As I discussed the paper, I invited people in the audience to come up to
the microphone and comment on the paper – “findings too good to be true”, “dual
blockade doesn’t work” were some of the remarks. Interestingly, no-one mentioned the fact that the paper
had been retracted by the Lancet because of mistakes or perhaps pure fabrication (as suggested by Kunz et al and others). Stuart Linas had a nice review in CJASN that reviews the evidence on dual blockade (combination therapy).
The most influential study on dual blockade is the ONTARGET study published in 2008 in NEJM. It showed increased risk with dual blockade.
Briefly, ONTARGET recruited 25,620
patients with either pre-existing vascular disease or diabetes and randomly
assigned patients to either ramipril, telmisartan, or the combination of ramipril
and telmisartan. Patients were followed for 4.5 years. The primary outcome was a
composite of CV death, MI, stroke, or hospitalized CHF. There were multiple
secondary endpoints.
The key findings of ONTARGET are as follows:
- No hint of benefit for any endpoint with combination therapy
- Trend toward increased risk of death, 1.07 (0.98-1.16), with combination therapy
- Many more adverse events with combination therapy
In 2011, there was a secondary analysis of ONTARGET
published in Circulation. The analysis comprised of 5623 patients with eGFR<60
and/or proteinuria
- Combination therapy reduced proteinuria to a greater degree than monotherapy
- Combination therapy significantly increased the incidence of ESRD or doubling of Scr (0.79 vs. 0.56 % per year)
- Combination therapy significantly increased the incidence of ESRD (2.7 vs. 1.6 % per year) among patients with both eGFR<60 and proteinuria
Why should there be increased risk with dual blockade? Non-one
really knows.
What should one do? For starters, avoid dual blockade
because it is unclear that it works and there appears to be increased risk.
However, the worse outcomes in ONTARGET turned out to be ONLY those sub-set of patients with advanced cardiovascular disease. Overall, the population risk was low, and events so few, that statistically, one would expect to ONLY find negative events. A fair discussion of this topic should include the following rebuttal: 1. Nat Rev Nephrol. 2009 Aug;5(8):436-7.
ReplyDeleteDr. Singh, do you apply ONTARGET conclusions to patients with proteinuric CKD whose proteinuria does not improve to <1g/d despite one agent (ACE or ARB) and have BP over 140/90?
ReplyDeleteUnfortunately the damage has been done with the COOPERATE study. I see rampant use of ACE- and ARB combination not only among nephrologists but also among primary care physicians.
ReplyDeleteNow with the availability of direct renin inhibitor I am seeing more patients on triple RAS blocake!
Also we all need to realize that proteinuria is a SURROGATE end point. It really is of no value to the patient if his or her proteinuria is under great control but they progress to ESRD with this combination.
ONTARGET has been widely touted as having "disproven" dual blockade on grounds of mortality and ESRD, but has NOT IN FACT done so, especially for the very population (proteinuric nephropathies) for whom this is most relevant.
ReplyDeleteIn ONTARGET "ESRD" is conflated through a confusing composite endpoint including *acute* dialysis. Further, the mortatility signal is associated with pre-established, advanced cardiovascular disease, i.e, likelihood of greater incidence of hypotension, and CHF.
Despite these confounders, ONTARGET makes a sweeping conclusion to dismiss *ALL* of dual-blockade based on incidents occuring in a mere .4% (POINT-FOUR) of the study population.
For the particular sub-population, dual blockade indeed presents additional risk. However, for *most*, it is perhaps at worst wasted effort. But--for proteinuric/hypertensive nephropathies which are inadequately responsive to single blockade, the concept still has merit until proven otherwise by a CLEARLY designed study whose outcomes can be accurately interpreted. ONTARGET is not that study.
doi:10.1038/nrneph/2009.109
Some will continue to progress despite controlling proteinuria, but FEWER progress, and fewer progress rapidly. Especially in non-diabetics. So . . . if a strategy doesn't *always* work, that means it NEVER works --so don't even bother to make the attempt? Faulty logic. The body of evidence supporting proteinuria reduction as generally renoprotective is solid. The challenge and responsibility is to do so safely.
ReplyDeleteWell, if in ONTARGET dual-therapy did not actually *achieve* superior albuminuria or HTN reductions VERSUS the mono-therapy arms, should we be surprised at such meaningless results?
ReplyDeleteThus, efforts to tweak statistical power through composite endpoints have merely served to achieve significance by which we "prove" a useless result. What a mess.
doi: 10.1161/CIRCULATIONAHA.111.01838