Yes, by Sr. Steven E. Nissen
PPAR Agonists should continue to play an important role in
clinical practice
-
Topics for Discussion
o
PPARs/TZDs are NOT a “drug class” in the
conventional sense, each has a unique pharmacogenetic profile
o
For some PPARs, the balance of benefits vs.
risks is adverse, while others are favorable
§
The Good, the Bad and the Ugly
o
PPARs play an essential role in controlling
blood sugar and lipids in some patients
o
PPAR agonists are unsurpassed in their
durability as oral glucose-lowering agents
o
PPAR agents have unique anti-inflammatory
properties
o
PPARs are the only diabetes drugs proven to slow
progression of atherosclerosis
o
Some PPARs reduce cardiovascular events
o
PPARs improve non-alcoholic fatty liver disease
o
Other diabetes drugs also have limitations/risks
o
Emerging PPARs have promising properties that
may yield important clinical benefits
-
Gene expression with PPAR gamma agents are
different for rosiglitazone, pioglitazone and troglitazone
o
Rosiglitazone activates matrix
metalloproteinase-3 that is involved in plaque rupture and increases levels of
LpPLA2 à
CV events; pioglitazone and troglitazone does not share the same gene
-
The Bad
o
Muraglitazar, the first PPAR agonist lowered
A1c, improved lipids, lower Tg, raised HDL, neutral LDL, but increased DEATH
(all cause mortality) nonfatal MI, or stroke
o
Muraglitazar never got approved
o
Learning point – not all PPARs are the same
o
Rosiglitazone (Dr. Buse’s concern?)
§
1999 Rosiglitazone Advisory Panel à born bad
·
adverse lipid effects, increase LDL by 18.6%
§
2004-2005 – overall, HR=1.31 for ischemic
cardiovascular events à
FDA agency or company did not tell the rest of the world
§
Rosiglitazone was born bad and should not have
been approved in the first place
§
OR for MI = 1.43
§
September 23, 2012 – Pharmageddon for
Rosiglitazone
·
European regulators ban the drug entirely, while
FDA proposes restricting use to patients who have failed all other diabetes
drugs
-
Pioglitazone
o
Lowers Tg, slight increase in non-HDL cholesterol
o
Very clear from the beginning that rosiglitazone
was different than pioglitazone
o
Proactive: Pioglitazone MACE Outcomes, Lancet
2005;366:1279
§
Major Adverse Cardiovascular Events (MACE) – 16%
RRR, does not meet regulatory standard, but gives us an informed data
indicating that pioglitazone reduced death, MI, stroke – HR=0.82 (0.72-0.94),
p=0.005, Lincoff 2007 pioglitazone meta-analysis
§
2010 meta-analysis: CV death, MI or stroke,
HR=0.80 (0.70-0.93, p=0.003) compared to comparators
o
Pioglitazone vs. Rosiglitazone – Forest Plots
§
Consistent HR for MI
·
Pioglitazone < 1.0, does not cross 1.0
·
Rosiglitazone – FDA, Point estimate = 1.80
o
Pioglitazone vs. Glimepiride
§
Mean posterior wall carotid intimal medial
thickness – lower with pioglitazone vs. glimepiride
§
Coronary atherosclerotic plaque volume, less
than baseline (anti-atherosclerotic effect), glimepiride is significantly
higher than baseline (PERISCOPE trial)
§
Lowers C-reactive protein, increase HDL, neutral
LDL, lower Tg (all biomarkers)
o
Pioglitazone or Vitamin E in NAFLD
(non-alcoholic fatty liver disease) – N Engl J Med 2010, 362:1675-85
§
In liver disease – pioglitazone may be
beneficial in patients with liver disease
o
Durability of Glucose-Lowering Efficacy
§
Chicago trial – glimepiride superior than
pioglitazone at 16 weeks, but by 72 weeks, pioglitazone was better than
glimepiride (which went back to baseline)
§
ADOPT trial – rosiglitazone better than
metformin, which is better than glyburide – 5 year duration of trial
§
PPAR gamma agonists can control glucose better
than other agents
o
Cancer
§
PROACTIVE
·
14 in pioglitazone vs 6 placebo (Bladder cancer)
& 3 breast cancer in pioglitazone group vs. 11 in placebo
§
retrospective cohort studies are all over the
map
§
largest study in Kaiser group à HR=0.98
§
extremely weak evidence
-
Safety issues with PPAR agonists
o
Heart failure à
avoid usage in patients with depressed EF
o
Fractures à
avoid using in elderly women at high risk of osteoporosis
o
Weight gain à
lifestyle interventions, counseling
-
Pioglitazone weight of evidence
o
Favourable
§
Reduces insulin resistance
§
Durable glucose lowering
§
Improved lipids
§
Anti-inflammatory
§
May reduce CV events
§
Improved NAFLD
o
Unfavourable
§
Weight gain
§
Congestive heart failure
§
Bone fracture
§
Bladder cancer?
All Diabetes Drugs have Limitations
|
Class
|
Limitations
|
Safety Concerns
|
|
Biguanides
|
GI A/E
|
Lactic acidosis
|
|
Sulfonylureas
|
Poor durability
|
Hypoglycemia, weight gain
|
|
DPP-4 inhibitors
|
Limited efficacy, particularly in patients close to goal
|
Pancreatitis?
|
|
Alpha-glucosidase inhibitors
|
Severe intestinal discomfort
|
Elevation of transaminases
|
|
GLP-1 agonists
|
Injectable only
|
Pancreatitis with some agents
|
Bottomline: All
agents have limitations, no perfect drug
Future
-
Aleglitazar, Dual PPAR alpha/gamma agonist
o
SYNCHRONY – Phase 2 trial
§
Looks about the same in efficacy as pioglitazone
§
Same favorable pattern as pioglitazone, may be
more favorable
§
Lower Tg by 30%, HDL increase by 25%, decrease
LDL by 10%
Summary
-
PPAR gamma agonists are not a class of drugs,
each agent has unique properties
-
PPAR agonists have important safety concerns,
but so do other diabetes drugs
-
Some PPAR agonists have favorable CV effects
-
We need a broad range of drugs to control blood
sugar in patients with diabetes
-
It is nihilistic to indict an entire approach to
treatment because of a couple of bad drugs (if Glaxo had stopped rosiglitazone
to be made, it would have solved a lot of problems)
Conclusions
-
Because of safety concerns, currently available
TZDs should not be used as first line agetns, but pioglitazone remains useful
in many patients not controlled on metformin
-
Future PPARs may find broader usage,
particularly if development efforts can uncouple efficacy from adverse effects
-
We still need PPAR agonists in the therapeutic
armamentarium
No, by Dr. George Grunberger
From the perspective of the prescriber…
Meet John S.
-
52 yo engineer
-
T2DM x 2 years (on metformin 500mg bid),
otherwise in good health (on low-dose ASA and a multivitamin)
-
Non-smoker, occasionally exercises (stationary
bike) on weekends, no specific diet, feels tired at the end of the day, had
normal stress test 2 years ago
-
Ht: 172 cm, wt: 89 kg, BMI: 30.1, WC: 41”, BP:
148/88
-
Random labs: normal electrolytes and LFTs,
glucose 184 mg/dl, creatinine 1.3 mg/dl, HDL-C 34, LDL-C 135, Tg 212 mg/dL,
microalbuminuria
-
HgbA1c: 7.9%
Decision making
-
Is TZD an option? Risk and benefits? Alternative Rx?
Promise of Thiazolidinediones
-
New mechanism of action (“insulin sensitizers”)
-
Great glucose reduction
-
Additional cardiovascular benefits (as surmised
from all the favorable direction of cardiovascular risk factor markers)
-
True “designer drugs”
Troglitazone (Rezulin®) background
-
banned in Great Britain in December 1997 – more
deaths
-
in 2000, FDA review of safety data: Rezulin is
more toxic to the liver than rosiglitazone and pioglitazone
-
Rosiglitazone and pioglitazone, both approved in
1999, offer the same benefits as Rezulin, without the same risk
-
FDA – withdrew troglitazone in 2000 (3 years
after Great Britain)
The story of Avandia
-
May 2, 2007, Dr. Nissen reports increased heart
risks for patients taking Avandia
-
(NEJM meta-analysis)
-
Avandia Harm
o
May 2007: use of rosiglitazone associated with a
significantly increased risk of MI (odds ratio = 1.43, CI 1.03-1.98, p=0.03)
and the risk of death from CV causes (odds ratio – 1.64) (Nissen & Wolski, NEJM)
o
FDA: 83,000 heart attacks from 1999 to 2007
linked to rosiglitazone
o
2009: RECORD study, an open label trial; no
increase in cardiovascular hospitalization or death with rosiglitazone compared
to metformin plus sulfonylurea, but the rate of CHF causing admission to
hospital or death was significantly increased
o
Retrospective study of 227,571 elderly American
patients, comparing rosiglitazone to pioglitazone
§
Rosiglitazone associated with “an increased risk
of stroke, heart failure, and all-cause mortality and an increased risk of the
composite AMI, stroke, heart failure, or all-cause mortality in patients 65
years or older
o
March 2011: Meta-analysis of 16 observational
studies (810,000 patients): rosiglitazone is associated with a higher risk of
heart failure, MI and death than pioglitazone (Lake et al, BMJ)
-
Can Avandia be Saved?
o
Yes, it will be vindicated in head-to-head trial
with Actos (“TIDE”)
§
By 2015, GSK is expected to release results of
this large trial that was requested by the FDS
§
This study will test the cardiovascular effects
of long-term treatment with rosiglitazone or pioglitazone when used as part of
standard of care compared to similar standard of care without rosiglitazone or
pioglitazone in patients with T2DM who have a history of or are at risk for
cardiovascular disease
o
No
§
This study has been terminated
§
FDA has placed the trial on full clinical hold
§
It is considered unethical to recruit given the
cloud over rosiglitazone (even though the study was designed in order to find
if those putative risks exist)
§
“After reading these documents, we would like to
know what stps the FDA has taken to protect patients in the TIDE trial, and why
this trial is allowed to continue.” Senators Baucus & Grassley
o
So, we will never find out!
-
The Fate of Avandia
o
Banned in Europe
o
US FDA – restricted access
§
Patients already being successfully treated with
these medicine (sign consent form)
§
Patients whose blood sugar cannot be controlled
with other anti-diabetic medicine and who, after consulting with their
healthcare provider, do not wish to use pioglitazone-containing medicines
-
How Bad is Avandia
o
In the Graham, 2010 study of Medicare recipients
(n=227,571) who started treatment with Avandia or Actos between July 2006 and
June 2009 (average age 75 years)
§
Avandia showed:
·
27% higher risk of stroke
·
25% higher risk of heart failure
·
14% higher risk of death
§
For every 60 older patients who take Avandia
instead of Actos for one year, there would be one extra heart attack, heart
failure, stroke or death
§
Avandia is much less safe than Actos
Actos
-
PROactive Study results
o
Primary endpoint NOT met
o
Secondary endpoints were not stated in the
design of the study; raising a concern that the secondary endpoints may have
been “cherry-picked” to obtain positive results. The
use of secondary endpoint in the presence of a negative primary endpoint has
been severely criticized
o
While there were 58 fewer primary end-point
events, there were 115 more incident congestive heart failure events in
patients treated with pioglitazone
o
4-fold increase in bladder cancers (16 vs. 4)
and more than 2-fold increase in fractures (5.1% vs. 2.5%)
-
Actos and Bladder Cancer
o
French National Health Insurance Plan
§
1.5 million patients followed for up to 4 years
(2006-2009)
§
Concluded statistically significantly more bladder
cancer in pioglitazone group compared to non-pioglitazone group
o
US: A 5-year interim analysis (1997-2008)
§
The risk of bladder cancer increased with
increasing dose and duration of pioglitazone use
§
Diabetes Care 2001;34:916
o
FDA Adverse Event Reporting System
§
Duration of therapy > 12 months was
associated with a 27.5 excess cases of bladder cancer per 100,000 person-years
follow-up, compared to never use of pioglitazone
o
UK general practice Health Improvement Network
(2000-2010): comparing tZDs (n=60 cases in 18,459) vs. SUs (n=137 cases in
41,396)
§
Risk of bladder cancer increased with time on
TZDs (but not with SUs)
·
3-4 years: HR: 1.15
·
4-5 years: HR: 1.40
·
> 5 years: HR: 1.72
§
Same increase for both PIO and RSG
§
Mamtani et al, ASCO 4 June 2012, abstract 1503
o
UK general practice research database
(1988-2009)
§
N=115,727 patients prescribed new antidiabetic
agent, nested case control study
·
376 new cases of bladder cancer matched with
6699 controls
·
use of pioglitazone incrased rate by 83%
(1.10-3.05)
·
Rate increased with duration of use (at > 2
yrs: HR 1.99)
·
And with higher doses (>28,000mg: HR 2.54)
o
Possible mechanism?
§
Unlikely direct interaction with the receptor
§
Changes in urine composition à solids
§
Cytotoxic to the urothelium à increased cell
proliferation
§
Speculation only
o
Dr. Ge’s (Contract physician of drug safety with
Takeda) allegations
§
Takeda knew in the 1990s when Actos was tested
in animals and bladder cancer was detected
§
Yale U. study enrolled more than 40 patients who
were diagnosed with bladder cancer while taking Actos. They were all reported as related to Actos by
the Yale clinical investigator
·
Takeda directed her to change her
assessment
§
More than 100 bladder cancers reported in
Takeda’s ARISg database, but only 72 found in FDA AERS – likely the result of
bladder cancer events getting classified as “not-related” by Takeda
§
Takeda amended the patient enrolment criteria in
the middle of the phase III trials and added the urine cytology test as an
exclusion criteria
§
PROactive trial results – in 2006, Takeda was
required by the FDA to include bladder cancers reported from the PROactive
study in the Actos label. Takeda never
issues a Dear Doctor letter
§
Analysis of the FDA’s AERs found that fully 20%
of the 138 bladder cancers reports for all drugs submitted between 2004 and
2009 were in patients taking Actos
o
How much safer is Actos?
§
Prior to the implementation of the Boxed
Warning, Takeda was reporting all CHF events (including hospitalized and
non-hospitalized cases) as serious events
§
Following the implementation of the Boxed
Warning, Takeda decided to report only hospitalized or fatal CHF events as
serious events, thus substantially reducing the number of CHF events
§
By classifying CHF cases as non-serious, Takeda
was able to publically make Actos appear much safer than it really was
§
Unlike Takeda, GSK continued to report
non-hospitalized cases of CHF as “serious” adverse events for Avandia
§
AERS data: the “Cardiac Failure Acute” reported
for Actos was nine times higher than Avandia: 43/23,532 (0.18%) vs 9/40,084
(0.02%)
§
Probably because acute CHF must be hospitalized
§
Because hospitalized events were automatically
classified as “serious” they could not be overwritten manually in the ARISg
(the software program used by Takeda to generate MedWatch reports)
§
FDA’s meta-analysis of all post-marketing
studies confirmed that CHF was higher in Actos than Avandia
·
CHF 1.2% in Actos vs. 0.9% in Avandia
·
CV death 0.4% in Actos vs. 0.2% in Avandia
·
MI 0.5% in Actos vs. 0.4% in Avandia
·
Serious ischemic event 1.3% in Actos vs. 1.2% in
Avandia
·
Total ischemic event 2.3% in Actos vs. 2.2% in
Avandia
TZDs and Fractures
-
ADOPT (RSG): 1.8 fold increase in women (foot
3.3x, hand 2.6x, prox humerus > 8x) – N Engl J Med 2006;355:2427
-
PIO: 1.9x increase fractures in women –
fda.gov/medwatch/safety/2007
-
TZD associated with more fracture than either SU
or MET (20,964 Medicare patients) – ICEM 2009;94:2792
-
Both PIO and RSG increase fracture risk in women
2.23x – CMAJ 2009;180:32
-
TZD associated with 57% increase (72% if > 65
years) compared with no-TZD use in women not men – JCEM 2010;95:592 (Detroit)
-
TZD use increased hospitalization for fractures
by 76% in women, not men (Taiwan) – Diabetologia 2010;53:489
-
PIO increased fractures by 77% in women (BC) –
Arch Int Med 2009;169:1395
-
Both PIO and RSG increased fracture risk in both
men and women identically by ~40% - Diab Obes Metab 2010;12:716
-
Possible Mechanisms?
o
Decreased osteoblast differentiation and
increased osteoclast formation (P1NP and bone alk phosphatase decrease, CTX-1
increase) – JCEM 2010;95:134
o
Increased sclerostin (negative regulator of bone
formation) and CTX after 24 weeks of pioglitazone in T2DM men (vs. metformin) –
Eur J Endocrinol 2012;166:711
Do you need more?
-
TZDs vs. other antidiabetic drugs (odds ratio):
o
RSG MI OR: 7.86
(7.34-8.34)
o
RSG macular
edema OR: 5.55 (3.94-7.79)
o
PIO multiple
fractures OR: 2.00 (1.70-2.35)
o
RSG/PIO bone
fractures OR: 1.73 (1.53-1.96)
-
Both MI and bone fracture signals appeared
before 2005 (arguing against publication bias) – Drug Safety 2012;35;315
-
21 cases of liver failure (11 rosi, 10
pioglitazone)
How good are the glycemic data?
-
TZDs lower glucose
-
Are they better than alternatives?
o
Efficacy
§
Not really
§
Most anti-diabetic agents have similar glucose
lowering efficacy and A1c reduction
o
Price
§
Not really
o
Prevention of DM?
§
Not really
·
They can mask development by glucose lowering
effect
·
Effect disappears after wash-out
Story of muraglitazar
-
Dual PPAR agonist
-
In 2005, Bristol-Myers Squibb received
regulatory approval for its new generation TZD
-
JAMA 2005;294:2581 – Dr. Nissen et al criticized
muraglitazar for excessive cardiovascular events
-
FDA requested BMS to conduct another study to
further prove muraglitazar’s cardiovascular safety
-
BMD withdrew the approved NDA for muraglitazar,
citing the excessive amount of time and money it would take to accomplish such
a study
RSG is a very potent PPAR gamma 1 and 2 agonist
PIO is a PPAR alpha ligand agonist, weaker PPAR gamma
agonist – which may explain the difference in lipids
TZDs
-
increased weight, increased intravascular
volume, peripheral edema, congestive heart failure
-
lower hemoglogin and hematocrit
-
macular edema
-
increased fracture
Back to John…
-
would you add TZD to his metformin?
-
NO, given the current state of knowledge about
potential long-term hazards, lack of any unique advantage, and the current
practice environment, the answer is NO
