Childhood- and adult-onset lupus: an update of similarities and differences

Eva D Papadimitraki, MD, PhD; David A Isenberg, MD, FRCP, FACS


Expert Rev Clin Immunol. 2009;5(4):391-403. 

In This Article

Epidemiological & Clinical Features of SLE

Epidemiological Data

One of the main difficulties in comparing data from published studies on cSLE is the lack of agreement about the definition of a 'child', with the cut off for inclusion as a 'childhood-onset case' varying between 14 and 20 years of age.[1–8] As disease expression in SLE is influenced by environmental factors and differs between racial and ethnic groups, when attempting comparisons, it is also important to use cohorts of adults and children with SLE from similar backrounds. Furthermore, juvenile SLE patients are invariably referred to adult clinics and thus are treated by different physicians when their age exceeds a certain limit, which might render their long-term enrollment to studies problematic or alternatively confound the studies they participate in. Despite these limitations, some conclusions regarding epidemiological differences between adult- and childhood-onset disease can safely be drawn. As already mentioned, in around 15% of individuals with SLE, the disease begins prior to 16 years of age. The diagnosis of lupus is uncommon before 10 years of age.[11,12] The median age of diagnosis of SLE in children is 12.1 years, with the female to male ratio ranging from 2.3:1 to 9:1, depending on the study.[13–17] In several series, the female to male ratio in children presenting with lupus before the age of 12 is 3–5:1, whereas this ratio for lupus presenting in the peri- or postpubertal age is approximately 5–7:1, approximately the same ratio as is seen in adults.[11] The incidence and severity of childhood-onset SLE varies among different ethnic groups. In Caucasians, the incidence of lupus onset prior to 19 years of age is between six and 18.9 cases per 100,000 individuals, whereas in pediatric populations of African–American ancestry, it reaches 30 cases per 100,000 individuals, emphasizing the striking impact of race on the incidence of lupus (up to a threefold increase in the prevalence of disease is observed in non-Caucasians), comparable to adult-onset disease.[12] Younger age, male sex, non-Caucasian race, low socioeconomic status, nephritis and CNS disease are considered to be risk factors for severe lupus; however, their association with a poorer prognosis in childhood- or adult-onset disease remains controversial.[13,18–20]

Clinical Manifestations

General Features Comparing the clinical features of childhood- and adult-onset disease reveals similarities as well as important differences. In general, children with lupus tend to have more severe and more aggressive disease than adult SLE patients, and childhood-onset SLE often presents with major organ system involvement, including renal and neuropsychiatric (NP) disease.[21,22] Among the general clinical characteristics, constitutional symptoms, such as unexplained fever, malaise and weight loss, present in 40–90% of cases, and are the most common manifestations in children and adolescents with SLE. These symptoms, especially in the context of an otherwise unexplained anemia and/or thrombocytopenia, should prompt a thorough investigation for lupus in this age group. Arthritis (60–70%) and rashes (40–60%) are also common clinical features of childhood-onset SLE; however, their prevalence is lower than in adult disease (80–95 and 60–80%, respectively). Raynaud's phenomenon is less common in childhood lupus patients and occurs in 10–20% of cases.[1–8] On the contrary, avascular necrosis is more common in children than in adults with lupus (10–15% of childhood lupus cases).[23] Table 2 summarizes the main clinical features of lupus in adult and child populations.

Renal Disease Nephritis is the primary disease manifestation in 60–80% of childhood-onset SLE patients, and represents the most significant determinant of prognosis and mortality within this population.[24,25] A higher frequency of aggressive renal disease, and thus a higher requirement for steroids and immunosuppressive drugs, has been reported among children with lupus compared with their adult counterparts.[3,4,6] To this end, a 78% incidence of renal disease in childhood-onset lupus compared with a 52% incidence in adults (p < 0.001), with a higher adjusted mean renal Systemic Lupus Erythematosus Activity Index (SLEDAI) and an increased use of immunosuppressive drugs and steroids within children with lupus, was reported by Brunner and colleagues in a cohort of 67 pediatric and 131 adult patients who were followed for 3.5 years.[7] Similarly, in the largest published cohort, which included 795 adult-onset and 90 childhood-onset lupus cases, with a follow-up of 16 and 13 years, respectively, 56% of children versus 37% of adults developed renal disease (p < 0.001).[8] This study reported an increased likelihood of children demonstrating active disease on all measures of disease activity (13 vs 8%, p < 0.001). Disease activity was measured using the Systemic Lupus Activity Questionnaire (SLAQ) and by the Medical Outcomes Study Short Form (PCS)-12. A higher prevalence of dialysis among childhood-onset patients compared with their adult counterparts were also reported (19 vs 5.7%; p < 0.001).[9]

As in adults, diffuse proliferative glomerulonephritis is the most frequent abnormality in children (present in 40–75% of cases). Proteinuria (60–70%) is the most common manifestation of renal disease, followed by microsopic hematuria (40–50%), hypertension (25%) and compromised renal function.[26] The most widely documented predictors of negative prognosis are renal scarring/increased chronicity index, impaired renal function at presentation, hypertension, diffuse proliferative lupus nephritis and nephrotic range proteinuria, and these apply to adult and pediatric lupus populations.[24,27–30] Importantly, studies conducted in developing countries have documented worse survival rates for pediatric lupus nephritis compared with developed countries.[31] Despite exhibiting a more aggressive course, pediatric lupus nephritis displays only slightly reduced 5-year survival rates compared with adult-onset disease (45–95 vs 70–95%, respectively) depending on ethnicity, patient selection and the severity of renal disease at presentation, a discrepancy that may be due to the lack of studies directly comparing renal survival between the two populations (Table 3).[24–31]

Mucocutaneous Manifestations Mucocutaneous manifestations are also common in children with lupus. Rash (40–60%), photosensitivity (35–50%) and oral ulcers (20–30%) appear to be equally frequent in children and adults, whereas alopecia tends to be more common in adults than in children (20–55 vs <30%, respectively). Vasculitic involvement of the hard palate may accompany the malar rash of pediatric lupus, and this is often used as a confirmatory sign to diagnose lupus when the diagnosis is questionable. Discoid lupus is unusual in childhood, and most children referred for discoid lupus erythematosus are found to have systemic disease manifestations when questioned appropriately. Approximately 25% of children with discoid lupus will progress to systemic disease.[32] The observed adult transition rates from discoid to systemic disease have been reported to be approximately 6%.[33]

Cardiopulmonary Manifestations Pulmonary involvement affects half of all patients at some point during their disease course and is part of the spectrum of the presenting symptoms in 4–15% of patients. Of note, according to some studies, up to 90% of adult lupus patients and 40% of children with SLE may develop subclinical pulmonary abnormalities.[34,35] Pleuritis causing dyspnea is the most common manifestation in both groups (35–50% of adults vs 15–40% of children) and bilateral small-to-moderate size pleural effusions are usually evident in this setting (35% of adults and 20% of children, respectively). The clinical spectrum of parenchymal pulmonary disease is wide and may include interstitial lung disease (ILD), diffuse alveolar hemorrhage (DAH), acute lupus pneumonitis (ALP) and acute respiratory distress syndrome (ARDS).[36] In a cohort of 60 Norwegian childhood lupus patients studied for 11 years, there was a 37% prevalence of abnormal pulmonary function tests (PFTs) and an 8% prevalence of abnormal high resolution computed tomography findings with no evidence of ILD, implying a low transition of asymptomatic functional and/or imaging abnormalities to substantial parenchymal alterations in childhood lupus.[37] In another cohort of 157 childhood lupus patients followed for 20 years in Taipei, the overall prevalence of pulmonary disease was 57%, with hospitalization requiring lung infection being the most prevalent pulmonary abnormality (46% of cases), which indicates the importance of prompt identification and treatment of infection in lupus patients who develop respiratory symptoms, particularly in the setting of immunomodulatory and/or high-dose steroid treatment. The same study reported a 14% overall prevalence of lupus pneumonitis, a 10% prevalence of pulmonary hemorrhages and a 4% prevalence of ILD. Positive anti-dsDNA antibodies at diagnosis were associated with pulmonary complications.[38] A restrictive pattern of pulmonary dysfunction, with characteristically reduced diffusing capacity, not necessarily correlating with symptoms or with other disease manifestations, is the most prevalent lung function test abnormality in both adults and children with SLE.[34,37] The prevalence of pulmonary arterial hypertension (PAH) in patients with lupus is largely unknown, but has been reported to approximate 6–15% in adult patients, in whom it is most commonly associated with Raynaud's phenomenon.[36] There have been very few studies addressing the prevalence of PAH in cSLE, which is reported to approximate 4–8% using transthoracic echocardiography.[38]

Pericarditis, myocarditis and mild valvular dysfunction are variably symptomatic in both populations.[39] A total of 10–15% of children and 20–25% of adults develop clinically evident myocarditis or pericarditis. Cardiac tamponade is a very rare complication (~2.5% in both populations).[40] Valvular abnormalities, arrhythmias/conduction defects and cardiomegaly are not uncommon. Of note, premature atherosclerosis is a major cause of morbidity and mortality in adult lupus patients, and Framingham-based calculated risk scores for cardiovascular disease (CVD) underestimate the real risk of CVD in patients with SLE.[41] In support of this, in a retrospective study, Esdaile et al. found that even after correction for all traditional risk factors, patients with SLE had a 10-fold higher risk of coronary artery disease (CAD) and a sevenfold higher risk of stroke compared with their sex- and age-matched counterparts.[42] Rahman et al. found that patients with SLE who developed CVD had fewer and less severe traditional risk factors than patients of similar age without SLE who developed CVD.[43] According to a study by Gazarian and colleagues, pediatric lupus patients exhibit a remarkably high prevalence of asymptomatic myocardial ischemia (16%).[44] One study of 157 patients, which was designed to describe and assess the incidence of cardiopulmonary disease in juvenile lupus, reported the incidence of established CAD to be approximately 4%, whereas atrial/ventricular hypertrophy and congestive heart failure were present in 11.5 and 7.5% of patients, respectively.[37] In a cohort of 137 pediatric lupus patients, the prevalence of lipid abnormalities was 100%, with increased lowdensity cholesterol levels evident in patients with high disease activity requiring high doses of steroids, and abnormal triglyceride levels mostly being linked to nephrotic range proteinuria.[45] Whether the cumulative effect of traditional risk factors, such as arterial hypertension, hyperinsulinemia, elevated levels of homocysteine, prolonged corticosteroid exposure and excessive inflammation or other as yet unidentified causes, is responsible for the increased CVD burden in lupus patients remains uncertain. Different factors may apply in the adult and pediatric age groups. The role of statins for primary prevention in both groups is being investigated.[46,47]

Neuropsychiatric Lupus The prevalence of NP involvement in SLE varies from 15–90% depending on diagnostic criteria and patient selection, with clinical symptoms varying from mild cognitive impairment to severe neurological and psychiatric features such as psychosis, stroke and seizures. NP involvement has been reported to be the third most common cause of mortality within the pediatric lupus population (mortality rates approximate 25%), and it is more common among pediatric lupus patients (20–45%) compared with the adult lupus population (10–25%).[48–51] NP events usually occur within 1 year of disease onset; however, 30% of children with lupus will develop neurological disease later on.[52] Psychosis with visual hallucinations is a hallmark of NP-SLE involvement in pediatric populations (12–40%), with seizures (20%) and antiphospholipid antibody-associated unilateral chorea being the most prevalent neurological signs. Cognitive dysfunction and mood disorders are also relatively common in NP childhood-onset SLE. Although headaches are frequent (50–75% of cases), the prevalence of the true 'lupus headache', that is, a headache associated with intracranial pathology of SLE, remains largely unknown. Importantly, MRI scanning may be normal in up to 50% of cases of children with psychosis, and electroencephalography is only helpful in the presence of seizures. The autoantibody profile maybe a helpful tool in assessing NP involvement in pediatric patients. Antiphospholipid antibodies have been associated with headaches, cerebrovascular accidents and transverse myelitis, and anticardiolipin antibodies have been associated with CNS vasculitis, cognitive impairment and psychosis. Antiribosomal P antibodies have been proposed as a surrogate marker for CNS disease, in particular depression and psychosis, but they may lack specificity. The association of antiphospholipid antibodies with seizures that has been documented in adult studies does not generally apply to pediatric populations, although the prevalence of these antibodies may be higher than that which had been documented in adult series.[53,54]

Studies in adults using the American College of Rheumatology (ACR) case definitions have collectively reported a wide spectrum of headaches (39–61%), seizures (8–18%), cerebrovascular disease (2–8%) and cranial neuropathy (1.5–2%) as being the most prevalent findings of NP lupus. The range in the prevalence of mood disorders and cognitive dysfunction is 12–75% and 5–40%, respectively (perhaps owing to methodological differences in their assessment), and no precise conclusions can therefore be drawn regarding their exact prevalence.[55] Small foci of high signal concentrated in subcortical and/or perivetricular white matter are the most common imaging findings in brain MRI scans, whereas reduced corpus callosum and cerebral volumes have been correlated with disease duration and cognitive impairment.[56] More recently, magnetic reasonance spectroscopy (MRS) has revealed neurometabolic abnormalities, even in white and grey matter appearing normal in conventional MRI, and these may correlate with cognitive impairment even in the absence of active NP-SLE. On the other hand, single photon emission computed tomography (SPECT) scanning may prove to be a useful imaging modality in assessing active NP disease in adults and children with SLE.[57]

Hematological Abnormalities & the Antiphospholipid Syndrome Autoimmune thrombocytopenia is present in 15% of cases of adults and children with lupus. It has been estimated, that 20–30% of children with idiopathic thrombopenic purpura (ITP) and positive antinuclear antibodies (ANAs) will eventually develop lupus.[58] However, anemia, usually the anemia of chronic disease, and leucopenia (notably lymphopenia) are more common in all age groups, whereas neutropenia is less frequent.

Primary antiphospholipid syndrome (APS) rarely progresses to lupus in adults (8% progression rate in a series of 128 patients who were followed-up for 9 years), although there is very limited data on the long-term outcome of pediatric APS.[59] Although the association of pathogenic antiphospholipid antibodies and clinical manifestations is clearest in adult patients with primary APS, there are no major differences in the clinical features between those patients and patients with underlying SLE. The most prominent clinical features of APS in adults include venous thrombosis (29–55%), pulmonary embolism (15–30%) and, less commonly, arterial thrombosis affecting the brain (50% of arterial thromboses), coronary arteries (27% of arterial events) or other otherwise infrequently affected vascular beds, such as the subclavian, renal, pedal, splenic and ocular arteries. Pregnancy loss during the fetal period (≥10 weeks of gestation) also occurs frequently in women with APS. The protean manifestations of APS in children include primarily immunologic thrombocytopenia and anemia, recurrent venous (60%) and arterial (30%) thrombosis and CNS manifestations (16%), mainly stroke and chorea. Comparable to adults, children with secondary APS also exhibit a higher likelihood of developing venous (vs arterial) thromboses, skin disease and nerurological complications.[60] Livedo reticularis, heart valve disease and pulmonary hypertension are less common in children than in adults with APS.[61] Pediatric patients with secondary APS tend to be older and to exhibit a higher frequency of venous versus arterial thrombotic events associated with skin and hematological manifestations compared with children with primary APS.[60] Several studies of adults with SLE have identified the presence of APS as an important predictor of irreversible organ damage and death, a finding that has not yet been confirmed in pediatric lupus populations.[62]

Uncommon Presentations A higher prevalence of more uncommon manifestations of SLE, such as gastrointestinal events with intestinal vasculitis and pancreatitis, have been described in pediatric lupus patients. We have reported an increased prevalence of biopsy-confirmed smooth muscle antibody-positive autoimmune liver disease that notably predated the disease onset in 9.8% of juvenile lupus patients compared with 1.3% of adults (p < 0.001). Interestingly, all the children exhibited histological changes consistent with autoimmune hepatitis, whereas, among adults, a variable autoantibody profile was documented, with one patient exhibiting histological evidence of primary biliary cirrhosis.[63]

Assessing Disease Activity, Disease Severity & Cumulative Disease Damage in Childhood- & Adult-onset SLE Fewer patients die from lupus and there is an overall increased life expectancy. Five-year survival rates have increased to 90% in both childhood and adult SLE, even allowing for ethnic group or socioeconomic differences in different studies. This fact, and the emerging and expanding new treatment options available, means that measuring the activity and the severity of the disease has become increasingly important. Furthermore, using objective indices to calculate cumulative disease damage or to identify treatment complications, is essential to assess morbidity across studies and in different centers. To this end, standardized measures of disease activity, severity and damage have been established and validated for both adult and child populations.

The SLEDAI, British Islet Lupus Assessment Group Index (BILAG), European Consensus Lupus Activity Measure (ECLAM) and Systemic Lupus Activity Measure (SLAM) indices are the most widely accepted measures for assessing disease activity in adults with SLE. In a study conducted by Brunner and colleagues, 35 childhood-onset lupus patients were followed up for 4 years, and the SLEDAI, BILAG and SLAM disease activity indices were tested for their reliability, validity and responsiveness (i.e., sensitivity to change). The authors concluded that all three measures of disease activity are highly sensitive to clinical change in children, with none showing an overall superiority.[64] More recently, the members of the Pediatric Rheumatology International Trials Organization (PRINTO) outlined a set of response criteria for use in trials of therapies for children with SLE. The final core set of criteria included a physician's global assessment of the disease activity, global disease activity measure, 24-h proteinuria, parent's global assessment of the patient's overall well being, and a health- related quality of life assessment, which were statistically validated as clinically and scientifically relevant.[65] Disappointingly, this group did not include analysis of the BILAG system. In our experience, the BILAG activity index, which measures activity and distinguishes eight different systems or organs and documents partial improvements or deteriorations, has proven to be the most useful instrument to guide clinical decisions. It captures change in a way that global scores do not. Thus, global scores are very 'black or white'; for example, if a patient has arthritis they will get the same number of points irrespective of whether the problem is getting better or worse. BILAG importantly distinguishes between partial improvement or deterioration.

The Systemic Lupus International Collaborative Clinics (SLICC)/ACR Damage Index (SDI) records damage in 12 organs and organ systems, with a score ranging from 0 to 47, and in order for an item to be endorsed as damaged, it must be persistently present for a minimum of 6 months. It has been shown to be a valid and reliable instrument of assessing permanent damage in adult SLE populations. Importantly, the renal and pulmonary scores have been shown to be predictors of adverse outcomes.[66] Rahman et al. showed that early damage as measured by the SDI score at 1 year postdiagnosis was associated with increased mortality in a prospective study of 230 patients who were followed-up for more than 10 years (25% of those with initial SDI damage died within 10 years compared with 7.3% of those with no initial SDI damage; p = 0.0002).[67] Total disease activity over time has the highest impact on the development of total damage, as demonstrated by Stoll et al. in a longitudinal study of 133 adult lupus patients who were followed prospectively for 4.6 years.[68] Analogous to adult studies, pediatric studies have revealed a relationship between cumulative disease activity over time and damage, confirming the validity of the use of the SDI score for childhood-onset SLE.[11,69] However, as discussed elsewhere, the SDI does not capture certain issues that are unique to children (e.g., retarded growth). In addition, the assessment of damage in childhood may also be made more difficult by the ability of children to recover and regenerate to a greater degree than adults, as is the case with (early) avascular necrosis or osteoporosis, both of which can be reversible once better control of the disease and restoration of the normal growth process are accomplished.[70] In view of these factors, a modified version of the SDI that includes additional criteria such as growth failure and delayed puberty has been proposed but requires further validation and improvement.[71,72]

Immunological Profiles of Childhood- versus Adult-onset Lupus

No laboratory feature of SLE in children or adults uniquely distinguishes the disease at different ages. In both, adults and children, the diagnosis is strongly suggested by the constellation of hypergammaglobulinemia, anemia, leucopenia, thrombocytopenia, hypoalbuminemia, increased erythrocyte sedimentation rate and a positive ANA test.

Most studies have shown no difference in the prevalence of ANA between adult- and childhood-onset lupus patients. We documented a higher prevalence of anti-dsDNA antibodies in children with SLE (85 vs 54%; p < 0.001),[4] a finding that was not, however, confirmed in some other studies directly comparing the serological profiles of adults and children with SLE.[3,6] Antibodies to ribosomal P protein, which have previously been implicated in SLE-associated psychosis and depression, have been shown to be more prevalent in childhood-onset compared with adult-onset disease (12–15 vs 40%).[73] Antineuronal, antiphospholipid antibodies and anti-Sm antibodies may also occur more frequently in childhood-onset SLE. Low C3 levels also appear to be more prevalent in childhood-onset lupus.[4]

Interestingly, a recent cluster autoantibody analysis study conducted in Toronto reported an increased prevalence of anti-U1RNP and anti-Sm antibodies within non-Caucasian populations. The same study identified three clusters of autoantibodies (anti-dsDNA, anti-dsDNA+antichromatin+antiribosomal P+anti-U1 RNP+anti-Sm+anti-Ro+anti-La and anti-dsDNA+anti-RNP+anti-Sm) associating with different clinical courses (mild disease with no major organ involvement, high frequency of nephritis/serositis/hemolytic anemia, and NP disease/nephritis, respectively).[74]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.