APC mosaicism in a young woman with desmoid type fibromatosis and familial adenomatous polyposis

In FAP, soft tissue tumours like desmoids and epidermoid cysts may present before the appearance of colon polyps [15, 19]. Several studies have demonstrated that asymptomatic patients who have extra-truncal desmoids, are less likely to have FAP [14, 15, 18], and that in FAP patients, most of the desmoids are located in the abdominal cavity/abdominal wall [14‐16]. Testing of the APC gene and/or colonoscopy to rule out FAP is therefore not performed routinely in patients with extra-abdominal desmoids. This patient initially only had extra-truncal fibromatosis and FAP was not considered before she had more extensive fibromatosis in her buttock and along her spine.

There are several reports of patients and families with desmoids, with none or a few polyps, but who harbour a mutation after codon 1900 in the 3′ end of the APC gene [20‐22]. Our patient had her first colonoscopy at 26 years of age and had fewer colon polyps than expected according to the site of her mutation. A colonoscopy done 10–15 years earlier would perhaps have missed the diagnosis, as she may not have developed polyps at that time. Our case illustrates that FAP may be suspected in patients with only extra-abdominal desmoids, and that not only colonoscopy but also genetic testing of the APC gene should be offered to these patients. Previous studies have already suggested testing for APC mutations in children with extracolonic manifestations as hepatoblastoma, osteoma and Gardner fibroma [11, 19, 23].

A de novo APC gene mutation was anticipated, as there was no family history of desmoids, polyps or CRC. Sanger sequencing revealed very weak signals representing the mutation in leukocyte DNA, and could have been missed. However, since mosaicism is not infrequent in FAP, sequencing DNA from one of her adenomas and her colonic mucosa followed up the analyses. This revealed stronger signals representing the mutation, suggests an early postzygotic mutational event involving both the endoderm (adenoma) and the mesoderm (blood) and shows the importance of including adenomas and colonic mucosa in the screening. Other screening methods including NGS have shown higher sensitivity than Sanger sequencing in detecting low level APC-mutation mosaicism [10, 24, 25]. A substantial number of mosaic cases may be missed when doing Sanger sequencing of blood (leukocyte DNA) [10, 24, 25]. Several studies have demonstrated that adenomas should be included in the screening for low level APC mosaicism as a high frequency of cases are not detectable in leukocyte DNA [10, 24, 25]. A recent publication underlines the importance of testing multiple adenomas as co-occurence of sporadic adenomas within a mosaic environment has been reported [26]. Our study also highlights the power of NGS in identifying and characterizing somatic mosaicism in various tissues.

The APC gene mutation in our patient affecting codon 1450 is in the mutation cluster region of the APC gene. This position is at the site for classical FAP, desmoids, osteomas and epidermoid cysts. Despite this, the patient had an attenuated colon polyposis with approximately 50 colon adenomas. But contrary to AFAP, the adenomas had a predominantly distal location in the colon. Her upper gastrointestinal findings were in accordance with what can be seen both in FAP and AFAP. Several studies have shown that somatic mosaicism for APC pathogenic variants usually associated with classic FAP may express a milder colon polyposis phenotype [10, 24, 25]. In these studies, there was only one patient with a desmoid tumour [10]. In our patient, the first fibromatous lesion was noted when she was 2 months old; she had several others over the years and massive fibromatosis in her thigh, buttock and back at the time she was diagnosed with FAP. Some reports have shown a higher incidence of epidermoid cysts and a more severe desmoid disease in patients with a mutation between codon 1445 and codon 1580 than in patients with mutations in other regions of the APC gene [27‐29]. Her mild colon polyposis and severe fibromatosis may therefore be explained by varying distribution of the mutation in different tissues. Additional genetic and environmental factors may also have an influence.

Prophylactic surgery at an early age is recommended in patients with FAP. Surgery is known to be a trigger of desmoid formation, and postponing surgery in patients at risk of developing desmoids could be advocated. In our opinion, such a policy cannot be justified as malignant transformation can occur even under close surveillance. Unfortunately, our patient developed intra-abdominal and abdominal wall desmoids shortly after surgery and she is now in a “wait and see” situation as desmoids can regress, remain stable or progress, in which case medical treatment could be considered.

Although identification of an APC mutation confirmed a FAP diagnosis for this patient, genetic counselling in this situation is difficult. For patients carrying germline mutations, offspring will have a 50% chance of inheriting the mutation. However, when there is somatic mosaicism the risk depends on the level of mosaicism in the patient´s germ cells [10, 30]. Germline mosaicism in an unaffected parent of children with colonic adenomatous polyposis has been reported [10, 31]. Another counselling issue is that the offspring will most likely have a more severe phenotype than the affected parent [25].

We counselled our patient not to become pregnant and give birth, primarily because her pelvic pouch would probably necessitate a caesarean section, which would put her at increased risk of abdominal desmoids. The natural history of her disease has proved this to be sound advice.