Abstract
Introduction
Lenalidomide, pomalidomide, and thalidomide are effective treatments for multiple myeloma but are teratogenic. To mitigate this risk, the US Food and Drug Administration (FDA) required risk evaluation and mitigation strategy (REMS) programs for these drugs, which include pregnancy testing among women of childbearing potential—twice before initiation, weekly in the first month on treatment, and every 2–4 weeks thereafter.
Objective
We evaluated dispensing trends of lenalidomide, pomalidomide, and thalidomide and assessed adherence to REMS pregnancy testing requirements among at-risk patients taking these drugs.
Methods
Using three US health insurance claims databases (Optum Clinformatics® [2004–2020], Merative Marketscan [2003–2019], and Medicaid [2000–2018]), we assessed monthly use of the drugs, patient characteristics and treatment persistence among drug initiators, and claims-based evidence for adherence to pregnancy testing requirements among initiators with child-bearing potential.
Results
Lenalidomide was the most prescribed agent following its approval in 2006 and through the end of the study period. A total of 48,311 lenalidomide (mean age = 59 years [standard deviation (SD) = 16]), 17,550 thalidomide (mean age = 65 years [SD = 12]), and 6560 pomalidomide initiators (mean age = 65 years [SD = 11]) were identified; 45% of initiators of each drug were women. Among initiators under follow-up on day 90, 70% were still on therapy. Initiators of childbearing potential comprised 3% (N = 1,920) of all initiators; among this cohort, 12% had evidence in claims data of two pregnancy tests before initiation, and 9% with at least 33 days of follow-up of four tests during the first month of treatment. By contrast, 52% who received a refill had claims-based evidence of a pregnancy test within 7 days of dispensing.
Conclusion
Although most patients who initiated lenalidomide, pomalidomide, and thalidomide were not of child-bearing potential, further investigation into actual non-adherence to pregnancy testing is needed.
1. Introduction
Multiple myeloma (MM) is a neoplasm of plasma cells that accounts for approximately 2% of all cancers and 17% of all hematologic malignancies in the USA [1, 2]. Each year, over 32,000 patients are diagnosed with MM, and almost 13,000 patients die from the disease [3]. However, the introduction of novel therapies in clinical practice has substantially improved overall survival of patients with MM, with many patients now living with the disease for more than 10 years [4]. Immunomodulatory drugs—including thalidomide, lenalidomide, and pomalidomide—are an important backbone of MM therapy. Thalidomide was developed in the 1950s as a mild sedative but was withdrawn from the market in Europe and not approved in the USA for this indication after the discovery of its teratogenic effects [5, 6]. Decades later, in June 1998, it was approved in the USA for the treatment of severe leprosy and in May 2006 for MM. Lenalidomide was approved for MM in June 2006 and pomalidomide in February 2013.
In the early 1960s, the thalidomide case inspired regulatory reform to strengthen US Food and Drug Administration (FDA) preapproval testing requirements, as well as systems to prevent fetal exposure to teratogenic drugs [7–9]. FDA regulatory authority relating to drug safety has been augmented numerous times since then, including by the FDA Amendments Act (FDAAA) of 2007, which improved prescription drug safety regulation following several instances in which drugs were removed from the market owing to serious adverse events [10]. One key provision of FDAAA granted FDA the authority to mandate that manufacturers implement risk evaluation and mitigation strategy (REMS) programs for certain drugs with serious adverse effects. A REMS program can entail a Medication Guide to highlight appropriate use to patients or can involve more complex “elements to assure safe use”, such as regular laboratory testing, for drugs with the most serious safety concerns. The main goal of the REMS program is to ensure that the benefits of drug use outweigh the risks.
Thalidomide, lenalidomide, and pomalidomide have had safety programs intended to prevent embryo-fetal exposure since their approval. For thalidomide and lenalidomide, these programs were initiated prior to FDAAA and implemented by the drugs’ manufacturer. They required prescriber certification and patient counseling, both of which included attesting to knowledge of drug risks and safe use requirements. For women of reproductive potential—we use the term “woman” to refer to biological sex, and it should be taken to include all people who can become pregnant or give birth—the programs mandated the use of two forms of contraception absent abstinence and pregnancy testing prior to, during, and after initiation.
These voluntary programs were converted to mandatory REMS programs after FDAAA. In 2010, the FDA approved slightly modified versions of the REMS programs that maintained their key elements, and in 2013 approved a similar REMS program for pomalidomide upon its marketing approval.
There are limited data on how REMS programs affect drug access and adherence to testing to ensure safe use, with prior studies primarily focused on patient knowledge and comprehension [11] or manufacturer’s perspectives of and experience with managing pregnancy prevention programs for thalidomide and lenalidomide [12]. There are also limited data on the relative use of the drugs over time. Therefore, the objectives of our study were to assess: (1) dispensing trends of thalidomide, lenalidomide, and pomalidomide; (2) drug initiation and medication persistence; and (3) claims-based evidence of adherence to REMS-required pregnancy testing among women of child-bearing potential [13].
2. Methods
2.1. Data Source
We used longitudinal administrative claims data from three large US health insurance databases—two from commercial insurers (Optum’s de-identified Clinformatics® Data Mart Database [2004–2020] and Merative Marketscan [2003–2019]) and one from a public insurance program (Medicaid [2000–2018]). The study period varied for the three drugs based on the dates of their original approvals and the availability of data (Online Supplementary Material (OSM) Appendix Table 1). All three databases contain demographic information on insurance enrollees and longitudinal information on their billed pharmacy use and clinical encounters (health insurance claims). Each database covers large, geographically diverse populations, and has been extensively used in pharmacoepidemiology research, including studies of drug use, adherence, and health outcomes [14]. In this study, we excluded Medicaid enrollees 65 years of age and older, as they qualified for Medicare and might have been missing claims. In addition, we required eligible Medicaid enrollees to be enrolled in fee-for-service plans or managed care plans with evidence of complete claims and dispensing data. This study was conducted using de-identified data and approved by the Mass General Brigham Institutional Review Board.
2.2. Trends in Monthly Drug Use
For each of the three study drugs, we evaluated changes in drug use over time in each database. Beginning on the drug-specific study start date (OSM Appendix Table 1), we identified the number of enrollees who had continuous enrollment in medical and pharmacy plans for at least 28 days of each calendar month, and among them, the number of enrollees who had at least one dispensing claim of a study drug. Monthly drug use, therefore, was defined as the number of individuals with at least one outpatient dispensing per million eligible enrollees in each month.
We plotted monthly drug use over time and in relation to regulatory events that occurred during the study period. These regulatory events included approval of supplemental indications, addition of boxed warnings, approval of the first generic version of the drug, and substantial changes to the REMS programs if they occurred (see detailed description of regulatory changes in OSM Appendix Table 2). Of note, pomalidomide had a REMS program in place at the time of its approval in February 2013. For thalidomide and lenalidomide, REMS programs were introduced in August 2010 but did not differ substantially from safety programs that existed beforehand. There were no changes to REMS pregnancy testing requirements for any of the study drugs during the study period.
The trends in drug use are presented in each database separately as the underlying population of insurance enrollees and study periods differ across the databases.
2.3. Initiator Characteristics
2.3.1. Initiator Cohort
To evaluate patient characteristics, treatment persistence, and claims-based evidence of adherence to REMS requirements, we identified a cohort of initiators within each database. Initiators were individuals with a drug dispensing with at least 6 months of insurance enrollment and no dispensing of the drug of interest over that period (baseline period). Individuals were allowed to use other immunomodulating therapies during the baseline period and were allowed to enter the cohort more than once if all inclusion and exclusion criteria were satisfied (OSM Appendix Fig. 1).
2.3.2. Patient Characteristics
Patient demographics, including age, sex, and region of residence, were assessed on the initiation date (index date). Other patient characteristics were assessed during the baseline period, including the presence of codes for MM and other FDA-approved indications for each study drug and use of other immunomodulating therapies and other common cancer therapies for MM, such as rituximab and bortezomib. Comorbid conditions considered included acute and chronic kidney disease, autoimmune diseases, deep vein thrombosis/pulmonary embolism, human immunodeficiency virus disease, ischemic heart disease, liver disease, and any diseases of skin, subcutaneous, musculoskeletal, or connective tissues. In addition, we assessed the burden of comorbidities, using a combined comorbidity score [15, 16]. The use of prescription medications such as bisphosphonates, opioids, and oral contraceptives, and healthcare use measures such as the number of hospitalizations, emergency room visits, and prescriptions were also evaluated during the baseline period. Patient characteristics were evaluated within each database and in the combined cohort of initiators, pooled across the three databases.
2.3.3. Treatment Persistence
Persistence was evaluated on days 90, 180, and 365 following initiation and was defined as continuous use of the drug among study individuals who were still under follow-up on the day of evaluation. Follow-up started on the day after initiation and ended on admission to a nursing facility or hospice, death, end of insurance enrollment, or the end of the study period (OSM Appendix Table 1 and Appendix Fig. 1). Patients were deemed to have stopped treatment if they did not receive a subsequent dispensing within 30 days following the end of days’ supply of a previous dispensing. Similar to patient characteristics, persistence was assessed within each database and in the combined cohort, pooled across the three databases. Additionally, for each drug, we assessed persistence among three sub-cohorts of initiators: (i) women of childbearing potential (defined below), (ii) all other women, and (iii) all individuals (men and women) who do not have childbearing potential. Finally, since the availability and use of the study drugs changed over time, we also assessed 90-day and 180-day treatment persistence for each drug in each year.
2.4. Evidence of Adherence to Pregnancy Testing
To evaluate evidence of adherence to the pregnancy testing schedule specified in the REMS programs, we created a sub-cohort of initiators comprising women of childbearing potential, defined as women aged 14–45 years on the index date, with no claims covering menopause or sterilization procedures any time prior to cohort entry based on all available data (these patients were censored during follow-up upon the occurrence of a sterilization procedure). We restricted the study period to index dates on or before 31 December 2019, as the FDA exercised enforcement discretion for REMS testing in 2020 because of the COVID-19 pandemic. Due to small numbers, women of childbearing potential were pooled across the three databases and adherence was assessed only in the combined cohort.
Under the REMS programs, pregnancy testing must occur within 10–14 days prior to the first prescription of thalidomide, lenalidomide, or pomalidomide. A second pregnancy test must be performed 24 h prior to the first prescription. While on therapy, pregnancy testing must occur weekly during the first 4 weeks and depending on whether the patient has regular or irregular menstrual cycles, every 2–4 weeks thereafter. In addition, a prescription can only be dispensed by the pharmacy within 7 days from the date of a negative pregnancy test. The REMS programs do not specify the type of pregnancy test that must be conducted.
We used Current Procedural Terminology (CPT) codes, revenue codes, and International Classification of Diseases (ICD)-9 and ICD-10 diagnosis codes to evaluate claims-based evidence for adherence to REMS pregnancy testing requirements (OSM Appendix Table 3). We assessed the proportion of eligible women of childbearing potential who had at least two tests within 14 days prior to and including the date of initiation. As a sensitivity analysis, we evaluated how many eligible women had at least one pregnancy test during the 14 days prior to and including the date of initiation. To evaluate weekly pregnancy testing on therapy, we assessed the number of eligible women with at least four tests during the first 33 days of follow-up among eligible women with at least 33 days of follow-up. In addition, among eligible women who had at least one refill of the study medication, we evaluated the number of women who had at least one pregnancy test within 7 days prior to or on the dispensing date of that first refill. Finally, since there could be time trends in how pregnancy tests were coded or billed over time, and the use of the drugs varied across time, we conducted a sensitivity analysis that limited the study years to 2013–2020, when all three study drugs were available.
3. Results
3.1. Trends in Drug Use
Figures 1A–C show the use of the three drugs over time relative to regulatory events that occurred during the study period, including supplemental indications, changes to boxed warnings, and approval of the first generic version. Use patterns were similar across the three databases. Thalidomide was the only drug available before 2006; its use decreased following the approval of lenalidomide. The use of lenalidomide increased substantially following its approval. In Clinformatics®, lenalidomide became more commonly used than thalidomide less than 2 years following its approval for MM, and its use increased throughout the study period, tripling from 59 to 183 users per 1,000,000 eligible enrollees between December 2010 and December 2020 (Fig. 1A). Among Marketscan (Fig. 1B) and Medicaid (Fig. 1C) enrollees, lenalidomide use similarly increased following its approval; however, the rate of increase slowed following the approval of pomalidomide in 2013. Among MarketScan enrollees, lenalidomide use increased from 54 to 83 lenalidomide users per 1,000,000 eligible enrollees between December 2010 and December 2019 (Fig. 1B). In Medicaid, lenalidomide use increased from nine to 21 lenalidomide users per 1,000,000 eligible enrollees from December 2010 to December 2018 (Fig. 1C). We did not observe substantial changes in the comparative use of the study drugs in relation to other documented events.
Fig. 1.
A Drug use over time among Clinformatics® enrollees. B Drug use over time among Marketscan enrollees. C Drug use over time among Medicaid enrollees. In Fig. 1C, The y-axis is restricted to a maximum of 100 individuals per million using the drug per 1 million eligible enrollees. MM multiple myeloma, REMS risk evaluation and mitigation strategy, HSCT hematopoietic stem cell transplantation
3.2. Initiator Characteristics and Treatment Persistence
We identified 17,550 thalidomide initiators, 48,311 lenalidomide initiators, and 6,560 pomalidomide initiators across the databases (OSM Appendix Table 4). Table 1 presents the characteristics of the pooled cohort (for database-specific characteristics, see OSM Appendix Tables 5A–C). Because patients over the age of 65 years were excluded from Medicaid analyses, Medicaid initiators were younger than initiators in commercial insurance data. In the pooled cohort, thalidomide initiators had a mean age of 59 years (standard deviation (SD) = 16 years). By contrast, the mean ages of lenalidomide initiators and pomalidomide initiators were both 65 years (lenalidomide SD = 12 years; pomalidomide SD = 11 years). The sex composition was similar across the three drugs: 45–46% of initiators were women.
Table 1.
Patient characteristics for all three databases combined (Clinformatics®, Marketscan, and Medicaid)
| Characteristics | Thalidomide initiators (N = 17,550) | Lenalidomide initiators (N = 48,311) | Pomalidomide initiators (N = 6560) |
|---|---|---|---|
| Demographics | |||
| Age (in years); mean (SD) | 58.6 (16.4) | 64.9 (12.2) | 65.3 (11.3) |
| Age groups, years; N (%) | |||
| 0–17 | 654 (3.7) | 112 (0.2) | < 11* |
| 18–54 | 5161 (29.4) | 9010 (18.6) | 1000–1500* |
| 55–64 | 5750 (32.8) | 15,806 (32.7) | 2225 (33.9) |
| 65–74 | 3078 (17.5) | 11,816 (24.5) | 1719 (26.2) |
| 75+ | 2907 (16.6) | 11,567 (23.9) | 1528 (23.3) |
| Sex; N (%) | |||
| Male | 9546 (54.4) | 26,495 (54.8) | 3611 (55) |
| Female | 8004 (45.6) | 21,816 (45.2) | 2949 (45) |
| Region; N (%) | |||
| Midwest / North-Central | 4038 (23) | 11,530 (23.9) | 1302 (19.8) |
| Northeast | 2768 (15.8) | 8065 (16.7) | 1256 (19.1) |
| South | 6962 (39.7) | 18,424 (38.1) | 2591 (39.5) |
| West | 3727 (21.2) | 10,033 (20.8) | 1386 (21.1) |
| Insurance type; N (%) | |||
| Commercial insurance | 13,350 (76.1) | 42,910 (88.8) | 5934 (90.5) |
| Medicaid | 4200 (23.9) | 5401 (11.2) | 626 (9.5) |
| FDA-approved indications; N (%) | |||
| Multiple myeloma | 10,594 (60.4) | 39,161 (81.1) | 6415 (97.8) |
| Hematopoietic cell transplant | 1415 (8.1) | 7157 (14.8) | 1919 (29.3) |
| Other FDA approved indications** (lenalidomide only) | N/A | 6575 (13.6) | N/A |
| Kaposi sarcoma (pomalidomide only) | N/A | N/A | 12 (0.2) |
| Leprosy (thalidomide only) | 84 (0.5) | N/A | N/A |
| Other blood cancers; N (%) | |||
| Monoclonal gammopathy | 2480 (14.1) | 12,700 (26.3) | 1,349 (20.6) |
| Chronic lymphocytic leukemia | 127 (0.7) | 833 (1.7) | 36 (0.5) |
| Blood cancer (any) | 12125 (69.1) | 45,774 (94.7) | 6459 (98.5) |
| Prior use of cancer drugs (not including the index date); N (%) | |||
| Other immunomodulating agents | 1208 (6.9) | 2530 (5.2) | 3288 (50.1) |
| MM therapy*** | 3249 (18.5) | 14,595 (30.2) | 4160 (63.4) |
| Rituximab | 158 (0.9) | 1296 (2.7) | 13 (0.2) |
| Bortezomib | 1844 (10.5) | 12,159 (25.2) | 2177 (33.2) |
| Common lymphoma therapies**** (lenalidomide only) | N/A | 13,826 (28.6) | N/A |
| Other comorbidities; N (%) | |||
| Autoimmune diseases | 1671 (9.5) | 3890 (8.1) | 469 (7.1) |
| Prior DVT/PE | 1121 (6.4) | 3695 (7.6) | 808 (12.3) |
| Ischemic heart disease | 2113 (12) | 7605 (15.7) | 965 (14.7) |
| Acute kidney injury | 1965 (11.2) | 6997 (14.5) | 1051 (16) |
| Chronic kidney disease | 2301 (13.1) | 9911 (20.5) | 1711 (26.1) |
| Liver disease | 1691 (9.6) | 4226 (8.7) | 510 (7.8) |
| HIV diagnosis or medications | 316 (1.8) | 269 (0.6) | 41 (0.6) |
| Combined comorbidity score | 3.6 (3.1) | 3.9 (3.2) | 4.7 (3.6) |
| Use of other medications during baseline; N (%) | |||
| Oral contraceptives | 244 (1.4) | 270 (0.6) | 26 (0.4) |
| Bisphosphonates | 2755 (15.7) | 8818 (18.3) | 2484 (37.9) |
| Opioids | 11,138 (63.5) | 29,823 (61.7) | 4012 (61.2) |
| Healthcare utilization; Mean (SD) | |||
| Number of hospitalizations | 0.8 (1.3) | 0.6 (1) | 0.6 (1.1) |
| Number of ER visits | 0.9 (3.1) | 0.8 (1.9) | 0.8 (1.9) |
| Number of office visits | 10.2 (7.8) | 10.7 (6.9) | 12.8 (7.3) |
| Number of unique drugs dispensed | 12.3 (6.8) | 11.8 (6.1) | 13.3 (6.2) |
| Hospitalization in 30 days prior to initiation; N (%) | 3267 (18.6) | 6,130 (12.7) | 677 (10.3) |
DVT deep vein thrombosis, ER emergency room, MM multiple myeloma, PE pulmonary embolism, HIV human immunodeficiency virus
Specific counts could not be provided due to the CMS cell Suppression Policy, which does not allow for cell counts between 1 and 10 to be reported
Other FDA approved indications for lenalidomide: Mantle cell lymphoma, follicular (nodular) lymphoma, marginal cell lymphoma, myelodys-plastic syndrome
Common multiple myeloma therapies: bortezomib, carfilzomib, cyclophosphamide, daratumumab, elotuzumab, isatuximab, ixazomib, and melphalan
Common lymphoma therapies: acalabrutinib, bendamustine, bortezomib, cisplatin, cytarabine, cyclophosphamide, doxorubicin, fludarabine, ibrutinib, idelalisib, vincristine, and vornistat
A diagnosis of MM was recorded for 60% (N = 10,594) of thalidomide initiators, 81% (N = 39,161) of lenalidomide initiators, and 98% (N = 6415) of pomalidomide initiators. Only a small number of thalidomide initiators (0.5%) had a leprosy diagnosis. A third of pomalidomide initiators had undergone hematopoietic stem cell transplantation during the 6 months prior to initiation compared to 15% of lenalidomide initiators and 8% of thalidomide initiators. Consistent with pomalidomide’s labeling, half of pomalidomide initiators had been previously treated with lenalidomide or thalidomide within 6 months prior to initiating pomalidomide, and 63% had received other MM therapy. The proportion of lenalidomide and thalidomide initiators who had used either other immunomodulators or other MM therapies during baseline was much lower. The mean comorbidity score was higher for pomalidomide initiators (4.7 [SD = 3.6]) than thalidomide initiators (3.6 [SD = 3.1]) and lenalidomide initiators (3.9 [SD = 3.2]) (Table 1).
Among patients who remained under follow-up on day 90 in our cohort (thalidomide = 14,871, lenalidomide = 40,776, pomalidomide = 5,116), persistence to the study drug was higher among lenalidomide (74%) and pomalidomide (74%) initiators than among thalidomide initiators (60%). At 180 days of follow-up, 44% of lenalidomide initiators, 51% of pomalidomide initiators, and 34% of thalidomide initiators were still on the drug (Table 2). At 1 year, these percentages were 24%, 29%, and 16%, respectively. Database-specific results for persistence for each drug are provided in OSM Appendix Table 6.
Table 2.
Treatment persistence among patients who initiated therapy
| Measure | Thalidomide (N = 17,550) | Lenalidomide (N = 48,311) | Pomalidomide (N = 6560) |
|---|---|---|---|
| Persistence measures | N (%) | N (%) | N (%) |
| Day 90 following treatment initiation | |||
| Patients under follow-up | 14,871 (84.7) | 40,776 (84.4) | 5116 (78.0) |
| Patients on treatment (% of under follow-up) | 8877 (59.7) | 30,135 (73.9) | 3801 (74.3) |
| Day 180 following treatment initiation | |||
| Patients under follow-up | 12,483 (71.1) | 34,774 (72.0) | 4016 (61.2) |
| Patients on treatment (% of under follow-up) | 4262 (34.1) | 15,116 (43.5) | 2027 (50.5) |
| Day 365 following treatment initiation | |||
| Patients under follow-up | 9474 (54.0) | 26,061 (53.9) | 2543 (38.8) |
| Patients on treatment (% of under follow-up) | 1522 (16.1) | 6179 (23.7) | 742 (29.2) |
For all three drugs, persistence at 90 days among the three sub-cohorts—women of childbearing potential, all other women, and all individuals who do not have childbearing potential—was comparable to persistence in the full cohort. At 180 days of follow-up, persistence among women of childbearing potential was lower compared to the full cohort for thalidomide (25% vs. 34%) and lenalidomide (38% vs. 44%) initiators. When stratified by calendar year, persistence at 90 days and 180 days increased over time for all three drugs, albeit more steadily for lenalidomide and pomalidomide than for thalidomide (OSM Appendix Figs. 2(a) and 2(b)). Additional persistence results are provided in OSM Appendix Table 8(i)–(iii).
3.3. Evidence of Adherence to Pregnancy Testing
The cohort of women of child-bearing potential who would be required to follow REMS pregnancy testing requirements comprised 1,920 women (3% of initiators): 1,004 initiators of thalidomide, 852 initiators of lenalidomide, and 64 initiators of pomalidomide (patient flowchart in OSM Appendix Table 7). Within this sub-cohort, 6% of thalidomide initiators, 19% of lenalidomide initiators, and 17% of pomalidomide initiators had claims for at least two pregnancy tests ordered within 14 days prior to initiation. The percentages increased to 42% for thalidomide, 62% for lenalidomide, and 66% for pomalidomide when restricted to one pregnancy test.
During the first month of treatment, 9% of 1,359 initiators with at least 33 days of follow-up, had claims for four ordered pregnancy tests. For each of the three drugs, there were more women with a claim for at least one pregnancy test ordered within 7 days prior to the dispensing of the first refill: 41% for thalidomide, 61% for lenalidomide, and 58% for pomalidomide (Table 3). In the sensitivity analysis limited to the years 2013–2020, claims-based evidence of pregnancy testing among thalidomide initiators increased and the difference between thalidomide and the other two agents was reduced (OSM Appendix Table 9).
Table 3.
Claims-based evidence of adherence to pregnancy testing* among women of childbearing potential
| Adherence measure | Thalidomide N (%) | Lenalidomide N (%) | Pomalidomide N (%) |
|---|---|---|---|
| Women of childbearing potential, N | 1004 | 852 | 64 |
| Testing during baseline | |||
| (1a) At least two tests within 14 days prior to and including initiation date | 62 (6.2) | 163 (19.1) | 11 (17.2) |
| (1b) At least one test within 14 days prior to and including initiation date | 428 (42.6) | 529 (62.1) | 42 (65.6) |
| Testing during therapy | |||
| Women of childbearing potential who had at least 33 days of follow-up, N | 637 | 672 | 50 |
| (2) At least four tests within the first month of treatment (33 days after initiation date) | 38 (6.0) | 83 (12.3) | < 11 |
| Women of childbearing potential with at least one dispensing after initiation, N | 622 | 661 | 50 |
| (3) At least one test 7 days prior to and including the first refill dispensing date | 259 (41.6) | 405 (61.3) | 29 (58.0) |
Defined by having an occurrence of a CPT code or ICD-9/ICD-10 diagnosis codes for a pregnancy test (Online Supplemental Material Appendix Table 3)
Exact counts could not be provided due to the CMS cell Suppression Policy, which does not allow for cell counts between 1 and 10 to be reported
4. Discussion
In this large study comprising commercially and publicly insured individuals, we found that the use of lenalidomide increased rapidly upon market entry, quickly surpassing that of thalidomide, and continued to increase after the approval of pomalidomide. Changes in the use of the drugs seemed to be driven more by the approval of new agents in the therapeutic class than by drug-specific regulatory events. For each drug, most patients remained on treatment 3 months after initiation, but persistence was higher for lenalidomide and pomalidomide than for thalidomide. Only a small proportion of thalidomide, lenalidomide, and pomalidomide initiators were women of childbearing potential. For thalidomide and lenalidomide, persistence was lower among this group than among other initiators. Additionally, in most women of childbearing potential, we did not observe claims-based evidence of complete adherence to REMS pregnancy testing requirements. However, there were a higher number of pregnancy testing claims in lenalidomide and pomalidomide initiators than in thalidomide initiators, and for pregnancy tests that were part of a dispensing requirement across all three drugs.
Several factors may have contributed to the increasing use of lenalidomide following pomalidomide approval and the higher persistence of lenalidomide and pomalidomide initiators compared to thalidomide initiators. As pomalidomide was approved as a third in class treatment for MM [17], most prescribers likely continued to use lenalidomide as an initial treatment. Additionally, over the study period, lenalidomide was increasingly used as a maintenance treatment [18] and a treatment for smoldering myeloma [19]. Prescribers’ familiarity and comfort with lenalidomide likely also played a role. Meanwhile, lenalidomide and pomalidomide have lower toxicity than thalidomide [17, 20].
Similar to different use patterns, we observed some differences in treatment persistence across the three drugs. Once initiated, most patients continued to take these drugs 3 months after initiation. However, treatment persistence was higher among lenalidomide and pomalidomide initiators than among thalidomide initiators, particularly at 6 months to 1 year post initiation. In addition, women of childbearing potential had lower persistence for both thalidomide and lenalidomide at 6 months post-initiation compared to the full study cohort. This could provide some evidence of the burden of the REMS pregnancy testing requirements, though other reasons, such as differences in indications, planned duration of treatment, and persistence by calendar year could also explain these findings.
The limited use of all three drugs among women of childbearing potential was not surprising. MM, the primary reason for the use of thalidomide, lenalidomide, and pomalidomide in the USA, is a disease that tends to occur later in life [21]. The majority of patients initiating these therapies—both in our data and the broader MM population—are over 55 years of age [21]. A majority are also men [22].
The low evidence of adherence to REMS required pregnancy testing based on insurance claims may have a number of explanations. REMS programs have been shown to be effective in enhancing patient understanding of the fetal toxicity risk associated with therapies for MM [11]. Pregnancy testing complements this risk communication, helping to ensure safe use of the drugs. However, pregnancy testing is time- and resource-intensive for prescribers and patients [23], which may lead to suboptimal adherence. Certified prescribes must complete a survey to verify the patient’s reproductive status, negative pregnancy test, and completion of counseling. Subsequently, the prescriber receives an authorization number from the REMS program, and this number along with the patient’s risk category is documented on the prescription before being sent to the pharmacy. Women who can become pregnant must also complete a survey, and receive counseling from the pharmacist before a prescription can be filled.
Variances in pregnancy testing could be explained in part by the structure of the REMS programs and differences in drug marketing. In our study, claims for pregnancy testing were three times higher for required pregnancy testing within 7 days before a dispensing than for required weekly testing during the first month of treatment. Under the REMS program, lack of adherence to the former requirement would automatically preclude a patient from receiving the medication, while the latter would not. Additionally, among women of childbearing potential, adherence to required pregnancy testing was lower for thalidomide than lenalidomide or pomalidomide. However, thalidomide was introduced several years earlier than lenalidomide or pomalidomide when data capture might have been poorer. The narrowing of the difference in adherence between the drugs when restricted to the same period supports this explanation.
It is possible that pregnancy tests were not recorded in our claims data. Claims data reflect billing processes and thus do not capture clinical or testing information that does not generate a health insurance claim. The REMS programs for the MM drugs do not specify the type of pregnancy testing required or where it must be performed. Therefore, women of child-bearing potential could have been tested through means that were not billed to insurance (e.g., a home pregnancy test). While missingness due to unbilled tests cannot be excluded, we observed substantial difference in claims-based adherence to testing between required testing within 7 days prior to pharmacy dispensing and required weekly testing during the first month on treatment that does not impact access to the drug. It is still possible that some healthcare providers required a laboratory pregnancy test prior to initiation to satisfy pharmacy dispensing requirement and then allowed at-home pregnancy tests (which are not captured in our data) while the patient was under treatment. To fully account for the observed difference, however, this differential testing practice must have been common. A component of the observed difference may also be attributable to eligible women stopping the drug before a full month of treatment and, thus, not needing four pregnancy tests. Further research into provider and patient practices to testing in the real world would offer greater insight into adherence to pregnancy testing.
Our study was subject to some important limitations. First, as with all insurance claims-based studies, it should be noted that we used dispensed prescriptions as a surrogate for use of drugs and in some cases patients who receive a dispensed drug may not actually take the prescribed drug. Second, in the event of non-persistence, we could not ascertain the reasons patients stopped therapy. Third, the population in each database was not constant during the study period. For example, in the Clinformatics® database the proportion of individuals over the age of 65 years increased over time, which could potentially affect population-level trends in drug use. Fourth, we were not able to assess how REMS influenced the use of these drugs during the study period, as there were no substantive changes to REMS testing requirements during the study period, and all three drugs had either REMS or pre-REMS safety programs in place at the time of their approval. Finally, as described earlier, pregnancy tests that are not billed to insurance are not captured in claims data, which could result in an underestimation of adherence to testing requirements.
5. Conclusion
This study provides a multiyear evaluation of the use of the MM treatments thalidomide, lenalidomide, and pomalidomide, patient persistence with the medication, and claims-based evidence of adherence to REMS program pregnancy testing requirements. Lenalidomide was the most widely used drug, and the persistence of lenalidomide and pomalidomide initiators was higher than that of thalidomide initiators. For each of the three drugs, women of childbearing potential accounted for a very small fraction of initiators. More of them had claims for pregnancy testing required prior to pharmacy dispensing than for required testing that had no impact on access to the drug. The relative lack of claims for pregnancy testing requires further investigation to assess the full extent of non-adherence to pregnancy testing, which is designed to ensure that the risk of potential teratogenic adverse outcomes is minimized while ensuring fair access to these important drugs to patients who need them.
Supplementary Material
Key Points.
Thalidomide, lenalidomide, and pomalidomide are effective treatments in the same therapeutic class for multiple myeloma that have been subject to risk mitigation measures in the USA, including scheduled pregnancy testing among women of childbearing potential, due to their teratogenicity.
The use of lenalidomide rapidly surpassed that of thalidomide upon market entry and continued to increase following the introduction of pomalidomide.
Only a small proportion of initiators were women of childbearing potential; among this cohort, overall adherence to required pregnancy testing was low based on insurance claims of receipt of physician-ordered pregnancy tests but higher for pregnancy tests required prior to dispensing.
Funding Information
This work was funded by the Center for Drug Evaluation and Research at the US FDA (Contract number 75F40120C0044). Ameet Sarpatwari and Aaron S. Kesselheim’s work was also supported by grant funding from Arnold Ventures. Katsiaryna Bykov’s effort was also supported by the National Institutes of Health (K01AG068365).
Conflict of Interest Statement
Ameet Sarpatwari reports having served as an expert witness on REMS-related litigation concerning mifepristone. Katsiaryna Bykov reports consulting fees from Alosa Health for unrelated work. Mufaddal Mahesri, Krista F. Huybrechts, Joyce Lii, Su Been Lee, Gita A. Toyserkani, Cynthia LaCivita, Esther H. Zhou, Gerald J. Dal Pan, and Aaron S. Kesselheim have no conflicts of interest to report.
Footnotes
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s40264-024-01443-3.
Ethics Approval Statement The study was conducted using de-identified data and approved by the Mass General Brigham Institutional Review Board (IRB # 2020P002183).
Code Availability Statement Programming codes are available from the corresponding author upon request.
Consent to Participate Statement Not applicable.
Consent to Publish Statement Not applicable.
Data Availability Statement
Data supporting the findings of the study are available from the Centers of Medicare & Medicaid Services (CMS), Optum’s de-identified Clinformatics® Data Mart Database, and Merative Marketscan. Restrictions apply to the availability of these data, which were used under license for this study.
References
- 1.Rajkumar SV, Dimopoulos MA, Palumbo A, et al. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. Lancet Oncol. 2014;15:e538–48. [DOI] [PubMed] [Google Scholar]
- 2.Rajkumar SV, Kumar S. Multiple myeloma current treatment algorithms. Blood Cancer J. 2020;10:94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2021. CA Cancer J Clin. 2021;71:7–33. [DOI] [PubMed] [Google Scholar]
- 4.Therneau TM, Kyle RA, Melton LJ III, et al. Incidence of monoclonal gammopathy of undetermined significance and estimation of duration before first clinical recognition. Mayo Clin Proc. 2012;87:1071–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Somers GF. Pharmacological properties of thalidomide (a-phthalimido glutarimide), a new sedative. Br J Pharmacol Chemother. 1960;15:111–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Smithells RW, Newman CG. Recognition of thalidomide defects. J Med Genet. 1992;29:716–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Greene JA, Podolsky SH. Reform, regulation, and pharmaceuticals—the Kefauver-Harris Amendments at 50. N Engl J Med. 2012;367:1481–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Uhl K, Cox E, Rogan R, et al. Thalidomide use in the US: experience with pregnancy testing in the S.T.E.P.S. programme. Drug Saf. 2006;29:321–9. [DOI] [PubMed] [Google Scholar]
- 9.Castaneda CP, Zeldis JB, Freeman J, et al. RevAssist: a comprehensive risk minimization programme for preventing fetal exposure to lenalidomide. Drug Saf. 2008;31:743–52. [DOI] [PubMed] [Google Scholar]
- 10.Avorn J, Kesselheim A, Sarpatwari A. The FDA Amendments Act of 2007—assessing its effects a decade later. N Engl J Med. 2018;379(12):1097–9. [DOI] [PubMed] [Google Scholar]
- 11.Brandenburg NA, Bwire R, Freeman J, Houn F, Sheehan P, Zeldis JB. Effectiveness of Risk Evaluation and Mitigation Strategies (REMS) for lenalidomide and thalidomide: patient comprehension and knowledge retention. Drug Saf. 2017;40(4):333–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bwire R, Freeman J, Houn F. Managing the teratogenic risk of thalidomide and lenalidomide: an industry perspective. Expert Opin Drug Saf. 2011;10(1):3–8. [DOI] [PubMed] [Google Scholar]
- 13.Sarpatwari A, Mitra-Majumdar M, Bykov K, et al. A multi-modal approach to evaluate the impact of risk evaluation and mitigation strategy (REMS) programs. Drug Saf. 2021;44(7):743–51. [DOI] [PubMed] [Google Scholar]
- 14.Schneeweiss S, Avorn J. A review of uses of health care utilization databases for epidemiologic research on therapeutics. J Clin Epidemiol. 2005;58(4):323–37. [DOI] [PubMed] [Google Scholar]
- 15.Gagne JJ, Glynn RJ, Avorn J, Levin R, Schneeweiss S. A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sun JW, Rogers JR, Her Q, Welch EC, Panozzo CA, Toh S, Gagne JJ. Adaptation and validation of the combined comorbidity score for ICD-10-CM. Med Care. 2017;55(12):1046–51. [DOI] [PubMed] [Google Scholar]
- 17.Chanan-Khan AA, Swaika A, Paulus A, et al. Pomalidomide: the new immunomodulatory agent for the treatment of multiple myeloma. Blood Cancer J. 2013;3(9): e143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Alonso R, Cedena MT, Wong S, Shah N, Ríos-Tamayo R, Moraleda JM, et al. Prolonged lenalidomide maintenance therapy improves the depth of response in multiple myeloma. Blood Adv. 2020;4(10):2163–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rajkumar SV, Kumar S, Lonial S, Mateos MV. Smoldering multiple myeloma current treatment algorithms. Blood Cancer J. 2022;12(9):129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Luo J, Gagne JJ, Landon J, Avorn J, Kesselheim AS. Comparative effectiveness and safety of thalidomide and lenalidomide in patients with multiple myeloma in the United States of America: A population-based cohort study. Eur J Cancer. 2017;70:22–33. [DOI] [PubMed] [Google Scholar]
- 21.Palumbo A, Anderson K. Multiple myeloma. N Engl J Med. 2011;364(11):1046–60. [DOI] [PubMed] [Google Scholar]
- 22.Huber JH, Mengmeng J, Shih YH, Wang M, Colditz G, Chang SH. Disentangling age, gender, and racial/ethnic disparities in multiple myeloma burden: a modeling study. Nat Commun. 2023;14(1):5768. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sarpatwari A, Brown BL, McGraw SA, Dejene SZ, Abdurrob A, Santiago Ortiz AJ, Kesselheim AS. Patient and caregiver experiences with and perceptions of risk evaluation and mitigation strategy programs with elements to assure safe use. JAMA Netw Open. 2022;5(1): e2144386. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data supporting the findings of the study are available from the Centers of Medicare & Medicaid Services (CMS), Optum’s de-identified Clinformatics® Data Mart Database, and Merative Marketscan. Restrictions apply to the availability of these data, which were used under license for this study.