© 2012 Informa UK Ltd. This provisional PDF corresponds to the article as it appeared upon acceptance. Fully formatted PDF and full text (HTML) versions will be made available soon. DISCLAIMER: The ideas and opinions expressed in the journal’s Just Accepted articles do not necessarily reflect those of Informa Healthcare (the Publisher), the Editors or the journal. The Publisher does not assume any responsibility for any injury and/or damage to persons or property arising from or related to any use of the material contained in these articles. The reader is advised to check the appropriate medical literature and the product information currently provided by the manufacturer of each drug to be administered to verify the dosages, the method and duration of administration, and contraindications. It is the responsibility of the treating physician or other health care professional, relying on his or her independent experience and knowledge of the patient, to determine drug dosages and the best treatment for the patient. Just Accepted articles have undergone full scientific review but none of the additional editorial preparation, such as copyediting, typesetting, and proofreading, as have articles published in the traditional manner. There may, therefore, be errors in Just Accepted articles that will be corrected in the final print and final online version of the article. Any use of the Just Accepted articles is subject to the express understanding that the papers have not yet gone through the full quality control process prior to publication.

Just Accepted by Journal of Medical Economics

Original article

Economic evaluation of tocilizumab combination in the treatment of moderate to severe rheumatoid arthritis in Italy

Alexander Diamantopoulos, Maurizio Benucci, Stefano Capri, Wolfgang Berger, Neil Wintfeld, Giovanni Giuliani, Walter Ricciardi

doi: 10.3111/13696998.2012.665110

Abstract

Objective: This study was designed to evaluate the cost utility of tocilizumab in RA patients, with inadequate responses to traditional disease-modifying anti-rheumatic drugs (tDMARDs) from a payer’s perspective in Italy.

Methods: An individual patient simulation model was used to project lifetime medical costs (payer’s perspective) and quality-adjusted life-years (QALYs). Treatment sequences starting with tocilizumab or the most commonly prescribed biologics (etanercept, adalimumab or infliximab) were compared. The addition of tocilizumab to standard of care, without the replacement of anti–tumor necrosis factor (TNF)-α treatments, was

also evaluated. Patient characteristics, treatment efficacy and quality of life data were based on three phase 3 tocilizumab clinical trials (TOcilizumab Pivotal Trial in Methotrexate Inadequate respONders [OPTION], Tocilizumab in cOmbination With traditional DMARD therapy [TOWARD], and TociLIzumab Safety and THE Prevention of Structural Joint Damage [LITHE]). Mixed-treatment comparison was used to estimate response probabilities. Resource utilisation, treatment acquisition, administration and monitoring costs were estimated using Italian secondary sources. Uncertainty in model parameters was evaluated by probabilistic sensitivity analysis.

Results: Replacement of anti–TNF-α treatments with tocilizumab reduced total costs over a patient’s lifetime (base-case analysis: tocilizumab sequence, €141,100vs standard of care sequence, €143,500). Patients receiving tocilizumab realised more QALYs than patients receiving standard of care (9.8881vs 9.3502QALYs). Therefore, according to the base-case analysis, the tocilizumab sequence dominated the standard of care. In a sensitivity analysis, the model base-case result was robust to input changes. When tocilizumab was added to standard of care, without replacing anti–TNF-α treatments, the incremental cost-effectiveness ratio was €17,100per QALY.

Conclusion: The analysis demonstrates that, in Italy, replacing another biologic DMARD with tocilizumab or adding tocilizumab to the current standard of care is a cost-effective strategy in the treatment of RA patients with inadequate responses to tDMARDs.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

1

Original article

Economic evaluation of tocilizumab combination in the treatment of moderate to severe

rheumatoid arthritis in Italy

Alexander Diamantopoulos1, Maurizio Benucci2, Stefano Capri3, Wolfgang Berger4, Neil Wintfeld5,

Giovanni Giuliani4, Walter Ricciardi6

1Symmetron Limited, Kinetic centre, London, UK; 2Section of Rheumatology, Nuovo Ospedale S. Giovanni

di Dio, Florence, Italy; 3Institute of Economics, Cattaneo–LIUC University, Castellanza, Italy; 4F.

Hoffmann-La Roche Ltd, Basel, Switzerland; 5former employee of Genentech, South San Francisco, CA

USA; 6Universita Cattolica del Sacro Cuore, Roma, Italy

Author for correspondence: Alex Diamantopoulos, Symmetron Limited, Kinetic centre

Theobald street, Elstree WD6 4PJ, United Kingdom. alexd@symmetron.net

Key words: cost-effectiveness; cost-utility analysis; disease-modifying anti-rheumatic drugs; health-

related quality of life; pharmacoeconomic modelling; rheumatoid arthritis; tocilizumab (additional)

biologics; disease severity; resource utilisation; standard of care; treatment-related costs

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

2

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

3

Abstract

Objective: This study was designed to evaluate the cost utility of tocilizumab in RA patients, with

inadequate responses to traditional disease-modifying anti-rheumatic drugs (tDMARDs) from a payer’s

perspective in Italy.

Methods: An individual patient simulation model was used to project lifetime medical costs (payer’s

perspective) and quality-adjusted life-years (QALYs). Treatment sequences starting with tocilizumab or

the most commonly prescribed biologics (etanercept, adalimumab or infliximab) were compared. The

addition of tocilizumab to standard of care, without the replacement of anti–tumor necrosis factor

(TNF)-α treatments, was also evaluated. Patient characteristics, treatment efficacy and quality of life

data were based on three phase 3 tocilizumab clinical trials (TOcilizumab Pivotal Trial in Methotrexate

Inadequate respONders [OPTION], Tocilizumab in cOmbination With traditional DMARD therapy

[TOWARD], and TociLIzumab Safety and THE Prevention of Structural Joint Damage [LITHE]). Mixed-

treatment comparison was used to estimate response probabilities. Resource utilisation, treatment

acquisition, administration and monitoring costs were estimated using Italian secondary sources.

Uncertainty in model parameters was evaluated by probabilistic sensitivity analysis.

Results: Replacement of anti–TNF-α treatments with tocilizumab reduced total costs over a patient’s

lifetime (base-case analysis: tocilizumab sequence, €141,100vs standard of care sequence, €143,500).

Patients receiving tocilizumab realised more QALYs than patients receiving standard of care (9.8881vs

9.3502QALYs). Therefore, according to the base-case analysis, the tocilizumab sequence dominated the

standard of care. In a sensitivity analysis, the model base-case result was robust to input changes. When

tocilizumab was added to standard of care, without replacing anti–TNF-α treatments, the incremental

cost-effectiveness ratio was €17,100per QALY.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

4

Conclusion: The analysis demonstrates that, in Italy, replacing another biologic DMARD with tocilizumab

or adding tocilizumab to the current standard of care is a cost-effective strategy in the treatment of RA

patients with inadequate responses to tDMARDs.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

5

Introduction

Rheumatoid arthritis (RA) is a chronic, progressive, systemic autoimmune disease that affects

joint function and leads to increased disability. The annual prevalence of RA in Italy is

approximately 272,004 cases (0.46% of the total population).1 The impact of RA on health care

resource use and productivity loss is considerable.2

Traditional disease-modifying anti-rheumatic drugs (tDMARDs; e.g. methotrexate, leflunomide)

are generally used as first-line treatments,3 primarily because they are available as generics and

are inexpensive. Biologic DMARDs (bDMARDs; e.g. etanercept, adalimumab, infliximab) are used

in combination with tDMARDs, usually methotrexate, as second-line treatments.3 Etanercept in

combination with methotrexate is the most commonly prescribed treatment after tDMARD

failure.

Tocilizumab is a new humanised interleukin-6 (IL-6) receptor monoclonal antibody with a novel

mechanism of action that provides a unique treatment option for patients with RA. The efficacy

and safety of tocilizumab in patients with an inadequate response to tDMARDs was

demonstrated in three phase 3 trials.4-6 The introduction of tocilizumab offers an alternative to

anti–tumour necrosis factor-α (anti–TNF-α) cycling, which is often considered an ineffective

treatment strategy.7

Several economic evaluations in Italy assessed the cost-effectiveness of bDMARDs, including

abatacept, etanercept and infliximab, compared with methotrexate or other tDMARDs.8-10

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

6

However, tocilizumab had not been considered in cost-effectiveness analyses in the Italian

health care environment. Given the clinical effectiveness of tocilizumab, this study was

conducted to assess the cost-utility of tocilizumab treatment (RoActemra*) in Italy compared

with other bDMARDs that represent standard of care.

*RoActemra; Roche, Basel, Switzerland

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

7

Methods

To synthesize available evidence and extrapolate clinical trial data, an economic model was

developed to assess the cost utility of tocilizumab in combination with methotrexate in RA

patients with an inadequate response to tDMARDs. The design of the economic analysis follows

guidelines set by the Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT)

Economics Working Group.11,12

The model attempts to predict patient outcomes until death. A patient lifetime horizon ensures

that all relevant costs and health-related quality of life (QoL) impairment are captured in the

analysis. Such a model timeframe is in line with most economic analyses in RA.13-18 The model

uses efficacy and QoL data from three phase 3 clinical trials that evaluated tocilizumab in

combination with methotrexate, compared with methotrexate alone or an alternative tDMARD

alone.4-6 Given that the clinical trial outcomes are available at 6 months, the model is designed

to evaluate patient transitions at 6-month intervals (model cycle length).

Patients entered the model after an inadequate response to tDMARDs and progressed through

a sequence of treatments that reflect real-life practice scenarios (Figure 1). In Italy, the standard

practice for patients whose tDMARD therapy fails is to prescribe a sequence of bDMARDs in

combination with methotrexate. It is assumed that etanercept would be the first bDMARD in the

standard-of-care sequence, followed by adalimumab, rituximab and abatacept. The comparator

sequence (with tocilizumab) assumes the replacement of etanercept with tocilizumab, with all

subsequent treatments remaining the same (Figure 1). Comparison of treatment sequences

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

8

offers an evaluation of patient outcomes mimicking real life and is often used when modelling in

RA.18-20

For completeness, the analysis considered three alternative scenarios. The first scenario

reversed the order of the first two biologic treatments, using adalimumab ahead of etanercept

as standard treatment. In this scenario, the evaluation compared the replacement of

adalimumab, rather than etanercept, by tocilizumab; etanercept followed in the sequence after

both treatment comparators. The second scenario was the same as the previous one but

assumed that the comparison was with infliximab instead of adalimumab. The third scenario

assumed no anti–TNF-α treatment replacement and evaluated the addition of tocilizumab in

combination with methotrexate at the start of the standard-of-care sequence.

American College of Rheumatology (ACR) response rates were used as a measure of initial

treatment efficacy. Based on ACR response rates, simulated individuals were allocated to one of

four categories: no response, ACR20 response, ACR50 response or ACR70 response. Those with

no response moved to the next treatment in the sequence, whereas those with a response

remained on treatment until withdrawal, with significant changes in QoL and costs.

Given the absence of head-to-head ACR response data and the sequential form of the model

structure, results from a Bayesian mixed-treatment comparison (MTC) were used to assign

appropriate ACR data according to the position of each treatment in the sequence. ACR

response probabilities for tocilizumab and etanercept, each in combination with methotrexate,

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

9

were derived from Bergman et al.21 Data from another MTC in an anti–TNF-α failure population

were used for ACR responses to rituximab and abatacept, each in combination with

methotrexate (Supplementary Digital Content Appendix A).Because of a lack of evidence about

efficacy after bDMARD (etanercept or tocilizumab) failure, the response rates of adalimumab in

combination with methotrexate were reduced by 30%.7

Disease severity and, consequently, QoL are reflected in the model by changes in patient Health

Assessment Questionnaire (HAQ) scores.18,20 Patient characteristics (such as age, sex, baseline

HAQ score and weight) were obtained from the clinical trials.4-6 These parameters were

maintained across the simulated cohort at the start of the model, assuming a homogeneous

group of patients. As patients progressed through treatment, individual simulation was used to

monitor HAQ score changes and projected costs and outcomes. Individual simulation was

preferred to cohort analysis for computational reasons; accrued health benefits and costs

depend on a large number of values for the HAQ score parameter.22,23

Response to treatment was assumed to have an impact on disease severity, as measured by

individual HAQ score. Data on the relationship between ACR responses and patient HAQ scores

were obtained from the clinical trials and corresponded to the first 24 weeks after baseline,

suggesting that the higher the observed response, the greater the drop in HAQ score (Table 1).

For the first 6 months of a new treatment, the model assumed a HAQ score reduction according

to the level of response: ACR20, ACR50 or ACR70. This HAQ score benefit was assumed to be

response related, not treatment related, and therefore was applied universally to all bDMARDs.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

10

After response to treatment, the model assumes no HAQ score progression.24,25 Moreover, the

model incorporates data from a recent analysis of the three clinical trials for HAQ score

progression during administration of tocilizumab in combination with methotrexate.26 The

analysis suggested a functional improvement for patients receiving tocilizumab, reflected by a

negative slope of the patient average HAQ score. Observations before week 24 were excluded

because initial response to treatment was expected during this period, and a decline in HAQ

score was already modelled. Because the data cut-off used in the analysis26 is at 4.5 years, the

model conservatively assumes no improvement after that time point for patients continuing on

tocilizumab combination therapy.24,25 Once patients exhaust all possible treatment options and

enter palliative care, they experience disease deterioration reflected by a significant HAQ score

increase of 0.03 every 6 months.7

Data from Geboreck et al.27 suggest that withdrawal rates at six months, for etanercept and

infliximab, are 8% and 12%, respectively. These estimates are congruent with data from

Bansback et al.28 The current model assumes a withdrawal risk equal to the average of the two

estimates for all treatments. At the point of withdrawal, patients were assumed to lose the

acquired benefit with response to treatment and rebound to a higher HAQ score.29 A summary

of HAQ score change from treatment response to withdrawal for each modelled individual is

presented in Figure 2.

Health-related QoL

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

11

A mechanism of mapping utility from patient HAQ score was incorporated into the model to

derive health-related QoL estimates (equation 1). This technique is similar to that of previously

published cost-utility studies and reimbursement submissions of biologic treatments in RA.28,29

The method is modified herein to reflect data from the clinical trials. The utility mapping model

is presented in detail in Supplementary Digital Content Appendix B.30

EQ5D = 0.82 – 0.11 × HAQ – 0.07 × HAQ^2 (1)

The base-case analysis uses equation 1, reflecting the assumption that decreases in HAQ level

are more valuable (as measured by change in utility) for severely disabled patients than for

patients who are less disabled. Sensitivity analyses tested alternative utility mapping scenarios

assuming a linear relationship between the two model parameters (Figure 3).

Treatment-related costs and resource utilisation

The economic evaluation reflects the Italian National Health Service perspective with all relevant

health care costs. The acquisition costs of bDMARDs were calculated based on hospital prices.

For methotrexate, which is available in public pharmacies, a public price that included

distribution margins was used. Dosage information was combined with acquisition prices to

estimate the annual costs of treatment (Table 2).31 All bDMARDs were considered in

combination with methotrexate.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

12

For infusion administrations (tocilizumab, abatacept, infliximab and rituximab; Table 3), health

care personnel time was estimated based on an Italian retrospective study, which assessed

resource consumption for 100 RA patients treated with infliximab:31 rheumatologist time per

cycle was 15 minutes, and nurse time was 20 minutes (calculated on a 180-minute infusion). For

tocilizumab, rituximab and abatacept, the model assumed the same rheumatologist time,

whereas nurse time was recalculated proportionally to the time of infusion. For subcutaneous

administration (etanercept and adalimumab), an average of 4.5 minutes was assumed.32 The

costs per minute for nurses and rheumatologists were assumed to be €0.43 and €0.97,

respectively.33

Direct medical costs, other than drug acquisition and administration costs, were obtained from

an Italian retrospective, prevalence-based, multicentre study of 200 patients (Table 4). 2 Direct

medical costs were associated with the four ACR response classes to reflect costs per patient

associated with disease severity. The HAQ score for each ACR response class was associated

with different HAQ scores based on the National Databank for Rheumatic Disease (Table 5); 9

2002 prices were inflated to 2009 prices using data from the Italian Institute for Statistics34.

Future costs and quality-adjusted life-years (QALYs) were discounted at an annual 3.0% rate.

All assumptions were presented to and validated by the involved clinical and health economic

experts. Several sensitivity analyses tested the robustness of the model to input changes.

Moreover, parameter uncertainty was addressed by fitting appropriate distributions to model

variables and running probabilistic sensitivity analysis (PSA). Supplementary Digital Content

Appendix C presents details on the sampling distributions used.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

13

Results

Ten thousand simulation trials were run. On average, the medical cost for standard of care was

estimated to be higher than the sequence with tocilizumab (€143,500 vs. €141,100). The drivers

of this cost difference were the lower price of tocilizumab and the decreased use of medical

resources. Moreover, patients in the tocilizumab sequence realised more utility benefits than

patients in the standard-of-care sequence, which translated to an increment of ~0.54 (9.3502 vs.

9.8881) QALYs. Therefore, given current prices in Italy, RA patients with an inadequate response

to tDMARDs incur lower costs and have higher QoL; in other words, the tocilizumab sequence

dominated the standard-of-care sequence (Table 6).

Several analyses tested the sensitivity of the model to input changes. (Note that sensitivity

analysis is presented with changes to the base-case scenario, as in Figure 1 [comparison with

etanercept]). Model estimates were fairly robust to input changes, and the tocilizumab

sequence dominated the standard-of-care sequence in most comparisons. The model was

sensitive to changes in HAQ score progression during treatment. Changes to this parameter had

an impact on patient QoL. As scenario C in Table 6 illustrates, when HAQ score progression was

assumed to be zero for tocilizumab, the incremental QALY benefit was reduced from 0.54 to

0.06 QALYs in the base case analysis. Other changes to the model structure (withdrawal,

scenario B) or parameter values (scenarios A, D, E, F, G, H, I) had little impact on the model

results (Table 6).

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

14

In PSA, more than 80% of the samples fell within the fourth quadrant, suggesting that the

tocilizumab sequence dominated the alternative (Figure 4). There were no samples represented

in the second quadrant, where tocilizumab would be dominated by the alternative (Figure 4).

The probability that the intervention was cost effective was 100% when considering a cost-

effectiveness threshold of €50,000 per QALY gained.

In the first alternative scenario, in which adalimumab was prescribed before etanercept, the

model produced similar results (Table 6). In the second scenario, in which infliximab was

prescribed before etanercept, the incremental cost-effectiveness ratio (ICER) was estimated at

€2,600 per QALY gained. In the third scenario, in which the analysis assumed no anti–TNF-α

treatment replacement and the addition of tocilizumab to the clinician’s treatment options, the

ICER was €17,100per QALY gained.

Discussion

Three large, multinational, randomised, phase 3 clinical trials demonstrated favourable results

for tocilizumab in combination with methotrexate, compared with methotrexate alone, in

patients with an inadequate response to at least one tDMARD. The present study synthesises

evidence from these studies with data from the published literature, extrapolates patient

outcomes to lifetime, and translates the clinical evidence to a cost-utility analysis.

Other economic studies in the treatment of RA focus on the initial clinical benefit and cost of

interventions (6-month time frame).8,10 In such cases, the analysis is based entirely on clinical

trial data, and no assumptions are necessary for the long-term disease progression in patients.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

15

However, given the chronic nature of RA, a long-term time frame of analysis is more relevant to

decision makers. Cost-effectiveness results from short-term analyses are likely to be

underestimated because not all relevant costs and QALYs are included.

Because no head-to-head comparisons between tocilizumab and anti–TNF-α agents

(adalimumab and etanercept) have been made, the model considers relative efficacy between

these biologics, as estimated by an MTC. The response probability of adalimumab is further

adjusted to reflect its anti–TNF-α-inadequate response (TNF-IR) position in the standard-of-care

sequence. Although the adjustment represents loss of treatment efficacy from anti–TNF-α

cycling, this adjustment is conservatively assumed to appear also after tocilizumab, where anti–

TNF-α cycling is not modelled. A second MTC among rituximab, abatacept and tocilizumab was

used to provide relative efficacy data in TNF-IR positions for these treatments. Given the

uncertainty surrounding these estimates, a sensitivity analysis based on ACR response rates was

conducted; ACR response was not a major driver of the model results.

Further assumptions on the lifetime disease severity of simulated individuals were necessary to

complete the analysis. The simulation of individual HAQ score progression was used to

represent changes in disease severity.18,20 Therefore, HAQ score change at treatment initiation

was based on clinical trial evidence, which reflected only tocilizumab. However, that the

observed improvement was response related, not treatment related, is a reasonable

assumption; hence, the benefit from response was applied universally to all treatments.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

16

To extrapolate HAQ score beyond clinical trial duration, an assumption was necessary. In the

past, similar economic evaluations have assumed a positive HAQ score progression, reflecting a

slow deterioration in disease severity.20 More recently, a distinction was made between the

levels of deterioration associated with bDMARDs and tDMARDs. Based on data from long-term

extension trials, patients receiving anti–TNF-α treatments were assumed to experience stable

HAQ score progression while on treatment.24,25 Our model uses data on HAQ score progression

for tocilizumab, which suggest a clear improvement of HAQ score in patients who continue to

receive tocilizumab for up to 4.5 years (for when data exist).

Moreover, recent observational data collected in patients treated with biologics other than

tocilizumab suggest a reduced impact on disease progression in real life compared with data

from clinical trials.35 The current analysis considers this by assigning stable HAQ progression to

tocilizumab in a sensitivity analysis. In this scenario, the tocilizumab regimen continues to

dominate the standard of care (case C in Table 6).

The QoL associated with RA and the benefits from each treatment were based a post-hoc

analysis of clinical trial data. A method to obtain QALYs from HAQ score was used. Similar

methods of HAQ and EQ-5D mapping are commonly used in RA modelling.28,29 The present study

improves this technique by introducing a non-linear relationship between the two parameters.

This reflects the assumption that improvements in QoL are more valuable for patients with

severe RA.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

17

Treatment costs are shown to be important drivers of the economic model, whereas the

implications to resource use of adopting tocilizumab as a standard of care are not large-scale

(sensitivity analysis G). Therefore, the generalisability of the model results to other countries

depends on the relative difference between tocilizumab and the other bDMARD prices.

The analysis does not include adverse events (AEs) from treatment. However, since the included

treatments have a similar safety profile, inclusion of AEs would not have any impact on the

incremental cost-effectiveness results. Although the implications of AEs on costs and patient

utility are not introduced into the model, they are indirectly reflected in the treatment

withdrawal rates. In addition, the disutility associated with treatment-related AEs has been

captured in the EQ-5D utility data collected in the tocilizumab phase III trials.

Further research is necessary to provide information about the relative efficacy of the

treatments in different populations (DMARD-IR and TNF-IR). The availability of head-to-head

comparisons would help in validating the MTC results this economic analysis uses. Moreover,

data on long-term disease progression would improve assumptions of the economic model and

provide the platform for long-term assessment of biologic treatments. Nevertheless, as

illustrated in the sensitivity analysis, although these developments would improve accuracy, a

significant change in the direction of the results is unlikely.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

18

Conclusion

The analysis demonstrates that, in Italy, replacing another bDMARD with tocilizumab or adding

tocilizumab to the current standard of care is a cost-effective strategy in the treatment of RA

patients with an inadequate response to tDMARDs.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

19

Transparency

Declaration of interest:

This study was funded by F. Hoffmann-La Roche Ltd. Roche was specifically involved in the

collection of drug costs and RA management costs from published sources, preparation of mixed

treatment comparisons, mapping Health Assessment Questionnaire (HAQ) data to EuroQoL-5D,

analysing clinical trial data for specific modelling purposes, running modelling analyses and

presenting and discussing results.

Declaration of financial/other interests:

M.B., S.C. and W.R. have disclosed that they received funds as consultants of F. Hoffmann-La

Roche Ltd, and A.D. reports that he received funds as a contractor from the same. W.B., N.W.

and G.G. have disclosed that they are employees of F. Hoffmann-La Roche Ltd.

Acknowledgment:

Maribeth Bogush, PhD, and David Murdoch, MB ChB provided writing assistance for this

manuscript, which was paid for by F. Hoffmann-La Roche Ltd.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

20

References

1. Salaffi F, De Angelis R, Grassi W; MArche Pain Prevalence INvestigation Group (MAPPING)

study. Prevalence of musculoskeletal conditions in an Italian population sample: results of a

regional community-based study, I: the MAPPING study. Clin Exp Rheumatol

2005;23(6):819-28

2. Leardini G, Salaffi F, Montanelli R, et al. A multicenter cost-of-illness study on rheumatoid

arthritis in Italy. Clin Exp Rheumatol 2002;20(4):505-15

3. Smolen JS, Landewé R, Breedveld FC, et al. EULAR recommendations for the management of

rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs.

Ann Rheum Dis 2010;69(6):964-75

4. Smolen JS, Beaulieu A, Rubbert-Roth A, et al.; OPTION Investigators. Effect of interleukin-6

receptor inhibition with tocilizumab in patients with rheumatoid arthritis (OPTION study): a

double-blind, placebo-controlled, randomized trial. Lancet 2008;371(9617):987-97

5. Genovese MC, McKay JD, Nasonov EL, et al. Interleukin-6 inhibition with tocilizumab reduces

disease activity in rheumatoid arthritis with inadequate response to disease-modifying

antirheumatic drugs: the tocilizumab in combination with traditional disease-modifying

antirheumatic drug therapy study. Arthritis Rheum 2008;58(10):2968-80

6. Fleischmann R, Burgos-Vargas R, Ambs P, et al. LITHE: tocilizumab inhibits radiographic

progression and improves physical function in rheumatoid arthritis (RA) patients (Pts) at 2

yrs with increasing clinical efficacy over time [abstract]. Arthritis Rheum 2009;60 Suppl.

10:637

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

21

7. National Institute for Health and Clinical Excellence. NICE technology appraisal guidance

(130): adalimumab, etanercept and infliximab for the treatment of rheumatoid arthritis.

Issued October 2007. http://www.nice.org.uk/nicemedia/pdf/TA130guidance.pdf. Accessed

30 March 2011

8. Benucci M, Li Gobbi F, Sabadini L, et al. The economic burden of biological therapy in

rheumatoid arthritis in clinical practice: cost-effectiveness analysis of sub-cutaneous anti-

TNFα treatment in Italian patients. Int J Immunopathol Pharmacol 2009;22(4):1147-52

9. De Portu S, Mantovani LG, Olivieri I. Analisi di costo-utilità di abatacept nell’artrite

reumatoide in Italia. Farmeconomia e percorsi terapeutici 2008;9(1):19-26

10. Leardini G, Ganguly R, Singh A. Cost-effectiveness analysis of etanercept versus infliximab

and anakinra in the treatment of rheumatoid arthritis in Italy [abstract 1152]. 67th Annual

Scientific Meeting of the American College of Rheumatology (ACR), 23-28 October 2003,

Orlando, Florida, USA

11. Gabriel S, Drummond M, Maetzel A, et al.; Patient Perspective Group. OMERACT 6

Economics Working Group report: a proposal for a reference case for economic

evaluation in rheumatoid arthritis. J Rheumatol 2003;30(4):886-90

12. Maetzel A, Tugwell P, Boers M, et al.; OMERACT 6 Economics Research Group.

Economic evaluation of programs or interventions in the management of rheumatoid

arthritis: defining a consensus-based reference case. J Rheumatol 2003;30(4):891-6

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

22

13. Kielhorn A, Porter D, Diamantopoulos A, et al. UK cost-utility analysis of rituximab in patients

with rheumatoid arthritis that failed to respond adequately to a biologic disease-modifying

antirheumatic drug. Curr Med Res Opin 2008;24(9):2639-50

14. Wailoo AJ, Bansback N, Brennan A, et al. Biologic drugs for rheumatoid arthritis in the

Medicare program: a cost-effectiveness analysis. Arthritis Rheum 2008;58(4):939-46

15. Vera-Llonch M, Massarotti E, Wolfe F, et al. Cost-effectiveness of abatacept in patients with

moderately to severely active rheumatoid arthritis and inadequate response to tumor

necrosis factor-(alpha) antagonists. J Rheumatol 2008;35(9):1745-53

16. Brennan A, Bansback N, Nixon R, et al. Modelling the cost effectiveness of TNF-(alpha)

antagonists in the management of rheumatoid arthritis: results from the British Society for

Rheumatology Biologics Registry. Rheumatology (UK) 2007;46(8):1345-54

17. Spalding JR, Hay J. Cost effectiveness of tumour necrosis factor-alpha inhibitors as first-line

agents in rheumatoid arthritis. Pharmacoeconomics 2006;24(12):1221-32

18. Chen YF, Jobanputra P, Barton P, et al. A systematic review of the effectiveness of

adalimumab, etanercept and infliximab for the treatment of rheumatoid arthritis in adults

and an economic evaluation of their cost-effectiveness [abstract]. Health Technol Assess

2006;10(42):248

19. Jobanputra P, Barton P, Bryan S, et al. The effectiveness of infliximab and etanercept for the

treatment of rheumatoid arthritis: a systematic review and economic evaluation. Health

Technol Assess 2002;6(21):1-110

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

23

20. Barton P, Jobanputra P, Wilson J, et al. The use of modelling to evaluate new drugs for

patients with a chronic condition: the case of antibodies against tumour necrosis factor in

rheumatoid arthritis. Health Technol Assess 2004;8(11):1-91

21. Bergman GJ, Hochberg MC, Boers M, et al. Indirect comparison of tocilizumab and other

biologic agents in patients with rheumatoid arthritis and inadequate response to disease-

modifying antirheumatic drugs. Semin Arthritis Rheum 2010;39(6):425-41

22. Brennan A, Chick SE, Davies R. A taxonomy of model structures for economic evaluation of

health technologies. Health Econ 2006;15(12):1295-310

23. Davies R, Roderick P, Raftery J. The evaluation of disease prevention and treatment using

simulation models. Eur J Operational Res 2003;150(1):53-66

24. Landewé RB, Boers M, Verhoeven AC et al. COBRA combination therapy in patients with

early rheumatoid arthritis: long-term structural benefits of a brief intervention. Arthritis

Rheum 2002 Feb;46(2):347-56.

25. Weinblatt ME, Keystone EC, Furst DE, et al. Change over time in the safety, efficacy, and

remission profiles of patients with rheumatoid arthritis receiving adalimumab for up to 7

years. Presented at: American College of Rheumatology 2007 Annual Scientific Meeting, 10-

11 November 2007, Boston, Massachusetts, USA

26. van Vollenhoven R, Ducournau P, Wintfeld N, et al. Health Assessment Questionnaire–

Disability Index scores in patients with rheumatoid arthritis treated with tocilizumab plus

conventional anti-rheumatic drugs. Presented at: ISPOR 12th Annual European Congress,

24-27 October 2009, Paris, France

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

24

27. Geborek P, Crnkic M, Petersson IF, et al. Etanercept, infliximab and leflunomide in

established rheumatoid arthritis: clinical experience using a structured follow up

programme in southern Sweden. Ann Rheum Dis 2002;61:793-8

28. Bansback N, Brennan A, Ghatnekar O. Cost effectiveness of adalimumab in the treatment of

patients with moderate to severe rheumatoid arthritis in Sweden. Ann Rheum Dis

2005;64(7):995-1002

29. Brennan A, Bansback N, Reynolds A, et al. Modelling the cost-effectiveness of etanercept in

adults with rheumatoid arthritis in the UK. Rheumatology 2004;43(1):62-72

30. Boggs R, Sengupta N, Ashraf T. Estimating health utility from a physical function assessment

in rheumatoid arthritis (RA) patients treated with adalimumab (HUMIRA) [abstract UT3].

Presented at: 7th International Society of Pharmacoeconomics and Outcomes Research, 19-

22 May 2002, Arlington, Virginia

31. Favalli EG, Marchesoni A, Colombo GL, et al. Pattern of use, economic burden and vial

optimization of infliximab for rheumatoid arthritis in Italy. Clin Exp Rheumatol

2008;26(1):45-51

32. Rubio-Terrés C, Ordovàs Baines JP, Pla Poblador R, et al.; Grupo de Investigadores del

Estudio PRAXIS. [Use and cost of biological disease-modifying anti-rheumatic drugs in Spain

(PRAXIS study)] [Spanish]. Farm Hosp 2007;31(2):71-92

33. Benucci M, Iannazzo S, Zaniolo O, et al. Rituximab in the treatment of rheumatoid arthritis

patients in Italy: a budget impact analysis. Clin Exp Rheumatol 2010;28(5):722-7

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

25

34. ISTAT Italian Institute of Statistics 2009. Coefficienti per tradurre valori monetari dei periodi

sottoindicati in valori del 2010 (a).

http://www.istat.it/prezzi/precon/rivalutazioni/val_moneta_2010.zip. Accessed 7 February

2011

35. Wolfe F, Michaud K. The loss of health status in rheumatoid arthritis and effect of biologic

therapy: a longitudinal observation study. Arthritis Res Ther 2010;12(2):R35

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

26

Figure 1. Treatment sequences modelled.

Figure 2. Change in individual HAQ scores. Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

27

Figure 3. Utility–HAQ score mapping models.

Figure 4. Cost-effectiveness scatter plot.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

28

Figure 5. Cost-effectiveness acceptability curve.

Table 1. Economic model parameters.

Patient characteristics4-6

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

29

Female, % 82

Mean weight, kg 72.82

Mean starting HAQ score 1.51

*ACR response, % of patients (2.5%CrL-97.5%CrL)21

ACR20

Tocilizumab pooled DMARD-IR population

(OPTION, TOWARD, LITHE)

65 (51-76)

TNF inhibitor (etanercept, adalimumab, infliximab) 63 (55-71)

Rituximab (TNF-IR) 46 (37-56)

Abatacept (TNF-IR) 43 (33-54)

ACR50

Tocilizumab pooled DMARD-IR population

(OPTION, TOWARD, LITHE)

44 (29-58)

TNF inhibitor (etanercept, adalimumab, infliximab) 39 (31-48)

Rituximab (TNF-IR) 23 (14-36)

Abatacept (TNF-IR) 22 (12-39)

ACR70

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

30

Tocilizumab pooled DMARD-IR population

(OPTION, TOWARD, LITHE)

29 (21-38)

TNF inhibitor (etanercept, adalimumab, infliximab) 16 (14-19)

Rituximab (TNF-IR) 14 (5-39)

Abatacept (TNF-IR) 8 (2-27)

HAQ score benefit for ACR response, mean (SD)4-6

No response –0.1357 (0.017)

ACR20 –0.4427 (0.018)

ACR50 –0.6680 (0.026)

ACR70 –0.9230 (0.032)

HAQ score change during treatment (SE)7,26

Tocilizumab (every 180 days) –0.01867 (0.00432)

Biologic treatments 0.0000

Palliative care (every 180 days) 0.0300

Withdrawal rate (every 180 days) 0.10

ACR, American College of Rheumatology; DMARD, disease-modifying anti-rheumatic drug; HAQ,

Health Assessment Questionnaire; IR, inadequate response; LITHE, TociLIzumab Safety and THE

Prevention of Structural Joint Damage; OPTION, TOcilizumab Pivotal Trial in Methotrexate

Inadequate respONders; TNF, tumour necrosis factor; TOWARD, Tocilizumab in cOmbination

With traditional DMARD therapy.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

31

* ACR response is presented as overlapping percentages. For the economic model ACR response

is transformed to non-overlapping to reflect mutually exclusive probabilities of response.

Table 2. Treatment acquisition costs (hospital acquisition prices).31

Annual treatment cost SmPC dosages

Tocilizumab +

methotrexate

€11,520.78 8 mg/kg every 4 weeks

Abatacept +

methotrexate

€12,527.22 750 mg infusion every 4 weeks

Adalimumab +

methotrexate

€12,134.02 40 mg every 2 weeks

Etanercept +

methotrexate

€12,469.13 50 mg weekly

Infliximab +

methotrexate

€11,210.76 4 mg/kg every 7 weeks

Rituximab +

methotrexate

€10,567.42 2 g every 6 months

Methotrexate €20.78 7.5 mg weekly

SmPC, summary of product characteristics.

Table 3. Treatment administration costs.31-33

Nurse time

(min/cycle)

Rheumatologist

time

(min/cycle)

Administration

cost/cycle

Annual

administration

cost

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

32

Tocilizumab +

methotrexate

6.66 15 €17.41 €208.92

Abatacept +

methotrexate

3.33 15 €15.98 €207.74

Adalimumab +

methotrexate

4.50 — €1.94 €50.44

Etanercept + methotrexate 4.50 — €1.94 €100.88

Infliximab + methotrexate 20 15 €23.15 €185.20

Rituximab + methotrexate 56.66 30 €53.46 €106.92

Table 4. Annual direct medical costs per patient based on ACR functional class.2

Direct medical costs

ACR functional class

I II III IV

Diagnostic examination

Physical €61.4 €79.0 €105.3 €114.6

Instrumental €55.8 €88.8 €77.0 €73.3

Laboratory €130.6 €162.6 €177.6 €206.5

Therapy

Physiotherapy €43.4 €65.6 €89.8 €157.0

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

33

Hospitalisation

Routine €399.6 €751.7 €1466.7 €1968.5

Rehabilitation €106.4 €94.0 €0.0 €0.0

Day €60.9 €101.7 €59.4 €13.4

Other costs

Transportation €113.1 €110.5 €247.8 €134.7

House help €50.1 €222.0 €840.0 €1689.7

Auxiliary devices €64.0 €160.0 €121.8 €236.4

Total €1085.3 €1835.9 €3185.4 €4594.1

ACR, American College of Rheumatology.

Table 5. HAQ score-related direct medical costs.9,34

ACR functional class HAQ score Direct medical costs

I <0.25 €1085.3

I 0.25–0.50 €1085.3

I 0.50–0.75 €1085.3

II 0.75–1.00 €1835.9

II 1.00–1.25 €1835.9

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

34

II 1.25–1.50 €1835.9

III 1.50–1.75 €3185.4

III 1.75–2.00 €3185.4

III 2.00–2.25 €3185.4

IV 2.25–2.50 €4594.1

IV 2.50–2.75 €4594.1

IV 2.75–3.00 €4594.1

ACR, American College of Rheumatology; HAQ, Health Assessment Questionnaire.

Table 6. Economic model results

Standard-of-care sequence Tocilizumab sequence

Total cost QALYs Total cost QALYs

ICER

(€ per QALY)

Replace etanercept

(Base case analysis) €143,547.04 9.3502 €141,181.56 9.8881 TCZ dominates

Replace adalimumab

(alternative scenario 1) €143,388.70 9.3502 €142,268.81 9.8881 TCZ dominates

Replace infliximab

(alternative scenario 2) €140,840.39 9.3502 €142,268.81 9.8881 €2,655.32

Add tocilizumab to

standard-of-care

(alternative scenario 3)

€143,547.04 9.3502 €164,086.87 10.5500 €17,119.29

Sensitivity analysis

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.

35

A

Use unadjusted-response

probabilities

€151,060.59 9.4758 €143,483.63 9.7945

TCZ dominates

B

Use fixed time on

treatment (3.25 years)

€114,794.31 8.5135 €112,892.57 8.9063

C

Use 0 HAQ score

progression on TCZ

€143,547.04 9.3502 €141,473.95 9.4121

D

Use Bansback et al28 QoL

mapping formula

€143,547.04 7.5848 €141,181.56 8.1721

E

Do not discount QoL €143,547.04 13.0244 €141,181.56 13.9314

F

Set rebound effect to 50% €138,372.22 11.6686 €135,611.44 12.1251

G

Remove resource use cost €112,736.42 9.3502 €110,950.95 9.8881

H

Set withdrawal risk to 8% €156,771.27 9.7187 €154,235.14 10.3298

I

Set withdrawal risk to 12% €133,638.79 9.0447 €131,421.93 9.5268

HAQ, Health Assessment Questionnaire; ICER, incremental cost-effectiveness ratio;

QALY, quality-adjusted life-year; QoL, quality of life; TCZ, tocilizumab.

Jour

nal o

f M

edic

al E

cono

mic

s do

wnl

oade

d fr

om in

form

ahea

lthca

re.c

omby

Hof

fman

n-L

a R

oche

AG

on

02/1

0/12

. For

per

sona

l use

onl

y.