The positive result for HIV in the otherwise asymptomatic girl aged 3 years, who had been born in a different country after inadequate monitoring during pregnancy, combined with the positive HIV test in the mother, confirmed vertical transmission of HIV. At the time of diagnosis, the disease was in its early stages with a normal CD4 count and mild immunosuppression. It is known that HIV-associated salivary gland disease tends to develop in the early stages of infection, in some cases before deterioration of the immune system becomes apparent, which makes parotitis an important warning sign for suspicion of HIV. Its pathogenesis is unclear, and some authors have proposed an autoimmune or chronic inflammatory aetiology as opposed to a purely infectious one.6 It is more frequent in children compared to adults, and tends to follow a benign course, as was the case in our patient. When JRP is detected, it is important to include HIV in the differential diagnosis, especially in children from foreign countries with unknown maternal HIV status or born to mothers in which HIV testing or monitoring was only performd in the first trimester of pregnancy.

Testing for detection of non-HIV viruses performed in recurrent episodes of parotitis8 did not identify any acute infections, so it may not be a priority in the initial workup of JRP due to the lack of a temporal association with the development of these episodes. However, the positive IgM results for non-HIV viral infection obtained in the first episode of parotitis in 4 of our patients in whom the test was ordered based on the clinician’s judgment suggests that an initial viral infection could cause immune changes prolong the period during which patients experience recurrent inflammatory episodes until their spontaneous resolution in puberty.9

There is also evidence in adults of a potential association between SS-like sialadenitis and chronic infection by hepatitis C virus.10

The detection in our series of 2 cases of selective IgA deficiency could be due to chance given the high documented prevalence of this disorder in the general population (1 case per 400–3000 donors). However, Fazekas et al.11 identified 5 cases in a total of 17 patients with JRP. Although the authors acknowledged that a clear association has yet to be established, they remarked on the growing interest in serological abnormalities in patients with JRP and the publication of individual case reports of patients with JRP and selective IgA deficiency. Other authors have since published case reports of patients with JRP and isolated IgA deficiency.12,13 In clinical practice, JRP could be considered a warning sign of potential selective IgA deficiency, and therefore of the disorders and abnormalities associated with this disease, such as selective IgG subclass deficiencies, common variable immunodeficiency, recurrent sinopulmonary infections, gastrointestinal disorders (especially coeliac disease), neurologic disorders, autoimmune diseases and even malignant tumours. The IgG subclass levels in our 2 patients were normal or elevated, and no associated diseases were identified in the initial workup or during the follow-up.

The study by Concheiro et al.,1 from which we obtained the laboratory and imaging test protocol for evaluation of JRP used in our study, included testing for markers of coeliac disease as part of the workup, which would be founded on the potential association of JRP with salivary gland disease resulting from malnutrition secondary to gluten enteropathy, cystic fibrosis or diabetes mellitus. However, none of the 3 patients with coeliac disease in our series, confirmed by the detection of villous atrophy (Marsh score III), 1 of whom received a diagnosis of diabetes mellitus 10 months after, nor another patient with normal villous architecture but with persistently positive tissue transglutaminase and endomysial antibody test results (possible coeliac disease) had any degree of malnutrition. In the 2 years of follow-up, these patients continued to experience episodes of parotitis while on a gluten-free diet. There is no previous description in the literature of a monosymptomatic form of coeliac disease presenting as JRP. We only found one article published in 198314 that described the case of a female patient aged 13 years with JRP and villous atrophy presenting in association with selective IgA deficiency, elevation of ANA, atopic dermatitis, hay fever and blepharitis that received a final diagnosis of gluten enteropathy and who achieved histological recovery and a reduction in the frequency of parotitis episodes with a gluten-free diet. The only reference to her nutritional status was growth delay at age 11 years, while the episodes of parotitis started 1 year after. The author interpreted the findings as immune changes associated with JRP and not secondary to undernutrition. More recently, Gungor et al.15 described a case of JRP associated with Sjögren syndrome and made a literature review of cases associated with immune disorders, including the case of coeliac disease documented by Friis et al.14

Our findings suggest that JRP could be a new indicator to rule out monosymptomatic coeliac disease. Based on the previous evidence, routine screening for coeliac disease could be useful in these patients due to the potential association of JRP with this disease. Routine testing for markers of coeliac disease in patients with JRP could bring forth evidence that this obscure association was not due to chance in future case series.

Sjögren syndrome is a progressive autoimmune disease that affects exocrine glands, mainly salivary and lacrimal glands, and presents with xerostomia and xerophthalmia. Although frequently diagnosed in adults, onset in childhood or adolescence is rare and often incomplete, in which context development of recurrent parotitis is a common warning sign. A multicentre study in 40 paediatric patients with a diagnosis of primary SS found that the most common manifestation at onset was JRP, present in 72.5% of cases, while the rest of initial signs and symptoms were nonspecific.16 The authors of this study reviewed smaller case series of similar characteristics published in the past. As was the case in our patients, all patients presented with elevated markers of autoimmune disease, especially ANAs, anti-Ro/SS and anti-La/SSB antibodies, rheumatoid factor and hypergammaglobulinemia. Patient 5 (Table 3) also tested positive of anti-dsDNA antibodies and lupus anticoagulants, had low C4 levels and leukopenia and had arthritis in more than 2 joints, leading to diagnosis of Sjögren syndrome secondary to or associated with systemic lupus erythematosus (SLE). Patient 8 (Table 3) only had elevation of ANAs (1/1280) and hypergammaglobulinemia, and the parotid scintigraphy found features compatible with early-stage SS. The absence of other signs and symptoms and negative results of other tests for immune markers typical of SS could be explained by the young age at diagnosis (4 years) compared to the mean age at diagnosis of cases reported in the literature (9.4–10.7 years).17–19 The heterogeneity of the applied diagnostic criteria and the lack of validation of these criteria in the paediatric age group pose a challenge to the diagnosis of SS at early ages, which requires additional tests, such as salivary gland scintigraphy or biopsy, and a longer follow-up. Gungor et al.15 described one patient that experienced recurrent episodes of parotitis from age 2 years and received a diagnosis of SS at age 4 years despite not meeting the full American-European consensus criteria, something that also occurs in other case series in the literature. Once again, our findings highlight the importance of performing tests for markers of autoimmune disease in patients with JRP, including those younger than 5 years (contrary to the recommendation of Concheiro et al.1), given the importance of early diagnosis of SS to initiate follow-up and treatment by a rheumatologist.

In 20 patients in our series, testing did not identify any serological abnormalities, and they received a final diagnosis of isolated JRP at variable times after onset. Of the 16 remaining patients, 10 have been described in previous sections, and in the other 6, the evaluation found isolated positive titres of ANA (1/40 to 1/120) with or without hypergammaglobulinemia and low levels of C4, with positive thyroid peroxidase antibodies accompanied by elevation of thyroid-stimulating hormone in one patient that received a final diagnosis of Hashimoto disease. A specific autoimmune disorder could not be diagnosed in these patients except the last one, so the remaining 5 patients received a final diagnosis of isolated JRP.

In one of the largest series published to date, Concheiro et al.1 found increased levels of IgE in 2 cases and decreased levels of IgA in 1 case out of a total of 30 cases that met the criteria for JRP. None of the cases was associated with an autoimmune disorder. In another large series published by Leerdam et al.,3 of a total of 50 paediatric patients identified retrospectively in a period of 21 years, 2 had increased levels of ANA, 1 tested positive for antithyroglobulin antibodies and 1 had levels of IgG and IgM in the lower limit of normal. These findings may be biased due to the low percentage of patients that underwent measurement of serum immunoglobulin levels (35.8%) or tests for detection of autoantibodies (24.5%). Of the other 3 patients, 2 received a diagnosis of congenital hypogammaglobulinemia, and 1 a diagnosis of SS with positive tests for ANA, rheumatoid factor and anti-Ro antibodies. In a retrospective 4-year review of cases that included 33 patients, Papadopoulou-Alataki et al.4 did not find any positive autoantibody test results or evidence of autoimmune disease. They only found one patient with selective IgA deficiency and another with isolated IgG4 deficiency.

We are unable to provide data on IgE levels and IgG subclasses because they were not measured routinely in the diagnostic workup. We also cannot conclude that autoimmune disorders and isolated antibody abnormalities are more frequent in patients with JRP compared to the general population, as the series was small and there was no control group, although overall we did find a greater than expected prevalence of final diagnoses related to autoimmunity.

In a prospective case-control study, Wu et al.20 assessed immune function in 100 paediatric patients per group and found significant differences with higher CD4 cell counts and lower CD8 counts and IgG, IgE, IgA and C3 levels in the case group compared to controls. The authors concluded that immune function was altered in patients with JRP compared to the general population, with a reduced cellular immune response and inadequate antibody production. The authors proposed that immunotherapy aimed at improving cellular immunity, and therefore the immune response at the systemic and parotid gland levels, should be a priority in these patients.

Lastly, López Picó et al.21 were the first to describe a case of JRP in association with natural killer (NK) cell deficiency in a girl aged 9 years, but given the benign course of the disease, they did not recommend routine performance of NK cell counts.