Childhood-onset primary Sjögren’s syndrome in a tertiary center in China: clinical features and outcome

Primary Sjögren’s syndrome (pSS) is a systemic autoimmune disease, predominantly affects middle-aged women, with a frequency ranging between 0.01 and 0.72% [1]. Patients with pSS mainly experience oral and ocular dryness and extraglandular manifestations (EGM), including interstitial pneumonitis, interstitial nephritis, isosthenuria or renal tubular acidosis, thyroiditis, central nervous system involvement, vasculitis, and an increased incidence of lymphoma [2].

In children, pSS is less well characterized in terms of clinical presentation and long-term outcomes [3]. When the disease starts before the age 18 of years, it is called SS with childhood onset or juvenile SS. Childhood-onset pSS is rarely reported, poorly defined and possibly underdiagnosed due to the different criteria used for diagnosis and the scarcity of available data [4]. A recent report of the Sjogren Big Data Consortium revealed that childhood-onset pSS involved approximately 1% of patients with pSS, with a clinical phenotype dominated by sicca features, parotid enlargement, and the highest frequencies of systemic disease in 5 (constitutional, lymphadenopathy, glandular, cutaneous and haematological) of the 12 European League Against Rheumatism (EULAR) Sjögren's syndrome disease activity index (ESSDAI) domains [5].

Neither of the currently most widely used classification criteria of pSS, the American-European Consensus Group (AECG) criteria [6] and the American College of Rheumatology/EULAR criteria [7], have been validated in a juvenile population. Child-specific criteria have been proposed [8]; however, even these criteria have a sensitivity of only 76% for diagnosing childhood SS [9, 10].

In the absence of established classification criteria specific for pSS in children, we chose to focus our study on children who met any of the current classification criteria [6, 7, 8]. Therefore, the aim of this study was to characterize the clinical features, immunological profiles and long-term outcomes of childhood-onset pSS in a cohort of Chinese patients from a tertiary paediatric medical centre.

The number of studies on and the number of reported cases of childhood-onset pSS are very limited. The data available to date suggest that pSS is a very rare disease in children. Manuel Ramos-Casals et al. [5] estimated that the frequency of childhood-onset pSS was approximately 1% of patients with primary SS based on the worldwide data-sharing cooperative merging of preexisting clinical SS databases from the five continents, which was to date the largest multicentre study investigating the clinical characteristics of childhood-onset pSS, confirming that pSS is a disease that is rarely diagnosed in children.

Currently available classification criteria [6, 7] have been developed for adult patients, but they are not well validated in childhood-onset SS. Houghton K et al. [9] reported that 39% of patients met the AECG criteria and 76% met the proposed paediatric criteria; the authors proposed that the AECG adult classification criteria for pSS should not be applied to children as the sensitivity is unacceptably low. In addition, a recent study by Basiaga ML et al. [12] showed that the majority of children diagnosed with SS in an international childhood SS cohort did not meet the 2016 ACR/EULAR classification criteria, and the authors emphasized the need for further research, including the creation of paediatric-specific classification criteria. In our cohort, we comprehensively evaluated the existing adult criteria as well as the proposed juvenile pSS criteria [8], 97.4% of patients fulfilled the ACR/EULAR criteria, 89.7% fulfilled the proposed juvenile pSS criteria and 69.2% fulfilled the AECG criteria. The ACR/EULAR criteria are applicable to any patient with at least one symptom of ocular or oral dryness and consider systemic signs and B-cell activation biomarkers (using the ESSDAI) as inclusion criteria [7], which allows the diagnosis of systemic and earlier forms of the disease when sicca features are not already present. All but one patient (97.4%) in our series had systemic activity at diagnosis (global ESSDAI score ≥ 1) and was diagnosed with SS according to the presence of SSA autoantibodies and LSG biopsy findings, and the last patient’s diagnosis of SS was made according to the proposed juvenile pSS criteria. The ACR/EULAR criteria performed best in our study and might be optimal for the early diagnosis of and clinical research on childhood-onset pSS when combined with the proposed juvenile pSS criteria.

Based on the findings of previous studies [5, 13, 14, 15, 16, 17], the clinical glandular phenotype of childhood-onset SS is mainly dominated by parotid involvement rather than by sicca features, and the symptoms of dry mouth and eyes mostly appear during the course of the disease and rarely as the first complaint. Data from the Sjögren Big Data Consortium [5, 18] confirm that the systemic phenotype of pSS at diagnosis is strongly influenced by personal determinants such as age, gender, and ethnicity. They revealed that patients with childhood-onset pSS showed the highest mean ESSDAI score and the highest frequencies of systemic disease in 5 (constitutional, lymphadenopathy, glandular, cutaneous and haematological) of the 12 ESSDAI domains in comparison with patients with adult-onset disease. In adults, Black/African American patients had the highest frequencies of activity in the lymphadenopathy, articular, neurological and biological domains; White patients in the glandular, cutaneous and muscular domains; Asian patients in the pulmonary, renal and haematological domains; and Hispanic patients in the constitutional domains. More than half of the patients in our study presented with EGMs, and skin, haematological, hepatic and renal involvement were more predominant than in other studies. As summarized in Table 4, ethnic differences might also play a role.

It is remarkable that one-third of patients in our cohort had abnormal hepatic biochemical indices, while only a few studies have reported the frequency (2.4–10.0%) of abnormal hepatic biochemical indices in children [16, 17]. Hepatic involvement is observed in 27 to 49% of cases of pSS in adults, and half of these patients have overt clinical liver disease. The most common liver diseases are primary biliary cholangitis (PBC) and autoimmune hepatitis (AIH) [19, 20]. Hepatic involvement is also frequently found in patients with childhood-onset pSS; therefore, children with pSS should be monitored to identify clinical liver disease early. Interestingly, one patient in our series was diagnosed with pSS combined with cholestasis, RTA, proteinuria (focal proliferative glomerulonephritis and mild TIN), lower limb peripheral neuropathy and 47,XXX. 47,XXX is a rare event and is significantly increased among patients with SLE and SS. The estimated prevalence of SS with 47,XXX in women was 0.29%, which was 2.9 times and 41 times higher than that in women with 46,XX and that in men with 46,XY [21]. This case supported the link between the presence of extra X chromosome(s) and the development of autoimmune diseases and enriched the phenotype of 47,XXX with autoimmune diseases.

Renal involvement in SS is rare, affecting < 10% of patients (range from ~ 1 to > 30% in different studies and ethnicities of adult patients [22] and 4.5–19.2% in paediatric patients). Renal involvement in pSS is heterogeneous and ranges from isolated electrolyte disturbances and nephrolithiasis to glomerulonephritis and TIN in both acute and chronic forms, and TIN is the most frequent renal complication of pSS. One-fifth of our case series had renal involvement, and four patients underwent renal biopsy (one was to confirm the severity of TIN, the second was to differentiate with SLE, and the last two were confirmed to have lupus nephritis by biopsy when SLE was suspected during the follow-up period). Kidney biopsy can confirm the severity of TIN to determine the treatment regimen and help in the differential diagnosis, especially when glomerular disease is suspected. A kidney biopsy should be performed as soon as possible to rule out other possible causes, such as lupus nephritis.

In addition, in view of ocular involvement, uveitis should be considered as well, similar to earlier reports [17, 23], and deserves comprehensive and continuously monitored systemic evaluation. SS can cause serious, vision-threatening extraglandular ocular manifestations in addition to dry eye.

When childhood pSS is suspected clinically, an appropriate workup should be performed. With respect to the study of exocrine gland dysfunction, some diagnostic tests used in adults are not validated in children [12], such as the Schirmer test, unstimulated whole saliva flows, and conjunctival lissamine green staining, due to the limited cooperation of the children and the unavailability of stains. We performed corneal fluorescein staining to determine the presence of lacrimal gland disease, and positive findings were noted in only nine (24.3%) of the 37 children evaluated, suggesting that positive ocular staining may be a later manifestation, consistent with the clinical features of sicca symptoms. The finding of focal lymphocytic sialadenitis on minor labial salivary gland biopsy is specifically suggestive of childhood pSS [24, 25]. In previous studies [5, 13, 14, 15, 16, 17], salivary gland biopsy was positive in 53–100% of children (Table 4), and the frequency in our study increased to 100% among the 97.4% of patients who underwent biopsy, which could explain the high rate of meeting the EUALR criteria in our study, while only 43.6% of children in the study of Basiaga ML et al. underwent LGS biopsies, and nearly half exhibited positive findings [12]. Histopathological analyses should also include the biopsy of other salivary or lacrimal glands if tissue was procured for other purposes, such as to rule out lymphoma.

Salivary gland ultrasound (SGUS) is a noninvasive test that may be used to evaluate the glandular damage of salivary glands in children with Sjögren’s syndrome [26]. Recently, Hammenfors et al. [13] reported pathological SGUS findings in 41 (61%) of 67 patients with childhood-onset SS and indicated that SGUS findings were associated with hyposalivation and autoantibodies. In the study of Manuel Ramos-Casals [5], the rate was even higher (94%). In our centre, although five of ten patients with parotid involvement underwent SGUS, abnormal findings were revealed in 100%. SGUS might be an interesting diagnostic tool for identifying patients with pSS in childhood, and further clinical application is needed.

The majority of our patients achieved a good response at a median follow-up time of nearly two years, while three patients (nearly 10%) had SLE at 13.3 months, 38.8 months and 63.8 months. Previous studies published that 7.5–10% of patients with long-standing pSS eventually developed a systemic disorder that fulfilled the criteria for SLE in adults [27, 28]. Similar to paediatric patients, Schuetz C et al. [10] reported that one in eight (12.5%) children with pSS developed overlapping lupus nephritis one year after the diagnosis of pSS. The coexistence of SS and SLE was first demonstrated in 1959 [29]. Compared with SLE alone, this combined disease of SS-SLE has distinct features with relatively less major internal organ involvement but a more specific immunological profile and more favourable clinical outcomes [30]. Yujiao Y et al. revealed that patients with Sjögren's syndrome that progresses to systemic lupus erythematosus (SS/SLE) had an earlier age of onset and higher incidences of proteinuria and low complement levels than patients with pSS in a retrospective case‒control study. Two of them experienced hypocomplementemia and lupus nephritis; thus, long-term follow-up and careful evaluation for SLE in patients with childhood-onset pSS are recommended, and we should be vigilant with patients with these mentioned clinical and laboratory characteristics, who have a higher risk of developing SS/SLE. No patients in our cohort progressed to lymphoma in a median follow-up of nearly 2 years. The complication of lymphoma is rare but is reported in some children with pSS [31, 32]. Whether the risk is similar to or greater than that in adults is not yet known, and attention to this complication is needed in children with pSS.

The study has some limitations. As this was a retrospective study, we analysed preexisting data obtained from electronic medical records; therefore, recall bias was inevitable. This was a single-centre study of Chinese children, and a larger cohort size based on a multicentre study or national registry system is necessary to better define and understand the natural history of pSS in children and to determine which features accurately predict a diagnosis and progression.

In conclusion, the main presentation of childhood-onset pSS were atypical, and extraglandular manifestations and systemic involvement were more common. Labial salivary gland biopsy is vital for patients with probable pSS. The 2016 ACR/EULAR criteria might be optimal for early diagnosis based on systemic assessment and labial salivary gland biopsy. Some patients with typical pSS may develop SLE over time, and attention is drawn to close monitoring and long-term follow-up.