Bone Quality: Understanding What Matters

1
Treatment of Osteoporosis Applying the Evidence
to Clinical Practice Chatlert Pongchaiyakul Divis
ion of Endocrinology and Metabolism Department of
Medicine, Faculty of Medicine Khon Kaen University
2
Expectations of an Agent for Treatment of
Osteoporosis

• Consistency across efficacy endpoints
• Increase in BMD at all sites
• Consistent fracture reduction
• Vertebral fracture (morphometric and clinical)
• Non-vertebral fracture
• Hip fracture
• Results reproducible and consistent across
• Subgroups
• Multiple trials
• Differing populations
• Established long-term efficacy and safety

3
What Is a Systematic Review?

• Summaries of the medical literature that address
  a focused clinical question
• Meta-analysis quantitative method to summarize
  the result
• Address issue of consistency of results
• Obtain overall estimate from several studies

4
What criteria should we consider when making
clinical decisions?

• Well conducted randomized placebo controlled
  clinical trials
• Direct comparison of drugs in the same class to
  demonstrate equal or superior efficacy
• Rigorous well defined meta-analysis (pitfalls of
  garbage in, garbage out)
• Clinical experience
• Physician/patient decision making

5
Data that can be misleading when making clinical
decisions

• Retrospective analysis
• Non pre-specified analysis
• Per-protocol vs. ITT
• Subgroup analyses
• Post-Hoc analyses
• Non-randomized, non-controlled
• Historical controls
• Anecdotal clinical experience

Downs SH, Black N J Epidemiol Community Health
1998 52377-384 Hauselmann HJ, Rizzoli R, Osteo
International 2003 142-12 / Freemantle, BMJ
2001, 322989-991 Assman SF, et alLancet
20003551064-1069 Sackett DL,et al Evidence
Based Medicine Second Edition Harcourt Publishers
Limited,London UK 2000.
6
Hierarchy of Evidence
Meta-analysis of RCTs systematic review of RCTs
Individual RCT
Observational studies patient-important outcomes
Basic research test tube, animal, human physiology
Clinical experience nonsystematic clinical
observation
7
Meta-analyses of Therapies for Postmenopausal
Osteoporosis

• Endocrine Reviews 200223495-578

8
Overview of Methodology

• Comprehensive literature search to 2000
• Explicit eligibility criteria
• 2 or more independent reviewers
• Assessment of validity of studies
• Concealment of randomization
• Blinding
• Completeness of follow-up
• Sophisticated analytic techniques

9
Evidence-based Review of Osteoporosis Trials

• Calcium Vitamin D
• HRT
• SERMs
• Raloxifene
• Calcitonin
• Bisphosphonates
• Alendronate
• Risedronate

Orag Cranney et al Endocrine reviews
2002 Hauselmann et al Osteopororosis Intnl
2003 Hochberg MC et al Arthritis Rheumatism 1999
10
Calcium Vitamin D
11
Incidence of Nonvertebral Fractures Calcium-Vitami
n D
Cumulative Percentage of Subjects with a First
Nonvertebral Fracture

• At 36 months (P0.02)
• 26 of 202 Placebo
• 11 of 187 Calcium-Vit D

Dawson-Hughes B, et al N Engl J Med 1997
12
Reduction in Hip Other Nonvertebral Fractures
Calcium and Vitamin D
Hip Fracture
Other Nonvertebral Fracture
Cumulative Probabilities of Fracture
Elderly institutionalized patients with high
prevalence of subclinical vitamin D deficiency
?? Placebo ?? Calcium and Vit D
Chapuy, MC etal, N Engl J Med 1992.
13
Effect of Calcitriol (1-25 Vitamin D) or Calcium
in the Treatment of Postmenopausal
Osteoporosis(single blinded study)
Proportion of Women Who Did Not Have New
Vertebral Fractures During the Three Years of
the Study
Plt0.01 Plt0.001
Tilyard MW, et al N Engl J Med, 1992
14
Evidence of Fracture Efficacy from Randomized
Clinical Trials for Calcium and Vitamin D Analogs
Hochberg MC. Drugs Aging 17 317-30,2000
15

• HRT

16
Effect of HRT on BMD Postmenopausal
Estrogen/Progestin Interventions Trial (PEPI)
Hip
Spine
Percent Change from Baseline
Placebo CEE Only CEE-MPA (cyc)
Placebo CEE Only CEE-MPA (cyc)
CEE-MPA (con) CEE-MP (cyc)
CEE-MPA (con) CEE-MP (cyc)
Baseline 12 mo
36 mo
Baseline 12 mo 36 mo

• P.E.P.I.
• 2,763 postmenopausal women
• Mean Age 56

• Mean Prevalent Fractures at baseline 0
• Drop Out Rate 22-44
• Primary Endpoint Hip and Spine BMD

Writing Group for the PEPI Trial. JAMA,
1996276138996
17
Effect of HRT On Vertebral Fracture
Vertebral Fractures

• No effect on fracture in PEPI
• Most data on vertebral fractures come from
  epidemiologic studies
• There exists limited conclusive prospective trial
  evidence to support vertebral fracture reduction
  with HRT
• NIH-sponsored Womens Health Initiative will
  study fracture efficacy in a prospective 812
  year trial

40
12/34
42 risk reduction p0.17
35
30
25
7/34
Women with New Vertebral Fractures
20
15
10
5
0
When analyzed using the numbers of fractures
method, a 61 risk reduction was observed p0.04)
Lufkin et al, Annals of Int Med., 19921171-9
18
4 Years of HRT Had No Effect on the Risk of
Non-spine Fractures
Type E P Placebo RH p-value Hip 12 11 1.1 .82
Other 47 48 0.9 .59 Any 130 138 1.0 .70
No effect on height loss
Hulley et al. HERS Study. JAMA. 1998 280(7)
August 19, 1998.
19
Hormone Replacement Therapy and Prevention of
Nonvertebral Fractures
A Meta-analysis of Randomized Trials
David J. Torgerson, PhD Sally E. M. Bell-Syer,
MSc JAMA, June 13, 2001-Vol 285, No. 22
20
HRTSummary

• Reduction in vertebral fracture in 1 small study
• Questionable evidence for reduction in
  non-vertebral fracture
• No evidence for hip fracture reduction
• Effectiveness not robust
• NIH Sponsored Womens Health Initiative studying
  fracture efficacy awaited

21
Risks and Benefits of Estrogen Plus Progestin in
Healthy Postmenopausal Women Principal Results
from the Womens Health Initiative Randomized
Controlled Trial
Writing Group for the Womens Health Initiative
Investigators JAMA July 17. 2002- Vol 288, No. 3
22
Womens Health Initiative (WHI) Overview

• Large and complex clinical investigation of a
  variety of prevention strategies for cancer,
  cardiovascular disease, and osteoporotic
  fractures
• The WHI was initiated in 1992, with a planned
  completion date of 2007
• Postmenopausal women aged 50 to 79 were enrolled
  into either a clinical trial (N?64,500) or an
  observational study (N?100,000)

The Womens Health Initiative Study Group.
19981961109.
23
Womens Health Initiative (WHI) Overview

• The clinical trial is a randomized, controlled,
  comparative trial, which includes three
  overlapping components
• Sustained low-fat eating pattern vs self-selected
  dietary behavior
• HRT or ERT vs placebo
• Calcium plus vitamin D (CaD) vs placebo
• Overall benefit-vs-risk is a central focus of the
  clinical trial components
• Evaluations will include risk of coronary heart
  disease and other cardiovascular diseases risk
  of hip and other fractures and risk of breast
  and colorectal cancer

The Womens Health Initiative Study Group.
19981961109.
24
HRT Component of the WHIA Randomized,
Controlled Primary Prevention Trial

• Trial Design
• Randomized, controlled primary prevention trial
  (planned duration, 8.5 years)
• 16,608 postmenopausal women aged 5079 (mean age,
  63.3) with an intact uterus at baseline
• Women received conjugated equine estrogens (CEE),
  0.625 mg/day, plus medroxyprogesterone acetate
  (MPA), 2.5 mg/day (n8,506) or placebo (n8,102)

The Writing Group for the Womens Health
Initiative Investigators. JAMA.
2002288(3)321333.
25
HRT Component of the WHIA Randomized,
Controlled Primary Prevention Trial

• Trial Design
• Primary outcome was coronary heart disease (CHD)
  (nonfatal myocardial infarction and CHD death),
  with invasive breast cancer as the primary
  adverse outcome
• The global index summarizing the
  risks-vs-benefits included the two primary
  outcomes plus stroke, pulmonary embolism,
  endometrial cancer, colorectal cancer, hip
  fracture, and death due to other causes

The Writing Group for the Womens Health
Initiative Investigators. JAMA.
2002288(3)321333.
26
HRT Womens Health Initiative Fracture
Outcomes
Writing Group for the Womens Health Initiative
Investigators JAMA July 17. 2002- Vol 288, No. 3
27
WHI Results Subanalysis of Fractures

• No attempt to include women with low BMD or prior
  fracture (89 normal or osteopenic)
• Approximately 1000 women had BMD measurements at
  baseline and every three years
• Drop out rate 30 -35
• Drop in rate 10
• Fracture rate significantly reduced if BMI (lt 25
  lean women)
• Efficacy of fracture reduction not influenced by
• prior fracture
• prior use of HRT
• prior osteoporosis

ASBMR 2002
28
HRT Womens Health Initiative Summary
Continued

• Trial was stopped early based on health risks
  exceeding benefits from combined use of HRT

Writing Group for the Womens Health Initiative
Investigators JAMA July 17. 2002- Vol 288, No. 3
29
HRT Cardiovascular Disease Results
Cardiovascular Disease Outcomes Relative to
Placebo Control at 5.2 Years
Increased Risk
Hazard Ratio (adjusted 95 CI)
Total Cardiovascular disease
CHD
CABG/ PTCA
Stroke
VTE disease
Decreased Risk
The Writing Group for the Womens Health
Initiative Investigators. JAMA.
2002288(3)321333.
30
HRT Cancer Results
Cancer Outcomes Relative to Placebo Control at
5.2 Years
Increased Risk
Hazard Ratio (adjusted 95 CI)
Decreased Risk
Invasive Breast
Colorectal
Total
Endometrial
The Writing Group for the Womens Health
Initiative Investigators. JAMA.
2002288(3)321333.
31
HRT Component of the WHISummary of Results at
5.2 Years (Early Termination)

• In postmenopausal women with an intact uterus,
  HRT was associated with
• 15 increase in global index (risks exceeded
  benefits)
• 29 increase in CHD events
• 22 increase in total cardiovascular disease
• 26 increase in invasive breast cancer
• 41 increase in stroke
• 111 increase in VTE disease

Nominal statistical significance at the 0.05
level. The Writing Group for the Womens Health
Initiative Investigators. JAMA.
2002288(3)321333.
32
HRT Component of the WHISummary of Results at
5.2 Years (Early Termination)

• In postmenopausal women with an intact uterus,
  HRT was associated with
• 37 decrease in colorectal cancer
• 34 decrease in hip and clinical vertebral
  fractures

Nominal statistical significance at the 0.05
level. The Writing Group for the Womens Health
Initiative Investigators. JAMA.
2002288(3)321333.
33
Relative Risk For Vertebral Fractures After
Treatment With HRT
Favours HRT Favours Control

(n 75)
Lufkin 0.63 (0.28,1.43)

(n193)
Greenspan 0.70 (0.06, 7.55)

(n32)
Wimalawansa 0.4 (0.09, 1.80)

(n2763)
Hulley 0.69 (0.34, 1.38)

Alexandersen 2.78 (0.12, 65.1)
(n52)
(n 1006)
Mosekilde, 2.01(0.61,6.63)

(n16,608)
WHI 0.65 (0.44, 0.97)

Pooled RR 0.70 (0.52, 0.94)
N20,729

0.01
0.1
1
10
100
(p0.02)
Wells Endo Reviews 2002
34

• SERMS-Raloxifene

35
Efficacy at the SPINEEffect of Raloxifene on
Radiographic Vertebral Fractures (MORE)
Radiographic Vertebral Fractures
25
Substudy 2 (n2,304) (BMD?-2.5 and
pre-existing vertebral fractures)
Substudy 1 (n4,524) (BMD?-2.5 and no
pre-existing vertebral fractures)
20
RRR 30
15
Percent of Patients withIncident Vertebral
Fracture
10
RRR 50
5
0
Placebo
Raloxifene 60 mg/d
Placebo
Raloxifene 60 mg/d
JAMA, August 18, 1999--Vol 282, No. 7, pp 637-645
36
Efficacy at the HIPEffect of Raloxifene on
Non-Vertebral and Hip Fracture MORE Pooled Data
(60 mg and 120 mg)
Hip Fractures
Non-Vertebral Fractures
3
15
Placebo
Percent of Patients with Incident Non-Vertebral
Fractures
2
10
RaloxifenePooled
RaloxifenePooled
1
5
Placebo
0
0
0
6
12
18
24
30
36
0
6
12
18
24
30
36
Months
Months
JAMA. 1999282637645
37
Relative Risk for Vertebral and Non-Vertebral
Fracture After Raloxifene Treatment
Favors Raloxifene
Favors Control
Vertebral Fractures
Ettinger 0.60 (0.50 to 0.70)
(N 6828)
?
Lufkin 1.16 (0.77 to 1.76)
?
( N 133)
Non-Vertebral Fractures
Ettinger 0.92 (0.79 to 1.07)
( N 6828)
?
Lufkin 0.52 (0.12 to 2.18)
(N 133)
?
All Trials Secondary Treatment
0.1
1
10
Vertebral fracture results from Lufkin trial
based on 15 cutoff in reduction of vertebrae
(baseline to 1 year)
38
Raloxifene Summary

• Reduces vertebral fracture risk
• No effect on hip or non-vertebral fractures

39

• Calcitonin

40
Calcitonin Nasal Spray Prevent Recurrence of
Osteoporotic Fractures (PROOF)
Number of patients 1,255 Most (79) had 1 to 5
Vertebral Fx 378 (30) completed the study Mean
Age 68 (postmenopausal) Mean T-Score lt 2.0 New
Vertebral Fx 15 Study Design 5 year, randomized,
double-blind, placebo-controlled Drug Placebo
(n 311), 100 IU (n 316), 200 IU (n 316,
marketed dose) or 400 IU (n 312) Calcium/Vitamin
D 1000 mg/400 IU daily Primary Endpoint Spine
BMD and new VFX Dropout rate 59 lost to follow-up
Chesnut CH et al., Am J Med. 2000, 109 267-276.
41
PROOFEffect of nasal calcitonin on risk of
vertebral fractures

• Women with Decreased Dose new
  fractures risk plt0.05?
• Placebo 26 -
• 100 IU 22 15 No
• 200 IU 18 33 Yes
• 400 IU 22 16 No

Non-vertebral fracture and hip fracture No
consistent trend with dose Non-significant
reductions at the approved dose (200 IU)
Chesnut CH et al., Am J Med. 2000, 109 267-276.
42
Trial quality is very important

• What PROOF Proves about Calcitonin and Clinical
  Trials
• Steven R. Cummings, MD
• Roland D. Chapurlat, MD
• Editorial from Am J Med 109 330-331, Sept 2000

43
CalcitoninSummary

• Morphometric vertebral fracture
• Small reduction in vertebral fracture rate
• No consistent trend with dose
• No reduction in non-vertebral or hip fracture
• Minimal effect on BMD and biomarkers
• Trial Issues
• High drop out rate
• Investigators and patients were not blinded to
  important outcomes (e.g., BMD)
• Effectiveness of nasal calcitonin not proven

44

• Bisphosphonates
• Alendronate
• Risedronate

45

• Etidronate

46
Four-Year Study of Intermittent Cyclic Etidronate
Treatment of Postmenopausal Osteoporosis Three
Years of Blinded Therapy Followed by One Year of
Open Therapy

• Steve Harris, MD, Nelson Watts, MD,
• Rebecca Jackson, MD,
• Harry Genent, MD, Richard Wasnich, MD,
• Philip Ross, MD, Paul Miller, MD,
• Angelo Licata, MD,Charles Chestnut, MD

47
Etidronate Treatment for Osteoporosis

(Per 1000 patient years)
Non-etidronate n34
etidronate n44
1 year etidronate n25
4 years etidronate n32
0-3 Years 4th
Year (Blinded) (Open
Label)
sign indicates a significant difference as
compared with non-etidronate treatment (plt0.05)
Harris, S., et al, Am J Med., Vol. 95, 1993
48

• By the 7th year of cumulative cyclical
  etidronate therapy, the spinal BMD was comparable
  to that observed with 3 years of continuous
  alendronate therapy.
• 
  Miller, PD et al.
• Cyclical Etidronate in the Treatment of
  Postmenopausal Osteoporosis Efficacy and Safety
  after Seven Years of Treatment.
  Am J Med 103 468476, Dec 1997

49

• Alendronate

50
Efficacy at the SPINEAlendronate Vertebral
Fracture Study1(Patients with pre-existing
vertebral fractures)
Reduction in Vertebral Fractures at Year 3
1Black DM. Lancet. 19963481535-1541.
51
Alendronate Reduced the Risk of Hip Fractures in
women with or without existing vertebral fracture
Reduction 51 p 0.047
Reduction 56 p 0.044
Placebo
Rate per 100 PYR
Alendronate
Without (T lt -2.5)
With (T lt -1.6)
Vertebral Fracture Status at Baseline
Cummings SR, et al. JAMA. 1998 2802077-2082.
Black DM, et al. Lancet. 1996 3481535-1541.
52
Evidence-Based Review of OP Trials Effect of
Alendronate in Reducing Vertebral Fractures
Favors alendronate
Favors control
Prevention Trials
McClung 0.34 (0.04 to 3.25)
(n 355)
(n 1355)
Pooled Prevention Estimate 0.45 (0.06 to 3.15)
Treatment Trials
Bone 0.68 ( 0.21 to 2.18)
(n 184)
Chesnut 0.25 (0.03 to 2.34)
(n 157)
Liberman (USA) 0.52 ( 0.24 to 1.15)
(n 478)
Liberman (Int) 0.52 ( 0.20 to 1.34)
(n 516)
Black 0.53 (0.41 to 0.69)
(n 2027)
Cummings 0.51 ( 0.31 to 0.84)
(n 4432)
Pooled Treatment Estimate 0.53 (0.43 to 0.65)
(n 8005)
48 reduction
Pooled Estimate 0.52 (0.43 to 0.65)
(n 9360)
0.01
0.1
1
10
Adami and Hoskings trials not included in figure
due to low vertebral fracture incidence.
53
Evidence-Based Review of OP Trials Effect of
Alendronate in Reducing Non-Vertebral Fractures
Cranney et al, Endocrine Reviews 23 (4)508-516
2002
Favors Alendronate Favors Control
Prevention Trials
(n 267)
McClung 0.79 (0.28 to 2.24)
Treatment Trials
Adami 0.36 (0.07 to 1.80)
(n 211)
Chesnut 0.43 (0.11 to 1.65)
(n 125)
Liberman (USA) 0.55 (0.31 to 0.97)
(n 380)
(n 412)
Liberman (Int) 0.65 (0.32 to 1.34)
Pols 0.47 (0.26 to 0.83)
(n 1908)
Rosen 0.35 (0.15 to 0.77)
(n 420)
Pooled Treatment Estimate 0.49 (0.36 to 0.67)
(n 3456)
49 reduction
Pooled Estimate 0.51 (0.38 to 0.69)
(n 3723)
0.01
0.1
1
10
54
The Effect of Alendronate on Hip Fracture
Favors Alendronate
Favors PBO
Primary Fracture Studies
Black vertebral osteo

Cummings T lt -2.5
Karpf Osteoporosis
BMD Primary Endpoint
PolsOsteoporosis
Long Term Care Osteoporosis

All Studies
52 (p 0.001)
Test for homogeneity p 0.996
Relative Risk
T lt 2.0 Quandt S., Annals Rheum Disease 2000,
59 83

55
Long-Term Effects of Bisphosphonates
56
Alendronate 10 Year Efficacy Data Lumbar Spine
BMD
Emkey et al. J Bone Miner Res 200217 (Suppl
1)S139
57
Alendronate 10 Year Efficacy Data Trochanter BMD
58
Percent of Patients with Non-vertebral Fractures
Years 1 to 3
Years 8 to 10
Liberman 1995, NEJM
59
Annual Height Loss
Years 1 to 3
Years 5 to 10
2.0
P0.005
1.5
Mean Height Loss (mm/year)
1.0
0.5
0
All ALN
Placebo
ALN 5mg
ALN 10 mg
ALN 20/5/PBO
Liberman 1995, NEJM
60
AlendronateSummary

• Most robust, consistent efficacy in reducing
  fractures at all sites
• Largest number of trials performed for the
  longest period of time across the entire spectrum
  of osteoporosis
• Long term 10 year data evidence of continued
  safety and efficacy

61
Effect of alendronate on trabecular microstructure
UNTREATED
ALENDRONATE
Hu et al. Bone Nov 2002
62
Alendronate Increases Trabecular Connectivity in
Elderly Osteoporotic Women Masarachia P, Howard
T, Santora A, Yates J, Rodan GA, Recker RR, and
Kimmel, DB ASBMR Poster M326 MicroCT Analysis of
Transilial Biopsy Specimens from Alendronate
(ALN) Phase III Studies Masarachia P,
Chavassieux P, Arlot M, Meunier PJ, Santora A,
Yates J, Rodan GA and Kimmel DB ASBMR Poster
M325 Monday 22 September 2003
63
m-CT Images representing respective BV/TVs for
PBO and ALN Groups

PBO Specimen 1618 BV/TV 13.9 BMD, 119.8 mg/cc
Euler 2.3
ALN Specimen 1776 BV/TV 19.0 BMD 156.1 mg/cc
Euler 3.1
64
Alendronate Increases Trabecular Connectivity in
Elderly Osteoporotic Women
Table 3. m-CT unique Endpoints (696yo) Table 3. m-CT unique Endpoints (696yo) Table 3. m-CT unique Endpoints (696yo)
PBO ALN
BMD (mg/cc) 112.0 55.4 146.2 53.1
Euler 3.29 1.7 4.89 2.33
DMB g/cc bone 7.96 2.34 8.29 1.03
ALN patients have higher Euler ALN patients
trend higher but without statistically
significant differences from PBO, for BMD or DMB.

ALN patients have higher Euler ALN patients
trend higher but without statistically
significant differences from PBO, for BMD or DMB.
65
Hip fractures are correlated to the number and
width of haversian canals (cortical porosity) in
femoral cortical bone
Bell et al. Bone, 2000
66
Alendronate Reduces Cortical Porosity
67
All patients had NTx levels above the lower limit
of the normal premenopausal range (n1023)
Garnero, JBMR 199611337-349 Alendronate Data
on File, Protocol 118
68
Risedronate
69
Risedronate Vertebral Fracture Studies
(VERT)(Patients with pre-existing vertebral
fracture)
Radiographic Vertebral Fractures at Year 3
49 Reduction P lt 0.001
41 Reduction P lt 0.003
of Patients with Fracture
PBO 678
RIS 696
PBO 346
RIS 344
n
NA(1)
MN(2)
1Harris et al. JAMA. 199928213441352. 2Reginste
r JY. Osteoporos Int. 2000118391.
70
Risedronate Non-Vertebral Fracture Studies (VERT)
Non Vertebral Fractures at Year 3
33 Reduction NS (0.063)
39 Reduction P 0.02
of Patients with Fracture
PBO
RIS
PBO/ RIS
RIS
NA(1)
MN(2)
1Harris et al. JAMA. 199928213441352. 2Reginste
r JY. Osteoporos Int. 2000118391.
71
Risedronate hip fracture efficacy data in
postmenopausal women with osteoporosis
HIP (70-79)2 With Prior VFx39 Hip T-score ?3.0
w/ risk factors or Hip T-score?4.0 Age
Range70-79
HIP (80)2 With Prior VFx45 Age Range80
Group 1 by dosage 2.5 mg 50 s.s 5 mg 30 n.s
(20) reduction at Year 3 p0.35
40 reduction at Year 3 p0.009
5
5
4
4
3
3
of patients with fracture
of patients with fracture
2
2
1
1
0
0
PBO n1,821
RIS n3,624
PBO n1,313
RIS n2,573
1. ACTONEL package insert. Cincinnati, OH
Procter Gamble Pharmaceuticals and Kansas City,
MO Aventis Pharmaceuticals, Inc. 2000. 2.
McClung MR, et al. NEJM. 2001344(5)333340.
The percentages are based on the number of women
for whom vertebral-fracture status was known.
72
Evidence-Based Review of OP Trials Effect of
Risedronate in Reducing Vertebral Fractures

• Cranney et al, Endocrine Reviews 23(4)517-523
  2002

Relative Risk with 95 CI
(Final Year, All Doses)
Favors Risedronate Favors control

Prevention Trials
?
(N 111)
Treatment Trials
?
(N 1374)
?
(N 132)
?
(N 297)
(N 690)
?
36 reduction
?
(N 2604)
0.1
1
10
73
Evidence-Based Review of OP Trials Effect of
Risedronate in Reducing Non-Vertebral Fractures

• Cranney et al, Endocrine Reviews 23(4)517-523
  2002

Relative Risk with 95 CI (Final Year, All Doses)
Favors Risedronate Favors Control
Prevention Trials
?
Mortensen (1998) 0.49 (0.12 to 2.03)
Treatment Trials
?
Harris (1999) 0.64 (0.42 to 0.98)
(N132)
?
Clemensen (1997) 1.52 (0.55 to 4.15)
?
McClung (2001) 0.76 (0.58 to 1.00)
Fogelman (2000) 0.62 (0.29 to 1.32)
?
?
Reginster (2000) 0.71 (0.47 to 1.06)
(N648)
?
McClung (1999) 0.71 (0.36 to 1.40)
27 reduction
?
Pooled Estimate 0.73 (0.61 to 0.87)
10
0
0.5
1
1.5
2
2.5
74
The Effect of Risedronate on Hip Fracture
Favors RIS
Favors PBO
McClung 70 -79 Osteoporosis (Group I)
Primary Hip Study

McClung Over 80 with T lt -2.5 (Group II)
Harris (with VFX)
Primary VFX Study
Reginster (with VFX)

30 (p 0.015)
All Studies
Test for homogeneity p 0.440
Relative Risk
T lt 2.0 Quandt S., Annals Rheum Disease 2000,
59 83

75
Risedronate 7 Year Data from VERT-MNVertebral
and Non-vertebral Fractures
Non-vertebral
Vertebral
Years 0-5
6-7
0-5
6-7
LS BMD increase 11.5 over 7 yrs Total hip BMD
increase 3.9 over 7 yrs
Sorensen et al. ECTS 2003
76
RisedronateSummary

• Reduction in vertebral fracture
• Reduction in non-vertebral fractures in North
  American cohort
• Reduction in hip fracture shown in women 70-79
  years of age
• 7 year data show sustained effect on vertebral
  fracture reduction

77
Meta Analysis Summary Antifracture Efficacy
-Vertebral Fractures
Number of
Relative
Trials
Relative Risk
Risk p
Hetero-
Intervention
(Patients)
(95 CI)
value
geneity
Alendronate
8 (9360)
0.52 (0.43 to 0.65)
lt 0.01
0.99
5 (2604)
0.64 (0.54 to 0.77)
0.01
0.89
Risedronate
Raloxifene
0.01
2 (6961)
0.65 (0.56 to 0.76)
0.01
Etidronate
9 (1076)
0.63 (0.44 to 0.92)
0.02
0.87
Calcitonin
4 (1404)
0.02
0.01
0.46 (0.25 to 0.87)
8 (1130)
Vitamin D
lt 0.01
0.16
0.63 (0.45 to 0.88)
0.57 (0.29,1.13)
0.12
HRT
3 (300)
0.86
Calcium
5 (576)
0.77 (0.54 to 1.09)
0.14
0.40
Fluoride (4yrs)
5 (646)
0.67 (0.38,1.19)
0.17
0.01
ORAG
78
Meta-Analysis Summary Antifracture Efficacy
-Non-vertebral Fractures
79
What criteria should we consider when making
clinical decisions?

• Well conducted randomized placebo controlled
  clinical trials
• Direct comparison of drugs in the same class to
  demonstrate equal or superior efficacy
• Rigorous well defined meta-analysis (pitfalls of
  garbage in, garbage out)
• Clinical experience
• Physician/patient decision making

80
Meta Analysis Summary Antifracture Efficacy
-Vertebral Fractures
Number of
Relative
Trials
Relative Risk
Risk p
Hetero-
Intervention
(Patients)
(95 CI)
value
geneity
Alendronate
8 (9360)
0.52 (0.43 to 0.65)
lt 0.01
0.99
5 (2604)
0.64 (0.54 to 0.77)
0.01
0.89
Risedronate
Raloxifene
0.01
2 (6961)
0.65 (0.56 to 0.76)
0.01
Etidronate
9 (1076)
0.63 (0.44 to 0.92)
0.02
0.87
Calcitonin
4 (1404)
0.02
0.01
0.46 (0.25 to 0.87)
8 (1130)
Vitamin D
lt 0.01
0.16
0.63 (0.45 to 0.88)
0.57 (0.29,1.13)
0.12
HRT
3 (300)
0.86
Calcium
5 (576)
0.77 (0.54 to 1.09)
0.14
0.40
Fluoride (4yrs)
5 (646)
0.67 (0.38,1.19)
0.17
0.01
81
Meta-Analysis Summary Antifracture Efficacy
-Non-vertebral Fractures
82
Are there direct comparisons with
anti-resorptive agents in assisting the
physicians clinical decision?
83
Combination Therapy with Hormone Replacement and
Alendronate for Prevention of Bone Loss in
Elderly Women
Greenspan SL et al JAMA 2003 289, 2525-2523
84
Combination Treatment of Raloxifene
Alendronate in PM Osteoporosis

• Randomized, double-blind, 12 month study
• Postmenopausal women with osteoporosis (FN -2.0),
  n331
• Greater spine and FN increases with ALN than RLX
• Similar tolerability

20
0
-20
-40

-60

-80
NTx/Cr
BSAP
P0.05 ALN vs. RLX
Johnell et al., JCEM 2002
85
EFFECT-International EFficacy of Fosamax vs.
Evista Comparison Trial
Sambrook et al, Calcif Tissue Int 2003
86
Introduction

• Both alendronate (ALN) and raloxifene (RLX) are
  used for postmenopausal osteoporosis
• Meta-analyses for both agents have been performed
  (Cranney et al., Endo Reviews 2002)
• Head-to-head clinical trial data provide the best
  means to evaluate comparative efficacy and
  tolerability
• Only 1 prior head-to-head trial (Johnell et al,
  JCEM 2002)
• EFFECT-US and EFFECT-International were conducted
  to provide additional head-to-head data for ALN
  vs. RLX in postmenopausal osteoporosis

87
ORAG Meta-analyses

• Raloxifene
• Small increases in spine BMD 2.5, combined hip
  2.1, forearm NS, and total body 1.3 (1-3 yrs)
• Reduces risk of vertebral fractures by 40, but
  results across trials inconsistent
• No effect on non-vertebral fractures
• Increased risk of withdrawal due to hot flashes

• Alendronate
• Increase in spine BMD 7.5, combined hip 4.2,
  forearm 2.1, total body 2.7 (2-4 yrs)
• Reduces risk of vertebral fracture by 48
• Reduces risk of non-vertebral fracture by 49
• No increase risk of withdrawal due to adverse
  effects

Cranney et al., Endocrine Reviews 2002
88
Introduction

• Both alendronate (ALN) and raloxifene (RLX) are
  used for postmenopausal osteoporosis
• Meta-analyses for both agents have been performed
  (Cranney et al., Endo Reviews 2002)
• Head-to-head clinical trial data provide the best
  means to evaluate comparative efficacy and
  tolerability
• Only 1 prior head-to-head trial (Johnell et al,
  JCEM 2002)
• EFFECT-US and EFFECT-International were conducted
  to provide additional head-to-head data for ALN
  vs. RLX in postmenopausal osteoporosis

89
EFFECT International Study Design

• Randomized, double-blind, double-dummy,
  active-comparator controlled, 12 month study
• 50 sites in 16 countries
• Separate study conducted in the US
• Patients evaluated at screening, randomization,
  3, 6, and 12 months
• Patients randomized 11 to
• ALN 70 mg once weekly and RLX placebo daily-or-
• RLX 60 mg daily and ALN placebo once weekly
• Key efficacy endpoints evaluated by central QA
  centers(BMD BioImaging markers of bone
  turnover - Quest)

Sambrook et al, Calcif Tissue Int 2003
90
EFFECT International Endpoints

• Primary endpoint
• change from baseline in lumbar spine BMD at 12
  months
• Secondary endpoints
• change from baseline in
• Total hip BMD at 12 months
• Hip trochanter BMD at 12 months
• BMD at 6 months
• Percentage of patients who maintained or
  increased BMD
• change in markers of bone turnover (NTx and
  BSAP)
• Tolerability, including upper GI and vasomotor
  events

Sambrook et al, Calcif Tissue Int 2003
91
EFFECT International Baseline Characteristics
Sambrook et al, Calcif Tissue Int 2003
92
Alendronate-Raloxifene Comparison12 Month
Percent Change in BMD relative to baseline
Plt0.001
Plt0.001
Plt0.001
Mean Change in BMD
Sambrook et al ECTS 2003
93
EFFECT InternationalMarkers of Bone Turnover
NTx
BSAP
20
0
-20
-29
Change (95 CI)
-40
-60
-68


-80
-100
0
6
12
Months
Months
Plt0.001 Alendronate vs. raloxifene
Sambrook et al, Calcif Tissue Int 2003
94
EFFECT-InternationalMarkers of Bone Turnover
Absolute Values
NTx
BSAP
Absolute Value




Months
Months
below upper limit Alendronate
Raloxifene NTX 96 60 BSAP 96 63
Plt0.001 Alendronate vs. raloxifene Premenopau
sal range Garnero et al, JBMR 1996
Sambrook et al, Calcif Tissue Int 2003
95
EFFECT InternationalTolerability
Possibly, probably or definitely drug-related as
determined by the investigator
Sambrook et al, Calcif Tissue Int 2003
96
EFFECT International Summary

• ALN produced 2-3 fold greater increases in hip
  and spine BMD than did RLX at 12 months
• ALN RLX
• Spine 4.8 2.2
• Total hip 2.3 0.8
• Trochanter 2.9 1.0
• More patients maintained or increased BMD with
  ALN than with RLX
• ALN produced significantly greater increases in
  BMD than did RLX at the early time point of 6
  months
• ALN decreased bone resorption (NTx) significantly
  more than did RLX at 6 and 12 months
• The decreases in markers of bone turnover with
  alendronate were to the middle of the
  premenopausal range
• Overall tolerability of ALN and RLX was similar,
  with more patients reporting vasomotor events on
  RLX

Sambrook et al, Calcif Tissue Int 2003
97
EFFECT International Conclusions

• Alendronate 70 mg OW produced larger increases in
  BMD and greater reduction in bone resorption at
  both 6 and 12 months than did raloxifene 60 mg
  daily
• These results are consistent with a similar study
  performed in the US (Kagan et al, ACOG 2003), the
  prior study by Johnell et al, and the
  meta-analysis by Cranney et al.
• The study was not powered for fracture endpoints.
  However, the greater reduction in bone
  resorption by ALN is consistent with its effect
  to reduce non-vertebral fracture incidence, which
  has not been shown for RLX
• These results are important to consider when
  evaluating treatment options for postmenopausal
  women with osteoporosis

Sambrook et al, Calcif Tissue Int 2003
98
Alendronate vs. RisedronateComparison Trial
99
Alendronate vs. Risedronate Comparison Trial
Aim To compare the efficacy of alendronate 70
mg once weekly with risedronate 5 mg daily in
post menopausal women with osteoporosis. Prima
ry Endpoint Rate of bone resorption (urine
NTx) at 3 months Secondary Endpoints Change
in BMD at spine and hip at 6 and 12 months Rate
of bone turnover (NTx and BSAP) at 6 and 12
months Safety and tolerability at 12 months
100
Alendronate vs. Risedronate Comparison
Trial Overview

• First head to head study comparing alendronate
  and risedronate for treatment of osteoporosis
• Endpoints
• Biochemical markers of bone turnover
• BMD spine and hip

101
Alendronate vs. Risedronate Comparison Trial
Study Design

• Randomized, double-blind, multicenter,
  multinational, placebo-controlled study
• 3 months bone turnover
• 6 and 12 months BMD
• 549 postmenopausal women with osteoporosis age gt
  60
• T-score lt -2.5 at either lumbar spine or total
  hip or
• T-score lt -2.0 at both lumbar spine and total
  hip
• Treatments (using approved dosing regimens)
• Alendronate 70 mg OW (standard am dosing) n219
• Risedronate 5 mg daily (post-meal dosing) n222
• Placebo n108

102
Approved Dosing Regimens

• Alendronate sodium (Fosamax)
• At least 1/2 hour before the first food,
  beverage, or medication of the day, upon arising
  for the day
• Risedronate (Actonel)
• At least 2 hours from any food or drink at any
  other time of the day, and at least 30 minutes
  before going to bed

Post-meal dosing approved outside the US
103
Risedronate Bioavailability
100
90
80
70
60
Relative Bioavailability ()
50
40
30
20
10
0
4 Hour Before Breakfast
1 Hour Before Breakfast
1/2 Hour Before Breakfast
2 Hour After Dinner
Absolute Bioavailability 0.63
Source Actonel Risedronate United States
Product Circular
104
Alendronate-Risedronate Comparison12 Month
Percent Change in BMD relative to baseline
Alendronate
Risedronate
Placebo
6

5
4


Change
3

2
1
0
-1
Lumbar Spine Femoral Neck
Trochanter Total Hip
p lt 0.001 p lt 0.05 Alendronate v
Risedronate
Hosking D I.O.F.meeting Lisbon 2002
105
Alendronate vs. Risedronate Comparison
Trial Markers of Bone Turnover
Plt0.001 Alendronate vs. risedronate
106
Alendronate vs. Risedronate Comparison Trial BSAP

0
-20

-40
Mean Percent Change


PlaceboRIS 5 mg DailyALN 70 mg Once Weekly
-60
0
3
6
12
Month
Plt0.001 Alendronate vs. Risedronate
107
Alendronate vs. Risedronate Comparison
TrialAdverse Experiences

• Similar tolerability seen in all three groups

Gastric or duodenal perforation, ulcer or bleed
108
Alendronate vs. Risedronate Comparison
TrialSummary

• Significantly greater increases in BMD at both
  the hip and spine with alendronate compared to
  risedronate over 12 months
• 70 greater at the lumbar spine (4.8 vs. 2.8)
• 3.7-fold greater at the hip trochanter (3.3 vs.
  0.9)
• 3-fold greater at the total hip (2.7 vs. 0.9)
• Significantly greater increases in BMD with
  alendronate seen early (6 months) at spine,
  trochanter, and total hip
• Significantly greater effect on markers of bone
  resorption with alendronate compared to
  risedronate
• Greater decrease in resorption with alendronate
  seen early (at one month)
• Similar tolerability was seen between alendronate
  and risedronate, including upper gastrointestinal
  adverse experiences

109
Conclusions

• In this first head-to-head double-blind
  comparison of alendronate and risedronate for
  treatment of osteoporosis in postmenopausal
  women

• Alendronate produced significantly greater
  increases in BMD at both spine and hip sites
• Alendronate produced significantly greater
  reduction in bone resorption
• Differences seen early, with greater effects of
  alendronate seen at 1 month for resorption and 6
  months for BMD

110
Alendronate-Risedronate Comparison12 Month
Percent Change in BMD relative to baseline
Alendronate
Risedronate
Placebo
6

5
4


Change
3

2
1
0
-1
Lumbar Spine Femoral Neck
Trochanter Total Hip
p lt 0.001 p lt 0.05 Alendronate v
Risedronate
Hosking D I.O.F.meeting Lisbon 2002
111
SummaryImpact of Therapeutic Trials in Clinical
Decision Making

• Prescribe the drug that has the most significant
  impact on reversing or stopping the consequences
  of osteoporosis.
• Fracture is the most serious outcome of
  osteoporosis Use the drugs that reduce fractures
  at all sites rapidly, consistently, and in a
  sustained fashion. This will afford the
  physician and the patient the best clinical
  benefit.

112
Historical Controls

• Randomized, controlled trials (RCTs) are the
  highest level of evidence.
• Analyses with historical controls or any other
  than the original randomization groups
• are NOT true RCTs
• are much more susceptible to bias
• are a lower level of evidence