Abstract: The therapeutic landscape for chronic myeloid leukemia (CML) has evolved substantially over the past decade, most notably with the introduction of second- and third-generation tyrosine kinase inhibitors (TKIs) and availability of generic imatinib. We examined 10-year US trends in treatment selection, switching, adherence, and molecular monitoring in conjunction with treatment switching. Methods: Using data from the OptumLabs Data Warehouse, 1449 commercially insured and Medicare Advantage enrollees initiating treatment with an indicated TKI for the treatment of CML between January 1, 2007, and October 31, 2017, were reviewed. We studied the impact of generic entry by comparing trends in treatment selection, switching, adherence, and molecular monitoring for patients initiating TKI treatment between 2007 and 2017. Results: Over time, the proportion of patients initiating imatinib (vs dasatinib or nilotinib) declined from 62.4% to 51.8% (P=.001). Among patients who initiated with imatinib in either period, 18.2% switched to a newer TKI within one year. Of patients starting on dasatinib or nilotinib, 8.6% switched to imatinib and 11.8% to another second-generation or ponatinib TKI. Of those who switched within the first year, 35.7% had evidence of molecular monitoring in the one month prior and 54.4% in the 90 days prior to switching. Conclusion: Second-generation TKIs are the most commonly used agents by the end of year one of treatment due to: (1) an increasing proportion of patients receiving dasatinib or nilotinib as first line; and (2) the nearly 20% switch rate in the first year (the majority from imatinib to second-generation TKIs). Molecular testing rates remain low, even prior to switching.
Key Words: TKI, tyrosine kinase inhibitors, adherence, molecular monitoring, generics
Download the PDF to view supplementary materials
Chronic myeloid leukemia (CML) treatment and clinical outcomes have improved considerably since the introduction of imatinib, the first oral tyrosine kinase inhibitor (TKI), in 2001.1 Second-generation TKIs (dasatinib and nilotinib) were initially approved as second-line TKI therapy following imatinib resistance or intolerance in 2006 and 2007, respectively and subsequently joined imatinib as approved for first-line treatment of newly diagnosed Philadelphia chromosome positive (Ph+) CML in chronic phase.2,3 In 2012, bosutinib and ponatinib were approved for treatment of CML but were not indicated for first-line treatment during our study period.4,5 Finally, generic imatinib was introduced in 2016. While TKI therapy has provided survival benefit for many patients, real-world prescribing of TKIs can be challenging to providers when balancing choosing first-line drug, maximizing patient adherence, and switching TKIs when appropriate. Further, studies have demonstrated that in “real-world” (nonclinical trial) settings, more than one-quarter of patients may not be adherent to TKI therapy6 with low adherence representing a major risk factor for poor treatment outcomes.6-9
Any of the three TKIs currently approved for first-line use may be selected as initial treatment for patients with CML. While dasatinib and nilotinib have demonstrated superiority over imatinib in complete cytogenetic or major molecular response, no survival benefits have been shown, with imatinib remaining highly effective in a large proportion of patients with CML.10 The DASISION trial found lower rates of major molecular response (MMR) at 5 years among patients treated with imatinib compared with dasatinib (64% vs 76%) but no difference in overall survival (OS) (90% vs 91%).11 Similarly, the ENESTnd trial found MMR rates were lower among the imatinib cohort (60%) vs patients treated with nilotinib (77%) at 5 years but found no difference in OS.12
Switching from one TKI to another is either reactive (eg, the patient experiences primary or secondary resistance or intolerance to current TKI) or proactive (ie, response is acceptable but not optimal).13 Response to TKIs can be measured as hematologic (level of blood cells), cytogenetic (proportion of Philadelphia chromosome-positive cells), or molecular (level of BCR-ABL transcripts). Monitoring helps determine whether a patient is responding optimally to treatment and provides evidence regarding the need to switch to a second-line TKI. Despite the inclusion of regular molecular monitoring (every 3 months after TKI initiation) in the US National Comprehensive Care Network (NCCN) guidelines since 2006, studies have found that many patients had no molecular monitoring test within the first year.14-17 Frequent monitoring is required for measurement of hematologic, cytogenetic, and molecular responses in order to evaluate therapy response and disease progression (ie, from chronic phase to accelerated or blast phase). Patients who do not respond should be considered for a change in therapy.10
Given these changes in the therapeutic landscape for CML, the objective of this study was to examine the impact the availability of newer TKIs and generic imatinib had on treatment selection. Does the potential for cost saving move patients to start with generic imatinib or are physicians starting patients on more expensive second-generation TKIs? Has the rate of switching from first-generation TKI to a second- or third-generation TKI decreased over time? Have lower priced generics had an impact on adherence rates given the possibility for lower out-of-pocket cost exposure? In addition to treatment patterns, we investigated the use of molecular monitoring over time and in relation to switching, which we would expect to be informed by molecular monitoring. This study used administrative claims data from 2007-2018 for commercial and Medicare Advantage (MA) health plan enrollees. The use of observational data in the study provides an alternate to the clinical trial perspective, offering insight into CML treatment patterns in real-world practice, where providers can choose between several TKIs for first-line treatment of CML.
Methods
Data Source
This was a retrospective analysis using de-identified administrative claims data from the OptumLabs Data Warehouse (OLDW), which includes medical and pharmacy claims, laboratory results, and enrollment records for commercial and MA enrollees. The database contains longitudinal health information on enrollees and patients, representing a diverse mixture of ages, ethnicities, and geographic regions across the United States, with greatest proportions of members in the Midwest and South regions.18 During the study period (July 1, 2006, through October 31, 2018), the number of covered annual lives with both medical and pharmacy benefits ranged between 13 and 15 million.
Study Population
Health plan enrollees who filled a prescription for an indicated TKI (imatinib, dasatinib, nilotinib, bosutinib, or ponatinib) were identified, and the sample was limited to those starting TKI treatment between January 1, 2007, and October 31, 2017. The index date was defined as the date of the first TKI fill during this period. Only those patients continuously enrolled in their health plan during the 6 months prior to the index date (ie, baseline period) and for the 12 months following the index date (ie, follow-up period) were retained, with follow-up censored after October 31, 2018. Patients were required to have a diagnosis of CML (ICD-9 code 205.1x or ICD-10 code C92.1x). Patients with stem cell transplantation prior to starting a TKI or medical claims for both CML and lymphoid leukemia were excluded from the study population (Supplemental Table 1).
Study Measures
Primary study measures included first-line TKI, switching, adherence, and molecular monitoring rates. For the purpose of this study, dasatinib, nilotinib, and bosutinib were categorized as a second-generation TKI; ponatinib as third generation. Adherence was defined by the medication possession ratio over the first year of TKI treatment (medication possession ratio=total days’ supply of any TKI filled during the year divided by 365; 1.0 indicates perfect adherence). For example, if a patient filled 60 days’ supply of imatinib and 270 days’ supply of dasatinib, without overlap, their MPR for the 1-year period would be approximately 90% ([270 + 60]/365). Patients were classified as switching if they had a filled prescription of a different TKI during the first year.
All dates of molecular monitoring were identified during the 90-day period immediately preceding the initiation of first-line TKI through one year following the initiation of first-line TKI. Molecular monitoring was identified using Current Procedural Terminology (CPT) codes for qPCR testing recorded on a medical claim. Because specific CPT codes to identify qPCR tests were not effective until January 1, 2013, we include a group of related procedure codes to identify molecular monitoring before 2013 (Supplemental Table 1). The identification criteria used to identify molecular monitoring in this study were less restrictive than criteria used by Latremouille-Viau, et al,17 because we prioritized mitigating type I error (classifying those who are tested as not being tested).
Covariates calculated during the baseline period included “care days” (the summation of total days in a hospital or long-term care facility and days with at least one ambulatory visit), hospitalizations, and the Charlson Comorbidity Index (CCI).19 We calculated the CCI excluding cancer-related categories to get a measure of patients’ noncancer comorbidities. Evidence of prior treatment with non-TKIs (interferon, cytarabine, busulfan, and hydroxyurea) and evidence of solid tumors during the baseline period were also identified.
Claims data does not include information on cancer staging. In the absence of cancer staging information when analyzing retrospective data, other investigators have used a proxy for cancer staging, where patients with evidence of metastases or lymph node involvement are classified as having advanced disease.20,21 Darkow et al22 built upon that methodology, using patients’ complete medical claims history to create a proxy variable for cancer complexity for inclusion in our econometric models that is more comprehensive than the aforementioned proxy for cancer staging used by other investigators. Their methodology defines complexity by diagnoses (as identified on claims) that suggest a patient will be more difficult to manage than the typical patient with a particular condition, whether because of (1) the intrinsic activity of the given condition or (2) the difficulty of managing associated complications, comorbidities, and adverse effects of the treatment(s) used for the management of that condition. Here, we adapt the methodology built by Darkow et al to include ICD-10 coding to create a proxy variable for cancer complexity (low, moderate, or high) for inclusion in our models.
Statistical Analysis
All variables, including patient demographics and clinical characteristics, utilization measures, and TKI treatment patterns, were compared via univariate analyses. Main comparisons of interest were patients who initiated treatment during 2007-2015 vs 2016-2017. This stratification of patients was selected because of the introduction of generic imatinib in 2016, which we hypothesized would affect prescribing patterns for patients with CML. MPR was modeled using ordinary least squares regression. Occurrence of treatment switching was modeled using logistic regression expressed via odds ratios (OR). All analyses were conducted using Stata/SE version 13.1 (College Station, TX) software.
Results
Out of 9928 patients identified with TKI prescription fills in the OLDW through October 31, 2017, we retained 3383 patients filling on or after 2007 who were enrolled for the entire baseline and follow-up period. Of those, 2626 patients were new starts and 1504 patients had a CML diagnosis (Figure 1). We removed 54 patients who were also diagnosed with acute lymphoblastic leukemia (ALL) or received a stem cell transplant prior to TKI treatment. The final sample of 1449 patients was categorized according to the year of first TKI prescription fill (2007-2015, 2016-2017). Over the study period, the demographic and clinical characteristics of TKI-treated CML patients changed (Table 1). The average age of treated patients increased (56.4 vs 59.9 years; P<.001) and the proportion of patients hospitalized prior to initiating TKI treatment decreased (22.4% vs 27.6%; P=.072), whereas the number of care days during baseline increased (12.5 vs 15.2 days; P<.001). While comorbidities as measured by the mean CCI excluding cancer worsened slightly overtime (0.60 vs 0.74; P=.037), the percent of patients classified as high-complexity CML declined from 34.7% to 24.8% (P<.001). The proportion of patients who had evidence of solid tumors and prior non-TKI treatments remained unchanged.
Treatment Choice, Adherence to TKI treatment, and Switching Patterns
Over time, the proportion of patients starting treatment with imatinib vs nilotinib or dasatinib declined from 62.4% to 51.8% (P=.001) (Table 2). Bosutinib and ponatinib were not used in first line. In 2016-2017, 104 of 157 (66%) imatinib starts were generic. MPR remained constant over the two study periods, with mean MPR=0.83 (SD=0.23) and 58.7% of patients classified as highly adherent (MPR > 0.90).
Approximately 1 in 5 switched from their initial TKI within the first year of treatment and this rate was similar for the 2007-2015 and 2016-2017 cohorts (18.2% vs 21.8%; P=.151). Among those who initiated treatment with imatinib, the proportion who switched to a newer TKI within the first year did not vary between the two periods and among those starting on dasatinib or nilotinib, the percent switching to a different TKI also remained constant (Table 2). Across both periods, of those starting on dasatinib or nilotinib and switching to another TKI, 42.4% (50 of 118) switched to imatinib and 57.6% (68) to another second-generation TKI or ponatinib. Over the entire study period, 28 (1.9%) patients switched to ponatinib.
Adherence to guidelines for molecular monitoring improved over time (Figure 2) from 82.4% to 89.1% having evidence of molecular monitoring in at least one quarter during the year following TKI initiation (P=.005). The proportion with testing in three or more quarters also increased (46.1% vs. 57.8%; P<.001). However, among those who switched from their initial TKI to another within the first year, those with evidence of molecular monitoring one month prior (35.1% vs 33.3%; P=.793) or 90 days prior to the switch (53.9% vs 56.1%; P=.753) did not vary over time.
Factors Associated With Treatment Adherence
Multivariate analyses indicated a number of patient characteristics that were significantly associated with adherence to TKI treatment in the first year (Table 3). In these analyses, period (2007-2015 vs 2016-2017) was not a factor associated with adherence nor did it significantly modify the effect of other factors on MPR (models with interaction terms not shown). Therapy adherence (MPR) was lower in those with a higher baseline CCI (P=.032) and in those with greater number of care days (P=.001); commercial enrollees were more likely to adhere to TKI treatment compared with MA enrollees. Further, patients with high cancer complexity had a significantly lower adherence (P=.010). The choice of initial TKI (imatinib or a second generation) was not associated with adherence.
Factors Associated With Switching
Similar to adherence, period was not a factor associated with the likelihood of switching (Table 3) nor did it significantly modify the effect of other factors on the likelihood of switching (models with interaction terms not shown). Men were significantly less likely to switch
(OR = 0.72; P=.017), and patients with high cancer complexity were 2.1 times as likely to switch to a different TKI compared with those with low cancer complexity (P=.03). The choice of initial TKI (imatinib or a second-generation) was not associated with the likelihood of switching.
Discussion
Patterns of TKI therapy among CML patients in the United States have shifted in the past decade—a period that included the approval of the first second- and third-generation TKIs and the availability of generic imatinib. The objective of this study was to examine the interplay between the availability of newer TKIs and generic imatinib had on treatment selection. Despite generic entry, we found that nearly half of patients (49% in 2016-2017) initiated TKI therapy with a second-generation TKI. By the end of the first year of treatment, 56% of patients were on a second- or third-generation TKI as approximately 23% of patients starting on imatinib switch to another TKI during the first year—consistent with 28% of patients switching reported by Banegas et al over a similar period.23 Our findings are more striking when one considers that US health plans often employ step therapy, which requires patients to try generic imatinib before coverage of a second-generation TKI. Patients must then fail on the imatinib before being allowed a “step up” to a second-generation TKI.
Molecular monitoring, used to determine whether a patient is responding, is not systematically done among patients who switch to a second-generation TKI. While we found that molecular monitoring rates increased over time, only 57.8% of patients were tested in at least three quarters following TKI initiation in the most recent years. This is consistent with previous studies suggesting that physicians do not fully adhere to guidelines for treatment response monitoring and therapy adjustment.17,24 Most recently, Latremouille-Viau et al17 found that 36% of patients had no qPCR tests during the first year of CML treatment, which is noticeably higher than the 11% to 18% with no testing in our study. This is possibly because the claims-based algorithm we used to identify molecular monitoring was less restrictive as we prioritized mitigating type I error. Most notably, we found an absence of molecular testing prior to the switch. While reasons for, or appropriateness of, switching could not be determined from the data used in this study, we found evidence of cryptogenic and molecular testing in the 30 days prior to switching for only 35% of patients. This suggests that switches were not being made based on NCCN guidelines for efficacy failure, which should be the most common reason, but rather because of tolerance or lack of understanding of what constitutes therapy failure. To the best of our knowledge, no other published study has investigated the relationship between monitoring and therapy switching.
The results of this study are subject to limitations relating to claims data and retrospective study design. Over time, the number of indications for TKIs, including imatinib, has grown. Because pharmacy claims did not include diagnosis codes, we relied on the presence of a diagnosis on a medical claim during the study period as evidence of CML as opposed to other conditions (eg, gastrointestinal stromal tumors, Ph+ ALL, and graft-vs-host disease). In addition, certain clinical information was not available in claims data, such as factors associated with clinical prognosis (phase of the disease, blood counts, and splenic enlargement) or reason(s) for discontinuation/switch of a medication.
Twelve months of continuous enrollment in a health plan following TKI treatment initiation were required for this study. While this is typical for retrospective analyses, it can be problematic when examining patients with potentially fatal diseases, such as cancer, because patients who disenroll for any reason (including death) are excluded from the analysis. Patients who died within the first 12 months following initiation of TKI therapy may be of special interest. In this dataset, deceased patients were not distinguishable from patients who disenroll for other reasons. This may have resulted in a sample of less severe patients with CML. However, we assumed the vast majority of patients excluded due to lack of continuous enrollment to have disenrolled due to the change of insurance and not due to death; studies have reported 5-year OS rates in patients with CML treated with TKIs to be between 90% and 96%.25,26
While the data we used were as up-to-date as possible (data through October 2018), we did not study the effect of drug price on adherence or switching. The initial entry of generic imatinib led to moderate price reductions, as generic imatinib was priced 30% below the brand-name price when first introduced.27 Based on Redbook pricing reported in November 2018, Shih et al note that mean price of all generic imatinib formulations in the US to be $35,000 (the lowest Redbook listed price was reported at $4400).28 More recently, prices have decreased further as more generics have entered the market. In fact, National Average Drug Acquisition Cost (NADAC) weekly reference data reported a NADAC value of $21.03 for a 400-mg tablet of imatinib on September 26, 2018 dropped to $10.81 by May 22, 2019.29 With these changes in mind and more time having elapsed following pricing changes, the impact of price on adherence and switching should be investigated as data become available. Finally, these results were most applicable to a managed care population and may not be generalizable to all populations as financial incentives, which may affect treatment decisions and adherence, may be different between populations (eg, uninsured, Medicaid, Medicare fee-for-service). For example, we found no impact of generic entry on medication adherence in a patient with relatively good insurance resulting in limited out-of-pocket exposure. A study of patients with no or less generous insurance may find lower cost TKIs are associated with improved adherence and generic use.
Conclusion
The availability of a generic has not resulted in a shift away from first-line use of second-generation TKIs, rather in contrary, use has increased over time. As noted by others, price increases of TKIs in the United States, including branded imatinib,30 have been significant, but this did not appear to have affected adherence rates during the period we examined. Second-generation TKIs were the most commonly used agents by the end of the first year of treatment due to an increasing proportion of patients receiving dasatinib or nilotinib as first line and the 20% switch rate in the first year (the majority from imatinib to a second-generation TKI). Among those who switched, we found a lack of molecular monitoring that would be utilized to inform this decision. While it is possible that adherence and switching rates will change as the number of generic manufactures of imatinib increases, the apparent preference for newer TKIs may nullify any effect of lower imatinib pricing.
Download the PDF to view supplementary materials
References
1. Hehlmann R, Berger U, Hochhaus A. Chronic myeloid leukemia: a model for oncology. Ann Hematol. 2005;84(8):487-497. doi:10.1007/s00277-005-1039-z
2. Sprycel (dasatinib) tablet. Prescribing information. Bristol-Myers Squibb Company; 2006. Accessed June 18, 2020. https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/021986s021lbl.pdf
3. Tasigna (nilotinib) capsules. Prescribing information. Novartis; 2001. Accessed June 18, 2020. https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/021588s053lbl.pdf
4. ICLUSIG (ponatinib) tablets. Prescribing information. ARIAD Pharmaceuticals, Inc; 2012. Accessed June 18, 2020. https://www.accessdata.fda.gov/drugsatfda_docs/label/2016/203469s022lbl.pdf
5. BOSULIF (bosutinib) tablets. Prescribing information. Pfizer; 2012. Accessed June 18, 2020. https://www.accessdata.fda.gov/drugsatfda_docs/label/2017/203341s009lbl.pdf
6. Noens L, van Lierde MA, De Bock R, et al. Prevalence, determinants, and outcomes of nonadherence to imatinib therapy in patients with chronic myeloid leukemia: the ADAGIO study. Blood. 2009;113(22):5401-5411. doi:10.1182/blood-2008-12-196543
7. Marin D, Bazeos A, Mahon FX, et al. Adherence is the critical factor for achieving molecular responses in patients with chronic myeloid leukemia who achieve complete cytogenetic responses on imatinib. J Clin Oncol. 2010;28(14):2381-2388. doi:10.1200/JCO.2009.26.3087
8. Ibrahim AR, Eliasson L, Apperley JF, et al. Poor adherence is the main reason for loss of CCyR and imatinib failure for chronic myeloid leukemia patients on long-term therapy. Blood. 2011;117(14):3733-3736. doi:10.1182/blood-2010-10-309807
9. de Almeida MH, Pagnano KB, Vigorito AC, Lorand-Metze I, de Souza CA. Adherence to tyrosine kinase inhibitor therapy for chronic myeloid leukemia: a brazilian single-center cohort. Acta Haematol. 2013;130(1):16-22. doi:10.1159/000345722
10. Jabbour E. Chronic myeloid leukemia: First-line drug of choice. Am J Hematol. 2016;91(1):59-66. doi:10.1002/ajh.24249
11. Cortes J, Saglio G, Baccarani M, et al. Final study results of the phase 3 dasatinib versus imatinib in newly diagnosed chronic myeloid leukemia in chronic phase (CML-CP) trial (DASISION, CA180-056) [abstract]. Blood. 2014;124(21):152. doi:10.1182/blood.V124.21.152.152
12. Larson RA, Kim DW, Jootar S, et al. ENESTnd 5-year (y) update: Long-term outcomes of patients (pts) with chronic myeloid leukemia in chronic phase (CML-CP) treated with frontline nilotinib (NIL) versus imatinib (IM) [abstract]. J Clin Oncol. 2014;32:7073.
13. Cortes J, Kantarjian H. Chronic myeloid leukemia: sequencing of TKI therapies. Hematology Am Soc Hematol Educ Program. 2016;2016(1):164-169. doi:10.1182/asheducation-2016.1.164
14. Chronic Myeloid Leukemia – v.1.2006. National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology. Accessed June 18, 2020.
15. Guérin A, Chen L, Dea K, Wu EQ, Goldberg SL. Economic benefits of adequate molecular monitoring in patients with chronic myelogenous leukemia. J Med Econ. 2014;17(2):89-98. doi:10.3111/13696998.2013.862251
16. Goldberg SL, Chen L, Guerin A, et al. Association between molecular monitoring and long-term outcomes in chronic myelogenous leukemia patients treated with first line imatinib. Curr Med Res Opin. 2013;29(9):1075-1082. doi:10.1185/03007995.2013.812034
17. Latremouille-Viau D, Guerin A, Gagnon-Sanschagrin P, Dea K, Cohen BG, Joseph GJ. Health care resource utilization and costs in patients with chronic myeloid leukemia with better adherence to tyrosine kinase inhibitors and increased molecular monitoring frequency. J Manag Care Spec Pharm. 2017;23(2):214-224. doi:10.18553/jmcp.2017.23.2.214
18. Wallace PJ, Shah ND, Dennen T, Bleicher PA, Crown WH. Optum Labs: building a novel node in the learning health care system. Health Aff (Millwood). 2014;33(7):1187-1194. doi:10.1377/hlthaff.2014.0038
19. Quan H, Li B, Couris CM, et al. Updating and validating the charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173(6):676-682. doi:10.1093/aje/kwq433
20. Chang S, Long SR, Kutikova L, Bowman L, Crown WH, Lyman GH. Burden of pancreatic cancer and disease progression: economic analysis in the US. Oncology. 2006;70(1):71-80. doi:10.1159/000091312
21. Thomas SK, Brooks SE, Mullins CD, Baquet CR, Merchant S. Use of ICD-9 coding as a proxy for stage of disease in lung cancer. Pharmacoepidemiol Drug Saf. 2002;11(8):709-713. doi:10.1002/pds.759
22. Darkow T, Henk HJ, Thomas SK, et al. Treatment interruptions and non-adherence with imatinib and associated healthcare costs: a retrospective analysis among managed care patients with chronic myelogenous leukemia. Pharmacoeconomics. 2007;25(6):481-496. doi:10.2165/00019053-200725060-00004
23. Banegas MP, Rivera DR, O’Keeffe-Rosetti MC, et al. Long-term patterns of oral anticancer agent adoption, duration, and switching in patients with CML. J Natl Compr Canc Netw. 2019;17(10):1166-1172. doi:10.6004/jnccn.2019.7303
24. Chen L, Guérin A, Xie J, et al. Monitoring and switching patterns of patients with chronic myeloid leukemia treated with imatinib in community settings: a chart review analysis. Curr Med Res Opin. 2012;28(11):1831-1839. doi:10.1185/03007995.2012.741577
25. Cortes JE, Saglio G, Kantarjian HM, et al. Final 5-year study results of DASISION: the dasatinib versus imatinib study in treatment-naïve chronic myeloid leukemia patients trial. J Clin Oncol. 2016;34(20):2333-2340. doi:10.1200/JCO.2015.64.8899
26. Hochhaus A, Saglio G, Hughes TP, et al. Long-term benefits and risks of frontline nilotinib vs imatinib for chronic myeloid leukemia in chronic phase: 5-year update of the randomized ENESTnd trial. Leukemia. 2016;30(5):1044-1054. doi:10.1038/leu.2016.5
27. Miller J. Sun hopes cut-price generic Gleevec will win third of U.S. market. Reuters. February 1, 2016. Accessed June 18, 2020. www.reuters.com/article/sun-gleevec-genericdrugs-idUSKCN0VA3M9
28. Shih YT, Cortes JE, Kantarjian HM. Treatment value of second-generation BCR-ABL1 tyrosine kinase inhibitors compared with imatinib to achieve treatment-free remission in patients with chronic myeloid leukaemia: a modelling study. Lancet Haematol. 2019;6(8):e398-e408. doi:10.1016/S2352-3026(19)30087-0
29. National Average Drug Acquisition Cost (NADAC) weekly reference data. Medicaid.gov. Updated May 22, 2018. Accessed June 18, 2020. https://www.medicaid.gov/medicaid/prescription-drugs/pharmacy-pricing/index.html
30. Chen CT, Kesselheim AS. Journey of generic imatinib: a case study in oncology drug pricing. J Oncol Pract. 2017;13(6):352-355. doi:10.1200/JOP.2016.019737
