Glucocorticoid Discontinuation in Patients With Systemic Lupus Erythematosus With Prior Severe Organ Involvement: a Single-center Retrospective Analysis


 Background:Long-term glucocorticoid use in systemic lupus erythematosus (SLE) may have significant side effects; however, glucocorticoid discontinuation is occasionally associated with disease flare-ups. Therefore, we evaluated the risk factors for disease flares and the flare rate upon glucocorticoid tapering in patients with prior severe organ involvement.Methods:Patients with SLE with glucocorticoid tapering at our institution were retrospectively analyzed. We divided the patients according to the presence of prior severe organ involvement and compared flare rates and time to first flare after glucocorticoid discontinuation. Furthermore, we determined risk and protective factors for flares after glucocorticoid discontinuation.Results:A total of 309 patients with SLE were screened; 298 had prednisolone tapered to less than 7.5 mg/day and 75 had glucocorticoid discontinuation. Overall, 73 patients met the inclusion criteria; 49 were classified as SLE with prior severe organ involvement. No statistical differences were noted in the 52-week flare rate and time to first flare after glucocorticoid discontinuation between patients with and without prior severe organ involvement (52-week flare rate: 16.7% vs. 18.2%, p = 1.0; time to first flare: 322 [280, 1169] vs. 385 [304, 2345] days, p = 0.33). Hypocomplementemia, elevated anti-dsDNA antibody titers of more than twice the upper limit of the laboratory reference range, and positive anti-Smith/anti-ribonucleoprotein antibody were negatively associated with flare-free remission.Conclusion:Glucocorticoid discontinuation can often be achieved in patients with SLE without increasing flare risk in most patients with prior severe organ involvement, especially when the disease is clinically and serologically stable.

classi ed as SLE with prior severe organ involvement. No statistical differences were noted in the 52- week are rate and time to rst are after glucocorticoid discontinuation between patients with and without prior severe organ involvement (52-week are rate: 16.7% vs. 18.2%, p = 1.0; time to rst are: 322 [280,1169] vs. 385 [304,2345] days, p = 0.33). Hypocomplementemia, elevated anti-dsDNA antibody titers of more than twice the upper limit of the laboratory reference range, and positive anti-Smith/antiribonucleoprotein antibody were negatively associated with are-free remission.

Conclusion:
Glucocorticoid discontinuation can often be achieved in patients with SLE without increasing are risk in most patients with prior severe organ involvement, especially when the disease is clinically and serologically stable.

Background
Glucocorticoids have been used as the mainstay of treatment for systematic lupus erythematosus (SLE), with a majority of patients undergoing long-term treatment. In a previous study of 215 patients with SLE, 214 patients used glucocorticoids for SLE treatment and 86% used glucocorticoids as maintenance therapy [1]. Glucocorticoids are associated with several long-term side effects, such as hypertension, diabetes, infection, and osteoporosis. As the cumulative glucocorticoid dosage increases, the rate of organ damage increases simultaneously [2]. It is noteworthy that even small amounts of glucocorticoids (prednisolone [PSL] equivalent > 2.5 mg/day) are associated with organ damage, including osteoporosis, fractures, and infection [3][4][5].
Treat-to-target in SLE aims to use the lowest possible glucocorticoid dose to control disease activity and to discontinue glucocorticoids as soon as feasible [6]. The European League Against Rheumatism (EULAR) also recommends reduction of glucocorticoid dose to < 7.5 mg/day (PSL equivalent) or discontinuation of glucocorticoid once disease activity has stabilized [7].
Many clinical trials have investigated the possibility of glucocorticoid discontinuation. Masthian et al. compared SLE are rates between patients who abruptly discontinued glucocorticoids and those who were maintained on low-dose glucocorticoids (PSL equivalent 5 mg/day) and found that the low-dose maintenance group had a lower are rate [8]. However, these data are di cult to apply in clinical practice, given that in our clinical practice, we follow stable patients every three months and gradually taper off low-dose glucocorticoid to avoid are. Although gradual glucocorticoid discontinuation has been suggested and can be successful [9][10][11], to the best of our knowledge, the predictive factors for subsequent are-free state have not been elucidated to date.
In addition, SLE patients with severe organ involvement tend to start treatment with high-dose glucocorticoids, and many clinicians hesitate to taper or discontinue glucocorticoids in such patients, so glucocorticoid-induced organ damage tend to be more severe. Little is known about glucocorticoid discontinuation in patients with prior severe organ involvement.
To address this gap in knowledge, this study investigated the difference in are-free remission rate and are-free duration in the presence of prior severe organ involvement and aimed to determine the factors contributing to a are-free state after glucocorticoid discontinuation.

Study design and participants
This study was a single-center retrospective analysis conducted using electronic health records of SLE patients who were followed up at a Japanese tertiary teaching hospital between January 2006 and March 2021. Patients were followed up for more than 52 weeks at our institution, and at least 6 months of follow-up was completed after glucocorticoid discontinuation.  [12][13][14]. We used three major classi cation criteria when enrolling SLE patients because we had previously established that diagnoses based solely on the 2019 EULAR-ACR classi cation criteria could occasionally miss SLE patients with a low anti-nuclear antibody titer [15].
Glucocorticoid tapering and data collection Glucocorticoids were gradually tapered off after the patients attained clinical remission. The timing and speed of glucocorticoid tapering was determined by the treating rheumatologists.
For each patient, we collected demographic information (age at glucocorticoid discontinuation, sex, Furthermore, we followed the included patients and monitored the occurrence of SLE ares until the time to rst are after glucocorticoid discontinuation or until the end of the study period (March 31, 2021).
We de ned patient with any of the following conditions as patient with prior severe organ involvement: renal manifestation (protein urea level > 0.5 g/24 h, cellular casts due to lupus, or biopsy proven lupus nephritis), neuropsychiatric SLE (delirium, psychosis, seizure, myelitis, or peripheral/cranial neuropathy due to lupus), prior treatment with mPSL pulse therapy, prior treatment with PSL 1 mg/kg/day, and prior treatment with rituximab (RTX) or cyclophosphamide (CY).
We divided the patients into the following two groups: patients with SLE with prior severe organ involvement and patients with SLE without prior severe organ involvement. Then, we compared the are rates and are-free duration after glucocorticoid discontinuation between the two groups.

Results
At our institution, 309 patients with SLE were treated with glucocorticoids and were followed up for more than 52 weeks between January 2006 and March 2021. Of these, 298 (96.4%) had their PSL tapered to ≤7.5 mg/day, and 75 (24.3%) discontinued glucocorticoids and completed at least 6 months of followup. Two patients were excluded due to missing data on prior treatment regimens. Finally, 73 patients were included in our study, of which, 49 were classi ed as having SLE with prior severe organ involvement and the remaining 24 were classi ed as having SLE without prior severe organ involvement ( Figure 1).
Baseline characteristics of each study group are summarized in Table 1. There were no statistically signi cant differences in age at the time of glucocorticoid discontinuation, female ratio, body mass index, and antibody pro les (anti-dsDNA antibody, anti-Sm antibody, anti-Ro/SSA antibody, anti-RNP antibody, lupus anticoagulant, anti-CL antibody, anti-CLβ2GPI antibody) between the two groups. Although disease duration before glucocorticoid discontinuation tended to be longer in patients with prior severe organ involvement, no statistical differences were noted (4367 [ [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] mg/day, p < 0.01). The detailed treatment regimen during the follow-up period is summarized in Table 2 and Additional le 1. On the day of glucocorticoid discontinuation, SLE patients with prior severe organ involvement tended to use hydroxychloroquine (HCQ), Tac, CyA, MMF, AZA, but no statistical difference was noted. MZR is an inhibitor of purine synthesis, with a mechanism of action similar to that of MMF, and is a widely used medication in Japan. Since we have proven the e cacy of MZR in IgG4-related disease with multiple organ involvement and of MZR/tacrolimus combination therapy in lupus nephritis [17,18], the use of MZR tended to be higher in patients with prior severe organ involvement than in those without, but no statistical differences were noted. As shown in Table 3 and Additional le 2, on the day of glucocorticoid discontinuation, median SLENA-SLEDAI was 0.0 [0.0, 2.0] in patients with prior severe organ involvement and 0.0 [0.0, 0.0] in patients without prior severe organ involvement.   Prior severe organ involvement did not affect the time to rst are after glucocorticoid discontinuation, as this time was 322 (280-1169) days in patients with prior severe organ involvement vs. 385 (304-2345) days in those without prior severe organ involvement (p = 0.33; Table 3).
As shown in Figure 2a, more than 80% of the patients achieved 52 weeks of are-free remission and more than 70% achieved 1000 days of are-free remission after glucocorticoid discontinuation. When we compared patients by the presence or absence of prior severe organ involvement, the are-free rate after glucocorticoid discontinuation in patients with prior severe organ involvement was lower than that in patients without prior severe organ involvement in the rst year. However, the are rate stabilized over time (Figure 2a), and the log rank test results showed that there was no statistical difference between the two groups by the end of the follow-up period (p = 0.73).
We de ned resuming glucocorticoids as another de nition of are and assessed glucocorticoid-free remission rate. As shown in Figure 2b, although patients with prior severe organ involvement tended to restart glucocorticoids, no statistical difference was noted between the two groups (p = 0.74).
SLE are rate (52 weeks) Of the 73 patients, 1 patient with prior severe organ involvement and 2 patients without prior severe organ involvement did not complete the entire 52-week follow-up after glucocorticoid discontinuation, and these patients could not be monitored for SLE ares during the follow-up. Therefore, we excluded these 3 patients from the 52-week evaluation for SLE are rates. Of the remaining 70 patients, 8 (16.7%) with prior severe organ involvement and 4 (18.2%) without prior severe organ involvement had ares during the 52 weeks of follow-up after glucocorticoid discontinuation; however, no statistical difference was noted between these patients (p = 1.0; Table 3).
A detailed electronic record search revealed that most ares were mild. However, 3 patients (6.3%) with prior severe organ involvement and 1 patient (4.5%) without prior severe organ involvement experienced a lupus nephritis are or a lupus are that necessitated re-initiation of glucocorticoid therapy at a dose of >0.5 mg/kg/day. Of the 3 patients with prior severe organ involvement, 1 experienced a lupus nephritis are 280 days after glucocorticoid discontinuation and was managed by restarting PSL at 25 mg/day, adding tacrolimus, and subsequently tapering prednisolone to 2.5 mg/day. The second patient experienced a are of lupus myopathy 313 days after glucocorticoid discontinuation. Although the patient had to start PSL at 40 mg/day, the dose was subsequently decreased to 6 mg/day. The third patient experienced a mild are of lupus nephritis 70 days after glucocorticoid discontinuation and was initiated on belimumab.
The one patient without prior severe organ involvement experienced fever, arthritis, leukocytopenia, and thrombocytopenia on day 301 after glucocorticoid discontinuation and was managed by restarting mPSL at 30 mg/day. Prednisolone was discontinued 251 days after the occurrence of rst are. The details of each are are summarized in Additional le 3.
Factors that tended to decrease the are-free ratio, but without statistical signi cance possibly due to the small sample size, were: SLE duration of more than 5000 days (patient with are 15.8% vs. patient without are 38.9%, p = 0.09), HCQ treatment on the day of glucocorticoid discontinuation (36.8% vs. 53.7%, p = 0.29), and LLDAS achievement on the day of glucocorticoid discontinuation (89.5% vs. 96.3%, p = 0.28). We then evaluated these variables using the Cox proportional hazard model, which showed that are rate was not in uenced by the presence of prior severe organ involvement (hazard ratio

Discussion
Our ndings showed that more than 95% (298 of 309 patients with SLE followed for more than 52 weeks at our institution) of patients with SLE could achieve a good response with a PSL dose of ≤7.5 mg/day at least once. Therefore, as recommended by EULAR, SLE treatment with PSL at ≤7.5 mg/day is a realistic goal for most patients with SLE [7]. Furthermore, glucocorticoid discontinuation was achieved in 75 of 309 patients (24.3%) with SLE. Of the 73 patients included in our study, more than 80% achieved 52 weeks of are-free remission and more than 70% achieved 1000 days of are-free remission after glucocorticoid discontinuation. These results were not in uenced by the presence of prior severe organ involvement. Furthermore, our results suggested that hypocomplementemia and elevated anti-dsDNA antibody titer at more than twice the upper limit of the laboratory reference range on the day of glucocorticoid discontinuation as well as positive anti-Sm antibodies and anti-RNP antibodies can be risk factors for ares after glucocorticoid discontinuation.  [11]. These are rates are in agreement with the rates noted for our cohort for both the patient groups.
This study showed that glucocorticoid doses can be safely tapered in patients with and without prior severe organ involvement. Regarding the risk factors for ares after glucocorticoid discontinuation, our results suggest that hypocomplementemia and elevated anti-dsDNA antibody at more than twice the upper limit of the laboratory reference range on the day of glucocorticoid discontinuation as well as positive anti-Sm antibodies and anti-RNP antibodies may be important factors in this regard.
As suggested by the SLICC classi cation criteria, we used anti-dsDNA antibody titers at more than twice the upper limit of the laboratory reference range as a risk factor for ares because low titers of anti-dsDNA antibody have low sensitivity for SLE activity [13].
A previous study on B-cell depression therapy illustrated that the one-year are rates were higher in SLE patients with anti-RNP or anti-Sm antibody positivity than in patients without anti-RNP or anti-Sm antibody positivity [19]. This is possibly due to the fact that anti-RNP/Sm antibodies are produced by long-lived plasma cells [20] that cannot be fully eliminated by rituximab, similar to that in the case with glucocorticoids [21][22][23].
LLDAS achievement on the day of glucocorticoid discontinuation was not statistically signi cant because more than 90% of the patients achieved LLDAS on the day of glucocorticoid discontinuation.
Nevertheless, on the day of glucocorticoid discontinuation, LLDAS achievement ratio was higher in patients without ares than in patients with ares (96.3% vs. 89.5%), thus LLDAS should be targeted before glucocorticoid discontinuation.
Previous reports have suggested that HCQ reduces the are risk and mortality of SLE patients [24][25][26].
However, HCQ use on the day of discontinuation was not a signi cant predictive factor of are-free state, probably because of the lower HCQ prescription ratio due to the formulary restrictions until 2016. Nevertheless, HCQ use was higher in patients in the are-free group (53.7%) than in patients in the are group (36.8%). In Japan, HCQ was approved for SLE in 2015, and the 2-week dosing period restrictions were removed in 2016. Of the 75 patients who discontinued glucocorticoid, 48 discontinued glucocorticoid between 2016 to 2020 (after the approval of HCQ) and 66.0% of them used HCQ on the day of glucocorticoid discontinuation. This prescription rate of HCQ was consistent with that stated in the Hopkins Lupus Cohort report, which included 2054 patients with SLE with an HCQ prescription rate of 64.4% [27].
In addition, our data showed that the number of patients who achieved disease stability after glucocorticoid discontinuation increased after HCQ approval. Of the 75 patients who discontinued glucocorticoids, 8 discontinued glucocorticoids between 2006 and 2010, 19 between 2011 and 2015, and 48 between 2016 and 2020. Therefore, the glucocorticoid discontinuation rate and are-free remission rate after glucocorticoid discontinuation are expected to increase in the future.
Currently, many clinicians, including our group, are attempting to reduce or discontinue the use of glucocorticoids, immunosuppressants, and biologic agents to avoid the associated adverse effects and reduce the medical burden and pregnancy-related complications [8][9][10][11][28][29][30].
Treatment with glucocorticoids is associated with many side effects, and even small amounts of glucocorticoids (prednisolone [PSL] equivalent > 2.5 mg/day) are associated with osteoporosis and an increased risk of fractures [5].  [31]. Furthermore, glucocorticoids not only impose a nancial burden on patients but also increase the rate of all-cause mortality as the cumulative glucocorticoid dosage increases [32].
In addition, glucocorticoid use during pregnancy is related to adverse pregnancy outcomes, including Glucocorticoid discontinuation can be achieved in an increasing number of patients with SLE with the current treatment regimens, and the presence of prior severe organ involvement does not in uence the 52week are-free remission after glucocorticoid discontinuation.
Rheumatologists should aim to taper glucocorticoid dosages as much as possible in patients with SLE.
We believe that our results support the idea that more patients with SLE, including those with prior severe organ involvement, can achieve freedom for glucocorticoid use, thereby reducing glucocorticoid-related adverse events, nancial burden, and adverse pregnancy outcomes.
The limitations of this study include its single-center retrospective cohort study design that included a limited number of patients and thus could not exclude cofounders. Second, since we included sulphasalazine, iguratimod, and bucillamine as immunosuppressants, we might have overestimated the are rate after glucocorticoid discontinuation. Finally, we could not include patients on belimumab therapy on the day of glucocorticoid discontinuation. Belimumab has only been available to patients in Japan since 2017; therefore, in the future we will need to re-analyze the data in patients treated with belimumab.

Conclusions
Our results suggest that more than 80% of patients who gradually discontinued glucocorticoid achieved 52 weeks of are-free remission even with prior severe organ involvement. Hypocomplementemia and elevated anti-dsDNA antibody titers of more than twice the upper limit of the laboratory reference range on the day of glucocorticoid discontinuation were risk factors for subsequent ares.

List Of Abbreviations
Ab: antibody Ethics approval and consent to participate: The study was approved by the Ethics Committee of St. Luke's International Hospital (approval no. 20-R-23). Written informed consent was obtained from all participants in this study.

Consent for publication:
Written informed consent was obtained from all participants in this study.
Availability of data and materials: All data generated or analyzed during this study are included in this published article [and its supplementary information les].
Competing interests: MO has received speaking fees and/or honoraria from Eli Lilly and Company, Santen Pharmaceutical, Mitsubishi Tanabe Pharma, P zer, and Abbott Japan. Other authors declare that they have no competing interests.

Funding:
No speci c funding was received from the public, commercial, or not-for-pro t sectors for the work described in this article.

Authors' contributions:
TN was involved in study conception and design, acquisition of the data, data analysis, drafting the manuscript, and nal approval of the article. SF was involved in acquisition of the data, statistical advice, and approval of the nal article. YI, MS, and HT were involved in acquisition of the data, and nal approval of the article. MO was involved in study conception and design, drafting the manuscript, and nal approval of the article. All authors read and approved the nal manuscript.  Figure 1 Patient owchart A total of 309 patients with SLE were treated with glucocorticoids and were followed up for > 52 weeks. Of these, 298 had their PSL dosage tapered to ≤ 7.5 mg/day and 75 discontinued glucocorticoids. Two patients were excluded from the study because of the lack of data on prior treatment regimen. Among the 73 patients nally included in our study, 49 were classi ed as having SLE