Epidemiology and geographic clustering of Erdheim-Chester disease in Italy and France

TO THE EDITOR:

Erdheim-Chester disease (ECD) is a multisystemic histiocytosis¹ characterized by heterogeneous phenotypes, ranging from an incidental localized disease to severe forms with multiorgan involvement.² ECD typically develops in adults, with an incidence peak of diagnosis between 46 and 60 years of age, and has a male-to-female ratio of approximately 3:1.³ Pediatric cases are exceptional.⁴ The growing awareness of this disease determined a considerable increase in the number of diagnoses: by 2006, only ∼240 cases had been reported in the English language literature,⁵ whereas the current estimated number of diagnoses worldwide exceeds 1500. ECD is driven by gain-of-function mutations of genes belonging to the MAPK and the PI3K-AKT pathways, with most patients harboring the BRAF^V600E mutation.^6,7 Irrespective of genetic aberrations, the MAPK cascade activation and an uncontrolled induction of the mechanistic target of rapamycin pathway are hallmarks of ECD.^8,9 Proinflammatory cytokines and chemokines are also crucial in the development of tissue lesions and systemic inflammation, but the pathogenic link between clonal histiocytic proliferation and inflammation remains unclear.² In line with this dual clonal-inflammatory pathogenesis, drugs targeting the mutated proteins as well as immunomodulatory agents (ie, BRAF/MEK inhibitors, interferon-α, and mechanistic target of rapamycin inhibitors) are effective in ECD and greatly improved survival, with 3-year mortality rates dropping from 60% to 20% in the last 2 decades.^3,10 

Germ line genetic aberrations and exposure to environmental factors confer susceptibility to clonal malignancies and inflammatory diseases.^11-15 In a recent genome-wide association study, genetic variations of SETBP1, a gene involved in myeloid cell biology, were associated with an increased susceptibility to develop ECD.¹⁶ Given the dual nature of ECD and since it typically develops in adulthood, it is reasonable to investigate whether exposure to environmental agents also plays a pathogenic role.

Epidemiological studies investigating environmental exposures as putative risk factors for a given disease often use geographic clustering and disease mapping as an initial approach.¹⁷ Herein, we report the results of a study including 356 adults with ECD, designed to assess the geographic origin of patients in Italy and France. We included all patients diagnosed as having ECD in adulthood who were born in Italy or France and whose city of birth was known. We considered patients diagnosed between 1996 and 2022 in Italy and between 1986 and 2022 in France. In retrospect, we checked that all diagnoses fulfilled the 2020 recommendations for ECD.¹⁸ Data concerning the region where the enrolled patients had lived for most of their lives were not available for all individuals. Among the 43 Italian and the 96 French patients for whom this information was available, we found that the region where they were born and the region where they spent the first 30 years of their lives coincided in 91% and 90% of the cases, respectively. Metrics regarding Italian and French region demographics (ie, resident adult population) were obtained from public reports produced by the Italian National Institute of Statistics and by the French National Institute of Statistics; these data were used to normalize the regional number of ECD diagnoses. A previously described statistical test assessing significance for geographic clusters of disease¹⁹ was used to analyze the observed clustering of ECD diagnoses in Italian and French regions. For each region the age-adjusted number of diagnoses per 1 000 000 adult residents was classified according to their ratio over the national average, which was taken as 100: (1) above national average (>115), (2) comparable to national average (85-115), (3) below national average (<85) (supplemental Table 1, available on the Blood website).

The main patient characteristics are described in Table 1. In Italy, data of 140 patients were obtained from clinical charts of the 2 Italian ECD referral care centers (ie, Meyer Hospital, Florence, and San Raffaele Hospital, Milan) and of the other centers belonging to the Italian ECD Network. Patients with childhood-onset ECD (n = 5), born in other countries (n = 13), or lacking data (n = 4) were excluded, leading to a final cohort of 118 individuals. In France, data of 354 patients were obtained from clinical charts of the French ECD Referral Care Centre (ie, Hôpital Universitaire Pitié Salpêtrière, Paris). Patients with childhood-onset ECD (n = 4), born in overseas French regions (n = 5), born in other countries (n = 96), or lacking data (n = 11) were excluded, leading to a final cohort of 238 individuals (supplemental Figure 1).

During the study period, we detected a progressive increase in the number of ECD diagnoses, with a fall after the COVID-19 pandemic in 2020 (supplemental Figure 2). If we assume that all new cases were centralized to the referral centers, the average disease incidence of the 2 countries considered together in 2018 to 2020 was 0.35 cases per 1 000 000 adult residents per year. This value probably underestimates the true disease incidence because other ECD cases might not have been referred to the referral care centers.

In Italy, the average number of diagnoses during the observation period (the 26 years from 1996 to 2022) was 2.37 per 1 000 000 adult residents. The Italian regions showing a frequency of ECD diagnoses over national average clustered in southern Italy (Figure 1). In France, the average number of diagnoses during the observation period (the 36 years from 1986 to 2022) was 4.68 per 1 000 000 adult residents. Regions exhibiting a rate of diagnosis over national average clustered in central-northern France (Figure 1). Statistical analysis demonstrated that the geographic aggregations observed in Italy and in France were statistically significant (P value for both < .01).

To explore whether the geographic distribution of environmental agents was associated with such a clustering, we assessed the national maps of air, soil, and water pollution but found no significant overlap with the areas where ECD diagnoses clustered. However, we found a striking overlap with the average Human Development Index (HDI) of the regions of interest, evaluated in the same study period (data obtained from GlobalDataLab.com, Subnational HD, v.7.0, 28 March 2023): in particular, the regions displaying the higher rates of ECD diagnoses had the lower HDI (Figure 1). The HDI summarizes the average accomplishment in 3 key determinants of human development: health (assessed by life expectancy), education (expressed as mean of years of schooling for adults aged ≥ 25), and standard of living (derived from the gross income per capita). This parameter is used by the United Nations Development Programme Human Development Report to assess the degree of development of a country.²⁰ The matching of the 2 mappings suggests that socioeconomic factors, productive activities, lifestyle, and environmental exposures could all be connected to risk factors contributing to ECD. Our study thus sets the stage for future epidemiological research aimed at identifying such risk factors. This observation, however, does not exclude that genetic factors in the same geographic areas may also be major determinants of the disease. It is worth noting that geographic patterns of disease were also described for other malignancies (ie, leukemia)²¹ and that there is proven evidence of an inverse association between some cancer types and HDI.²² 

In summary, ECD is increasingly diagnosed in Italy and France, possibly thanks to the growing awareness about this rare condition. In these 2 countries, we identified a geographic clustering of the disease with a distribution that appears inversely proportional to that of HDI, a socioeconomic parameter expressing the degree of health, education, and income. Further research is needed to identify environmental exposures to agents potentially involved in the development of the disease.

Contribution: F. Peyronel designed the study and analyzed the data; F. Peyronel, A.V., and F.C.-A. wrote the manuscript; and all authors followed the patients, collected data, and reviewed the final version of the manuscript.

Conflict-of-interest disclosure: E.A. reports receiving grants from the Italian Ministry of Health (GR-2019-12368506) during the conduct of the study and personal fees from Kiniksa Pharmaceuticals and Cytokinetics outside the submitted work. P.L.Z. reports being a consultant for MSD, Euspharma, and Novartis; a member of the speakers’ bureau of Celltrion, Gilead, Janssen-Cilag, BMS, Servier, MSD, AstraZeneca, Takeda, Roche, Eusapharma, Kyowa Kirin, Novartis, Incyte, and Beigene; and a member of the advisory board of Secura Bio, Celltrion, Gilead, Janssen-Cilag, BMS, Servier, Sandoz, MSD, AstraZeneca, Takeda, Roche, Eusapharma, Kyowa Kirin, Novartis, ADC Therapeutics, Incyte, and Beigene. G.C. currently works for Novartis. The remaining authors declare no competing financial interests.

Correspondence: Augusto Vaglio, Nephrology and Dialysis Unit, Meyer Children’s Hospital IRCCS, Viale Gaetano Pieraccini, 24, 50139 Florence, Italy; e-mail: augusto.vaglio@unifi.it.