RESUMEN
Cerebral cavernous malformation (CCM) is a hemorrhagic cerebrovascular disease where lesions develop in the setting of endothelial mutations of CCM genes, with many cases also harboring somatic PIK3CA gain of function (GOF) mutations. Rapamycin, an mTORC1 inhibitor, inhibited progression of murine CCM lesions driven by Ccm gene loss and Pik3ca GOF, but it remains unknown if rapamycin is beneficial in the absence of induction of Pik3ca GOF. We investigated the effect of rapamycin at three clinically relevant doses on lesion development in the Ccm3-/-PDGFb-icreERPositive murine model of familial CCM disease, without induction of Pik3ca GOF. Lesion burden, attrition, and acute and chronic hemorrhaging were compared between placebo and rapamycin-treated mice. Plasma miRNome was compared to identify potential biomarkers of rapamycin response. Outlier, exceptionally large CCM lesions (> 2 SD above the mean lesion burden) were exclusively observed in the placebo group. Rapamycin, across all dosages, may have prevented the emergence of large outlier lesions. Yet rapamycin also appeared to exacerbate mean lesion burden of surviving mice when outliers were excluded, increased attrition, and did not alter hemorrhage. miR-30c-2-3p, decreased in rapamycin-treated mouse plasma, has gene targets in PI3K/AKT and mTOR signaling. Progression of outlier lesions in a familial CCM model may have been halted by rapamycin treatment, at the potential expense of increased mean lesion burden and increased attrition. If confirmed, this can have implications for potential rapamycin treatment of familial CCM disease, where lesion development may not be driven by PIK3CA GOF. Further studies are necessary to determine specific pathways that mediate potential beneficial and detrimental effects of rapamycin treatment, and whether somatic PIK3CA mutations drive particularly aggressive lesions.
RESUMEN
Cerebral cavernous malformations (CCMs) are ectatic capillary-venous malformations that develop in approximately 0.5% of the population. Patients with CCMs may develop headaches, focal neurologic deficits, seizures, and hemorrhages. While symptomatic CCMs, depending upon the anatomic location, can be surgically removed, there is currently no pharmaceutical therapy to treat CCMs. Several mouse models have been developed to better understand CCM pathogenesis and test therapeutics. The most common mouse models induce a large CCM burden that is anatomically restricted to the cerebellum and contributes to lethality in the early days of life. These inducible models thus have a relatively short period for drug administration. We developed an inducible CCM3 mouse model that develops CCMs after weaning and provides a longer period for potential therapeutic intervention. Using this new model, three recently proposed CCM therapies, fasudil, tempol, vitamin D3, and a combination of the three drugs, failed to substantially reduce CCM formation when treatment was administered for 5 weeks, from postnatal day 21 (P21) to P56. We next restricted Ccm3 deletion to the brain vasculature and provided greater time (121 days) for CCMs to develop chronic hemorrhage, recapitulating the human lesions. We also developed the first model of acute CCM hemorrhage by injecting mice harboring CCMs with lipopolysaccharide. These efficient models will enable future drug studies to more precisely target clinically relevant features of CCM disease: CCM formation, chronic hemorrhage, and acute hemorrhage.
