Comparison of renal remission and relapse‐free rate in initial‐ and delayed‐onset lupus nephritis

Abstract Introduction Lupus nephritis (LN) is a major manifestation of systemic lupus erythematosus (SLE) which contributes to significant morbidity and mortality. It is unclear whether the timing of LN onset influences renal outcome. This study aimed to investigate differences in clinical features—particularly the relapse‐free rate—in remission duration from induction therapies for LN and the onset timing of LN after the development of SLE. Methods We enrolled 66 LN patients from January 2004 to March 2020. We collected the following: demographic data, laboratory data, renal histology data, and LN induction therapy data. Renal remission and relapse‐free rates were calculated for each group. Results Patients were first divided into early remission group (achieved renal remission in <12 months [n = 44]) and others (n = 22). There were no significant differences in clinical data, treatments, and relapse‐free rate of LN. Patients were then divided into initial‐onset LN (<12 months after development of SLE [n = 49]) and delayed‐onset LN (≥12 months after development of SLE [n = 17]). Kaplan–Meier analysis showed that the relapse‐free rate was significantly higher in all patients with initial‐onset LN than those with delayed‐onset LN (P = .0094). Conclusion The relapse‐free rate was significantly higher in the initial‐onset LN group than the delayed‐onset LN group of patients with LN of various histopathological backgrounds. These data suggest that delayed‐onset LN is a risk factor for the relapse of LN.


| INTRODUC TI ON
Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by a loss of self-tolerance and formation of nuclear autoantigens and immune complexes, resulting in inflammation of multiple organs. 1 Lupus nephritis (LN) is a major manifestation of SLE that contributes to significant morbidity and mortality. 2 Patients with LN have higher mortality rates than SLE patients without LN. 3  Association (EDTA), 6 and Asian Lupus Nephritis Network. 7 EULAR updated the management recommendation for SLE in 2019. 8 Those clinical guidelines recommend performing a renal biopsy in order to obtain renal histology unless strongly contraindicated. Also, treatment should be based on the type of LN, as classified by the International Society of Nephrology/Renal Pathology Society (ISN/ RPS) criteria. 9,10 Additionally, in induction therapies for classes Ⅲ or Ⅳ LN, the use of mycophenolate mofetil (MMF) or intravenous cyclophosphamide (IVCY), along with glucocorticoids, was recommended based on various clinical trials including the Aspreva Lupus Management Study. 11 Further, it has been reported that renal response after 6 months of treatment with IVCY can predict long-term renal outcomes based on data obtained in the Euro-Lupus Nephritis Trial. 12 Reports also suggest that, in Japan, achieving early renal remission of classes Ⅲ or Ⅳ LN using glucocorticoids with immunosuppressants might predict good outcomes, such as reduced organ damage and a low incidence of disease flare. 13,14 On the other hand, some patients develop SLE and LN simultaneously, while others develop LN after an SLE diagnosis without renal involvement. According to several reports, the number of SLE patients who developed LN later was fewer compared to those who were diagnosed with LN simultaneously or within a few years. [15][16][17][18] It is unclear whether the timing of the onset of LN influences renal outcome. Reports from Ugolini-Lopes et al 19

and
Dlifino et al 20 showed that no major differences were noted when disease profile or treatment outcome of early-and late-onset LN were compared. On the other hand, several reports from Japan showed that SLE patients with early-onset LN had better renal outcomes when compared to those with late delayed-onset LN. 18,21,22 Because the results were different, it is unclear how the timing of LN onset affects its pathophysiology. Although a renal biopsy is recommended for diagnosing and treating LN, it is difficult to achieve in patients at risk of bleeding, in poor general condition, or who refuse it. In clinical settings, it is often difficult to treat and predict the prognosis of patients with LN

| Statistical analysis
Continuous values are shown as median and interquartile range (IQR). A nonparametric Mann-Whitney U test was used for intergroup comparisons of multiple variables. Fisher's exact test was used to investigate a possible association between each variable.
Kaplan-Meier method was used to calculate the rate of remission and relapse-free rate, while a log-rank test was used to assess differences between the 2 groups. GraphPad Prism 5 software (GraphPad Software, San Diego, CA) was used to perform all of the statistical analyses. The significance level was set at P <.05.

| Ethics approval and consent to participate
The study was approved by the Ethics Committee of Fukushima Medical University (No. 30155).

| Patient characteristics
The demographic and disease-related features of the enrolled 66 patients are as follows ( Table 1)

| Comparison of LN patients according to time of renal remission
Enrolled SLE patients were initially divided into 2 groups based on the time of renal remission. Since the Euro-Lupus Nephritis Trial results showed that early response to therapy at 6 months was the best predictor of good long-term renal outcomes, 12 we categorized the patients as follows. Patients who achieved renal remission in <6 months were defined as early remission, and others (ie, patients who achieved renal remission at ≥6 months and those who did not achieve remission) were defined as late remission. The demographic, disease-related features, renal histopathology, and induction therapies were summarized in each group and compared ( Table 2): 44 patients (66.7%) achieved early remission, while 22 (33.3%) achieved late remission. There were no significant differences between the 2 groups regarding disease-related features at baseline. No significant difference was observed in terms of the observation period between the 2 groups, and 1 patient in the late remission group did not achieve partial remission during the observation period. There were no significant differences regarding kidney histopathology of LN, although 14 patients of early remission and 7 patients of late remission did not have renal biopsies conducted for various reasons. In induction therapies for LN, more patients were treated with mPSL pulse therapy in the early remission group, but there was no significant difference. A Kaplan-Meier analysis showed that the relapse-free rate after induction therapies was not significantly different between the early and late remission groups (P = .1202, log-rank test) ( Figure 1A).
Regarding medication of induction therapies, there were no sig-

| Comparison of LN patients according to timing of LN development from SLE onset
Enrolled SLE patients were then divided into another 2 groups based on timing of LN development from SLE onset. Jacobsen et al 16 and Seligman et al 17 demonstrated that renal manifestation in most SLE patients was observed within a year, and therefore, we divided the patients as follows. Patients who were diagnosed with SLE and LN simultaneously, or LN within <12 months of diagnosis with SLE, were defined as initial-onset LN, and others (ie, patients who were diagnosed with LN after ≥12 months of diagnosis with SLE, and who had already achieved remission of SLE before being diagnosed with LN) were defined as delayed-onset LN (Table 3). There were 49 patients (74.2%) with initial-onset LN, and 17 (25.8%) with delayed-onset LN. Median duration from onset of SLE to onset of LN in delayed-onset LN patients was 41.0 months. No significant difference in observation period was observed between the 2 groups, and 1 patient in the initial-onset LN group did not achieve partial remission during the observation period.
Of the disease-related features at baseline, higher serum albumin levels (P = .0086) and higher serum C3 levels (P = .0477) were significantly related to delayed-onset LN, and significantly higher

| D ISCUSS I ON
This study demonstrated that the relapse-free rate of LN in delayedonset LN patients was significantly low when compared to that of initial-onset LN patients. We also showed that delayed-onset LN patients who were administered immunosuppressants experienced more relapses of LN when compared to initial-onset LN patients who were administered immunosuppressants. We could not show those differences in the study of time of remission (early remission and late remission) in our patients.
In delayed-onset LN patients, serum albumin and C3 were significantly high when compared to initial-onset LN patients.
However, there was no difference regarding the distribution of renal

F I G U R E 1
The relapse-free rate from achieving renal remission between early and late remission patients of lupus nephritis. A: There is no significant difference in the relapse-free rate between early (n = 44) and late (n = 21) remission patients of lupus nephritis (log-rank test, P =.1202). B: There is no significant difference in the relapsefree rate between early (n = 22) and late (n = 14) remission patients of lupus nephritis treated with prednisolone plus immunosuppressants (log-rank test, P =.1429) and 1 cases of global glomerulosclerosis and fibrous crescents were observed among the 6 patients in the no-recurrence group, whereas 3 cases of global glomerulosclerosis and fibrous crescents each were observed among the 7 patients in the recurrence group although the comparison was based only on the presence or absence of findings without considering differences in degree. In this study, we could not show the differences in chronicity findings between the initialand delayed-onset LN groups. Nakano et al stated that the relatively worse long-term renal outcome in delayed-onset LN was primarily because of failure to achieve sustained remission in these patients, and delayed-onset LN might be a potential predictor of poorer treatment outcomes. 18 Some reports state that a poor renal response to initial treatment and renal flares were strongly associated with future renal damages. [26][27][28] It was thought that delayed-onset LN patients were difficult to treat because they had already received immunosuppressive treatments for SLE before development of LN.
As previously reported, one reason why delayed-onset LN was hard to treat was that kidneys of delayed-onset LN patients had more chronic damaged lesions compared to kidneys of initial-onset LN pa- initial manifestation of proteinuria and hematuria, the inclusion of initial-onset LN patients in the delayed-onset LN group could not be ruled out. Therefore, we must focus on and interpret the results of similar studies reported from various countries, and it is necessary to conduct a prospective large-scale, multicenter international collaborative study to verify the findings described here.
In conclusion, our study demonstrated that the relapse-free rate was significantly higher in the initial-onset LN group when compared to the delayed-onset LN group of LN patients of various histopathological backgrounds. These data suggest that delayed-onset LN is a risk factor for relapse of LN. Therefore, it is important that SLE patients who are not complicated by LN also be carefully monitored regarding renal function and urinalysis. When LN develops, a renal biopsy should be conducted and immunosuppressive therapies commenced as soon as possible.

ACK N OWLED G EM ENTS
We would like to thank Enago (www.enago.jp) for the English language review.

CO N FLI C T O F I NTE R E S T
The authors of this work have nothing to disclose.