Editorial Note: Since this week's NephJC is somewhat unusual in having already spawned a bit of online conversation, we invited author Laurie Tomlinson to pre-emptively comment on the implications of their work. Don't miss the NephJC summary from Tom Oates, read on for Dr Tomlinson's commentary, and join us April 25/26 for the live discussion.
What does this paper show?
In the UK, current guidelines recommend that renal function should be measured before and two weeks after starting an ACE Inhibitor or an Angiotensin Receptor Blocker (ACEI/ARB). If creatinine rises by >30% or eGFR falls by >25% then prescribers should consider stopping the drugs. The rationale for this is somewhat mysterious. Using real-world data (from routinely-collected general-practice electronic health records) from over 120,000 UK patients who were started on treatment with an ACEI/ARB, we showed that there was a progressively increased risk of end-stage renal disease (ESRD), heart attack, heart failure or death for every 10% increase in creatinine after starting the drug.
So are we suggesting that the drugs cause adverse outcomes?
Overall, our results likely reflect that creatinine rises more in sicker patients. Large and small vessel renovascular disease, or cardiac failure with volume overload may well underly the fall in renal function after initiating the drugs. Therefore creatinine rise is a marker of pathology likely to be associated with a greater risk of an adverse outcome. In addition, patients with >30% creatinine rise were sicker in general and taking more drugs including NSAIDs and diuretics than those with smaller rises.
Despite this, it is possible that in a proportion of patients the drugs do directly cause adverse outcomes: I am sure that most nephrologists can recall seeing patients with unsuspected severe renovascular disease who have a profound acute drop in eGFR after starting ACEI/ARB which never recovers.
So should the drugs be stopped?
Some have interpreted our results as though the comparison group is those who were not treated with ACEI/ARB and therefore that the drugs should be stopped if renal function declines. However, the comparison group is those with no decline in renal function after initiation. We tried to be clear that from our results we cannot state that the risks outweigh the benefits of these drugs in any group. There is a vast amount of evidence from RCTs that these drugs are effective for a range of conditions. As we discuss, for some trials, investigators have undertaken post-hoc analyses of outcomes among subgroups defined by post-initiation change in creatinine. These show that ACEI/ARB are beneficial compared to placebo, even among those with the largest increases in creatinine. However, this evidence is predominantly confined to trials of cardiac failure and may not be relevant to current clinical care if patients are different from those in the RCTS (e.g. more comorbidities).
So what should we do if renal function declines?
Our work only relates to the immediate post-initiation period. We are not suggesting that the drugs should be stopped in all people whose renal function deteriorates after commencing ACEI/ARB: local guidelines and the clinical situation should be considered. For those in the UK, The NHS Think Kidneys programme has some useful guidance here. If there is progressive decline in patients with stage 4 or 5 CKD, in the UK they should be randomised to the STOP-ACE trial which will hopefully provide a clearer idea of how to optimally manage these patients in the future.
What is the key message of the paper?
Simply that patients whose renal function declines after commencing an ACEI/ARB are at increased risk of adverse events, even below a 30% rise in creatinine. We should consider why this has happened, whether any further investigations are needed and arrange for ongoing testing of their renal function. We have previously shown that <10% of patients have guideline recommended blood tests around initiation of an ACEI/ARB so a positive outcome would be renewed focus on the importance of monitoring these patients.
In addition, I hope that this paper will make us think about the limits of the evidence we use in our everyday practice. Despite our attempts to be measured in our wording and to discourage alarmist responses and deprescribing, this paper has been called dangerous and widely criticized. But, how confident are you that existing RCTs enable you to consider the change in the risk: benefit profile for prevention of ESRD if a patient has a substantial decline in renal function after initiation of ACEI/ARB?