Background
Prevention of prenatal exposures to teratogenic drugs is a significant clinical and public health concern. With the enactment of the US Food and Drug Administration Amendments Act in 2007, the US Food and Drug Administration has begun to require manufacturers to implement Risk Evaluation and Mitigation Strategies to prevent prenatal exposures. Among 12 risk evaluation and mitigation strategy drugs, several had predecessor risk mitigation plans (eg, isotretinoin) and some were newly required (eg, mycophenolate). Only a small proportion of teratogenic drugs are currently subject to Risk Evaluation and Mitigation Strategies, and the extent of prenatal exposure to the universe of teratogenic drugs compared with drugs subject to Risk Evaluation and Mitigation Strategies is unknown. Moreover, the effectiveness of such advanced risk mitigation programs in preventing prenatal exposure is not clear.
Objective
This study aimed to characterize the epidemiology of prenatal exposures to definite and potential teratogens during the risk evaluation and mitigation strategy era.
Study Design
We constructed a time-series of pregnancies identified from a national private insurance claims database (IBM MarketScan) to estimate prenatal exposures to teratogenic drugs (2006–2017). Pregnancy outcomes, gestational age, and the onset of pregnancy were determined with previously validated algorithms. The Teratology Information Service and Clinical Pharmacology databases were used to identify drugs with definite (n=141) or potential (n=65) teratogenic effects, and drugs with debatable risks such as benzodiazepines, statins, tetracyclines, sex hormones, infertility treatments, and gonadotropin-releasing hormone analogs were excluded. We defined prenatal exposure as ≥1 prescription fill or medical encounter involving administration of drugs with a definite teratogenic risk (including 12 for which there is a “current or discontinued” risk evaluation and mitigation strategy) or a potential teratogenic risk. We evaluated secular trends and modeled the effects of age, preconception exposure, and state healthcare quality rankings on prenatal exposure, adjusting for demographic factors and clinical conditions.
Results
The cohort included 3,445,612 pregnancies (2,532,444 live deliveries). Prenatal exposures to definite teratogens decreased slightly during the study years from 1.86 to 1.24 per 100 pregnancies between 2006 and 2017, whereas exposure increased for potential teratogens from 3.40% to 5.33%. Prenatal exposure prevalences were higher during the first trimester and for pregnancies that ended in nonlive outcomes. Drugs subject to Risk Evaluation and Mitigation Strategies had low background utilization and contributed to a small proportion of prenatal exposures (15.1 per 100,000 pregnancies). We also observed fewer prenatal exposures to risk evaluation and mitigation strategy drugs among women of childbearing age who used these treatments (0.14% vs 0.36% for any definite teratogen). Age extremes and low state-level healthcare quality rankings were independent predictors of prenatal exposure.
Conclusion
Fetuses in more than 1 in 16 pregnancies continued to be exposed to teratogenic drugs during the past decade. Drugs with Risk Evaluation and Mitigation Strategies imposed a small burden of prenatal exposure because of the low background utilization rates and lower pregnancy prevalence among women of childbearing age who used these drugs. Although the declining exposure rates to teratogenic drugs with definite risk are encouraging, the rising prenatal exposure to drugs with potential risk calls for more assessments. Future research is needed to elucidate the health outcomes of fetuses exposed to potential risk drugs, understand the effectiveness of risk evaluation and mitigation strategy programs, and prioritize teratogenic drugs for advanced risk mitigation.
Introduction
Prenatal exposure to drugs with known teratogenic risks is a significant concern for regulatory agencies globally. Learning from the thalidomide tragedy in the 1960s, risk management strategies utilized by the US Food and Drug Administration (FDA) to mitigate prenatal exposure have evolved from package inserts implemented in the 1970s to medication guides and risk minimization action plans (RiskMAP) introduced in the 1990s. In an effort to provide more comprehensive information and avoid misinterpretation, the FDA modernized the format according to which pregnancy and lactation risks are communicated within the prescribing label, from designating letter categories (A-X) to a narrative. Furthermore, to strengthen regulatory efforts involving risk mitigation, the FDA Amendments Act (2007) authorized the agency to require companies to implement Risk Evaluation and Mitigation Strategies (REMSs) for known safety risks, including teratogenicity. , REMSs may include educational interventions, patient registries, and drug distribution restrictions to ensure appropriate prescribing and monitoring.
Why was this study conducted?
It is unknown whether prenatal exposure to teratogenic drugs has changed during the past decade since the United States law authorized the US Food and Drug Administration to implement advanced risk mitigation programs (ie, Risk Evaluation and Mitigation Strategies [REMSs]).
Key findings
Prenatal exposures to definite teratogens decreased slightly during the study years from 1.86 to 1.24 per 100 pregnancies between 2006 and 2017 (absolute difference, 0.62), whereas exposure increased for potential teratogens from 3.40% to 5.33% (absolute difference, 1.93). Drugs with REMSs contributed a small proportion to prenatal exposure (15.1 per 100,000 pregnancies).
What does this add to what is known?
Fetuses from >1 in 16 pregnancies continue to be exposed to definite or potential teratogenic drugs during pregnancy, and drugs with REMSs contribute only marginally to the overall prenatal exposure.
The FDA has approved 12 REMS programs to mitigate teratogenicity risk, including 6 programs with legacy risk management programs such as for isotretinoin; 2 of the programs have been discontinued, leaving 10 active programs. Several studies have focused on the effectiveness of individual REMS programs in terms of their effects on patient or provider risk awareness and adherence to recommended or required pregnancy tests and contraception use. A few studies have also evaluated the prevalence of prenatal exposure to specific drugs, including our recent study of mycophenolate, which showed that the REMSs had an insignificant effect on the risk for pregnancy during drug treatment. ,
For a sizable number of teratogenic drugs commonly utilized by women of childbearing age, mitigation of prenatal exposure relies solely on the information in the drug label and, in some instances, information in required medication guides, such as for valproate and topiramate. , In the present study, we aimed to quantify the prevalence of prenatal exposure to drugs with a definite or potential teratogenic risk, as defined in the Materials and Methods section, during the REMS era and stratify the prevalence by the presence of a REMS program. We also explored determinants of prenatal exposure to these medications.
Materials and Methods
Study design
We constructed a multiyear time series (2006–2017) to evaluate the prevalence of exposure to potentially harmful medications during pregnancy (teratogenic, fetotoxic, and embryotoxic, which are referred to as teratogenic drugs from here onward). We used the IBM MarketScan Commercial Claims Databases (2006–2018) that contained in- and outpatient encounter and outpatient pharmacy claims data and enrollment information for a national sample of individuals with employer-sponsored health insurance. The institutional review board of the University of Florida exempted the study because de-identified data were used.
Pregnancy identification
We employed a previously developed algorithm to identify pregnancy episodes from claims data, including both pregnancies ending in live births (full-term, preterm, postterm) and those ending in nonlive births (ectopic pregnancy termination, spontaneous and induced abortions, and stillbirths). , , This algorithm uses previously validated measures of pregnancy endpoints and estimates gestational age by calculating from the date of the last menstrual period. Technical details about this algorithm are provided in Appendix 1 , Supplemental Figure . We required continuous health plan enrollment from 90 days before conception until 30 days after pregnancy end date to capture drug exposures and the relevant information to calculate gestational age.
Drug exposure measurement
We used the Teratogen Information System (TERIS) and Clinical Pharmacology databases to create lists of definite and potential teratogens based on the available evidence for teratogenic risk. The final drug list included 206 unique generic drug names, with 141 of those designated as having a definite risk (including 12 drugs with current or discontinued REMS) and 65 drugs designated as having potential risk. Prominent drug classes in the definite risk group included agents acting on the renin-angiotensin system, certain antiepileptic agents, systemic antimycotics, and antineoplastics. Isotretinoin, sulfamethoxazole, and warfarin were also included. Drugs with potential risk included prescription nonsteroidal anti-inflammatory drugs (NSAIDs), systemic antivirals, beta-blockers, aminoglycosides, antiarrhythmics, and immunosuppressants, along with several individual agents (eg, duloxetine, gabapentin, teriflunomide). Details on the drug selection criteria and drug names are provided in Appendix 1 , Supplemental Tables 1 and 2 . Determination of drug utilization was based on outpatient pharmacy dispensing claims using the National Drug Codes and medical encounter claims using the Healthcare Common Procedure Coding System codes. Time of exposure was set based on the pharmacy dispensing or outpatient encounter administration date and determined for the following time windows: preconception (90 days before conception), trimester 1 (from conception to week 12), trimester 2 (week 13–26), trimester 3 (≥27 weeks), and any time during pregnancy. Although the time between last menstrual period and conception is typically counted in the first trimester, it was omitted from this analysis to focus on the time period that could lead to teratogenic effects.
Statistical analysis
We anchored each pregnancy to a calendar year based on the estimated conception date ( Appendix 1 , Supplemental Figure ). Data from the last year in the database (2018) were only used to identify pregnancy end dates and prenatal exposures for pregnancies that started in 2017. The annual prevalence of prenatal exposures was calculated separately for drugs with definite teratogenic risks and those with potential teratogenic risks as the proportion of pregnancies exposed to the relevant study drugs over the total number of observed pregnancies in each study year. Because of the small number of pregnancies exposed to REMS drugs, prevalence estimates were determined only for the preconception and overall pregnancy time windows (omitted trimester-specific estimates).
For the visual examination of exposure trends, we standardized the prevalence estimates based on drug utilization in the baseline year (2006) to control for factors that could alter the probability of prenatal drug exposure over the study years (eg, drug approvals or withdrawals, changes in health plan formularies, changes in the database). The standardized annual prevalence was calculated as follows:
observed prenatal teratogen exposure prevalence inYeark×teratogen exposure prevalence among women of childbearing age in2006teratogen exposure prevalence among women of childbearing age inYeark
We fitted trend lines across the annual prevalences using polynomial functions for illustrative purposes. For context, we also reported the annual prevalence of prenatal exposure among women of childbearing age who used these drugs in a given year.
We explored the association with 2 factors that may provide some insight into the context of prenatal exposure, including women’s age and healthcare quality rankings of the resident state. We assumed that women at both extremes of the age spectrum would be at the highest risk for unplanned pregnancy , ; thus, we expected a bimodal association between age and prenatal exposure if inadvertent exposure was an important mechanism. Second, we assumed that failure to consider teratogenic risk in treatment decisions would be the main underlying mechanism for prenatal exposure. Thus, we expected that states with poor overall healthcare quality would have higher prenatal exposure rates based on the observation that healthcare quality measures are typically correlated. The reference and categorization of state-level healthcare quality are provided in Appendix 1 , Supplemental Table 3 .
We used generalized linear mixed models to investigate these determinants of prenatal exposure to account for clustering introduced by women contributing multiple pregnancies. We adjusted this analysis for patient-level preconception exposure to teratogenic drugs, calendar year, and the number of inpatient admissions and outpatient visits in the 6 months preceding conception as a global measure of healthcare utilization. In an expanded analysis, we used the Clinical Classification System, developed by the Agency for Healthcare and Quality, to summarize women’s clinical history based on diagnoses provided during medical encounters within the same 6-month review period. Models were fitted separately for prenatal exposure to drugs with definite and those with potential teratogenic risks or with REMS.
Sensitivity analysis
The accuracy of the gestational age estimates varies across pregnancy endpoints, with >93% of live birth gestational age estimates being within 1 to 2 weeks of the clinical estimate of gestational age when validated against birth certificates, whereas nonlive birth episodes have shown modest accuracy (36%–77%) for gestational age estimates based on in vitro fertilization records. We therefore conducted a sensitivity analysis of the effect of misclassified gestational age estimates on prenatal exposure prevalences for terminations and stillbirths by varying the estimated conception date by 2-week increments (−2,+2,+4,+6,+8). This range was chosen because the estimated gestational age for abortions (most of the non-live deliveries) is 10 weeks, and gestational ages of −2 to +8 weeks will cover the most common scenarios that are encountered in clinical care. All data management and statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc, Cary, NC).
Results
The final cohort included 3,445,612 pregnancy episodes of which 2,532,444 (73.5%) episodes had live birth outcomes and 913,168 ended in abortions (elective or spontaneous), stillbirths, or ectopic pregnancy terminations. Most women were 25 to 39 years old (77.6%), employees (51.3%), and resided in the Southern region of the United States (39.8%) ( Table 1 and Appendix 1 , Supplemental Tables 4 and 5 ).
| Characteristic a | Preconception | Trimester 1 | During pregnancy | |||
|---|---|---|---|---|---|---|
| Unexposed (n=3,364,812) | Exposed (n=80,800) | Unexposed (n=3,412,129) | Exposed (n=33,483) | Unexposed (n=3,386,802) | Exposed (n=58,810) | |
| Age | ||||||
| <20 | 4.4 | 7.8 | 4.3 | 8.2 | 4.3 | 7.6 |
| 20–24 | 12.4 | 17.0 | 12.5 | 16.6 | 12.5 | 16.1 |
| 25–29 | 27.0 | 26.4 | 27.1 | 24.5 | 27.1 | 24.6 |
| 30–34 | 32.9 | 27.7 | 32.8 | 26.0 | 32.9 | 25.8 |
| 35–39 | 18.0 | 15.2 | 17.9 | 16.2 | 17.9 | 15.6 |
| 40–44 | 4.7 | 4.4 | 4.7 | 5.7 | 4.7 | 5.7 |
| ≥45 | 0.7 | 1.4 | 0.7 | 2.8 | 0.7 | 4.7 |
| Region | ||||||
| Northeast | 17.2 | 14.3 | 17.2 | 14.8 | 17.2 | 13.3 |
| North Central | 22.5 | 22.9 | 22.5 | 21.7 | 22.5 | 21.4 |
| South | 39.6 | 45.2 | 39.7 | 46.3 | 39.6 | 49.2 |
| West | 19.6 | 16.5 | 19.5 | 16.1 | 19.6 | 15.0 |
| Unknown | 1.0 | 1.1 | 1.0 | 1.1 | 1.0 | 1.0 |
| State-level healthcare quality ranking | ||||||
| High rank | 16.8 | 14.2 | 16.7 | 13.7 | 16.8 | 12.7 |
| Middle rank | 67.1 | 67.0 | 67.1 | 66.9 | 67.1 | 66.8 |
| Low rank | 16.1 | 18.8 | 16.2 | 19.5 | 16.1 | 20.6 |
| Inpatient admission | 2.6 | 5.6 | 2.6 | 6.0 | 2.6 | 5.3 |
| Outpatient visits b | ||||||
| No visit | 20.1 | 2.8 | 19.8 | 10.0 | 19.8 | 11.9 |
| 1–3 | 40.1 | 31.2 | 40.8 | 32.7 | 40.8 | 34.8 |
| 4–6 | 18.0 | 25.3 | 18.2 | 21.4 | 18.1 | 21.2 |
| 7–12 | 13.0 | 23.6 | 13.1 | 20.3 | 13.1 | 18.7 |
| >12 | 8.0 | 17.2 | 8.1 | 15.6 | 8.1 | 13.6 |
| Beneficiary status | ||||||
| Employee | 51.3 | 51.3 | 51.3 | 51.8 | 51.3 | 52.3 |
| Spouse | 38.0 | 30.9 | 37.9 | 30.3 | 38.0 | 31.3 |
| Child or other | 10.7 | 17.8 | 10.8 | 17.9 | 10.8 | 16.5 |
| Health-plan type | ||||||
| Comprehensive | 1.1 | 1.5 | 1.2 | 1.5 | 1.2 | 1.5 |
| Health maintenance organization | 16.3 | 16.0 | 16.3 | 15.9 | 16.3 | 15.6 |
| Preferred provider organization | 58.1 | 58.8 | 58.1 | 59.3 | 58.1 | 59.6 |
| Point-of-service | 7.4 | 7.8 | 7.4 | 8.0 | 7.4 | 8.1 |
| Consumer driven | 6.8 | 6.6 | 6.8 | 6.2 | 6.8 | 6.4 |
| Other plan types | 10.2 | 9.3 | 10.2 | 9.1 | 10.2 | 8.9 |
a The numbers reported in the table are the percentages of the total pregnancies with or without prenatal exposure in each column
b All outpatient claims in 1 calendar day were considered as 1 outpatient visit.
Overall exposure to teratogenic drugs
Across all study years, exposure to drugs with a definite teratogenic risk during the preconception period was identified for 80,800 pregnancies (2.35%; 95% confidence interval [CI], 2.33–2.36), exposure during the first trimester was identified for 33,483 pregnancies (0.97%; 95% CI, 0.96–0.98), and exposure at any time during pregnancy was identified for 58,810 pregnancies (1.71%; 95% CI, 1.69–1.72) ( Table 2 ). For drugs with a potential teratogenic effect, we found 136,740 pregnancies with exposure during the preconception period (3.97%; 95% CI, 3.95–3.99), 75,249 with exposure during the first trimester (2.18%; 95% CI, 2.17–2.20), and exposure at any time during gestation was identified for 170,106 pregnancies (4.94%; 95% CI, 4.91–4.96). Exposure prevalences for REMS drugs were appreciably smaller, with only 952 pregnancies with exposure during the preconception period (0.03%; 95% CI, 0.03–0.03) and 519 with exposure during gestation (0.02%; 95% CI, 0.01–0.02). We contrasted prenatal exposure during pregnancy with exposure during the preconception period by calculating the prevalence ratio for each drug group and constructing the diagram in Figure 1 to facilitate interpretation.
| Pregnancy outcome subgroups | Preconception n (%) | Trimester 1 n (%) | Anytime during gestation n (%) |
|---|---|---|---|
| All pregnancies (N=3,445,612) | |||
| Definite teratogenic drugs | 80,800 (2.35) | 33,483 (0.97) | 58,810 (1.71) |
| Potential teratogenic drugs | 136,740 (3.97) | 75,249 (2.18) | 170,106 (4.94) |
| REMS drugs | 952 (0.03) | Not assessed | 519 (0.02) |
| Live births (n=2,532,444) | |||
| Definite teratogenic drugs | 56,203 (2.22) | 21,169 (0.84) | 45,959 (1.81) |
| Potential teratogenic drugs | 94,060 (3.71) | 50,316 (1.99) | 144,176 (5.69) |
| REMS drugs | 459 (0.02) | Not assessed | 229 (0.01) |
| Nonlive births (n=913,168) | |||
| Definite teratogenic drugs | 24,597 (2.69) | 12,314 (1.35) | 12,851 (1.41) |
| Potential teratogenic drugs | 42,680 (4.67) | 24,933 (2.73) | 25,930 (2.84) |
| REMS drugs | 493 (0.05) | Not assessed | 290 (0.03) |
When contrasting by outcome ( Table 2 ), pregnancies that ended in a live delivery had lower exposure prevalences to drugs with definite or potential teratogenic risks than pregnancies that ended in abortion or stillbirth (eg, 0.84% vs 1.35% for exposure to definite teratogenic drugs). This difference changed when exposure during gestation was considered, partly because of the longer observation periods, which continued for most live deliveries into the third trimester. When contrasting exposure across trimesters for live birth deliveries, we noted a slight decline in exposure prevalence with progression of gestation for drugs with a definite teratogenic risk (0.8%, 0.7%, and 0.6% during the first, second, and third trimester, respectively). The opposite pattern was found for potential teratogenic drugs for which exposure prevalences increased across trimesters (2.0%, 2.3%, and 2.9%). The trimester-specific annual prevalences stratified by pregnancy outcome, period, and teratogenic drug category are shown in Appendix 1 , Supplemental Tables 6 to 14 . An interactive version of these tables can be found at https://tabsoft.co/3BfNK47 .
Prenatal exposure among women of childbearing age with relevant teratogenic medication use
Considering the background utilization rates, we found that 0.48% of women of childbearing age who utilized a definite teratogenic drug during an examined calendar year did so within 3 months before conception and 0.36% did so during pregnancy ( Table 3 ). The prevalences were higher for potential teratogenic drugs, with 0.70% using such a drug during the preconception period and 0.87% using the drug during pregnancy. In contrast, there was a lower prevalence for the use of REMS drugs with 0.25% of women of childbearing age who used REMS drugs during the calendar year doing so during the 3 months before conception and 0.14% of women using the relevant drug during pregnancy.
| Period | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | Total a |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Exposure to any definite teratogenic drug | |||||||||||||
| Prepregnancy (%) | 4204 (0.48) | 5163 (0.51) | 7238 (0.48) | 8418 (0.48) | 8578 (0.5) | 9286 (0.51) | 8214 (0.46) | 7470 (0.51) | 6516 (0.43) | 5719 (0.49) | 5124 (0.46) | 4870 (0.46) | 80,800 (0.48) |
| Any time during pregnancy (%) | 3609 (0.42) | 4267 (0.43) | 5993 (0.4) | 6527 (0.38) | 6630 (0.4) | 6876 (0.38) | 6009 (0.34) | 5156 (0.36) | 4409 (0.3) | 3622 (0.32) | 3021 (0.28) | 2691 (0.26) | 58,810 (0.36) |
| Any WCBA b | 871,326 | 1,013,924 | 1,508,505 | 1,751,411 | 1,708,046 | 1,829,956 | 1,804,059 | 1,462,860 | 1,524,108 | 1,164,048 | 1,122,367 | 1,055,384 | 16,815,994 |
| Exposure to any teratogenic drug with REMS | |||||||||||||
| Prepregnancy (%) | 27 (0.19) | 42 (0.23) | 63 (0.22) | 77 (0.22) | 82 (0.25) | 80 (0.24) | 80 (0.22) | 100 (0.28) | 90 (0.22) | 115 (0.34) | 104 (0.31) | 92 (0.27) | 952(0.25) |
| Any time during pregnancy (%) | 14 (0.10) | 24 (0.13) | 42 (0.14) | 40 (0.11) | 37 (0.11) | 43 (0.13) | 41 (0.11) | 44 (0.12) | 51 (0.12) | 62 (0.18) | 60 (0.18) | 61 (0.18) | 519(0.14) |
| Any WCBA | 14,520 | 17,896 | 29,172 | 35,635 | 32,941 | 33,048 | 36,538 | 35,306 | 40,839 | 34,207 | 33,401 | 33,744 | 377,247 |
| Exposure to any potential teratogenic drug | |||||||||||||
| Prepregnancy (%) | 6541 (0.66) | 7760 (0.68) | 11,141 (0.66) | 13,058 (0.67) | 13,919 (0.73) | 15,706 (0.75) | 13,944 (0.66) | 12,773 (0.75) | 11,458 (0.63) | 10,719 (0.74) | 9996 (0.7) | 9725 (0.71) | 136,740 (0.70) |
| Any time during pregnancy (%) | 6600 (0.67) | 8248 (0.72) | 12,631 (0.75) | 15,171 (0.78) | 17,218 (0.9) | 20,083 (0.95) | 18,213 (0.87) | 17,345 (1.02) | 16,070 (0.88) | 14,762 (1.03) | 12,179 (0.86) | 11,586 (0.85) | 170,106 (0.87) |
| Any WCBA | 986,830 | 1,139,455 | 1,684,668 | 1,956,266 | 1,909,407 | 2,106,503 | 2,105,403 | 1,696,652 | 1,818,563 | 1,439,086 | 1,422,507 | 1,366,360 | 19,631,700 |
a The total column includes the sum of observations in each year. Individual women could be present in several study years
b WBCA, women of childbearing age with relevant drug exposure, that is, use of a definite, REMS, or potential teratogenic drug.
Secular trends
Among women of childbearing age, the overall annual utilization rates increased slightly for drugs with a definite risk (from 7.8% in 2006 to 8.7% in 2017) and for drugs with a potential risk (from 16.5% to 19.8%). The annual prevalence of exposure at any time during pregnancy for drugs with a definite risk decreased slightly: prenatal exposure prevalences among women of childbearing age who utilized these drugs during a calendar year changed from 0.42% in 2006 to 0.26% in 2017 ( Table 3 ). Likewise, prenatal exposure prevalences across all pregnancies decreased from 1.9% to 1.2% across the study period ( Figure 2 , A). The odds of prenatal exposure in 2017 was 34% lower than in 2006 after adjustment for prepregnancy exposure and other covariates (odds ratio [OR], 0.58; 95% CI, 0.55–0.61) ( Table 4 ).
| Factor | Prevalence | Base model | Enhanced model a | |||
|---|---|---|---|---|---|---|
| % of unexposed (n=3,386,802) | % of exposed (n=58,810) | Odds ratio | 95% confidence interval | Odds ratio | 95% confidence interval | |
| Healthcare quality ranking b | ||||||
| High rank | 16.8 | 12.7 | Ref | Ref | Ref | Ref |
| Middle rank | 67.1 | 66.8 | 1.30 | 1.27–1.33 | 1.26 | 1.23–1.30 |
| Low rank | 16.1 | 20.6 | 1.63 | 1.58–1.68 | 1.55 | 1.51–1.60 |
| Age (y) | ||||||
| <20 | 4.3 | 7.6 | 1.94 | 1.88–2.01 | 1.84 | 1.77–1.90 |
| 20–24 | 12.5 | 16.1 | 1.55 | 1.51–1.59 | 1.48 | 1.44–1.52 |
| 25–29 | 27.1 | 24.6 | 1.12 | 1.10–1.16 | 1.12 | 1.10–1.15 |
| 30–34 | 32.9 | 25.8 | Ref | Ref | Ref | Ref |
| 35–39 | 17.9 | 15.6 | 1.10 | 1.07–1.13 | 1.08 | 1.06–1.11 |
| 40–44 | 4.7 | 5.7 | 1.51 | 1.45–1.57 | 1.44 | 1.39–1.50 |
| ≥45 | 0.7 | 4.7 | 8.17 | 7.82–8.55 | 6.83 | 6.51–7.17 |
| Inpatient admission | 2.6 | 5.3 | 1.40 | 1.36–1.47 | 1.01 | 0.97–1.06 |
| Outpatient visits | ||||||
| No visit | 19.8 | 11.9 | Ref | Ref | Ref | Ref |
| 1–3 | 40.8 | 34.8 | 1.27 | 1.23–1.30 | 1.19 | 1.16–1.23 |
| 4–6 | 18.1 | 21.2 | 1.54 | 1.49–1.58 | 1.29 | 1.24–1.33 |
| 7–12 | 13.1 | 18.7 | 1.76 | 1.69–1.80 | 1.29 | 1.24–1.34 |
| >12 | 8.1 | 13.6 | 1.97 | 1.88–2.01 | 1.18 | 1.13–1.24 |
| Exposure during preconception | 2.0 | 21.5 | 10.84 | 10.60–11.08 | 8.69 | 8.48–8.90 |
| Study year | ||||||
| 2006 | 5.6 | 6.1 | Ref | Ref | Ref | Ref |
| 2007 | 6.5 | 7.3 | 1.01 | 0.97–1.06 | 1.00 | 0.96–1.05 |
| 2008 | 9.1 | 10.2 | 1.02 | 0.98–1.06 | 1.00 | 0.96–1.05 |
| 2009 | 9.9 | 11.1 | 1.01 | 0.97–1.05 | 0.98 | 0.94–1.02 |
| 2010 | 10.3 | 11.3 | 1.00 | 0.96–1.04 | 0.96 | 0.92–1.00 |
| 2011 | 11.3 | 11.7 | 0.93 | 0.90–0.97 | 0.89 | 0.85–0.93 |
| 2012 | 10.0 | 10.2 | 0.93 | 0.89–0.97 | 0.88 | 0.85–0.92 |
| 2013 | 9.1 | 8.8 | 0.88 | 0.84–0.92 | 0.83 | 0.79–0.87 |
| 2014 | 8.0 | 7.5 | 0.85 | 0.81–0.89 | 0.79 | 0.75–0.83 |
| 2015 | 7.2 | 6.2 | 0.78 | 0.74–0.82 | 0.70 | 0.67–0.74 |
| 2016 | 6.6 | 5.1 | 0.71 | 0.67–0.74 | 0.63 | 0.60–0.67 |
| 2017 | 6.3 | 4.6 | 0.66 | 0.62–0.69 | 0.58 | 0.55–0.61 |
a In the enhanced model, clinical conditions that were determined from diagnoses codes on medical in- and outpatient encounters within 6 months before conception and aggregated based on the Agency for Healthcare and Quality’s Clinical Classification System were included as covariates in the model (222 variables). Beneficiary status (employee, spouse, children or other) and type of health plan were also included in this model
b The healthcare quality ranking was extracted from the US News portal ( https://www.usnews.com/news/best-states/rankings/health-care ) and the top 10 states were categorized as high rank, the last 10 states as low rank, and the rest as middle rank.
In contrast, for drugs with a potential teratogenic risk, we observed an increasing trend for prenatal exposure across all pregnancies from 3.4% in 2006 to 5.3% in 2017 ( Figure 2 , B) (OR, 1.56; 95% CI, 1.51–1.61) ( Appendix 1 , Supplemental Table 16 ). A similar trend was seen when considering all women of childbearing age who utilized these drugs during a study year ( Table 3 ).
Although exposure was small overall, we also observed an increasing trend for prenatal exposure to drugs with a REMS program across all pregnancies, starting from 0.007% in 2006 to 0.028% in 2017 ( Figure 2 , C) with an OR of 2.63 (95% CI, 1.40–4.95) ( Appendix 1 , Supplemental Table 17 ). Again, this trend was similar for prenatal exposure prevalences when using the number of women of childbearing age who utilized REMS drugs in a given year as the denominator ( Table 3 ).
Determinants of prenatal exposure
Exposure to definite teratogenic drugs during the preconception period showed a strong association with the probability for exposure during pregnancy (OR, 8.69; 95% CI, 8.48–8.90). When compared with middle-aged women (30–34 years), teenagers and older women had a higher risk for prenatal exposure (OR, 1.84; 95% CI, 1.77–1.90 for patients aged <20 years, and OR, 6.83; 95% CI, 6.51–7.17 for patients aged ≥45 years). Women living in states ranked in the middle or low healthcare quality category had higher odds for prenatal exposure than those living in states with high healthcare quality rankings (OR, 1.26; 95% CI, 1.23–1.30 and OR, 1.55; 95% CI, 1.51–1.60, respectively) ( Table 4 ). The ORs for these determinants were similar for the probability of prenatal exposure to potential teratogenic drugs and REMS drugs, although the CIs for the REMS model became wide because of the small numbers ( Appendix 1 , Supplemental Tables 16 and 17 ).
Sensitivity analysis
When we varied the estimated gestational age for pregnancy episodes that ended with nonlive deliveries, the prevalence of prenatal exposure to definite teratogens at any time during gestation varied from 1.04% to 3.19% (original estimate, 1.41%) and exposure to potential teratogens varied from 2.15% to 5.41% (original estimate, 2.84%) ( Appendix 1 , Supplemental Table 15 ).
Comment
Principal findings
Four key study findings deserve highlighting. First, we found a slight decreasing trend for prenatal exposure to definite teratogenic risk drugs, which is in contrast with the increasing trend seen for drugs with potential teratogenic risks. Second, REMS drugs comprised a very small portion of the overall prenatal exposure, which was a function of the overall small background utilization among women of childbearing age and lower prenatal exposure rates. Third, prenatal exposure prevalences were higher in the first trimester and among pregnancies with nonlive outcomes. Lastly, we observed that state healthcare quality rankings and age extremes were associated with a higher probability of prenatal exposure, suggesting that there are shortcomings in pregnancy prevention when women use teratogenic drugs.
Results in the context of what is known
Previous studies on prenatal exposure to medications have focused on specific drug classes (eg, cardiovascular agents) and/or limited the analyses to cohorts of pregnancies with live birth outcomes. , In a retrospective study in which data from 8 health maintenance organizations (1996–2000) were used, the authors estimated that 1.1% of pregnant patients had prescription filled for 23 drugs or drug classes with teratogenic effects during the 270 days before delivery, whereas 5.8% of pregnant patients had a prescription filled for any medication in the FDA’s previously used teratogenicity categories D or X. Another study of 1.1 million pregnancies with live birth outcomes in the Medicaid database (2000–2007) reported that 42% of pregnant patients had at least 1 prescription filled with category D or X medications. Opioids were the most common category D medication and hormonal contraceptives were the most common category X medication in the aforementioned study, which were not included in our study based on our more stringent definition of teratogenic risk. Instead of employing the FDA’s now discontinued pregnancy risk classification system, we screened the TERIS database as a specialized source of information on teratogenic drugs and supplemented the list with a standard drug information source to capture drugs with explicit contraindications, allowing some confidence in terms of the potential harm imposed on pregnancy outcomes.
Clinical implications
We observed that drugs with potential teratogenic risks contribute significantly to prenatal exposure, although conclusive evidence about the risk-benefit is lacking. Multiple reasons could lead to this higher prevalence of exposure to these drugs, including a higher overall background utilization among women of childbearing age, uncertainty about the risk-benefit based on clinical judgment and the available therapeutic options, and more exposure because the pregnancy was unintended. It is prudent to consider these safety concerns in treatment decisions for women of childbearing age and pregnancy planning consultations until further clinical evidence is available to confirm or rule out a risk for harm.
Preconception use was the strongest predictor of prenatal exposure, but both a younger (<20 years) and older (≥45 years) maternal age were significantly associated with prenatal exposure, which follows the distribution of reported unintended pregnancy rates and suggests that some of these exposures might have been inadvertent. , The notion that in some instances the potential for teratogenic risk was unknown or not considered is corroborated by the finding that states with lower healthcare quality rankings had a higher prenatal exposure prevalence. Failure to follow best practices has been associated with overall healthcare quality for various medication use issues.
Patients’ awareness of the potential teratogenic risk for medications is often limited, and the information available in the public domain, especially safe drug lists found in internet-based resources, is often inconsistent and may not reflect the current state of knowledge. The evidence on the teratogenic risk of many drugs and, even more importantly, on the risk-benefit of use during pregnancy is fragmented and more concerted approaches are needed to provide and disseminate treatment guidance. From a regulatory perspective, effective risk communication to healthcare providers is necessary to facilitate decision-making at the point of care. Although the FDA modernized the format of communicating pregnancy and lactation risk in the prescribing label (2015), a recent survey reported that most clinicians in the study sample were unaware of the label format change and continued to rely on letter categories. Little evidence exists on whether providing an evidence summary is effective in improving the assessment of risk-benefit, but concerns about the utility of these summaries in busy clinical practices and for integration into automated clinical decision support have been raised. ,
REMS programs aim to enhance risk communication beyond warnings in the package insert and to enforce adherence to safe-use recommendations. Therefore, they might require dissemination of additional educational materials, patient registration, prescriber and pharmacy certification, or laboratory tests (eg, pregnancy tests) before and/or during drug utilization, with significant variation in the required components across REMS programs. Our study showed that prenatal exposure still occurs in the real world for drugs with a REMS program despite these additional measures. We also observed that preconception use of teratogenic drugs and frequent medical visits before conception are strong determinants of prenatal exposure, suggesting that conception during treatment (as opposed to treatment initiation during pregnancy) is an important pathway for prenatal exposure. In our study on the impact of the new REMS for mycophenolate in comparison with its predecessor strategy (black box warning), we observed that the REMS effectively prevented treatment initiation during pregnancy. However, there was no difference in pregnancy occurrence rates during treatment, which demonstrated effectiveness in changing provider behavior (in not initiating mycophenolate during pregnancy) but limited effect on patient behavior (in avoiding conception during use). These findings suggest that encouraging and facilitating appropriate pregnancy prevention methods (eg, effective contraceptive options) are critical for avoiding prenatal exposures. Despite the possible failures, even of drugs subject to REMS programs, we see opportunities to improve prenatal exposure to drugs with definite (or probable) teratogenic risks and no current REMS.
Research implications
There is an urgent need for population-based studies on drugs with potential teratogenic risks to determine whether the risk-benefit of treatment during pregnancy is favorable. Although our study was not designed to quantify the effect of prenatal exposure on pregnancy loss, the observation that exposure prevalences during the first trimester were higher for nonlive birth pregnancies than for live births warrants further investigation. Moreover, studies are needed to understand whether implementing additional risk minimization measures is warranted to prevent prenatal exposure for definite and potential teratogenic risk drugs. Unfortunately, high-quality evidence on the incremental effectiveness of different REMS elements is sparse, and the quality of evidence and reporting in REMS evaluation studies is substandard. , Our study showed that the prevalence of prenatal exposure to drugs with REMS in 2017 was higher than during the baseline year (2006) primarily because of increased overall utilization. More research is needed to understand the root causes of REMS program failures and what criteria should be employed to decide whether and which REMS elements should be used for teratogenic medications.
Strengths and limitations
Our study provides up-to-date estimates on maternal exposure to teratogenic drugs during the REMS era using a national sample of women with private insurance plans. We included pregnancies with nonlive outcomes, which is in contrast with previous studies that were mostly limited to live births.
The study findings are subject to several limitations. First, our drug exposure measure relied on insurances claims; no information on actual drug consumption was available. To mitigate the risk for misclassification bias, we considered pregnancies as exposed only if the dispensing or in-clinic administration date occurred during a given pregnancy period and did not consider drug supply that may have been available from previous prescription fills. Second, validation studies have shown high accuracy for measuring the type of pregnancy endpoint and end date (>95%). However, although the timing of conception for live deliveries is quite accurate, the gestational age estimation for cases ending in nonlive births has shown only modest accuracy. We therefore conducted a sensitivity analysis in which we varied the estimated gestational age for nonlive deliveries to gestational ages that fall within the most common range that would be encountered for spontaneous abortions in clinical practice (8–18 weeks). Changes in the estimated conception date would especially affect prenatal exposure measures during the first trimester if women discontinued teratogenic medications shortly before conception. We did notice a drop in both prenatal exposure to definite and potential teratogenic drugs when reducing the gestational age and thus extending the conception date, but the magnitude of exposure remained significant, which is corroborated by the observation that exposure commonly extended into the second and third trimester. Third, pregnancies with exposures to teratogenic medications ending in early spontaneous abortion or elective abortions without medical encounters would not be captured in our database and thus would lead to an underestimate of prenatal exposure. Finally, our study findings may not be generalizable to the entire population in the United States because the database is a sample of the privately insured population, whereas Medicaid covers approximately 45% of deliveries.
Conclusions
More than 1 in 16 pregnancies continue to be exposed to drugs with teratogenic potential during the REMS era. Although exposure to teratogenic drugs with a definite risk has decreased slightly from 2006 to 2017, exposure to drugs with potential risk has increased. Drugs with REMS imposed a small burden on prenatal exposure because of the low background utilization rates and lower pregnancy prevalence among women of childbearing age who used these drugs. Future research is needed to elucidate the health outcomes of pregnancies exposed to drugs with potential risks, to understand the effectiveness of REMS programs, and to prioritize teratogenic drugs for advanced risk mitigation.
FDA: US Food and Drug Administration
Potential teratogen: a drug with a plausible teratogenic effect given drug labeling, pharmacologic mechanism, controlled population-based studies, or registry analysis
Prenatal exposure: drug use within the period starting at the last menstrual period plus 2 weeks (ie, conception) until pregnancy outcome.
REMS: risk evaluation and mitigation strategy
A drug safety program that the FDA can require for certain medications with serious safety concerns to help ensure that the benefits of the medication outweigh its risks. A REMS is designed to reinforce medication use behaviors and actions that support the safe use of that medication.
RiskMAP: risk minimization action plan
A strategic safety program designed to meet specific goals and objectives in minimizing known risks of a product while preserving its benefits. A RiskMAP targets 1 or more safety-related health outcomes or goals and uses 1 or more tools to achieve those goals. These programs were REMS predecessors and were conducted voluntarily. The FDA had no authority to require pharmaceutical manufactures to implement them if deemed necessary.
Definite teratogen: a drug with high certainty of teratogenic effect based on the totality of evidence available from pharmacology, case reports, and epidemiologic studies.
TERIS: Teratogen Information System
Supplementary Data
Sarayani et al. Prenatal exposure to teratogenic drugs during the risk evaluation and mitigation strategies era. Am J Obstet Gynecol 2022.
Sarayani et al. Prenatal exposure to teratogenic drugs during the risk evaluation and mitigation strategies era. Am J Obstet Gynecol 2022.
Appendix 1
Description of the pregnancy identification algorithm
From a healthcare data perspective, “pregnancy” is an episode of medical care that may generate encounter data from the initial visits for preconception care to prenatal visits and through the pregnancy endpoint (eg, normal birth of an infant, premature birth, abortion). Three parameters define each pregnancy episode, namely (1) start date (date of clinical conception or last menstrual period), (2) pregnancy duration, and (3) pregnancy end date. Although the pregnancy end date is typically well specified by the relevant clinical encounter, conception is not. Clinically, estimates of gestational age are derived based on ultrasound findings or women’s reports of their last menstrual period, neither of which is systematically captured in billing records (claims data). However, coding of gestational age on pregnancy-related encounters and other relevant diagnoses codes allows estimation.
The pregnancy identification algorithm employed in this study relies on 3 steps, namely (1) identify claims for pregnancy endpoint encounters to anchor the pregnancy end date; (2) estimate gestational age at the endpoint and back-calculate to the start date; and (3) apply a prioritization rule to resolve overlapping episodes. Endpoints include live or mixed birth, delivery (unclassified birth status), stillbirth, abortion (induced or spontaneous), and ectopic pregnancy termination. Pregnancies with an unknown endpoint, that is pregnancies only identified based on prenatal visits but no encounter indicating termination or delivery, are not considered. After identifying each episode’s endpoint date, the algorithm estimates gestational age based on the coding on the observed claims on or close to the endpoint day (for live, mixed, or unclassified delivery) or assigns fixed-values determined from the literature (ie, 10 weeks for abortions, 8 weeks for ectopic pregnancy terminations, and 28 weeks for stillbirths). The start date of each episode is back-calculated from the end date based on the estimated gestational age for each episode. After establishing all pregnancy episodes in the database, which may lead to multiple episodes for each woman, a hierarchical approach is used to reconcile overlapping episodes. For example, an episode ending in live delivery is prioritized over an episode ending in abortion, and the latter is removed from the data set if its end date falls within the gestational period assigned for a live delivery.
In Supplemental Figure 2 , example of two pregnancy episodes with clinically plausible sequence (top) and two episodes for which the end date of the first episode falls within the gestational period of the second are illustrated. Because both the validity of the endpoint measurement and the accuracy of gestational age estimations are superior for live deliveries, the second pregnancy episodes are prioritized, and the episode ending in abortion is removed from the data set.
For more detail on the algorithm implementation, refer to the following citation:
Sarayani A, Wang X, Thai TN, Albogami Y, Jeon N, Winterstein AG. Impact of the Transition from ICD–9–CM to ICD–10–CM on the Identification of Pregnancy Episodes in US Health Insurance Claims Data. Clin Epidemiol. 2020; 12:1129-1138
Approach to selecting teratogenic medications and categorizing teratogenic risk
We created 2 lists of teratogenic medications, based on the level of evidence, with 1 listing definite teratogens and the other listing potential teratogens.
Data sources and initial screening procedure
We used the Teratogen Information System (TERIS) and Clinical Pharmacology databases to identify candidate teratogenic drugs. From each database, only drugs approved and marketed in the United States during the study period (2006–2018) were considered, and other agents such as dietary supplements, vaccines, and nonpharmaceutical products were excluded. Nonsystemic pharmaceutical agents were excluded if there was no significant evidence of harm.
Categorization of teratogenicity risk
Two authors independently reviewed each drug monograph in TERIS to confirm the potential for harm and categorize the risk as described below.
Definite teratogens: we selected all drugs with a risk designation magnitude of small to moderate, moderate, or high. If the magnitude of risk was categorized as undetermined, but the comment section had plausible arguments in favor of significant teratogenicity based on pharmacology (eg, cytotoxic or antineoplastic agent), a known teratogen in the same drug class, or reference to a substantial level of risk, the agent was included in this category.
Potential teratogens: we selected those drugs classified in TERIS as having an undetermined magnitude of risk with a narrative or comment section that included either a controlled population-based study or registry analysis that found an increased risk for birth defects. All drugs with an absolute contraindication in the Clinical Pharmacology database (n=185) were cross-checked with TERIS. Drugs with an absolute contraindication were included in the potential teratogen category if no TERIS summary was available or the information in the summary section was not adequate to support conclusive assessments. If a drug had an absolute contraindication but the TERIS summary ruled out risk, the drug was excluded.
To determine the trimester of concern, we utilized TERIS and the Clinical Pharmacology databases to group drugs based on risk during first trimester and/or second and third trimester. A drug was considered to be of concern in both the first trimester and second and third trimester unless there was evidence to suggest minimal risk during some part of pregnancy.
Final harmonization
To remove drugs with debatable fetal risk, we removed the following drug groups: benzodiazepines, obesity drugs if the mechanism of harm was solely attributed to weight loss during pregnancy, statins, tetracyclines, sex hormones (estrogens, progesterone or progestins, testosterone), drugs with infertility treatment indication (ganirelix, lutropin alfa, menotropins, letrozole, etc.), gonadotropin-releasing hormone analogs, abortion treatments (short-term misoprostol and mifepristone), and postpartum or abortion hemorrhage treatments (ergonovine and methylergonovine). These medications are either frequently used during pregnancy for medical purposes or have minimal risk. ,
We confirmed that the 12 teratogenic drugs with a “current or discontinued” REMS program and cancer medications were included in the definite risk group. Our final drug list included 206 unique generic drug names with 141 of those designated as having definite and 65 designated as having potential risk, including 12 drugs with current or discontinued REMS.
| Generic name | First–second or third trimester risk (yes/no) | Generic name | First–second or third trimester risk (yes/no) |
|---|---|---|---|
| Acenocoumarol | Yes-yes | Lisinopril | No-yes |
| Acetohydroxamic Acid | Yes-yes | Lomustine | Yes-yes |
| Acitretin | Yes-yes | Losartan | No-yes |
| Afatinib | Yes-yes | Macitentan a | Yes-yes |
| Aflibercept | Yes-yes | Mechlorethamine | Yes-yes |
| Alitretinoin (systemic administration) | Yes-yes | Melphalan | Yes-yes |
| Altretamine | Yes-yes | Methotrexate | Yes-yes |
| Ambrisentan a | Yes-yes | Methsuximide | Yes-yes |
| Amiodarone | No-yes | Methylene blue (intra-amniotic administration) | Yes-yes |
| Asparaginase | Yes-yes | Misoprostol | Yes-yes |
| Axitinib | Yes-yes | Mitomycin | Yes-yes |
| Azacitidine | Yes-yes | Mitotane | Yes-yes |
| Azilsartan | No-yes | Moexipril | No-yes |
| Benazepril | No-yes | Mycophenolate mofetil a | Yes-yes |
| Bendamustine | Yes-yes | Nelarabine | Yes-yes |
| Bevacizumab | Yes-yes | Nilotinib | Yes-yes |
| Bexarotene (systemic administration) | Yes-yes | Nintedanib | Yes-yes |
| Bleomycin | Yes-yes | Olmesartan | No-yes |
| Bortezomib | Yes-yes | Onartuzumab | Yes-yes |
| Bosentan a | Yes-yes | Oxaliplatin | Yes-yes |
| Bosutinib | Yes-yes | Panitumumab | Yes-yes |
| Brequinar | Yes-yes | Panobinostat | Yes-yes |
| Cabazitaxel | Yes-yes | Pazopanib | Yes-yes |
| Cabozantinib | Yes-yes | Pemetrexed | Yes-yes |
| Candesartan | No-yes | Penicillamine | Yes-yes |
| Capecitabine | Yes-yes | Perindopril | No-yes |
| Captopril | No-yes | Pertuzumab | Yes-yes |
| Carbamazepine | Yes-yes | Phenprocoumon | Yes-yes |
| Carmustine | Yes-yes | Phenytoin or fosphenytoin | Yes-yes |
| Ceritinib | Yes-yes | Pomalidomide a | Yes-yes |
| Cetuximab | Yes-yes | Ponatinib | Yes-yes |
| Cladribine | Yes-yes | Pralatrexate | Yes-yes |
| Clofarabine | Yes-yes | Primidone | Yes-yes |
| Cytarabine | Yes-yes | Quinapril | No-yes |
| Dabrafenib | Yes-yes | Ramipril | No-yes |
| Dacarbazine | Yes-yes | Ramucirumab | Yes-yes |
| Danazol | Yes-yes | Regorafenib | Yes-yes |
| Dasatinib | Yes-yes | Riociguat a | Yes-yes |
| Daunorubicin | Yes-yes | Romidepsin | Yes-yes |
| Decitabine | Yes-yes | Ruxolitinib | Yes-yes |
| Dicumarol | Yes-yes | Sirolimus | Yes-yes |
| Divalproex | Yes-yes | Sorafenib | Yes-yes |
| Docetaxel | Yes-yes | Sulfamethoxazole with trimethoprim | Yes-yes |
| Doxorubicin | Yes-yes | Sunitinib | Yes-yes |
| Enalapril | No-yes | Talimogene laherparepvec | Yes-yes |
| Eprosartan | No-yes | Tamoxifen | Yes-yes |
| Eribulin | Yes-yes | Tasosartan | No-yes |
| Erlotinib | Yes-yes | Telavancin a | Yes-yes |
| Ethotoin | Yes-yes | Telmisartan | No-yes |
| Etoposide | Yes-yes | Temozolomide | Yes-yes |
| Everolimus | Yes-yes | Thalidomide a | Yes-yes |
| Fingolimod a | Yes-yes | Tofacitinib | Yes-yes |
| Floxuridine | Yes-yes | Topiramate a | Yes-yes |
| Fludarabine | Yes-yes | Topotecan | Yes-yes |
| Fluorouracil | Yes-yes | Trabectedin | Yes-yes |
| Fosinopril | No-yes | Trametinib | Yes-yes |
| Gefitinib | Yes-yes | Trandolapril | No-yes |
| Hydroxyurea | Yes-yes | Trastuzumab | Yes-yes |
| Ibrutinib | Yes-yes | Tretinoin (systemic administration) | Yes-yes |
| Ifosfamide | Yes-yes | Trimethadione or paramethadione | Yes-yes |
| Imatinib | Yes-yes | Valproic acid | Yes-yes |
| Iodine-131 | Yes-yes | Valsartan | No-yes |
| Irbesartan | No-yes | Vandetanib | Yes-yes |
| Irinotecan | Yes-yes | Vedolizumab | Yes-yes |
| Isotretinoin a | Yes-yes | Vemurafenib | Yes-yes |
| Ixabepilone | Yes-yes | Vincristine | Yes-yes |
| Lapatinib | Yes-yes | Vismodegib | Yes-yes |
| Leflunomide | Yes-yes | Vorinostat | Yes-yes |
| Lenalidomide a | Yes-yes | Warfarin | Yes-yes |
| Lenvatinib | Yes-yes | Zonisamide | Yes-yes |
| Lestaurtinib | Yes-yes |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
