View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Volume 13 • Number 8 • 2010 VALUE IN HEALTH

Cost-Effectiveness of Maraviroc for Antiretroviral

Treatment-Experienced HIV-infected Individuals in Mexicovhe_798 903..914

Iris Contreras-Hernandez, MD, MSc,1 Debbie Becker, MSc,2 Jeremy Chancellor, MSc,3 Felicitas Kühne, MSc,4 Joaquin Mould-Quevedo, PhD, MSc, MBA,5 Gabriela Vega, MSc,6 Shalaka Marfatia, BPharm, MS, MPH7

1Health Economics Research Unit, Social Security Mexican Institute, Mexico City, Mexico; 2Health Economics and Outcomes Research, i3 Innovus, Burlington, ON, Canada; 3Health Economics and Outcomes Research, i3 Innovus, Uxbridge, United Kingdom (at the time of the conduct of this study); 4Health Economics and Outcomes Research, i3 Innovus; PharmacoConsult, Uxbridge, United Kingdom; 5Pfizer Mexico, Mexico City, Mexico; 6Pfizer Mexico, Mexico City, Mexico (at the time of the conduct of this study); 7Pfizer Inc., New York, NY, USA (at the time of the conduct of this study)

ABSTRACT

Objective: Maraviroc is the first approved drug in a new class of antiret- undiscounted life expectancy and discounted quality-adjusted life expect- rovirals, the CCR5 antagonists. The objective of this study was to predict ancy from 7.54 to 8.71 years and 4.42 to 4.92 quality-adjusted life years the long-term clinical impact and cost-effectiveness of maraviroc in (QALYs), respectively, at an incremental cost of $228,215 (US$21,329). treatment-experienced adults with HIV/AIDS in Mexico. The resultant incremental cost-effectiveness ratio (ICER) was $453,978 Methods: The AntiRetroviral Analysis by Monte Carlo Individual Simu- (US$42,429) per QALY gained. The ICER was somewhat lower when lation (ARAMIS) model was adapted to the Mexican context to predict maraviroc was modeled in individuals susceptible to Յ2 components of clinical and economic outcomes of treating with optimized background OBT ($407,329; US$38,069), while the ICER was higher in individuals therapy (OBT) versus testing for viral tropism status and treating with susceptible to Ն3 OBT components ($718,718; US$67,171). OBT Ϯ maraviroc accordingly in treatment-experienced adults in Mexico. Conclusion: In treatment-experienced individuals with HIV/AIDS in Baseline characteristics and efficacy were from the MOTIVATE trials’ Mexico, maraviroc may be cost-effective, particularly in individuals with screening cohort. Costs and population mortality data were specific to limited options for active antiretroviral therapy (ART). Mexico. Results were reported from the perspective of health care payers Keywords: AIDS, anti-HIV agents, cost-effectiveness, economic model, in 2008 Mexican pesos (converted to 2008 US$ in parentheses). HIV, maraviroc, Mexico. Results: Compared to treatment with OBT alone, treatment with OBT Ϯ maraviroc contingent on tropism test result increased projected

Introduction strains after long exposure to ART therapies, resulting in treat- ment failure. Acquired immunodeficiency syndrome (AIDS) continues to be a The newly developed drug maraviroc is the first of a class major health epidemic with an estimated 39.5 million people known as CCR5 antagonists. Maraviroc selectively and revers- living with human immunodeficiency virus (HIV) worldwide and ibly binds to CCR5 receptor on CD4+ cells, preventing interac- 2.9 million people dying from AIDS in 2006 [1]. Based on data tion with the gp120 protein of CCR5-tropic virions and from the Mexican National Center for the Prevention and the subsequent entry. In double-blind, placebo-controlled clinical Control of the HIV/AIDS, it was estimated in 2006 that 182,000 trials, maraviroc in addition to optimized background therapy people in Mexico were living with the disease [2,3]. In 2007, the (OBT) were compared to OBT alone in treatment-experienced World Health Organization (WHO) estimated HIV prevalence in adults infected with CCR5-tropic HIV-1 [7]. Patients treated Mexico, among adults aged 15 years and older, to be 244 per with maraviroc achieved a substantially greater reduc- 100,000 population [4]. From 1992 to 1997, only social security tion than did patients treated with OBT alone (-1.84 and -0.78 beneficiaries had free access to antiretroviral therapies (ARTs) log10 copies/mL, respectively), with over twice the proportion of and HIV care in Mexico [5]. However, by 2006, accessibility to patients achieving undetectable levels of HIV-1 (56.1% and ART had increased to 78% of the infected Mexican population 22.5% with <400 copies/mL at 48 weeks) [8,9]. [6]. To predict the long-term clinical and economic impact of Currently, highly active antiretroviral therapy (HAART) rep- maraviroc as add-on therapy to OBT, we used the Anti Retroviral resents the mainstay among treatments for HIV-infected indi- Analysis by Monte Carlo Individual Simulation (ARAMIS) viduals, which suppress viral load, improve immune function, model [10] to analyze the cost-effectiveness of maraviroc for the and delay the progression of the disease. Since its introduction in management of individuals infected with only CCR5-tropic 1996, HAART has dramatically improved survival rates for indi- HIV-1 detectable, and who have evidence of viral replication and viduals with HIV infection. However, new classes of antiretrovi- HIV-1 strains resistant to multiple antiretroviral agents (The ral agents with novel mechanisms of action are now receiving ARAMIS cost-effectiveness model is the property of Pfizer Inc., greater consideration due to side effects and poor tolerability and is protected by US and international copyright laws. Unau- with existing agents and the emergence of drug-resistant viral thorized reproduction, modification, republication, distribution, or display of all, or any portion, of the ARAMIS cost- Address correspondence to: Debbie Becker, i3 Innovus, 1016 Sutton Drive, effectiveness model is prohibited). The analysis was conducted Burlington, ON, Canada L7L 6B8. E-mail: [email protected] from the perspective of the health care payer in Mexico to a 10.1111/j.1524-4733.2010.00798.x lifetime horizon.

© 2010, International Society for Pharmacoeconomics and Outcomes Research (ISPOR) 1098-3015/10/903 903–914 903 904 Contreras-Hernandez et al.

Methods Analytic Overview The analysis was conducted using ARAMIS, a previously vali- dated and published computer-based microsimulation model of HIV disease progression and treatment [10]. The ARAMIS model was designed to be applicable to countries with well- developed systems for the care of individuals with HIV/AIDS, and was adapted to include Mexico-specific inputs. The model was used to extrapolate 48-week results from the MOTIVATE-1 and -2 studies and predict clinical and economic outcomes over a lifetime, using disease progression data derived from epidemio- logical cohort studies. Individuals screened for entry to the MOTIVATE trials (the “MOTIVATE screening cohort”) included adults with an HIV-1 RNA of greater than 5000 copies/mL despite at least 6 months of prior therapy with at least one agent from three ART classes or documented resistance or intolerance to at least one member of each class [8,9]. All indi- Figure 1 Markov state transition diagram for the ARAMIS model. OI, oppor- viduals in the MOTIVATE screening cohort were tested for viral tunistic infection. ↓ Denotes allowable transitions. ↑ Denotes transitions only tropism status, and only those in whom only CCR5 tropic virus allowable with treatment. was detectable were eligible for entry into the trial. The primary economic analysis was a comparison of two treatment strategies: 1) a “contingent maraviroc” strategy in untreated patients is known to be strongly related to the HIV which individuals would be treated with OBT, with or without viral load set point (i.e., the stabilized viral load in the postacute the addition of maraviroc according to viral tropism test results phase) (Table 1) [15]. As set point data were unavailable for the (“OBT Ϯ maraviroc”) versus 2) an “OBT alone” strategy of study population, baseline viral load was deemed to be a reason- treating the entire population with OBT alone, without testing able proxy for modeling the rate of decline of CD4+ cells in for tropism. Hence, the economic comparison was specified patients failing treatment. The loss of CD4+ cells in HIV-infected slightly differently from the design of the MOTIVATE trials, in individuals who are untreated or failing treatment drives transi- order to represent the ex ante policy decision on the use of tion through successively lower CD4+ states, while successful maraviroc in a treatment-experienced population, without prior antiretroviral treatment results in reconstitution of CD4+ cells knowledge of individuals’ tropism status. The primary outcomes and transitions from lower to higher CD4+ states. The CD4+ cell of the analysis were lifetime costs, life expectancy and quality- count trajectory and other aspects of each individual’s simulated adjusted life years (QALYs) resulting from each treatment strat- lifetime clinical history are recorded within the model. egy. Incremental cost effectiveness ratios (ICERs), including cost Independently of CD4+ cell count, the risk of an opportunis- per life year and cost per QALY gained, were calculated and tic infection and chronic HIV mortality is modified by history of reported in 2008 Mexican pesos (with values converted to 2008 an opportunistic infection and current exposure to antiretroviral US$ reported in parentheses). The analysis was conducted using treatment [13,14]. In the model, causes of death include chronic a lifetime time horizon and a Mexican payer perspective. HIV disease, acute fatal opportunistic infection, and all-cause mortality. The ARAMIS Model Structure and Global Parameters The probabilities of progression between health states each The ARAMIS model has previously been described within the cycle depend on each individual’s baseline characteristics, includ- context of an analysis of maraviroc conducted from a US payer ing current CD4+ status, viral load set point and history of perspective [10]. The microsimulation model simulates the pro- opportunistic infections. The risk of events does not remain gression of a cohort, one individual at a time, through a set of fixed through the model, but depends on each individual’s health states representing the natural history of HIV/AIDS from accumulated history of events, particularly the occurrence of initial infection to death. Six mutually exclusive health states are opportunistic infections and antiretroviral treatment status. Con- defined by CD4+ cell count categories (>500, 351–500, 201–350, currently, individuals progress through successive lines of ART 101–200, 51–100, and <50 cells/mL), and mortality is modeled as contingent on the occurrence of virologic failure or adverse a seventh, absorbing state (Fig. 1). In a given 1-month cycle, an events. With successful treatment, viral load is assumed to drop individual may remain in the same CD4+ state, progress to an below detectable levels (<400 copies/mL) at 48 weeks and is adjacent CD4+ state, enter an opportunistic infection transition described as “suppressed.” Although tests with a lower detection state, or die. limit of 50 copies/mL are now in common use, the 400 copies/mL Disease progression is modeled as CD4+ cell decline, the occurrence of an acute opportunistic infection, or death. CD4+ cell count is the marker of current disease state (which is tracked Ta b l e 1 Mean monthly CD4+ cell count decline by HIV RNA stratum at an individual patient-level using a tracker variable) [11]. [15] Depending upon treatment status and history of opportunistic infections, CD4+ cell count determines the subsequent risk of CD4+ count 3 opportunistic infections and HIV-related mortality that does not HIV RNA (copies/mL) (cells/mm ) occur within the acute phase of an opportunistic infection. Mor- >30,000 6.375 tality within the month immediately after a new OI (referred to 10,001–30,000 5.400 as “acute HIV mortality”) is captured separately in the model. 3,001–10,000 4.600 501–3,000 3.733 Data on disease progression were based on published cohort Յ500 3.025 studies [12–16]. The natural rate of decline of CD4+ cells in Cost-Effectiveness of Maraviroc for HIV in Mexico 905

Ta b l e 2 Baseline characteristics of cohorts from the MOTIVATE trials experience acute opportunistic infections, which are additive to [8,9] the monthly routine care cost. Each CD4+ state and opportunistic infection state is associated with a utility value ona0to1scale,by Screening Trial population Characteristic population (CCR5 monotropic) which survival is quality-adjusted to allow calculation of QALYs. Opportunistic infections are modeled as transitions that may be Mean age at baseline (years) 45.3 45.6 encountered at each cycle. In this manner, individuals incur tran- Gender (% male) 86 87 sition tolls (cost of treatment and quality of life decrements) when Mean CD4+ count (cells/mL) 183.94 217.15 opportunistic infections occur. Viral load set point (log10 copies/mL 4.72 4.76 HIV-1 RNA) Proportion receiving with 58 56 Study cohort. The distributions of the baseline characteristics of optimized background therapy (%) the cohort simulated in this study were directly derived from Proportion exclusively CCR5 tropic (%) 51 100 Proportion entering with history of individual level data for the MOTIVATE screening cohort, as opportunistic infection (%) previously described (mean values are shown in Table 2, Bacterial 7 8 n = 3244 patients screened) [7–9]. A secondary analysis was Fungal 43 41 performed simulating the subset of individuals with only CCR5- Viral 20 19 Protozoal 3 3 tropic HIV-1 detectable after tropism testing, and who were thus Other illness 30 28 eligible to enroll in MOTIVATE (the “MOTIVATE trial cohort”). Any opportunistic infection 67 63 ART intervention. The treatment effects for individuals in the OBT alone strategy and for individuals in the contingent maravi- roc strategy who were not CCR5-tropic were derived from criterion is used in ARAMIS because many of the clinical trials pooled 48-week data from the placebo arms of the MOTIVATE used to calibrate the model reported only suppression rates to -1 and -2. The treatment effects for CCR5 positive individuals in <400 copies/mL [16,17]. Forty-eight weeks has become estab- the maraviroc arm were derived from 48-week pooled data from lished as a standard length of follow-up for clinical trials of ART. the 300 mg twice-daily maraviroc arms from the same studies In ARAMIS, when HIV is suppressed, the decline in CD4+ cell (Table 3) [8].] Within the model, CD4+ cell count increased in count is arrested. If suppression is sustained, treatment specific each cycle that viral suppression was sustained. Consistent with CD4+ count recovery is modeled with time. the normal pattern observed in trials of ARTs [8,18–22], a rapid Each cycle, ARAMIS updates its account of each individual’s CD4+ increase was modeled during the first two months of accumulated survival time, along with the associated costs and effective antiretroviral therapy, and a slower rate thereafter. The quality-adjusted survival associated with his or her pathway modeled rate of CD4+ cell recovery to 48 weeks was treatment- through the modeled health states and treatment. Specific monthly specific and based on observed data from the MOTIVATE trials. costs are assigned to treatment status and to each CD4+ state, the Beyond 48 weeks, CD4+ increases were conservatively latter representing the costs of routine health care in HIV/AIDS, as assumed not to differ between treatments and were estimated by described later. Additional cost tolls are imposed when individuals linear extrapolation from the observed trial data. Individuals

Ta b l e 3 Efficacy parameters from the MOTIVATE trials

MVC + OBT (CCR5+) OBT alone First regimen Mean Mean

First year virologic suppression rate (<400 copies/mL) Pooled over both trials [8]* 56.1% 22.5% Monthly late failure rate† Week 48 to week 96 0.75% 1.06% Thereafter 1.06% 1.06% Monthly CD4+ increase when successful (cells/mL)‡ First 2 months 54.1 39.3 Months 2 to 12§ 5.4 5.3 Thereafter§ 5.6 5.6 Monthly adverse event risk (grade 3 and 4)* 1.2% 1.0%

Subsequent regimen, given failure of first regimen 20%# 80% 20%# 80% (reoptimized background therapy; rOBT) [42,43]¶ OBT + T20 OBT OBT + T20 OBT

First year virologic suppression rate (<400 copies/mL) 30.4% 12% 30.4% 12% Monthly late failure rate 1.87% 1.87% 1.87% 1.87% Monthly CD4+ cell increase when successful (cells/mL)‡ First 2 months 33.64 16.63 33.64 16.63 Months 2 to 12 3.39 1.68 3.39 1.68 Thereafter 3.50 1.73 3.50 1.73

*Pooled 48-week data from trials A4001027 and A4001028 (MOTIVATE-1 and -2). Maraviroc efficacy taken from 300 mg twice-daily maraviroc arms of trials. Pfizer Global Research and Development, data on file. †The late failure rates in fully suppressed patients are estimated by projecting the rate of viral rebound between week 16 and week 48. ‡Success is defined as viral load dropping below detectable levels (<400 copies/mL) at 48 weeks. §Estimated by linear regression on pooled (MVC and placebo arms) CD4+ count data from week 8 to 48. ¶Failure on either of the compared regimens results in a switch to an unspecified reoptimized background therapy regimen; a four-month time lag was assumed between failure and switching, to mimic the likely delay in discovering and confirming virologic failure in clinical practice. #20% of patients were assumed to receive T20 in their subsequent OBT regimen based on unpublished information for the MOTIVATE open-label continuation phase. MVC, maraviroc; OBT, optimized background therapy. 906 Contreras-Hernandez et al. who experienced virologic failure after 48 weeks were considered Ta b l e 4 Utility values by model health state “late failures.” The late failure rate was treatment-specific and derived by linear regression from the observed failure rate from Health state Utility value Duration* the peak of suppression at week 8 through week 48 in the Chronic HIV by CD4+ category [26] MOTIVATE trials. As no long-term studies provide information CD4+ >500 0.870 Present month + on sustainability of suppression beyond 10 years, all patients CD4 351–500 0.860 Present month CD4+ 201–350 0.860 Present month modeled as suppressed at 10 years were conservatively assumed CD4+ 101–200 0.850 Present month to fail at this point in time. CD4+ 51–100 0.850 Present month After an individual failed either of the compared regimens, CD4+ <50 0.832 Present month they were assumed to switch to an unspecified reoptimized back- Acute OI [27] Bacterial 0.561 3 months ground therapy regimen (rOBT), based on treatment history and Protozoal 0.561 3 months the results of resistance testing. A 4-month time lag was assumed Viral 0.652 3 months between failure and switching, to mimic the likely delay in dis- Fungal 0.652 3 months covering and confirming virologic failure in clinical practice. Other illness 0.561 3 months Post-acute OI [26] Based on unpublished data from the open-label continuation Bacterial 0.735 6 months phase of the MOTIVATE studies, it was assumed that for 20% of Protozoal 0.731 Rest of life these switched patients, the rOBT regimen would include enfu- Viral 0.760 Rest of life virtide. After virologic failure, the CD4+ cell count was modeled Fungal 0.743 6 months Other illness 0.770 Rest of life to remain static for 12 months [23] and thereafter to decline at a Disutility of AEs 0.140 Present month rate determined by the viral load set point [23]. For initial fail- ures (i.e., those whose CD4+ count never reaches <400 *When more than one criterion applies, the lowest utility value is applied. copies/mL during their lifetime), CD4+ cell count was modeled to AEs, adverse events; OI, opportunistic infection. remain static for 12 months after the start of the model, consis- tent with the assumption of virologic failure to the treatment received prior to entering the analysis. Failing patients were tality data for risk of deaths due to non-HIV-related causes [29], assumed to remain on ART, and, consistent with evidence from and a discount rate for costs and outcomes of 5% [30]. the literature [13,14], they were assumed to benefit from a The scope of Mexico-specific costs included direct monthly reduced risk of opportunistic infection incidence conferred by health care costs for routine HIV care (for both inpatients and treatment in the absence of suppression. outpatients) including ART, costs of tropism testing, adverse ART-related adverse events were modeled based on pooled events associated with ART, acute and prophylactic treatment of rates of drug-related grade 3 and 4 adverse events from the opportunistic infections, CD4+ cell count tests, HIV RNA tests, MOTIVATE studies, adjusted for exposure time. Monthly risks treatments, and HIV- or opportunistic infection-related palliative of an adverse event for individuals receiving OBT plus maraviroc care preceding deaths. The daily cost of maraviroc was $181.50 and OBT alone were 1.2% and 1.0%, respectively. (US$16.96) based on 300 mg twice daily dosing. The complete set of cost inputs is presented in Table 5. Mexican pesos were Routine care and testing. Prophylaxis against Pneumocystis inflated [31] and converted to US$ using a published rate of $1 (PCP) and Mycobacterium avium complex (MAC) was assumed Mexican Peso = US$0.09346 [32]. to be initiated and ended when the CD4+ cell count passed Patterns of resource use were estimated based on a review of thresholds of 200 and 50 cells per ml, respectively, and consisted patient records from nine hospitals in Mexico City belonging to of trimethoprim-sulfamethoxazole 800 mg/160 mg (TMP-SMX) Social Security Mexican Institute (IMSS). The data collection [24]. Routine testing was assumed to include: CD4+ cell count, study gathered information from 637 patients treated for HIV/ HIV-1 RNA level, and, after treatment failure, genotypic resis- AIDS at one of the participating hospitals during 2004 to 2005. tance testing, and was assumed to be performed according to US The study sample had an average age of 49 years (standard clinical guidelines [25]. The cost of one viral tropism test was deviation [SD] 9 years) and was 87% male. At the time of data applied to all individuals in the maraviroc treatment strategy collection, patients had been infected with HIV for an average of arm. Although the tropism test, as it was available and used in 8.8 years (SD 2.3), with 46% of the sample being diagnosed with MOTIVATE trials, is known to be imperfect, sensitivity and AIDS and 78% having failed at least one line of therapy. specificity were set to 100% in the analysis to avoid double- Through review of clinical records, the number of physician counting, because individuals with false-positive test results con- visits, laboratory tests, and utilization of retroviral, non- tributed to the reported results for the OBT+ maraviroc arms of retroviral, and opportunistic infection-related therapies were the trials. The frequency of the remaining tests followed clinical identified to define patterns of routine care for individuals with guidelines. HIV/AIDS. Data related to hospitalizations for HIV/AIDS- Quality of life. With the exception of the impact of adverse related treatment were also collected, including length of stay events, utility values used in the model for calculating quality- and frequency of inpatient visits, tests, surgeries, and medical adjusted life expectancy were based on published estimates procedures. Monthly routine care costs by CD4+ cell count for (Table 4) [26,27]. In the absence of published utility estimates of individuals with and without a history of opportunistic infection the impact of adverse events in HIV infection, a decrement of were determined by dividing the sum of all resources used within 0.14 was applied for one cycle, based on expert opinion, for all a year by 12 months. To determine costs associated with acute patients experiencing a grade 3 or 4 adverse event. This is some- opportunistic infection treatment, HIV/AIDS-related resources what less than the utility decrement of 0.20 reported for gas- utilized 30 days prior and 60 days after the diagnosis were trointestinal adverse events in chemotherapy [28]. considered. Because the metropolitan area from where the par- ticipating hospitals are located shows the highest HIV prevalence Adaptation of ARAMIS to Mexico in Mexico, it is likely that the patterns of resource utilization Adaptation of ARAMIS to the Mexican setting involved use of observed are representative of the average treatments used in Mexico-specific costs, age- and gender-specific population mor- Mexico. Drug costs were obtained from the IMSS [33], while Cost-Effectiveness of Maraviroc for HIV in Mexico 907

Ta b l e 5 Summary of cost inputs mean unit costs for other HIV/AIDS-related health care resources were based on costs collected from the hospital databases of the 2008 Mexican nine participating sites. It was assumed that patients experiencing pesos (US$) grade 3 or 4 adverse events due to ART incur the additional cost Daily ART drug costs [33] of one hospital inpatient day per event; for simplicity, efforts MVC (300 mg twice daily) 181.50 (16.96) were not made to model individual types of adverse events. OBT 144.80 (13.53) Enfuvirtide (90 mg twice daily subcutaneous 796.41 (74.43) injection) Results Routine care cost (per month)*† No OI history CD4+ >500 1,577 (147) Base-Case Analysis CD4+ 351–500 1,406 (131) Applying the base-case assumptions to the MOTIVATE screening CD4+ 201–350 1,512 (141) cohort, the model predicted that the contingent maraviroc strat- CD4+ 101–200 1,737 (162) CD4+ 51–100 1,940 (181) egy would lead to an increase in mean undiscounted life expect- CD4+ <50 2,094 (196) ancy of 14 months (15.5% increase; 8.7 vs. 7.5 years), a OI history 1,687 (158) 10-month increase in AIDS-free survival time (38%; 35.3 vs. OI prophylaxis (per month) [33] 25.5 months), a decrease in the proportion of individuals expe- M. avium complex 223 (21) P. carinii 9.3 (1) riencing AIDS of 2.4% (87.4% vs. 89.8%), and a decrease in the Acute OIs (per episode)† proportion of deaths caused by HIV (87.9% vs. 90.2%) relative Bacterial 82,285 (7,960) to the OBT alone strategy. The contingent maraviroc strategy Fungal 53,928 (5,040) produced a decrease in the incidence of all types of opportunistic Other 41,703 (3,898) Protozoal 64,732 (6,050) infections (19.2% vs. 20.0%) and a corresponding increase in the Viral 78,811 (7,366) mean time to the first opportunistic infection event among inci- Last month of life costs dent cases (10.2 vs. 8.1 years). The average time to treatment HIV death (assumed to be 69% of acute OI 56,784 (5,307) failure was 28.5 months for the contingent maraviroc strategy death cost [44]) Acute OI death (assumed same cost as bacterial 82,285 (7,690) compared to 18.7 months for the OBT alone strategy (Table 6). OI episode) Corresponding outcomes for the comparison between treating Test costs (per test) with OBT + maraviroc versus treating with OBT alone in CCR5 Tropism test 7,700 (720) monotropic individuals (i.e., based on the MOTIVATE trial CD4+ cell count 1,249 (117) Viral load test 2,458 (230) cohort) are shown in Table 6. Treatment-related AE‡ 4,463 (417) Figures 2 and 3 show the impact of treatment strategy on overall survival and on the evolution of the CD4+ cell count *Includes inpatient and outpatient care cost during one month. distribution at baseline and after 3, 5, and 7 years. The stacked † = Resource patterns and unit costs collected through review of patient records (n 637) at + nine hospitals in Mexico. bar segments in Figure 3 represent the CD4 categories of those ‡Represents cost of a 1-day hospital stay. individuals in the study population who are still alive, while the US$ = 2008 US$. AE, adverse event; MVC, maraviroc; OI, opportunistic infections; OBT,optimized background height of the bars represents the surviving proportion of the therapy. starting population. The upper graphs show the CD4+ cell count

Ta b l e 6 Undiscounted clinical outcomes predicted for the base-case scenario

MOTIVATE screening cohort MOTIVATE trial cohort (CCR5 monotropic) Treat with Test and treat with Treat with Treat with Clinical outcomes OBT alone OBT Ϯ MVC OBT alone OBT+MVC

Life expectancy (years) 7.54 8.71 8.12 10.46 Proportion experiencing AIDS 89.8% 87.4% 88.0% 83.2% AIDS-free survival (months) 25.49 35.25 31.13 50.30 Causes of death HIV 90.2% 87.9% 89.2% 84.6% Acute OIs 1.2% 1.2% 1.3% 1.1% Other non-HIV/AIDS-related 8.6% 10.9% 9.5% 14.3% Time spent on each line of therapy (months) MVC + OBT or OBT 18.71 28.51 19.05 38.45 Subsequent reoptimized OBT 71.75 76.04 78.45 87.12 Proportion of patients experiencing first or new OI Bacterial 4.9% 4.7% 5.1% 4.7% Fungal 7.1% 6.8% 7.5% 6.7% Viral 2.2% 2.1% 2.3% 2.0% Protozoal 3.4% 3.1% 3.6% 3.0% Other 4.8% 4.7% 5.0% 4.9% Mean time to OI among patients with incident OI (years) Bacterial 9.04 10.54 9.67 12.54 Fungal 8.70 10.09 9.20 12.12 Viral 9.28 10.72 9.92 12.76 Protozoal 9.45 10.74 10.08 12.92 Other 8.50 9.76 8.98 11.45 Any 8.81 10.17 9.36 12.10

MVC, maraviroc; OI, opportunistic infection; OBT, optimized background therapy. 908 Contreras-Hernandez et al.

The incremental cost-effectiveness analysis results for the MOTIVATE screening cohort are shown in the upper section of Table 7. The estimated average discounted quality-adjusted and unadjusted life expectancies with the OBT alone strategy were 4.4 QALYs and 5.4 years, respectively. The contingent maraviroc strategy increased the discounted quality-adjusted and unad- justed life expectancies by 0.5 QALYs and 0.6 years to 4.9 QALYs and 6.0 years, respectively. These benefits were gained at an incremental discounted lifetime cost of $228,215 (US$21,329), yielding ICERs of $453,978 (US$42,429) per QALY gained and $383,904 (US$35,880) per life year gained. The incremental cost-effectiveness results for the MOTIVATE trial cohort are shown in lower section of Table 7. ICERs of $443,823 (US$41,480) per QALY gained and $190,211 (US$17,777) per life year gained were predicted when comparing a strategy of treating with OBT + maraviroc to treating with OBT alone.

Sensitivity and Subgroup Analyses Figure 2 Survival Curves (showing only first 10 years of lifetime analysis) for A series of sensitivity analyses were conducted to explore the MOTIVATE screening cohort and MOTIVATE trial cohort by treatment strat- impact on the ICERs of alternative assumptions for the values of egy.The solid curves show the predicted survival for the screening cohort.The dotted curves represent the individuals that tested CCR5-tropic.The maraviroc key input parameters. The following parameters and ranges were (MVC) treatment strategies are shown in grey and the OBT alone strategies are tested: shown in black. 1. OBT + maraviroc efficacy: the mean 48-week virological suppression for OBT + maraviroc of 56.1% was replaced distribution resulting from the use of OBT alone and by the upper and lower bounds of the 95% confidence OBT Ϯ maraviroc strategies, respectively, in the MOTIVATE interval (60.8% and 51.4%); screening cohort, while the lower two graphs present the CD4+ 2. OBT + maraviroc efficacy: the mean 48-week virological cell count distribution resulting from the use of OBT alone and suppression rate (56.1%), CD4+ cell count increase (0 to 8 OBT + maraviroc in the MOTIVATE trial cohort, respectively. weeks 54.1, 8 to 48 weeks 5.4, 48 to 96 weeks 5.6), and After 3, 5, and 7 years, 22.5%, 28.1%, and 28.1% of individuals monthly late failure rate (-0.75% to 96 weeks then -1.06% in the MOTIVATE trial cohort treated with OBT + maraviroc thereafter) for OBT + maraviroc were replaced by the upper were projected to have a CD4+ cell count greater than 500, as (suppression rate 60.8%; CD4+ increase 60.3, 6.1, 6.3; late compared to 10.2%, 11.9%, and 11.9% of those treated with failure rate -0.54% then -0.72%) and lower bounds (sup- OBT alone. pression rate 51.4%; CD4+ increase 47.9, 4.8, 4.9; late

Figure 3 CD4+ cell development based on model results. The upper two charts present the CD4+ evolution in the screening cohort; the lower charts present the same data for the trial cohort. The charts on the left represent treatment with OBT alone; the charts on the right represent the maraviroc (MVC) treatment strategies. In each chart, the lower and darker segments of the stacked bars represent lower CD4+ categories, while the upper, lighter segments represent higher CD4+ categories. The total height of the stacked bars reflects the proportion of patients alive at the indicated time points. Cost-Effectiveness of Maraviroc for HIV in Mexico 909

Ta b l e 7 Total discounted cost, quality-adjusted life years (QALYs), incremental cost-effectiveness ratios (ICERs) for the base-case analysis in 2008 Mexican pesos (US$)*

Cost ($) QALY Life years ICER Strategy Total (US$) Incr. (US$) Total Incr. Total Incr. Cost per QALY Cost per life year (US$)

MOTIVATE screening cohort Treat with OBT alone 1,002,664 (93,709) 4.4209 5.4033 Test and treat with MVC 1,230,879 (115,038) 228,215 (21,329) 4.9236 0.5027 5.9977 0.5945 453,978 (42,429) 383,904 (35,880) contingent on test result MOTIVATE trial cohort Treat with OBT alone 1,042,604 (97,442) 4.7282 8.1249 Test and treat with MVC 1,487,511 (139,023) 444,904 (41,581) 5.7306 1.0024 10.4639 2.3390 443,823 (41,480) 190,211 (17,777)

*Discount rate was 5% for costs and outcomes per annum. MVC, maraviroc; OBT, optimized background therapy.

failure rate -0.96% then =1.40%) of the 95% confidence already reflects false positives. Therefore, the results of this intervals; sensitivity analysis in which the sensitivity and specificity 3. OBT and OBT + maraviroc efficacy: the efficacy parameters values were lowered should be considered with caution for both OBT and OBT + maraviroc were changed to their because this approach allows for double-counting; upper and then to their lower bounds simultaneously. The 12. Cost of maraviroc: $181.50 per day was changed to $165 parameters for OBT + maraviroc were as described above. and $220 per day (US$: $16.96, $15.42, and $20.56, The mean parameters (upper, lower 95% CI) for OBT were: respectively) reflecting potential variation in the drug cost suppression rate 22.5% (28.1%, 16.8%); CD4+ cell count within the Mexican market; increase 39.3 (53.3, 25.2), 5.3 (7.8, 2.2), and 5.6 (6.3, 4.9), 13. Cost of OBT: $144.80 per day was increased and decreased respectively; and late failure rate -1.06% (-0.72%, by 50% reflecting the fact that OBT is based on individual -1.40%); resistance profiles and can vary greatly among patients; 4. OBT and OBT + maraviroc efficacy: the late failure rates for 14. Variation in utility values: was assessed by altering the OBT and OBT + maraviroc were changed to their upper and base-case values by Ϯ10% that covers generally the ranges then to their lower bounds assuming the values described of utility values published by other researchers [34,35]; previously; 15. Discount rate: for costs and outcomes was set to zero to 5. Rates of CD4+ cell increase after 48 weeks for suppressed assess the impact of discounting. patients: were changed from the base-case values of 5.6 cells per month (OBT + maraviroc), 5.6 cells per month (OBT In addition to the deterministic and multiway sensitivity alone), 3.5 cells per month (rOBT + T20), and 1.75 cells per analyses, a subgroup analysis was conducted. In the MOTIVATE month (rOBT alone) to zero (i.e., no CD4+ increase) to test trials, phenotypic and genotypic resistance to protease inhibitors the most conservative scenario; (PIs), nucleoside reverse transcriptase inhibitors (NRTIs) and 6. Viral load set point: which was estimated using the non-nucleoside reverse transcriptase inhibitors (NNRTIs) were observed viral load of the MOTIVATE trial population at evaluated using the Monogram Biosciences PhenoSense GT baseline was changed to the lowest category (<50 copies/ (PSGT) assay at screening, weeks 24 and 48, time of treatment mL) and to the highest category (>100,000 copies/mL) failure and early termination visit. These results were used to because set point data were not available and the accuracy calculate genotypic (GSS), phenotypic (PSS), and overall (OSS) of our base-case assumption is therefore not known; susceptibility scores for the OBT activity in each patient. OSS is 7. For patients that are still suppressed, the time at which they a composite score, which takes into account both genotypic and are assumed to fail treatment: was assumed to be 10 years in vitro phenotypic susceptibility of the virus to each component in the base-case. For sensitivity analysis, this cutoff point of the OBT regimen [7]. For each drug in OBT, a score of “1” was set to 5 and 100 years; was assigned if the virus from the patient was sensitive to the 8. The time to treatment failure and failure rate for patients on drug and a score of “0” if the virus showed reduced susceptibility all treatments: was set to 2, 4, 6, 8, and 10 years after to the drug. Primary and secondary end points were evaluated treatment initiation, assuming a constant rate of failure within patient subgroups defined by GSS, PSS, and OSS of 0, 1, from 1 year for all patients using the DEALE approach (i.e., 2, and Ն3 at baseline. A treatment benefit for maraviroc for all 1/life expectancy). Failure two years after treatment initia- end points was observed for each OSS group. The MOTIVATE tion, for example, means that everyone fails treatment studies showed that the clinical benefit of maraviroc in patients between week 48 and 96. with CCR5-tropic virus, when used as add-on therapy to OBT 9. Lag time between treatment failure and switch to OBT: was compared to OBT alone, is greater in patients in whom fewer of changed from its base-case assumption of 4 months to 2 and the drugs comprising the OBT regimen are virologically active. 6 months, reflecting a window of -/+ 50%; To explore the economic implications of this clinical finding, the 10. Proportion of CCR5 monotropic patients: was changed analysis was repeated according to OSS subgroup. Observed from its baseline value of 51% to 30% and 80% to reflect virologic suppression rates for the two treatment arms for the variation that may be observed in treated populations; patients with an OSS of 2 or less and individuals with an OSS of 11. Sensitivity and specificity of the tropism test: were changed 3 or greater were substituted for the base-case values, holding all from the base-case values of 100% and 100%, respectively, other efficacy inputs constant (Table 8). to 85% and 94%, respectively. The base-case values of The results of these analyses are shown in Figure 4 and 100% were selected based on the fact that the efficacy data Table 9, which demonstrates that the results are relatively for patients in the OBT + maraviroc arms of the trials insensitive to the majority of the variables tested, with the model 910 Contreras-Hernandez et al.

Ta b l e 8 Efficacy parameters from the MOTIVATE trials for OSS sub- observed virologic suppression rates for the two treatment arms group analysis for patients with an OSS of 2 or less and individuals with an OSS of 3 or greater were substituted for the base-case values, the MVC + OBT (CCR5+)* OBT alone† respective ICERs were $407,329 (US$38,069) and $718,718 24 weeks 48 weeks 24 weeks 48 weeks (US$67,171) per QALY gained. Probability of treatment Although probabilistic sensitivity analysis (PSA) is very success at 24 and helpful in assessing the uncertainties in all model parameters 48 weeks‡§ simultaneously, a PSA was not conducted because of the compu- Յ OSS 2 56.80% 53.40% 13.90% 9.50% tational expense of the microsimulation approach when com- OSS Ն 3 70.60% 62.70% 58.20% 49.30% bined with PSA. While methods have been proposed to overcome Monthly probability of success during and this problem and were explored during model development, we after the first six months‡§ concluded that PSA was not feasible within ARAMIS. It was felt OSS Յ 2 0.910 0.990 0.719 0.939 by the authors that this trade off in favor of a microsimulation Ն OSS 3 0.944 0.980 0.914 0.973 approach was desirable to enable the modeling of switching

*Number of MOTIVATE subjects receiving maraviroc plus OBT in OSS Յ 2 was 296 and in because of sequential treatment failure over several lines of OSS Ն 3 was 126 (data pooled over both trials). therapy and to allow memory of individual patient progression † Number of MOTIVATE subjects receiving OBT in OSS Յ 2 was 137 and in OSS Ն 3 was 67 to impact subsequent probabilities. These capabilities are gener- (data pooled over both trials). ‡Pooled 48-week data from trials A4001027 and A4001028 (MOTIVATE-1 and -2). Pfizer ally not possible to incorporate using a cohort approach which is Global Research and Development, data on file. more amenable to PSA. Instead, a wide range of deterministic §Success is defined as viral load dropping below detectable levels (<400 copies/mL). MVC, maraviroc; OBT, optimized background therapy. and multiway sensitivity analyses were conducted to explore the impacts of uncertainty in the model. being most sensitive to changes in the discount rate and to the Discussion assumptions regarding the CD4+ increase in suppressed patients after 48 weeks. In these particular sensitivity analyses, results According to the National Center for the Prevention and Control ranges from as low as $343,714 (US$32,124) when no discount- of HIV/AIDS, over 182,000 HIV-infected people live in Mexico ing was applied to costs and outcomes, to as high as $598,397 [2]. As in the rest of the Latin American countries, the HIV/AIDS (US$55,926) when it was assumed that there was no increase in epidemiology in Mexico is concentrated in risk practice groups. CD4+ cell count in suppressed patients after week 48. The results In the late 1990s, the Mexican Health System began scaling up were most sensitive to the OSS subgroup analyses; when the access to ART. Unfortunately, increased access to HIV therapies

2008 US$

OSS Subgroups $407,329 $718,718 $38,069 / $67,171 (≤2; all; ≥3) CD4+ increase after 48 weeks when suppressed for all treatments $598,397 $42,429 / $55,926 (base; zero) Discount rate $343,714 $32,124 / $42,429 (0%; 5%) All efficacy parmeters for OBT+MVC $418,550 $515,620 $39,118 / $48,190 (upper 95% CI; mean; lower 95% CI) Alternative utility values $412,637 $504,556 $38,565 / $47,156 (110%; 100%; 90%) Cost per day of maraviroc $437,775 $491,828 $40,914 / $45,966 ($165; $181.5; $220) Cost per day of OBT $423,168 $484,813 $39,549 / $45,311 (50%; 100%; 150%) Time to treatment failure in suppressed patients $434,087 $477,720 $40,570 / $44,648 (5; 10; 100 years) Proportion CCR5 monotropic $455,161 $472,512 $42,539 / $44,161 (80%; 51%; 30%) Virologic suppression $440,675 $471,621 $41,186 / $44,078 95% CI OBT+MVC (60.8%; 56.1%; 51.4%) All efficacy parameters for OBT+MVC and OBT $444,849 $467,533 $41,576 / $43,696 (lower 95% CI; mean; upper 95% CI) Sensitivity/Specificity of tropism test $463,821 (100%/100%; 85%/94%) $42,429 / $43,349 Lag time between failure and switch to OBT (2; 4; 6 months) $452,567 $463,482 $42,297 / $43,317 Late failure rates for OBT+MVC and OBT (upper 95% CI; mean; lower 95% CI) $452,904 $462,073 $42,328 / $43,185 Treatment failure time after treatment initiation (constant rate)* $395,062 $459,342 (2; base; 10 years) $36,922 / $42,930 Viral load set point $448,867 $452,162 (<50; MOTIVATE baseline; >100,000) $41,951 / $42,259

$100,000 $250,000 $400,000 $550,000 $700,000 $850,000 Incremental cost per QALY gained (2008 Mexican Pesos)

Figure 4 Tornado diagram of sensitivity and subgroup analyses on discounted ICER for test and treat with maraviroc strategy compared to treat with OBT strategy for MOTIVATE screening cohort. Base-case ICER = $453,978/QALY (US$42,429) gained. CI, confidence interval; MVC, maraviroc; OBT, optimized background therapy; OSS, overall susceptibility score; *values for 2 years and 10 years shown; results for 4, 6, and 8 years were as follows: $424,455 ($39,670), $437,607 ($40,899), and $447,498 ($41,823), respectively. otEfcieeso aaio o I nMexico in HIV for Maraviroc of Cost-Effectiveness

Ta b l e 9 Total discounted cost, quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs) for the sensitivity and subgroup analyses in 2008 Mexican pesos (US$)*

Total Cost ($) Total QALYs ICER Test and treat with OBT Test and treat Cost/QALY Analysis OBT alone (US$) MVC (US$) Incr. (US$) alone with MVC Incr. Gained (US$)

Base-case 1,002,664 (93,709) 1,230,879 (115,038) 228,215 (21,329) 4.4209 4.9236 0.5027 453,978 (42,429) OSS subgroup Յ 2 906,931 (84,762) 1,170,956 (109,438) 264,025 (24,676) 4.0698 4.7179 0.6482 407,329 (38,069) OSS subgroup Ն 3 1,198,977 (112,056) 1,364,632 (127,538) 165,655 (15,482) 5.1473 5.3778 0.2305 718,718 (67,171) CD4+ increase after 48 weeks = 0 983,082 (91,879) 1,166,551 (109,026) 183,468 (17,147) 4.2468 4.5534 0.3066 598,397 (55,926) Discount rate 0% 1,380,485 (129,020) 1,719,848 (160,737) 339,364 (31,717) 6.1459 7.1333 0.9873 343,714 (32,124) All efficacy parameters for OBT + MVC upper 95% CI 1,002,664 (93,709) 1,295,482 (121,076) 292,818 (27,367) 4.4209 5.1205 0.6996 418,550 (39,118) All efficacy parameters for OBT + MVC lower 95% CI 1,002,664 (93,709) 1,177,717 (110,069) 175,053 (16,360) 4.4209 4.7604 0.3395 515,620 (48,190) Alternative utility values 110% 1,002,664 (93,709) 1,230,879 (115,038) 228,215 (21,329) 4.8645 5.4175 0.5531 412,637 (38,565) Alternative utility values 90% 1,002,664 (93,709) 1,230,879 (115,038) 228,215 (21,329) 3.9773 4.4296 0.4523 504,556 (47,156) Cost per day of maraviroc $165 1,002,664 (93,709) 1,222,727 (114,276) 220,063 (20,567) 4.4209 4.9236 0.5027 437,775 (40,914) Cost per day of maraviroc $220 1,002,664 (93,709) 1,249,899 (116,816) 247,235 (23,107) 4.4209 4.9236 0.5027 491,828 (45,966) Cost per day of OBT 50% 861,833 (80,547) 1,074,554 (100,428) 212,721 (19,881) 4.4209 4.9236 0.5027 423,168 (39,549) Cost per day of OBT 150% 1,143,495 (106,871) 1,387,204 (129,648) 243,708 (22,777) 4.4209 4.9236 0.5027 484,813 (45,311) Time to treatment failure in suppressed patients 5 years 960,608 (89,778) 1,140,189 (106,562) 179,582 (16,784) 4.2766 4.6903 0.4137 434,087 (40,570) Time to treatment failure in suppressed patients 100 years 1,019,064 (95,242) 1,270,010 (118,695) 250,946 (23,453) 4.4556 4.9809 0.5253 477,720 (44,648) Treatment failure 2 years after treatment initiation (constant rate) 884,212 (82,638) 980,963 (91,681) 96,751 (9,042) 4.0095 4.2544 0.2449 395,062 (36,922) Treatment failure 4 years after treatment initiation (constant rate) 926,520 (86,593) 1,064,553 (99,493) 138,033 (12,901) 4.1596 4.4848 0.3252 424,455 (39,670) Treatment failure 6 years after treatment initiation (constant rate) 961,792 (89,889) 1,135,785 (106,150) 173,993 (16,261) 4.2879 4.6855 0.3976 437,607 (40,899) Treatment failure 8 years after treatment initiation (constant rate) 992,051 (92,717) 1,194,902 (111,676) 202,851 (18,958) 4.3960 4.8493 0.4533 447,498 (41,823) Treatment failure 10 years after treatment initiation (constant rate) 1,017,106 (95,059) 1,245,767 (116,429) 228,661 (21,371) 4.4820 4.9798 0.4978 459,342 (42,930) Proportion CCR5 monotropic 80% 1,002,664 (93,709) 1,368,181 (127,870) 365,517 (34,161) 4.4209 5.2239 0.8031 455,161 (42,539) Proportion CCR5 monotropic 30% 1,002,664 (93,709) 1,142,202 (106,750) 139,538 (13,041) 4.4209 4.7162 0.2953 472,512 (44,161) Virologic suppression rate for OBT + MVC upper 95% CI 1,002,664 (93,709) 1,255,510 (117,340) 252,846 (23,631) 4.4209 4.9946 0.5738 440,675 (41,186) Virologic suppression rate for OBT + MVC lower 95% CI 1,002,664 (93,709) 1,206,750 (112,783) 204,086 (19,074) 4.4209 4.8536 0.4327 471,621 (44,078) All efficacy parameters for OBT + MVC and OBT lower 95% CI 943,047 (88,137) 1,150,036 (107,482) 206,988 (19,345) 4.1860 4.6513 0.4653 444,849 (41,576) All efficacy parameters for OBT + MVC and OBT upper 95% CI 1,080,511 (100,985) 1,334,101 (124,685) 253,590 (23,701) 4.7294 5.2718 0.5424 467,533 (43,696) Sensitivity/Specificity of tropism test 85%/94% 1,002,664 (93,709) 1,200,206 (112,171) 197,542 (18,462) 4.4209 4.8468 0.4259 463,821 (43,349) Late failure rates for OBT + MVC and OBT upper 95% CI 988,865 (92,419) 1,182,301 (110,498) 193,435 (18,078) 4.3778 4.8049 0.4271 452,904 (42,328) Late failure rates for OBT + MVC and OBT lower 95% CI 1,018,293 (95,170) 1,241,381 (116,020) 223,089 (20,850) 4.4702 4.9530 0.4828 462,073 (43,185) Lag time between failure and switch to OBT 2 months 1,001,205 (93,573) 1,205,132 (112,632) 203,927 (19,059) 4.4217 4.8723 0.4506 452,567 (42,297) Lag time between failure and switch to OBT 6 months 1,004,782 (93,907) 1,213,858 (113,447) 209,077 (19,540) 4.4220 4.8731 0.4511 463,482 (43,317) Viral load set point < 50 copies/mL 1,065,832 (99,613) 1,292,914 (120,836) 227,082 (21,223) 4.8426 5.3485 0.5059 448,867 (41,951) Viral load set point > 100,000 copies/mL 988,056 (92,344) 1,216,443 (113,689) 228,387 (21,345) 4.3139 4.8190 0.5051 452,162 (42,259)

*Discount rate was 5% for costs and outcomes per annum. MVC, maraviroc; OBT, optimized background therapy. 911 912 Contreras-Hernandez et al. resulted in prescription practices that were largely inconsistent dicted ICER under base-case assumptions of approximately with published guidelines, with patients receiving anywhere from $454,000 (US$42,400) per QALY compares favorably to ICERs 4 to 13 changes in their treatment regimens [5]. Due to these low reported for other ARTs in treatment-experienced patients [39]. levels of adherence to prescribing recommendations, it has been Although no official declarations have been made on Mexican speculated that the rate of development of multidrug resistant payers’ valuations of a QALY, the ICER obtained falls below HIV strains has increased in Mexico, though studies are required $50,000 per QALY gained, which has been suggested as a thresh- to confirm this hypothesis [5]. In order to prevent disease pro- old applicable in the neighboring United States [40]. Alterna- gression and death in these individuals, new treatments with tively, using the criterion for cost-effectiveness recommended by novel mechanisms of action are urgently needed. the World Health Organization’s Commission on Macroeconom- As an emerging country, Mexico spends only approximately ics and Health of less than three times the country’s gross domes- 6.6% of the gross domestic product on health services, resulting tic product per capita (i.e., 3 ¥ US$10,751 [41] =US$32,253 per in limited resources being available for drug funding. Since 2003, QALY gained), the ICER falls slightly above this threshold. Nev- all antiretroviral-indicated HIV patients were granted free access ertheless, in the absence of an official Mexican threshold, the to treatment [36], with funding being provided by the Mexican reality is that additional issues, such as the complexity and sever- government and health institutions. In this context, economic ity of the disease state, for example, may be considered in the assessments for new antiretroviral medications are important to Mexican decision-making process. allow Mexican decision-makers to optimize use of limited health Consistent with the findings, the benefits associ- care resources for an increasing HIV/AIDS population challenged ated with maraviroc treatment were augmented in the subgroup of by the emergence of drug-resistant viral strains after exposure to patients who had two or less active agents in their OBT regimen multiple treatment regimens. (i.e., OSS score Յ 2). Although the virological suppression rate The purpose of this study was to predict the long-term clinical associated with maraviroc for the OSS Յ 2 subgroup was similar impact and cost-effectiveness of maraviroc in treatment- to that of the full trial population, the suppression rate associated experienced adults with HIV/AIDS in Mexico. The analysis was with OBT was substantially lower in the OSS Յ 2 subgroup. conducted by adapting the previously published and validated Consequently, greater increases in mean life expectancy and ARAMIS model to the perspective of health care payers in quality-adjusted life expectancy of 0.77 years and 0.65 QALYs, Mexico to compare a strategy in which individuals are treated respectively, were predicted relative to treatment with OBT alone. with OBT, with or without the addition of maraviroc contingent The resultant ICER associated with this subgroup defined by OSS on viral tropism test results, versus a strategy of treating the score was considerably lower than for the base-case population at entire population with OBT alone (without testing for tropism $407,329 (US$38,069) per QALY gained for the contingent status) in treatment-experienced patients with HIV/AIDS. maraviroc strategy versus OBT alone, indicating that maraviroc The antiviral effect of maraviroc is the result of binding to the may provide the most value when used in this scenario. human receptor CCR5, preventing CCR5-tropic The choice of antiretrovirals continues to grow as new agents HIV-1 from entering cells. Because its mechanism of action are introduced. Because this analysis and the MOTIVATE trials differs from that of all existing classes of ART, viruses that are were completed, new agents introduced include , resistant to existing agents may remain susceptible to maraviroc. , , and , all of which are indicated in In the MOTIVATE studies, patients treated with maraviroc plus treatment-experienced patients. In this analysis, the modeled OBT experienced a significant reduction in viral load and greater population is prescribed OBT regimens similar to those used in increases in CD4+ cells at 48 weeks than those who received the MOTIVATE trials. Further, it relies on the results of the OBT alone, in spite of the fact that patients had extensive pre- TORO-1 and -2 trials [42,43] for efficacy assumptions for existing resistance to multiple antiretroviral agents [8,9]. reoptimized background therapy. These assumptions may no As with most clinical trials of antiretrovirals that are designed longer be as valid as they were at the time of analysis, because of with surrogate marker end points to meet regulatory require- the introduction of new treatment options. ments for accelerated approval, the 48-week follow-up period of Our analysis has reported results for two different scenarios: a the MOTIVATE trials was not designed to demonstrate differ- contingent maraviroc strategy versus treatment with OBT alone ences between treatments in clinical end points. However, eco- modeled on the MOTIVATE screening cohort, and a comparison nomic evaluations should employ a time horizon sufficiently of maraviroc + OBT versus OBT alone based on the MOTIVATE distant to capture all the clinical consequences and costs likely to trial population (i.e., CCR5 tropic individuals only). Although our result from the studied interventions. In HIV/AIDS, the relation- discussion has focused primarily on the contingent strategy ships between immunologic and virologic markers, treatment, involving the MOTIVATE screening cohort, the analysis based on and longer-term clinical outcomes have been well documented the MOTIVATE trial population is also of interest for Mexican through longitudinal cohort studies [15,18,37,38]. Making use decision-makers. Economic evaluations are valuable tools in the of this information, the ARAMIS model of HIV/AIDS disease decision-making process, especially in developing countries where progression was used to project important long-term benefits of resources are scarce and where cost containment policies are the maraviroc treatment associated with the proven short-term viral priority in government agendas. AIDS is a high prevalence disease load reductions and CD4+ cell increases. ARAMIS has been worldwide and Mexico is no exception. In this sense, results such validated against observational trials and results of the previ- as the ones presented in this research could help to allocate more ously published CEPAC model [10]. efficiently the limited budget available to Mexicans Institutions. By adapting ARAMIS with Mexico-specific natural mortality data and HIV-related resource use and costs, which differ mark- Conclusions edly from corresponding US data, the present study provides results that are applicable to the Mexican setting. Our analysis Based on the substantial virologic, immunologic, and projected predicted increases in AIDS-free survival time of approximately clinical improvements associated with maraviroc, a treatment 10 months (38% increase) and average life expectancy of 14 strategy including this unique therapy is valuable and, from a months (approximately 15.5% increase) for the contingent Mexican payer perspective, may be cost-effective in patients with maraviroc strategy relative to the OBT alone strategy. The pre- advanced HIV/AIDS and limited therapeutic options for active Cost-Effectiveness of Maraviroc for HIV in Mexico 913

ART, whose prognosis is otherwise poor. Screening for tropism 16 Lalezari JP, Henry K, O’Hearn M, et al. Enfuvirtide, an HIV-1 status prior to maraviroc utilization will help to ensure it is used fusion inhibitor, for drug-resistent HIV infection in North and appropriately and that health care dollars are invested in those South America. N Engl J Med 2003;348:2175–85. who are most likely to benefit from its efficacy. 17 Lazzarin A, Clotet B, Cooper D, et al. Efficacy of enfuvirtide in patients infected with drug-resistant HIV-1 in Europe and Aus- tralia. N Engl J Med 2003;348:2186–95. Source of financial support: This study was sponsored by Pfizer Inc. 18 Raffi F, Katlama C, Saag M, et al. Week-12 response to therapy as Editorial support and the development of the model was provided by i3 a predictor of week 24, 48, and 96 outcome in patients receiving Innovus and was funded by Pfizer Inc. The authors from i3 Innovus did the HIV fusion inhibitor enfuvirtide in the T-20 versus Optimized not receive individual payments from Pfizer for their involvement in the Regimen Only (TORO) trials. Clin Infect Dis 2006;42:870–7. model and article development. 19 Albrecht MA, Bosch RJ, Hammer SM, et al. Nelfinavir, , or both after the failure of nucleoside treatment of HIV infection. References N Engl J Med 2001;345:398–407. 20 Baxter JD, Mayers DL, Wentworth DN, et al. A randomized 1 United Nations Programme on HIV/AIDS and World Health study of antiretroviral management based on plasma genotypic Organization (UNAIDS/WHO). Epidemiological Fact Sheets on antiretroviral resistance testing in patients failing therapy. AIDS HIV/AIDS and Sexually Transmitted Infections. Mexico, 2006. 2000;14:F83–93. Available from: http://www.who.int/globalatlas/predefined 21 Gulick RM, Ribaudo HJ, Shikuma CM, et al. Three- vs four-drug Reports/EFS2006/EFS_PDFs/EFS2006_MX.pdf [Accessed May antiretroviral regimens for the initial treatment of HIV-1 infec- 5, 2008]. tion: a randomized controlled trial. JAMA 2006;296:769–81. 2 Centro Nacional para la Prevención y el Control del VIH/SIDA. 22 Hammer SM, Squires KE, Hughes MD, et al. A controlled trial of AIDS in numbers. Available from: http://www.salud.gob.mx/ two nucleoside analogues plus in persons with human conasida/ [Accessed August 12, 2008]. immunodeficiency virus infection and CD4 cell counts of 200 per 3 Bravo-Garcia E, Magis-Rodriquez C, Saavedra J. New estimates cubic millimeter or less. AIDS Clinical Trials Group 320 Study in Mexico: more than 180 000 people living with HIV. Abstract Team. N Engl J Med 1997;337:725–33. CDC0411. XVI International AIDS Conference; 2006 August 23 Deeks SG, Barbour JD, Grant RM, Martin JN. Duration and 13–18; Toronto, ON, Canada. predictors of CD4 T-cell gains in patients who continue combi- 4 World Health Organization (WHO). HIV prevalence among nation therapy despite detectable plasma viremia. AIDS 2002; adults aged Ն15 years (per 100 000 population). WHO Statistical 16:201–7. Information System page. Available from: http://www.who.int/ 24 US Department of Health and Human Services. U.S. Public Health whosis/indicators/2007HIVPrevAdults/en/index.html [Accessed Service and the Infectious Diseases Society of America. Guidelines May 5, 2008]. for the Prevention of Opportunistic Infections Among HIV- 5 Bautista-Arredondo S, Mane A, Bertozzi SM. Economic impact of Infected Persons—2002. June 14, 2002. Available from: http:// antiretroviral therapy prescription decisions in the context of www.aidsinfo.nih.gov/default.aspx [Accessed April 18, 2008]. rapid scaling-up of access to treatment: lessons from Mexico. 25 Panel on Antiretroviral Guidelines for Adults and Adolescents. AIDS 2006;20:101–9. Guidelines for the use of antiretroviral agents in HIV-1-infected 6 World Health Organization (WHO). Significant growth in access adults and adolescents. US Department of Health and Human to HIV treatment in 2006: More efforts needed for universal Services. 2008;1–128. Available from: http://www.aidsinfo. access to services. Media Centre page. 2007. Available from: nih.gov/ContentFiles/AdultandAdolescentGL.pdf [Accessed April http://www.who.int/mediacentre/news/releases/2007/pr16/en/ 18, 2008]. index.html [Accessed May 5, 2008]. 26 Schackman BR, Goldie SJ, Freedberg KA, et al. Comparison of 7 Gulick RM, Lalezari J, Goodrich J, et al. Maraviroc for previ- health state utilities using community and patient preference ously treated patients with R5 HIV-1 infection. N Engl J Med weights derived from a survey of patients with HIV/AIDS. Med 2008;359:1429–41. Decis Making 2002;22:27–38. 8 Hardy D, Reynes J, Konourina I, et al. Efficacy and safety of 27 Paltiel AD, Scharfstein JA, Seage GR, III, et al. A Monte Carlo maraviroc plus optimized background therapy in treatment- simulation of advanced HIV disease: application to prevention of experienced patients infected with CCR5-tropic HIV-1: 48-week CMV infection. Med Decis Making 1998;18(Suppl.):S93–105. combined analysis of the MOTIVATE studies. Abstract of poster 28 Zbrozek AS, Cantor SB, Cardenas MP, Hill DP, Jr. Pharmaco- presentation 792. 15th Conference on Retroviruses and Oppor- economic analysis of ondansetron versus metoclopramide for tunistic Infections (CROI); 2008; Boston, MA, USA. cisplatin-induced nausea and vomiting. Am J Hosp Pharm 1994; 9 Pfizer Inc. Selzentry (maraviroc). Prescribing information (US 51:1555–63. Label). 2007. 29 Consejo Nacional de Población. Mexico en Cifras: Indicadores 10 Kuhne FC, Chancellor J, Mollon P, et al. A microsimulation of Demográficos Básicos. Available from: http://www.conapo. the cost-effectiveness of maraviroc for antiretroviral treatment- gob.mx/00cifras/proy/ [Accessed May 5, 2008]. experienced HIV-infected individuals. HIV Clin Trials 2010; 30 Consejo de Salubridad General. Guía para la conducción de estu- 11:80–99. dios de evaluación económica para la actualización del Cuadro 11 Henry K, Tebas P, Lane C. Explaining, predicting, and treating Básico de Insumos del Sector Salud en México. Dirección General HIV-associated CD4 cell loss: after 25 years still a puzzle. JAMA Adjunta de Priorización. Comisión interinstitucional del Cuadro 2006;296:1523–5. Básico de Insumos del Sector Salud. 2008. Available from: http:// 12 Multicenter AIDS Cohort Study (MACS) Public Data Set: Release www.csg.salud.gob.mx/descargas/pdfs/cuadro_basico/GUxA_ P04. 1995. Springfield, VA: National Technical Information EVAL_ECON25082008-2_ech.pdf [Accessed June 20, 2010]. Service. 31 Diario Oficial de la Federación. Costos Unitarios de Atención 13 Cole SR, Hernan MA, Robins JM, et al. Effect of highly active Médica. Instituto Mexicano del Seguro Social (IMSS). Martes 9 antiretroviral therapy on time to acquired immunodeficiency syn- de Marzo de 2004. drome or death using marginal structural models. Am J Epidemiol 32 OANDA the Currency Site. FX history: historical currency 2003;158:687–94. exchange rates. Available from: http://www.oanda.com/convert/ 14 d’Arminio MA, Sabin CA, Phillips A, et al. The changing inci- fxhistory [Accessed July 18, 2008]. dence of AIDS events in patients receiving highly active antiret- 33 Instituto Mexicano del Seguro Social (IMSS). Portal de Transpar- roviral therapy. Arch Intern Med 2005;165:416–23. encia. Available from: http://transparencia.imss.gob.mx/html/ 15 Mellors JW, Munoz A, Giorgi JV, et al. Plasma viral load and bienes.htm [Accessed May 5, 2008]. CD4+ lymphocytes as prognostic markers of HIV-1 infection. 34 Tsevat J, Sherman SN, McElwee JA, et al. The will to live among Ann Intern Med 1997;126:946–54. HIV-infected patients. Ann Intern Med 1999;131:194–8. 914 Contreras-Hernandez et al.

35 Simpson KN, Luo MP, Chumney E, et al. Cost-effectiveness of 40 Grosse SD. Assessing cost-effectiveness in health care: history of / versus nelfinavir as the first-line highly active the $50,000 per QALY threshold. Expert Rev Pharmacoeconom- antiretroviral therapy regimen for HIV infection. HIV Clin Trials ics Outcomes Res 2008;8:165–78. 2004;5:294–304. 41 PNUD. Informe sobre desarrollo humano 2007–2008. La lucha 36 Cordova-Villalobos JA, Ponce de León-Rosales S, Valdespino JL, contra el cambio climático: Solidaridad frente a un mundo divi- authors, eds. 25 años de SIDA en México. Logros, desaciertos y dido. New York: Grupo Mundi-Prensa, 2007. retos. Primera edición. Cuernavaca, Mexico: Instituto Nacional 42 Nelson M, Arasteh K, Clotet B, et al. Durable efficacy of enfu- de Salud Pública, 2008. virtide over 48 weeks in heavily treatment-experienced HIV-1- 37 Cameron DW, Heath-Chiozzi M, Danner S, et al. Randomised infected patients in the T-20 versus optimized background placebo-controlled trial of ritonavir in advanced HIV-1 disease. regimen only 1 and 2 clinical trials. J Acquir Immune Defic Syndr The Advanced HIV Disease Ritonavir Study Group. Lancet 2005;1:404–12. 1998;351:543–9. 43 Raffi F. [Clinical efficacy and tolerance of enfuvirtide (Fuzeon), 38 Fiscus SA, Hughes MD, Lathey JL, et al. Changes in virologic new antiretroviral inhibitors of intracellular penetration of markers as predictors of CD4 cell decline and progression of human immunodeficiency virus (HIV) type 1]. Med Mal Infect disease in human immunodeficiency virus type 1-infected adults 2004;34:8–17. treated with nucleosides. AIDS Clinical Trials Group Protocol 44 Schackman BR, Gebo KA, Walensky RP, et al. The lifetime cost 174 Team. J Infect Dis 1998;177:525–633. of current human immunodeficiency virus care in the United 39 Sax PE, Losina E, Weinstein MC, et al. Cost-effectiveness of enfu- States. Med Care 2006;44:990–7. virtide in treatment-experienced patients with advanced HIV disease. J Acquir Immune Defic Syndr 2005;39:69–77.