Corneal Infantile Myofibromatosis Caused by Novel Activating Imatinib-Responsive Variants in PDGFRB

Abstract

Purpose: To investigate the genetic cause, clinical characteristics, and potential therapeutic targets of infantile corneal myofibromatosis.

Design: Case series with genetic and functional in vitro analyses.

Participants: Four individuals from 2 unrelated families with clinical signs of corneal myofibromatosis were investigated.

Methods: Exome-based panel sequencing for platelet-derived growth factor receptor beta gene (PDGFRB) and notch homolog protein 3 gene (NOTCH3) was performed in the respective index patients. One clinically affected member of each family was tested for the pathogenic variant detected in the respective index by Sanger sequencing. Immunohistochemical staining on excised corneal tissue was conducted. Functional analysis of the individual PDGFRB variants was performed in vitro by luciferase reporter assays on transfected porcine aortic endothelial cells using tyrosine kinase inhibitors. Protein expression analysis of mutated PDGFRB was analyzed by Western blot.

Main outcome measures: Sequencing data, immunohistochemical stainings, functional analysis of PDGFRB variants, and protein expression analysis.

Results: We identified 2 novel, heterozygous gain-of-function variants in PDGFRB in 4 individuals from 2 unrelated families with corneal myofibromatosis. Immunohistochemistry demonstrated positivity for alpha-smooth muscle actin and β-catenin, a low proliferation rate in Ki-67 (< 5%), marginal positivity for Desmin, and negative staining for Caldesmon and CD34. In all patients, recurrence of disease occurred after corneal surgery. When transfected in cultured cells, the PDGFRB variants conferred a constitutive activity to the receptor in the absence of its ligand and were sensitive to the tyrosine kinase inhibitor imatinib. The variants can both be classified as likely pathogenic regarding the American College of Medical Genetics and Genomics classification criteria.

Conclusions: We describe 4 cases of corneal myofibromatosis caused by novel PDGFRB variants with autosomal dominant transmission. Imatinib sensitivity in vitro suggests perspectives for targeted therapy preventing recurrences in the future.

Financial disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Keywords: Corneal myofibroma; Imatinib; Infantile myofibromatosis; PDGFRB; Receptor tyrosine kinase inhibitor.

Figure 1

Osseous myofibroma lesion at the left mandible of patient 1. At 9 years of age, individual F1-III-13 was diagnosed with the incidental finding of a myofibroma at the left mandible. The cone-beam computed tomography imaging shows a hypodense, oval-shaped cystic lesion of the left mandible with thinning of the cortical bone (yellow ring).

Figure 2

Pre- and postoperative clinical presentation of corneal myofibroma of individuals F1-III-13 and F2-III-4. A.1, Pre-operative slit-lamp photography of individual F1-III-13 depicts a white corneal tumor of the left eye with superficial vascularization. A.2, In the preoperative slit-lamp-OCT (SL-OCT), the extension and structural composition of the tumor can be acknowledged. A.3, Postoperative slit-lamp photography after lamellar keratectomy shows a clear cornea and (A.4) the postoperative SL-OCT proves the absence of the tumor. Note early postoperative bandage contact lens on top of cornea. A.5, Recurrence of disease noted 18 months after the surgical excision is shown on slit lamp and (A.6) SL-OCT images. B.1, Preoperative slit-lamp photography of individual F2-III-4 shows a white lesion at the corneal center. B.2, The preoperative SL-OCT shows the non-homogenous structure and extension of the tumor. B.3, Post-operative slit-lamp photography after lamellar keratectomy shows a clear cornea. A contact lens for surface protection is present (as in D, too). B.4, In the post-operative SL-OCT, the tumor was absent. B.5, Two years and 2 months after the second surgical excision of the lesion, recurrence of disease was noted on the slit lamp and (B.6) SL-OCT. C.1, Individual F2-II-2 with bilateral lesions of the cornea (right eye [OD], C.1–C.3) and of both conjunctiva and cornea (left eye [OS]; C.4–C.7). C.1–C.3, The corneal lesion of the right eye seems stable over time. C.4, Pterygium-like lesion of the left eye with growth of corneal tissue onto the cornea (C.5) which was mostly stable over a time course of 2 years. C.6, Three months after the surgical excision of the left eye in February 2022, recurrence of disease with a corneal lesion was noted. It is depicted 1 year and 7 months after surgery on slit lamp imaging and (C.7) Casia 2 anterior segment OCT imaging. PDGFRB = platelet-derived growth factor receptor beta.

Figure 3

Family pedigrees and sequencing data of patient 1 and 2. A, Pedigree of family 1 with individuals F1-III-13 (black arrowhead) and F1-III-15 with pathogenic platelet-derived growth factor receptor beta gene (PDGFRB) variants. Sanger sequencing screenshots by Sequence Pilot (SeqPilot) of the affected individuals show PDGFRB c.1766A>G (p.Y589C) in exon 12 of PDGFRB. Clinically affected but not genetically tested female family members of the maternal side (F1-I-4, F1-II-6, F1-II-7, and F1-III-14) with a “pterygium”-like phenotype, a lesion affecting the conjunctiva and cornea, are highlighted in oblique stripes. B, Pedigree of family 2 with affected individuals F2-II-2 and F2-III-4. Exome-based panel sequencing of PDGFRB revealed the variant c.1949C>G (p.S650W), which was confirmed by Sanger sequencing of PDGFRB exon 14.

Figure 4

Histopathologic evaluation of resected tumors of individual F1-III-13 and F2-III-4. A, E, Hematoxylin–eosin (H&E) staining at 20x (A) and 10x (E) magnification, respectively, shows spindle-shaped cells. B, F, Immunohistochemical staining (10x magnification) demonstrates a positive staining for alpha smooth muscle actin (α-SMA) in both specimens. C, G, Ki-67 staining shows a low proliferation rate of < 5%. D, H, Desmin staining was marginally positive (see red arrow heads).

Figure 5

Characterization of platelet-derived growth factor receptor beta (PDGFRB) variants. A, Localization of PDGFRB variants in the kinase domain structure. The structure is based on the highly conserved PDGFR alpha intracellular part (Protein Data Bank reference 5K5X) and was visualized using Pymol. The juxtamembrane domain is represented in yellow (with tyr589 in blue) and the tyrosine kinase domain in green (with ser650 in red). B, Platelet-derived growth factor receptor beta variant p.Y589C and p.S650W are constitutively active and sensitive to imatinib. Luciferase reporter assays were performed in porcine aortic endothelial cells after transient transfection with the indicated receptors. Cells were left untreated (black bars), stimulated with platelet-derived growth factor (PDGF; white bars) or treated with both PDGF and imatinib (grey bars). The constitutively activated p.P584R variant was used as a positive control. The mean of 6 independent experiments is shown with standard error of the mean (Student t test; ∗P ≤ 0.05; ∗∗P ≤ 0.01). C, Expression of PDGFRB variants in transfected HEK293T cells. Lysates of transiently transfected cells were analyzed by western blot using anti-PDGFRB and anti-β-actin antibodies. WT = wild-type.

Figure 6

Substitution of serine 650 by histidine or tyrosine also activates platelet-derived growth factor receptor beta (PDGFRB). A, Serine 650 was mutated in the indicated amino-acid. Luciferase reporter assays were performed in porcine aortic endothelial cells after transient transfection with the indicated receptors. Cells were stimulated by platelet-derived growth factor (PDGF) as positive control (white bars). The mean of 3 independent experiments is shown with standard error of the mean (Student t test; ∗P ≤ 0.05; ∗∗P ≤ 0.01; ∗∗∗P ≤ 0.001). B, Expression of PDGFRB variants in transfected HEK293T cells. Lysates of transiently transfected cells were analyzed by western blot using anti-PDGFRB and anti-β-actin antibodies. WT = wild-type.