ARTHRITIS & RHEUMATISM
Vol. 62, No. 10, October 2010, pp 3059–3063
DOI 10.1002/art.27598
© 2010, American College of Rheumatology
Increased Prevalence of M694V in Patients With
Ankylosing Spondylitis
Additional Evidence for a Link With Familial Mediterranean Fever
Nurullah Akkoc,1 Ismail Sari,1 Servet Akar,1 Omer Binicier,1 Mark G. Thomas,2
Michael E. Weale,3 Merih Birlik,1 Yusuf Savran,1 Fatos Onen,1
Neil Bradman,4 and Christopher A. Plaster4
controls, P ⴝ 0.046 for AS patients versus RA patient
controls, and P ⴝ 0.008 for AS patients versus healthy
and RA patient control groups). The carriage rate of
M694V was also significantly higher in the AS patient
group than in the combined control group (odds ratio
7.0, P ⴝ 0.014). Neither M694V nor any other MEFV
variant showed a correlation with most of the diseaserelated measures examined.
Conclusion. We found an increased frequency of
MEFV variants in AS patients as compared with healthy
controls and with RA patient controls. This was primarily due to the presence of M694V. The roles of other exon
10 variants, as well as the relationship between the
variant status and the severity and clinical course of the
disease, need to be explored in further studies that
include sufficiently large sample sizes.
Objective. To assess whether there is a statistically significant difference in the frequency of common
MEFV allele variants in patients with ankylosing spondylitis (AS) as compared with control patients with
rheumatoid arthritis (RA) and with healthy control
subjects.
Methods. Sixty-two patients with AS, 50 healthy
control subjects, and 46 patients with RA were assessed
for the presence of MEFV variants. Exon 10 was analyzed by direct sequencing. E148Q was analyzed by
restriction endonuclease enzyme digestion (REED) or
by direct sequencing when REED analysis failed.
Results. The allele frequency of all MEFV variants
in the AS group was significantly higher than that in the
pooled control group of healthy subjects plus RA patients (15.3% versus 6.8%; P ⴝ 0.021). M694V was the
only variant that was significantly more common in the
AS group than in the combined or individual control
groups (P ⴝ 0.026 for AS patients versus healthy
Familial Mediterranean fever (FMF) is an autoinflammatory disease that occurs predominantly in Jewish, Armenian, Turkish, and Middle Eastern Arab populations. It is characterized by recurrent attacks of fever
and serositis. The disease is caused by MEFV, which
encodes an immunoregulatory protein called pyrin, or
marenostrin, and is inherited in an autosomal-recessive
manner (1,2).
We previously identified 3 patients with ankylosing spondylitis (AS) carrying 2 MEFV variants, 2 of
whom were homozygous for M694V (3). None of the
patients had any current or previous symptoms suggestive of FMF, and all were HLA–B27 negative. These
cases are of particular interest in light of recent reports
indicating an increased frequency of sacroiliitis or spondylarthritis in patients with FMF (4,5).
In the present study, we assessed the prevalence
Supported in part by the Melford Charitable Trust.
1
Nurullah Akkoc, MD, Ismail Sari, MD, Servet Akar, MD,
Omer Binicier, MD, Merih Birlik, MD, Yusuf Savran, MD, Fatos
Onen, MD: Dokuz Eylül University School of Medicine, Izmir,
Turkey; 2Mark G. Thomas, PhD: University College London, London,
UK, and Uppsala University, Uppsala, Sweden; 3Michael E. Weale,
PhD: King’s College London, Guy’s Hospital, and University College
London, London, UK; 4Neil Bradman, DPhil, MSc, Christopher A.
Plaster, MSc: University College London, London, UK.
Dr. Bradman is a trustee of the Melford Charitable Trust,
which has no intellectual property or other rights whatsoever with
respect to this research.
Address correspondence and reprint requests to Nurullah
Akkoc, MD, Dokuz Eylül Universitesi Tip Fakultesi, Romatoloji Bilim
Dali, 35340 Balcova, Izmir, Turkey. E-mail: nurullah.akkoc@
deu.edu.tr.
Submitted for publication July 17, 2009; accepted in revised
form June 1, 2010.
3059
3060
AKKOC ET AL
of common MEFV variants (E148Q in exon 2 and exon
10 variants) and compared their frequencies in AS
patients with those in healthy control subjects as well as
in patients with a different inflammatory rheumatic
disease. We also evaluated whether the variant status
was correlated with clinical findings and disease-related
measures.
PATIENTS AND METHODS
Selection of patients and controls. Sixty-four consecutive patients with AS diagnosed according to the modified
New York criteria (6) were enrolled in the study. For comparison, 51 healthy control subjects and 46 disease controls with
rheumatoid arthritis (RA) diagnosed according to the 1987
revised criteria of the American College of Rheumatology
(formerly, the American Rheumatism Association) (7) were
also included in the study. One AS patient (positive for
M694V) and 1 RA patient (negative for MEFV variants) were
excluded because they had first-degree relatives who were
already enrolled in the study. The healthy control subjects were
volunteers who had no history, symptoms, or signs of rheumatic disease. These subjects were recruited from blood
donors, hospital staff, and another outpatient clinic associated
with Dokuz Eylül University.
None of the study subjects reported any symptoms
suggestive of FMF, nor did any of them have a family history
of FMF. All study subjects gave their informed consent to
participate in the study, and ethics committee approval was
obtained.
Results of the baseline evaluations of the 62 AS
patients (demographic, clinical, and laboratory features) are
provided in Table 1.
Genetic analysis. Peripheral venous blood (10 ml) was
drawn into tubes containing EDTA and was stored at –20°C.
MEFV exon 10 sequencing method. MEFV (exon 10)
was amplified using the primers MEFV_Ex10B_U (5⬘TCCTTGTCTTTCCTTGTTGTCAG-3⬘) and MEFV_
Ex10C_L (5⬘-CGTGGGCACAGTAACTATTTTGT-3⬘). Se-
Table 1. Baseline demographics and clinical features of the 62 study
patients with ankylosing spondylitis*
Age, mean ⫾ SD years
Sex, no. male/female
Disease duration, mean ⫾ SD years
BASFI, mean ⫾ SD
BASDAI, mean ⫾ SD
Schober test, mean ⫾ SD cm
Chest expansion, mean ⫾ SD cm
Anti-TNF therapy, no. (%) of patients
Uveitis, no. (%) of patients
Hip prosthesis, no. (%) of patients
HLA–B27 status, no. (%) of 45 patients tested
Erythrocyte sedimentation rate, mm/hour
C-reactive protein, mg/liter
42.8 ⫾ 12.5
41/21
10.9 ⫾ 7.2
4.4 ⫾ 2.5
3.4 ⫾ 2.2
2 ⫾ 1.5
3.5 ⫾ 1.5
16 (26)
7 (11)
4 (6)
37 (82)
46.4 ⫾ 28.4
45.7 ⫾ 55
* BASFI ⫽ Bath Ankylosing Spondylitis Functional Index; BASDAI ⫽
Bath Ankylosing Spondylitis Disease Activity Index; anti-TNF ⫽
anti–tumor necrosis factor.
quencing was undertaken in both forward and reverse
directions using the sequencing primers MEFV_Ex10U (5⬘TCCCTGTGCTCTCCCCTACCA-3⬘), and MEFV_Ex10Lseq
(5⬘-GTCGGCATTCCGTGACTATT-3⬘), and the products
were run on an ABI 3730xl DNA Analyzer (Applied Biosystems).
Forward and reverse sequencing chromatograms were
aligned with the human MEFV exon 10 Ensembl reference
sequence using Sequencher 4.7 software (Gene Codes). A
polymorphism was recognized if it appeared in both the
forward and reverse sequences. Ambiguous chromatograms
and nucleotides called from sequences in one direction that
was not compatible with its complement in the other direction
were resolved by reamplifying and resequencing the sample. If
after 3 rounds of amplification and sequencing the genotype of
the sample was unresolved, the sample was excluded.
MEFV exon 2 E148Q genotyping assay and quality
control. Polymerase chain reaction (PCR) amplification of a
277-bp fragment containing the E148Q polymorphism was
performed by using the primers MEFV-EXON2-13F (5⬘ATCATTTTGCATCTGGTTGTCCTTCC-3⬘) and MEFVEXON2-17R (5⬘-GAGGCTTGCCCTGCGCG-3⬘). PCR
products were digested by restriction enzyme Bst NI. Digestion
product sizes were resolved by electrophoresis through a 3%
agarose gel. Digestion of the 277-bp PCR product yields
fragments of 160, 79, 21, and 17 bp if the G allele (glutamic
acid codon) is present and fragments of 97, 79, 63, 21, and 17
bp if the C allele (glutamine codon) is present.
To assess genotyping accuracy, 22 chromosomes (17 G
and 5 C) were confirmed by sequencing. No errors in genotyping were observed. Samples that failed the initial E148Q
restriction fragment length polymorphism assay more than 3
times had their genotypes determined by direct sequencing of
the PCR fragment in both the forward and reverse directions
(20 chromosomes), using the PCR primers on an ABI 3730xl
DNA Analyzer. No differences in forward and reverse sequences were observed. One sample could not be successfully
sequenced in either direction.
Statistical analysis. A chi-square test or Fisher’s exact
test was used, when appropriate, to asses the difference in the
prevalence of MEFV variants between AS patients and healthy
controls or RA patient controls. Odds ratios were calculated as
a measure of the strength of the association. Spearman’s rho
test was used to describe correlations. All P values were
2-tailed, and confidence intervals (CIs) were set at 95%. P
values less than 0.05 were considered significant.
RESULTS
Genotyping of E148Q failed for 1 AS patient,
and genotyping of exon 10 failed for 1 healthy control.
Consequently, the statistical analysis was performed on
the remaining 158 subjects with complete data for
E148Q and exon 10: 62 AS patients (41 men and 21
women with a mean ⫾ SD age of 42.8 ⫾ 12.5 years), 46
RA patients (13 men and 33 women with a mean ⫾ SD
age of 54.4 ⫾ 12 years), and 50 healthy controls (34 men
and 16 women with a mean ⫾ SD age of 32.0 ⫾ 10.0
years).
PREVALENCE OF M694V IN AS AND EVIDENCE OF A LINK WITH FMF
3061
Table 2. Distribution of MEFV variants in patients with AS as compared with control groups of healthy
subjects and patients with RA*
Controls
No. homozygous or compound heterozygous
E148Q/E148Q
M694V/M694V
M694V/M680I
M680I/V726A
M694V/E148Q
No. heterozygous
E148Q
M694V
V726A
M680I(G/A)
P760P
K695R
Total patients with MEFV gene variants
AS
patients
(n ⫽ 62)
Healthy
subjects
(n ⫽ 50)
RA
patients
(n ⫽ 46)
All
controls
(n ⫽ 96)
5
0
1
2
1
1
9
2
4
2
0
1
0
14
2
2
0
0
0
0
5
0
1
2
1
0
1
7
0
0
0
0
0
0
4
1
1
1
0
1
0
4
2
2
0
0
0
0
9
1
2
3
1
1
1
11
* AS ⫽ ankylosing spondylitis; RA ⫽ rheumatoid arthritis.
The distribution of alleles according to disease
status is shown in Table 2. There were 14 AS patients
with 19 allelic variants, 4 RA patients with 4 allelic
variants, and 7 healthy controls with 9 allelic variants.
The prevalence of total allelic variants was significantly
increased in the AS patient group in comparison with
the RA control patients (P ⫽ 0.013, OR 3.98 [95% CI
1.31–12.14]) but not in comparison with the healthy
controls (P ⫽ 0.222, OR 1.83 [95% CI 0.79–4.24])
(Table 3). The allele frequency of MEFV variants in
the healthy control subjects as compared with the RA
patient controls was very similar (P ⫽ 0.75); therefore,
the AS group was also compared with the combined
control group. AS patients had a significantly increased
variant allele frequency as compared with the combined
control groups (P ⫽ 0.021, OR 2.5 [95% CI 1.2–5.2]).
There was also a significant excess of AS patients
carrying 2 exon 10 variants as compared with the combined controls (P ⫽ 0.008) (Table 2).
Analysis of individual variant rates revealed a
significant increase in the frequency of M694V in the AS
patient group as compared with the combined control
groups (P ⫽ 0.008) or with the individual control groups
(P ⫽ 0.026 versus healthy controls and P ⫽ 0.046 versus
Table 3. Allele frequencies and carriage rates of MEFV variants in patients with AS as compared with control groups of healthy subjects and
patients with RA
Allele frequency, %
E148Q
M694V
Other exon 10 variants
Total exon 10 variants
Total MEFV variants
Carriage rate, %
E148Q
M694V
Any exon 10 variants
Total MEFV variants
AS
patients
(n ⫽ 62)
Healthy
subjects
(n ⫽ 50)
P*
RA patients
(n ⫽ 46)
P†
All controls
(n ⫽ 96)
P‡
OR
(95% CI)§
2.4
7.3
5.6
12.9
15.3
4
1
4
5
9
0.703
0.026
0.758
0.063
0.222
1.1
1.1
2.2
3.3
4.3
0.638
0.046
0.307
0.015
0.013
2.6
1.0
3.1
4.2
6.8
0.99
0.008
0.385
0.008
0.021
0.9 (0.2–3.9)
7.4 (1.6–35)
1.9 (0.6–5.6)
3.4 (1.4–8.2)
2.5 (1.2–5.2)
4.8
12.9
19.4
22.6
4.0
2.0
10.0
14.0
0.99
0.041
0.196
0.331
2.2
2.2
6.5
8.7
0.635
0.075
0.089
0.069
3.1
2.1
8.3
11.5
0.68
0.014
0.052
0.075
1.6 (0.3–8.1)
7.0 (1.4–34)
2.6 (1–6.9)
2.3 (0.9–5.4)
* For the comparison of ankylosing spondylitis (AS) patients versus healthy control subjects.
† For the comparison of AS patients versus rheumatoid arthritis (RA) control patients.
‡ For the comparison of AS patients versus the combined group of controls.
§ Odds ratios (ORs) and 95% confidence intervals (95% CIs) are relative to the 2 control groups combined.
3062
AKKOC ET AL
RA patient controls). The carriage rate of M694V was
also higher in AS patients than in the combined control
group (P ⫽ 0.014). The prevalence of other exon 10
variants and E148Q was not different between the
patients and controls (P ⬎ 0.05). However, we noted
that the power to detect individual variant effects was
low because of the low numbers observed. The observed
allele frequencies for individual variants in the combined
control group ranged from 0.5% to 2.7%. Power calculations based on these observed control allele frequencies in 62 patients and 96 controls indicated that allelic
odds ratios in the range of 4.1 to 11.2 are required to
achieve 80% power. The observed allelic odds ratio for
M694V was 7.4 (Table 3).
The HLA–B27 status was available for 45 of the
62 AS patients, of whom 37 were B27 positive (82%).
Sixteen AS patients had been treated with an anti-TNF
agent (Table 1). Clinical and laboratory parameters did
not correlate with the presence of MEFV variants (P ⬎
0.05).
DISCUSSION
MEFV is located on chromosome 16p (1,2). Recent studies have indicated that pyrin participates in a
complex interplay with the pyrin-domain protein superfamily and lipopolysaccharides via the Toll-like receptor
family and procaspase 1 activation. Pyrin is implicated in
the control of inflammation through leukocyte apoptosis
and interleukin-1␤ and NF-␬B activation. Therefore, it
may be speculated that MEFV might have a modifying
effect on the expression of certain inflammatory diseases. Consistent with this hypothesis, MEFV variants
have been reported to be associated with susceptibility
to a number of inflammatory rheumatic disorders, such
as Behçet’s disease (8) and palindromic rheumatism (9).
In the present study, we found an increased
prevalence of MEFV variants in AS patients in comparison with a pooled control group of healthy subjects and
RA patients. RA patients were chosen to serve as
controls, because the prevalence of MEFV variants in
RA has previously been shown to be similar to that in
healthy subjects (10), a finding that was confirmed in this
study.
The difference in the frequency of variants between AS patients and controls was a consequence of
M694V, which was the only MEFV variant that was
significantly more frequent in the AS patients than in the
individual control groups or in the control groups combined. The allele frequency and the carriage rate of
M694V in the AS group were 7.3% and 12.9%, respec-
tively, which were significantly higher than the corresponding values in the pooled control group (1.0% and
2.1%, respectively).
Other than M694V, our results did not reveal any
significant difference in the prevalence of exon 10
variants between the AS patients and the controls.
However, our sample size may not have been large
enough to establish the contribution of other exon 10
variants, all of which are less common than M694V. It
should be noted that the number of patients with any 2
exon 10 variants was significantly greater in the group
with AS than in the combined control group.
E148Q in exon 2 is the most common MEFV
variant in healthy subjects in populations in which FMF
is prevalent. However, some individuals who are homozygous for this variant are known to be disease-free,
which was true for 2 of the healthy subjects included in
this study. Thus, it is as yet unclear whether E148Q is a
disease-causing variant with low penetrance or an allelic
variant with no pathologic consequences (11). No significant difference in the frequency of E148Q between the
AS patients and the controls was observed in this study.
The presence of MEFV variants in AS patients
has been evaluated in some recent studies from other
regions of Turkey. One study from the middle region of
Turkey analyzed the 8 most common MEFV variants and
detected an increased M694V prevalence in a group
of 95 AS patients (12). Unfortunately, that study did
not include a control group, but the finding of a 23%
carriage rate of M694V was markedly higher than the
previously published 3% carriage rate of M694V in a
Turkish population (13). One major concern about that
study (12) is the very low prevalence of HLA–B27 in the
patient population (32%). Moreover, 8% of the patients
in that study had no evidence of radiologic sacroiliitis,
even though the modified New York criteria were used
as the case definition.
Another study from the Black Sea region of
Turkey revealed no significant difference in the frequency of MEFV variants between AS patients and
unrelated, population-matched healthy controls (14).
The statistical analysis, however, was based on all variants taken together, which led the investigators to miss
a significantly higher frequency of M694V in the AS
patients, with an OR of 3.9 (95% CI 1.03–14.75, P ⬍
0.045) (15).
A study from Istanbul (16) found a significantly
higher prevalence of M694V in 193 AS patients as
compared with 103 healthy controls (OR 4.73 [95% CI
1.39–16.12]). That study, which tested the same hypothesis in a different set of patients, provides independent
PREVALENCE OF M694V IN AS AND EVIDENCE OF A LINK WITH FMF
support for our study, the results of which, in isolation,
may not be considered conclusive because of its small
sample size and limited power.
The relationship between the presence of an
MEFV variant and clinical features was assessed in all 4
studies in Turkey (the present study and the above 3
studies), with somewhat contradictory results. However,
these studies have low statistical power because of their
small sample sizes and the low frequency of MEFV
variants. Further studies with larger populations will be
required to confirm these findings suggesting an association of M694V with AS and to clarify its influence on
the disease course.
Our findings are of particular interest in the
context of the existing literature (mostly in the form of
case reports) suggesting a link between AS and FMF. An
increased frequency of sacroiliitis, which is a hallmark of
AS, has been reported in different studies assessing
patients with FMF. In the most recent study (4), sacroiliitis was found in 18 of the FMF patients (32.7%), but
the most striking finding was a 94% prevalence of
M694V in patients with sacroiliitis as compared with a
45% prevalence in those without sacroiliitis.
The results of our present study, along with the
published literature, indicate that M694V may be an
additional susceptibility factor for AS. Its contribution to
the development of AS will obviously be limited to
populations that carry M694V. However, it should be
noted that in some populations in which FMF is uncommon, some novel MEFV variants, such as P646L and
L649P, have been identified among patients with seronegative polyarthropathies (9). In this particular study, 4
uncommon MEFV variants were detected with higher
allele frequencies in patients with palindromic rheumatism than in either Spanish FMF patients or healthy
subjects. It must also be borne in mind that MEFV may
not be a susceptibility factor for AS, but could be in
linkage disequilibrium with an unknown variant located
close to exon 10 on chromosome 16p. Interestingly, in a
study investigating genetic determinants of clinical manifestations of AS, Brown et al (17) found strong linkage
with chromosome 16p and disease activity, as determined by the Bath Ankylosing Spondylitis Disease Activity Index. Future studies with sufficiently large sample
sizes are needed to investigate the role of M694V and
other MEFV variants in susceptibility to AS and the
course of the disease.
AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it
critically for important intellectual content, and all authors approved
3063
the final version to be published. Dr. Akkoc had full access to all of the
data in the study and takes responsibility for the integrity of the data
and the accuracy of the data analysis.
Study conception and design. Akkoc, Sari, Onen, Bradman.
Acquisition of data. Akkoc, Sari, Thomas, Binicier, Savran, Onen,
Bradman, Plaster.
Analysis and interpretation of data. Akkoc, Sari, Akar, Weale, Birlik,
Onen, Bradman, Plaster.
REFERENCES
1. French FMF Consortium. A candidate gene for familial Mediterranean fever. Nat Genet 1997;17:25–31.
2. International FMF Consortium. Ancient missense mutations in a
new member of RoRet gene family are likely to cause familial
Mediterranean fever. Cell 1997;90:797–807.
3. Akar S, Birlik M, Sari I, Onen F, Akkoc N. M694V mutation may
have a role in susceptibility to ankylosing spondylitis. Rheumatol
Int 2009;29:1259–60.
4. Kasifoglu T, Calisir C, Cansu DU, Korkmaz C. The frequency of
sacroiliitis in familial Mediterranean fever and the role of HLAB27 and MEFV mutations in the development of sacroiliitis. Clin
Rheumatol 2009;28:41–6.
5. Langevitz P, Livneh A, Zemer D, Shemer J, Mordechai P.
Seronegative spondyloarthropathy in familial Mediterranean fever. Semin Arthritis Rheum 1997;27:67–72.
6. Van der Linden S, Valkenburg HA, Cats A. Evaluation of
diagnostic criteria for ankylosing spondylitis: a proposal for modification of the New York criteria. Arthritis Rheum 1984;27:361–8.
7. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF,
Cooper NS, et al. The American Rheumatism Association 1987
revised criteria for the classification of rheumatoid arthritis.
Arthritis Rheum 1988;31:315–24.
8. Touitou I, Magne X, Molinari N, Navarro A, Quellec AL, Picco P, et
al. MEFV mutations in Behçet’s disease. Hum Mutat 2000;16:271–2.
9. Canete JD, Arostegui JI, Queiro R, Gratacos J, Hernandez MV,
Larrosa M, et al. An unexpectedly high frequency of MEFV
mutations in patients with anti–citrullinated protein
antibody–negative palindromic rheumatism. Arthritis Rheum
2007;56:2784–8.
10. Rabinovich E, Livneh A, Langevitz P, Brezniak N, Shinar E, Pras
M, et al. Severe disease in patients with rheumatoid arthritis
carrying a mutation in the Mediterranean fever gene. Ann Rheum
Dis 2005;64:1009–14.
11. Ben-Chetrit E, Lerer I, Malamud E, Domingo C, Abeliovich D.
The E148Q mutation in the MEFV gene: is it a disease-causing
mutation or a sequence variant? Hum Mutat 2000;15:385–6.
12. Cinar M, Dinc A, Simsek I, Erdem H, Koc B, Pay S, et al. The rate
and significance of Mediterranean fever gene mutations in patients with ankylosing spondylitis: a three-month, longitudinal
clinical study. Rheumatol Int 2008;29:37–42.
13. Yilmaz E, Ozen S, Balci B, Duzova A, Topaloglu R, Besbas N, et
al. Mutation frequency of familial Mediterranean fever and evidence for a high carrier rate in the Turkish population. Eur J Hum
Genet 2001;9:553–5.
14. Durmus D, Alayli G, Cengiz K, Yigit S, Canturk F, Bagci H.
Clinical significance of MEFV mutations in ankylosing spondylitis.
Joint Bone Spine 2009;76:260–4.
15. Akkoc N, Gul A. Comment on the article by Durmus, et al.
Clinical significance of MEFV mutations in ankylosing spondylitis
[letter]. Joint Bone Spine 2010;77:281.
16. Cosan F, Ustek D, Oku B, Duymaz-Tozkir J, Cakiris A, Abaci N, et
al. Association of familial Mediterranean fever–related MEFV variations with ankylosing spondylitis. Arthritis Rheum. In press.
17. Brown MA, Brophy S, Bradbury L, Hamersma J, Timms A, Laval
S, et al. Identification of major loci controlling clinical manifestations of ankylosing spondylitis. Arthritis Rheum 2003;48:2234–9.