Hemifacial myohyperplasia is due to somatic muscular PIK3CA gain-of-function mutations and responds to pharmacological inhibition

Abstract

Hemifacial myohyperplasia (HFMH) is a rare cause of facial asymmetry exclusively involving facial muscles. The underlying cause and the mechanism of disease progression are unknown. Here, we identified a somatic gain-of-function mutation of PIK3CA in five pediatric patients with HFMH. To understand the physiopathology of muscle hypertrophy in this context, we created a mouse model carrying specifically a PIK3CA mutation in skeletal muscles. PIK3CA gain-of-function mutation led to striated muscle cell hypertrophy, mitochondria dysfunction, and hypoglycemia with low circulating insulin levels. Alpelisib treatment, an approved PIK3CA inhibitor, was able to prevent and reduce muscle hypertrophy in the mouse model with correction of endocrine anomalies. Based on these findings, we treated the five HFMH patients. All patients demonstrated clinical, esthetical, and radiological improvement with proof of target engagement. In conclusion, we show that HFMH is due to somatic alteration of PIK3CA and is accessible to pharmacological intervention.

Figure 1.

Patients with hemifacial myohyperplasia. Representative photographs of the five patients and affected muscles assessed with MRI. Scale bar: 5 cm.

Figure S1.

Workflow for the assessment of 3D pictures. First, the template (blue mesh) was rigidly matched with the “target” head surface of each patient (cyan mesh) by an ICP (iterative closest point) alignment. The template was then non-rigidly matched with the target surface by means of an NICP (non-rigid iterative closest point) alignment, which consists of the local deformation of the template following thin-plate splines to adopt the conformation of the target. This resulted in a registered template. The process was repeated for each subject at each age. Then, the facial region was extracted and symmetrized against the non-affected facial side. A Procrustes superimposition was performed to align all faces to tackle undesirable effects of position and orientation. Finally, all subsequent analyses including the generation of heatmaps were computed on the Procrustes coordinates exclusively.

Figure 2.

A mouse model of PIK3CA-related skeletal muscle overgrowth. (A) Male and female body weights of PIK3CA^WT and PIK3CA^HSA-CreER (n = 15 per group) mice following Cre recombination. (B) Coronal whole-body T2-weighted magnetic resonance images of PIK3CA^WT and PIK3CA^HSA-CreER mice. Scale bar: 1 cm. (C) Adipose tissue and skeletal muscle volume quantification. (D) Strength measured using grip test (n = 3 per group). (E) Representative pictures of PIK3CA^WT and PIK3CA^HSA-CreER mice. Scale bar: 1 cm. (F) Western blot of p110α and GFP in skeletal muscles of PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 3 per group). (G) Representative GFP immunofluorescence in skeletal muscles of PIK3CA^WT and PIK3CA^HSA-CreER mice. Scale bar: 10 μm. (H) Representative H&E staining of striated muscles of PIK3CA^WT and PIK3CA^HSA-CreER mice. Scale bar: 20 μm. (I) Western blot of P-AKT^Ser473, total AKT, P-S6RP, and S6RP in skeletal muscles of PIK3CA^WT and PIK3CA^HSA-CreER mice and quantification at 8 and 24 wk of age (n = 5–10 mice per group). (J) Representative immunofluorescence of P-AKT^Thr308 and P-S6RP in skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice. Scale bar: 10 μm. (K) Representative immunofluorescence of KI67 in skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice. Scale bar: 10 μm. (L) Quantification of skeletal muscle cell area of PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 5 mice per group). (M) 12-h fasted glycemia in PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 5–13 mice per group). (N) Insulin circulating levels in PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 10 per group). (O) Circulating IGF-1 levels in PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 10–11 per group). (P) Oral tolerance test (GTT) in PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 6 mice per group). (Q) 18F-FDG uptake in skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 4 per group). (R–T) (R) TMRE, (S) Mitotracker, and (T) 10 NAO staining in skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice (n = 4 per group). Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 (two-tailed unpaired t test). Each dot represents one mouse. Data are representative of at least two independent experiments. Of note, all blots from this figure originate from the same gel. Source data are available for this figure: SourceData F2.

Figure S2.

Mouse model characterization and quantitative histological analysis of tibialis anterior (TA) muscle changes in PIK3CA^WT and PIK3CA^HSA-CreER mice treated with either vehicle or alpelisib. (A) Representative immunofluorescence of PAX7 satellite cells and quantification. (B and C) Digital pictures of TA muscle cross-section area allowing an automated image quantification of (B) major and minor (C) myofiber diameters in PIK3CA^WT and PIK3CA^HSA-CreER mice treated with either vehicle or alpelisib. (D) Quantification. (E) Complete blood count in PIK3CA^WT and PIK3CA^HSA-CreER mice. (F) β Islet area in PIK3CA^WT and PIK3CA^HSA-CreER mice. Data are shown as mean ± SEM. *P < 0.05 and ***P < 0.001 (two-tailed unpaired t test for A, E, and F, and ANOVA, followed by the Tukey–Kramer post hoc test for D). Each dot represents one mouse. Data are representative of at least two independent experiments.

Figure S3.

Metabolic changes observed in PIK3CA^WT and PIK3CA^HSA-CreER. Graphic example of metabolite level changes observed. AU: Arbitrary units. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 (two-tailed unpaired t test).

Figure 3.

Alpelisib prevents and reverses skeletal muscle overgrowth in PIK3CA^HSA-CreER mice. (A) Male body weights of PIK3CA^WT (n = 6) and PIK3CA^HSA-CreER treated with either vehicle (n = 12) or preventive (n = 12) or curative alpelisib (n = 12). (B) Skeletal muscle and adipose tissue volume quantification. (C) Whole-body T2-weighted magnetic resonance images of PIK3CA^WT treated either with vehicle (n = 4) or alpelisib (n = 4), PIK3CA^HSA-CreER vehicle-treated (n = 5), PIK3CA^HSA-CreER-treated with preventive alpelisib (n = 4), and PIK3CA^HSA-CreER-treated with therapeutic alpelisib (n = 4) mice. Scale bar: 1 cm. (D) Representative H&E staining of skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice treated with either vehicle or preventive or curative alpelisib. Scale bar: 10 μm. (E) Representative immunofluorescence of P-AKT^Thr308 and P-S6RP in skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice treated with either vehicle or alpelisib. Scale bar: 10 μm. (F) Western blot and quantification of P110, P-AKT^Ser473, AKT, P-S6RP, and S6RP in skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice treated with either vehicle or alpelisib (n = 4–5 mice per group). (G) 12-h fasted glycemia in PIK3CA^WT and PIK3CA^HSA-CreER mice treated with vehicle, preventive, or curative alpelisib (n = 4 mice per group). (H and I) (H) Insulin and (I) IGF-1 circulating levels in PIK3CA^WT and PIK3CA^HSA-CreER mice treated with vehicle, preventive, or curative alpelisib (n = 4 mice per group). Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 (ANOVA, followed by the Tukey–Kramer post hoc test). Each dot represents one mouse. Data are representative of at least two independent experiments.

Figure S4.

Metabolic changes observed in PIK3CA^WT and PIK3CA^HSA-CreER treated with either vehicle or alpelisib. (A) TMRE, Mitotraker, and 10 NAO staining in skeletal muscle of PIK3CA^WT and PIK3CA^HSA-CreER mice treated with either vehicle or alpelisib (n = 4 per group). (B) Graphic example of metabolite level changes observed (n = 4–7 per group). AU: Arbitrary units. Data are shown as mean ± sem. *P < 0.05 and ***P < 0.001 (ANOVA, followed by the Tukey–Kramer post hoc test).

Figure 4.

Alpelisib improves hemifacial myohyperplasia phenotype in patients. (A) Photographs of the five patients before and 6 mo after alpelisib introduction. (B) 2D mapping of facial asymmetry in 3/5 patients. Blue colors are associated with expanded regions relative to contralateral control. Red colors are associated with regions that display shrinkage compared to contralateral control. Distances (cm) indicate the absolute distance between each node of the affected side and the mirrored non-affected side. Affected side: right. (C) Volumetric changes of facial muscles in patients before and after drug introduction. (D and E) (D) Representative immunofluorescence of P-AKT^Thr308 and P-S6RP in healthy contralateral skeletal muscle and (E) in the affected skeletal muscle before and 6 mo after alpelisib introduction. (F) Quantification of immunofluorescence intensity. AU: Arbitrary units. Scale bar: 10 μm. Data are shown as mean ± SEM. *P < 0.05 and ***P < 0.001 (two-tailed unpaired t test for C and ANOVA, followed by the Tukey–Kramer post hoc test for F). Each dot represents one mouse. Data are representative of at least two independent experiments.

Figure S5.

Covariation between treatment duration and facial asymmetry using PLS2B. PLS2B assessing the covariation between facial morphology (Block 1) and treatment duration (Block 2) in 3/5 patients. rPLS: coefficient of covariation; p: P value; each point represents a 3D face mesh annotated with the duration, in days, since treatment onset. White faces represent the theoretical face deformation along with the covariation axis (from the left to the right: from the start of the treatment, until the most recent 3D picture). Rightmost is a heatmap representation of the main facial regions corrected by the treatment. Colder colors pinpoint expanded regions, whereas hotter colors pinpoint shrunk regions.

Figure 5.

Metabolic changes observed in patients before and 6 mo after alpelisib introduction. Graphic example of metabolite level changes observed. AU: Arbitrary units. Data are shown as mean ± sem. *P < 0.05, **P < 0.01, ***P < 0.001 (two-tailed unpaired t test).