Proteus syndrome is a life-threatening segmental overgrowth syndrome caused by a mosaic gain-of-function AKT1 variant. There are no effective treatments for Proteus syndrome. Miransertib is an AKT1 inhibitor that, prior to this study, has been evaluated only in adult oncology trials. We designed a non-randomized, phase 0/1 pilot study of miransertib in adults and children with Proteus syndrome to identify an appropriate dosage starting point for a future efficacy trial using a pharmacodynamic endpoint. The primary endpoint was a 50% reduction in the tissue levels of AKT phosphorylation from biopsies in affected individuals. We also evaluated secondary efficacy endpoints. We found that a dose of 5 mg/m 2 /day (1/7 the typical dose used in oncology) led to a 50% reduction in phosphorylated AKT (pAKT) in affected tissues from five of six individuals. This dose was well tolerated. Two of the six efficacy endpoints (secondary objectives) suggested that this agent may be efficacious. We observed a decrease in a cerebriform connective tissue nevus and a reduction in pain in children. We conclude that 5 mg/m 2 /day of miransertib is an appropriate starting point for future efficacy trials and that this agent shows promise of therapeutic efficacy in children with Proteus syndrome.

Phenotype‐based diagnostic criteria were developed for Proteus syndrome in 1999 and updated in 2006. Subsequently, the causative mosaic gene alteration was discovered, the c.49G>A p.E17K variant in AKT1. As well, a number of overlapping overgrowth disorders attributable to mosaic PIK3CA variants have now been characterized, leading to the designation of PIK3CA‐related overgrowth spectrum (PROS). Finally, ongoing work to better characterize Proteus syndrome has led to identification of additional features of that disorder that could be useful in diagnostic criteria. We have taken the opportunity of these discoveries to re‐evaluate the Proteus syndrome diagnostic criteria. Here we propose a new set of diagnostic criteria that establishes a weighted, point‐based system for the phenotypic attributes and then integrates that with the potential molecular test results to result in one of two designations: AKT1‐related Proteus syndrome or AKT1‐related overgrowth spectrum. A patient whose only manifestation is an AKT1 c.49G>A‐positive tumor would receive neither of these designations. Here we review the rational basis of diagnostic criteria and argue that a unitary diagnostic entity is a distinct gene‐phenotype dyad and that this should be the model for all mendelian disorders. The gene‐alone or phenotype‐alone approach is inadequate to rigorously delineate a unitary diagnostic entity.

Increased risk of thromboembolism has been recognized in individuals with mosaic overgrowth disorders, Proteus syndrome (PS) and PIK3CA‐related overgrowth spectrum (PROS), including Klippel–Trenaunay syndrome and CLOVES syndrome. PS and PROS have distinct, yet overlapping clinical findings and are caused by somatic pathogenic variants in the PI3K/AKT gene signaling pathway. PS is caused by a single somatic activating AKT1 c.49G > A p.E17K variant while PROS can be caused one of multiple variants in PIK3CA. The role of prothrombotic factors, endothelial cell adhesion molecules, and vascular malformations in both PS and PROS have not been previously investigated. A pilot study of prospective clinical and laboratory evaluations with the purposes of identifying potential risk factors for thrombosis was conducted. Doppler ultrasounds and magnetic resonance angiogram/ venography (MRA/MRV) scans identified vascular malformations in PS and PROS that were not appreciated on physical examination. Abnormal D‐dimers (0.60–2.0 mcg/ml) occurred in half of individuals, many having vascular malformations, but no thromboses. Soluble vascular endothelial markers, including thrombomodulin, soluble vascular adhesion molecule (sVCAM), soluble intercellular adhesion molecule (sICAM), E‐selectin, and P‐selectin were significantly higher in PS and PROS compared to controls. However, no single attribute was identified that explained the risk of thrombosis. Predisposition to thrombosis is likely multifactorial with risk factors including chronic stasis within vascular malformations, stasis from impaired mobility (e.g., following surgery), decreased anticoagulant proteins, and effects of AKT1 and PIK3CA variants on vascular endothelium. Based on our findings, we propose clinical recommendations for surveillance of thrombosis in PS and PROS.