Introduction
End-stage kidney disease (ESKD) is a major health burden worldwide, associated with high rates of cardiovascular mortality, morbidity, and healthcare utilization [1,2]. Over 500,000 people in the United States have been diagnosed with ESKD, and its prevalence continues to increase worldwide [3,4]. South Korea is among the countries with the fastest-growing ESKD rates in the world, ranking fourth, behind Taiwan, the United States, and Singapore, respectively [3]. Given this alarming trend, identifying and modifying risk factors that contribute to the progression of chronic kidney disease (CKD) to ESKD is of utmost importance in South Korea [5].
Acid-suppressive medications, particularly proton pump inhibitors (PPIs), are widely prescribed in patients with CKD due to the high prevalence of gastrointestinal symptoms in this population [6]. However, growing evidence suggests that long-term PPI use may increase the risk of CKD progression, raising concerns about their safety in patients with impaired renal function [7,8]. Histamine-2 receptor antagonists (H2RAs) are often considered a safer alternative to PPIs in patients with CKD, as doses can be adjusted in cases of renal impairment [9]. However, their anti-acid effects are generally weaker than those of PPIs, potentially limiting their clinical use.
Potassium-competitive acid blockers (P-CABs), such as tegoprazan, have emerged as a novel class of acid-suppressive medications that can provide rapid and sustained acid inhibition through a mechanism of action distinct from PPIs and H2RAs [10]. Since its approval in South Korea in 2018, tegoprazan has been increasingly prescribed, and recent national prescription data show that its use continues to rise in various patient populations [11]. Despite its growing use, little is known about the renal safety of tegoprazan, particularly in patients with moderate to severe CKD. To date, no large-scale population-based studies have compared the risk of ESKD from tegoprazan to other anti-acid medications.
This study investigated whether the use of tegoprazan for a minimum of 90 days in patients with stage 3 or 4 CKD was associated with a higher risk of ESKD progression compared to PPIs or H2RAs.
Methods
Study design and source of data
This retrospective real-world study used national claims data from South Korea. The Health Insurance Review and Assessment Service (HIRA) database was used—specifically, entries recorded from January 2013 to December 2022, encoded in the Observational Medical Outcomes Partnership Common Data Model (OMOP-CDM) version 5.3.1—and the data were analyzed using the ATLAS platform [12–14]. Three independent pairwise analyses were conducted, comparing the risk of ESKD between tegoprazan and H2RAs (analysis 1), tegoprazan and PPIs (analysis 2), and PPIs and H2RAs (analysis 3).
Network meta-analysis was performed to estimate the risk of ESKD across the three classes of acid-suppressive medications (Fig. 1).
The study was reviewed by the Institutional Review Board of the Pusan National University Hospital and certified as exempt from ethics committee review (PNUH IRB 2502-011-148).
Study population and definitions
Adult patients with stage 3 or 4 CKD prescribed tegoprazan, PPIs, or H2RAs for more than 90 days were included in the study. Patients with a prior history of dialysis or kidney transplantation were excluded. Index date was defined as the first day of prescription of each drug, and patients exposed to more than two categories of acid-suppressive medications during the study period (365 days before the index date–1,095 days after the index date) were excluded (Fig. 1). CKD-tegoprazan, CKD-PPIs, and CKD-H2RAs cohorts were defined as patients with CKD 3 or 4 prescribed tegoprazan, PPIs, or H2RAs for more than 90 days without cross medications. PPIs included dexlansoprazole, esomeprazole, ilaprazole, lansoprazole, omeprazole, pantoprazole, and rabeprazole; H2RAs included cimetidine, famotidine, lafutidine, nizatidine, ranitidine, and roxatidine. CKD status was determined based on International Classification of Diseases, 10th Revision (ICD-10) codes. Histories of hemodialysis, peritoneal dialysis, and kidney transplantation were obtained via procedure codes. Prescriptions for each drug were determined by respective OMOP-CDM concept IDs (Supplementary Table 1, available online).
Primary and secondary outcomes
The primary composite outcome was death or ESKD within 3 years of the index date. ESKD was identified using ICD-10 and procedure codes, as evidenced by the presence of both N18.5 and a dialysis- or kidney-transplantation-associated procedure (Supplementary Table 1, available online). To more accurately evaluate the impact of using each acid-suppressive medication for 90 days or longer, and to minimize protopathic bias and early events unrelated to prolonged exposure, any events occurring within 90 days of the index date were excluded from analyses. The secondary outcomes were death and ESKD.
Statistical analysis
All analyses were conducted using OMOP-CDM-standardized data. Analyses were performed by HIRA personnel within a secure, Docker-based R environment (version 3.5.1) provided by HIRA, independent of the study investigators. Continuous variables were summarized as means with standard deviations, and categorical variables were summarized as counts and percentages. The number of events (death or ESKD) was divided by the cumulative observation time and reported as incidence rates (IRs) per 100 person-years. Hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated using Cox proportional hazards regression models.
Propensity scores were estimated via L1-regularized logistic regression, adjusted for sex, age group, index year, condition group, drug group within 365 days before the index date, and the Charlson Comorbidity Index (CCI). Each comparison incorporated more than 8,000 covariates (range, 8,450–9,470). Covariate balance was evaluated using standardized mean differences (SMDs), with an SMD of <0.20 considered indicative of acceptable balance.
To compare risk estimates across the three treatment groups, a frequentist random-effects network meta-analysis was conducted. Heterogeneity and inconsistency were assessed using Cochran’s Q statistic, and between-study variance (τ2) was estimated via the DerSimonian–Laird method. The I2 statistic was calculated to quantify heterogeneity, with values close to zero indicating low variability and high consistency across comparisons.
In sensitivity analyses, outcome event rates at 1 and 2 years after the index date were compared between the CKD-tegoprazan and CKD-H2RAs cohorts, between the CKD-tegoprazan and CKD-PPIs cohorts, and between the CKD-PPIs and CKD-H2RAs cohorts. All statistical analyses were conducted in R version 4.4.2 (R Foundation for Statistical Computing) with the meta, netmeta, and ggplot2 packages. All tests were two-sided, and p-values <0.05 were considered statistically significant.
Results
Construction of the study population
The system retrieved 1,005, 12,465, and 17,009 patients in the CKD-tegoprazan, CKD-PPIs, and CKD-H2RAs cohorts, respectively. Among these patients, 22, 34, and 412 individuals who were prescribed other types of anti-acid medications during the study period were excluded. Additionally, 4% of tegoprazan users, 1.7% of PPIs users, and 0.5% of H2RAs users whose events occurred within 90 days of the index date were also excluded. Patient data totals of 943 vs. 16,896, 932 vs. 12,215, and 11,844 vs. 16,507 were available for crude estimation in analyses 1, 2, and 3, respectively. After 1:1 propensity score matching (PSM), 674, 902, and 4,583 patient data pairs were available for each analysis (Fig. 2).
During the 3-year study period, prescription patterns for long-term acid-suppressive medications changed. The use of H2RAs decreased, whereas the use of PPIs increased until 2021, then slightly decreased in 2022. In South Korea, the prescription of tegoprazan in patients with CKD 3 or 4 began in 2019 and increased until 2022 (Supplementary Fig. 1, available online).
Baseline characteristics of propensity score-matched cohorts
Baseline characteristics of patients in each PSM analysis are summarized in Table 1. Nearly 50% were aged 75 years or older, and approximately 60% were male. Approximately 80% of the study population had gastritis. Over 85%, 45%, and 28% had hypertension, diabetes, and ischemic heart disease, respectively, and approximately 70% were prescribed antithrombotic agents. Over 70%, 50%, and 30% of the population were prescribed renin-angiotensin-aldosterone system inhibitors (RAASi), diuretics, and dipeptidyl peptidase-4 inhibitors, respectively. In analyses 1 and 2, approximately 5% of patients were prescribed sodium-glucose cotransporter-2 inhibitors (SGLT2i), whereas in analysis 3, they were prescribed to less than 2% of patients. Glucagon-like peptide-1 receptor agonists (GLP1RAs) were prescribed to 0.6%–2.4% of the patients. CCIs were 8.5, 8.4, and 7.1 in analyses 1, 2, and 3, respectively. In each analysis, the SMD was less than 0.2 (Supplementary Fig. 2, available online), representing adequate covariate balance for outcome analysis. Baseline characteristics of the study population before PSM are shown in Supplementary Table 2 (available online).
Death or end-stage kidney disease progression within 3 years in patients with chronic kidney disease 3 or 4 in the tegoprazan, proton pump inhibitors, and histamine 2 receptor agonists groups
Tegoprazan compared to histamine 2 receptor agonists (analysis 1)
A total of 943 patients in the CKD-tegoprazan cohort were followed for 1,220 person-years, and 147 patients experienced death or ESKD, yielding an IR of 12.05 per 100 person-years (95% CI, 10.11–14.00). Of the 16,896 patients in the CKD-H2RAs cohort who were followed up for 38,168 person-years, 5,407 events occurred, with an IR of 14.17 (95% CI, 13.79–14.54)/100 person-years. The incidences of death or ESKD tended to be lower in tegoprazan users (HR, 0.85 [95% CI, 0.72–1.00]; p = 0.06); however, in PSM analysis, there was no significant difference between the two cohorts (HR, 0.91 [95% CI, 0.66–1.26]; p = 0.57) (Table 2).
Tegoprazan compared to proton pump inhibitors (analysis 2)
A total of 932 patients in the CKD-tegoprazan cohort were observed for 1,204 person-years, and events occurred in 147 (IR, 12.21 [95% CI, 10.24–14.19]/100 person-years). Of the 12,215 patients in the CKD-PPI cohort who were observed for 23,208 person-years, events occurred in 3,665 (IR, 15.79 [95% CI, 15.28–16.30]/100 person-years). In the crude analysis, the risk of death or ESKD development was significantly lower in the CKD-tegoprazan cohort (HR, 0.75 [95% CI, 0.63–0.88]; p < 0.001); however, significance did not increase in the PSM cohorts (HR, 0.87 [95% CI, 0.66–1.15], p = 0.32) (Table 2).
Proton pump inhibitors compared to histamine 2 receptor agonists (analysis 3)
The risks of ESKD or death were compared in patients with CKD 3 or 4 in the CKD-PPI and CKD-H2RA groups to validate the soundness of the study population and analytical methods. A total of 11,844 PPI users were followed up for 22,404 person-years, and 16,507 H2RA users were followed up for 37,146 person-years. In the crude analysis ESKD or death was more frequent in PPI users (PPI: IR, 15.95 [95% CI, 15.43–16.47]/100 person-years; H2RA: IR, 14.38 [95% CI, 13.99–14.76]/100 person-years), and the HR was significantly higher in PPI users than in H2RA users (HR, 1.11 [95% CI, 1.06–1.15]; p < 0.001), although statistical significance was diminished in PSM analysis (HR, 1.08 [95% CI, 0.99–1.18]; p = 0.09) (Table 2).
Network meta-analysis
Between-study heterogeneity and inconsistency were negligible in the crude (τ2 = 0, I2 = 0%, Q = 0.04, p = 0.84) and PSM (τ2 = 0, I2 = 0%, Q = 0.02, p = 0.90) analyses. In the crude analysis, the risk of death or ESKD was highest in the PPIs group (HR, 1.31 [95% CI, 1.17–1.48]), followed by the H2RAs (HR, 1.19 [95% CI, 1.06–1.34]) and tegoprazan (reference) groups. Similar trends were observed in the PSM analysis, although statistical significance was not noted (Fig. 3).
Secondary outcomes
End-stage kidney disease progression
IR and HRs for analyses 1, 2, and 3 are summarized in Table 3. According to the network meta-analysis using the unadjusted crude data, the risk of ESKD progression in the CKD-tegoprazan group was similar to that in the CKD-H2RAs group. The CKD-PPIs group showed a 28% higher risk of ESKD progression compared to either the CKD-tegoprazan or CKD-H2RAs groups in the crude analysis. The trends were similar in PSM analysis; nevertheless, statistical significance was not observed (Fig. 3).
Death
In the network meta-analysis using crude data, both CKD-H2RAs and CKD-PPIs groups exhibited 54% and 66% higher risks of death, respectively, compared to the CKD-tegoprazan group; however, these risks decreased in the PSM analysis (Fig. 3). Detailed results for analyses 1, 2, and 3 are presented in Table 3.
Sensitivity analysis
The risk associated with the composite outcome of death or progression to ESKD, as well as the individual outcomes observed within 1 and 2 years from the index date, were evaluated and compared. In crude analyses, the HRs consistently indicated a trend toward lower risk of adverse outcomes in the CKD-tegoprazan group compared to those receiving CKD-H2RAs or CKD-PPIs. Notably, when comparing tegoprazan to PPIs, a reduction in mortality risk was evident at 1, 2, and 3 years, with statistical significance sustained after PSM at both 2 years (HR, 0.64 [95% CI, 0.42–0.97]) and 3 years (HR, 0.55 [95% CI, 0.35–0.85]). No adverse safety signals were detected for ESKD or the composite outcome. Consistent with previous findings, the risk of ESKD progression remained higher in the CKD-PPIs group than in the CKD-H2RAs group (Fig. 4).
Discussion
In the current study, the HIRA database, a nationwide claims database in South Korea, was used to evaluate the risk of death or ESKD progression in patients with CKD 3 or 4 receiving tegoprazan for at least 90 days. Due to the extremely high mortality risk among patients with CKD [15]—which is a significant competing outcome compared to ESKD progression—we selected death or ESKD as the composite primary outcome. Patients were followed for 3 years, and outcomes were compared among the CKD-tegoprazan and CKD-H2RAs, CKD-tegoprazan and CKD-PPIs, and CKD-PPIs and CKD-H2RAs cohorts. Network meta-analysis was performed, and the results of the three independent analyses were merged. The risk of death or ESKD was not increased in the CKD-tegoprazan cohort compared to the CKD-H2RAs or CKD-PPIs cohorts. In sensitivity analyses, the risk of events at 1 and 2 years was compared, and the results yielded the same trend. The effect estimates remained directionally stable even with a small cohort, suggesting that prescribing tegoprazan for over 90 days did not add additional risk of ESKD progression or death in patients with CKD stage 3 or 4.
Several previous studies have reported concerns with respect to possible nephrotoxic effects of P-CABs. Ishida et al. [16] reported an association between vonoprazan and an increased risk of tubule-interstitial nephritis, with a similar rate to that associated with PPIs. Suzuki et al. [17] reported higher rates of AKI among vonoprazan users. In another cross-sectional study [11] based on South Korean national claims data, an association was reported between tegoprazan and an increased risk of AKI development, with a similar rate to that associated with PPIs. A risk of hypomagnesemia, which is associated with an increased risk of death or ESKD progression in patients with CKD via endothelial dysfunction or oxidative stress [18–20], has also been reported in tegoprazan users [21,22]. Our finding of no additional risk of death or ESKD in tegoprazan usage does not exclude these possibilities, and continued pharmacovigilance studies are warranted.
The enrollment of patients with significant comorbidities may have lessened the visible effect of tegoprazan on mortality or ESKD progression. Nearly half of the study participants were over 75 years old and had a median CCI of 7 or higher, indicating multiple health issues. Given that medically ill hospitalized patients with a CCI ≥5 have an 85% 1-year mortality rate [23], CCIs higher than 7 reflect a substantial risk of death. Additionally, having CKD is itself associated with risks of death and ESKD progression. Data from the KNOW-CKD (KoreaN Cohort Study for Outcomes in Patients With Chronic Kidney Disease) registry in South Korea, which followed patients for 4 years, showed a mortality rate of 9.6 per 1,000 person-years [24] and the estimated annual risk of ESKD progression of 5.4% [25].
In our previous studies, long-term PPI prescription was associated with a 14% increase in the risk of ESKD progression in patients with stage 3 or 4 CKD. Nevertheless, the nephrotoxic effects of PPIs differed depending on the patient’s condition. In a subgroup of non-diabetic patients or patients with stage 3 CKD, the risk of ESKD progression was 36% or 39% higher than that in H2RA users, respectively. In a subgroup of patients with stage 4 CKD, diabetes, and liver cirrhosis, long-term PPI prescription was not associated with an increased risk of ESKD progression [7]. We speculated that long-term PPI prescription contributed to ESKD progression in low-risk groups, whereas it may not add additional risks in high-risk groups. Similarly, associations between tegoprazan and death or ESKD progression may have been mitigated in patients with CKD and multiple known risk factors for adverse kidney outcomes. To clarify tegoprazan’s nephrotoxicity, these factors should be studied in low-risk patients.
The relatively small patient numbers and short follow-up period in the CKD-tegoprazan cohort reduced the statistical power of the current study. Since tegoprazan has been available in South Korea since 2019, and data were collected up to 2022, some patients had follow-up periods of less than 3 years. To overcome this limitation, we adjusted the index year in PSM and performed sensitivity analysis, counting outcomes at 1 and 2 years. The findings were consistent with those of the main analysis, indicating that tegoprazan did not cause extra harm for patients with CKD who already have multiple risk factors for mortality or ESKD progression.
The current study had several limitations. First, the retrospective design may introduce selection bias and unmeasured confounding factors. True adherence to each acid-suppressive medication was not definitively confirmed. Although over 8,000 covariates were adjusted for in each analysis, key clinical variables, including lifestyle factors (e.g., smoking, alcohol consumption), laboratory measures, proteinuria severity, blood pressure, and over-the-counter medication use, were not adjusted for, which may introduce bias toward the null and warrant cautious interpretation of the findings. Second, potential interactions between acid-suppressive medications and RAASi, SGLT2i, and GLP1RAs were not evaluated. Third, the relatively small CKD-tegoprazan cohort compared to the PPIs or H2RAs groups reduced statistical power. Finally, as the study included only patients with CKD in South Korea, differences in race and ethnicity may limit the generalizability of these results to broader populations.
The current study had several strengths. To our knowledge, it is the first to evaluate real-world data pertaining to the influence of tegoprazan on death or ESKD progression in patients with CKD. Using nationwide data and extensive PSM adjusting for over 8,000 variables enhanced the reliability of our results. Network meta-analysis enabled the estimation of direct comparisons among the three categories of anti-acid medications. Although the tegoprazan cohort was smaller than those of the PPIs or H2RA groups, this is the largest nationwide study, providing the most extensive real-world evidence available.
Although the evidence remains insufficient to draw firm conclusions, the prescription of tegoprazan for more than 90 days may not add additional risk for ESKD progression or death in patients with CKD stage 3 or 4. The influence of tegoprazan on kidney outcomes in low-risk groups or non-Asian populations may differ; therefore, further investigations are required in diverse populations.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement table 1
Print
