aspirin in low-risk diabetics???

A recent study (the ASCEND trial) found that the benefits of low-dose aspirin in diabetic patients was not much more than the risk of major bleeding (see dm aspirin ASCEND nejm2018 in dropbox, or DOI: 10.1056/NEJMoa1804988​)

Details:

-- 15,480 participants in the UK with diabetes (any type), at least 40 years old, and no known cardiovascular disease

-- mean age 63 (36% <60yo), 63% male, 96% white, BMI 31, 8% current smokers/46% former, 62% hypertension, 36% on aspirin prior to study (but then stopped for the study), 75% statin, median duration of diabetes 7 yrs, SBP 136, 5-yr risk of vasc event: 40% <5% risk/43% 5-10% risk/17% >10% risk

-- randomly assigned to aspirin 100 mg daily vs placebo and followed mean 7.4 years

-- primary outcome was the first serious vascular event (i.e. myocardial infarction, stroke or transient ischemic attack, or death from any vascular cause, excluding any confirmed intracranial hemorrhage).

-- primary safety outcome was the first major bleeding event (i.e. intracranial hemorrhage; sight-threatening bleeding event in the eye; gastrointestinal bleeding; or other serious bleeding leading to hospitalization, transfusion or death).

-- Secondary outcomes included gastrointestinal tract cancer.

-- there was a 70% adherence rate to both the aspirin and the placebo

Results:
--serious vascular events:

    -- 658 (8.5%) with aspirin vs 743 (9.6%) with placebo; a 12% decrease, rate ratio 0.88 (0.79-0.97), p = 0.01

--major bleeding events:

    --314 (4.1%) with aspirin vs 245 (3.2%) with placebo; a 29% increase, rate ratio 1.29 (1.09-1.52), p = 0.003

        -- additional cases of major bleeding with aspirin: 36 cases of gastrointestinal bleeding, 31 of “other major bleeding”, 10 intracranial hemorrhage (though no difference in hemorrhagic stroke), but 7 fewer sight-threatening bleeds

--though there was no clear benefit of continuing aspirin after the first 5 years of the study (ie, the cardiovasc disease incidence curves were then parallel), there was no attenuation of the adverse bleeding events over time [ie, over time, the continued benefit of aspirin will likely be less than the increasing bleeding risk]

--no difference in outcomes in patients considered low, moderate or high vascular risk

--GI tract cancer:

    --157 (2.0%) with aspirin and 158 (2.0%) with placebo [no difference]

-- all cancers

    -- 897 (11.6%) with aspirin and 887 (11.5%) with placebo [no difference]

--long-term follow-up for these outcomes is planned.

Commentary:

--the protective effect of aspirin in diabetics from previous studies is unclear: several well-conducted studies (eg POPADAD, see cad popadad apirin pad dm bmj 2008 in dropbox) did not find any benefit. And this POPADAD study was of patients with established atherosclerotic disease (peripheral artery disease). This lack of benefit in diabetics has been attributed to the increased stickiness of platelets or aspirin resistance, for example. Though this does raise also the question of whether an adequate dose of aspirin was given (see http://gmodestmedblogs.blogspot.com/2018/07/aspirin-one-dose-does-not-fit-all.html ). A difference in aspirin effect by weight was not confirmed in post-hoc analysis of this ASCEND trial, but the vast majority of patients were heavy (only 903 patients, 12%, were <70kg and for them there were 118 cardiovascular events in the aspirin group vs 108 in the placebo group). but a post-hoc analysis may involve biases.

--the 12% cardiovascular risk reduction in this study was about the same as in the 2009 Antithrombotic Trialists’ Collaboration meta-analysis of 95K patients in 6 primary prevention trials. ----As a perspective here, the absolute percentage differences: 1.1 percentage point lower risk of vascular events, but at the cost of 0.9 percentage points higher risk of bleeding events (and, again, the incidence of bleeding events continued to increase throughout the study, but the protective benefits of aspirin leveled off after 5 years)

    --looking at the actual numbers, there were 85 aspirin-preventing events but 69 aspirin-inducing events.

    --and, if one does not consider TIA as a severe vascular event (TIAs were not considered in the original study design, but were added later), the rate ratio was a nonsignificant benefit from aspirin at 8%, RR 0.92 (0.82-1.03). Without TIAs included, there were only 45 vascular events prevented by aspirin but 69 serious adverse events…

    --perhaps related to the prior lack of definitive evidence in aspirin’s favor, the Am Diabetes Assn guidelines state “aspirin therapy (75-162 mg/day) may be considered as a primary prevention strategy in those with type 1 or type 2 diabetes who are at increased cardiovascular risk” (my emphasis added).

--and, it is still pretty striking to me that only 75% of these diabetics were on statins, which might have decreased the numbers of cardiovascular events further in both groups, and perhaps unequally (maybe the group on aspirin would have had less protection by aspirin if they were on a statin???)

--the lack of effect of aspirin on GI cancers is not so surprising: the 30-40% reductions in other trials was found with longer exposure to aspirin, as in follow-up up to 20 yrs later.

So, based on this trial, adding 100mg of aspirin to the typically large medication regimen of diabetics without known cardiovasc disease is a pretty mixed bag: some likely benefit but also some considerable risk (and, if one does not include TIAs as an outcome, a statistically non-significant one at that, the benefit-risk ratio is even worse for aspirin). And, as mentioned above, the POPADAD study found no aspirin benefit even in those with established peripheral arterial disease.

Though, that being said, there may be differential values placed in these outcomes: there were 27 fewer nonfatal strokes with aspirin, though 10 more intracranial bleeds (but no difference in hemorrhagic strokes).  Patients may be much more concerned about a non-fatal stroke than the potential for an upper GI bleed… Also, perhaps optimizing statin therapy might have decreased the benefit from aspirin (including protecting against stroke). On the other hand, perhaps using a weight-based aspirin dosing might have increased the benefit from aspirin. 

And, the potential benefit from aspirin on colorectal cancers is another unknown from this study (others have been pretty consistent on this), since this study was likely too short to show benefit. More data later, on longer follow-up

Bottom line, lots of questions, but the impetus for a pretty toxic drug over the long-term especially in lower-risk diabetic patients is not so clear and it seems that these uncertainties should be discussed with the patients

geoff​

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org

to get access to blogs since 8/15/17:

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just type in a name in the search box and get all the blogs with that name in them

to access older blogs from the BMJ website, from October 2013 until 8/15/17: go to http://blogs.bmj.com/bmjebmspotlight/category/archive/ 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list