Treatment patterns in paediatric and adult patients with SLE: a retrospective claims database study in the USA

Abstract

Objective

To assess real-world treatment regimens and patterns in childhood-onset SLE (cSLE) and adult-onset SLE (aSLE) cohorts, including similarities in treatments, duration of use and adherence.

Methods

This retrospective study utilised data from Merative L.P. MarketScan Research Databases (USA). Index date was the date of first SLE diagnosis (2010–2019). Patients aged <18 years (cSLE) and ≥18 years (aSLE) at index date with confirmed SLE diagnosis and ≥12 months continuous enrolment during pre-index and post-index periods were included. The cohorts were stratified based on the presence (existing) or absence (new) of pre-index SLE. Primary outcomes (post-index period) included treatment regimens (all patients), and adherence (proportion of days covered (PDC)) and discontinuation of therapies initiated within 90 days of diagnosis (new patients). Univariate comparisons between cSLE and aSLE cohorts were performed using Wilcoxon rank-sum and χ² (or Fisher’s exact) tests.

Results

cSLE cohort included 1275 patients (mean age=14.1 years) and aSLE cohort included 66 326 patients (mean age=49.7 years). Antimalarials and glucocorticoids were commonly used among new (cSLE=64.4%/62.0%; aSLE=51.8%/49.7%) and existing (cSLE=68.6%/58.9%; aSLE=63.8%/51.3%) patients in both cohorts. Median oral glucocorticoid dose (prednisone equivalent) was higher in cSLE vs aSLE (new=22.1 vs 14.0 mg/day; existing=14.4 vs 12.3 mg/day; p<0.05). Mycophenolate mofetil use was higher in patients with cSLE vs aSLE (new=26.2% vs 5.8%; existing=37.6% vs 11.0%; p<0.0001). Compared with aSLE, more patients used combination therapies in cSLE (p<0.0001). Median PDC was higher in cSLE vs aSLE for antimalarials (0.9 vs 0.8; p<0.0001) and oral glucocorticoids (0.6 vs 0.3; p<0.0001). Treatment discontinuation was lower in cSLE vs aSLE for antimalarials (25.0% vs 33.1%; p<0.0001) and oral glucocorticoids (56.6% vs 71.2%; p<0.0001).

Conclusions

Management of cSLE and aSLE includes the same medication classes; differences include more intensive use of therapy in cSLE, warranting the need for approved safe medications for cSLE.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• SLE pathophysiology is similar in children and adults. However, patients with childhood-onset SLE (cSLE) often present with higher disease activity and more severe clinical manifestations. Existing treatments for adult-onset SLE (aSLE) are used off-label in cSLE, given the limited number of approved drugs for children.

WHAT THIS STUDY ADDS

• Our study identified treatment patterns in cSLE from a claims database. It provides an overview of SLE disease management patterns and the types of therapies initiated among newly diagnosed patients in both cSLE and aSLE cohorts. Although the same medication classes are being used for treatment of cSLE and aSLE, treatment patterns may differ between the two cohorts.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The more common severe clinical course of cSLE may explain higher overall use of therapy in patients with cSLE compared with aSLE in our study. The results further support the need for approved drugs and the use of extrapolation approach in paediatric development to improve labelled treatment options in cSLE.

Introduction

SLE is a chronic autoimmune disease impacting multiple organ systems with heterogeneous manifestations ranging from fatigue and mild skin rash to end-stage renal failure.^{1 2}

The clinical course of SLE is marked by periods of quiescence followed by flares and active disease.^{3 4} Childhood-onset SLE (cSLE), defined as disease onset in patients aged <18 years, affects approximately 15%–20% of patients with SLE.^{5 6} The reported incidence of cSLE ranges between 0.36 and 2.5 per 100 000 children, with a prevalence estimate of 1.89–34.1 per 100 000.^6–8 cSLE and adult-onset SLE (aSLE) are reported to have similar underlying pathophysiology.^{9 10} However, they may differ in the type and severity of manifestations at the time of disease onset.^{9 10} While children are more likely to present with renal and neuropsychiatric involvement at onset, aSLE has less involvement of major organ systems and a greater likelihood of appearance of subacute cutaneous and discoid lupus.^{11 12} Due to the more aggressive course of cSLE, children encounter greater long-term morbidity and mortality compared with aSLE, with a higher risk of organ damage.¹²

The treatment goals in SLE are to maintain low disease activity, achieve remission, prevent flares, minimise the use of glucocorticoids and prevent organ damage.^{13 14} The standard of care treatment includes combinations of hydroxychloroquine (HCQ) and glucocorticoids, with or without immunosuppressive agents with variability among different geographic areas.^{13 15 16} Currently, biologics are recommended as an add-on therapy in patients with persistent disease activity and inadequate response to standard therapy. HCQ is recommended for nearly all patients with SLE, including adults and children, to decrease flares and improve survival.^{13 17} However, glucocorticoids remain a mainstay of treatment in both children and adults with SLE. The prolonged use of glucocorticoids can have long-term adverse effects.¹³ Therefore, it is important to monitor treatment response and modify regimen as required to achieve treatment goals, while minimising adverse glucocorticoid effects.

While the Single Hub and Access point for paediatric Rheumatology in Europe (SHARE) initiative provided treatment recommendations for cSLE,^{18 19} there is limited research on the treatment patterns in SLE, particularly among children. Most treatments used to manage cSLE are off-label and derived from treatments used to manage aSLE. This study assessed the real-world treatment regimens and patterns in cSLE and aSLE cohorts in patients with newly diagnosed and existing SLE using a large US-based administrative claims database. The study also explored the degree of similarity in treatments, adherence, and duration of use between children and adults with SLE.

Methods

Study design and population

This retrospective cohort study utilised data from the Merative L.P. (formerly IBM) MarketScan Claims (Commercial and Medicare Supplemental) Databases in the USA between 1 January 2009 and 31 December 2020 (study period). Study variables were identified using enrolment records; service dates; International Classification of Diseases, 9th and 10th Revision, Clinical Modification (ICD-9/10-CM) codes; Current Procedural Technology 4th edition codes; Healthcare Common Procedure Coding System codes; and National Drug Codes, as needed.

The index date was defined as the date of the first SLE diagnosis in the identification period (1 January 2010 to 31 December 2019). The study considered data up to 12 months prior to the index date (ie, pre-index period) until 12 months following the index date (ie, post-index period). Based on the age at index date, patients aged <18 years and ≥18 years with a confirmed SLE diagnosis and continuous medical and pharmacy benefit enrolment of ≥12 months during the pre-index and post-index period were included in the cSLE and aSLE cohorts, respectively. SLE diagnosis in both cohorts was confirmed by ≥1 inpatient SLE diagnosis code (ICD-9-CM: 710.0 and/or ICD-10-CM: M32.10–M32.19, M32.8 and M32.9) or ≥2 non-diagnostic (not laboratory or radiology) outpatient claims at least 30 days but less than 2 years apart in the study period (figure 1).^20–23 Within the cSLE and aSLE cohorts, patients were stratified as (1) newly diagnosed if they did not have SLE diagnosis in the pre-index period and (2) having existing SLE if they had SLE diagnosis in the pre-index period.

Figure 1.

Study disposition. aSLE, adult-onset SLE; cSLE, childhood-onset SLE; ICD-9-CM and ICD-10-CM, International Classification of Diseases, 9th and 10th Revision, Clinical Modification.

Data sources

The Merative L.P. (formerly IBM) MarketScan Commercial Database contains individual-level health insurance claims including healthcare utilisation; expenditures; and insurance enrolment for inpatient, outpatient, prescription drug and carve-out services from individuals and their dependents with employer-sponsored health insurance. The Medicare Supplemental Database included the same healthcare data elements for retirees with employer-sponsored Medicare supplemental insurance.^{24 25}

Study measures

Demographic and clinical characteristics were summarised on the index date and during the pre-index period, respectively, for the cSLE and aSLE cohorts.

Treatment regimen stratified by diagnosis (ie, newly diagnosed or existing SLE) was compared between the cSLE and aSLE cohorts and was reported by medication class during the 12-month post-index period. Medication classes were (1) antimalarials (chloroquine and HCQ), (2) glucocorticoids (intravenous (IV; methylprednisolone) and oral (betamethasone, budesonide, cortisone, deflazacort, dexamethasone, hydrocortisone, methylprednisolone, paramethasone, prednisolone, prednisone and triamcinolone)), (3) immunosuppressants (methotrexate, azathioprine (and mercaptopurine), mycophenolate mofetil, cyclophosphamide, cyclosporine, leflunomide, tacrolimus and IV immunoglobulin) and (4) biologics (belimumab (IV and subcutaneous) and rituximab). The number of medication classes used over the 12-month post-index period was reported in newly diagnosed and existing patients with cSLE vs aSLE. Average oral glucocorticoid dose in milligrams (mg) per day of prednisone equivalent (except for budesonide and deflazacort) was also reported, along with the number of days of glucocorticoid use in the post-index period.

Treatment patterns were examined among newly diagnosed patients with cSLE vs aSLE who initiated therapy within 90 days of first diagnosis. Treatment patterns included identification of therapies initiated within 90 days of diagnosis (‘initial treatment’), adherence and discontinuation of initial treatment during the 9-month post-index period (since the total post-index period was 12 months). Adherence was measured using the proportion of days covered (PDC; range: 0–1) metric, recognised by the Pharmacy Quality Alliance as a preferred method for measuring adherence, calculated as below:^{26 27}

sum of days covered by the index drug during the follow-up (post-index period) ÷ length of follow-up (ie, 9 months of the post-index period) of 270 days

PDC≥80% was defined as high treatment adherence.^{26 28} Discontinuation of initial therapy was defined as a gap of >60 days between subsequent refills. The date of discontinuation was set as the last day’s supply before the 60-day gap. Treatment adherence and discontinuation were evaluated by medication class for antimalarials, biologics and glucocorticoids, and individually for immunosuppressant therapies methotrexate, azathioprine and mycophenolate mofetil.

Sensitivity analysis

The study utilised a 1-year pre-index period to categorise patients as being newly diagnosed with SLE. To assess the magnitude of potential miscategorisation, we used a 2-year pre-index period to identify these patients in a post hoc sensitivity analysis.

Compliance with ethics guidelines

The study used fully de-identified databases and data accessed were compliant with the US patient confidentiality requirements, including the Health Insurance Portability and Accountability Act of 1996 regulations. Thus, this study was exempted from institutional review board approval.

Statistical analyses

Continuous variables were presented as mean and SD and/or median and IQR. Categorical variables were summarised as frequencies and percentages. Univariate comparisons between cSLE and aSLE cohorts were performed using Wilcoxon rank-sum test for continuous or ordinal variables, and χ² test (or Fisher’s exact test, where appropriate) for categorical variables. Two-tailed p values <0.05 were considered statistically significant. As the databases included only patients with complete claims, enrolment and demographic data, imputation of missing data was not performed. All analyses were performed using Instant Health Data software (Panalgo, Boston, Massachusetts, USA), R V.3.2.1. (R Foundation for Statistical Computing, Vienna, Austria) and SAS V.9.4 (SAS Institute Inc, Cary, North Carolina, USA).

Results

Demographic and clinical characteristics

Of the 368 515 patients with SLE diagnosis during the identification period, a total of 1275 patients in the cSLE cohort and 66 326 patients in the aSLE cohort met the inclusion criteria (figure 1). The number of newly diagnosed patients were 1017 (79.8%) and 48 415 (73.0%) in the cSLE and aSLE cohorts, respectively (table 1). In both the SLE cohorts, most patients were female (cSLE: 84.7%; aSLE: 89.4%). Mean (SD) age in cSLE cohort was 14.1 (3.1) years and that in aSLE cohort was 49.7 (13.8) years. Arthritis (cSLE: 9.0%; aSLE: 14.4%; p<0.0001) and psoriasis-associated diagnoses (cSLE: 0.9%; aSLE: 2.0%) were frequently reported comorbidities in both cohorts. In the pre-index period, emergency room/department visits were more frequent in patients with cSLE (36.4% vs 29.1%; p<0.0001) and hospitalisation rates were comparable between the two cohorts (cSLE: 19.8%; aSLE: 19.7%; p=0.9229) (table 1).

Table 1.

Demographic and baseline clinical characteristics of patients with SLE

Variables	cSLE (N=1275)	aSLE (N=66 326)	P value
Newly diagnosed SLE,* n (%)	1017 (79.8)	48 415 (73.0)	<0.0001
Gender, n (%)
Female	1080 (84.7)	59 300 (89.4)	<0.0001
Age at index (years), mean (SD)	14.1 (3.1)	49.7 (13.8)
Age (years), n (%)
0–11	210 (16.5)	–
12–17	1065 (83.5)	–
18–64	–	58 510 (88.2)
≥65	–	7816 (11.8)
Comorbidities, n (%)
Arthritis†	115 (9.0)	9534 (14.4)	<0.0001
Plaque psoriasis	4 (0.3)	866 (1.3)	0.0019
Psoriatic arthritis	8 (0.6)	475 (0.7)	0.7095
Ulcerative colitis	3 (0.2)	561 (0.8)	0.0176
Crohn’s disease	8 (0.6)	600 (0.9)	0.2991
Ankylosing spondylitis	4 (0.3)	311 (0.5)	0.4203
Any of the above autoimmune diseases	131 (10.3)	11 384 (17.2)	<0.0001
Diabetes	22 (1.7)	8007 (12.1)	<0.0001
Vitiligo	2 (0.2)	132 (0.2)	1.00‡
Thyroiditis	24 (1.9)	1354 (2.0)	0.6905
All-cause healthcare resource utilisation, n (%)
Emergency room/department visit in pre-index period	464 (36.4)	19 325 (29.1)	<0.0001
Hospitalisations in pre-index period	252 (19.8)	13 037 (19.7)	0.9229

P values calculated using χ² test unless specified otherwise.

*No diagnosis codes of SLE in the 1-year pre-index period.

†Includes patients with rheumatoid arthritis or juvenile idiopathic arthritis.

‡P value calculated using Fisher’s exact test based on the assumptions of the χ² test.

aSLE, adult-onset SLE; cSLE, childhood-onset SLE.

Treatment regimen during the post-index period among patients with SLE

Medications used during the 12-month post-index period stratified by SLE diagnosis are presented in figure 2. Use of monotherapy was observed in more patients with aSLE compared with cSLE (newly diagnosed: 35.5% (n=17 164) vs 23.0% (n=234); existing SLE: 37.1% (n=6645) vs 22.9% (n=59); both p<0.0001), whereas the use of two or more therapies was higher in cSLE vs aSLE (newly diagnosed: 56.2% (n=572) vs 38.3% (n=18 566); existing SLE: 64.0% (n=165) vs 46.8% (n=8391); both p<0.0001) (data not shown). Use of all medication classes (except antimalarials in patients with existing SLE) was higher in the cSLE cohort vs aSLE cohort (p<0.05). The difference in immunosuppressant use was driven by higher mycophenolate mofetil use in patients with cSLE (p<0.0001) (figure 2). Antimalarials and glucocorticoids were the most used therapies among existing and newly diagnosed patients in the cSLE and aSLE cohorts (figure 2). Daily oral glucocorticoid dosage was higher in newly diagnosed vs existing patients in the cSLE and aSLE cohorts. The median daily oral glucocorticoid dose was higher in cSLE than aSLE in the newly diagnosed (22.1 mg vs 14.0 mg; p<0.0001) and existing (14.4 mg vs 12.3 mg; p=0.01) patients. Duration of oral glucocorticoid use during the post-index period was higher in the cSLE cohort vs aSLE cohort (newly diagnosed SLE: 153 vs 91 days, p<0.0001; existing SLE: 151 vs 130 days, p=0.01) (table 2).

Figure 2.

Treatment regimen during the 12-month post-index period: (A) Patients with existing SLE; (B) newly diagnosed patients with SLE. P values were calculated using χ² test, unless specified otherwise. ^aImmunosuppressants include azathioprine (and mercaptopurine), cyclophosphamide, cyclosporine, leflunomide, methotrexate, mycophenolate mofetil, tacrolimus and immunoglobulin (IV); ^bBiologics include belimumab and rituximab; we have not reported use for individual biologics due to low sample size; ^cP value calculated using Fisher’s exact test based on the assumptions of the χ² test. aSLE, adult-onset SLE; cSLE, childhood-onset SLE; IV, intravenous.

Table 2.

Average daily steroid dose in the 12-month post-index period

Variables		Existing cSLE (n=258)	Existing aSLE (n=17 911)	P value	Newly diagnosed cSLE (n=1017)	Newly diagnosed aSLE (n=48 415)	P value
Treatment with IV or oral glucocorticoids, n (%)		152 (58.9)	9195 (51.3)	0.02	631 (62.0)	24 045 (49.7)	<0.0001
Treatment with oral glucocorticoids, n (%)		151 (58.5)	9087 (50.7)	0.01	621 (61.1)	23 607 (48.8)	<0.0001
Number of treated days, mean (SD)*		151.4 (104.8)	130.3 (114.8)	0.01	153.0 (104.7)	91.0 (101.1)	<0.0001
Treatment with IV glucocorticoids, n (%)		18 (7.0)	603 (3.4)	0.002	115 (11.3)	2219 (4.6)	<0.0001
Number of treated days, mean (SD)*†		5.4 (10.7)	2.4 (3.1)	0.01	5.7 (8.0)	2.4 (3.2)	<0.0001
Additional dosing information for oral glucocorticoids only‡
N		149	9062		596	23 544
Average daily dose (mg)	Mean (SD)	20.1 (19.6)	16.4 (20.8)		26.1 (16.8)	21.5 (30.8)
	Median (IQR)	14.4 (7.7–30.0)	12.3 (6.2–20.0)	0.01	22.1 (14.0–35.1)	14 (10.0–25.1)	<0.0001
Average dose range (mg/day), n (%)
<7.5		34 (22.8)	2664 (29.4)	0.001	46 (7.7)	3918 (16.6)	<0.0001
7.5–14		41 (27.5)	3188 (35.2)		122 (20.5)	8540 (36.3)
15–24		34 (22.8)	1621 (17.9)		162 (27.2)	4900 (20.8)
>25		40 (26.8)	1589 (17.5)		266 (44.6)	6186 (26.3)

P values calculated using χ² (categorical variables) and Wilcoxon rank-sum (continuous and ordinal variables) tests.

*Patients without treated days were excluded.

†Duration of IV steroid use (ie, pulse methylprednisolone use) was calculated based on unique days with HCPCS code for IV steroids

‡Steroid dosing was available for a subgroup of the entire sample, doses mentioned here are prednisone equivalent doses (except for budesonide and deflazacort).

aSLE, adult-onset SLE; cSLE, childhood-onset SLE; HCPCS, Healthcare Common Procedure Coding System; IV, intravenous.

Treatment patterns

Therapy used within 90 days of diagnosis of newly diagnosed SLE

A total of 733 (72.1%) newly diagnosed patients with cSLE and 29 588 (61.1%) newly diagnosed patients with aSLE initiated therapy within 90 days of SLE diagnosis. While use of antimalarials as monotherapy was more common in aSLE vs cSLE cohort (23.7% vs 15.7%; p<0.0001), use of combination therapies was more frequent in cSLE vs aSLE cohort (antimalarials+glucocorticoids: 21.7% vs 12.6%, p<0.0001; antimalarials+glucocorticoids+immunosuppressant/biologic: 17.6% vs 4.7%, p<0.0001) (table 3).

Table 3.

Treatment patterns among newly diagnosed patients during the 9-month post-index period

Variables		cSLE (n=1017)	aSLE (n=48 415)	P value
Initial treatment by class—use at any time in the first 90 days, n (%)
Antimalarials only		160 (15.7)	11 466 (23.7)	<0.0001
Immunosuppressants/biologics* only		16 (1.6)	1362 (2.8)
Glucocorticoids only		90 (8.8)	5212 (10.8)
Antimalarials+immunosuppressants/biologics*		21 (2.1)	1168 (2.4)
Antimalarials+glucocorticoids		221 (21.7)	6077 (12.6)
Immunosuppressants/biologics*+glucocorticoids		46 (4.5)	2037 (4.2)
Antimalarials+glucocorticoids+immunosuppressants/biologics*		179 (17.6)	2266 (4.7)
No therapy within 90 days		211 (20.7)	12 685 (26.2)
Discontinuation† of initial treatment, n (%)
Antimalarial		145 (25.0)	6946 (33.1)	<0.0001
Immunosuppressant therapies
Azathioprine		14 (35.0)	637 (45.4)	0.19
Methotrexate		22 (39.3)	1299 (43.1)	0.57
Mycophenolate mofetil		36 (24.5)	623 (36.0)	0.005
Oral glucocorticoids		297 (56.6)	10 877 (71.2)	<0.0001
Adherence †‡ to initial treatment, PDC
Antimalarial	Mean (SD)	0.8 (0.2)	0.7 (0.3)
	Median (IQR)	0.9 (0.6–1.0)	0.8 (0.5–0.9)	<0.0001
	N	581	20 977
Immunosuppressant therapies
Azathioprine	Mean (SD)	0.7 (0.3)	0.6 (0.3)
	Median (IQR)	0.8 (0.5–0.9)	0.7 (0.3–0.9)	0.75
	N	40	1402
Methotrexate	Mean (SD)	0.7 (0.3)	0.6 (0.3)
	Median (IQR)	0.8 (0.5–0.9)	0.7 (0.4–0.9)	0.21
	N	56	3015
Mycophenolate mofetil	Mean (SD)	0.8 (0.3)	0.7 (0.3)
	Median (IQR)	0.8 (0.7–0.9)	0.8 (0.4–0.9)	0.005
	N	147	1729
Oral glucocorticoids	Mean (SD)	0.5 (0.3)	0.4 (0.3)
	Median (IQR)	0.6 (0.3–0.8)	0.3 (0.1–0.7)	<0.0001
	N	525	15 278
Adherence† to initial treatment, PDC ≥80%, n (%)
Antimalarial		333 (57.3)	10 376 (49.5)	0.0002
Immunosuppressants therapies
Azathioprine		21 (52.5)	582 (41.5)	0.16
Methotrexate		29 (51.8)	1265 (42.0)	0.14
Mycophenolate mofetil		83 (56.5)	833 (48.2)	0.054
Oral glucocorticoids		101 (19.2)	2333 (15.3)	0.01

P values calculated using χ² test (categorical variables) and Wilcoxon rank-sum (continuous and ordinal variables) test, unless specified otherwise.

*Biologics include belimumab and rituximab.

†Treatment discontinuation and adherence are not reported for biologics due to low sample size.

‡Adherence measured as PDC ranging from 0 to 1. Drug could be monotherapy or part of a combination therapy.

aSLE, adult-onset SLE; cSLE, childhood-onset SLE; PDC, proportion of days covered.

Discontinuation of therapy initiated within 90 days of diagnosis during the 9-month post-index period

Discontinuation rates were lower in cSLE vs aSLE cohorts for antimalarials (25.0% vs 33.1%; p<0.0001), mycophenolate mofetil (24.5% vs 36.0%; p=0.005) and oral glucocorticoids (56.6% vs 71.2%; p<0.0001) (table 3). Discontinuation rates for biologics are not reported due to low sample size.

Adherence to therapy initiated within 90 days of diagnosis during the 9-month post-index period

The median PDC was higher in the cSLE vs aSLE cohort for antimalarials (0.9 vs 0.8; p<0.0001), mycophenolate mofetil (0.8 for both; p=0.005) and oral glucocorticoids (0.6 vs 0.3; p<0.0001) (table 3). Treatment adherence (PDC≥80%) was higher in cSLE compared with aSLE cohort for antimalarials (57.3% vs 49.5%; p=0.0002), oral glucocorticoids (19.2% vs 15.3%; p=0.01) and immunosuppressants (51.8%–56.5% vs 41.5%–48.2%; p=not significant) (table 3). For biologics, median PDC and treatment adherence are not reported due to low sample size.

Sensitivity analysis

A total of 747 patients with cSLE and 44 669 patients with aSLE who had 2-year pre-index and 1-year post-index enrolment were included in the post hoc sensitivity analysis. The proportion of newly diagnosed patients in 2-year vs 1-year pre-index period was marginally lower in the cSLE (71.75% vs 74.83%) and aSLE (57.01% vs 66.76%) cohorts.

Discussion

This study utilised de-identified patient-level data from the MarketScan Claims Databases in the USA to provide an overview of SLE disease management patterns and therapies initiated among newly diagnosed patients with cSLE and aSLE; in consideration that there is lack of data on general treatment patterns in adults and few in children.^29–31

Antimalarials and glucocorticoids were commonly used therapies among existing and newly diagnosed patients with cSLE and aSLE. Compared with patients with newly diagnosed SLE, patients with existing SLE reported higher use of antimalarials, immunosuppressants and combination therapy in cSLE and aSLE cohorts. Although majority of newly diagnosed patients with cSLE and aSLE initiated therapy within 90 days of diagnosis, we noted low adherence and high discontinuation of initial treatment.

Despite similarities in the underlying pathophysiology of cSLE and aSLE, a higher initial disease activity is commonly reported in cSLE.^{9 10} As cSLE management is limited by lack of high-quality evidence from multinational clinical trials, treatment strategies are derived from adult protocols and clinical experience.³² This explains the use of same medication classes in the cSLE and aSLE cohorts in our study. The need for evidence-based treatments for cSLE management is well-recognised and clinical trials are identified as one of the top research priorities in cSLE.^{33 34} Indeed, several methodological strategies are being explored to expedite the approval of new drugs for children with SLE.³⁵

Owing to differences in disease severity, glucocorticoids are more commonly used to treat cSLE than aSLE in clinical practice.^{11 36} In our study, the average daily dose and use of oral glucocorticoids were higher in cSLE than in aSLE in both patients with newly diagnosed and existing SLE (p<0.05). Results from this study are congruent with the findings from a retrospective cohort study where patients with early-onset cSLE (age <12 years) required a significantly higher daily glucocorticoid dose than those diagnosed with cSLE during adolescence (0.6 mg/kg prednisone equivalent vs 0.2 mg/kg prednisone equivalent, p<0.05).³⁷ In a longitudinal study assessing steroid-related damage, patients with cSLE were at increased risk of damage (OR 1.7, 95% CI: 1.1 to 2.8; adjusted analysis) compared with patients with aSLE.³⁸ Thus, use of steroid-sparing treatment regimens should be considered, especially in cSLE, to prevent damage accrual in adulthood.³⁸ In clinical practice, glucocorticoids are prescribed intermittently for shorter periods to control symptoms and are usually tapered off. This could be the reason for higher discontinuation rates reported for oral glucocorticoids in the cSLE and aSLE cohorts in this study. Since information on glucocorticoid dose tapering was not captured in the database, it may be difficult to interpret adherence to oral glucocorticoids.

Mycophenolate mofetil is commonly recommended as induction and maintenance therapy in patients with cSLE with proliferative lupus nephritis or in the setting of neuropsychiatric involvement.^{16 32} This study reported higher use of immunosuppressants in cSLE than in aSLE, predominantly due to the higher use of mycophenolate mofetil among patients with cSLE. Thus, more common use of mycophenolate mofetil is consistent with the multiorgan involvement often seen in cSLE compared with aSLE.^{16 32}

Antimalarial drugs are the backbone of the immunomodulatory regimen used to treat children and adults with SLE.^{13 39} We found that discontinuation of antimalarials was low in both cSLE and aSLE cohorts, conforming with the quality indicators for cSLE and aSLE treatment.13 18 40 41

The US Food and Drug Administration approved belimumab, the first targeted therapy for aSLE, in 2011.⁴² However, belimumab was not approved for cSLE until 2019 and remains the only approved drug to treat cSLE.⁴³ Since majority of patients (cSLE: 66.4%; aSLE: 71.8%) were enrolled in the initial years of the study period (2010–2013), the use of biologics in our study was low. Among patients with cSLE, only five patients received belimumab, while 26 patients received rituximab within 90 days of index date. Rituximab is commonly used as a rescue therapy in children for treating flare or refractory cSLE.⁴⁴ Thus, there remains scarcity of approved treatment options for both aSLE and cSLE.

While the similar treatment approaches and drug adherence data for children and adults may suggest similar treatment response, the database used for this study does not allow direct assessment of comparative effectiveness. In contrast to previous studies that have reported poor treatment adherence in patients with cSLE,^{32 45 46} we found similar adherence in children and adults. However, a decline in treatment adherence during the course of the disease remains a major concern and further research is needed to develop interventions to maintain treatment adherence over time. More studies with head-to-head comparisons between patients with cSLE and aSLE for adherence to and efficacy of therapeutic interventions are needed. Nonetheless, similarity of treatment approach and drug adherence supports the use of extrapolation in paediatric development to improve labelled treatment options in children.

The results of the study should be interpreted with consideration of several limitations. As with other claims-based studies, our results cannot be generalised to the uninsured or patients insured by a non-commercial programme. Presence of juvenile idiopathic arthritis or idiopathic arthritis as a comorbidity could be a misdiagnosis of SLE. Information on clinical variables, flares and disease severity was not available. A causal relationship between treatment and response could not be established owing to the observational study design. Glucocorticoids in patients with cSLE are administered based on patient weight, which was unavailable in claims data, and dosing (mg/kg) could not be computed. We used a 12-month pre-index period to identify patients with newly diagnosed SLE. As patients with stable disease without diagnosis codes in the pre-index may be misclassified as ‘newly diagnosed’, the proportion of these patients may have been overestimated. However, in the post hoc sensitivity analysis, the proportion of newly diagnosed patients in 2-year pre-index period was marginally lower in the cSLE and aSLE cohorts.

In summary, the current study showed that although a majority of patients initiate therapy within 90 days of diagnosis, they tend to exhibit low adherence and high discontinuation to the initial therapies. Even though same medication classes are being used for management of aSLE and cSLE, differences in treatment patterns exist, with higher overall use of therapy in cSLE. Medical management of SLE remains to be defined and established in children.

Data availability statement

Data are available upon reasonable request. The datasets generated and/or analysed during the current study are not publicly available due to individual data privacy but may be available from the corresponding author on reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

The study used fully de-identified databases and data accessed were compliant with the US patient confidentiality requirements, including the Health Insurance Portability and Accountability Act of 1996 regulations. Thus, this study was exempted from institutional review board approval.