The majority of studies reporting the long-term outcome of jSLE have examined data collected from patients attending paediatric rheumatology clinics, with no direct comparison with adult-onset patients. Although, some of these studies report 5- and 10-year outcomes at the time of last follow-up, most patients in these studies still remain in late adolescence at the time the data had been analysed. This study is a unique description of outcome of adolescents with SLE, who have been followed up in adulthood. Using a nested case-control study design, we examined a multiethnic longitudinal cohort of adult lupus patients and assessed whether there were differences in clinical course and outcomes when SLE begins in adolescence. These adult patients were not taken from paediatric rheumatology clinical programmes, but rather were followed by adult rheumatologists at the time of study entry and thus represent an exceptional study population. There were no patients in the childhood-onset group whose disease began prior to age 13 years. In the LUMINA study, patients have to be 16 years of age to enter the cohort, and cannot have more than 5 years of disease prior to entry; therefore, the earliest age at disease onset could have been 11 years. Patients with a younger onset of SLE would likely have been referred to paediatric programmes and are, therefore, not represented in this study.
In this study, we found that patients with adolescent-onset SLE had more active disease during the entire follow-up period as measured by the SLAM-R and physician rating of disease activity, although these differences were not statistically significant. Moreover, patients with adolescent-onset SLE were found to have significantly higher occurrence of renal and neurological involvements at time of diagnosis when compared with adult-onset lupus patients. These findings are consistent with other studies of paediatric SLE cohorts.19,20
Font, et al
described a higher proportion of patients with renal involvement at diagnosis in childhood-onset SLE patients as compared to patients with adult disease (20% vs. 9%); in contrast, they found neurological involvement to be more frequent among patients with adult-onset lupus. Increased SLE activity is likely to result in greater requirement of aggressive treatments to control disease, and in the longer term more disease damage. Our findings substantiate this with significantly higher rates of renal damage and ESRD amongst SLE patients with adolescent-onset disease. In our study, adolescent-onset lupus patients were almost two times more likely to be hospitalized due to SLE-associated causes and four times more likely to receive intravenous cyclophosphamide when compared with adult-onset patients; these findings demonstrate the significant persistent impact of SLE on the health of affected youth.
In our study, patients with adolescent-onset SLE accrued more damage with significantly more renal damage within the first years of their disease when compared with adult-onset disease. In a multicentre, multinational cross-sectional study from paediatric rheumatology centers in Europe, USA, Mexico and Japan where disease damage in 387 patients with SLE was examined, patients had a mean damage score of 1.1 after mean follow-up of 5.7 years.21
An extension of this multicentre study to include 1015 patients showed a mean damage score of 0.8 (mean follow-up 4.0 years), with a progressive increase in damage score over time with increasing disease duration.22
Three smaller single institution studies from paediatric rheumatology centers in Norway, USA and Canada showed similar results with mean damage scores of 1.3,23
. Although, the mean damage scores of patient groups in these studies are similar to damage scores reported for aSLE patients, it is difficult to compare these results taken from very different cohorts. The comparison of adolescent-onset to adult-onset patients from the same cohort in this study showed a trend towards higher damage scores in the adolescent-onset SLE patients at last follow-up. Longer follow-up and larger patient numbers may help to further clarify this issue. Of interest, the mean damage index of our adolescent-onset patients in this study was 2.3, which is higher than that reported in all other paediatric rheumatology clinic population studies. This may be explained by factors, such as patient ethnicity, access to care, compliance, or socioeconomic factors which impact disease outcome in the LUMINA cohort, because the length of patient follow-up in our study is similar to the other paediatric studies.
In addition to clinical outcomes, SLE has broad impact on socioeconomic, educational and behavioural functioning of patients. Patients whose disease begins during childhood and adolescence are at risk of having interference with the normal developmental tasks of late adolescence and young adulthood, and this may result in problems in establishing healthy independent adult behaviour. 24
There have been very few long-term prospective studies of children with SLE examining these types of outcomes. Candell Chalom, et al.6
described the long-term educational, vocational and socioeconomic status and quality of life of 64 lupus patients with disease onset in childhood from their paediatric rheumatology center, followed up after a mean of 13 years. These authors found that most patients were living on relative low incomes (< $30 000 USD per year), and many still lived with their parents. Our study also shows that patients with adolescent SLE were more likely, although not significantly, to be living below the poverty line than the matched adult-onset group. This finding may be partially explained by younger age individuals being less established in permanent jobs or just beginning in pay scales of their careers. However, this may indicate a potential risk factor for juvenile lupus patients which should have further follow-up, because poverty has been identified as a negative long-term prognostic factor in SLE by us and others.25,26
A two-fold increase in mortality of adolescent-onset SLE patients seen in this study was unexpected and disturbing. Overall survival for children and adolescents with SLE has improved from 5- and 10-year survival of 60–78%27,28
in the 1970s, to 94–100% 5-year and 85% 10-year survival3,6
in 2000. However, these data come from patient cohorts cared for in paediatric rheumatology multidisciplinary clinics, which provide coordinated careful patient routine follow-up, patient education and transition to adult healthcare services; this type of care may have a positive impact on long-term survival. In contrast, Levy, et al.29
studied the outcome of 118 children with SLE followed in paediatric units of public hospitals in Paris and suburbs as well as some children seen in adult medical units. They found that 23 (19%) of these patients died, a mean of 7.6 years after disease onset. The authors suggest that the higher rate of mortality in their study may be related to poor socioeconomic status and treatment non-compliance among patients who died. A more recent study of 31 children with SLE from Chandigarh, India, demonstrated a very high mortality rate of 32% after 10 years of follow-up, related to late referrals and delay in diagnosis of disease.30
The issue of increased mortality amongst children and adolescents with SLE is concerning, and deserves further study in larger cohorts, with consideration of factors of healthcare services, socioeconomic background and continuity of care through adolescent transition.
This study has a number of limitations. Due to the inclusion criteria of the LUMINA study (age ≥16 years and ≤5 years of disease duration at entry) and the fact that all the LUMINA patients were seen in adult rheumatology clinics, we were not able to study children with SLE onset prior to 11 years of age, thus the data focus on adolescent-onset disease. The numbers of adolescent-onset patients in the LUMINA study were relatively small, which represents an issue of power for some of the analyses presented. The higher number of African-American patients in the adolescent-onset SLE patients may influence the results, as some studies have indicated differences in outcomes based on patient ethnicity. The current patient follow-up time among studied patients is relatively short [4.4 (2.9) years overall], preventing a more comprehensive assessment of longer term outcomes. We would hope to continue to capture these patients as the LUMINA study continues, allowing for 10- and 15-year follow-ups.
In conclusion, the severity of adolescent-onset SLE, as shown by persistence of active disease over long periods of time, results in more frequent disease damage, an increase in renal morbidity and may impact mortality. Young adults with adolescent-onset SLE were two times more likely to be living in poverty, which may represent an important negative prognostic factor for these patients. These findings suggest that an aggressive approach to the treatment of adolescent-onset SLE, coupled with educational, vocational and transition support might be important measures to improve outcomes.