Targeted-release Budesonide (Nefecon) versus systemic glucocorticoids for IgA nephropathy: a placebo-anchored indirect treatment comparison of randomized controlled trials Authors: Jianbo Qing, Wenjie Wei, Junnan Wu Published: 2026 ABSTRACT AbstractBackgroundIgA nephropathy (IgAN) is a leading cause of end-stage renal disease, particularly in East Asia. Although systemic glucocorticoids are commonly used to delay progression, they are associated with significant adverse effects. Targeted-release budesonide (Nefecon) offers localized immunosuppression, targeting the gut-associated lymphoid tissue (Peyer’s patches), with the aim of reducing systemic exposure. This study evaluates the efficacy and safety of Nefecon compared to systemic glucocorticoids through a placebo-anchored indirect treatment comparison.MethodsPubMed, Embase, and the Cochrane Library were systematically searched for randomized controlled trials evaluating Nefecon (16 mg/day), systemic glucocorticoids, or placebo in adults with IgAN. The primary efficacy outcome was annualized preservation of estimated glomerular filtration rate (eGFR), and secondary outcomes included annualized proteinuria reduction and serious infections. Placebo-anchored indirect treatment comparisons were performed using the Bucher difference-of-estimates method. This study was registered in PROSPERO (CRD420261357809).ResultsThree independent randomized controlled trials were included. Nefecon improved the annualized eGFR trajectory by 2.95 mL/min/1.73 m2 per year compared with placebo (95% CI, 1.68–4.58), whereas systemic methylprednisolone produced an improvement of 2.46 mL/min/1.73 m2 per year (95% CI, 0.94–3.99). The placebo-anchored indirect comparison showed no statistically significant difference in eGFR preservation between the two treatments (MD, 0.49 mL/min/1.73 m2 per year; 95% CI, −1.62 to 2.59; Pindirect = 0.651). For annualized absolute proteinuria reduction, the pooled estimate for Nefecon versus placebo was −0.26 g/g per year (95% CI, −0.34 to −0.18), compared with −0.69 g/g per year (95% CI, −0.97 to −0.41) for systemic methylprednisolone. The indirect comparison favored systemic methylprednisolone (MD, 0.43 g/g per year; 95% CI, 0.13–0.73; Pindirect = 0.004), although this analysis was exploratory because proteinuria measures and follow-up durations required harmonization. The indirect comparison of serious infections was not statistically significant (RR, 0.55; 95% CI, 0.04–7.16; Pindirect = 0.650).ConclusionIn this indirect comparison, both Nefecon and systemic glucocorticoids were associated with improved kidney outcomes versus placebo in IgAN. No statistically significant indirect difference was observed for eGFR preservation, while proteinuria and safety findings require cautious interpretation. Head-to-head trials are needed to further define their relative efficacy and safety.Supplementary informationThe online version contains supplementary material available at https://doi.org/10.1186/s12882-026-05251-y.Keyword: IgA Nephropathy, Nefecon, Systemic Glucocorticoids, Network Meta-analysis, Efficacy and Safety Introduction IgA nephropathy (IgAN) is the most common primary glomerulonephritis worldwide and represents a leading cause of chronic kidney disease among young and middle-aged adults. It is estimated that more than 10 million individuals are affected globally, with substantial geographic heterogeneity in incidence and prevalence [1, 2]. Despite its often indolent onset, up to 30–40% of patients with IgAN progress to (ESKD) within 20–30 years of diagnosis, imposing a considerable burden on patients, healthcare systems, and society [3]. Advances in the understanding of IgAN pathogenesis have culminated in the “four-hit hypothesis,” which describes a multistep process involving the overproduction of galactose-deficient IgA1 (Gd-IgA1), the formation of anti–Gd-IgA1 autoantibodies, the generation of circulating immune complexes, and their mesangial deposition with subsequent activation of inflammatory and fibrotic pathways [4]. These mechanistic insights have historically supported the use of immunosuppressive therapies in patients with high-risk disease. For decades, systemic glucocorticoids have remained the cornerstone of immunosuppressive treatment for IgAN, particularly in patients with persistent proteinuria despite optimized supportive care [5]. Randomized controlled trials (RCTs) have demonstrated their ability to reduce proteinuria and slow kidney function decline. However, these benefits are counterbalanced by a well-documented risk of serious adverse events, including infections, metabolic complications, and cardiovascular toxicity, which has led to substantial controversy regarding the risk–benefit profile of systemic steroid therapy in IgAN [6]. Peyer’s patches in the distal ileum have been identified as a major source of aberrantly glycosylated IgA1 [7]. This paradigm shift has catalyzed the development of targeted therapeutic strategies aimed at modulating intestinal immune responses while minimizing systemic immunosuppression. Targeted-release budesonide (Nefecon) is a locally acting therapy designed to deliver high concentrations of glucocorticoid to the Peyer’s patches of the distal ileum, where it can suppress aberrant mucosal B-cell activation at a key upstream site in the multi-hit pathogenesis of IgAN, while minimizing systemic exposure through extensive first-pass hepatic metabolism [8]. Phase 2 and phase 3 randomized controlled trials have demonstrated that targeted-release budesonide significantly reduces proteinuria and preserves kidney function in patients with IgAN, with a more favorable safety profile compared with conventional systemic glucocorticoids [9]. Nevertheless, direct head-to-head comparisons between targeted-release budesonide and systemic glucocorticoids are lacking, and the relative efficacy and safety of these therapeutic strategies remain uncertain. Therefore, we conducted a systematic review and placebo-anchored indirect treatment comparison of RCTs to compare the efficacy and safety of targeted-release budesonide and systemic glucocorticoids in patients with IgAN receiving background supportive care. By comparing their respective treatment effects versus placebo, this study aimed to provide an exploratory assessment of the comparative therapeutic profiles of gut-targeted and systemic glucocorticoid strategies and to inform future clinical research and treatment decisions in IgAN. Materials and methods Data sources, search strategy, and selection criteria A systematic literature search was conducted to identify RCTs evaluating targeted-release budesonide (Nefecon), systemic glucocorticoids, or placebo in patients with IgAN. PubMed, Embase, and the Cochrane Central Register of Controlled Trials were searched from database inception to January 2025. This study was registered in PROSPERO under the registration number CRD420261357809. The search strategy combined controlled vocabulary terms and free-text keywords related to IgAN and the interventions of interest, including: “IgA nephropathy,” “IgA nephritis,” “budesonide,” “targeted-release budesonide,” “Nefecon,” “glucocorticoids,” “corticosteroids,” and “randomized controlled trial.” No restrictions were applied with respect to publication status. Only studies published in English were considered. Eligible studies met the following criteria:1) RCT design; 2)Adult patients with biopsy-proven IgAN; 3) Comparison of targeted-release budesonide, systemic glucocorticoids, or placebo on a background of optimized supportive care; 4) Reporting at least one prespecified efficacy or safety outcome. Studies enrolling pediatric populations, observational studies, single-arm trials, and conference abstracts without full data were excluded. 5). For trials with multiple publications, only the most complete and non-overlapping report from each parent RCT was included in this study. Interim analyses, subgroup reports, and post hoc or regional cohort analyses derived from the same parent trial were not treated as independent studies, in order to avoid double counting of participants. Two reviewers independently screened titles and abstracts for eligibility, followed by full-text review of potentially relevant articles. Discrepancies were resolved by consensus or consultation with a third reviewer. Data extraction and quality assessment Data were independently extracted by two investigators using a standardized extraction form. For each included trial, the following information was recorded: 1) Study characteristics (author, year, journal, study design, region, and duration of follow-up); 2) Baseline patient demographics (age, sex, and race/ethnicity, with a specific focus on the Asian population and Chinese cohorts); 3) Baseline clinical profiles; 4) Concomitant background therapy, particularly the proportion and optimization of renin-angiotensin system (RAS) blockade. The risk of bias was assessed using the Cochrane Risk of Bias 2.0 tool, covering five domains: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. The TESTING trial included two clinically distinct methylprednisolone treatment phases. Participants enrolled before the protocol amendment received the original full-dose regimen, whereas participants enrolled after the amendment received a reduced-dose regimen together with sulfamethoxazole–trimethoprim prophylaxis. Because the published reports did not provide complete phase-specific estimates and measures of uncertainty for all prespecified efficacy and safety outcomes, the two regimens could not be represented as separate nodes in the quantitative analysis. The combined TESTING estimate was therefore used and was interpreted as representing the overall trial strategy rather than a single homogeneous methylprednisolone regimen. Outcome measures Kidney function outcome: eGFR Kidney function was assessed according to changes in estimated glomerular filtration rate (eGFR). For the primary quantitative analysis, we preferentially used published, directly comparable estimates of absolute treatment effect, including differences in annualized eGFR decline (ml/min/1.73 m2 per year) or annualized effects derived from cumulative between-group absolute differences during follow-up. For NefIgArd and TESTING, comparable published absolute effect estimates and their corresponding standard errors were extracted directly or derived from reported data. For NEFIGAN, the original publication reported arm-specific baseline eGFR values, percentage changes from baseline at 12 months, and a between-group ratio for eGFR, but did not report a directly comparable absolute between-group difference or its standard error. We reconstructed the absolute change within each group by multiplying the reported baseline eGFR by the corresponding percentage change at 12 months. The approximate absolute treatment effect was then calculated as the difference between the reconstructed changes in the Nefecon 16 mg and placebo groups. To approximate the uncertainty of this reconstructed effect, the standard error of the reported 12-month eGFR ratio was first calculated on the logarithmic scale as [ln(upper 95% CI) − ln(lower 95% CI)]/(2 × 1.96). This value was subsequently scaled using the placebo-group baseline eGFR to obtain an approximate standard error on the absolute scale. All inputs, formulas, intermediate calculations, and derived estimates are provided in Supplementary Table S1. Because the NEFIGAN estimate required approximate reconstruction, whereas the NefIgArd phase 3 estimate was directly reported, the NefIgArd estimate was used as the primary Nefecon estimate in the indirect comparison of eGFR preservation. The reconstructed NEFIGAN estimate was included in sensitivity analyses. Because the included trials assessed eGFR outcomes over different observation periods, between-group eGFR differences were converted to annualized estimates to facilitate comparison on a common scale. This transformation was intended only for quantitative harmonization and does not imply a constant rate of eGFR change throughout follow-up or equivalence between active-treatment and post-treatment periods. Proteinuria outcome: UPCR Because the definitions and reporting formats for proteinuria outcomes differed across the included trials, we harmonized treatment effects, where possible, to urine protein-to-creatinine ratio (UPCR; g/g) to improve comparability across studies. For trials reporting absolute between-group differences, the published treatment-effect estimates and corresponding 95% CIs were extracted directly. For studies that did not report absolute between-group differences but instead presented proteinuria outcomes as percentage changes from baseline, geometric least-squares mean ratios, or relative reductions, we derived approximate absolute between-group differences in UPCR using the published baseline UPCR values and relative change estimates. These estimates were then converted to annualized effects, expressed as g/g per year, according to the duration of follow-up to obtain a common effect scale. Corresponding standard errors were derived from the published 95% CIs on the ratio scale and then approximately transformed to the absolute scale. Because 24-hour urinary protein excretion in g/day and UPCR in g/g are commonly used as clinically comparable measures of proteinuria burden in adults, with UPCR numerically approximating 24-hour urinary protein excretion under the assumption of an average daily creatinine excretion of approximately 1 g/day [10, 11], the 24-hour urinary protein values or time-averaged proteinuria reported in TESTING were harmonized to an approximate UPCR-equivalent scale for cross-trial comparison. This conversion was used only for the exploratory comparison of absolute proteinuria effects. Safety outcomes The principal quantitative safety outcome was the incidence of serious infections, including infections requiring hospitalization or resulting in death. Serious infections were treated as a specific category of serious adverse events and were not considered equivalent to the overall incidence of all serious adverse events. Secondary safety outcomes included selected steroid-related treatment-emergent adverse events, such as acne, hypertension, peripheral edema, insomnia, and muscle cramps. Adverse events potentially related to local or mucosal exposure, including rhinitis and upper respiratory tract infection, were also extracted. Because the definitions and reporting of individual adverse events were not sufficiently consistent across trials for formal quantitative synthesis, these secondary safety outcomes were compared descriptively. Certainty of evidence assessment The certainty of evidence for the principal efficacy and safety outcomes was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach. Two reviewers independently assessed the certainty of evidence, with disagreements resolved through discussion or consultation with a third reviewer. For each direct comparison with placebo, evidence from RCTs initially started at high certainty and was evaluated for potential downgrading according to risk of bias, inconsistency, indirectness, imprecision, and publication bias. For the placebo-anchored indirect comparison between Nefecon and systemic glucocorticoids, the initial certainty was determined by the lower certainty rating of the two direct comparisons contributing to the indirect estimate. The certainty was then further assessed for intransitivity across trials and for imprecision of the final indirect estimate. Potential intransitivity was evaluated by comparing clinically relevant effect modifiers across the included trials, including baseline eGFR, baseline proteinuria, racial and ethnic composition, optimization of renin–angiotensin system blockade, glucocorticoid regimen, follow-up duration, and the definition and timing of outcome assessment. Because no direct head-to-head comparison between Nefecon and systemic glucocorticoids was available and the evidence network contained no closed loop, incoherence between direct and indirect evidence could not be formally assessed. Statistical analysis All statistical analyses were performed in R software (version 4.3.0), primarily using the meta (version 8.2–1), metafor (version 4.8–0) and netmeta (version 3.2–0) packages. Continuous and binary outcomes were analysed using distinct approaches. Because the available evidence formed a simple anchored network with placebo as the common comparator and contained no direct comparison between Nefecon and systemic methylprednisolone, formal global and local inconsistency tests were not applicable. We therefore assessed the transitivity assumption by comparing clinically relevant potential effect modifiers across trials, including baseline proteinuria, baseline eGFR, use and optimization of renin–angiotensin system blockade, follow-up duration, glucocorticoid regimen, and racial and ethnic composition. Transitivity was assessed using a prespecified trial-level comparison matrix rather than by statistical hypothesis testing across studies. For each trial, we extracted and compared the reported values or distributions of baseline eGFR, baseline proteinuria, racial and ethnic composition, geographic region, use and optimization of renin–angiotensin system blockade, glucocorticoid regimen, treatment duration, duration of follow-up, and the definition and timing of outcome assessment. These variables were selected because they could plausibly modify the renal efficacy or safety effects of glucocorticoid treatment in IgAN. The comparability of each potential effect modifier was classified qualitatively as broadly comparable, partially comparable, or materially different across trials. No formal statistical tests were performed to compare baseline characteristics between trials because only aggregate trial-level data were available, the reporting formats differed across studies, and the number of included trials was insufficient for reliable meta-regression or treatment-effect modifier analysis. The detailed comparison and its implications for the transitivity assumption are presented in Supplementary Table S2. Transitivity was considered more plausible for comparisons involving the direction of renal treatment effects than for comparisons of their exact magnitude. In particular, differences in baseline proteinuria, baseline kidney function, racial and ethnic composition, systemic glucocorticoid exposure, and follow-up duration were regarded as potential sources of intransitivity. These factors were therefore considered when interpreting the indirect estimates and when grading the certainty of the evidence. Particular attention was paid to differences in ethnic composition between TESTING, which predominantly enrolled Chinese patients, and NEFIGAN and NefIgArd, which enrolled predominantly White populations. Because these differences may be associated with variation in IgAN phenotype, baseline disease risk, treatment response, and adverse-event risk, complete transitivity across the trials could not be assumed. The indirect comparisons were therefore considered exploratory and were interpreted in conjunction with the clinical comparability of the included populations. Given the small number of included trials, statistical tests for heterogeneity had limited power. Where pooling was performed, heterogeneity was assessed using Cochran’s Q and I2, together with consideration of clinical characteristics, methods of outcome derivation, and the direction and magnitude of the treatment effects. For continuous outcomes, including annualized between-group differences in eGFR and proteinuria change, treatment effects were expressed as mean differences (MDs) with 95% CIs. Published treatment-effect estimates and corresponding 95% CIs were extracted directly when available. For studies reporting 95% CIs but not standard errors, standard errors were calculated as the difference between the upper and lower confidence limits divided by (2 \times 1.96). Evidence for Nefecon versus placebo was first considered at the individual-trial level. For annualized UPCR reduction, the estimates from NEFIGAN and NefIgArd were nearly identical, with no observed statistical heterogeneity, and were therefore combined using a common-effect inverse-variance model before the indirect comparison. Because these absolute proteinuria estimates required harmonization of measurement units, effect scales, and follow-up durations, the resulting comparison was considered exploratory. Relative proteinuria effects were additionally compared descriptively without formal indirect hypothesis testing. For eGFR preservation, the primary indirect comparison used the directly reported phase 3 NefIgArd estimate. The NEFIGAN estimate required approximate reconstruction and differed from the NefIgArd estimate in both magnitude and method of derivation; therefore, the two estimates were not pooled for the primary analysis. Sensitivity analyses were performed using the common-effect pooled estimate from NEFIGAN and NefIgArd, the NEFIGAN estimate alone. Placebo-anchored indirect comparisons were performed using the Bucher difference-of-estimates method. For continuous outcomes, the indirect effect for Nefecon versus systemic methylprednisolone was calculated by subtracting the effect of systemic methylprednisolone versus placebo from the effect of Nefecon versus placebo. Assuming independence between the contributing comparisons, the standard error of the indirect estimate was calculated as the square root of the sum of their variances. The 95% CI was calculated as the indirect estimate ±1.96 times its standard error. For serious infections, direct treatment effects were expressed as risk ratios (RRs) with 95% CIs and were calculated from the reported event counts and total sample sizes. NefIgArd contributed the direct estimate for Nefecon versus placebo, and TESTING contributed the direct estimate for systemic methylprednisolone versus placebo. NEFIGAN reported no serious infections in either treatment group. Because a conventional RR cannot be estimated from a study with zero events in both groups, NEFIGAN was included in the descriptive comparison of absolute event rates but did not contribute to the relative-risk or indirect-comparison calculations. No continuity correction was applied to include this double-zero study. The indirect comparison of serious infection risk was performed on the logarithmic RR scale. The indirect log RR was calculated by subtracting the log RR for systemic methylprednisolone versus placebo from the log RR for Nefecon versus placebo. Its standard error was calculated as the square root of the sum of the corresponding log-RR variances. The indirect estimate and its confidence limits were then exponentiated to obtain the RR and 95% CI. Given the small number of serious infection events, interpretation focused on the width of the confidence interval and the absolute event rates rather than on the point estimate alone. Forest plots were generated in R, with RRs displayed on a logarithmic scale. All statistical tests were two-sided, and a p value < 0.05 was considered statistically significant. Results Literature search and study selection Using the prespecified search strategy and eligibility criteria, we identified 401 records from PubMed, Embase and the Cochrane CENTRAL databases. After removal of duplicates and screening of titles and abstracts, potentially eligible full-text articles were assessed in detail. Because some publications represented interim reports, regional cohort analyses or subgroup analyses from the same parent trial, we included, for the primary quantitative analysis, only the report from each parent trial with the most complete data, the longest follow-up and no overlap in study population, to avoid double counting of participants. Ultimately, three independent randomized controlled trials were included in the primary analysis: NEFIGAN [12], NefIgArd [9] and TESTING [13] (Fig. 1). PRISMA flow diagram of literature search and study selection. The diagram summarizes the identification, screening, full-text eligibility assessment, and inclusion process. Seven full-text reports were assessed, of which three independent parent RCTs were included in the primary quantitative analysis NEFIGAN and NefIgArd evaluated targeted-release budesonide (Nefecon) at a dose of 16 mg per day versus placebo on top of optimized supportive care, whereas TESTING evaluated systemic glucocorticoid therapy with methylprednisolone versus placebo. Together, these three trials constituted the principal evidence base for an anchored indirect comparison using placebo as the common comparator (Fig. 2). Conceptual network of treatment comparisons included in the primary quantitative analysis. NEFIGAN and NefIgArd directly compared Nefecon with placebo, whereas TESTING directly compared systemic glucocorticoids with placebo. No head-to-head randomized trial directly compared Nefecon with systemic glucocorticoids; therefore, their relative effects were estimated through a placebo-anchored indirect comparison All three included studies were RCTs. NEFIGAN and NefIgArd were double-blind, placebo-controlled trials, and TESTING was an international, multicentre, double-blind RCT. The racial and ethnic composition differed substantially across the included trials. Approximately 75–76% of participants in TESTING were reported as Chinese, whereas only 22–24% of participants in NefIgArd were categorized as Asian, and 96% of participants in NEFIGAN were categorized as Caucasian. All trials enrolled adults with biopsy-confirmed IgAN who were receiving optimized supportive care, including a high proportion of patients treated with renin–angiotensin system blockade. Overall, baseline kidney function and proteinuria were broadly comparable across studies, with baseline eGFR generally around 55–65 ml/min/1.73 m2; however, baseline proteinuria was higher in TESTING. The key design characteristics and baseline patient characteristics of the included trials are summarized in Supplementary Table 3. Overall, the three trials included in the primary analysis were of high methodological quality and used randomized allocation. The placebo-controlled trials were double-blinded, supporting the internal validity of the subsequent indirect comparison. Comparative efficacy in renal preservation and proteinuria reduction In the primary placebo-anchored indirect comparison of renal function outcomes, the phase 3 NefIgArd estimate was used to represent the effect of Nefecon. Nefecon improved the annualized eGFR trajectory by 2.95 mL/min/1.73 m2 per year compared with placebo (95% CI, 1.68–4.58), whereas systemic methylprednisolone produced an improvement of 2.46 mL/min/1.73 m2 per year compared with placebo (95% CI, 0.94–3.99). The indirect comparison showed no statistically significant difference between Nefecon and systemic methylprednisolone in annualized eGFR preservation (MD, 0.49 mL/min/1.73 m2 per year; 95% CI, −1.61 to 2.59; Pindirect = 0.651; Fig. 3). Effects of Nefecon and systemic methylprednisolone on renal function preservation. Forest plot showing direct and placebo-anchored indirect comparisons of annualized eGFR preservation, expressed as the mean difference in eGFR trajectory in mL/min/1.73 m2 per year. The primary indirect comparison used the phase 3 NefIgArd estimate for Nefecon. Estimates based on the common-effect pooling of NEFIGAN and NefIgArd are presented as sensitivity analyses. Positive values favor the first-named treatment. Error bars indicate 95% confidence intervals Because the eGFR estimates differed between NEFIGAN and NefIgArd, their pooled estimate was examined as a sensitivity analysis rather than the primary estimate. Under a common-effect inverse-variance model, the pooled effect of Nefecon versus placebo was 3.17 mL/min/1.73 m2 per year (95% CI, 1.75–4.60). The corresponding indirect comparison with systemic methylprednisolone remained non-significant (MD, 0.71 mL/min/1.73 m2 per year; 95% CI, −1.37 to 2.80; Pindirect = 0.502). For proteinuria, both treatments reduced annualized UPCR compared with placebo. The effects of Nefecon in NEFIGAN and NefIgArd were nearly identical, with no observed heterogeneity (I2 = 0%; P for heterogeneity = 0.993; Figure S2). The common-effect pooled estimate for Nefecon versus placebo was −0.26 g/g per year (95% CI, −0.34 to −0.18), compared with −0.69 g/g per year for systemic methylprednisolone versus placebo. The placebo-anchored indirect comparison yielded an MD of 0.43 g/g per year for Nefecon versus systemic methylprednisolone (95% CI, 0.13–0.73; Pindirect = 0.004; Fig. 4). This exploratory harmonized analysis yielded a statistically significant difference in the estimated absolute proteinuria effect, numerically favoring systemic methylprednisolone. However, because the comparison required harmonization across UPCR, 24-hour urinary protein excretion, relative treatment effects, and different follow-up periods, it should not be interpreted as a definitive demonstration of superior antiproteinuric efficacy. Comparative efficacy of Nefecon and systemic steroids on proteinuria reduction. Forest plot illustrating the annualized MD in urine UPCR (g/g per year) compared with placebo. Data points represent the pooled MD for Nefecon (derived from NEFIGAN and NefIgArd trials) and the MD for systemic steroids (derived from the TESTING trial). Error bars indicate 95%CIs However, this finding should be interpreted with caution, as baseline proteinuria levels and definitions of proteinuria endpoints were not fully consistent across studies; in particular, baseline proteinuria was substantially higher in the TESTING cohort than in NEFIGAN and NefIgArd. To reduce the influence of baseline differences on comparisons of absolute effects, we further performed a descriptive comparison of relative changes in proteinuria. In TESTING, mean proteinuria during follow-up was 1.70 g/day in the methylprednisolone group and 2.39 g/day in the placebo group, corresponding to an approximately 28.9% lower proteinuria level. In NefIgArd, Nefecon reduced UPCR by 30.1% versus placebo at 24 months (95%CI, 16.4–41.5). In NEFIGAN, the UPCR ratio for Nefecon 16 mg/day versus placebo at 12 months was 0.68 (95%CI, 0.57–0.96), corresponding to an approximately 32% relative reduction. Overall, although analyses based on absolute differences in proteinuria suggested a larger apparent antiproteinuric effect with systemic glucocorticoids, descriptive comparisons based on relative effects indicated that the magnitude of proteinuria reduction was broadly similar between the two treatment strategies. Serious infections and steroid-related adverse events Serious infection rates differed numerically across the included trials. In TESTING, serious infections occurred in 17 of 257 participants receiving systemic methylprednisolone (6.6%) and in 3 of 246 participants receiving placebo (1.2%), corresponding to an RR of 5.42 (95% CI, 1.61–18.28; p = 0.006). In NefIgArd, serious infections occurred in 3 of 182 participants receiving Nefecon (1.65%) and in 1 of 182 participants receiving placebo (0.55%), corresponding to an RR of 3.00 (95% CI, 0.31–28.57; p = 0.339). No serious infections were reported in either treatment group in NEFIGAN. Because the study contained zero events in both groups, NEFIGAN was included in the descriptive comparison of absolute event rates but did not contribute to the conventional RR calculation. The placebo-anchored indirect comparison yielded an RR of 0.55 for Nefecon versus systemic methylprednisolone (95% CI, 0.04–7.16; p = 0.650; Fig. 5). Although the point estimate was numerically below 1, the confidence interval was extremely wide and included both a potentially lower and a potentially higher risk with Nefecon. The indirect comparison therefore did not establish a statistically significant difference in serious infection risk between the two treatment strategies. Direct and placebo-anchored indirect comparisons of serious infection risk. The forest plot shows the direct comparisons of Nefecon and systemic methylprednisolone with their respective placebo groups and the placebo-anchored indirect comparison of Nefecon versus systemic methylprednisolone. Risk ratios and 95% confidence intervals are displayed on a logarithmic scale. The indirect comparison was calculated on the log-risk-ratio scale using the Bucher method. NEFIGAN was included in the descriptive event-rate comparison but did not contribute to the RR calculation because no serious infections occurred in either treatment group Formal indirect comparisons of other treatment-emergent adverse events were not performed because the event definitions and reporting formats were not sufficiently consistent across the trials. Across NEFIGAN and NefIgArd, acne occurred in 11.0%–18.4% of participants receiving Nefecon, hypertension in 10.2%–12.1%, and peripheral edema in 12.2%–17.0%, with generally lower incidences in the corresponding placebo groups. Insomnia and muscle cramps were also reported more frequently in some Nefecon groups. By contrast, the incidences of rhinitis and upper respiratory tract infection were generally similar between the Nefecon and placebo groups (Supplementary Table 4). These findings indicate that targeted-release budesonide does not completely eliminate glucocorticoid-related adverse effects. However, differences in treatment exposure, follow-up duration, adverse-event definitions, and reporting methods, together with the imprecision of the serious infection comparison, preclude firm conclusions regarding comparative safety. In addition, the certainty of evidence was rated as low for the indirect comparisons of eGFR preservation and annualized absolute proteinuria reduction, and as very low for serious infections, primarily because of intransitivity and imprecision (Supplementary Table S5). Discussion IgAN remains one of the most prevalent primary immune-mediated glomerular diseases, particularly in East Asia [14]. Despite its strong heterogeneity, the central pathogenic role of abnormal mucosal immunity in IgAN has been consistently confirmed [15]. This has led to the exploration of novel therapeutic strategies targeting this immune dysregulation. Currently, treatment options for IgAN are limited, and immunosuppressive agents, especially systemic glucocorticoids, remain the cornerstone of therapy [16]. However, these treatments are not without significant drawbacks, especially in young patients who often require long-term use, leading to concerns about cumulative adverse effects. While systemic glucocorticoids (e.g., methylprednisolone) have shown efficacy in reducing proteinuria and slowing renal function decline in IgAN, they are associated with a wide range of adverse effects, including weight gain, hypertension, diabetes, osteoporosis, and increased susceptibility to infections [17]. These adverse effects are particularly concerning in younger patients, who often require prolonged treatment courses, further complicating the risk–benefit ratio of glucocorticoid therapy. The need for safer and more mechanism-based alternatives has led to the development of targeted-release budesonide, a locally acting therapy designed to deliver glucocorticoid selectively to the Peyer’s patches of the distal ileum. By suppressing aberrant mucosal B-cell activation within the gut-associated lymphoid tissue (GALT), Nefecon is intended to intercept the pathogenic cascade of IgAN at an upstream source, while minimizing the broad systemic effects associated with conventional glucocorticoid [18, 19]. The present placebo-anchored indirect comparison integrated evidence from the phase 2b NEFIGAN trial, the phase 3 NefIgArd trial, and the TESTING trial to provide an exploratory comparison of Nefecon and systemic methylprednisolone. The principal finding was an apparent divergence between the proteinuria and kidney function outcomes. In the harmonized absolute analysis, systemic methylprednisolone was associated with a greater reduction in proteinuria than Nefecon. In contrast, the indirect comparison did not identify a statistically significant difference between the two treatment strategies in eGFR preservation. The safety analyses showed different observed patterns of serious infection, although the indirect comparison was imprecise and did not establish a definitive comparative safety advantage. The proteinuria findings varied according to the analytical scale used. In TESTING, methylprednisolone was associated with an approximately 28.9% lower time-averaged proteinuria level than placebo. Nefecon was associated with a 30.1% reduction in UPCR versus placebo at 24 months in NefIgArd and an approximately 32% lower UPCR level at 12 months in NEFIGAN. Thus, the relative proteinuria effects were broadly similar in magnitude across the trials. However, after conversion to a harmonized absolute scale, systemic methylprednisolone appeared to produce a greater absolute reduction in proteinuria. This apparent difference should be interpreted cautiously. Absolute proteinuria reductions are strongly influenced by baseline proteinuria levels, and the included trials differed in baseline disease severity, proteinuria measurement, endpoint definition, and follow-up duration. In particular, NEFIGAN enrolled participants with lower baseline proteinuria than TESTING. The conversion of 24-hour urinary protein excretion and trial-specific relative effects to an approximate UPCR-equivalent absolute scale also required several assumptions. Therefore, the apparent advantage of systemic methylprednisolone in absolute proteinuria reduction may partly reflect cross-trial differences rather than an intrinsically greater treatment effect. For this reason, the absolute proteinuria comparison should be considered exploratory and interpreted together with the relative effects reported in the individual trials. With respect to kidney function, both Nefecon and systemic methylprednisolone were associated with attenuation of eGFR decline compared with placebo. The primary indirect comparison did not show a statistically significant difference between the two treatment strategies. However, the absence of a statistically significant indirect difference should not be interpreted as evidence of equivalence. The number of contributing trials was small, and the Nefecon and TESTING studies differed in baseline kidney function, patient characteristics, treatment regimen, and duration of follow-up. Nevertheless, the consistency in the direction of the effects suggests that both strategies may preserve kidney function in patients at risk of IgAN progression. These findings also raise the possibility that gut-targeted treatment may achieve favorable kidney outcomes without requiring the same degree of broad systemic immunosuppression, although this hypothesis requires confirmation in a direct head-to-head trial. The safety findings require cautious interpretation. In TESTING, systemic methylprednisolone was associated with a significantly increased risk of serious infection compared with placebo. In the available Nefecon trials, serious infections were uncommon, and the direct comparison in NefIgArd was not statistically significant. However, the placebo-anchored indirect comparison between Nefecon and systemic methylprednisolone was highly imprecise and did not demonstrate a statistically significant difference. Therefore, the results indicate different observed serious-infection signals across the trials but do not provide definitive evidence that Nefecon has a lower serious infection risk than systemic methylprednisolone. Nefecon was also associated with higher incidences of acne, hypertension, and peripheral edema than placebo, indicating that targeted intestinal delivery and extensive first-pass metabolism do not completely eliminate systemic glucocorticoid-related adverse effects. Direct quantitative comparison of these events with systemic methylprednisolone was not possible because adverse-event definitions, ascertainment periods, and reporting formats differed across trials. An important consideration in interpreting all indirect estimates is the transitivity of the contributing trials. The explicit trial-level assessment identified both similarities and clinically relevant differences. All three trials enrolled adults with biopsy-proven primary IgAN and persistent proteinuria despite renin–angiotensin system blockade, and all used placebo groups receiving background supportive care. These features support the use of placebo as a common comparator. However, the common comparators were not fully exchangeable. NEFIGAN enrolled participants with better baseline kidney function and lower proteinuria, whereas TESTING enrolled a higher-risk population and used two clinically distinct systemic methylprednisolone regimens. The trials also differed in glucocorticoid exposure, infection prophylaxis, treatment duration, follow-up length, and the timing and definition of renal outcomes. Transitivity was therefore considered only partially supported. The indirect estimates are best interpreted as comparisons of trial-defined treatment strategies rather than estimates that isolate the pharmacological difference between targeted-release budesonide and systemic methylprednisolone. Differences in racial and ethnic composition were a particularly important potential source of intransitivity. TESTING predominantly enrolled Chinese participants, whereas NEFIGAN enrolled an almost entirely White population and NefIgArd enrolled a predominantly White population with a smaller Asian subgroup. East Asian patients with IgAN may differ from Western populations in clinical presentation, histopathological characteristics, baseline progression risk, treatment responsiveness, and susceptibility to glucocorticoid-related adverse events. In addition, although Nefecon is designed for targeted ileal delivery with low systemic exposure, some systemic budesonide exposure exposure may occur, and its multi-target glucocorticoid-mediated anti-inflammatory effects may be particularly relevant in Chinese patients with a greater burden of active inflammatory lesions. Consequently, the larger absolute proteinuria reduction and the magnitude of benefit observed with the TESTING treatment strategy may partly reflect the baseline-risk profile and disease phenotype of its enrolled population. Individual participant-level data on baseline risk, Oxford MEST-C lesions, genetic background, and other potential effect modifiers were not available for the present analysis. It was therefore not possible to determine the extent to which ethnic and clinical differences contributed to the observed treatment effects. The results should not be interpreted as definitive evidence that systemic glucocorticoids possess intrinsically greater efficacy than targeted-release budesonide, nor should they be generalized across ethnic populations without appropriate caution. Several additional limitations should be acknowledged. First, no head-to-head randomized controlled trial has directly compared Nefecon with systemic methylprednisolone. The principal conclusions were therefore based on placebo-anchored indirect comparisons, which provide less certain evidence than direct randomized comparisons. Second, only three independent parent trials were available, and the evidence structure consisted of a simple anchored network without a closed loop. Formal assessment of inconsistency between direct and indirect evidence was therefore not possible, and estimates of heterogeneity were imprecise. Third, the included trials differed in the definition, timing, and duration of efficacy outcome assessment. This was particularly relevant for eGFR. TESTING assessed kidney function over a median follow-up of 3.5 years, including both active treatment and post-treatment observation, whereas the Nefecon trials assessed eGFR over shorter and differently structured treatment and follow-up periods. Annualization placed the estimates on a common numerical scale but could not account for potential differences in the temporal pattern of treatment effects. Phase-specific eGFR estimates from TESTING were unavailable, preventing a reliable comparison restricted to the active-treatment period. Fourth, the absolute proteinuria analysis required harmonization of UPCR, 24-hour urinary protein excretion, relative reductions, and follow-up durations. Although these transformations facilitated an exploratory cross-trial comparison, they introduced additional uncertainty and should not be regarded as equivalent to analyses based on directly reported, identically defined outcomes. Finally, the safety analyses were limited by the small number of serious infection events and by differences in glucocorticoid dose, infection prophylaxis, follow-up duration, and adverse-event reporting. The observed safety patterns are therefore clinically informative but do not establish definitive comparative safety. In conclusion, both Nefecon and systemic methylprednisolone were associated with favorable renal effects compared with placebo in patients with IgAN. Systemic methylprednisolone appeared to produce a greater absolute reduction in proteinuria, although relative proteinuria effects were broadly similar and the absolute comparison was sensitive to cross-trial differences in baseline proteinuria and outcome definition. No statistically significant difference was identified in eGFR preservation or serious infection risk in the indirect comparisons, but the estimates were limited by imprecision and only partial transitivity. These findings should be regarded as hypothesis-generating and support the need for direct comparative trials with standardized outcomes and adequate representation of different ethnic populations. Conclusion In this indirect comparison, both Nefecon and systemic glucocorticoids were associated with improved kidney outcomes compared with placebo in patients with IgAN. No statistically significant indirect difference was observed between the two strategies in preserving eGFR. Although systemic glucocorticoids showed a larger reduction in absolute proteinuria, the relative reductions in proteinuria were broadly comparable between treatments. Compared with placebo, systemic glucocorticoids were associated with a significantly increased risk of serious infection, whereas serious infections were infrequent with Nefecon. However, given the limited indirect evidence comparing Nefecon with systemic glucocorticoids, the relative efficacy and safety of these strategies require further evaluation in head-to-head trials. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Supplementary Material 5 Acknowledgements We thank all individuals who provided assistance and support during the conduct of this study. Abbreviations Confidence interval Estimated glomerular filtration rate End-stage kidney disease Gut-associated lymphoid tissue Galactose-deficient immunoglobulin A1 Immunoglobulin A nephropathy Mean difference Network meta-analysis Randomized controlled trial Risk ratio Serious adverse event Treatment-emergent adverse event Targeted-release formulation Urine protein-to-creatinine ratio Author contributions Conceptualization: J.Q.; Investigation: J.Q.; Visualization: J.Q.; Resources: W.W.; Funding acquisition: J.W. Project administration: J.W. Supervision: J.W. Writing—original draft: W.W. and J.Q.; Writing–review & editing: W.W. All authors have read and approved the article. Funding This study was supported by funding from Nefecon Therapeutics. The funder had no role in the study design, literature search, study selection, data extraction, statistical analysis, interpretation of the results, manuscript preparation, or the decision to submit the manuscript for publication. Data availability The data generated or analyzed in this study are provided in the table within the manuscript. Any additional data related to this study are available from the corresponding author upon reasonable request. Declarations Ethical approval Not applicable. Consent for publication Not applicable. Consent for participation Not applicable. Contributors not qualifying for authorship Not applicable. AI assistance Not applicable. Competing interests Dr. Junnan Wu has received funding from Nefecon Therapeutics. The remaining authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Supplementary Material 5 Data Availability Statement The data generated or analyzed in this study are provided in the table within the manuscript. Any additional data related to this study are available from the corresponding author upon reasonable request.