Severe disease presentation and poor outcomes among pediatric systemic lupus erythematosus patients in South Africa

Abstract

Background

Systemic lupus erythematosus (SLE) is a life-threatening multisystem autoimmune disease that is more severe in patients of African ancestry and children, yet pediatric SLE on the African continent has been understudied. This study describes a cohort of pediatric SLE (PULSE) patients in South Africa.

Methods

Patients with a diagnosis of SLE (1997 American College of Rheumatology criteria) diagnosed prior to age 19 years in Cape Town, South Africa, were enrolled in this cross-sectional study from September 2013 to December 2014. Information on clinical and serological characteristics was extracted from medical records. Results were compared to a well-described North American pediatric SLE cohort.

Results

Seventy-two South African patients were enrolled in the study; mean age 11.5 years; 82% were girls. The racial distribution was 68% Coloured, 24% Black, 5% White and 3% Asian/Indian. Most patients presented with severe lupus nephritis documented by renal biopsy (61%). Of patients with lupus nephritis, 63% presented with International Society of Nephrology/Renal Pathology Society class III or IV. Patients in the PULSE cohort were more likely to be treated with cyclophosphamide, methotrexate and azathioprine. The PULSE cohort had high disease activity at diagnosis (mean Systemic Lupus Erythematosus Disease Activity Index-2K (SLEDAI-2K) 20.6). The SLEDAI-2K at enrolment in the PULSE cohort (5.0) did not differ from the North American pediatric SLE cohort (4.8). Sixty-three per cent of the PULSE cohort had end organ damage with Systemic Lupus International Collaborating Clinics Damage Index (SLICC-DI) score >0 (mean SLICC-DI 1.9), compared to 23% in a previously reported US cohort. Within the PULSE cohort, nine (13%) developed end-stage renal disease with six (8%) requiring transplant, strikingly higher than North American peers (transplant rate <1%).

Conclusions

The PULSE cohort had highly active multiorgan disease at diagnosis and significant disease damage at enrolment in the South African registry. South African patients have severe lupus nephritis and poor renal outcomes compared to North American peers. Our study revealed a severe disease phenotype in the PULSE cohort resulting in poor outcomes in this high-risk population.

Keywords

Systemic lupus erythematosus pediatric lupus Africa lupus nephritis

Introduction

Systemic lupus erythematosus (SLE) is a life-threatening multisystem autoimmune disease. In North America and Europe, SLE is more common and more severe with increased mortality risk in people of African extraction compared with Caucasians.^1–4 Despite the disproportionate prevalence of SLE among individuals of African descent, literature describing the epidemiology, diagnosis and treatment of SLE among individuals living on the African continent is sparse.^5–12 Adult cohort studies from South Africa show poor outcomes, with five year survival rates ranging from 69% to 78%, compared to 88–95% in high-income countries. The only published pediatric SLE studies in sub-Saharan Africa have been reported from South Africa.^13–16 The paucity of data has led to perceptions that SLE is rare in Africa,^17,18 although more recent studies suggest under-recognition, under-diagnosis, under-reporting and poor access to care rather than rarity of disease.^19–21

Children are at higher risk for severe SLE, typically presenting with higher disease activity and continuing with higher activity over time than adults. Presentation in childhood occurs in 15–20% of all SLE patients, with median diagnosis age for pediatric SLE (pSLE) at 12–14 years.^22–24 Children are more likely to have serious renal, neurological and hematological involvement than their adult counterparts.²⁵ Mortality rates are higher in pSLE patients than adults, despite having fewer comorbidities.²⁶

Lupus nephritis (LN) is a serious complication of SLE, influencing prognosis and life expectancy more than any other organ system involvement.^27,28 SLE patients with LN show a 15–24 year decrease in life expectancy compared to SLE patients without nephritis.²⁹ In high income countries, patients of African ancestry are two to three times more likely to develop LN, with a more severe renal disease course,³⁰ and those with end-stage renal disease (ESRD) have higher mortality rates than Caucasians with ESRD.^31,32 Furthermore, LN is common in children.³³ pSLE patients have higher rates of renal damage and proceed to ESRD more frequently than their adult counterparts.²⁵ Treatment resistance disproportionately affects African American children.³⁴ In addition, LN is a major cause of morbidity and mortality for SLE patients in low income nations.³⁵

Thus, pSLE patients in Africa are potentially at high risk for poor outcomes based on race and age, yet strikingly little research has addressed this population.^9,20,21,36 To date, only one cohort study of pSLE in sub-Saharan Africa has been published and there has not been an updated report on pSLE in South Africa over the past 10 years.¹⁶

The aim of the current study is to determine if pSLE patients in South Africa have statistically significant and clinically important differences in disease severity, rates or severity of renal involvement, and disease damage compared to pSLE patients in a well-characterized North American cohort. We initiated a pSLE registry in South Africa and have performed an initial cross-sectional analysis at enrolment of pSLE patients presenting to care since 1988 at several hospitals in Cape Town, South Africa (referred to hereafter as the PULSE cohort). Here, we describe this cohort, compare it to patients in the Childhood Arthritis and Rheumatology Research Alliance (CARRA) Registry, a database that includes over 900 pSLE patients from 60 sites in North America,³⁷ and attempt to identify modifiable risk factors that could improve treatment and reduce organ damage in this population.

Methods

This retrospective study enrolled patients diagnosed with SLE prior to age 19 years at two tertiary care, government-funded university hospitals (Red Cross War Memorial Children’s Hospital; Groote Schuur Hospital) and one private medical center (Winelands Rheumatology Centre) in Cape Town, South Africa. The study was conducted as a collaboration between Duke University, USA and the University of Cape Town, South Africa, and approved by the Duke Institutional Review Board and the University of Cape Town Ethics Committee.

Participants

To develop the PULSE cohort, we identified eligible patients by searching the computerized databases of the two hospital centers where the study was conducted using International Classification of Disease version 10 codes for SLE or LN from hospital admissions and clinic visits. In addition, we polled physicians on site to generate a comprehensive list of pSLE patients. To include patients seen in private practice, we presented at national South African rheumatology meetings and sent recruitment flyers to all private rheumatology offices in the Western Cape. Inclusion criteria encompassed patients aged <19 years at SLE diagnosis who met the American College of Rheumatology (ACR) classification criteria for SLE,³⁸ and who had received care in the Western Cape Province, South Africa.

Once identified, living patients were approached for consent by their physician or study personnel. For patients less than 18 years old, we obtained parental consent; for patients aged 13 to 18 years, child assent was also obtained. Consent was conducted in the patient’s language of choice (Afrikaans, isiXhosa, or English). Deceased pSLE patients meeting enrolment criteria were included in review, for which a waiver of consent was obtained.

The CARRA Registry

For comparison, we used the existing CARRA Registrty. The CARRA Registry is a multicenter collaborative database that collects standardized information from 60 expert centers in North America.³⁹ Thus, we refer to the North American cohort as the total CARRA cohort. As of July 2012, there were over 900 children with pSLE enrolled. Patients with pSLE were eligible for inclusion in the CARRA Registry if they met at least four of 11 ACR criteria for SLE classification, developed pSLE prior to age 18 years, and were less than 21 years at the time of enrolment.⁴⁰

The following demographic features were obtained via chart review: age at disease diagnosis, gender, disease duration and self-reported race.⁴¹ In South Africa, race was reported as one of the five racial groups employed in South African population surveys: Coloured, Black, White, Indian/Asian, and other (see Appendix). As most of the PULSE cohort is of African descent, we compared those patients to the African American subjects of the CARRA cohort (AA CARRA) to match more evenly on race/ethnicity between cohorts.

Clinical features

We measured the presence of organ system involvement at time of enrolment. To measure disease activity, we employed a well-validated SLE disease activity index measure (the Systemic Lupus Erythematosus Disease Activity Index-2K (SLEDAI-2K)).⁴² We describe catastrophic clinical presentation in the South African patients. This is a term we created to capture the severe illness with which the majority of the PULSE cohort presented, including stroke, seizure, blindness, pericarditis causing tamponade, or acute renal failure. There are no data available on the frequency of such presentations in CARRA patients for comparison.

Laboratory tests and renal biopsy

We captured rates of positive tests for ANA, anti-Sm and anti-dsDNA antibodies. The anti-Ro, anti-La and anti-RNP antibodies have only recently become part of the routine assessment in South Africa, and were not included in this analysis.

In South Africa, only International Society of Nephrology and Renal Pathology Society (ISN/RPS)⁴³ histological classification is utilized; we therefore excluded CARRA patients with only World Health Organization biopsy classification from this comparison (n = 148).

Therapy

Inpatient and outpatient records were reviewed for current and prior therapy. All medications are reported as binary variables. The medication was scored as positive if the patient was receiving or had received the medication ever at study enrolment.

Outcomes

Through chart review, we extracted data on ESRD, dialysis, transplant and mortality. To measure chronic disease damage, we calculated Systemic Lupus International Collaborating Clinics/ACR Damage Index for systemic lupus erythematosus (SLICC-DI) scores.⁴⁴

Statistical analysis

We compared age of SLE diagnosis, race, presenting features, rate and class of LN, laboratory features, SLEDAI-2K and SLICC-DI scores, treatment and outcomes between the South African and North American patient cohorts. Characterization of the patient population was summarized using descriptive statistics with 95% confidence intervals of means and proportions for cohort comparison.

We evaluated enrolment SLEDAI-2K scores of the PULSE cohort compared to the CARRA cohort. Where appropriate, Kruskall–Wallis, Pearson, χ² or Fisher’s exact test were used to evaluate differences in both clinical and demographic features and disease scores of the two cohorts. For better comparison of subjects of African ancestry between cohorts, we compared the PULSE cohort to the African American subjects in the CARRA cohort (AA CARRA).

As an exploratory aim, we attempted to identify risk factors for poor outcomes within the PULSE cohort. Unfortunately, the PULSE cohort was not large enough to provide sufficient power for this sub-analysis.

All calculations were performed using STATA statistical software (Stata Corp., College Station, TX, USA).

Results

Enrolment characteristics are summarized in Table 1. As expected for patients from the African continent, 92% of the patients in the PULSE cohort are of African descent, versus 35% of the CARRA cohort. A larger proportion of the CARRA cohort (64% vs. 46%) had a diagnosis of SLE for at least 2 years at the time of enrolment. Seventy-three per cent of the PULSE cohort had age of onset less than or equal to 13 years of age at diagnosis, compared to only 48% of the CARRA cohort.

Table 1

Demographic features at enrolment

	PULSE N = 72	Total CARRA N = 982	P value
Mean age of SLE diagnosis, years (SD)	11.5 (3.5)	12.4 (3.2)	0.027
Mean disease duration, years (SD)	2.4 (3.2)	3.5 (3.0)	0.001
% Female (n)	82 (59)	83 (810)	0.873
% White (n)	6 (4)	47 (459)	<0.001
% African descent (n)	92 (66)	35 (339)	<0.001
% Other (n)	3 (2)	19 (184)	<0.001
% Age ≤13 at diagnosis (n)	73 (53)	48 (475)	<0.001
% Disease duration ≥2 years (n)	46 (33)	64 (632)	0.003

CARRA: Childhood Arthritis and Rheumatology Research Alliance Registry.

There are important differences in clinical and serological features of the cohorts (Figure 1, Table 2). Notably, a much higher percentage of the PULSE cohort had LN at enrolment compared to the AA CARRA group (61% vs. 42%, P < 0.001). The PULSE cohort demonstrated high disease activity at the time of diagnosis, represented by an average SLEDAI-2K score of 20.5; however, the score was recorded for only 44 of 72 PULSE subjects, and must be interpreted with caution due to the large amount of missing data. The CARRA Registry does not capture the SLEDAI-2K score at time of diagnosis. On chart review of the PULSE subjects, we noted that more than half of the patients (57%) presented with emergencies, such as acute renal failure, pericarditis causing tamponade, or stroke. There were 51 catastrophic events recorded among 72 patients. Seven patients (10%) had a severe disease manifestation in more than one category (i.e. presenting with stroke and acute renal failure). More than one third of patients in the PULSE cohort (37%) had acute kidney injury with decreased renal function at the time of diagnosis, 5% stroke, 4% deep vein thrombosis, 5% pericarditis requiring intervention and 3% seizure. At time of enrolment, the average SLEDAI-2K score of the PULSE cohort was not different from that of the total CARRA subjects or the AA CARRA subjects.

Figure 1

Clinical Features at Enrollment.

Table 2

Catastrophic presentation and SLEDAI at enrolment

	PULSE N = 72	Total CARRA N = 982	P value	AA CARRA N = 339	P value
% Catastrophic presentation (n)	57 (40)^a	NR	–	NR	–
Average SLEDAI-2K at diagnosis^b (SD)	20.6 (9.9)	NR	–	NR	–
Average SLEDAI-2K at enrolment^b (SD)	5.0 (5.2)	4.8 (5.7)	0.437	5.4 (6.5)	0.960

Catastrophic presentations include acute kidney injury, seizure, deep vein thrombosis, pericarditis requiring intervention, stroke, cytopenia requiring therapy, and blindness.

Range 0–105.

NR: not reported; SLEDAI-2K: Systemic Lupus Erythematosus Disease Activity Index-2K; CARRA: Childhood Arthritis and Rheumatology Research Alliance Registry.

PULSE patients displayed an increased rate of severe renal disease compared to North American peers (Figures 1 and 2). The rate of biopsy confirmed LN (61%) in the PULSE cohort is higher than the CARRA cohort (42%). In comparison with CARRA, nearly all biopsied patients in the PULSE cohort (93%) had ISN/RPS class III, IV, V, or VI LN. Notably, 63% of those biopsied in the PULSE cohort had class III or IV LN, while only 48% of the CARRA cohort had these findings on biopsy. Moreover, 7% of the PULSE cohort had class VI LN on renal biopsy at enrolment, while none of the total or AA CARRA cohort demonstrated advanced sclerosis.

Figure 2

Features of Lupus Nephritis at Enrollment.

We compared the medical management of patients in each cohort (Table 3). Notably, a much larger number of patients in the CARRA cohort had been treated with an antimalarial and mycophenolate mofetil at the time of enrolment. The PULSE cohort had a higher percentage of patients treated with azathioprine and methotrexate. The PULSE cohort also had a higher percentage of patients treated with intravenous immunoglobulin (IVIg).

Table 3

Treatment history at enrolment

	PULSE N = 72	Total CARRA N = 982	P value	AA CARRA N = 339	P value
% Treated with steroids (n)	92 (66)	94 (917)	0.725	94 (320)	0.538
% Treated with antimalarial (n)	75 (54)	87 (853)	0.001	88 (297)	<0.001
% Treated with cyclophosphamide (n)	42 (30)	27 (267)	0.022	30 (101)	0.061
% Treated with mycophenolate mofetil (n)	38 (28)	57 (556)	0.004	64 (205)	<0.001
% Treated with azathioprine (n)	56 (41)	15 (146)	0.001	14 (47)	<0.001
% Treated with methotrexate (n)	26 (19)	19 (184)	0.019	17 (59)	0.082
% Treated with rituximab (n)	6 (4)	10 (99)	–	14 (49)	–
% Treated with IVIg (n)	11 (8)	2 (17)	–	6 (21)	–

AA: African American; CARRA: Childhood Arthritis and Rheumatology Research Alliance Registry.

We identified a high rate of irreversible damage in the PULSE cohort (Table 4), with an average SLICC-DI score of 1.9. The PULSE cohort demonstrates poor renal outcomes, with 15% requiring dialysis as renal replacement therapy (RRT), versus 1% in the CARRA cohort, and 1.5% in the AA CARRA cohort. South African pSLE patients with LN also had much higher rates of renal transplant than both the total and AA CARRA subjects.

Table 4

Measures of disease damage and renal outcomes

	PULSE N = 72	Total CARRA N = 982	P value	AA CARRA N = 339	P value
% ESRD (n)	13 (9)	NR	–	NR	–
% Requiring dialysis (n)	15 (11)	1 (11)	<0.001	1.5 (5)	<0.001
% Requiring transplant (n)	8 (6)	0.7 (7)	<0.001	0.3 (1)	<0.001
SLICC score, median	1.9	NR	–	NR	–

AA: African American; CARRA: Childhood Arthritis and Rheumatology Research Alliance Registry; ESRD: end-stage renal disease; NR: not reported; SLICC: Systemic Lupus International Collaborating Clinics (Damage Index).

Discussion

This study describes a cohort of 72 pSLE patients diagnosed before the age of 19 years in two medical centers and one office practice in Cape Town, South Africa (PULSE). This is the largest cohort of pSLE patients described in Africa to date. The only previous study of pSLE from sub-Saharan Africa was a retrospective cohort of 36 patients who presented to a Johannesburg hospital between 1974 and 2000.¹⁶ The previous report had a much higher proportion of white patients (42%) than the current PULSE cohort, probably reflecting shifts in recognition, diagnosis and access to subspecialty care for black and coloured patients in South Africa since the end of apartheid in 1994.⁴⁵ The PULSE cohort reflects the racial composition of Western Cape Province (32% Black, 49% Coloured, 16% White, 1% Indian/Asian, 2% other) but is different from South Africa as a whole (80% Black, 9% Coloured, 2.5% Indian/Asian, 8.5% White).⁴⁶

The PULSE cohort is similar to the CARRA cohort in gender distribution, but differs in important measures of race, age, specific antibody profile and disease duration. Previous work suggests that young age and black race³⁷ predispose to increased disease activity in pSLE patients, and may contribute to the severe disease phenotype seen in our PULSE cohort. We compared the PULSE cohort to the African American patients in the CARRA cohort for a comparison based on self-reported race. Although the CARRA registry uses the eligibility criterion of SLE diagnosed prior to age 18 years, the present study includes all South African patients diagnosed prior to age 19 years. We extended the cut-off age to include those diagnosed at age 18 years to accommodate South African physician opinion that diagnosis was often delayed (C. Scott, June 2013; personal communication).

We observed increased clinical severity at presentation, higher rates of LN, and worse outcomes, as measured by SLICC-DI and the need for renal dialysis and transplant, among pSLE patients in South Africa. The PULSE cohort had higher rates of every disease manifestation examined, but the largest differences were seen in rates of renal and central nervous system involvement, which have the most impact on patient health and prognosis,^34,47,48 and heavily influence treatment choice in pSLE.^49,50 The high SLEDAI-2K score at diagnosis in the PULSE cohort indicates high disease activity at the time of presentation. The CARRA cohort study does not capture the SLEDAI-2K score at presentation, and data on SLEDAI-2K at presentation among diverse races/ethnicities are limited. One previous study of a well-described cohort of North American patients reports a mean SLEDAI-2K score of 13.8 in Caucasian patients and 12.6 in non-Caucasians at diagnosis, although black patients comprised only 15% of patients in that study.⁵¹ The high SLEDAI-2K score and increased rates of all disease manifestations suggest a more severe phenotype in the PULSE cohort. Notably, the majority of South African patients had a catastrophic presentation, presenting with severe life-threatening manifestations such as acute renal failure requiring dialysis, blindness, or hemorrhagic stroke. These findings add to the evidence that South African patients have severe active SLE at diagnosis. It is not possible to determine if barriers to diagnosis and care influence the severity of disease on presentation in this cross-sectional cohort study; however, in-depth interviews with patient families suggest that this is likely to be a contributing factor (unpublished data). The SLEDAI-2K score at diagnosis was missing for more than half of PULSE subjects. For those in South Africa who had an enrolment SLEDAI score recorded, it did not differ from that of the CARRA cohort. Although we cannot infer causality in this retrospective study, this finding suggests that PULSE patients respond to therapy despite high activity at presentation.

The reported rate of LN in children with SLE is high (ranging from 50% to 67%) and patients of African descent in high income nations are two to three times more likely to develop lupus, more likely to have severe disease, and more likely to progress to ESRD than Caucasian counterparts.^31,32,47 We found the rate of biopsy confirmed LN to be 61% in the PULSE cohort. The majority of biopsied patients in the PULSE cohort had class III–IV LN on biopsy at enrolment, while fewer than half of biopsied CARRA patients had class III or IV LN. Also, a percentage of biopsied patients in the PULSE cohort had advanced sclerosis, while no one in the CARRA cohort demonstrated this at the time of enrolment. One third of the PULSE cohort had acute kidney injury (AKI) at diagnosis. There is no information from the CARRA cohort on AKI for comparison.

The severity of LN in the PULSE cohort directly translates into poor outcomes. The PULSE cohort had high rates of ESRD, dialysis use and renal transplant at enrolment, in stark contrast to the much lower rate requiring dialysis and transplant in the AA CARRA cohort. Higher rates of dialysis and transplant in the PULSE cohort are not surprising given the higher rates of LN. However, the South African medical system is resource-limited with renal replacement and transplant therapy tightly rationed, requiring application and approval.⁵² In context of the hurdles for access to these treatments, the increased rate in the PULSE cohort is especially striking.

Disease management differs between the PULSE and CARRA cohorts. Although antimalarial therapy is standard care worldwide, the PULSE cohort demonstrated lower use than the AA CARRA cohort. Antimalarial therapy is associated with higher rates of remission, fewer relapses and reduced organ damage.⁵³ It is unclear if this difference is due to lack of provider knowledge, drug cost, or adherence. The PULSE cohort was more frequently treated with cyclophosphamide, methotrexate and azathioprine, and less often with mycophenolate mofetil than the North American comparators. Treatment differences are impacted by medication rationing practices in South Africa. Therefore, the increased use of IVIg is surprising. However, 20% of the PULSE cohort initially presented with severe idiopathic thrombocytopenic purpura (ITP), which is commonly managed with IVIg in both North America and South Africa. It is not possible to determine if medication management differences resulted in differences in disease damage or outcomes in this cross-sectional study.

The PULSE cohort had significant organ damage, as reflected in SLICC-DI scores. The SLICC-DI score was not recorded in the CARRA database, so direct comparison between the two cohorts is not possible. Instead, we compared the PULSE SLICC-DI scores to those published in the Atherosclerosis Prevention in Pediatric Lupus Erythematosus (APPLE) cohort study, in which 25% of the 221 patients had a SLICC-DI score greater than zero.⁵⁴ This is vastly less than the 63% of the PULSE cohort with a SLICC-DI score greater than zero at enrolment. Higher SLICC-DI scores correlate with mortality in populations in high and low income nations.⁵⁵ Thus, high SLICC-DI scores indicate high morbidity and predict high mortality risk in the PULSE cohort.

The mortality rate in the PULSE cohort was 9.7%. Five year survival for pSLE patients has been reported as greater than 95%;³³ however, reports from other lower income nations show a poorer prognosis.^56,57 Most deaths (six/seven) in the PULSE cohort were related to renal disease, whereas in the USA, SLE deaths are most commonly due to infections.⁵⁸ This contrast reinforces the importance of LN and poor access to RRT and transplant on prognosis in South Africa.

This study is limited by the relatively small number of patients enrolled in the PULSE cohort. Although it is the largest cohort in South Africa to date, the small total number of patients does not provide statistical power to look for meaningful subgroup differences within the cohort. The study is cross-sectional and limited by the documented medical chart. In addition, the study is restricted to one city in South Africa, and most patients were recruited from two major academic medical centers. This may limit generalizability, although the majority of CARRA participants are recruited from academic centers as well. The study uses standard local definitions of self-reported race; however, race is a social, not a biological construct, and patients are not readily separated into distinct categories. Our study included patients of African ancestry from two continents. The admixture of Africans, Caucasians and patients of other ancestry is arbitrary and inexact,⁵⁹ therefore comparison between populations is difficult. The specific racial contributions of South African black and coloured patients are different from African Americans. The South African black population is derived from Bantu tribes and the coloured population has indigenous Khoisan roots, with an admixture of Dutch, British and Asian ancestry.⁶⁰ The majority of African Americans have ancestral links to West Africa⁶¹ and the genetic diversity of Africa is reportedly vast.⁶² Nevertheless, SLE research has shown race to be a risk factor for severe disease.^3,4,32,37,47 The only pediatric rheumatologists in sub-Saharan Africa are in South Africa; therefore, we rely on these data for the first glimpse into pSLE in the larger region.²⁰

In summary, the PULSE cohort demonstrates high disease activity at diagnosis and increased rates of LN, ESRD and organ damage at enrolment compared to North American cohorts. The high percentage of patients presenting with AKI suggests delayed diagnosis may play a role in the poorer outcomes seen in the PULSE cohort. The differences in outcome may be due to racial/ethnic predisposition to severe disease, late presentation due to barriers to health care access, or a combination of these and other factors. These differences were apparent even when compared to a North American cohort of African descent. Improving early recognition, diagnosis and access to treatment for pSLE patients in South Africa may lead to reduced organ damage. In addition, early aggressive therapy of LN may lower rates of ESRD requiring either RRT or renal transplant. Given the limited resource of dialysis and transplant, preventing primary damage in this cohort is especially important.

Further prospective research is necessary to gain a better understanding of the burden of pSLE in South Africa. A larger, prospective study is necessary to increase the cohort of patients with pSLE, determine management, and identify risk factors for poor outcomes. Translational research investigating biomarkers of lupus severity or genetic markers for increased disease severity may help to identify patients at high risk for severe disease in both low and high income nations. Future qualitative studies examining this population could help determine the role of barriers to care, which may in part explain disease severity at onset. Identifying risk factors and barriers to care is necessary to inform interventions to improve outcomes in this high-risk pediatric population.

Footnotes

Acknowledgements

The authors thank Melissa Watt and Coleen Cunningham for their thoughtful review of this work. Larry Park and Joseph Egger of Duke Global Health Institute assisted with statistical analysis. CARRA provided access to the North American SLE database. The authors would like to acknowledge the research team at Red Cross War Memorial Children’s Hospital, including Dr Kate Webb, Dr Lawrence O’kongo, Dr Nicola Brice, Sr Dorothy Brown, Sr Caroline and Zodwa Sam. Bradley Otterson, biomedical librarian/informationist, NIH Library at National Institutes of Health provided critical reading of this manuscript. This article was prepared while Laura B. Lewandowski was employed at Duke University Medical Center. The opinions expressed in this article are the author's own and do not reflect the view of the National Institutes of Health, the Department of Health and Human Services, or the United States government.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: LL was funded by the Lupus Foundation of America Early Career Award, T32 Training grant 5T32 AI0007217-31, Duke Global Health Institute Fieldwork grant and the Fogarty International Center of the NIH R25TW009337. The CARRA Registry was supported by grants from NIAMS RC2AR058934, Friends of CARRA and the Arthritis Foundation.

Appendix

References

Uribe

McGwin

Jr. Reveille

Alarcon

. What have we learned from a 10-year experience with the LUMINA (Lupus in Minorities; Nature vs. nurture) cohort? Where are we heading? Autoimmun Rev 2004; 3: 321–329.

Cervera

Doria

Amoura

. Patterns of systemic lupus erythematosus expression in Europe. Autoimmun Rev 2014; 13: 621–629.

Danchenko

Satia

Anthony

. Epidemiology of systemic lupus erythematosus: a comparison of worldwide disease burden. Lupus 2006; 15: 308–318.

Gomez-Puerta

Barbhaiya

Guan

Feldman

Alarcon

Costenbader

. Racial/ethnic variation in all-cause mortality among United States medicaid recipients with systemic lupus erythematosus: a Hispanic and asian paradox. Arthritis Rheum 2015; 67: 752–760.

Taylor

Stein

. Systemic lupus erythematosus in Zimbabwe. Ann Rheum Dis 1986; 45: 645–648.

Adelowo

Oguntona

. Pattern of systemic lupus erythematosus among Nigerians. Clin Rheumatol 2009; 28: 699–703.

Okpechi

Ayodele

Jones

Duffield

Swanepoel

. Outcome of patients with membranous lupus nephritis in Cape Town South Africa. Nephrol Dial Transplant 2012; 27: 3509–3515.

Diallo

Kane

. Systemic lupus erythematosus and lupus syndromes in Senegal. A retrospective study of 30 patients seen over 10 years. Rev Rhum Engl Ed 1998; 65: 471–476.

el-Garf

Salah

. Juvenile systemic lupus erythematosus among Egyptian children. J Rheumatol 1990; 17: 1168–1170.

10.

Bakr

. Epidemiology treatment and outcome of childhood systemic lupus erythematosus in Egypt. Pediatr Nephrol 2005; 20: 1081–1086.

11.

Salah

Lotfy

Sabry

El Hamshary

Taher

. Systemic lupus erythematosus in Egyptian children. Rheumatol Int 2009; 29: 14–18.

12.

Abdel-Hafez

Abdel-Nabi

. Juvenile systemic lupus erythematosus: onset patterns and short-term outcome in Egyptian children, a single-center experience. Lupus 2015; 24: 1455–1461.

13.

Ransome

Thomson

. Systemic lupus erythematosus with nephritis in children. A report of 6 cases. South Afr Med J: Suid-Afrikaanse tydskrif vir geneeskunde 1986; 69: 629–633.

14.

Rovers

Coovadia

. Systemic lupus erythematosus in children. A report of 3 cases. South Afr Med J: Suid-Afrikaanse tydskrif vir geneeskunde 1981; 60: 711–713.

15.

Coovadia

Hussain

Mwelase

. Systemic lupus erythematosus in a black South African child. First documented case report. South Afr Med J: Suid-Afrikaanse tydskrif vir geneeskunde 1991; 79: 101–103.

16.

Faller

Thomson

Kala

Hahn

. Demographics and presenting clinical features of childhood systemic lupus erythematosus. South Afr Med J: Suid-Afrikaanse tydskrif vir geneeskunde 2005; 95: 424–427.

17.

Kumar

Chambers

Gordon

. Challenges of ethnicity in SLE. Best Practice & Res Clin Rheumatol 2009; 23: 549–561.

18.

Clark

al-Yaman

Cowden

Rockett

. Does malarial tolerance, through nitric oxide, explain the low incidence of autoimmune disease in tropical Africa? Lancet 1996; 348: 1492–1494.

19.

Bae

Fraser

Liang

. The epidemiology of systemic lupus erythematosus in populations of African ancestry: a critical review of the “prevalence gradient hypothesis”. Arthritis Rheum 1998; 41: 2091–2099.

20.

Tiffin

Hodkinson

Okpechi

. Lupus in Africa: can we dispel the myths and face the challenges? Lupus 2014; 23: 102–111.

21.

Pan

. A bibliometric study of literature on SLE research in PubMed (2002–2011). Lupus 2013; 22: 772–777.

22.

Brunner

Huggins

Klein-Gitelman

. Pediatric SLE – towards a comprehensive management plan. Nat Rev Rheumatol 2011; 7: 225–233.

23.

Watson

Leone

Pilkington

. Disease activity, severity, and damage in the UK Juvenile-Onset Systemic Lupus Erythematosus Cohort. Arthritis Rheum 2012; 64: 2356–2365.

24.

Papadimitraki

Isenberg

. Childhood- and adult-onset lupus: an update of similarities and differences. Exp Rev Clin Immunol 2009; 5: 391–403.

25.

Mina

Brunner

. Pediatric lupus – are there differences in presentation, genetics, response to therapy, and damage accrual compared with adult lupus? Rheum Dis Clin North Am 2010; 36: 53–80,vii–viii.

26.

Ardoin

Schanberg

. Paediatric rheumatic disease: lessons from SLE: children are not little adults. Nat Rev Rheumatol 2012; 8: 444–445.

27.

Maroz

Segal

. Lupus nephritis and end-stage kidney disease. Am J Med Sci 2013; 346: 319–323.

28.

Inda-Filho

Neugarten

Putterman

Broder

. Improving outcomes in patients with lupus and end-stage renal disease. Semin Dial 2013; 26: 590–596.

29.

Mok

Kwok

Yip

. Effect of renal disease on the standardized mortality ratio and life expectancy of patients with systemic lupus erythematosus. Arthritis Rheum 2013; 65: 2154–2160.

30.

Ward

. Recent clinical trials in lupus nephritis. Rheum Dis Clin North Am 2014; 40: 519–535,ix.

31.

Sule

Fivush

Neu

Furth

. Increased risk of death in African American patients with end-stage renal disease secondary to lupus. Clin Kidney J 2014; 7: 40–44.

32.

Nee

Martinez-Osorio

Yuan

. Survival disparity of African American versus non-African American patients with ESRD due to SLE. Am J Kidney Dis 2015; 66: 630–637.

33.

Hiraki

Benseler

Tyrrell

Hebert

Harvey

Silverman

. Clinical and laboratory characteristics and long-term outcome of pediatric systemic lupus erythematosus: a longitudinal study. J Pediatr 2008; 152: 550–556.

34.

Gibson

Gipson

Massengill

. Predictors of relapse and end stage kidney disease in proliferative lupus nephritis: focus on children, adolescents, and young adults. Clin J Am Soc Nephrol: CJASN 2009; 4: 1962–1967.

35.

Tikly

Navarra

. Lupus in the developing world – is it any different? Best Practice & Res Clin Rheumatol 2008; 22: 643–655.

36.

Borchers

Naguwa

Shoenfeld

Gershwin

. The geoepidemiology of systemic lupus erythematosus. Autoimmun Rev 2010; 9: A277–A287.

37.

Rullo OJ, McKurdy DM, Yadin O, et al. Race, ethnicity, and gender affect the severity of renal outcomes in patients with pediatric systemic lupus erythematosus: an analysis of the CARRAnet data at baseline visit [abstract]. American College of Rheumatology/ARHP Scientific Meeting. Chicago, IL: Arthritis and Rheuma, 2011; 63: S781–S782.

38.

Hochberg

. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1997; 40: 1725–1725.

39.

Natter MD, Winsor J, Fox K, et al. The Childhood Arthritis &Rheumatology Research Alliance Network Registry: Demographics and Characteristics of the Initial One Year Cohort [abstract]. American College of Rheumatology/AHRP Scientific Meeting. Chicago, IL: Arthritis Rheum, 2011.

40.

Boneparth

Ilowite

. CARRA Registry Investigators. Comparison of renal response parameters for juvenile membranous plus proliferative lupus nephritis versus isolated proliferative lupus nephritis: a cross-sectional analysis of the CARRA Registry. Lupus 2014; 23: 898–904.

41.

Lin

Kelsey

. Use of race and ethnicity in epidemiologic research: concepts, methodological issues, and suggestions for research. Epidemiol Rev 2000; 22: 187–202.

42.

Hawker

Gabriel

Bombardier

Goldsmith

Caron

Gladman

. A reliability study of SLEDAI: a disease activity index for systemic lupus erythematosus. J Rheumatol 1993; 20: 657–660.

43.

Markowitz

D’Agati

. The ISN//RPS 2003 classification of lupus nephritis: an assessment at 3 years. Kidney Int 2007; 71: 491–495.

44.

Stoll

Seifert

Isenberg

. SLICC/ACR Damage Index is valid, and renal and pulmonary organ scores are predictors of severe outcome in patients with systemic lupus erythematosus. Br J Rheumatol 1996; 35: 248–254.

45.

Lewandowski L, Scott C. Apartheid and health care access for pediatric systemic lupus erythematosus patients in South Africa. South Afr J Child Health. 2015; 9: 36.

46.

Statistics South Africa. Mid-year Population Estimates 2014. 2014; Statistical release P0302, www.statssa.gov.za/publications/P0302/P03022014.pdf.

47.

Hagelberg

Lee

Bargman

. Longterm followup of childhood lupus nephritis. J Rheumatol 2002; 29: 2635–2642.

48.

Zirkzee

Huizinga

Bollen

. Mortality in neuropsychiatric systemic lupus erythematosus (NPSLE). Lupus 2014; 23: 31–38.

49.

Mina

von Scheven

Ardoin

. Consensus treatment plans for induction therapy of newly diagnosed proliferative lupus nephritis in juvenile systemic lupus erythematosus. Arthritis Care Res (Hoboken) 2012; 64: 375–383.

50.

Hanly

. Diagnosis and management of neuropsychiatric SLE. Nat Rev Rheumatol 2014; 10: 338–347.

51.

Hiraki

Benseler

Tyrrell

Harvey

Hebert

Silverman

. Ethnic differences in pediatric systemic lupus erythematosus. J Rheumatol 2009; 36: 2539–2546.

52.

Moosa

Kidd

. The dangers of rationing dialysis treatment: the dilemma facing a developing country. Kidney Int 2006; 70: 1107–1114.

53.

Ruiz-Irastorza

Khamashta

. Hydroxychloroquine: the cornerstone of lupus therapy. Lupus 2008; 17: 271–273.

54.

Ardoin

Schanberg

Sandborg

. Laboratory markers of cardiovascular risk in pediatric SLE: the APPLE baseline cohort. Lupus 2010; 19: 1315–1325.

55.

Nived

Jonsen

Bengtsson

Sturfelt

. High predictive value of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology damage index for survival in systemic lupus erythematosus. J Rheumatol 2002; 29: 1398–1400.

56.

Singh S, Devidayal, Kumar L, Joshi K. Mortality patterns in childhood lupus – 10 years’ experience in a developing country. Clin Rheumatol 2002; 21: 462–465.

57.

Appenzeller

Marini

Costallat

. Damage did not independently influence mortality in childhood systemic lupus erythematosus. Rheumatol Intl 2005; 25: 619–624.

58.

Ravelli

Ruperto

Martini

. Outcome in juvenile onset systemic lupus erythematosus. Curr Opin Rheumatol 2005; 17: 568–573.

59.

Osborne

Feit

. The use of race in medical research. JAMA 1992; 267: 275–279.

60.

Wit

Delport

Rugamika

. Genome-wide analysis of the structure of the South African coloured population in the Western Cape. Human Genet 2010; 128: 145–153.

61.

Yaeger

Avila-Bront

Abdul

. Comparing genetic ancestry and self-described race in African Americans Born in the United States and in Africa. Cancer Epidemiol Biomarkers & Prev 2008; 17: 1329–1338.

62.

Tishkoff

Williams

. Genetic analysis of African populations: human evolution and complex disease. Nat Rev Genet 2002; 3: 611–621.

63.

Posel

. Race as common sense: racial classification in twentieth-century South Africa. African Studies Rev 2001; 44: 87–113.