Discussion

First described by Marshall et al. in 1989,Reference Marshall, Edwards and Lawton¹ PFAPA syndrome is one of the less common paediatric recurrent fever syndromes.Reference Lee, Yang, Lee, Chen, Lin and Yeh² The cardinal symptom is recurrent, abrupt episodes of fever lasting several days, followed by an entirely asymptomatic interval period.Reference Marshall, Edwards and Lawton¹ Other common symptoms include aphthous stomatitis, pharyngitis, cervical lymphadenopathy, abdominal pain, headache, nausea, myalgia and diarrhoea.Reference Tasher, Somekh and Dalal³ Classically, recurring episodes begin before the fifth birthdayReference Marshall, Edwards and Lawton¹ and recurrences occur every three to eight weeks,Reference Marshall, Edwards and Lawton^1, Reference Long^4, Reference Scholl⁵ initially with clockwork regularity in early childhood but tending to become less frequent and less severe over time.Reference Long⁴ The syndrome rarely persists into adulthood.Reference Marshall, Edwards and Lawton^1, Reference Padeh, Brezniak, Zemer, Pras, Livneh and Langevitz^6–Reference Parikh, Reiter, Kenna and Roberson⁸ The child's growth and development are normal, and there are no recognised long term sequelae.

The diagnosis is generally one of exclusion, requiring differentiation from other recurrent fever syndromes, including cyclical neutropenia, familial Mediterranean fever and hyper-immunoglobulin D (IgD) syndrome. Cyclical neutropenia is a disorder associated with fever, malaise, adenopathy and abdominal pain, and is characterised by a severe fall in peripheral neutrophil count every 21 days,Reference Marshall, Edwards and Lawton^1, Reference Lee, Yang, Lee, Chen, Lin and Yeh^2, Reference Long^4, Reference Scholl⁵ making it readily distinguishable from PFAPA syndrome. Familial Mediterranean fever is an autosomal recessive disease largely limited to four Mediterranean ethnic groups (Turks, Arabs, Armenians and non-Ashkenazi Jews) and characterised by recurring episodes of self-limiting fever.Reference Lee, Yang, Lee, Chen, Lin and Yeh^2, Reference Long^4, Reference Scholl⁵ A common gene, MEFV, has been identified.Reference Scholl^5, Reference Cazeneuve, Genevieve, Amselem, Hentgen, Hau and Reinert⁹ Hyper-immunoglobulin D syndrome is a recessively inherited condition typified by recurrent fevers, arthritis, adenopathy and rash, in association with persistent, polyclonal elevation of serum IgD levels.Reference Lee, Yang, Lee, Chen, Lin and Yeh^2, Reference Long^4, Reference Scholl^5, Reference Frenkel and Kuis¹⁰ Other, less common periodic fever syndromes (familial Hibernian fever, familial periodic fever, etc) must also be excluded. Due to this diagnostic complexity, the diagnosis is generally made in the secondary or tertiary paediatric setting, often by those with an interest in infectious disease. Our patient had been extensively investigated to rule out the above conditions before the diagnosis of PFAPA syndrome had been made.

The cause of PFAPA syndrome remains undetermined. Initial theories based around genetic mutation and chronic infection have been largely unsupported by subsequent work.Reference Long^4, Reference Cazeneuve, Genevieve, Amselem, Hentgen, Hau and Reinert^9–Reference Yoshihara, Imamura, Yokoi, Shibata, Kano and Osone¹² More recent literature has raised the possibility of immune dysregulation as a significant factor.Reference Frye^13–Reference Leong, Karkos and Apostolidou¹⁶

The PFAPA syndrome may come to the attention of the otolaryngologist through referral for recurrent stomatitis or for recurrent pharyngitis with a view to tonsillectomy,Reference Long^4, Reference Galanakis, Papadakis, Giannoussi, Karatzanis, Bitsori and Helidonis¹⁷ with or without a definitive diagnosis.

There has been debate in the recent literature over the benefit of various treatment strategies; steroid treatment,Reference Tasher, Somekh and Dalal^3, Reference Hernandez-Bou, Giner, Plaza, Sierra and Martin Mateos¹⁸ cimetidineReference Lee, Yang, Lee, Chen, Lin and Yeh^2, Reference Dahn, Glode and Chan^7, Reference Parikh, Reiter, Kenna and Roberson⁸ and thalidomideReference Pinto, Lindemeyer and Sollecito¹⁵ all have their proponents and detractors. Most recently, however, particular attention has focussed on debating the role of tonsillectomy.Reference Tasher, Somekh and Dalal^3, Reference Long^4, Reference Parikh, Reiter, Kenna and Roberson^8, Reference Galanakis, Papadakis, Giannoussi, Karatzanis, Bitsori and Helidonis¹⁷ Various studies have quoted success rates from 0 to 100 per cent,Reference Leong, Karkos and Apostolidou¹⁶ and a corresponding spectrum of opinion has been voiced within the otolaryngology community regarding the benefits of this approach.

Whilst several of the case series cited above have shown impressive results for one or other of these therapeutic strategies, their numbers have been small, their results not easily reproducible, and there has been a paucity of properly controlled studies. As a consequence, the majority of recommendations for various treatment modalities – tonsillectomy included – are based on low levels of evidence.

Deep neck space infection is well known as a complication of acute tonsillar and para-tonsillar infection, and it is generally accepted that removal of the tonsils substantially reduces the risk of this condition developing. Our patient had undergone adenotonsillectomy one year prior to the episode described above, and had not suffered a PFAPA syndrome type episode in the meantime. Why she developed a recurrence (and subsequently an abscess) after this time is not known.

Deep neck space infection has not previously been reported in connection with PFAPA syndrome. The current report widens the spectrum of potential complications of this syndrome, as well as adding to the ongoing debate over the merits of tonsillectomy in the treatment of this poorly understood condition.