records. Because the dataset was derived from routine diagnostic practice, inclusion was based on test referral rather than on a uniform clinical classification system.

Laboratory work was performed using the MEFV FragPure Kit v2.1 (Maygen, Ankara, Türkiye) (catalog no. F2756), a fragment analysis assay performed on a capillary electrophoresis platform. Variant nomenclature was reported according to the MEFV reference transcript NM_000243.3, and the assay targeted the following 18 variants across exons 2, 3, 5, and 10: E148Q (c.442G > C; p.(Glu148Gln)); R202Q (c.605G > A; p.(Arg202Gln)); S179I (c.536G > T; p.(Ser179Ile)); E167D (c.501G > C; p.(Glu167Asp)); P369S (c.1105 C > T; p.(Pro369Ser)); P350R (c.1049 C > G; p.(Pro350Arg)); Y471X (c.1413 C > A; p.(Tyr471)); F479L (c.1437 C > G; p.(Phe479Leu)); M694V (c.2080 A > G; p.(Met694Val)); A744S (c.2230G > T; p.(Ala744Ser)); M680I (c.2040G > C / c.2040G > T; p.(Met680Ile)); G632A (c.1895G > C; p.(Gly632Ala)); V726A (c.2177T > C; p.(Val726Ala)); K695R (c.2084 A > G; p.(Lys695Arg)); K695N (c.2085G > C; p.(Lys695Asn)); M694I (c.2082G > A; p.(Met694Ile)); R761H (c.2282G > A; p.(Arg761His)); and I692del (c.2076_2078del; p.(Ile692del)); for M680I, the amino-acid change may correspond to either c.2040G > C or c.2040G > T, both of which encode p.(Met680Ile) in NM_000243.3.

De-identified demographic and laboratory data were retrieved from the institutional laboratory information system. Extracted variables included age, sex, and the final reported MEFV genotype result. For individuals with more than one reported variant, all common findings were reviewed at the patient level and categorized according to the final laboratory result. The analysis was based on a targeted MEFV assay routinely used in the diagnostic laboratory during the study period. Accordingly, the study reflects the distribution of recurrent variants included in the routine panel and does not capture rare variants outside the scope of that assay.

For cohort-level analysis, genotype results were classified as wild-type (WT) when no variant was detected, monoallelic when a single heterozygous variant was present, and multi-allelic when more than one variant finding was identified in the same individual. Multi-allelic results were further subdivided into homozygous, compound heterozygous, and complex multi-site configurations in order to preserve clinically relevant differences in zygosity and allelic composition.

Genotype classification was performed based on the presence and combination of pathogenic MEFV variants. Homozygous genotypes were defined as the presence of two identical pathogenic variants. Compound heterozygous genotypes were defined as the presence of two different pathogenic variants. Variants with high population frequency and low or uncertain pathogenicity, such as R202Q and E148Q, were not used as primary determinants of genotype classification but were recorded separately. In cases where more than two variants were detected (e.g., three or four variants), classification was based on the pathogenic exon 10 variants. For instance, genotypes such as R202Q–M694V / M694V were classified as homozygous for M694V, while additional variants were considered secondary findings.

The analysis was primarily descriptive. Continuous variables were summarized using mean, standard deviation, median, and range, while categorical variables were presented as counts and percentages. Variant frequencies were calculated across the full cohort, and genotype category distributions were summarized for the overall population and according to sex. Additional descriptive analyses examined the heterozygous and homozygous distribution of the most frequent variants. The purpose of the analysis was to define a regional reference profile of MEFV variation in a large real-world referral population rather than to perform formal inferential testing.

The study was conducted using fully de-identified laboratory data generated during routine clinical practice. No direct patient contact or intervention was involved. Data handling was performed in accordance with institutional confidentiality principles and applicable ethical standards. Ethics committee approval and consent procedures should be stated in the final manuscript according to the requirements of the relevant institution and target journal.

Results

A total of 8,252 individuals underwent MEFV analysis during the study period. The median age was 17 years (mean 22.0 ± 16.9 years; range 0–84), and 53.8% of the cohort were female (4,442/8,252). Cohort characteristics are shown in Table 1. Overall, 60.4% of individuals (4,985/8,252) were classified as wild-type, 28.2% (2,327/8,252) had a monoallelic finding, and 11.4% (940/8,252) had a multi-allelic genotype. Among the multi-allelic group, 311 individuals (3.8%) were homozygous, 569 (6.9%) were compound heterozygous, and 60 (0.7%) showed complex multi-site configurations (Table 2).

Table 1.

Cohort characteristics

Variable	Value
Total cohort size	8,252
Median age (years)	17
Mean age ± SD (years)	22.0 ± 16.9
Age range (years)	0–84
Female, n (%)	4,442 (53.8%)
Male, n (%)	3,810 (46.2%)

Table 2.

Genotype category distribution

Category	N	%
Wild-type (WT)	4985	60.4
Monoallelic	2327	28.2
Multi-allelic (total)	940	11.4
- Homozygous	311	3.8
- Compound heterozygous	569	6.9
- Complex multi-site	60	0.7

Note: This table reflects full panel-based variant distribution. A separate restricted analysis including only pathogenic exon 10 variants is presented in the Results section

The genotype distributions presented in Tables 2 and 3 reflect the full panel-based results, including variants with variable or uncertain clinical significance.

Table 3.

Summary of common variant zygosity patterns

Variant	Non-WT	HET	HOMO	Frequency
M694V	1199	1012	187	14.53
E148Q	1158	1112	46	14.03
V726A	426	408	18	5.16
M680I (G>C)	337	295	42	4.08
P369S	221	214	7	2.68
R761H	192	176	16	2.33
A744S	144	142	2	1.75
M694I	116	100	16	1.41
R408Q	58	54	4	0.70
K695R	43	43	0	0.52
Total	3894	3556	338	47.19

HOMO/non-WT (%) was calculated as homozygous carriers divided by total carriers for each variant

To address this, an additional restricted analysis was performed considering only clearly pathogenic exon 10 variants (M694V, M680I, V726A, M694I, and R761H). Under this definition, 282 individuals (3.42%) were classified as homozygous and 252 (3.05%) as compound heterozygous. This contrasts with the broader dataset, in which compound heterozygous configurations appear more frequent due to the inclusion of low-penetrance variants. Under this stricter definition, homozygous genotypes slightly outnumbered compound heterozygous genotypes.

The variant spectrum was dominated by M694V, which was detected in 1,199 individuals (14.5%), followed closely by E148Q in 1,158 (14.0%). The next most frequent variants were V726A (426, 5.2%) and M680I (G > C) (337, 4.1%), followed by P369S (221, 2.7%), R761H (192, 2.3%), A744S (144, 1.7%), M694I (116, 1.4%), R408Q (58, 0.7%), and K695R (43, 0.5%) (Table 5). Together, these findings indicate that the cohort was largely shaped by M694V and E148Q, with additional contribution from classical exon 10 variants, particularly V726A and M680I (G > C).

Table 5.

Comparison of our cohort with selected Turkish MEFV studies

Study	n	M694V	E148Q	V726A	M680I (G > C)	General pattern
Current study	8,252	14.53% (1,199)	14.03% (1,158)	5.16% (426)	4.08% (337)	M694V and E148Q were nearly equal; V726A and M680I followed
Akin et al., 2010	1,201	47.60% (572)	16.75% (201)	12.95% (156)	11.94% (143)	Strong M694V predominance
Bilge et al., 2019	2246	31% (718)	9.5% (214)	9.2% (315)	12.3% (423)	Marked exon 10 predominance
Aksoy et al., 2024	2,984	25.1% (1,004)	11.5% (463)	7.9% (319)	8.8% (354)	Similar core spectrum; M694V predominant
Yilmaz et al., 2016	1,435	48.79% (700)	13.70% (197)	12.35% (177)	14.86% (213)	Strong M694V predominance

Note. Values in parentheses indicate counts

Across the most frequent variants, heterozygous calls clearly predominated, including for exon 10 variants, contrary to an overly simplified interpretation of their clinical severity. When examined quantitatively, the combined proportion of homozygous genotypes for the major pathogenic exon 10 variants (M694V, M680I, V726A, M694I, and R761H) corresponded to 282 individuals (3.42% of the total cohort). In contrast, the exon 2 variant E148Q, which is generally considered context-dependent or of lower penetrance, was identified in 1,158 individuals (14.0%), with a strong predominance of heterozygous carriers (1112/1158, 96.0%). These findings indicate that, although exon 10 variants are central to FMF pathogenesis, their homozygous forms represent a relatively small fraction of the overall tested population, whereas E148Q constitutes a large proportion of variant-positive results, primarily in the heterozygous state.

Sex-stratified analysis showed only modest differences in genotype distribution. Among females, 61.4% were wild-type, 27.8% had monoallelic findings, and 10.7% had multi-allelic genotypes. Among males, the corresponding proportions were 59.2%, 28.6%, and 12.2%, respectively (Table 4). Despite these small differences in category frequencies, the overall variant ranking remained similar between the sexes, with M694V and E148Q remaining the leading findings in both groups.

Table 4.

Distribution of genotype categories stratified by sex

Sex	WT (n, %)	Monoallelic (n, %)	Multi-allelic (n, %)
Female	2728 (61.4%)	1237 (27.8%)	477 (10.7%)
Male	2257 (59.2%)	1090 (28.6%)	463 (12.2%)

Discussion

The present cohort provides a detailed regional picture of MEFV variation in Eastern and Southeastern Anatolia. In 8,252 consecutive referrals, most individuals were wild-type, while a smaller but clinically important subset carried monoallelic or multi-allelic findings. The variant profile was led by M694V, closely followed by E148Q, with V726A and M680I (G > C) forming the next most frequent group. This pattern places exon 10 variants at the center of the regional mutation landscape, while also showing that a considerable fraction of routine laboratory positivity is attributable to variants whose clinical meaning depends heavily on context.

Importantly, when applying a stricter definition limited to biallelic pathogenic exon 10 variants, the distribution of genotypes differed from the overall descriptive analysis. In this restricted framework, homozygous genotypes were slightly more frequent than compound heterozygous genotypes, indicating that the apparent predominance of compound heterozygosity in the full dataset is largely influenced by the inclusion of low-penetrance or context-dependent variants. This distinction highlights the importance of differentiating between raw laboratory outputs and clinically meaningful genotype classifications. While the full dataset reflects real-world testing results, inclusion of variants such as E148Q may inflate the apparent proportion of compound heterozygous genotypes. When analysis is restricted to clearly pathogenic exon 10 variants, the genotype distribution shifts, with homozygous genotypes slightly exceeding compound heterozygous ones.

This distribution is compatible with the broader Turkish FMF literature, where M694V repeatedly appears as the dominant allele in both pediatric and adult cohorts [3, 5–18, 21]. Studies from East Anatolia, southeast Turkey, and other Turkish regions have consistently shown that M694V, M680I, and V726A represent the most clinically relevant part of the classical FMF spectrum [5–7, 10–16, 18]. Our results fit well within that framework. In particular, the prominence of M694V in the overall cohort and its relative enrichment among homozygous results support the long-standing view that exon 10 variants remain the most important contributors to genetically supported FMF in high-prevalence Turkish populations.

Overall, this comparison shows that our cohort shares the core Turkish MEFV mutation pattern reported in previous studies. In all compared cohorts, M694V was the leading variant, confirming its central role in the Turkish FMF spectrum. Our data also overlap with earlier reports in showing that E148Q, V726A, and M680I (G > C) constitute the main accompanying variants. The main distinction of our series is not a completely different variant profile, but rather the more balanced distribution between M694V and E148Q, whereas several previous Turkish studies demonstrated a more pronounced M694V-dominant and exon 10–weighted pattern. In this respect, our findings support the overall national mutation framework while also suggesting that regional cohorts may differ in the relative contribution of individual variants. These results further underline the value of region-specific reference data for the interpretation of MEFV testing in routine clinical practice (Table 5).

At the same time, the present data show why local datasets remain necessary even in the presence of large national studies. The National Genetics Consortium analysis and multicenter Anatolia-wide reports have been invaluable in defining the overall Turkish mutation profile, yet they also demonstrate that allele frequencies are not distributed uniformly across the country [8–9, 12]. Studies from the Aegean region, Central Anatolia, the Black Sea region, Van, Adıyaman, and Hatay all point to measurable regional shifts in allele ranking and relative burden [6, 11, 13–18]. Our findings reinforce that message. The mutation spectrum in this catchment overlaps with national Turkish experience, but the balance between M694V, E148Q, V726A, and M680I (G > C) is best understood as a regional pattern rather than as a simple reflection of a pooled national average. For clinicians and diagnostic laboratories, that distinction is not merely epidemiologic; it directly affects how often certain variants are encountered, how they are prioritized in targeted testing, and how cautiously they should be interpreted.

Beyond descriptive variant frequencies, the regional MEFV repertoire observed in this cohort has important practical implications for variant interpretation, screening strategies, and genetic counseling in Eastern and Southeastern Anatolia. In this high-prevalence setting, the penetrance of debated variants such as E148Q is unlikely to be determined solely by the variant itself, but rather by a combination of factors including zygosity, co-occurrence with clearly pathogenic exon 10 variants, clinical phenotype, and regional background frequency. In our cohort, the high frequency of E148Q—predominantly in the heterozygous state—supports its interpretation as a low-penetrance or context-dependent variant rather than a stand-alone disease-defining mutation.

These findings argue for a calibrated approach to MEFV interpretation in this region. Specifically, clearly pathogenic exon 10 variants (M694V, M680I, V726A, M694I, and R761H) should carry the greatest diagnostic weight, whereas isolated E148Q should be interpreted cautiously and in conjunction with clinical findings. This distinction is particularly important in referral-based cohorts, where background variant frequency may be high and overinterpretation of low-penetrance variants may lead to diagnostic uncertainty. In this cohort, E148Q was the second most frequent variant and was seen predominantly in the heterozygous state. Similar observations have been reported in several Turkish and international series [6, 8, 13, 17–18, 20, 22–23]. Yet, unlike the major exon 10 alleles, E148Q does not carry the same interpretive weight when encountered in isolation. Previous work has shown that genotype-symptom correlations are more robust for well-established pathogenic variants, particularly exon 10 mutations [20]. Clinical reviews have also emphasized that FMF cannot be reduced to genotype alone, especially in patients with atypical presentations or borderline variants [1, 4]. Our results support that position. The frequency of E148Q in routine testing is too high to ignore, but its predominance in heterozygous form argues against automatic causal interpretation without clinical correlation, family history, segregation data, or supportive co-variation with clearly pathogenic alleles.

A similar caution applies when comparing our results with Turkish studies that highlight R202Q. Several recent series have reported substantial frequencies of that variant [17–18, 23], whereas R202Q was not prioritised in the current study, because this variant is widely regarded as benign or likely benign due to its high allele frequency in the general population, including both Turkish and European cohorts. Similarly, E148Q is considered a low-penetrance or context-dependent variant. For this reason, we adopt a calibrated approach to the interpretation of R202Q and E148Q, in which these variants are evaluated in the context of zygosity, co-occurrence with known pathogenic exon 10 variants, and clinical findings. Isolated presence, particularly in the heterozygous state, should not be overinterpreted as disease-causing.

From a practical perspective, our data suggest that region-adapted screening strategies should prioritize clinically meaningful discrimination rather than maximal variant detection. In patients with strong clinical suspicion but inconclusive panel results, broader sequencing-based approaches may be more informative than reliance on limited variant panels. Similarly, genetic counseling in this region should explicitly address variable penetrance, the limited predictive value of certain variants in isolation, and the importance of integrating genetic findings with clinical context.

Finally, the variability of FMF expression in this cohort also supports the role of non-MEFV modifiers in shaping disease intensity. Region-specific factors such as family structure, possible consanguinity patterns, diagnostic delay, environmental inflammatory triggers, and access or adherence to colchicine therapy may all contribute to clinical variability. Although our study was not designed to directly evaluate these factors, acknowledging their potential influence is essential for a realistic and clinically applicable interpretation of MEFV results in this setting.

Additionally, we acknowledge that comparisons with previously published Turkish cohorts must be interpreted cautiously, as some studies are based on clinically confirmed FMF patients, whereas others, including ours, are referral-based genetic testing cohorts. These differences inherently influence variant heterogeneity and frequency distributions. Notably, when R202Q is excluded from the analysis, the remaining variant distribution in our cohort shows substantial similarity to previously reported national MEFV profiles, supporting the notion that the observed heterogeneity is largely driven by high-frequency benign variants in referral populations.

The internal structure of genotype categories in our cohort is also noteworthy. That pattern broadly agrees with previous Turkish reports showing that a large proportion of referred patients carry only a single detectable variant, whereas a smaller subgroup carries more diagnostically significant multi-allelic genotypes [3, 6–9, 14–18, 21]. From a practical standpoint, this matters because disease behavior is influenced not only by the identity of the variant, but also by its zygosity and its allelic partner. Earlier studies have repeatedly linked M694V, especially in homozygous form, to more typical and sometimes more severe disease [3, 5, 7], and work on FMF-associated AA amyloidosis has underscored the importance of pathogenic exon 10 genotypes in long-term risk assessment [19]. Although our study was not designed to evaluate outcomes, the zygosity distribution observed here supports reporting practices that clearly distinguish monoallelic, homozygous, and compound heterozygous states rather than collapsing all positive findings into a single category.

When compared with non-Turkish studies, our cohort shows both shared regional characteristics and clear population-specific differences. The overall structure of our data, particularly the predominance of M694V together with the contribution of other classical exon 10 variants such as V726A and M680I (G > C), appears closer to reports from Iran and other nearby high-prevalence populations than to more genetically distant cohorts [1, 24]. Iranian registry data likewise identify M694V as a major component of the FMF mutation spectrum, supporting the view that exon 10-centered profiles remain a common feature in populations around the eastern Mediterranean and adjacent regions [24]. In contrast, Egyptian studies have reported a different hierarchy, with M694I and E148Q occupying a more prominent position, indicating that even geographically related populations may show substantial shifts in allele distribution [22]. An even clearer contrast is seen in Japanese cohorts, where FMF is relatively uncommon and the reported genetic pattern is characterized more by E148Q/M694I-related genotypes than by a dominant M694V background [25]. These differences suggest that the variant profile observed in our cohort is not simply a universal FMF pattern, but rather a regionally shaped distribution reflecting the population background of Eastern and Southeastern Anatolia. At the same time, the partial overlap with Iranian data indicates that neighboring high-burden populations may share a common pathogenic core, even though the relative contribution of individual variants differs across countries [1, 24]. Taken together, these comparisons place our findings within the broader international FMF literature while also underscoring the importance of population-specific reference data for the accurate interpretation of MEFV results in clinical practice [1, 22, 24–25].

The study has clear strengths. It represents, to our knowledge, the largest single-center MEFV cohort centered specifically on Eastern and Southeastern Anatolia. The size of the cohort allows more stable estimation of both common and less frequent variants, including alleles such as A744S, R408Q, and K695R, which are often difficult to assess in smaller reports. In addition, the use of a consistent laboratory workflow across the entire series reduces inter-laboratory variability and makes the data immediately relevant for real-world regional diagnostics.

The limitations are also important. This was a retrospective laboratory-based study, and standardized phenotypic information was not available for all referrals. As a result, genotype-phenotype conclusions must remain conservative. The use of common variant testing also means that rare coding variants, deep intronic changes, and structural alterations were outside the scope of the assay. In patients with strong clinical suspicion but limited findings, broader sequencing-based approaches will remain necessary. Finally, comparisons with other Turkish cohorts should always be made with attention to panel design, inclusion criteria, and reporting practice, because methodological differences can create apparent discrepancies even when the underlying population burden is similar.

An important limitation of this study relates to the referral-based nature of the cohort. The population analyzed here does not represent a clinically confirmed FMF cohort, but rather a real-world group of individuals referred for genetic testing with varying levels of clinical suspicion. As a result, the dataset may be influenced by referral bias. Patients with highly typical FMF presentations may be diagnosed clinically and therefore underrepresented, whereas individuals with nonspecific or overlapping symptoms may be overrepresented. In regions such as ours, where other inflammatory disorders are relatively prevalent, such overlap may further affect the composition of the tested population. In addition, the tertiary-care setting may enrich for diagnostically unresolved or atypical cases.

For these reasons, our findings should not be interpreted as defining a definitive regional genetic makeup of FMF. Instead, they are better understood as reflecting the spectrum of MEFV variants encountered in routine diagnostic practice in a high-prevalence region. While this approach has inherent limitations for genotype–phenotype inference, it provides practical value for clinicians and laboratories by illustrating how genetic results are distributed in real-world testing scenarios. Future studies incorporating clinically well-defined FMF cohorts will be essential to refine population-specific genotype–phenotype relationships.

In summary, this cohort provides a regional reference for MEFV testing in a referral-based cohort for Eastern and Southeastern Anatolia, marked by the predominance of M694V and other exon 10 variants, together with a high background frequency of E148Q. The relative distribution of genotype categories depends on the analytical framework. While compound heterozygous configurations appear more frequent in the full panel-based dataset, a restricted analysis limited to clearly pathogenic exon 10 variants shows that homozygous genotypes slightly exceed compound heterozygous genotypes. Read alongside regional Turkish studies, national consortium data, and international series [1–26], these results provide a practical reference for counseling, targeted testing, and cautious interpretation of context-dependent variants in daily FMF practice.

Conclusion

In conclusion, this large single-center cohort defines a reference MEFV testing profile for Eastern and Southeastern Anatolia. The data show that M694V and other exon 10 variants remain the dominant components of the local mutation spectrum, while E148Q represents a frequent background finding that should be interpreted with appropriate clinical caution. Taken together, these findings provide a useful regional reference for clinicians and diagnostic laboratories and support the use of population-specific genetic baselines in the evaluation of FMF. Future studies integrating broader sequencing approaches with standardized phenotypic and longitudinal clinical data will be essential for improving genotype-phenotype resolution and guiding more precise FMF management.

Author contributions

All authors contributed to study design, data interpretation, manuscript drafting or revision, and approved the final version. U.G. performed data extraction and statistical analyses. A.I.G. performed the control of the relevance of the literatüre data and last checks.

Funding

This research received no external funding. All analyses were performed using institutional resources at the Gaziantep City Hospital Department of Medical Genetics. The funders had no role in study design, data collection, analysis, interpretation, or manuscript preparation.

Data availability

The dataset analyzed in this study consists of de-identified laboratory records generated as part of routine MEFV common variant testing at a single tertiary genetics laboratory. Due to institutional and national regulations governing the sharing of clinical genetic data, the raw individual-level dataset cannot be made publicly available. However, aggregated data supporting the findings of this study—including variant frequencies, genotype category counts, and sex-stratified summaries—are available from the corresponding author upon reasonable request, pending approval by the institutional data governance board. All analyses, figures, and summary tables reported in this manuscript were derived from these de-identified data.

Declarations

Ethics approval

This retrospective study was conducted using fully de-identified laboratory data. Ethics committee approval and a waiver of informed consent were obtained from the Gaziantep City Hospital Ethics Committee prior to the conduct of the study, in accordance with national regulations and institutional policy. All procedures adhered to the ethical standards of the Helsinki Declaration and applicable regulatory requirements.

Competing interests

The authors declare no competing interests.