This retrospective study from the FOAD between 2000 and 2021 describes a cohort of 54 patients suffering from pSLE. This is one of the largest cohorts of pediatric lupus patients of Afro-Caribbean origin. One of the strengths of this work is the multicenter inclusions, with the participation of all the reference pediatricians in the FOAD. Our methodology allowed us to analyze patient therapies by consulting the registries of referring physicians and examining computerized hospital archives in each center using broad inclusion criteria. This explains why, from the original, wide selection of 2148 patients identified, only 54 patients with a diagnosis of pSLE were finally included. This methodology ensured exhaustive identification of patients, with cross-checking of multiple data sources to minimize the loss of patients and data. It should be noted that there is potential for recruitment bias in the hospital computer archives, because they only index patients requiring hospitalization. Nevertheless, most patients with pSLE usually have a short hospitalization at the initial phase. There was also an imbalanced distribution across the departments of origin, with most of the patients registered coming from Martinique (59%). This can be at least partly explained by the fact that the regional competence center for childhood autoimmune diseases is based in Martinique. A second potential explanation is a disparity in identification of patients by the computerized archives system in Guadeloupe.

For pSLE, with a reported incidence of 0.3—0.9 per 100,000 children per year [16], and a total population of 330,000 children aged under 18 in Martinique, Guadeloupe and French Guyana combined, the expected number of pediatric lupus patients over the study period would have been between 21 and 62. We found 54 patients with pSLE, which is in line with the higher frequency of pSLE and adult SLE reported among the Afro-Caribbean populations in Martinique/Guadeloupe [2, 17,18,19]. The prevalence in our study was likely underestimated because of the memory bias inherent to retrospective studies, and a defect in the computerized hospital archives of one center in Guadeloupe. Since the populations of Martinique and Guadeloupe are similar in number, ethnic original, and environmental exposures, incidence in both departments would be expected to be approximately the same. Despite these limitations, which likely resulted in underestimation of the true number of cases, our cohort nevertheless found a high prevalence of pSLE, notably higher than previously reported in European countries or the Pacific islands [16, 20, 21]. To the best of our knowledge, there is no high-quality study of SLE incidence in cohorts of Afro-Caribbean descent. In one study in a population of US Medicaid beneficiaries (of whom 40% were African American), the authors found an incidence of 2.2 per year per 100,000 children [22], which seems consistent with our results. Similarly, in a study from Northern Italy, Tsioni et al. reported an annual incidence rate of 2.0 (95%CI 0.9–3.8) per 100,000 individuals overall, and of 3.8 (95%CI 0.5–13.8) in children [23].

Our study shows an increase in the number of pSLE over the years in FOAD. Given that cross-checking found 86% agreement between the physician registry and computer records, potentially lost data likely did not significantly affect the trend towards increasing cases over the years but may have underestimated the true number. Moreover, this was observed for 2010–2020, for which memory bias should be at its lowest. The genetic factors and other prominent environmental factor(s) explaining this increasing trend deserve to be explored [24, 25]. This increase is of similar magnitude to that of the two other main chronic autoimmune diseases in children, namely type 1 diabetes, and inflammatory bowel disease (unpublished, personal data). In our cohort, 30% of patients had a family history of SLE, and this strong heritability has previously been described, especially for lupus patients [26]. Conversely, the environmental factors have not been widely studied among the Afro-Caribbean population, especially for those living in the FOAD, and deserves further prospective investigation.

The average length of follow-up was quite long in our study, which enabled detailed study of the progression of these children during childhood. There was an high rate of infectious morbidity requiring hospitalization in our cohort, but none of the children died from infectious complications, which differs from African or non-White patients in emerging countries [27,28,29]. This infectious risk during childhood has previously been described in pSLE cohorts, with an increased risk in Afro-American populations [30].

The proportion of patients diagnosed after 10 years old was high in our study (85%). This differs from White populations, where the distribution is more balanced [16, 31]. The median duration of symptoms before diagnosis was short (3 months), and mortality in this pSLE cohort was low, with a median follow-up of 7 years; only one patient died from accidental causes (2%). These data show better results than those from emerging countries, and are comparable to statistics from cohorts in Western countries [19, 27, 28]. The children in our cohort had some florid symptoms at onset that were not associated with chronic sequalae (i.e., SDI score). This confirms the clinicians’ impression of greater clinical severity at diagnosis in this population. The classic symptoms at diagnosis were no different from those described in other cohorts, with a triad of febrile skin and joint involvement present in 9 out of 10 patients [27, 28, 31, 32]. The presence of mucosal involvement with oral ulcers (70%) was substantially higher in our cohort than previously described in pSLE or adult cohorts [33]. The proportion of patients with neurological or renal involvement at onset and during childhood corresponds to series described in Western countries, and remains somewhat lower than reports from emerging countries [5, 28, 31]. The proportion of cardiac involvement was significantly higher than in most cohorts described [5, 28, 31]. Despite this, there were no cardiovascular events or chronic, symptomatic impairment of cardiac function in our cohort. This can be explained by the low proportion of patients still being followed up beyond the age of 25 years [34] (16%).These findings nonetheless confirm greater organ involvement at onset and during pSLE [31]. Although the clinical symptoms and findings of laboratory tests are similar to those seen in adults, patients with pSLE tend to have a higher rate of major organ involvement and a more aggressive clinical course [31, 32]. In studies from North America or Africa, ethnic status as Afro-Caribbean seems to be associated with worse prognosis in pSLE. However, the French healthcare system is universal and free of charge, and therefore, the bias related to socioeconomic status and access to healthcare seems to be less significant in our cohort [35, 36]. Thus, with some clinical and biological specificities, patients with pSLE in our cohort had similar rates of organ involvement during childhood and similar overall outcomes to patients described in Western countries [27, 28, 32]. Despite suffering from several flares during childhood, SLICC/ACR Damage Index (SDI) scores were mostly nil in our cohort of patients. Conversely, organ damage associated with childhood flares occurred during adulthood (4 patients on dialysis).

The proportion of patients with anti-DNA, anti-Ssa, anti-RNP, and anti-Sm antibodies was very high in our cohort, and higher than reported in previous studies of White patients, highlighting increased autoantibodies in the population of African descent [37]. This is also consistent with previous reports indicating higher percentages of pediatric patients with anti-Sm and anti-RNP antibodies compared to adult patients [38,39,40]. The proportion of patients with antiphospholipid antibodies (30%) was also above adult standards [40]. Most of our patients had a marked increase in ESR at onset (80–90), in contrast with a moderately increased CRP level. CRP in pSLE is usually associated with seritis or infection rather than disease flares [41]. In our population, at the age of transition to adult care, boys seemed to have better disease control than girls, which could be explained by the puberty-related increase in testosterone, which has immunosuppressive activity [42].

One of the strengths of this study is the description of the severity of the disease during childhood. Our definition of a flare was based on a marked degree of severity, as opposed to a slight exacerbation of symptoms leading to minor therapeutic adjustment. With a median number of relapses of 3.4 per child, our patients suffer from a substantial flare every 2 years, with almost a third (28%) having more than 5 flares through childhood. In our cohort, patients who were younger at pSLE diagnosis had more severe disease during childhood, and more infectious problems. This has already been described as a criterion for severity for these diseases [43], but not for infectious issues, which can raise the question of the frontier between dysimmunity and immune deficiency in these young patients [44, 45]. Information about pubertal status at onset was poorly and unequally described in the medical records. Since puberty seems to begin earlier in Afro-Caribbean populations [46], we were better able to classify patients according to their age at onset. In addition, puberty starts earlier for girls [46], so classification strictly by pubertal stage would probably have led to a more balanced sex ratio in the prepubertal pSLE group. Organ damage, disease control, morbidity, and mortality in pSLE have previously been linked to risk factors such as young age at diagnosis, male sex, and Afro or Hispanic ethnicity [35, 43] that are not considered in the disease activity score (SLEDAI, EULAR).