Expenditures for First- and Second-Generation Tyrosine Kinase Inhibitors Before and After Transition of Imatinib to Generic Status

Key Points

Question  How did the entry of imatinib into the generic market change health plan spending and patient cost responsibility associated with imatinib and comparable second-line drugs for chronic myeloid leukemia (ie, nilotinib and dasatinib)?

Findings  In this retrospective cohort study, health plan and patient spending for 1301 commercially insured or Medicaid-insured patients with chronic myeloid leukemia demonstrated a modest decline in imatinib spending after generic entry, and a continued increase in nilotinib and dasatinib spending from pregeneric to postgeneric entry for imatinib.

Meaning  Although further price changes may occur with generic imatinib, spending remains high for second-line therapies, suggesting a need for measures to facilitate cost reduction as well as standardization of treatment of chronic myeloid leukemia.

Importance  Imatinib introduced a paradigm shift in the treatment of patients with chronic myeloid leukemia (CML), allowing a lifespan that is almost comparable to the general population. However, the health care expenditures associated with imatinib have increased steadily since its introduction in 2001. Since the generic market entry of imatinib on February 1, 2016, it became possible to determine the effect of generic entry on the health care expenditures associated with imatinib, along with the concurrent pricing trends of second-generation tyrosine kinase inhibitors (TKIs).

Objective  To compare health care expenditure related to imatinib treatment for patients with CML prior to its generic status with expenditures after, in a real-world setting.

Design, Setting, and Participants  A retrospective cohort study using data from the Truven Health MarketScan Database was carried out including 1301 commercially insured patients or patients with Medicaid between the ages of 18 and 64 years with a CML diagnosis between September 1, 2014 and December 31, 2017. A single interrupted time series (ITS) method was used to evaluate monthly expenditures associated with imatinib for the health plan from September 1, 2014 to December 31, 2017, with imatinib switching to generic on February 1, 2016, as the interruption. The initial 6-month period postinterruption was excluded to allow for the new price structure to stabilize. Nilotinib and dasatinib were evaluated using a comparative ITS design. The analysis took place between September 1, 2014 to December 31, 2017.

Main Outcomes and Measures  Commercial and Medicaid health plan expenditure and patient cost responsibility for 30-day blocks per patient for imatinib, dasatinib, and nilotinib were calculated from September 1, 2014 through December 31, 2017. Pricing was adjusted via 2017 consumer price index for medical services.

Results  The sample included a total of 1301 patients (1102 commercially insured and 199 insured through Medicaid) with a median (range) age at diagnosis of 50 (18-62) years for commercially insured patients and 50 (47-52) years for patients with Medicaid. Of the 1301 patients, 704 (54.1%) were men. There was a significant pregeneric increase in expenditure (commercially insured: $143 per patient per month, P < .001; Medicaid: $152 per patient per month, P = .001). There was a significant reduction over the 6-month interruption period (February 2016 through August 2016) between the pregeneric and generic phase (commercial: $−3097 per patient, P < .001; Medicaid: $−2077 per patient, P = .002). Controlling for secular trends, this was followed by a small decline in expenditure during the generic phase (commercially insured: −$93 per patient per month; P = .03; Medicaid: −$182 per patient per month; P = .001). Nilotinib and dasatinib maintained an increasing trend in both the pregeneric and generic phase of imatinib, resulting in a significant difference-in-difference slope for commercially insured patients (nilotinib: −$186, P = .006; dasatinib: −$175, P = .007) and patients with Medicaid (nilotinib: −$299, P = .002; dasatinib: −$329, P < .001). There were no significant trends for patient cost responsibility.

Conclusions and Relevance  The modest decline in expenditure associated with imatinib after generic entry accompanied by the increasing expenditure trends associated with the second-generation TKIs suggest a need for measures to facilitate cost reduction as well as standardization of CML treatment.

As the prototype of tyrosine kinase inhibitors (TKIs), imatinib mesylate has been a phenomenal success for patients with chronic myeloid leukemia (CML), allowing them to have a lifespan approaching that of the general population.¹ This success, however, is limited to those who can afford to adhere to lifelong treatment,² at least until discontinuation studies can prove otherwise.³ At its introduction in 2001, the list price of imatinib was $26 400 per year in the United States; in February 2016, just prior to transition to generic status, the price had reached $142 000 per year.⁴ Simulation data projected the decline in cost after transition of imatinib to generic status⁵; however, real-world data are lacking. Further, the pricing patterns of second-generation TKIs (dasatinib and nilotinib) from pretransition to posttransition of imatinib to the generic status are unknown. We address these gaps in this report.

Truven Health MarketScan Database was used to identify 1301 patients with a CML diagnosis between September 1, 2014 and December 31, 2017, aged between 18 and 64 years. The sample included patients who were commercially insured (n = 1102) or insured through Medicaid (n = 199) for 12 months or longer prior to diagnosis and 6 months or more after diagnosis, and were receiving imatinib, nilotinib, or dasatinib. This study was deemed exempt by the University of Alabama at Birmingham institutional review board, because all data used were deidentified.

Commercial and Medicaid health plan expenditure and patient cost responsibility for 30-day blocks per patient for imatinib, dasatinib, and nilotinib were calculated from September 1, 2014 through December 31, 2017. Pricing was adjusted via 2017 consumer price index for medical services.⁶

A single interrupted time series (ITS) method evaluated monthly expenditures for the health plan from September 1, 2014 to December 31, 2017, with imatinib entering the generic market on February 1, 2016 as the interruption. The initial 6-month period postinterruption was excluded to allow for the new price structure to stabilize, resulting in a postinterruption observation period from August 1, 2016 to December 31, 2017 (generic phase). Nilotinib and dasatinib were evaluated using comparative ITS, allowing for a difference-in-difference estimation between slopes’ pregeneric and postgeneric phases (eg, imatinib pregeneric-postgeneric vs dasatinib pregeneric-postgeneric; imatinib pregeneric-postgeneric vs nilotinib pregeneric-postgeneric). The analysis took place between September 1, 2014 to December 31, 2017 using SAS statistical software (version 9.4; SAS Institute, Inc).

Table 1 summarizes the patient characteristics. The median age at CML diagnosis was 50 years (range, 18-62 years), median follow-up was 22 months (range, 6-59 months). After February 1, 2016, among 441 patients taking Gleevec (Novartis) prior to the generic switch, 300 (68%) received at least one 30-day supply of generic imatinib, whereas 225 (51%) received at least one 30-day supply of Gleevec among the commercially insured; among the Medicaid population of 111 patients, the corresponding proportions were 82 (74%) and 37 (33%), respectively. For imatinib, only generic medication prescriptions were included in the cost analysis in the generic phase. Among commercially insured patients and patients with Medicaid taking generic imatinib, 56 (18.7%) and 11 (13.4%), respectively switched to nilotinib or dasatinib after imatinib entered the generic market.

There was a significant (P < .001) pregeneric increase in expenditure (commercially insured: $143 per patient per month; 95% CI, $90-$196; P < .001; Medicaid: $152 per patient per month; 95% CI, $54-$250; P = .001) (Table 2) (Figure, A and B). There was a statistically significant reduction in expenditure over the 6-month interruption (February 2016 through August 2016) immediately after the pregeneric phase (commercially insured: −$3097 per patient; 95% CI, −$4117 to $2077; P < .001; Medicaid: −$2077 per patient; 95% CI, −$3294 to −$861; P = .002). Controlling for secular trends, this was followed by a continued small decline in expenditure during the generic phase (commercially insured, −$93 per patient per month; 95% CI, −$16 to −$20; P = .01; Medicaid −$182 per patient per month; 95% CI, −$248 to −$117; P = .001). The 2017 expenditure for imatinib was estimated at $109 008 per patient per year for commercially insured and $93 378 per patient per year for Medicaid. The eTable in the Supplement shows comparable trends when evaluating patients treated with imatinib, nilotinib, or dasatinib as single agents.

There was a significant difference-in-difference slope for commercially insured patients (nilotinib: −$186; 95% CI, −$313 to −$60; P = .006; dasatinib: −$175; 95% CI, −$299 to −$53; P = .007) and patients with Medicaid (nilotinib: −$299; 95% CI, −$478 to −$120; P = .002; dasatinib: −$329; 95% CI, −$469 to −$188; P < .001), indicating that nilotinib and dasatinib had a greater magnitude of increase in expenditure from 2014 to 2017 than the change in expenditure for imatinib that occurred over time (Figure) (Table 2).

The mean patient cost responsibility of imatinib in the pregeneric phase was $147 per patient per month for commercially insured patients and $14 per patient per month for patients with Medicaid; the corresponding costs in the generic phase were $85 per patient per month and $7 per patient per month. There were no statistically significant pregeneric trends (commercially insured, $1.60; P = .09 vs Medicaid, −$0.02; P = .10), level change (commercially insured, −$52; P = .40 vs Medicaid, −$0.16; P = .20), or postgeneric trends (commercially insured, −$4.20; P = .50 vs Medicaid, $0.01; P = .50) for either the commercially insured or patients with Medicaid, respectively, for imatinib. The postgeneric trends remained nonsignificant after controlling for secular trends (commercially insured: −$2.60, P = .50; Medicaid: −$0.01, P = .20). Similarly, there were no significant trends for cost for nilotinib or dasatinib, with the exception of a pregeneric negative slope (−$0.02, P = .01).

After generic entry, the monthly health plan expenditure for imatinib declined for both commercially insured patients and patients with Medicaid. However, the rate of decline was slower than the increasing trend observed prior to generic entry. Health plan expenditures for nilotinib and dasatinib demonstrated a steady increase after imatinib became generic. There were no significant trends for patient cost responsibility associated with the TKIs before or after generic entry of imatinib.

Barriers to effectively reducing costs⁷ can include lack of transition to generic product, smaller price reductions than anticipated, fewer market entrants, and changes in prescribing patterns. Cole and Dusetzina⁷ found that generic imatinib accounted for 58% of all such prescriptions 1 year after generic entry; our findings support this (51% of commercially insured and 33% of patients with Medicaid had at least 1 prescription of brand-name Gleevec after February 1, 2016). Further, they identified an increase in the use of nilotinib and dasatinib, such that the use of nilotinib and dasatinib accounted for over half of all TKI initiations. These findings, combined with our observation of a continued increase in second-generation TKI expenditure speak to a need for standardizing the treatment of patients with CML. Several European countries have successfully accomplished generic medication use by generic substitution and/or international nonproprietary name prescribing.^8,9 The United States primarily operates on a state-by-state basis for implementing policies to increase generic medication use.¹⁰ Generic substitution is implemented either as mandatory or a recommended policy, where it can be overruled by the patient. Rates of generic substitution were significantly lower in states with policies requiring patient consent prior to generic substitution.¹¹

Commercial and Medicaid spending were similar, likely because the Medicaid data did not accurately reflect the Medicaid Drug Rebate Program (MDRP) spending adjustment. However, regardless of rebates, imatinib reached almost $52 million dollars in spending for Medicaid in 2018.¹² Medicaid programs must rely on the MDRP to abate costs.¹³ The MDRP’s negotiating power for pricing and rebates is severely limited when there are few competitors.

The patient cost responsibility for imatinib declined from $147 per month to $85 per month for commercially insured patients. Prior studies emphasize that imatinib’s financial burden increases the likelihood of drug discontinuation.^14,15 Given the nature of claims-based analysis, we were unable to determine the financial effect of the decline noted on the patient.

This study was limited to those younger than 65 years. Although the conclusions are likely to be similar for those older than 65 years, similar analyses are warranted. Further, we may not have a full representation of all patients aged 18 to 64 years; MarketScan commercial data are a convenience sample primarily from large employers. The small Medicaid sample limits our ability to provide robust findings. Finally, patient cost responsibility does not account for brand-name drug manufacturers assistance.

Despite limitations, we provide real-world evidence for health plan and patient spending associated with imatinib and second-generation TKIs after the generic entry of imatinib, providing evidence for measures that need to be adopted to facilitate cost reduction as well as standardization of treatment of CML.

Corresponding Author: Kelly M. Kenzik, PhD, University of Alabama at Birmingham, 1600 Seventh Ave S, Lowder 500, Birmingham, AL 35223 (kkenzik@uab.edu).

Accepted for Publication: November 13, 2019.

Published Online: January 30, 2020. doi:10.1001/jamaoncol.2019.6390

Author Contributions: Dr Kenzik had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Kenzik, S. Bhatia.

Critical revision of the manuscript for important intellectual content: Kenzik, R. Bhatia.

Statistical analysis: Kenzik.

Study supervision: R. Bhatia, S. Bhatia.

Conflict of Interest Disclosures: None reported.

Funding/Support: Dr Kenzik was funded by the American Cancer Society Mentored Research Scholar Grant in Applied and Clinical Research (MRSG-18-020-01 CPPB). Funding from the American Cancer Society for Dr Kenzik’s Mentored Research Scholar Grant was used to obtain the Truven MarketScan Healthcare Database.

Role of the Funder/Sponsor: The American Cancer Society had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.