1

Controversies in Long‐term Care of the Patient With Osteoporosis

North American Menopause SocietyOctober 16, 2014

Elizabeth Shane, M.D.Professor of Medicine

Columbia University Medical Center,New York, NY

Disclosures

• Industry Research Support (To Institution) – Amgen– Eli Lilly

• Consultancies, Speakers Bureaus, Financial holdings– Radius Pharmaceuticals

• Discussion of unlabeled use of drugs– None

Osteoporosis Medications

ANTIRESORPTIVE• Bisphosphonates

• RANK Ligand inhibitors (Denosumab)• Selective Estrogen Receptor Modulators (SERMs)• Calcitonin• Hormone therapy• Cathepsin K inhibitors

ANABOLIC• Parathyroid hormone analogs• Sclerostin inhibitors

Effects dissipate after stopping

Bisphosphonates  (BPs) have dominated the therapy of osteoporosis for the last 2 decades.

• What are their benefits?• What are their side effects and complications?

– Osteonecrosis of the Jaw, Atypical Femur Fractures

• How long should we to treat to maximize benefit and minimize risk?

• Which patients can safely stop and which should continue?

2

Bisphosphonates Available in the US

• Alendronate (Fosamax)– Oral weekly

• Zoledronate (Reclast)  – IV annually

• Risedronate (Actonel, Altelvia)– Oral weekly, monthly

• Ibandronate (Boniva)– Oral daily, monthly– IV every 3 months

Bisphosphonates Have Long Half‐Life

• “The gift that keeps on giving”

• Long residence in bone with release and possibly recycling of drug

• Sufficient drug retained so that benefit continues after therapy is stopped

• Half‐life estimated in years – but differs for different bisphosphonates– Longer for alendronate and zoledronic acid– Shorter for risedronate, ibandronate and pamidronate

EXPECTED Benefits of Bisphosphonates

Bisphosphonates for Postmenopausal Osteoporosis

• BPs increase bone mineral density– At the spine by 5‐8%– At the hip by 3‐5%

• In multiple, high quality RCTs, they reduce– Vertebral fractures by 40‐77%  – Hip fractures by 40‐60%– Non‐vertebral fractures by 30‐40%

• Reduction in fracture risk

– Sustained when continued beyond 5 years

• In randomized clinical trials, very safeBlack DM, Lancet 1996; Cummings SR, JAMA, 1998; McClung M, New Engl J Med 2001; Chesnut III CH, J Bone Miner Res 2004; Black DM, N Engl J Med. 2007; Lyles KW, N Engl J Med. 2007; Wallach, Calcif Tiss Int 2000; Adachi, Arthr Rheum 2001

3

UNEXPECTED Benefits of Bisphosphonates

Bisphosphonates Associated with LOWER MORTALITY

• After hip fracture 1,2,3

• In community‐dwelling elderly 4

• In institutionalized elderly 5

1. Lyles et al, N Engl J Med 2007. 2. Boonen et al, J Am Geriatr Soc 2011. 3. Beaupre et al, Osteoporos Int 2011. 4. Center et al, J Clin Endocrinol Metab 2011. 5. Sambrook et al, Osteoporos Int 2011.

Bisphosphonates Associated with Decreased Risk of Cancer

Breast cancer 1‐5

Colon cancer 6,7

Gastric cancer 8

1. Rennert et al, J Clin Oncol 2010. 2. Newcomb et al, Br J Cancer 2010 3. Chlebowski et al, J Clin Oncol 2010.4. Cardwell et a., Int J Cancer 2011. 5. Monsees et al, J Natl Cancer Inst 2011.6. Rennert et al, J Clin Oncol 2011. 7. Cardwell et al, Int J Cancer 20118. Abrahamsen et al., J Bone Miner Res 2012

Risks of Bisphosphonates

4

Osteonecrosis of the Jaw (ONJ)

AKA Bisphosphonate‐Related Osteonecrosis of the Jaw (BRONJ)

In 2014, renamedMedication‐Related Osteonecrosis of the 

Jaw (MRONJ)

Updated AAOMS Case Definition of ONJ

• Current or previous treatment with antiresorptive or antiangiogenic drugs

• Exposed bone or bone that can be probed through an intraoral or extraoral fistula in the maxilla or mandible

• No history of craniofacial radiation and no obvious jaw bone metastases

• No healing after 8 weeks

Ruggiero et al., AAOMS Position Paper: Medication-Related ONJ 2014

Osteonecrosis of the Jaw 

• Reported with all BPs 1 and denosumab (Prolia) 2‐4 

• and angiogenic agents

• >90% of cases occur in cancer patients on long‐term, high dose IV BPs for prevention of skeletal‐related events• Incidence 1‐16% 1

• Very rare in patients on bisphosphonates for osteoporosis 5‐11

• 1‐10/10,000 5, 6

• 0.4/10,000 7

• 1/250,000 8

1. Walter C et al. Head and Face Med 2010 2. Stopeck et al., J Clin Oncol 2010 3. Henry et al., J Clin Oncol 20114. Fizazi et al., Lancet 20115. Mavrokokki et al., J Oral Maxillofac Surg, 20076. Lo JC, et al., J Oral Maxillofac Surg 20107. Malden et al., J Bone Miner Metab, 20128. Felsenberg et al., Deutsches Arzteblatt, 20069. Khosla S et al., J Bone Miner Res, 200710. Hellstein et al., JADA, 201111. AAOMS Position paper 2014

Risk Factors for ONJ

• Dental extraction or surgery• Poorly fitting dentures• Periodontal disease• Potency of antiresorptive drug

• Denosumab =  Zoledronic acid > Pamidronate >>>> Alendronate, Risedronate, Ibandronate

• Long duration of BP use ‐ >3‐4 yrs

Ruggiero et al., AAOMS Position paper, 2014.

5

Prevalence of ONJ in Patients on Oral BPs by Duration of Use

21 per 10,000 at > 4 yrs

2011 Position Statement:American Dental Association * 

Benefits of osteoporosis therapy outweigh low risk of developing ONJ

•*Hellstein et al., Managing the care of patients receiving antiresorptive therapy for prevention andtreatment of osteoporosis. JADA.org, 10/31/2011

Atypical Femur FracturesTYPICAL (Ordinary) Hip Fractures

Femoral Neck (FN) Fracture Intertrochanteric (IT) Fracture

BPs reduce these types of fractures by 40-60%

6

TYPICAL and Atypical ST/FS FracturesTypical SPIRAL Subtrochanteric

FractureCOMPLETE Atypical

TRANSVERSE SubtrochantericFracture

Courtesy of Eve Donnelly, PhD

Incomplete Atypical Femur Fracture

Courtesy of Fergus McKiernan, MD; McKiernan et al. J Clin Densitom. 2011;14:354-

Plain X-rays

MRI

Bone Scan

Even DXA

Risk Factors for Atypical Femur Fractures

• Younger age• Obesity• Asian race 1,2,3

• Hispanic ethnicity 3

• Low vitamin D levels, low serum calcium levels• Multiple antiresorptive drugs• Glucocorticoids• Rheumatoid arthritis• Diabetes

• Long duration of BP use – usually > 5 yrs1. Lo et al., Bone 2012;51:181-4 2. Dell et al., JBMR 2012; 27:2544-50 3. Marcano et al., Clin Orthop Relat Res epub Oct 29, 2013

Most Studies WITH Radiograph Review Suggest They Are.

Are atypical fractures associated with bisphosphonate use?

7

Swiss Case‐Control StudyMeier et al., Arch Intern Med 2012

• 477 patients > age 50 admitted with ST or FS fractures 1999‐2010

• 438 had classic fractures and 6% used BPs

• 39 had atypical fractures and 82% used BPs

0

50

100

150

200

OR of Atypical Fracture and Years of BP Use

Adjusted OR for BP use - 69.1 95% CI 22.8-209.5

Atypical Fractures per 100,000 BP‐exposed Patients by Duration of Use ‐ Kaiser Permanente

Duration of Bisphosphonate Use, Years

Num

ber o

f AFF

s pe

r 100

,000

Exp

osed

Pat

ient

s

P=NS P< 0.0001

Dell et al., J Bone Miner Res 2012

The RELATIVE RISK of a patient with an AFF being on BP therapy is HIGH.

HOWEVER,The ABSOLUTE RISK of AFFs in Patients on Bisphosphonates is VERY LOW.

1. Meier et al., Arch Intern Med 2012 2. Schilcher et al., New Engl J Med 20113. Feldstein et al., J Bone Miner Res 2012

8

The Kaiser California Study

• 4 years of BP use, risk of AFF – 30 per 100,000 patient years

• 10 years of BP use, risk of AFF – 125 per 100,000 patient years

White US women age 65 with T-score -2.5 and no risk factors:

Risk of classic hip fracture - 280 per 100,000 patient yrsOther osteoporotic fractures - 1,200 per 100,000 patient yrs

Dell et al., J Bone Miner Res 2012

• Stress or insufficiency fractures

• 75% have prodromal pain

• 25‐50% are bilateral

–often located at same level in the opposite femur

Important Clinical Features of AFFs

Shane et al., ASBMR Task Force Report, J Bone Miner Res 2010

Proposed Pathogenetic Mechanisms for Antiresorptive Therapy and Atypical Femur Fractures

Prolonged suppression of bone turnover may adversely affect bone material 

properties and strength      brittle failure.

Prevent healing of stress fractures.

Likely applies to both Bisphosphonatesand Denosumab

AFF Risk Declines After Stopping BPs Schilcher et al., New Engl J Med, 2011

• Swedish case‐control study– 59 cases, 269 matched controls

• Most AFFs occurred within 1 yr of last BP Rx• 70% risk reduction for every yr since last use

– Adjusted OR 0.28 (95% CI 0.21, 0.38)

• Risk reduction similar for those taking BPs for < or > 2 yrs

9

ONJ and Atypical Femur Fractures:Summary

• Strongly associated with BP therapy, though BPs not proven to be causal

• Directly related to duration of BP treatment

• ONJ is extremely rare in patients treated with the doses of BPs used for osteoporosis ‐‐‐‐‐ basically a NON‐ISSUE

• Atypical Femur Fractures are a greater concern ‐‐‐‐‐unusual before 5 years of use

1. Khosla et al., J Bone Miner Res, 2007 2. Shane et al., J Bone Miner Res, 2010

Long‐term Side Effects of BPs: Influence on Management of 

Osteoporosis

Recognition of Long Term Risks Has Influenced How We Treat Osteoporosis

• Avoid bisphosphonates and denosumab in patients at low risk of fracture

• Avoid combination anti‐resorptive therapies– e.g., estrogen + BPs, SERMs + BPs, denosumab + BPs

• As 75% of patients with AFFs have prodromal pain– Ask patients about thigh or groin pain– Image patients with unexplained thigh or groin pain

• Reassess patients after 3‐5 years of BP treatment and consider drug holiday where appropriate

Shane et al., ASBMR Task Force Report, J Bone Miner Res 2010

How Long To Treat Osteoporosis With Bisphosphonates?

• The concept of a drug holiday

• Very controversial, though has become more generally accepted over time

• But for whom is it appropriate?

• How long should it last?

10

Bisphosphonate Long‐term Efficacy Analysis Predominantly Based Upon

• ALENDRONATE – FIT and FLEX data

• ZOLEDRONIC ACID – HORIZON and HORIZON EXTENSION

FIT Long-Term Extension of Alendronate (FLEX)*

FIT N = 6,459

Placebo N = 3,223 Alendronate N = 3,236

Randomized in FLEXN = 1,099

Alendronate, 5 mg N = 329 Placebo N = 437 Alendronate, 10 mg

N = 333

FIT (3 to 4.5 yrs)

Post-FIT (1-2 yrs)

FLEX (5 yrs)

40% 30% 30%

Primary endpoint: Change in hip BMD

* Black, et al, J Bone Miner Res 2012

Zoledronic Acid HORIZON Extension Study

N = 7,736

Placebo N = 3,876

Zoledronate 5 mg IV for 3 yearsN = 3,867

Randomized in ExtensionN = 1,233

Zoledronate, 5 mg IV for 3 years N = 616Z6

Placebo N = 617

Z3P3

* Black et al, JAMA 2006

Primary endpoint: Change in hip BMD

EXTENSION STUDIES: Key Limitations

• Not powered for fracture outcomes

• Looked only at efficacy (for vertebral fractures)• No analysis of benefits versus risks

• Included only femoral neck BMD

11

Survival Curve for Time to First NONVERTEBRAL Fracture in FLEX

0

5

10

15

20

0 1 2 3 4

Cum

ulat

ive

Inci

denc

e, %

5

Time to First Fracture, years

Placebo ALN (pooled)

RH=1.00 (0.76, 1.32)Similarly, no reduction in risk of nonvertebral fracture in

HORIZON EXTENSION

Black et al. JAMA 2006; 296:2927 Black et al. JBMR 2012; 27:240

Summary of FLEX and Horizon Extension Studies:Long Term BP vs Switch to Placebo

• FLEX demonstrates that ALN for 10 yrs• Maintained BMD at all sites versus loss in PBO, p<0.001• Lowered risk of clinical vertebral fractures• RR = 0.45 (95% CI 0.24-0.85)

• HORIZON extension demonstrates that ZLN for 6 yrs• Maintained BMD at all sites versus loss in PBO, p<0.001• Lowered risk of morphometric vertebral fractures• RR = 0.51 (95% CI 0.26,0.95)

Black et al. JAMA 2006; 296:2927 Black et al. JBMR 2012; 27:240

Predictors of Vertebral and Non‐vertebral Fractures OFF Therapy

• Age– ~50% increase for any clinical fracture per 5 year increase– In FLEX 1 but not HORIZON 2

• Fem Neck or Total Hip T score < ‐2.5– CLINICAL Vert FX risk: FLEX 2‐fold increase 3

– MORPHOMETRIC Vert FX risk: HORIZON 3‐4 fold increase 2

• Incident fractures during core HORIZON trial 2– Morphometric Vert FX – 5‐fold increase in Morphometric Vert FX – Non‐vert FX – 2‐fold increase in incident Non‐vert FX

• Having more risk factors increased fracture risk (HORIZON) 2– Low T score, prevalent fracture, incident fracture

1. Bauer D et al JAMA Int Med 2014;doi101.1001 2. Black D et al. JAMA 2006; 296(24)3. Cosman F et al JCEM 2014; in press

UCSF Opinion on BP Continuation(Black et al., New Engl J Med 2012)

• No consistent reduction in non‐vertebral fractures

• Reasonably consistent ~52‐55% reduction in vertebral fractures 

• Therefore, base recommendations on reductions in vertebral fracture risk

• Those at higher absolute risk will have larger benefit• Lower “Number Needed to Treat” (NNT)

• Use hip BMD  and other risk factors to define groups at higher absolute vertebral fracture risk• Hip BMD related to vertebral fracture risk in FLEX and HORIZON 

12

UCSF Analysis Summary:Groups at High Absolute Risk of Fracture May Benefit From Continued Therapy After 5 Years

After 5 years of alendronate or 3 years of zoledronate:• Many women can discontinue with little loss of efficacy• Those at higher risk of vertebral fracture may benefit from continuing• FN BMD T‐score below ‐2.5 ‐ NNT 21• Prevalent vertebral fractures and somewhat higher FN BMD T‐score (up to ‐2.0) ‐ NNT 17

• Cannot generalize to risedronate or ibandronate

How Long to Treat with Bisphosphonates?

• 3‐5 years appears to be safe for most patients

• After 3 years of Zoledronate or 5 years of Alendronate, assess for ongoing fracture risk– Clinical assessment (Fx history), DXA, Vertebral Fracture Assessment

Watts NB and Diab D. J Clin Endocrinol Metab. 2010;95(4):1555-1565.

Drug Holiday After 3-5 years

Drug Holiday After 10 years

Higher RiskLower Risk

My Approach to Long‐term Management of Postmenopausal Osteoporosis

Lower Risk• BMD > ‐2.0 at hip and spine• No incident fractures on Rx• No prevalent vertebral 

fractures• Relative youth

– < 70

• No other major risk factors– GCs, aromatase inhibitors

Higher Risk• BMD < ‐2.5 at FN or TH• Hip BMD between ‐2.5 & ‐

2.0 BUT < ‐2.5 at spine• Prevalent vertebral fracture• Incident fractures on Rx• Older age > 75• Other major risk factors

– GCs, aromatase inhibitors

Treat with oral BPs for >5 yrs, IV BPs for >3 yrsReevaluate

My Approach to Long‐term Management of LOW Risk Women

• Drug Holiday• Reassess every 2‐3 years• Reassessment includes

• Clinical evaluation• DXA• +/‐ Bone markers (Serum CTx)

Treat with oral BPs for >5 yrs, IV BPs for >3 yrs

13

My Approach to Long‐term Management of HIGH Risk Women

• Continue antiresorptive Rx for up to 10 yrs……or……..

• Consider alternative Rx (Teriparatide)• Reassess every 2‐3 years

• Earlier if fracture or other high risk situation associated with rapid bone loss

• Reassessment includes• Clinical evaluation• DXA• +/‐ Bone markers (Serum CTx)

Treat with oral BPs for >5 yrs, IV BPs for >3 yrs

10‐Year Probabilities

Fracture risk typical of patient with osteoporosis; MVA and murder data from the CDC at http://www.cdc.gov/nchs/data/nvsr/nvsr56/nvsr56_10.pdf ; ONJ estimate is 1/100,000 patient-treatment-years from the ADA at JADA. 2006;137:1144-1150. AFF estimate is 5/10,000 patient-treatment-years from Schilcher J et al. N Engl J Med. 2011;364:1728-1737.

FRAX 10-year probability of major osteoporotic fracture for untreated 72 year-old woman with FN T-score of -3.0 is 25%

For patients in whom antiresorptive therapy is indicated because they are at high risk of fracture,

the benefits of treating virtually always outweigh the risks.

Courtesy of Michael Lewiecki, M.D.

Thank you!