Which commercial Chinese polyherbal preparations combined with ACEI/ARB is effective and safe for IgA nephropathy? A systematic review and network meta-analysis Authors: Li Zheng, Xiaotong Gu, Lin Zhang, Chumeng Yang, Xuelin Sun, Weina Zhang, Songtao Yang Published: 2026 ABSTRACT BackgroundThis systematic review and network meta-analysis (NMA) evaluates and compares the efficacy and safety of six commercial Chinese polyherbal preparations (Bailing Capsule (BL), Huangkui Capsule (HK), Tripterygium Glycosides (TG), Jinshuibao Capsule (JSB), Huobahuagen Tablet (HBHG), and Shenyan Kangfu Tablet (SYKF)) combined with angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers (ACEI/ARB) for treating Immunoglobulin A nephropathy (IgAN), offering updated evidence to support clinical decision-making.MethodsWe conducted a systematic literature search in PubMed, Embase, the Cochrane Library, ClinicalTrials.gov, SinoMed, China National Knowledge Infrastructure (CNKI), Wanfang Data Knowledge Service Platform, and VIP Database for Technology Periodicals (up to 20 October 2025) for randomized controlled trials (RCTs) that enrolled patients with IgAN and compared the combination of six commercial Chinese polyherbal preparations with ACEI/ARB against ACEI/ARB therapy alone. To synthesise the evidence from the network of treatments, we performed frequentist network meta-analyses and subsequently assessed the certainty of evidence through the systematic application of the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach. The study is registered in the PROSPERO database (CRD420251175045).ResultsA total of 70 RCTs involving 4855 patients were included in this systematic review and NMA. The overall risk of bias was high among the included studies. Compared with ACEI/ARB alone, BL combined with ACEI/ARB probably improves overall response rate (OR: 3.27, 95% CI: 2.21 to 4.84; moderate certainty) and reduces 24‐h urine protein quantification (MD: −0.83 g, 95% CI: −1.21 to −0.44; moderate certainty). HBHG combined with ACEI/ARB may reduce serum creatinine (MD: −26.63 μmol/L, 95% CI: −47.91 to −5.35; low certainty), while TG combined with ACEI/ARB may reduce blood urea nitrogen (MD: −2.53 mmol/L, 95% CI: −4.54 to −0.52; moderate certainty). BL combined with ACEI/ARB may also improve hemoglobin (MD: 9.41 g/L, 95% CI: 8.18 to 10.64; low certainty). No convincing differences were observed in adverse drug reactions between most commercial Chinese polyherbal preparations combined with ACEI/ARB (very low to moderate certainty), although HK combined with ACEI/ARB showed a relatively higher incidence. Sensitivity analyses supported the robustness of findings. Egger’s test suggested potential publication bias for overall response rate and 24‐h urine protein quantification.ConclusionIn patients with IgAN, BL combined with ACEI/ARB may outperform other Chinese- Biomedicine regimens in reducing 24‐h urine protein quantification, with all combinations potentially superior to ACEI/ARB alone. However, evidence certainty remains low, highlighting the need for more rigorous, long-term RCTs for validation.Systematic Review Registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420251175045, PROSPERO: CRD420251175045. Introduction Immunoglobulin A nephropathy (IgAN), the most common primary glomerular disease globally (Jarrick et al., 2019; Schena and Nistor, 2018), progresses to end-stage renal disease (ESRD) in approximately 30% of patients. Treatment options remain limited and often yield unsatisfactory outcomes (Coppo, 2017; Wang et al., 2025). The cornerstone of conventional IgAN management involves renin-angiotensin system (RAS) inhibitors, namely, angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin II receptor blockers (ARBs) (Kidney Disease: Improving Global Outcomes CKD Work Group, 2024; Cheng et al., 2009). Additionally, traditional immunosuppressive therapies, including glucocorticoids, cyclophosphamide, and tacrolimus, are available (Chen et al., 2022). However, their long-term adverse reactions and the risk of disease relapse after withdrawal have become new clinical challenges (Li et al., 2023; Zhang et al., 2023). Commercial Chinese polyherbal preparations (CCPPs) can significantly compensate for the limitations of Western therapy in IgAN. Despite this potential, clinicians lack evidence for selecting a first-choice agent. Some systematic reviews have summarized the benefits of CCPPs combined with ACEI/ARB. Nevertheless, these pairwise meta-analyses are typically constrained by small sample sizes and, more importantly, have failed to apply the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework to evaluate the certainty of the supporting evidence (Chen et al., 2022; Li et al., 2025). To our knowledge, no network meta-analysis (NMA) has evaluated the efficacy and safety of six commonly used CCPPs (Bailing Capsule (BL), Huangkui Capsule (HK), Tripterygium Glycosides (TG), Jinshuibao Capsule (JSB), Huobahuagen Tablet (HBHG), and Shenyan Kangfu Tablet (SYKF)) in combination with ACEI/ARB therapy for IgAN. The optimal integrative treatment strategy combining Chinese and Biomedicine for IgA nephropathy remains undetermined. Therefore, we conducted a systematic review and NMA of relevant randomized controlled trials (RCTs) to evaluate the efficacy and safety of integrated CCPPs and Biomedicine (CCPPs-B) for IgAN, aiming to provide objective evidence-based findings for clinical practice. Methods This systematic review was registered in PROSPERO (CRD420251175045), and it was reported in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) (Hutton et al., 2015). To improve transparency in the reporting of plant-derived medicinal interventions, we assessed the reporting completeness of the included commercial Chinese medicinal products using the ConPhyMP framework. Search strategy and selection criteria We searched PubMed, Embase, the Cochrane Library, the ClinicalTrials.gov, CNKI, Wanfang, VIP and SinoMed from database inception up to 20 October 2025, and screened the reference lists of relevant systematic reviews for additional trials. The search was conducted using a combination of controlled vocabulary (e.g., MeSH) and free-text terms. The search terms included “iga nephropath”, “tripterygium glycosides”, “bailing”, “jinshuibao”, “huangkui”, “huobahuagen”, “shenyankangfu”, and “randomized controlled trials” (Supplementary Appendix 1). RCTs were included in this NMA if they met the following eligibility criteria: 1) Patients with IgA nephropathy confirmed by clearly defined diagnostic criteria; 2) The intervention group received treatment with CCPPs, including BL, JSB, HK, TG, HBHG, and SYKF combined with ACEI/ARB, while the control group received ACEI/ARB treatment alone; 3) Outcome measures included overall response rate (refers to the effective rate plus the marked effective rate, where “effective” means a 50%–70% reduction in 24-h urinary protein, and “marked effective” means 24-h urinary protein less than 0.3 g/day), 24-h urinary protein quantity (24hUPQ), serum creatinine (Scr), blood urea nitrogen (BUN), hemoglobin (Hb), and adverse drug reactions (ADRs). We excluded animal studies, studies without clearly defined diagnostic criteria, and studies for which full text was unavailable. Study selection and data extraction Using Endnote X9, pairs of reviewers (LZ, XTG, XLS, CMY and LZ*) independently screened titles and abstracts of all references and corresponding full-text documents initially deemed relevant. The identification of a retraction in any of the eligible publications led to the study’s exclusion from our analysis (Brown et al., 2022). Two reviewers (CMY and WNZ) independently extracted data on study characteristics, patient characteristics, intervention measures, and outcomes of interest. The process for resolving discrepancies involved discussion between the initial reviewers, with an additional reviewer (LZ*) being brought in to make a final decision if required. Risk of bias assessment Pairs of reviewers (LZ, XTG, XLS, CMY and LZ*) independently evaluated the risk of bias of included RCTs by using the Cochrane Risk of Bias Tool 2.0 (RoB 2.0) (Sterne et al., 2019) based on the following domains: bias arising from the randomisation process, bias due to deviations from intended interventions, bias due to missing outcome data, bias in measurement of the outcome, bias in selection of the reported result and other risk of bias. If the risk of bias assessment in all domains is judged as “low risk,” then the overall risk of bias is considered “low risk”. Conversely, if some domains are assessed as having “some concerns” and no domain is rated as “high risk,” the overall risk of bias is classified as “some concerns.” As long as one domain is evaluated as “high risk,” the overall risk of bias is deemed “high risk.” Data analysis Pairwise meta-analyses for all direct comparisons of each outcome were conducted utilizing the Hartung-Knapp-Sidik-Jonkman (HKSJ) random-effects model, implemented in R software (version 4.2.1). Using network packages in R (version 4.5.2), we performed random-effects NMA. For each outcome, the design-by-treatment model (global test) was employed to evaluate the coherence assumption across the entire network (Higgins et al., 2012). Indirect estimates were derived through node-splitting and back-calculation methods (König et al., 2013). To assess the inconsistency, we computed corresponding P values and deviance information criterion (DIC) for inconsistency testing (Dias et al., 2013). For dichotomous outcomes, we calculated odds ratio (ORs) with 95% CIs for overall response rate and ADRs. For continuous outcomes, we calculated mean differences (MDs) with 95% CIs. In cases of missing SDs, these were imputed following the guidelines of the Cochrane Handbook (Higgins et al., 2022). Between-study heterogeneity was assessed using the I2 statistic and through visual inspection of forest plots. For meta-analyses comprising ten or more studies, we performed the Egger’s test for the outcomes by using the statistical software R (version 4.5.2) to assess publication bias, and also used the test to identify outliers (Egger et al., 1997). To assess the robustness of the pooled results and examine the sources of heterogeneity between studies, we conducted a sensitivity analysis. The relative performance of the treatments was inferred from the Surface Under the Cumulative Ranking Curve (SUCRA). SUCRA values range from 0% to 100%, where higher values represent a more favorable ranking. We used the GRADE approach to assess certainty of evidence for the critical outcomes (Puhan et al., 2014; Brignardello-Petersen et al., 2018). By considering the risk of bias, inconsistency, indirectness, imprecision, publication bias, intransitivity, and incoherence, we graded the certainty of evidence into four level (Brignardello-Petersen et al., 2019; Guyatt et al., 2008): 1) very low, the estimate is subject to extreme uncertainty; 2) low, it is highly probable that further research will exert a considerable influence on our confidence in this assessment and is expected to revise it; 3) moderate, it is probable that further research will have an important impact on our confidence and might lead to a revised estimate; 4) high, it is improbable that further research will alter our confidence in the current estimate. Description of investigated preparations The interventions included in this review were CCPPs used in combination with ACEI/ARB-based biomedicine, including BL, HK, TG, JSB, HBHG, and SYKF. To improve reproducibility and transparency, we extracted all formulation-related information reported in the original trials, including dosage form, source of composition information, medicinal ingredients, single-herb or polyherbal, consistency of composition across included studies. We summarize this information in Supplementary Appendix 2. The interventions included in this review were marketed commercial Chinese medicinal products with fixed formulations, rather than individualized herbal prescriptions or study-specific mixtures. Under routine manufacturing and regulatory practice, a given product is intended to have a stable composition under the same product name. Therefore, interventions were grouped according to the reported commercial product name and treated as the same marketed preparations across studies. In this respect, these products are conceptually closer to standardized marketed drug products than to flexible investigator-defined herbal formulations. However, because this review was based on published randomized controlled trials, the level of formulation detail available for extraction depended on the reporting quality of the original articles. Although the products themselves are fixed commercial preparations, details such as manufacturer, batch number, and quality-control documentation were not consistently reported across all included studies. Role of the funding source The funder played no role in the design, conduct, analysis, or reporting of this study. All aspects of the research were performed independently by the authors. Results Search results and study characteristics A total of 1419 records were identified through the electronic search. After 690 titles and abstracts underwent screening, 103 articles underwent full-text review. This process culminated in the final inclusion of 70 articles, which represented 70 unique RCTs (Zeng, 2021; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Zhou, 2002; Zhang and Yang, 2016; Guan and Wu, 2005; Guo and Du, 2023; Wang and Lai, 2005; Zhao, 2013; Ma and Zhao, 2016; Li and Piao, 2017; Chen and Shou, 2015; Chen, 2019; Li et al., 2012; Han and Qiu, 2010; Su and Zhang, 2014; Tang, 2018; Zhang, 2021; Tang et al., 2009; Xu et al., 2009; Liang et al., 2017; Zhang and Xia, 2011; Zhang and Xia, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Guo et al., 2019; Lu, 2020; Yang, 2010; Peng et al., 2010; Zhou et al., 2012; Yin, 2023; Yang and Chen, 2010; Lin et al., 2014; Sun, 2015; Zhang et al., 2010; Yang et al., 2016; Yang et al., 2013; Shi et al., 2015; Zhang et al., 2013; He et al., 2016; Huang et al., 2012; Wei et al., 2019; Yu and Huang, 2012; Fan et al., 2013; Lu et al., 2016; Yu, 2016; Yang et al., 2014; Liang et al., 2019; Hu, 2021; Fang et al., 2017; Li and Huang, 2021; Feng, 2020; Cai, 2018; Guan et al., 2015; Wang and Liu, 2018; Wei, 2019; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Guo, 2017; Xiang et al., 2014; Cao and Chen, 2012; Wang, 2016; Xu, 2020; Shi et al., 2005) (Figure 1). Study selection. The basic information for each included RCTs was performed in Table 1. A total of 4855 patients were included, with a median age ranged from 21 to 70 years, the median duration time varied between 1 and 12 months, and almost of included RCTs reported the disease course of the patients spanning from 1 to 50 months. TABLE 1Basic characteristics of included studies in the systematic review.ComparationStudyNo. (I/C)Age (year)Disease course (month)Treatment duration (month)ICICBL + ACEI/ARB vs. ACEI/ARBZengcong 2021 (Zeng, 2021)25/2540.23 ± 4.1839.75 ± 4.22NRNR2Liulin 2018 (Liu et al., 2018)30/3035.78 ± 4.7535.75 ± 4.723.32 ± 1.453.32 ± 1.46NRLuoli 2023 (Luo et al., 2023)40/4047.84 ± 6.3548.95 ± 7.263.14 ± 0.523.08 ± 0.453Liping 2017 (Li, 2017)43/4340.58 ± 6.3915.46 ± 3.482Yuxuelei 2015 (Yu and Zhu, 2015)50/5037.9 ± 3.938.5 ± 4.29.9 ± 1.510.3 ± 1.62Huangqin 2011 (Huang and He, 2011)33/17NRNR3Huoyan 2016 (Huo and Wu, 2016)31/3536.5 ± 6.5NR4Zhouhongxing 2002 (Zhou, 2002)18/14NRNR3Zhangnan 2016 (Zhang and Yang, 2016)48/4836.59 ± 15.2235.31 ± 13.571.80 ± 1.401.87 ± 1.351HBHG + ACEI/ARB vs. ACEI/ARBGuanxiaodong 2005 (Guan and Wu, 2005)32/3031.2 ± 6.830.4 ± 6.213.2 ± 9.112.8 ± 8.73Guomin 2023 (Guo and Du, 2023)50/5042.06 ± 3.8441.16 ± 3.3372.12 ± 1.5268.28 ± 1.456Wangkunming 2005 (Wang and Lai, 2005)54/1829.8 ± 8.52.8 ± 1.42HK + ACEI/ARB vs. ACEI/ARBZhaojian 2013 (Zhao, 2013)41/4134 ± 933 ± 87 ± 38 ± 32Mafei 2016 (Ma and Zhao, 2016)40/4036.2 ± 12.413.6 ± 2.33Limingyu 2017 (Li and Piao, 2017)35/3535.0 ± 11.132.0 ± 12.114.2 ± 3.614.3 ± 3.73Chenjianjun 2015 (Chen and Shou, 2015)30/3039.2 ± 11.221.21 ± 10.6438.8 ± 10.621.47 ± 10.562Chenxuelian 2019 (Chen, 2019)40/4028.58 ± 6.3627.84 ± 5.3610.69 ± 2.169.56 ± 2.982Liliusheng 2012 (Li et al., 2012)30/3032.8 ± 12.714.4 ± 3.834.5 ± 11.714.5 ± 3.73Hanyarong 2010 (Han and Qiu, 2010)29/2637 ± 17.438 ± 16.3NR3Supofeng 2014 (Su and Zhang, 2014)24/2434.4 ± 7.934.2 ± 8.113.4 ± 5.413.1 ± 7.63Tangyi 2018 (Tang, 2018)34/3437.5 ± 3.637.2 ± 3.211.8 ± 1.611.2 ± 1.23Zhangchunzhao 2021 (Zhang, 2021)30/3046.22 ± 4.2646.23 ± 4.36NR3Tangweigang 2009 (Tang et al., 2009)70/5030.18 ± 16.2631.25 ± 15.363.26 ± 2.853.58 ± 2.682​Xuke 2009 (Xu et al., 2009)42/3836.4 ± 10.835.8 ± 12.4NR4Liangyan 2017 (Liang et al., 2017)26/2637.1 ± 11.738.7 ± 12.1NR6Zhangning 2011 (Zhang and Xia, 2011)32/3038 ± 1037 ± 11NR3Zhangning 2010 (Zhang and Xia, 2010)52/5044 ± 16.742 ± 17.8NR3Zhangchen 2014 (Zhang et al., 2014)30/30NRNR2Luhuiqin 2021 (Lu, 2021)78/7849.67 ± 6.1651.06 ± 5.3410.28 ± 2.8611.26 ± 3.026Zhangwei 2014 (Zhang, 2014)36/3635.9 ± 10.219.2 ± 0.92Guoyuqiang 2019 (Guo et al., 2019)30/3044.5 ± 6.145.2 ± 6.05.1 ± 3.35.3 ± 3.03Luwenjuan 2020 (Lu, 2020)25/2558.2 ± 9.157.5 ± 8.9NR3Yangjuhong 2010 (Yang, 2010)31/30NRNRNR6Pengtao 2010 (Peng et al., 2010)28/2539.9 ± 7.941.7 ± 8.7NR3Zhouhuilan 2012 (Zhou et al., 2012)30/30NRNR2Yinyong 2023 (Yin, 2023)45/4535.68 ± 7.1236.09 ± 6.8836.54 ± 8.3337.11 ± 7.953Yangliu 2010 (Yang and Chen, 2010)21/2126.8 ± 8.3NR3JSB + ACEI/ARB vs. ACEI/ARBLinlijuan 2014 (Lin et al., 2014)30/3035.2 ± 10.820.4 ± 10.221.0 ± 10.32Sunxuehui 2015 (Sun, 2015)53/5032 ± 733 ± 6NR6Zhangyongxiu 2010 (Zhang et al., 2010)53/4935.9 ± 8.737.1 ± 10.2NR3SYKF + ACEI/ARB vs. ACEI/ARBYangzhongmin 2016 (Yang et al., 2016)60/6042.8 ± 10.643.7 ± 11.2NR4Yangpo 2013 (Yang et al., 2013)22/2131.6 ± 10.464.8 ± 1.33Shifeng 2015 (Shi et al., 2015)20/2022.3​NR2Zhangxinzhi 2013 (Zhang et al., 2013)32/3038.6 ± 4.284.5 ± 20.52Helijuan 2016 (He et al., 2016)20/2026.7 ± 3.433.4 ± 6.25.2 ± 3.14.2 ± 3.12Huangjing 2012 (Huang et al., 2012)25/25NRNR3TG + ACEI/ARB vs. ACEI/ARBWeilan 2019 (Wei et al., 2019)51/4731.2 ± 4.332.4 ± 5.020.4 ± 0.218 ± 0.33Yuxufeng 2012 (Yu and Huang, 2012)20/2236.9 ± 9.837.3 ± 10.5NR3Fanye 2013 (Fan et al., 2013)20/2052.3 ± 12.1​NRNRLuxiaomei 2016 (Lu et al., 2016)55/5441.36 ± 11.7637.89 ± 11.14NR12Yutao 2016 (Yu, 2016)30/3046.1 ± 9.445.2 ± 5.7NR3Yangzhongmin 2014 (Yang et al., 2014)32/3150.01 ± 10.1249.3 ± 10.6NR3Liangyan 2019 (Liang et al., 2019)46/4045.5 ± 12.746.3 ± 11.438.7 ± 16.835.8 ± 15.33Hujiane 2021 (Hu, 2021)48/4849.40 ± 9.7248.74 ± 9.5625.41 ± 8.1225.61 ± 8.132Fanghui 2017 (Fang et al., 2017)40/4066.2 ± 4.165.2 ± 4.011.90 ± 0.2211.80 ± 0.3212Liyanfeng 2021 (Li and Huang, 2021)55/5541.47 ± 12.5141.52 ± 12.3327.6 ± 0.9526.2 ± 0.976Fengwei 2020 (Feng, 2020)45/4553.53 ± 9.52NR3Caiyuping 2018 (Cai, 2018)34/3446.12 ± 9.0545.78 ± 8.83NR6Guanyibiao 2015 (Guan et al., 2015)34/3137.25 ± 6.7336.72 ± 5.9223.02 ± 3.5622.18 ± 2.733Wangdan 2018 (Wang and Liu, 2018)33/3337.24 ± 6.7237.71 ± 5.91NRNRWeijiwei 2019 (Wei, 2019)35/3539.57 ± 5.1637.65 ± 5.581.35 ± 1.281.28 ± 1.233Wangyanhua 2020 (Wang, 2020)17/1742.61 ± 4.2243.75 ± 3.92NR6Chendu 2021 (Chen et al., 2021)40/4037.12 ± 2.1336.99 ± 2.1418.11 ± 1.5218.26 ± 1.556Shenshuijuan 2009 (Shen et al., 2009)26/2632.48 ± 10.12NR6Guobaoqin 2017 (Guo, 2017)15/1554.1 ± 8.953.7 ± 8.5NR3Xiangqiong 2014 (Xiang et al., 2014)30/3050.3 ± 9.651.3 ± 8.22.7 ± 1.92.6 ± 1.73Caoshan 2012 (Cao and Chen, 2012)12/11NRNR3Wangxiaoyan 2016 (Wang, 2016)41/4133.52 ± 5.3035.21 ± 4.46NR3Xulei 2020 (Xu, 2020)31/3149.07 ± 8.0448.57 ± 7.88NR2​Shiyongjun 2005 (Shi et al., 2005)19/1337.5NR12NR, not application; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BL, bailing capsule; HBHG, huobahuagen tablet; HK, huangkui capsule; TG, tripterygium glycosides; JSB, jinshuibao capsule; SYKF, shenyan kangfu tablet. Basic characteristics of included studies in the systematic review. NR, not application; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BL, bailing capsule; HBHG, huobahuagen tablet; HK, huangkui capsule; TG, tripterygium glycosides; JSB, jinshuibao capsule; SYKF, shenyan kangfu tablet. Risk of bias assessment The risk of bias for the included RCTs was provided in Supplementary Appendix 3. Sixteen studies (Han and Qiu, 2010; Su and Zhang, 2014; Liang et al., 2017; Zhang and Xia, 2011; Zhang, 2014; Guo et al., 2019; Yang, 2010; Zhou et al., 2012; Sun, 2015; Zhang et al., 2013; Huang et al., 2012; Yu and Huang, 2012; Liang et al., 2019; Shen et al., 2009; Cao and Chen, 2012; Shi et al., 2005) were classified as high risk of bias due to inadequate allocation concealment and lack of blinding. Nineteen studies (Zeng, 2021; Liu et al., 2018; Luo et al., 2023; Huang and He, 2011; Huo and Wu, 2016; Zhou, 2002; Zhang and Yang, 2016; Guan and Wu, 2005; Ma and Zhao, 2016; Li and Piao, 2017; Tang, 2018; Xu et al., 2009; Zhang and Xia, 2010; Zhang et al., 2014; Peng et al., 2010; Zhang et al., 2010; Yang et al., 2013; Shi et al., 2015; He et al., 2016) were rated as probably high risk of bias primarily due to deviations from intended intervention. Thirty-three studies (Li, 2017; Yu and Zhu, 2015; Guo and Du, 2023; Wang and Lai, 2005; Zhao, 2013; Chen and Shou, 2015; Chen, 2019; Li et al., 2012; Zhang, 2021; Tang et al., 2009; Lu, 2021; Lu, 2020; Yin, 2023; Yang and Chen, 2010; Lin et al., 2014; Yang et al., 2016; Wei et al., 2019; Fan et al., 2013; Yu, 2016; Yang et al., 2014; Hu, 2021; Fang et al., 2017; Feng, 2020; Cai, 2018; Guan et al., 2015; Wang and Liu, 2018; Wei, 2019; Wang, 2020; Chen et al., 2021; Guo, 2017; Xiang et al., 2014; Wang, 2016; Xu, 2020) were categorized as probably low risk of bias, while only two studies (Lu et al., 2016; Li and Huang, 2021) were rated as low risk. Overall, the quality of the included studies was not high. However, in terms of “Bias due to missing outcome data”, only two studies (Huang and He, 2011; Zhou, 2002) were assessed as probably high risk of bias, indicating virtually no data integrity issues identified in the included research. Network meta-analysis (NMA) Seventy studies were included in the NMA (Zeng, 2021; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Zhou, 2002; Zhang and Yang, 2016; Guan and Wu, 2005; Guo and Du, 2023; Wang and Lai, 2005; Zhao, 2013; Ma and Zhao, 2016; Li and Piao, 2017; Chen and Shou, 2015; Chen, 2019; Li et al., 2012; Han and Qiu, 2010; Su and Zhang, 2014; Tang, 2018; Zhang, 2021; Tang et al., 2009; Xu et al., 2009; Liang et al., 2017; Zhang and Xia, 2011; Zhang and Xia, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Guo et al., 2019; Lu, 2020; Yang, 2010; Peng et al., 2010; Zhou et al., 2012; Yin, 2023; Yang and Chen, 2010; Lin et al., 2014; Sun, 2015; Zhang et al., 2010; Yang et al., 2016; Yang et al., 2013; Shi et al., 2015; Zhang et al., 2013; He et al., 2016; Huang et al., 2012; Wei et al., 2019; Yu and Huang, 2012; ; Fan et al., 2013; Lu et al., 2016; Yu, 2016; Yang et al., 2014; Liang et al., 2019; Hu, 2021; Fang et al., 2017; Li and Huang, 2021; Feng, 2020; Cai, 2018; Guan et al., 2015; Wang and Liu, 2018; Wei, 2019; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Guo, 2017; Xiang et al., 2014; Cao and Chen, 2012; Wang, 2016; Xu, 2020; Shi et al., 2005). Network plots for the major outcomes are shown in Figure 2. Between-study heterogeneity varied across outcomes and comparisons and was mainly observed in some continuous outcomes (Supplementary Appendix 4), whereas no significant inconsistency was detected for any outcome (Supplementary Appendix 5). The GRADE summary of findings was performed in Supplementary Appendix 6. We judged the certainty of evidence to be low or very low for most of outcomes. Network plot for overall response rate (a), 24hUPQ (b), blood urea nitrogen (c), hemoglobin (d), adverse drug reactions (e) (Notes: Circle size corresponds to participant count; connecting lines and their width indicate direct comparisons and study volume, respectively). The NMA of overall response rate included 51 studies (Zeng, 2021; Liu et al., 2018; Luo et al., 2023; Yu and Zhu, 2015; Huang and He, 2011; Zhou, 2002; Guo and Du, 2023; Wang and Lai, 2005; Zhao, 2013; Ma and Zhao, 2016; Li and Piao, 2017; Chen, 2019; Li et al., 2012; Han and Qiu, 2010; Su and Zhang, 2014; Tang, 2018; Zhang, 2021; Tang et al., 2009; Xu et al., 2009; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Guo et al., 2019; Lu, 2020; Yang, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Guo et al., 2019; Lu, 2020; Yang, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Guo et al., 2019; Lu, 2020; Yang, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Guo et al., 2019; Lu, 2020; Yang, 2010; Shi et al., 2015; Zhang et al., 2013; He et al., 2016; Wei et al., 2019; Yu and Huang, 2012; Fan et al., 2013; Lu et al., 2016; Yu, 2016; Yang et al., 2014; Liang et al., 2019; Wei et al., 2019; Yu and Huang, 2012; Fan et al., 2013; Lu et al., 2016; Yu, 2016; Yang et al., 2014; Liang et al., 2019; Cai, 2018; Guan et al., 2015; Wang and Liu, 2018; Wei, 2019; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Guo, 2017; Xiang et al., 2014; Cao and Chen, 2012; Wang, 2016; Cai, 2018; Guan et al., 2015; Wang and Liu, 2018; Wei, 2019; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Guo, 2017; Xiang et al., 2014; Cao and Chen, 2012; Wang, 2016) of BL combined with AECI/ARB, HBHG combined with ACEI/ARB, HK combined with ACEI/ARB, JSB combined with ACEI/ARB, SYKF combined with ACEI/ARB or TG combined with ACEI/ARB, involving 3518 patients. The results showed that compared with ACEI/ARB alone, CCPPs combined with ACEI/ARB significantly improved the overall response rate, with the effect of BL combined with ACEI/ARB being the most significant (OR: 3.27, 95% CI: 2.21–4.84), and the certainty of evidence was moderate (detailed results can be seen in Supplementary Appendix 6.1). When compared with each other, the OR ranged from 0.72 to 1.26, and the certainty of evidence was very low (see Supplementary Appendix 6.1). According to their respective SUCRA values, it can be observed that regardless of whether compared with ACEI/ARB alone or with each other, the BL combined with ACEI/ARB regimen demonstrated the most significant therapeutic effect (see Supplementary Appendix 7.1, Supplementary Appendix 8). The NMA of 24hUPQ included 52 studies (Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Han and Qiu, 2010; Su and Zhang, 2014; Tang, 2018; Zhang, 2021; Tang et al., 2009; Xu et al., 2009; Liang et al., 2017; Zhang and Xia, 2011; Zhang and Xia, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Lu, 2020; Yang, 2010; Peng et al., 2010; Yin, 2023; Yang and Chen, 2010; Lin et al., 2014; Sun, 2015; Zhang et al., 2010; Yang et al., 2016; Yang et al., 2013; Shi et al., 2015; Zhang et al., 2013; He et al., 2016; Huang et al., 2012; Yin, 2023; Yang and Chen, 2010; Lin et al., 2014; Sun, 2015; Zhang et al., 2010; Yang et al., 2016; Yang et al., 2013; Shi et al., 2015; Zhang et al., 2013; He et al., 2016; Huang et al., 2012; Yu, 2016; Yang et al., 2014; Liang et al., 2019; Hu, 2021; Fang et al., 2017; Feng, 2020; Cai, 2018; Guan et al., 2015; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Wang, 2020; Chen et al., 2021; Shen et al., 2009) of BL combined with AECI/ARB, HBHG combined with ACEI/ARB, HK combined with ACEI/ARB, JSB combined with ACEI/ARB, SYKF combined with ACEI/ARB or TG combined with ACEI/ARB, involving 3629 patients. The results showed that compared with ACEI/ARB alone, CCPPs combined with ACEI/ARB significantly reduced the 24hUPQ, with the effect of BL combined with ACEI/ARB being the most significant (MD: −0.83 g, 95% CI: −1.21 to −0.44), and the certainty of evidence was moderate (detailed results can be seen in Supplementary Appendix 6.2). When compared with each other, the MD ranged from −0.17 g to 0.35 g, and the certainty of evidence was very low (see Supplementary Appendix 6.2). BL combined with ACEI/ARB regimen demonstrated the most significant therapeutic effect based on their respective SUCRA values (see Supplementary Appendix 7.2, Supplementary Appendix 8). The NMA of Scr included 40 studies (Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Guan and Wu, 2005; Guo and Du, 2023; Wang and Lai, 2005; Zhao, 2013; Li et al., 2012; Han and Qiu, 2010; Zhang, 2021; Tang et al., 2009; Liang et al., 2017; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Yang, 2010; Yang and Chen, 2010; Lin et al., 2014; Sun, 2015; Yang et al., 2016; Yang et al., 2013; Zhang et al., 2013; He et al., 2016; Huang et al., 2012; Yu and Huang, 2012; Liang et al., 2019; Hu, 2021; Fang et al., 2017; Li and Huang, 2021; Feng, 2020; Cai, 2018; Guan et al., 2015; Wei, 2019; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Xiang et al., 2014; Shi et al., 2005) of BL combined with AECI/ARB, HBHG combined with ACEI/ARB, HK combined with ACEI/ARB, JSB combined with ACEI/ARB, SYKF combined with ACEI/ARB or TG combined with ACEI/ARB, involving 2867 patients. The results showed that compared with ACEI/ARB alone, CCPPs combined with ACEI/ARB significantly reduced the Scr, with the effect of HBHG combined with ACEI/ARB being the most significant (MD: −26.63 μmol/L, 95%CI: −47.91 to −5.35), and the certainty of evidence was low (detailed results can be seen in Supplementary Appendix 6.3). When compared with each other, the MD ranged from −12.38 μmol/L to 18.60 μmol/L, and the certainty of evidence was very low (see Supplementary Appendix 6.3). However, it is worth noting that BL combined with ACEI/ARB regimen demonstrated the most significant therapeutic effect based on their respective SUCRA values (see Supplementary Appendix 7.3, Supplementary Appendix 8). The NMA of BUN included 20 studies (Liu et al., 2018; Luo et al., 2023; Huang and He, 2011; Guo and Du, 2023; Wang and Lai, 2005; Zhao, 2013; Li et al., 2012; Han and Qiu, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Yang et al., 2016; Huang et al., 2012; Liang et al., 2019; Hu, 2021; Fang et al., 2017; Li and Huang, 2021; Cai, 2018; Wang, 2020; Xiang et al., 2014) of BL combined with AECI/ARB, HBHG combined with ACEI/ARB, HK combined with ACEI/ARB, SYKF combined with ACEI/ARB or TG combined with ACEI/ARB, involving 1548 patients. The results showed that compared with ACEI/ARB alone, CCPPs combined with ACEI/ARB significantly reduced the BUN, with the effect of TG combined with ACEI/ARB being the most significant (MD: −2.53 mmol/L, 95% CI: −4.54 to −0.52), and the certainty of evidence was moderate (detailed results can be seen in Supplementary Appendix 6.4). When compared with each other, the MD ranged from −2.22 mmol/L to 1.46 mmol/L, and the certainty of evidence was very low (see Supplementary Appendix 6.4). Based on their respective SUCRA values, TG combined with ACEI/ARB regimen demonstrated the most significant therapeutic effect (see Supplementary Appendix 7.4, Supplementary Appendix 8). The NMA of Hb included 29 studies (Li, 2017; Yu and Zhu, 2015; Guan and Wu, 2005; Li et al., 2012; Han and Qiu, 2010; Su and Zhang, 2014; Tang, 2018; Zhang, 2021; Tang et al., 2009; Liang et al., 2017; Zhang and Xia, 2010; Zhang et al., 2014; Zhang, 2014; Yang, 2010; Yin, 2023; Lin et al., 2014; Sun, 2015; Zhang et al., 2013; He et al., 2016; Yu and Huang, 2012; Liang et al., 2019; Fang et al., 2017; Cai, 2018; Guan et al., 2015; Wei, 2019; Chen et al., 2021; Shen et al., 2009; Xiang et al., 2014; Shi et al., 2005) of BL combined with AECI/ARB, HBHG combined with ACEI/ARB, HK combined with ACEI/ARB, SYKF combined with ACEI/ARB or TG combined with ACEI/ARB, involving 1994 patients. The results showed that compared with ACEI/ARB alone, CCPPs combined with ACEI/ARB significantly improved the Hb, with the effect of BL combined with ACEI/ARB being the most significant (MD: 9.41 g/L, 95% CI: 8.18–10.64), and the certainty of evidence was low (detailed results can be seen in Supplementary Appendix 6.5). When compared with each other, the MD ranged from −6.96 g/L to 4.73 g/L, and the certainty of evidence was very low (see Supplementary Appendix 6.5). Based on their respective SUCRA values, BL combined with ACEI/ARB regimen demonstrated the most significant therapeutic effect (see Supplementary Appendix 7.5, Supplementary Appendix 8). The NMA of ADRs included 29 studies (Huang and He, 2011; Zhang and Yang, 2016; Guan and Wu, 2005; Ma and Zhao, 2016; Li and Piao, 2017; Chen and Shou, 2015; Chen, 2019; Han and Qiu, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Zhou et al., 2012; Yin, 2023; Lin et al., 2014; Wei et al., 2019; Yu and Huang, 2012; Lu et al., 2016; Yang et al., 2014; Liang et al., 2019; Fang et al., 2017; Li and Huang, 2021; Guan et al., 2015; Wang and Liu, 2018; Wei, 2019; Shen et al., 2009; Xiang et al., 2014; Cao and Chen, 2012; Xu, 2020; Shi et al., 2005) of BL combined with AECI/ARB, HBHG combined with ACEI/ARB, HK combined with ACEI/ARB, JSB combined with ACEI/ARB or TG combined with ACEI/ARB, involving 2059 patients. There were no convincing differences in any ADRs among these drugs. The certainty of evidence was ranged from very low to moderate (see Supplementary Appendix 6.6). Based on their respective SUCRA values, it can be observed that the incidence of ADRs varies little among the different treatment measures, with the HK group showing a relatively higher incidence of ADRs (see Supplementary Appendix 7.6, 8). The most commonly reported ADRs across all groups were mild gastrointestinal discomfort (nausea, abdominal distension, diarrhea), transient dizziness, and mild elevation of liver transaminases. All reported ADRs were mild to moderate in severity and resolved spontaneously without discontinuation of treatment. No fatal or life-threatening serious adverse events were reported in any of the included trials (see Supplementary Appendix 9). Sensitivity analysis and publication bias A sensitivity analysis was conducted to evaluate the robustness of the NMA results. This was achieved by iteratively removing each of the 70 included studies (Zeng, 2021; Liu et al., 2018; Luo et al., 2023; Li, 2017; Yu and Zhu, 2015; Huang and He, 2011; Huo and Wu, 2016; Zhou, 2002; Zhang and Yang, 2016; Guan and Wu, 2005; Guo and Du, 2023; Wang and Lai, 2005; Zhao, 2013; Ma and Zhao, 2016; Li and Piao, 2017; Chen and Shou, 2015; Chen, 2019; Li et al., 2012; Han and Qiu, 2010; Su and Zhang, 2014; Tang, 2018; Zhang, 2021; Tang et al., 2009; Xu et al., 2009; Liang et al., 2017; Zhang and Xia, 2011; Zhang and Xia, 2010; Zhang et al., 2014; Lu, 2021; Zhang, 2014; Guo et al., 2019; Lu, 2020; Yang, 2010; Peng et al., 2010; Zhou et al., 2012; Yin, 2023; Yang and Chen, 2010; Lin et al., 2014; Sun, 2015; Zhang et al., 2010; Yang et al., 2016; Yang et al., 2013; Shi et al., 2015; Zhang et al., 2013; He et al., 2016; Huang et al., 2012; Wei et al., 2019; Yu and Huang, 2012; ; Fan et al., 2013; Lu et al., 2016; Yu, 2016; Yang et al., 2014; Liang et al., 2019; Hu, 2021; Fang et al., 2017; Li and Huang, 2021; Feng, 2020; Cai, 2018; Guan et al., 2015; Wang and Liu, 2018; Wei, 2019; Wang, 2020; Chen et al., 2021; Shen et al., 2009; Guo, 2017; Xiang et al., 2014; Cao and Chen, 2012; Wang, 2016; Xu, 2020; Shi et al., 2005). The stability of the overall results is evidenced by the fact that omitting any individual study did not significantly alter the drug’s therapeutic efficacy, further affirming the consistency of the findings (see Supplementary Appendix 10). Additionally, Egger’s test was performed to assess publication bias for the following outcomes: overall response rate, 24hUPQ, Scr, BUN, Hb, and ADRs. The results indicated significant publication bias for overall response rate and 24hUPQ (Egger’s test p < 0.05). For all other outcomes, the Egger’s test results were non-significant (p > 0.05) (see Table 2). TABLE 2The results of Egger’s test.OutcomesOverall response rate24hUPQScrBUNHbADRsP0.0010.010.060.140.890.2324hUPQ, 24-h urinary protein quantification; Scr, serum creatinine; BUN, blood urea nitrogen; Hb, hemoglobin; ADRs, adverse drug reactions. The results of Egger’s test. 24hUPQ, 24-h urinary protein quantification; Scr, serum creatinine; BUN, blood urea nitrogen; Hb, hemoglobin; ADRs, adverse drug reactions. Discussion This NMA systematically compared the efficacy and safety of six CCPPs combined with ACEI/ARB in 4,855 IgAN patients. The BL combined with ACEI/ARB regimen demonstrated the highest probability of being the best for improving the overall response rate, reducing 24hUPQ, and increasing Hb level, with SUCRA values consistently high. In contrast, the TG combined with ACEI/ARB regimen showed a unique advantage in reducing BUN level. It is noteworthy that none of the combination regimens showed clinically significant differences in safety profiles, although the quality of evidence was generally low or very low. To some extent, our findings corroborate and extend the existing body of evidene on integrated CCPPs-B for IgAN. The study confirms this trend utilizing NMA and further quantifies the relative efficacy among different CCPPs, thereby providing more nuanced ranking information for clinical decision-making. However, similar to the findings of Gao et al. (2024) in the context of anti-influenza viral drugs, the core conclusions of our study are also plagued by very low quality of evidence. Although studies have shown that combination therapy of CCPPs-B may benefit kidney protection (Chen et al., 2022; Li et al., 2025), the included studies did not undergo long-term follow-up, resulting in limited understanding of the true effects of these combination regimens on hard endpoints such as delaying estimated glomerular filtration rate (eGFR) decline and reducing the incidence of end-stage renal disease. Compared with previous reviews, a major strength of this study is that we evaluated multiple CCPPs combined with ACEI/ARB within a single network meta-analysis, rather than focusing on one specific intervention or relying only on pairwise comparisons. This allowed a more comprehensive comparison of the relative efficacy and safety of different treatment options. In addition, some previous pharmacological and experimental studies provide mechanistic support for our findings. TG have been reported to exert anti-inflammatory and immunomodulatory effects and to alleviate renal injury (Chen et al., 2022). HK capsule, derived from Abelmoschus manihot, may have anti-inflammatory, anti-oxidative, and anti-fibrotic properties and has shown renoprotective potential in kidney disease (Pei and Li, 2021; Yuan et al., 2023). BL capsule and JSB capsule, both related to Cordyceps sinensis, have also been reported to improve renal function and regulate immune and inflammatory responses (Wang et al., 2024; Fan et al., 2025). Given the important roles of inflammation, immune dysregulation, and fibrosis in IgA nephropathy, these findings provide a plausible biological basis for the observed clinical benefits. However, these mechanistic data are supportive rather than confirmatory, and the conclusions of this study remain primarily based on comparative clinical evidence. This study had three key methodological strengths. First, the application of a NMA transcended the limitations of conventional pairwise comparisons. By constructing an evidence network containing six intervention measures, we quantified and ranked the relative efficacy of different integrated CCPPs-B regimens, thereby providing more refined evidence for clinical decision-making. Second, the study rigorously adhered to the PRISMA guidelines (Hutton et al., 2015), employing the Cochrane Risk of Bias tool (Sterne et al., 2019) to assess the included studies and the GRADE framework to evaluate the certainty of evidence (Puhan et al., 2014; Brignardello-Petersen et al., 2018). Finally, the combined presentation of SUCRA values with traditional effect sizes not only illustrates the comprehensive ranking probabilities of each regimen but also provides the magnitude of clinical benefit, thereby enhancing the practical applicability of the findings. However, the limitations of this study should not be overlooked. First and foremost, the methodological quality of the included studies was generally low. Among the 70 included studies, only two were rated as having a low risk of bias. The vast majority exhibited a high or unclear risk of bias in key domains such as random sequence generation, allocation concealment, and blinding, which likely led to an overestimation of the treatment effect. Second, we observed pronounced statistical heterogeneity. For key continuous outcomes, such as 24hUPQ and Scr, high I2 values were consistently observed. This heterogeneity likely stems from intrinsic variations in the production processes, chemical composition, and dosage of the CCPPs themselves, as well as differences in the etiology, disease stages, and concomitant treatments among the enrolled patients. Although no significant statistical inconsistency was detected, the clinical and methodological heterogeneity nevertheless undermines the robustness of the pooled results. Third, Egger’s test suggested possible publication bias for two outcomes, both of which were important endpoints favoring BL. This raises the possibility that the observed benefit of BL may have been exaggerated by small-study effects or selective publication. Therefore, the superiority of BL should be interpreted with caution. Furthermore, the results of Egger’s test for the overall response rate suggest the possible presence of publication bias, indicating that positive results might have been more likely to be published. Finally, there were important limitations regarding exploration of heterogeneity. Although sensitivity analyses showed that the main findings were robust to the exclusion of individual studies, we did not perform formal subgroup analyses based on factors such as risk of bias, treatment duration, or sample size. This decision was made for methodological reasons. Specifically, the number of studies available for several treatment comparisons and outcomes was limited, and further stratification would have resulted in sparse subnetworks, unstable estimates, and reduced interpretability. In addition, key stratification variables were incompletely and inconsistently reported across the included trials, and simple study-level subgrouping would have been unlikely to adequately explain the observed heterogeneity. Therefore, we considered it more appropriate to address these issues through cautious interpretation, sensitivity analyses, and certainty rating using the GRADE framework. Nevertheless, the lack of formal subgroup analyses limits our ability to determine whether specific patient subgroups may benefit more from particular interventions. In clinical practice, given the current low-certainty evidence, any therapeutic decision must be approached with caution. For IgAN patients considering integrated CCPPs-B therapy, the BL combined with ACEI/ARB regimen demonstrates potential in reducing proteinuria and improving Hb, yet the level of evidence remains low. Clinicians should recognize that the existing evidence falls far short of establishing any combination therapy as a standard treatment plan. This NMA underscores the imperative for future large-scale, high-quality, long-term RCTs. Future investigations should prioritize: 1) enhance methodological rigor through strict implementation of randomization, allocation concealment, and blinding, particularly for outcome assessors; 2) focus on patient-important long-term hard endpoints, such as renal failure, cardiovascular events and all-cause mortality; 3) Standardize the intervention measures and clarify the standardized ingredients, dosage and course of treatment of CCPPs and simple preparations to reduce heterogeneity; 4) conduct pre-specified subgroup analyses to explore treatment effect variations across different clinical characteristics and traditional Chinese medicine syndrome types, thereby facilitating precision medicine. Conclusion Based on current evidence from randomized trials, in patients with IgAN, BL + ACEI/ARB appeared to show favorable effects in some key outcomes, however, given the potential publication bias detected for outcomes supporting BL, this apparent advantage should be interpreted cautiously and requires confirmation in future high-quality studies. While all combined regimens may offer advantages over ACEI/ARB alone in terms of overall response rate and laboratory parameters, the certainty of evidence for these findings is very low, resulting in substantial uncertainty. These findings primarily highlight the weakness and limitations of the current evidence base rather than providing clear recommendations for any specific regimen. There is a pressing need for more rigorous, high-quality RCTs with longer follow-up periods to elucidate the true efficacy and place of these combination regimens in IgAN management.