Acute Rheumatic Fever (ARF) is a multisystemic autoimmune response to infection of the group A streptococcus (GAS). The incidence of ARF is low in high-income countries such as North America (1, 2) and Western Europe (3), but persists high in low-income countries (4, 5), and among marginalized sections of the society of high-income countries (6–8) and in the South Pacific (9–11), becoming an important cause of morbidity and mortality. Jones criteria for ARF diagnosis were first established in 1944, then revised in 1992 (12) and afterward in 2015 (13) by the American Heart Association (AHA), leading to a distinction between moderate to high-risk (HR) and low-risk (LR) populations (incidence cut-off <2 out 100,000 per year), to avoid underdiagnoses in LR populations and overdiagnoses in HR populations. Therefore, these new criteria incorporate three major changes: the above-mentioned risk stratification based on disease endemicity, inclusion of polyarthralgia or monoarthritis and monoarthralgia among major and minor criteria of HR population respectively, and subclinical carditis as a major manifestation in both LR and HR populations. Consequently, the knowledge of the incidence of ARF is an essential need for the application of the new diagnostic criteria in each country.

The incidence of ARF in Italy has been scarcely investigated and limited to the main regions of Northern and to a single region of Central Italy. We performed a multicenter retrospective study to describe the clinical picture of ARF and to estimate its incidence in Tuscany, a region of the Central Italy, focusing on the epidemiologic impact of the new diagnostic criteria for HR populations.

The childhood population (2–18 years) at risk of GAS infection was 3,300,928 children distributed from 2010 to 2019, as calculated from the data provided by ISTAT. A total of 125 patients was included. Among them, 96 patients were assigned to LR group (pre-2015 n = 35, post-2015 n = 61), while 29 children were assigned to HR group (pre-2015 n = 8, post-2015 n = 21). The main characteristics of the cohort are reported in Table 1. Age distribution showed a higher prevalence in children aged between 5 and 10 years and a mild preponderance in male patients (1.4:1). No significant differences were observed for age (HR median age = 9.5 95%CI = 7.4–10.2, LR median age = 9.4 95%CI = 8.1–9.7, P = 0.91) and gender (Chi squared with Yates'correction χ² 0.001, P = 0.98) between group HR and LR patients (Table 1).

The median range time between the streptococcal infection and the onset of ARF symptoms was 12 days (IQR 17). ASO and Anti-DNase B titres of both groups did not differ significantly (HR median ASO = 1,270 95%CI = 805–1,610, LR median ASO = 1,378 95%CI = 1130–1563, P = 0.58; HR median anti-DNase B = 1,385 95%CI = 738–2200, LR median anti-DNase B = 927 95%CI = 730–1218, P = 0.24) (Table 1).

Most of the cases were diagnosed during spring (42% of total cases) with the peak of incidence in March (18% of total cases) and a late peak in June (15% of total cases) (Figure 1). ARF cases presented with chorea showed a delayed peak in June.

Our results highlight that the annual incidence of ARF in Tuscany was above the cut-off (<2 out 100,000 per year), when calculated in a period of 10 years (2010-2019), with an increasing trend from 2012 to 2019. The mean incidence was over the limit for most of the years considered in the selected period. Interestingly, it remained above the threshold value either after or before the application of HR criteria leading to an increase of 30% on total ARF diagnoses. Therefore, Tuscany should be considered an HR geographic area and, according to the latest guidelines, the criteria for HR populations should be used in this region.

Investigations on the incidence of ARF along the Italian country are few and limited to the Northern Italy. Recently, Licciardi et al. (15) estimated that ARF incidence in children living in the Province of Turin (Northern Italy) in the period between the years 2007–2016 ranged from 3.8 to 12.7/100,000, when calculated using LR and HR criteria respectively. Moreover, the study concluded that the application of HR criteria led to a 20.7% increase in the ARF diagnosis. In the period between the years 2014 and 2016 the Province of Turin showed an ARF rate incidence higher than Lombardy in the Northern Italy (4.24/100,000) (16). Other investigations performed along the territory of the North of Italy were conducted in Emilia Romagna in the period between 2000 and 2014 (before the publications of the AHA guidelines), showing that the ARF annual incidence in children presented with and without carditis were 0.8–2.4/100,000 and 1.5–3.8/100,000/year, respectively (17). Recently, a retrospective analysis of the childhood population living in the province of Bologna (Emilia Romagna) in the period between 2012 and 2017 showed ARF annual incidence of above two out of 100,000 (18). The unique report about the ARF incidence in the Central Italy refers to the Breda's study which evidenced in the period 2002–2009 an ARF incidence of 4.1/100,000 (19). Comparing the above-mentioned Italian studies we observed the same overall ARF rate incidence with regional variations in the same periods analyzed. In particular, Tuscany represented the unique region in which the incidence, calculated with HR criteria, was under the limit in 2016, reaching value over the limit in the next years. This indicates that the use of HR criteria is mandatory to avoid ARF underdiagnoses. Indeed, the ARF incidence in 2017 would have changed from 6.04 to 4.23/100,000, with the loss of six cases out of 20 if the use of HR criteria had not been continued. In an attempt to explain the inter-regional differences in the ARF rate incidence, we focused on the strains of GAS-causing infections. GAS M types have been highlighted as an important determinant to clarify the distribution and development of ARF outbreaks. Genetic variation in the emm gene, which encodes the M protein, has been correlated with varying ARF incidence, particularly the occurrence of GAS serotype M18 isolates (20–22). The association of several outbreaks of ARF with the circulation of mucoid GAS strains in the community has been reported (23). Therefore, we might argue that the variability of ARF incidence observed in the same periods in our country could be related to the heterogeneity and resurgence of rheumatogenic strains and to the different distribution of M types, due to the mobility of GAS strains on a global scale (24).

Concerning the peak of incidence observed in our cohort in 2019, an increase in throat GAS infections untreated or not correctly treated could be the explanation. Recently Di Muzio et al. reported knowledge gaps of GAS pharyngitis diagnosis and low adherence rates to the treatment in our country, resulting in difficulties to reduce the incidence of ARF (25). This observation strongly indicates the need to improve the pediatrician's approach to diagnosis and management of GAS infection, as the decline in the incidence of ARF in developed countries in previous decades, has reduced the attention to either diagnosis or treatment of streptococcal pharyngitis.

It is known that first episodes of ARF can rarely occur in children younger than 5 years, the highest incidence being reported among children aged between 10 and 14 years, followed by those ones aged between 5 and 9 years (4, 6, 26, 27), with episodic occurrence in adolescence. Furthermore, our population showed equal distribution.

According to previous studies (15), ARF diagnosis in our cohort showed a typical seasonal distribution pattern, time between the triggered infection and the onset of clinical manifestations being after about 2–3 weeks, predominantly in March, except for chorea which developed after about 40 days from the infection, compatible with a greater latency on developing symptoms.

Arthritis was the most frequent major criterion found in our cohort. Interestingly, in LR group it presented as polyarthritis without being the more relevant, while it is always found in HR group, mainly as monoarthritis, proving to be the most inclusive sign for ARF diagnosis. This observation agrees with a previous study which underlies the clinical relevance of monoarthritis in high incidence population (28), confirming its value as major criteria only in HR population.

Clinical carditis and arthritis in our cohort showed a frequency similar to that reported in studies from the other regions of Northern Italy (15, 29).

After arthritis, carditis was the second represented Jones's criterion in our cohort. It was present in more than half of the overall cohort, significantly predominant in LR patients and mainly involving mitral valve. Similarly, subclinical carditis showed a prevalence in both groups with significant predominance in LR group. This outcome indicated a higher frequency of silent carditis (without a reduction in the frequency of total carditis) in our region than in the regions previously investigated (15, 29). Indeed, in most of the patients of our cohort the presence of polyarthralgia and monoarthritis, as first signs, oriented to a suspicious of ARF, leading to discover a silent carditis. It is known that ARF clinical features can resolve completely in weeks or months, while valvular lesions may persist and lead to RHD (30–32). The evolution of the silent carditis has not been completely clarified; however, a prospective study demonstrated that subclinical lesions can persist and sometimes even worsen (33). Since it has been observed that subclinical RHD prevalence correlates closely with ARF incidence, being considered as its surrogate measure in HR population (34), we highlight the relevance of the echocardiographic evaluation to estimate ARF incidence in HR populations.

Chorea was the third represented criterion in our cohort, females being more affected, as described in other reports (35). As opposed to our results and previous investigations (29), Pastore et al. (30) reported a higher prevalence of chorea (54%). Interestingly, we enlightened a higher frequency of carditis among patients with chorea either at presentation or during the follow-up; of note, chorea was usually associated with mild cardiac involvement, confirming that patients with chorea hardly develop severe carditis (36, 37). Therefore, we empathize that echocardiographic assessment should be routinely performed in each patient with suspected or confirmed chorea in order to identify the presence of valvulitis (even subclinical) at the time of diagnosis or developing later.

Among minor criteria, the prolongation of PR interval [occurring in over 2% of the normal population (38)] was detected in a limited number of our cases without a positive correlation with carditis or with other major criteria.

The present study has some limitations because of its retrospective design and a prospective study will be necessary to confirm our findings. Due to the lack of nationwide epidemiological data, especially in the unexplored South of Italy, an Italian registry of all cases of ARF may be a relevant aim. Finally, we point out the need of a critical approach in the application of the new ARF criteria in children presenting with fever, polyarthralgia and increase of ESR and/or CRP, particularly in HR population.

The epidemiological impact of HR criteria is relevant in Tuscany, where their application led to an increase in ARF diagnoses of 30%. Thus, Tuscany should be considered an HR area and the new criteria for HR populations should be used in this region. Moreover, if the observed trend were confirmed, it would represent an inversion with respect to the past century, when a decline in the incidence of ARF in developed countries has been described as consequence of improved public hygiene and widespread use of antibiotics.

The datasets generated for this study are available on request to the corresponding author on reasonable request.

AMQA conceptualized and designed the study, designed the data collection instruments, contributed to collect the clinical data, conducted the final database, and drafted the manuscript. FP conducted the final database, performed the statistical analyses, and drafted the initial manuscript. ADG designed the data collection instruments, performed the statistical analyses, contributed to interpretation of data, and drafted the initial manuscript. SC, GB, SA-R, AB, and AO contributed to collect the clinical data, carried out the initial analyses, and critically reviewed the manuscript for important intellectual content. DP, NA, CG, and GS conceptualized and designed the study, supervised data collection, contributed to interpretation of data, and critically reviewed the manuscript for important intellectual content. RC conceptualized and designed the study, coordinated and supervised data collection, contributed to interpretation of data, and critically reviewed the manuscript for important intellectual content. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.