COMPLICATIONS—HYPOGLYCEMIA COMPLICATIONS—HYPOGLYCEMIA & 390-P The Novel Glucagon Analogue ZP-GA-1 has Superior Physico- Guided Audio Tour: Predictors, Morbidity, and Mortality (Posters: 389-P to chemical Properties while Maintaining the Pharmacokinetic and 395-P), see page 17. Pharmacodynamic Profi le of Native Glucagon & 389-P DITTE RIBER, ANDERS VALEUR, METTE SVENDGAARD, FRANCESCA MACCHI, Benefi t-Risk Relationship between A1C and Hypoglycemia for LISE GIEHM, KELD FOSGERAU, PIA NOERREGAARD, Copenhagen, Denmark Glucagon is used in diabetic patients for treatment of severe episodes of Addition of Exenatide BID to Insulin Glargine in Patients with T2DM POSTERS JARET MALLOY, MING ZHOU, JENNY HAN, San Diego, CA, Hopewell, NJ hypoglycemia. Pharmaceutically, however, native glucagon possesses poor Complications Exenatide BID (ExBID) improves postprandial glucose and A1C with low physicochemical properties, making convenient dosing in a ready-to-use Acute and Chronic hypoglycemia (hypo) risk. Two 30-wk studies evaluated insulin glargine (IG; rescue pen or the development of an artifi cial pancreas diffi cult. Accordingly titrated per algorithm based on fasting glucose [goal <100 mg/dL]) + ExBID. In the development of novel glucagon analogues such as ZP-GA-1 is pursued. Study 1 (N=627), ExBID or insulin lispro (IL) was added to IG + metformin. In Study The solubility of ZP-GA-1 at physiological pH was shown to be >25 mg/mL 2 (N=259), ExBID or placebo was added to IG ± metformin and/or pioglitazone. and thus highly superior to that of native glucagon ~0.2 mg/mL. In addition, IL was titrated based on pre-meal glucose (goal =100 mg/dL with no hypo). We the stability data suggests that ZP-GA-1 is suitable for long term storage as quantifi ed the benefi t-risk relationship between A1C lowering and hypo risk a liquid formulation. The pharmacokinetic (PK) and pharmacodynamic (PD) versus comparators using Poisson regression models to evaluate hypo exposure- properties of ZP-GA-1 and native human glucagon were investigated in dogs. adjusted event rates (EAER) adjusted for the lowest A1C observed over 30 wks. In Animals were administered either subcutaneously (SC, 20 and 120 nmol/ both studies, IG + ExBID signifi cantly reduced A1C and weight. Study 1: With the kg) or intravenously (IV, 75 nmol/kg). Our data demonstrated overall similar same A1C achieved, hypo EAER was 58% less with IG + ExBID vs. IG + IL (risk ratio PK profi les as well as blood glucose (BG) profi les of ZP-GA-1 and glucagon. [RR] 0.42; p<.0001; Figure 1A). IG + ExBID reduced daytime (RR 0.19; p<.0001) and Further, the SC effect of ZP-GA-1 on BG in a rat model of hypoglycemia nocturnal hypo EAER (0.86; p=.076) vs. IG + IL. Overall, being female, lower BMI was investigated (Figure). Both ZP-GA-1 and glucagon restored BG to or longer diabetes duration resulted in signifi cantly higher hypo EAER. Study 2: baseline levels or above in a dose-dependent manner during insulin-induced hypo EAER was slightly lower for IG + ExBID than IG + placebo (RR=0.75; p=.107; hypoglycemia. In conclusion, ZP-GA-1 displays improved physicochemical Figure 1B). IG + ExBID signifi cantly reduced A1C and weight, and the modeled properties while maintaining similar PK and PD profi les compared to native A1C and hypo ben efi t-risk relationship further supported use of ExBID with IG glucagon. for patients with T2DM not at goal with titrated IG. & 391-P Naltrexone for Treatment of Hypoglycemia Unawareness in Type 1 Diabetes: A Randomized Clinical Trial AMIR MOHEET, SILVIA MANGIA, ANJALI KUMAR, NOLA TESFAYE, LYNN E. EBERLY, YUN BAI, ELIZABETH R. SEAQUIST, Minneapolis, MN Hypoglycemia unawareness (HU) is a limiting factor in the treatment of type 1 diabetes (T1D) and is a challenging problem to reverse. Previous studies have suggested that the opioid receptor antagonist naltrexone might be of benefi t in treatment of HU. The objective of this study was to test the hypothesis that naltrexone therapy in subjects with T1D and HU will improve counterregulatory (CR) hormone response, recognition of hypoglycemia (HG) symptoms and increase brain activation during HG in regions involved in the regulation of CR response to HG as measured by cerebral blood fl ow (CBF) using fMRI. We performed a pilot randomized double blind trial of 4 weeks of naltrexone 50 mg bid (n=11) or placebo (n=12) given orally in subjects with T1D and HU. Outcome measures included HG symptom scores, CR hormone levels and CBF measurement obtained during experimental HG in all subjects before and after 4 weeks of intervention. Subjects were defi ned as having HU based on Cox questionnaire. Baseline subject characteristics were similar between the naltrexone (A1C 7.0 ±0.6, Cox score 5.3±0.9) and placebo (A1C 6.5±0.7, Cox score 5.8±0.9) arms. After 4 weeks of therapy with naltrexone or placebo, no signifi cant differences were seen in any outcomes of interest within each group or between the treatment and placebo arms. At the end of study HG symptom scores (10±7 vs. 14±12, P=0.47) and CR hormone response including epinephrine (116±154 vs. 97±99 pg/ml, P=0.75), glucagon (65±40 vs. 47±12 pg/ml, P=0.25) and cortisol (19±7 vs. 20±8 mg/dl, P=0.35) were similar between the 2 groups. No differences were also seen in the brain activation pattern in response to HG between the 2 groups. In this small pilot study, treatment with naltrexone did not improve recognition of HG symptoms or CR hormone response during experimental HG in subjects with T1D and HU. Future studies with larger sample sizes and different dosing regimens will be necessary to fully evaluate the effi cacy of naltrexone as a treatment of HU. Supported By: ADA (7-07-DCS-02) Supported By: Bristol-Myers Squibb/AstraZeneca ADA-Funded Research & Guided Audio Tour poster For author disclosure information, see page A743. A103 COMPLICATIONS—HYPOGLYCEMIA & 392-P & 394-P Recent vs. Remote History of Severe Hypoglycemia (SH) as a Predictors of Hypoglycemia in Insulin-Treated Type 1 (T1) and Type Predictor of Mortality in Type 1 Diabetes (T1D) 2 (T2) Diabetes Patients: Analysis of Self-Reported Hypoglycemic GEORGIA PAMBIANCO, TINA COSTACOU, TREVOR J. ORCHARD, Pittsburgh, PA Events in 10 European Countries A history of SH has been associated with increased mortality in type 2 DOMINGO OROZCO-BELTRAN, PEDRO MEZQUITA-RAYA, WERNER KERN, BERN- diabetes. While in T1D up to 10% of deaths can be attributable to acute HARD KULZER, JORMA LAHTELA, ULRIK PEDERSEN-BJERGAARD, RAIMUND hypoglycemia, the role of a history of SH is unclear. We thus examined WEITGASSER, PETRONELLA GEELHOED-DUIJVESTIJN, CHANTAL MATHIEU, POSTERS Complications data from the ongoing Epidemiology of Diabetes Complications study of HENRIK HOLM JENSEN, ANTONIO RAMIREZ DE ARELLANO, MARIE MARKERT Acute and Chronic childhood onset T1D (mean age 28 and duration 19 years, n=658 at baseline JENSEN, San Juan de Alicante, Spain, Almería, Spain, Ulm, Germany, Bad (1986-88)) to determine if a self reported history of SH in the prior 2 years, Mergentheim, Germany, Tampere, Finland, Hillerød, Denmark, Salzburg, Austria, The is associated with total (n=76), CAD (no renal, n=27), renal (no CAD, n=14) or Hague, Netherlands, Leuven, Belgium, Holte, Denmark, Madrid, Spain, Copenhagen, CAD+renal (n=19) mortality, in 443 participants attending the 1990-92 exam Denmark when these questions were fi rst asked. SH was defi ned as unconsciousness, Unpredictability of hypoglycemia is an ever-present threat of insulin requiring assistance and/or having a low blood sugar (<50 mg/dl) without therapy and measures to prevent its incidence depend on knowledge of recognition. Death was classifi ed according to Diabetes Epidemiology predictors. The identifi cation of clinical and socio-demographic factors as Research International (DERI protocol) by a physician committee. Baseline markers of hypoglycaemia may improve diabetes management. (1990-1992) and most recent history of SH analyses were performed. The “Hypoglycemia in Insulin Treated diabetes” (HIT) Study included 5483 Separate multivariable Cox models were determined for each outcome based T1 and insulin-treated T2 patients in 10 European countries who reported on univariate signifi cance of previously established predictors. Baseline or rates of non severe hypoglycaemic events (NSHE) and related matters via 4 recent history of SH was forced into each model. The median time from online questionnaires. last follow up to death was 1.5 years. SH 2 years before baseline was not Risk markers were identifi ed in regression models. Weighting ensured independently related to total, CAD, renal, or CAD+renal mortality. SH in sample size and demographic features were similar in the 10 countries. the prior 2 years at most recent exam was not related to CAD or CAD+renal Markers of high frequency of NSHE for T1 patients were: female gender, mortality but was independently related to total mortality (HR=1.79, 1.1-2.9); longer duration of insulin treatment, >1 daily injections, Body Mass Index other predictors were diabetes duration, estimated glomerular fi ltration rate and perceived poor advice on hypoglycemia from GP (Table). Markers for T2 (eGFR), HbA1c, hypertension, smoking, albumin excretion rate (AER), and patients were the same, except for BMI. No statistical signifi cant association nonHDLc. A univariate association with renal mortality (HR=3.84, 1.3-10.9 was found with living status, primary education and work for pay. was weakened (HR=2.57, .89-7.4, p=0.08) with the addition of eGFR and AER. Gender, insulin duration and daily injections were found to be strong We conclude that SH in the last 1-4 years doesn’t increase CAD mortality predictors of NSHE in T1 and T2 diabetes.