Portal hypertension is the main complication and prognostic marker of liver cirrhosis, and it results in gastroesophageal varices, hepatic encephalopathy, and ascites [
Several large-scale, randomized, placebo-controlled clinical trials have discovered that ACEis and ARBs have renoprotective effects in diabetic [
In this population-based nationwide cohort study, we used the Taiwan National Health Insurance Research Database (NHIRD) and the Registry for Catastrophic Illness Patient Database (RCIPD), a subsystem of the NHIRD, to investigate the renal effects of long-term ACEis/ARBs use in patients with liver cirrhosis.
The data of this population-based cohort study were derived from the NHIRD, which contains prospectively collected nationwide health-care data, including demographic data, all records of outpatients’ visits and hospitalizations, details of prescriptions, operation codes, expenditure amounts, and diagnostic codes according to the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) from January 1, 1997, to December 31, 2013 [
Identification of patients with liver cirrhosis was based on specific codes (571.2, 571.5, and 571.6) once at admission or more than three times at the outpatient clinic. We also identified cirrhotic patients with ascites based on specific codes of liver cirrhosis and ascites (789.5) or a reimbursement code for ascites analysis (16002C). This study enrolled patients with liver cirrhosis who had continuously used hypotensive regimens of ACEi/ARB or calcium channel blockers (CCBs) for more than 63 days in the first 90 days. To avoid indication and immortal time biases, patients who continuously used ACEi or ARB (defined as the ACEi/ARB cohort) were randomly matched with those continuously using CCB (defined as the CCB cohort) at a ratio of 1 : 1. To optimize comparability among the study cohorts, patients were not enrolled if they had cancer, as identified in the RCIPD, before the index date [
Patients who attained major outcomes before enrolment or the index date were excluded. Those who used ACEi/ARB and CCB at the same time or these two kinds of drugs in any sequential orders for more than 30 days per year were also excluded.
All patients were followed up for the occurrence of the outcomes from the last day of ACEi/ARB or CCB administration in the first 90 days to death or December 31, 2013, whichever came first. The major outcome was end-stage renal disease (ESRD). ESRD was defined as irreversible renal failure requiring long-term dialysis and was ascertained by the certification in RCIPD [
The use of certain medications, including nonsteroidal anti-inflammatory drugs (NSAIDs) or cyclooxygenase-2 inhibitors (COX-2), aspirin, statins, metformin, and BBs, which may influence renal function, was analyzed. Drug users were defined as patients who used more than one tablet per month during the study period. The propensity score was measured using logistic regression analysis consisting of demographic factors including age, sex, comorbidities, and concomitant medications. Comorbidities, identified based on ICD-9-CM codes, included viral hepatitis B (070.2, 070.3, and V02.61), viral hepatitis C (070.41, 070.44, 070.51, 070.54, 070.70, 070.71, and V02.62), alcoholic liver disease (571.0–571.3), other chronic hepatitis (571.40, 571.41, and 571.49), hypertension (401–405, A260, and A269), diabetes mellitus (249–250), congestive heart failure (428), and hyperlipidemia (272) (Table
Because death for patients with liver cirrhosis led us to apply informative censoring when calculating the major outcome, mortality in enrolled patients was regarded as a competing risk event and was adjusted for through competing risk analyses.
Continuous variables are reported as median (25%–75% interquartile range) and categorical variables as number (percentage). The modified Kaplan–Meier method and Gray’s method were used to calculate and compare the cumulative incidence rates of ESRD [
Between January 1, 1997, and December 31, 2013, we identified 311,361 newly diagnosed cirrhotic patients. We excluded 284,597 patients who used neither ACEi/ARB nor CCB for more than 63 days in the first 90 days; 32,943 who simultaneously used ACEi/ARB and CCB for more than 30 days per year; 1,972 who had ESRD before the index date; 7,589 who had cancer before the index date; and 134 without records of sex or age. Of note, more than one exclusion criteria could overlap in a patient. Finally, 9,475 cirrhotic patients (4,208 in the ACEi/ARB cohort and 5,267 in the CCB cohort) were enrolled. After propensity score matching of patients in the two cohorts who showed no differences in demographic factors, viral hepatitis B, viral hepatitis C, hypertension, diabetes mellitus, congestive heart failure, hyperlipidemia, and concomitant use of BBs, statin, metformin, aspirin, NSAID, and COX-2, 4,376 patients (2,188 patients in each cohort) were eligible for comparison (Figure
Flowchart of the enrollment process for patients with liver cirrhosis.
The median patient age was 67.75 (58.41–75.52) years in the ACEi/ARB cohort and 67.97 (58.58–75.30) years in the CCB cohort (
Baseline characteristics of study patients.
Characteristics | ACEi/ARB ( |
CCB ( | |
---|---|---|---|
Age, y, median (IQR) | 67.75 (58.41–75.52) | 67.97 (58.58–75.30) | 0.914 |
Gender | >0.999 | ||
Female | 873 (39.9) | 873 (39.9) | |
Male | 1315 (60.1) | 1315 (60.1) | |
Cause of cirrhosis | |||
Hepatitis B virus infection | 274 (12.5) | 274 (12.5) | >0.999 |
Hepatitis C virus infection | 312 (14.3) | 283 (12.9) | 0.217 |
Alcoholic liver disease | 145 (6.6) | 147 (6.7) | 0.952 |
Other chronic hepatitis | 942 (43.1) | 922 (42.1) | 0.561 |
Comorbidity | |||
Hypertension | 1982 (90.6) | 2000 (91.4) | 0.369 |
Diabetes mellitus | 845 (38.6) | 856 (39.1) | 0.756 |
Congestive heart failure | 222 (10.2) | 210 (9.6) | 0.577 |
Hyperlipidemia | 504 (23.0) | 501 (22.9) | 0.943 |
Drug exposure | |||
Beta-blockers | 528 (24.1) | 542 (24.8) | 0.648 |
Statin | 159 (7.3) | 193 (8.8) | 0.067 |
Metformin | 542 (24.8) | 561 (25.7) | 0.531 |
Aspirin | 464 (21.2) | 478 (21.9) | 0.633 |
NSAIDs or COX-2 | 862 (39.4) | 879 (40.2) | 0.621 |
ESRD | 29 (1.3) | 28 (1.3) | >0.999 |
Competing mortality | 739 (33.8) | 928 (42.4) | <0.001 |
Follow-up year (IQR) | 2.95 (1.26–5.78) | 3.14 (1.24–6.19) | 0.089 |
ACEi: angiotensin-converting enzyme inhibitor; ARB: angiotensin receptor blocker; CCB: calcium channel blocker; COX-2: cyclooxygenase-2 inhibitors; ESRD: end-stage renal disease; IQR: interquartile range; NSAIDs: nonsteroidal anti-inflammatory drugs.
Using the modified Kaplan–Meier method and Gray’s method, the 10-year cumulative incidence rates of ESRD were 2.32% (95% CI: 1.45–3.20) and 1.70% (95% CI: 1.03–2.36) in the ACEi/ARB and CCB cohorts after adjustment for competing mortality (
Cumulative incidence of ESRD in patients with liver cirrhosis that was analyzed using the modified log-rank test with death adjusted as a competing risk event.
Table
Multivariate Cox proportional hazards model analysis of risk of ESRD after adjustment for competing mortality.
HR (95% CI) | ||
---|---|---|
ACEi/ARB vs. CCB users | 1.15 (0.69–1.94) | 0.591 |
Age | 0.98 (0.96–1.00) | 0.040 |
Male vs. female | 1.02 (0.59–1.80) | 0.932 |
Hepatitis B virus infection | 0.71 (0.28–1.82) | 0.481 |
Hepatitis C virus infection | 0.73 (0.28–1.90) | 0.522 |
Alcoholic liver disease | 0.20 (0.03–1.40) | 0.104 |
Other chronic hepatitis | 0.73 (0.42–1.27) | 0.267 |
Hypertension | 1.82 (0.57–5.74) | 0.309 |
Diabetes mellitus | 2.49 (1.29–4.82) | 0.007 |
Congestive heart failure | 1.31 (0.53–3.27) | 0.559 |
Hyperlipidemia | 1.25 (0.69–2.25) | 0.465 |
Beta-blockers | 0.71 (0.38–1.32) | 0.278 |
Statin | 1.14 (0.50–2.61) | 0.760 |
Metformin | 0.86 (0.44–1.68) | 0.664 |
Aspirin | 0.94 (0.49–1.81) | 0.858 |
NSAIDs or COX-2 | 0.51 (0.29–0.91) | 0.022 |
ACEi: angiotensin-converting enzyme inhibitor; ARB: angiotensin receptor blocker; CCB: calcium channel blocker; CI: confidence interval; COX-2: cyclooxygenase-2 inhibitor; ESRD: end-stage renal disease; HR: hazard ratio; NSAIDs: nonsteroidal anti-inflammatory drugs.
Enrolled patients who received ACEis/ARBs did not exhibit an increased risk of ESRD compared with patients who received CCB (
Multivariate stratified analyses for the association between ACEi/ARB or CCB therapy and ESRD risk in patients with liver cirrhosis. ACEi: angiotensin-converting enzyme inhibitor; ARB: angiotensin receptor blocker; COX-2: cyclooxygenase-2 inhibitor; CI: confidence interval; HR: hazard ratio; HBV: patients with hepatitis B virus infection; HCV: patients with hepatitis C virus infection; NSAID: nonsteroidal anti-inflammatory drug.
We also identified a subgroup of cirrhotic patients with ascites (
The median age was 69.93 (57.31–78.38) years in the ACEi/ARB cohort and 69.68 (58.49–78.44) years in the CCB cohort (
We also identified a subgroup of patients with decompensated liver cirrhosis (
This population-based nationwide cohort study is the first study to investigate the renal effects of ACEis and ARBs in patients with liver cirrhosis. The results revealed that long-term ACEis/ARBs use was safe in cirrhotic patients without ascites. However, ACEis and ARBs potentially but nonsignificantly (
Managing renal failure in patients with liver cirrhosis is challenging [
Randomized controlled trials have shown that ACEis/ARBs are renoprotective [
Diabetes mellitus is a well-known risk factor for ESRD [
Schepke et al. [
This study had several limitations. First, dynamic changes in renal function were unknown. Although only a few patients showed ESRD occurrence, which was the major outcome in this study, ESRD was a definite diagnosis in comparison with reversible chronic kidney disease. Furthermore, it is unethical to design a prospective study to investigate the side effects of ACEis/ARBs in cirrhotic patients. Second, although proteinuria is a key factor predicting renal function deterioration [
In conclusion, long-term ACEis/ARBs use did not increase the risk of ESRD in patients with liver cirrhosis, but they tended to increase the risk of ESRD in cirrhotic patients with ascites. ACEis and ARBs should be used with caution in cirrhotic patients with ascites, and future multicenter retrospective studies should collect detailed information on hepatic and renal biochemistries of cirrhotic patients taking ACEis or ARBs.
The data used to support the findings of this study may be released upon application to the Bureau of National Health Insurance, Department of Health, which can be contacted at
This study was conducted using the NHIRD provided by the Bureau of National Health Insurance, Department of Health, and managed by National Health Research Institutes. The interpretation and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health, or National Health Research Institutes.
The authors declare no conflict of interest.
Wei-Fan Hsu and Cheng-Yuan Peng contributed to the study concept and design. Wei-Fan Hsu and Chun-Ying Wu acquired the data. Wei-Fan Hsu drafted the manuscript. Chun-Ying Wu and Cheng-Yuan Peng critically revised the manuscript for important intellectual content. Wei-Fan Hsu performed the statistical analysis. Wei-Fan Hsu and Chun-Ying Wu interpreted the data. All authors reviewed and approved the final version of the manuscript. Chun-Ying Wu and Cheng-Yuan Peng contributed equally to this work and should be considered as co-corresponding authors.
This study was supported in part by the Taiwan Ministry of Health and Welfare Clinical Trial Center (MOHW106-TDU-B-212-113004) and by a grant (DMR-108-225) from the China Medical University Hospital, Taichung, Taiwan.
Table S1: antihypertensive agents. Table S2: ICD-9-CM codes for the diagnoses of comorbidities. Table S3: baseline characteristics of cirrhotic patients with ascites. Table S4: multivariate Cox proportional hazards model analysis of risk of ESRD in cirrhotic patients with ascites after adjustment for competing mortality. Table S5: baseline characteristics of the patients with decompensated liver cirrhosis. Table S6: baseline characteristics of patients with liver cirrhosis taking ACEi/ARB or CCB before propensity score matching. Figure S1: flowchart of the enrollment process for cirrhotic patients with ascites.