Contraception for Women With Medical Problems
Siri L. Kjos
Table Of Contents
Siri L. Kjos, MD
CARDIAC LESIONS: CONGENITAL AND ACQUIRED VALVULAR LESIONS, VALVULAR REPLACEMENTS
ASYMPTOMATIC MITRAL VALVE PROLAPSE
DIABETES MELLITUS: TYPE 1 AND TYPE 2
PRIOR HISTORY OF GESTATIONAL DIABETES MELLITUS
OTHER CARDIOVASCULAR RISK FACTORS: HYPERTENSION, HYPERLIPIDEMIA, SMOKING, OBESITY, LATER CHILDBEARING AGE
SYSTEMIC LUPUS ERYTHEMATOSUS
SEIZURE DISORDERS (EPILEPSY)
SICKLE CELL DISEASE
HUMAN IMMUNODEFICIENCY VIRUS
Women with medical disease often face greater obstacles in obtaining effective contraception based on their doctor’s belief that hormonal or intrauterine methods of contraception may place them at increased risk of infectious, metabolic, or vascular complications. Although research and educational efforts addressing contraceptive use in women with medical problems are limited, much of the current evidence in either healthy or diseased women does not support a significant increase in contraceptive risk. Paradoxically, pregnancy in women with medical disease in many instances exposes both the woman and her offspring to definite morbidity and possible mortality. Women with medical disease should be offered the most effective methods of contraception that also preserve their existing health status.
The selection of a contraceptive method for a woman with a medical disease must address the possible adverse or beneficial interactions with the state and complications of the disease and possible drug interactions with her medical therapy. It must be recognized that a risk-free and efficacious contraceptive option may not exist. However, the successful avoidance of pregnancy may be a lower net risk, and a further reduction of any risk for the contraceptive method may be achieved with close surveillance for complications.
This chapter evaluates contraceptive methods with a low failure rate, specifically combination oral contraceptives (COCs), progestin-only oral contraceptives (PO-OCs), injectable and implanted hormonal methods, and the intrauterine device (IUD) in women with various medical problems. Barrier methods are not addressed because they are metabolically neutral and have no medical contraindication to their use except for their higher failure rates. The IUD is metabolically neutral and highly efficacious and has minimal risks when patients are properly selected to be at low risk for sexually transmitted diseases and an aseptic insertion technique is followed. The metabolic and endocrinologic effects of hormonal contraception become important considerations when selecting the best formulation, dose, and route of delivery for their use in women with medical conditions. When a hormonal method is considered in women with medical disease, the first question to consider is whether the prescription of an estrogen, currently only available as in the COC, is beneficial or should be avoided. Second, if a progestin-only method is considered, the various effects of oral, intramuscular, or subcutaneous1,2,3,4,5,6,7,8,9,10,11,12,13,14,15 routes of delivery should be evaluated (Table 1).
HDL-Chol, high-density lipoprotein cholesterol; LDL-Chol, low-density lipoprotein cholesterol.
|CARDIAC LESIONS: CONGENITAL AND ACQUIRED VALVULAR LESIONS, VALVULAR REPLACEMENTS|
Women with congenital cardiac lesions may lead a virtually asymptomatic life with minimal morbidity or may experience severe limitations on their activity with life-threatening disease. The New York Heart Association classification is generally used to assess functional cardiac capacity. Corrective surgery and medical therapy have provided many women with an excellent prognosis for pregnancy. However, several uncorrected lesions including Eisenmenger’s syndrome, primary pulmonary hypertension, cardiomyopathy, congestive heart failure, and Marfan’s syndrome with aortic involvement carry a mortality of up to 25% to 50%. For this reason, pregnancy is not recommended in these women and contraception is mandatory, especially until corrective surgery can be accomplished.
Combination Oral Contraceptives
Improved medical and surgical treatment of congenital heart disease has resulted in an increasing number of young, reproductive age women with congenital cardiac lesions who need effective contraception. Women with acquired valvular disease (e.g. secondary to rheumatic heart disease or intravenous drug abuse, or symptomatic mitral valve prolapse) face similar contraceptive considerations as those with congenital lesions. Both categories of lesions place women at risk of thrombosis owing to increased risk of coagulation from cardiac flow disturbances. Estrogen produces a dose-dependent increase in coagulation factors, further increasing their risk for a thrombotic event. For this reason, a COC containing estrogen should not be prescribed.
Women with cardiac disease or valve replacements who are chronically receiving anticoagulating therapy are an exception to the use of estrogen-containing COCs. Chronic anticoagulation places these women at an increased risk of intraperitoneal hemorrhage from a ruptured corpus luteum cyst, which may require emergency surgery. The use of COCs reduces the risk of functional ovarian cysts and thereby cyst rupture. All women with functional cardiac lesions should be monitored for signs and symptoms of fluid retention or increased blood pressure, especially with the use of hormonal contraception.
Progestin-only methods, including PC-OCs, intramuscular progestins, and subcutaneous implants are acceptable for use in women with functional cardiac disease because of their lack of effect on coagulation and blood pressure. Again, these patients should be monitored for signs and symptoms of fluid retention or increased blood pressure, which may be unrelated to their contraceptive method, but a sign of their cardiac disease.
The copper IUD is metabolically neutral and therefore is an ideal method in women with cardiac disease, owing to its low pregnancy rate and 10 years or more of potential use after insertion. Evidence for its safety can be found in a study by Abdalla and colleagues,16 who followed a large series of women with surgically treated cardiac lesions using copper IUDs. All underwent antibiotic prophylaxis intramuscular ampicillin and gentamicin17 at the time of insertion and removal, and all received daily iron supplementation. None developed bacterial endocarditis. About half of the women using the IUD were receiving long-term anticoagulating therapy, and there was no significant difference in the mean hemoglobin levels of the women who were treated and those who were not. the American Heart Association18 changed their recommendations regarding antibiotic prophylaxis for subacute endocarditis to the insertion and removal of the IUD as well as vaginal delivery, which are no longer considered procedures requiring antibiotic prophylaxis. Studies have not examined the safety of this approach with respect to IUDs, nor have there been any randomized, controlled human trials in patients with structural heart disease examining endocarditis prophylaxis with gynecologic procedures. Patients at highest risk for endocarditis are those with prosthetic heart valves, previous history of endocarditis, cyanotic heart disease, or surgically constructed pulmonary shunts or conduits.18 A consultation with the patient’s cardiologist regarding the need for prophylaxis before insertion of the IUD may be helpful. Standard gynecologic care includes negative cervical examinations for chlamydia, gonorrhea, and cervicitis prior to insertion, patient education regarding pelvic inflammatory disease (PID), and iron supplementation for those women receiving chronic anticoagulation. Giving a nonsteroidal anti-inflammatory agent prior to insertion, using local anesthesia, and reassuring the patient during the procedure can minimize the risk of a vasovagal response during the insertion of the IUD.
|ASYMPTOMATIC MITRAL VALVE PROLAPSE|
Asymptomatic mitral valve prolapse rarely presents any substantial risk to pregnancy, delivery, or contraceptive prescription. Newer data suggest that mitral valve prolapse has been overdiagnosed and is relatively uncommon, being present equally in approximately 2.4% of men and women.19
Women with asymptomatic mitral valve prolapse may use low-dose estrogen COCs. Progestin-only methods, including PO-OCs, intramuscular injectables, or subcutaneous implants, are all acceptable. Theoretically, progestin-only methods are preferable because they have no effect on coagulation.
The IUD is an excellent contraceptive choice for women with asymptomatic mitral valve prolapse. Although the American Heart Association18 no longer recommends prophylaxis for subacute bacterial endocarditis at time of IUD insertion or removal in women with mitral valve prolapse, prophylaxis has traditionally been recommended without any reported infectious complications from IUD use in women with valvular lesions.16
Several medical conditions are associated with an increased risk of thrombosis. These include medical conditions with altered vascular structures such as cardiac valvular disease and previous history of venous or arterial thrombosis and coagulopathies with altering clotting mechanism such as protein S or C deficiency, activated protein C resistance, antithrombin III deficiency, and antiphospholipid syndrome. These patients are often receiving long-term anticoagulating therapy to decrease their risk of thrombosis.
Combination Oral Contraceptives
COCs are contraindicated in women with medical conditions producing a hypercoagulable state secondary to the estrogen dose-dependent increase in coagulation factors. This caution extends to women with a prior history of deep venous thrombosis. Women with hypercoagulable conditions who are receiving long-term anticoagulating therapy may use COCs because this decreases their risk of hemorrhage from ruptured corpus luteal cysts.
Progestin-only forms of contraception are the preferred types of hormonal contraception owing to their neutral effect on coagulation factors. If increased menses or breakthrough bleeding occurs, iron supplementation should be prescribed and the patient should be reassured that this is not unusual or harmful and that she is still protected from pregnancy.
IUDs are metabolically neutral with respect to their effect on coagulation. Their use in women with hypercoagulable conditions is not contraindicated. A retrospective study of copper-medicated IUD use in a large number of women requiring chronic anticoagulation did not find any significant difference in their mean hemoglobin levels compared with a similar cohort of IUD users who were not anticoagulated.16 A progestin-containing IUD, which is associated with decreased menstrual flow, may be beneficial.
|DIABETES MELLITUS: TYPE 1 AND TYPE 2|
Diabetes mellitus is characterized by absolute (type 1) or relative lack (type 2) of insulin, resulting in hyperglycemia that ranges from asymptomatic to severely symptomatic with ketoacidosis, hyperosmolar coma, and protein wasting. End-stage organ damage generally occurs with prolonged disease, including microvascular disease (retinopathy, peripheral neuropathy and nephropathy) and macrovascular disease (hypertension and coronary artery disease). The presence and the increasing severity of diabetic vascular sequelae determine the patient’s prognosis, quality of life, and risk of pregnancy complications. Similarly, the presence of vascular sequelae influences contraceptive choices while increasing the need for effective contraception. Pregnancies in diabetic women should be planned such that good glycemic control is achieved prior to conception and through the first 8 weeks of gestation, to reduce the risk of major congenital malformations that occur in up to 20% to 25% of pregnancies in which glycemic control is poor.20,21 Euglyemic control prior to and early in pregnancy, utilizing multiple daily insulin injections and glycemic monitoring, has been shown to reduce the rate of congenital malformations.22
Combination and Progestin-Only Oral Contraceptives
The lowest dose and potency of progestin formulation, whether the newer, less androgenic progestins or the older preparations that are found in COCs, should be selected to minimize deterioration in glucose tolerance2,23 and lipid metabolism.3 Short-term, prospective studies for up to 1 year have shown combination preparations with lower doses of the older progestins, norethindrone (≤0.50 mg mean daily dose) or triphasic levonorgestrel preparation, or the newer progestins (gestodene, desogestrel) to have minimal effect on diabetic control, lipid metabolism,24,25,26 and cardiovascular risk factors.27,28 All preparations examined in these studies also contained a low estrogen dosage. The PO-OC containing 0.35 mg of norethindrone has also been similarly studied in diabetic women and found to have no significant effect on carbohydrate or lipid metabolism.26
Currently, there are no long-term, prospective studies evaluating the effect of oral contraceptive use on diabetic sequelae. One older retrospective study did report 3 cases of cerebral thrombosis and 1 case of myocardial infarction, which all occurred in the group of 120 women with type 1 diabetes who used oral contraceptives as opposed to 159 diabetic women who never used oral contraceptives.29 Newer retrospective, cross-sectional studies30 and case-control trials31 in women with type 1 diabetes that have not found any increased risk or progression of diabetic sequelae (retinopathy, renal disease, or hypertension) with past or current use of oral contraceptives. In a retrospective case-control study following, for up to 7 years, two groups of young women with type 1 diabetes who had either used or never used oral contraception, there was no difference in the mean glycosylated hemoglobin levels, the mean albumin excretion rates, or retinopathy scores.31 Similarly, in a cross-sectional study of 384 women with type 1 diabetes, no association was found between the use of oral contraceptives, either current or past, and the severity of retinopathy, hypertension, or glycosylated hemoglobin when the known risk factors for diabetic sequelae were controlled for.30 Currently, there are no studies examining either the retrospective or the prospective use of oral contraceptives in women with type 2 diabetes.
The perceived risk of oral contraceptives in diabetic women stems from the drugs’ ability to increase cardiovascular complications and blood pressure. A large cross-sectional study3 as well as short-term prospective trials32,33,34,35 in women with type 1 diabetes have found no association between risk factors for cardiovascular markers and use of current low-dose preparations of oral contraceptives. Furthermore, in large, prospective, cohort trials in healthy women, no evidence can be found for any excess risk of myocardial infarction associated with the use of low-dose oral contraceptives.36,37,38,39 However, a multicenter, case-control study examining acute myocardial infarction in women aged 20 to 44 years, the use of COCs was associated with increased risk of acute myocardial infarction in women with known cardiovascular risk factors, especially in those with hypertension.40 Therefore, COCs should not be prescribed in diabetic women with hypertension.
A woman with type 1 or 2 diabetes using oral contraceptives should be monitored more frequently for changes in blood pressure and weight. A baseline evaluation of weight, blood pressure, and fasting lipids is recommended. Her internist or primary care physician should be consulted and her gynecologist should be aware of her diabetic therapy and home glucose-monitoring regimen. She should return after the initial cycle of therapy and thereafter every 4 to 6 months to evaluate for blood pressure, weight, and glycemic control. Lipids should be reassessed annually in diabetic women, and more frequently if abnormal values are detected, following standard guidelines.41
Owing to the absence of clinical studies examining either depo-medroxyprogesterone acetate (DMPA) or levonorgestrel (Norplant) use in diabetic women, neither can be recommended as first-line methods. If contraindications to estrogens exist, the PO-OC would be preferable because short-term studies have demonstrated its safety.26 However, in select patients for whom compliance is a problem, a long-acting progestin may be considered. Theoretically, levonorgestrel may be preferred over DMPA4,5 because its lack of effect on carbohydrate metabolism was demonstrated in healthy women.7 Recommendations for monitoring are similar to the monitoring recommended for diabetic women using oral contraceptives.
Physicians have been reluctant to prescribe the IUD in diabetic women secondary to the possible increase in risk for PID associated with the IUD, which in diabetic women, may precipitate life-threatening ketoacidosis. The newer, copper-medicated IUDs have not been associated with any increase risk of PID after the postinsertion period. In a large meta-analysis of several prospective World Health Organization (WHO) trials, the overall incidence of PID associated with copper-medicated IUDs was 1.6 per 1000 women-years of IUD use.42 Analysis of prior studies has found that the development of PID and subsequent tubal infertility was not related to the use of the IUD per se, but to the exposure risk for sexually transmitted disease.43,44,45 Similarly, none of the prospective studies examining copper-medicated IUD use in diabetic women with either type 146,47 or type 248 diabetes have found any support for increased risk of PID. In a controlled trial examining the use of the copper-medicated IUD in 103 women with type 1 diabetes and 119 healthy women for 1 year, no pelvic infections or perforations occurred, and the incidence of failure, expulsion, and pain was similar in the two groups.46 The continuation rates at 1 year were similar (87% versus 88%). In another 3-year trial involving 59 women with type 1 diabetes and 1043 nondiabetic women using the copper-medicated IUD, no cases of PID were found after 1754 cumulative months of use in the diabetic cohort.47 In an uncontrolled study examining the Copper T380A IUD in 160 women with type 2 diabetes for up to 3 years, no cases of PID occurred with 3066 cumulative months of use.48 In fact, in all the studies examining IUD use in diabetic women,46,47,48,49,50,51 no demonstrable increase in PID has been found. However, caution must be exercised because the risk of PID is extremely low with use of medicated IUDs in the general population, making it highly unlikely that large enough studies can ever be conducted in diabetic women to confirm the absence of any increase in risk of pelvic infection.52
The copper-medicated IUD, rather than the progesterone-medicated IUD, would be a preferable choice in diabetic women based on its proven low association with PID.42 General gynecologic principles should be followed in proper patient selection, insertion, and monitoring of IUD use. None of the studies in diabetic women used routine antibiotic prophylaxis with insertion or removal, and therefore it appears unlikely to add any benefit.
|PRIOR HISTORY OF GESTATIONAL DIABETES MELLITUS|
Women with prior gestational diabetes mellitus (GDM) have a substantial risk of developing diabetes after pregnancy.53,54 The prevalence of diabetes after GDM varies among certain ethnic groups (e.g., Latinos and American Indians), generally paralleling background rates of type 2 diabetes in these groups.55,56 Periodic testing for diabetes is recommended in women with prior GDM.57 This is especially important prior to a subsequent pregnancy because undiagnosed diabetes and untreated hyperglycemia,58,59 which is often asymptomatic, increase the risk of major congenital malformations. In additional, evidence suggests that a subsequent pregnancy after an initial pregnancy complicated by GDM appears to increase the risk of subsequent diabetes. In a cohort of Latino women, a subsequent pregnancy was found to triple the risk of subsequent diabetes after adjusting for known predictive factors for diabetes.60. For these reasons, an effective contraceptive method that does not accelerate their risk of diabetes is an important consideration.
Combination and Progestin-Only Oral Contraceptives
As in women with overt diabetes, the lowest dose and potency progestin (and estrogen) combination should be selected for use in women with prior GDM to minimize adverse deterioration in glucose tolerance, lipid metabolism, and blood pressure effects. Short-term studies have demonstrated the absence of adverse effects of low-dose progestin and estrogen COCs on glucose and lipid metabolism.61,62,63 In a retrospective cohort of 904 women with prior GDM in whom diabetes was excluded postpartum, the long-term use of low-dose COCs was shown not to influence the development of diabetes.64 The cumulative incidence rates for diabetes were virtually identical for women with 3 years of uninterrupted use of COCs (25.4%) and of nonhormonal forms of contraception (26.5%). The predominant formulations of COCs used in this study and the previous studies61,62,63,64 were the low-dose monophasic norethindrone preparation (0.40 mg) and the triphasic levonorgestrel preparation (0.050 to 0.125 mg). However, breastfeeding women who were using PO-OCs had almost a threefold increase in risk (adjusted relative risk [RR], 2.87; 95% confidence interval [CI], 1.57 to 5.27) in the development of diabetes after delivery, and this risk was increased with the duration of uninterrupted use.64 Thus, the PO-OC should not be given to breastfeeding women with prior GDM. Either nonhormonal methods or starting with low-dose COCs 6 to 8 weeks’ postpartum after the establishment of lactation is recommended. The use of PO-OCs outside of breastfeeding has not been examined in this population. Regardless of the contraceptive method, all women with previous GDM should undergo annual evaluation of glucose tolerance with a fasting plasma glucose57 A confirmed fasting plasma glucose level greater than 126 mg/dL is diagnostic of diabetes, and a fasting plasma glucose level greater than110 mg/dL and less than 126 mg/dL is diagnostic of impaired fasting glucose.65 Effective contraception and pregnancy prevention may decrease the risk of subsequent diabetes in women with previous GDM.
Currently, no studies exist evaluating the metabolic effects of long-acting progestin use in women with prior GDM, and therefore these contraceptive methods cannot be recommended as first-line choices in these women. Deterioration of carbohydrate tolerance with the use of DMPA has been demonstrated in healthy women4 but not with use of levonorgestrel.6 However, in patients with hypertension or in whom daily compliance may be problematic, a long-acting progestin may be desirable. Levonorgestrel may be preferred over DMPA because of its lack of effect on carbohydrate metabolism and the option of removal, should unwanted effects occur. Annual testing for diabetes should be performed.
No contraindications exist for IUD use in women with prior GDM. General guidelines regarding IUD use should be followed.
Generally, pregnancy is not recommended in women with known cardiovascular disease (prior myocardial infarction, angina, prior cerebrovascular accidents) owing to the increased cardiac demands during pregnancy and high maternal morbidity. Women without cardiovascular disease but with cardiovascular risk factors, that is obesity, smoking, maternal age older than 40 years, hypertension, and diabetes also face increased risks with pregnancy and require effective contraception that does not additionally increase their cardiovascular morbidity.
Combination Oral Contraceptives
Estrogen-containing COCs generally are contraindicated in women with cardiovascular disease secondary to the increase in coagulation factors produced by estrogen, increasing the risk of vascular occlusion. Older studies using high-dose estrogen COCs reported an increase in myocardial infarction that was related to increased arterial thrombosis.66,67Accelerated atherosclerosis, which is involved in the pathophysiology of coronary artery disease, has never been demonstrated in either primate68 or human studies or in former users of oral contraceptives.36 A multicenter, case-control study found very few cases of acute myocardial infarction occurring in women aged 20 to 44 years without cardiovascular risk factors and who had blood pressure monitored prior to oral contraceptive prescription.40 This study found that the use of COCs was associated with increased risk of acute myocardial infarction in women with known cardiovascular risk factors, especially in those with hypertension and smoking. Interestingly, preliminary results from a second international case-control study suggest that the use of third-generation COCs (containing gestodene or desogestrel) appears to be associated with a nil effect or reduced risk for myocardial infarction compared to second-generation COCs (containing less than 50 μg of estrogen).69 Whether this reduction can be attributed to the new progestin components or the decreased estrogen dose is not clear.
The emerging data from large case-control studies regarding COC use and stroke suggests that there may be additional risk in users with cardiovascular risk. In a two large case-control studies from the United States70 and Europe,71 the adjusted risk for ischemic or hemorrhagic stroke was not increased with current or former use of low-estrogen COCs. However, the risk was increased when cardiovascular risk factors, that is hypertension, diabetes, smoking, migraine, or prior thrombotic disease, were present. The presence of hypertension appears to be a particularly strong risk factor, being associated with hemorrhagic stroke in hypertensive women using COCs.72,73 In the Collaborative Study for Stroke in Young Women, the risk for hemorrhagic stroke was also shown to be increased with increasing blood pressure.72 Another international case-control study also found an increased risk in ischemic stroke related to the use of oral contraceptives, primarily in hypertensive women, smokers, and women whose blood pressure was not measured prior to oral contraceptive prescription.4 The risk of stroke has also been shown to be higher with the use of COC preparations containing 50 μg or more of estrogen.74,75 Although there is a twofold to fourfold increase in relative risk of thromboembolic stroke in women using COCs, the absolute risk of occlusive stroke for women using low-dose COCs is low, approximately 1 per 12,000 women.76 The use of first-generation COCs (> 50 μg of estrogen) was associated with a greater risk compared with second- (< 50 μg estrogen) and third-generation (containing gestodene or desogestrel), which appear to be similar in risk.76
Although prospective studies have not examined COC use in women with either cardiovascular disease or cardiovascular risk factors, particularly with hypertension and smoking after age 35 years, prescription of COCs are not advised in these women.69,74,75
Progestin-Only Oral Contraceptives and Long-Acting Progestins
Progestin-only methods are the oral contraceptive of choice in women with cardiovascular disease or risk factors owing to their neutral effect on coagulation factors and blood pressure.13 Lidegaard75 found that the use of PO-OCs was not associated with any increased risk of cerebral thromboembolism.
The metabolic neutrality of the IUD makes it an excellent choice in women with cardiovascular disease or risk factors. However, no studies have examined the use of the IUD in this population. General guidelines regarding IUD use should be followed.
|OTHER CARDIOVASCULAR RISK FACTORS: HYPERTENSION, HYPERLIPIDEMIA, SMOKING, OBESITY, LATER CHILDBEARING AGE|
In addition to diabetes, hypertension, dyslipidemia, smoking, or increasing age are risk factors for the development of cardiovascular disease. They are often present concurrently, and the presence of more than one risk factor further increases cardiovascular risk. Similarly, the treatment of one factor, such as obesity, will often improve another coexisting risk factor such as hypertension, hyperlipidemia, or diabetes. The first line of therapy is to implement a healthy lifestyle—daily exercise, weight loss, and dietary changes (weight reducing, low salt, and low fat), and to stop smoking. Second-line therapy includes medical management. Untreated risk factors over time produce end-organ damage and atherosclerotic changes in the vasculature as well as continued risk of a cardiovascular event. Contraception should be effective without increasing cardiovascular risk.
Combination Oral Contraceptives
No prospective studies have been published examining the safety of COCs specifically in women with hypertension, hypercholesterolemia, obesity, or other cardiovascular risk factors. Most of our information regarding the interactions between cardiovascular risk factors, such as smoking and hypertension, and COCs has come from long-term cohort studies in the general population or case-control studies.
The Royal College of General Practitioners’ Oral Contraception Study in 1981 first documented the interactive effect of smoking in women,over the age of 35 years and the use of older high-dose COCs.77 They reported a 20-fold increase in risk of myocardial infarction, which was increased in heavy smokers compared with nonsmokers who did not use COCs.78 This association was later corroborated by the 1997 WHO study,79 which found a substantial increase in risk of myocardial infarction in older women who smoke (approximately 400 per 106 woman-years) compared with similar age nonsmokers (approximately 30 per 106 woman-years). A Women’s Health Consensus Conference addressed the current literature and continued to recommend that COCs should be denied to women older that 35 years who smoke heavily (more than 15 cigarettes per day).80 Preliminary data suggest that the use of a low-estrogen COC (20 μg ethinyl estradiol) may be preferable in women who smoke because of the reduction in renin substrate and coagulation parameters.81
As reviewed in the previous section (Cardiovascular Disease), use of COCs in the presence of hypertension and significant smoking increases the risk for a cardiovascular event. The use of COCs in healthy women has been associated with mild increases in baseline blood pressure.13,82 In older studies, high-estrogen dose COC use induced hypertension in up to 5% of women83,84 An analysis from the ongoing, prospective Nurse’s Health Study II85 found that the adjusted risk for the development of hypertension in healthy women with current COC use was increased (RR,1.8; 95% CI, 1.5 to 2.3) compared with nonusers. This risk was increased with 6 or more years of use (RR, 2.1; 95% CI, 1.6 to 2.7) and decreased shortly after cessation of COC use (RR, 1.9; 95% CI, 0.9 to 1.1). However, the actual cases attributable to COCs were low, 41.5 cases per 10,000 person-years.
In a case-control study of 61 women with past history of hypertension, currently well controlled without medication, Tsai and associates86 did not find a significant increase in mean blood pressure from baseline in the majority of patients with previous hypertension followed for variable lengths of time. During the study, 5 (8.2%) women with a prior history of hypertension redeveloped significant hypertension mandating discontinuation of COCs. Extrapolation from this study to support the safety of COC use for hypertensive women in general is difficult. First, the subjects were not hypertensive or taking antihypertensive medication at study entry. Second, the number of subjects was small, with only 10 subjects still using COCs at 24 months. Lastly, only 17 of the 61 subjects had essential hypertension, with the remaining subjects having prior hypertension associated with pregnancy or prior COC use.
In summary, evidence from case-control studies and large cohort studies have consistently demonstrated adverse associations between the use of COCs and hypertension. However, many experts do not consider the use of low-dose COCs in women with well-controlled hypertension to be contraindicated.86,87,88,89 Should COCs be prescribed in hypertensive women, a program of frequent surveillance of blood pressure and serum lipids is mandatory. Owing to the dose-dependent effect of estrogen on renin substrate production, the prescription of a COC with very low estrogen dose (20 μg) may be preferable in women with controlled hypertension.81
Abnormal serum lipoprotein levels and metabolism can lead to atherosclerosis and cardiovascular disease. Whereas atherosclerosis and cardiovascular disease are rare in women younger than 35 years, at-risk levels of serum lipids are more common.90,91 Elevated total cholesterol, low-density lipoprotein (LDL) cholesterol and decreased high-density lipoprotein (HDL) cholesterol levels are independent predictors of cardiovascular risk. The National Cholesterol Education Program41 recommends screening every 5 years in all healthy adults. In those with abnormal total cholesterol and HDL levels, a stratified treatment plan of diet and medical therapy is recommended, which is based on cardiovascular risk factors and lipid levels. This strategy targets individuals with two or more cardiovascular risk factors for more intensive therapy at lower lipid levels.
No studies have examined the metabolic effects with COC use in women with hyperlipidemia. Generally, many, if not the majority of, women are unaware of their cholesterol levels. Epidemiologic data assessing the cardiovascular risk associated with COC use for women with dyslipidemia are not available as it is for hypertension and smoking. The case-control studies examining OCs did not evaluate the effect of serum lipid levels as risk cofactors for either acute myocardial infarction,69,78,79 or stroke.70,71,74 Knopp and colleagues92 have proposed a schema for COC prescription in women with dyslipidemia that parallels the National Cholesterol Education Program Guidelines. They recommend COC prescription in women who have less than one risk factor and LDL cholesterol levels below 160 mg/dL and in women with two or more risk factors and LDL cholesterol level below 130 mg/dL. They estimate that this approach would exclude very few women below the age of 35 years; however, over the age of 35 years, approximately 40% of women would be excluded by the presence of two or more risk factors. In healthy women, the use of COCs formulated with less androgenic new progestins results in significant increases in HDL cholesterol and triglyceride levels and decreases in LDL cholesterol levels.8 Thus, they might be a preferable choice for women with elevated LDL and decreased HDL cholesterol levels. However, estrogen-dominant COCs may exacerbate hypertriglyceridemia (greater than 250 mg/dL) and either a less estrogenic COC or a PO-OC is preferable. Monitoring of serum lipid levels should be frequent enough to ensure successful control of lipid abnormalities.
Progestin-Only Methods: Oral Contraceptives, Injectables, and Implants
A 1-year prospective study in healthy women found blood pressure to be increased with COCs, but not with PO-OCs.13 In another cross-sectional study, PO-OCs were found to have no effect on blood pressure when controlling for cofactors such as hypertension therapy, body mass index, physical activity, and alcohol use.93 Whereas no studies have examined the use of either injectable or implanted progestins in women with cardiovascular risk factors, prospective studies in healthy women found no increase in blood pressure after 113 to 2 years94 of use. Similarly, subdermal levonorgestrel implants (Norplant) had no effect on blood pressure in women followed for up to 5 years.95
No studies exist examining PO-OCs in women who smoke. Owing to the neutral effect of progestin-only methods on coagulation factors and globulin production, the use of progestin-only methods is not felt to be contraindicated in women who smoke.
Again, no studies have examined progestin-only contraception in women with dyslipidemia. In the cross-sectional study by Godsland and colleagues3, PO-OC users had favorable lipid and carbohydrate metabolic markers for coronary artery disease. The effect of DMPA injections on lipid metabolism does not appear to be beneficial. In a review of this topic, Westhoff96 found that in all studies in which HDL or LDL cholesterol levels were measured, the mean HDL cholesterol levels were decreased and LDL cholesterol levels were increased.10,14,97 In contrast, levonorgestrel appears to have minimal, if any, overall effect on serum lipid levels.98 In summary, in the absence of data in women with dyslipidemia who use PO-OCs, lipid levels should be monitored periodically, with a re-evaluation within 3 to 6 months of initiating hormonal contraceptives.92 Progestin-only methods may be preferable in women with overt hypertriglyceridemia, avoiding estrogen-induced increases in triglycerides.
Although the IUD has not been specifically studied in women with cardiovascular risk factors, the metabolic neutrality of the IUD makes it an excellent choice, especially in women with multiple cardiovascular risk factors. General guidelines regarding IUD use should be followed. Irregular or anovulatory vaginal bleeding and endometrial hyperplasia are more frequent in obese women and should not be attributed to IUD use. Endometrial biopsy should be used to establish the etiology and a progesterone IUD may be considered to decrease the risk of hyperplasia.
|SYSTEMIC LUPUS ERYTHEMATOSUS|
Systemic lupus erythematosus (SLE) is a chronic and progressive disease primarily attacking young women during their reproductive years. It has a variable course, which is characterized by exacerbations and remissions. Autoantibodies bind to tissues, producing immune complex inflammation of various organs (causing nephritis, arthritis, dermatitis) and specific cytotoxic damage (causing hemolytic anemia, thrombocytopenia). Progressive damage leads to end-organ disease, including renal failure and hypertension. Exacerbations are treated with high-dose immunosuppressive and/or cytotoxic therapy. A pregnancy should be planned when the disease has been in remission for at least 6 months to reduce the risk of perinatal mortality and morbidity. If pregnancy occurs during or shortly after active disease, the risk of complications, primarily hypertension and nephropathy, can make the pregnancy course very stormy with up to a threefold increase in abortion, stillbirth, prematurity, and preeclampsia.
Combination Oral Contraceptives
The predilection of SLE for women, with onset during their reproductive years and exacerbation during pregnancy, has suggested an influence of sex hormones, particularly estrogen, on development of the disease or its activity. In a case-control study exploring possible etiologic factors for SLE, no association was found with the prior usage of oral contraceptives, but a significant association with a family history of autoimmune disease, history of shingles, and medication allergies was found.99 In the prospective Nurses’ Health Study, the relative risk for SLE, defined by displaying at least the four criteria of the American College of Rheumatology with previous use of oral contraceptives was 1.4 (95% CI, 0.9 to 2.1) compared with never users after adjusting for age and exposure to postmenopausal hormone.100 When the researchers required a certified rheumatologist to confirm the diagnosis, the association with past oral contraceptive use became significant (RR, 1.9; 95% CI, 1.1 to 3.3), but this risk was not found to be associated with duration of oral contraceptive use or interval since past oral contraceptive use. In summary, any increase in risk of developing SLE from exposure to oral contraceptives is likely to be very small, if present at all, in healthy women. Several case reports have reported lupus flares associated with estrogen-containing COCs.101,102,103 Thrombosis and hypertension, both of which can be associated with COC use, are also serious sequelae of SLE. One retrospective study in women with severe forms of SLE reported that 9 of 21 women (43%) developed lupus flares with COCs containing high- and low-dose estrogen preparations.104 A second group of 11 patients in this study who received PO-OCs had no exacerbation whereas 6 of 30 (20%) control patients with SLE had lupus flares during the study period. Evidence from an animal model for SLE, murine lupus, suggests a deleterious effect of estrogens and a protective effect of androgens on the progression of lupus and production of autoantibodies.105,106. In humans, anti-ethinyl estradiol antibodies were detected in 30% of healthy COC users, and 57% of female COC users with SLE.107 Interestingly, no anti-ethinyl estradiol antibodies were found in female non-COC users, but were paradoxically found in 13% of males. Lastly, one cross-sectional review of contraceptive methods in SLE patients, reported a nonsignificant increase in the risk of deep venous thrombosis while using COCs, (RR 2.3; 95% CI, 0.5 to 10.3).108 In summary, estrogen-containing COCs are generally not advised in women with SLE.
Progestin-Only Methods: Oral Contraceptives, Injectables, and Implants
Progestin-only hormonal contraception avoids blood pressure increases and exacerbation of hypercoagulability associated with SLE. The use of PO-OCs has been reported in several retrospective studies104,108 without apparent increase in lupus flares or complications. In a prospective, controlled trial in women with SLE in remission, 10 women received intramuscular norethisterone enanthate (200 mg) every 3 months, 15 women received levonorgestrel-only oral contraceptives, and 18 women used IUD or barrier methods in the control group.109 The frequency and severity of lupus flares occurred equally in the 10 women receiving progestin during 170 months of exposure and in 9 women in the control group during 198 months of follow-up. Although morbid events occurred in both groups, including one acute myocardial infarction in a norethisterone-treated subject with chronic hypertension and glomerulonephritis controlled by daily corticosteroid therapy, and two cases of pulmonary thromboembolism, occurring one each in the progestin treated and control groups. Whereas this study found no increase in adverse morbidity with progestin-only methods in women with inactive disease, the numbers are small. Close surveillance for complications is advised. The PO-OC would be preferable to either DMPA or levonorgestrel because it has been better studied and is easily discontinued. However, in select patients in whom compliance is a problem, a long-acting progestin may be considered. Levonorgestrel has the advantage of being able to be immediately discontinued by the removal of the implant, whereas the effects of DMPA are not reversible and persist for variable periods of time.
As in all women with chronic disease, consultation with her internist to develop a contraceptive plan is appropriate.
In a cross-sectional, controlled, retrospective study of 85 women with SLE interviewed about their past history of contraception, 44 women reported either currently or previously having used a copper-medicated IUD.108 There were no reported complications of severe uterine bleeding or pelvic infection, and the 3-year continuation rate was 43%. The concurrent use of corticosteroid therapy was not addressed in this study.
In women with SLE on immunosuppressive therapy, careful aseptic technique and prophylaxis with azithromycin or doxycycline during insertion and removal are recommended. Consideration may be given to the levonorgestrel IUD if the patient is anemic.
|SEIZURE DISORDERS (EPILEPSY)|
Epilepsy is characterized by chronic, recurrent paroxysmal changes in neurologic function produced by abnormalities in the brain cortex. Seizures are variable as to their etiology and are classified by their presentation—generalized tonic-clonic, absence, simple or complex partial, or status epilepticus. Antiepileptic drugs are prescribed in monotherapy or in combination and are titrated to achieve the lowest dose effective to control the seizures. Several of the antiepileptic drugs are associated with a twofold increased risk of congenital malformations, presenting as associated fetal syndrome (hydantoin and trimethadoin) or isolated malformations (valproic acid). Women with reproductive potential must be counseled regarding the teratogenic potential as well as their absolute need to continue with antiepileptic therapy until advised to do otherwise by their neurologist.
Combination Oral Contraceptives
Several antiepileptic drugs, phenobarbital, phenytoin, paramethadione, and carbamazepine, but not benzodiazepines and valproate have been shown to induce the hepatic microsomal oxidase system and accelerate the metabolism of several drugs.110 Animal models when treated with phenobarbital111 and phenytoin110 demonstrate accelerated catabolism of exogenously administrated oral steroids. The degree to which enzyme induction occurs is not predictable. Monitoring of serum steroid levels in response to antiepileptic drugs such as phenobarbital112 shows highly variable serum levels, which may or may not correlate with suppressed ovulation. Thus, there exists a biologic mechanism for possible increase in contraceptive failures and breakthrough bleeding when oral contraceptives are used in women taking antiepileptic drugs that induce hepatic enzymes. Several case reports have dealt with oral contraceptive failures in women using antiepileptic drugs and high-dose estrogen COCs (50 to 100 μg ethinyl estradiol). In one retrospective study of epileptic women, there were 3 COC failures in 41 women taking concomitant antiepileptic drugs with 955 months of COC exposure compared with no pregnancies in 41 epileptic women not treated with antiepileptic drugs with 2278 months of COC exposure.113 No information on daily compliance with COC use or the types of antiepileptic therapy was provided in this study. To date, there have been no prospective, controlled studies examining COC use and antiepileptic drugs.
The use of COCs does not increase seizure frequency or severity.114,115 Women on antiepileptic therapy who desire to use COCs should be so advised that the concomitant use of COCs and antiepileptic drugs may be associated with an increased rate of COC failure and breakthrough bleeding. Initial formulations of COCs containing 30 to 40 μg of ethinyl estradiol are appropriate.116 A higher-estrogen (50 μg) (and thereby higher progestin dose) COC may be considered if an unplanned pregnancy would pose special problems to the patient. Education of the patient and communication with her neurologist are crucial to relate changes in antiepileptic medication and how to handle breakthrough bleeding, missed pills or menses.
Progestin-Only Oral Contraceptives
The therapeutic use of PO-OCs to control menstrual epilepsy has been investigated. A case report documented the use of a PO-OC containing norethindrone to successfully treat a woman with menstrual epilepsy.117 A follow-up controlled trial failed to show any difference in seizure frequency in women receiving norethindrone compared with placebo.118 No studies have addressed the efficacy of PO-OCs in women with epilepsy. PO-OCs contain lower doses of progestin than their COC counterparts. Thus, without evidence of their efficacy, they are not generally recommended. If preparations containing estrogen are contraindicated for other medical reasons, DMPA should be considered.
In women, seizure frequency has long been noted to be affected by the menstrual cycle, decreasing during the midluteal phase and increasing at the end of the menstrual cycle, suggesting that high levels of progesterone were protective from seizures.119 This has given rise to the entity of catamenial or “menstrual” epilepsy. Backstrom120 correlated the decrease in seizure frequency to high serum progesterone levels and the increase in seizures during the follicular phase with increasing estrogen levels. A case report described the use of DMPA to successfully treat intractable seizures.121 In a study of 14 women with uncontrolled seizures on antiepileptic medication, the administration of DMPA resulted in a significant decrease in seizure frequency in half of the women.122 Of these 7 women, all developed amenorrhea after DMPA administration and had a 52% reduction in seizure frequency. This study correlated the DMPA therapy with a cessation of menstrual cycling and found no change in the levels of antiepileptic drug levels after treatment with DMPA. Injections were given more frequently, every 6 to 12 weeks, to subjects with breakthrough bleeding. Although this study was limited in subjects, DMPA should be considered as first-choice hormonal contraceptive method for epileptic women secondary to its antiepileptic properties and high efficacy.
In contrast, the use of levonorgestrel is not recommended in epileptic women on antiepileptic drugs secondary to a high failure rate. In a prospective, controlled 1-year study of levonorgestrel implants in epileptic women, 2 of 9 women using anti-epileptic drugs became pregnant compared with no pregnancies occurring in a control group of 10 healthy women.123 No change in seizure activity was noted with the use of levonorgestrel. The serum levels of levonorgestrel were noted to be significantly lower during the study in the 6 patients taking phenytoin either alone or in combination with other antiepileptic drugs Other researchers have also documented lower circulating levels of levonorgestrel in Norplant users with concurrent phenytoin ingestion.124
Although the IUD has not specifically been studied in women with epilepsy, there are no contraindications to its use. Because of its metabolic neutrality, lack of drug interaction, and high efficacy, it should be considered a first-line method.
Migraines or vascular headaches are divided into classic or common migraines. Classic migraines are characterized by well-defined symptoms of neurologic dysfunction such as scotomata, hemianosmia, unilateral paresthesias, aphasia, hemiparesis, and hemisensory defects that typically precede the headache. The common migraine is similar to the classic migraine but without neurologic symptoms. It often begins unilaterally and is associated with malaise, nausea, vomiting, or photophobia. In addition, it can be brought on by “triggers” such as menses, red wine, cheese, chocolate, lack of sleep, or stress.
Combination Oral Contraceptives
Migraine headache has been identified as a risk factor for stroke in premenopausal women,71,125,126 with some studies suggesting that classic migraines (with aura) pose a much higher risk for stroke compared with simple migraines (without aura).127,128 The association between timing of migraine with menstrual cycling and the higher prevalence of migraines in women has suggested a link between female hormones and migraines. By extension, Bickerstaff129 proposed that oral contraceptive use in migraine suffers may predispose users to strokes, and he recommended that women should discontinue oral contraceptives if their migraine changed from simple to classic in its presentation. The association, if any, between migraine, COC use and stroke is complex. In an early study, the Collaborative Group for Study of Stroke130 found the combination of migraine and (high-dose) COC use to have a similar increase risk in thromboembolic stroke as would be expected from COC usage alone. In a later case-control study a using multivariate analysis to examine risks for thromboembolic stroke in young women, COC use was not found to have a significant impact but diabetes, migraines, hypertension, and a previous history of thrombotic disease were independently associated with an increased risk for stroke.71 Another case-control study examining ischemic stroke in young women found a significant association between ischemic stroke and migraine that significantly increased with the use of high-dose COCs or smoking, particularly in women with classic migraines.128 Lastly, in a multinational WHO case-control study131 designed to examine the association of migraine with stroke in young women, found migraine, both classic and simple, to be associated with ischemic stroke (adjusted odds ratio [OR], 3.54; 95% CI, 1.30 to 9.61) but not with hemorrhagic stroke (adjusted OR, 1.10; 95% CI, 0.63 to 1.94). Furthermore, the current use of COCs with either hypertension or smoking in migraine sufferers had a multiplicative effect on the OR of ischemic stroke. Lastly, if migraine sufferers smoked and used COCs concurrently, the OR for ischemic stroke was 34.4 (CI, 3.3 to 361) compared with nonmigraine sufferers who did not smoke or use COCs. It should be recognized that the rate of cases of ischemic stroke in young women is very low, estimated at 5.5 cases per 100 000 woman-years, and was found to be increased approximately threefold in young women with migraines.131 This latest study challenges the concept that COCs can be used in women with simple but not classic migraines because they appear to have similar rates of ischemic stroke. Women with migraines should be counseled regarding their increased risk for stroke and strongly advised not to smoke. Blood pressure should be closely monitored if COCs are prescribed. Prescription of COCs in migrainous women who smoke should be avoided based on over a 30-fold increased risk in stroke.131 The lowest estrogen dosage of COCs would appear to be the best choice.
Lastly, menstrual migraines have been reported to be successfully treated with monophasic low-dose COCs and may be given continuously without pill-free intervals.132 Several studies have found no evidence that the use of COCs increases the severity or frequency of migraine headaches.133,134
No studies have examined progestin-only contraceptive methods in migraine uses. Extrapolating from the studies that have examined oral contraceptive use in relation to stroke, the risks of thromboembolic stroke are not increased with PO-OCs.75 Progestin-only methods also do not affect blood pressure or smoking risks, both of which appear to multiplicatively increase the risk of ischemic stroke in migraine uses. Although their use has not been specifically studied in migraine sufferers, PO-OCs would seem to be an ideal oral contraceptive in this population.
The effect of hormonal contraception on headache frequency is difficult to predict. Both PO-OCs and levonorgestrel can be discontinued if headaches worsen after use, whereas the effects of DMPA may be long-lasting. Again with menstrual migraines, the continuous hormone release of the long-acting progestins may provide uninterrupted relief. The PO-OCs can also be given continuously in women with menstrual migraines. In summary, there are no apparent contraindications to progestin-only methods in migraines.
Although the IUD has not specifically been studied in women with migraines, there are no contraindications to its use. Because of its metabolic neutrality, lack of drug interaction, and high efficacy, it should be considered a first-line method.
|SICKLE CELL DISEASE|
Several hemoglobinopathies producing sickling, including homozygous sickle cell disease (HbSS), sickle cell hemoglobin C disease (HbSc) and sickle thalassemia disease (HbS ~ thalassemia). Integral to the pathophysiology of sickle cell disease is the morphologic change in the red blood cell, or sickling, which produces a rigid, crescent-shaped blood cell that is unable to transverse the microcirculation, resulting in obstruction, thrombosis, and tissue hypoxia. Sickling is generally triggered by hypoxia associated with infection, acidosis, or dehydration and may produce severe hemolytic anemia. Over time, macroinfarcts may occur, which can result in end-organ damage, that is renal failure, impaired pulmonary function, or congestive heart failure.
Combination and Progestin-Only Oral Contraceptives
Although both COCs and sickle cell disease are associated with thrombotic complications, their etiology is different. Thrombotic events associated with COCs are related to the estrogen dose-dependent increase in coagulation factors, whereas in sickle disease, the sickling of red blood cells produces destruction of trapped erythrocytes, which release thromboplastin and cause thrombus formation. Although there are older, isolated case reports of thromboembolism with COC use in women with sickle cell disease, there is little, if any, evidence of an additive or synergistic effect to increase the risk of thromboembolism.135 The literature does not support an increase in thrombus formation or an increased incidence or severity of sickle crisis136]with COC use in women with sickle disease. Furthermore, in vivo studies on HbSS erthyrocytes did not find an increase in red cell deformability when exposed to therapeutic concentrations of estradiol and progesterone.137 In summary, low-dose COCs offer a safe and effective contraceptive method for women with sickle disease.135
By extension, there are no contraindications to the use of PO-OCs in women with sickle disease. Studies have not specifically examined the use of PO-OCs separate from COCs. They offer an alternative to women with sickle disease who desire oral contraceptives but also have a contraindication to estrogen, such as hypertension.
Long-Acting Progestins: Injectables, and Implants
Several steroid compounds, including testosterone, progesterone, and estradiol, have been shown in in vivo and in vitro to stabilize the red blood cell membrane, rendering them less fragile.138 In fact, treatment of sickle cell disease with steroids has been shown to be of clinical benefit. Weekly doses of intramuscular progesterone (10 mg) in women and testosterone (10 mg) in men were shown to decrease the severity and duration of sickle pain and crisis in 80% of subjects compared with a 20% reduction in a control group on placebo injections.139 In an elegant, 2-year study, 23 subjects with sickle cell disease received either DMPA (150 mg every 3 months for three cycles) or placebo in a blinded, crossover trial with a 6-month washout between medication.140 During the treatment phase with DMPA, there were significant increases in the levels of fetal and total hemoglobin, the count, mass, and survival of red cells, decreases in reticulocyte and irreversibly sickled cell counts, and serum total bilirubin, and a decrease in episodes of bone pain compared with the placebo-free periods. The improved survival of red cells was independent of the decreased menstruation produced by DMPA. A 1-year study randomized 27 women with sickle cell disease to receive either DMPA injections (150 mg monthly for the first 3 months, then every 3 months) or a COC (0.15 mg levonorgestrel and 30 μg ethinyl estradiol) with a surgically sterilized control group.141 By the end of the study, 70% of the DMPA users were pain-free, compared with 45.5% of the COC users and 50.5% of the controls. This study, in contrast to the preceding De Ceulaer study,140 did not find any significant improvement in hematologic-parameters with either hormonal method. This may be explained in part from a different study design and relatively small numbers in each group compared with the increased power from a controlled, blinded trial with paired analysis. In summary, DMPA is a first-choice contraceptive method for women with sickle cell disease that not only provides excellent pregnancy protection but also decreases complications from sickle disease.
Levonorgestrel use also provides a long-acting and effective alternative to DMPA, although it has not demonstrated the clinical therapeutic benefits in decreasing painful crisis or improving hematologic parameters. In an uncontrolled study of 25 women with sickle cell disease, each woman received levonorgestrel implants and was followed for a mean of 1 year.142 No change from baseline of hematologic or biochemical markers was found, and levonorgestrel was well tolerated.
The potential side effects of PID and anemia have been suggested as contraindicating the use of the IUD in women with sickle disease. However, there has been no evidence of any increase in risk of pelvic infection with IUD use in women with HbSS disease. In a large Jamaican clinic treating sickle cell disease, the IUD has been the long-preferred method of contraception, based on efficacy, long-term protection, cost, and low rate of complications.143
|HUMAN IMMUNODEFICIENCY VIRUS|
In women, the human immunodeficiency virus (HIV) is most frequently acquired through heterosexual intercourse and has become a leading cause of death in women of reproductive age. The last few years have witnessed treatment breakthroughs, notably, for women the opportunity to substantially decrease the risk of vertical transmission to their newborn by azidothymidine (AZT) prophylaxis during pregnancy and the newborn period. Multiple drug therapy strategies using both antiretroviral and protease inhibitors have changed the face of HIV infection toward a managed but long-term, chronic disease, thereby increasing the need for effective prevention of HIV transmission and effective and safe contraceptive options.
Human Immunodeficiency Virus Transmission: Prevention and Risks
Consistent condom use has been shown to be highly effective in the prevention of HIV infection in heterosexual couples in which one partner is infected.144 Condom use also reduces the transmission of most other sexually transmitted diseases, including gonorrhea, chlamydia, and trichomoniasis, and genital ulcerative infections, all of which have been shown to further increase the risk of HIV seroconversion.145 The efficiency of male-to-female transmission of HIV infection appears to be approximately twofold greater than female-to-male transmission146 after adjusting for known covariants.147 Currently, condoms are the only contraceptive method shown to reduce infectivity and susceptibility to HIV. Spermicidal agents have not been shown to reduce viral transmission. Thus, an essential component of HIV care is education and reinforcement about the importance of consistent condom use to reduce transmission. Little data are available on contraceptive use in HIV-positive women. In a survey of 412 HIV-infected women in France followed in an HIV program from 1988 to1993, only a small minority reported using oral contraceptives, 10.2% used oral contraceptives with condoms, and 6.5% used oral contraceptives without consistently using condoms.148 In contrast, 50% systematically used condoms, 28% were sexually inactive, 2% used the IUD or diaphragm, and approximately 20% were not using any contraception. The elective abortion rate in this study was high, suggesting many unplanned pregnancies. The lack of other effective contraceptive prescription in addition to the condom, as documented in this study, likely reflects the emphasis that has been placed on HIV transmission and the medical community’s fear that if other effective methods are provided, the use of the condom will be less consistent. Whereas condoms do reduce the risk of pregnancy, the issue of contraception should be addressed separately, allowing the woman to choose a more effective method that she has control over in addition to using condoms for prevention of HIV transmission. A continued educational effort is needed. The recent Women’s Interagency HIV Study found that only 63% of HIV-positive women always used condoms compared with 28% HIV-negative, at-risk women.149 Furthermore, this study found that when HIV-positive women used an additional contraceptive method, they were slightly less consistent condom users than the HIV-negative cohort (57% versus 67%, respectively).
In some cross-sectional studies, the use of hormonal contraception has been positively associated with increased viral shedding from the cervix and vagina. In a study of African women attending a sexually transmitted disease clinic, increased viral shedding was associated with the use of oral contraceptives, the presence of cervical mucopus, cervical ectopy, and pregnancy.150 This study did not control for the infection status, for example the serum CD4 count or viral load, and only a small number of subjects used oral contraceptives (8 of 84). The authors suggested that cervical ectopy, which was present in 6 of the 8 oral contraceptive users, increases in response to states of elevated reproductive hormone levels, for example in pregnancy and with use of COCs, and it may increase the rate of cervical viral shedding. More recently, data from another cross-sectional study in Africa that adjusted for the CD4 count also found cervical viral shedding significantly increased with the use and dosage of oral contraceptives (OR, 3.8; 95% CI, 1.4 to 9.9 for low-dose COCs; and OR, 12.3; 95% CI, 1.5 to 101 for high-dose COCs).151 This study also reported a similar increase in cervical viral shedding with the use of DMPA injections (OR, 2.9; 95% CI,1.5 to 5.7), vitamin A deficiency, gonococcal cervicitis, and vaginal candidiasis.
Other studies have examined the risk of seroconversion of HIV-negative women and hormonal contraception. An Italian cross-sectional trial examining the risk factors for man-to-woman transmission of HIV, awareness of partner’s seropositivity, condom use- and oral contraceptive use (OR, 0.3; 95% CI, 0.1 to 1.0) were negatively associated with HIV transmission.152 Frequent intercourse, increased time (2 to 5 years) of exposure to HIV, low partner CD4 count, anal sex, vaginitis, genital warts, and the use of IUDs (OR, 3.1; 95% CI, 1.4 to 7.1) were positively associated with increased risk of HIV transmission. A meta-analysis using data from 28 articles to estimate the association between HIV seroprevalence and oral contraceptive use found an overall summary OR of 1.19 (95% CI, 0.99 to 1.42).153 The summary OR for the eight best studies was higher for oral contraceptive use (OR, 1.60; 95% CI, 1.05 to 2.44). Another prospective study examining seroprevalence in African female sex workers found that the use of DMPA was associated with an increased hazard ratio (2.0; 95% CI, 1.3 to 3.1) after adjusting for demographic variables and significant covariates but not high dose of oral contraceptives (OR, 2.6; 95% CI, 0.8 to 8.5).154 Overall, these findings suggest that women using hormonal contraception and IUDs may be at increased risk of HIV infection and underscore the importance of condom use in addition to any more effective contraceptive method for women at risk of HIV infection.
In summary, there is discordance between the results of these studies examining the risks of HIV transmission and viral shedding in relation to hormonal contraception. The different results and associations are likely due to the different study populations, for example prostitutes versus women in stable monogamous unions, and the many covariant risk factors and behaviors that may or may not be similarly or readily identified on cross-sectional or observational studies. Current data do not support avoidance of effective contraceptive methods. Instead, efforts should be directed toward encouraging consistent condom use to decrease HIV transmission and venereal infections, frequent gynecologic and cervical cytologic examinations with prompt treatment of infections, and effective contraception in the gynecologic care of HIV-infected women.
Oral contraceptives, both combination and progestin-only formulations, are felt to be acceptable methods of birth control in HIV-positive women. Studies have shown that when HIV-positive women are educated about hormonal contraception, not only does their use of hormonal contraception increase, but also substantially more of them continue hormonal contraception after 1 year and have lower pregnancy rates.155 Although many studies, as mentioned, have addressed exposure to oral contraceptives, no studies have addressed the use of oral contraceptives in HIV-positive women or the effect of oral contraceptives on the course of HIV infection.
Long-Acting Progestins: Injectables and Implants
Long-acting progestin-only contraceptives are acceptable and effective methods of contraception because they are less user-dependent and of assured compliance. Again, no studies have addressed their effect on disease progression. A 1-year study following 41 HIV-positive women who received levonorgestrel implants after delivery or abortion found excellent acceptance, with no pregnancies or significant changes in weight, hemoglobin levels, or blood pressure from baseline after 1 year of use.156 Disrupted menstrual patterns were the primary side effect, with irregular bleeding and amenorrhea occurring in 56% and 37% of cases, respectively. Levonorgestrel appears to be well tolerated in asymptomatic HIV-positive women. No studies have specifically addressed DMPA use. DMPA use is associated with a high degree of amenorrhea after 1 year of use, which is a desirable side effect becauses decreased menses may decrease infectivity and prevent anemia.
The use of the IUD is contraindicated in women with an active sexually transmitted disease,157,158 and generally has not been recommended in HIV-positive women. Experts have also cited the theoretical increase risk of transmission secondary to the increased duration and flow of menses with the copper-medicated IUDs.159 A theoretical increased risk of PID in an immunocompromised individual represents another concern. A study in Africa examined copper IUD use in 156 HIV-positive women meeting standard criteria for IUD use.160 Compared with noninfected similar women with IUD insertions after 4 months of follow-up, the authors found a similar rate of complications (7.6% versus 7.9%), PID (1.4% versus, 0.2%), and infection-related symptoms (6.9% versus 5.7%). The authors also did not find an increase in complications in more immunocompromised women with lower CD4 counts. Multivariate analysis suggested no association of HIV status with IUD complications. Although these findings are preliminary, the authors suggest that the IUD may be safe for appropriately selected women with HIV infection.
In summary, most women with medical disease have effective contraceptive options that do not accelerate their disease process or affect medical therapy. These women do require a careful review of their medical history and a complete assessment of their current disease status. They also require closer and more frequent follow-up, often with special programs to monitor their metabolic and disease parameters for possible contraceptive or medical complications. In consultation with the patient, the physician should develop an individualized reproductive health plan that addresses her specific disease process and prognosis, her risks from pregnancy, possible teratogenicity of her medical therapy, and the optimal time to plan for a future pregnancy. Patient participation and education regarding both her disease and contraceptive method are crucial to achieve maximal compliance and health.
Complications that arise in women who have medical diseases may not be due to the contraceptive method, but possibly to the many aspects of the disease process. Effective contraception, although not necessarily risk-free, represents in the majority of cases a much safer option than pregnancy.
1. Spellacy WN, Buhi WC, Birk SA: The effect of estrogens on carbohydrate metabolism: glucose, insulin and growth hormone studies on one hundred seventy-one women ingesting Premarin, mestranol and ethinyl estradiol for six months. Am J Obstet Gynecol 114:388, 1971
5. Liew DFM, Ng CSA, Yong YM, et al: Long-term effects of Depo-Provera on carbohydrate and lipid metabolism. Contraception 31:51, 1985
11. Fajumi JO: Alterations in blood lipids and side effects induced by Depo=Provera in Nigerian women. Contraception 27:161, 1983
13. Wilson ES, Cruickshank J, McMaster M, et al: A prospective controlled study of the effect on blood pressure of contraceptive preparations containing different types of dosages and progestogen. Br J Obstet Gynaecol 91:1254, 1984
22. Fuhrmann K, Reiher H, Seemler K, et al: The effect of intensified conventional insulin therapy before and during pregnancy on malformation rate in offspring of diabetic mothers. Exp Clin Endocrinol 83:173, 1984
24. Skouby SO, Jensen BM, Kuhl C, et al: Hormonal contraception in diabetic women: Acceptability and influence on diabetes control of a nonalkylated estrogen/progestogen compound. Contraception 32:23, 1985
26. Radberg T, Gustafson A, Skryten A, et al: Oral contraception in diabetic women. Diabetes control, serum and high-density lipoprotein lipids during low-dose progestogen, combined oestrogen/progestogen and non-hormonal contraception Acta Endocrinol 98:246, 1981
32. Loke DFM, Ng CSA, Samsioe G, et al: A comparative study of the effects of a monophasic and a triphasic oral contraceptive containing ethinyl estradiol and levonorgestrel on lipid and lipoprotein metabolism. Contraception 42:535, 1990
34. Runnebaum B, Grunwald K, Rabe T: The efficacy and tolerability of norgestimate/ethinyl estradiol (250 μg of norgestimate/35 μg of ethinyl estradiol): Results of a open, multicenter study of 59,701 women. Am J Obstet Gynecol 166:1963, 1992
40. WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception: Acute myocardial infarction and combined oral contraceptives: Results of an international multicenter case-control study. Lancet 349:1202, 1997
54. Metzger BE, Bybee DE, Freinkel N, et al: Gestational diabetes mellitus: Correlations between the phenotypic and genotypic characteristics of the mother and abnormal glucose tolerance during the first year postpartum. Diabetes 40(Suppl 2):111, 1985
56. Pettitt DJ, Knowler WC, Baird HR, et al: Gestational diabetes: Infant and maternal complications of pregnancy in relation tothird-trimester glucose tolerance in Pima Indians. Diabetes Care 3:458, 1979
57. Metzger BE, Coustan DM, and the Organizing Committee: Summary and recommendations of the Fourth International Workshop-Conference on Gestational Diabetes Mellitus. Diabetes Care 21(Suppl 2):B161, 1998
62. Skouby SO, Anderson O, Saurbrey N, et al: Oral contraception and insulin sensitivity: In vivo assessment in normal women and women with previous gestational diabetes. J Clin Endocrinol Metab 64:519, 1987
63. Kjos SL, Shoupe D, Douyan S, et al: Effect of low-dose oral contraceptives on carbohydrate and lipid metabolism in women with recent gestational diabetes: Results of a controlled, randomized, prospective study. Am J Obstet Gynecol 163:1822, 1990
71. Lidegaard O: Oral contraceptives, pregnancy and the risk of cerebral thromboembolism: The influence of diabetes, hypertension, migraine and previous thrombotic disease. Br J Obstet Gynaecol 102:153, 1995
74. WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception: Ischaemic stroke and combined oral contraceptives: Results of an international, multicentre, case-control study. Lancet 348:498, 1996
76. Heimemann LA, Lewis MA, Thorogood M, et al: Case-control study of oral contraceptives and risk of thromboembolic stroke: Results from International Study on Oral Contraceptives and Health of Young Women. BMJ 315:1502, 1997
79. WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception:Acute myocardial infarction and combined oral contraceptives: Results of an international multicentre case-control study. Lancet 349:1202, 1997
81. Basdevant A, Conard J, Pelissier C, et al: Hemostatic and metabolic effects of lowering the ethinyl-estradiol dose from 30 mcg to 20 mcg in oral contraceptives containing desogestrel. Contraception 48:193, 1993
91. Lipid Research Clinics Program: The Lipid Research Clinics Coronary Primary Prevention Trial results. II. The relationship of reduction in the incidence of coronary heart disease to cholesterol lowering JAMA 251:365, 1984
94. World Health Organization Expanded Programme of Research, Development and Research Training in Human Reproduction Task Force on Long-Acting Systemic Agents for the Regulation of Fertility: Multinational comparative clinical evaluation of two long-acting injectable contraceptive steroids: Norethisterone enanthate and medroxyprogesterone acetate. Final report Contraception 18:1, 1983
97. Fahmy K, Khairy M, Allam G, et al: Effect of depo-medroxyprogesterone acetate on coagulation factors and serum lipids in Egyptian women. Contraception 44:431, 1991
112. Back DJ, Bates M, Bowden A, et al: The interaction of phenobarbital and other anticonvulsants with oral contraceptive therapy. Contraception 22:495, 1980
117. Hall SM: Treatment of menstrual epilepsy with a progesterone-only oral contraceptive. Epilepsia 18:235, 1977
120. Backstrom T: Epileptic seizures in women related to plasma estrogen and progesterone during the menstrual cycle. Acta Neurol Scand 54:321–, 1976
122. Mattson RH, Cramer JC, Caldwell BV, et al: Treatment of seizures with medroxyprogesterone acetate: Preliminary report. Neurology 34:1255, 1984
124. Odlind V, Olsson SE: Enhanced metabolism of levonorgestrel during phenytoin treatment in a woman with Norplant implants. Contraception 33:257, 1986
141. deAbood M, deCastillo Z, Guerrero F, et al: Effect of Depo-Provera® or Microgynon® on the painful crises of sickle cell anemia patients. Contraception 56:313, 1997
147. Nicolosi A, Correa Leite ML, Musicco M, et al: The efficiency of male-to-female transmission of the human immunodeficiency virus: A study of 730 stable couples. Italian Study Group on HIV Heterosexual Transmission Epidemiology 6:565, 1994
151. Mostad SB, Overbaugh J, DeVange DM, et al: Hormonal contraception, vitamin A deficiency, and other risk factors for shedding of HIV-1 infected cells from the cervix and vagina. Lancet 350:922, 1998
152. Lazarin A, Saracco A, Musicco M, et al: Man-to-woman sexual transmission of the human immunodeficiency virus. Risk factors related to sexual behavior, man’s infectiousness, and woman’s susceptibility. Italian Study Group on HIV Heterosexual Transmission Arch Intern Med 151:2411, 1991
154. Martin HL, Nyange PM, Richardson BA, et al: Hormonal contraception, sexually transmitted diseases, and risk of heterosexual transmission of human immunodeficiency virus type 1. J Infect Dis 178:1053, 1998