Full text of DOI:10.1038/sj.ejhg.5200725

European Journal of Human Genetics (2001) 9, 849 ± 854
ã 2001 Nature Publishing Group All rights reserved 1018-4813/01 $15.00
www.nature.com/ejhg
ARTICLE
Familial Mediterranean Fever: association of elevated
IgD plasma levels with specific MEFV mutations
Myrna Medlej-Hashim1, Isabelle Petit2, Salim Adib3, Eliane Chouery1, Nabiha Salem1,
1
4
5
6
Â
Â
Valerie Delague , Mohammed Rawashdeh , Issam Mansour , Gerard Lefranc ,
Roger Naman5, Jacques Loiselet1, Jean-Claude Lecron2, Jean-Louis Serre7 and
Andre Megarbane*
,1
Â
Â
Â
1
2
Â
 Â
Â
Â
Â
Â
Â
Â
Unite de Genetique Medicale, Faculte de Medecine, Universite Saint Joseph, Beirut, Lebanon; Faculte de Medecine
3
Â
et de Pharmacie. Laboratoire cytokines, IBMIG, Universite de Poitiers, France; Department of Community
Medicine, Kuwait University, Kuwait; Division of Gastroenterology, Department of Pediatrics, Faculty of Medicine,
Jordan University of Science and Technology, Irbid, Jordan; Service d'Immunologie, Hotel-Dieu de France, Beirut,
Lebanon; Laboratoire d'Immunogenetique Moleculaire, Institut de Genetique Humaine, CNRS UPR 1142 et
4
5
Ã
6
 Â
Â
 Â
 Â
7
Â
 Â
Â
Â
Universite Montpellier II, France; Laboratoire de Cytogenetique et de Genetique Moleculaire Humaine, Universite
de Versailles/Saint Quentin en Yvelines, France
Familial Mediterranean Fever (FMF) is a recessively inherited disorder, characterized by episodic fever,
abdominal and arthritic pain, as well as other forms of inflammation. Some FMF patients present higher IgD
serum levels, and it is not yet known whether such an elevation is related to specific genotypes or correlated
with a specific phenotype. In order to evaluate the association between known FMF-related mutations and
IgD levels in confirmed patients, as well as the correlation between those levels and the presence of specific
clinical signs, genotypic analysis and IgD plasma measurements were performed for 148 Lebanese and
Jordanian FMF patients. Most common mutational patterns were M694V heterozygotes (19%) and
homozygotes (17%), and V726A heterozygotes (18%) and homozygotes (5%), with an additional 11%
combining both mutations. Twenty-one patients had higher IgD levels (superior to 100 mg/ml). The risk for
higher IgD levels was significantly associated with M694V homozygote status (OR = 6.25) but not with
heterozygotic one (OR = 1). Similarly, the risk for higher IgD was also found with V726A homozygotes
(OR = 2.2) but not with heterozygotes (OR = 1.05). The use of colchicine was not statistically associated with
IgD levels. Clinically, hyper IgD was also found significantly associated with arthritis (OR = 18). Thus,
homozygotic status for M694V, and to a lesser extent V726A, is associated with increased risk for higher IgD
plasma levels, regardless of colchicine use. Elevated IgD plasma levels are also correlated with the severity of
FMF manifestations, and especially with arthritis. European Journal of Human Genetics (2001) 9, 849 ± 854.
Keywords: FMF; genotype; IgD; mutation; phenotype; arthritis
Introduction
abdominal pain and/or arthritic problems, and various other
more specific clinical and genetic characteristics.1 ± 3 Of those
disorders, Familial Mediterranean Fever (FMF), frequently
found in populations of the Mediterranean area,4 and the
Hyperimmunoglobulinemia D Syndrome (HIDS), diagnosed
mainly in Western European countries,1,5 are recessively
inherited disorders. MEFV, the gene responsible for the
Mediterranean disorder, was cloned on chromosome
16p13.3 in 1997.6,7 Mutations in the mevalonate kinase
Periodic fever syndromes are a family of diseases character-
ized by common symptoms, namely, episodic crisis of fever,
Â
Â
Â
 Â
*Correspondence: Andre Megarbane MD, PhD, Unite de Genetique
Â
Â
Â
Â
Medicale, Faculte de Medecine, Universite Saint-Joseph, 42, rue de
Grenelle, 75007 Paris, France. Fax: 00 961 1 614 054;
E-mail: megarban@ dm.net.lb
Received 23 May 2001; revised 21 August 2001; accepted 28 August 2001

MEFV mutations and IgD plasma levels
M Medlej-Hashim et al
850
gene, MVK, on chromosome 12q24, were found responsible
for HIDS.8,9 Changes in serum immunoglobulin D (IgD)
levels are also features of these diseases.2
tion refractory mutation screening (ARMS) with special PCR
conditions (100 ng genomic DNA, 0.2 mM dNTPs, 4 mM
MgCl2, 1 pmol of each primer, 16 PCR buffer and 0.04 U
AmpliTaq Gold in 25 ml mix).
Immunoglobulins D are a class of antibodies discovered in
myeloma patients in 196410 and, subsequently detected as
membrane bound Ig, present predominantly on the mature
lymphocyte B cells surface in association with IgM.11 The
function of secreted IgD is still unclear. The distribution of
IgD in normal adult sera is non-gaussian and extremely
heterogeneous (0.1 ± 213 mg/ml, mean: 36 mg/ml).12
While most HIDS patients have elevated IgD serum levels,
some FMF patients show elevated IgD sera levels, as reported
by a previous study.2 However, since that study was
conducted prior to the cloning of the MEFV gene, it is not
possible to ascertain whether various mutations leading to
FMF had varying effects on IgD levels. Similarly, elevated
plasma IgD was not correlated with any specific phenotypic
FMF sign.
Statistical analysis
Chi-squared test was performed to test the significance of
differences in the prevalence of higher IgD values in various
comparison groups. A P-value of 0.05 or less meant that the
difference was statistically significant. The association
between a putative risk factor and higher IgD levels was
measured using the odds-ratio (OR) and its corresponding
95% confidence interval (95% CI). While an OR41 suggests
the existence of an association, the inclusion of 1 in the 95%
CI indicates that the estimation of risk is not significant.
Computations were conducted using SPSS and EPI-6.
The present study aims at determining the association of
various FMF mutations and IgD plasma levels and correlating
those levels with specific phenotypic characteristics.
Results
Both alleles were identified for 99 patients, 42 homozygous
and 57 compound heterozygous, whereas 33 patients were
heterozygous for one mutation and 16 had none of the tested
mutations. Most commonly found mutations within 147
unrelated patients were M694V (32%), V726A (20%), M694I
(11.6%), M680I (6.8%) and E148Q (5%), while other rare and
non-identified mutations totaled 24.5%.
Materials and methods
Participants and materials
The sample included in this analysis was composed of 148
Lebanese and Jordanian FMF patients (all unrelated except
one), diagnosed according to Heller's criteria.13 Data
obtained included clinical symptoms and colchicine intake
at the time of the study. The clinical severity score was
evaluated for each participant according to the Tel Hashomer
classification.14 Thirteen patients had an incomplete clinical
investigation form, and one could not then evaluate their
severity scores.
Using these allele frequencies, expected genotypes were
calculated under Hardy ± Weinberg panmixia hypothesis
(Table 1) and showed a significant departure. Since 23
patients out of 147 unrelated participants were issued from
first cousins parents, allele frequencies and expected number
of genotypes were also calculated once homozygote inbred
patients had been removed. The comparison between
observed and expected genotypes showed that homozygotes
for M694V are in excess, and M694V carriers are less than
expected in FMF patients, even when inbred patients were
removed.
Blood samples were collected from patients with their full
consent. At the time of blood donation, none of the patients
was undergoing an FMF attack. For comparison, blood was
also drawn from 81 unrelated consenting individuals,
matched relatively to sex and whose ages vary between 18
and 50 years old, after ensuring they were free of any
apparent inflammatory symptoms or allergy. Plasma was
obtained by instant centrifugation and stored at 7808C, then
DNA was extracted from leucocytes by standard techni-
ques.15 IgD measurements in plasma were performed using a
previously described12 ELISA technique with a detection
threshold of 0.1 mg/ml.
IgD levels in the patients group varied between 0.07 and
337 mg/ml, with a median of 28 mg/ml, and only 14% (n = 21)
had IgD levels superior to 100 mg/ml. Remarkably, in the
control group, values varied between 0.2 and 218 mg/ml, with
a median of 25 mg/ml. The measurements in the latter group
included eight values superior to 100 mg/ml (10%). As the
distributions of IgD levels in both the patients and control
series are non-gaussian, these values were transformed to
logarithms and the histograms obtained (Figure 1) showed a
distribution quite similar to the multimodal distribution
previously observed in sera of healthy subjects.18 Some
prevalence of higher IgD values in apparently healthy
subjects is not unusual.10,12
FMF mutations screening
Five most frequent FMF mutations (M694V, V726A, M680I,
M694I, E148Q), as well as eight rarer ones (E167D, T267I,
R761H, P369S, A744S, F479L, I692del, M694del) were
investigated as described elsewhere.16 E230K, a newly
discovered mutation,17 was tested using PCR same proce-
dures as for other mutations of exon 2 and the restriction
enzyme Taq I. Mutation K695R was also tested by amplifica-
All patients with higher IgD levels had two identified
mutations. Detailed clinical, genotypic and serological
features of those 21 FMF patients are presented in Table 2.
The association between specific genotypes and higher IgD
levels was analysed and presented in Table 3. Compared to
European Journal of Human Genetics

MEFV mutations and IgD plasma levels
M Medlej-Hashim et al
851
Table 1 Observed and expected number (frequencies) of genotypes among unrelated FMF patients with either removing or
not homozygote inbred patients
M694V/
M694V
M694V/
V726A
M694V/
others
V726A/
V726A
V726A/
others
Other
genotypes
Chi-square
(df=3)*
n (%)
Total
Observed genotypes
25 (17)
16 (11)
28 (19)
8 (5.5)
27 (18.5)
43 (29)
147 (100)
Expected under Hardy ± Weinberg 15.05 (10) 18.91 (13) 45.06 (31) 5.94 (4) 28.31 (19) 33.73 (23) 147 (100) 16.80 (p*0.001)
Observed genotypes once
inbred removed
20 (15)
16 (12)
28 (21)
7 (5)
27 (20.5) 35 (26.5) 133 (100)
Expected genotypes once
homozygotes inbred removed
13.28 (10) 17.99 (13.5) 39.50 (30) 6.09 (4.5) 26.75 (20) 29.39 (22) 133 (100) 8.17 (p*0.004)
*df: degrees of freedom.
Figure 1 Plasma IgD values distribution among 148 FMF patients and 81 healthy individuals.
those not carrying any M694V mutation, the homozygotic
pattern of this mutation was associated with highest risk for
higher IgD (OR = 6.25), followed by the combination M694V/
V726A (OR = 2.3), while other heterozygotic patterns were
not associated with higher IgD levels. Compared to those not
carrying any V726A mutation, the homozygotic pattern of
V726A was associated with some risk for higher IgD
(OR = 2.2), while heterozygotic patterns were not. Patients
with a combined M694V/V726A genotype were about five
times more likely to present higher IgD levels than all other
genotypes (OR = 4.7).
with higher levels were more likely to suffer from articular
crisis (OR = 18, CI(2.33 ± 139.31)), amyloidosis (OR = 3.44,
CI(0.78 ± 15.13)), thoracic crisis (OR = 1.5, CI(0.55 ± 3.93)),
and erysipelas-like erythema (OR = 1.25, CI(0.25 ± 6.19))
(Data not shown).
About 58% of the sample were using colchicine at the time
of the study. The proportion of colchicine users having
elevated IgD levels (67%) was more important than that of
colchicine users in the group with lower IgD levels (57%).
However, no significant association was found between
colchicine and lower IgD levels.
Average severity scores of the high and low IgD plasma
levels group show significant differences (8.1+5.1 for the
former versus 5.5+4.9 for the latter). Moreover, differences
in specific clinical features were detected between the two
groups. Compared to those with lower IgD levels, patients
Discussion
The aim of this study was mainly to clarify the association of
specific genotypes with IgD levels in FMF patients, and to
European Journal of Human Genetics

MEFV mutations and IgD plasma levels
M Medlej-Hashim et al
852
Table 2 Clinical and serological characteristics of 21 FMF patients with higher IgD plasma
Abdominal
attacks
Thoracic
attacks
Articular
attacks
Erysipela-like
erythema
Severity
score
IgD plasma
level (mg/ml)
Patients*
Fever
Amyloidosis
Genotype
1
2
3
4
5
6
7
8
+
+
+
+
+
+
+
+
+
7
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
7
7
+
+
+
+
+
+
+
+
+
+
+
+
?
7
7
+
7
7
7
+
7
7
7
7
?
7
?
7
7
+
7
7
7
7
7
7
7
7
7
7
+
7
7
12
8
11
6
12
8
10
7
10
Ð
4
Ð
6
4
M694V/M694V
M694V/M694V
M694V/M694V
M694V/M694V
M694V/M694V
M694V/M694V
M694V/M694V
M694V/M694V
M694V/M694V
M694V/M694V
M694V/V726A
M694V/V726A
M694V/V726A
M694V/E148Q
V726A/V726A
V726A/V726A
V7826A/M680I
V726A/M680I
V726A/T267I
182
236
169
201
162
250
149
139
105
106
198
301
123
247
284
281
175
224
337
108
111
+
7
7
+
7
+
?
7
?
+
7
+
7
+
+
+
9
7
7
7
?
7
?
7
7
7
7
7
7
7
10
11
12
13
14
15
16
17
18
19
20
21
+
?
+
+
+
+
+
+
7
12
8
8
9
5
M694I/M694I
M694I/E148Q
+
*All these patients, except 2 (numbers 2 and 4), are non consanguineous.
Patients' numbers 3, 4, 5, 17, 18 and 19 are Jordanian and the others are Lebanese.
Table 3 Risk for higher IgD serum levels associated with selected types of mutations in FMF patients
n (%)
IgD5100 mg/ml
IgD4100 mg/ml
Total
OR (95% CI)*
Total
127 (85.8)
21 (14.2)
148
Ð
M694V carriers
M694V/M694V
M694V/other
16 (61.5)
27 (96.4)
13 (81.3)
40 (90.9)
56 (80)
10 (38.5)
1 (3.6)
3 (18.7)
4 (9.1)
14 (20)
7 (9)
26
28
16
44
70
78
6.25 (1.8 ± 25)
0.4 (0.02 ± 3.3)
2.3 (0.4 ± 12.2)
1 (0.24 ± 4.31)
2.5 (0.96 ± 6.71)
reference
M694V/V726A
All M694V heterozygotes
All M694V carriers
All others**
71 (91)
V726A carriers
V726A/V726A
V726A/other
M694V/V726A
All V726A heterozygotes
All others**
6 (75)
2 (25)
3 (11.1)
3 (18.7)
6 (14)
8
2.2 (0.3 ± 14)
0.8 (0.2 ± 3.4)
1.5 (0.3 ± 6.8)
1.05 (0.4 ± 3)
reference
24 (88.9)
13 (81.3)
37 (86)
27
16
43
97
84 (86.6)
13 (13.4)
Mixed carriers
M694V/V726A
Other carriers (excluding M694V or V726A)
13 (81.3)
41 (95.3)
3 (18.7)
2 (4.7)
16
43
4.7 (0.6 ± 46.7)
reference
*Limits of confidence intervals are not always precise because of the small numbers involved.
**All those not carrying this specific mutation.
determine whether a specific phenotype characterises
patients with an elevated plasma IgD level. Higher IgD values
were found in 14% of the FMF patients involved in this study.
In contrast, the distribution of IgD levels in patients and
controls in this study were largely similar (Figure 1), which is
not in favour of a direct effect of MEFV mutations on serum
IgD levels. However, the analysis of the association between
specific mutational genotypes and IgD levels suggests that
homozygotic genotypes for M694V and to a lesser extent
V726A increase the risk of higher IgD levels when compared
to other genotypes (Table 3). But the association with the risk
for higher IgD decreased considerably when all M694V
carriers were considered against non-M694V carriers
(OR = 2.5). Such a result is especially interesting for it is in
accordance with the significantly high proportion of M694V/
M694V on a genetic populational point of view.16,19 Indeed,
although Lebanese and Jordanian people are known to
display a high level of consanguinity (mean value of 1% in
our sample),20 ± 22 the difference between observed and
expected number of homozygotes is far too high to be
explained only by consanguinity. On the other hand, the
genetic differences between Lebanese and Jordanian people,
European Journal of Human Genetics

MEFV mutations and IgD plasma levels
M Medlej-Hashim et al
853
although present,16,19 are too slight to induce a strong
Walhund effect. Thus, due to the correlation between
M694V/M694V and clinical status of FMF, especially through
the high level of IgD, one can postulate that the excess of
these homozygotes is mainly due to an assessment bias and
partially due to consanguinity and Walhund effect. In fact,
the M694V mutation was previously correlated with severe
phenotypes,19,23 and the recruitment of more patients whose
genotypes carry M694V, especially homozygous ones, may be
due to the severity of their crises.
Acknowledgments
We thank Mrs Adriana Delwail and Dr Franck Morel for their help, as
well as Professors Josue Feingold for his review and Patrice Bouvagnet
for his suggestions. This work was supported by Scientific Research
grants from the Saint Joseph University and the University of Poitiers,
and the Lebanese-NCSR.
The estimated association of V726A with higher IgD values
did not achieve statistical significance, most likely because of
the small numbers involved. However, when M694V
homozygotes were excluded from the reference group, the
homozygotic pattern of V726A was found associated with a
high risk for higher IgD (OR = 7.5, CI(1.05 ± 54.41)), which
suggests that M694V homozygotes were overwhelming the
V726A homozygotic effect. As for M694I, small numbers also
precluded any meaningful statistical association between this
mutation and IgD levels.
References
1 Centola M, Aksentijevich I, Kastner DL: The hereditary periodic
syndromes: molecular analysis of a new family of inflammatory
diseases. Hum Mol Genet 1998; 7: 1581 ± 1588.
2 Livneh A, Drenth JPH, Klasen IS et al: Familial Mediterranean
Fever and Hyperimmunoglobulinemia D Syndrome: two dis-
eases with distinct clinical, serologic, and genetic features. J
Rheumatol 1997; 24: 1558 ± 1563.
3 McDermott MF, Aksentijevich I, Galon
mutations in the extracellular domains of the 55 kDa TNF
receptor, TNFR1, define family of dominantly inherited
autoinflammatory syndromes. Cell 1999; 7: 133 ± 144.
4 Sohar E, Gafni J, Pras M, Heller H: Familial Mediterranean Fever.
A survey of 470 cases and review of the literature. Am J Med
1967; 43: 227 ± 253.
J et al: Germline
a
Colchicine is, since 1972, the only therapeutic agent used
to decrease FMF attacks frequencies and prevent amyloidosis,
the most severe complication of this disease.24 According to
the results, colchicine use may be discarded as a potential
confounding variable.
5 Drenth JPH, Haagsma CJ, van der Meer JWM: The International
Hyper-IgD Study Group. Hyperimmunoglobulinemia D and
Periodic Fever Syndrome: The clinical spectrum in a series of 50
patients. Medicine 1994; 73: 133 ± 144.
The clinical significance of higher IgD levels is not clear. In
the present study, higher IgD levels were associated with high
severity scores and particular clinical signs or symptoms,
namely articular pain, amyloidosis, and to a lesser extent
thoracic crisis, and erysipelas-like erythema. Indeed, all
except one patient (95%) with higher IgD levels reported
articular pain crises, whereas this type of pain was present in
only 50% of patients with lower IgD plasma level. In a
previous study,2 78% of FMF patients having an IgD value
superior to 100 IU/ml (equivalent of 141 mg/ml)25 had
attacks of monoarthritis. As a matter of fact, arthralgia is a
common feature of HIDS crisis, as well as arthritis involving
large joints.2,5 Thus, the severity of attacks was noticed in
FMF patients having high IgD plasma levels, especially
regarding joints pain, whereas the absence of relationship
between serum IgD level and the frequency and severity of
attacks had been reported in HIDS patients.5
6 French FMF Consortium:
A candidate gene for familial
Mediterranean Fever. Nat Genet 1997; 17: 25 ± 31.
7 International FMF Consortium: Ancient missense mutations in
a new member of the RoRet family are likely to cause familial
Mediterranean fever. Cell 1997; 90: 797 ± 807.
8 Drenth JPH, Cuisset L, Grateau G et al: Mutations in the gene
encoding mevalonate kinase cause hyper-IgD and periodic fever
syndrome. Nat Genet 1999; 22: 178 ± 181.
9 Houten SM, Kuis W, Duran M et al: Mutations in MVK, encoding
mevalonate kinase, cause hyperimmunoglobulinaemia D and
periodic fever syndrome. Nat Genet 1999; 22: 175 ± 177.
10 Rowe DS, Fahey JL: A new class of human immunoglobulins. J
Exp Med 1965; 121: 171 ± 199.
11 Preud'homme JL, Petit I, Barra A et al.: Review Structural and
functional properties of membrane and secreted IgD. Mol
Immunol 2001; 37: 871 ± 887.
12 Levan-Petit I, Cardonna J, Garcia M et al: Sensitive ELISA for
human immunoglobulin D measurement in neonate, infant,
and adult sera. Clin Chem 2000; 46: 876 ± 878.
13 Heller H, Sohar E, Sherf L: Familial Mediterranean Fever. Arch
Intern Med 1958; 102: 50 ± 71.
14 Sohar E, Gafni J, Pras M, Freund (eds): Proceedings 1st
International Conference on FMF, Jerusalem. 1997; pp. 207 ±
208.
15 Miller SA, Dynes DD, Polesky F: A simple salting out procedure
for extracting DNA from human nucleated cells. Nucleic Acids
Res 1988; 16: 1215.
16 Medlej-Hashim M, Rawashdeh M, Chouery E et al: Genetic
screening of fourteen mutations in Jordanian Familial Medi-
terranean Fever patients. Hum Mutat (Online) 2000; 15: 384.
17 Torosyan T, Aksentijevich I, Sarkisian T et al: Role of complex
alleles and gender in susceptibility to Familial Mediterranean
fever in the Armenian population. Am J Hum Genet 2000;
67(supp.2): 404.
18 Dunette SL, Gleish GJ, Drew Miller R, Kyle RA: Measurement of
IgD by a double antibody radioimmunoassay: demonstration of
an apparent trimodal distribution of IgD levels in normal
human sera. Journal Immunol 1977; 119: 1727 ± 1731.
In conclusion, the most solid finding here is the association
between M694V homozygotic genotypes and higher IgD
levels. The fact that this association seems confined only to
one specific genotype may explain the low though apparently
persistent incidence of higher IgD in FMF patients. The
mechanism that governs the effect of M694V mutation on
the serum IgD level is currently unknown and may probably
be indirect. On the other hand, the correlation established
between high IgD levels and the severity of FMF manifesta-
tions, especially arthritis may reveal a novel aspect of IgD
effects. However, the functions of IgD remain to be
investigated as well as the mechanism of the IgD-pheno-
type-genotype association revealed in this analysis.
European Journal of Human Genetics

MEFV mutations and IgD plasma levels
M Medlej-Hashim et al
854
19 Mansour I, Delague V, Cazeneuve C et al: Familial Mediterra-
nean fever in Lebanon: mutation spectrum, evidence for cases
in Maronites, Greek catholics, Syriacs and Chiites and for an
association between amyloidosis and M694V and M694I
mutations. Eur J Hum Genet 2001; 9: 51 ± 55.
23 Dewalle M, Domingo C, Rozenbaum M et al: Phenotype-
genotype correlation in Jewish patients suffering from familial
Mediterranean fever (FMF). Eur J Hum Genet 1998; 6: 95 ± 97.
24 Goldfinger SE: Colchicine for familial Mediterranean fever. N
Engl J Med 1972; 287: 1302.
20 Khlat M: Consanguineous marriage and reproduction in Beirut,
Lebanon. Am J Hum Genet 1988; 43: 188 ± 196.
21 Al-Salem M, Rawashdeh N: Consanguinity in North Jordan:
prevalence and pattern. J Biosoc Sci 1993; 25: 553 ± 556.
22 Khoury SA, Massad D: Consanguineous marriage in Jordan. Am J
Med Genet 1998; 43: 769 ± 775.
25 Rowe DS, Anderson SG, Tackett L: A research standard for
human serum immunoglobulin D. Bull WHO 1970; 43: 607 ±
609.
European Journal of Human Genetics