Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Cervical Adenitis (PFAPA) Syndrome in Children—From Pathogenesis to Treatment Strategies: A Comprehensive Review

Abstract

Periodic fever, aphthous stomatitis, pharyngitis, and cervical adenitis (PFAPA) syndrome is the most frequent periodic fever syndrome in non-Mediterranean children, usually manifesting before the age of 5 years. It is characterized by clockwork episodes of fever lasting 3–7 days, accompanied by aphthous stomatitis, pharyngitis, and/or cervical adenitis. Typically, patients with PFAPA are generally well between episodes and exhibit normal growth and development. Although PFAPA often resolves spontaneously, its recurrent nature can significantly impact quality of life, and symptoms may persist into adulthood. This narrative review aimed to consolidate current knowledge on PFAPA epidemiology, pathogenesis, clinical presentation, diagnostic considerations, and therapeutic options. A structured literature search was performed using PubMed, Cochrane Library, and Scopus, focusing on relevant articles specifically addressing PFAPA. Increasing evidence suggests multifactorial pathogenesis involving innate immune dysregulation, activation of the NLRP3 inflammasome, and Th1-driven inflammation. Genetic analysis studies suggest a polygenic inheritance of PFAPA, linking it to immune pathways shared with familial Mediterranean fever and Behçet’s disease. Diagnosis remains clinical, though genetic testing may be warranted in specific cases. Management strategies vary owing to the absence of standardized guidelines. Oral corticosteroids are highly effective for acute episodes but may shorten the interval between flares. Among preventive therapies, colchicine appears to reduce attack frequency, although evidence of its efficacy is limited, while tonsillectomy is often considered curative but recommended for patients with refractory disease or when there is a concurrent otolaryngologic indication. Further research is needed to refine diagnostic criteria and optimize treatment strategies, ultimately improving patients’ and caregivers’ quality of life.

Key Points

Periodic fever, aphthous stomatitis, pharyngitis, and cervical adenitis (PFAPA) syndrome is the most common periodic fever syndrome in non-Mediterranean children, presenting with regular episodes of fever, pharyngitis, cervical adenitis, and aphthous ulcers. It is self-limiting but can persist into adulthood and affect quality of life.

Pathogenesis involves innate immune dysregulation, NLRP3 inflammasome activation, and Th1-driven inflammation, with genetic overlaps with familial Mediterranean fever (FMF) and Behçet’s disease.

Diagnosis is clinical; differential diagnoses include monogenic autoinflammatory diseases (e.g., FMF, tumor necrosis factor receptor-associated periodic syndrome (TRAPS), mevalonate kinase deficiency (MKD)), syndrome of undifferentiated recurrent fever (SURF), cyclic neutropenia, and recurrent infections. Genetic testing and selected biochemical markers may aid in the differential diagnosis.

Treatment remains variable owing to the lack of standardized guidelines. Corticosteroids are effective for aborting flares but may shorten intervals; colchicine may help in MEFV carriers; and tonsillectomy can be curative. Management should follow a shared decision-making process or a personalized case-by-case approach.

Background

Periodic fever, aphthous stomatitis, pharyngitis, and cervical adenitis (PFAPA) syndrome is considered to be the most common periodic fever syndrome among children, especially in non-Mediterranean countries [1]. PFAPA is mainly characterized by clockwork-like, regularly occurring episodes of high fever, lasting 3–7 days and recurring every 2–8 weeks, along with the associated hallmark features of aphthous stomatitis, pharyngitis, and cervical adenitis. In addition to these cardinal symptoms, abdominal pain, headache, and myalgia can also be observed [2]. The vast majority of cases (90%) appear by the age of 5 years, although adult-onset PFAPA cases have also been reported [3, 4]. Patients typically remain symptom-free between episodes and maintain overall normal growth and development. PFAPA in children generally has a favorable outcome, resolving spontaneously before adulthood without long-term sequelae [5]. However, it can significantly impact both the children’s and their caregivers’ quality of life [6]. Since its first description by Marshall et al. in 1987 [1], the pathophysiology of PFAPA syndrome remains largely misunderstood, as no monogenic origin has been identified to date. Longitudinal observations and analyses of specific immunological markers suggest that PFAPA may share pathophysiological mechanisms with both rare and more common inflammatory disorders. This overlap underscores its complexity and may contribute to the challenges in establishing clear clinical criteria, as seen in other multifactorial diseases [7, 8].

The management of PFAPA can be particularly challenging when symptomatic treatment alone proves insufficient. Moreover, the lack of standardized, objective assessments of therapeutic efficacy limits the validity of universally applicable management recommendations [9] .

Our aim was to examine PFAPA syndrome in terms of its epidemiology, pathogenesis, clinical manifestations, diagnostic criteria, and differential diagnosis, with a particular focus on the current treatment strategies

Literature Search Methods

We searched the databases PubMed, Cochrane Library, and Scopus from 2010 to 2025, following established guidelines for narrative reviews. We utilized the keywords: “periodic fever, aphthous stomatitis, pharyngitis, and cervical adenitis syndrome,” “PFAPA,” “Marshall’s syndrome,” and “children.” We evaluated case reports, original research studies, and review articles specifically addressing PFAPA syndrome. In addition, we examined relevant publications cited within the selected articles. The search was restricted to English-language publications.

Epidemiology

Population-based studies of PFAPA remain scarce. Forsvoll et al. [10] reported an annual incidence of 2.3/10,000 in children up to 5 years of age. Consistent with this, Renko et al. [11] reported an annual incidence in Finland of 2/10,000 in children < 5 years. Rydenman et al. [12] showed an annual incidence of 2.6/10,000 of PFAPA in children aged < 5 years and an incidence rate of 0.86/10,000 in children aged 0–17 years. A slight predominance of PFAPA in boys has been described [13].

Pathogenesis

The clinical presentation of periodic fevers, along with evidence showing the innate immune system activation in PFAPA, has placed the syndrome among the autoinflammatory disorders, even though the exact etiology of the syndrome remains unclear [14]. Increasing evidence [7, 8, 15, 16] suggests that PFAPA results from a loop-like association between raised inflammatory cytokines, dysregulated immune system, environmental triggers, and genetic susceptibility, ultimately leading to a multifactorial disease.

Inflammatory Cytokines and Immune Dysregulation

A feature of PFAPA pathogenesis is the aberrant activation of the NLRP3 inflammasome, as suggested by the increased serum Caspase-1 and interleukin-1β (IL-1β) levels during febrile attacks [17]. In addition, patients with PFAPA show elevated acute phase reactants such as interleukin-6 (IL-6) and tumor necrosis factor (TNF) levels, central to the acute-phase response [15]. IL-6, in particular, induces the production of C-reactive protein (CRP) and serum amyloid A (SAA), thereby contributing to systemic inflammation in PFAPA.

Moreover, during PFAPA flares, monocytes and neutrophils exhibit heightened activity, as indicated by the increased expression of S100A12 and MRP8/14, both markers of early stage neutrophil and macrophage activation [16]. Although PFAPA has traditionally been considered an innate immune disorder, increasing evidence suggests the involvement of the adaptive immune system. In fact, IL-1β acts as a co-stimulator of T cell activation; activated T cells secrete interferon gamma (IFN-γ), which counteracts the antiinflammatory cytokines (IL-4 and IL-10) and induces the synthesis of T cell chemokines such as CXCL9 and CXCL10, which were elevated in the serum of patients with PFAPA [7]. PFAPA flares are thus driven by a Th1-mediated inflammatory response. Similarly, tonsillar biopsies from patients with PFAPA revealed an increased presence of cytotoxic CD8⁺ T cells, along with overexpression of Th1-associated chemokines (CXCL9 and CXCL10) [8]. Infiltration of activated T cells into inflamed tissues may contribute to the prolonged inflammatory state observed during PFAPA flares [16]. Another critical aspect of immune dysregulation in PFAPA is the disrupted balance between proinflammatory and regulatory signals. Studies have described reduced levels of IL-4, a cytokine associated with Th2 regulatory immune responses, in both serum and tonsillar tissue of patients with PFAPA [15]. This reduction of IL-4, coupled with overall decreased regulatory T cell activity, likely contributes to the persistent inflammatory response. In addition, lower eosinophil counts observed in patients with PFAPA further support the hypothesis of impaired antiinflammatory mechanisms [15].

In line with this hypothesis, in a pioneering proteomic analysis of tonsillar tissue in children with PFAPA syndrome, Mutlu et al. [18] revealed that while histological and immunohistochemical comparisons between patients with PFAPA and controls showed no significant differences, proteomic profiling identified marked downregulation of proteins involved in mitochondrial oxidative phosphorylation, particularly within complexes II, III, and IV and the pyruvate dehydrogenase complex. These findings suggest a shift from oxidative metabolism to glycolysis during active inflammation—an established hallmark of immune cell activation. The study proposes that this metabolic switch, characterized by impaired pyruvate oxidation and reduced acetyl-CoA production, may intensify activation of the NLRP3 inflammasome and downstream IL-1β secretion. Moreover, this shift could disrupt neutrophil chemotaxis and influence T cell differentiation, favoring proinflammatory subsets over regulatory T cells. Although limited by small sample size and the collection of tonsillar samples outside of acute episodes, this study provides compelling evidence that mitochondrial dysfunction and immunometabolism are central to the pathophysiology of PFAPA and may offer new therapeutic targets to modulate excessive IL-1β-driven inflammation.

Genetics

Unlike other monogenic autoinflammatory syndromes, PFAPA does not exhibit highly penetrant mutations in a single gene. Genetic analysis studies have revealed that PFAPA is a polygenic disorder, where multiple common variants contribute to disease susceptibility​ [19–21].

Mediterranean Ancestry and Ethnic Susceptibility

In countries where FMF is endemic, differentiating PFAPA from familial Mediterranean fever (FMF) can be challenging [22]. An ethnic predisposition for PFAPA has been observed in populations of Mediterranean descent, with a higher prevalence among Sephardic Jews, Israeli Arabs, and Turkish individuals. The distribution is similar to that of FMF, caused by mutations in the MEFV gene​. Although MEFV mutations have not been consistently identified in patients with PFAPA, some studies suggest that heterozygous carriers of MEFV variants may have an increased risk of developing PFAPA​ [22–24]. Berkun et al [20] found that 52% of patients with PFAPA carried at least one MEFV mutation, a significantly higher prevalence than the general Israeli population (20–25%). None of the patients with PFAPA carrying single MEFV mutation fulfilled the clinical criteria of FMF. Interestingly, patients with PFAPA with MEFV mutations experienced shorter fever episodes, fewer cyclic episodes, and a lower frequency of aphthous ulcers. Moreover, they required lower corticosteroid doses to control fever attacks​. In addition, colchicine was ineffective in these patients, further distinguishing these two conditions. These findings suggest that MEFV mutations may modulate the severity of PFAPA through an unclear mechanism. Another study by Butbul Aviel et al. [24] in Israeli patients found that almost 20% of patients with PFAPA were also diagnosed with FMF, with 93.3% of patients carrying at least one MEFV variant, suggesting that a shared genetic predisposition may contribute to overlapping inflammatory mechanisms in both diseases. In contrast to what was reported by Berkun et al. [20], most patients were responsive to colchicine. However, the fact that 32% of patients with PFAPA also met the clinical and genetical criteria for FMF may have introduced a bias in the interpretation of the results.

In a cohort of 131 Turkish patients with PFAPA, Konte et al. [22] identified MEFV mutations in 80.9% of participants, with heterozygous M694V being the most common variant (29%), underscoring that children in FMF-endemic regions can display overlapping features of both PFAPA and FMF. Approximately 62% of these patients responded favorably to colchicine monotherapy. However, as mentioned in the study by Butbul et al., the inclusion of patients fulfilling both clinical and genetic criteria for FMF (29%) may have led to misclassification of PFAPA and consequently limited the validity of the findings. Further studies are needed to clarify the role of MEFV mutations in the pathogenesis of PFAPA.

Genetic Overlap with Behçet’s Disease and Recurrent Aphthous Stomatitis

A large-scale study conducted on European–American and Turkish PFAPA cohorts has identified several genetic risk loci that PFAPA shares with Behçet’s disease (BD) and recurrent aphthous stomatitis (RAS). Among these, the IL12A gene variant (rs17753641) stands out as the strongest genetic risk factor for PFAPA. It encodes the p35 subunit of the cytokines IL-12 and IL-35. IL-12 is known to promote T helper1 (Th1) cell polarization and IFN-γ production. This suggests that patients with PFAPA carrying this variant are prone to excessive Th1 cell activation. Similarly, another risk allele variant found in patients with PFAPA was STAT4 (rs7574070). STAT4 is a transcription factor involved in T cell differentiation and Th17 activity; this variant presumably enhances signaling from IL-12, IL-23, and type-1 IFN stimulation in CD4⁺ T cells. However, IL-10, an antiinflammatory cytokine, is underexpressed in patients carrying the PFAPA risk variant IL10 (rs1518110), leading to impaired immune suppression and excessive inflammation. Another significant genetic marker, CCR1-CCR3 (rs7616215)—involved in the migration of monocytes and neutrophils—has also been identified in patients with PFAPA. The presence of these variants reinforces the hypothesis that both innate and adaptive immune dysfunction contribute to PFAPA pathogenesis [19].

Interestingly, the human leukocyte antigen (HLA) risk factors associated with PFAPA are distinct from those associated with BD, despite their overlapping genetic background. For instance, while HLA-B*51 is a well-established risk allele for BD, it is not significantly associated with PFAPA. Instead, HLA-DQB106:03, HLA-DRB113:01, and HLA-DQA1*01:03 have been identified as key risk factors for PFAPA. These findings suggest that while PFAPA, BD, and RAS share core inflammatory pathways, HLA differences may determine their distinct clinical manifestation explaining why some patients with PFAPA may develop atypical symptoms, such as genital ulcers, or why patients with BD report episodes similar to PFAPA in childhood, before developing the complete clinical presentation of the disease later in life [25]. Given these genetic and clinical overlaps, a group of researchers has proposed grouping PFAPA, BD, and RAS under a new category: Behçet’s spectrum disorders (BSDs) [19]. This framework positions PFAPA as an intermediate phenotype, with RAS representing the mildest form and BD being the most severe.

Beyond genetic background, a recent comparative study by Gürel Bedir and Kilic [26] explored the clinical features in pediatric patients with PFAPA syndrome and BD, raising the question of whether PFAPA could represent an early or incomplete form of BD. The study included 150 children (60 with PFAPA, 30 with BD, and 60 healthy controls) and found that although the two diseases share overlapping symptoms—most notably recurrent fever and oral ulcers—they remain distinct clinical entities. PFAPA was diagnosed at a significantly younger age and more frequently associated with cervical lymphadenopathy and a family history of periodic fever, while BD was characterized by multisystem involvement, genital ulcers, ocular inflammation, and joint symptoms. During flares, patients with PFAPA showed higher levels of inflammatory markers. Interestingly, colchicine appeared effective in reducing attack frequency in both groups, with reported benefits in over 85% of cases [26].

So far, genome-wide association studies (GWAS) have only been conducted in BD, identifying risk variants near or within loci such as IL10, IL23R-IL12RB2, CCR1-CCR3, STAT4, and IL12A [19]. Performing a GWAS in PFAPA could provide valuable insights into its genetic architecture, allowing comparisons of HLA and non-HLA susceptibility loci, and help to identify genetic factors that may differentiate PFAPA from other disorders within the BSDs.

Role of Rare Variants of Unknown Significance as Emerging Genetic Markers

Emerging studies have identified rare genetic variants that may predispose individuals to more severe or familial forms of PFAPA. ALPK1 is a cytoplasmic kinase involved in innate immune responses. Mutations in ALPK1 have been found in ROSAH syndrome. A study on familial cases of PFAPA [27] identified one rare missense variant in ALPK1, (c.2770T>C, p.S924P) that co-segregated with the disease, following an autosomal dominant pattern. The same authors found two additional heterozygous ALPK1 variants in patients with sporadic PFAPA. Notably, one of them presented with ROSAH syndrome clinical phenotype. In addition, some patients with PFAPA carry Q703K variants in NLRP3 gene [28], which has been already described in patients with cryopyrin-associated periodic syndromes (CAPS) and is implicated in excessive IL-1β production. However, although the NLRP3 Q703K variant is more prevalent in patients with mild CAPS and PFAPA, it was found at similar frequencies in patients with autoinflammatory diseases and healthy individuals, suggesting that it is unlikely to be a primary disease-causing mutation. Consistently, Cosson et al. [29] showed that Q703K is a variant of unknown significance (VUS) that exclusively increases pyroptosis through the NLRP3 pathway, often leading to misdiagnosis. A low-penetrance missense mutation, R92Q, in TNFRSF1A has been observed in children with periodic fever syndromes, including PFAPA [30]. Unlike structural TNFRSF1A mutations, R92Q does not significantly alter TNFR1 conformation and function, and is associated with a milder disease course. A comparative study of children with TRAPS-associated TNFRSF1A mutations, PFAPA, and R92Q-positive patients found that R92Q carriers had shorter fever episodes and a higher rate of spontaneous resolution. In fact, the clinical presentation of R92Q carriers resembled PFAPA more than TRAPS, suggesting that R92Q may act as a genetic modifier rather than a causative mutation​. Interestingly, long-term follow-up studies indicate that many R92Q carriers experience disease remission over time, whereas those with TRAPS mutations often develop progressive, chronic inflammation. These findings suggest that TNFRSF1A R92Q may contribute to an intermediate phenotype, modulating the inflammatory response rather than directly causing PFAPA​. A study by Cheung et al. [21] has identified a genetic association between PFAPA syndrome and a frameshift variant in the CARD8 gene (CARD8-FS). This mutation results in a truncated protein lacking the FIIND and CARD domains, which are essential for binding the NLRP3 inflammasome. The study analyzed 82 unrelated patients with PFAPA and identified a CARD8-FS variant in 13.9% of them, compared with only 3.2% in healthy controls, suggesting a strong genetic association between this mutation and PFAPA​. Functional studies revealed that this defective CARD8 protein fails to interact with NLRP3, leading to increased inflammasome activity and excessive IL-1β production​. Patients carrying the CARD8-FS variant exhibited a distinct PFAPA phenotype, characterized by more severe and frequent symptoms outside of febrile episodes, higher prevalence of aphthous stomatitis, and stronger familial clustering, with many carriers reporting a family history of recurrent fever or pharyngitis [21]. However, Manthiram et al. [19] did not find a significant association between the CARD8 variant and patients with PFAPA.

Role of Genetic Testing in the Diagnostic Process

Currently, no genetic test can definitively confirm a diagnosis of typical PFAPA syndrome, which remains a clinical diagnosis. Furthermore, the identification of a variant of unknown significance (VUS) does not necessarily influence treatment decisions. However, genetic testing may be warranted in cases of atypical PFAPA presentations or when there is clinical suspicion of an alternative autoinflammatory disease, such as FMF, mevalonate kinase deficiency (MKD), or A20 haploinsufficiency (HA20) ​[31].

Environmental Factors and the Role of the Oral and Gut Microbiome

Although PFAPA does not appear to have a direct infectious cause, environmental factors may act as triggers for immune activation. The oral microbiome has been suggested as a potential contributing factor, given its possible role in the pathogenesis of PFAPA, particularly considering that flares frequently involve aphthous stomatitis and pharyngitis [32]. A study by Tejesvi et al.[33] found significant differences in the composition and relative abundance of specific taxa in the tonsillar tissue of patients with PFAPA. Notably, Cyanobacteria, known for their ability to produce inflammatory microcystins, were significantly more prevalent in patients with PFAPA. Conversely, Streptococcus spp., commonly found in healthy nasopharyngeal flora, were less abundant in PFAPA tonsils. Moreover, children with PFAPA display a distinct proinflammatory immune profile in the tonsils, even in the absence of clinical symptoms, such as elevated numbers of CD8⁺ T cells and memory B cells, as well as increased expression of IL-1β and TNF, suggesting a baseline immune activation and dysregulation of innate immune responses in the palatine tonsils​ [14]. Dysbiosis in the tonsillar microbiota could lead to persistent immune activation, possibly triggering inflammasome activity and Th1-driven inflammation [14]. Moreover, no specific pathogen has been consistently linked to PFAPA, reinforcing the idea that it is primarily an immune-mediated rather than an infectious disease.

A retrospective study of 150 children with PFAPA [34] found that those who were breastfed for more than 6 months had a significantly higher rate of spontaneous resolution of symptoms compared with children who were breastfed for a shorter duration. Breastfeeding is known to promote the colonization of beneficial bacteria, such as Bifidobacteria and Lactobacilli, which play a key role in regulating inflammation and immune activation. This aligns with the hypothesis that tonsillar and gut microbiome alterations may act as a trigger for PFAPA​.

In addition, the study found that maternal education level was independently associated with PFAPA resolution, possibly reflecting differences in dietary practices, healthcare access, and overall early life immune exposures​ [34].

Vitamin D Deficiency

Another emerging risk factor for PFAPA is vitamin D deficiency, which has been widely implicated in immune-mediated diseases. Mahamid et al. [35] were the first to report that vitamin deficiency (< 20 ng/ml) is a significant risk factor for PFAPA occurrence. A systematic review by Fahdy et al. [36] found that vitamin D levels were significantly lower in patients with PFAPA compared with healthy controls. Oner et al. [37] reported that 66.2% of their cohort of patients with PFAPA had vitamin D deficiency (< 50 nmol/L), while 33.8% had vitamin D insufficiency. Moreover, patients with lower vitamin D levels experienced more frequent and longer PFAPA attacks, and higher CRP values, suggesting a potential role in disease severity and inflammation regulation [38].

Clinical Presentation and Diagnostic Criteria

The diagnosis of PFAPA is primarily based on clinical characteristics (history and physical examination), and there is no definitive diagnostic test available. PFAPA episodes typically last between 3 and 7 days and recur every 2–8 weeks, with a common interval of 3–6 weeks. In approximately 60% of cases, patients experience prodromes such as fatigue prior to the onset of fever. A key clinical feature is the clockwork regularity of fever episodes; as children grow older, the episodes tend to become less severe and more widely spaced [39]. Unlike upper respiratory tract infections, which are more common in winter, PFAPA episodes do not follow a seasonal pattern, although some patients may notice fewer episodes during the summer months [2]. During a typical flare, the key clinical features include high fever (between 39 °C and 40.5 °C, usually resistant to antibiotics and antipyretics), erythematous or exudative pharyngitis (> 90% of cases), cervical adenitis (up to 75%), and oral aphthosis (50%) [39]. The aphthous lesions are generally less than 1 cm in diameter, painful, shallow, round, and display well-defined red margins on non-masticatory areas of the oral cavity; small clusters of these ulcers may also be seen [40]. The cervical lymph nodes are usually enlarged bilaterally, moderately tender, and measure less than 5 cm in diameter [39]. Some patients may also experience other symptoms during attacks, including abdominal pain, arthralgia/arthritis, headache, rash, diarrhea, nausea, or vomiting [41].

An overview of the proposed classification criteria for PFAPA is provided in Table 1. Traditionally, the modified Marshall’s criteria [42], proposed in 1999, have been widely used in clinical practice owing to their high sensitivity. Despite their widespread use, these criteria do not account for cases of late-onset PFAPA, nor do they fully differentiate PFAPA from other recurrent fever syndromes. Given the limitations of the modified Marshall’s criteria, Vanoni et al. [43] and the Eurofever/ Paediatric Rheumatology International Trials Organization (PRINTO) consortium proposed, in 2018, new classification criteria aimed at improving diagnostic specificity. However, their application in clinical practice showed that they are too restrictive, excluding nearly 50% of previously diagnosed patients with PFAPA. Therefore, while they may be valuable for research purposes, they may not be ideal for routine clinical diagnosis. The Cantarini et al. criteria specifically address late-onset PFAPA, defining cases occurring in patients ≥ 16 years of age [4]. The Eurofever/PRINTO classification criteria [2] further improved the identification of PFAPA by incorporating both inclusion and exclusion criteria. Unlike previous criteria, the revised criteria do not impose an age threshold and do not consider oral aphthosis and normal growth and development as a mandatory feature. This approach has high sensitivity (93.4%) and specificity (91.7%), making it useful for identifying children and adult PFAPA cases.

Table 1.

PFAPA classification criteria

Criteria	Description	Sensitivity*	Specificity*
Modified Marshall’s (1999) [42]	- Periodic fever episodes (onset typically < 5 years old) - Pharyngitis - Aphthous stomatitis - Cervical adenitis - Symptom-free intervals between attacks	Moderate	Low
Cantarini et al. (2017) (Late-onset PFAPA) [4]	- Onset at ≥ 16 years - Recurrent fever episodes - Pharyngitis and cervical adenitis (may or may not include aphthous stomatitis) - Absence of other periodic fever syndromes	High	High
Vanoni et al. (2018) (Eurofever/PRINTO) [43]	1. Periodic fever for at least 6 months: a. Fever ≥ 38.5 °C (axillary) for 2–7 days b. At least five regularly recurring fever episodes with a maximum interval of 2 months between them 2. Pharyngitis, cervical adenitis, or oral aphthae: at least one in every episode and at least two out of three in the majority of episodes 3. Exclusion of other causes of recurrent fever (clinical or laboratory evaluation) 4. Exclusion of infections, immunodeficiency, and cyclic neutropenia 5. Disease onset before 6 years of age 6. Complete resolution between episodes 7. Normal linear growth	High	High
EUROFEVER/PRINTO (2019) (revised criteria) [2]	At least 7 of the following 8 criteria: - Pharyngotonsillitis - Fever lasting 3–6 days - Cervical lymphadenitis - Clockwise periodicity of fever episodes - No diarrhea - No chest pain - No skin rash - No arthritis	0.97	0.93

^*Sensitivity and specificity are classified as high (≥ 90%), moderate (70–89%), or low (< 70%) on the basis of their ability to correctly identify or exclude PFAPA cases in validation studies

Beyond formal criteria, additional clinical clues can help differentiate PFAPA from infectious or genetic autoinflammatory conditions, such as:

• A dramatic resolution of fever following a short course (1–2 doses) of corticosteroids;

• Clockwork periodicity of fever episodes;

• A family history of recurrent pharyngitis and tonsillectomy;

• Overall well-being between episodes;

• Overall normal growth in height and weight.

Laboratory Findings

Laboratory findings during febrile episodes typically show elevated CRP and erythrocyte sedimentation rate (ESR) [44] but no significant increase in procalcitonin, distinguishing it from bacterial infections [17]. The increased levels of S100 proteins (S100A8/9 and S100A12) [17] indicate activation of both innate and adaptive immune responses. Between attacks, inflammatory markers normalize, which further supports the diagnosis of PFAPA.

A study examining immune cell profiles in patients with PFAPA [7] have revealed that during flares, patients exhibited marked leukocytosis, characterized by neutrophilia and monocytosis, along with a relative lymphopenia and eosinopenia. These changes were statistically significant when compared both to asymptomatic intervals in patients with PFAPA and to individuals experiencing non-PFAPA febrile episodes. Further immunophenotyping of lymphocyte populations showed a notable reduction in T lymphocytes during PFAPA flares, affecting both CD4⁺ and CD8⁺ subsets. Among these, activated CD4⁺ T cells (HLA-DR⁺ and CD25⁺ T cells) were significantly decreased, whereas activated CD8⁺ T cell levels remained unchanged. No significant fluctuations were observed in B cell, natural killer (NK), or natural killer T (NKT) cell counts. Taken together, these findings indicate that PFAPA flares are associated with complex and dynamic perturbations not only in neutrophils but also in monocytes, eosinophils, and specific T lymphocyte subsets, suggesting broad involvement of both innate and adaptive immunity during disease activity.

Differential Diagnosis

Monogenic Autoinflammatory Diseases

One of the main challenges in diagnosing PFAPA is differentiating it from other monogenic autoinflammatory diseases, such as FMF, MKD, tumor necrosis factor receptor-associated periodic syndrome (TRAPS), and cryopyrin-associated periodic syndromes (CAPS)​​. While all these conditions share recurrent fever as a central symptom, their episode duration, accompanying symptoms, and genetic markers help distinguish them.

FMF typically presents with shorter fever episodes (1–3 days), often accompanied by severe abdominal pain, arthritis, and pleuritis. Unlike PFAPA, FMF attacks can be triggered by stress or exertion. Genetic testing for MEFV mutations finally confirms the diagnosis​ [23]. MKD is characterized by longer fever episodes (4–15 days), usually associated with diarrhea, rash, and hepatosplenomegaly, which are not seen in PFAPA. Genetic testing for MVK, as well as urinary mevalonic acid can support the diagnosis [2]. TRAPS features prolonged fever episodes (7–21 days), often accompanied by severe myalgia, periorbital swelling, and migratory skin rashes. Genetic testing for TNFRSF1A mutations may confirm the diagnosis [2]​. CAPS, unlike PFAPA, presents with persistent low-grade inflammation and chronic symptoms rather than distinct fever episodes. Patients may have urticarial-like rash, neurological symptoms, and progressive hearing loss, making CAPS clinically distinct​. Genetic testing for NLRP3 mutations may confirm the diagnosis [2]. HA20 is characterized by inconstant and not strictly periodic fever attacks; oral, genital, or gastrointestinal ulcerations; and frequent systemic involvement (arthritis/arthralgias, skin rashes, uveitis, neurologic involvement) [45].

The clinical response to corticosteroids during flares of monogenic autoinflammatory diseases such as MKD can sometimes mimic PFAPA and complicate the differential diagnosis. However, while PFAPA typically shows a rapid and dramatic resolution of symptoms following a single dose of corticosteroids, patients with MKD and other hereditary fever syndromes often require a short course of corticosteroids to achieve similar control of the episode. According to Soriano et al. [46] corticosteroids provide partial benefit in approximately 73% of MKD cases, but complete resolution is uncommon, highlighting a key therapeutic distinction between PFAPA and monogenic periodic fever syndromes.

In addition to the previously described periodicity, other key features distinguishing PFAPA from monogenic autoinflammatory diseases include the complete clinical and biological resolution between attacks, the aspect of oral aphthae (small, barely visible), and the normal growth and psychomotor development observed in affected children.

SURF

In the differential diagnosis of PFAPA syndrome, increasing attention has been given to the syndrome of undifferentiated recurrent fever (SURF), which includes patients with recurrent inflammatory episodes who do not meet criteria for PFAPA or monogenic periodic fever syndromes, and whose genetic testing is inconclusive. Although both PFAPA and SURF share features such as periodic fever, tonsillar involvement, and systemic symptoms, several key clinical and biochemical differences have been identified [47]. Clinically, patients with SURF often exhibit more heterogeneous and multi-system symptoms—including abdominal pain, myalgia, and fatigue—and are less likely to present with the classic PFAPA triad of aphthous stomatitis, pharyngitis, and cervical adenitis. In addition, SURF tends to persist beyond the typical pediatric age range of PFAPA and is often less responsive to corticosteroids or tonsillectomy. Moreover, while colchicine may partially relieve symptoms in PFAPA, patients with SURF often exhibit a higher prevalence of complete response [48].

From a biochemical standpoint, a recent study by Palmeri et al. [48] showed that pyrin inflammasome activation patterns in SURF differ significantly from those in FMF and PFAPA. Specifically, monocytes from patients who are colchicine-naïve and with active SURF exhibited pyrin activation in response to Clostridium difficile toxin A (TcdA), which normalized after colchicine treatment. In contrast, patients with PFAPA exhibited minimal inflammasome activation under similar conditions, suggesting limited involvement of the pyrin pathway in PFAPA pathogenesis. These findings support the hypothesis that SURF is a biologically and clinically distinct autoinflammatory condition and underscore the importance of functional assays and detailed clinical phenotyping in evaluating patients with atypical or refractory PFAPA-like presentations.

Infections

Because PFAPA primarily affects young children and presents with recurrent fever and pharyngitis, it is frequently misdiagnosed as recurrent streptococcal pharyngitis, leading to unnecessary antibiotic treatments​. Several features help differentiate the two conditions; streptococcal pharyngitis typically presents with exudative tonsillitis, fever, and positive throat cultures for Group A Streptococcus, whereas patients with PFAPA consistently test negative for bacterial infections. Moreover, PFAPA follows a predictable fever pattern, whereas streptococcal infections occur sporadically and often coincide with sick contacts or seasonal outbreaks​. Unlike bacterial infections, PFAPA episodes resolve dramatically with corticosteroids, whereas streptococcal pharyngitis requires antibiotic treatment ​[39].

Cyclic Neutropenia

Another important condition to consider is cyclic neutropenia, a hematological disorder that also presents with periodic fevers. However, cyclic neutropenia is characterized by severe neutropenia (< 500/µL), which is absent in PFAPA, regular neutrophil count fluctuations every 21 days, leading to cycles of deep mucosal ulceration, increased infection risk, and a tendency for bacterial infections (e.g., gingivitis, cellulitis, otitis media​). Genetic testing for mutations in the ELANE gene can confirm the diagnosis [49].

Treatment Strategies for PFAPA Syndrome

The primary goal of PFAPA treatment is to mitigate inflammation by controlling acute febrile episodes and preventing recurrences or, at least, reducing the severity and frequency of attacks, thereby improving the child’s quality of life. The therapeutic approach should be tailored to the evolving characteristics of the disease and the patient’s individual clinical and social context.

Treatment of Acute Attacks

Antipyretics and Nonsteroidal Antiinflammatory Drugs (NSAIDs)

Antipyretics, such as paracetamol, may reduce fever during PFAPA attacks, but their effectiveness is partial and variable [9]. Studies indicate that NSAIDs, particularly indomethacin and ibuprofen, are slightly more effective than paracetamol in managing symptoms​ [50, 51]. However, despite their widespread use, antipyretics and NSAIDs do not prevent PFAPA attacks and do not alter their natural course.​ They may be considered for patients affected by mild PFAPA cases where corticosteroid therapy is not warranted.

Corticosteroids

Corticosteroids, particularly prednisolone (1–2 mg/kg) or betamethasone (0.1–0.2 mg/kg, single dose), are the most effective treatment for aborting acute PFAPA attacks, with 85–95% of patients experiencing rapid fever resolution within hours [50, 52]. Other symptoms can take longer to resolve [9, 53]. Some authors have reported efficacy with lower doses of prednisolone (0.5 mg/kg) [54]. In cases where a single corticosteroid dose does not resolve symptoms, a second dose may be administered the following day [9]. Manthiram et al. [55], in a survey conducted among CARRA and Pediatric Infectious Disease Society (PIDS) practitioners, observed that 72% of physicians reported feeling comfortable with administering ≥ 2 mg/kg/month of prednisolone before switching to alternative treatments, while the remaining physicians preferred to stay within ≤ 1 mg/kg/month.

So far, corticosteroids are used differently across clinical settings, with some clinicians prescribing steroids for every episode to minimize fever burden and some reserving them for episodes occurring at inconvenient times (e.g., travel, school exams). In some cases, corticosteroids are used as a diagnostic tool, given that PFAPA flares respond dramatically to steroid treatment [39]. However, it is well known that while corticosteroids abort febrile episodes, they are not effective in preventing attacks and may increase attack frequency in 25–50% of cases [50, 52, 55]. Moreover, although the on-demand use of single-dose corticosteroids is considered safe, symptoms such as restlessness, mood change, or sleep disruption are noted in several studies [39, 53, 56], and so far, no well-designed longitudinal studies have systematically investigated the possible long-term corticosteroid-related adverse effects following intermittent administration.

While corticosteroids are widely recognized as effective for PFAPA flares in children, their efficacy in adult-onset PFAPA appears to be more variable. Cattalini et al. [57] reported that, although 85% of adults treated with prednisone (50–60 mg/day) achieved complete remission, 12% had a partial response and 3% did not respond at all. This variability suggests potential differences in disease expression or immune regulation between pediatric and adult patients with PFAPA.

Given the lack of evidence that corticosteroids modify the long-term course of the disease and the risk of shortened intervals between flares, in our daily clinical practice we prefer to reserve their use for exceptional circumstances, such as when episodes occur at particularly disruptive times (e.g., exams, travel). In most cases, we manage flares with NSAIDs and consider preventive therapy in patients with very frequent episodes. However, the CARRA consensus treatment plans [9] suggest treating flares with a starting dose of 1 mg/kg of prednisone or prednisolone, with the possibility of increasing to 2 mg/kg in cases of inadequate response (flares occurring between 14 and 21 days). If flares are very frequent (≤ 14 days), a switch to another treatment arm is recommended.

Anti-IL-1

IL-1β inhibition has been proposed as a targeted therapy since it reduces Th1-driven inflammation and cytokine overproduction​ [58]. Few studies report a prompt clinical response in patients receiving treatment with anti-IL-1 biologics (anakinra or canakinumab), with fever and inflammatory symptoms ceasing within hours of the injection. Cantarini et al. [59] reported a case of adult-onset PFAPA refractory to conventional treatments successfully treated with anakinra 100 mg/day, achieving complete symptom resolution, inflammatory marker normalization, and no relapses at 6-month follow-up. Stojanovic et al. [7] showed that five children receiving a single dose (1–2 mg/kg/day) of anakinra at the onset of a PFAPA attack experienced a prompt fever resolution​ within hours. Two patients had relapses within 1–2 days, requiring a second dose. The short-term follow-up (only immediate response measured) and the absence of control group limit conclusions on long-term efficacy compared with spontaneous remission. Lopalco et al. [60] reported a case of adult-onset PFAPA patient refractory to conventional treatment and anakinra experiencing a complete symptom resolution and long-lasting remission (at 14-month follow-up) after treatment with canakinumab 150 mg every 8 weeks. Soylu et al.​ [58] reported the case of a 2-month-old patient with PFAPA carrying a heterozygous M694V mutation in MEFV refractory to colchicine who significantly improved under canakinumab (2 mg/kg every 8 weeks). It is worth mentioning that tonsillectomy was finally resolutive in this patient.

To the best of our knowledge, there are no published reports on the use of rilonacept in patients with PFAPA.

Although adverse effects of IL-1 inhibitors—such as infections, injection site reactions, mild immunosuppression, hematologic abnormalities, and hepatotoxicity—have been described in literature, none of the previously cited studies observed any significant side effects. These findings suggest a favorable safety profile in the short term, although larger studies and long-term follow-up are needed to confirm tolerability in patients with PFAPA. Owing to the absence of randomized clinical trials and to the limited evidence available in children, there is no established consensus regarding the use of anti-IL-1 drugs in PFAPA. Given their high cost, these therapies should be reserved for patients who experience persistent breakthrough episodes and are unable to tolerate alternative treatment options.

Table 2 summarizes the principal therapeutic strategies for acute attacks in PFAPA, including pharmacological classes, dosing approaches, reported efficacy, and adverse effects.

Table 2.

Treatment of acute attacks in patients with PFAPA

Drug class		Active ingredients	Mechanism of action	Dosage		Route of administration	Efficacy			Adverse effects
Corticosteroids		Prednisolone Betamethasone	Immune modulation, inflammation suppression	1–2 mg/kg; 0.1–0.2 mg/kg (single dose); second dose if needed		Oral	Rapid fever resolution (85–95% of cases)			Mood changes, sleep disturbances, increased attack frequency (25–50% of cases)
Antipyretics		Paracetamol	Fever reduction	15 mg/kg/dose; repeat after 4-6 h if needed		Oral	Partial and inconsistent symptom relief			Minimal with short-term use
NSAIDs		Ibuprofen Indomethacin	COX inhibition	10-15 mg/kg/dose; variable		Oral	More effective than paracetamol but does not prevent attacks			Gastrotoxicity, nephrotoxicity
Anti-IL-1 biologics		Anakinra Canakinumab	IL-1 inhibition	1–2 mg/kg (second dose may be needed); 2 mg/kg every 8 weeks		Subcutaneous	Rapid fever resolution in refractory cases			Injection site reactions, risk of infections

NSAIDs nonsteroidal antiinflammatory drugs, COX cyclooxygenase, IL-1 interleukin-1

Preventive Therapy

Colchicine

Colchicine has emerged as a potential prophylactic treatment for PFAPA syndrome, particularly in patients experiencing frequent or severe attacks. Its proposed mechanisms in PFAPA include the inhibition of IL-1β activation via inflammasome modulation, the stabilization of neutrophil activity, and microtubule disruption, reducing the monocyte and neutrophil migration to inflamed tissues and preventing excessive immune responses. However, its exact role in controlling the symptoms remains under investigation [39, 61]. Butbul Aviel et al. [62] conducted a randomized controlled trial showing that the colchicine group experienced a significant reduction in attack frequency compared with patients without colchicine treatment. Welzer et al. [63] evaluated the effectiveness and safety of colchicine (0.5–2 mg/day) in 27 pediatric patients with genetically negative PFAPA syndrome and moderate-to-high disease activity. The primary outcome, defined as a reduction of ≥ 2 points on both physician and parent global assessment scales, was achieved in 63% of patients at last follow-up (13 months). In addition, 52% reported complete resolution of flares, while the majority of the remaining patients experienced a reduction in attack duration and frequency. According to Dusser et al. [61] patients with MEFV gene mutations tended to respond better. Among 20 patients treated, 45% had a favorable response, defined as either complete resolution or a > 50% reduction in attack frequency. In addition, Özaslan et al. [64] found that the presence of the MEFV M694V mutation correlated with colchicine responsiveness, whereas patients presenting with pharyngitis and arthralgia, as well as those experiencing higher attack frequency prior to treatment, were more likely to be unresponsive to colchicine. In the cohort study by Otar Yener et al. [65], 46.7% of patients receiving colchicine achieved complete or near-complete resolution, but in contrast with what was reported by other studies, colchicine efficacy was not significantly influenced by the presence of MEFV mutations.

Overall, only Welzer et al. [63] reported side effects of colchicine such as mild-to-moderate gastrointestinal symptoms (abdominal pain and diarrhea) at doses ≥ 1 mg/day. A few patients developed mild asymptomatic elevations in liver enzymes or leukopenia, but no severe adverse events were reported and no patients discontinued treatment owing to side effects. In contrast, the other previously mentioned studies did not mention any adverse events, focusing instead on colchicine’s efficacy or predictive factors of response.

To date, colchicine could be used for symptom control in patients with frequent or severe attacks of PFAPA, to prevent overuse of corticosteroids, and/or in patients carrying MEFV mutations, as they may have a higher likelihood of responding. More studies on a larger scale are needed to confirm the long-term efficacy of colchicine, clarify its role as a preventive treatment in PFAPA, and determine the optimal duration of therapy.

Histamine Receptors Antagonists

Cimetidine, a H2-histamine receptor antagonist, has been investigated as a prophylactic treatment option in PFAPA syndrome owing to its potential immunomodulatory effects. Although the exact mechanism remains unclear, it is hypothesized that cimetidine may modulate T cell activity and reduce proinflammatory cytokine release. Several studies have reported variable efficacy. The CARRA Consensus Treatment Plans [9] identified cimetidine (20–40 mg/day) as one of the four major treatment arms for PFAPA and recommended its use particularly in patients with frequent episodes or parental reluctance toward steroids or surgery​. Although large-scale controlled trials are lacking, retrospective studies suggest that daily cimetidine may lead to partial or complete remission in approximately 25–50% of patients, especially those with milder disease courses​. In their review, Gaggiano et al. [66] concluded that while cimetidine may be considered in milder forms of PFAPA or in patients in whom corticosteroids or surgery are not appropriate, its efficacy is clearly inferior to tonsillectomy and corticosteroids, and further validation is needed.

The 2021 review by Wang et al. [3] also highlighted that cimetidine may be especially useful in children unresponsive to tonsillectomy or not suitable for surgery, although it is generally considered less effective than corticosteroids or surgical intervention. As such, cimetidine represents a low-risk, noninvasive option in selected patients with PFAPA; however, further studies are needed to elucidate its mechanisms of action and effectiveness.

Ketotifen is an H1-histamine receptor antagonist that may exert its effects in PFAPA through mast cell stabilization, leading to reduced histamine and cytokine release, while also contributing to the regulation of the Th1/Th2 balance. Kapustova et al. [67], in their prospective study, treated 111 children with PFAPA with ketotifen (0.08 ± 0.01 mg/kg/day), observing a significant increase in the attack-free interval, with mild side effects being reported in four patients (restlessness, irritability, agitation, and constipation). However, the study lacked a control group, and patients received multiple concomitant treatments, which did not allow for drawing definitive conclusions.

Regarding sides effects, although the known side effects of cimetidine—headache, dizziness, gastrointestinal discomfort, and, more rarely, gynecomastia or liver enzyme elevation—are well documented in broader clinical use, none of the reviewed PFAPA studies have reported adverse events related to cimetidine in their patient populations. Kapustova et al. [67] reported mild side effects in 3.6% of patients with PFAPA receiving ketotifen, including restlessness, irritability, agitation, and constipation. These effects did not require permanent treatment discontinuation, and symptoms improved either with dose reduction or cessation of concomitant medications. No serious adverse events were observed, supporting the overall safety of ketotifen in this setting.

Probiotics

Probiotics are defined as “live microorganisms which when administered in adequate amounts confer a health benefit on the host” by the World Health Organization (WHO) [68].

Emerging evidence suggests that Streptococcus salivarius K12 (SSK12), an oropharyngeal probiotic, may help prevent PFAPA flares. It produces bacteriocins that inhibit the growth of proinflammatory bacteria, such as Streptococcus pyogenes, Streptococcus pneumoniae, and Moraxella catarrhalis, considered as key triggers of oropharyngeal inflammation. Moreover, it is described that SSK12 modulates cytokines responses, by reducing plasma IL-8 levels, increasing salivary interferon-γ production [69, 70]. A multicenter study involving patients with PFAPA treated with SSK12 found a 50% reduction in febrile attacks per year, a shortened duration of fever episodes, and a steroid-sparing effect, with annual corticosteroid use decreasing significantly [69].

In another retrospective cohort study [68], 20 pediatric Turkish patients received a probiotic preparation containing Lactobacillus plantarum HEAL9 and Lactobacillus paracasei 8700:2, alongside low doses of vitamin D3, B12, and folic acid. In total, 95% of patients experienced a reduction in attack frequency, with the median number of episodes dropping from three to one over a 3-month period. Notably, 40% had no attacks at all during this time, and nearly half of those who still experienced flares reported milder symptoms. These findings suggest that probiotics may exert beneficial effects in PFAPA through their antiinflammatory properties or by modulating the tonsillar and gut microbiota, both of which have been implicated in the disease’s pathophysiology. Further studies are required to confirm their long-term efficacy and their role as a standalone or adjunct preventive therapy.

Vitamin D

Vitamin D is increasingly recognized as a modulator of immune function, particularly in T cell function and cytokine modulation. Several immunomodulatory patterns, including the balance of Th1–Th2 immune responses, the suppression of proinflammatory cytokines, and the enhancement of T-reg function, have been proposed for Vitamin D treatment in PFAPA, potentially reducing flare frequency and severity [36]. Stagi et al. [38] conducted a prospective study in which patients with PFAPA received 400 IU/day of vitamin D for 7 months. After supplementation, fever episodes decreased in duration and frequency. However, the study did not report on concurrent PFAPA treatments, and the small sample size (25 patients) limits generalizability. Oner et al. [37] reported that vitamin D supplementation (2000 IU/day) for 2–4 months resulted in a statistically significant reduction in attack frequency and duration in patients with vitamin D deficiency (< 50 nmol/L). The cutoff value for significant attack reduction was determined to be ≥ 29.7 nmol/L of vitamin D. However, 70.8% of patients were also receiving colchicine, which could have independently contributed to the observed improvement; moreover, the relationship between treatment differences and vitamin D levels was not evaluated. ​

Vitamin D is generally considered safe and well tolerated when used within recommended doses, with known side effects in the general population being rare and typically limited to hypercalcemia, hypercalciuria, or gastrointestinal symptoms at high dosages. However, none of the PFAPA-focused studies [37, 38] reported any adverse effects in patients receiving vitamin D supplementation. This supports the conclusion that vitamin D supplementation is a safe intervention in this population, although broader studies with formal safety monitoring would help confirm these findings.

In conclusion, in patients with PFAPA, especially in those with documented vitamin D deficiency, supplementation may reduce attack frequency and severity. It is worth mentioning that so far there is no standardized dosage recommendation for vitamin D supplementation in PFAPA, and further high-quality clinical trials are needed to establish definitive treatment guidelines.

Tonsillectomy

Tonsillectomy, with or without adenoidectomy, remains one of the most effective long-term interventions for PFAPA, with multiple studies demonstrating symptom resolution or significant reduction in attack frequency [39]. The therapeutic effect is thought to reside in the removal of a potential source of chronic immune activation, leading to reduced systemic inflammation and symptom resolution [3]. A Cochrane Review by Burton et al. [71] reported complete resolution of PFAPA symptoms or a reduction in attack frequency in most patients after tonsillectomy. Rydeman et al. [72] reported a 90% rate of complete symptom resolution following tonsillectomy in children with PFAPA, with a consequent significant increase in health-related quality of life (HRQOL) scores post-surgery. However, no randomization was performed, and the control group consisted of healthy children rather than patients with PFAPA not undergoing surgery, thus limiting direct comparison. In a prospective long-term study, Licameli et al. [73] followed 102 pediatric patients who had undergone surgery, with a mean follow-up period of 43 months. Remarkably, 97% of patients experienced complete resolution of PFAPA symptoms immediately after surgery.

So far, two randomized controlled trials have provided evidence supporting tonsillectomy as an effective and often curative treatment for PFAPA. Renko et al. [74] found that tonsillectomy was curative in 100% of children in the surgery group, with no symptoms reported after 6 months after tonsillectomy compared with the 50 % resolution of control group. However, its main limitations were the inclusion of fever as the sole symptom, which may have compromised diagnostic accuracy, and the short duration of follow-up (only 12 months). Garavello et al. [75] reported a 63% complete resolution rate in the surgery group versus 5% in the control group, along with a significant reduction in episode frequency and corticosteroid use (50% versus 88%). The stringent PFAPA criteria and the longer follow-up (18-months) further strengthen their findings.

Supporting this evidence, Lantto et al. [76], in a retrospective study with a longer follow-up of 8.9 years, reported complete symptom resolution in 97% of children fulfilling PFAPA criteria and 100% of those with atypical presentations (fever-only, symptom onset after age 5 years) following tonsillectomy. This suggests that tonsillectomy may be effective even in older children (> 5 years) or in those presenting fever as the only symptom. However, the retrospective design and the lack of a control group may have affected the reliability of the results.

A recent, retrospective, cross-reactional study investigating the long-term outcomes of patients with PFAPA following tonsillectomy [77] reported a complete remission in 63% of patients after a median follow-up period of 8 years, increased intervals between flares in 17%, and persistent non-febrile PFAPA-related symptoms in 20%, with recurrent aphthous stomatitis being the most common residual symptom. This aligns with other previous studies reporting a recurrence of aphthous stomatitis or cervical lymphadenopathy without febrile episodes after tonsillectomy in patients with PFAPA [51, 78, 79].

Interestingly, a recent multicenter study by Manthiram et al. [80] investigated the long-term efficacy of tonsillectomy in children with PFAPA and proposed clinical predictors of treatment response. Among 97 patients followed for a median of 49 months, approximately half experienced complete resolution of flares following surgery, with a notably higher response rate observed in children presenting with typical PFAPA features. The strongest predictors of a favorable outcome included a rapid resolution of fever following one or two doses of corticosteroids, the presence of exudative pharyngitis, and the absence of rash or musculoskeletal symptoms. In children with incomplete response, flares were often milder, and certain symptoms such as pharyngitis, cervical adenitis, and vomiting decreased in frequency post-surgery.

In conclusion, tonsillectomy appears to be effective in resolving PFAPA symptoms in the majority of patients; however, owing to the self-limiting nature of PFAPA, the potential surgical risks and the risk of long-term symptom recurrence, the decision to proceed with surgery should be considered for refractory or severe patients. Further controlled studies are needed to determine patient selection criteria and the optimal timing for surgery.

Other preventive therapies, such as thalidomide, pidotimod, and montelukast [40, 81–83], have been suggested by some physicians; however, these results have not been consistently confirmed in current studies.

Table 3 summarizes the principal preventive therapeutic strategies for PFAPA, including pharmacological classes, dosing approaches, reported efficacy, and adverse effects.

Table 3.

Preventive therapy in patients with PFAPA

Drug class/intervention	Active ingredients/treatment		Mechanism of action	Dosage	Route of administration	Efficacy	Adverse effects
Colchicine	Colchicine		IL-1β inhibition, neutrophil modulation	0.5–1 mg/day	Oral	Reduces attack frequency (especially if MEFV mutation)	Gastrointestinal discomfort
Probiotics	SSK12 Lactobacillus plantarum et paracasei		Modulates cytokines, inhibits proinflammatory bacteria	Variable	Oral	Reduces attack frequency	None reported
Vitamin D	Vitamin D		Immunomodulation, cytokine balance	400–2000 IU/day	Oral	Decreases frequency and severity of attacks in deficient patients	Hypercalcemia
Histamine receptor antagonists	Ketotifen Cimetidine		Mast cell stabilization, H1 histamine receptor inhibition H1 histamine receptor inhibition	0.08 mg/kg/day 20–40/mg/kg/day	Oral Oral	Increases attack-free interval	Restlessness, irritability, constipation Headache, dizziness, gastrointestinal discomfort
Tonsillectomy	Surgical removal of tonsils		Eliminates source of immune activation	N/A	Surgical	Complete resolution in 63–100% of cases	Surgical risks, possible recurrence of aphthous stomatitis

IL-1β interleukin-βeta, MEFV Mediterranean FeVer gene, SSK12 Streptococcus salivarius K12

Consensus Treatment Plans

In 2020, the CARRA PFAPA Working Group developed consensus treatment plans (CTPs) for PFAPA [9] to establish standardized response criteria for assessing treatment effectiveness and to provide a framework for future comparative effectiveness studies.

The Working Group defined specific outcome measures. The primary outcome was fever resolution, categorized as complete response (no fever episodes over a 3-month period), partial response (a reduction in the total number of fever days within 3 months), and no response (no change or an increase in fever days during that period). It was agreed that a treatment regimen should be tried for three febrile episodes before determining its effectiveness and switched to an alternative approach if deemed ineffective by the physician and/or the family. Quality of life was assessed on the basis of the number of missed school days and a parent-reported global assessment using a visual analog scale (VAS). On the basis of both an extensive review of literature and a worldwide survey, they reached a consensus on four main treatment arms for PFAPA:

1. Antipyretic arm: the use of NSAIDs or paracetamol is intended for mild cases or families avoiding medications or surgery.

2. Abortive arm (corticosteroids): prednisone or prednisolone (1–2 mg/kg, max 60 mg) at fever onset is considered effective, but may shorten intervals between flares. The recommended starting dose is 1 mg/kg, with the option to increase to 2 mg/kg in cases of inadequate response. An interval of 21 days or more was considered an adequate response to corticosteroids, whereas frequent flares occurring every 14 days or less prompted a recommendation to switch to another treatment arm. For intervals between 14 and 21 days, an increase in the steroid dose (from 1 mg/kg to 2 mg/kg) was advised. If this adjustment led to an interval of ≥ 21 days, continuation of steroids at the higher dose was recommended, otherwise, a transition to an alternative treatment arm was suggested.

3. Prophylactic arm: colchicine (0.5–1 mg/day) or cimetidine (20–40 mg/kg/day) are intended for patients with frequent attacks to reduce flare frequency, but with limited evidence of efficacy; they could be chosen as a first choice or after a failure of another arm.

4. Surgical treatment (tonsillectomy): considered curative in many cases, though evidence is of moderate quality. Tonsillotomy was not included owing to lack of efficacy data.

IL-1 inhibitors and vitamin D were not included in the CTPs owing to limited and mainly anecdotal evidence, with data on IL-1 blockade restricted to small case series and uncontrolled studies.

Future Research The discovery of shared genetic pathways between PFAPA and BD has significant therapeutic implications. Given that IL-12 plays a central role in PFAPA pathogenesis, biologic agents targeting IL-12, such as ustekinumab (IL-12/IL-23 inhibitor), could be investigated as potential treatments for severe or refractory cases. In addition, immune-modulating drugs such as apremilast, which has proven effective in BD and RAS, may offer therapeutic benefits in patients with PFAPA who are unresponsive to standard therapies [19].

Conclusions

PFAPA syndrome is a heterogeneous autoinflammatory disease whose pathogenic mechanisms are not yet fully understood, making both diagnosis and treatment challenging. Although much progress has been made in the understanding of the genetic, immune, and environmental factors, additional research is needed to clarify exactly how these elements interact to drive PFAPA’s hallmark symptoms. Consequently, no single gold standard exists for diagnosing PFAPA, and clinical practice varies widely, ranging from supportive care to surgical options, without a clear consensus on the ideal approach. Given the generally self-limiting nature of PFAPA and its lack of significant long-term sequelae, it seems reasonable to optimize the use of corticoids and to reserve tonsillectomy for patients with a genuine otolaryngologic indication or for refractory cases. However, given the importance of quality of life and in accordance with ACR/EULAR treat-to-target principles, a shared decision-making process or a personalized case-by-case approach is strongly advised when selecting a therapeutic strategy.

Declarations

Funding

Open access funding provided by Université Paris-Saclay.

Conflicts of Interest:

The authors declare that they have no conflicts of interest.

Ethical Approval:

This article does not contain any studies with human participants or animals performed by the authors.

Authors’ Contributions

Federica Anselmi (F.A.) and Isabelle Kone-Paut (I.K.P.) designed the structure of the article; F.A. drafted the text; F.A., P.D. and I.K.P. approved the final version of the manuscript.

Consent to Participate/Publication

Not applicable.

Availability of Data/Code Availability

Not applicable.