Periodic fever is often considered a rare disease in east Asian populations. That's true when it comes to hereditary periodic fevers such as familial Mediterranean fever, hyperimmunoglobulin D syndrome and Muckle-Wells syndrome etc. Another type of periodic fever, PFAPA, which was thought to have no racial difference in incidence [6], is common in this region. There is no clear data on the incidence of PFAPA in China. A Norwegian population survey showed an incidence of 2.3 per 10,000 children under 5 years of age [7], while a study in the United States found that the incidence of PFAPA in siblings of normal control children was 2% [8], which is close to our estimate of a rough incidence of about 1% in China. However, the diagnosis rate of PFAPA in China is obviously much lower. Only a small part of the children had been suspected or confirmed a diagnosis of periodic fever before entering our study, and most of them had experienced a long course of disease when finally diagnosed. We believe that the main reason for missed diagnosis is that in China most doctors have not yet recognized this disease, and the diagnosis procedure still faces some problems to be solved.
Unlike hereditary recurrent fevers, which can be diagnosed by genetic testing, the diagnosis of PFAPA is primarily based on clinical presentation. As the name suggests, PFAPA is characterized by recurrent fever, aphthous stomatitis, pharyngitis tonsillitis, and cervical lymphadenitis. The periodicity of fever is undoubtedly the most fundamental characteristic of PFAPA. The cycle duration and episode duration in this study were 30 days and 4 days, respectively, which was similar to most reports [9, 10].
In addition to fever, pharyngitis is the most common manifestation of an attack in both adults and children. Tonsil exudation is uncommon in adult onset patients [11]. Whereas in this study pharyngitis with tonsil exudation is the most common symptom in these children. It is reported that even in the intermission phase, the activity of immune cells in the tonsils of PFAPA patients was different from that of the control group, as well as the bacterial flora [11, 12]. And tonsillectomy had complete resolution the attacks in most patients [4, 13, 14]. The tonsillar microenvironment seems to be an important factor in the pathogenesis of PFAPA. Leukocytosis and CRP elevation are frequently seen during attacks in most patients, which tends to mislead to bacterial infection.
Although the onset of PFAPA is more frequent in younger children, onset in school-ages and even in adults have also been reported [11, 15]. We also accepted patients with adult onset (not included in this study). Some patients had experienced a long remission period in disease evolution [4, 15]. It is reasonable to remove the age limit of onset < 5 years.
Most researchers believe that the growth and development of children with PFAPA are not affected [16–18]. whereas we observed abnormal growth indexes in children with PFAPA compared with normal children of the same age and sex. The growth failure may be due to long-term recurrently uncomfortable, unnecessary antibiotic use and inadequate nutritional intake derived from the practice of light diet during fever, which is very popular in China. Growth and development were not included in this two-step diagnostic criteria because other types of periodic fever do not affect growth and development within a short period from onset, and children with delayed growth and development cannot exclude the possibility of suffering from PFAPA.
The pathogenesis of PFAPA is not yet fully understood. No single pathogenic gene mutation has been identified to independently account for PFAPA. Genetic detection has never been used as a diagnostic element for PFAPA in the accepted diagnostic criteria. Although more than one-third of the first-degree relatives ever has periodic fever and almost half of children identify genetic abnormalities associated with auto-inflammation, phenotype conformity of the parents carrying the same mutations cannot be established. We agree that PFAPA is genetically predisposed but not inherited, and that PFAPA may be caused by the combination of various degree of genetic penetrance and environmental factors [10, 19, 20].
In all current PFAPA diagnostic criteria, response to steroids is one of the key elements [21]. Fever with leukocytosis and elevated CRP always reveals a picture of bacterial infection. In this case, treatment with corticosteroids alone are obviously a concern. In the two-step diagnostic method, indications for glucocorticoid use are established once suspected PFAPA is diagnosed. When using prednisone to abort an attack, we observed that smaller doses often resulted in recurrent fever requiring a second dose. To avoid misjudgment, in diagnosing stage, we recommend a single sufficient dose of prednisone.
In general, four episodes are sufficient to describe the characteristics of periodicity. The fourth or last episode, and the next, are launched for prednisone trial. In this study the vast majority of patients who met the proposed criteria were diagnosed with PFAPA after prednisone tests, except for one child. The CARRA-PFAPA criteria is a good diagnostic method, which includes: ① recurrent fever, ② more than 6 episodes occurring with regularity, ③ normal growth and development, ④ good response to steroid, ⑤ excluding other known autoimmune/autoinflammatory disorder, cyclic neutropenia, malignancy, infection, etc [17]. We used the CARRA-PFAPA diagnostic criteria to verify our two-step method. Most of patients met the CARRA-PFAPA criteria when the criteria “normal growth and development” were eliminated. Those who had only 5 episodes when diagnosing finally fulfilled the CARRA-PFAPA criteria in follow-up. The effectiveness of different diagnostic criteria varied and led to the concept of undifferentiated periodic/recurrent fever [18, 22, 23]. Dosage of steroid or growth development may contribute to the difference between our method and others.
In this study, we found that manifestations of PFAPA children in China was somewhat different from that reported abroad. The two-step diagnostic method provides the indication of prednisone use and is feasible for primary pediatricians. Although PFAPA is a benign disease, its repeated attacks still cause trouble to the children, their families and the national health insurance system [24–26]. Therefore, it is necessary to develop a diagnostic method suitable for local condition and raise awareness of PFAPA among primary health care professionals.