Real‐World Data as External Control in Assessing the Efficacy of Allogeneic Adipose‐Derived Stem Cells Therapy for Advanced Chronic Kidney Disease

ABSTRACT

Chronic kidney disease is a progressive condition with limited therapeutic options in its advanced stages. Adipose‐derived stem cell therapy has shown potential in preclinical studies for renal repair. This study evaluated the short‐term stability of renal function in patients with moderate to severe chronic kidney disease who received adipose‐derived stem cell therapy, using a matched control group derived from real‐world clinical data for comparison. A total of 34 treated patients were matched in a one‐to‐five ratio with 170 control patients based on key clinical characteristics. The primary outcomes included the mean percentage change in estimated glomerular filtration rate and the incidence of renal function decline exceeding defined thresholds. To enhance the robustness of treatment effect estimation, real‐world data were utilized to construct an external control group that closely resembled the clinical trial population. This approach allowed indirect treatment comparisons and strengthened the internal validity of findings in the absence of randomization. Results demonstrated that the treated group exhibited a more stable renal function trajectory and a significantly lower risk of deterioration compared to the control group, particularly in patients with more advanced disease. Among dose groups, the low‐dose group showed the greatest stability in renal function. These findings support the feasibility of using real‐world data to construct external comparators and suggest that stem cell therapy may offer a short‐term stabilizing effect on renal function. Further research is needed to validate these findings and explore their long‐term clinical implications.

Trial Registration: ClinicalTrials.gov identifier: NCT02933827 (registered October 13, 2016. https://clinicaltrials.gov/study/NCT02933827)

Study Highlights.

• What is the current knowledge on the topic?

  • ○
    Chronic kidney disease is a progressive condition with limited treatment options at advanced stages. Mesenchymal stem cells (MSCs), particularly adipose‐derived stem cells (ADSCs), have shown promise in preclinical studies for renal repair, but clinical trials are often single‐arm and lack comparator groups.
• What question did this study address?

  • ○
    This study investigated whether ADSC therapy is associated with short‐term renal function stability in CKD patients by comparing trial participants with a matched real‐world control group.
• What does this study add to our knowledge?

  • ○
    By integrating real‐world data as an external control, this study found that ADSC therapy was associated with a slower decline in eGFR and a lower risk of deterioration, particularly among patients with CKD Stage 4.
• How might this change clinical pharmacology or translational science?

  • ○
    These findings support the use of RWD in evaluating investigational therapies and suggest that ADSC‐based interventions may offer a novel option for stabilizing renal function in advanced CKD.

1. Introduction

Chronic kidney disease (CKD) is a common disease with a high incidence, affecting approximately 10% of the worldwide population [1]. CKD is typically defined as a chronic and irreversible condition. Impaired kidney function is challenging to restore and progressively leads to complications such as cardiovascular diseases, hypertension, and cognitive impairment [2, 3, 4]. CKD is defined as abnormalities of kidney structure or function for over 3 months. After kidney injury, persistent inflammation can cause tubulointerstitial fibrosis and glomerular loss, leading to diminished kidney function. Because tubulointerstitial fibrosis is irreversible, early intervention is crucial to reduce inflammation and promote tubular cell regeneration, thereby preventing the progression of chronic kidney disease [5]. Despite recent advances in preventing CKD progression with RAAS inhibitors, SGLT2 inhibitors, and mineralocorticoid receptor antagonists, rapid kidney function decline in advanced CKD stages remains inevitable [6, 7, 8]. Thus, there is a critical need for innovative treatment modalities. Mesenchymal stem cells (MSCs), which have garnered significant attention for their ability to promote kidney repair and regeneration, have emerged as a potential treatment to prevent CKD progression [9, 10].

Recent studies have demonstrated that mesenchymal stem cells secrete cytokines, chemokines, and growth factors via paracrine mechanisms to prevent apoptosis in neighboring cells and to promote cell proliferation, angiogenesis, and tissue regeneration [11, 12]. Investigations using mouse models of kidney injury have shown that MSC‐derived extracellular vesicles can suppress pro‐inflammatory factors while enhancing anti‐inflammatory responses, thereby exerting immunoregulatory effects, inhibiting tubular epithelial cell (TEC) pyroptosis and inflammation, reducing renal fibrosis, and facilitating cellular repair [13, 14, 15]. Among various MSC sources, adipose tissue‐derived stem cells (ADSCs) have been found to exhibit superior anti‐inflammatory and immunomodulatory properties compared to bone marrow‐derived MSCs (BMSCs) [16].

Ethical and financial limitations have led many clinical trials of MSC‐based therapies for CKD to be conducted without a control group, which hinders robust assessment of therapeutic efficacy and external validity [17, 18, 19, 20]. The application of real‐world data (RWD) has proven to be an effective strategy to complement these single‐arm trials, by establishing external control groups that enhance comparability and strengthen the reliability of study findings. By integrating RWD with clinical trial data, researchers can mitigate the absence of randomization and provide a more rigorous framework for evaluating treatment effects [21, 22, 23].

By employing target trial emulation methodologies, RWD can facilitate the construction of an RWD external control group that closely mirrors the inclusion criteria and baseline characteristics of clinical trial participants. This approach is particularly valuable for MSC‐based therapies, as it minimizes design biases, enhances comparability, and provides a robust framework for evaluating treatment efficacy in real‐world settings [24, 25, 26, 27]. Additionally, in the context of MSC‐based therapies for CKD, RWD external controls provide critical insights into treatment effects, and safety profiles across diverse populations. This integration strengthens the evidence base of single‐arm trials by offering external control comparisons, thereby enhancing clinical decision‐making and supporting regulatory approvals [22].

The Phase I and II clinical trials of allogeneic adipose tissue‐derived stem cells were conducted as single‐arm studies, making it essential to establish an RWD external control group to provide a comparative reference for evaluating treatment outcomes. This study aims to establish a comparable RWD external control group to evaluate the efficacy of adipose‐derived stem cell therapy (ADSC therapy) in slowing CKD progression. The primary objective of this study is to construct a comparable RWD external control group using the Taipei Medical University Clinical Research Database (TMUCRD), which integrates electronic health records from multiple affiliated hospitals and provides longitudinal information on patients with CKD, to evaluate the efficacy of ADSC therapy in slowing CKD progression.

2. Materials and Methods

2.1. Data Resource

2.1.1. Phase I/II Clinical Trial Data of ADSC Therapy

The clinical trial data of ELIXCYTE, an allogeneic adipose‐derived stem cell (ADSC) therapy, is derived from a multicenter, open‐label Phase I and II clinical trial. This trial was conducted at three medical centers including Taipei Medical University Shuang Ho Hospital, LinKou Chang Gung Memorial Hospital, and Taichung Veterans General Hospital. A total of 12 subjects participated in Phase I, and 27 subjects participated in Phase II. Among these participants, 12 received a low dose (ADSCs 6.4 × 10⁷ cells) of ADSC therapy, 12 received a moderate dose (ADSCs 19.2 × 10⁷ cells), and 15 received a high dose (ADSCs 32.0 × 10⁷ cells).

2.1.2. Taipei Medical University Clinical Research Database (TMUCRD)

Taipei Medical University Clinical Research Database (TMUCRD) serves as a repository for real‐world data on patients receiving conventional treatments, aggregating patient conditions and medical test data from approximately 24,687 patients with CKD treated at Taipei Medical University Shuang Ho Hospital and Taipei Medical University Hospital from 2018 to 2022. The identities of patients in TMUCRD are encrypted to protect patient confidentiality. The database provides long‐term follow‐up data including electronic medical records, physiological parameters, and biochemical test results for patients with CKD.

The ELIXCYTE Phase I/II clinical trial was approved by the institutional review boards of Taipei Medical University Shuang Ho Hospital (N201710032), Linkou Chang Gung Memorial Hospital (201900020A0), and Taichung Veterans General Hospital (SF20136B). Written informed consent was obtained from all participants, and the trial was conducted in accordance with the Declaration of Helsinki. The study was registered on ClinicalTrials.gov (NCT02933827). The real‐world data external control was derived from the Taipei Medical University Clinical Research Database (TMUCRD), and its use was separately approved by the Taipei Medical University Joint IRB (N202310001). As all data were de‐identified, individual consent was not required.

2.2. Target Population

Participants in the Phase I/II clinical trial of ADSC therapy were designated as the case group, while patients recorded in TMUCRD served as the RWD external control group. Eligible patients in both groups were aged 20–80 years with CKD Stages 3b to 4, and an estimated glomerular filtration rate (eGFR) of 15–44 mL/min/1.73 m², determined using the Modification of Diet in Kidney Disease (MDRD) formula.

Patients in the RWD external control were selected according to the exclusion criteria defined in the Phase I/II clinical trial of ADSC therapy (see Supporting Information, Section S2), ensuring that the characteristics of the control group were aligned with those of the case group and thereby minimizing potential confounding effects on the study outcomes. Patients were excluded from the study if they had hypersensitivity to any component used in the study, inadequate hematologic or hepatic function, uncontrolled diabetes mellitus with glycated hemoglobin A1c (HbA1c) > 8.0%, human immunodeficiency virus infection, any type of hepatitis, autoimmune diseases, or cystic kidney disease, or if they required any form of dialysis.

Patients in both the case and control groups were prohibited from using Amphotericin B, vancomycin, and amikacin. Those who inadvertently used these medications during the study were withdrawn from participation. Concurrently, patients could maintain their prescribed medications and treatments for CKD throughout the trial. The regimen for these medications needed to remain stable throughout the trial period. This study collected eGFR data from the RWD control group corresponding to the visit schedules of the case group to ensure comparability between the two groups. Detailed visit schedules and frequencies are provided in Supporting Information Table S1.

2.3. Study Design

Figure 1 illustrates the research procedure of this study. First, we utilized the Phase I/II clinical trial data of ADSC therapy and the TMUCRD to construct the target populations for the case and control groups, comprising 39 and 3558 individuals, respectively. Subsequently, the RWD external control group underwent patient screening based on the exclusion criteria, resulting in 2384 eligible patients. Meanwhile, the number of eligible patients in the case group was 34. After propensity score matching (PSM) at a 1:5 case‐to‐control ratio, a total of 34 cases and 170 RWD controls were selected. Efficacy analyses were first conducted in the overall matched cohort, and then further stratified by CKD Stage (3B and 4) as well as by dose subgroups (low, moderate, and high), as illustrated in Figure 1.

FIGURE 1.

Flow chart of study.

2.4. Efficacy Outcome Measures

The primary objective of this study was to evaluate the short‐term stability of renal function following treatment. To assess and compare renal function dynamics between the ADSC therapy and external control groups, two principal measures were employed.

The first measure was the mean percentage change in eGFR from baseline across scheduled follow‐up time points. This longitudinal metric reflects temporal variation in renal function and serves as an indicator of treatment‐associated stability or fluctuation.

The second measure was the incidence of a decline in eGFR of 30% or more from baseline. A reduction exceeding 30% at any time point was considered a significant marker of renal function deterioration. This threshold was widely recognized in clinical trials and clinical practice guidelines as a standard measure of CKD progression [28, 29]. To further characterize renal function decline, the study also examined the incidence of eGFR reductions of 15%, 20%, 25%, and 35% as secondary indicators.

2.5. Statistical Methods

This study employed PSM to improve baseline comparability between the ADSC therapy and the RWD EC groups. The selection of covariates was guided by clinical practice guidelines [29] and prior literature [30] identifying well‐established prognostic factors for CKD progression. Propensity scores were estimated using a logistic regression model that incorporated demographic variables (sex, age), baseline renal function (eGFR), major comorbidities (hypertension, hyperlipidemia, glomerulonephritis, diabetes mellitus, and coronary heart disease), and concomitant medications (proteinuria‐reducing agents, renin–angiotensin–aldosterone system [RAAS] inhibitors, lipid‐lowering agents, and urate‐lowering agents). A greedy nearest‐neighbor algorithm without replacement was applied, using the logit of the propensity score as the distance metric. A caliper width of 0.5 standard deviations was adopted to retain all treated cases, maintain the intended matching ratio, and ensure covariate balance. PSM was performed for both the overall cohort and subgroup analyses stratified by CKD Stage (IIIB and IV) and ADSC therapy dose groups (low, moderate, high). A 1:3 ratio was applied for the CKD stage IIIB subgroup to avoid case loss, while all other analyses used a 1:5 ratio.

Balance was assessed using the standardized mean difference (SMD). An absolute SMD < 0.1 was considered stringent [31], and 0.1–0.2 acceptable [32]; in this study, an absolute SMD < 0.2 was regarded as a broadly acceptable criterion for adequate balance [32, 33, 34, 35]. As shown in Table 1, most covariates achieved SMDs < 0.1, indicating satisfactory comparability between groups. Further details are provided in Supporting Information Tables S2 and S3.

TABLE 1.

Baseline characteristics of subjects in the ADSC therapy and RWD EC groups before and after propensity score matching.

	Subjects prior to PS matching			Subjects following PS matching
	ADSC therapy group (N = 34)	RWD EC group (N = 2384)	SMD a	ADSC therapy group (N = 34)	RWD EC group (N = 170)	SMD a
Sex
Female	13 (38.24%)	999 (41.90%)	−0.2660	13 (38.24%)	63 (37.06%)	0.0257
Male	21 (61.76%)	1385 (58.10%)		21 (61.76%)	107 (62.94%)
Age (years)
Mean ± SD	50.82 ± 12.69	65.96 ± 10.94	−1.2781	50.82 ± 12.69	53.38 ± 14.57	−0.1869
CKD Stage
IIIB	11 (32.35%)	1332 (55.87%)	−0.2660	11 (32.35%)	55 (32.35%)	< 0.0001
IV	23 (67.65%)	1052 (44.13%)		23 (67.65%)	115 (67.65%)
Baseline eGFR (mL/min/1.73 m2)
Mean ± SD	28.36 ± 7.81	30.68 ± 9.51	−0.2660	28.36 ± 7.81	28.14 ± 9.28	0.0257
Comorbidities
Hypertension	26 (76.47%)	1914 (80.29%)	−0.0928	26 (76.47%)	138 (81.18%)	−0.1154
Hyperlipidemia	23 (67.65%)	1319 (55.33%)	0.2552	23 (67.65%)	112 (65.88%)	0.0375
Glomerulonephritis	16 (47.06%)	2384 (19.76%)	0.6047	16 (47.06%)	74 (46.53%)	0.0709
Diabetes mellitus	5 (14.71%)	1583 (66.40%)	−1.2383	5 (14.71%)	22 (12.94%)	0.0511
Coronary heart disease	4 (11.76%)	835 (35.03)	−0.5714	4 (11.76%)	30 (17.65%)	−0.1667
Medicaments b
Proteinuria‐reducing medication	28 (82.35%)	1669 (70.01%)	0.2929	28 (82.35%)	134 (78.82%)	0.0893
RAAS inhibitors	30 (88.24%)	1798 (75.42%)	0.3370	30 (88.24%)	145 (85.29%)	0.0869
Anti‐Hyperlipidemic agents	18 (52.94%)	1522 (63.84%)	−0.2225	18 (52.94%)	90 (52.94%)	< 0.0001
Anti‐hyperuricemic agents	27 (79.41%)	1278 (53.61%)	0.5684	27 (79.41%)	128 (75.29%)	0.0985

^a The standardized mean difference (SMD) was used to assess the balance of baseline characteristics between the ADSC therapy and RWD EC groups. An absolute SMD < 0.1 was considered evidence of negligible imbalance, while values between 0.1 and 0.2 indicated minor differences that were still consistent with acceptable group comparability.

^b The category of proteinuria‐reducing medications included agents such as dipyridamole and pentoxifylline. RAAS inhibitors referred to angiotensin‐converting enzyme inhibitors (ACEIs) and angiotensin II receptor blockers (ARBs). Anti‐hyperlipidemic agents primarily consisted of statins, while anti‐hyperuricemic agents encompassed drugs that either inhibited uric acid production or promoted uric acid excretion.

Statistical analyses were conducted to assess short‐term changes in renal function between the two groups. Linear mixed‐effects models were applied to compare the mean percentage change in eGFR over time, accounting for the correlation introduced by matching and providing unbiased between‐group estimates.

The risk of a decline in eGFR of 30% or more from baseline in the ADSC therapy group relative to the RWD EC group, along with risks associated with other secondary thresholds, was estimated using the Cox proportional hazards model with a robust sandwich variance estimator, adjusting for sex, age, comorbidities (hypertension, hyperlipidemia, glomerulonephritis, diabetes mellitus, coronary heart disease), and concomitant medications (proteinuria‐reducing agents, RAAS inhibitors, lipid‐lowering agents, urate‐lowering agents). The decline in eGFR was treated as a time‐to‐event outcome, defined as the first occurrence of a specified percentage decrease in eGFR during follow‐up, with time calculated from baseline to the event or censoring at the last available visit.

The cumulative incidence and progression rate of renal function deterioration was further evaluated using the Kaplan–Meier method with a stratified log‐rank test to account for the matched design. All statistical analyses were performed using SAS version 9.4.

3. Results

3.1. Baseline Characteristics of Subjects

This study applied PSM to adjust for baseline differences between the ADSC therapy group and the RWD EC group, as detailed in Table 1. Before matching, significant differences were observed in comorbidities and medication use; however, after PSM, the baseline characteristics were balanced, with no statistically significant differences, ensuring comparability between groups. In the matched cohort, 62% were male, with an average age of 51–53 years. Stage IV CKD was more prevalent (67.65%), and most participants had hypertension, hyperlipidemia, and glomerulonephritis, with over 85% receiving RAAS inhibitors for CKD treatment. Overall, PSM effectively balanced baseline characteristics, enhancing the validity of subsequent efficacy analyses.

3.2. Comparative Trends in Short‐Term Renal Function Stability Between ADSC Therapy and RWD EC Groups

The ADSC therapy group demonstrated a more stable eGFR trajectory with a slower decline over time compared to the RWD EC group, as shown in Figure 2 Part A. Stratification by CKD stage revealed that in CKD Stage 3, eGFR remained stable in both groups, whereas in CKD Stage 4, the RWD EC group experienced a significantly greater decline. Additionally, the ADSC therapy group showed higher variability in eGFR, as indicated by a larger standard deviation. Figure 2 Part B presents the percentage change in eGFR from baseline, showing that the RWD EC group deviated more significantly from 0% change, reflecting greater fluctuations in renal function over time, while the ADSC therapy group remained relatively stable. In CKD Stage 3, the percentage change in eGFR did not significantly differ between groups, indicating general stability. However, in CKD Stage 4, the RWD EC group experienced a statistically significant decline compared to the ADSC therapy group, suggesting a more stable short‐term renal function trajectory in the ADSC therapy group among patients with advanced CKD.

FIGURE 2.

Comparison of eGFR trends and percentage changes over time between ADSC therapy and real‐world data external control groups across CKD stages. A1, B1, and C1 illustrate the trend lines representing the mean eGFR at each time point. A2, B2, and C2 display stem‐and‐leaf plots depicting the distribution of eGFR values at each time point. A3, B3, and C3 present the trend of the mean change in eGFR compared to the baseline, calculated as (eGFR at each specific time point—baseline eGFR)/baseline eGFR. A4, B4, and C4 depict the distribution of eGFR changes at each time point. The mean values or mean percentage change in eGFR for each group are displayed below the corresponding time points. The p‐values were calculated using linear mixed‐effects models to compare eGFR between the two groups at each time point. ADSC‐TH, ADSC therapy; RWD‐EC, real‐world data external control.

3.3. Comparative Risk and Rate of Renal Function Decline to Assess Short‐Term Stability

The risk of renal function decline was assessed in both the ADSC therapy and RWD EC groups across various thresholds of eGFR reduction, stratified by the overall population, CKD Stage 3B, and CKD Stage 4 (Figure 3). Renal function decline events were defined as the first occurrence of a specified percentage decrease in eGFR during the study period, with follow‐up calculated from baseline to the initial event. In the overall population, the ADSC therapy group showed a consistently lower risk of eGFR decline across all thresholds (adjusted HR: 0.276–0.478), with statistical significance observed for most outcomes (p < 0.05), except for the ≥ 25% decline threshold, where the difference was not significant. No significant differences were observed between the groups in CKD Stage 3B (p > 0.05), suggesting comparable short‐term risk in this subgroup. In CKD Stage 4, patients in the ADSC therapy group exhibited a significantly lower risk of eGFR decline at the ≥ 15% threshold (adjusted HR: 0.442, p < 0.05) and the ≥ 30% threshold (adjusted HR: 0.276, p < 0.05). Although results at the other thresholds did not reach statistical significance, the hazard ratios consistently favored the ADSC therapy group, indicating a trend toward reduced risk of renal function decline.

FIGURE 3.

Comparison of renal function decline risk across different thresholds of eGFR reduction between ADSC therapy and RWD control groups in overall and CKD Stage 3B and 4 patients. Renal function decline event is defined as the occurrence of a specified percentage decline in renal function for the “first time” during the study period. The follow‐up duration is calculated from the baseline to the “first occurrence” of the event. If no event occurs, the follow‐up duration is determined by the length of the study period. Adj. HR (adjusted hazard ratio) quantifies the risk of renal function decline in patients treated with ELIXCYTE compared to those in the RWD external control group. It is estimated using a Cox proportional hazards model, which incorporates key covariates, including age, sex, comorbidities, and medication use, to adjust for potential confounding factors.

The cumulative risk of renal function decline was estimated using the Kaplan–Meier method to evaluate group differences. Decline events were defined as the first occurrence of a percentage decrease in eGFR from baseline. At the 30% decline threshold, the RWD EC group exhibited a significantly faster risk accumulation in both the overall population (Figure 4 A4, p = 0.0153) and CKD Stage 4 patients (Figure 4 C4, p = 0.0317). Similar trends were observed at the 15% and 20%, thresholds (all p < 0.05) in the overall population, while differences at the 25% and 35% thresholds were not statistically significant. In CKD Stage 3B (Figure 4 part B), cumulative risk accumulation did not significantly differ between the groups at any threshold. In contrast, in CKD Stage 4 (Figure 4 part C), the RWD EC group showed a consistently faster accumulation of risk across multiple thresholds (p < 0.05 for 15%, 20%, 25%, and 30% declines). These results suggest that the ADSC therapy group experienced more gradual changes in renal function over time, reflecting greater short‐term renal stability in patients with more advanced CKD.

FIGURE 4.

Cumulative risk analysis of renal function decline comparing ADSC therapy and RWD control groups across CKD stages. Renal function decline event is defined as the occurrence of a specified percentage decline in renal function for the “first time” during the study period. The follow‐up duration is calculated from the baseline to the “first occurrence” of the event. If no event occurs, the follow‐up duration is determined by the length of the study period. ADSC‐TH, ADSC therapy; RWD‐EC, real‐world data external control.

3.4. Short‐Term Renal Function Trends Across ADSC Therapy Dose Groups Compared to the RWD EC Group

Trends in mean eGFR varied across ADSC therapy dose groups in comparison to the RWD EC group, as shown in Figure 5. In the low‐dose group, the ADSC therapy group maintained a more stable mean eGFR, whereas the RWD EC group showed a gradual decline. In the moderate‐dose group, both groups followed a similar trajectory without significant differences. In the high‐dose group, the ADSC therapy group appeared to exhibit slightly more stability, but the difference was not statistically significant. Figure 5 Part B shows the percentage change in eGFR from baseline, where the ADSC therapy low‐dose group consistently maintained a positive trend, indicating stable renal function, while the RWD EC group showed a progressive decline. In moderate‐ and high‐dose groups, both groups demonstrated reductions in eGFR percentage, and differences between them were not statistically significant. Overall, these findings suggest that the low‐dose ADSC therapy group exhibited relatively greater short‐term renal function stability compared to other dose groups.

FIGURE 5.

Comparison of eGFR trends and percentage changes over time between ADSC therapy doses and RWD control groups. A1, B1, and C1 illustrate the trend lines representing the mean eGFR at each time point. A2, B2, and C2 display stem‐and‐leaf plots depicting the distribution of eGFR values at each time point. A3, B3, and C3 present the trend of the mean change in eGFR compared to the baseline, calculated as (eGFR at each specific time point—baseline eGFR)/baseline eGFR. A4, B4, and C4 depict the distribution of eGFR changes at each time point. The mean values or mean percentage change in eGFR for each group are displayed below the corresponding time points. The p‐values were calculated using linear mixed‐effects models to compare eGFR between the two groups at each time point. ADSC‐TH, ADSC therapy; RWD‐EC, real‐world data external control.

3.5. Cumulative Risk of Renal Function Decline by ADSC Therapy Dose: A Comparative Analysis With RWD EC Group

Cumulative risk trajectories of renal function decline varied across ADSC therapy dose groups compared to the RWD EC group (Figure 6). In the low‐dose group, the ADSC therapy group showed slower risk accumulation across all eGFR decline thresholds. Notably, the risk of a decline exceeding 30% or 35% was significantly lower in the ADSC therapy group. In the moderate‐ and high‐dose groups, no significant differences were observed for eGFR declines of 20%, 25%, 30%, or 35%. For the 15% threshold, a minor difference was observed in the high‐dose group, whereas the moderate‐dose group showed no significant difference. These results suggest that low‐dose ADSC therapy may be associated with greater short‐term renal function stability compared to moderate and high doses.

FIGURE 6.

Cumulative risk analysis of renal function decline comparing ADSC therapy and RWD control groups across ADSC therapy doses. Renal function decline event is defined as the occurrence of a specified percentage decline in renal function for the “first time” during the study period. The follow‐up duration is calculated from the baseline to the “first occurrence” of the event. If no event occurs, the follow‐up duration is determined by the length of the study period. ADSC‐TH, ADSC therapy; RWD‐EC, real‐world data external control.

4. Discussion

We conducted a comparative analysis of ADSC therapy and an external control group derived from real‐world data to assess short‐term renal function stability in patients with CKD. As the original clinical trial was designed as a single‐arm study without a built‐in comparator, an external real‐world cohort receiving standard‐of‐care therapies was constructed to enable indirect treatment comparisons. To reduce baseline confounding and simulate key features of a randomized controlled trial, we applied PSM to generate two cohorts with comparable baseline characteristics including eGFR, comorbidities, and concomitant medications. Both groups received standard CKD management in accordance with routine clinical practice, with ADSC therapy being the only investigational intervention administered. This methodological approach strengthens the interpretability of treatment effects and approximates the counterfactual conditions necessary for drawing causal inferences in non‐randomized settings.

Given this design, the observed treatment effects should be interpreted in the context of short‐term renal outcomes. In patients with moderate‐to‐advanced CKD, where disease progression can occur rapidly, the preservation of renal function—even over a relatively brief period—can have important clinical implications. Prior studies have demonstrated that even transient declines in eGFR are associated with higher risks of dialysis initiation, progression to end‐stage renal disease and mortality [36, 37]. Therefore, stabilization of eGFR over a 48‐week period may reflect effective disease control in the absence of long‐term follow‐up data. Moreover, considering the regenerative and immunomodulatory mechanisms of ADSCs, short‐term stability may signal the initiation of renal repair processes that require longer durations to fully manifest. Achieving this early stability not only suggests potential therapeutic activity but may also lay the groundwork for improved long‐term outcomes, which should be confirmed in future studies.

This clinical relevance of short‐term renal function stability was further supported by our findings. Over the 48‐week observation period, the ADSC therapy group exhibited a more stable trajectory in mean percentage change in eGFR compared to the real‐world control group, in which eGFR declined progressively. When evaluating the incidence of renal function deterioration across various eGFR reduction thresholds (> 15%, 20%, 25%, 30%, and 35%), the ADSC therapy group consistently showed a lower risk across all levels. The differences were especially pronounced among patients with CKD Stage 4. In addition, cumulative risk progression analyses demonstrated a significantly slower rate of renal function decline in the ADSC therapy group. Notably, among patients with a baseline eGFR of less than 30—indicating advanced CKD, ADSC therapy was associated with greater stability in renal function. These findings support the potential utility of ADSCs in maintaining short‐term renal function in patients with moderate to advanced CKD.

Findings from Phase I and Phase II trials of ADSC therapy further support this observation. When stratified by a baseline eGFR threshold of 30, outcomes varied substantially. In patients with eGFR below 30, corresponding to severe CKD, the ADSC therapy group demonstrated greater renal function stability relative to the control group.

Traditional treatments in the control group, such as renin‐angiotensin‐aldosterone system (RAAS) inhibitors, are known to reduce intraglomerular pressure and consequently lower renal filtration pressure. However, their effect may be limited in advanced CKD due to irreversible structural damage to the kidneys [38, 39]. In this setting, the potential mechanisms of ADSCs—including anti‐inflammatory properties, reduction of renal fibrosis, and promotion of tissue regeneration—may play a critical role in stabilizing renal function by restoring damaged nephron structures [38, 39, 40].

The association between ADSC dose and short‐term renal function stability did not demonstrate a clear dose–response trend [41, 42]. Notably, patients receiving low‐dose ADSC therapy showed the most stable renal function, whereas those on moderate doses experienced outcomes comparable to the RWD external control group, and those on higher doses showed modest improvement but without a clear advantage over the low‐dose group.

These observations are consistent with findings from Papazova et al. [42], who also identified the absence of a dose‐dependent effect in stem cell‐based interventions. A supplementary analysis of baseline characteristics (Supporting Information, Section S6) further showed that dose groups were generally well balanced, indicating that the lack of a clear dose–response relationship is unlikely to be driven by baseline imbalances. Instead, this phenomenon may be linked to the underlying mechanism of ADSCs, which primarily stimulate the body's intrinsic repair systems rather than acting as a sustained external cellular input [43]. Additionally, higher doses might potentially compromise therapeutic benefits by provoking immune reactions or over activating inflammatory cascades, thereby offsetting their protective effects. Collectively, these insights highlight the need for future studies to delineate the optimal therapeutic window for stem cell–based treatments in kidney disease.

The therapeutic effects of ADSCs are thought to be mediated via paracrine and autocrine secretion of cytokines and growth factors that promote endogenous renal repair [43, 44]. The optimal stem cell concentration may vary depending on the severity and nature of kidney damage. In certain cases, a relatively small number of cells may be sufficient to activate reparative pathways, while others may require a higher dose to achieve similar effects [45, 46, 47]. Current evidence suggests that moderate‐dose ADSC therapy is adequate to trigger this reparative response and maintain renal function.

Additionally, the observed efficacy of low‐dose ADSC therapy highlights the therapeutic potential of ADSCs in CKD management. The absence of a strict dose–response relationship suggests that lower doses may be sufficient for clinical benefit, potentially improving treatment safety and cost‐effectiveness. Further studies are warranted to refine dose selection and better understand the underlying biological mechanisms of ADSC therapy in CKD.

In addition to efficacy, safety profiles were assessed in the Phase I/II trial of ADCS therapy. Across 39 participants, the overall incidence of adverse events was low and evenly distributed among the dose groups, with no evidence of dose‐dependent clustering. The most common events were infections, predominantly COVID‐19, while other categories, such as gastrointestinal, systemic, and laboratory findings, were infrequent and balanced. Importantly, no severe or dose‐related adverse reactions were identified, suggesting that ADCS therapy was well tolerated across all dose groups (Supporting Information Table S5).

Several limitations of this study should be acknowledged. First, although PSM and comprehensive adjustment for clinical variables were applied to minimize confounding, residual confounding due to unmeasured factors such as lifestyle, diet, and adherence to therapy cannot be excluded. Nonetheless, the baseline characteristics between the two groups were well balanced, providing a reasonable basis for comparison.

Second, the relatively short observation period, constrained by the clinical trial protocol, precludes assessment of long‐term renal outcomes. However, short‐term changes in eGFR are clinically meaningful, as they are associated with the risk of dialysis initiation and mortality in patients with advanced CKD [36, 48]. In this study, the ADSC therapy group exhibited a slower rate of short‐term eGFR decline compared to the control group, suggesting a potentially lower risk of future adverse renal outcomes.

At last, as Phase I and II studies are designed primarily to explore safety and short‐term effects, larger Phase III trials are needed to validate these findings and establish definitive clinical efficacy. Future research should incorporate extended follow‐up periods, larger patient cohorts, and evaluation of long‐term renal and survival outcomes.

In conclusion, this study utilized real‐world data to construct a comparable control group for evaluating the effects of ADSC therapy. The findings suggest that ADSC therapy may help stabilize short‐term renal function in patients with CKD, particularly those with advanced disease. Further studies are needed to confirm these preliminary observations and inform clinical practice.

Author Contributions

Y.‐C.L., Y.‐P.H., Y.‐C.T., M.‐J.W., H.‐C.L., S.‐Y.C., M.‐S.W. and H.‐Y.C. wrote the manuscript; Y.‐C.L., Y.‐P.H. and H.‐Y.C. designed the research; Y.‐P.H., Y.‐C.T., M.‐J.W., and M.‐S.W. performed the research; Y.‐C.L., H.‐C.L., S.‐Y.C. and H.‐Y.C. analyzed the data.

Ethics Statement

• ELIXCYTE Phase I/II clinical trial

The trial entitled “A Phase I/II Study to Evaluate the Safety and Efficacy of Allogeneic Infusion of Adipose‐Derived Stem Cells in Moderate to Severe Chronic Kidney Disease” was reviewed and approved by the institutional review boards (IRBs) of all participating hospitals:

• ○Taipei Medical University Shuang Ho Hospital, TMU‐Joint IRB approval no. N201710032 (November 24, 2017).
• ○Linkou Chang Gung Memorial Hospital, IRB approval no. 201900020A0 (April 24, 2019).
• ○Taichung Veterans General Hospital, IRB I & II approval no. SF20136B (July 24, 2020).
• •Real‐world data (RWD) external control

The external control group was constructed using de‐identified patient information obtained from the Taipei Medical University Clinical Research Database (TMUCRD). The use of these data for research purposes was independently reviewed and approved by the Taipei Medical University Joint IRB (approval no. N202310001, October 20, 2023). Since all data were anonymized prior to access, individual patient consent was not required in accordance with local regulations and ethical guidelines.

Consent

Written informed consent was obtained from all trial participants prior to enrollment. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki.

Conflicts of Interest

Y.‐P.H. was employed by UnicoCell BioMed Co. Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflicts of interest. The authors declare that this study received funding from UnicoCell BioMed Co. Ltd. The funder had the following involvement with the study: design and analysis.

Supporting information

Data S1: cts70415‐sup‐0001‐supinfo.docx.

Lin Y.‐C., Hung Y.‐P., Tian Y.‐C., et al., “Real‐World Data as External Control in Assessing the Efficacy of Allogeneic Adipose‐Derived Stem Cells Therapy for Advanced Chronic Kidney Disease,” Clinical and Translational Science 18, no. 12 (2025): e70415, 10.1111/cts.70415.