Periodic fevers enter the era of molecular diagnosisBMJ 2000; 320 doi: https://doi.org/10.1136/bmj.320.7242.1091 (Published 22 April 2000) Cite this as: BMJ 2000;320:1091
And they are throwing some light on inflammatory mechanisms
- Joost P H Drenth, resident in internal medicine (, )
- Jos W M van der Mee, professor of internal medicine
Periodic fever syndromes are a group of disorders characterised by attacks of fever separated by symptom free intervals. So far four types, all inherited, have been described. Familial Mediterranean fever and hyper-IgD and periodic fever syndrome (hyper-IgD syndrome) are transmitted as autosomal recessive traits while familial Hibernian fever and Muckle-Wells syndrome are autosomal dominant periodic fevers. Over the past decade the gene for Muckle-Wells syndrome has been localised and those for the other three disorders identified. This work has resulted in better means of diagnosing these rare conditions and has also thrown new light on the molecular basis of inflammation.
Familial Mediterranean fever is the commonest periodic fever disorder, occurring mainly in people originating from the Mediterranean basin.1 In Israel alone over 5000 people are estimated to suffer from this disorder. Attacks start before the age of 20 and are characterised by short (1-4 days) attacks of fever and serositis. Most patients have recurrent peritonitis, but pleuritis also occurs. Asymmetric monoarthritis of the large joints is common, while an erysipelas-like rash develops less often. Familial Mediterranean fever can be complicated by nephropathic amyloidosis of the AA type, but in most patients colchicine up to 1–2 mg/day prevents both the acute attacks and the development of amyloidosis. Diagnosis is based on clinical manifestations, ethnicity, family history, and response to colchicine.
After initial localisation on the short arm of chromosome 16 (16p 13) the gene for familial Mediterranean fever, designated MEFV, was cloned in 1997.2 3 It encodes a hitherto unknown protein named pyrin-marenostrin. The gene consists of 10 exons, and the genomic sequence of MEFV predicts a protein of 781 amino acid residues which are mainly expressed in granulocytes. Homology studies suggest that pyrin-marenostrin has a role in the autoregulation of inflammation. So far at least 15 mutations have been detected mostly in exon 10 but also in exon 2. The M694V mutation is the most prevalent, being found in about 80% of the familial Mediterranean fever chromosomes.4
Hyper-IgD syndrome occurs as recurrent fever in patients of Western European origin and to date some 140 patients are known. Attacks mostly start under the age of 1 year and the raised serum IgD serves as a biological marker of this syndrome. During episodes lasting 3–7 days patients suffer from abdominal distress (not peritonitis), lymphadenopathy, skin lesions, and arthralgias. Surprisingly, amyloidosis has not been described. The diagnosis is made from the typical clinical picture and a high IgD value.5
The familial occurrence made it possible to use linkage analysis to localise the gene for hyper-IgD syndrome on the long arm of chromosome 12 (12q24).6 This finding and other investigations, which showed a mild mevalonic aciduria with attacks, led to the identification of mevalonate kinase as the hyper-IgD gene.6 7 Mevalonate kinase follows 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase in cholesterol biosynthesis. In the hyper-IgD syndrome we and others detected several missense mutations in the gene encoding for mevalonate kinase, 6 7 which all result in decreased enzyme activity. It is surprising that a defect in cholesterol metabolism can cause a periodic inflammatory disease, and the connection remains a puzzle.
Familial Hibernian fever presents as a vertically transmitted syndrome with fever, abdominal pain, and localised myalgias. Episodic erythematous patches, conjunctivitis, unilateral periorbital oedema, and amyloidosis may occasionally be found. Attacks tend to last longer than in the hyper-IgD syndrome or familial Mediterranean fever. The condition was initially described in one family of Irish descent,8 but similar findings were later found in other families. A breakthrough came when in two separate families linkage was detected with chromosome 12 (12p 13).9 10 The type I tumour necrosis factor receptor appeared a good candidate gene because serum levels are grossly diminished during and between attacks. Indeed in seven different families, six missense mutations in the type I tumour necrosis factor receptor gene were identified, establishing this gene as the causative gene11 and suggesting that attacks result from decreased shedding of tumour necrosis factor receptor from the membrane. This discovery also led to a change of name to tumour necrosis factor-receptor associated periodic syndrome and opened the avenue to treatment with injectable soluble tumour necrosis factor receptor; preliminary experience suggests it might work (Drewe L, Kastner D, 2nd international workshop on hyper-IgD and periodic fever syndrome, Nijmegen, December 1999).
Patients with autosomal dominant Muckle-Wells syndrome suffer from acute febrile attacks with abdominal pain, arthritis, and urticaria sometimes complicated by progressive nerve deafness and multiorgan AA-type amyloidosis. With linkage analysis the gene for Muckle-Wells syndrome was recently localised at chromosome 1q44 and efforts are under way to clone the gene.12
The identification of the periodic fever genes has clear implications. Firstly, it sheds important new light on the molecular basis of inflammation and illustrates the exhaustive repertoire of inflammatory response in humans. Secondly, knowledge on correlates between genotypes and phenotypes influences patient care. For example, homozygosity for the M694V mutation is strongly associated with the development of amyloidosis, emphasising the need for early treatment with colchicine in these patients.13 Lastly, until recently the diagnosis of a periodic fever syndrome was made solely on clinical grounds, which can be difficult in some (non-familial) cases. With the discovery of the implicated genes, molecular testing becomes possible. Several service laboratories in the United Kingdom provide testing for familial Mediterranean fever. Nevertheless, rapid screening tests screen only for known mutations, and in some patients with a very typical clinical picture even advanced molecular tools may fail to detect mutations. In familial Mediterranean fever, for example, a genetic diagnosis will be made in only 65-94% of typical cases, even after thorough testing.14 Until the full mutational spectrum of these periodic fever genes are known clinicians should continue to make the diagnosis on clinical grounds and use the molecular test only as a useful adjunct.