Extrarenal symptoms associate with worse quality of life in patients enrolled in the AMP RA/SLE Lupus Nephritis Network

Abstract

Objective

Lupus nephritis (LN) can occur as an isolated component of disease activity or be accompanied by diverse extrarenal manifestations. Whether isolated renal disease is sufficient to decrease health-related quality of life (HRQOL) remains unknown. This study compared Patient-Reported Outcomes Measurement Information System 29-Item (PROMIS-29) scores in LN patients with isolated renal disease to those with extrarenal symptoms to evaluate the burden of LN on HRQOL and inform future LN clinical trials incorporating HRQOL outcomes.

Methods

A total of 181 LN patients consecutively enrolled in the multicentre multi-ethnic/racial Accelerating Medicines Partnership completed PROMIS-29 questionnaires at the time of a clinically indicated renal biopsy. Raw PROMIS-29 scores were converted to standardized T scores.

Results

Seventy-five (41%) patients had extrarenal disease (mean age 34, 85% female) and 106 (59%) had isolated renal (mean age 36, 82% female). Rash (45%), arthritis (40%) and alopecia (40%) were the most common extrarenal manifestations. Compared with isolated renal, patients with extrarenal disease reported significantly worse pain interference, ability to participate in social roles, physical function, and fatigue. Patients with extrarenal disease had PROMIS-29 scores that significantly differed from the general population by >0.5 SD of the reference mean in pain interference, physical function, and fatigue. Arthritis was most strongly associated with worse scores in these three domains.

Conclusion

Most patients had isolated renal disease and extrarenal manifestations associated with worse HRQOL. These data highlight the importance of comprehensive disease management strategies that address both renal and extrarenal manifestations to improve overall patient outcomes.

Rheumatology key messages.

• Extrarenal symptoms, especially arthritis, are associated with significantly worse HRQOL in patients with lupus nephritis.

• Patients with lupus nephritis, but without extrarenal disease, may not report an increased HRQOL burden.

• Lupus nephritis clinical trials evaluating HRQOL outcomes should account for baseline extrarenal symptoms.

Introduction

Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by heterogeneous clinical manifestations and affects the kidney in up to 60% of patients [1, 2]. Lupus nephritis (LN) can occur as an isolated component of disease activity or be accompanied by diverse extrarenal symptoms captured on the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) such as arthritis, serositis and/or rash [3]. While SLE has been shown to adversely affect a patient’s perceived health-related quality of life (HRQOL), whether LN is sufficient to decrease HRQOL in the absence of activity in other organs has not been fully explored [4]. Favourable HRQOL in those with isolated renal disease may place patients at risk for delayed diagnosis of LN and/or medication nonadherence which could be life-threatening since LN carries the highest standardized mortality ratio in SLE and early detection and treatment are associated with better outcomes [2, 5, 6]. In addition, assessing the influence of extrarenal disease on HRQOL in those with active LN would aid the interpretation of LN clinical trials that incorporate HRQOL outcomes.

Several instruments have been used to measure HRQOL in rheumatic diseases [7]. The Patient-Reported Outcomes Measurement Information System (PROMIS) 29-Item (PROMIS-29) Profile was developed as part of an initiative from the National Institutes of Health to improve and standardize the quantification of patient-reported outcomes [8]. PROMIS-29 combines scales across seven domains evaluating physical and social functioning, anxiety, depression, fatigue, sleep, and pain to quantify HRQOL. Prior studies have demonstrated the cross-sectional and longitudinal validity of PROMIS measures in SLE including across racial/ethnic groups [7, 9–12].

The Accelerating Medicines Partnership (AMP) study is a public-private network formed to identify new targets for LN diagnostics and drug development with a multimodal approach combining clinical, histologic, proteomic, transcriptomic, and genetic evaluations [13]. The cohort resulting from this consortium represents a large well-characterized population of multi-racial/ethnic LN patients enrolled across many clinical sites in real-world settings. This study leveraged the AMP cohort to compare PROMIS-29 scores between LN patients with isolated renal activity to those with extrarenal symptoms to evaluate the burden of LN on HRQOL and to inform future clinical trials that include HRQOL outcomes.

Methods

Study design and patient population

Detailed descriptions of AMP have been provided previously [14, 15]. All patients fulfilled the American College of Rheumatology (ACR) or Systemic Lupus International Cooperating Clinics (SLICC) classification criteria for SLE and were consecutively enrolled in Phase 2 of AMP [16, 17]. Patients required a urine-protein-to creatinine ratio > 0.5 g/g on a random or > 500 mg of protein in a 24-h urine collection and underwent a clinically indicated percutaneous renal biopsy as part of routine care. Patients with biopsies showing proliferative, membranous, or mixed International Society of Nephrology/Renal Pathology Society (ISN/RPS) Classes were included [18]. Receipt of rituximab within six months preceding renal biopsy, pregnancy, and/or a history of kidney transplant were excluded from AMP. At the time of kidney biopsy, demographics, clinical characteristics including a completed printed PROMIS-29 v2.1 questionnaire, disease activity assessed by the hybrid Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA)-SLE Disease Activity Index (SLEDAI), laboratory measures, and medications were recorded [3]. Patients were considered to have extrarenal clinical disease activity if any of the clinical SLEDAI descriptors listed in Fig. 1 were present. In compliance with the Declaration of Helsinki, institutional review boards and ethics committees at individual sites (Cedars-Sinai; University of Cincinnati; Albert Einstein College of Medicine; Johns Hopkins; University of Michigan; Medical University of South Carolina; Northwell Health; NYU Grossman School of Medicine; University of Rochester; Texas Tech; University of California sites including San Francisco, Los Angeles and San Diego; University of North Carolina; and University of Texas Health) approved the protocol and consent forms; written informed consent was obtained from all patients. Of the 252 patients fulfilling inclusion criteria for AMP Phase 2, this analysis included all 181 patients who had completed baseline PROMIS-29 questionnaires.

Figure 1.

Map of the clinical SLEDAI domains affecting each of the 75 patients with extrarenal clinical disease activity. Shaded green represents the presence of that manifestation

Raw PROMIS-29 scores were converted to standardized T scores (with the exception of pain intensity which is scored on a 0–10 scale) with a general population mean of 50 and standard deviation of 10 using the online HealthMeasures scoring service (https://www.healthmeasures.net/score-and-interpret/calculate-scores). PROMIS Profile v2.1 was set as the calibration sample. PROMIS scales are designed such that higher scores represent more of the domain being measured. Thus, higher scores on pain, anxiety, depression, fatigue, and sleep and conversely lower scores on physical and social functioning indicate a worse outcome. A prior study evaluating SLE patients estimated minimally important differences to be approximately two points across PROMIS scales supporting the clinical meaningfulness of these findings [11].

Statistical analysis

Descriptive statistics are presented as the mean and standard deviation for continuous variables and frequencies for categorical variables. A two-sample 2-tailed t test or Pearson’s chi-squared test was used to compare patients with extrarenal vs isolated renal disease. A one-sample 2-tailed t test was used to compare PROMIS-29 T scores to the general population reference mean. Univariate linear regressions were performed to evaluate associations between clinical variables and PROMIS-29 domains. Multivariable linear regressions were completed using a backward stepwise procedure with PROMIS-29 scores as the dependent variable and clinical parameters with p-value <0.1 in univariate analyses as candidate covariates. All data were analyzed using R Studio version 4.0.3.

Results

Of the 181 LN patients included in the analysis, 106 (59%) had isolated renal disease and 75 (41%) had extrarenal disease. Of the latter, the most common SLEDAI disease manifestations were rash (n = 34, 45%), arthritis (n = 30, 40%), alopecia (n = 30, 40%), mucosal ulcers (n = 10, 13%), and serositis (pericarditis and/or pleurisy, n = 9, 12%) (Table 1 and Fig. 1). The presence of extrarenal disease was not influenced by age (isolated renal: mean age 36.43 vs extrarenal mean age: 33.89), sex (isolated renal: 82% female vs extrarenal: 85% female), or race (isolated renal: 25% white vs extrarenal: 27% white). (Table 1). A significantly higher percentage of patients with extrarenal disease had lower C3 and were more likely to be at first biopsy (Table 1). Patients with extrarenal disease also had significantly more proteinuria (Table 1). However, anti-double-stranded DNA antibody positivity, biopsy class, and NIH activity/chronicity indices did not differ from those patients with isolated renal disease (Table 1). Extrarenal disease was associated with significantly more belimumab and prednisone use (Table 1).

Table 1.

Clinical characteristics in patients with isolated renal compared with extrarenal disease

Baseline Clinical Characteristic	Isolated Renal n = 106	Extrarenal n = 75	P-value
Age, years, mean (S.D.)	36.43 (12.70)	33.89 (9.83)	0.148
Sex: Female, N (%)	87 (82.1)	64 (85.3)	0.706
Race, N (%)			0.723
Asian	17 (16.0)	10 (13.3)
Black	45 (42.5)	36 (48.0)
White	26 (24.5)	20 (26.7)
Other/Unknown	18 (17.0)	9 (12.0)
Ethnicity: Hispanic, N (%)	28 (26.4)	20 (26.7)	1.000
NIH Activity index, mean (S.D.)	4.30 (4.35) n = 77	5.54 (4.92) n = 56	0.128
NIH Chronicity index, mean (S.D.)	3.14 (2.65) n = 78	2.69 (2.37) n = 54	0.313
First biopsy, N (%)	37 (34.9)	41 (54.7)	0.013
Positive anti-dsDNA, N (%)	65 (67.7) n = 96	54 (74.0) n = 73	0.475
Low C3, N (%)a	51 (50.0) n = 102	55 (73.3) n = 75	0.003
Low C4, N (%)a	50 (49.0) n = 102	47 (62.7) n = 75	0.099
Serum creatinine, mg/dl, mean (S.D.)	1.24 (1.02)	1.10 (0.71) n = 74	0.31
Urine protein: creatinine ratio g/g, mean (S.D.)	2.66 (2.19)	3.47 (2.89) n = 74	0.035
Biopsy Class, N (%)			0.408
Proliferative	42 (39.6)	35 (46.7)
Mixed	29 (27.4)	22 (29.3)
Membranous	35 (33)	18 (24)
Extrarenal Clinical SLEDAI, mean (S.D.)		4.72 (2.80)
Extrarenal Clinical SLEDAI Domains, N (%)
Visual		1 (1)
Vasculitis		3 (4)
Seizure		0
Rash		34 (45)
Psychosis		0
Pleurisy		9 (12)
Pericarditis		6 (8)
Organic Brain Syndrome		0
Myositis		1 (1)
Mucosal Ulcers		10 (13)
Lupus Headache		1 (1)
Fever		4 (5)
Cranial Nerve Disorder		1 (1)
Cerebrovascular Accident		0
Arthritis		30 (40)
Alopecia		30 (40)
Medications
Hydroxychloroquine	93 (87.7)	58 (77.3)	0.099
Mycophenolate	60 (56.6)	39 (52.0)	0.645
Azathioprine	11 (10.4)	7 (9.3)	1.000
Belimumab	0 (0.0)	6 (8.0)	0.011
Tacrolimus	9 (8.5)	3 (4.0)	0.372
Cyclophosphamide	2 (1.9)	1 (1.3)	1.000
Prednisone/methylprednisolone	67 (63.2)	60 (80.0)	0.023
Prednisone or equivalent dose > 20 mg	32 (30.2)	39 (52.0)	0.005

Data represented as mean (SD) or N (%), n is specified where it differed from the overall sample size, Two-tailed t test was used for continuous variables and Pearson’s chi-squared was used for categorical variables. Bold text highlights significant P-values.

^aClassified by local laboratory cutoffs.

Overall PROMIS-29 scores significantly differed from the reference population in the domains of pain interference, physical function, fatigue, anxiety, and sleep irrespective of extrarenal manifestations (Table 2). While those with isolated renal disease had PROMIS-29 scores that significantly differed from the reference population, scores in all domains for these patients remained within the generally accepted ‘normal range’ of 45–55, or 0.5 standard deviations of the general population mean (Table 2) [19, 20]. Patients with extrarenal disease had T scores outside this range in the domains of pain interference, physical function, and fatigue (Table 2). Compared with isolated LN, patients with extrarenal disease reported significantly worse pain interference, ability to participate in social roles, physical function, and fatigue (Table 2).

Table 2.

Mean PROMIS-29 T scores in patients compared with the general population and in patients with isolated renal compared with extrarenal disease

PROMIS Domain	All Patients n = 181		Extrarenal n = 75		Isolated Renal n = 106		Extrarenal vs Isolated Renal
	T Score mean (SD)	P-value c	T Score mean (SD)	P-value c	T Score mean (SD)	P-value c	P-value d
Pain Interference	56.6 (10.8)	<0.001	60.01 (10.40)	<0.001	54.11 (10.46)	<0.001	<0.001
Ability to Participate in Social Rolesb	49.6 (9.2) n = 179	0.529	46.47 (8.77)	<0.001	51.80 (8.91) n = 104	0.042	<0.001
Physical Functionb	45.3 (9.52)	<0.001	41.78 (9.43)	<0.001	47.83 (8.80)	0.013	<0.001
Fatigue	56.1 (10.9)	<0.001	58.98 (10.14)	<0.001	54.00 (11.00)	<0.001	0.002
Anxiety	52.7 (10.4)	<0.001	52.67 (10.53)	0.03	52.72 (10.34)	0.008	0.978
Sleep	54.3 (8.34)	<0.001	54.53 (9.47)	<0.001	54.14 (7.48)	<0.001	0.755
Depression	49.4 (9.50)	0.386	49.72 (9.79)	0.802	49.15 (9.32)	0.352	0.695
Pain Intensitya	4.1 (3.0) n = 180	–	5.19 (3.01) n = 74	–	3.42 (2.86)	–	<0.001

Data represented as mean (SD), n is specified where it differed from the overall sample size. Bold text highlights significant P-values.

^aPain intensity scored separately on a 0–10 scale and is not converted to a standardized T Score.

^bLower value indicates a worse score.

^cTwo-tailed one-sample t test comparing PROMIS-29 T Score mean to the mean of the reference general population (mean of reference: 50, standard deviation of reference: 10).

^dTwo-tailed two-sample t test comparing PROMIS-29 T Score mean between extrarenal and isolated renal disease.

For pain interference, physical function, and fatigue which were outside the ‘normal range’ of the reference population in extrarenal LN patients, a univariate linear regression analysis was used to evaluate whether relevant clinical variables were associated with worse scores in these areas and to assess which specific extrarenal disease manifestations associated with PROMIS-29 domains. PROMIS-29 scores were not associated with age, positive anti-double-stranded DNA antibodies, or nephrotic range proteinuria (Table 3). Female sex was associated with worse physical function and fatigue (Table 3). White/non-Hispanic race/ethnicity trended toward an association with worse physical function and significantly associated with worse fatigue but did not associate with pain interference (Table 3). Prednisone dose >20 mg was significantly associated with worse pain interference, physical function, and fatigue. Low complement (C3 or C4) was only associated with worse fatigue (Table 3). Arthritis was the strongest significant predictor of pain interference, physical function, and fatigue (Table 3). Mucosal ulcers and rash demonstrated less strong but similarly significant associations (Table 3). Alopecia trended toward an association with pain interference and physical function but was only significantly associated with fatigue. Serositis was only significantly associated with worse physical function (Table 3).

Table 3.

Univariate linear regressions analysis of clinical variables with PROMIS-29 T scores

Predictor	Pain Interference n = 181		Physical Function b n = 181		Fatigue n = 181
	Estimate (95% CI)	P-value	Estimate (95% CI)	P-value	Estimate (95% CI)	P-value
Age	−0.02 (−0.15, 0.12)	0.820	−0.04 (−0.16, 0.08)	0.472	−0.04 (−0.18, 0.10)	0.584
Sex: Female	0.70 (−3.58, 4.97)	0.748	−4.16 (−7.88, −0.45)	0.028	5.69 (1.46, 9.92)	0.009
Non-Hispanic White	2.67 (−1.78, 7.11)	0.238	−3.49 (−7.39, 0.41)	0.079	7.10 (2.72, 11.48)	0.002
First biopsy	3.35 (0.18, 6.52)	0.038	−3.33 (−6.12, -0.55)	0.019	2.22 (−1.01, 5.44)	0.176
Low Complement	1.66 (−1.73, 5.05) n = 177	0.335	−0.76 (−3.77, 2.25) n = 177	0.618	3.67 (0.23, 7.11) n = 177	0.037
Anti-dsDNA Positive	2.52 (−1.05, 6.1) n = 169	0.166	−1.46 (−4.62, 1.70) n = 169	0.362	1.30 (−2.36, 4.95) n = 169	0.485
Prednisone >20mg	4.12 (0.92, 7.32)	0.012	−5.60 (−8.35, −2.85)	<0.001	3.75 (0.51, 6.98)	0.024
Nephrotic range proteinuria	−0.69 (−4.22, 2.83) n = 180	0.698	−2.45 (−5.54, 0.63) n = 180	0.119	−0.21 (−3.76, 3.35) n = 180	0.909
Arthritis	9.28 (5.23, 13.33)	<0.001	−10.35 (−13.79, −6.91)	<0.001	8.72 (4.60, 12.83)	<0.001
Rash	4.73 (0.73, 8.74)	0.021	−4.03 (−7.57, −0.50)	0.026	4.70 (0.65, 8.74)	0.023
Alopecia	3.69 (−0.55, 7.93)	0.088	−3.46 (−7.19, 0.27)	0.069	5.20 (0.95, 9.44)	0.017
Serositisa	5.71 (−0.39, 11.81)	0.066	−5.68 (−11.04, −0.32)	0.038	2.05 (−4.15, 8.25)	0.515
Mucosal Ulcers	8.40 (1.55, 15.25)	0.017	−8.10 (−14.11, −2.09)	0.009	7.38 (0.45, 14.31)	0.037

n is specified where it differs from the overall sample size. Bold text highlights significant P-values.

^aPleurisy and/or pericarditis.

^bLower value indicates a worse score.

A stepwise multivariable linear regression analysis confirmed that the associations of arthritis with pain interference, physical function, and fatigue were independent of other clinical parameters (Table 4). Rash, serositis, alopecia and mucosal ulcers were not significantly associated with these PROMIS-29 domains in multivariable models (Table 4). Prednisone dose >20 mg remained significantly associated with worse physical function and White Non-Hispanic race/ethnicity remained significantly associated with worse fatigue in adjusted models (Table 4).

Table 4:

Multivariable linear regression of clinical variables with PROMIS-29 T scores

Predictor	Pain Interference		Physical Function b		Fatigue
	Adjusted Estimate (95% CI)	Adjusted P-value	Adjusted Estimate (95% CI)	Adjusted P-value	Adjusted Estimate (95% CI)	Adjusted P-value
Sex: Female			−2.29 (−5.68, 1.10)	0.185	3.29 (−0.89, 7.47)	0.122
White/Non-Hispanic			−2.23 (−5.74, 1.28)	0.212	5.81 (1.60, 10.03)	0.007
First Biopsy
Low Complement					2.37 (−0.91, 5.64)	0.157
Prednisone >20mg	2.44 (−0.72, 5.59)	0.129	−3.74 (−6.37, −1.11)	0.006
Arthritis	7.56 (3.30, 11.82)	<0.001	−8.05 (−11.63, −4.47)	<0.001	7.06 (2.91, 11.21)	<0.001
Rash					2.32 (−1.78, 6.42)	0.265
Alopecia					2.26 (−1.99, 6.51)	0.296
Serositisa	4.05 (−1.87, 9.96)	0.179	−3.41 (−8.32, 1.51)	0.173
Mucosal Ulcers	4.70 (−2.16, 11.55)	0.178	−3.56 (−9.25, 2.14)	0.219

Bold text highlights significant P-values.

^aPleurisy and/or pericarditis.

^bLower value indicates a worse score.

Discussion

In this prospective study a majority of LN patients had isolated renal disease and patients with extrarenal manifestations reported significantly worse pain interference, ability to participate in social roles, physical function, and fatigue on PROMIS-29. While patients with isolated renal disease had scores within 0.5 standard deviations of the reference population on all PROMIS domains, patients with extrarenal disease had scores outside this generally accepted ‘normal’ range in the areas of pain interference, physical function, and fatigue [19–21]. Arthritis was most strongly associated with decreased HRQOL in these three domains.

LN has previously been linked to decreased HRQOL, but prior studies on this association are conflicting and a recent systematic review concluded that data on the influence of extrarenal symptoms on HRQOL in LN are lacking [4]. Jolly et al. found patients with active LN reported significantly worse HRQOL using the lupus patient-reported outcome tool (LupusPRO) compared with SLE patients without active LN [22]. Kim et al. and Appenzeller et al. also found that patients with LN had worse physical component scores on the Short-form 36 (SF-36) health survey compared with SLE patients without LN [23, 24]. Chaing and colleagues performed a cross-sectional assessment of the Swiss Systemic Lupus Erythematosus Cohort Study and observed SF-36 subscales reflecting physical and mental health associated with both musculoskeletal and active renal disease [25]. Most recently, Hashemi et al. found renal and skin involvement but not arthritis to be related to decreased HRQOL in an Iranian SLE cohort [26]. In contrast, Clarke et al. did not report a difference in annual change of SF-36 summary scales between SLE patients with chronic renal damage and those without [27]. In the Systemic Lupus International Collaborating Clinics inception cohort, Hanly et al. showed that HRQOL between LN and non-LN SLE did not differ overall, but an association between LN and decreased SF-36 physical component was observed when stratifying by eGFR [28]. Consistent with our finding that arthritis was most highly associated with decreased HRQOL in LN, Doria et al. found that in an Italian cohort arthritis-arthralgia, but not other components of disease activity, including renal, associated with worse HRQOL in SLE patients [29]. Similarly, Zhu et al. observed musculoskeletal flares but not renal flares in the preceding year were associated with significantly worse SF-36 physical and mental health summary scales [30]. While some studies on LN did adjust for extrarenal SLEDAI or musculoskeletal symptoms, our study was unique in that it specifically addressed the influence of extrarenal symptoms on a background of active lupus nephritis [23, 24, 26–28].

Routine surveillance for renal involvement is a critical aspect of SLE disease management and medication adherence is essential for patients with LN. Most patients had isolated renal disease and mean PROMIS-29 scores in this group did not fall outside the ‘normal range’ supporting the clinical suspicion that patients are unlikely to sense ongoing renal involvement. This observation is consistent with a recent study by Sun et al. who demonstrated the limited ability of PROMIS-29 to detect type 1 SLE (characterized by active inflammation) compared with type 2 SLE (characterized by fatigue, pain, and cognitive disturbances) [21]. Given that a lack of extrarenal symptoms was common, the additional absence of impaired HRQOL could potentially contribute to a disconnect between LN patients and their treating physicians regarding the severity of their disease and the need for treatment [31–33]. This may pose a barrier to a positive therapeutic relationship and place patients at risk for treatment and medication nonadherence [34, 35]. Through patient surveys and in-depth interviews, Sloan et al. demonstrated that rheumatologists’ listening skills were significantly associated with medication adherence in SLE patients [34]. As such, physicians should be especially mindful to listen to the SLE patients who otherwise feel well and take care to educate them about the need for monitoring of kidney disease even in times of seemingly better health to optimize compliance with LN surveillance and treatment.

Despite recent treat-to-target guidelines recommending HRQOL optimization, only a few LN trials have incorporated HRQOL outcomes [4, 36, 37]. Using SF-36, Grootscholten et al. found significant improvements in HRQOL at one year of treatment with either cyclophosphamide or azathioprine with methylprednisolone but found few differences between these two arms [38]. Rovin et al. reported no difference in SF-36 physical function scores after one year of rituximab compared with placebo [39]. Furie et al. did not identify an improvement in any SF-36 domain with abatacept over placebo at one year [40]. Similarly, Askanase et al. found no significant difference in SF-36 physical and mental component summary scores between abatacept and placebo at 24 weeks [41]. In contrast, belimumab resulted in significant improvements in the SF-36 physical component summary score in the phase 3 trial of this drug, an effect that was sustained over six years in a continuation study [42, 43]. Other randomized trials such as the recent phase 3 voclosporin and phase 2 obinutuzumab trials did not include HRQOL as a secondary outcome [44, 45]. Importantly none of these studies examined whether extrarenal manifestations confounded HRQOL outcomes, though the belimumab study did adjust for overall SLEDAI. The association of extrarenal symptoms with PROMIS-29 scores in AMP suggests that future trials incorporating HRQOL outcome measures should consider stratifying on extrarenal disease, specifically arthritis.

The strengths of this study were that it was prospective, multicentre, and included a diverse multiracial/ethnic cohort with a range of disease activities who were observed in a real-world setting. Moreover, this study used PROMIS-29 which is likely to be one of the preferred patient-reported outcome measures moving forward including in clinical trials [46, 47]. Several limitations are also acknowledged. The PROMIS-29 questionnaire may not be sensitive enough to detect diminished HRQOL in pure renal patients, and the possibility that these patients report subjective symptoms not captured on PROMIS-29 to physicians during clinical encounters cannot be discounted. Patients were evaluated at academic centres by rheumatologists specializing in SLE which likely helped to ensure the accuracy of SELENA-SLEDAI assessments, however, this may limit the generalizability of the findings when considering patients followed in other settings. The association between HRQOL and component extrarenal manifestations was limited by a small sample size. We did not correct for multiple statistical comparisons due to the exploratory nature of the study and limited sample size raising the possibility of type I error. Data on precise disease duration, education level, socioeconomic status, and comorbidities including fibromyalgia which have been shown to be associated with HRQOL outcomes were not available [4, 48–50]. This analysis relied on first vs repeat biopsy as a proxy for a measure of disease duration because data on other factors that capture longstanding illness such as damage indices were not recorded. Missing information on very rare SLE manifestations not captured on the SELENA-SLEDAI (such as gastrointestinal involvement) might misclassify patients, but this is unlikely. Due to a lack of sufficient follow-up PROMIS-29 data, a reliable longitudinal assessment could not be performed. Since this was a cross-sectional analysis, only associations and not causal inferences can be obtained. Future longitudinal studies are needed to further understand the relationship between extrarenal manifestations and HRQOL in LN patients.

In summary, this study showed that isolated renal disease is common in patients with LN and that these patients may not report decreased HRQOL. Comparatively, patients with extrarenal disease, specifically arthritis, report significantly worse HRQOL as measured by PROMIS-29. These results reinforce the critical importance of routine laboratory surveillance for nephritis and education concerning medication adherence even in SLE patients with seemingly quiescent nonrenal clinical disease. These findings also suggest that the presence of extrarenal manifestations is associated with a higher burden of illness including increased physical limitations and pain interference, highlighting the importance of comprehensive disease management strategies that address both renal and extrarenal manifestations to improve overall patient outcomes.

Supplementary material

Supplementary material is available at Rheumatology online.

Data availability

Data underlying this article will be deposited in the public domain by the AMP RA/SLE Network at a future date.

Funding

This work was supported by the Accelerating Medicines Partnership (AMP) in Rheumatoid Arthritis and Lupus Network. AMP is a public-private partnership (AbbVie Inc., Arthritis Foundation, Bristol-Myers Squibb Company, Foundation for the National Institutes of Health, GlaxoSmithKline, Janssen Research and Development, LLC, Lupus Foundation of America, Lupus Research Alliance, Merck Sharp & Dohme Corp., National Institute of Allergy and Infectious Diseases, National Institute of Arthritis and Musculoskeletal and Skin Diseases, Pfizer Inc., Rheumatology Research Foundation, Sanofi and Takeda Pharmaceuticals International, Inc.) created to develop new ways of identifying and validating promising biological targets for diagnostics and drug development Funding was provided through grants from the National Institutes of Health (UH2-AR067676, UH2-AR067677, UH2-AR067679, UH2-AR067681, UH2-AR067685, UH2- AR067688, UH2-AR067689, UH2-AR067690, UH2-AR067691, UH2-AR067694 and UM2- AR067678).

Disclosure statement: The authors have declared no conflicts of interest.