Comparative cardiovascular effects of thiazolidinediones: systematic review and meta-analysis of observational studiesBMJ 2011; 342 doi: https://doi.org/10.1136/bmj.d1309 (Published 17 March 2011) Cite this as: BMJ 2011;342:d1309
All rapid responses
Corresponding author: Prof .Dr. Jogenananda Pramanik MD Senior
Associate Dean, Allianze university College of Medical Sciences, Kepala
Batas-13200, Pulau Penang, Malaysia. E-mail:
We applaud the insightful research review and meta-analysis of Yoon K
L et al,(1) and appreciate the effort of the authors to explore scientific
basis of the adverse cardiac events occurring due to indiscriminate use of
synthetic PPAR gamma agonists for glycemic control in type 2 diabetic
patients without monitoring potentially hazardous consequences.
The furore rose worldwide shortly after introduction of first generation
of synthetic PPAR gamma agonists receiving series of reports from the
clinicians about the hepatotoxic adverse effects of these drugs, and
therefore, what we urgently need is independent post-marketing
surveillance (BMJ 2010; 340:c1922) to put a check on this kind of menace
Our past experience says that a series of potentially hepatotoxic drugs
like bromfenac, trovafloxacin, ebrotidine, nimesulide, nefazodone,
ximelagatran and pemoline etc (http://en.wikipedia.org/wiki/Bromfenac)
were introduced in the market prematurely and were withdrawn in a short
while, when clinicians started reporting about Drug Induced Liver Injury
(DILI) cases in different medical journals and conferences.
Contrary to our concerns about the class effects of thiazolidinediones
(2), we noted that recent regulatory actions taken by the US Food and Drug
administration (FDA) and European Medicines Agency (EMEA) to restrict the
use of rosiglitazone ( second generation thiazolinedinedione ) was on the
basis of its cardiac risks only, while hepatotoxicity was not considered
as an issue during these deliberation (3).Therefore, we emphasized here
about the importance of synthetic PPAR gamma agonist induced
hepatocellular injury, related mechanisms of cytotoxicity and availability
of toxicogenomic tools for independent research investigations.
Thiozolindinediones are synthetic PPAR gamma agonists with
thiozolindinedione ring as a common structural characteristic.
Citiglitazone, the first PPAR gamma agonist that was put into clinical
trials, was not marketed because of liver toxicity. Troglitazone, the
first PPAR gamma agonist approved by the FDA for the treatment of type 2
diabetes in 1997,was found to be hepatotoxic and was withdrawn in March
2000 after receiving several reports of severe liver failure causing death
(4). Although less frequent and less severe incidences of hepatotoxicity
was observed with rosiglitazone and pioglitazone compared to troglitazone,
hepatic failure in patients taking rosiglitazone and pioglitazone have
been reported (5). However, due to the relative lack of reliable
biomarkers to accurately predict liver injury prior to its occurrence in
the clinical trial and post marketing setting, majority of the clinicians
remain dependent on ALT (Alanine aminotransferase) and other liver
associated enzymes (LAEs) like aspartate aminotransaminase (AST), alkaline
phosphatase (ALP), gamma glutamyl transferase (GGT) etc to identify
patients in whom injury has already occurred (3).
It is also a matter of our major concern that a variety of synthetic
analogs targeting the PPAR isomers and their modulators are under
development at a rapid space in several drug industries, and therefore, it
is of great importance to investigate the mechanisms of toxicities of
various PPAR-gamma agonists in order to safeguard the consumers from life
In recent years, toxicogenomics is emerging as a new field of research
that applies "Omics" technologies in toxicological research which permits
correlation between toxicity endpoints and changes in gene expression
profiles (5). Microarray is a powerful tool for such research in
toxicogenomics that can be used to simultaneously examine the expression
levels of thousands of genes in a sample. Microarray analysis can be
applied to cluster toxins based on similarities in gene expression of
similar mechanisms of toxicity. It can be potentially used as a tool to
determine the degree of toxicity, elucidate the mechanism of toxicity and
show high impact during drug safety evaluation (6).
The mechanism whereby troglitazone caused severe hepatotoxicity continues
to be explored, despite its withdrawal from the market 10 years ago.
Recent studies suggested that reactive metabolites play a role mediated by
increased mitochondrial permeability transition, inhibition of the bile
salt export pump and possibly related to the sulphur moiety of the
thiazlolinedinedione ring which induces apoptosis via mRNA damage (3).
However, some other researchers reported that rosiglitazone does not
necessarily share the hepatotoxic profile of troglitazone (Diabetes
metabolism 2001,27,305-313). Meanwhile, a critical analysis by Martyn T.
Smith revealed that one of the major reasons for missing the hepatotoxic
responses to PPAR gamma agonists may be that the animals employed in the
toxicity studies were physiologically normal whereas the humans using the
drug were diabetic and usually obese (Chemical research in toxicology.,
vol. 16, No.6) Therefore, the disease control groups should also be
included in pre-clinical studies and post marketing surveillances.
Moreover, in contrast to mouse in which only two isoforms of PPAR-gamma
have been described so far, while in human three isoforms of PPAR-gamma
have been identified ( Diabetologia 1999,42:1033-1049)
One of our pre-clinical disease control studies on rodent models
demonstrated that biliary obstructive jaundice in rodents causes down
regulation of PPAR gamma expression which in turn contributes to enhanced
susceptibility to endotoxin-induced hepatotoxicity and rosiglitazone does
not show any protective effect when administered in jaundiced rat models
with endotoximia (  and Xin Lv et al., unpublished experimental
In this connection, it is important to mention that the patients who
developed hepatic dysfunction while taking one of the thiazolinedinedione
products should not be rechallenged with any other drug of this class (5)
to avoid cumulative adverse effects.
At the end of the long continued controversy, FDA responded with an
announcement for a strict regulation to curve indiscriminate sale of
rosiglitazone, which states that from November 18, 2011 the federal
government of USA will not allow rosiglitazone (Avandia) to be sold
without a prescription from certified doctors. Patients will be required
to be informed of the risks associated with the use of rosiglitazone
(Avandia), and the drug will be required to be purchased by mail order
through specified pharmacies (Most Popular E-mail Newsletter. USA Today.
Hence, we concluded that when a pharmocotherapeutic product is left in the
market imposing such a strong restriction from the regulatory bodies like
FDA, it is imperative to continue active exploratory research on these
products with narrow therapeutic index. In the current situation, we may
use all available research tools like tissue selective knockout mice, gene
microarray analysis, proteonomic techniques etc, to carefully monitor the
full spectrum of gene expression and protein effects of more selective
compounds of this class until we obtain detail insight into the roles of
PPAR agonists in individual cell types for complete understanding of their
potential utility and pitfalls. We wish if we all can actively participate
in global collaboration for pharmacovigilance (Pharmacovigilance. Mann RD,
Andrews EB, eds. John Wiley & Sons Ltd, Chichester, 2002).
1.Yoon,K. L.,Chun S. K.,and Singh,S.: Research: Comparative cardiovascular
effects of thiazolidinediones: systematic review and meta-analysis of
observational studies ;BMJ 2011;342:doi:10.1136/bmj.d1309 (Published 17
2.Tolman K.G.,Thiazolinedinedione hepatotoxicity: a class effect? Int
J. Clin. Pract Suppl.2000 Oct; (113) : 29-34.
3.Drug-induced liver injury: a summary of recent advances: Expert opinion,
Drug Metabolism, Toxicology (2011) 7 (7): 875-890(Published by Yale
University 20th June 2011)
4.Marcy TR, Britton ML, Blevin SM. Second-generation Thiazolidinediones
and Hepatotoxicity: A Case Report and Review of Current Literature. Annals
of Pharmacotherapy. 2004;38:1419-23.
5.Guo, Lei, Zhang, et al.: Differences in hepatotoxicity and gene
expression profiles by anti-diabetic PPAR gamma agonists on rat primary
hepatocytes and human HepG2 cells: Molecular Diversity, Vol. 10, No. 3.
(August 2006), pp. 349-360.
6. Thomas, R.S., Rank, D.R., et al.: Identification of
toxicologically predictive genes sets using cDNA microarrays. Mol.
Pharmaco., 60(2001) 1189-1194.
7..Mian Zhou, Rongqian Wu, Weifeng Dong, Endotoxin downregulates
peroxisome proliferator-activated receptor-gamma via the increase in TNF-
alpha release. Am J Physiol Regul Integr Comp Physiol 294: R84-R92, 2008
1. This review work was supported by China Postdoctoral Science Foundation?Grant No. 20080440626)awarded to the first author Dr.Xin Lv MD.
2. We sincerely acknowledge the critical comments received from Dr.Alan T Remaley, NIH, Bethesda, USA and other collaborative researchers.
3. Prof.Dr.J.Pramanik,the corresponding author, is a member of the research committee of Key Open Laboratory for tissue regeneration of Henan Universities at Xinxiang Medical University, Xinxiang City, P.R.China.
Competing interests: No competing interests
Dr. Heazlewood describes the modest evidence for pioglitazone's
beneficial effect on a composite cardiovascular endpoint, but fails to
mention the inconclusive data for pioglitazone on specific individual
endpoints such as myocardial infarction (RR 0.81, 95% CI 0.64 - 1.02,
p=0.08), or death (RR 0.92, 95% 0.76 - 1.11, p =0.38), when compared to
non-thiazolidinedione therapies or placebo. In our study, the
unfavourable findings on rosiglitazone as compared to pioglitazone could
stem either from pioglitazone being relatively less harmful or neutral in
its effects, or from some yet unproven benefit for pioglitazone (when
compared to non-thiazolidinedione therapies) in reducing myocardial
infarction or death. Given the absence of compelling data for such
benefit with pioglitazone, we should avoid speculative conclusions on the
virtues of pioglitazone. Interpretation of the pioglitazone meta-analysis
is hampered by the fact that the pharmaceutical company trial data was
made available to only to a single group of investigators,  and some
of the data are still not in the public domain, thus precluding any
replication of the findings. The arguments for the efficacy of
pioglitazone are driven mainly by the findings of a trial (ProActive)
where a new composite outcome had been introduced during the study
period. Finally, the evidence of impact on the surrogate outcome of
coronary atheroma volume should not be construed as compelling evidence of
benefit relating to patient-centred outcomes.
Given these uncertainties, we are not persuaded that pioglitazone has
significantly greater beneficial effects than other treatments for type II
diabetes mellitus, although we do agree that pioglitazone is likely to be
relatively less harmful than rosiglitazone.
1.Lincoff AM, Wolski K, Nicholls SJ, et al. Pioglitazone and risk of
cardiovascular events in patients with type 2 diabetes mellitus: a meta-
analysis of randomized trials. JAMA 2007; 298(10): 1180-1188
2.Guillausseau PJ. PROactive study. Lancet. 2006;367(9504):23
3.Nissen SE, Nicholls SJ, Wolski K, et al. Comparison of pioglitazone
vs glimepiride on progression of coronary atherosclerosis in patients with
type 2 diabetes: the PERISCOPE randomized controlled trial. JAMA 2008;
Competing interests: No competing interests
Loke et al cursorily entertain the alternative postulate that
pioglitazone yields cardiovascular benefits, but readily dismiss it.
However, this is an equally plausible explanation of the analysis, given
that it is a comparison between pioglitazone and rosiglitazone rather than
placebo or control therapy. There is moderate evidence that the composite
endpoint of death, myocardial infarction and stroke is reduced by
pioglitazone in high risk vascular patients (PROactive study HR 0.84 CI
0.72-0.98) (1) and lower risk patients (HR 0.75 CI 0.55-1.02) which, when
combined (n=16390 patients) results in an 18% reduction (HR 0.82 CI 0.72-
0.94, p=0.005) (2). Furthermore, the reduced rate of progression of
coronary atherosclerosis associated with pioglitazone, compared with
glimepiride, in the PERISCOPE study, is also compelling evidence for its
beneficial cardiovascular effects (3).
The order of magnitude of rosiglitazone's harmful vascular effect
compared to pioglitazone (14-22%), as assessed by Loke et al, is similar
to that of pioglitazone's beneficial effect compared to control therapy.
It therefore seems unreasonable to unilaterally interpret the findings
negatively towards rosiglitazone where they could equally be interpreted
positively for pioglitazone, avoiding further demonization of
rosiglitazone. It is nevertheless appreciated that class adverse effects
of the PPAR gamma agonists, such as non-fatal congestive heart failure,
apply generally to all agents in this class of drugs.
1. Dormandy JA, Charbonnel B, Eckland DJA, et al. Secondary
prevention of macrovascular events in patients with type 2 diabetes in the
PROactive Study (PROspective pioglitAzone Clinical Trial in macroVascular
Events): a randomised controlled trial. Lancet 2005; 366(9493): 1279-
2. Lincoff AM, Wolski K, Nicholls SJ, et al. Pioglitazone and risk
of cardiovascular events in patients with type 2 diabetes mellitus: a
meta-analysis of randomized trials. JAMA 2007; 298(10): 1180-1188
3. Nissen SE, Nicholls SJ, Wolski K, et al. Comparison of
pioglitazone vs glimepiride on progression of coronary atherosclerosis in
patients with type 2 diabetes: the PERISCOPE randomized controlled trial.
JAMA 2008; 299(13): 1561-1573
Competing interests: No competing interests