Glomuvenous Malformations

A number sign (#) is used with this entry because of evidence that glomuvenous malformations are caused by heterozygous mutation in the glomulin gene (601749) on chromosome 1p22.

Description

Glomuvenous malformations, also known as 'venous malformations with glomus cells' or glomangiomas, are similar to mucocutaneous venous malformations (VMCM; 600195), but clinically are distinguishable: they have a cobble-stone appearance, have a consistency harder than that of venous malformations, and are painful on palpation. Histologically, GVMs are distinguishable by the presence of pathognomonic rounded cells (glomus cells) around the distended vein-like channels. The term glomus (Latin for ball) stems from the morphologically similar contractile cells of the Sucquet-Hoyer arteriovenous anastomoses in glomus bodies that are involved in cutaneous thermoregulation. Glomus cells in GVMs appear to be incompletely or improperly differentiated vascular smooth muscle cells, since they stain positively with smooth muscle cell alpha-actin (102620) and vimentin (193060) (summary by Brouillard et al., 2002). The genetic distinctness of glomuvenous malformations from mucocutaneous venous malformations is indicated by the fact that mutations have been found in the TIE2/TEK gene (600221) in mucocutaneous venous malformations and not in glomuvenous malformations.

Clinical Features

Glomus tumors are benign cutaneous neoplasms that are derived from specialized arteriovenous shunts that occur normally in many parts of the body. Gorlin et al. (1960) reported 5 affected members in 2 generations of a family. The lesions tend to resemble cavernous hemangiomas. The distinctive feature is the presence of multiple layers of glomus cells lining the blood-filled cavities. The tumors are present at birth or appear in the first 2 decades. Isolated glomus tumor usually develops later (at about age 33 years on the average), is more frequently subungual than is the case with multiple tumors, and has no particular familial occurrence. Reed (1970) presented a pedigree of 4 persons with multiple glomus tumors in 2 generations.

Beasley et al. (1986) reported 4 patient with GVMs in 3 generations. The 9-year-old proposita had had 6 soft, blue-black skin lesions from birth, on the forearm, thigh, and buttocks. All but one were raised.

As pointed out by Boon et al. (1999), glomus tumors, or glomangiomas, are a clinical and radiologic subtype of venous malformations. Their pathognomonic characteristic is the presence of undifferentiated smooth muscle cells (glomus cells) surrounding convoluted venous channels. Although clinically they look like any venous malformation, they are more painful on palpation, only partially compressible, and usually not found in mucosa. In addition, familial aggregation is more common than in venous malformations generally, and several pedigrees showing autosomal dominant inheritance have been reported. Iqbal et al. (1998) estimated that penetrance rises from 70% at age 5 years to 100% by age 30 years.

Inheritance

The inheritance of the cutaneous disorder discussed here is uncomplicated autosomal dominant with many instances of male-to-male transmission. It is not to be confused with multiple paragangliomata (168000), which is often referred to as glomus tumors.

In various autosomal dominant skin disorders, segmental forms reflecting mosaicism have been reported. Happle (1997) delineated 2 different types of mosaic manifestation. Type 1 reflects heterozygosity for the underlying mutation and shows the same degree of severity as that observed in the corresponding nonmosaic phenotype. In the case of cutaneous disorders, the skin other than that in the segmental area is normal. Type 2 originates from loss of heterozygosity and shows an excessively severe involvement in a segmental area, usually superimposed on the disseminated lesions of the ordinary trait. Happle and Konig (1999) surveyed the literature on multiple glomus tumors and found 5 cases suggesting a type 2 segmental involvement. In all of these cases, a unilateral band-like or patchy arrangement of excessively pronounced glomus tumors was associated with disseminated lesions corresponding to the ordinary phenotype, and in 3 cases other family members were affected with disseminated glomus tumors. The unilateral agminated (i.e., gathered in clusters) lesions were reported to be present in early childhood, whereas the disseminated lesions appeared later.

Mapping

Boon et al. (1999) demonstrated that 5 families with inherited cutaneous venous malformations with glomus cells showed linkage to 1p22-p21 (lod score = 12.7 at recombination fraction = 0.00). They designated the locus VMGLOM. Irrthum et al. (2001) reported 7 additional families with glomangioma showing linkage to 1p22-p21. Combined with the families reported by Boon et al. (1999), they found a lod score of 18.41 at theta = 0.0 at marker D1S188. Haplotype analysis revealed evidence for a founder effect in some of the families. In 4 glomangioma families, Calvert et al. (2001) found linkage of the trait to 1p22-p21.

Brouillard et al. (2000) reported a physical map based on 18 overlapping YAC clones spanning the 5-Mb VMGLOM locus. They also reported a sequence-ready PAC map of 46 clones covering 1.48 Mb within the YAC contig, a region to which they restricted VMGLOM. They identified several positional candidate genes within a narrowed region on 1p and found that one of these, designated originally FAP48 (601749), contained mutations in cases of GVMs.

Molecular Genetics

In connection with the demonstration of mutations in glomulin as the cause of this disorder, Brouillard et al. (2002) found a somatic 'second hit' mutation (601749.0002) in affected tissue of a patient with an inherited genomic deletion (601749.0001). Furthermore, since all but one (601749.0004) of the 14 different germline mutations identified in patients with GVMs resulted in premature stop codons, and since the localized nature of the lesions could be explained by the Knudson 2-hit model, GVMs are likely caused by complete loss of function of glomulin.

Amyere et al. (2013) hypothesized that a Knudson '2-hit' model could explain the multifocality and partial penetrance of GVMs, and performed a systematic analysis using multiple approaches, including a sensitive allele-specific pairwise SNP-chip method. Amyere et al. (2013) identified 16 somatic mutations, most of which were not intragenic but were cases of acquired uniparental isodisomy (aUPID) involving chromosome 1p. The breakpoint of each aUPID is located in an A- and T-rich high DNA flexibility region (1p13.1-1p12). This region corresponds to a possible fragile site. Occurrences of these mutations render the inherited glomulin variant in 1p22.1 homozygous in the affected tissues without loss of genetic material. Amyere et al. (2013) concluded that this finding demonstrates that a double hit is needed to trigger formation of a GVM.

Nomenclature

Strauchen (2002) noted 'a common point of confusion,' namely the interchangeable use of 'paraganglioma' and 'glomus tumor.' He emphasized that the glomus tumor is a tumor of modified perivascular smooth muscle, which frequently presents as a painful subungual mass, and is unrelated to tumors of the adrenal and extraadrenal paraganglia. Jugulotympanic paraganglioma is often referred to as a 'glomus jugulare tumor.' This tumor arises from minute, anatomically dispersed paraganglia located at the base of the skull and temporal bone and is closely related to similar tumors of the carotid body and other extraadrenal paraganglia. It is unrelated to the much more common glomus tumor of skin and soft tissue.