Pioglitazone vs sulfonylurea as add-on DM therapy

a recent study assessed the risk of cardiovascular events in diabetic patients inadequately controlled with metformin monotherapy, then randomized them to pioglitazone vs a sulfonylurea, finding no overall outcome difference, though pioglitazone conferred a reduction in ischemic cardiac events (see http://dx.doi.org/10.1016/S2213-8587(17)30317-0 ).

Details:

--Multicenter (57 centers in Italy), randomized, pragmatic study of 3028 patients 50-75 yo, with inadequate diabetic control on metformin 2-3g/d. Study from 2008-2014.

--mean age 62, 59% men, BMI 30, 18% smokers, LDL 103/HDL 46, BP 134/80, 11% with prior cardiovascular event (6% prior MI, 1% prior stroke, 3% prior ACS, 7% prior coronary revascularization), 70% on antihypertensives, 57% lipid lowering agents, 40% antiplatelet drugs

--mean diabetes duration 8.5 years, A1c 7.7%

--exclusions: patients with cardiovascular event in prior 6 months, chronic heart failure, serum creatinine >1.5

--randomized to pioglitazone 15-45mg (mean dose 23mg) vs a sulfonylurea (2% were on 5-15mg of glibenclamide with mean dose 7.6 mg, 48% were on 2-6 mg glimepiride with mean dose 2.5 mg, and 50% were on 30-120 mg of gliclazide with mean dose 42mg, decision on which drug per local practice). unblinded trial, but event adjudicators were unaware of treatment used [it is good that only a small number of patients were on glibenclamide, since it has been associated with increased cardiovascular risk]

--primary outcome: composite of 1st occurrence of all-cause death, nonfatal MI, nonfatal stroke, or urgent coronary revascularization

--mean follow-up 57 months, study stopped early because of analysis indicating futility of trial

Results:

--HgbA1c was 7.20 in the pioglitazone group vs 7.34 in the sulfonylurea group; fewer patients had treatment failure (defined as 2 consecutive A1c's >8%) with pioglitazone (13% vs 20%), so fewer had rescue insulin therapy (11% vs 16%)

--primary outcome occurred in:

    --pioglitazone: 105 patients (1.5/100 person-years)

    --sulfonylurea: 108 patients (1.5/100 person-years)

--no difference in outcomes controlling for sex, age, BMI, diabetes duration, A1c, and eGFR

--secondary outcomes: pioglitazone, in on-treatment analysis for combination of sudden death, fatal and nonfatal MI (including silent MI), fatal and nonfatal stroke, major leg amputation above the ankle, any revascularization: a significant 33% decrease, HR 0.67 (0.47-0.96), p=0.03, from 48 cases (3%) vs 79 cases (5%) with sulfonylureas 

--adverse events: 

    ​--hypoglycemia: 10% pioglitazone, 34% sulfonylurea (p<0.0001)

    --moderate weight gain (average <2kg), LDL, waist circumference: no difference, though HDL was significantly higher with pioglitazone

    ​--heart failure, bladder cancer, fractures: no difference

--28% of those on pioglitazone prematurely discontinued therapy, vs 16% on sulfonylurea

Commentary:

--pretty healthy population overall. still probably not on optimal cardiovascular protection, with only half on lipid-lowering agents (and the low rate of heart failure caused by pioglitazone may be related to this being a healthy population explicitly excluding those with chronic heart failure)

--piogligazone was associated with: better, durable A1c control, less hypoglycemia (though more patients on sulfonylureas were also on insulin as a rescue med) and higher HDL levels

--other pioglitazone trials (PROACTICE, see Lancet 2005; 366: 1279; and IRIS, see N Engl J Med. 2106; 374: 1321) have supported pioglitazone's role in decreasing ischemic cardiovascular events, though these studies had patients with higher baseline cardiovascular risk than the current study.

--though there was no statistical benefit for pioglitazone for any endpoint by intention-to-treat analysis, there was a significant 33% decrease in major ischemic cardiovascular events in the on-treatment analysis, and i think this is important given that the overall population in this study was quite healthy and had limited (lower than expected) clinical outcomes overall, and that there was such a high but asymmetric attrition rate where 28% of the pioglitazone vs 16% of sulfonylurea patients stopped taking their assigned drug.  Likely a large component of the pioglitazone withdrawals was that safety concerns about its reported linkage with bladder cancer were raised in 2012, and both France and Germany withdrew it from the market

So, this brings up, yet again, my major concern with using A1c as the target for therapy (though the FDA does accept this surrogate marker as adequate in assessing the value of a diabetic med!!!). the reality is that 80% of diabetics die from cardiovascular causes, and there are clear meds that work well for diabetes yet either increase cardiovascular events (rosiglitazone), probably increase cardiovascular events (sulfonylureas, insulin), or may decrease them (GLP-1 agonists, perhaps SGLT-2 antagonists, likely pioglitazone). So, seems pretty clear to me that diabetes meds should have documented clinical cardiovascular benefit as a major criterion for approval.

And, with an emphasis on cardioprotection in diabetic patients, my bias (as per many blogs over the past few years) is to use GLP-1 agonists as my second-line drug after metformin.  Several patients are really resistant to needles, so I try to make the case that the newer GLP-1 agonists are weekly injections with very small needles and hurt less than fingersticks, and that these are different meds (and probably better) than insulin (many patients seem to be afraid of insulin itself: in many cases they associate their relative being put on insulin as their diabetes got worse and that it was the insulin leading to their morbidity or mortality). In cases where either my logic/promotional efforts are unsuccessful (also, sometimes because of resistance by insurers to allow GLP-1’s as second-line med), I do try pioglitazone next, though its diabetes control benefit pales in comparison to GLP-1’s.

see:

-- http://gmodestmedblogs.blogspot.com/2016/04/diabetes-update-ahrq-and-pioglitazone.html  for the AHRQ analysis of the effectiveness of the different diabetes meds, including macrovascular (eg atherosclerotic) benefits and noting that sulfonylureas have a 50-70% higher risk of cardiovascular mortality than metformin; this blog also reviews the IRIS and PROACTIVE studies

--http://gmodestmedblogs.blogspot.com/2015/12/empagliflozin-good-and-bad.html  , http://gmodestmedblogs.blogspot.com/2016/06/canagliflozin-decreases-macrovasc.html  , and  http://gmodestmedblogs.blogspot.com/2016/05/another-fda-alert-about.html for my concerns about the quality of the SGLT-2 inhibitor studies in suggesting cardiovascular benefit

--http://gmodestmedblogs.blogspot.com/search/label/diabetes​ for the more recent diabetes articles