An update to the cost-effectiveness of posaconazole vs fluconazole or itraconazole in the prevention of invasive fungal disease among neutropenic patients in the United States

Abstract

Objectives:

Posaconazole has shown superior clinical efficacy in the prevention of invasive fungal disease (IFD) among neutropenic patients as well as cost-effectiveness in the US healthcare setting vs fluconazole or itraconazole (FLU/ITRA) based on oral suspension formulations of each therapy. This study aims to provide an update on the cost-effectiveness of posaconazole in the current US healthcare setting to reflect bioequivalent tablet formulations of posaconazole and fluconazole, as well as changes in healthcare and drug costs.

Methods:

An existing model was used to assess the cost-effectiveness of posaconazole vs FLU/ITRA in the prevention of IFD among patients with acute myelogenous leukemia (AML) or myelodysplastic syndrome (MDS) and chemotherapy-induced neutropenia. Drug efficacy, mortality related to IFD, and death from other causes were estimated for tablet formulations using data from a randomized clinical trial of oral suspensions based on bioequivalence. IFD treatment costs were updated using the average inflation rate over 8 years (2006–2014) and drug costs were based on 2014 Analysource data.

Results:

Trial data show a lower IFD probability over 100 days of follow-up with posaconazole compared to standard azole therapy (0.05 vs 0.11). The treatment duration on posaconazole is 29 days compared to 24 days for FLU and 29 days for ITRA. The average cost of prophylaxis is higher in the posaconazole group compared to FLU/ITRA ($4673 vs $353); however, the costs associated with treating the IFD are lower in the posaconazole group compared to FLU/ITRA ($2205 vs $5303). The incremental cost effectiveness ratio of IFD avoided for posaconazole is $18,898 vs FLU/ITRA.

Conclusions:

In the current healthcare cost environment where both drug costs and overall IFD treatment costs have increased since 2007, posaconazole tablets are a cost-effective alternative to fluconazole or itraconazole in the prevention of IFD among neutropenic patients with AML and MDS in the US.

Keywords::

• Cost-effectiveness
• Decision analysis
• Invasive fungal disease
• Prophylaxis

Introduction

The incidence of invasive fungal disease (IFD) has increased over the past few decades, and one of the key risk factors for development of IFD is neutropenia1^,2. Given the high morbidity and mortality of IFD, guidelines recommend anti-fungal prophylaxis for high-risk patients, including neutropenic patients undergoing chemotherapy for acute myelogenous leukemia (AML) or myelodysplastic syndrome (MDS)3^,4. Patients with IFD also have greater overall healthcare costs. IFD patients stayed in the hospital twice as long and cost triple to treat compared to matched controls1. Thus, anti-fungal prophylaxis may reduce the costs attributable to IFD which totaled ∼$2 billion for hospitals in the US in 20041.

Posaconazole is an extended-spectrum triazole approved for use in the US and Europe for prophylaxis of IFD in immunocompromised patients5^,6. In a randomized, open-label, evaluator-blinded multi-center trial of patients with prolonged neutropenia resulting from chemotherapy for AML/MDS, posaconazole has been shown to be more effective than standard fluconazole or itraconazole (FLU/ITRA) prophylaxis in preventing IFD, while also improving overall survival7. Recent retrospective and prospective studies have supported the efficacy of posaconazole beyond the randomized controlled trial setting8–16. Several pharmacoeconomic analyses have modeled cost-effectiveness of prophylaxis with posaconazole in neutropenic AML/MDS patients. Most of these studies have found posaconazole to be dominant (cost-saving) over FLU and ITRA, while the other studies have found posaconazole to be cost-effective17–35.

A cost-effectiveness analysis of posaconzaole vs FLU/ITRA for the prevention of IFD among patients with AML or MDS at high-risk for IFD due to chemotherapy-induced neutropenia in the US was previously conducted based on 2006 healthcare costs and drug prices27. The acquisition cost of posaconazole is relatively high compared with that of FLU/ITRA and drug costs, along with the cost of healthcare for IFD treatment, have changed over time. Previous work had been based on oral suspension formulations of each drug, but the newly available tablet formulation of posaconazole offers advantages over the old formulation. The oral suspension is dosed at 200 mg 3-times a day with a full meal, while the tablet is bioequivalent at 300 mg once a day with food36. The mean half-life values are similar for each formulation with higher plasma drug exposure and less variability in exposure observed with the tablets compared to the oral suspension37^,38. An update of the previous cost-effectiveness model is presented to reflect these changes in cost and examine posaconazole tablets vs FLU/ITRA in IFD prevention in high-risk neutropenic AML/MDS patients in the current US healthcare setting.

Methods

Model structure

A decision analytic model was developed to estimate the cost-effectiveness of anti-fungal prophylaxis among AML/MDS patients with chemotherapy-induced neutropenia at high-risk of IFD. The initial model was developed by Schering-Plough Corp. (subsequently acquired by Merck Sharp & Dohme Corp., Whitehouse Station, NJ) and i3 Innovus (now OptumInsight, Waltham, MA) and detailed by O’Sullivan et al.27. Patients in the model are assumed to receive prophylaxis with oral suspension formulations of posaconazole or standard azole therapy (FLU or ITRA) as reported in the clinical trial of Cornely et al.7. Patients received either posaconazole (n = 304), FLU (n = 240), or ITRA (n = 58) with each cycle of chemotherapy until recovery from neutropenia and complete remission, until occurrence of an IFD, or for up to 12 weeks after randomization, whichever came first. The total follow-up period was 100 days after randomization. Clinical events in the first 100 days were modeled with chance nodes reflecting probabilities of IFD, IFD-related death, and other death. One-month Markov cycles were used to extend the decision tree to a lifetime horizon.

The structure of the model is presented in Figure 1. Based on the patient population of the clinical trial where the median age was 53 years and ∼53% of patients were male, patients enter the model as a cohort of men and women 50 years of age, with AML or MDS, who are at high-risk of IFD due to chemotherapy-induced neutropenia. The model begins with the choice of anti-fungal prophylaxis (posaconazole or FLU/ITRA). Patients who survive the 100-day follow-up period following initiation of prophylaxis enter the Markov sub-tree, in which their monthly risk of death is specific to their underlying condition (i.e., AML or MDS), but independent of whether or not they had an IFD. Outcomes for this analysis include the proportion of patients experiencing IFD events, life expectancy, costs of prophylaxis and IFD treatment, and incremental cost per IFD avoided and per life-year saved (LYS), over a lifetime horizon.

Figure 1. Structure of the model of prophylaxis for the prevention of IFD.

Model estimation

Clinical inputs were estimated from the clinical trial data and published sources, as presented in Table 1. The case-fatality rates were obtained from those observed from Cornely et al.7 as 0.14 for those treated with posconazole and 0.17 for those treated with FLU/ITRA. The probabilities of IFD-related death were also obtained from the clinical trial data as 0.36 for posaconazole and 0.48 for FLU/ITRA. Survival beyond the 100-day follow-up period was dependent upon the underlying condition. The National Cancer Institute’s Surveillance, Epidemiology, and End results (SEER) registry data was used to estimate the 5-year survival rate for AML patients (24.9%)39. The 5-year survival rate for MDS patients (17.6%) was estimated from the International Working Group for Prognosis in MDS (IWG-PM) project database using the weighted average of younger MDS patients at higher risk groups, as defined by the revised International Prognostic Scoring System (IPSS-R)40. The proportion of patients with AML (86%) and MDS (14%) was used as weighted coefficients for the relative survival rates.

Table 1. Model parameters.

Model parameter	Estimate	Data source
Efficacy of prophylaxis
Probability of IFD with fluconazole/itraconazole	0.11	Cornely et al. (2007)7
Probability of IFD with posaconazole	0.05	Cornely et al. (2007)7
Survival following prophylaxis
Probability of IFD-related death, given IFD; posaconazole	0.36	Cornely et al. (2007)7
Probability of IFD-related death, given IFD; fluconazole/itraconazole	0.48	Cornely et al. (2007)7
Probability of other death within 100 days; posaconazole	0.14	Cornely et al. (2007)7
Probability of other death within 100 days; fluconazole/itraconazole	0.17	Cornely et al. (2007)7
5-year relative survival for AML patients after 100 days	0.249	Howlander et al. (2014)39
5-year relative survival for MDS patients after 100 days	0.176	Greenberg et al. (2012)40
Costs (2014 USD)
Mean IFD treatment costs	$47,890	HCUP-NIS analysis from 2007 inflation-adjusted to 201427^,41
Drug costs (per day)
Posaconazole tablet	$161	Data from Merck&Co.,Inc.
Fluconazole tablet	$5.31	Analysource database, data on file, Merck&Co.,Inc., 2014
Itraconazole oral suspension	$45.30	Analysource database, data on file, Merck&Co.,Inc., 2014
Mean duration of prophylaxis (days)
Posaconazole	29	Cornely et al. (2007)7
Fluconazole	24	Cornely et al. (2007)7
Itraconazole	29	Cornely et al. (2007)7
Discount rate	3%	Gold et al. (1996)46

IFD, invasive fungal disease; AML, acute myelogenous leukemia; MDS, myelodysplastic syndromes.

The cost of treating an IFD in a US hospital was estimated from the 2004 Healthcare Cost and Utilization Project-Nationwide Inpatient Sample data (unpublished data from analysis of HCUP-NIS, Schering-Plough Corp., 2007) as calculated by O’Sullivan et al.27 for patients with blood cancers who experienced an IFD and for matched controls who did not experience an IFD. The matching was based on age, sex, hospital region, hospital teaching status, and high-risk conditions (i.e., cancer, HIV, chronic obstructive pulmonary disease, diabetes, and solid organ, hematopoietic stem cell, or bone marrow transplant). Candidiasis (40%), other mycoses (36%), aspergillosis (16%), and cryptococcosis (9%) were the most common types of infection among the cases of IFD. Charges were converted to costs using hospital-specific cost-to-charge ratios, and the difference in total costs between IFD cases and controls was assumed to represent the excess cost attributable to the IFD. This cost of treating IFD was updated and adjusted for inflation from 2007 to 2014 using US inflation data from the Bureau of Labor Statistics41.

Acquisition costs for posaconazole were obtained from Merck&Co., Inc., while those for FLU and ITRA were updated from 2006 with Analysource data based on average price of generics (data on file, Merck&Co., Inc., 2014). Previous work was based on costs for the oral suspension formulations of 400 mg generic FLU daily, 200 mg generic ITRA twice daily, and 600 mg posaconazole daily. A delayed-release tablet formulation of posaconazole has become available since the previous cost-effectiveness analysis by O’Sullivan et al.27 was conducted. The daily drug costs were updated for 300 mg of posaconazole tablets daily, which is equivalent to the 600 mg oral suspension dose37^,38. Fluconazole tablets have bioequivalence to the oral suspension42^,43; however, itraconazole oral suspension and capsules are not interchangeable44^,45. Thus, acquisition costs in the updated model were based on 300 mg of posaconazole tablets daily ($161 per day), 400 mg generic FLU tablets daily (average cost of $5.31 per day), and 200 mg generic ITRA oral suspension twice daily (average cost of $45.30 per day). Total drug costs were estimated based on the average duration of prophylaxis. Total drug costs for FLU and ITRA were also weighted by the proportions receiving each drug (81% FLU and 19% ITRA).

Analyses

Cumulative numbers of IFD over a 100-day period and life-years over the lifetime horizon were estimated for the posaconazole and FLU/ITRA strategies. Cost-effectiveness was assessed in terms of incremental cost per IFD avoided and incremental cost per LYS. All costs and health outcomes were discounted at 3% per year46. A sensitivity analysis for incremental cost per LYS was conducted using a probabilistic sensitivity analysis (PSA) based on 1000 samples. This was done to assess the effects of parameter uncertainty on the study findings, particularly with respect to treatment efficacy and the costs of prophylaxis and treating an IFD. The parameters and distributions of the variables included in the probabilistic analysis are summarized in Table 2. PSA results are presented in the form of a cost-effectiveness acceptability curve (CEAC), which shows the fraction of the 1000 simulations in which posaconazole prophylaxis was incrementally cost-effective over a range of incremental cost-effectiveness ratio (ICER) thresholds.

Table 2. Distributions and parameters used in the probabilistic sensitivity analysis.

Variable	Distribution	Parameters	Data source
Probability of IFD
Fluconazole/Itraconazole	Beta	a = 33, b = 265	Cornely et al. (2007)7
Posaconazole	Beta	a = 14, b = 290	Cornely et al. (2007)7
Probability of IFD-related death, given IFD
Fluconazole/Itraconazole	Beta	a = 21, b = 26	Cornely et al. (2007)7
Posaconazole	Beta	a = 21, b = 26	Cornely et al. (2007)7
Probability of other death within 100 days	Beta	a = 95, b = 507	Cornely et al. (2007)7
Mean duration of prophylaxis (days)
Posaconazole	Gamma	Mean = 29, SE = 1.20	Cornely et al. (2007)7
Fluconazole	Gamma	Mean = 24, SE = 1.03	Cornely et al. (2007)7
Itraconazole	Gamma	Mean = 29, SE = 2.80	Cornely et al. (2007)7
IFD treatment costs	Gamma	Mean = $47,890, SE = $3636	HCUP-NIS analysis from 2007 inflation-adjusted to 201427^,41

IFD, invasive fungal disease; HCUP-NIS, Healthcare Cost and Utilization Project-Nationwide Inpatient Sample; SE, standard error.

Alternative scenario

Treatment-specific estimates of IFD related fatality as observed by Cornely et al.7 are used in the base-case scenario. The alternative scenario considers a pooled case-fatality rate of 0.16 for both posaconazole and FLU/ITRA. Similarly, the probability of IFD-related death in the base-case scenario is as observed from the clinical trial, but is assumed equal at 0.45 for both treatment groups in the alternative scenario.

Results

The costs and effects of posaconazole vs FLU/ITRA in terms of IFD avoided and LYS are presented in Table 3. Posaconazole was found to be associated with fewer IFD per patient compared to FLU/ITRA (0.05 vs 0.11) in the clinical trial of anti-fungal prophylaxis for the prevention of IFD among high-risk neutropenic AML/MDS patients7. The average duration of treatment with posaconazole was 29 days compared to 24 days for FLU and 29 days for ITRA. The total costs of prophylaxis with standard azole therapy with FLU/ITRA and posaconazole are $5656 and $6879, respectively. Although the average drug acquisition cost of prophylaxis was higher for those treated with posaconazole ($4673) compared to those treated with FLU/ITRA ($353), the costs associated with IFD treatment were lower for the posaconazole group ($2205 vs $5303). The incremental cost-effectiveness ratio (ICER) for posaconazole vs FLU/ITRA was estimated to be $18,898 per IFD avoided and $6638 per life-year saved. Although posaconazole is more costly ($1222), it is also more effective than standard azole therapy.

Table 3. Cost-effectiveness of posaconazole versus fluconazole or itraconazole in the prevention of IFDs among high-risk neutropenic patients in the United States.

Analysis of 100-day clinical trial period (per patient results)
Treatment arm	Drug costs	IFD treatment costs	Total costs	Incremental costs	IFD events	IFD events avoided	ICER (IFD avoided)	Cost effective?
FLU/ITRA	$353	$5303	$5656	–	0.111	–	–	–
POS	$4673	$2205	$6879	$1222	0.046	0.065	$18,898	Yes
Analysis of lifetime model (per patient results)
Treatment arm	‘Tree’ costs	AML treatment costs	Total costs	Incremental costs	Life years	Life years saved	ICER (LYS)	Cost effective?
FLU/ITRA	$5,656	$0	$5656	–	2.68	–	–	–
POS	$6879	$0	$6879	$1222	2.86	0.184	$6,638	Yes

Sensitivity analyses

The probabilistic sensitivity analysis (PSA) included the probability of developing an IFD, the probability of death conditioned on having an IFD, the probability of death not related to IFD, and the duration of treatment. Each of these parameters is included in the PSA separately for the FLU/ITRA and posaconazole groups. The resulting cost-effectiveness acceptability curve (CEAC) is presented in Figure 2. Posaconazole was dominant (cost saving) in 12.2% of the scenarios and dominated by FLU/ITRA in 0.1% of the scenarios. At a threshold of $50,000 per quality-adjusted life-year gained, there was a 93.7% probability that posaconazole would be cost-effective.

Figure 2. Probability that posaconazole is cost-effective vs fluconazole or itraconazole for the prevention of IFD among high-risk neutropenic patients.

Alternative scenario

The use of pooled estimates is a more conservative approach. It is based on the assumption that absence of a statistically significant difference in probabilities of IFD and non-IFD related deaths indicates equality. Using pooled estimates for survival following prophylaxis treatment, the ICER per life-year saved for posaconazole vs FLU/ITRA was estimated to be $15,500. PSA results showed that posaconazole was dominant in 13.9% and dominated by FLU/ITRA in 1.5% of the scenarios. At a threshold of $50,000 per quality-adjusted life-year gained, there was an 85.4% probability that posaconazole would be cost-effective.

Discussion

The present study aimed to evaluate the costs and outcomes of IFD prophylaxis treatment with posaconazole tablets compared to standard azole therapy with fluconazole or itraconazole in high-risk neutropenic patients undergoing chemotherapy for acute myelogenous leukemia or myelodysplastic syndrome in the current US healthcare setting. A previous study had found that use of posaconazole was cost-saving compared to FLU/ITRA in preventing IFD27. After updating the costs for drug acquisition and for treatment of IFD, as well as updating the 5-year survival rates for MDS and AML, this analysis has found that posaconazole, in a tablet formulation, remains cost-effective compared to FLU/ITRA. These results are consistent with previously published cost-effectiveness analyses where posaconazole was found to be either cost-saving or cost-effective when compared to standard azole therapies17–35.

In 2006, the acquisition cost for oral suspension formulations of posaconazole was $72 per day, while the cost for fluconazole was $2.74 per day and the cost for itraconazole was $1.57 per day27. Although the original clinical trial by Cornely et al.7 used all oral suspension formulations, tablet prices were used in the present study where bioequivalence has been demonstrated. This was done in order to reflect current practice where convenience may favor the use of tablets in the US. These changes in the current US healthcare setting warrant an update to the cost-effectiveness analysis. A new formulation of posaconazole as a delayed-release tablet was introduced in 2014 at a cost of $161 per day. Based on 2014 Analysource data, generic tablet formulations of fluconazole had an average price of $5.31 per day, generic oral suspension formulations of itraconazole had an average price of $45.30 per day, and generic capsule formulation of itraconazole had an average price of $34.29 per day. If itraconazole capsule and oral suspension formulations were interchangeable, the results of the model would be similar (ICER: $19,836 per IFD avoided and $6967 per LYS).

These cost-effectiveness analyses for the updated price of posaconazole show that, at $161 per day, posaconazole was still a cost-effective alternative in the prevention of IFD compared to FLU/ITRA. Treatment costs for IFD have increased since 2006 from $40,583 to $47,890, although there may be uncertainty and under-estimation in the updated costs. Costs for IFD treatment may be even greater considering the high costs associated with prolonged hospital stays and stays in the intensive care unit, along with the recent increase in the use of more expensive anti-fungal treatments such as echinocandins47.

The original cost-effectiveness model by O’Sullivan et al.27 noted some limitations which include the assumption that risk of death after 100 days was equal for those who did and did not develop IFD and use of equal rates of IFD-related death for both prophylaxis groups. This study has used the observed point estimates from the clinical trial and considered the equality assumption in an alternative scenario. However, the present study is still limited by the inability to detect differences between posaconazole and fluconazole or itraconazole separately due to insufficient power.

The present study examines the use of a newly introduced formulation of posaconazole, which is limited in that this formulation was not directly compared with standard azole therapy. The oral suspension formulation of posaconazole is taken with a full meal to maximize systemic absorption, but the solid tablet formulation has improved bioavailability, less variability, and minimal food effects on its pharmacokinetics37^,38. Patients at risk for low absorption of posaconazole oral suspension due to food intolerance may benefit from treatment with the tablet formulation.

Posaconazole is more costly but also more effective in terms of IFD events avoided and life years saved, making it cost-effective in the prevention of invasive fungal disease among neutropenic patients compared to fluconazole or itraconazole. Given the need to take posaconazole oral suspension with a full meal to improve bioavailability, the new tablet formulation of posaconazole may also be viewed as a cost-effective option in patients who cannot tolerate food and who were previously unable to undergo prophylaxis with posaconazole.

Conclusion

Posaconazole tablets are cost-effective compared to fluconazole or itraconazole in the prevention of invasive fungal disease among neutropenic patients with acute myelogenous leukemia or myelodysplastic syndrome in the current US healthcare setting.

Transparency

Declaration of funding

This study was funded by Merck.

Declaration of financial/other relationships

All authors are employees of Merck.