Congestive heart failure and cardiovascular death in patients with

prediabetes and type 2 diabetes given thiazolidinediones: a meta-

analysis of randomised clinical trials

Background
The overall clinical beneIit oI thiazolidinediones (TZDs) as a treatment Ior hyperglycaemia can
be diIIicult to assess because oI the risk oI congestive heart Iailure due to TZD-related Iluid
retention. Since prediabetic and diabetic patients are at high cardiovascular risk, the outcome and
natural history oI such risks need to be better understood. We aimed to examine the risk oI
congestive heart Iailure and oI cardiac death in patients given TZDs.
Methods
We used a search strategy to identiIy 3048 studies. 3041 were excluded, and we did a systematic
review and meta-analysis oI the seven remaining randomised double-blind clinical trials oI drug-
related congestive heart Iailure in patients given TZDs (either rosiglitazone or pioglitazone). We
calculated pooled random-eIIects estimates oI the risk ratios Ior development oI congestive heart
Iailure in patients given TZDs compared with controls. The main outcome measures were
development oI congestive heart Iailure and the risk oI cardiovascular death.
Findings
360 oI 20 191 patients who had either prediabetes or type 2 diabetes had congestive heart Iailure
events (214 with TZDs and 146 with comparators). Results showed no heterogeneity oI eIIects
across studies (2228; p Ior interaction026), which indicated a class eIIect Ior TZDs.
Compared with controls, patients given TZDs had increased risk Ior development oI congestive
heart Iailure across a wide background oI cardiac risk (relative risk |RR| 172, 95 CI 121
242, p0002). By contrast, the risk oI cardiovascular death was not increased with either oI the
two TZDs (093, 067129, p068).
nterpretation
Congestive heart Iailure in patients given TZDs might not carry the risk that is usually associated
with congestive heart Iailure which is caused by progressive systolic or diastolic dysIunction oI
the leIt ventricle. Longer Iollow-up and better characterisation oI such patients is needed to
determine the eIIect oI TZDs on overall cardiovascular outcome.










Pathogenesis of sudden unexpected death in a clinical trial of
patients with myocardial infarction and left ventricular dysfunction,
heart failure, or both.
!ouleur AC, Barkoudah E, Uno H, Skali H, Finn !', ZelenkoIske SL, Belenkov YN, Mareev ',
'elazquez EJ, Rouleau JL, Maggioni A!, Kober L, CaliII RM, McMurray JJ, !IeIIer MA,
Solomon SD; 'ALIANT Investigators.
Cardiovascular Division, Brigham and Women's Hospital, Boston, MA 02115, USA.
Abstract
BACKGROUND: The Irequency oI sudden unexpected death is highest in the early post-
myocardial inIarction (MI) period; nevertheless, 2 recent trials showed no improvement in
mortality with early placement oI an implantable cardioverter-deIibrillator aIter MI.
METHODS AND RESULTS: To better understand the pathophysiological events that lead to
sudden death aIter MI, we assessed autopsy records in a series oI cases classiIied as sudden death
events in patients Irom the 'ALsartan In Acute myocardial inIarctioN Trial ('ALIANT).
Autopsy records were available in 398 cases (14 oI deaths). We determined that 105 patients
had clinical circumstances consistent with sudden death. On the basis oI the autopsy Iindings, we
assessed the probable cause oI sudden death and evaluated how these causes varied with time
aIter MI. OI 105 deaths considered sudden on clinical grounds, autopsy suggested the Iollowing
causes: 3 index MIs in the Iirst 7 days (2.9); 28 recurrent MIs (26.6); 13 cardiac ruptures
(12.4); 4 pump Iailures (3.8); 2 other cardiovascular causes (stroke or pulmonary embolism;
1.9); and 1 noncardiovascular cause (1). FiIty-Iour cases (51.4) had no acute speciIic
autopsy evidence other than the index MI and were thus presumed arrhythmic. The percentage oI
sudden death due to recurrent MI or rupture was highest in the Iirst month aIter the index MI. By
contrast, aIter 3 months, the percentage oI presumed arrhythmic death was higher than recurrent
MI or rupture (chi(2)23.3, !0.0001).
CONCLUSIONS: Recurrent MI or cardiac rupture accounts Ior a high proportion oI sudden
death in the early period aIter acute MI, whereas arrhythmic death may be more likely
subsequently. These Iindings may help explain the lack oI beneIit oI early implantable
cardioverter-deIibrillator therapy.