The results of our present study reveal increased prevalence of SpA and AS among FMF patients. Our results thus confirm previous observations. The first case of FMF in association with AS was described in 1963 . Thereafter, this association was reported in numerous other case reports. However, there are only a few studies to date that have systematically evaluated the relationship between FMF and SpA. In the first such study of this nature, 3,000 FMF patients were initially screened for the presence and manifestations of SpA , and among them, 160 patients with chronic arthritis were included in the subsequent study. In the aforementioned study, SpA was defined as the presence of chronic arthritis, inflammatory back/neck pain and sacroiliitis. Of the 160 FMF patients with chronic arthritis, 11 met the criteria for SpA of FMF (and actually all the subjects in this previous cohort fulfilled the New York criteria for AS). Moreover, the prevalence of SpA was estimated as 0.4% among the 3,000 initial patients. The HLA-B27 test was negative in all cases of SpA among these FMF patients. The authors excluded three other patients who had bilateral sacroiliitis, bamboo spine or were HLA-B27 positive. Although these cases met the authors' criteria of SpA of FMF they were interpreted as suffering from AS that coincided with their FMF, and were therefore clinically and genetically different from the other 11 patients. However, they also considered the possibility that these three patients might represent a more severe form in the clinical spectrum of FMF-related SpA, perhaps associated with the presence of HLA-B27.
In another earlier study from Turkey, the clinical and demographic features of 503 FMF patients were evaluated and the phenotypic differences between patients with and without amyloidosis were analyzed . In this previous study, the prevalence of clinically and radiologically proven sacroiliitis was 6% (3/50) in patients with amyloidosis, and was 11% (50/453) in patients without amyloidosis. Overall, the frequency of sacroiliitis was 10.5%. However, the authors did not provide detailed definitions of clinical or radiographic sacroiliitis or information regarding the MEFV or HLA-B27 status of their FMF patients with sacroiliitis.
Recently in another study from Turkey , the authors retrospectively reviewed the medical records of 256 FMF patients to evaluate the presence of one or more musculoskeletal manifestations (inflammatory low back pain, arthritis and enthesitis). Of the 70 FMF patients with musculoskeletal findings 55 agreed to participate in the study. Direct radiographs of the SIJs were used to grade the sacroiliitis and MRI analysis of the SIJs was performed in patients with sacroiliitis grade 0, 1 or 2 as determined on direct radiography. In the abovementioned study there were eight patients (3.1%) diagnosed with a grade 3 to 4 sacroiliitis by direct radiography. All of these cases were male and were HLA-B27-positive, and none had signs of vertebral involvement. There were an additional 10 patients (3.1%) with sacroiliitis identified on MRI, all of whom were HLA-B27-negative. Overall the frequency of sacroiliitis among all FMF patients was reported to be 7%.
The main difference between our current study and previous reports such as those discussed here is that we evaluated all of the cases in a cohort of 201 patients rather than a select group of patients with articular or musculoskeletal involvement. In addition to the increased prevalence of SpA and AS found among our unselected FMF patients, we show from our current analysis that all of our tested patients with AS were negative for HLA-B27. This finding suggests that factors other than HLA-B27 play a role in the association of FMF and AS. MEFV itself may be the link between these two disorders since we also revealed an association between the development of AS in FMF patients and the M694V variation in the MEFV gene. This is in line with a previous observation from Turkey  which showed that M694V is the most common variation in FMF patients with both radiographic and MRI evidence of sacroiliitis. However, in the latter study most of the patients with radiographic sacroiliitis had only one MEFV variant and all of them were HLA-B27 positive. Moreover, in our previous report, we reviewed 22 adult case descriptions , of which MEFV gene variants were analyzed in 15 cases. Twelve patients were homozygous for M694V, two patients were compound heterozygotes (M694V/M680I; M680I/V726) and one patient had a single polymorphism (M680I/-).
To further delineate the role of MEFV in the susceptibility to SpA and AS, we also assessed the FDRs of FMF patients who were predicted to have a higher carrier frequency. We also analyzed the parents of our probands for obligate carrier status. We found significantly increased frequency of AS and SpA in the FDRs of FMF patients. In addition, the frequency of AS among the parents of our FMF patients was also high in comparison with the general population.
In FMF patients, mutations are found throughout the MEFV gene; however, those producing the most severe phenotype are clustered in exon 10, which encodes the B30.2/SPRY domain (PRYSPRY), at the C terminus of the pyrin protein. The exact function of pyrin still remains somewhat controversial. N-terminal pyrin appears to activate nuclear factor-κB (NF-κB) through the increased calpain mediated degradation of inhibitor of NF-κB (IκB)-α . Recently, it was demonstrated that pyrin can interact with the apoptosis-associated speck-like protein (ASC), which has a caspase-recruitment domain (CARD) . In addition to its role in apoptosis, ASC also nucleates inflammasome complexes through the homotypic interactions of its pyrin domain and CARD with NLRP proteins and inflammatory caspases, respectively . Thus, ASC may mediate the activation of IL-1. The direct interaction of pyrin with ASC also uncovers potential molecular mechanisms for the abrupt onset inflammatory attacks associated with FMF .
Although the association between AS susceptibility and the class I molecule HLA-B27 is one of the strongest known HLA disease associations, the molecular mechanisms underlying disease pathogenesis still require clarification. In fact, the inability to explain this association on the basis of antigen presentation and major histocompatibility complex region, which reveal only half of the genetic susceptibility to AS, has led to alternative hypotheses . The IL-1 pathway might be one of the pathogenetic mechanisms involved in AS. Candidate gene analyses have implicated the IL-1 cluster of genes as an AS susceptibility locus  and subsequent meta-analysis of whole genome linkage scans supported the linkage of chromosome 2q (IL-1 gene cluster) with AS . Further, endoplasmic reticulum associated aminopeptidase 1 (ERAP1), a gene that was shown to be the strongest non-MHC gene associated with AS , also modulates the proinflammatory cytokines IL-1, IL-6, and TNF by cleaving their receptors at the cell surface .
In our present study, consecutive patients were included in order to minimize a possible bias for recruitment of patients with a history of SpA or AS, and a large number of relatives were also assessed. We evaluated both the patients and their FDRs using the same standardized protocol, which included a screening questionnaire, clinical examinations and radiography. One of the main limitations of our study may be its hospital-based nature. The frequency of more severe disease and allied conditions may be higher in those patients attending hospital visits regularly. Similarly, the frequency of symptomatic individuals may be higher in those relatives who agreed to participate in the study.
Unfortunately, we could not perform genetic analysis in the FDRs due to lack of resources, thus we could not use Amor, or the HLA-B27 arm of the ASAS classification criteria. Another limitation of this study was that only 319 (45%) of the available FDRs could be examined. However, our calculated prevalence rates represent the minimum prevalence, since we assumed that all of the potential subjects who did not participate in this study did not have SpA.