Forgotten compartment: Impact of tubulointerstitial inflammation and damage on renal outcomes in lupus nephritis from Saudi Arabia

Abstract

BACKGROUND

Lupus nephritis (LN) is a significant cause of kidney morbidity in patients with systemic lupus erythematosus. While glomerular lesions are the primary focus of current classifications, tubulointerstitial involvement may significantly influence renal outcomes. There is limited data on tubulointerstitial inflammation (TII) and tubulointerstitial damage (TID) in Middle Eastern populations.

AIM

To examine the clinical, pathological, and prognostic significance of coexisting TII and TID in patients with biopsy-proven LN from Saudi Arabia.

METHODS

We retrospectively analyzed 100 patients with biopsy-confirmed LN. Patients were stratified into those with TII + TID (n = 48) and those without (n = 52). Baseline demographics, clinical features, laboratory and immunological data, and histopathological findings, including modified National Institutes of Health (NIH) activity and chronicity scores, were collected. Multivariable logistic regression identified predictors of TII + TID. Renal response during follow-up was evaluated using Kaplan-Meier analysis, and predictors of adverse composite outcomes were assessed using Cox regression.

RESULTS

Patients with TII + TID had lower baseline estimated glomerular filtration rate (77.7 ± 37.4 mL/minute/1.73 m² vs 98.2 ± 52.7 mL/minute/1.73 m²; P = 0.028) and higher low-density lipoprotein cholesterol (LDL-C; 3.01 ± 1.88 mmol/L vs 1.78 ± 1.82 mmol/L; P = 0.006). They also had higher modified NIH activity (6.17 ± 3.90 vs 3.13 ± 3.18; P < 0.001) and chronicity scores (4.21 ± 1.74 vs 1.85 ± 2.44; P < 0.001). Proliferative classes, especially class III (35.4% vs 15.4%) and class IV + V (20.8% vs 15.4%; P = 0.024), were more common in this group. Independent predictors of TII + TID included class III (OR = 150.42; P = 0.006), class IV + V (OR = 44.11; P = 0.03), LDL-C (OR = 2.26; P = 0.004), fibrinoid necrosis [Exp(B) = 2.29; P = 0.041], and interstitial inflammation [Exp(B) = 73.29; P < 0.001]. Patients with TII + TID had slower and reduced rates of achieving renal remission. Higher baseline serum creatinine, older age, elevated C-reactive protein (CRP), and the presence of hyaline deposits were associated with worse composite renal outcomes. In contrast, total glomerulosclerosis showed an inverse association in the subgroup but not in the overall cohort.

CONCLUSION

TII and damage are common in LN and are closely associated with proliferative glomerular lesions, fibrinoid necrosis, and adverse renal outcomes. The presence of TII + TID, along with baseline creatinine, CRP, and hyaline deposits, identifies patients at higher risk for poor renal prognosis. These findings highlight the importance of evaluating the tubulointerstitial compartment for risk stratification and tailored management in LN, particularly in the Saudi Arabian population.

Key Words: Lupus nephritis; Tubulointerstitial inflammation; Tubulointerstitial damage; Renal biopsy; Renal outcomes; Prognosis; Saudi Arabia

Core Tip: Although current lupus nephritis (LN) classifications focus mainly on glomerular lesions, tubulointerstitial involvement is frequent and clinically relevant. In this Saudi cohort, the presence of combined tubulointerstitial inflammation and damage was associated with proliferative disease, higher activity and chronicity scores, and poorer renal outcomes. These patients achieved remission more slowly and less often during follow-up. Baseline kidney function, systemic inflammation, and selected histologic features further influenced prognosis. Detailed assessment of the tubulointerstitial compartment may therefore add meaningful prognostic information in LN.

INTRODUCTION

Systemic lupus erythematosus (SLE) is a multisystem autoimmune disorder that affects multiple organs[1]. Lupus nephritis (LN) affects 40%-60% of SLE patients, notably within 5 years of SLE diagnosis[2]. Despite advances in immunosuppressive therapy, a substantial proportion of patients with LN progress to chronic kidney disease or end-stage kidney failure. Long-term follow-up studies of patients with LN have shown that approximately one-third develop chronic kidney disease, nearly 10% progress to end-stage kidney disease, and about 14% die over two decades of follow-up[3]. Current histopathological classification systems for LN, including the International Society of Nephrology/Renal Pathology Society (ISN/RPS) classification, focus primarily on glomerular lesions[4]. Much of the existing literature on prognostication and management in LN is based on assessments of glomerular lesions. Histopathological classification systems and therapeutic decisions have traditionally relied on glomerular findings[4-6]. As a result, the prognostic significance of tubulointerstitial inflammation (TII) and tubulointerstitial damage (TID) is often underrecognized, limiting the ability to predict renal outcomes and tailor therapy accurately.

Tubulointerstitial involvement in LN, including TII and TID such as interstitial fibrosis and tubular atrophy, is common but remains relatively understudied. Studies report that these lesions are present in approximately 66%-69% of kidney biopsies[7,8]. Immune complex deposition in the tubular basement membrane and interstitium recruits inflammatory cells[9]. In contrast, tubular epithelial cells can be activated to secrete pro-inflammatory cytokines and chemokines[10], further amplifying local inflammation. Proteinuria also contributes by stimulating tubular cells to produce inflammatory mediators, promoting tubulointerstitial injury[11]. TID appears to play an essential role in kidney function and long-term outcomes, highlighting the need to pay closer attention to this part of the kidney. Although tubulointerstitial lesions are frequently seen in LN, they are often overlooked in clinical decision-making despite their essential role in kidney function and long-term outcomes. Despite their frequency, the contribution of tubulointerstitial lesions to renal prognosis and treatment response has not been systematically integrated into clinical decision-making. Several studies have demonstrated that tubulointerstitial lesions correlate strongly with reduced renal function, poor response to therapy, and progression to chronic kidney disease, independent of glomerular class[12,13].

Importantly, data on tubulointerstitial involvement in LN from Middle Eastern populations, particularly from Saudi Arabia, are scarce, and similar to global experience, TII and damage remain relatively understudied[14]. Ethnic, genetic, and environmental factors may influence disease severity, histologic patterns, and treatment response[15]. In Saudi Arabia, the contribution of tubulointerstitial pathology to outcomes in this population has not been systematically evaluated. In this study, we aimed to examine the clinical, pathological, and prognostic significance of coexisting TII and damage in patients with biopsy-proven LN from Saudi Arabia. Specifically, we evaluated their association with baseline clinical and laboratory features, histopathological characteristics including modified National Institutes of Health (NIH) activity and chronicity scores, and renal response during follow-up. By focusing on the tubulointerstitial compartment, this study seeks to provide additional prognostic insight beyond traditional glomerular-based classifications and to highlight the importance of comprehensive histologic assessment in LN.

MATERIALS AND METHODS

Study design and setting

This retrospective study was carried out at two tertiary referral centers in Saudi Arabia: King Fahd Armed Forces Hospital in Jeddah and King Fahd Medical City in Riyadh. The Ethics Review Committee of King Fahd Armed Forces Hospital, Jeddah, approved the study. It was conducted in accordance with the principles of the Declaration of Helsinki. Adult patients (≥ 18 years) with biopsy-confirmed LN between 2003 and 2024 were identified from the pathology databases. Patients were included if they had complete baseline clinical and laboratory data and had received standard induction and maintenance therapy. Exclusion criteria were inadequate biopsy specimens, overlapping kidney diseases (e.g., diabetic nephropathy), end-stage renal disease at the time of biopsy, incomplete follow-up, or age younger than 18 years. As this was a retrospective observational study, no formal sample size calculation was performed. The study included all consecutive patients who met the predefined inclusion criteria during the study period. Complete clinical, laboratory, and histopathological data were available for the variables included in the analysis. The primary objective of the study was to examine the clinical, laboratory, and histopathological features associated with coexisting TII and TID in Saudi patients with LN. Secondary objectives were to identify independent clinical and pathological predictors of TII/TID and to assess the effect of TII/TID on renal treatment response and outcomes.

Clinical and laboratory data

Baseline demographic and clinical information was obtained from chart reviews and electronic medical records at the time of kidney biopsy. Continuous variables included age, systolic and diastolic blood pressure, serum creatinine, estimated glomerular filtration rate (eGFR), hemoglobin, white blood cell count, platelet count, serum albumin, lipid profile, electrolytes, uric acid, C-reactive protein (CRP), complement levels (C3 and C4), and autoantibodies, including antinuclear antibody (ANA), anti-double-stranded DNA, anti-Smith, and anti-ribonucleoprotein. Categorical variables comprised sex, presence of hypertension or diabetes, induction therapy (cyclophosphamide or mycophenolate mofetil), maintenance therapy (mycophenolate mofetil, azathioprine, or other immunosuppressants), use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers, sodium-glucose cotransporter-2 inhibitor therapy, statin therapy, and composite renal outcomes.

Renal histopathology

Kidney biopsy specimens were examined using light microscopy, immunofluorescence, and, in selected cases, electron microscopy. For light microscopy, tissue samples were fixed in 10% formalin, embedded in paraffin, and stained with hematoxylin and eosin, Periodic acid-Schiff, Masson’s trichrome, and Jones methenamine silver. Immunofluorescence was performed on frozen tissue to detect IgG, IgA, IgM, C3, C1q, kappa and lambda light chains, and fibrinogen, with staining intensity scored from 0 to 3 + based on location and pattern. Electron microscopy was used selectively to evaluate electron-dense deposits and podocyte changes. Biopsies were classified according to the 2003 ISN/RPS criteria and the 2018 revisions. Histopathological severity was assessed using the modified NIH activity and chronicity indices. The activity (range 0-24) captures active inflammation, including endocapillary hypercellularity, leukocyte infiltration, karyorrhexis or fibrinoid necrosis (weighted × 2), cellular or fibrocellular crescents (weighted × 2), and hyaline deposits. The chronicity (range 0-12) reflects irreversible injury, including glomerular sclerosis, fibrous crescents, tubular atrophy, and interstitial fibrosis. TII was defined as inflammation involving more than 25% of tubules and/or interstitium, and TID was defined as tubular atrophy and/or interstitial fibrosis affecting more than 25% of the biopsy specimen. Consistent with prior studies[12,13,16], patients with coexisting TII and TID were analyzed as a combined group (TII + TID) to capture the overall burden of tubulointerstitial injury. Although these studies assessed the clinical impact of coexisting tubulointerstitial lesions, a standardized quantitative threshold for defining TII and TID has not been established. We therefore selected a ≥ 25% cutoff to represent moderate-to-severe involvement, allowing evaluation of the composite effect of tubulointerstitial lesions on clinical outcomes while acknowledging partial overlap with elements of the modified NIH activity and chronicity indices.

Treatment and renal outcomes

Treatment decisions were made by the attending nephrologist and generally involved glucocorticoids, with or without intravenous methylprednisolone pulses, combined with either high-dose intravenous cyclophosphamide or oral immunosuppressive agents such as mycophenolate mofetil or azathioprine. This was typically followed by maintenance therapy at the clinician's discretion. For patients with class III, IV ± V, or class V LN, induction therapy generally lasted around 6 months. Renal response was evaluated in accordance with the KDIGO 2024 guidelines. A complete response was defined as a urine protein-creatinine ratio < 0.5 g/g, accompanied by stabilization or improvement in kidney function (within ± 10%-15% of baseline) within 6-12 months, without the need for rescue therapy. A partial response was characterized by at least a 50% reduction in proteinuria to less than 3 g/g, with kidney function remaining stable or improving[17]. The composite renal outcome was defined as the occurrence of any of the following: EGFR < 20 mL/minute/1.73 m², ≥ 20% decline in eGFR, doubling of serum creatinine, initiation of dialysis, or kidney transplantation. The eGFR threshold of < 20 mL/minute/1.73 m² was chosen due to its clinical relevance, as this is the level at which, under current practice guidelines and recommendations, patients are typically referred for preemptive kidney transplant evaluation and vascular access planning[17].

Statistical analysis

Continuous variables were reported as mean ± SD or median [interquartile range (IQR)]. Categorical variables were reported as n (%). Comparisons between groups were performed using Student’s t-test or Mann-Whitney U test for continuous variables and χ² or Fisher’s exact test for categorical variables. Kaplan-Meier survival curves were generated to analyze time to treatment response and differences were assessed using the log-rank test. Multivariable logistic regression was used to identify predictors of coexisting TII and TID, and Cox proportional hazards regression was applied to determine predictors of treatment response over time. Statistical significance was defined as P < 0.05. Analyses were performed using SPSS version 30 (IBM Corp., Armonk, NY, United States) and R version 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria).

Several steps were taken to minimize potential bias. Consecutive eligible cases within the study period were included to reduce selection bias. Clinical and histopathological data were obtained from institutional electronic records using predefined criteria to minimize information bias. Multivariable Cox regression analyses were used to adjust for potential confounding variables associated with graft outcomes.

RESULTS

Baseline characteristics

A total of 100 patients with biopsy-proven LN were included. Forty-eight had coexisting TII and TID, and 52 had no TII or TID. Patients were followed for up to 15 years, with a median follow-up of 2 years (IQR: 1-5 years). Patients with TII and TID had lower baseline eGFR (77.7 ± 37.4 mL/minute/1.73 m² vs 98.2 ± 52.7 mL/minute/1.73 m²; P = 0.028) and higher low-density lipoprotein cholesterol (LDL-C; 3.01 ± 1.88 mmol/L vs 1.78 ± 1.82 mmol/L; P = 0.006). Histologically, TII and TID patients had higher modified NIH activity (6.17 ± 3.90 vs 3.13 ± 3.18; P < 0.001) and chronicity scores (4.21 ± 1.74 vs 1.85 ± 2.44; P < 0.001). Class III was more common in TII and TID (35.4% vs 15.4%), while class V/VI predominated in the non-TII/TID group (25.0% vs 10.4%; P = 0.024). Electron microscopy showed more mesangial deposits in TII + TID (30.4% vs 6.0%; P = 0.007). Podocyte effacement ≥ 50% occurred in 50% of patients (P = 0.083). Other demographic characteristics, laboratory results, and medications were similar between groups. These baseline differences set the stage for exploring predictors of TII and TID and subsequent outcomes. Table 1 shows baseline characteristics of the two cohorts.

Table 1 Baseline characteristics of the study population, mean ± SD/n (%).

Variable	TII/TID, n = 52	TII + TID, n = 48	Total, n = 100	P value
Age (year)	26.08 ± 14.89	29.79 ± 10.69	27.86 ± 13.12	0.158
Systolic BP (mmHg)	125.10 ± 24.87	133.74 ± 30.75	129.24 ± 28.03	0.132
Diastolic BP (mmHg)	74.30 ± 15.92	80.28 ± 17.65	77.17 ± 16.95	0.084
Urine protein (g)	3.35 ± 4.52	2.97 ± 2.71	3.17 ± 3.75	0.615
Serum creatinine (µmol/L)	121.27 ± 162.65	126.75 ± 120.38	123.92 ± 143.02	0.852
eGFR (mL/minute/1.73 m2)	98.21 ± 52.71	77.67 ± 37.37	88.35 ± 46.91	0.028
Hemoglobin (g/dL)	10.51 ± 2.05	13.87 ± 16.90	12.12 ± 11.85	0.161
WBC count (× 109/L)	7.17 ± 3.89	7.04 ± 4.02	7.10 ± 3.93	0.888
Platelet count (× 109/L)	230.67 ± 135.81	236.26 ± 150.98	233.53 ± 142.94	0.859
PTH (pmol/L)	8.09 ± 14.35	9.33 ± 10.09	8.72 ± 12.30	0.674
C3 (g/L)	0.70 ± 0.46	0.67 ± 0.30	0.69 ± 0.39	0.648
C4 (g/L)	0.17 ± 0.15	0.14 ± 0.09	0.16 ± 0.13	0.248
Total protein in serum (g/L)	52.37 ± 17.04	51.50 ± 16.02	51.93 ± 16.42	0.825
Albumin in serum (g/L)	28.50 ± 7.59	27.66 ± 7.06	28.09 ± 7.32	0.571
Total cholesterol (mmol/L)	4.75 ± 4.32	4.70 ± 2.36	4.72 ± 3.44	0.951
LDL-C (mmol/L)	1.78 ± 1.82	3.01 ± 1.88	2.40 ± 1.94	0.006
HDL-C (mmol/L)	0.91 ± 0.86	0.94 ± 0.56	0.93 ± 0.72	0.843
TG (mmol/L)	1.32 ± 0.96	1.57 ± 0.86	1.45 ± 0.91	0.253
HA1C at diagnosis (%)	3.74 ± 3.00	4.70 ± 1.61	4.19 ± 2.48	0.069
Sodium (mmol/L)	137.80 ± 4.06	135.53 ± 5.72	136.65 ± 5.07	0.058
Potassium (mmol/L)	4.23 ± 0.60	4.42 ± 0.70	4.32 ± 0.65	0.226
Calcium (mmol/L)	2.05 ± 0.49	2.09 ± 0.24	2.07 ± 0.39	0.626
Phosphorus (mmol/L)	1.30 ± 0.65	1.22 ± 0.38	1.26 ± 0.54	0.517
Uric acid (µmol/L)	313.23 ± 193.36	383.15 ± 178.02	349.14 ± 187.68	0.11
CRP (mg/L)	10.98 ± 17.35	12.85 ± 28.66	11.88 ± 23.40	0.703
ANA	1.20 ± 0.40	1.07 ± 0.26	1.14 ± 0.35	0.077
ANA titer or dilution	621.71 ± 584.77	687.78 ± 492.12	655.21 ± 536.93	0.608
Anti ds-DNA	1.24 ± 0.65	1.45 ± 0.54	1.34 ± 0.61	0.085
Anti ds-DNA titer	215.29 ± 512.00	125.32 ± 355.03	172.14 ± 443.78	0.319
Anti SM antibody	16.59 ± 81.41	1.92 ± 7.11	9.15 ± 57.43	0.285
Anti RNP antibody positivity	0.69 ± 4.06	0.00 ± 0.00	0.34 ± 2.85	0.314
Modified NIH Activity score	3.13 ± 3.18	6.17 ± 3.90	4.59 ± 3.84	0
Chronicity score	1.85 ± 2.44	4.21 ± 1.74	2.98 ± 2.43	0
Gender
Female	43 (82.7)	41 (85.4)	84 (84.0)	0.71
Male	9 (17.3)	7 (14.6)	16 (16.0)
Class
I/II	7 (13.5)	1 (2.1)	8 (8.0)	0.024a
III	8 (15.4)	17 (35.4)	25 (25.0)
III and V	5 (9.6)	8 (16.7)	13 (13.0)
IV	11 (21.2)	7 (14.6)	18 (18.0)
IV and V	8 (15.4)	10 (20.8)	18 (18.0)
V/VI	13 (25.0)	5 (10.4)	18 (18.0)
Hypertension
Yes	33 (64.7)	38 (79.2)	71 (71.7)	0.11
No	18 (35.3)	10 (20.8)	28 (28.3)
DM
Yes	6 (12.0)	9 (19.1)	15 (15.5)	0.33
No	44 (88.0)	38 (80.9)	82 (84.5)
Induction drug: CYC	15 (28.8)	9 (18.8)	24 (24.0)	0.292
MMF	36 (69.2)	39 (81.3)	75 (75.0)
MMF + CYC	1 (1.9)	0 (0.0)	1 (1.0)
Maintenance (yes)	5 (9.6)	2 (4.2)	7 (7.0)	0.502
MMF (maintenance)	38 (73.1)	42 (87.5)	80 (80.0)
Azathioprine (AZA)	3 (5.8)	3 (6.3)	6 (6.0)
MMF + CNI	3 (5.8)	1 (2.1)	4 (4.0)
MMF or AZA	1 (1.9)	0 (0.0)	1 (1.0)
MMF+ AZA + CNI	1 (1.9)	0 (0.0)	1 (1.0)
Podocyte effacement ≥ 50	27 (52.9)	22 (46.8)	49 (50.0)	0.083
Podocyte effacement < 50	14 (27.5)	7 (14.9)	21 (21.4)
No comment or not done	10 (19.6)	18 (38.3)	28 (28.6)
EM mesangium deposits	3 (6.0)	14 (30.4)	17 (17.7)	0.007a
EM
Yes	26 (52.0)	16 (34.8)	42 (43.8)
No	21 (42.0)	16 (34.8)	37 (38.5)
Full house	30 (63.8)	21 (52.5)	51 (58.6)	0.285
Partial	17 (36.2)	19 (47.5)	36 (41.4)
Statin
Yes	17 (34.0)	17 (36.2)	34 (35.1)	0.823
No	33 (66.0)	30 (63.8)	63 (64.9)
ACEI
Yes	24 (55.8)	27 (64.3)	51 (60.0)	0.425
No	19 (44.2)	15 (35.7)	34 (40.0)
ARB
Yes	22 (47.8)	23 (56.1)	45 (51.7)	0.441
No	24 (52.2)	18 (43.9)	42 (48.3)
SGLT-2
Yes	4 (7.8)	4 (8.5)	8 (8.2)	0.904
No	47 (92.2)	43 (91.5)	90 (91.8)

^aP < 0.05.

ACEI: Angiotensin-converting enzyme inhibitor; ANA: Antinuclear antibody; ds-DNA: Double-stranded DNA; RNP: Ribonucleoprotein; SM: Smith; ARB: Angiotensin receptor blocker; AZA: Azathioprine; BP: Blood pressure; CNI: Calcineurin inhibitor; CRP: C-reactive protein; DM: Diabetes mellitus; eGFR: Estimated glomerular filtration rate; EM: Electron microscopy; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; MMF: Mycophenolate mofetil; NIH: National Institutes of Health; PTH: Parathyroid hormone; SGLT-2: Sodium-glucose cotransporter-2; TG: Triglycerides; TII: Tubulointerstitial inflammation; TID: Tubulointerstitial damage; WBC: White blood cell; CYC: Cyclophosphamide.

Predictors of TII and TID

On multivariate analysis, histologic class III (OR = 150.42; 95%CI: 4.28-5285.89; P = 0.006) and class IV + V (OR = 44.11; 95%CI: 1.45-1347.07; P = 0.03) were strong predictors of TII and TID. Among biochemical factors, ANA positivity was associated with a lower risk (OR = 0.011; 95%CI: 0.001-0.199; P = 0.002), whereas higher LDL-C levels were associated with an increased risk (OR = 2.26; 95%CI: 1.29-3.96; P = 0.004). Analysis of individual activity index features showed that interstitial inflammation was the strongest histologic predictor [Exp(B) = 73.29; 95%CI: 8.24-652.25; P < 0.001], with fibrinoid necrosis also significant [Exp(B) = 2.29; 95%CI: 1.03-5.09; P = 0.041]. Other activity index features, including endocapillary hypercellularity, neutrophil/karyorrhexis, cellular/fibrocellular crescents, and hyaline deposits, were not significantly associated. These results highlight the histologic and biochemical factors most closely linked to coexisting tubulointerstitial pathology in LN. Table 2 shows predictors of LN with coexisting TII and TID.

Table 2 Multivariable logistic regression analyses to identify clinical predictors of coexisting tubulointerstitial inflammation and tubulointerstitial damage.

Variable	OR (95%CI)	P value
Age at time of diagnosis	1.044 (0.984-1.109)	0.157
Class
III	150.420 (4.280-5285.889)	0.006a
III and V	14.361 (0.744-277.264)	0.078
IV	8.181 (0.388-172.397)	0.177
IV and V	44.114 (1.445-1347.071)	0.03a
V/VI	2.301 (0.139-38.101)	0.561
ANA	0.011 (0.001-0.199)	0.002a
LDL-C	2.259 (1.288-3.963)	0.004a
eGFR at zero	0.995 (0.979-1.012)	0.579
Endocapillary hypercellularity	0.461 (0.203-1.046)	0.064
Neutrophil and/or karyorrhexis	1.744 (0.720-4.225)	0.218
Fibrinoid necrosis × 2	2.293 (1.034-5.085)	0.041a
Cellular/fibrocellular crescents score × 2	0.897 (0.584-1.378)	0.619
Hyaline deposit (wire loop and/or hyaline thrombi)	1.319 (0.530-3.285)	0.552
Interstitial inflammation (interstitial leukocytes)	73.287 (8.235-652.247)	< 0.001

^aP < 0.05.

ANA: Antinuclear antibody; eGFR: Estimated glomerular filtration rate; LDL-C: Low-density lipoprotein cholesterol.

In our multivariable analyses, the number of outcome events relative to the number of covariates included in the models was limited. For the logistic regression predicting TII and TID, there were 48 events and 12 covariates, yielding an events-per-variable (EPV) of approximately 4, which is below the commonly recommended threshold of 10 for stable estimates. Similarly, in the Cox regression for the composite renal outcome, the number of events was likely fewer than 48, given the similar number of covariates, resulting in an EPV of less than 4. This limited event count likely contributed to the extremely high odds and hazard ratios, with wide confidence intervals, observed in our models. Consequently, these estimates should be interpreted with caution, reflecting potential associations rather than precise effect sizes.

Impact of tubulointerstitial lesions on treatment response

Next, we assessed the association between TII and TID and treatment response. During follow-up, patients with TII and TID were less likely to achieve a favorable treatment response. Kaplan-Meier analysis demonstrated a significantly lower cumulative probability of remission over time in this group (P = 0.0047; Figure 1). These findings suggest that combined tubulointerstitial lesions are associated with poorer treatment response, underscoring the clinical relevance of these histologic abnormalities.

Figure 1 Kaplan-Meier survival shows treatment response over time. The orange curve shows patients without coexisting tubulointerstitial inflammation (TII) and tubulointerstitial damage (TID), and the blue curve shows those with coexisting TII and TID. TII: Tubulointerstitial inflammation; TID: Tubulointerstitial damage.

Predictors of composite outcome

Finally, we examined predictors of the composite outcome in patients with TII and TID. Higher baseline serum creatinine was associated with worse outcomes [HR = 1.010 (1.003-1.018); P = 0.006], and older age was also associated with increased risk [HR = 1.123 (1.025-1.230); P = 0.013]. CRP predicted a higher risk in both the subgroup [HR = 1.062 (1.025-1.100); P = 0.001] and the total cohort [HR = 1.018 (1.003-1.034); P = 0.022]. Among chronic lesions, total glomerulosclerosis was inversely associated with outcome in the TII and TID subgroup [HR = 0.344 (0.135-0.874); P = 0.025] but not in the total cohort [HR = 0.616 (0.323-1.174); P = 0.141]. This may reflect subgroup-specific factors, interactions with other lesions, or limited events, rather than an actual protective effect. Hyaline deposits in acute lesions were associated with a higher risk of the composite renal outcome [HR = 3.923 (1.197-12.853); P = 0.024]. No other clinical or histologic features were significantly associated with outcomes. Other features were not significant. These results suggest that both baseline clinical factors and specific histologic features influence outcomes in patients with combined tubulointerstitial lesions. Table 3 presents the multivariable Cox regression analysis of predictors of the composite outcome in LN with TII and TID.

Table 3 Multivariable Cox-proportional regression analysis to identify predictors for the composite outcome in lupus nephritis patients with coexisting tubulointerstitial inflammation and tubulointerstitial damage, and all the patients.

Variable	TII + TID		Total
	HR (95%CI)	Sig.	HR (95%CI)	Sig.
Baseline characteristics
Age	1.123 (1.025-1.230)	0.013	1.037 (0.999-1.076)	0.055
Gender	0.641 (0.093-4.418)	0.652	1.460 (0.521-4.090)	0.472
Baseline urine protein (g)	0.969 (0.763-1.231)	0.796	0.916 (0.759-1.105)	0.359
Baseline serum creatinine	1.010 (1.003-1.018)	0.006	1.005 (1.002-1.007)	< 0.001
C3	0.001 (0.000-0.134)	0.007	0.260 (0.050-1.336)	0.107
CRP	1.062 (1.025-1.100)	0.001	1.018 (1.003-1.034)	0.022
ANA titer	0.998 (0.997-1.000)	0.051	0.999 (0.998-1.000)	0.205
Chronic pathological features
Total glomerulosclerosis	0.344 (0.135-0.874)	0.025	0.616 (0.323-1.174)	0.141
Fibrous crescents	3.112 (0.899-10.771)	0.073	3.091 (1.063-8.984)	0.038
Interstitial fibrosis	0.002 (approximately 0-approximately ∞)	0.961	0.798 (0.134-4.745)	0.804
Tubular atrophy	4155.207 (approximately 0-approximately ∞)	0.945	1.424 (0.222-9.131)	0.709
Acute pathological features
Endocapillary hypercellularity	0.318 (0.069-1.468)	0.142	1.192 (0.650-2.187)	0.57
Neutrophil and/or karyorrhexis	1.005 (0.338-2.989)	0.993	0.743 (0.366-1.507)	0.41
Fibrinoid necrosis	1.225 (0.737-2.038)	0.434	0.828 (0.554-1.237)	0.357
Cellular/fibrocellular crescents	0.928 (0.458-1.883)	0.837	1.401 (0.947-2.073)	0.091
Hyaline deposit	3.923 (1.197-12.853)	0.024	1.301 (0.573-2.953)	0.53
Interstitial inflammation	2.158 (0.678-6.868)	0.193	0.844 (0.447-1.593)	0.601

Sig.: Significance; TII: Tubulointerstitial inflammation; TID: Tubulointerstitial damage; CRP: C-reactive protein; ANA: Antinuclear antibody.

DISCUSSION

In our study, we identified a notable trend in LN. The presence of TII and TID was closely associated with both baseline renal severity and poorer treatment outcomes. Nearly half of the patients exhibited TII and TID. TII and TD was associated with higher modified NIH activity and chronicity scores, as well as lower baseline eGFR. Histologic classes III and IV + V were strong predictors of TII and TID, whereas higher LDL-C levels were associated with increased risk. Importantly, patients with TII and TID had a slower and reduced rate of achieving a favorable treatment response. Among these patients, higher baseline creatinine, older age, hyaline deposits, and elevated CRP predicted worse outcomes, underscoring the significant influence of tubulointerstitial pathology on both therapeutic response and renal prognosis.

To further contextualize these findings, we analyzed coexisting TII and TID as a combined entity to reflect the overall burden of tubulointerstitial injury. This approach is consistent with prior reports[12,13,16], which demonstrated that evaluating TII and TID together captures clinically meaningful prognostic information, despite partial overlap with components of the modified NIH activity and chronicity indices. While certain elements of TII and TID are embedded within these indices, assessing the composite injury provides valuable insights into renal outcomes and highlights the importance of considering both inflammation and chronic damage. Our findings support the concept that the overall burden of tubulointerstitial lesions, rather than isolated features, is strongly associated with adverse renal outcomes. Therefore, this approach reflects the total pathophysiologic impact of tubulointerstitial lesions without implying complete statistical independence from NIH indices. However, we acknowledge that partial overlap with components of the NIH activity and chronicity indices may introduce collinearity, limiting the independence of the observed associations.

Although LN has been studied in Saudi Arabia and the broader Middle East, most reports focus on general epidemiology. They also primarily describe histologic class distribution and long-term renal outcomes[14,18]. To date, no studies from the region have specifically examined the prognostic significance of coexisting TII and TID or its impact on treatment response. Our study addresses this critical gap. We demonstrated that TII and TID is associated with more severe baseline renal injury, slower and reduced treatment response, and a higher risk of treatment failure in patients with LN. These findings provide novel insights into the clinical relevance of tubulointerstitial pathology in the Middle Eastern population. It also signifies the need to incorporate detailed interstitial evaluation into routine histopathologic assessment.

In our cohort, proliferative glomerular lesions, particularly class III and IV + V, were strongly associated with coexisting TII and TID. This supports the concept that severe glomerular injury can drive tubulointerstitial pathology. It is important to note that, in our multivariable models, the number of outcome events relative to the number of covariates was limited. In the logistic regression predicting TII and TID, there were 48 events for 12 covariates (EPV = approximately 4), and in the Cox regression for the composite renal outcome, the number of events was likely fewer than 48 with a similar number of covariates (EPV < 4). This limited event-per-variable ratio may have contributed to the very high odds and hazard ratios with wide confidence intervals observed in our analyses. Therefore, these estimates should be interpreted cautiously, serving as indicative signals of association rather than precise effect sizes. Despite this limitation, the identified associations remain clinically meaningful and highlight important histologic and biochemical predictors of tubulointerstitial injury and adverse renal outcomes in LN.

Fibrinoid necrosis is another marker of aggressive glomerular injury. It also predicted TII and TID, further reinforcing this link. These lesions may lead to more TII and TID by triggering spillover of immune complexes and inflammatory mediators into the interstitium, with proteinuria and tubular stress further promoting tubular injury and interstitial fibrosis[19,20]. This is supported by the observation that patients with TII and TIN have a higher activity index in our cohort than those without interstitial involvement. In addition, the association between higher LDL-C levels and TII + TID suggests that dyslipidemia may exacerbate endothelial and tubular injury, thereby amplifying inflammatory and fibrotic pathways within the tubulointerstitial compartment, as reported in previous studies[21,22]. Further studies are warranted to clarify the role of LDL-C in promoting tubulointerstitial injury and its consequences. We found an inverse association between ANA positivity and coexisting TII and TID, which may reflect underlying immunologic heterogeneity in LN. Differences in autoantibody profiles, immune activation pathways, and disease phenotypes can influence the pattern of renal injury observed. Variability in serologic expression at the time of biopsy is a significant contributing factor. This association should be interpreted cautiously and considered exploratory, reflecting possible disease heterogeneity rather than a true protective or causal effect of ANA positivity.

The presence of combined TII and TID was associated with a significantly lower likelihood of achieving remission over time. This finding underscores the prognostic relevance of tubulointerstitial pathology. This reflects more advanced and less reversible renal injury. Persistent interstitial inflammation and early fibrotic changes may limit responsiveness to immunosuppressive therapy. These findings are consistent with prior evidence linking TII to poorer treatment response in LN. Increased CD68⁺ macrophage infiltration in the tubulointerstitium has been shown to independently predict a lower likelihood of remission at 12 months[23]. Hsieh et al[12] demonstrated that TII and chronicity independently predict worse renal outcomes in LN, regardless of glomerular class. Severe interstitial injury was associated with reduced renal survival and provided more prognostic information than glomerular lesions. Collectively, these results suggest that TII and TID identifies a higher-risk subgroup requiring closer monitoring and tailored management strategies.

In patients with combined TII and TID, baseline clinical factors were important determinants of adverse renal outcomes. We found that older age in TII and TID was associated with poorer composite outcomes, likely reflecting the decline in renal reserve with aging[24]. Consistent with this, Jeong et al[16] reported that patients with coexisting TII and TID in LN were significantly older than those without these lesions, underscoring the association between age and tubulointerstitial injury. Prior studies in non-lupus TII and TID cohorts also show that older age is a risk factor for adverse renal outcomes, including progression to chronic kidney disease or renal failure[25]. Together, these observations provide biological plausibility for our finding that, among patients with TII and TID in LN, older age is associated with a higher risk of the composite renal outcome.

In our cohort, proliferative glomerular lesions were significantly associated with coexisting TII and TID. Higher baseline serum creatinine in patients with TIN and TID was associated with worse composite outcomes, likely reflecting reduced renal reserve or more advanced kidney injury at the time of biopsy. Similar findings have been reported in LN and other tubulointerstitial disease cohorts, where elevated baseline creatinine consistently predicts adverse renal outcomes. Jeong et al[16] showed that patients with LN who had coexisting TII and TID had more severe disease, proliferative histology, reduced eGFR, higher proteinuria, and poorer renal response. O’Dell et al[26] demonstrated that TII was associated with elevated serum creatinine at biopsy and an increased risk of doubling baseline creatinine. Yu et al[13] reported a progressive increase in serum creatinine with greater severity of interstitial inflammation, tubular atrophy, and fibrosis, with prognostic significance independent of glomerular class. Similarly, Hsieh et al[12] found that moderate-to-severe TII and scarring were associated with higher baseline creatinine and independently predicted long-term renal decline.

Interestingly, we found that higher CRP levels were associated with worse composite outcomes in both the TIN and TID subgroup and the total cohort. Although CRP is often normal in lupus patients, elevated levels in our cohort likely reflect more severe renal inflammation or tissue injury rather than systemic disease activity alone. Supporting this concept, Pesickova et al[27] reported that baseline antiCRP antibodies, which reflect immune reactivity to CRP, were associated with active renal disease and predicted unfavorable composite outcomes, including nonresponse, renal flare, or ESRD in LN during long-term follow-up. Similarly, Yuan et al[28] demonstrated that CRP-related markers, such as urinary modified CRP, closely correlate with the severity of tubulointerstitial lesions in LN and are associated with worse composite renal outcomes. Together, these findings suggest that CRP-related processes, whether systemic or local, may serve as markers of high-risk renal involvement and poor prognosis in LN. In contrast, among chronic lesions, total glomerulosclerosis was inversely associated with adverse outcomes in the TII and TID subgroup (HR = 0.344, 95%CI: 0.135-0.874; P = 0.025). However, when Cox regression was applied to the entire cohort, this effect was lost (HR = 0.616, 95%CI: 0.323-1.174; P = 0.141). This apparent protective association should be interpreted strictly as a subgroup-specific statistical observation and is likely a consequence of the limited number of events, interactions with other lesions, or baseline renal reserve, rather than representing a true protective effect. Additionally, hyaline deposits in acute lesions were associated with a higher risk of adverse composite outcomes (HR = 3.923, 95%CI: 1.197-12.853; P = 0.024), indicating more severe active renal injury. Although direct studies of hyaline deposits are limited, they are components of established LN activity indices that correlate with poorer prognosis[4]. Further studies are needed to explore the prognostic values of hyaline deposits in patients with TII and TID.

Our study has a few strengths. This study provides a detailed assessment of LN in a well-characterized cohort from two tertiary referral centers in Saudi Arabia. All patients underwent comprehensive evaluation of kidney biopsies, with classification according to ISN/RPS criteria and scoring of activity and chronicity indices. The study examines explicitly coexisting TII and TID, which has not been previously studied in the region. By integrating both histologic and biochemical predictors, including proliferative lesions, fibrinoid necrosis, hyaline deposits, CRP, and LDL-C, the study provides a thorough analysis of factors associated with adverse renal outcomes. Additionally, the use of KDIGO 2024 criteria for renal response and composite outcomes enhances the clinical relevance and applicability of the findings.

This study is limited by its retrospective design, which may introduce selection bias and unmeasured confounding. The relatively small sample size, particularly in specific subgroups, resulted in wide confidence intervals and uncertainty in some estimates. In several multivariable models, the number of outcome events relative to the number of covariates was limited, resulting in a low EPV ratio. Additionally, some components of TII and TID are embedded within the modified NIH activity and chronicity indices, which may introduce partial collinearity and limit the statistical independence of the observed associations. Given the exploratory nature of the study and limited EPV, these models should be considered hypothesis-generating rather than definitive estimates of effect size. Penalized regression techniques were not applied, and future studies with larger cohorts are needed to validate these findings and improve model stability. Induction and maintenance immunosuppressive therapies were administered at the discretion of treating clinicians over the extended study period. Although drug classes were recorded, detailed data on dosing, cumulative exposure, and treatment intensity were not available. This may have introduced variability in treatment effects and outcomes. Finally, as a two-center study, the findings may have limited generalizability to other populations, and serial measurements of CRP and LDL-C were not available to assess their longitudinal impact.

CONCLUSION

In conclusion, TII and TID coexist in approximately half of patients with LN. It is associated with more severe baseline renal injury. Proliferative glomerular lesions, fibrinoid necrosis, and higher LDL-C levels were independent predictors of TII and TID. Patients with TII and TID have slower and lower rates of treatment response than those without these lesions. Baseline factors, including higher serum creatinine, older age, elevated CRP, and the presence of hyaline deposits, were associated with a higher risk of adverse composite renal outcomes in this subgroup. Overall, our findings highlight the clinical and prognostic significance of TII and TID and underscore the need for close monitoring and tailored management of high-risk patients.