Original Research

Real-World Analysis of Dispensed International Units of Coagulation Factor VIII and Resultant Expenditures for Hemophilia A Patients: A Comparison Between Standard Half-Life and Extended Half-Life Products


Amit Chhabra, MBBS, MPH
Pfizer Inc., New York, N.Y.
Patrick F. Fogarty, MD
fizer Inc. Medical Affairs, Collegeville, Pa.
Bartholomew J. Tortella, MD, MBA
fizer Inc. Medical Affairs, Collegeville, Pa.
Dean Spurden, PhD
Pfizer Ltd., Patient and Health Impact, Tadworth, Surrey, U.K.
José Alvir, DrPH
Pfizer Inc. Statistical Research and Data Science Center, Global Product Development, New York, N.Y.
Margaret McDonald, PhD
Pfizer Inc. Patient and Health Impact, New York, N.Y.
Jennifer Hodge, PhD
Pfizer Inc., New York, N.Y.
Andreas M. Pleil, PhD
Pfizer Inc. Patient and Health Impact, San Diego, Calif.

Abstract

Purpose: To identify international units (IUs) dispensed and consequent expenditures for standard half-life (SHL) versus extended half-life (EHL) recombinant factor VIII (rFVIII) replacement products in hemophilia A patients in a real-world setting.

Design: Two U.S. claims databases were analyzed.

Methodology: Number of IUs dispensed and quarterly expenditures for rFVIII products were collected from the Optum Clinformatics Data Mart and Truven Health MarketScan Databases. Truven claims were also analyzed for factor IUs dispensed and expenditures for patients with data for ≥3 months before and after switching to an EHL product.

Results: The Optum and Truven databases, respectively, included 276 (SHL, n=243; EHL, n=33) and 500 (SHL, n=409; EHL, n=91) hemophilia A patients. Median quarterly factor IUs dispensed in Optum were 10% higher with EHL versus SHL products over nine quarters, and 45% higher with EHL versus SHL products in Truven over 10 quarters. Median quarterly expenditures in the EHL cohort were 51% (individual quarterly medians range, 1%–101%) higher than in the SHL cohort in Optum and 122% higher (individual quarterly medians range, 1%–189%) in Truven. Twenty-nine Truven patients switched to an EHL product; median factor IUs dispensed varied quarterly. The lowest SHL and highest EHL values occurred in the quarter immediately before switching and the first quarter post-switch, respectively. Overall median quarterly expenditures were higher post-switch; this was consistent over seven quarters.

Conclusion: We found higher expenditures over two years for hemophilia A patients using EHL versus SHL products. Switching to an EHL rFVIII product was associated with variable factor IUs dispensed and consistently higher expenditures.

Keywords: blood coagulation factor VIII, cost analysis, drug prescriptions, drug switching

Introduction

Hemophilia A, an X-linked congenital bleeding disorder, is caused by a deficiency of coagulation factor VIII (FVIII) that results from a mutation of the gene for FVIII (Srivastava 2013). Representing 80% to 85% of the total hemophilia population, hemophilia A occurs in approximately one in 5,000 live male births (Franchini 2013, Srivastava 2013). The severity of hemophilia A is determined by the level of FVIII coagulant activity; approximately 60% of hemophilia A patients have severe disease, with plasma FVIII activity <1% (Castro 2014, George 2015). The majority of bleeding events (70%–80%) occur in the joints and muscles, resulting in hemar­throsis and progressive arth­ropathy that can lead to chronic pain and disability (Castro 2014, George 2015, Srivastava 2013). The primary goals of care are the treatment and prevention of bleeding events with replacement FVIII products and managing complications.

In the early 1980s, plasma-derived factor concentrates were associated with the transmission of hepatitis B and C viruses and human immunodeficiency virus (Castro 2014, Franchini 2013). The rapid progress in DNA technology and the ability to sequence the FVIII gene allowed for production of recombinant-based clotting factor concentrate replacement products. This development advanced the treatment of hemophilia beyond the viral inactivation techniques used to improve the safety of plasma-derived products. Continued improvements in manufacturing processes for rFVIII products have reduced or eliminated animal or human components that were used in the cell culture medium or to stabilize the final formulation (Franchini 2013, George 2015).

Traditional rFVIII products have half-lives of approximately 12 hours, leading to recommended prophylactic dosing three to four times weekly (Advate 2016, George 2015, Kogenate 2016, Miners 2016, Xyntha 2015). Strategies to extend the half-lives of rFVIII products include covalently linking the human FVIII to the human immunoglobulin G1 Fc domain or conjugating with polyethylene glycol (PEG) (Adynovate 2017, Eloctate 2017, Miners 2016). The resulting extended half-life (EHL) products (t½=14–20 hours) are dosed less frequently at twice weekly or every four days. Pricing of EHL products sought parity with standard half-life (SHL) products by balancing the lower expected need for factor international units (IUs) dispensed with a higher per-unit price, assuming similar efficacy based on annualized bleeding rates demonstrated in clinical studies (Konkle 2015, Mahlangu 2014, McMullen 2017, Miners 2016, Shapiro 2014).

This study aims to assess the real-world utilization of SHL and EHL rFVIII products on the population and patient level. At the population level, this analysis provides data regarding factor IUs dispensed and expenditures for SHL versus EHL rFVIII products based on specialty pharmacy claims (cross-sectional analysis). At the patient level, it examines patients who switched from an SHL to EHL product and were followed in an administrative claims database at the individual patient level (switch analysis).

Methods

Study design

This retrospective cohort study included males diagnosed with hemophilia A using SHL or EHL rFVIII replacement products. De-identified claims data were obtained from two sources to assess factor IUs dispensed and expenditures for overall rFVIII replacement products (cross-sectional analysis) and for those who switched from SHL to EHL products (switch analysis). Data were collected for individuals using SHL products, including rurioctocog alfa, octocog alfa, moroctocog alfa, turoctocog alfa, and simoctocog alfa, and for those using EHL products, including rurioctocog alfa pegol and efmoroctocog alfa.

Cross-sectional analysis
Optum Clinformatics Data Mart

A proprietary research database, Clinformatics Data Mart, contains pharmacy and medical claims and enrollment data for approximately 66.5 million individuals. The information is geographically diverse, with claims from across the United States. Data were collected for quarterly factor IUs dispensed and related expenditures for patients with hemophilia A (diagnosis code ICD-9 286.0 or ICD-10 D66) who used FVIII replacement products between Aug. 1, 2014, and Dec. 31, 2016. The variables available in the Optum database included baseline characteristics (e.g., age, race/ethnicity, and geographic region), the factor product and units dispensed, and direct pharmacy expenditures. The analysis of data from this database reports differences between groups and does not provide data on patients who switched from one therapy to another.

Truven Health MarketScan Research Databases

Clients of MarketScan Research Databases include U.S. employers, health plans, and hospitals, and the database contains detailed information about expenditures, factor IUs dispensed, and outcomes for both inpatient and outpatient settings. Coverage for approximately 100 million members is provided under a variety of fee-for-service, fully capitated, and partially capitated health plans, including preferred provider organizations, point-of-service plans, indemnity plans, and health maintenance organizations. Data were collected to identify quarterly factor IUs dispensed and related expenditures for all patients with hemophilia A (diagnosis code: ICD-9 286.0 or ICD-10 D66) who used rFVIII replacement between Aug. 1, 2014, and Jan. 31, 2017.

Switch analysis
Truven Health MarketScan Research Database

For analysis of patients who switched from SHL to EHL rFVIII replacement products, claims were identified from the Truven Health MarketScan Research Database based on the following criteria: time period from June 1, 2014, the month during which the first EHL FVIII replacement product was approved in the U.S., to Jan. 31, 2017, for patients who had a diagnosis code for hemophilia A, continuous pharmacy enrollment, at least one prescription for an EHL rFVIII product, a prior prescription for an SHL rFVIII product, claims data for at least three months before and three months after switching, no use of an rFVIII product other than the index SHL or EHL products, and were not enrolled in a capitated insurance plan. These within-switcher analyses, while restricted to a small number of patients, mitigate indication bias owing to unobserved factors that could drive product selection. Further, restrictions to individuals who had complete data for two and three quarters pre- and post-switch allowed us to make comparisons in groups who had the same amount of follow-up.

Assessments

Analyses were conducted separately for each database. The key outcome measures were factor IUs dispensed and direct expenditures for the rFVIII products. Outcomes were measured over quarterly (i.e., three months/90 days) periods; incomplete quarters (those with patients who did not have 90 days of follow-up) were not included. Overall data for all available quarters comprised the sum of available-quarter data divided by the number of quarters with available data. For the cross-sectional analysis, up to nine quarters were assessed for the Optum database, and up to 10 quarters were assessed for the Truven database.

For the switch analysis, the first switch to an EHL product occurred in September 2014, as defined by the first month that an EHL claim was paid. Data for up to nine quarters pre-switch were assessed. The switch analysis was limited to the Truven data­base because of the fewer number of patients in the Optum database.

For the Optum data, standard pricing algorithms were applied to the claims data to account for differences in pricing across health plans and provider contracts. These algorithms were designed to create standard costs that reflect allowed payments across all provider services. If a data cut contained more than one value for standard costs, the costs were adjusted to account for the difference. Costs were normalized to 2014 dollars.

For the Truven data, expenditures were calculated using the Truven variable that measures gross payments to a provider for a service in which payment equals the amount eligible for payment under the medical plan terms, after applying rules such as discounts but before applying coordination of benefits, copayments, and deductibles.

All data were summarized descriptively. Median values (and their corresponding interquartile ranges [IQRs]) were used to characterize the distribution of factor IUs dispensed and related expenditures, as determined by the product class indicated on the insurance claim. These values are more robust with small data sets because of the presence of outliers and skewed data distribution that are common in data for factor IUs dispensed and expenditures.

In keeping with the descriptive nature of these analyses, we did not perform statistical significance tests, opting for the numbers and the real-world graphs to speak for themselves. Means, standard deviations (SD), minimums, and maximums were calculated. Although we present median values in the text to avoid skewing, which can occur with high numeric outliers, we present both mean and median data in the tables. Analyses were completed using SAS Version 9.4 on Red Hat Enterprise Linux.

Results

Cross-sectional analysis of two databases

Quarterly factor IUs dispensed and expenditures were analyzed for a total of 276 and 500 patients identified as having hemophilia A in the Optum and Truven databases, respectively. In the Optum database, 243 (88%) patients had an SHL product dispensed and 33 (12%) had an EHL product dispensed. In the Truven database, 409 patients (82%) received an SHL product and 91 (18%) received an EHL product. Some patients received both an SHL and EHL product during the study period. The majority of patients were younger than 35 years of age and received rurioctocog alfa (SHL) or efmoroctocog alfa (EHL) (Table 1).

Table 1
Demographics and characteristics at index datea based on data from the Optum Clinformatics Data Mart and Truven Health MarketScan Databases
Characteristic Optum database Truven databaseb
SHL
(n=243)
EHL
(n=33)
SHL
(n=409)
EHL
(n=91)
Age, mean (SD) 25.0 (17.4) 24.3 (17.5) 20.5 (14.6) 24.1 (13.6)
Age group, n (%)
0–17 99 (41) 17 (52) 192 (47) 33 (36)
18–34 85 (35) 7 (21) 144 (35) 39 (43)
35–44 23 (9) 6 (18) 39 (10) 11 (12)
45–54 15 (6) 0 22 (5) 5 (5)
55–64 13 (5) 1 (3) 11 (3) 3 (3)
≥65 8 (3) 2 (6) 1 (<1) 0
Race/ethnicity, n (%)
White 137 (56) 12 (36) n/a n/a
Hispanic 33 (14) 4 (12) n/a n/a
Black 17 (7) 4 (12) n/a n/a
Asian 10 (4) 3 (9) n/a n/a
Missing 37 (15) 10 (30) n/a n/a
Unknown 9 (4) 0 n/a n/a
Region of U.S., n (%)
Northeast 17 (7) 6 (18) 76 (19) 17 (19)
North Central 53 (22) 10 (30) 80 (20) 26 (29)
South 115 (47) 13 (39) 179 (44) 40 (44)
West 56 (23) 4 (12) 68 (17) 8 (9)
Unknown 2 (<1) 0 6 (1) 0
rFVIII replacement product
SHL, n (%)
Rurioctocog alfa 173 (71) 272 (67)
Octocog alfa 40 (16) 82 (20)
Moroctocog alfa 29 (12) 49 (10)
Simoctocog alfa 1 (<1) 1 (<1)
Turoctocog alfa 0 5 (1)
EHL, n (%)
Efmoroctocog alfa 30 (91) 68 (75)
Rurioctocog alfa pegol 3 (9) 23 (25)
aDemographic information represents values during the first quarter available for each patient, based on the factor VIII product they were using at that time.
bPatients who switched between drug categories during the first quarter were not excluded.
EHL=extended half-life, n/a=not available, SD=standard deviation, SHL=standard half-life
Factor IUs dispensed

The median factor IUs dispensed for all patients per quarter in the Optum database ranged from 61,920 to 75,108 IU for SHL products, and from 50,315 to 96,864 IU for EHL products (Supplemental Table 1). Median factor IUs dispensed for all patients over the study period was 10% higher in the EHL cohort (by 6,332 IU; Figure 1; Table 2). However, differences by quarter were variable, ranging from 31% lower to 36% higher compared with the SHL cohort (median quarterly use difference, –22,722 to 25,294 IUs) (Supplemental Table 1).

Figure 1

Median and mean quarterly factor international units (IUs) dispensed and expenditures for standard half-life and extended half-life recombinant factor VIII replacement products: Optum Clinformatics Data Mart Database and Truven Health MarketScan Databases

Table 2
Cross-sectional analysis of quarterly IUs of factor dispensed and expenditures of recombinant factor VIII replacement products: Overall data from the Optum Clinformatics Data Mart and Truven Health MarketScan Databasesa
Parameter Optum database (n=276) Truven database (n=500)b
IUs dispensed Expenditures IUs dispensed Expenditures
EHL cohort, n 33 89
Median 67,662 $113,876 67,375 $135,519
IQR 44,480–99,035 $74,860–$195,760 $50,524–$98,264 $100,320–$186,557
Mean (SD) 73,696 (40.875) $126,725 ($71,037) $72,746 ($35,254) $148,238 ($70,135)
SHL cohort, n 243 415
Median 61,330 $75,298 46,409 $61,152
IQR 30,000–109,275 $37,113–$133,545 12,760–87,670 $18,593–$115,845
Mean (SD) 75,255 (61,847) $92,337 ($75,242) 60,198 (60,846) $79,462 ($77,863)
Median difference, EHL vs. SHL (%) ↑$6,332 (10) ↑$38,578 (51) ↑$20,966 (45) ↑$74,367 (122)
aOverall for all available quarters is the sum of available quarter data divided by the number of quarters with available data. For the Optum data, to account for differences in pricing across health plans and provider contracts, standard pricing algorithms were applied to the claims data. These algorithms were designed to create standard prices that reflect allowed payments across all provider services. For the Truven data, expenditures were calculated using the Truven variable that measures gross payments to a provider for a service in which payment equals the amount eligible for payment under the medical plan terms, after applying rules such as discounts but before applying coordination of benefits, copayments, and deductibles.
bSome individuals received an EHL and an SHL factor VIII replacement product during the study period and are included in both cohorts.
EHL=extended half-life, IQR=interquartile range, IU=international unit, SD=standard deviation, SHL=standard half-life

In the Truven database, the median quarterly factor IUs dispensed for all patients ranged from 37,284 to 83,592 IU for SHL products and from 59,024 to 83,080 IU for EHL products (Supplemental Table 1). The median factor IUs dispensed for all patients over the study period were 45% higher in the EHL cohort (by 20,966 IUs; Figure 1; Table 2). Substantial variability in quarterly usage differences was also observed in the Truven database, ranging from 25% lower to 110% higher compared with the SHL cohort (median quarterly use difference, –21,308 to 41,114 IUs) (Supplemental Table 1).

Expenditures

The overall median factor expenditures per patient per quarter over the Optum study period were 51% higher (by $38,578) in the EHL cohort ($113,876) compared with the SHL cohort ($75,298; Figure 1; Table 2). The median quarterly expenditures for all patients ranged from $77,729 to $93,209 in the SHL cohort and from $84,860 to $163,022 in the EHL cohort (Supplemental Table 2). Individual quarterly median differences in expenditures for all patients ranged from $1,195 (1% difference between the SHL and EHL cohorts) to $79,878 (101% difference between the SHL and EHL cohorts) per quarter.

In the Truven database, the overall median factor expenditures per patient per quarter over the study period were 122% higher (by $74,367) in the EHL cohort ($135,519) compared with the SHL cohort ($61,152; Figure 1; Table 2). The median quarterly expenditures for all patients ranged from $51,338 to $106,741 in the SHL cohort, and from $107,429 to $162,694 in the EHL cohort (Supplemental Table 2). Individual quarterly median differences in expenditures for all patients ranged from $688 (1% difference between the SHL and EHL cohorts) to a high of $106,314 (189% difference between the SHL and EHL cohorts) per quarter.

Similar results were observed in both the Optum and Truven analyses of mean factor expenditures (Figure 1; Table 2; Supplemental Table 2).

Switch analysis

Overall, 29 patients who switched from an SHL product to an EHL product were identified from the Truven Database. Mean (SD) age in the immediate pre- and post-switch periods was 21.0 (12.5) years (range, <1–52). The majority of patients were from the southern (41%) and north central (41%) regions of the United States. Most patients had received the SHL rurioctocog alfa (83%) prior to the switch to an EHL drug; 55% of patients switched to efmoroctocog alfa and 45% switched to rurioctocog alfa pegol. Of these 29 patients, 17 had complete data for two quarters pre- and post-switch, and 12 had complete data for three quarters pre- and post-switch.

Factor IUs dispensed

Overall, median quarterly factor IUs dispensed increased from pre-switch to post-switch by 9,438 IUs and mean quarterly factor IUs dispensed increased by 18,053. Median quarterly factor IUs dispensed for an SHL product declined to their lowest value in the quarter immediately preceding the switch to an EHL product (Figure 2, Table 3). Among the 17 individuals with complete data for two quarters pre- and post-switch, the median quarterly factor IUs dispensed over this period increased from 57,308 IUs to 82,941 IUs. Among the 12 individuals with complete data for three quarters pre- and post-switch, factor IUs dispensed increased from 52,691 IUs to 73,719 IUs.

Figure 2

Median IUs dispensed and expenditures associated with switching from a standard half-life to an extended half-life recombinant factor VIII replacement product (Truven Health MarketScan Databases), by quarterly intervals pre- and post-switch

Box=25th and 75th percentiles, crossbar=median.

Table 3
Switch analysis of factor concentrate IUs dispensed and expenditures by quarter pre- and post-switch from a standard half-life to an extended half-life product (Truven Health MarketScan Databases)
PRE-SWITCH Q1
(n=29)
Q2
(n=26)
Q3
(n=25)
Q4
(n=24)
Q5
(n=23)
Q6
(n=20)
Q7
(n=15)
Q8
(n=13)
Q9
(n=12)
IUs dispensed
Mean 45,228 58,687 71,284 56,846 62,626 59,564 64,692 58,860 54,844
SD 30,913 34.244 40,017 31,054 32,403 40,339 34,726 28,915 24,405
Lower quartile
(25th percentile)
23,745 31,980 41,160 31,523 40,280 38,488 35,240 43,422 43,171
Median 40,764 57,404 64,515 54,756 57,960 51,304 67,760 61,230 60,374
Upper quartile
(75th percentile)
60,975 80,496 97,140 77,325 86,940 74,052 95,310 83,430 73,218
Expenditures
Mean $61,192 $79,264 $100,543 $80,920 $89,972 $81,646 $88,149 $78,777 $73,694
SD $39,014 $42,670 $53,128 $41,450 $47,341 $49,816 $45,240 $29,272 $31,308
Lower quartile
(25th percentile)
$30,683 $40,743 $60,861 $42,530 $56,569 $53,869 $49,461 $51,731 $46,396
Median $57,342 $77,337 $84,116 $73,576 $76,700 $65,831 $79,908 $84,150 $77,418
Upper quartile
(75th percentile)
$78,433 $113,743 $130,723 $112,753 $112,034 $93,669 $116,600 $103,683 $91,097
POST-SWITCH Q1 (n=29) Q2 (n=20) Q3 (n=16) Q4 (n=10) Q5 (n=8) Q6 (n=5) Q7 (n=4) Q8 (n=1) Q9 (n=1)
IUs dispensed
Mean 80,421 70,551 68,799 68,124 49,176 59,197 48,049 17,188 12,944
SD 47,612 37,373 51,617 40,044 30,063 31,391 25,947 - -
Lower quartile
(25th percentile)
45,767 45,733 32,326 44,792 17,656 46,851 32,310 17,188 12,944
Median 73,464 72,668 57,354 57,547 57,789 72,808 55,696 17,188 12,944
Upper quartile
(75th percentile)
115,712 91,895 86,804 82,971 72,708 80,199 63,788 17,188 12,944
Expenditures
Mean $171,562 $139,297 $141,714 $129,943 $94,592 $118,266 $95,892 $28,364 $21,222
SD $88,871 $68,183 $95,815 $73,415 $60,133 $64,685 $54,997 - -
Lower quartile
(25th percentile)
$99,450 $90,593 $75,421 $83,123 $33,006 $93,107 $61,147 $28,364 $21,222
Median $169,717 $137,379 $124,732 $103,478 $106,034 $151,062 $110,689 $28,364 $21,222
Upper quartile
(75th percentile)
$215,633 $176,022 $175,889 $164,164 $145,858 $155,749 $130,638 $28,364 $21,222
IU=international unit, SD=standard deviation
Expenditures

The median overall quarterly expenditures for rFVIII replacement product were higher ($64,548) after the switch from an SHL to an EHL product. Higher median and mean expenditures were associated with the switch to an EHL product compared with expenditures before the switch for each corresponding quarter with the exception of Quarters 8 and 9, for which data from only one patient was evaluated post-switch (Figure 2, Table 3). As expected, clotting factor concentrate expenditures accounted for the majority of total health care expenditures both before (89%) and after the switch (90%). Median quarterly expenditures increased from $81,178 to $170,428 among the 17 individuals with two complete quarters of data, and increased from $74,477 to $143,051 among the 12 patients with three complete quarters of data.

Discussion

Hemophilia is a rare chronic disease that places a substantial economic burden on patients, their families, and the health care system (Aledort 2000, Zhou 2015). The advances in coagulation factor development and manufacturing have contributed to more expensive products, and the newer generation of recombinant products constitutes the majority of expenses for patients with hemophilia (Chen 2016). In 2015, data from the Hemophilia Utilization Group Study part Va were examined and showed that mean annual direct medical costs for all hemophilia A patients without inhibitors were $185,256 (Zhou 2015). Clotting factors accounted for 92% of direct medical costs.

Prophylactic therapy is recommended by the National Hemophilia Foundation for patients with severe hemophilia because it reduces cumulative arthropathic joint damage and health care utilization (MASAC 2016, Zhou 2015). When patients with severe hemophilia were cate­gorized by prophylactic and other kinds of treatment, the mean annual direct medical costs were substantially higher for those receiving prophylactic treatment ($292,525) versus episodic treatment ($184,518). This finding is consistent with earlier studies that showed that replacement factor use with prophylactic treatment is 50% greater than with on-demand therapy (Aledort 2000, Carlson 1999, Rogoff 2002). However, a recent finding by Sun (2017) showed only a modest 7.3% increase in factor use in patients with severe hemophilia A (n=18) who switched from on-demand to prophylactic therapy.

Higher treatment expenditures associated with the use of newer factor replacement products that carry a higher international per-unit price are subject to value-added scrutiny (Aledort 2002, Kletter 2002). While EHL products offer less-frequent injections with similar annualized bleeding rates compared with SHL prophylactic regimens (Konkle 2015, Mahlangu 2014), the expectation for EHL products is that fewer factor IUs would be required to achieve similar efficacy, such that a higher per-unit price would be balanced by lower usage (Croteau 2015). Our analysis provides real-world information that contradicts that expectation. Using two databases, the cross-­sectional analysis identified 10% to 45% higher median factor IUs dispensed per patient per quarter and 51% to 122% higher median expenditures for EHL products compared with SHL products. Between the two databases, overall median factor IUs dispensed and expenditure values for the EHL products were consistent; however, factor IUs dispensed for SHL products were higher in the Optum data­base than in the Truven database. The number of patients in the Optum database was approximately half of that in the Truven database. That there were fewer patients in the Optum database may contribute to the greater variability in the Optum values. In addition, as the quarters progressed, a decline was noted in the number of patients included in each quarter. A review of quarterly median and mean values for IUs dispensed and expenditures over time (data not shown) revealed greater variability for EHL than SHL products. Despite these differences, the expenditures were consistently higher for patients in the EHL group compared with the SHL group, with the exception of a few later quarters in which there were five or fewer patients in the EHL cohort.

The Truven database identified 29 patients who served as their own controls and were followed over time as they switched from SHL to EHL factor replacement products. These data provide similar findings of greater-than-anticipated factor IUs dispensed for an EHL product, compared with an SHL product, in the initial two quarters after switching products. Median factor IUs dispensed for EHL products were higher than that for SHL products in the first two quarters, and thereafter failed to meet substantial reductions from baseline. This finding is consistent with clinical trial data. In the A-LONG study of the EHL product efmoroctocog alfa, patients were able to decrease the number of infusions per week from three times weekly to every three days (i.e., twice weekly) or even every five days when switching from rFVIII or FVIII to efmoroctocog alfa, but their weekly factor use remained constant (Shapiro 2014). In a study of pegylated FVIII, the same doses of SHL rFVIII and EHL pegylated rFVIII were administered to patients with hemophilia A, and 70.4% were able to reduce their dosing frequency (Konkle 2015). As with all new products, there may be a learning curve involved for switching from an SHL to EHL. Pharmaco&w173;kinetic models indicate that patients exhibit a broad range of times for factor degradation rates. In addition, multiple variables contribute to bleeding tendency, including adherence, joint health, lifestyle, and activity level (Gringeri 2015). Thus, achieving goals for FVIII levels requires knowledge of a patient’s pharmacokinetics on their current standard factor regimen and monitoring these levels to design an optimal EHL regimen (Croteau 2015). Patients with pharmaco­kinetic half-lives at the shorter end of the range may benefit from increased factor activity with an EHL product, but they may not always be able to reduce the dosing interval. The transition from an SHL to EHL rFVIII product generally can use a standardized half-life extension of 1.5 times that of the SHL product. Interestingly, the dosing strategy that gives pricing parity for the Fc conjugated rFVIII EHL product does not align with the recommended starting dose per the product labelling (Croteau 2015, Eloctate 2017).

The need for higher doses and the variability in patients’ pharmaco­kinetic parameters provide possible explanations for the consistently higher expenditures associated with EHL products documented in our analyses. The variability in expenditures and factor dispensation observed in this analysis calls into question the predictability of future costs. Based on our findings, it cannot be assumed that use of an EHL rFVIII product will necessarily result in cost-neutrality compared with an SHL rFVIII product. Additional studies are warranted to evaluate whether the increase in factor IUs dispensed and related expenditures observed with EHL rFVIII products in our study are associated with improved clinical outcomes and downstream cost savings. Until such data are available, patients, health care providers, and payers cannot assume that the higher per-IU cost of EHL rFVIII products is justified.

Limitations

This analysis was retrospective in nature and provided descriptive data. As a result, the predictive power of the results is limited. There is also a potential for bias as a result of the small sample sizes and the lack of random selection.

For rare diseases such as hemophilia, the challenges associated with data acquisition are well documented. In addition to small sample sizes, populations may lack geographic diversity populations, and there may be gaps in evidence related to the natural history and course of the disease (Spannheimer 2016). Moreover, claims payments do not provide information typically used in research to customize a study, such as disease characteristics (mild, moderate, or severe), patient weight, treatment regimen (prophylaxis or on-demand), number of bleeding events (if any), and patient adherence to the regimen. As a result, these variables were either not known or not provided in all databases. In addition, there was no measure of the value of convenience associated with the EHL products or an assessment of other patient-related factors. The prescription data on factor IUs dispensed are a proxy measure for usage in this analysis, but there is no confirmation that these factor IUs were actually administered. Also, the contribution of other subsequent discounts (if any) or distribution expenditures not already captured in the databases to the overall expenditures and IUs dispensed could not be determined. A prospective, diary-based observational study of real-world patient use and outcomes would address some of the hypotheses generated from the current analysis.

Conclusions

This analysis of real-world administrative data showed that switching from an SHL to an EHL product for hemophilia A was associated with variable factor IUs dispensed and higher expenditures over two years following the switch. Expenditures were consistently higher after the switch to an EHL product, even in periods when the number of rFVIII IUs dispensed was lower compared with a corresponding period before the switch. In the initial two quarters after switching to an EHL product, most patients had substantially more factor IUs dispensed compared with prior SHL product IUs. This finding stands in contrast with the less-frequent dosing schedules anticipated for EHL products that were the rationale given for pricing EHL products to create parity with existing SHL products. These real-world data may challenge assumptions regarding typical factor usage and expenditures associated with product switching in patients with hemophilia A. Further analyses incorporating larger numbers of patients and assessment of clinical outcomes are needed.

Corresponding Author

Amit Chhabra, MBBS, MPH
Pfizer Inc.
235 E. 42nd St.
4th Floor, Office 131
New York, NY 10017
Fax: (646) 563-1634
E-mail: Amit.Chhabra@pfizer.com

Acknowledgments. The authors thank Emily Rubinstein of Pfizer Inc. for her contributions to this analysis. Medical writing and editorial support were provided by Bina J. Patel, PharmD, and Michelle McDermott, PharmD, of Peloton Advantage, Parsippany, N.J., and were funded by Pfizer Inc.

Author Disclosures. This study was sponsored by Pfizer Inc. Pfizer is the manufacturer of Xyntha, a product discussed in this article. Chhabbra, Fogarty, Tortella, Spurden, Alvir, McDonald, and Hodge are employees of Pfizer and may own stock or stock options in the company. Pleil was an employee of Pfizer at the time of this analysis.

Author Roles. All authors participated in study design, data interpretation, and manuscript review and revisions, and granted approval for the submission of the manuscript. Chhabra, Alvir, McDonald, and Tortella also participated in data analysis.

References

  • Advate [package insert]. Westlake Village, Calif.: Baxalta US Inc. (Shire plc); 2016.
  • Adynovate [package insert]. Westlake Village, Calif.: Baxalta US Inc. (Shire plc); 2017.
  • Aledort LM. Economics of hemophilia care. Haemostasis. 2000;30(6):333–336.
  • Aledort LM. Can costs of hemophilia products be curtailed? Not as we do business today! Thromb Haemost. 2002;88(4):541.
  • Carlson DL, Heaton A, Portuese E. Improving the care of hemophiliacs by managing the drug benefit. Am J Health Syst Pharm. 1999;56(13):1303–1305.
  • Castro HE, Briceno MF, Casas CP, Rueda JD. The history and evolution of the clinical effectiveness of haemophilia type A treatment: a systematic review. Indian J Hematol Blood Transfus. 2014;30(1):1–11.
  • Chen SL. Economic costs of hemophilia and the impact of prophylactic treatment on patient management. Am J Manag Care. 2016;22(5 Suppl):s126–s133.
  • Croteau SE, Neufeld EJ. Transition considerations for extended half-life factor products. Haemophilia. 2015;21(3):285–288.
  • Eloctate [package insert]. Cambridge, Mass.: Biogen Inc.; 2017.
  • Franchini M. The modern treatment of haemophilia: a narrative review. Blood Transfus. 2013;11(2):178–182.
  • George LA, Camire RM. Profile of efraloctocog alfa and its potential in the treatment of hemophilia A. J Blood Med. 2015;6:131–141.
  • Gringeri A, Wolfsegger M, Steinitz KN, Reininger AJ. Recombinant full-length factor VIII (FVIII) and extended half-life FVIII products in prophylaxis—new insight provided by pharmacokinetic modelling. Haemophilia. 2015;21(3):300–306.
  • Kletter SD, Rankin PJ. Cost trends in the treatment for hemophilia: another view. Thromb Haemost. 2002;88(4):542–544.
  • Kogenate FS [package insert]. Whippany, N.J.: Bayer HealthCare LLC; 2016.
  • Konkle BA, Stasyshyn O, Chowdary P, et al. Pegylated, full-length, recombinant factor VIII for prophylactic and on-demand treatment of severe hemophilia A. Blood. 2015;126(9):1078–1085.
  • Mahlangu J, Powell JS, Ragni MV, et al. Phase 3 study of recombinant factor VIII Fc fusion protein in severe hemophilia A. Blood. 2014;123(3):317–325.
  • MASAC Document #241. MASAC Recommendation Concerning Prophylaxis (regular administration of clotting factor concentrate to prevent bleeding). MASAC Document #241 (replaces #179). 2016. https://www.hemophilia.org/sites/default/files/document/files/241Prophylaxis.pdf. Accessed Aug. 17, 2018.
  • McMullen S, Buckley B, Hall E 2nd, et al. Budget impact analysis of prolonged half-life recombinant FVIII therapy for hemophilia in the United States. Value Health. 2017;20(1):93–99.
  • Miners AH, Krishnan S, Pasi KJ. Predicting the outcomes of using longer-acting prophylactic factor VIII to treat people with severe hemophilia A: a hypothetical decision analysis. J Thromb Haemost. 2016;14(11):2141–2147.
  • Rogoff EG, Guirguis HS, Lipton RA, et al. The upward spiral of drug costs: a time series analysis of drugs used in the treatment of hemophilia. Thromb Haemost. 2002;88(4):545–553.
  • Shapiro AD, Ragni MV, Kulkarni R, et al. Recombinant factor VIII Fc fusion protein: extended-interval dosing maintains low bleeding rates and correlates with von Willebrand factor levels. J Thromb Haemost. 2014;12(11):1788–1800.
  • Srivastava A, Brewer AK, Mauser-­Bunschoten EP, et al. Guidelines for the management of hemophilia. Haemophilia. 2013;19(1):e1–e47.
  • Spannheimer A. The value of real-world evidence in rare disease research. CenterWatch Online. 2016. https://www.centerwatch.com/news-online/2016/03/08/value-real-world-evidence-rare-disease-research/. Accessed Sept. 7, 2018.
  • Sun HL, McIntosh KA, Squire SJ, et al. Patient powered prophylaxis: a 12-month study of individualized prophylaxis in adults with severe haemophilia A. Haemophilia. 2017;23(6):877–883.
  • Xyntha [package insert]. Philadelphia, Pa.: Wyeth Pharmaceuticals (Pfizer Inc); 2015.
  • Zhou ZY, Koerper MA, Johnson KA, et al. Burden of illness: direct and indirect costs among persons with hemophilia A in the United States. J Med Econ. 2015;18(6):457–465.
Some abbreviations used in this article

EHL. Extended half-life

FVIII. Factor VIII

IQR. Interquartile range

IU. International unit

PEG. Polyethylene glycol

rFVIII. Recombinant factor VIII

SD. Standard deviation

SHL. Standard half-life

Supplementary Table 1
Cross-sectional analysis of median and mean IUs dispensed of recombinant factor VIII replacement products (Optum Clinformatics Data Mart and Truven Health MarketScan Databases)
Optum database  
Parameter Q1
(n=276)
Q2
(n=169)
Q3
(n=125)
Q4
(n=86)
Q5
(n=71)
Q6
(n=54)
Q7
(n=49)
Q8
(n=35)
Q9
(n=27)
-
EHL cohort n=33 n=26 n=18 n=12 n=5 n=4 n=1 n=1 n=1 -
Median 79,332 59,824 58,551 67,671 50,315 51,678 96,864 94,624 64,648 -
IQR 47,351–118,070 46,690–105,790 48,994–96,814 55,410–112,909 32,064–56,490 37,377–63,905 - - - -
Mean (SD) 81,084 (44,403) 73,972 (47,549) 72,093 (39,228) 83,989 (47,450) 46,994 (20,405) 50,641 (21,291) 96,864 94,624 64,648 -
SHL cohort n=243 n=136 n=101 n=70 n=63 n=47 n=45 n=31 n=24 -
Median 65,306 61,920 75,108 75,108 70,138 74,400 72,439 69,330 69,620 -
IQR 30,240–125,100 38,290–115,277 41,600–122,670 45,360–109,695 44,145–114,525 46,755–94,500 46,908–112,560 46,244–92,060 42,741–104,346 -
Mean (SD) 85,008 (73,758) 81,080 (60,399) 90,339 (71,504) 81,892 (46,745) 83,324 (46,790) 77,820 (41,709) 81,078 (43,808) 77,108 (39,065) 91,096 (66,382) -
Median difference, EHL vs. SHL (%) 14,026 (21) –2,096 (–3) –16,557 (–22) –7,437 (–9) –19,823 (– 28) –22,722 (–31) 24,425 (34) 25,294(36) –4,972(–7)
Truven database
Parameter Q1
(n=500)
Q2
(n=301)
Q3
(n=237)
Q4
(n=189)
Q5
(n=152)
Q6
(n=120)
Q7
(n=101)
Q8
(n=75)
Q9
(n=58)
Q10
(n=6)
EHL cohort n=86 n=67 n=54 n=38 n=27 n=18 n=11 n=6 n=4 n=1
Median 83,080 65,660 65,653 62,549 62,181 79,496 59,024 62,284 69,141 78,398
IQR 52,638–105,776 46,848–90,640 41,760–94,840 40,380–82,761 33,221–87,213 55,280–94,280 50,888–83,364 17,188–75,376 30,856–98,097 -
Mean (SD) 82,533 (42,106) 70,334 (38,747) 71,779 (42,306) 64,277 (29,766) 61,586 (31,223) 78,025 (37,996) 60,725 (27,909) 53,762 (32,536) 64,477 (42,073) 78,398
SHL cohort n=409 n=232 n=183 n=151 n=125 n=102 n=90 n=69 n=54 n=5
Median 42,680 51,201 61,080 59,135 63,550 64,774 64,208 83,592 61,070 37,284
IQR 11,808–90,900 24,480–90,726 33,060–95,040 30,720–102,096 33,540–99,548 37,260–101,239 43,860–96,096 43,356–116,622 37,404–104,730 35,115-59,160
Mean (SD) 61,983 (65,623) 65,092 (59,568) 73,685 (59,752) 75,299 (63,654) 77,028 (67,657) 75,754 (60,069) 79,294 (51,875) 85,563 (54,039) 75,067 (46,980) 56,269 (38,221)
Median difference, EHL vs. SHL (%) 40,400 (95) 14,459 (28) 4,573
(7)
3,414
(6)
–1,369 (–2) 14,722 (23) –5,184 (–8) –21,308 (–25) 8,071 (13) 41,114 (110)

One quarter=90 days.
EHL=extended half-life, IQR=interquartile range, IU=international unit, SD=standard deviation, SHL=standard half-life

Supplementary Table 2
Cross-sectional analysis of median and mean quarterly expenditures of recombinant Factor VIII replacement products (Optum Clinformatics Data Mart and Truven Health MarketScan Databases)
Optum database
Parameter Q1
(n=276)
Q2
(n=169)
Q3
(n=125)
Q4
(n=86)
Q5
(n=71)
Q6
(n=54)
Q7
(n=49)
Q8
(n=35)
Q9
(n=27)
-
EHL Cohort n=33 n=26 n=18 n=12 n=5 n=4 n=1 n=1 n=1 -
Median 133,516 100,684 98,541 113,889 84,680 86,974 163,022 159,252 108,803 -
IQR 79,692–199,234 78,579–182,572 82,457–164,193 93,255–190,026 53,964–95,073 62,905–107,552 - - - -
Mean (SD) 139,298 (77,120) 126,559 (81,219) 122,791 (66,928) 141,354 (79,859) 79,091 (34,341) 85,229 (35,833) 163,022 159,252 108,803 -
SHL Cohort n=243 n=136 n=101 n=70 n=63 n=47 n=45 n=31 n=24 -
Median 79,379 77,729 93,209 89,945 79,359 85,779 89,024 79,374 86,398 -
IQR 37,230–152,351 47,518–142,325 51,626–151,203 56,396–127,081 56,396–142,126 58,023–116,772 59,303–137,312 56,70–115,156 52,826–129,493 -
Mean (SD) 103,637 (89,581) 99,379 (71,594) 111,014 (87,076) 100,637 (57,418) 102,322 (57,793) 95,124 (51,048) 98,944 (53,065) 94,178 (49,099) 104,804 (58,146) -
Median difference, EHL vs. SHL (%) $54,137 (68) $22,955 (30) $5,332 (6) $23,944 (27) $5,321 (7) $1,195 (1) $73,998 (83) $79,878 (101) $22,405 (26) -
Truven database
Parameter Q1
(n=500)
Q2
(n=301)
Q3
(n=237)
Q4
(n=189)
Q5
(n=152)
Q6
(n=120)
Q7
(n=101)
Q8
(n=75)
Q9
(n=58)
Q10
(n=6)
EHL cohort n=86 n=67 n=54 n=38 n=27 n=18 n=11 n=6 n=4 n=1
Median 162,694 122,111 124,310 120,761 116,179 153,405 114,892 107,429 135,766 152,819
IQR 107,174–209,194 86,200–167,972 86,885–183,774 77,172–164,164 57,718–166,427 93,107–179,312 94,763–151,389 28,364–150,101 61,805–184,701 -
Mean (SD) 168,942 (81,620) 137,337 (73,178) 141,082 (78,848) 122,455 (55,620) 115,757 (60,221) 147,326 (74,143) 113,317 (55,008) 97,002 (63,518) 123,253 (79,334) 152,819
SHL cohort n=409 n=232 n=183 n=151 n=125 n=102 n=90 n=69 n=54 n=5
Median 56,380 68,775 83,939 85,357 81,895 83,843 86,903 106,741 79,100 51,338
IQR 16,917–119,228 34,178–119,184 47,256–132,132 50,356–134,366 49,165–137,923 53,897–148,263 58,129–134,484 56,095–162,264 53,887-152,483 45,294–78,385
Mean (SD) 79,928 (82,019) 89,527 (78,028) 101,930 (80,973) 105,933 (88,191) 106,977 (94,326) 107,331 (91,782) 109,213 (76,255) 119,859 (82,594) 105,635 (71,497) 77,219 (53,873)
Median difference, EHL vs. SHL (%) $106,314 (189) $53,336 (78) $40,371 (48) $35,404 (41) $34,284 (42) $69,562 (83) $27,989 (32) $688 (1) $56,666 (72) $101,481 (198)
One quarter = 90 days. EHL=extended half-life, IQR=interquartile range, SD=standard deviation, SHL=standard half-life