Renin–Angiotensin System Blocker Discontinuation and Adverse Outcomes in Chronic Kidney Disease

Carl P. Walther; Wolfgang C. Winkelmayer; Peter A. Richardson; Salim S. Virani; Sankar D. Navaneethan


Nephrol Dial Transplant. 2021;36(10):1893-1899. 

In This Article


In this large, older, predominantly male veteran cohort with CKD and new ACEI/ARB use we found ACEI/ARB discontinuation to be associated with a substantially increased risk of death and ESKD. The risk of all-cause death approximately doubled following ACEI/ARB discontinuation and the risk of ESKD increased ~50%. The risk of death was found to decrease with more prolonged discontinuation, whereas the risk of incident ESKD showed no clear pattern over the different durations. The strength of the association, especially with all-cause death, and the decrease in the association with more prolonged discontinuation suggest that the observed association is primarily due to factors driving discontinuation events rather than being caused by the discontinuation events themselves.

These results add new information to what is currently known about outcomes following ACEI/ARB discontinuation. Causal information about the effect of ACEI/ARB use patterns on the risk of all-cause death or ESKD in CKD can only be obtained from randomized trials. Randomized trials in people with CKD, with and without diabetes or albuminuria, comparing ACEI/ARB therapy to other antihypertensives or to placebo have shown varying degrees of benefit for all-cause mortality and ESKD.[20] Benefits for other outcomes (such as GFR change, doubling of serum creatinine and cardiovascular endpoints) and among certain subgroups (such as people with diabetes or albuminuria) have been more consistent and pronounced.[20] The benefit and safety of closely monitored ACEI/ARB use even in advanced CKD has been demonstrated.[21–23] However, trial results are not available to answer questions about the effects of temporary or permanent discontinuation of ACEIs/ARBs, as randomized studies comparing stoppage to continuation have not been performed, although one is ongoing in people with advanced CKD.[24]

Observational evidence on the effects of ACEI/ARB discontinuation in CKD, either temporarily or permanently, is limited.[25] Using clinical data from 53 912 patients on ACEIs and ARBs, Qiao et al.[26] reported that ACEI/ARB discontinuation was common among those with Stage 4 CKD. Furthermore, hyperkalemia, acidosis, hypotension and hospitalization were associated with ACEI/ARB discontinuation. Furthermore, a recent analysis of 3909 patients with CKD cared for at a comprehensive health system found that ACEI/ARB discontinuation for >60 days within 6 months following an eGFR decline to <30 mL/min/1.73 m2 was associated with a 39% increased risk of death, 37% increased risk of major adverse cardiac events and a nonstatistically significant finding of 19% increased risk of ESKD.[27] The relative strength and direction of these findings accord with the findings of our study. In a propensity-matched cohort study of Chronic Renal Insufficiency Cohort (CRIC) study participants with eGFR <30 mL/min/1.73 m2, the pattern of renin–angiotensin system inhibitor use over 1 year (never, always, new and dynamic) was not associated with the risk of ESKD or death.[28] A recently published study analyzed ACEI/ARB use in CKD patients treated at the VA during a 3-year period before transition to dialysis.[17] The authors found that longer and continuous ACEI/ARB use prior to hemodialysis was associated with better postdialysis survival. This accords well with the findings of our study. Given the consistency of the findings, our data support and extend the notion that ACEIs/ARBs should generally be continued unless specific clinical situations (such as hypotension or AKI) preclude their use.

Possible factors driving discontinuation events and the associated mortality—only partially captured in our covariate data—include the severity of kidney disease, variability of kidney function, cause of kidney disease, frailty and limited projected life expectancy. It is notable that a significant portion of discontinuation events had no clearly identifiable cause based on lab results or severe intercurrent illness reflected in hospitalization. This suggests that most discontinuation events are related to other, more nuanced and difficult to discern factors, including those discussed above. Differentiating and understanding the burden of physiological and nonphysiological adverse events would aid clinicians in focusing on CKD patients who could be safely restarted on these agents. This is critical, as ACEI/ARB use in CKD appears to have plateaued in recent years and many people with CKD may be missing out on the important cardiovascular and kidney benefits of these agents.[29] It is also plausible that some of the low use of ACEIs or ARBs in CKD patients is related to clinical inertia (i.e. failure to initiate or intensify therapy when clinically indicated). Indeed, recent studies have shown that even among patients admitted with AKI, ACEI or ARB therapy can be initiated or reinitiated in most patients after discharge and is associated with improved mortality.[30] ACEI and ARB use remains foundational to CKD management, even as new effective therapies emerge for management of CKD.

The strengths of this analysis include the large sample size, the stringent CKD definition requiring sustained reduced eGFR and the availability of pharmacy data and key outcome measures. This analysis also expands the current knowledge regarding the association of ACE/ARB discontinuations with long-term outcomes. Limitations for this study include the inability to determine the exact circumstances of ACEI/ARB discontinuation, including what factors led to the discontinuations and what the intent of discontinuation events, and the inability to determine whether reinitiation of these agents is safe and offers kidney and cardiovascular benefits. Future investigations applying techniques such as natural language processing to large national registries may be able to shed additional light on these issues. Additionally, we were unable to investigate cause-specific deaths. Furthermore, we were unable to study the association of discontinuation with specific outcomes (such as major atherosclerotic cardiovascular events and CHF) due to a lack of comprehensive data for these outcomes. The observational nature of the study prevents any conclusions on causality. Due to the nature of the study cohort, whether these data are applicable to other settings is unclear. Proteinuria data were not available for the majority of people and were excluded from the model. We were unable to differentiate ACEI/ARB discontinuation with replacement with alternative antihypertensive medications from discontinuation without replacement.

In conclusion, we have demonstrated that ACEI/ARB discontinuation, even for a short period, is associated with a substantially increased risk of death and ESKD in a large cohort of patients with CKD. It remains to be determined what factors are driving the high frequency of discontinuations and, more importantly, what the appropriate stopping and restarting criteria are for ACEI/ARBs to optimize outcomes.