Contraception and pregnancy in cystic fibrosis

The increasing adult population of those affected by cystic fibrosis (CF) has inevitably increased the number of women with the condition capable of falling pregnant. Indeed, as in the non‐CF population, it should be assumed that all women with the disease should be considered fertile unless proven otherwise. Similarly although 95% of men with CF are infertile due to congenital absence of the vas deferens, active spermatogenesis does occur and it is essential that all men are counselled about their potential fertility before they consider themselves infertile.

Contraception

Although the first pregnancy in CF was reported in 1960, for many years it had generally been accepted that fertility in the female population with CF was about 20% less than in a comparable group of healthy women.^1–3 However the data to support this are far from convincing. There is evidence that menarche in the CF population is related to general wellbeing and that those with more severe disease, often resulting in a low BMI, have a later menarche and are therefore less likely to be ovulating.⁴ In another study, the average age of menarche was 14.4 years, as compared to 12.9 years for American women.⁵ There have also been suggestions that the finding of CFTR in the hypothalamus and the cervix may have an effect on gonadotrophin‐releasing hormone function and cervical mucus consistency.^6,7 These data and their possible effect on fertility do not provide any predictive information about an individual's fertility and it is important to counsel all young women with CF regarding the issues of contraception. The importance of understanding the implications of pregnancy in view of its possible detrimental affect on maternal health means that the responsibility for discussing contraception should not fall only on the parents, but should be a shared responsibility with the CF team.

In a study from Aberdeen, 59% of women with CF were found to be using contraception, but this compared unfavourably with the national rate of 65%.⁸ In a more recent study from the US, 70% of CF women were using contraception, an improvement on the national US rate in the non‐CF population of 62%. The aim of CF centres must be to ensure that all women of fertile age are aware of the importance of using contraception and the options available considering the needs and beliefs of the individual. While it is entirely understandable that informed women may choose not to use contraception, failure to have educated CF women regarding its importance and the choices available must be seen as a failure of care.

The choice of contraception is potentially difficult and needs to be individualized. If one type proves for whatever reason to be unsuitable another option should be considered. Barrier methods remain a very effective option. In addition to their contraceptive effect they also help to reduce the risks of sexually transmitted disease including cervical disease. The role of some types of human papilloma virus (HPV) infection in the development of cervical intra‐epithelial disease, cancer and genital warts is now well established. Barrier methods have been associated with a reduced incidence of these conditions. The current recommendation to use HPV vaccines in all adolescents will also reduce the incidence of cervical disease in the years ahead. In circumstances when unprotected intercourse has occurred it is important to remind couples of the availability of emergency postcoital contraception.

Oral contraception remains the commonest form of contraception used in women with CF and in one centre 26% of those questioned used the combined preparation.⁹ It has some theoretical disadvantages including concerns regarding the possible worsening of pre‐existing diabetes, malabsorption and liver dysfunction. The progesterone component may act as a respiratory stimulant and adversely effect mucus production and viscosity. In addition the use of broad spectrum antibiotics may adversely affect absorption and efficacy. Despite these potential disadvantages there is now strong evidence supporting the efficacy of the combined contraceptive preparations.

In a study of 10 CF women with moderate to severe lung disease, use of a combined oral contraceptive preparation (OCP) was found to cause no significant deterioration in their clinical status or pulmonary lung function.¹⁰ The pharmacokinetics of the commonly used contraceptive steroids ethinylestradiol and levonorgestrel was investigated in six CF women. Although there was an increased total body clearance of ethinylestradiol, this was compensated for by increased oral bioavailability of the oestrogen. Reassuringly the pharmacokinetics of levonorgestrel did not significantly alter. These studies suggest that women with CF using oral contraception are likely to receive a similar contraceptive effect as do healthy women.¹¹ Careful monitoring remains important however in view of the possible changes in clinical status and treatment regimens, such as the use of new antibiotics. The interaction of antibiotics and the OCP is often a concern in CF. Recent pharmacokinetic data has demonstrated that the plasma levels of ampcillin, ciprofloxacin, clarithromycin, doxycline, methotrexate, ofloxacin, roxithromycin, temafloxacin and tetracycline are all unchanged following OCP use.⁹ This is very reassuring and allows women to have a considerable degree of confidence in using the OCP. Liver disease remains a firm contraindication to the use of the combined OCP.

The intrauterine contraceptive device is also a common form of contraception and is highly acceptable for many women. The traditional copper devices may be associated with heavier menstrual loss for the first 2–3 cycles and if this persists the device may be unsuitable for women with borderline iron reserve. There is a small risk of infection, approximately 2%, but this can be reduced by the use of prophylactic antibiotics at the time of the device insertion.¹² More recent advances have seen the introduction of the progestogen‐only intrauterine system which releases levonorgestrel directly into the uterine cavity. It is the treatment of choice for many women with heavy menses and offers excellent contraception. Since the progestogen is released close to the site of action, progestogenic side‐effects and interactions are minimal. There is some evidence that the risk of pelvic inflammatory disease is also reduced. Patients should be informed however to expect some irregular bleeding for up to 6 months following insertion. The system may be a little more difficult to insert in the non‐parous population, but in skilled hands and with the use of local anaesthesia it can usually be achieved.

Physiology of pregnancy

Normal pregnancy is associated with significant respiratory, cardiovascular and metabolic changes. Although healthy women usually adjust easily to these changes, in CF they may impose a considerable strain for women whose physiology is already severely compromised.

In normal pregnancy the significant increase in resting lung volume may represent a haemostatic response to the increased oxygen consumption and carbon dioxide burden. Increased progesterone levels may sensitize the medulla oblongata to carbon dioxide and this may partly explain the respiratory stimulant effect that occurs. This appears to exceed that which is required to meet metabolic demands and provide adequate gas exchange. In the normal pregnant woman there is increased diaphragmatic function and inspiratory lung capacity. Overall the vital capacity and total lung capacity change little due to a reduction in the functional residual capacity (FRC) of 10–25%.¹³ It is possible that the reduced FRC results in closure of small airways at the lung bases during normal breathing, contributing to impaired gas exchange.¹⁴ These changes in pulmonary function are usually of little importance in the normal pregnant woman. However in a patient with CF, with borderline pulmonary function, they may significantly reduce the efficiency of gas exchange, inducing a degree of hypoxemia and contribute to pulmonary decompensation. Chronic hypoxia is associated with fetal growth restriction and preterm delivery.

Significant cardiovascular changes occur in normal pregnancy. The first response is a fall in vascular resistance. In a singleton pregnancy the plasma volume at term is approximately 50% greater than that seen in a non‐pregnant individual. The cardiac output begins to rise from as early as 10 weeks gestation, reaching a plateau at the end of the second trimester, 30–50% above the non‐pregnant value.^15,16 This rise is a consequence of both an increased stroke volume and heart rate, the latter increasing from as early as 5 weeks gestation to 10–15 beats per minute over the baseline pulse. Once again these changes are usually well tolerated in the normal individual but in women with CF, particularly those who have developed pulmonary hypertension, the increased cardiovascular demands may cause right ventricular decompensation and cardiovascular collapse. This phenomenon is more likely to be seen immediately following delivery due to the process of auto‐transfusion which occurs following the delivery of the placenta. Pulmonary hypertension associated with pregnancy is a grave condition carrying with it a maternal mortality as high as 50%.

The normal weight gain in pregnancy is approximately 12–16 kg. The total energy costs of pregnancy have been calculated to be between 80,000 and 124,000 kcal. The three major components of energy expenditure are growth of the fetus and reproductive tissues, new maternal fat stores and increased maternal metabolism. This represents a rise of approximately 50–100 kcal/day in early pregnancy rising to 200–300 kcal/day from 32 weeks gestation,¹⁷ even in a healthy population though it is recognized that there may be enormous individual variation. Approximately one‐quarter of adults with CF have a BMI < 19 kg/m².^18,19 Low BMI in pregnancy is associated with fetal growth restriction and preterm delivery as poorly‐nourished women try to maintain fetal growth at the expense of laying down body fat and by reducing their metabolic rate. The additional calorific requirements in women with CF may be very difficult to achieve and supplemental feeding may be required.^20–22 In a survey of North American CF centres maternal weight gain in pregnancy was less than 4.5 kg in 41% of the patients.²³ In a study of 26 term and 22 preterm live deliveries in women with CF the mean maternal weight gain of those pregnancies reaching term was 8.9 kg and only 2.6 kg in those delivering prematurely.²⁴

Careful consideration should be given to appropriate vitamin supplementation. Folic acid should ideally be used for 3 months preconceptually and for the first trimester of pregnancy to reduce the risk of fetal neural tube defects.²⁵ Vitamin D supplementation is also recommended, but vitamin A supplements should only be administered at doses <10,000 IU/day as excess use has been associated with fetal anomalies.²⁶

In normal pregnancy a degree of impaired glucose tolerance occurs due to changes in oestrogens, progesterone and human placental lactogen. The renal threshold for glucose is reduced and an increased glomerular filtration rate in pregnancy results in glycosuria in 5–50% of pregnant women. Glucose tolerance is already impaired in many women with CF. In one study 24% of CF patients aged 20 years had CF‐related diabetes.²⁷ Therefore the development of gestational diabetes will be more common, particularly in the second half of pregnancy when women become increasingly resistant to insulin. In a study of 92 pregnancies in women with CF the prevalence of gestational diabetes was 14%, considerably higher than the normal population.²⁸ Screening for diabetes in CF pregnancies should be vigilant. It has been suggested that an oral glucose tolerance test should be performed at 20 weeks gestation in the CF population and again at 28 weeks gestation if the initial investigation was normal.²⁹

The consequences of impaired glucose tolerance on pregnancy are debated, however insulin‐dependent diabetes is certainly associated with a worse pregnancy outcome for both mother and baby. Prepregnancy diabetes is associated with an increased risk of miscarriage, fetal abnormality (particularly cardiac), fetal macrosomia, growth restriction, preterm delivery, polyhydramnios, respiratory distress, intrauterine fetal death and fetal trauma at delivery.³⁰ Postnatally the risks of jaundice, polycythaemia, hypocalcaemia and hypoglycaemia are all increased. Mothers are at greater risk of pre‐eclampsia, infection, maternal trauma at delivery and Caesarean section. Optimization of diabetic control preconceptually and knowledge of the implications of poor control are therefore extremely important.

Prepregnancy counselling

The responsibilities that pregnancy brings should never be undertaken without careful thought. The tragically high rate of abortion in the UK in particular and in many Western societies suggests that all too often though this is not the reality. While many women with CF are now able to fall pregnant, the outcome for mother and baby is not universally successful. The unfavourable physiological changes in pregnancy and the potential significant consequence of pregnancy require careful consideration. Without doubt this is best undertaken prior to pregnancy. The list of confounding factors is long and each one needs to be addressed. Some will be of little relevance for any given individual, while others such as insulin‐dependent diabetes will potentially significantly affect the outcome of the pregnancy ( Table 1).

Table 1.

Issues to consider in prepregnancy counselling

Issue	Discussion
Partner CF carrier status	Prenatal diagnosis
Prepregnancy lung function	Prognosis
Prepregnancy diabetes	Optimize control
Prepregnancy nutrition	Folic acid supplementation
Current medication	Review of therapy
Increased risk of premature delivery	Neonatal consequences
Increased risk of fetal growth restriction	Ultrasound assessment
Possible deterioration lung function	Long‐term prognosis
Increased hospital visits
Risk of hospital admission
Increased risk of Caesarean section
Breastfeeding

Genetic counselling should be offered and the genotyping of the partner undertaken. If the partner is identified as a CF mutation carrier the couple may have to consider the difficult decision as to whether to have invasive prenatal diagnosis which carries a small risk of miscarriage. If the partner is not a carrier it is important that the couple should understand that this does not totally exclude the possibility that the baby will have CF as there are many rarer mutations that will not have been screened, although the risk of having an affected baby will be very small.

Drug therapy should be carefully reviewed. The likely changes in physiotherapy and nutritional requirements will place an increased treatment burden on the pregnant mother and these additional commitments may be more than an individual is able to cope with. Each pregnancy will provide a different challenge and for some women with poor prepregnancy lung function, the issue of their own mortality needs to be frankly discussed. They may need to consider the implications of their death for their family, partner and of course the baby. In those women with severe disease in whom preterm delivery is more common, the women also need to have an understanding of the possible implications, including long‐term handicap, of premature babies. It is essential that the woman's partner is also involved at this stage. In a worse‐case scenario he may potentially be faced with the prospect of his partner's death and the responsibility of caring for a handicapped child. While clearly there is no right or wrong answer under these circumstances, honest prenatal counselling will help focus the thoughts of couples considering pregnancy. It will make them aware not only of the risks but also the huge commitment in terms of time, monitoring and hospital visits that a pregnancy is likely to necessitate. Whatever decisions prospective parents make, it is important that they are reassured that they will be supported by their CF and obstetric team.

Inevitably situations will arise when pregnancy occurs unexpectedly. Although this is far from ideal, urgent referral for counselling and assessment is required. The option of terminating the pregnancy either for social or medical reasons should be explained, and if it is the mothers wish not to continue the pregnancy, the future importance of reliable contraception stressed. In early pregnancy medical termination has been shown to be a safe option and potentially avoids the need of general anaesthesia in the very sick woman.³¹

Pregnancy and cystic fibrosis

The first reported case of CF in pregnancy in 1960 provides a salutatory reminder of the difficulties involved in such pregnancies. A positive sweat test confirmed the diagnosis of CF in a 20-year‐old during pregnancy. She remained well until the 34th week of gestation when following worsening respiratory symptoms she went into preterm labour. The patient died six weeks postpartum.¹ This report together with other early cases were reviewed and although the data made scanty reference to objective respiratory function measurements, two distinct populations appeared to emerge. In five out of 10 women the pregnancy did not appear to adversely affect the maternal condition. None of the mothers in this group were diabetic. Four pregnancies went to term and none of the women suffered any significant decline in their lung function. However the other five women, two of whom suffered from diabetes, all delivered prematurely and all suffered a loss of lung function postnatally. One baby was stillborn and one died in the neonatal period in this group. Furthermore two women died from cor pulmonale in the immediate postpartum period and two others died within 18 months of delivery. The review concluded that although pregnancy was possible in women with CF, there appeared to be an increased risk of fetal prematurity, the possibility of a progressive loss of lung function and even loss of life in those with a more severe pre‐existing disease.³² Other small case series have also reported an increased incidence of preterm delivery and maternal death in women with worse pre‐existing lung function.^33,34

In view of the potentially devastating consequences of an unwanted pregnancy for those with severe CF some authors have advocated termination of pregnancy in women with poor lung function (FVC less than 50% predicted value) or progressive pulmonary deterioration.³⁵ However successful pregnancies have been reported even in those with poor lung function. It is therefore difficult for clinicians to recommend termination of pregnancy in these circumstances, although most obstetricians would certainly support it if that is the mother's considered wish after she has been extensively counselled. It has been suggested that stable prepregnancy lung function may be the key.³⁶ These very difficult issues reinforce the crucial importance of prepregnancy counselling by clinicians directly involved in the management of such cases.

There are a number of reviews examining retrospective data of the effects of pregnancy on CF. One of the largest early series of data relating to pregnancy outcome in women with CF was collected by the US Cystic Fibrosis Foundation Registry. In 1990 of 111 pregnancies reported, 47.7% of the pregnancies were completed, 31.6% of women underwent a therapeutic termination and at the time of reporting 30.6% of the pregnancies were still progressing.³ Approximately one‐quarter of the completed pregnancies delivered prematurely. Diabetes mellitus was reported in 4.5% of the pregnant group. In a subsequent study the outcomes of 217 pregnancies in 162 women were described.³⁷ The rate of miscarriage was 4.6%. Termination of pregnancy was undertaken in 13.8% of cases, although it is possible that this figure may have been under-reported. In over 80% of women, the pregnancy progressed past 20 weeks gestation. A total of 24.3% delivered preterm, mainly as a consequence of spontaneous preterm labour rather than induction of labour. It has been suggested that this may be due to the effects of chronic hypoxia in the fetus or poor maternal nutrition and pancreatic insufficiency. A total of 7.9% of mothers died within 6 months of delivery and 13.6% within 2 years. This study confirmed the association between poor prepregnancy lung function and maternal death. The fetal perinatal death rate was 7.9%, the commonest cause being extreme prematurity.

Edenborough and colleagues reported the outcome of 72 pregnancies in 55 women identified in seven UK CF centres over a 20-year period.²⁴ Fourteen (20%) pregnancies were terminated. The data showed that the prepregnancy %FEV₁ and %FVC predicted were significantly associated with gestational age, maternal weight gain and infant birth-weight. The authors concluded that when the FEV₁ is >70% predicted pregnancy is safe. However in severe disease (FEV₁ <50%), or in women with pulmonary hypertension or cor pulmonale, they advised against pregnancy in view of the poor prognosis. The median survival of the women who delivered prematurely was 7.6 years.

A number of studies have attempted to address the issue of the possible long‐term effects of pregnancy on CF. Although the number of pregnancies in women with CF is growing, the total number of cases remains small. Many of the studies have reported series of patients managed in multiple centres and over a prolonged period (20–30 years). Despite these drawbacks the conclusions and comments of the different studies are very similar. Frangolias et al. matched seven pregnant women with CF to a non‐pregnant CF group.³⁸ The rate of decline of %FEV₁ during the pregnancy period was similar in both groups however the mean prepregnancy FEV₁ was 75.3%, suggesting a relatively healthy population. In spite of this the pregnancy group was unable to regain the weight lost in the postpartum period even after 2 years. In addition two of the seven pregnant women did suffer a significant decline in lung function. Gilljam et al. analysed the course of 92 pregnancies over a 37-year period.²⁸ Although this study has the benefit of long‐term follow‐up (mean follow‐up was 11 years), it is likely that both the obstetric and cystic fibrosis management of the mothers would have varied considerably over the years. The mean prepregnancy FEV₁ was 68% predicted. The study concluded that pregnancy did not appear to affect the natural decline in lung function, but that women with FEV₁ <50% had a poorer outcome with a 50% 10-year postpregnancy survival as compared with the 89% survival in the group whose FEV₁ was >50% predicted.

Data from the French CF Registry were reported in 2002.³⁹ In contrast to other studies the pregnant CF population was matched on the same genotype and cystic fibrosis care centres network as compared to clinical parameters. Ninety pregnancies were reported over a 20-year period. The mean prepregnancy FEV₁ was 61%. All three women who died in the year following delivery had an FEV₁ <50% before pregnancy. The authors concluded that pregnancy probably did have an adverse effect on the health of the mothers, but this effect was only slight if the women were in good general health before pregnancy. In addition pregnant women lost weight after the pregnancy while the controls gained weight.

In contrast in a large matched‐parallel cohort study, using data from the US Cystic Fibrosis Foundation National Patent Registry, of 680 pregnant women with CF, matched to over 3000 control women with CF, it was concluded that pregnancy was not harmful in any group of CF patients.⁴⁰ The study concluded that women with CF who became pregnant tended to be healthier and had an improved 10-year survival rate than those who did not become pregnant. In further analysis the study did not confirm the findings of previous groups that diabetes, B cepacia infection or prepregnancy FEV₁ <40% were associated with poorer outcomes. Overall 20% of mothers with CF died within 10 years of the pregnancy and this increased to 40% if the FEV₁ <40% predicted. More recently Madge et al. reported a significant deterioration in maternal lung function in pregnant women with CF wellbeing in the first year postnatally.⁴¹

One further significant risk factor that relates to maternal health in pregnancy is infection with Burkholderia cepacia.³⁸ This organism is associated with reduced long‐term survival, but does not always cause a decline in pulmonary function. Nonetheless in one report of four cases of CF pregnancy where the prepregnancy FEV₁<50% predicted, three women died and all were infected with B cepacia.⁴² In a larger group of women the 10-year postdelivery survival rate was 95% in B cepacia negative women but only 67% in those positive for the organism.²⁸ These findings though are in contrast to those of Goss et al. who reported no increased risk to the pregnancies with B cepacia infection.⁴⁰

Obstetric management

The role of the obstetrician is to become a part of the multidisciplinary CF team providing individualized care for the pregnant woman. In addition an experienced obstetric anaesthetist is essential. The anaesthetic team should have the chance to assess the mother during the antenatal period. If the local team is not confident to provide an integrated service, referral to specialized centres should be considered. Edenborough and colleagues have produced a thorough set of guidelines for the management of pregnancy in women with CF.²⁹ Co‐ordinating these services and ensuring good interprofessional communication is crucial as individual experience of CF and pregnancy will remain small in absolute terms. Once pregnancy is diagnosed consideration is given to ensuring all aspects of maternity care are catered for, including normal maternal screening for maternal blood group, possible congenital infections such as rubella, hepatitis, syphilis and HIV. Screening for toxoplasmosis is routine in Europe and the USA, but not in the UK. Screening for chromosomal anomalies should be discussed and offered, and prenatal diagnosis considered if appropriate. This is undertaken at 12–13 weeks gestation using a combination of nuchal translucency scanning and maternal biochemistry. Serial ultrasound scans and fetal Doppler assessment provides useful information about the maternal and fetal wellbeing and should be routinely performed in view of the increased risk of growth restriction. Evidence of fetal compromise may result in a recommendation for earlier delivery. If delivery is anticipated before 34 weeks gestation, the administration of maternal steroids, betamethasone, should be recommended to aid fetal lung maturation.⁴³ Care should be taken to ensure adequate glucose monitoring in pancreatic insufficient women and an insulin pump or sliding scale dose regimen may be necessary at this time.

Careful liaison throughout the pregnancy with the lead chest physician permits appropriate outpatient or inpatient management. Evidence of infective exacerbations should be aggressively treated with intravenous antibiotics. Changes of physiotherapy regimes are common, particularly in women with poor pre‐existing lung function.⁴⁴ Women should be encouraged to attend antenatal classes.

The timing of delivery will depend on the maternal and fetal wellbeing as the pregnancy advances. The most common fetal complication is preterm delivery and in the majority of cases this is iatrogenic. Deteriorating maternal lung function will frequently lead to a decision to undertake earlier delivery and approximately 25% of deliveries are preterm. In pregnancies that reach term (>37 weeks), many are induced electively if there is evidence of maternal compromise. Having made a decision regarding the timing of delivery the mode of delivery must be determined. It is reasonable in most cases, in the absence of obstetric indications, to aim for a vaginal delivery. Epidural anaesthesia allows for a relatively stress‐free labour and consideration should be given to limiting the length of the second stage, depending on the clinical status of the mother.⁴⁵ Elective assistance in the second stage is by forceps or ventouse and may reduce the risk of overburdening the respiratory system, although if the mother is well and coping there is every chance she will deliver the baby spontaneously. Caesarean section may be indicated for obstetric considerations such as breech presentation. In women with particularly poor lung function, although the ideal remains vaginal delivery, often the mothers are clearly unable to cope with the demands of a labour and operative delivery may be required. The skill in management is in trying to judge the optimal time for delivery in the mother with poor lung function, before there is a rapid decline in reserve. It is also important to involve the women in these decisions as she will often have a preference for the mode of delivery and her views should be respected.

Care should be taken, particularly if an assisted delivery is undertaken, to minimize the risk of blood loss. Postnatal anaemia will contribute to a slow recovery. If intravenous fluids are administered at any stage, normal saline should be avoided. A careful study of breast milk concluded that CF breast milk does not contain elevated concentrations of sodium and is safe for infant feeds.⁴⁶ Breastfeeding is therefore a reasonable option and this may be particularly important for preterm babies as expressed breast milk is undoubtedly their best form of nutrition. However, breastfeeding is potentially an exhausting process. There is a risk that new mothers will sacrifice the time required to maintain their own health to provide for their baby and in so doing compromise their long‐term prognosis. In those women who are struggling to cope in the postnatal period, it may be sensible to suggest a change to bottle feeding and suppress lactation.

Summary

The continuing advances in the management of CF have meant that parenthood is a realistic aim for many women and increasingly for affected men. Although the outlook for pregnancy in the CF population is generally good, it seems likely that those with poor prepregnancy lung function are at greater risk of complications and long‐term sequelae and must be informed of the unpredictable but possible irreversible deterioration in their own health. Therefore the issues regarding fertility and contraception should be addressed in the teenage years, and a responsible attitude to contraception generated. Prepregnancy counselling is a crucial component of overall obstetric care. Perhaps a standard ‘discussion’ or forum should be arranged at 16–18 years of age to ensure that all CF patients have the chance to be appropriately informed. During pregnancy a multidisciplinary approach is crucial and meticulous antenatal care are crucial in order to ensure the best possible outcome for both mother and baby.

Footnotes

DECLARATIONS —

Competing interests None declared

Funding None

Ethical approval Not applicable

Guarantor JGT-B

Contributorship JGT-B is the sole contributor