Endogenous hormones, particularly estrogen and progesterone, have been hypothesized to influence CM behavior (26–28). Early studies suggested that women are more likely to present with SH during periods of significant hormonal change, such as pregnancy (21, 24), due to factors such as hyperdynamic circulation, increased turbulent blood flow, and higher estrogen levels leading to endothelial cell degeneration. Also, placental production of growth factors such as VEGF, bFGF, and placental growth factors were hypothesized to contribute to proliferation and angiogenesis, potentially exacerbating CM lesional activity (29) Hormonal changes during pregnancy have also been proposed to cause intralesional CM or DVA thrombosis, leading to venous hypertension and subsequent CM hemorrhage (29, 30). More recent studies, including those by Kalani, Witiw, and Joseph, suggest that the risk of hemorrhage during pregnancy may not differ significantly from the non-pregnant state, but have limitations (31–33). These studies are discussed in more detail under the section “Pregnancy and CM Management.” Although the literature has not specifically addressed the effects of endogenous estrogen on CMs, it has been suggested that estrogen plays a crucial role in maintaining blood–brain barrier integrity and supporting vascular repair. Estrogen may enhance the repair of endothelial cells and help restore the tight junctions within the blood–brain barrier (34, 35), which could potentially mitigate the risk of hemorrhage. However, this protective effect might decrease with aging (36, 37), leading to greater permeability of the blood–brain barrier and an increased susceptibility to hemorrhagic stroke.

Zhang et al. suggest that progesterone may also negatively affect the cerebral cavernous malformation (CCM) complex, which stabilizes the endothelium (26, 28, 38). Recent research has further elucidated the role of progesterone receptors in CCMs, indicating that activation of progesterone receptors can disrupt the blood–brain barrier and exacerbate CM pathology by promoting vascular permeability and inflammation (38). Additionally, the absence or dysregulation of progesterone receptor signaling has been associated with structural changes in the blood–brain barrier, further contributing to the progression of CMs (27).

These data limitations underscore the need for more robust, prospective studies to better understand the relationship between endogenous hormonal fluctuations and CM behavior. Moreover, there remains a need for further investigation into how other hormonal changes, such as those during the menstrual cycle and menopause, may influence CM behavior, as the current literature has predominantly focused on pregnancy and hormone use.

Concerns have been raised about the impact of exogenous estrogen and progesterone, such as those found in oral contraceptives and oral hormone replacement therapy (HRT) used for perimenopausal symptoms on CM behavior (39).

A prospective cohort study from the Mayo Clinic found that brainstem location and estrogen use in female patients were associated with an initial presentation of SH, although the study had a small sample size and did not differentiate between hormone delivery methods (40). A larger study by Zuurbier and colleagues analyzed 722 female patients, 137 of whom had taken exogenous estrogen and/or progesterone for contraception or HRT. The study found that 33.6% of patients taking hormones experienced a prospective CM hemorrhage, compared to 15.6% of those not taking hormones. The increased hemorrhage risk was particularly significant for those using oral contraceptives, especially when combined with smoking. In the HRT group, oral hormone replacement, but not other hormone delivery methods were associated with an increased risk of SH. While estrogen-containing contraceptives were associated with an increased risk, the study grouped oral progesterone-only contraceptives together with combined estrogen-progesterone options, rather than evaluating them separately. Additionally, progesterone-only methods delivered via intrauterine devices (IUDs) were not evaluated due to differences in the ascertainment of data between included cohorts. This grouping approach presents a limitation, as it does not allow for a clear assessment of the risks associated with progesterone-only contraceptives.

While the above study raises concerns about exogenous oral estrogen and progesterone in patients with CM, many questions remain. If estrogen increases SH risk by its influence on intralesional thrombosis, then the effect should be dose dependent, dependent on which generation progesterone it is combined with and may be dependent on the mode of delivery (transdermal versus vaginal versus oral). The potential negative effects of progesterone on the CCM complex should still be considered given the work by Zhang and colleagues, but further research is necessary to accurately distinguish the risks of progesterone separated from the estrogen effect (39). A women’s health specialist should be consulted to discuss alternative options such as copper IUDs for pregnancy prevention or consider the lowest risk estrogen/progesterone combination in patients with medical conditions requiring hormones.

For transgender patients undergoing hormone therapy, similar considerations apply. Counseling for transgender patients should address both the need for gender-affirming care and the risks associated with hormone therapy in the context of CM location, symptoms, and treatment options.

In summary, estrogen-containing contraceptives and oral hormone replacement therapies should generally be avoided in women with CM due to the increased risk of thrombosis and CM hemorrhage associated with estrogen-containing therapies. In addition, given the potential effects of progesterone on the CCM complex, these options should be chosen carefully and monitored closely (26, 28). However, given the many uncertainties of the current data and the need for hormonal therapy in certain conditions, a personalized approach should be emphasized in clinical practice. An individual with CM considering hormone therapy should be evaluated to determine their individual risk of SH and the morbidity associated with potential hemorrhage and compare that to the risks, benefits and alternatives to hormone therapy. For example, there are familial CM patients who inherit the gene, but have no lesions or have only lesions that appear on hemosiderin sensitive sequences. The risk in such patients may not be similar to those with a recently symptomatic hemorrhagic event, brainstem lesion, or enlarging CM lesion. Hormone therapy use should be carefully monitored, with individualized care plans to minimize risk while effectively addressing the patient’s symptoms (39).

A neurologist or obstetrician should guide patients on the optimal timing to consider conceiving. Ideally, pregnancy should be planned when the risk of CM hemorrhage is lowest, and if the patient has seizures, they should be well-controlled. Natural history studies indicate that the risk of recurrent hemorrhage is highest within the first year following an initial event and begins to plateau around 2.5 years thereafter (16). Consequently, patients who have recently experienced SH are generally advised to wait at least 1 year before conceiving, if feasible, considering maternal age and other factors. Similarly, since frequent and/or convulsive seizures during pregnancy can negatively affect both the mother and fetus, achieving adequate seizure control prior to conception is ideal.

For female patients with a history of seizures, managing ASMs involves important considerations, especially when planning for pregnancy. First, ASMs may have teratogenic potential, posing risk to the developing fetus. The fetus is most vulnerable to congenital malformations early in development, often before the pregnancy is recognized (41–43). To mitigate these risks, folate supplementation in addition to discussions on whether the ASM should be continued, held or changed to an alternative should take place. It is recommended that women of childbearing age on ASMs take at least 0.4 mg of folic acid daily, as this supplementation may help reduce the risk of neural tube defects associated with ASM use (47). Prospective registry studies indicate that valproic acid has the highest teratogenic potential, while lamotrigine and levetiracetam are associated with the lowest risk. The teratogenic effects of newer ASMs remain less well understood (57–62). Additionally, achieving optimal seizure control before pregnancy is crucial, as poorly controlled seizures can increase the risk of preterm labor and pose significant risks to the fetus. Therefore, baseline ASM levels should be obtained and maintained at effective therapeutic levels to ensure both maternal and fetal safety during pregnancy (44, 63–65).

Patients with FCCM should receive genetic counseling regarding the risk of passing the mutated gene to their offspring. The genes CCM1, CCM2, and CCM3 are inherited in an autosomal dominant manner. However, not all individuals with the gene develop CM lesions, and among those who do, symptoms may vary significantly (66, 67). Data suggests that by age 80, approximately 60.4% of patients with FCCM will experience at least one seizure (68), and up to 64% will experience at least one symptomatic cerebral or spinal hemorrhage within the first 10 years after diagnosis. It is also important to note that as many as 20% of patients may remain asymptomatic for many years (69).

For patients with familial CCM, reproductive decisions can be complex and may require input from genetic counselors, fertility experts, and obstetricians (69). Patients may choose to proceed with natural conception, consider using donor sperm or eggs, explore in vitro fertilization (IVF) with or without embryo selection (70), or consider adoption. While not studied in CM patients, there is an increased risk of venous thromboembolism in patients undergoing IVF (71). Given that CM hemorrhage may be the result of intralesional thrombosis, concern has been raised about the potential for IVF medications to cause CM hemorrhage. Thus, the decision to pursue IVF should be made with careful consideration of the individual’s CM hemorrhage risk and in consultation with a healthcare provider experienced in managing CM. Genetic testing of embryos and chorionic villus sample require prior knowledge of the parental mutation. Chorionic villus sampling (CVS) for genetic testing can be performed between 10 and 12 weeks of pregnancy (72, 73).

Early case reports and series initially raised concerns that pregnancy might increase the risk of CM hemorrhage. However, more recent studies from large tertiary care centers suggest that the risk of CM hemorrhage during pregnancy may be comparable to the non-pregnant state (31–33). For example, Kalani et al. followed 64 female patients with 168 pregnancies and found that four patients experienced five hemorrhages during pregnancy, resulting in an annual hemorrhage rate of 3.4% per year pregnant. The study concluded that this rate was similar to the natural history of CM hemorrhage in non-pregnant patients, although the average age of diagnosis was not specified (31).

In a study from Toronto, Witiw and colleagues assessed 186 female patients with 349 pregnancies. They estimated the risk of CM hemorrhage by assuming the patient was at risk between the ages of 15 and 45 (childbearing years). The rate of hemorrhage during pregnancy was calculated as 1.15% per patient-year, compared to 1.01% per patient-year during the non-pregnant state. However, this methodology assumed that CM lesions were present throughout the reproductive years, despite the unclear timing of cavernoma development, potentially underestimating the risk during pregnancy, particularly given that the average age of diagnosis in the Toronto study was 42.7 years (32).

Joseph and colleagues addressed this potential confounder by assessing the risk of CM hemorrhage in female patients diagnosed with CM prior to age 46. Of the 90 patients, 136 pregnancies occurred before CM diagnosis, with four cases of CM hemorrhage during pregnancy leading to the initial diagnosis. Of 21 patients with 32 pregnancies after the CM diagnosis, no hemorrhages occurred during subsequent pregnancies. The rate of hemorrhage during the non-pregnant state in the same group was 10.4% per patient-year. The authors concluded that the risk of CM hemorrhage was similar in both the pregnant and non-pregnant states. However, they acknowledged the small sample size and the possibility of selection bias, as patients who experienced SH might have deferred future pregnancies, further complicating the interpretation of results (33).

Patients should be counseled that the absolute risk of complications during pregnancy is low, but close monitoring for clinical symptoms is important. Patients should be educated about their individual potential symptoms to be concerned about, which can vary depending on CM location—whether in the spine, brainstem, or cortex. Symptoms of concern, such as persistent, severe headaches or focal neurological deficits, may necessitate further investigation to rule out conditions like venous thrombosis, pre-eclampsia, or CM-related complications. MRI without contrast is considered safe during pregnancy if such concerns arise.

In the non-pregnant state, surgery is generally favored for CM patients with recurrent seizures unresponsive to medication or SH with mass effect in non-eloquent regions, as well as in cases of repeated hemorrhages and increasing disability in eloquent regions. When symptomatic or hemorrhagic CMs present with acute neurological decline, urgent surgical intervention is typically warranted (74). Most experts agree that surgical resection during pregnancy should be reserved for those with rapidly progressive symptoms and that it should be performed by a multidisciplinary team at a specialized center (29, 75–77).

Most patients with well-controlled seizures before pregnancy will remain seizure-free during pregnancy (64). However, 20–30% of patients may experience increased seizure frequency during pregnancy. This is often due to a decline in ASM levels, which can drop significantly during pregnancy, particularly with medications like lamotrigine, levetiracetam, oxcarbazepine, topiramate, and zonisamide. This decline is attributed to changes in maternal renal and hepatic metabolism, volume of distribution, and interference by placental enzymes. Therefore, it is recommended to establish baseline ASM levels prior to pregnancy and to monitor levels throughout pregnancy, adjusting doses as necessary to maintain pre-pregnancy levels. Depending on the ASM, levels may need to be checked every 4 weeks (47), although no standardized testing frequency protocols have been published. Other factors contributing to seizure exacerbation during pregnancy may include stress, sleep deprivation, and infection. If a previously well-controlled seizure disorder worsens without an apparent cause, MRI might be considered to assess for CM hemorrhage.

Pushing during childbirth has raised concerns about the potential for intracerebral hemorrhage in patients with vascular malformations. However, there have been only rare reports of hemorrhage from a CM during delivery (58). Vaginal delivery is generally considered safe and reasonable for stable patients with CM (29). In cases where the patient has a recent neurologic issue, such as a significant focal neurologic deficit or a recent CM hemorrhage, a Cesarean section may be considered.

The postpartum period, typically defined as the first 4–6 weeks after birth, requires ongoing monitoring of neurological symptoms and ASM levels, as hormonal fluctuations can persist for up to 6 months. For women whose ASM dose was increased during pregnancy, serum levels may rise shortly after delivery, necessitating dose reductions to avoid toxicity. In ASMs cleared by glucuronidation, such as lamotrigine, tapering should begin as early as 3 days postpartum, while ASMs with renal clearance may return to pre-pregnancy metabolism within 2–3 weeks postpartum (41, 44).

Breastfeeding offers numerous health benefits, but there are specific considerations for patients with CM. If an MRI is needed postpartum, it should be performed without contrast, or the patient should pump and discard breast milk for 24–48 h after the procedure. ASMs can be present in breast milk, though at lower concentrations than a fetus is exposed to during pregnancy. Studies have not demonstrated adverse effects of maternal ASM use during breastfeeding on the IQ, social or motor skills, or behavior of breastfed children (78, 79). Therefore, women on ASMs can breastfeed but should monitor their babies for signs of sedation, particularly if taking benzodiazepines or barbiturates (41).