Systemic lupus erythematosus (SLE) is a severe, multi-systemic rheumatic disease. It is most prevalent among women of childbearing age, but can occur in all ages. In about 10–20% of cases it begins before 16 years.1,2

Several studies have reported that age at onset has a modifying effect on disease expression. Juvenile-onset SLE (jSLE) tends to have a more aggressive presentation and course, high rates of organ involvement and increased need for long-term immunosuppressive medications.3–5

Most studies have found a higher prevalence of lupus nephritis and haematological involvement in jSLE. Among adults, mild forms prevail, being the discoid lupus and arthritis more common in this group. However, in respect of many other clinical manifestations, including neuropsychiatric lupus (NPSLE), serositis, and autoantibody profiles, reports are conflicting, which is probably due to the small number of patients analysed in these studies.4–7

These observations are of high importance since a high prevalence of severe organ involvement, as well as an early and prolonged exposure to high dose of steroid therapy and immunosuppressive drugs, raises concerns regarding the long-term morbidity and early mortality.

Although patients with jSLE present more often with severe forms of disease, differences in cumulative long-term damage, according to age of onset, is not well known. Few studies have proved the existence of a greater cumulative damage juvenile SLE patients. In 2008, Tucker and colleagues found a higher renal damage in these patients. In the same year, Ramirez Gomez and colleagues and Brunner and colleagues, showed that juvenile SLE patients had higher rates of activity (assessed by SLEDAI) but also may a higher chronic damage (assessed by SLICC/ACR index).8–10

The objectives of the study were to compare patients with juvenile-onset SLE with those with adult-onset subset, regarding demographic, clinical and immunologic manifestations and to compare the cumulative organ damage between the groups.

Cross-sectional study was performed.

A total of 204 patients diagnosed with SLE were included, comprising 38 (18.6%) jSLE and 166 (81.4%) adult-onset SLE patients. 187 (91.7%) patients were female and had a mean age of 46.1±15.4 years and a mean of disease duration of 17.1±10 years (Table 1).

The most prevalent clinical manifestations in this cohort of SLE patients were: arthritis (88.2%) and muco-cutaneous manifestations such as photosensitivity (73%) and malar rash (57.8%). The other systemic organ involvement was less common (Table 1).

Regarding immunologic profile, all patients had positive ANA and 92.6% had positive anti-dsDNA (Table 1).

Comparison between juvenile and adult-onset SLE patients (Table 2) showed no statistically significant differences regarding demographic characteristic, except for age at diagnosis, as it was expected.

Juvenile SLE patients had more often muco-cutaneous manifestations, such as malar rash (78.9% vs 53%; p=0.001) and oral ulcers (45.5% vs 17.5%; p=0.001). Arthritis was more prevalent in adult-onset patients (70.9% vs 90%; p=0.04).

Neurological involvement was more commonly observed in jSLE group (13.1% vs 3.6%; p=0.02), as well as nephritis (50% vs 33.9%), p=0.04) and haematological manifestations such as haemolytic anaemia (23.6% vs 5.4%; p=0.002) and leukopenia (46.1% vs 4.2%; p<0.001).

Other manifestations, like photosensitivity, serositis and thrombocytopenia occurred in comparable percentages in both groups. Regarding immunologic profile, we did not find any statistically significant differences.

During the follow-up, the number of deaths was comparable in both groups. Were observed three deaths, two in the group of adults (one due to septic arthritis and septic shock and the other one in the context of pulmonary thromboembolism) and one in the group of juvenile-SLE (due to a severe pulmonary arterial hypertension). The mortality rate observed was 0.05 cases per 100 person-years in adult-onset SLE group and 0.15 in juvenile-SLE.

The disease activity score (SLEDAI) and the damage index SLICC/ACR were comparable in both groups. However, the percentage of the patients with irreversible damage (SLICC≥1) was superior in adult SLE patients group (24% vs 20%; p=0.04).

The study population consisted in patients from a public tertiary-level centre in Portugal. The clinical and demographic characteristic of the patients included in this study are similar to other Portuguese cohorts which were previously described.3,12

Consensus regarding the cut-off age for defining jSLE is still missing in literature. In our study, and also in most published studies, 16 years was considered the upper limit to define juvenile-Lupus. However, there are some studies that consider 18 years. Such variations may contribute to differences in results across studies.1,2

So far, few centres have deal with the comparison of paediatric and adult-onset lupus and only one meta-analysis has been published.4,7,13–15 In our study we found a higher prevalence of malar rash, oral ulcers, haemolytic anaemia, nephritis and CNS involvement in patients with juvenile onset, whereas arthritis was more prevalent in adult patients.

In fact, our data support the 2011 Livingstone meta-analysis results, where malar rash, nephritis, thrombocytopenia, haemolytic anaemia, seizure, fever, and lymphadenopathy were more common in jSLE whereas Raynaud's phenomenon and secondary Sjögren's syndrome were more prevalent in adult-onset group.13

Despite many controversies in the results of various studies, the most consistent among them is the higher prevalence of lupus nephritis. The most discordant results are with respect to CNS involvement probably because it is a rare manifestation and cohorts are too small, but we managed to prove in our cohort.13

Few studies have reported differences in chronic organ damage, which is one of the main concerns in the management of paediatric patients.8–10 In our study, SLICC/ACR damage index were comparable in both groups. However, the percentage of the patients with irreversible damage (SLICC≥1) was superior in adult SLE patients group. In fact, while some components of the SDI may be related to the use of corticoids (cataracts, avascular necrosis, osteoporosis, muscle atrophy/weakness) and cyclophosphamide (premature gonad failure, malignancy), some components may also be related to age (cataracts, diabetes mellitus, neoplasia, osteoporosis, stroke, and acute myocardial infarction) being expected a higher value in the adult group.

Yet it must be emphasised that 20% of paediatric patients had some chronic damage (SLICC/ACR DI≥1) which already shows a high percentage of morbidity in this group of patients.

Our study has a number of limitations. The use of a cross-sectional design, with retrospective analysis of medical records, may have underestimated the frequency of some clinical manifestation.

In summary, our study confirms the existence of differences in clinical phenotype, according to age at diagnosis of SLE. Although juvenile-SLE patients have high percentages of disease damage (20% of patients), we could not show differences comparing to adults groups, perhaps because many SLICC/ACR damage index items reflect age. It is clear that rheumatologists caring for children with SLE must be aware of the greater risk of major haematological, renal and CNS involvement as well as important long-term morbidity.

This research received no specific grant from any funding agency in the public, commercial, or not for-profit sectors.

The authors declare that they have no conflict of interest.