Attenuated Familial Adenomatous Polyposis: A Case Report With Mixed Features and Review of Genotype-Phenotype Correlation

Colorectal cancer is the third most commonly diagnosed cancer in both men and women, and the second leading cause of cancer death. In 2003, it was estimated that 147 500 new cases of colorectal cancer would be diagnosed, and 57 100 deaths from colorectal cancer would occur.1 

Familial adenomatous polyposis (FAP), also known as adenomatous polyposis coli (APC), is an autosomal-dominant disorder affecting roughly 1 in 8000 individuals in the United States and represents approximately 1% of all colorectal cancers.2 The hallmark of this disease is the development of hundreds of adenomatous polyps in the colon and rectum that usually emerge during the second and third decade of life and harbor a high risk of malignant transformation (nearly 100% by age 60 years).

It is now recognized that FAP has a broad spectrum of clinical manifestations and, in addition to classic FAP, includes 3 phenotypes: attenuated FAP (AFAP), Gardner syndrome, and Turcot syndrome. Patients with classic FAP present with more than 100 colorectal adenomatous polyps or fewer than 100 adenomatous polyps and a relative diagnosed with FAP. Patients with AFAP present with fewer than 100 colonic adenomatous polyps that are found more proximally in the colon than in classic FAP. The average age of colon cancer diagnosis in individuals with attenuated FAP is 50 to 55 years, which is 10 to 15 years later than that observed in classic FAP. A predominance of right-sided colorectal adenomas and rectal polyp sparing have been observed in these patients. Attenuated FAP also has been described as “hereditary flat adenoma syndrome,” since colonic adenomas may present with a flat (rather than polypoid) morphology. Gastric fundic polyps and duodenal adenomas are also seen, and, in contrast to classic FAP, congenital hypertrophy of retinal pigment epithelium lesions and desmoid tumors rarely have been described in AFAP.3,4 Gardner syndrome is the association of colonic adenomatous polyposis, osteomas, and soft tissue tumors (epidermoid cysts, fibromas, desmoid tumors) caused by mutations in the APC gene. Turcot syndrome is the association of colon cancer and central nervous system tumors, usually medulloblastoma. Surgical pathologists should be aware of these variants and keep them in mind when evaluating biopsies or resections from patients with multiple upper and lower gastrointestinal tract adenomas.

Mutations 5′ of codon 158 are associated with the attenuated phenotype.5 We describe a patient with a mutation in codon 161, who displays a phenotype with intermediate features between the attenuated and classic forms of FAP.

A white woman in her early 40s had been in good health until she developed episodic pain in the right upper quadrant and back. The patient's family history was positive for first-degree relatives with breast cancer (>50 years old) and renal cancer (>40 years old). She underwent laparoscopic cholecystectomy, and on pathologic examination, chronic follicular cholecystitis and two 7-mm, jet black, irregular gallstones were found.

The patient continued to experience occasional atypical right upper quadrant pain with persistent minor elevation of aspartate aminotransferase and alanine aminotransferase. An upper endoscopy showed small, benign-appearing fundic polyps and a hypertrophied, friable ampulla of Vater. Endoscopic ultrasound showed a mass in the ampulla, measuring 20 mm and confined to the mucosa with possible extension into the distal bile duct. Brushings of the ampulla disclosed clusters of atypical columnar epithelial cells, and a biopsy revealed a tubular adenoma without high-grade dysplasia (Figure 1, A and B). Colonoscopy was performed and demonstrated numerous colonic polyps measuring 3 to 5 mm in diameter, and a 1-cm sigmoid polyp. Histopathologic examination of the polyps revealed that the sigmoid polyp was a tubulovillous adenoma and the smaller polyps were tubular adenomas.

On exploratory laparotomy, a transduodenal ampullectomy and a total abdominal colectomy with end-to-end ileorectal anastomosis was performed. The colonic mucosa was tan and involved with innumerable (around 100), but not confluent, small, tan, round, sessile polyps, which were distributed throughout the entire colon; all were less than or equal to 0.2 cm in diameter (Figure 1, C). On microscopic examination, diffuse adenomatous changes of the colonic mucosa were seen, but there was no evidence of high-grade dysplasia or carcinoma. The ampullary tumor was a tubular adenoma without high-grade dysplasia (Figure 1, D).

The mutational alterations at the APC locus usually manifest as allelic loss (loss of the nonmutated allele in the tumor), also known as loss of heterozygosity, which is generally regarded as a measure of tumor suppressor gene deletion. Manual microdissection genotyping from serial 4-μm histologic sections of gastrointestinal tract biopsies was used to determine loss of heterozygosity at the APC locus. Three to 4 unstained histologic sections were microdissected under stereomicroscopic visualization, as previously described.6 Six different areas were selected for microdissection genotyping as follows: normal colonic epithelium, colonic tubulovillous adenomatous polyp, colonic tubular adenoma, colonic mucosa with early adenomatous change, ampullary adenomatous polyp, and gastric mucosa. Microdissected tissue was collected in appropriately labeled Eppendorf tubes containing 50 μL of buffer (Tris-HCl, pH 7.0) and digested with proteinase K, as described previously.6 

Each microdissected tissue target was assessed for allelic loss pattern with extragenic adjacent microsatellite markers for APC loss using markers MCCE10 (5q23.1), 5q21D5S592 (5q23.3), and 5q21D5S615(5q23.3). Markers 5q21D5S592 and 5q21D5S615, situated at 135.574 centimorgans and 135.659 centimorgans from the p terminal of chromosome 5, were informative. No consistent pattern of allelic loss was identified. The allelic loss analysis was then extended to the APC gene itself. An informative single nucleotide sequence polymorphism in exon 15 (A/G NCBI SNP cluster identification: rs42427) was evaluated in microdissected nonneoplastic and polyp tissue (Figure 2). In the nonneoplastic tissue, both alleles were represented in approximately equal amounts; in the microdissected polyp tissue (T1 and T2), there was an imbalance in the representation of the individual nucleotides in keeping with allelic loss. Of note was the finding that all polyps analyzed (T1–T6) manifested the same pattern of allelic loss, supporting the concept of a germline alteration in the APC gene. This finding provided further support for more detailed mutational screening.

A mutation scanning technique was used to determine the presence of mutations in the APC gene in peripheral blood (germline). A polymerase chain reaction–based assay was used to amplify all 15 exons of the APC gene and intron/exon boundaries. The amplified products were scanned for mutations using denaturing high-performance liquid chromatography, as described by Wu et al7 (Figure 3, A).

Amplicons showing variations were subsequently sequenced for mutation/polymorphism confirmation. A single base pair change of a cytosine (C) to a thymine (T) was detected at nucleotide position 481 (481 C→T) (Figure 3, B), predicted to result in a change of a glutamine residue to a stop codon at codon position 161 (Q161X) (GenBank mRNA accession number NM_000038, nucleotide n1 corresponds to the A residue of the start ATG).4,7 The Q161X mutation identified in this patient is predicted to produce a truncated protein and is interpreted as a disease-causing mutation. These findings are consistent with the diagnosis of FAP. Family studies revealed the presence of this same mutation in the patient's asymptomatic daughter.

Familial adenomatous polyposis is caused by germline mutations in the APC gene.8 The APC gene spans approximately 150 kb of genomic DNA and has 15 exons, which are spliced into a messenger RNA transcript of about 8 kb. The gene encodes for a protein of 2843 amino acids that is expressed in many adult tissues. Almost all APC mutations identified to date are nonsense or frameshift mutations leading to the synthesis of truncated proteins. Most mutations are located within the first 2000 codons of the APC gene, and several mutational hot spots have been identified at codons 1061 and 1309.7 

The relative location of the germline mutation in the APC gene in patients with polyposis may be associated with the number of polyps and partially predicts specific phenotypic expression.9 Mutations associated with the attenuated phenotype are found predominantly in the 5′ region of the gene (exons 3 and 4), or in the last third (3′ end of exon 15), although AFAP cases have been reported with mutations in exons 6 and 9.10 Phenotypic expression in these 3 groups was shown to be variable but milder than that in classic FAP. Greater phenotype variability in polyp number with predominance of polyps located on the right side and increased occurrence of duodenal adenomas has been observed in AFAP families carrying 5′ germline APC mutations.11 Extracolonic features may be more common in patients with mutations in the 3′ end.

Attenuated FAP patients can be difficult to diagnose, since their clinical phenotype might go unrecognized for some time. Pathologists should be aware of the associations between different central nervous system, soft tissue, endocrine, skin, and gastrointestinal tumors and the clinical phenotypes of FAP patients, and could recommend genetic testing when a suspicious case is identified. Molecular-genetic testing has been suggested as an aid in decision making with respect to the management of patients with the attenuated phenotype.12 Several factors need to be considered in treatment of AFAP patients, including age at AFAP diagnosis, number of adenomatous polyps (>20), location of polyps (right-side colon), frequency of polyp recurrence, and polyp morphology (confluent vs scattered, histologic appearance, and presence of high-grade dysplasia). This information needs to be correlated with the molecular diagnosis of AFAP in affected presymptomatic individuals and family members.

In our case, the patient presented with the clinical picture of an attenuated form of FAP, that is, negative family history, late presentation (older than 40 years old), no evidence of cancer, and sparing of the rectum by polyps, but the colectomy showed more than 100 small, sessile, adenomatous colonic polyps. Based on polyp number, this case should be classified as classic FAP, contrary to the strong clinical evidence of this being AFAP. However, the Q161X mutation identified in this patient is situated in close proximity to the codons that are usually mutated in the attenuated form of the disease (5′ to codon 158). Furthermore, this mutation has been described in the literature in only 1 other patient, who presented with an attenuated phenotype of FAP and serrated adenoma.13,14 These contradictory findings underscore the variability in AFAP phenotypes and the need for clinicopathologic correlation. The fact that 2 different patients identified as carriers of the same mutation (Q161X) have phenotypes representing variations in the attenuated form of FAP would suggest that this mutation should be regarded as causative of AFAP.

It is important to remember that differences in the APC mutation sites alone cannot completely account for intrafamilial and interfamilial variation in the polyposis phenotypes. Other modifying genes, epigenetic mechanisms, or environmental factors not yet identified could play a role in determining the phenotype a patient with FAP will have. Pathologists need to be aware of the variability in FAP phenotypes, including AFAP, and should perform a thorough histopathologic evaluation in cases in which features suggestive of this disease are present. Molecular testing of these patients can help confirm the presence of FAP, and the detection of a specific mutation can have significant impact on the clinical follow-up of these patients and the strategy for screening other family members.

The authors acknowledge Sydney Finkelstein, MD, PhD, for performing the allelic loss studies and for providing the image of the allelic loss at the APC locus.