GNAQ/GNA11 Mosaicism Is Associated with Abnormal Serum Calcium Indices and Microvascular Neurocalcification

Abstract

Mosaic mutations in genes GNAQ or GNA11 lead to a spectrum of diseases including Sturge-Weber syndrome and phakomatosis pigmentovascularis with dermal melanocytosis. The pathognomonic finding of localized "tramlining" on plain skull radiography, representing medium-sized neurovascular calcification and associated with postnatal neurological deterioration, led us to study calcium metabolism in a cohort of 42 children. In this study, we find that 74% of patients had at least one abnormal measurement of calcium metabolism, the commonest being moderately low serum ionized calcium (41%) or high parathyroid hormone (17%). Lower levels of ionized calcium even within the normal range were significantly associated with seizures, and with specific antiepileptics despite normal vitamin D levels. Successive measurements documented substantial intrapersonal fluctuation in indices over time, and DEXA scans were normal in patients with hypocalcemia. Neurohistology from epilepsy surgery in five patients revealed not only intravascular, but perivascular and intraparenchymal mineral deposition and intraparenchymal microvascular disease in addition to previously reported findings. Neuroradiology review clearly demonstrated progressive calcium deposition in individuals over time. These findings and those of the adjoining paper suggest that calcium deposition in the brain of patients with GNAQ/GNA11 mosaicism may not be a nonspecific sign of damage as was previously thought, but may instead reflect the central postnatal pathological process in this disease spectrum.

Figure 1

Phenotypic and radiological features in GNAQ/GNA11 mosaicism. (a) Clinical features of a patient with SWS with a capillary malformation of the face involving the critical forehead area (Waelchli et al., 2014), associated with glaucoma in the right eye. This patient had hypocalcemia with low levels of ionized and total calcium and increased PTH. (b and c) Clinical features of a patient with PPV-DM exhibiting a capillary malformation (with naevus anaemicus) on the face, upper trunk, and extensive areas of dermal melanocytosis. (d and e) Contrast enhanced fluid-attenuated inversion recovery and T1-weighted magnetic resonance images of a patient with SWS show left frontal, parietal, and occipital pial angiomatosis (vascular malformation, arrows). (f) Susceptibility-weighted imaging (SWI) and (g) SWI phase map depicting vascular calcifications (arrow). This patient had low levels of ionized calcium and total calcium with PTH in the normal range. (h) Axial DWI b0 image does not show calcifications on the baseline magnetic resonance imaging study. (i–k), images from 8-years later. (i) Coronal fluid-attenuated inversion recovery after administration of a gadolinium-based contrast agent shows left parietal and occipital (and some contralateral) sulcal enhancement in keeping with pial angioma. (j) Axial postcontrast T1-weighted image depicts the sulcal enhancement and a prominent draining vein. (k) DWI b0 image at follow-up shows foci of increased susceptibility which are suggestive of calcification. Patients and/or parents/guardians provided written consent for the publication of images. Ca/Cr, calcium to creatinine ratio; PTH, parathyroid hormone.

Figure 2

Fluctuating abnormalities of calcium metabolic profile in patients with GNAQ/GNA11 mosaicism and association with seizures and status epilepticus. (a) Graphical representation of abnormal results in calcium profiling investigations in the cohort of patients at different time points, demonstrating intra and interpatient variability typical in mosaic disease. Where no coloured marker is shown the result was normal. (b) Significant correlation between age and serum calcium corrected to albumin from whole patient cohort. Red and blue dots correspond to serum calcium measurements below or above normal range, respectively, black dots are normal measurements. Linear regression analysis showed a statistically significant negative correlation (P < .001). (c) Correlation between occurrence of seizures and serum ionized calcium corrected to pH from the patients’ cohort. The scatter plot shows the mean of the two groups, and red dots correspond to ionized calcium measurements below normal range. Linear regression analysis showed a statistically significant correlation (P = .013), independent of the effect of age. (d) Correlation between status epilepticus and serum ionized calcium corrected to pH in the patients’ cohort. The scatter plot shows the mean of the two groups, and red dots correspond to ionized calcium measurements below normal range. Linear regression analysis showed statistically significant correlation (P = .017), independent of the effect of age.

Figure 3

Localized intravascular, perivascular, and parenchymal patterns of mineral (calcium) deposition. (a) Image of the cortex with extensive foci of mineralization. (b) A small cortical vessel (likely to be a capillary) with granular mineralization of the wall. (c) A white matter vessel encircled by mineral and fibrosis. (d) A white matter vessel with perivascular deposits and granular parenchymal mineral deposits. In each image, the arrow indicates an example of the mineral deposits. The scale bar represents 500 micrometres (a) and 50 micrometres (b–d).

Supplementary Figure S1

FGF23 (intact molecule and C-terminal) correlation to 1,25-hydroxy vitamin D, serum calcium corrected to albumin and serum phosphate. Note the opposite relationships between 1,25-hydroxyvitamin D and C-terminal FGF23 and its biologically active form intact FGF23. (a and b) Correlation between intact and C-terminal FGF23 and 1,25-hydroxyvitamin D. (c and d) Correlation between intact and C-terminal FGF23 and albumin-corrected serum calcium. (e and f) Correlation between intact and C-terminal FGF23 and serum phosphate. Statistical correlations were performed in Prism as detailed in the methods. FGF23, fibroblast growth factor 23; 1,25OH, 1,25-hydroxy.

Supplementary Figure S2

Calcium homeostatic relationships governed by the parathyroid and kidney behave normally in the GNAQ/GNA11 mosaic cohort. (a) Significant inverse correlation between serum calcium corrected to albumin and serum PTH. (b) Significant correlation between serum calcium corrected to albumin and urine calcium/creatinine ratio. (c) Significant correlation between 1,25-hydroxyvitamin D and intact FGF23. (d) Correlation between 1,25-hydroxyvitamin D and PTH. Statistical correlations were performed in Prism as detailed in the methods. FGF23, fibroblast growth factor 23; 1,25OH, 1,25-hydroxy; PTH, parathyroid hormone.