The association between exogenous sex hormones and meningiomas has long been a subject of debate, with epidemiological studies showing an absence of link or a weak link between oral contraceptive pills (OC) or hormone replacement therapy (HRT) and the risk of meningiomas. While no studies clearly established an association between contraceptives and meningioma[23–26], several studies suggested a positive correlation between HRT and an increased risk of meningiomas[26–28]. In a large cohort study, Jhawar et al. reported that the risk of meningioma associated with current hormone use was 2.48-fold for premenopausal women when compared with postmenopausal women who had never used HRT. The relative risk for postmenopausal women taking HRT was 1.86[28]. Conversely, two hospital-based case-control studies and one population-based case-control study found no association between HRT and meningioma[24, 29, 30]. A meta-analysis published in 2013 by Qi et al., concluded that HRT in postmenopausal women and an increased number of births were associated with an increased risk of meningioma, while oral contraception, age at menarche/menopause or age at first birth do not contribute significantly to the risk of developing meningiomas [26]. In a more recent meta-analysis, Benson et al. reported,, that there is an increased risk of meningioma among HRT users with a relative risk of 1.35 (95% confidence interval 1.21 to 1.49)[27].
Since 2008, attention has focused more specifically on CA, following a report from Froelich et al. that showed evidence of tumor regression after treatment discontinuation[10]. Since then, multiple case reports and small series were published suggesting a positive association between CA and the risk of meningioma. Only recently, a larger study based on the French social security database published by Weill et al., proved the link between CA and meningioma, with a risk of having a meningioma treated with surgery or radiotherapy in women pursuing high doses of CA multiplied by 6.6. The study also confirmed a strong dose-effect and a noticeable reduction in risk after treatment discontinuation[15].
While the link between CA and meningioma was thoroughly studied, there is far less available data regarding the link between CHA-NA and risk of intracranial meningioma [20, 31–33]. The first suspicion of a link between NA and meningioma was published in 2004 by Cottin et al. who described the case of a patient diagnosed with a large meningioma after being treated with NA for 16 years for pulmonary lymphangioleiomyomatosis [34]. In 2008 Shimizu et al reported the case of an 80 years old male incidentally diagnosed with a meningioma on a CT scan following a head injury [33]. The patient was treated with CHA for a benign prostate hypertrophy. The authors noticed a significant volume reduction on follow-up imaging after his treatment was changed to an alpha-2-blocker. In 2019, Passeri et all described a series of 3 patients presenting with meningiomas and treated with NA, that showed spontaneous tumor regression after treatment discontinuation [20]. Champagne et al. also presented a case of a patient in which CHA and NA showed similar effect on meningiomas: growth under treatment and shrinkage after cessation of treatment[31]. In a recently published studying a consecutive series of patients between 1995 and 2008, Malaize et al. reported an increased number of patients diagnosed with progestin associated meningiomas over time. The rate of conservatively managed meningiomas also increased over time, with tumor regression seen in 29.6% of cases and tumor stability in 68.5% of patients [32].
In the present series, a clear favorable outcome of meningioma was observed after NA and CHA discontinuation, with most of the patients showing a regression in total tumor volume (50%) or volume stabilization (39.3%). By analyzing each of the 65 meningiomas, there was a higher percentage of tumor growth for midline and convexity lesions although these results were not statistically significant. Furthermore, 89% of our patients were asymptomatic at the end of the follow-up period with only one patient that worsened with a slight volumetric evolution on MRI.
Although there is a clear tendency towards regression or stabilization after NA and CHA discontinuation, similar to what is observed with CA, the percentage of tumor regression seems to be less important after cessation of CHA or NA [11, 15, 18, 32]. Bernat et al. reported a series of 12 CA-related patients, with tumor regression seen in over 90% of the patients. Malaize et al. showed that CA-related meningioma had a better answer to treatment discontinuation compared to CHA and NA-related tumors[11, 32].
Regarding tumor location, a predominance of anterior and middle skull base meningioma was observed, in accordance with previous findings[15, 18, 36]. Prolonged treatment (more than 10 years of progestin intake) seems to be linked to the occurrence of multiple meningioma and these patients appear to have a better clinical course than patients with single lesions. This is also in favor of the gradual appearance of new meningiomas under the influence of progestin intake. Discordant tumor evolution was seen in 4 patients with multiple meningiomas. Interestingly, in these patients, most of the growing meningiomas were either convexity or midline lesions, with a rather posterior location, while the majority of skull base and anterior convexity lesions showed significant regression.
A recent study by Peyre et al. that compared gene sequencing in progestin-associated meningioma vs a control group, showed a specific mutational landscape in progestins-related meningiomas, with a higher frequency of PIK3CA mutation (known to be associated with a more aggressive course in meningioma[37]) in hormone-related meningiomas[38]. When looking at NF2 and TRAF7 mutations (the 2 most frequent gene mutations in meningioma[37]), there was a significantly lower incidence of NF2 and a higher frequency of TRAF7 (40% vs 26%) when compared to the control group. Also, while TRAF7 were found to be associated with AKT1 and KLF4 mutations in the control group, in hormone-associated meningiomas there was a higher rate of isolated TRAF7 mutations.
Our results confirm the strong association between CHA-NA and meningioma. However, the results seem to be less encouraging than previously observed in patients treated by CA, with more patients showing tumor growth despite treatment discontinuation. Still, an important percentage of patients has a favorable response after treatment discontinuation. Therefore CHA-NA discontinuation and close observation should be, in our opinion, the first measure taken if urgent surgery is not needed. While the relationship between hormonal treatment and meningioma seems rather clear, the exact mechanisms of action are still to be determined. The long-term evolution after discontinuation of the drug is also unknown and requires larger scale studies to be properly determined.