Optimal treatment targets for lupus nephritis using per-protocol repeat kidney biopsy findings at 2 years and clinical data up to 5 years: a single-center observational study

Abstract

Background:

The criteria for remission in both clinical and pathological contexts for lupus nephritis (LN) remain controversial.

Objectives:

To identify optimal short-term goals (remission criteria) for LN predicting long-term success at 5 years, using repeat kidney biopsy (Biopsy 2) and clinical data.

Design:

Single-center observational study.

Methods:

Twenty-three consecutive LN patients undergoing Biopsy 2 2 years post-induction therapy were evaluated. Two ideal long-term goals at 5 years were defined as: “A,” Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) = 0 and prednisolone (PSL) ⩽5 mg/day, and “B,” proteinuria ⩽0.2 g/day with a normal serum creatinine level and PSL ⩽5 mg/day. Histologically, the electron-dense deposit (EDD) score grades immune deposits based on their intensity, amount, and location. A score of ⩽1 was defined as “electron microscopy remission (ER).”

Results:

Conventional renal indices failed to predict long-term goals. The short-term goals with an accuracy (area under the curve: 95% confidence interval) of ⩾0.8 predicted long-term goals: “A at 5 years” (A-5y), A-2y (0.91: 0.79–1.00), DORIS-R-2y (0.87: 0.72–1.00), EDD score (0.85: 0.70–1.00), B-2y (0.83: 0.66–0.99), and SLEDAI-R-2y (0.82: 0.66–0.98) as well as “B at 5 years” (B-5y), A-2y (0.87: 0.73–1.00), B-2y (0.87: 0.73–1.00), EDD score (0.85: 0.69–1.00), and DORIS-R-2y (0.83: 0.67–0.99). EDD scores predicted A-5y, B-5y, and PSL dose at 5 years in proportion to the score. The clinical and histological goals aligned.

Conclusion:

The best predictive short-term goal was A-2y. Concordance between clinical remission (A-2y, B-2y, and DORIS-R-2y) and histological remission (ER) at 2 years suggests optimal short-term goals for LN.

Keywords

electron microscopy long-term goals lupus nephritis per-protocol repeat kidney biopsy remission criteria short-term goals

Introduction

Lupus nephritis (LN), one of the most intractable conditions in systemic lupus erythematosus (SLE), can lead to unsatisfactory long-term, renal, and quality-of-life outcomes, even with the current standard of care.^1–3 Accurately determining the diagnosis and severity of LN by clinical indicators alone is difficult⁴; thus, histological confirmation with an initial kidney biopsy (Biopsy 1) is indispensable. Predicting long-term outcomes is also difficult before starting treatment and is limited to the chronicity index (CI) at the time of biopsy 1^5,6 and serum creatinine (Cr) level.^7,5 Therefore, reassessing induction therapy at the “appropriate” timing with “appropriate” short-term goals to modify treatment accordingly is an imperative step. Based on many reports that proteinuric response and serum Cr levels at 6–12 months correlate with long-term outcomes,^7–10 various short-term goals (i.e., remission criteria) have been devised with these items. There are conflicting opinions on using urinary sediment and serological indices as remission criteria.^11,12 However, one-third of patients with proteinuria ⩽0.5 g/day at 1 year will still have progressive renal damage¹³; thus, the optimal clinical short-term goals for LN have yet to be established.

Given that clinical indicators alone cannot accurately predict renal histology after the start of induction therapy, a repeat kidney biopsy (Biopsy 2) is performed as needed.^14,15 Due to the low incidence of end-stage renal kidney disease in patients with an activity index (AI) of 0 in Biopsy 2,¹⁶ this was conventionally defined as “histological remission.” However, histological and clinical responses to treatment do not match at any time of Biopsy 2 from 6 months to 3 years later; poor renal prognosis is still observed even in cases with AI = 0.^16–19 AI = 0 is thus not a short-term goal. The optimal histological short-term goal for LN has not been established. Regarding treatment modification, there are two proven benefits of Biopsy 2. First, higher AI correlates with worsening renal function²⁰ and early relapse,¹⁸ findings that can be useful in deciding to intensify therapy. Second, AI = 0 in the maintenance phase at 3–4 years after clinical remission is a determinant for discontinuing maintenance therapy without relapse.^17,21,22

In our experience, proteinuria improvement reached a plateau after 1 year of induction therapy. In contrast, improvement of the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) and glucocorticoid reduction reached a plateau after 2 years.²³ Based on these results, we performed Biopsy 2 at 2 years in patients who provided consent, regardless of clinical remission or nonremission, to histologically confirm the therapeutic response and help determine whether subsequent treatments need to be modified. In clinical practice, Biopsy 2 is often performed in patients with a poor clinical evolution,²⁰ and information on patients in clinical remission at 2 years is minimal. In addition, our department performs electron microscopy on all patients. Thus, this study aimed to identify the optimal short-term goals (remission criteria) and their timing to predict achieving ideal long-term goals at 5 years, using per-protocol repeat kidney biopsy findings at 2 years post-induction therapy and clinical data up to 5 years.

Materials and methods

Patient selection

This is a single-center observational study, conducted and reported in accordance with the STROBE statement. Forty patients were diagnosed with LN by Biopsy 1 at our hospital between July 2009 and September 2015. Excluding four cases deemed inappropriate by their attending physician (classified as Class II, rapidly progressive glomerulonephritis, aged 79, and with coexisting other nephritis), all 36 received remission induction therapy with a first-line combination treatment of mizoribine (MZB), tacrolimus (TAC), and glucocorticoids according to our protocol.²³ Of these, 24 patients consented to undergo Biopsy 2 at a median of 27 (quartile range: 25–34) months after the start of treatment. Twenty-three patients with adequate renal tissue (⩾10 glomeruli) were included in the analysis. All patients continued on their first-line strategy therapy at the time of Biopsy 2.

Two ideal long-term goals (“A” and “B” at 5 years)

There were no deaths at 5 years among 23 patients, and 1 patient underwent hemodialysis. Given the current inadequate long-term prognosis of patients with LN, we considered the conventional “nondialysis” or “serum Cr level of ⩽1”^8,9 as inappropriate goals at 5 years. Thus, we defined the following two criteria from different perspectives as ideal long-term goals at 5 years: “A,” SLEDAI-2K = 0 and prednisolone (PSL) ⩽5 mg/day and “B,” proteinuria ⩽0.2 g/day and normal serum Cr levels and PSL ⩽5 mg/day.

Kidney biopsy and histological outcomes

Using standard techniques, kidney biopsy tissues were processed for light, immunofluorescence, and electron microscopy. Histological evaluation was performed by the following personnel blinded to the patient’s clinical data: light microscopy by our nephrologists, immunofluorescence by our pathologists, and electron microscopy by outsourced pathologists. Light microscopy was used to classify disease types according to the ISN/RPS 2003 classification²⁴ and evaluate disease severity according to the modified NIH scoring system²⁵ of AI and CI. AI ⩽ 1 was defined as “light microscopy remission (LR).” Immunofluorescence microscopy was performed to evaluate the presence and pattern of immunoglobulin (Ig) and complement (C) deposits. An intensity of <1+ was defined as negative. Overall, negative staining (IgG, IgA, IgM, C3, and C1q) was defined as “immunofluorescence remission (IR).” Electron microscopy was performed to evaluate the presence and distribution of high electron-density deposits. Electron-dense deposit (EDD) scores (0–3) were defined as follows: 0, complete resolution of every deposit; 1, obsolete deposits or deposits in the mesangial area only; 2, mild-to-moderate deposits in the epithelial and/or endothelial areas; 3, severe deposits in the epithelial and/or endothelial areas. An EDD score of ⩽1 was defined as “electron microscopy remission (ER).”

Clinical data collection and clinical outcomes

Clinical data at baseline (start of remission induction therapy), at 6 and 12 months after starting induction therapy, the time of Biopsy 2, and at 5 years were collected. The following remission criteria were used: “complete renal response (CRR),” proteinuria ⩽0.5 g/day and estimated glomerular filtration rate ⩾60 or a ⩽20% decrease from baseline²⁶; “complete remission (CR),” proteinuria ⩽0.4 g/day, normal urinary sediment and serum albumin (Alb) level ⩾3.5 g/dl, and normal serum Cr level²⁷; “proteinuria ⩽0.2 g/day”; “SLEDAI remission (SLEDAI-R),” SLEDAI-2K = 0; “clinical SLEDAI remission (cSLEDAI-R),” clinical SLEDAI-2K = 0¹²; “DORIS remission (DORIS-R),” clinical SLEDAI-2K = 0, Evaluator’s Global Assessment <0.5 (0–3), and PSL ⩽5 mg/day with a stable dose of antimalarials, immunosuppressants, or biologic agents¹²; our defined “A,” SLEDAI-2K = 0 and PSL ⩽5 mg/day and our defined “B,” proteinuria ⩽0.2 g/day, normal serum Cr level, and PSL ⩽5 mg/day. The urinary protein/Cr ratio was used as a substitute for 24-h urinary protein excretion when 24-h urine excretion could not be measured. Normal serum Cr levels were defined as ⩽1 mg/dl in men and ⩽0.8 mg/dl in women. Complement and anti-ds-DNA antibodies were determined according to the institutional criterion. Owing to the retrospective study design, the Evaluator’s Global Assessment data were unavailable and thus omitted from the evaluation. Data at 5 years for one case lost to follow-up was substituted with values obtained at the last visit. There were no missing data concerning the above parameters.

Treatment protocol

Between 2009 and 2015, before the approval of mycophenolate mofetil (MMF) for patients with LN in Japan in 2016, the first-line combination strategy of MZB, TAC, and PSL was our institution’s standard therapy.^23,28,29 MZB was developed in Japan and approved for LN in this country in 1990. Its mode of action is similar to that of MMF. The treatment protocol was described in detail in a previous publication.²³ Briefly, patients initially received intravenous methylprednisolone pulse therapy (0.5 g/day for 3 days) followed by oral PSL. The daily dose of PSL was started at 60 mg/day (80 mg/day for patients weighing >60 kg) and then reduced by 10 mg/day every week to reach 30 mg/day, which was followed by further tapering by 5 mg/day at 2-week intervals until the dose reached 20 mg/day. Further tapering to 5 mg/day was allowed if the patient’s condition was stable. The target doses of PSL at months 2, 6, and 12 were 20, 10, and 5 mg/day, respectively.

The initial MZB dose was 300 mg/day once daily for 3 days per week.^28,29 Peak blood concentration (C3) was measured once or twice during the hospital stay to assess the therapeutic range.²⁹ The initial TAC dose was 3 mg/day once daily. Blood trough concentration was measured every week during the hospital stay and at every outpatient visit, and the dose was reduced if the blood trough concentration was >10 ng/ml or the serum Cr level was dangerously elevated.²⁹ The maximum doses of MZB and TAC were 300 and 3 mg/day, respectively.

When a treated patient achieved SLEDAI remission (=0) and/or the PSL dose could be tapered to 5 mg/day, either MZB or TAC was discontinued, while the remaining drug was continued as maintenance therapy. In case of treatment failure or relapse, second-line therapy (intravenous cyclophosphamide [IVCY] or MMF plus TAC) was initiated.

After 2 years, Biopsy 2 was recommended for all patients, whether in clinical remission or not, to confirm the histological response. If high disease activity, which could not be overlooked, remained (a high AI score), the patient was transitioned to second-line therapy. If immunological activity (non-IR or non-ER) alone remained, we refrained from further reduction of the PSL dose and/or immunosuppressant.

Statistical analysis

Data are presented as median (interquartile range (IQR)) or number of patients (percentages) unless otherwise specified. Continuous variables were compared using the Student’s t-test or the Mann–Whitney U test depending on data distribution. Categorical variables were compared using Fisher’s direct probability test. The area under the receiver-operating characteristic curve (AUC) was used to evaluate the predictive performance of potential short-term goals for predicting long-term goals. A p-value <0.05 indicated a statistically significant difference. All statistical analyses were performed using EZR ver. 1.60 (Saitama Medical Center, Jichi Medical University, Saitama, Japan).

Results

Patient background at baseline and at the time of Biopsy 2

Patient background characteristics at baseline and at the time of Biopsy 2 are shown in Table 1. The ISN/RPS classifications at the time of Biopsy 1 were as follows: Class III, 3; IV, 8; V, 2; and III + V and IV + V, 10. Biopsy 2 was performed at a median of 27 (quartile range: 25–34) months after starting treatment. At the time of Biopsy 2, proteinuria ⩽0.2 g/day was present in 20 patients (87%), the median serum Cr level was 0.70 (0.66–0.86) mg/dl, and the median PSL dose was 5 (5–7.5) mg/day. At the time of Biopsy 2, SLEDAI remission was achieved in 14 patients (61%), “A” in 12 (52%), and “B” in 12 (52%). Histologically, 17 patients (74%) had “LR,” 8 (35%) had “IR,” and 13 (57%) had “ER.”

Table 1.

Patient background characteristics at baseline and at the time of Biopsy 2 and outcomes at 5 years.

Characteristic	Baseline, n = 23	Biopsy 2, n = 23	5 years, n = 23
Age, years	45 (32–60)	–	–
Female sex, no. (%)	20 (87%)	–	–
Duration of treatment until Biopsy 2, months	–	27 (25–34)	–
Cr, mg/dl	0.69 (0.57–0.90)	0.70 (0.66–0.86)	0.68 (0.64–0.81)
eGFR, ml/min/1.73 m²	80 (62–90)	74 (57–81)	72 (61–79)
Proteinuria, g/day	2.7 (1.4–5.3)	0.05 (0.02–0.14)	0.06 (0.03–0.09)
PSL, mg/day	60 (60–60)	5 (5–7.5)	5 (5–7)
Immunosuppressants, no., MZB or TAC/MZB, and TAC/second-line therapy	0/23/0	10/13/0	15/6/2
Hydroxychloroquine, no. (%)	0 (0)	0 (0)	3 (13)
Antihypertensive therapy, no. (%)	6 (26)	15 (65)	11 (48)
CR, no. (%)	–	14 (61)	16 (70)
CRR, no. (%)	–	20 (87)	19 (83)
Proteinuria ⩽0.2 g/day	–	20 (87)	20 (87)
SLEDAI remission, no. (%)	–	14 (61)	16 (70)
DORIS remission, no. (%)	–	13 (57)	12 (52)
“A,” SLEDAI-2K = 0 and PSL ⩽5 mg/day	–	12 (52)	12 (52)
“B,” proteinuria ⩽0.2 g/day and normal serum Cr levels and PSL ⩽5 mg/day	–	12 (52)	11 (48)
AI	2 (1–6)	1 (1–2)	–
CI	3 (2–4)	4 (3–5)	–
LR	–	17 (74)	–
IR	–	8 (35)	–
ER	–	13 (57)	–

Data are shown as the median (IQR) or the number of patients (percentage).

AI, activity index; CI, chronicity index; Cr, creatinine; CR, complete remission; CRR, complete renal response; DORIS, definitions of remission in SLE; eGFR, estimated glomerular filtration rate; ER, electron microscopy remission; IQR, interquartile range; IR, immunofluorescence remission; LR, light microscopy remission; MZB, mizoribine; PSL, prednisolone; SLEDAI, systemic lupus erythematosus disease activity index; TAC, tacrolimus.

Potential indicators of the ideal long-term goals

Clinical and histological indices up to the time of Biopsy 2 that predict the ideal long-term goals (“A” and “B” at 5 years, A-5y and B-5y) are shown in Table 2. The indices that were correlated with the achievement of both A-5y and B-5y simultaneously are none at baseline; SLEDAI-2K score at 6 months; C3 levels, SLEDAI-2K score, and PSL dose at 12 months; C3 levels, SLEDAI-2K score, SLEDAI-R, DORIS-R, A-2y, B-2y, PSL dose, ER, and EDD score at the time of Biopsy 2.

Table 2.

Potential indicators that predict the ideal long-term goals (“A” and “B” at 5 years).

Characteristic	A-5y			B-5y
Characteristic	No, n = 11	Yes, n = 12	p-Value	No, n = 12	Yes, n = 11	p-Value
Baseline
Age, years	45 (39–54)	45 (28–64)	0.878	46 (41–59)	43 (28–59)	0.735
Female sex, no. (%)	9 (82)	11 (92)	0.590	9 (75)	11 (100)	0.217
Cr, mg/dl	0.70 (0.56–1.15)	0.69 (0.60–0.71)	0.423	0.86 (0.57–1.15)	0.68 (0.58–0.70)	0.218
eGFR (ml/min/1.73 m²)	73 (44–90)	82 (67–89)	0.424	72 (45–89)	83 (74–90)	0.218
Proteinuria (g/day)	2.1 (1.6–5.2)	2.9 (1.5–5.1)	0.951	2.9 (1.8–5.1)	2.7 (1.3–4.9)	0.712
C3 (mg/dl)	36 (30–52)	59 (34–82)	0.295	39 (31–57)	62 (33–82)	0.355
Anti-dsDNA titer (IU/ml)	123 (28–310)	91 (47–205)	1.000	129 (29–284)	78 (40–194)	0.805
SLEDAI-2K scores	21 (18–25)	22 (17–27)	0.643	21 (19–24)	23 (16–28)	0.643
Biopsy class, no. (Class III/IV/V/III + V and IV + V)	1/5/2/3	2/3/0/7	0.443	2/5/2/3	1/3/0/7	0.443
AI	2 (1–6)	2 (1–5)	0.975	2 (1–5)	2 (1–6)	1.000
CI	3 (2–6)	2.5 (2–3)	0.482	3 (2–5)	3 (2–4)	0.701
At 6 months
Cr (mg/dl)	0.80 (0.62–0.99)	0.73 (0.62–0.78)	0.229	0.86 (0.62–1.05)	0.73 (0.60–0.76)	0.069
eGFR (ml/min/1.73 m²)	66 (50–79)	76 (59–85)	0.325	64 (50–78)	78 (60–85)	0.176
Proteinuria (g/day)	0.19 (0.06–0.90)	0.10 (0.03–0.35)	0.195	0.18 (0.06–0.87)	0.09 (0.03–0.48)	0.218
Normal urine sediment, no. (%)	8 (73)	11 (92)	0.317	9 (75)	10 (91)	0.590
CRR, no. (%)	6 (55)	8 (67)	0.680	7 (58)	7 (64)	1.000
CR, no. (%)	3 (27)	7 (58)	0.214	3 (25)	7 ( 64)	0.100
C3 (mg/dl)	78 (67.5–89)	94 (82.3–102)	0.157	81 (70–93)	95 (82–103)	0.207
Anti-dsDNA titer (IU/ml)	3.9 (3.3–24)	3.4 (1.8–6.4)	0.166	4.4 (3.4–23)	3.3 (1.6–5.1)	0.079
SLEDAI-2K scores	4 (2–7)	0 (0–2.5)	0.013	4 (2–6.5)	0 (0–3)	0.046
cSLEDAI-R, no. (%)	3 (27)	9 (75)	0.039	4 (33)	8 (73)	0.100
SLEDAI-R, no. (%)	1 (9)	7 (58)	0.027	2 (17)	6 (55)	0.089
PSL dose (mg/day)	13 (11–14)	11 (10–12)	0.114	13 (10–14)	11 (10–12)	0.269
At 12 months
Cr (mg/dl)	0.77 (0.70–1.04)	0.71 (0.64–0.77)	0.196	0.78 (0.70–1.04)	0.68 (0.64–0.75)	0.056
eGFR (ml/min/1.73 m²)	65 (50–75)	74 (63–78)	0.325	64 (47–75)	75 (68–80)	0.140
Proteinuria (g/day)	0.12 (0.04–0.24)	0.05 (0.03–0.06)	0.073	0.12 (0.03–0.21)	0.05 (0.03–0.06)	0.194
Normal urine sediment, no. (%)	10 (91)	10 (83)	1.000	10 (83)	10 (91)	1.000
CRR, no. (%)	9 (82)	12 (100)	0.217	10 (83)	11 (100)	0.478
CR, no. (%)	6 (55)	9 (75)	0.400	6 (50)	9 (82)	0.193
C3 (mg/dl)	81 (65–89)	95 (89–101)	0.006	82 (66–89)	96 (89–104)	0.013
Anti-dsDNA titer (IU/ml)	7.0 (2.2–16)	3.3 (2.0–6.3)	0.139	6.4 (2.3–15)	3.0 (1.6–5.8)	0.109
SLEDAI-2K scores	2 (1–5)	0 (0–0.5)	0.030	2 (1.5–4.5)	0 (0–0)	0.008
cSLEDAI-R, no. (%)	6 (55)	10 (83)	0.193	6 (50)	10 (91)	0.069
SLEDAI-R, no. (%)	3 (27)	9 (75)	0.039	3 (25)	9 (82)	0.012
DORIS-R, no. (%)	0 (0)	0 (0)	NA	0 (0)	0 (0)	NA
A-12 m, no. (%)	0 (0)	0 (0)	NA	0 (0)	0 (0)	NA
B-12 m, no. (%)	0 (0)	0 (0)	NA	0 (0)	0 (0)	NA
PSL dose, mg/day	10 (8–11)	7 (6–8)	0.002	10 (8–11)	7 (6–8)	0.012
At Biopsy 2 (2 years)
Cr (mg/dl)	0.73 (0.68–1.12)	0.69 (0.56–0.74)	0.103	0.79 (0.68–1.11)	0.68 (0.56–0.72)	0.039
eGFR (ml/min/1.73 m²)	66 (46–76)	79 (65–84)	0.065	66 (48–75)	80 (73–86)	0.027
Proteinuria (g/day)	0.05 (0.03–0.31)	0.04 (0.02–0.08)	0.477	0.05 (0.03–0.24)	0.04 (0.01–0.09)	0.421
Normal urine sediment, no. (%)	8 (73)	11 (92)	0.317	4 (33)	5 (46)	0.680
CRR, no. (%)	9 (82)	11 (92)	0.590	10 (83)	10 (91)	1.000
CR, no. (%)	5 (46)	9 (75)	0.214	5 (42)	9 (82)	0.089
C3 (mg/dl)	79 (72–85)	94 (83–108)	0.010	80 (72–88)	96 (83–108)	0.018
Anti-dsDNA titer (IU/ml)	3.3 (1.6–8.9)	2.9 (1.6–4.3)	0.355	3.3 (1.8–8.2)	2.9 (1.5–4.3)	0.309
SLEDAI-2K scores	2 (1–4)	0 (0–0)	0.004	2 (0–4)	0 (0–0)	0.009
cSLEDAI-R, no. (%)	5 (46)	11 (92)	0.027	6 (50)	10 (91)	0.069
SLEDAI-R, no. (%)	3 (27)	11 (92)	0.003	4 (33)	10 (91)	0.009
DORIS-R, no. (%)	2 (18)	11 (92)	0.001	3 (25)	10 (91)	0.003
A-2y, no. (%)	1 (9)	11 (92)	<0.001	2 (17)	10 (91)	0.001
B-2y, no. (%)	2 (18)	10 (83)	0.003	2 (17)	10 (91)	0.001
PSL dose (mg/day)	8 (7–9)	5 (5–5)	<0.001	8 (7–9)	5 (5–5)	0.001
AI	1 (1–3)	1 (1–1)	0.162	1 (1–2)	1 (1–1)	0.227
CI	4 (3–6) \|	4 (3–4)	0.594	4 (3–6)	4 (3–5)	0.754
LR, no. (%)	7 (64)	10 (83)	0.371	8 (67)	9 (82)	0.640
IR, no. (%)	2 (18)	6 (50)	0.193	3 (25)	5 (46)	0.400
ER, no. (%)	3 (27)	10 (83)	0.012	4 (33)	9 (82)	0.036
EDD score	2 (2–3)	0 (0–1)	0.003	2 (1–3)	0 (0–1)	0.004

Data are shown as the median (IQR) or the number of patients (percentage). The indicators that can predict both the ideal 5-year goals, A-5y and B-5y, are highlighted in bold. A-5y, “A” goal at 5 years defined as SLEDAI-2K = 0 and PSL ⩽5 mg/day; B-5y, “B” goal at 5 years defined as proteinuria ⩽0.2 g/day and normal serum Cr levels and PSL ⩽5 mg/day; A-2y, “A” goal at 2 years; B-2y, “B” goal at 2 years; A-12m, “A” goal at 12 months; B-12m, “B” goal at 12 months.

AI, activity index; CI, chronicity index; Cr, creatinine; CR, complete remission; CRR, complete renal response; cSLEDAI-R, Clinical Systemic Lupus Erythematosus Disease Activity Index Remission; DORIS, definitions of remission in SLE; EDD, electron-dense deposit; eGFR, estimated glomerular filtration rate; ER, electron microscopy remission; IQR, interquartile range; IR, immunofluorescence remission; LR, light microscopy remission; PSL, prednisolone; SLEDAI-R, Systemic Lupus Erythematosus Disease Activity Index Remission.

Time course of the predictive performance of candidate short-term goals toward the ideal long-term goals

The time course of the predictive performance of the candidate’s short-term goals toward the ideal long-term goals (A-5y and B-5y) by comparing the AUC is shown in Figure 1. The short-term goals with an accuracy (AUC: 95% confidence interval) of ⩾0.8 to predict the long-term goals were as follows: “A-5y,” A-2y (0.91: 0.79–1.00), DORIS-R-2y (0.87: 0.72–1.00), EDD score (0.85: 0.70–1.00), B-2y (0.83: 0.66–0.99), and SLEDAI-R-2y (0.82: 0.66–0.98) as well as “B-5y,” A-2y (0.87: 0.73–1.00), B-2y (0.87: 0.73–1.00), EDD score (0.85: 0.69–1.00), and DORIS-R-2y (0.83: 0.67–0.99). The best predictive short-term goal was A-2y.

Figure 1.

Time course of the predictive performance of candidate short-term goals toward ideal long-term goals (“A” and “B”).

The short-term goals, such as proteinuria, CR, and CRR, consisting solely of renal indices, had low predictive performance for long-term goals.

Usefulness of pathological targets on Biopsy 2 for predicting long-term outcomes

The interrelationships among pathological targets on Biopsy 2 (AI ⩽1, IR, EDD score ⩽1) and between short- and long-term goals (A-5y and B-5y) are shown in Figure 2. The electron microscopy evaluation had a higher predictive ability (AUC of 0.78 and LR+ of 3.06 for A-5y; AUC of 0.74 and LR+ of 2.46 for B-5y) than light or immunofluorescence microscopy. EDD scores and grading of immune deposits based on their intensity, amount, and location were useful in predicting A-5y, B-5y, and PSL doses at 5 years in proportion to the score (Figure 3 and Table 3).

Figure 2.

The usefulness of Biopsy 2 pathological targets for predicting long-term outcomes.

Figure 3.

The association between EDD score and long-term outcomes.

Table 3.

Patient background characteristics and outcomes at 5 years based on the EDD scores at the time of Biopsy 2.

EDD score	0	1	2	3	p-Value
EDD score	n = 8	n = 5	n = 6	n = 4	p-Value
At 5 years
A-5y, no. (%)	7 (88)	3 (60)	2 (33)	0 (0)	0.025
B-5y, no. (%)	7 (88)	2 (40)	2 (33)	0 (0)	0.025
Cr, mg/dl	0.65 (0.62–0.70)	0.72 (0.68–0.88)	0.64 (0.61–0.67)	1.61 (1.09–2.78)	0.021
eGFR (ml/min/1.73 m²)	78 (72–80)	65 (53–71)	77 (71–89)	45 (21–64)	0.029
Proteinuria (g/day)	0.04 (0.03–0.08)	0.03 (0–0.09)	0.08 (0.06–0.12)	0.46 (0.03–1.60)	0.369
PSL dose (mg/day)	5 (5–5)	5 (5–5)	6 (5–7)	10 (7–13)	0.008
At biopsy 2 (2 years)
A-2y, no. (%)	7 (88)	3 (60)	2 (33)	0 (0)	0.025
B-2y, no. (%)	6 (75)	3 (60)	3 (50)	0 (0)	0.104
Cr (mg/dl)	0.70 (0.65–0.72)	0.73 (0.66–0.75)	0.65 (0.57–0.68)	1.32 (1.05–1.53)	0.032
eGFR (ml/min/1.73 m²)	79 (70–81)	66 (59–80)	84 (69–91)	43 (34–57)	0.108
Proteinuria (g/day)	0.03 (0.01–0.06)	0.05 (0.03–0.05)	0.06 (0.04–0.15)	0.62 (0.34–1.08)	0.178
PSL dose (mg/day)	5 (5–5)	5 (5–5)	6 (5–7)	11 (10–12)	0.004
Baseline
Age, years	43 (28–52)	46 (36–63)	47 (28–53)	54 (38–66)	0.871
Female sex, no. (%)	8 (100)	4 (80)	6 (100)	2 (50)	0.068
SLEDAI-2K scores	21 (12–24)	21 (18–27)	21.5 (16–24)	23.5 (21–26)	0.869
Cr (mg/dl)	0.69 (0.58–0.70)	0.70 (0.61–0.77)	0.59 (0.46–0.68)	1.35 (1.12–1.73)	0.023
eGFR (ml/min/1.73 m²)	82 (77–86)	71 (58–88)	94 (77–108)	37 (26–53)	0.065
Proteinuria (g/day)	2.3 (0.60–3.9)	1.7 (0.84–3.6)	4.2 (2.5–5.5)	4.0 (2.1–7.2)	0.327

Data are shown as the median (IQR) or the number of patients (percentage).

A-5y, “A” goal at 5 years defined as SLEDAI-2K = 0 and PSL ⩽5 mg/day; B-5y, “B” goal at 5 years defined as proteinuria ⩽0.2 g/day, normal serum Cr levels, and PSL ⩽5 mg/day; A-2y, “A” goal at 2 years; B-2y, “B” goal at 2 years.

EDD score, electron-dense deposit score; Cr, creatinine; eGFR, estimated glomerular filtration rate; IQR, interquartile range; PSL, prednisolone; SLEDAI-2K, Systemic Lupus Erythematosus Disease Activity Index 2000.

Interrelationships among candidate short-term goals and between short- and long-term goals

The interrelationships among candidate short-term goals (A-2y, B-2y, DORIS-R-2y, and ER) and between short- and long-term goals (A-5y and B-5y) are shown in Figure 4. The clinical goals of A-2y, B-2y, and DORIS-R-2y were highly consistent. The histologically defined ER was also well-matched with the abovementioned three clinical goals. These measures efficiently predicted A-5y and B-5y. Despite being defined from different perspectives, the patients meeting A-5y and B-5y were nearly identical.

Figure 4.

Interrelationship among the candidate’s short-term goals and between the short- and long-term goals.

Discussion

This study demonstrated the usefulness of electron microscopic evaluation of Biopsy 2 at 2 years. The ER and EDD scores predicted the long-term goals at 5 years. These results are reasonable and consistent with the pathogenesis of LN. The five findings that characterize LN are “full-house” staining by immunofluorescence, intense C1q staining, extraglomerular deposits, combined subendothelial and subepithelial deposits, and endothelial tubuloreticular inclusions.³⁰ Even though the abovementioned findings can be obtained by immunofluorescence and electron microscopy, conventional histological evaluations were based on light microscopic findings alone, defining “AI = 0 as histological remission.”

In addition, we first demonstrated a concordance between clinical and histological remissions as defined clinically by “A,” “B,” and “DORIS remission,” and histologically by “ER,” suggesting optimal short-term goals for LN.

Conventional renal indices, such as proteinuria levels, CR, and CRR, could not predict the ideal long-term goals. This finding is consistent with that of studies reporting that proteinuria and serum Cr levels do not correlate quantitatively with AI and do not accurately reflect the response to induction therapy.³¹ If renal indices are used, the PSL dose should be added as observed in “B,” with proteinuria ⩽0.2 g/day, normal Cr level, and PSL ⩽5 mg/day. As the PSL dose is carefully reduced while monitoring the treatment response, achieving “PSL ⩽5 mg/day” represents a favorable clinical course.

Setting ideal long-term goals first was beneficial in identifying the short-term goals. The short-term goals to be achieved depend on the long-term goals set. For example, a short-term goal of “proteinuria ⩽0.8 or 0.7 g/day at 12 months” was identified for the long-term goal of “serum Cr ⩽1 mg/dl at 7 years;” however, its prediction accuracy was not very good.^8,9 We defined two long-term goals at 5 years: “A,” which would be considered ideal for rheumatologists, and “B,” for nephrologists. Even though both criteria were defined from different perspectives, as shown in Figure 2, the patient groups who met each goal were concordant at 2 and 5 years. This finding indicates that both “A” and “B” are ideal short- and long-term goals. Additionally, in our previous report,²³ cases that achieved “A” maintained remission without relapse for up to 5 years, suggesting that “A” is an excellent short-term goal.

SLEDAI-R-2y was more predictive than cSLEDAI-R-2y, and A-2y was more predictive than DORIS-R-2y, suggesting target superiority, including serologic indices. A-2y was a better predictor than SLEDAI-R-2y, and DORIS-R-2y was a better predictor than cSLEDAI-R-2y, suggesting target superiority, including the PSL dose. However, considering that the difference in predictive performance between “A” and “DORIS-R” was small in this study, the merits and demerits of aiming for the normalization of serological indices should be carefully judged in each case. Further prospective comparisons of the long-term outcomes are needed to determine the optimal goal.

The present study has some limitations. First, the dosing of PSL is susceptible to variations in prescribing behavior and practices among different clinicians. This variability could potentially influence the attainment of outcomes “A” and “B.” Second, future prospective studies are warranted to determine how maintenance therapy can be tapered in patients meeting ER at 2 years and how treatment should be modified for patients presenting with residual activity of non-ER. Third, our results need to be validated with various induction regimens. Each treatment regimen may have a different short-term goal and timing. Regarding maintenance therapy, no difference was observed in histological findings between the MMF and AZA groups post-induction with IVCY and the subsequent 2 years of maintenance therapy with these drugs.³² Since the publication of the Japanese Guideline for the Management of Systemic Lupus Erythematosus, our department has switched to MMF as a first-line treatment; thus, we plan to compare the results of the MMF groups with our present findings. Finally, this is a single-center, retrospective study with a small number of cases and without sample size calculation. This may limit the generalizability of the findings and increase the risk of type II errors, potentially leading to underpowered results that might miss detecting significant differences or associations. Nevertheless, the fact that the results were as expected with statistically significant differences suggests that the present findings are truly clinically important.

Conclusion

The best predictive short-term goal was A-2y, that is, SLEDAI-R and PSL ⩽5 mg/day at 2 years. Thus, we recommend adjusting medications with the aim of SLEDAI-R in the early stage, subsequently aiming for “A” by 2 years. B-2y, that is, proteinuria ⩽0.2 g/day, normal Cr level, and PSL ⩽5 mg/day, DORIS-R-2y, and ER can be alternative short-term goals. The concordance between clinical and histological remissions at 2 years was demonstrated for the first time with these newly defined remission criteria. If only the renal indices are targeted, the ideal long-term prognosis will not be achieved. The comprehensive criteria, including the PSL dose, are useful but somewhat subjective. Achieving ER on repeat kidney biopsy indicates that the patient is in good histological remission. Future prospective studies are needed to determine how the treatment should be modified for patients presenting residual activity on electron microscopy.

Footnotes

Acknowledgements

The authors would like to thank Dr. Hirofumi Makino and Dr. Ken-ei Sada for their helpful advice. We are also grateful to our team members for their collaboration.

Declarations

ORCID iD

Hidetoshi Kagawa

References

Hanly

O’Keeffe

, et al. The frequency and outcome of lupus nephritis: results from an international inception cohort study. Rheumatology (Oxford) 2016; 55: 252–262.

Reppe Moe

Molberg

Strøm

, et al. Assessing the relative impact of lupus nephritis on mortality in a population-based systemic lupus erythematosus cohort. Lupus 2019; 28: 818–825.

Duarte-García

Hocaoglu

Valenzuela-Almada

, et al. Rising incidence and prevalence of systemic lupus erythematosus: a population-based study over four decades. Ann Rheum Dis 2022; 81(9): 1260–1266.

Soliman

Mohan

Lupus nephritis biomarkers. Clin Immunol 2017; 185: 10–20.

Moroni

Porata

Raffiotta

, et al. Beyond ISN/RPS lupus nephritis classification: adding chronicity index to clinical variables predicts kidney survival. Kidney360 2021; 3: 122–132.

Nakagawa

Toyama

Iwata

, et al. The relationship between the modified National Institute of Health activity and chronicity scoring system, and the long-term prognosis for lupus nephritis: a retrospective single-center study. Lupus 2021; 30: 1739–1746.

Houssiau

Vasconcelos

D’Cruz

, et al. Early response to immunosuppressive therapy predicts good renal outcome in lupus nephritis: lessons from long-term followup of patients in the Euro-Lupus nephritis trial. Arthritis Rheum 2004; 50: 3934–3940.

Tamirou

Lauwerys

Dall’Era

, et al. A proteinuria cut-off level of 0.7 g/day after 12 months of treatment best predicts long-term renal outcome in lupus nephritis: data from the MAINTAIN nephritis trial. Lupus Sci Med 2015; 2: e000123.

Dall’Era

Cisternas

Smilek

, et al. Predictors of long-term renal outcome in lupus nephritis trials: lessons learned from the Euro-Lupus Nephritis cohort. Arthritis Rheumatol 2015; 67: 1305–1313.

10.

Mackay

Dall’Era

Fishbein

, et al. Establishing surrogate kidney end points for lupus nephritis clinical trials: development and validation of a novel approach to predict future kidney outcomes. Arthritis Rheumatol 2019; 71: 411–419.

11.

Bose

Silverman

Bargman

JM.

Ten common mistakes in the management of lupus nephritis. Am J Kidney Dis 2014; 63: 667–676.

12.

van Vollenhoven

Bertsias

Doria

, et al. 2021 DORIS definition of remission in SLE: final recommendations from an international task force. Lupus Sci Med 2021; 8: e000538.

13.

Weeding

Fava

Magder

, et al. One-third of patients with lupus nephritis classified as complete responders continue to accrue progressive renal damage despite resolution of proteinuria. Lupus Sci Med 2022; 9: e000684.

14.

Parodis

Tamirou

Houssiau

FA.

Prediction of prognosis and renal outcome in lupus nephritis. Lupus Sci Med 2020; 7: e000389.

15.

Tamirou

Houssiau

FA.

Management of lupus nephritis. J Clin Med 2021; 10: 670.

16.

Alsuwaida

Husain

Alghonaim

, et al. Strategy for second kidney biopsy in patients with lupus nephritis. Nephrol Dial Transplant 2012; 27: 1472–1478.

17.

Malvar

Pirruccio

Alberton

, et al. Histologic versus clinical remission in proliferative lupus nephritis. Nephrol Dial Transplant 2017; 32: 1338–1344.

18.

Parodis

Adamichou

Aydin

, et al. Per-protocol repeat kidney biopsy portends relapse and long-term outcome in incident cases of proliferative lupus nephritis. Rheumatology (Oxford) 2020; 59: 3424–3434.

19.

De Rosa

Azzato

Toblli

, et al. A prospective observational cohort study highlights kidney biopsy findings of lupus nephritis patients in remission who flare following withdrawal of maintenance therapy. Kidney Int 2018; 94: 788–794.

20.

Moroni

Porata

Raffiotta

, et al. Predictors of increase in chronicity index and of kidney function impairment at repeat biopsy in lupus nephritis. Lupus Sci Med 2022; 9: e000721.

21.

Malvar

Alberton

Lococo

, et al. Kidney biopsy-based management of maintenance immunosuppression is safe and may ameliorate flare rate in lupus nephritis. Kidney Int 2020; 97: 156–162.

22.

Das

Patel

Guditi

, et al. Correlation between the clinical remission and histological remission in repeat biopsy findings of quiescent proliferative lupus nephritis. Lupus 2021; 30: 876–883.

23.

Kagawa

Yamanaka

Hiromasa

First-line combination strategy provides favorable 5-year outcomes for patients with lupus nephritis: a single-center observational study. Acta Med Okayama 2022; 76: 547–555.

24.

Weening

D’Agati

Schwartz

, et al. The classification of glomerulonephritis in systemic lupus erythematosus revisited. J Am Soc Nephrol 2004; 15: 241–250.

25.

Bajema

Wilhelmus

Alpers

, et al. Revision of the International Society of Nephrology/Renal Pathology Society classification for lupus nephritis: clarification of definitions, and modified National Institutes of Health activity and chronicity indices. Kidney Int 2018; 93: 789–796.

26.

Rovin

Teng

YKO

Ginzler

, et al. Efficacy and safety of voclosporin versus placebo for lupus nephritis (AURORA 1): a double-blind, randomised, multicentre, placebo-controlled, phase 3 trial. Lancet 2021; 397: 2070–2080.

27.

Liu

Zhang

Liu

, et al. Multitarget therapy for induction treatment of lupus nephritis: a randomized trial. Ann Intern Med 2015; 162: 18–26.

28.

Kagawa

Hiromasa

Hara

, et al. Mizoribine, tacrolimus, and corticosteroid combination therapy successfully induces remission in patients with lupus nephritis. Clin Exp Nephrol 2012; 16: 760–766.

29.

Kagawa

Hiromasa

Yamanaka

, et al. The first year results of mizoribine/tacrolimus-based multitarget treatment for consecutive patients with lupus nephritis. Clin Exp Nephrol 2018; 22: 1371–1378.

30.

Kudose

Santoriello

Bomback

, et al. Sensitivity and specificity of pathologic findings to diagnose lupus nephritis. Clin J Am Soc Nephrol 2019; 14: 1605–1615.

31.

Katsuyama

Miyawaki

Sada

, et al. Association of explanatory histological findings and urinary protein and serum creatinine levels at renal biopsy in lupus nephritis: a cross-sectional study. BMC Nephrol 2020; 21: 208.

32.

Stoenoiu

Aydin

Tektonidou

, et al. Repeat kidney biopsies fail to detect differences between azathioprine and mycophenolate mofetil maintenance therapy for lupus nephritis: data from the MAINTAIN nephritis trial. Nephrol Dial Transplant 2012; 27: 1924–1930.