Patients fulfilling the following criteria were eligible and included: (1) radiological and/or histological diagnosis of CCM, (2) available clinical baseline characteristics, (3) medication history, and (4) follow-up data with available magnetic resonance imaging and consultation with a board-certified neurosurgeon or neurologist. Radiological and histological diagnoses of CMs were confirmed either by a board-certified neuro- radiologist and/or neuropathologist.

Patient records were reviewed to assess baseline characteristics: date and age at initial diagnosis, sex, date of surgery for CCM, medical history, cardiovascular and non-cardiovascular comorbidities, and date of last follow-up.

Magnetic resonance imaging (MRI) was examined to determine the location of the CCM, which includes supratentorial, brainstem, cerebellum, and spinal cord locations, as well as the presence of multiple CCMs (mCCM). A diagnosis of familial cerebral cavernous malformation (FCCM) was established in patients meeting one or both of the following criteria: (1) having multiple CCMs or one CCM along with at least one other family member with one or more CCMs; and/or possessing a pathogenic variant in KRIT1, CCM2, or PDCD19 (10).

The cardiovascular comorbidities were differentiated in (1) hypertension, (2) coronary heart disease (CHD) / myocardial ischemia (MI), (3) transient ischemic attack (TIA) / stroke, (4) atrial fibrillation (AF) and other cardiac arrythmias, and (5) artificial heart valves.

Medical history from initial medical records was gathered and treatment and outcomes were tracked through yearly prospective monitoring of hospital records and any available primary care practitioner records.

Retrospectively data for female hormone therapy use was collected from these data sources. In this study female hormone therapy is used as an overarching term for the intake of either birth control treatment in reproductive age [birth control pill, hormonal intrauterine device (IUD), and other hormonal contraceptives] or hormone replacement therapy in postmenopausal women. As birth control pills we included combined estrogen contraception and progesterone only pills. Female hormone therapy use was determined as the prescription and receipt of it at any point after the starting point, but before the first outcome event or the end of follow-up (if an outcome event did not occur).

The starting point was defined as the time of the initial clinical presentation. In the final analysis, we included every female individual with a first-in- a-lifetime diagnosis of CCM which fulfilled the abovementioned inclusion criteria and categorized the type of the first clinical presentation by the symptoms that led to the initial diagnosis into (1) incidental finding, (2) focal neurological deficits, or (3) seizures (regardless of earlier events that might, in retrospect, have been attributable to a CM). We noted as well whether hemorrhage did or did not occur at the initial diagnosis. All results are adjusted for age, brainstem location, initial presentation with hemorrhage and brainstem location.

The primary outcome was defined by a new hemorrhage (confirmed by acute/subacute hemorrhage on brain imaging consistent with the time of symptom onset) or a new focal neurological deficit attributable to the location of the CM (but without evidence of a new hemorrhage on brain imaging). Focal neurological deficits without evidence of a new hemorrhage on brain imaging were included as primary outcome since they might be due to an undetected hemorrhage on imaging or triggered by thrombosis and, thus, might be affected by medical treatment (6, 9). The type of new focal neurological deficit during follow-up was further categorized into (1) sensory deficit, (2) motoric deficit, (3) seizures, (4) cranial nerve deficit, and (5) headache.

Dates and numbers of outcome events (new ICH or FND) were recorded during the follow-up period. Three researchers, identified as SW, SN, and VitS, evaluated these outcome events based on the clinical, radiological, and pathological data available, while remaining unaware of the patients’ use of female hormone therapy. Patients who had undergone surgery were considered until the day of their CCM removal, and those who had lost follow-up were considered up to their last recorded follow-up day. Additionally, a review of the electronic patient chart was conducted to investigate the causes and instances of death.

All statistical analyses were performed using R 4.3.2 (R Core Team, 2023). Descriptive statistics are presented as absolute numbers (n) and proportions (%) for categorical variables, whereas continuous variables are shown as mean and standard deviation (SD). The primary endpoint was defined by a new hemorrhage or a new focal neurological deficit (ascribable to the CM location). Time-to-event probabilities were plotted using the Kaplan–Meier method (11). For multivariable survival analysis, the Cox proportional hazards regression model was used (12). The strength of association between the number of outcome events (ICH, FND) and patient age, female hormone therapy, the location of the CCM and mCCM was assessed using a multivariable Poisson regression model with an offset for length of follow-up. A two-tailed p value of ≤ 0.05 was considered statistically significant.

Three hundred twenty-eight female patients diagnosed with CCM met the criteria and were included in the final analysis. The median age at first presentation leading to a CCM diagnosis was 45.3 years (SD 18.1) (Table 1). CCMs were located in the supratentorial compartment in 227 patients (69.2%), the brainstem in 85 patients (25.9%), the cerebellum in 49 patients (14.9%), and the spinal cord in 19 patients (5.8%). Additionally, 56 female patients (17.1%) had CCMs in multiple anatomical compartments.

In 125 patients (38.1%), CCM was detected incidentally, while 151 patients (46.0%) presented with focal neurological deficits (FND), and 52 patients (15.9%) presented with seizures. Among the 328 patients, surgical resection of at least one CCM was performed in 126 patients (38.4%). Additional details on the cardiovascular and non-cardiovascular comorbidities of our patient cohort can be found in Table 1.

Among the 328 patients, 56 (17.1%) were receiving female hormone therapy. This included 37 patients (11.3%) on birth control treatments and 19 patients (5.8%) on hormone replacement therapy. Birth control treatments comprised the use of birth control pills in 28 cases (8.5%), hormonal intrauterine devices (IUDs) in 5 cases (1.5%), and other hormonal contraceptives (transdermal estrogen patch and subdermal contraceptive implant) in 3 cases (0.9%) (see Table 2).

Patients on female hormone therapy were younger (37.8 vs. 46.9 years, p < 0.001) and had a longer follow-up time compared to the control group (2025.6 vs. 1400.7 days, p < 0.03). The likelihood of ending follow-up due to surgery was higher in patients on female hormone therapy than in the control group (35.7% vs. 30.9%).

There were no significant differences in the number of CCMs, symptoms at initial diagnosis, or comorbidities between the female hormone therapy group and the control group (Table 1).

A comparable number of ICH and FND events were observed in patients with CCM who were on female hormone therapy compared with those who were not (see Table 3).

A cohort of 328 female patients was monitored from the initial diagnosis of CCM to assess the primary outcomes of ICH or FND associated with CCM. Over the follow-up period of 2025.6 days (FHT) and 1400.7 days (nFHT), there was no significant difference in the number of ICH between patients receiving female hormone therapy (FHT) and those in the control group (nFHT) (mean 0.43; SD 1.11 vs. 0.38; SD 0.8, p = 0.1). Similarly, the number of FND was comparable between the two groups, with no significant difference (FHT vs. nFHT: mean 0.36; SD 0.84 vs. 0.28; SD 0.66, p = 0.58, see Table 3).

The likelihood of experiencing a first ICH or FND was comparable for both groups of CCM patients, regardless of whether they were on female hormone therapy or not.

On average, patients in the female hormone therapy group experienced a hemorrhage after 1631.5 days of follow-up (SD 2324.6), compared to 1161.4 days (SD 1650.8) in the control group (p = 0.35) (Figure 1). Regarding the risk of developing a FND in the female hormone therapy group, the average time until the first FND was 1601.0 days (SD 1934.1), while in the control group, it was 1208.1 days (SD 1649.5) (p = 0,86) (Figure 2).

While there is a trend indicating that ICH and FND tend to occur later in patients receiving FHT compared to the control group, these differences were not statistically significant. The time to the first hemorrhage (p = 0.35) and the time to the first FND (p = 0.86) did not show a significant association with the use of FHT.