Clinical Genetic Testing of Periodic Fever Syndromes

Periodic fever syndromes (PFSs) are a wide group of autoinflammatory diseases. Due to some clinical overlap between different PFSs, differential diagnosis can be a difficult challenge. Nowadays, there are no universally agreed recommendations for most PFSs, and near half of patients may remain without a genetic diagnosis even after performing multiple-gene analyses. Molecular analysis of periodic fevers' causative genes can improve patient quality of life by providing early and accurate diagnosis and allowing the administration of appropriate treatment. In this paper we focus our discussion on effective usefulness of genetic diagnosis of PFSs. The aim of this paper is to establish how much can the diagnostic system improve, in order to increase the success of PFS diagnosis. The mayor expectation in the near future will be addressed to the so-called next generation sequencing approach. Although the application of bioinformatics to high-throughput genetic analysis could allow the identification of complex genotypes, the complexity of this definition will hardly result in a clear contribution for the physician. In our opinion, however, to obtain the best from this new development a rule should always be kept well in mind: use genetics only to answer specific clinical questions.


Introduction
Periodic fever syndromes (PFSs) represent a wide group of diseases characterized by recurrent attacks of apparently unprovoked in�ammation and are thus considered among the so-called autoin�ammatory diseases.
For the clinician, the question whether a patient suffers from a PFS usually arises aer the evaluation and exclusion of more common clinical problems associated with fevers and in�ammation, such as chronic infections, systemic autoimmune diseases, and paraneoplastic in�ammatory conditions [1]. However, the pattern of associated clinical manifestations, the age at disease onset and, above all, the stereotypic recurrence of attacks can induce the suspicion of a PFS. A key to diagnose PFSs appears to simply be the consideration of its evenience [2,3]: this can diminish the delay in time to diagnosis, avoiding in some cases repeated invasive and unsuccessful investigations performed to exclude more common multifactorial disorders.
A de�nite diagnosis is made easier today thanks to improved feasibility of genetic analysis for most PFSs: Familial Mediterranean Fever (FMF), Mevalonate Kinase De�ciency (MKD), Tumor Necrosis Factor Receptor-Associated Periodic Syndrome (TRAPS), and Familial Cold Auto-in�ammatory Syndromes (FCAS). Indeed, each of these diseases can be identi�ed and diagnosed by the detection of mutations in speci�c genes (MEFV, MVK, TFRSF1A, and NLRP3/NLRP12, resp.) [4] (Table 1).
A genetic cause is not yet known for another PFS, called PFAPA (Periodic Fever, Aphthous stomatitis, Pharyngitis, and cervical Adenitis). is is a common and benign condition that can be easily diagnosed from clinical data, not requiring specialized laboratory investigations [5].
All these autoin�ammatory diseases are characterized by recurrent �ares of systemic in�ammation, presenting sudden fever episodes associated with elevation of acute phase reactants and with a number of clinical manifestations that might include in�ammation of serosal surfaces and �oints,  skin rashes of unknown origin, lymphadenopathy, arthritis, as well as the involvement of other organs such as muscles and the central nervous system [6]. Rheumatic manifestations are extremely common and highly variable in their presentation and course in PFSs. In most cases, however, the typical recurrence of attacks with symptoms-free intervals can help in differentiating the disorder from a chronic rheumatic disease. Furthermore, the pattern of musculoskeletal involvement and concomitant manifestations can help to formulate the correct diagnosis, which can be con�rmed by genetic testing [1]. A family history of similar problems, when present, enforces the idea of a genetic origin of the disease. Due to some clinical overlap between different PFSs, differential diagnosis can be a difficult challenge that can require sequential or simultaneous analysis of different genes. In fact, there are no universally agreed recommendations for most PFSs. Various attempts have been made in order to develop clinical criteria or guidelines to identify which of these patients should be considered for genetic testing [7,8]. Databases such as INFEVERS (http://fmf.igh.cnrs.fr/ISSAID/infevers/) and the Eurofever Project (http://www.printo.it/eurofever/autoin�ammatory� diseases.asp) have been developed to collect clinical and genetic information on patients with all monogenic autoin�ammatory syndromes, including HPFs. is has led to improved recognition of the spectrum of clinical problems associated with these conditions as well as development of targeted treatment strategies [9].
In this paper, we focus our discussion on the effective usefulness of genetic diagnosis. Even if a molecular diagnosis is not always fundamental for the choice of an effective treatment, it is of great utility in the followup of patients, avoiding inadequate diagnostic procedures and focusing on the risk of speci�c complications, such as amyloidosis. e aim of this paper is to establish how much can the diagnostic system improve, in order to increase the success of PFS diagnosis.

Clinical Considerations and Treatment
Hereditary PFSs encompass a rare group of diseases that share lifelong recurrent episodes of in�ammatory symptoms and an acute phase response. Periodic fevers are typically present in the pediatric population, but a disease onset in adult age is not a rare occurrence for some conditions [10]. is paper will focus on four hereditary PFSs, which include two autosomal recessive conditions, FMF and MKD, and a group of autosomal dominant diseases, including TRAPS and FCAS [11]. Moreover, in pediatrics, these pathologies have been distinguished from the more common and usually self-limiting PFAPA [11]. �ther autoin�ammatory syndromes, associated with fever, show a continuous in�ammatory phenotype but not necessarily with a periodic feature. is category includes diseases such as MWS (Muckle-Wells Syndrome), CINCA (Chronic Infantile Neurologic Cutaneous and Arthritis) syndrome, PAPA (Pyogenic Arthritis, Pyoderma gangrenosum, and Acne), Behçet's disease, adult Still's disease, and SoJIA (Systemiconset Juvenile Idiopathic Arthritis) syndrome.
Not all inherited PFSs are rare diseases and for most of them the prevalence can vary depending on the ethnology [8,12]. As an example, FMF has a high prevalence, from 1 : 500 to 1 : 256 in different ethnic groups of non-Ashkenazi Jews [13], while MKD is a rare disease, and it is found predominantly in north-western Europe, with a carriage rate of MVK V377I of 1 : 350 [14]. Among the autosomal dominant PFSs, TRAPS is thought to be the most common form of dominant PFS in Europe [14].
e onset of symptoms occurs generally in childhood, usually in the �rst year of life for MKD and with more variability for other PFSs. Besides, the diagnosis of PFSs can be formulated later due to mild or unclear symptoms that overlap among different diseases and sometimes remain for long time misinterpreted or ignored [15]. In fact, the current approach for the diagnosis of PFSs in children may not be adequate for adults, in which the diseases are oen underdiagnosed, and there is the need for speci�c diagnostic algorithms [16,17].
e diagnosis of most periodic fevers in children is based on clinical history and physical examination. e management of the acute attacks is nonspeci�c and poorly standardized in PFSs. Similarly, long treatments are specifically designed to reduce the number and severity of the in�ammatory attacks, as well as to prevent the development of amyloidosis. [18]. It predominantly affects populations from Mediterranean descent in which the frequency of carriers is high [19,20]. In the vast majority of patients FMF becomes clinically apparent before the age of 20. Signs of painful serositis accompanying fever are the hallmark of the disease [21]. Acute attacks are characterized by a high-grade fever and in�ammation of one or more serous membranes (peritoneum in 90% of cases, pleura, synovial membrane, tunica vaginalis, and pericardium). Fever usually lasts less than three days and in most cases a single pick is observed. ere are usually no symptoms between acute attacks, although the joint involvement and myalgia may persist for several weeks.

Familial Mediterranean Fever (FMF). e FMF (MIM 249100) is the most frequent hereditary recessive PFS
Colchicine is the �rst-choice treatment for FMF, but a small percentage (5-10%) of patients does not show a complete response [22,23] to this drug and needs additional treatments such as interleukin-1beta (IL-1 ) inhibitors [24]. [25]. e �rst clinical episode usually begins before the end of the �rst year of life. Attacks start with chills, followed by a sharp rise in body temperature accompanied by cervical lymphadenopathy and abdominal pain. Hepatomegaly, splenomegaly, arthralgia, skin rash, diarrhea, and vomiting are common symptoms [26]. Mean attack duration is 5-7 days. Intervals between attacks usually range from 4 to 8 weeks and tend to increase with age aer adolescence.

Meval�nate �ina�e �e��ien�� (M��). MKD (MIM 260920) is a PFS identi�ed in 1984
Suspect of MKD is usually con�rmed by elevated concentrations of mevalonic acid in urines during the �ares and/or by a decreased enzymatic activity of mevalonate kinase during asymptomatic intervals. e diagnosis is consequently de�ned by genetic testing of the MVK gene. A marked elevation of polyclonal immunoglobulin D is found in the serum, so that the disease is also called Hyper-IgD Syndrome (HIDS), but this is neither speci�c, as it can also be found in some patients with other PFSs, nor sensitive, as younger patients may have normal IgD values.
In MKD, steroids are administered during febrile attacks, but for some patients with long-lasting �ares, the treatment becomes continuative. Other treatments (colchicine, cyclosporine, thalidomide, and statins) are of little bene�ts and biological drugs (anakinra and etanercept) have been used with some success as steroid-sparing agents [24]. e heterogeneous results obtained by novel biologic treatments [27,28] remain unclear since the molecular events leading to in�ammatory phenotype are still unknown, and a role of cell apoptosis has been recently proposed in the MKD pathogenesis [29,30].

Tumor-Necrosis-Factor-(TNF-)
Receptor-Associated Periodic Syndrome (TRAPS). �RAPS was �rst described in 1982 in a family of Irish and Scottish descent and was initially called "familial Hibernian fever" [31]. Since then, cases have been identi�ed in other populations [32]. Mean age at onset is approximately 10 years. Each attack lasts few days to few weeks. Sometimes, in addition to fever, the attacks are characterized by severe abdominal pain, which may mimic a surgical emergency. Cutaneous manifestations are present in 87% of patients during attacks [33]. Most patients exhibit localized painful erythematous macules and patches that tend to migrate to the distal parts of the extremities.
High-dosage corticosteroids are administered during febrile attacks. Colchicine and immunosuppressant are ineffective. Etanercept, to prevent febrile attacks and to avoid long-term renal complications, is effective in a subgroup of patients, while IL-1 inhibitors are effective in most. On the contrary anti-�NF monoclonal antibodies (in�iximab or adalimumab) have been shown to worsen the in�ammatory condition [16].

NOD-Like-Receptor-Protein (NLRP-) Related Diseases.
NLRP-related disease comprises phenotypically distinct autosomal dominant syndromes, such as NLRP3-associated Familial Cold Autoin�ammatory Syndrome (also named FCAS1) and NLRP12 associated periodic fever (also named FCAS2). Syndromes-1 (FCAS1). FCAS1, formerly known as familial cold urticaria, was �rst reported in 1940 [34], and, since then, only 20 families have been described worldwide. It is a rare autosomal dominant syndrome caused by a mutation in the NLRP3 gene, located on chromosome 1p44 that encodes a pyrin-like protein, known as cryopyrin [6]. ere are three distinct clinical disorders related to NLRP3 mutations (FCAS1, MWS, and CINCA) that can now be seen as a single disorder with variable phenotypic expression [35][36][37]. FCAS1 is characterized by attacks of fever, urticarial skin rash, arthralgia, and conjunctivitis brought on by exposure to cold. An episode starts 2-3 hours aer exposure and generally subsides within 24 hours. Attacks are accompanied by an intense acute phase response, as evidenced by high leukocyte counts in peripheral blood [38,39].

Familial Cold Autoin�ammatory
In patients with FCAS1, treatment with IL-1 antagonists (i.e., anakinra) allows a complete response for almost all patients, but it has limited effects on hearing loss, bone dysplasia, and mental retardation [24].

Familial Cold Autoin�ammatory Syndrome-2 (FCAS2).
FCAS2 is a more recently characterized condition, �rstly described in two families from Guadeloupe [37]. e disease is secondary to mutations in the NLRP12 gene with an autosomal dominant inheritance. It was the �rst example of an NLRP capable of negatively regulating NF-B activation [40], and it could be classi�ed as a NF-B activation disorder [41]. e phenotype is characterized by recurrent episodes of fever, secondary to cold exposure, associated with other signs, as skin rash, lymphadenopathy, aphthous ulcers, and abdominal pain [42]. A few patients, so far, have been presenting a good response to anti-in�ammatory drugs, given at occurrence, while treatment with anakinra induced only an initial good response, that decreased over time [24].

Periodic Fever, Aphthous Stomatitis, Pharyngitis, and
Cervical Adenitis (PFAPA). PFAPA syndrome was �rstly described in 1987 by Marshall et al. [43] and is characterized by episodes of fever, generally higher than 39 ∘ C, lasting for 3-6 days with recurrences every 3-8 weeks. Fever attack is associated with at least one of three main signs: aphthous stomatitis, cervical adenitis, and pharyngitis. e onset of the disease usually occurs before the age of 5 years and generally resolves by adolescence. Diagnosis is established on the basis of clinical criteria by which cardinal signs and symptoms must be carefully observed for a differential diagnosis [44,45]. Periodic fever with tonsillitis can be part of the clinical phenotype also in some of the previous mentioned monogenic PFSs, even if these diagnoses should be considered only in the presence of atypical signs, such as clinical relapse aer tonsillectomy, complaint of severe symptoms not referred to ear-nose and throat, and absent response to glucocorticoids [5].
Corticosteroids (prednisone, betamethasone) are successfully used in PFAPA syndrome during the febrile �ares, as they can dramatically switch off fever in few hours. Other accompanying symptoms, however, take longer to resolve, and corticosteroid therapy sometimes can lead to shorten the interval between attacks. Colchicine, used as prevention of febrile attacks, can induce an increased interval between fever attacks but not a complete remission. Cimetidine, a histamine type 2 receptor blocker with immunomodulating properties, was shown to be effective in only one-third of patients [45]. Although the role of tonsillectomy is still controversial, it remains the most effective intervention for long-term resolution of PFAPA syndrome symptoms [44,46].
e in�ammasomopathies include intrinsic in�ammasomopathies (such as FCAS1 [47,48]), extrinsic in�ammasomopathies (such as FMF [49,50] and MKD [51,52]), and Complex/Acquired In�ammasomopathies. is category was characterized by an activation of NLRP3 (originally denoted cryopyrin or CIAS1) that processes pro-IL-1 into an active form, mature IL-1 . e difference between intrinsic and extrinsic forms depends on whether the mutation is on NLRP3 constituents complex or on proteins that regulate the production of IL-1 , respectively.
e NF-B activation syndromes are mediated predominantly by improper regulation of NF-B within the innate immune system (i.e., FCAS2). In NF-B/FCAS2, although clinically similar to FCAS1, the mutational screening in NLRP3 and other known periodic fever genes is negative [37]. A subsequent examination of NLRP12 (NALP12, PYPAF7, and MONARCH-1), chosen because of its similarity to NLRP3 and because of its expression in myelomonocytic cells, revealed dominantly inherited nonsense and splice-site mutations in the two families. Moreover, NLRP12 was the �rst example of an NLR protein capable of negatively regulating NF-B activation [40]. It is not known whether NLRP12 can participate in an in�ammasome complex regulating IL-1 production.
e TRAPS syndrome instead is an example of a protein misfolding disorder. is kind of disorder is caused by multiple mechanisms leading to several problems, that can vary from misfolding of proteins to production of pro-in�ammatory cytokines by innate immune cells [53].
Finally, there are a number of disorders that appear clearly to be autoin�ammatory, but for which there are insufficient data to place them into one of the six categories. An example is the PFAPA, which is not inherited as a mendelian trait, but does exhibit some familial tendency [54,55]. While the exact cellular determinants and pathogenic molecular mediators remain elusive, the clinical picture and cytokine pro�le of PFAPA outline a very complex, heterogeneous, but nevertheless autoin�ammatory, disease [55].

Genetic Testing
Molecular analysis of periodic fevers' causative genes can dramatically improve patient quality of life by allowing early and accurate diagnosis and the administration of appropriate treatments.
However, the molecular genetic analysis of these diseases based solely on the candidate gene has low efficiency (close to 20%) and is time consuming and expensive. Moreover, since most of the causative mutations are located in speci�c regions of the genes mainly involved in these disorders, only some exons are analyzed for genetic diagnosis [56,57] (Figure  1). e greatest difficulty is to decide which gene should be �rst screened, based on patient's clinical features. is is true in particular for patients with low penetrance mutations in whom the clinical manifestations of the disorders can show wide overlapping. Aimed at ameliorating the efficiency of the genetic diagnosis, a set of clinical parameters that can predict the probability of carrying mutations in one of the genes associated with hereditary autoin�ammatory syndromes, has been proposed by Gattorno et al. [58]. In particular, the study allowed to identify some clinical variables (family history, age of onset, presence of abdominal, and chest pain, diarrhea) that are strongly associated with the probability of detecting relevant mutations in known PFS genes, thus also suggesting the order in which the genes should be screened. is diagnostic score revealed high sensitivity (87%) and good speci�city (72%) for the identi�cation of genetically positive and negative patients. Moreover, Federici et al. [59] performed a similar study in a large group of adult patients screened for three genes (MEFV, MVK, and TNFRSF1A) and obtained results comparable with the study carried out on children.
Unfortunately, despite the application of this validated diagnostic system, literature data support the evidence that PFSs are underdiagnosed or not correctly classi�ed. As reported by Lainka et al. [60], in Germany FMF is the most frequent PFS, but its incidence is low and, at present, genetic testing should be considered aer an accurate clinical history including ethnic ancestry and family history [60,61]. Besides, the overlap and the wide range of clinical signs shared by different PFSs and/or other diseases may cause a delayed or incorrect diagnosis [62]. In these cases, clinicians can be supported in a faster and correct identi�cation by appropriate genetic tests.
An ordinary medium-throughput genotyping method based on a 96-well sequencing plate can also be developed to analyze the most frequently PFS causing genes by automating several processes required for direct sequencing, using a commercially available robotic systems and routinely used instruments [63]. is medium-throughput genotyping method may improve diagnostic success in patients with overlapping phenotype but cannot completely solve all the lacks of correct evaluation that still exist in PFS diagnostics.
High-throughput DNA sequencing (HTS) could be also used for sequencing patients' whole genomes or targeted regions such as all exonic regions (i.e., the exome) [64]. e ob�ective is the identi�cation of genetic variants such as single nucleotide polymorphisms (SNPs) that can in�uence the clinical phenotype. e extraction of SNPs from the raw genetic sequences involves many processing steps and the application of a diverse set of tools, and thus its use is mainly limited to research [65]. ese second-generation sequencing technologies, including resequencing microarray, allow for multiple gene tests in a single experiment and thus hold considerable promise for routine molecular diagnosis of heterogeneous disorders [66]. Recent data con�rmed the key role of this technology in the diagnosis of autoin�ammatory diseases [67].
e expectation is that, in the very near future, this technology will enable us to identify all the variants in an individual's personal genome and, in particular, clinically relevant alleles. Beyond this, whole genome sequencing is also expected to bring a major shi in clinical practice in terms of diagnosis and understanding of diseases, ultimately enabling personalized medicine based on one's genome [68]. However, increasing the potential of genetics can bring to the attention new problems linked to the right interpretation of a huge amount of data.

Conclusions
PFSs are increasingly recognized in current clinical practice. is is probably due to a diminished clinical impact of infections at present times and to a wider awareness of the problem among physicians. PFAPA is by far the most common of PFSs and in most cases it can be easily diagnosed on the basis of the typical clinical presentation, usually in young children and with complaints mainly limited to pharynx, mouth, and neck. On the contrary, atypical PFAPA and other kinds of periodic fever can be a diagnostic challenge, even for experienced physicians. In some cases, the clinical picture may suggest a speci�c PFS, and the diagnosis can be con�rmed on the basis of speci�c clinical features, response to drugs, and ultimately molecular analyses. However, patients may present incomplete phenotypes that can be compatible with different PFSs, making it difficult for the choice of the correct analysis to perform. In such cases, the analysis of a set of different genes, thanks to the availability of new technological platforms, can help the clinician. Near half of patients may remain without a genetic diagnosis even aer performing multiple-gene analysis. In few of these cases, high-throughput genetic analysis will probably allow to identify novel monogenic disorders, but it is reasonable to hypothesize that a consistent percentage of patients will still remain without a de�nite diagnosis. e application of bioinformatics to high throughput genetic analysis could allow the identi�cation of complex genotypes, but the complexity of this de�nition will hardly result in a clear contribution for the physician.
In conclusion, great improvements have been obtained in recent years in the �eld of PFSs thanks to the identi�cation of a genetic cause for different disorders, and novel knowledge, including novel doubts, will come from high-throughput molecular analysis. In our opinion, however, to obtain the best from these new developments, a rule should always be kept in mind: use genetics only to answer speci�c clinical questions.