Familial Mediterranean Fever

This book, written by very well known opinion leaders in the field, covers all aspects of familial Mediterranean fever, the most common monogenic autoinflammatory disease. The opening chapters explain the genetic basis of the disease and provide insights into the pathogenesis derived from recent experimental studies. A large part of the book is then devoted to a detailed description of the typical and atypical clinical presentations, the disease course, and potential complications in both pediatric and adult patients. Guidance is provided on the measurement of disease severity and the management of patients in daily practice. The advice regarding treatment is based on the best currently available evidence and attention is also paid to important emerging treatments.

The book is part of Springer’s series Rare Diseases of the Immune System, which presents recently acquired knowledge on pathogenesis, diagnosis, and therapy with the aim of promoting a more holistic approach to these conditions. Monogenic autoinflammatory diseases are hereditary disorders that are caused by single-gene defects in innate immune regulatory pathways and are characterized by a clinical and biological inflammatory syndrome in which there is limited, if any, evidence of autoimmunity. Familial Mediterranean fever itself is due to a mutation in the MEFV gene, which codes for the protein pyrin; it is characterized by periodic fever and episodes of painful inflammation in the abdomen, chest, and joints. Familial Mediterranean Fever will be an invaluable source of up-to-date information for all practitioners involved in the care of patients with the disease.

• Language: English
• Publisher: Springer
• Release date: Mar 19, 2015
• ISBN: 9783319146157

Book preview

© Springer International Publishing Switzerland 2015

Marco Gattorno (ed.)Familial Mediterranean FeverRare Diseases of the Immune System310.1007/978-3-319-14615-7_1

1. Genetics

Guillaume Sarrabay¹ and Isabelle Touitou², ³, ⁴  

(1)

Laboratory of genetics of rare and autoinflammatory diseases, CHRU of Montpellier, University of Montpellier, INSERM, 1183 Montpellier, France

(2)

Laboratory of genetics of rare and autoinflammatory diseases, CHRU of Montpellier, Montpellier, France

(3)

University of Montpellier, Montpellier, France

(4)

INSERM, 1183 Montpellier, France

Isabelle Touitou

Email: [e-mail]

1.1 Introduction

Familial Mediterranean fever (FMF) was first known as periodic disease or recurrent polyserositis. It was described by Janeway and Mosenthal in 1908 [1]. Heller gave this disease its definitive appellation in 1955 [2]. All these names refer to the key symptoms of this disease: recurrent fever with polyserositis (although this is not specific to FMF). The Mediterranean adjective is more relevant, since the prevalence is higher in the countries around this sea [3]. For example, in the non-Ashkenazi Jews, the prevalence ranges from 1/250 to 1/500 [4] and in Turks, from 1/1,073 to 1/395 [5].

FMF is a rare monogenic disease (p = 1–5 in 10,000), i.e., the genetic impact on the disease occurrence is more important than the environment. Since the discovery of the causative gene MEFV in 1997 [6, 7], it has been possible to perform genetic diagnosis. This chapter will describe our most recent knowledge on the genetic lesions responsible for this disease and the technical resources involved to identify and interpret them.

1.2 Mode of Transmission

1.2.1 Classical Autosomal Recessive Transmission

FMF is the prototype of autoinflammatory diseases (AID), and it is the first that has been identified in this group. It classically involves an autosomal recessive transmission [8]. As expected in recessive diseases, it is not rare to find sporadic cases due to the small size of siblings, and this should not eliminate the FMF diagnosis.

The theoretical risk for a carrier couple to have an affected child is one in four and has to be adjusted according to the penetrance of mutations. The penetrance is defined as the proportion of individuals with one mutation (dominant disease) or two (recessive disease) who exhibit clinical symptoms.

1.2.2 Pseudodominance and True Dominance

It is not rare to see pseudodominance , i.e., a vertical transmission (Fig. 1.1). It is important to look for consanguinity or endogamy (preferential union within a population at risk), these two causes significantly increasing the likelihood of coexistence of multiple mutated alleles in such families.

A323505_1_En_1_Fig1_HTML.jpg

Fig. 1.1

Pseudodominant transmission in an Armenian family with familial Mediterranean fever. Both high carrier frequency in this population (one in three to one in five) and consanguinity account for the remarkable vertical distribution of the affected individuals in three successive generations (II, III, and IV). By deduction, the II2 and II3 sisters inherited the same p.M680I mutation, which was transmitted to both III4 and III5. This mutation is finally found in double dose in individual IV1. Black symbols depict patients, and open symbols show asymptomatic individuals. The three obligatory carriers of the p.M680I mutation are in green. The double line indicates consanguinity

A true dominant transmission was however suggested in several studies [9–12]. Aldea et al. [11] described a three-generation Spanish kindred with a severe dominantly inherited periodic inflammatory disorder complicated with renal AA amyloidosis. This phenotype was associated with a p.H478Y mutation in the MEFV gene. In a study from Stoffels et al., whole-exome sequencing revealed a novel missense sequence variant, c.1730C>A; p.T577N, perfectly co-segregating with the disease in this family. Another mutation at the same amino acid (c.1730C>G; p.T577S) was found in a family of Turkish descent. These two mutations segregated with autosomal dominant inheritance, which suggested a fundamental role of the 577 threonine at this position.

1.3 The MEFV Gene

1.3.1 Discovery

MEFV (MEditerranean FeVer) was discovered in 1997 by two distinct consortia, through a positional cloning approach [6, 7]. The 781-amino-acid protein encoded was given two names: pyrin (International FMF consortium) and marenostrin (French FMF consortium).

1.3.2 Structure of the Gene

1.3.2.1 Location and Structure

The MEFV gene (NM_000243.2) is located on the short arm of chromosome 16 (16p13.3). It contains ten exons and is 14 kb long (Fig. 1.2). The coding sequence is around 3,500 bp.

A323505_1_En_1_Fig2_HTML.jpg

Fig. 1.2

Schematic structure of the MEFV gene. Squares illustrate the ten exons; the thin line shows the introns. Exons and introns are roughly at scale

MEFV is a medium-size gene. Exon 2 is GC-rich and thus can be difficult to amplify or to sequence.

Pyrin is expressed mainly in neutrophils and macrophages and modulates the production of the potent pro-inflammatory cytokine interleukin-1β through regulation of nuclear factor-κB and caspase-1 [13].

1.3.2.2 Mutation Type and Distribution

Mutations have been found in all exons of the MEFV gene (for a fairly exhaustive list, please refer to InFevers, a database dedicated to AID (http://​fmf.​igh.​cnrs.​fr/​infevers/​) [14]). However, most mutations are found in specific exons (2, 3, 5, and 10). Three mutational hot spots are remarkable. The first one is in codon 694 in exon 10, in which five different variants have been reported: p.M694V the most prevalent, p.M694I, p.M694K, p.M694L, and p.M694del. The second hot spot is also in exon 10, in codon 680: p.M680I (c.2040G>C), p.M680I (c.2040G>A), and p.M680L. The last one is in exon 2, in codon 148: p.E148Q and p.E148V.

Most mutations are missense type, small deletions are quite rare, and only one nonsense mutation has been described (p.Y688*) [15]. To date, large rearrangements have not been identified. One hypothesis is that such defect would be lethal.

1.4 Population Genetics

1.4.1 Mediterranean Founder Effect

Haplotypes analysis has shown that most FMF chromosomes originate from common ancestors, dating back to prebiblical times [16]. This founder effect is responsible for the very high prevalence of the most common mutations in four main Mediterranean populations: Arabs , Armenians , Jews , and Turks [17, 18].

The five most frequent mutations are p.E148Q in exon 2 and p.M680I, p.M694V, p.M694I, and p.V726A in exon 10. The spectrum of MEFV mutations in FMF patients differs among countries and populations. p.M694V is overrepresented in North African Jews (>70 %), while in East European Jews (Ashkenazim), a milder mutation, p.V726A, is the most frequent (38 %) [18]. The prevalence of these two mutations in Oriental Jews is in between that of North African Jews and Ashkenazim. These data suggest that p.M694V and p.V726A likely spread from the Middle East more than 2,500 years ago.

Some mutations are almost pathognomonic of certain populations such as p.M694I and p.A744S mutations in the Maghreb and p.F479L mutation in Armenians and Greeks.

1.4.2 Other Ethnicities

FMF can also be found worldwide, mostly in other Mediterranean groups, such as Italians, Spanish, and Greeks, but patients from England, India, China, Afghanistan, Hungary, and Japan have also been described [12, 19, 20]. In Japan, the p.M680I and p.E148Q variants are the most prevalent.

1.5 Genetic Diagnosis

1.5.1 Sanger Sequencing

Most laboratories providing genetic diagnosis of FMF use Sanger sequencing (Fig. 1.3). Developed by Frederick Sanger and colleagues in 1977, this approach is still the most widely used sequencing method as the gold standard in molecular diagnosis. Sanger sequencing is based on the selective incorporation of chain-terminating dideoxynucleotides by DNA polymerase during in vitro DNA replication [21].

A323505_1_En_1_Fig3_HTML.jpg

Fig. 1.3

An example of two close MEFV mutations (c.2040G>C;p.M680I and c.2080A>G;p.M694V) visualized on an electropherogram (SeqScape®, Life Technologies)

1.5.1.1 Benefits

Sanger sequencing is well suited for small-scale projects (MEFV-targeted genetic diagnosis of recurrent fever in Mediterranean patients) and to generate long contiguous DNA sequence reads (>500 nucleotides). Its main advantage is its relative low cost and reliability. It is very efficient for missense mutation detection.

1.5.1.2 Limitations

The main pitfall of the Sanger method resides in the risk of monoallelic amplification giving rise to an apparent homozygosity . This can happen in the event of a heterozygous deletion or if one primer hybridization site contains a single nucleotide polymorphism (SNP).

Another issue is the fact that Sanger sequencing is not a quantitative method; thus, low-level mosaicism can be missed.

1.5.1.3 Consensus Strategy

Since MEFV mosaicism and large rearrangements have not been reported, Sanger method remains a very good approach, though exon 2 is still difficult to amplify. The diagnosis route taken by most laboratories is a two-step strategy.

The entire gene sequence is not screened routinely because most of the variants are located in specific exons. Therefore, laboratories usually first analyze at least exon 10 , then possibly the other exons.

1.5.2 Other Techniques

Other molecular techniques have been used for the genetic diagnosis of FMF: restriction fragment length polymorphism (RFLP) and denaturizing gradient gel electrophoresis (DGGE), although they also tend to be supplanted by sequencing. RFLP can still be useful in determining the phasing when two mutations occur on the same codon.

Commercial kits targeted on the search for the 5–12 most frequent mutations are still used by some laboratories. They are based on immunochemistry and are cheaper and quicker than exon sequencing.

The use of next-generation sequencing (NGS ) for FMF is not yet in current practice, but will indubitably be developed shortly. This high-throughput multiparallel approach allows rapid generation of large amounts of sequence data. Panels of targeted genes and whole exomes will facilitate differential diagnosis with other autoinflammatory diseases [22] and precipitate the discovery of new FMF-like and modifier genes, respectively.

1.6 MEFV Variants

1.6.1 The Spectrum of MEFV Variants

1.6.1.1 Clearly Pathogenic Variants

The vast majority of MEFV mutations are substitutions.

A typical example is p.M694V in exon 10, substituting methionine at position 694 into valine. This mutation is the most frequent in all Mediterranean mutations and shows quasi-full penetrance.

Another well-known variant is p.V726A, which is common in Ashkenazi Jews. Though it is considered a clearly pathogenic mutation, carriers exhibit a milder clinical picture.

To date, two small in-frame deletions and three mutations leading to the creation of a stop codon have been reported (p.I692del, p.M694del [12, 19], and p.Y688* [15]).

1.6.1.2 Variants of Unknown Significance

Those are MEFV variants found in or associated with FMF, but not necessarily causative. They can have a high prevalence in the general population. For example, p.E148Q in exon 2 is a well-known one. This mutation is alternately considered a true mutation or a polymorphism (over 20 publications to date) [23]. Variants with no reliable information or new variants are also in this category.

1.6.1.3 Benign Variants

Simple polymorphisms, changing (p.R202Q) or not (p.P706P) the encoded amino acid, have been described in the MEFV gene. These variants are insufficient to trigger the disease but may act as phenotype modifiers.

1.6.2 Databases

1.6.2.1 Generalist

Generalist databases such as NCBI, Ensembl, EVS, and 1000 Genomes Project can be used for FMF genetic diagnosis. They help in evaluating the mutation frequencies.

1.6.2.2 FMF Specific

Loci-dependent databases have been created (ClinVar, LOVD) and record the mutations described by genetic laboratories in the world. As for FMF, the most exhaustive and comprehensive is InFevers [14, 24, 25] (Internet Fevers; http://​fmf.​igh.​cnrs.​fr/​ISSAID/​infevers) (Fig. 1.4). This website dedicated to mutations responsible for hereditary autoinflammatory diseases was created in 2002. Twenty-three genes are referenced in InFevers, and this website is a very useful tool for genetic laboratories involved in those disorders as it includes specific functions such as graphical mutation maps that display all submitted sequence variants.

A323505_1_En_1_Fig4_HTML.jpg

Fig. 1.4

The MEFV gene with all the mutations as reported in InFevers: an online database for autoinflammatory mutations (Copyright. Available at http://​fmf.​igh.​cnrs.​fr/​ISSAID/​infevers/​ Accessed (2014-07-17))

1.6.3 Clinical Interpretation

1.6.3.1 Guidelines

A committee gathering recognized geneticist and clinician experts has worked on guidelines for the interpretation of HRF gene variants. A draft was prepared based on current practice and was disseminated through the European Molecular Genetics Quality Network. Then a workshop was held in Bruges on 2011 to obtain a final consensus. An agreed set of practice guidelines was proposed for genetic diagnosis, reporting of results, and their clinical interpretation [26].

This committee has established a classification of gene variants based on the expertise of HRF diagnostic laboratories and on the review of current publications:

Clearly pathogenic variant (e.g., p.M694V)

Variants of uncertain significance

Controversial mutations (e.g., p.E148Q)

Unknown variants (e.g., p.L384P)

Variants that are clearly not the genetic cause (e.g., R202Q)

These guidelines highlight the importance of parental allele study for mutation phasing as complex alleles, usually including p.E148Q, have been recurrently identified.

1.6.3.2 Simple Heterozygotes

Interpretation of cases with a single demonstrated mutation is problematic. It happens frequently in patients meeting the clinical criteria for FMF. A second mutation elsewhere in the MEFV gene or in another gene should be searched for. In those heterozygous patients, the single mutation may simply act as a susceptibility factor. The relative risk for heterozygotes to develop FMF, compared to noncarriers, is estimated between 6.3 and 8.1 [27].

The consensus is then to rely on the clinics and if necessary to treat patients to prevent the possible lethal consequences of the FMF, e.g., renal amyloidosis . Furthermore, a colchicine trial if effective may further support FMF diagnosis [28].

1.7 Genotype-Phenotype Correlations

Different clinical presentations for patients with the same mutations, often within the same family or ethnic group, have been observed.

1.7.1 Phenotypic Heterogeneity

Patients with identical genotype can present with either mild or marked clinical pictures (frequency, duration, intensity of the attacks, occurrence of renal complications, etc.). Expression of the MEFV gene is probably impacted by environmental factors [29]. It is well known that the Armenians living in Armenia have more crises than those living in the USA [30] and that the risk for renal amyloidosis is higher [31].

Other genes encoding proteins involved in innate immunity may also act as modifier on the FMF phenotype. The list of patients with mutations in more than one AID gene is expanding rapidly, raising the hypothesis of possible oligogenism [32].

1.7.2 p.M694V and Disease Severity

p.M694V is the most frequent pathological MEFV variant in all Mediterranean populations: Jews [29, 33], Arabs [34], and Armenians [35]. It is also the most severe although this remains controversial among Turks [36]. Patients homozygous for this mutation generally exhibit more frequent crises and evolve more often to renal amyloidosis.

1.7.3 E148Q and Variable Penetrance

This variant has been the subject of numerous publications, but no consensus has yet emerged to classify it as clearly pathogenic or nonpathogenic [23]. This sequence variation was described as a disease-causing mutation with low penetrance and mild symptoms [37, 38]. It could promote a subclinical inflammatory state [38]. Furthermore, a quantum chemistry-based model suggested that its effect is low but not zero [37].

On the other hand, 50 % of E148Q homozygotes are asymptomatic [18], and there is a high prevalence of this mutation in certain populations (Japan, China) contrasting with a low FMF prevalence.

1.7.4 Modifier Genes

By definition, modifier genes can modulate the phenotype but are not necessary for the occurrence of the disease. The gene encoding the protein MICA (MHC class I chain A) is one of them. Patients having the MICA A9 allele along with the p.M694V MEFV mutation have increased risk of early onset of the disease. In contrast, patients with allele A4 have fewer attacks [39]. It seems that this gene has no impact on the occurrence of amyloidosis [34]. A second gene, SAA , encoding serum amyloid A, a protein of inflammation, is also a modulator. Homozygous patients for the alpha allele have a higher risk to develop renal amyloidosis [34, 40–42].

1.8 Genetic Consultation in FMF

Genetic counseling for patients with FMF is based on several elements. Patient’s family clinical data have to be specified. A family tree is drawn and related symptomatic or asymptomatic subjects mentioned. Ethnicity and consanguinity must be documented. The geneticist has to report biological or clinical elements supporting FMF, such as Mediterranean ancestry, a high number of unexplained autoinflammatory episodes lasting 2–3 days, high levels of CRP during attacks, and age of onset less than 30 years. The interview collects clinical signs to guide the diagnosis and finally prescribe the appropriate genetic test.

1.9 Conclusion

Since its discovery in 1997, the MEFV gene has been widely explored in FMF patients. This outstanding milestone, by elucidating the molecular basis of this hereditary disease and providing us with inestimable genetic data, dramatically improved FMF diagnosis, treatment, and prognosis. Initially, only the most common variants due to the Mediterranean founder effect were identified and the first diagnosis strategies were based on the search for these highly prevalent mutations. To date, Sanger sequencing is the most commonly used method, and numerous rare variants, many of unknown clinical significance, have been reported (InFevers currently records nearly 300 variants). Yet, the five classical mutations are still the most represented in genetics reports.

The family history, to be thoroughly collected during the genetic consultation, and the parental analysis of the identified variants remain two major prerequisites for FMF accurate genetic diagnosis. However, interpretation issues can occur when only one mutation is identified. In addition, VUS are common (e.g., p.E148Q) and their clinical implication is still in question. In both cases, clinical conclusion must prevail.

NGS is the future option for genetic diagnosis. With targeted-panel approaches, diagnosis algorithms will change, and several AID genes will possibly be sequenced in FMF patients. New challenges for the next years could be the identification of a new MEFV locus (FMF-like clinical pictures with no mutation in MEFV), through whole-exome sequencing. Our understanding of FMF and AID in general will undoubtedly be dramatically revised.

References

1.

Janeway TC, Mosenthal H. Unusual paroxysmal syndrome, probably allied to recurrent vomiting, with a study of nitrogen metabolism. Trans Assoc Am Physicians. 1908;23:504–18.

2.

Heller H, Kariv J, Sherf L, Sohar E. Familial Mediterranean fever. Harefuah. 1955;48:91–4.PubMed

3.

Heller H, Sohar E, Pras M. Ethnic distribution and amyloidosis in familial Mediterranean fever (FMF). Pathol Microbiol (Basel). 1961;24:718–23.

4.

Daniels M, Shohat T, Brenner-Ullman A, Shohat M. Familial Mediterranean fever: high gene frequency among the non-Ashkenazic and Ashkenazic Jewish populations in Israel. Am J Med Genet. 1995;55:311–4. doi:10.​1002/​ajmg.​1320550313.CrossRefPubMed

5.

Ozen S, Karaaslan Y, Ozdemir O, et al. Prevalence of juvenile chronic arthritis and familial Mediterranean fever in Turkey: a field study. J Rheumatol. 1998;25:2445–9.PubMed

6.

French FMF Consortium. A candidate gene for familial Mediterranean fever. Nat Genet. 1997;17:25–31. doi:10.​1038/​ng0997-25.CrossRef

7.

The International FMF Consortium. Ancient missense mutations in a new member of the RoRet gene family are likely to cause familial Mediterranean fever. Cell. 1997;90:797–807.CrossRef

8.

Shohat M, Livneh A, Zemer D, et al. Twin studies in familial Mediterranean fever. Am J Med Genet. 1992;44:179–82. doi:10.​1002/​ajmg.​1320440212.CrossRefPubMed

9.

Stoffels M, Szperl A, Simon A, et al. MEFV mutations affecting pyrin amino acid 577 cause autosomal dominant autoinflammatory disease. Ann Rheum Dis. 2014;73:455–61. doi:10.​1136/​annrheumdis-2012-202580.CrossRefPubMed

10.

Fukushima Y, Obara K, Hirata H, et al. Three Japanese patients (mother and two children) with familial Mediterranean fever associated with compound heterozygosity for L110P/E148Q/M694I and an autosomal true dominant inheritance pattern. Asian Pac J Allergy Immunol Launched Allergy Immunol Soc Thail. 2013;31:325–9. doi:10.​12932/​AP0244.​31.​4.​2013.

11.

Aldea A, Campistol JM, Arostegui JI, et al. A severe autosomal-dominant periodic inflammatory disorder with renal AA amyloidosis and colchicine resistance associated to the MEFV H478Y variant in a Spanish kindred: an unusual familial Mediterranean fever phenotype or another MEFV-associated periodic inflammatory disorder? Am J Med Genet A. 2004;124A:67–73. doi:10.​1002/​ajmg.​a.​20296.CrossRefPubMed

12.

Booth DR, Gillmore JD, Booth SE, et al. Pyrin/marenostrin mutations in familial Mediterranean fever. QJM Mon J Assoc Physicians. 1998;91:603–6.CrossRef

13.

Grandemange S, Aksentijevich I, Jeru I, et al. The regulation of MEFV expression and its role in health and familial Mediterranean fever. Genes Immun. 2011;12:497–503. doi:10.​1038/​gene.​2011.​53.CrossRefPubMed

14.

Sarrauste de Menthière C, Terrière S, Pugnère D, et al. INFEVERS: the registry for FMF and hereditary inflammatory disorders mutations. Nucleic Acids Res. 2003;31:282–5.CrossRefPubMedCentralPubMed

15.

Notarnicola C, Manna R, Rey JM, Touitou I. Y688X, the first nonsense mutation in familial Mediterranean fever (FMF). Hum Mutat. 2001;17:79. doi:10.​1002/​1098-1004(2001)17:​1<79:​:​AID-HUMU23>3.​0.​CO;2-2.CrossRefPubMed

16.

Balow JE, Shelton DA, Orsborn A, et al. A high-resolution genetic map of the familial Mediterranean fever candidate region allows identification of haplotype-sharing among ethnic groups. Genomics. 1997;44:280–91.CrossRefPubMed

17.

Yilmaz E, Ozen S, Balci B, et al. Mutation frequency of familial Mediterranean fever and evidence for a high carrier rate in the Turkish population. Eur J Hum Genet EJHG. 2001;9:553–5. doi:10.​1038/​sj.​ejhg.​5200674.CrossRef

18.

Touitou I. The