metformin in CKD dec CV events/mortality
Geoff A. Modest, M.D.
Tue 3/5/2019 7:22 AM
Another study reinforced the benefit of metformin in those with stage 3 chronic kidney disease (see dm metformin ckd dec death dmobesmetab2019 in dropbox, or doi.org/10.1111/dom.13642).
Details:
-- the study was a subsequent analysis of diabetic patients in the TREAT study (Trial to Reduce Cardiovascular Events with Aranesp: darbepoetin-alfa)
-- there were 591 individuals who used metformin at baseline, and 3447 nonusers
-- mean age 67, 36% male, 65% white/10% black, blood pressure 137/73, BMI 31, median duration of diabetes 15 years, 57% cardiovascular disease/38% coronary disease/23% heart failure/18% MI/10% stroke/17% PVD, 3% current smoker
-- overall metformin users were less likely to be male and more likely to be white and have a shorter duration of diabetes, and tended to have fewer comorbid conditions and less frequent history of heart failure, coronary disease, and peripheral vascular disease [ie, those on metformin were healthier]
-- not much difference in diabetes control at baseline: 6.8% vs 7.0% nonusers, though metformin users were less likely to utilize insulin (29% vs 53%)
-- meds: insulin 29%, sulfonyureas 49%, TZDs 25%; statins 61%, antiplatelet 10%, vitamin K antagonists 7%, ACE/ARB 85%, beta blockers 47%
-- kidney disease (mean eGFR: 39)
-- stage 1- 2 (eGFR > 60 mL/minute): 5% of patients
-- stage 3a (eGFR 45-59): 27%
-- stage 3b (eGFR 30-44): 44%
-- stage 4-5 (eGFR<30): 23%
-- metformin users were less likely to have stage 4-5 CKD
-- researchers used propensity matching between the metformin users and nonusers in order to mathematically adjust for the level of comorbidities etc.
Results:
-- among propensity matched metformin users vs nonusers:
-- all-cause mortality: 51% decreased, HR 0.49 (0.36-0.69), 4.6 vs 8.5/100 patient-years, p<0.001
-- cardiovascular death: 51% decreased, HR 0.49 (0.32-0.74), 2.8 vs 5.2/100 patient-years, p=0.004
-- cardiovascular composite (heart failure hospitalization, MI, stroke, myocardial ischemia, or death): 34% decreased, HR 0.66 (0.51-0.86), 8.6 vs 11.9/100 patient-years, p=0.01
-- kidney disease composite (end-stage renal disease ESRD or death): decreased 23%, HR 0.77 (0.61-0.98), 8.0 vs 10.9/100 patient-years, p=0.02
-- ESRD was more common in metformin users (4.0 vs 3.6/100 patient-years), though not statistically significant
-- metformin users vs nonusers, by CKD stage:
-- stage 1-3:
-- overall mortality: 39% decrease, HR 0.61 (0.44 - 0.85), 3.7 vs 7.6/100 person-years
-- CV death: 41% decrease, HR 0.59 (0.38 - 0.90), 2.3 vs 4.9/100 person-years
-- cardiovascular disease composite: 30% decreased risk, HR 0.70 (0.53 - 0.90), 7.0 vs 13.0/100 person-years
-- kidney disease composite: 30% decrease, HR 0.70 (0.53 - 0.92), 5.5 vs 10.2/100 person-years
-- ESRD: 2.2 vs 3.3/100 person-years
-- stage 4-5 (these were not statistically significant, but all showed a beneficial trend from metformin):
-- overall mortality: 17% decrease, HR 0.83 (0.54 - 1.27), 7.9 vs 10.4/100 person-years
-- CV death: 20% decrease, HR 0.80 (0.46 - 1.39), 4.6 vs 6.4/100 person-years
-- cardiovascular disease composite: nonsignificant 14.0 vs 17.0/100 person-years
-- kidney disease composite: 5% decrease, HR 0.95 (0.70 - 1.29), 17.4 vs 22.9/100 person-years
-- ESRD: 10.8 vs 16.1/100 person-years
-- lactic acidosis: 2 patients taking metformin (0.3%), one of whom had an eGFR of 40.8 and the other 36.2 measured within 1 to 2 months of the lactic acidosis
Commentary:
-- 43% of people with type 2 diabetes have evidence of CKD, increasing to 61% of those over 65 years old
-- this retrospective analysis did find a few substantial outcomes:
-- metformin was associated with significantly lower risk of all-cause mortality, cardiovascular death, and kidney and cardiovascular disease composites
-- metformin was not associated with evident renal harm, such as ESRD or doubling of creatinine levels
-- metformin was most effective in those with stage 1 to 3 CKD (eGFR > 30) and less effective in those with stage 4 to 5 [this is also true for statins, by the way]
-- a large VA study found that sulfonylureas, vs metformin, were associated with a 20-25% increase in ESRD, sustained decreases of eGFR, and in the combined endpoint of ESRD/sustained decrease in eGFR/or death
-- it should be noted that there have been other studies suggesting that there is an increased risk of ESRD in those on metformin, e.g. a large Danish registry study
-- I did contact the author of the above study regarding the specifics of the 2 cases of lactic acidosis reported (eg, were they symptomatic, were there other potential causes of lactic acidosis like infections, what was the dose of metformin). He replied that he would try to find out, though the details of these cases were likely buried in storage. So, not sure what to make of the association with lactic acidosis. of course, the background issue is that metformin’s kissing cousin, phenformin, was used in the US for many years and had huge numbers of cases of almost universally fatal lactic acidosis. The data on metformin has been remarkably free of cases, which is notable since it has been around for past 70 years!! (see blogs listed at end, esp http://gmodestmedblogs.blogspot.com/2015/01/metformin-in-renal-failure.html
-- a relatively recent metformin dosing study found that the following daily doses of metformin all had stable blood levels that were below the accepted upper limit of 5.0 mg/L, and none had hyperlactatemia (see Lalau JD. Diabetes Care 2018; 41(3): 547)
-- CKD stage 3a: 1500 mg
-- CKD stage 3b: 1000 mg
-- CKD stage 4: 500 mg
-- [my personal feeling here is to use even lower doses of metformin when the eGFR <45, since metformin even in quite low doses works well, it is excreted unchanged in the urine (which may be compromised in CKD), and if there are intervening changes in renal function (e.g. dehydration in an older person in the summer) I am not sure the risk of a higher dose is worth it]
-- one important piece of information lacking in the above study is the actual metformin dose used. Did they exceed these doses in general? Were the 2 patients with lactic acidosis on a supratherapeutic dose of metformin?
-- Some of the newer agents do have evident renal protection, but as noted, this may not be much different from those of metformin, given that metformin seems to provide renal protection safely down to an eGFR of 30:
-- empagliflozin was associated with less renal deterioration even when eGFR <45. This was especially true in those not on metformin; 79% of those with eGFR >60 and 60% of those with eGFR <60 were on metformin, though there was no specific breakdown of renal events by metformin use/who that group was, and no comment on dose (see Wanner C. N Engl J Med 2016; 375:323). So, hard to draw full conclusions. Of note, SGLT-2 inhibitors are not recommended if eGFR <45
-- Liraglutide also provided renal protection; their baseline eGFR was 80, though 21% had an eGFR of 30 to 59. Benefit was limited to those with an eGFR >60. no comment on the concommitant use of metformin in this article orin its predecessor on cardiovasc outcomes (see Mann JFE. N Engl J Med 2017; 377: 839). There is no lower limit of renal function for GLP-1 agonist usage
-- The current FDA guidelines are to check in eGFR prior to starting metformin, okay to continue to use it if the eGFR is above 30, but it is not recommended to start metformin if the eGFR is as 30-45.
-- there are evident limitations of this study: it is a retrospective observational study, so unable to attribute causality. we do know that the baseline demographics/illness severity were pretty difflerent between those on metformin vs not (and the former were healthier). they did do propensity matching, but this mathematical tool is not a great substitute for a real RCT (eg, they can only match those items that they have information about)...
So, the bottom line issue is that not only does metformin seem safe in those with stage 3 CKD, but that it still seems to decrease the overall risk of death and cardiovascular events (and metformin is still likely safer and better than sulfonylureas). I do keep bringing up articles on the safety of metformin with CKD because metformin is such a great drug, has been around for so long and not found to have serious adverse effects, and should still be the mainstay primary therapy for diabetics with CKD with eGFR down to the 30 to 45 range. It is an open question whether adding a GLP-1 agonist is beneficial on top of metformin, though this likely is and GLP-1’s can be used without regard to eGFR.
prior blogs on metformin:
http://gmodestmedblogs.blogspot.com/2016/04/fda-changes-metformin-guidelines.html has the new FDA guidelines, which mirror the prior ones from Europe
http://gmodestmedblogs.blogspot.com/2015/06/metformin-ckd-and-death.html of patients with very high creatinine levels (stage 5 ckd, gfr<15), with increase in all-cause mortality but not in the group on <=500mg/d, and no significant increase in lactic acidosis
http://gmodestmedblogs.blogspot.com/2015/01/metformin-in-renal-failure.html for a systematic review of metformin and lactic acidosis
http://gmodestmedblogs.blogspot.com/2014/10/heart-failure-microbiome.html has a section on the effects of metformin on the microbiome, finding that metformin may mediate much of its hypoglycemic effect through bacterial changes
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:
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
Comments
Post a Comment
if you would like to receive the near-daily emails regularly, please email me at gmodest@uphams.org