Muir–Torre syndrome is a rare autosomal dominant condition, first described in 1967 in an individual who had multiple benign sebaceous adenomas and keratoacanthoma of the skin and multiple internal malignancies (large bowel, duodenum, and larynx) (Muir et al. 1967). The skin stigmata of this condition include sebaceous hyperplasia, adenoma, and carcinoma, with keratoacanthoma and basal cell cancer (BCC). The internal neoplasias include Tumours of the colon, stomach and esophagus, breast, uterus, ovaries, bladder, and larynx and squamous cell carcinomas of the mucous membranes (Grignon et al. 1987). The syndrome is defined by dermatologists as a combination of at least one sebaceous gland Tumours and a minimum of one internal malignancy (Cohen et al. 1991). The condition usually becomes manifest from the fifth decade of life, and multiple skin lesions develop.
It has become apparent that the main hereditary cause of Muir–Torre syndrome is mutations in the MMR genes MSH2, MLH1 and MSH6. The presence of identical mutations in families with HNPCC/Lynch syndrome and Muir–Torre syndrome confirms that this syndrome is nearly always a variant of Lynch syndrome (Sieber et al. 2003a, b; Aretz et al. 2013; Bartkova et al. 2008; Kolodner et al. 1994). The implication of this conclusion is that individuals with sebaceous carcinoma and a mutation in a MMR gene should be considered to be at an elevated lifetime risk of all the cancers that are known to occur in Lynch syndrome. In a large, single-institution study, 27 of 41 patients with Muir–Torre syndrome had mutations in MLH1 or MSH2 (Mangold et al. 2004). Interestingly, 25 of the 27 mutations were in MSH2, confirming multiple earlier observations that found far more mutations in MSH2 than in MLH1 (Lucci-Cordisco et al. 2003). Because the ascertainment criteria in the recent study of Mangold et al. were a sebaceous gland neoplasm and at least one internal neoplasm in the same patient (regardless of site of cancer or family history), it is not surprising that not all individuals with germline MMR gene mutations fell within the Bethesda guidelines for gene testing. Probably all sebaceous neoplasms occurring in the context of Muir–Torre syndrome are associated with MSI (Kruse et al. 2003), or loss of expression of MSH2 (or rarely MLH1 and MSH6) (Fiorentino et al. 2004), so all patients with sebaceous neoplasms should be offered either or both of these tests. If either test suggests a MMR gene mutation, analysis should be carried out, even in the absence of an internal malignancy or a positive family history. The situation may be particularly challenging on the rare occasion when a child is diagnosed with a sebaceous neoplasm (which are usually periocular) (Omura et al. 2002); as in the absence of family history, testing of a minor may be warranted. From a laboratory standpoint, the spectrum of mutations in MSH2 is similar to that seen in non-Muir–Torre-associated Lynch syndrome, no genotype-phenotype associations within MSH2 are seen (Mangold et al. 2004), and a search for genomic deletions is indicated if sequenced-based analysis is negative (Barana et al. 2004).
Screening for cancer in relatives of individuals with this disorder is obviously important and should be similar to that outlined for Lynch syndrome, with extracolonic surveillance, particularly of the skin. It is interesting to note that in many families with Muir–Torre syndrome, only one person is affected by a sebaceous cancer, suggesting that other modifying factors may be present. In one Quebec family, which was later identified to carry a germline deletion in MLH1, the one person who developed sebaceous carcinoma only did so after receiving a heart transplant and subsequent immunosuppression (Paraf et al. 1995). The generality of this finding has been confirmed in a more recent study (Harwood et al. 2003).
When exposed to N-nitrosomethylbenzylamine, mice heterozygous for the gene encoding the fragile histidine triad gene, Fhit, developed a syndrome akin to human Muir–Torre syndrome (Fong et al. 2000). In human sebaceous carcinomas, FHIT protein was absent only from those Tumours that were microsatellite stable, suggesting an alternative pathway to skin Tumours in this syndrome (Holbach et al. 2002). No germline mutations in FHIT have been identified in Muir–Torre (or any other human cancer) syndrome. On the other hand, there have been reports of sebaceous carcinomas in MUTYH- associated polyposis and thus not all Muir-Torre syndrome is due to MMR gene mutations (Vogt et al. 2009).