Periodicals Home

Search

User Pref

Help

Logout

J Peds Home

Table of Contents

All Issues

Order

About this Journal

<< Issue

>> Issue

Shai Padeh, MD [MEDLINE LOOKUP] Naphtali Brezniak, MD, DMD, MSD [MEDLINE LOOKUP] Debora Zemer, MD [MEDLINE LOOKUP] Elon Pras, MD [MEDLINE LOOKUP] Avi Livneh, MD [MEDLINE LOOKUP] Pnina Langevitz, MD [MEDLINE LOOKUP] Amyel Migdal, MD [MEDLINE LOOKUP] Mordechai Pras, MD [MEDLINE LOOKUP] Justen H. Passwell, MD [MEDLINE LOOKUP] Sections Abstract Patients and Methods Results Discussion References Publishing and Reprint Information

• Previous article in Issue • Next article in Issue • View print version (PDF) • Drug links from Mosby's DrugConsult • Genetic information from OMIM • Citation of this Article    • View on PubMed    • Download in citation manager format    • Download in Medlars format • Related articles in PubMed

Abstract

TOP

We report 28 patients (20 male) with a syndrome characterized by abrupt onset of fever, malaise, aphthous stomatitis, tonsillitis, pharyngitis, and cervical adenopathy (PFAPA syndrome). Episodes of fever occurred at intervals of 5.1 ± 1.3 weeks beginning at the age of 4.2 ± 2.7 years. Fever, malaise, tonsillitis with negative throat cultures, and cervical adenopathy were reported in all 28 patients, aphthae in 19, headache in 5, abdominal pain in 5, and arthralgia in 3. Mild hepatosplenomegaly was observed in 6 patients. Mild leukocytosis, elevation of the erythrocyte sedimentation rate, and fibrinogen were found during attacks. These episodes of illness resolved spontaneously in 4.3 ± 1.7 days. Serum IgD was found elevated (>100 U/mL) in 12 of the 18 patients tested (140.2 ± 62.4 U/mL). Affected children grow normally, have no associated diseases, and have no long-term sequelae. Attacks were aborted by a single dose of oral prednisone (2 mg/kg) at the beginning of the attack in all 15 patients in whom this medication was prescribed. In 9 patients the syndrome has completely resolved (beginning at the age of 2.9 ± 1.3 and lasting 8 ± 2.5 years). In 3 other patients complete resolution of the attacks occurred after tonsillectomy was performed. PFAPA is sporadic, and no ethnic predilection was found. Increased awareness of the clinical syndrome has resulted in more frequent diagnosis and adequate treatment. (J Pediatr 1999;135:98-101)

FMF Familial Mediterranean Fever PFAPA syndrome Periodic fever, aphthous stomatitis, pharyngitis, and adenopathy syndrome

See related articles, p. 1 and p. 15.

Recurrent fevers in children are common, and in most instances specific infections account for the fever. “Periodic fever” refers to recurrent bouts of fever with no apparent infection. Causes of periodic fever in children include systemic onset juvenile rheumatoid arthritis, cyclic neutropenia, Behçet’s disease, and other rheumatic diseases that are characterized by acute exacerbations and remissions. In Israel, the most common periodic disease is Familial Mediterranean Fever.¹ This disease is confined to those of Sephardic Jewish or Arab ethnicity and is characterized by recurrent episodes of fever and polyserositis. The molecular basis of the disorder has recently been defined.² To date, 3 point mutations in the sequence of the pyrin gene are found in ~90% of patients with FMF.² Recently, 2 new syndromes of periodic fever were described: the hyper IgD syndrome and the periodic fever, aphthous stomatitis, pharyngitis, and adenopathy syndrome. The hyper IgD syndrome is a familial disorder of periodic fever characterized by recurrent febrile attacks accompanied by abdominal distress, arthralgia or arthritis, headache, skin lesions, and an elevated serum IgD level (>100 U/mL); pharyngitis is not a feature of this syndrome.^3,4 Serum levels of IgD are elevated in these patients. PFAPA syndrome was first described by Marshall et al⁵ in 12 patients. We describe the presentation and course of the disease in 28 patients with PFAPA syndrome. This syndrome has particular relevance in Israel, where it must be distinguished from FMF. A single dose of corticosteroids aborted the febrile episodes.

PATIENTS AND METHODS

TOP

The inclusion criteria for children with PFAPA syndrome were recurrent fevers, exudative tonsillitis, recurrent oral ulcers, cervical lymph node enlargement, negative throat cultures, and failure of antibiotic treatment during the acute episode or as prophylactic treatment. Over the 4-year period from 1994 to 1998, 28 patients met the criteria for PFAPA syndrome and were monitored at the Sheba Medical Center. Children who initially fulfilled the criteria for PFAPA syndrome but did not have a trial of prophylactic antibiotics (oral penicillin VK or amoxycillin) and were later found to respond well to this therapy were not included in our series.³

Ten patients were identified and monitored at the national FMF outpatient clinic and were initially given the diagnosis of FMF.

FMF Gene Analysis

DNA was extracted from 2 mL blood with a commercial kit (Puregene, Gentra Inc, Plymouth, Minn) and was screened for 3 FMF mutations, M680I, M694V, and V726A, by polymerase chain reaction with a commercial kit (Gamidigen, Rehovot, Israel).

RESULTS

TOP

Of the 28 patients, 20 were male and 8 were female. Clinical features are summarized in Table I.

Table I. Clinical presentation of 28 patients with PFAPA

---

	No	%
Fever	28	100
Exudative tonsillitis	28	100
Malaise	28	100
Cervical adenopathy	28	100
Aphthae	19	68
Headache	5	18
Abdominal pain	5	18
Arthralgia	3	11

Symptoms resolved irrespective of treatment with antipyretics or antibiotics. The persistent fevers usually prohibited the children from pursuing normal activities.

Medical histories were unremarkable; all patients had completed their immunization schedule. No family history of other diseases was reported, except for the 6 of 10 children from the FMF group whose family history was compatible with autosomal inheritance of FMF (see following text). No children had affected siblings. Physical examination between episodes was normal. Children had normal growth and development.

The age of onset of the disease was 4.2 ± 2.7 years with a range of 6 months to 7 years of age. The average duration of the attack was 4.3 ± 2.7 days with a range of 2 to 8 days of fever without treatment. The average duration of the interval between attacks was 5.1 ± 1.3 weeks, or 11 attacks per year, at the peak of the disease. The average follow-up of these patients was 5.0 ± 3.4 years (range 1.8 to 9.8 years). In 9 patients symptoms had ceased by the time of this report. In these, symptoms began at the age of 2.9 ± 1.3 and had stopped by the age of 10.9 ± 2.5 years. In 10 patients the attacks recurred at precise intervals, with an exact time interval between episodes. However, in most children the intervals varied between 3 to 6 weeks apart. Four patients noticed that the attacks became more periodic only after prednisone was prescribed.

Twenty-seven patients were Jewish, and 1 was an Arab child. No particular predilection of ethnicity was seen among patients with PFAPA syndrome.

None of the 10 patients in whom FMF was initially diagnosed was homozygote or double heterozygote for the defined mutations of this disease. Six patients were heterozygote for the M694V mutation and were therefore carriers of the FMF gene (Table II).

---

Table II. The FMF gene mutations in the PFAPA group

---

Patient	Sex	Origin	Mutation	FMF in family*
1	F	Turkey-Iran	Heterozygote -M694V	3
2	F	Morocco-Morocco	Heterozygote -M694V	0
3	M	Iraq-Iraq	Heterozygote -M694V	0
4	M	Egypt-Egypt	No mutation†	0
5	M	Rumania-Serbia	Heterozygote -M694V	3
6	M	Algeria-Libya	Heterozygote -M694V	1
7	M	Morocco-Morocco	No mutation†	0
8	M	Iraq-Iraq	Heterozygote -M694V	2
9	M	Morocco-Morocco	No mutation†	0
10	F	Libya-Morocco	No mutation†	0
*FMF in family = number of siblings with FMF. †680I, M694V, and V726A were not found.

---

In the other 4 patients no mutation was found. Four out of the 6 patients in the heterozygote group had siblings with FMF, whereas the 4 who had no mutations had no family members with FMF. In retrospect the clinical presentation of these 10 patients was similar to that of the other described patients in our series. Colchicine, which had been prescribed in these patients, had only a partial effect and was discontinued and replaced by prednisone with a dramatic effect on the febrile episodes in 5 children who continued to have febrile episodes.

In the total group of 28 patients, laboratory investigation at onset of the fever showed a mean hemoglobin level of 128 ± 2 g/L, mild leukocytosis of 13.0 ± 2.6 × 10-9 mm leukocytes/mm³, moderate elevation of the sedimentation rate 56 ± 19 mm/1st h, and platelet count of 225 ± 15.8 × 10⁹/L (Table III).

---

Table III. Laboratory investigation

---

	Mean ± SD
Hemoglobin	128 ± 2 g/L
Leukocyte	13.0 ± 2.6 × 109/L
Platelet	225 ± 15.8 × 109/L
ESR	56 ± 19 mm 1st h
Fibrinogen	370 ± 117 mg/dL
IgG	9.26 ± 2.52 g/L
IgM	1.15 ± 0.38 g/L
IgA	1.63 ± 0.68 g/L
IgD*	140.2 ± 62.4 U/mL
ESR, Erythrocyte sedimentation rate. *Significantly elevated compared with con-trol group (P < .001 analysis of variance).

---

Repeated leukocyte counts between attacks were obtained to exclude cyclic neutropenia and were found within the normal range in all children.

Serum IgD levels were elevated in 12 of the 18 patients in whom they were measured. The levels were >100 U/mL, which is the cutoff level for the hyper IgD syndrome. The serum IgD levels (140.2 ± 62.4 U/mL) were significantly higher than those found in healthy children in an age-matched control group (16.5 ± 15.8 U/mL, P < .0001 by analysis of variance) or children with juvenile rheumatoid arthritis (85.9 ± 47.4 U/mL, P < .001 by analysis of variance).

The patients’ serum was obtained when they were in remission and therefore does not reflect a nonspecific acute phase reactant characterizing the febrile attack. The other serum immunoglobulin values were within the normal range (Table III), indicating that the elevated serum IgD levels were specifically elevated in these patients.

Corticosteroids were not prescribed in 5 patients in whom the syndrome had terminated by the time of this report. A single dose of prednisone (2 mg/kg) resulted in abrogation of each of the recurrent episodes in 23 children treated at the onset of the attack. All patients reported a dramatic improvement in their general well-being, return of their temperature to normal, and complete abortion of the attack within 2 to 4 hours after the ingestion of the corticosteroids. The patients were re-examined on the next day, and all had a normal physical examination with clearing of the tonsillar exudates and oral ulcers. Four children reported an increasing rate and periodicity of the attacks after the initiation of the corticosteroid therapy, which also recurred exactly at 3-week intervals.

DISCUSSION

TOP

We have identified 28 children with the clinical criteria for PFAPA syndrome. Ten patients originally had a diagnosis of FMF, which we have excluded by genetic analysis. The clinical response of abrogation of all symptoms after a single dose of prednisone prevented considerable morbidity in our patient population. The dramatic response to a single oral dose of corticosteroids is unique to this syndrome; therefore we suggest that this therapeutic measure be used as a diagnostic criterion in a patient thought to have PFAPA syndrome; albeit the mechanism of action is speculative. The effect of the use of steroids regarding possible rebound phenomenon and long-term efficacy still requires further investigation in a controlled study. Successful prophylactic therapy with cimetidine has been reported in 6 patients with PFAPA syndrome.⁶ We believe that the intermittent use of a single oral corticosteroid dose should be the preferred first line of treatment because it is safe, convenient, and cost-effective.

We used the recommended cutoff levels of 100 U/mL for identification of hyper IgD levels; 66% of our patients had an increase in serum IgD levels between attacks. These levels were significantly higher than normal control levels (P < .001) and those of a group of patients with juvenile rheumatoid arthritis (P < .0001). However, these levels are less than those observed in the hyper IgD syndrome.³ The role of IgD in the syndrome is unknown. The syndrome we report is apparently quite different from the hyper IgD syndrome. In hyper IgD syndrome, symptoms often start at an early age and persist throughout life, attacks are severe, abdominal pain, arthralgia, and arthritis and skin lesions are common, whereas pharyngitis is not a feature of the disease. The course of PFAPA syndrome may be prolonged, but it is self-limited.

In our population FMF was mistakenly diagnosed in 10 of our patients. Retrospective review of these patients’ clinical symptoms showed that their symptoms were indeed atypical for FMF. Clinical response to colchicine was variable; in 5 of these patients the febrile attacks responded to steroids, and genetic analysis strongly indicated that FMF was not the correct diagnosis. Nevertheless 6 of 10 of these patients were heterozygotes for the FMF gene. We believe that this result most likely reflects the normal distribution of heterozygotes in our population; however, because FMF is also characterized by recurrent episodes of sterile inflammation, it remains to be determined whether the heterozygote may be more susceptible to the PFAPA syndrome.

The diagnosis of PFAPA must be considered in any child with recurrent unexplained fevers. We have found that PFAPA syndrome is apparently not as rare as believed and is not well recognized by pediatricians or otorhinolaryngologists. Increased awareness has resulted in more frequent diagnosis, prevented unnecessary investigations, provided considerable reassurance to patients and parents that they do not have a serious disease, and in our series considerable reduction in morbidity after steroid therapy.

We thank Drs Joost P. H. Drent and Ina S. Klasen from the University Hospital St. Radboud, Nijmegen, The Netherlands, for their help with immunoglobulin determinations in patients’ sera.

REFERENCES

TOP

1.  Sohar E, Gafni J, Pras M, Heller H. Familial Mediterranean fever. A survey of 470 cases and review of the literature. Am J Med 1967;43:227-53.

MEDLINE


2.  The International FMF Consortium. Ancient missense mutations in a new member of the RoRet gene family are likely to cause Familial Mediterranean Fever. Cell 1997;90:797-807.

MEDLINE


3.  Drenth JP, Denecker NE, Prieur AM, Van der Meer JW. Hyperimmunoglobulin D syndrome. Presse Med 1995;24:1211-3.

MEDLINE


4.  Drenth JP, Haagsma CJ, van der Meer JW. Hyperimmunoglobulinemia D and periodic fever syndrome. The clinical spectrum in a series of 50 patients. International Hyper-IgD Study Group. Medicine 1994;73:133-44.

MEDLINE


5.  Marshall GS, Edwards KM, Butler J, Lawton AR. Syndrome of periodic fever, pharyngitis, and aphthous stomatitis. J Pediatr 1987;110:43-6.

MEDLINE


6.  Feder HM Jr. Cimetidine treatment for periodic fever associated with aphthous stomatitis, pharyngitis and cervical adenitis. Pediatr Infect Dis J 1992;11:318-21.

MEDLINE

Publishing and Reprint Information

TOP

• From the Department of Pediatrics and the Department of Medicine C and F, Heller Institute of Medical Research, Sheba Medical Center, affiliated with the Tel-Aviv University, Sackler School of Medicine, Tel Hashomer, Israel.
• Presented at the 108th Annual Meeting of the Pediatric Academic Society, New Orleans, Louisiana, May 1-5, 1998.
• Submitted for publication Sept 9, 1998.
  
• Revision received Dec 21, 1998.
  
• Accepted Jan 11, 1999.
  
• Reprint requests: S. Padeh, MD, Pediatric Department, Sheba Medical Center, Affiliated with the Tel-Aviv University, Sackler School of Medicine, Tel Hashomer 52621, Israel.
  

• Copyright © 1999 by Mosby, Inc.
  

• 0022-3476/99/$8.00 + 0  9/22/97123