Page 1 of 2
`
`AstraZeneca Exhibit 2012
` Mylan v. AstraZeneca
` IPR2015-01340
`
`

`
`
`
`Abstracts from the ADA 60111 Scientific Sessions
`
`(including monotherapy and combination with SU or metformin), the rate
`ofALT levels > 3x the ULN is 0.30 cases per 100 patient years (see Table
`below). This compares to 0.59 cases per 100 patient years for placebo-
`treated patients and 0.73 cases per 100 patient years for SU- or metformin-
`_treated patients. These are similar to the rates seen 1 year prior.
`Rates of ALT Levels > 3x ULN iu the Rosiglitazone Trial Program
`(expressed as cases per 100 patient years)
`SU or Metformin
`Rosiglitazoue*
`Placebo
`0.78
`0.35
`0.59
`November 1998
`0.73
`0.30
`0.59
`November 1999
`*Includes monotherapy and combination with SU or metformin
`I11 addition, rosiglitazone has been prescribed to over 250,000 patients,
`and thus far the clinical trial experience has been predictive of the rosigli-
`tazone safety experience in the marketplace.
`In conclusion,
`the current clinical trial and postmarketing experience
`with rosiglitazone indicate no evidence of troglitazone-like hepatotoxicity.
`
`160-OR
`Treatment with a DPP-IV Inhibitor, NVP-DPP728, Increases
`Prandial Intact GLP-1 Levels and Reduces Glucose Exposure in
`Humans
`PAUL ROTHENBERG," 2 JYOTI KALBAG, HAROLD SMITH,
`RONALD GINGERICH, JERRY NEDELMAN, EDWIN VILL-
`HAUER, JAMES MCLEOD, THOMAS HUGHES, East Hanover;
`NJ,‘ St. Charles, MO
`NVP-DPP728 is a highly selective, orally active inhibitor of dipeptidyl
`peptidase-IV (DPP-IV) designed to augment the glucose-lowering activity
`of endogenously secreted GLP-1. Recent studies have demonstrated that
`NVP-DPP728 prevents N-terrninal degradative inactivation of GLP-1 and
`improves glucose tolerance in insulin-1'esistant rats. The present crossover
`trial evaluated the single dose pharmacodynamics of NVP-DPP728 admin-
`istered to 12 healthy normoglycemic volunteers. Afier an overnight fast,
`subjects were administered 100 mg NVP-DPP728 or placebo, followed 30
`minutes later by a 1000 kcal solid meal (23g protein, 42g fat, 36g carbo-
`hydrate, standard FDA breakfast). Blood samples were obtained predose
`V and for up to 24 hours afier each dose for analysis of plasma glucose and
`insulin levels and also for active, undegraded GLP-1 levels by direct
`ELISA (Linco Research). NVP-DPP728 increased peak plasma levels of
`active GLP-1 (l5i2 vs. 9i2 pmol/1, p = 0.018) and also increased active
`GLP-1 prandial exposures, AUCGu,_,(04,,) (28i4’vs. l4i4 pmol.l‘l.h'1, p <
`0.0001). Prandial glucose excursions above baseline were reduced by
`NVP-DPP728 relative to placebo (l2i3 vs. 20i4 mg.dl".h", respectively,
`p = 0.04), while glucose excursions below baseline were unchanged (-
`22i4 vs. -16:4, p=0.3). Insulin excursions were not affected. No clinical-
`ly significant adverse events were observed. In addition, administration of
`NVP-DPP728 to fasting individuals was unaccompanied by-clinically sig-
`nificant changes in plasma glucose levels.
`Thus NVP-DPP728 increased prandial active GLP-1 levels with con-
`comitant reduction in prandial glucose exposure in normal subjects with-
`out causing hypoglycemia. These results provide the first direct clinical
`demonstration that DPP-1V inhibition is a viable new pharmacological
`approach for potentiating endogenous GLP-1 activity, and support the
`investigation of the glucose-lowering potential of NVP-DPP728 for the
`treatment of type 2 diabetics.
`
`decreases of 2 -0.7% for the 5mg or 10mg doses. The overall tolerability
`profile of GI262570 was similar to that observed with other PPAR7 ago-
`nists, with dose-related weight gain, hemoglobin decreases, and peripher-
`al edema. These data suggest that GI262570 will have clinical 11tility in the
`treatment of T2DM.
`5mg 10mg
`lmg 2mg
`P
`Parameter
`
`Baseline (B) FSG (mg/dL)
`201
`205
`208
`204
`206
`FSG A from B at 4 weeks
`+21
`-101
`-20+
`-34+
`-54+
`FSG A from B at 12 weeks
`+22
`-8+
`-281
`-48’
`-66+
`% of subjects with 2 -30 mg/dL FSG A from B 11% 33%‘ 48%+ 74%+ 85%+
`Baseline HbAlc (%) '
`8.1
`7.8
`8.0
`8.1
`8.1
`HbAlc A vs. P at 12 weeks
`— -0.4‘‘‘’
`-0.81 -1.41 -1.9+
`HbAlc A from B at 12 weeks
`+1.1 +0.7 +0.3
`-0.3
`-0.7
`% ofsub'ects with 2 -0.7% HbAlc A from B
`3% 9% 19%‘ 41%‘ 54%+
`Significance level's vs. P: V = < 0.05 * = < 0.005 J“ = < 0.001
`
`158-OR
`Monotherapy with GI262570, a Tyrosine-Based Non-Tliiazolidinedione
`PPAR'yAgonist, Significantly Reduces Triglyceride and Increases HDL-
`C Concentrations in Patients with Type 2 Diabetes Mellitus
`GREG G. WlLSON,"2 MARTHA ABoU—DoN1A,'-2 LUCY FRlTH,1'2
`JAI PATEL,“ 2 FRED T. FIEDOREK,“ 2 STUDY GROUP- PPA20005,‘
`Research Triangle Park, NC,‘ Greenfmd, Middlesex, United Kingdom
`GI262570 is a novel, non-thiazolidinedione, L-tyrosine-based peroxi-
`some proliferator-activated receptor gamma (PPAR7) agonist
`that
`is
`~l000-fold more selective for human PPAR7 compared to the human
`PPAROL isoform. A total of 376 Type 2 diabetes mellitus (T2DM) patients
`were randomized to receive either placebo (P) or one of 5 daily doses of
`GI262570 (0.25, 1, 2, 5, or 10mg) in a 12-week double-blind placebo-con-
`trolled study. Mean baseline fasting triglyceride (TG) and high-density
`lipoprotein cholesterol (HDL-C) concentrations are listed in the table.
`Approximately 60% of T2DM patients had mild fasting hypertriglyc-
`eridemia (HT) defined as TG > 150 mg/dL. Twelve weeks of treatment
`with GI262570 resulted in significant metabolic improvement in fasting
`TG and HDL-C (see table) along with changes in glycemic parameters that
`are described in a separate abstract. The maximum reduction in TG was
`achieved within 4 weeks and the effect was maintained over 12 weeks.
`While metabolic improvement in TG and HDL-C levels \vas observed
`across the range of T2DM patients regardless of baseline TG status, the
`greatest reductions were observed in subjects with mild/moderate HT.
`Apolipoprotein B levels fell by 8-14% for the 5mg and ,l0mg doses and
`there was also a non-significant decrease in LDL-C at these two GI262570
`doses. These findings suggest that GI262570 will enhance overall meta-
`bolic control for T2DM patients by improving the high TG/low HDL-C
`dyslipidemia associated with T2DM as \vell as by improving hyper-
`glycemia.
`10mg
`5mg
`2mg
`1mg
`P
`PARAMETER
`67
`58
`61
`59
`67
`Number of subjects (Intent-to-Treat)
`179
`199
`185
`200
`170
`Baseline (B) TG (mg/dL)
`-30%+ —43%+
`+3% -6% -13%)”
`% TG A from B at 12 weeks
`322
`268
`269
`302
`261
`Baseline TG (mg/dL) HT group
`% TG A from B at 12 wks HT group -7% -18% -19% -44% -53%
`Baseline HDL-C (mg/dL)
`42
`42
`42
`40
`42
`% HDL-C A from B at 12 weeks
`0% +4% +13"/o* +l2%* +l5%+
`Significance levels vs. P: ‘l’ = p < 0.05 * = p < 0.005 J’ = p < 0.001
`
`159-OR
`
`Rosiglitazone Liver Safety Update
`HAROLD E. LEBOVITZ,“ 2 ALAN sALzMAN,‘» 2 Brooklyn, NY;
`Collegeville, PA
`Rosiglitazone is a potent thiazolidinedione for the treatment of type 2
`diabetes. The thiazolidinedione troglitazone has been associated with
`hepatotoxicity, including liver failure and hepatic-related deaths. To date,
`no signal of hepatotoxicity has been seen with rosiglitazone, which has
`been extensively evaluated.
`The incidence in rosiglitazone-treated patients ofALT elevations greater
`than 3x the upper limit of normal (ULN) \vas low and similar to place-
`bo/comparators at the time of FDA filing in November 1998, at which time
`total exposure to rosiglitazone was 3600 patient years. Subsequently, expo-
`sure to rosiglitazone in clinical trials has substantially increased and as of
`November 1999 comprised over 5000 patient years including more than
`1000 patients treated for Z 2 years. For all rosiglitazone-treated patients
`
`A numeral beside an author's name indicates a duality of interest. See page 93.
`
`A39
`
`Page 2 of 2
`
`BMS-Onglyza 007425
`
`Page 2 of 2