Objective
The objective of the study was to estimate the potential reduction of neural tube defects (NTDs) through the use of Metafolin-fortified oral contraceptives (OCs) in the United States.
Study Design
A population-based decision analytic model was developed to estimate the benefits of increased red blood cell (RBC) folate levels through the use of Metafolin-fortified OCs on NTD risk during pregnancy. We modeled women who began the year taking Metafolin-fortified or traditional OCs. Folate levels were derived from the National Health and Nutrition Examination Survey and clinical trial data. NTD risk was estimated by applying a published risk equation to respective RBC folate levels.
Results
The number of predicted NTD cases declined by 23.7% to 31.4%, depending on median baseline folate levels in women taking a fortified OC compared with taking a traditional OC.
Conclusion
Metafolin-fortified OCs have the potential to reduce the number of folate-dependent NTDs among current and recent OC users.
Neural tube defects (NTDs) are the second most common group of major structural birth defects; they are associated with high mortality and short- and long-term morbidity as well as significant humanistic and economic burdens to society. Spina bifida, anencephaly, and encephalocele are the 3 most common types of NTDs, affecting as many as 1 in 1000 pregnancies each year in the United States.
Infants with anencephaly are stillborn or die within days of birth. Survival rates for infants with spina bifida have improved over the years, but survivors are likely to have long-term physical and psychosocial disabilities, often requiring special education and substantial caregiver time. In addition, the lifetime costs associated with spina bifida were estimated for 2002 to be $279,210 per person in direct cost or $635,210 in direct and indirect costs associated with morbidity and premature mortality.
Low folate levels are associated with increased risk of NTDs. In addition, women who have had a previous NTD-affected pregnancy, are white or Hispanic, are obese, have insulin-dependent diabetes, are taking certain antiseizure medications, or have had high temperatures during early pregnancy are at increased risk for NTDs. Periconceptional use of folic acid to achieve adequate folate levels has been demonstrated to reduce the risk of NTDs by as much as 50-70%.
The US Preventive Services Task Force guidelines recommend that all women capable of pregnancy take an oral supplement containing folic acid (400-800 μg per day) and that women planning pregnancy should initiate daily folic acid supplementation 1 month prior to conception and continue until at least 2-3 months following conception. Unfortunately, estimates based on surveys suggest that only 26-40% of women of child-bearing age are taking oral supplements with rates lower among women aged 18-24 years. Women not on folate supplementation who become pregnant unintentionally might initiate oral supplements too late to achieve adequate folate levels to reduce the risk of folate dependent NTDs, even if they begin oral supplementation as soon as their pregnancy is confirmed.
The United States and Canada are among countries that have instituted mandatory folic acid food-fortification programs that require cereal-grain products to be enriched with folic acid. Evaluation of the mandatory fortification program in the United States suggests that folic acid food fortification is associated with a reduction of 26-27% in the prevalence of NTDs. Despite this, some women consuming fortified foods still fail to achieve folate levels assumed to be associated with minimum NTD risk (red blood cell [RBC] folate levels of ≥906 nmol/L) as demonstrated by Daly et al.
One method targeted specifically to women of child-bearing age for raising folate levels involves the use of Metafolin (Merck & Cie, Schaffhausen, Switzerland; registered trademark, Merck KGaA, Darmstadt, Germany) fortified oral contraceptives (OCs). This approach recognizes that approximately half of all pregnancies in the United States are unplanned and that women who are planning pregnancies may not begin taking folic acid until sometime after the critical first month following conception. A large percentage of women of child-bearing age, who therefore are at risk for becoming pregnant, use OCs prior to becoming pregnant, whether or not the pregnancy was intentional. Approximately 5-8% of women in the United States on OCs become pregnant each year under conditions of typical use, as defined by self-report.
In addition, for women who discontinue OC use to become pregnant, folate levels can be expected to diminish but remain above baseline for some weeks following discontinuation of Metafolin-fortified OCs. Thus, fortifying OCs with Metafolin to raise folate levels would target those women who may become pregnant unintentionally or shortly after discontinuing their fortified OC.
The concept of using OCs as a vehicle for elevating folate levels among women of child-bearing age was unanimously endorsed at a US Food and Drug Administration Advisory Committee for Reproductive Health Drugs meeting in 2003. The committee concluded that increased folate levels could be beneficial to women who become pregnant shortly after discontinuing OCs or who unintentionally become pregnant while taking OCs. With approximately 1 million pregnancies annually among OC users in the United States, fortifying OCs to elevate folate levels may have considerable potential for further reducing the burden of NTDs.
Two products currently approved, Beyaz and Safyral (Bayer HealthCare Pharmaceuticals Inc, Wayne, NJ), combine an ethinylestradiol/drospirenone-containing OC with 0.451 mg levomefolate calcium (Metafolin), a dose equivalent to 0.4 mg of folic acid. Two clinical trials have examined the effects of these Metafolin-fortified OC on RBC folate levels. The European Longterm Folate Study, conducted in Germany, compared the combination of Metafolin in an OC containing drospirenone (3 mg) and ethinyl estradiol (30 μg), given orally in a cyclic regimen for 24 weeks (6 cycles), with the same OC administered separately with folic acid tablets. After the 24 week active treatment phase, women were followed up for an additional 20 weeks to evaluate the elimination of folate from the body over time.
The US Benefit study compared an OC containing drospirenone (3 mg) and ethinyl estradiol (20 μg) plus Metafolin with the non–Metafolin-fortified version of the OC. RBC folate levels were measured every 2 weeks for 24 weeks (6 cycles). Both trials demonstrated significant increases in RBC folate levels associated with the fortified OC. In the European Longterm Folate Study, folate levels were assessed in an additional follow-up phase for 20 weeks after discontinuation during which they were found to diminish but remained elevated over baseline levels.
Mandatory food fortification programs and efforts to promote dietary supplement recommendations have been major factors in the decline of NTD cases in the United States since around 1990. Despite these efforts, NTDs remain an important public health concern. Because of the low incidence of NTDs, it would be impractical to implement a clinical trial to directly assess the impact of Metafolin-fortified OC use on NTD risk.
The objective of this study was to apply clinical trial data derived from Metafolin-fortified OC studies to population-based NTD risk estimates and assess the potential reduction in NTDs through the use of such an OC among women using OCs in the United States.
Materials and Methods
Model overview
We developed a population-based model ( Figure 1 ) to estimate the potential reduction in NTDs in the United States that would occur if women of child-bearing age taking OCs were switched to Metafolin-fortified OCs. The model estimated the risk of NTDs as a function of RBC folate levels, using data for an Irish population reported by Daly et al. The model considered only spina bifida and anencephaly; we judged that the data were too limited to adequately model the occurrence of other NTDs, such as encephalocele. We assumed that adverse events associated with OC use were the same, whether fortified or not; therefore, adverse events were omitted from the analysis.
The model compared outcomes associated with 2 scenarios: one in which all women currently on OCs received a Metafolin-fortified OC and one in which all women received a traditional OC (current practice). The population for this model was all women who were taking OCs in the United States. The model estimated outcomes for women using an OC at the beginning of the year. Women on fortified OCs were assumed to have been on the fortified OC for at least 6 months to assure that they had achieved steady-state folate levels upon entry into the model. Pregnancy outcomes (live birth, stillbirth, or elective termination) were estimated for all women who conceived in the year, even if the pregnancy outcome occurred in the subsequent year.
The model was developed in Microsoft Excel (Microsoft, Redmond, WA) and was quality checked by individuals not directly involved with this project. This study was exempt from an institutional review board because the study did not use human subjects.
Model inputs
Base-case model input values are displayed in Table 1 . We applied age-specific estimates of OC use obtained from the 2002 National Survey of Family Growth to US population estimates for 2008 to estimate the number of women using OCs.
| Model parameter | Base-case value | Reference |
|---|---|---|
| Population of women of child-bearing age using OCs | 12,228,111 | |
| Annual probability of pregnancy with continued typical use of OCs | 5.0% | |
| Annual probability of discontinuing OC use | 23.0% | |
| Percentage of women discontinuing OC use who desire pregnancy | 18.3% | |
| Probability of pregnancy within 20 weeks, given discontinued OC and desired pregnancy | 75.7% | |
| Probability of pregnancy among women who discontinued OC who were not planning on becoming pregnant | 3.3% | |
| Percentage of NTDs that are spina bifida | 57.4% | |
| Spina bifida pregnancy outcomes | ||
| Live birth a | 48.0% | |
| Elective termination | 43.1% | |
| Fetal death | 8.9% | |
| Anencephaly pregnancy outcomes | ||
| Live birth a | 28.7% | |
| Elective termination | 46.1% | |
| Fetal death | 25.2% | |
| Non-NTD pregnancy outcomes | ||
| Live birth a | 77.0% | |
| Elective termination | 22.4% | |
| Fetal death | 0.6% |
Women who have continued use of OCs under typical-use situations may experience unintended pregnancies because of either OC failure or less-than-perfect regimen compliance. We assumed that women who experienced an unintended pregnancy under typical OC use received the full benefits of Metafolin, although they may have missed several doses of their OC.
Some women discontinue OC use each year. We estimated the percentage of women who voluntarily discontinued OC use in 1 year to be 23%, using a linear extrapolation of estimates from 2 sources. For women who discontinued OC use, the model differentiated between those who discontinued OCs with the intention of becoming pregnant and those who discontinued for other reasons and did not desire a pregnancy. We estimated the percentage of women discontinuing OC use who were seeking pregnancy by using age-specific data from the National Survey of Family Growth.
For women who become pregnant in the months following discontinuation, the model assumed a partial benefit of Metafolin fortification for up to 20 weeks, as reported in the discontinuation phase of the European Longterm Folate Study. To estimate the probability of pregnancy within 20 weeks of discontinuing OC use, we utilized estimates reported by Cronin et al, who reported monthly pregnancy rates among women discontinuing OC use because of a planned pregnancy. For women who discontinue OC use and do not intend to become pregnant, age-specific pregnancy rates were estimated from Henshaw. We assumed that discontinuations and pregnancies were uniformly distributed throughout the year.
NTD risk
The impact of folate levels on NTD risk was modeled using the relationship between RBC folate levels and NTD risk reported by Daly et al. In the sample population used to derive the Daly equations, there were few cases of RBC folate levels greater than 1300 nmol/L. We therefore conservatively limited the maximum folate value in the model to 1300 nmol/L, effectively placing a floor on the estimated NTD risk at approximately 8.3 NTDs per 10,000 pregnancies.
Because we were interested in extrapolating data on the change in RBC folate levels reported in the US Benefit Study to the US population of women taking OCs as a whole, we used 2 approaches to estimate baseline RBC folate levels. First, we used median RBC folate levels as reported in the US Benefit Study. Second, median RBC folate levels were estimated using data from the 2005-2006 National Health and Nutrition Examination Survey (NHANES).
The NHANES is a nationally representative survey conducted by the National Center for Health Statistics. NHANES gathers data from laboratory tests, physical examinations, and questionnaires and includes information on participant demographics, vital statistics, other examination measures, medical history, medication and supplement use, and laboratory values (including RBC folate levels). Sampling weights provided with the data allow for generation of nationally representative estimates of measured parameters. From the original 10,122 participants in the 2005-2006 NHANES (after excluding men, women not of child-bearing age [ie, younger than 14 or older than 45 years], and observations missing folate data), the final study sample included 1710 participants.
In estimating median folate levels using NHANES data, we adjusted reported folate levels upward by 30% to account for underestimation associated with the microbiologic assay used by NHANES at the time of the 2005-2006 survey. Both the US Benefit Study and the NHANES survey data reflect folate levels in a population also receiving folate via food fortification implemented in the United States in 1998.
Impact of Metafolin on RBC folate levels
The change in the median RBC folate level from baseline to week 24 reported in the US Benefit Study was used to estimate the effect of using a Metafolin-fortified OC on RBC folate levels of women in both analysis sets, using baseline data from both the US Benefit Study and the NHANES data. Because the US Benefit Study did not assess RBC folate levels following OC discontinuation, we used data from the discontinuation phase of the European Longterm Folate Study and assumed the same percentage reduction in RBC folate levels over each of the 20 weeks following discontinuation. Because the European Longterm Folate Study followed subjects only for 20 weeks after discontinuation, it was conservatively assumed that RBC folate levels of women who became pregnant more than 20 weeks after discontinuation had returned to pretreatment baseline levels.
Pregnancy outcomes
Both NTD and non-NTD pregnancies were assumed to result in 1 of 3 outcomes: elective termination, fetal death, or a live birth. Spontaneous terminations were not modeled because no data were available. The probability of elective termination for non-NTD terminations was derived from Jones et al. The probabilities of elective termination for NTD cases were modeled separately for spina bifida and anencephaly as a weighted average of termination rates in 6 states. Fetal death rates for non-NTD pregnancies, spina bifida, and anencephaly were estimated separately from Hoyert and were adjusted for elective terminations. All anencephaly live births were identified as such in the model, but these infants were assumed to die shortly after birth.
Sensitivity analysis
One-way sensitivity analyses were performed by varying the base-case input values around the lower and upper limits of their 95% confidence intervals (or 20% for those parameters for which confidence intervals were not available) to assess how robust the results were to changes in the input parameter assumptions. For the sensitivity analysis, the base case assumed baseline folate levels are based on the US Benefit Study.
In addition, the number needed to treat (NNT) to prevent 1 NTD pregnancy was calculated for both the Metafolin-fortified OC intervention based on model outcomes and compared to an NNT calculated for oral supplements, based on either the baseline RBC folate levels from the NHANES data or the US Benefit Study, for comparison purposes. To calculate the NNT for oral supplements, the absolute risk of an NTD pregnancy was calculated, assuming 100% use of oral supplements compared with no use of oral supplement.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
