BAD for the Bones Skeletal Woes from Commonly Prescribed Medications

BAD for the Bones Skeletal Woes from Commonly Prescribed Medications A Case of Skeletal Cruelty Phenobarbitol Arimidex Heparin Jonathan Graf, MD Asst. Professor Medicine, UCSF Div. Rheumatology, SFGH Prednisone Zoloft Lupron Prilosec Outline of Today s Talks Part 1: Corticosteroids Definitely bad for the bones! Lunch Part 2: Other acts of skeletal cruelty Cancer therapies (anti-hormonal therapies) Proton Pump Inhibitors Surprise guest! Glucocorticoids: Some toxicities Diabetes Cataracts HTN Weight gain Fluid retention PUD Myopathy Psychiatric OSTEOPOROSIS & OSTEONECROSIS 50% of patients 1

GLUCOCORTICOIDS: Potential Mechanisms Contributing to Osteoporosis Bone OB bone formation OB lifespan, apoptosis, osteocyte apoptosis Pituitary estrogen testosterone adrenal androgens OC bone resorption Gut Renal GI Ca absorption Ur- Ca excretion Serum Ca PTH GLUCOCORTICOIDS: Double Whammy to Bones Pro-resporbtive effects (early) Decrease OPG Increase RANK-L osteoclast #, activity, lifespan Anti bone-formation effects (late) Decrease osteoblast and osteoclast formation Decrease osteoblast lifespan Enhance apoptosis in osteocytes OSTEOPOROSIS Muscle (myopathy) GLUCOCORTICOID - INDUCED OSTEOPOROSIS Glucocorticoid Effects on Remodeling/Strength Early phase of rapid bone loss (Anti- Resorption) As early as 2 mos into therapy Resorption markers elevated Pts on high dose prednisone can lose 15-20% of trabecular bone (spine) in 5-7 mos Slower phase of bone loss (Anti-Formation) Continues indefinitely Trabecular bone especially vulnerable X many Manolagas, JBMR, 2000 2

Fracture risk increases with AGE, GENDER, and Dose HIGH Van Staa et al, JBMR, 1998 MEDIUM Steroids WOMEN MEN CONTROL * Relative Rate (± 95% CI) of Non-vertebral Fractures : D/C Steroids (5 years) - REVERSAL PATIENTS -- INHALED STEROIDS RR Retrospective, case-control cohort (UK Gen Practice Database) Assessed fracture risk in users of inhaled GCs vs bronchodilators vs controls (100,000-170,000 pts) No one on systemic steroids; 54% female, average age -- 45 RRs: NON-VERT 1.15 (CI 1.1-1.2) (c/w controls) HIP 1.22 (CI 1.04-1.43) VERTEBRAL 1.51 (CI 1.22-1.85) Modest increases: Does not vary depending upon which inhaled steroid used?possibly related to underlying pulmonary disease Van Staa et al, 2001, JBMR 3

SUMMARY Relatively low doses of po GCs (2.5-7.5 mg Pred) increase fracture risk Increase in risk is quick -- within 3 mos of starting therapy Vulnerable population: postmenopausal women, elderly pts Fracture risk decreases if stop therapy More falls in steroid-treated patients (more frailty, less mobility, less activity) van Staa et al, JBMR, 2000 Management of GIO American College of Rheumatology Guidelines 2001 ACR Recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis Patient begining therapy with glucocorticoid (prednisone equivalent of 5 mg/day) with plans for treatment dur. of 3 months Smoking cessation or avoidance Reduction of alcohol consumption if excessive Instruct in weight-bearing physical exercise Initiate calcium with vitamin D (plain or activated form) supplementation Prescribe bisphosphonate (caution with premenopausal women) ACR Recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis Patient receiving long-term glucocorticoid therapy (prednisone equivalent of 5 mg/day): Same as for patients initating therapy, EXCEPT: Measure bone mineral density (BMD) at lumbar spine and/or hip If BMD is not normal (i.e., T-score below -1) - Prescribe bisphosphonate (caution with premenopausal women) - Consider calcitonin as second-line agent only If BMD is normal, follow up and repeat BMD measurement either annually or biannually 4

Drug Management of GIO Gonadal steroids: testosterone, HRT Calcitonin Bisphosphonates Alendronate** Risedronate** Zoledronic Acid ** Ibandronate Pamidronate Anabolic therapy: rh-pth (1-34)** ** FDA approved GIO: Alendronate Saag et al, NEJM, 1998 477 men and women, ages 17-83 RCT: placebo vs ALN (5 or 10 mg/d) x 48 wk ALL: 800-1000 mg Ca + 250-500 vitamin D Diseases: rheumatic, derm, pulm, GI + others GCs: > 7.5 mg prednisone or equiv Stratified: duration of previous GC therapy Less than 4 mos 4 to 12 mos > 12 mos Women on HRT continued it Outcomes: BMD, markers, vert fractures (semiquantitative, digitized techniques) Change in BMD over 48 weeks * * * * GIO and BMD: increase in L-Spine BMD (%) across all groups of patients Placebo ALN 10 mg GROUP (all) - 0.4 + 2.9** Men - 0.7 + 2.9 Premenopausal - 0.3 + 2.0 Postmenopausal - 0.6 + 1.5 on estrogen Postmenopausal - 0.1 + 4.0 no estrogen Duration of GC < 4 mos - 1.0 + 3.0 4-12 mos - 0.6 + 2.8 > 12 mos + 0.2 + 2.8 ** statistically significant Saag et al, NEJM, 1998 Saag et al, NEJM, 1998 5

GIO & Fracture: No overall difference in morphometric spinal fractures: 48 wk. Quantitative morphometry: Fracture = 20% or > or 4 mm decrease in vert ht no significant difference -- ALN-treated vs placebo Semi-quantitative assessment (48 wk): Placebo ALN 5 or 10 mg All patients 8/135 8/268 (NS) Men 1/48 2/75 Premenopausal 0 0 Postmenopausal 7/54 (13%) 6/135 (4.4%)** GIO and fracture: 2yr. Alendronate Extension Study 12 mo extension in 208 pts (66 men, 142 women) on continued GCs 24 mos total Placebo vs ALN 5 or 10 mg/d ALL: 800-1000 mg Ca + 250-500 vitamin D Outcomes: BMD, markers, vert fractures Adachi et al, Arth Rheum, 2001 ** statistically significant Saag et al, NEJM, 1998 GIO and fracture: Alendronate Extension Study Morphometric vertebral fractures at 24 months Placebo 4/59 (6.8%) ** Alendronate 1/143 (0.7%) Year 1 Placebo 1/59 (1.7%) Alendronate 1/143 (0.7%) Year 2 Placebo 3/59 (5.1%) Alendronate 0/143 (0) **statistically significant Adachi et al, Arth Rheum, 2001 Adachi et al, Arth Rheum, 2001 6

GIO: RISEDRONATE 2 Studies Both randomized, MC, DB, PC 12 mo trials Placebo vs 2.5 or 5 mg risedronate 500 mg Ca +/- vit D (1) + vit D (2) BMD, fractures (1) 224 men and women starting long term GC Cohen et al, Arth Rheum, 1999 (2) 290 men and women on chronic > 7.5 mg prednisone for > 6 mos Reid et al, JBMR, 2000 #1 (N=224) 5 P 5 mg placebo Morphometric fracture reduction in patients starting GC s: RISEDRONATE Number of pts with new vertebral fractures (quantitative morphometry) At 1 year Placebo 9/52 (17.3%) Risedronate (5 mg) 3/53 (5.7%)** ** p = 0.072 Cohen et al, Arth Rheum, 1999 Chronic Glucocorticoid users: Fracture reduction with RISEDRONATE 290 patients > 6 mos New vertebral fractures: 9/60 patients in placebo vs 3/60 in each RIS treatment group ** ** p = 0.042 (combined groups vs placebo) Adverse events: NSD from placebo RIS protects bone in high-risk, chronic GC-treated pts (FDA approved) RIS also sig. increased BMD at Hip and Spine (data not shown here) Reid et al, JBMR, 2000 7

What about Zoledronic Acid? Better than other bisphosphonates?: Horizon Reid et al. Lancet. 2009 Apr 11;373(9671):1253-6363 Horizon Design Reid et al. Lancet. 2009 Apr 11;373(9671):1253-6363 1 year randomized double blind, double dummy, non- inferiority 833 patients Subdivided into treatment groups based on duration of steroid therapy (>< 3 months) IV ZA 5mg vs. PO Risedronate 5mg Primary endpoint: BMD LS spine Horizon Demographics: A representative population Horizon Results Majority of patients were on more than 7.5mg prednisone a day! Majority of patients had rheumatoid arthritis or SLE Figure 2. Change in mean bone mineral density of lumbar spine and femoral neck for (A) treatment and (B) prevention subgroups Error bars=95% CI. *p=0 0005. p=0 0001. p=0 0050. p<0 0001. p=0 0156. p=0 0049. 8

PTH vs. bisphosponates: Beneficial for GIO? (2008) 36 Month randomized double blinded placebo controlled ongoing 18 month interim analysis 428 patients studied 22-89 years of age Treated with GC s for at least three months Prednisone equivalent of 5 mg/day or more 20 mcg/d PTH vs. 10 mg/d alendronate Everyone continued Ca/VitD Saag et al. NEJM 2007;357:2028-3939 Who were the Patients? BMD<-2.0 or <1.0 + fragility fracture (more severe) Exclusions: Standard for PTH Use Two treatment groups similar (n=214 both) Alendronate PTH Age 57.3 56.1 Prednisone dose 7.8 7.5 Non Vert Frag Fx 20.1% 19.6% BMD T score LS Spine -2.6-2.5 BMD T score Hip -1.9-2.0 Patients Underlying Disease PTH vs. Alendronate More than 25% drop out for both arms 9

PTH vs. Alendronate: Fracture Results PTH vs. Alendronate 36 month Follow Up: BMD results Saag et al. Arthritis Rheum. 2009 Nov;60(11):3346 Saag et al. Arthritis Rheum. 2009 Nov;60(11):3346-55 PTH PTH PTH PTH vs. Alendronate: 36 month fracture follow up Subgroup analysis from phase 2 trial data: Patients receiving Donosumab or PBO without exclusions for bisphosphonate use Dore et Al. Ann Rheum Dis 2010;69:872-875 Saag et al. Arthritis Rheum. 2009 Nov;60(11):3346-55 10

PTH for GC induced Osteoporosis: Summary PTH appears to improve BMD in GC - osteoporosis Evidence suggests superior increases in BMD vs. alendronate at hip, femoral neck, and LS spine However, high attrition rate in this study Nearly 50% drop out in both arms by 36 months Appears to be of benefit both for clinical and radiographically defined fractures at 36 months BUT.overall rate of clinical fracture is low in both groups Decrease in fractures limited to vertebral fractures but non-significant for non-vertebral fractures Perhaps consider PTH for those with the most severe GC induced OP and those at highest risk for fracture (such as previous fragility fracture) Actual SFGH Case in 2007 Very prominent retired VIP UCSF physician admitted after suffering multiple severe spinal and hip fractures after a fall. In fact, the fractures were so severe and unstable, that he was transferred from UCSF to SFGH for special handling by the trauma and neurosurgery services. (First time I had ever heard of a transfer going from the university to county hospitals and not the other way around!) His past medical history was notable for cardiac disease and prostate cancer. His x-rays, in addition to demonstrating multiple severe and unstable hip and spinal fractures showed profund osteopenia. Rather than baby sit the patient while in the hospital, the medical service did a more thorough review of his history and discovered: The patient had received many years of lupron therapy for prostate CA Had never received a DXA Had never been offered any osteoporosis prevention or therapy (not even calcium and vitamin D) Patient had gotten VIP care from the very best doctors in the system Osteoporosis (OP) and Prostate Cancer High incidence of OP in men with prostate cancer even before ADT (10-25%) May be due to advanced age, hypogonadism, and diminished vitamin D Risk of OP in androgen deprivation therapy (ADT) patients is markedly increased, and doesn t depend upon the regimen used 11

Androgen Deprivation Therapy Effects on Bone Mineral Density Comparing Surgical and Chemical Orchiectomy, Effects on BMD Daniels et al. In one prospective trial of 62 patients, BMD decreased 7.6% at two years with surgical or chemical castration (+/- antiandrogen therapy) Most studies concur in showing declining BMD post Androgen Deprivation Therapy Chemical Castration Surgical Castration Medical castration is very effective!! Incidence of First OP fracture after Orchiectomy Daniels et al. Fracture Free Survival over Time Shahinian, Goodwin, et al. NEJM 2005;. 352:154-164164 Restrospective study of 50,000 patients with prostate cancer 1992-9797 Diminishes with increasing dosages of ADT Incidence of first fracture is 15% at 4 years in castrated patients vs. 1.5% in prostate ca patients without ADT Rises to 50% vs. 8% At 9 years!!!! At 9 years, only 50% patients receiving 9 doses of therapy remained fracture free! 12

Number of Patients needed to harm to cause a fracture: Synergistic effects of Age and cululative ADT dose Shahinian, Goodwin, et al. NEJM 2005;. 352:154-164164 Treatment of ADT Osteoporosis Treatment initiation is recommended for both iatrogenic and naturally occurring hypogonadism in males In cases involving Prostate CA androgen replacement therapy not an option! Therefore, recommendations of the American Cancer society: Use of bisphosphonate Non Randomized, Uncontrolled Trial: BMD in Patients on ADT: Effects of risedronate 2.5 mg/d for 6m Ishizaka et al. Int. J. Urology. December 2007 Summary of studies using IV Bisphosphonates in ADT: through 2003 They work! 13

Clinical Guildelines Singer F., et al. Cancer 2004;100(5): 892-899899 Donosumab increases BMD up to 36 months Smith et al. NEJM 2009; 361 (8) Main Point: Be alert to osteoporosis in these patients!!! Donosumab lowers rate of vert. fractures at 12, 24, and 36 months Androgen Deprivation Therapy Take Home Points ADT is very bad for the bones RAPID, Dramatic loss in BMD BMD and Fx incidence worse with advancing age and cumulative ADT dose Surgical or chemical castration the same Close monitoring of BMD (at least 1-2 years) Bisphosphonates recommended, especially in lower BMD individuals Smith et al. NEJM 2009; 361 (8) 14

Aromatase Inhibitors Breast Cancer and Osteoporosis Chemotherapy causes gonadal ablation, premature menopause and premature osteoporosis May be possible direct anti-metabolic effects of chemotherapy on bone (Greep et al. Am J Medicine 2003;114:653-9) Increasing use of Aromatase inhibitors Comparing AI s to Tamoxifen: Different effects on BMD Use of 3 rd generation non-steroidal: anastrozole and letrozole cause 96-99% 99% aromatase inhibition very effective Used more widely for metastatic or high risk disease because of superiority vs. tamoxifen REBBeCa: Risedronate Effect on Bone Loss in Breast Cancer Greenspan et al. J. Clin Endocrinology 2007 Randomized double blinded placebo controlled trial 12 months follow up with 12 month extension 87 newly post- menopausal women after chemotherapy for breast cancer Risedronate weekly vs. placebo Primary outcome: changes in hip and spine BMD Patient Characteristics Fewer than 20% of patients were taking an aromatase inhibitor 15

FIG. 2. Mean (SEM) percent change in bone mineral density from baseline to 12 months Use of Bisphosphonates to Prevent Bone Loss in Breast Cancer Treated with AI s Greenspan et al. J Clinical Oncology June 2008 Since publication of trial, standard of care has shifted away from tamoxifen and towards aromatase inhibitors 12 month extension of REBBeCa to 24 months (roughly 10% drop out in both arms) Greenspan, S. L. et al. J Clin Endocrinol Metab 2007;92:131-136 Use of AI s increased from less than 20% to 44% in second year Copyright 2007 The Endocrine Society Fig 3. (A) Mean (SE) percent changes in bone mineral density from baseline to 24 months Fig 4. (A) Mean (SE) percent change in bone mineral density from baseline to 24 months Copyright American Society of Clinical Oncology Greenspan, S. L. et al. J Clin Oncol; 26:2644-2652 2008 Copyright American Society of Clinical Oncology Greenspan, S. L. et al. J Clin Oncol; 26:2644-2652 2008 16

Z-Fast: Zoledronic Acid and AI s Brufsky et al. J Clinical Oncology 2007 Fig 2. Mean (SEM) percent change in bone mineral density of the lumbar spine and the total hip at months 6 and 12 in women with early-stage breast cancer administered upfront or delayed zoledronic acid Open label, randomized, un- blinded study Patients receiving letrozole received either A. upfront or B. delayed Z.A. (if their t-score dropped <2.0) In group B: 4% at 6 months, 8% at 12 months received Z.A. Group B: mean time to Rx =8.8 months Brufsky, A. et al. J Clin Oncol; 25:829-836 2007 Copyright American Society of Clinical Oncology Zo-Fast Fast (not to be confused with Z-fast) Bundred et al. Cancer July 2008 Nearly identically designed study looking at BMD in 931 patients 17

Proton Pump Inhibitors They re ubiquitous They re over the counter Rumor has it, they may even be in the water supply! But are they bad for the bones, too??? Hadji et al. Ann Oncol. 2008 Aug;19(8):1407-16. 16. *Novartis Funded PPIs: Mechanism of Action PPIs: Theoretically can have beneficial and detrimental effects on bone Block acid secretion in stomach Could effect calcium absorption in intestine Block acid secretion by osteoclasts May also interfere with bone resorption by inhibiting the vacuolar proton pumps of osteoclasts (osteoclasts pump out acid to erode bone) Proton Pump Inhibitors: Are they Really Bad for the Bones, Too?? Yang et al. JAMA 2006;296:2947-29532953 Nested Case Control Study 13,556 hip fracture cases 135,386 controls Age >50 OR Hip Fx (>1yr therapy): 1.44 OR Hip Fx (>1.75 dose): 2.65 OR of fracture rises with dose and duration of therapy with PPI 18

Women s Health Initiative Gray et al. Archives of Internal Medicine: May 2010 160,000 postmenopausal women from Women s health initiative (age 50-79) 2010 study from Kasier San Francisco 33,000 patients and 130,000 control Reaches similar conclusion Risk of hip fracture increases with dose and duration of PPI therapy 1,000,000 years of patient follow up Many receiving calcium +/or Vit D Examined self-reported fractures and medication records Study did its best to confirm accuracy of both Risk for Fracture According to PPI and H2RA Use at Baselinea Three-Year Changes in Mean Bone Mineral Density (BMD) According to Baseline Proton Pump Inhibitor (PPI) Use Gray, S. L. et al. Arch Intern Med 2010;170:765-771. Gray, S. L. et al. Arch Intern Med 2010;170:765-771. Copyright restrictions may apply. Copyright restrictions may apply. 19

Proton pump inhibitor (PPI) use and bone mineral density (BMD) at hip, spine, and total body Hip Spine Total Body PPIs: Summary No increased risk of hip fracture Possible increase risk of spine and or wrist fractures Self reported spine fracture data called into question (50% accuracy when audited) Marginal change in BMD at 3 years Jury is still out!!! Gray, S. L. et al. Arch Intern Med 2010;170:765-771. Mystery Guest Medication 20

Bisphosphonates: Bad for the bones???? We ve just spent the past hour lauding the benefits of biphosphonates How can they be bad for the bones too?? Heresy!! Sub IT/FN Reviewed 41 cases of ST fractures Radiographic Findings insub- Trochanteric Fractures Lenart et al. Osteoporosis Intl 2009 Thickened corticies Fracture near thickest cortex Thinner corticies More more complex fractures 21

Bisphosphonates and ST Fx 14,000+ patient records reviewed from 3 randomized bisphosphonate trials 284 fractures identified (12 ST fractures in 10 patients) No significant difference in ST fracture rates in patients using bisphosphonates followed up to 10 years Number of fractures very small Study underpowered (confidence intervals crossed 1.0) Nonetheless, given widespread use of bisphosphonates, low rate of fracture reassuring!! 22