Frequency and predictors of the Lupus Low Disease Activity State in Chinese patients with systemic lupus erythematosus: a prospective observational cohort study

Background Systemic lupus erythematosus (SLE) is a chronic autoimmune disorder with significant risk of organ damage. Lupus Low Disease Activity State (LLDAS) was generated as a potential treatment target for SLE. Previous studies had validated that the maintenance of LLDAS can predict reduced damage accrual or flare. The objective of this study was to estimate the time to LLDAS and the predictors of time to LLDAS in our prospective observational cohort of Chinese patients with SLE. Methods Patients were from Peking University First Hospital SLE cohort and those having not fulfilled LLDAS at enrolment were included in this study. The time to LLDAS and annual cumulative probabilities of LLDAS achievement were estimated by the Kaplan-Meier approach. The predictors of time to LLDAS were identified by univariate and multivariable Cox proportional hazards. Results A total of 574 patients with SLE were included and 435 (75.8%) of them achieved LLDAS during a median 4.2 years of follow-up. The median time to LLDAS was 19.0 months and the cumulative probabilities at 1, 2, 3, 5 and 10 years were 19.8%, 57.6%, 72.0%, 85.1% and 98.0%, respectively. In multivariable Cox models, older age at disease onset, treatment-naïve and hydroxychloroquine prescription were found to be independent predictors of shorter time to LLDAS, after adjusted by daily prednisone dose, SLE Disease Activity Index 2000 and physician’s global assessment. Finally, we developed a matrix model based on the identified independent predictors to present the time to LLDAS in patients with respective characteristics. Conclusions Our study proved that LLDAS is attainable as an early treatment target for SLE in Chinese patients. The older age at disease onset, treatment-naïve and

Long-term observational studies have demonstrated that disease activity was positively associated with the risk of subsequent organ damage and mortality [5,6].
In accordance with these findings, remission and the lowest possible disease activity were set as the treatment targets of SLE in the treat-to-target recommendations released in 2014 [7].
Recently, a framework for remission in SLE has been established by a large international task force, however, the framework contained four different definitions and no specific definition of remission has been widely accepted by now [8].
Until now, the only one study which estimated the attainability of LLDAS was the Hopkins Lupus Cohort study consisting of Caucasian and African-American SLE patients [26]. Several studies found that Asians had significant differences with Caucasians in terms of disease activity, renal involvement, autoantibody-positive rate, organ damage rate and mortality rate [27][28][29][30][31][32]. Moreover, studies by Asia-Pacific Lupus Collaboration showed that the possibility of LLDAS ≥ 50% time during follow-up was lower in Asian population than that in Caucasians, though it did not reach statistical difference [22,23]. No data about the attainability of LLDAS achievement in Asian ethnicity has been reported so far.
In this study, we estimated the time to LLDAS and the predictors of time to LLDAS in our prospective observational cohort of Chinese patients with SLE.

Patients
The Peking University First Hospital SLE (PKUFHS) cohort is a single-centre prospective observational cohort since 2007. Patients were diagnosed according to the revised American College of Rheumatology (ACR) criteria or the Systemic Lupus International Collaborating Clinics (SLICC) criteria. After the recruitment, patients were regularly followed up every 1 to 3 months when the disease was active, or every 3 to 6 months when the disease was stable, which was determined by the 6 rheumatologists.
In this study, patients from the PKUFHS cohort recruited between January 2007 and December 2018 were included if they meet all of the following criteria: (a) did not fulfil LLDAS at baseline; (b) had no other connective tissue diseases except Sjogren's syndrome (SS) or antiphospholipid syndrome (APS), inflammatory arthropathy or other diseases exerting obvious effect on disease activity assessment and GCs tapering; (c) had at least 3 follow-up visits and 6 months of follow-up duration; and (d) had no interval of more than 12 months between two consecutive visits.
This study was approved by the ethics committee of Peking University First Hospital and all patients signed the informed consent at enrolment.

Data collection
Data were collected at outpatient visits and hospital admissions from January 2007 to June 2019. Gender, age at disease onset and disease duration were recorded with baseline manifestations, laboratories and complications, including SS and APS.
Age at disease onset was defined as the age when the symptom or laboratory abnormality related to SLE was initially presented. Disease duration was defined as the time from disease onset to the recruitment. Early SLE was defined when the disease duration was no more than 1 year. Treatment-naïve was defined as no previous glucocorticoid or immunosuppressants (IMs) therapy. Clinical manifestations and laboratories were defined and recorded with the reference of the SLE Disease Activity Index 2000 (SLEDAI-2K) [33,34]. Additionally, nephritis was defined when either of proteinuria, haematuria, sterile pyuria or urinary casts was presented. Serositis was defined as pleurisy or pericarditis. Anaemia was defined when the haemoglobin was less than 12 g/dl in women and 13 g/dl in men according to the WHO definition [35].
The disease activity of SLE was assessed by 1 rheumatologist using the SLEDAI-2K and the physician's global assessment (PGA) (range 0-3.0) at baseline and each visit [33,36]. Organ damage of each patient was determined by the SLICC/ACR Damage Index (SDI) at baseline [37]. The daily prednisone (or equivalent) dose, and hydroxychloroquine (HCQ) and IMs usage at baseline and each visit were also collected.
During follow-up, 435 (75.8%) patients achieved LLDAS at least once. The baseline characteristics and the comparison between patients who achieved and did not achieve LLDAS during follow-up were shown in Table 1. The patients who achieved LLDAS had younger age at disease onset (p=0.027), shorter disease duration (p=0.004), higher frequency of treatment-naïve (p<0.001), lower frequency of nephritis (p=0.005), lower SDI (p=0.004), less IMs usage (p=0.005), and lower level of SLEDAI (p<0.001) and PGA (p=0.010) at baseline compared with those who never achieved LLDAS during follow up.

Predictors of time to LLDAS
The predictors at baseline of time to LLDAS were identified by the univariable and multivariable Cox regression analyses.
In univariable analysis, baseline variables including male, age at disease onset, early SLE, treatment-naïve, complicating with SS, arthritis and HCQ prescription were positively associated with LLDAS achievement, while nephritis, serositis and hypocomplementemia, high SLEDAI, PGA and SDI, as well as IMs prescription were negatively associated with LLDAS achievement (  This study demonstrated that LLDAS is attainable as a treatment target for Chinese patients with SLE. In our SLE cohort, the time to LLDAS was 1.7 years, which was longer than that in the Johns Hopkins Lupus Cohort (1.1 years) mainly consisting of Caucasian and African-American patients [26]. The difference may be partly due to the different disease activity at baseline. In their cohort, 50.7% of patients had PGA ≤1 and 53.8% had SLEDAI ≤4 at baseline, while the percentages in our patients were only 28.7% and 24.0%, respectively.
Our study confirmed that baseline disease activity could strongly influence the LLDAS achievement. Some previous studies found that patients with LLDAS ≥50% of follow-up time had a lower level of SLEDAI at baseline [14,22,23], and other studies found that baseline SLEDAI ≤4 was a predictor of achieving LLDAS and maintaining LLDAS for 5 years [15,26]. The present study showed that patients with lower disease activity (by PGA or SLEDAI-2K) at baseline were more quickly to achieve LLDAS. Particularly, some components of SLEDAI including active nephritis, serositis and hypocomplementemia were found to be predictors of achieving LLDAS with longer time, while arthritis was found to be predictors of earlier LLDAS achievement.
Our study indicated that the age at disease onset of SLE could influence the LLDAS achievement. Previous researches had demonstrated that patients with juvenile-onset SLE had higher disease activity, received more intensive prednisone and IMs therapy, had more organ damage and a two-fold higher mortality rate, compared with adult-onset patients [38][39][40][41][42]. Recently, some studies reported that patients who were in LLDAS ≥50% time or observations during follow-up had older age at diagnosis of SLE [14,22,23]. In the present study, patients with older age at disease onset were demonstrated to have shorter time to LLDAS after adjusting other factors, including daily prednisone dose, SLEDAI and PGA.
It should be noted that early treatment was a strong positive predictor for LLDAS achievement. Babaoglu et al found that patients with disease duration ≤1 year had shorter time to LLDAS [26]. Our study confirmed the association and further found that treatment-naïve was also a positive predictor of LLDAS achievement. Moreover, patients with organ damage before recruitment were found to require longer time to achieve LLDAS. These evidences indicated early treatment could provide more opportunities for quick achievement of LLDAS.
Our study revealed that the patients who received HCQ as an initial treatment were able to achieve LLDAS faster. HCQ has been used in SLE for over half a century. was a strong independent predictor of shorter time to LLDAS, even after adjusted by disease activity factors. Apart from HCQ, other treatments were also evaluated in the present study. We found that those patients who used IMs at baseline required longer time to achieve LLDAS. A study from Tsang et al also showed that patients using IMs at baseline had a lower possibility of maintaining LLDAS ≥50% time during follow-up [14].
We found that complicating SS in SLE patients predicted a shorter time to LLDAS. A study by Baer et al showed that patients with both SLE and SS were a distinct subset with older age at SLE diagnosis and lower frequency of renal involvement [50], and our study revealed this overlap was also a factor of earlier achievement of LLDAS.
There were some limitations in this study. As an observational study, our data can hardly be used to evaluate the effect of different IMs on achieving LLDAS in SLE.
Data from prospective intervention studies were needed to determine which medication could help to achieve LLDAS with less time. In addition, as all of the patients in this cohort were Chinese ethnicity, the conclusions may not be extrapolated to other ethnicities.

Conclusions
Our study proved that LLDAS is attainable as an early treatment target for SLE in

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
None.

Authors' contributions
Zhuoli Zhang and Yanjie Hao contributed to the design of the cohort and this study,