Bad to the Bones: Bad to the Bones: Diabetes and Thiazolidinediones - PowerPoint PPT Presentation

Bad to the Bones: Bad to the Bones: Diabetes and Thiazolidinediones Diabetes and Thiazolidinediones 9/9/2010 9/9/2010 Steven Ing, MD, MSCE Assistant Professor Division of Endocrinology, Diabetes & Metabolism

“… Any reduction of bone mass in diabetics that is revealed by sophisticated analysis is of no medical or economic importance … Further extensive studies of bone metabolism in diabetics are unlikely to yield positive results of practical importance …” Heath NEJM 1980

Effects of DM on Bone T1DM T2DM Insulin deficiency Insulin resistance/ Pathology hyperinsulinemia Age of Younger age affects peak Older age after peak bone bone mass mass achieved onset BMI Often low Often high: loading, padding, more E2 Hyperglycemia increases Hyperglycemia increases Mechanism urinary calcium loss and urinary calcium loss and inhibits bone formation inhibits bone formation Low bone turnover Low bone turnover AGE AGE BMD May be lower May be higher

Potential Factors Affecting Fracture Risk in Diabetes Hypercalciuria Lower 25 OH Vitamin D T2DM have increased BMD Lower IGF ‐ 1 (anabolic for bone) Increased loading from obesity AGE’s Anabolic effect of hyperinsulinemia Inflammation Obesity associated with lower fracture risk DM nephropathy Cushioning during falls Fall risk greater: retinopathy, Lower bone turnover neuropathy, foot problems, Lower PTH levels cerebrovascular disease, hypoglycemia

Metanalysis: Effect of DM on Fracture & BMD Fracture Site T1DM T2DM Hip 6.94* (3.25 ‐ 14.78) 1.38* (1.25 ‐ 1.53) Wrist ‐‐‐ 1.19* (1.01 ‐ 1.41) Spine ‐‐‐ 0.93 (0.63 ‐ 1.37) Any Fracture ‐‐‐ 1.19* (1.11 ‐ 127) Spine Z ‐ score ‐ 0.22* ± 0.01 0.41* ± 0.01 Hip Z ‐ score ‐ 0.37* ± 0.16 0.27* ±0.01 Vestergaard, Osteoporos Int 2007;18:427

Women’s Health Initiative RR (95% CI) in RR (95% CI) in Multivariate Multivariate without BMD with BMD Any fracture 1.20 (1.11, 1.30) 1.24 (0.96, 1.63) Hip/pelvis/upper leg 1.46 (1.17, 1.83) 1.82 (0.90, 3.64) Lower leg/ankle/knee 1.13 (0.95, 1.34) 1.31 (0.76, 2.24) Foot 1.32 (1.07, 1.62) 1.27 (0.61, 2.64) Upper arm/shoulder/elbow 1.13 (0.90, 1.41) 0.90 (0.39, 2.07) Lower arm/wrist/hand 1.02 (0.85, 1.22) 1.27 (0.71, 2.25) Spine/tailbone 1.27 (1.00, 1.61) 1.57 (0.72, 3.44) Bonds, JCEM 2006;91(9):3404

Health Aging and Body Composition: TZDs associated with bone loss in women � 2006: No published data of BMD in clinical trials of ROSI and PIO � Prospective cohort community dwelling, 3075 men and women, 70 ‐ 79 yrs � 69 TZD users among 666 diabetics Annualized % Change in BMD per Year of TZD Use in Women % Change 95% CI P value Whole body ‐ 0.67 ‐ 1.03, ‐ 0.30 < 0.001 Lumbar spine ‐ 1.14 ‐ 1.90, ‐ 0.37 0.004 Total hip ‐ 0.38 ‐ 0.93, 0.17 0.178 Femoral neck ‐ 0.26 ‐ 0.86, 0.34 0.391 Trochanter ‐ 0.50 ‐ 1.02, 0.003 0.063 Schwartz, JCEM 2006; 91;3349

A Diabetes Outcome Progression Trial (ADOPT) and Fractures December, 2006 Kahn, SE, et al.NEJM 2006;355(23):2427-43

“Dear Doctor …” February, 2007 � February, 2007: GlaxoSmithKline (Avandia) reports the increased fracture risk in women in upper arm, hand, or foot � Based on 4360 subjects in ADOPT (1840 women) � Fracture incidence in ROSI ‐ treated = 2.74 per 100 patient ‐ years � Fracture incidence in MET ‐ treated = 1.54 per 100 patient ‐ years � Fracture incidence in GLY ‐ treated = 1.29 per 100 patient ‐ years

“Dear Doctor …” March, 2007 � Analysis of Takeda (Actos) clinical trials database � N=8100 PIO ‐ treated; N=7400 comparator ‐ treated � No increase in fracture risk in men � In women, there was higher incidence of fracture � Distal upper limb (forearm, hand, wrist) � Distal lower limb (foot, ankle, fibula, tibia) � Fracture incidence � PIO: 1.9 fractures per 100 pt ‐ yrs � CON: 1.1 fractures per 100 pt ‐ yrs

ADOPT Cont’d � Among 2511 men, fractures were not different by tx group # subjects with Incidence HR fracture ROSI 32 (4.0%) 1.16/100 pt ‐ yr metformin 29 (3.4%) 0.98/100 pt ‐ yr NS glyburide 28 (3.4%) 1.07/100 pt ‐ yr NS Kahn, Diabetes Care 2008;31(5)845

ADOPT Cont’d Among 1840 women, ROSI doubled fracture risk # subjects with Cumulative HR fracture Incidence @ 5 yr (95% CI) ROSI 60 (9.3%) 15.1% (11.2 ‐ 19.1) metformin 30 (5.1%) 7.3% (4.4 ‐ 10.1) 1.81 (1.17 ‐ 2.80) p= 0.008 glyburide 21 (3.5%) 7.7% (3.7 ‐ 11.7) 2.13 (1.30 ‐ 3.51) p=0.0029 Kahn, Diabetes Care 2008;31(5)845

ADOPT Cont’d � Increased fracture risk after 1 st yr � No placebo arm � Fracture not a specified endpoint (only Adverse Event reporting) � No spinal x-rays Kahn, Diabetes Care 2008;31(5)845

ROSI Decreases BMD & Bone Formation 14 week RCT in 50 postmenopausal women without DM Grey, JCEM 2006;92(4):1305

PIO Decreases BMD & Alkaline Phosphatase 16 week RCT in 30 premenopausal women with PCOS PIO Pre PIO Post CON Pre CON Post ALP 186 173 174 186 U/L (110 ‐ 315) (104 ‐ 288) (109 ‐ 280) (113 ‐ 186) Glintborg JCEM 2008; 93(5):1696

Mechanism: PPAR- γ Alter Lineage Allocation of Precursors C/EBP Preadipocyt Adipocyt e e Osteoclast Mesenchyma PPAR ‐ γ l TZD Stem Cell RANKL Preosteoclas t Preosteobla Osteoblas st t Runx2 OSX OPG Hematopoieti c Stem Cell

Unclear Mechanism by ∆ BTMs in ADOPT Women Men ROSI MET GLY ROSI MET GLY CTX 6.1 ‐ 1.3* ‐ 3.3* ‐ 1.0 ‐ 12.7* ‐ 4.3 (3.7, 8.7) (3.8, 1.2) ( ‐ 6.0, ‐ 0.6) ( ‐ 3.0, 1.0) ( ‐ 14.4, ‐ 10.9) ( ‐ 6.0, ‐ 2.5) P1NP ‐ 4.4 ‐ 14.4* ‐ 5.0 ‐ 14.4 ‐ 19.3* 0.2* ( ‐ 6.2, ‐ 2.6) ( ‐ 16.4, ‐ 12.4) ( ‐ 7.1, ‐ 2.8) ( ‐ 15.9, ‐ 13.0) ( ‐ 20.7, ‐ 18.0) ( ‐ 1.7, 2.1) BSAP ‐ 12.6 ‐ 15.7 ‐ 11.6 ‐ 13.6 ‐ 16.4 ‐ 6.8 ( ‐ 15.3, ‐ 9.9) ( ‐ 17.8, ‐ 13.6) ( ‐ 14.7, ‐ 8.3) ( ‐ 15.8, ‐ 11.3) ( ‐ 18.9, ‐ 13.8) ( ‐ 9.4, ‐ 4.0) Zinman, JCEM 2010;95(1):134

Metanalysis: TZD and Fracture � 10 RCTs, N=13,715 � IGT or T2DM � 1 ‐ 4 years of TZD exposure � TZDs increased overall fracture risk � Both: OR 1.45 (1.18 ‐ 1.79, p <0.001) � Women: OR 2.23 (1.65 ‐ 3.01, p < 0.001) � Men: OR 1.00 (0.73 ‐ 1.39, p = 0.98) Loke, CMAJ 2009;180(1):32

Is Fracture Risk Increased Only in Appendicular Skeleton? � Underlying hip fracture risk in RCT population was low � UKGPRD, 1020 cases of incident fracture, 3728 matched controls among 66,696 diabetics Adjusted OR Adjusted OR (95% CI) (95% CI) 1 ‐ 7 Rx ≥ 8 Rx Hip/femur 1.40 (0.31 ‐ 6.30) 4.54 (1.28 ‐ 16.10) Humerus 0.28 (0.04 ‐ 1.92) 2.12 (0.62 ‐ 7.26) Wrist/forearm 0.74 (0.23 ‐ 2.35) 2.90 (1.19 ‐ 7.10) Meier, Arch Intern Med 2008;168(8):820

Is Increased Fracture Risk only in Early Menopausal Women? � In UKGPRD: � For men: OR 2.50 (0.84 ‐ 7.41) � For women: OR 2.56 (1.43 ‐ 4.58) � For < 70 years: OR 2.96 (1.40 ‐ 6.25) � For ≥ 70 years: OR 2.57 (1.22 ‐ 5.4) � For PIO: OR 2.59 (0.96 ‐ 7.01) � For ROSI: 2.38 (1.39 ‐ 4.09) Meier, Arch Intern Med 2008;168(8):820

Number Needed to Harm Dormuth, Arch Intern Med 2009;169(15):1395

Are Men Susceptible? Annualized % No ROSI ROSI P Change at N=128 N=32 Lumbar Spine 2.3 ± 2.9 0.69 ± 2.4 0.03 Total Hip ‐ 0.137 ± 1.9 ‐ 1.19 ± 1.8 0.006 Femoral Neck ‐ 0.20 ± 1.25 ‐ 1.22 ± 1.3 0.0001 Yaturi, Diabetes Care 2007;30(6):1574

Practical Tips � Be aware of potential for bone loss and increased fracture risk in T2DM patients who initiate or continue TZD treatment � A doubling of fracture risk by TZD for older diabetic women � 4.3 � 8 ‐ 9% � Comparable to 1 SD decrease in T ‐ score � DXA in postmenopausal women ≥ 60 yrs � Other fracture risk factors � Age � Prevalent fragility fracture � Family history of fragility fracture � Low body weight or BMI � Cigarette smoking � Corticosteroids � Consider pharmacologic osteoporosis therapy in those with increased risk for fracture