Download - Economic Evaluation of Clodronate and Zoledronate in Patients Diagnosed With Metastatic Bone Disease From the Perspective of Public and Third Party Payors in Brazil

N

S

an

Clinical Therapeutics/Volume 33, Number 11, 2011

Economic Evaluation of Clodronate and Zoledronate inPatients Diagnosed With Metastatic Bone Disease From thePerspective of Public and Third Party Payors in Brazil

Marcelo Cunio Machado Fonseca, MD, MSc1,2;Gabriela Tannus Branco de Araújo, MSc1,2; Helder Etto, MSc2; Alexandre Schiola, MD3;

atália Santoni, PharmD3; and Marcio Machado, PharmD, PhD4

1Federal University of São Paulo, São Paulo, Brazil; 2Axia.Bio Consulting, São Paulo, Brazil; 3Bayerhering Pharma Brazil, São Paulo, Brazil; and 4University of Toronto, Toronto, Ontario, Canada

ABSTRACTBackground: Metastatic bone disease (MBD) is re-

sponsible for �99% of malignant tumors that affectthe bone. MBD patients have increased risk of skeletalcomplications that are often dramatic and result in lossof function or disability, leading to rapid deteriorationof quality of life. Bisphosphonates have become thestandard therapy for the treatment and prevention ofskeletal-related events (SREs).

Objective: The objective of this study was to evalu-te the cost-effectiveness of zoledronate and clodro-ate in the prevention of SREs in patients with MBD.

Methods: A pharmacoeconomic analysis was per-formed for a hypothetical cohort of patients with MBDto compare the costs and consequences of the use ofclodronate and zoledronate for treatment and preven-tion of SREs in MBD in Brazil. The model was con-structed using decision analysis techniques. Costs weredescribed in 5 categories—drugs, physician visits, hos-pitalizations, surgical/medical care, and laboratorytests—and were reported in 2008 Brazilian reais (1BRL � 0.54 US dollar). Quality-adjusted life yearsgained was considered as an outcome. Sensitivity anal-yses tested model robustness.

Results: The total cost of treatment of MBD inBrazil for a 5-year time-horizon was R$46,313 withclodronate and R$50,319 with zoledronate. The es-timated number of quality-adjusted life years was2.00 and 1.90 for clodronate and zoledronate, re-spectively. Cost-effectiveness ranking was un-changed when model time-horizon was changed to 1or 10 years. Univariate analysis revealed the incidenceof osteonecrosis as a sensitive parameter in the model.Multivariate analysis confirmed base-case results, inwhich �60% of model iterations favored clodronate
over zoledronate.
November 2011

Conclusion: The present pharmacoeconomic evalu-ation, under the premises presented, found that clodro-nate was dominant over zoledronate from both thepublic and the private health care perspectives inBrazil. (Clin Ther. 2011;33:1769–1780) © 2011Elsevier HS Journals, Inc. All rights reserved.

Key words: clodronate, economic evaluation, meta-static bone disease, zoledronate.

INTRODUCTIONMetastatic bone disease (MBD) is responsible for�99% of the malignant tumors that affect the bone,and every malignant tumor can eventually metastasizeto the bone.1

MBD originates more frequently from breast, lung,prostate, kidney, and thyroid cancers, and the preva-lence is higher in those with prostate and breast carci-nomas (�80% of MBD). Approximately 65% to 75%of patients with breast or prostate cancer will developbone metastases during the course of their disease.2

Bone metastases are more commonly located in thevertebrae, ribs, pelvis, and femur. Eventually the loca-tion of the primary tumor remains unknown, and onlythe presence of MBD is identified.3

MBD patients have increased risk of skeletal com-plications that are often dramatic and result in loss offunction or disability, leading to rapid deterioration ofquality of life. As a result, patients live with the con-stant possibility of developing skeletal morbidity for along period of time, which is linked with emotional

Accepted for publication September 20, 2011.doi:10.1016/j.clinthera.2011.09.0250149-2918/$ - see front matter

© 2011 Elsevier HS Journals, Inc. All rights reserved.

1769

Mcse

ttcicms

cafveti

Clinical Therapeutics

distress.4 Moreover, patients with MBD have an aver-age survival ranging from �6 months to 48 months;depending on the type of tumor, nearly 5% to 40% ofpatients reach a 5-year survival.2

Clinically, bone pain is the main symptom of MBDand may be accompanied by bone-related events suchas local swelling, pathologic fractures, vertebral com-pression, and, in some cases, hypercalcemia. The in-jury, however, can progress asymptomatically and be-come evident only when a pathologic fracture occurs orlocal swelling appears, which often is associated mistak-enly with venous thrombosis.4,5 The clinical staging of

BD includes laboratory tests such as complete bloodount, electrolyte measurement, enzymes dosage, mea-urement of specific tumor markers, immunoglobulin lev-ls, specific proteins, and hormone dosage.6

MDB treatment includes orthopedic follow-up, radio-therapy, and systemic treatment (ie, endocrine therapy, che-motherapy, and bisphosphonates). The main goals of localtreatmentarepainrelief,maintenanceorrestorationof func-tion, neurologic decompression, and, when possible, localcontrol of tumor growth. The use of radiotherapy alone orin combination with surgical procedures is recommendedfor the last-mentioned goal, which should take into accountthe prognosis of the primary tumor, life expectancy, and thepatient’s general health status.7,8

Often used as an adjunct to radiotherapy, bisphos-phonates have become part of the standard therapy forthe treatment and prevention of skeletal-related events(SREs).9 However, owing to increased risk of renaloxicity caused by bisphosphonates (predominantlyhe intravenous forms) and their association with avas-ular osteonecrosis of the jaw and/or maxilla, monitor-ng of renal function, performance of an early oralomprehensive examination before the start of treat-ent, good oral hygiene to prevent infections, and pos-

ible tooth extraction are recommended.10

Zoledronate and clodronate are 2 bisphosphonates cur-rentlyapproved inBrazil for the treatmentandpreventionofSREs in MBD; each shows different pharmacologic proper-ties. Clodronate is administered orally in daily doses,whereas zoledronate is administered via intravenous infu-sions every 4 weeks. A recent meta-analysis of randomizedclinical trials showed minimal differences between the 2treatment options with regard to the reduction in the risk ofSREs but suggested that zoledronate is a potentially moreefficacious alternative.11 However, evidence has shown thatzoledronate is associated with the occurrence of osteonecro-
sis of the jaw, which may result in increased health care
1770

resource use and costs, as well as a negative impact on pa-tients’ quality of life.12 In terms of treatment costs in Brazil,lodronate and zoledronate are equivalent. All those char-cteristics lead to uncertainty regarding the relative cost-ef-ectiveness of the 2 comparators in both the public and pri-ate markets in Brazil. Thus, the aim of this study was tovaluate the cost-effectiveness of the pharmacologic alterna-ives zoledronate and clodronate in the prevention of SREsn patients with MBD.

METHODSA pharmacoeconomic analysis was performed for ahypothetical cohort of patients with MBD to comparethe costs and consequences of using clodronate andzoledronate for treatment and prevention of SREs inMBD patients in Brazil. The simulation model wasconstructed using decision analytic techniques.13

The study was set in the Brazilian health care systemin 2008. In Brazil there are 2 analytical perspectives toconsider, the public from the Ministry of Health andthe private from the supplementary medicine system,as these 2 markets show different characteristics andpeculiarities, especially cost weights. For both eco-nomic perspectives, direct costs presented in 2008 reais(1 BRL � 0.54 USD) were included. The clinical andhumanistic consequences considered in the pharmaco-economic model were life years (LYs) free of SREs andquality-adjusted life years (QALYs), respectively.

A Markov model was designed to represent a cohortof patients diagnosed with MBD. Figure 1 shows theMarkov model used. The pharmacoeconomic modelwas developed based on the clinical events of the dis-ease in question. Consequently, we identified 4 basichealth states for the mathematical model, describedlater: (1) No SREs: The patient has no SREs and, there-fore, is in a state of symptom remission and diseasecontrol. In this health state bisphosphonate is admin-istered along with medical follow-up. (2) SREs: Thepatient has at least 1 of the 3 studied SREs: pathologicfracture, radiation or orthopedic surgery, or hypercal-cemia. This health state is characterized by high utili-zation of medical resources, including surgery, hospi-talization, drugs, diagnostic tests, medical follow-up,and others. Each skeletal event was linked to a partic-ular use of medical resources for each specific event. (3)Osteonecrosis: The patient presents an osteonecrosisthat may or may not be related to the pharmacologic
alternative in use. Treatment ranges from drugs to hy-
Volume 33 Number 11

M.C.M. Fonseca et al.

perbaric chamber sessions. (4) Death (absorbing state):death from any cause, including disease complications.

Three time horizons were analyzed: 1, 5 (base-case),and 10 years with 1-year time cycles. At the 10-yeartime horizon, more than 90% of the patients reached

Cost-effectiveness ofclodronate and zoledronatein Brazil

Clodronate

Zoledronate

No SREs

No SREs

With SREs

With SREs

With osteonecrosis

With osteonecrosis

Dead

Dead

N

W

D

Pf

Rs

H

W

O

D

N

W

D

Pf

Rs

H

W

O

D

p

p

p

p

p

p

p

p

p

p

p

p

p

p

p

p

p

p

0

1

0

0

0

1

0

0

Figure 1. Schematic representation of the Markov mnate and zoledronate in Brazil. SRE � skele

the absorbing health state. The model was initiated

November 2011

with patients with MBD who were starting therapywith clodronate or zoledronate. The therapeutic doseswere 1600 mg per day for oral clodronate and 4 mgevery 4 weeks, intravenously, for zoledronate. Treat-ment with one of these therapies stopped only when the

No SREs

With SREs

No SREs

With SREs

No SREs

With SREs

No SREs

With SREs

With SREs

With SREs

Es

No SREs

With SREsic

n oro bone

lcemia

Es

crosis

With osteonecrosis

With Osteonecrosis

Osteonecrosis

DeathDead

Dead

With osteonecrosis

Dead

With osteonecrosis

Dead

Dead

Dead

No SREs

With SREs

Osteonecrosis

Death

Es

ic

n oro bone

lcemia

Es

crosis

[+]

[+]

[+]

[+]

SREs_clodronate

prob_No_SREs_clodronate

s_clodronate

prob_SREs_clodronate

th_clodronate

prob_death_clodronate

ture_clodronate

iosurg_clodronate

ercalc_clodronate

s_clodronate

onec_clodronate

prob_osteonec_clodronate

th_clodronate

SREs_zoledronate

s_zoledronate

th_zoledronate

ture_zoledronate

iosurg_zoledronate

ercalc_zoledronate

s_zoledronate

prob_No_SREs_zoledronate

prob_SREs_zoledronate

onec_zoledronate

prob_osteonec_zoledronate

th_zoledronate

prob_death_zoledronate

3

1

esigned to compare the cost-effectiveness of clodro-lated event.

o SREs

ith SR

eath

athologractures

adiatiourgery t

yperca

ith SR

steone

eath

o SREs

ith SR

eath

athologractures

adiatiourgery t

yperca

ith SR

steone

eath

rob_No_

rob_SRE

rob_dea

rob_frac

rob_rad

rob_hyp

rob_SRE

rob_oste

rob_dea

rob_No_

rob_SRE

rob_dea

rob_frac

rob_rad

rob_hyp

rob_SRE

rob_oste

rob_dea

odel dtal-re

patient died.

1771

eSaemaa

tPcbbst


Clinical data to populate the Markov model (ie,transition probabilities) derived from a previouslypublished meta-analysis11 and results are presented inTable I. In the case of osteonecrosis, the transitionprobabilities were obtained from Krueger et al.12 Allprobabilities were reweighted in fixed proportions to100% transition probability of events in the phar-macoeconomic model.

The utility values used to calculate the number ofQALYs at a given time horizon were taken from thestudy by Dranitsaris and Hsu.14 In this study, it wasstimated that the utility of the “no SREs” and “withREs” health states were 0.64 (95% CI, 0.53–0.76)nd 0.46 (95% CI, 0.37–0.54), respectively. Adversevents were not included in the pharmacoeconomicodel. In this case, it was assumed that both compared

lternatives had similar incidence rates and clinical char-cteristics of adverse events (eg, severity, duration).

For both the public and private markets, the costs ofhe pharmacologic alternatives came from the Drugrice List from the Pharmaceutical Price Counsel. Suchosts represent the ex-factory wholesaler list price ofranded and generic drugs in Brazil with equal weightetween the products. All other types of medical re-ources for the public perspective were obtained fromhe DATASUS database15 and from the Outpatient

and Hospital Brazilian Unified Health System lists.16

The unit costs for these same resources, except for theBrazilian supplementary medicine perspective, werederived from the database of Impact Medical AuditCompany, which audits 170,000 lives for Brazilian pri-vate medicine companies. Both cost databases arebased on medical/service claims from public or private

Table I. Data on the incidence of skeletal-related evecost-effectiveness of clodronate and zoledro

Events

Pharmac

Clodronate

All bone events 31.5%Fractures 20.0%Radiotherapy/bone surgery 21.3%Hypercalcemia 9.3%Death 19.0%Osteonecrosis 0.0%

health care providers in Brazil and represent the actual

1772

cost of resources paid by each system. All costs in-cluded in the pharmacoeconomic model were de-scribed in 5 categories: drugs, physician visits, hospi-talizations, surgical/medical care, and laboratory tests.Table II shows all costs included in the pharmacoeco-nomic model for the public health care perspective.

Univariate and multivariate sensitivity analyseswere used to determine the robustness of the phar-macoeconomic model. The univariate analysis assessedand identified the most influential parameters for themodel results (eg, cost of pharmacologic alternatives).Monte Carlo simulations were used with 10,000 iter-ations among all variables of the model. Then, in themultivariate analysis, log-normal distributions wereused for costs, beta distributions for transition proba-bilities, and triangular distributions for utilities. Whereparameter variability was not available, it was as-sumed a value of 10% of the original parameter toserve as standard deviation.

The analysis did not require submission or approvalfrom an ethics review board, as it used only secondaryor published data from the literature. Thus, the re-search does not impose any risk to individuals, as onlyhypothetical patients were considered.

RESULTSCost Analysis—Public Perspective

The total cost of treatment of MBD in Brazil in thebase-case time horizon was R$46,313 with clodronateand R$50,319 with zoledronate.

The analysis of influence showed that the cost ofmedications represented �90% of the total cost forboth therapeutic alternatives. Surgeries and medical care

sed in the pharmacoeconomic model to evaluate thein Brazil.

ic Alternatives

SourceZoledronate

24.4% Machado et al, 200915.5% Machado et al, 200914.1% Machado et al, 2009

2.6% Machado et al, 200922.9% Machado et al, 2009

7.0% Krueger et al, 2007

nts unate

olog

were the second most influencial cost parameters (aver-

Volume 33 Number 11


Table II. Costs included in the pharmacoeconomic model for the Brazilian public health care system perspective.

Comparator Disease State Substate Resource Used Description Units Cost per Unit

Cost in DiseaseState (per

Cycle) Cost References

Clodronate No SREs N/A Drugs Clodronate (1600mg/d)

Per mg R$0.021 R$12,396 ANVISA, 2007a

Physician visit Monthly physicianvisit

Per visit R$8 R$91 Procedure SIA (0701204)b

With SREs N/A Drugs Clodronate(1600mg/d)

Per mg R$0.021 R$12,396 ANVISA, 2007a



Pathologicfractures

Surgical/medicalcostsc

Endoprosthesis(66%)

Per procedure R$1870 R$1234 Procedure SIH (39703231)b


Osteosyntheses(13%)



Hartshill rectangle(5.8%)

Per procedure R$764 R$44 Procedure SIH(’’93396406)b


Amputation(4.4%)

Per procedure R$281 R$12 Average procedures SIH(’39001083, ’39001105,’39001172, ’39003108,’39003159, ’39004139,’39005062, ’39036103)b


Pedicular screws(3%)



Harrington-Luqueinstrumentation(3%)

Per procedure R$610 R$18 Procedure SIH(’93398476)b


External fixators(3%)

Per procedure R$655 R$20 Average procedures SIH(’93399103, ’93399111,’93399120, ’93399138,’93399146)b


Girdlestonearthroplasty(1.5%)

Per procedure R$628 R$9 Average procedures SIH(’39045170, ’39003051,’39028143, ’39016056,’39015050, ’39014053,’39027120, ’39028127,’39026140, ’39025128,’39003078, ’39021076,’39023079, 39022072,39014096, 39018059,39017052, 39008100,39009106, 39029123,39003124, 39026124,39027147, 39018121,39016129, 39022145,39003140b

Hospitalizationd Average of 5.7 d Per hospitalization R$76 R$432 Procedure SIH (68500017)b

Radiation orsurgery

Surgical/medicalcostse

Radiation (50%) Per session R$19 R$116 Procedure SIH (85300829)b


Bone fracturesurgery (50%)


Hospitalizationf Average of 3.6 d Per hospitalization R$76 R$136 Procedure SIH (68500017)b

Hypercalcemia Diagnostic costg Serum calcium(monthly tests)

Per test R$2 R$22 Procedure SIA (1101109)b

Diagnostic costg TSH test (monthlytest)


Hospitalizationh 3 d in hospital forpatient hydration

Per hospitalization R$76 R$227 Procedure SIH (68500017)b

(continued)

November 2011 1773


Table II (continued).

Comparator Disease State Substate Resource Used Description Units Cost per Unit



Zoledronate No SREs N/A Drugs Zoledronate (4 mgevery 4 wk)

Per mg R$249 R$11,947 ANVISA, 2007a

Drugs IV infusion Per infusion R$174 R$2093 Procedure SIH (79700888)b

With SREs N/A Drugs Zoledronate (4 mgevery 4 wk)

Per mg R$249 R$11,947 ANVISA, 2007a

Drugs IV infusion Per infusion R$174 R$2093 Procedure SIH (79700888)b



Pathologicfractures


Endoprosthesis(66%)



Osteosyntheses(13%)



Hartshill rectangle(5.8%)



Amputation(4.4%)

Per procedure R$281 R$12 Average procedures SIH(’39001083, ’39001105,’39001172, ’39003108,’39003159, ’39004139,’39005062, ’39036103)b


Pedicular screws(3%)



Harrington-Luqueinstrumentation(3%)



External fixators(3%)

Per procedure R$655 R$20 Average procedures SIH(’93399103, ’93399111,’93399120, ’93399138,’93399146)b


Girdlestonearthroplasty(1.5%)

Per procedure R$628 R$9 Average procedures SIH(’39045170, ’39003051,’39028143, ’39016056,’39015050, ’39014053,’39027120, ’39028127,’39026140, ’39025128,’39003078, ’39021076,’39023079, 39022072,39014096, 39018059,39017052, 39008100,39009106, 39029123,39003124, 39026124,39027147, 39018121,39016129, 39022145,39003140b

Hospitalizationd Average of 5.7 d Per hospitalization R$76 R$432 Procedure SIH (68500017)b

Radiation orsurgery


Radiation (50%) Per session R$19 R$116 Procedure SIH (85300829)b


Bone fracturesurgery (50%)


Hospitalizationf Average of 3.6 d Per hospitalization R$76 R$136 Procedure SIH (68500017)b

Hypercalcemia Diagnostic costg Serum calcium(monthly tests)


Diagnostic costg TSH test (monthlytest)


Hospitalizationh 3 d in hospital forpatient hydration

Per hospitalization R$76 R$227 Procedure SIH (68500017)b

(continued)

1774 Volume 33 Number 11


age of 4.5% of total cost). Therefore, the cost of bispho-sphonates was considered the driver of the total cost ofMDB treatment in Brazil. Table III presents the analysisof influence on the total cost of all medical resources used.

Cost Analysis—Private PerspectiveSimilar results were seen in the private perspective.

The total cost of MBD treatment in Brazil from theperspective of the supplementary system for 5 years oftreatment was R$49,004 with clodronate and

Table II (continued).

Comparator Disease State Substate Resource Used Description

Osteonecrosis N/A Drugs Zoledronate (4every 4 wk)

Drugsi Ciprofloxacin(1000 mg/d)

Drugs IV infusion

Physician visit Monthly physicvisit

Diagnostic costsi Facialradiographies

Surgical/medicalcostsi

Hyperbaricchamber (15sessions)

Surgical/medicalcostsi

Sequestrectom(60%)

Hospitalizationj 3 d in hospitalpatient hydrati

N/A � not applicable; R$� reais; SIA/SUS Outpatient Inforskeletal-related event; TSH � thyroid-stimulating hormone.aNational Agency of Sanitary Surveillance—medical productbAvailable at http://sigtap.datasus.gov.br/cResources used and utilization rate were obtained from Nfactors in pathologic fractures secondary to metastatic tum

dAverage days of hospitalization obtained from DATASUSwww.datasus.gov.br.

eResources used and utilization rate were assumed at 50% efAverage days of hospitalization obtained from DATASUSgov.br

gResources used and utilization rates were obtained at Fariastreatment [in Portuguese]. Arq Bras Endocrinol Metabol. 2005

hResources used and utilization rate assumed were 3 d of hoiResources used and utilization rate were obtained at Júnassociated with use of biphosphonates. Rev Soc Bras Cir CranBuco_julho-setembro.pdf.jResources used and utilization rate assumed were 3 d of hokTemporal WF. Aspectos econômicos da oxigenoterapia hipmil.br/revistas/2002/04_02.pdf.

R$53,076 with zoledronate.

November 2011

The analysis of influence for the private systemshowed comparable results to those of the public per-spective, with slight numerical variation (Table III andSupplemental Table in the online version atdoi:10.1016/j.clinthera.2011.09.025).

Consequence Analysis—Public and PrivatePerspectives

Using the base-case time horizon of 5 years, zole-dronate generated on average 1.76 life years without

Units Cost per Unit



r mg R$249 R$11,947 ANVISA, 2007a

r mg R$0.008 R$2782 ANVISA, 2007a

r infusion R$174 R$2093 Procedure SIH (79700888)b

r visit R$8 R$91 Procedure SIA (0701204)b

r test R$6 R$78 Procedure SIA (1010201)b

r session R$329 R$4935 Temporal, 2002k

r procedure R$554 R$332 Procedure SIH (39003310)b

r hospitalization R$76 R$227 Procedure SIH (68500017)b

n System; SIH/SUS � Hospital Information System; SRE �

ttp://www.anvisa.gov.br/

i DK, de Alverga Neto CC, Baptista AM, et al. Prognosticlinics (Sao Paulo). 2006;61:313–320.0 description: malignant neoplasm of bone and cartilage.

0 description: fracture of other limb bones. www.datasus.

ypercalcemia of malignancy: clinical features, diagnosis and16–824.lization to treat dehydration.C, Rocha RI, Massarollo LCB. Mandibular osteonecrosislofac. 2007;10:93–96. http://www.abccmf.org.br/Revi/pdf/

ization for rehabilitation.ica (OHB). RMAB 2002; 52:20–24. http://www.dirsa.aer.

mg Pe

Pe

Pe

ian Pe

Pe

Pe

y Pe

foron

Pe

matio

list. h

arazakors. CICD-1

ach.ICD-1

ML. H;49:8spita

ior Wiomaxi

spitalerbár

SREs and clodronate 1.81 years. The QALYs were

1775

http://dx.doi.org/10.1016/j.clinthera.2011.09.025

http://www.anvisa.gov.br/

http://sigtap.datasus.gov.br/

http://www.datasus.gov.br



http://www.abccmf.org.br/Revi/pdf/Buco_julho-setembro.pdf

http://www.abccmf.org.br/Revi/pdf/Buco_julho-setembro.pdf

http://www.dirsa.aer.mil.br/revistas/2002/04_02.pdf



2.00 and 1.90, respectively, for clodronate andzoledronate.

Cost-effectiveness—Public and Private PerspectivesTable IV illustrates the results of the cost-effective-

ness analysis of clodronate and zoledronate in treat-

Table III. Analysis of influence on costs of all medicalmetastatic bone disease in Brazil, in public

AlternativeTherapies Cost Type

Pub

Costs

Clodronate Drugs R$44,068Physician visits R$324Hospitalizations R$494Surgical/medical R$1390Laboratory tests R$37TOTAL R$46,313

Zoledronate Drugs R$47,615Physician visits R$307Hospitalizations R$479Surgical/medical R$1896Laboratory tests R$21TOTAL R$50,318

Table IV. Incremental analysis of costs per life years frin Brazil, from the public health care persp

Perspective Time-Horizon (y) Comparato

Public 1 ClodronateZoledronat

5 ClodronateZoledronat


Private 1 ClodronateZoledronat



ICER � incremental cost effectiveness ratio.*Effectiveness is represented by life years free of skeletal-rela

1776

ment and prevention of MBD in Brazil. In the analysisof costs per life years free of SREs, zoledronate is dom-inated by clodronate in both perspectives and time ho-rizons. When we analyzed incremental costs perQALYs gained (Table V), zoledronate was also domi-nated by clodronate (ie, higher cost and lower effec-

urces used (in a period of 5 years) in the treatment ofrivate perspectives.

rspective Private Perspective

Influence (%) Costs Influence (%)

95.15 R$44,068 89.930.70 R$1280 2.611.07 R$792 1.623.00 R$2750 5.610.08 R$113 0.23

100.00 R$49,004 100.0094.63 R$47,615 89.71

0.61 R$1215 2.290.95 R$718 1.353.77 R$3428 6.460.04 R$101 0.19

100.00 R$53,076 100.00

skeletal-related events of clodronate and zoledronate.

Cost Effectiveness* ICER

R$13,536 0.58R$15,313 0.54 DominatedR$46,313 1.81R$50,318 1.76 DominatedR$64,995 2.86R$67,140 2.64 DominatedR$14,752 0.58R$16,632 0.54 DominatedR$49,004 1.81R$53,076 1.76 DominatedR$68,533 2.86R$70,565 2.64 Dominated

ents.

resoand p

lic Pe

ee ofective

r

e

e

e

e

e

e

ted ev

Volume 33 Number 11


tiveness) in both perspectives. The cost-effectivenessranking remained unchanged when model time hori-zon was changed to 1 or 10 years.

Sensitivity Analysis—Public and PrivatePerspectives

Results from the univariate sensitivity analysis high-lighted sensitive parameters in the Markov model. Theincidence of osteonecrosis was identified as the mostsensitive parameter in the model. When using the samerate of osteonecrosis in both treatment arms, the cost-effectiveness ranking is inverted and zoledronate be-comes the dominant strategy.

The multivariate sensitivity analysis showed nochanges compared with the initial results of the phar-macoeconomic model, in which �60% of model iter-ations favored clodronate over zoledronate. In otherwords, the pharmacoeconomic model was robust tosimultaneous changes in all input parameters, for boththe public and private perspectives. Figure 2 depicts theincremental cost-effectiveness per QALY gained result-ing from the Monte Carlo simulation for the publichealth care perspective. Results for the private healthcare perspective were very similar (data not shown).

DISCUSSIONThe present pharmacoeconomic analysis of clodronate

Table V. Incremental analysis of costs per quality-adjfrom the private health care perspective.

Perspective Time-Horizon (y) Comparato

Public 1 ClodronateZoledronat



Private 1 ClodronateZoledronat



ICER � incremental cost effectiveness ratio.*Effectiveness is represented by quality-adjusted life years.

and zoledronate for the treatment of MBD showed, as

November 2011

in a recent meta-analysis,11 that the alternative thera-pies are very similar in relation to their clinical conse-quences. The resulting difference in terms of QALYsand LYs free of SREs was considered small. Clodro-nate compared with zoledronate has only 18 more daysfree of bone events within a 5-year treatment period.Therefore, such difference may not be considered rele-vant to some clinicians.

However, in economic terms, the cost of the phar-macologic alternatives came to represent nearly85% and 90% of the total cost of treatment andprevention of MBD, as shown by the analysis ofinfluence of costs, for the private and public markets,respectively. Therefore, with similar clinical out-comes and drug costs as drivers of the economicresults, the drug with lowest price can be consideredcost effective in Brazil. In this case, the treatmentcost of clodronate was estimated to be about 7% lessexpensive than that of zoledronate because of itsassociated infusion costs.

Hence, the cost difference between zoledronate andclodronate may be even greater in Brazil, both in thepublic and private markets. In the public market thecosts tend to be lower than in the private market, asthe drugs are purchased through auctions, making theprices well below the factory price. In the private mar-

life years free of clodronate and zoledronate in Brazil,

Cost Effectiveness* ICER

R$13,536 0.46 Cost-savingR$15,313 0.46R$46,313 2.00R$50,318 1.90 DominatedR$64,995 2.89R$67,140 2.62 DominatedR$14,752 0.46R$16,632 0.46 DominatedR$49,004 2.00R$53,076 1.90 DominatedR$68,533 2.89R$70,565 2.62 Dominated

usted

r

e

e

e

e

e

e

ket, on the other hand, the drugs are acquired accord-

1777

tnbahetemagtcotrAtpidbic

tfoncabh

cfbttpccat


ing to published factory prices on which commercialdiscounts may apply.

A similar publication by Botteman et al17 evaluatedhe economic consequences of clodronate, ibandro-ate, pamidronate, and zoledronate in patients withreast cancer regarding the prevention of SREs. Theylso used a decision analytic model with a 10-year timeorizon. The results favored zoledronate as the cost-ffective alternative from the perspective of the Na-ional Health Service in the United Kingdom. How-ver, the clinical data used in the pharmacoeconomicodel derived from only 1 trial for each therapeutic

lternative. A systematic review of the literature re-arding the clinical data was not used, and thereforehe presence of selection bias for the clinical input dataould not be overruled. In this case, the clinical resultsf zoledronate were largely superior to their compara-ors, thus deflating the incremental cost-effectivenessatio of zoledronate over the other bisphosphonates.nother important point was the cost of the compared

reatment alternatives, which showed no difference inrice between zoledronate and pamidronate (accord-

ng to the British National Formulary) but a large priceifference between zoledronate and clodronate. Thus,oth the drug cost difference and the clinical data used

n the Botteman et al17 study clearly determined its

R$22K

R$17K

R$12K

R$7K

R$2K

–R$3K

–R$8K

–R$13K

–R$18K

–R$23K

–R$28K–0.210 QALYs –0.060 QALYs 0.09

Incre

Incr

emen

tal C

ost

Figure 2. Incremental cost-effectiveness scatterplot cpublic health care perspective. Note: the etiveness results. QALY � quality-adjusted l

ost-effectiveness results.

1778

As confirmed by the univariate sensitivity analysis,he main clinical parameter in our model responsibleor the cost-effectiveness results was the probability ofsteonecrosis. The analysis considered that there waso incidence of osteonecrosis in patients treated withlodronate. In fact, since the first osteonecrosis casesssociated with bisphosphonates were reported in theeginning of this decade,18,19 more than 3000 casesave been described to date in the medical literature.20

The majority of the cases were associated with amino-bisphosphonate treatments, mainly zoledronate. Dataon osteonecrosis of the jaw due to alkylbisphospho-nates (such as clodronate) are rare. Actually, in a recentsystematic review of the literature,21 of 27 identifiedases of alkylbisphosphonate-induced osteonecrosisrom 1966 to 2010, 10 of the patients were on alkyl-isphosphonate monotherapy, and of these only 3 pa-ients had MBD and were on clodronate therapy. Also,here is evidence (in animals) that various aminobis-hosphonates have necrotic functions; clodronate, inontrast, has antinecrotic effects on the bone. Actually,lodronate lessened or averted the necrosis induced byn aminobisphosphonate when both were adminis-ered in combination. Thus, the authors22 concluded

that “clodronate could be useful as a combination drugwith aminobisphosphonates for preventing their ne-

Ys

al Effectivenss0.240 QALYs 0.390 QALYs

aring clodronate and zoledronate in Brazil, from therepresents the 95% CI of the incremental cost-effec-r.

0 QAL

ment

ompllipseife yea

crotic actions while retaining their anti–bone-resorp-

Volume 33 Number 11


tive effects (ABREs)” and that “clodronate could alsobe useful as a substitution drug for nitrogen-containingbisphosphonates (NBPs) in patients at risk of osteone-crosis of jaw bones.”

This pharmacoeconomic evaluation has many limi-tations. The most important relates to the use of adecision analytic model. In this case, the model at-tempts to mimic the treatment and prevention of SREsin patients with MBD. As in all the model simulations,not all data being input are from a robust source ofevidence. For use in the current pharmacoeconomicmodel, assumptions were made for some missing infor-mation, such as the use of medical resources, adverseevents, and rate of osteonecrosis. To that extent, weapplied univariate and multivariate sensitivity analysisto test model robustness.

In addition, the pharmacoeconomic model did notuse data such as measures of patient adherence to thecompared alternatives, which may generate conse-quences in the clinical and economic results found. Ad-herence, compliance, and persistence to prescribedtreatments are linked to many inherent characteristicsof the pharmacotherapy and/or patient behaviors. It iswell known that administration routes and dosingschemes, both differentiated among bisphosphonatessuch as clodronate and zoledronate, can have a signif-icant impact on findings. Thus, the results reportedhere should be restricted to patients who are able toadhere to bisphosphonate therapy according to theirprotocol established in the clinical trial setting.

Furthermore, adverse drug reactions were also notincluded in the cost-effectiveness analysis. For all ad-verse drug reactions except those with osteonecrosis, asimilar incidence and severity in both the clodronateand the zoledronate groups was assumed, and also as-sumed was that their clinical and economic impactwould be similar. The literature frequently reports gas-trointestinal intolerance with oral bisphosphonatesand hypocalcemia, acute phase reactions, osteonecro-sis, and renal toxicity with intravenous bisphospho-nates.23–26 To that extent, the lack of adverse drugreactions in the analysis seems to favor zoledronateover clodronate, and, consequently, the present cost-effectiveness findings are likely conservative.

Another limitation of the present study refers to theprojected cumulative incidence of SREs. Model inputdata derived from a meta-analysis of 1-year SRE inci-dence and reliable data for extrapolation were not
found in the literature. Therefore, a continuous SRE
November 2011

incidence was assumed across model cycles. Future re-search could identify whether there is an increasing, adecreasing, or even a stable risk of SREs overtime in thestudied population.

Finally, the applicability of the present findings toother jurisdictions is dependent on several factors, suchas availability of both clodronate and zoledronate inother markets, similarity in medical resource types,clinical practice and health care system properties,price weights between jurisdictions, and others. There-fore, data transferability between jurisdictions shouldbe performed cautiously.

CONCLUSIONClodronate demonstrated equivalent or higher benefitsat a lower cost for the treatment of MBD in Brazil. Thepresent pharmacoeconomic evaluation, under thepremises presented, found that clodronate was domi-nant over zoledronate from both the public and theprivate health care perspectives.

CONFLICT OF INTERESTThe present study was sponsored by Bayer ScheringPharma Brazil, the maker of Bonefos (sodiumclodronate).

ACKNOWLEDGMENTSDrs. Cunio Machado Fonseca and Machado were re-sponsible for the development of the model used in thisstudy. Ms. Tannus Branco de Araújo, Mr. Etto, andDr. Santoni were responsible for the clinical and eco-nomic data research. Drs. Fonseca, Schiola, Machadoand Ms. Tannus Branco de Araújo were responsible forthe development and final revision of the manuscript.

SUPPLEMENTARY MATERIALA supplemental table accompanying this article can befound in the online version at doi:10.1016/j.clinthera.2011.09.025.

REFERENCES1. Diel IJ, Kauffamann M, Bastert G. Metastatic Bone Diseases,

Fundamental and Clinical Aspects. Berlin, Germany: SpringerVerlag; 1994:1–11.

2. Coleman RE. Skeletal complications of malignancy. Cancer.1997;80:1588–1594.

3. Donato J. Tumores ósseos.In: Metástase de Carcinoma. Rio deJaneiro, Brazil: Rocca; 2001.

4. British Association of Urological Surgeons. Systemic
management of metastatic bone disease—Guidelines on
1779



1

1

1

1

2

2L. Osteonecrosis of the jaw induced

2

2

2

2

2

2011;71:791–814.


Treatment. www.pchurology.co.uk/bausca/met_cap.pps. AcessedMarch 6, 2008.

5. Lichtenstein JR. Osteogenesis imper-fecta with floppy mitral valve syn-drome. Birth Defects Orig Artic Ser.1975;11:349.

6. Brown HK, Healey JH. Metastaticcancer to bone. In: DeVita VT Jr,Hellman S, Rosenberg AS. Principlesand Practice of Oncology. 6th ed. Phila-delphia, Pa: Lippincott Raven; 2001.

7. British Association of Surgical On-cology. The management of meta-static bone disease in the UnitedKingdom. Eur J Surg Oncol. 1999;25:3–23.

8. Healey JH, Brown HK. Complica-tions of bone metastases. Cancer.2000;88:2940–2950.

9. Brown JE, Neville-Webbe H, Cole-man RE. The role of bisphospho-nates in breast and prostate cancers.Endocr Relat Cancer. 2004;11:207–224.

10. Hungria VTM. Doença óssea emmieloma múltiplo [Bone disease inmultiple myeloma]. Rev Bras HematolHemoter. 2007;29:60–66.

11. Machado M, Cruz LS, Tannus G,Fonseca M. Efficacy of clodronate,pamidronate, and zoledronate inreducing morbidity and mortality incancer patients with bone metasta-sis: a meta-analysis of randomizedclinical trials. Clin Ther. 2009;31:962–979.

12. Krueger CD, West PM, Sargent M, etal. Biphosphonate-induced osteone-crosis of the jaw. Ann Pharmacother.2007;41:276–284.

13. Weinstein MC, Fineberg HV. ClinicalDecision Analysis. Philadelphia, Pa:WB Saunders; 1980:228–265.

14. Dranitsaris G, Hsu T. Cost-utilityanalysis of prophylactic pamidro-nate for the prevention of skeletalrelated events in patients with ad-vanced breast cancer. Support CareCancer. 1999;7:271–279.

15. Banco de dados do Sistema Únicode Saúde—DATASUS. www.datasus.
gov.br. Accessed January 22, 2008.
1780

6. Sistemas Ambulatoriais e Hospital-ares do Sistema Único de Saúde—SIA/SIH SUS. w3.datasus.gov.br/siasih/siasih.php. Accessed January22, 2008.

7. Botteman M, Barghout V, StephensJ, et al. Cost effectiveness of bisphos-phonates in the management ofbreast cancer patients with bonemetastases. Ann Oncol. 2006;17:1072–1082.

8. Marx KA, O’Neil P, Hoffman P,Ujwal ML. Data mining the NCIcancer cell line compound GI(50)values: identifying quinone subtypeseffective against melanoma and leu-kemia cell classes. J Chem Inf ComputSci. 2003;43:1652–1667.

9. Ruggiero SL, Mehrotra B, Rosen-berg TJ, Engroff SL. Osteonecrosis ofthe jaws associated with the use ofbisphosphonates: a review of 63cases. J Oral Maxillofac Surg. 2004;62:527–534.

0. Wimalawansa SJ. Insight into bisphos-phonate-associatedosteomyelitisofthejaw:pathophysiology,mechanismsandclinical management. Expert Opin DrugSaf.2008;7:491–512.

1. Crépin S, Laroche ML, Sarry B, Merle

[email protected]

by clodronate, an alkylbisphos-phonate: case report and literaturereview. Eur J Clin Pharmacol. 2010;66:547–554.

2. Oizumi T, Yamaguchi K, FunayamaH, et al. Necrotic actions of nitrogen-containing bisphosphonates andtheir inhibition by clodronate, a non-nitrogen-containing bisphospho-nate in mice: potential for utilizationof clodronate as a combinationdrug with a nitrogen-containing bis-phosphonate. Basic Clin PharmacolToxicol. 2009;104:384–392.

3. Papapetrou PD. Bisphosphonate-associated adverse events. Hormones.2009;8:96–110.

4. Diel IJ, Bergner R, Grötz KA. Adverseeffects of bisphosphonates: currentissues. J Support Oncol. 2007;5:475–482.

5. Strampel W, Emkey R, Civitelli R.Safety considerations with bisphos-phonates for the treatment of osteo-porosis. Drug Saf. 2007;30:755–763.

6. Lewiecki EM. Safety of long-termbisphosphonate therapy for themanagement of osteoporosis. Drugs.

Address correspondence to: Gabriela Tannus Branco de Araújo, RuaSetembrino Woitechumas 38/4, Brooklin São Paulo – SP – Brazil. E-mail:

Volume 33 Number 11

http://www.pchurology.co.uk/bausca/met_cap.pps

http://www.pchurology.co.uk/bausca/met_cap.pps



http://w3.datasus.gov.br/siasih/siasih.php

http://w3.datasus.gov.br/siasih/siasih.php

mailto:[email protected]


Supplemental Table. Costs included in the pharmaco

DrugsHealth

ConditionsSubhealthConditions

MedicalResources Descriptio

Clodronate Without boneevents

N/A Drugs Clodronate (16mg/d)

Medical visit Monthly medic

With boneevents

N/A Drugs Clodronate (16mg/d)

Medical visit Monthly medic

Pathologicfractures

Surgical/doctorc Endoprosthesis(66%)

Surgical/doctorc Osteosynthesis

Surgical/doctorc Hartshill rectan(5.8%)

Surgical/doctorc Amputation (4.

Surgical/doctorc Pedicular screw

Surgical/doctorc Harrington-Luqinstrumentation

Surgical/doctorc External fixators

Surgical/doctorc Girdlestonearthroplasty (1.

Hospitalizationd 5.7 days averag

Radiation orbone surgery

Surgical/doctorcf Radiation (50%

Surgical/doctorcf Bone fracture s(50%)

Hospitalizationg 3.6 days averag

Hypercalcemia Diagnostictesth

Serum calcium(monthly tests)

Diagnostictesth

TSH (monthly t

Hospitalizationi 3 d hospitalizatfor patient hydr

Zoledronate Without boneevents

N/A Drugs Zoledronate (4every 4 wk)

Drugs Infusion IV

With boneevents

N/A Drugs Zoledronate (4every 4 wk)

Drugs Infusion IV

Medical visits Monthly medicvisits

Pathologicprocedures

Surgical/doctorc Endoprosthesis(66%)

Surgical/doctorc Osteosynthesis

Surgical/doctorc Hartshill rectan(5.8%)

Surgical/doctorc Amputation (4.

Surgical/doctorc Pedicular screw

economic model—private market.

n UnitCost per

Unit

Costs onHealth

Condition Sources

00 By mg R$0.0214 R$12,512 Kairos Magazine, May 2008a

al visit By visit R$39 R$468 AMB procedures tableb

00 By mg R$0.0214 R$12,512 Kairos Magazine, May 2008a

al visit By visit R$39 R$468 AMB procedures tableb

By procedure R$3378 R$2229 AMB procedures tableb

(13%) By procedure R$1547 R$201 AMB procedures tableb

gle By procedure R$1300 R$75.40 AMB procedures tableb

4%) By procedure R$1943,41 R$85.51 AMB procedures table, mean value52050017; 52070018; 52090027;5220244; 52140032; 52160017;52090019b

s (3%) By procedure R$540 R$16.20 AMB procedures tableb

ue(3%)

By procedure R$1700 R$51 AMB procedures tableb

(3%) By procedure R$1100 R$33 AMB procedures tableb

5%)By procedure R$1547.62 R$23.21 AMB procedures tableb

e By day ofhospitalization

R$200 R$1140 PROAHSA Bulletin no. 35 (categoryB hospital)e

) By section R$520 R$260 AMB procedures tableb

urgery By procedure R$2451.89 R$1225.94 AMB procedures tableb

e By day ofhospitalization


By test R$3.75 R$41.25 AMB procedures tableb

ests) By test R$24.30 R$302 AMB procedures tableb

ionation

By day ofhospitalization


mg By mg R$258,34 R$12,400 Kairos Magazine, May 2008a

By infusion R$35 R$420 AMB procedures tableb

mg By mg R$258.34 R$12,400 Kairos Magazine, May 2008a

By infusion R$35 R$420 AMB procedures tableb

al By visit R$39 R$468 AMB procedures tableb

By procedures R$3378 R$2229 AMB procedures tableb

(13%) By procedures R$1547 R$201 AMB procedures tableb

gle By procedures R$1300 R$75.40 AMB procedures tableb

4%) By procedures R$1943.41 R$85.51 AMB procedures table, mean value52050017; 52070018; 52090027;5220244; 52140032; 52160017;52090019b

s (3%) By procedures R$540 R$16.20 AMB procedures tableb

(continued)

November 2011 1780.e1


Supplemental Table (continued).

DrugsHealth

ConditionsSubhealthConditions

MedicalResources Description Unit

Cost perUnit

Costs onHealth

Condition Sources

Surgical/doctorc Harrington-Luqueinstrumentation (3%)

By procedures R$1,700 R$51 AMB procedures tableb

Surgical/doctorc External fixators (3%) By procedures R$1,100 R$33 AMB procedures tableb

Surgical/doctorc Girdlestonearthroplasty (1.5%)

By procedures R$1547.62 R$23.21 AMB procedures tableb

Hospitalizationd 5.7 d average By day ofhospitalization


Radiation orbone surgery

Surgical/doctorc,f Radiation (50%) By section R$520 R$260 AMB procedures tableb

Surgical/doctorc,f Bone fracture surgery(50%)

By procedure R$2451.89 R$1225.94 AMB procedures tableb

Hospitalizationg 3.6 d average By day ofhospitalization


Hypercalcemia Diagnostictesth

Serum calcium(monthly tests)

By test R$3.75 R$41.25 AMB procedures tableb

Diagnostictesth

TSH (monthly tests) By test R$24.30 R$302 AMB procedures tableb

Hospitalizationi 3 d hospitalizationfor patient hydration



Ostenecrosis N/A Drugs Zoledronate (4 mgevery 4 wk)

By mg R$6.43 R$6953 Kairos Magazine, May 2008—generica

Drugsj Ciprofloxacin (1000mg/d)

By mg R$0.008 R$2782 Kairos Magazine, May 2008—generica

Drugs Infusion IV By infusion R$35 R$420 AMB procedures tableb

Medical visits Monthly medicalvisits

By visit R$39 R$468 AMB procedures tableb

Diagnostictestsj

Facial radiographs By test R$53.01 R$689 AMB procedures tableb

Surgical/doctorc,j Hyperbaric chamber(15 sessions)

By section R$329 R$4935 Temporal, 2002l

Surgical/doctorc,j Sequestrectomy(60%)

By procedures R$1626.88 R$976.12 AMB procedures tableb

Hospitalizationk 3 d hospitalizationfor patient hydration



AMB � Brazilian Medical Association; N/A � not applicable; PROAHSA � Programa de Estudos Avançados em Administra-ção Hospitalar e de Sistemas de Saúde, da Faculdade de Medicina da USP e da Escola de Administração de Empresas daFundação Getulio Vargas; R$� Brazilian reais; THS � thyroid-stimulating hormone.aKairos Magazine.http://brasil.kairosweb.com/.bAvailable at http://www.amb.org.br/.cResources used and utilization rate were obtained from Narazaki DK, de Alverga Neto CC, Baptista AM, et al. Prognosticfactors in pathologic fractures secondary to metastatic tumors. Clinics (Sao Paulo). 2006;61:313–320.

dAverage days of hospitalization obtained from DATASUS ICD-10 description: malignant neoplasm of bone and cartilage.ePROAHSA Bulletin no. 35. http://www.hcnet.usp.br/proahsa/indicadores/.fResources used and utilization rate were assumed at 50% each.gAverage days of hospitalization obtained from DATASUS ICD-10 description: fracture of other limb bones.hResources used and utilization rates were obtained at Farias ML. Hypercalcemia of malignancy: clinical features, diagnosis andtreatment [in Portuguese]. Arq Bras Endocrinol Metabol. 2005;49:816–824.

iResources used and utilization rate assumed were 3 d of hospitalization to treat dehydration.jResources used and utilization rate were obtained at Júnior WC, Rocha RI, Massarollo LCB. Mandibular osteonecrosisassociated with use of biphosphonates. Rev Soc Bras Cir Craniomaxilofac. 2007;10:93–96.kResources used and utilization rate assumed were 3 d of hospitalization for rehabilitation.lTemporal WF. Aspectos econômicos da oxigenoterapia hiperbárica (OHB). RMAB 2002; 52:20–24. http://www.dirsa.aer.mil.br/revistas/2002/04_02.pdf.

1780.e2 Volume 33 Number 11

http://brasil.kairosweb.com/

http://www.amb.org.br/

http://www.hcnet.usp.br/proahsa/indicadores/

