Role of the R92Q TNFRSF1A mutation in patients with familial Mediterranean Fever




To define the frequency of the R92Q tumor necrosis factor receptor–associated periodic syndrome (TRAPS) mutation in patients with familial Mediterranean fever (FMF) and to study the role of this mutation in FMF.


Ninety-two FMF patients and 250 controls were genotyped for the R92Q mutation. The frequency of R92Q was assessed among 5 groups of FMF patients.


R92Q was found in 6% of the controls, with an especially high carrier rate among Moroccan Jews (8%). R92Q was found in 3 (3.2%) of the 92 FMF patients, 1 homozygous for the MEFV M694V mutation and 2 heterozygous for M694V. All 3 patients showed partial response to colchicine. R92Q was not found in patients unresponsive to colchicine, nor was it found in patients with amyloidosis or in patients with FMF-like disease without MEFV mutations.


The frequency of the R92Q mutation in FMF patients is comparable with that of controls. Despite the fact that TRAPS and FMF share common biochemical pathways, we found no evidence for an interaction between these two genes.


Familial Mediterranean fever (FMF) is an inherited autoinflammatory disorder characterized by recurrent attacks of fever accompanied by peritonitis, pleuritis, arthritis, and a typical rash named erysipelas-like erythema (1, 2). In the past, the development of renal amyloidosis was the major cause of morbidity and mortality in FMF patients; however, introduction of colchicine treatment in the early 1970s changed the natural history of the disease (3), completely abolishing the attacks in 60% of the patients, substantially decreasing them in 35%, and preventing the appearance of amyloidosis in almost all of them. Approximately 5% of the patients do not respond to colchicine despite the use of maximal doses (4). High carrier rates of FMF have been described among the Mediterranean and Middle Eastern population, ranging from 1:5 in North African Jews, Arabs, and Turks to 1:3 in Iraqi Jews and Armenians (5–8). The disease is caused by mutations in the MEFV gene encoding a 781 amino acid protein named pyrin (9, 10). Pyrin affects the inflammatory response by regulating the processing of mature interleukin-1β (IL-1β), a potent pyrogenic cytokine. Depending on the experimental system used, pyrin has been shown to act both as an inhibitor and an activator of IL-1β processing (11–13). Pyrin also participates in regulating the NF-κB pathway (11, 14). To date, more than 60 disease-associated mutations have been identified, most of which are extremely rare (Infevers database, online at:

Tumor necrosis factor receptor–associated periodic syndrome (TRAPS) is another autoinflammatory disorder inherited in an autosomal dominant fashion. TRAPS was initially reported in Irish and Scottish families and is now known to be more common than previously thought (15). TRAPS, characterized by episodes of fever, abdominal pain, pleurisy, arthritis, myalgia, and localized tender skin lesions, is caused by mutations in the TNFRSF1A gene encoding the 55 kd receptor for tumor necrosis factor α (TNFα) (16). TRAPS pathophysiology involves a number of biochemical pathways, including NF-κB and IL-1β (17). Thus far, more than 50 different TNFRSF1A mutations have been associated with TRAPS (Infevers database, online at: Treatment with steroids, etanercept, a TNFα blocker, and anakinra, an IL-1 receptor antagonist, was found to be effective in most patients.

R92Q is a low-penetrance TNFRSF1A mutation that has been found to have a population frequency varying between 2% and 5%. We have recently encountered a patient with FMF homozygous for the M694V mutation who continued to experience residual symptoms despite maximal dose treatment with colchicine. The finding of a heterozygous R92Q TNFRSF1A mutation in this patient and 2 similar previously reported patients (18, 19) prompted us to examine what role this mutation plays in determining the final phenotype among FMF patients. Does it play a role in the development of amyloidosis or in response to colchicine treatment, and can it explain the disease in patients who carry only a single FMF mutation?


Ninety-two patients were recruited at the Sheba Medical Center, Israel. The Institutional Review Board approved the study and the participants gave informed consent. The diagnosis was based on the Tel Hashomer criteria (4). Patients from 5 categories were selected: 1) patients with definite FMF with 2 MEFV mutations; 2) patients with definite FMF with only 1 MEFV mutation; 3) FMF patients with 1 or 2 mutations who did not respond to colchicine treatment (nonresponsiveness to colchicine has previously been defined as an attack frequency of at least twice a month at any typical site while compliant with an oral dosage ≥2.0 [range 2–3] mg/day of colchicine) (20); 4) FMF patients with renal amyloidosis; and 5) patients with atypical disease and no detectable MEFV mutations. Five milliliters of heparinized blood was drawn from each participant. MEFV mutations were determined as previously described (5). The segment containing the R92Q TNFRSF1A mutation was amplified using the primers forward 5′-CACTGCATGGATGTGAGTGTGTAT-3′ and reverse 5′-GTTGGTTGTCAGACCCACAGAATA-3′. Amplification was carried out in a 25-μl reaction containing 50 ng of DNA, 13.4 ng of each primer, and 1.5 mM dNTPs, in a 1.5 mM MgCl2 polymerase chain reaction (PCR) buffer with 1.2 units of Taq polymerase (Bio-Line). After an initial denaturation of 5 minutes at 95°C, 30 cycles were performed (94°C for 30 seconds, 58°C for 30 seconds, and 72°C for 30 seconds), followed by a final extension of 10 minutes at 72°C. The PCR product was then cut twice. The Bst NI restriction enzyme cut the 380-bp product to yield additional 153-bp and 227-bp alleles in carriers of this mutation. The results were confirmed by using a different restriction enzyme, Nci I, which cut the normal allele to yield a 153-bp and a 227-bp product. We also assessed the frequency of R92Q in 250 controls of various ethnic origins.


Case study.

The index patient was a 38-year-old woman first seen at the FMF clinic at age 5 years due to recurrent febrile episodes of a duration of 1–2 days, accompanied by severe abdominal or chest pain occurring 2–3 times a month. In addition, she experienced arthritis of the ankles and knees that lasted up to 1 week and recovered spontaneously without any residual symptoms. Her parents were healthy and of North African Jewish descent. One of her 9 siblings experienced similar symptoms, although of lesser intensity. The diagnosis of FMF was made and 1 mg of daily colchicine was initiated. In 1998, genetic analysis showed that she was homozygous for the M694V MEFV mutation. Thirty-three years of supervision and treatment at the FMF clinic revealed that her disease was only partially responsive to colchicine despite a gradual increase of up to 3 mg/day. She continued to experience febrile episodes that appeared every 4–6 weeks, in addition to diffuse joint pain and recurrent rashes. At age 36 years, these episodes intensified and continuous elevations of both erythrocyte sedimentation rate and C-reactive protein (CRP) level were observed. Considerable improvement was noted on initiation of prednisone 20 mg/day; however, the symptoms and laboratory tests worsened once prednisone was tapered. In 2009, genetic testing revealed that she was heterozygous for the R92Q TNFRSF1A mutation.

Assessment of R92Q in controls and patients.

The frequency of R92Q in controls of different ethnic origins is shown in Table 1. A surprisingly high carrier rate was noted among North African Jews, especially among Moroccan Jews (8%). The overall frequency of R92Q in the control group was 6%. The ethnic origin of the patients is shown in Table 2. Since the majority of patients were of mixed origin (53%), we chose to present the data according to the origin of each of the MEFV alleles. Sixty percent of the MEFV alleles were of North African Jewish decent and 13.7% were of Iraqi Jewish decent. The most frequent MEFV mutation was M694V, which appeared in 85% of the carrier chromosomes, a finding consistent with the ethnic origin of these patients. R92Q was found in our index patient and only in 2 other patients, both with definite FMF but with only 1 MEFV mutation (Table 3). The 2 patients were heterozygous for M694V. Further questioning after finding the R92Q mutation revealed that the response to colchicine was not complete in these 2 patients. One of the patients, a 17-year-old female, continued to experience attacks of fever and abdominal pain, averaging once every 6 weeks, while the second patient, a 13-year-old male, reported 2 episodes of fever and 1 episode of severe ankle pain over a 1-year period. During one of these attacks, the patient presented with an increased leukocyte count (15.1 × 103) and an elevated CRP level (8.4 mg/liter). At the time, both of the patients were receiving 2 mg/day of colchicine.

Table 1. The frequency of R92Q in different Jewish groups in Israel
OriginTotal no. of controlsNo. (%) of carriers
Ashkenazim502 (4)
Tunisians502 (4)
Moroccans13611 (8)
Other North Africans140 (0)
Total North African Jews20013 (6.5)
Total25015 (6)
Table 2. Ethnic origin of the MEFV alleles
OriginNo. (%) of chromosomes
North African Jews110 (60.4)
Iraqi25 (13.7)
Ashkenazi10 (5.4)
Yemenite5 (2.7)
Turkish6 (3.2)
Syrian2 (1.0)
Arab12 (5.4)
Others14 (7.6)
Table 3. R92Q in patients with familial Mediterranean fever (FMF)
GroupNo. of patientsNo. of R92Q carriers
  • *

    The index patient.

  • 0/0: no mutations, 148/0: patients heterozygote for E148Q.

1. Patients with definite FMF with 2 mutations131*
2. Patients with definite FMF with only 1 mutation202
3. Colchicine nonresponders430
4. FMF patients with renal amyloidosis40
5. Others120
Total923 (3.2%)

Thirty patients in group 3 (nonresponders) had 2 MEFV mutations and 13 of them had only 1 detectable mutation. Patients in group 5 most likely experience an “FMF-like” condition(s). We also included in this group patients who were heterozygous for E148Q with no other detected mutations. This missense substitution was initially regarded as a low-penetrance mutation, but recent studies doubt this assumption, and currently it is not clear whether E148Q represents a true mutation or a benign polymorphism (21–23). Except for the index patient and the 2 patients in group 2, R92Q was not found in any of the other patients.

The overall frequency of R92Q in the patients was 3.2%, not significantly different from the control group (6%; P = 0.42).


This study was conducted in order to assess the role of the R92Q TNFRSF1A mutation in FMF. As an initial step we studied the frequency of R92Q in a group of North African Jewish and Ashkenazi controls. Surprisingly, we found a very high carrier rate for this mutation in North African Jews, especially in those of Moroccan origin. The overlap in symptoms between FMF and TRAPS and the association between this mutation and other inflammatory conditions (24–26) led us to hypothesize that unexplained FMF phenomena may involve this mutation. Theoretically, the combination of FMF mutations with R92Q may lead to an increased inflammatory state that could manifest itself as a periodic fever in the spectrum of TRAPS and FMF. If there is an interaction between the two genes, an increased prevalence of R92Q should be detected among the patients. The high frequency of both M694V and R92Q in North African Jews make this population ideal to look for such an interaction.

FMF is classically regarded as an autosomal recessive disorder and most of the patients diagnosed with this disorder carry 2 MEFV mutations. However, recently we and others have shown that in up to 25% of the patients, a single MEFV mutation is probably enough to manifest the disease (27–29). In these patients, an interaction with another mutated inflammatory gene may provide an explanation for the pathogenesis of the disease, and R92Q is an ideal candidate. Among definite FMF patients with a single MEFV mutation, we found 2 who were heterozygous for R92Q. The frequency of R92Q in this group is similar to what was found in other studies: Booty at al (28) found R92Q in 1 of 14 patients, and Koné-Paut et al (29) found this mutation in 3 of 21 single-mutation FMF patients. Taken together, the carrier rates for R92Q are very similar in all 3 studies and do not significantly differ from the carrier rate in controls.

R92Q was not found in any of the 4 FMF patients with renal amyloidosis, all of whom were homozygous for M694V. Known risk factors for amyloidosis in FMF patients include homozygosity for M694V, country of origin, a positive family history of amyloidosis, male sex, and the SAA1 α/α genotype (30, 31). Two recent studies have suggested R92Q as a possible risk factor for the development of amyloidosis (32, 33). Our study does not support this observation.

Colchicine nonresponsiveness, which is found in approximately 5% of FMF patients, is heterogeneous in its etiology and poses a major clinical problem (20). Even though the 3 carriers of the R92Q mutation experienced residual symptoms while receiving colchicine, none of the 43 nonresponder patients (group 3) that were described in a previous study dealing with this issue carried this mutation (20). Therefore, in most FMF patients, the lack of response to colchicine cannot be explained by the R92Q mutation.

The mutation also does not explain periodic fever or “FMF-like syndromes” in patients that do not display any mutations. In the last decade, additional periodic fever syndromes were characterized and new autoinflammatory genes were cloned; however, the molecular basis of the disease remains elusive in many of the patients with periodic fever.

Despite sharing common biochemical pathways, the similar frequency of R92Q in the patients and controls supports the notion that an interaction between TNFRSF1A and MEFV is minimal or nonexistent. Yet, we cannot rule out that the residual symptoms in the 3 patients heterozygous for R92Q are TRAPS associated.

Recently, a number of patients with heterozygous mutations in 2 autoinflammatory genes have been described. In addition to MEFV and TNFRSF1A, such combinations have been detected in MVK and TNFRSF1A (34, 35) and CIAS1 and MEFV (36). These patients often have overlapping symptoms and do not respond well to standard therapy. The scarcity of such patients provides support to our conclusion that these two autoinflammatory genes do not interact.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Pras 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. Marek-Yagel, Berkun, Padeh, Lidar, Langevitz, Livneh, Pras.

Acquisition of data. Marek-Yagel, Shinar, Bar-Joseph, Reznik-Wolf, Langevitz, Livneh, Pras.

Analysis and interpretation of data. Marek-Yagel, Berkun, Padeh, Lidar, Shinar, Bar-Joseph, Reznik-Wolf, Pras.