Familial Adenomatous Polyposis 1

A number sign (#) is used with this entry because familial adenomatous polyposis-1 (FAP1) and its variant Gardner syndrome are caused by heterozygous mutation in the APC gene (611731) on chromosome 5q22.

See also hereditary desmoid disease (135290), an allelic disorder considered by some (e.g., Lynch, 1996) to be a variant of FAP.

Description

Familial adenomatous polyposis-1 is an autosomal dominant disorder characterized by predisposition to cancer. Affected individuals usually develop hundreds to thousands of adenomatous polyps of the colon and rectum, a small proportion of which will progress to colorectal carcinoma if not surgically treated. Gardner syndrome is a variant of FAP in which desmoid tumors, osteomas, and other neoplasms occur together with multiple adenomas of the colon and rectum (Nishisho et al., 1991).

Rustgi (2007) reviewed the genetics of hereditary colon cancer, including APC.

Genetic Heterogeneity of Familial Adenomatous Polyposis

See also autosomal recessive FAP2 (608456), caused by mutation in the MUTYH gene (604933) on chromosome 1p34; autosomal recessive FAP3 (616415), caused by mutation in the NTHL1 gene (602656) on chromosome 16p13; and autosomal recessive FAP4 (617100), caused by mutation in the MSH3 gene (600887) on chromosome 5q11.

Nomenclature

Early terms for this disorder include multiple polyposis of the colon, hereditary polyposis coli, familial multiple polyposis, and familial polyposis of the colon (FPC). The designation familial adenomatous polyposis (FAP) is most often used today, particularly in Great Britain, based in part on the appreciation that the polyps are not confined to the colon. FAP has also been used as an acronym for familial amyloid polyneuropathy (176300) and for fibroblast activation protein (600403).

Clinical Features

Gardner (1951) reported a large Utah family with intestinal polyposis that appeared to be a predisposing factor for carcinoma of the colon and rectum. Inheritance was autosomal dominant. In ensuing years, affected family members developed other abnormal growths, including intestinal polyps, osteomas, fibromas, and sebaceous cysts. Desmoid tumors, dental abnormalities, carcinoma of the ampulla of Vater, and thyroid carcinoma were also reported (Gardner and Plenk, 1952; Gardner, 1962). In a follow-up of this original family, Naylor and Gardner (1977) concluded that the mutant gene shows high penetrance and variable expressivity. Danes and Gardner (1978) noted that some branches of the original Utah family had the full syndrome, including both colonic and extracolonic lesions, whereas other branches had only extrabowel lesions.

Gorlin and Chaudhry (1960) described familial association of multiple intestinal polyposis, multiple osteomata, fibromas, lipomas, and fibrosarcomas of the skin and mesentery, epidermoid inclusion cysts of the skin, and leiomyomas, and suggested that it was a heritable disorder of connective tissue.

Savage (1964) reported a woman with Gardner syndrome who had multiple colorectal adenomas and rectal carcinoma, desmoid tumors, multiple sebaceous cysts, an osteoma of the forehead, and 2 subcutaneous lipomata.

Although FAP patients with extracolonic features have been referred to in the past as having a distinct phenotype labeled 'Gardner syndrome,' detailed evaluation has shown that a majority of FAP patients have one or more extracolonic features (Krush et al., 1988). In addition, Gardner syndrome and FAP may occur in sibships, and both disorders are associated with pathologic mutations in the APC gene. Thus, Gardner syndrome is best described as a variant of FAP (Nishisho et al., 1991).

Pierce et al. (1970) provided follow-up of a large Canadian kindred with FAP originally reported by Kelly and McKinnon (1961). Pierce et al. (1970) concluded that the kindred actually had Gardner syndrome, which they referred to as a 'triad' of colonic polyposis, soft tissue abnormalities such as dermoid and epidermal cysts and desmoid tumors, and hard tissue abnormalities like osteomas. Of 71 affected family members, 37 had polyposis only, 10 had only soft tissue abnormalities, and 1 had only bone abnormalities. Nineteen family members manifested 2 components, and 4 had the complete triad.

Butson (1983) reported a patient with FAP who had almost every recorded manifestation of the syndrome, including carcinomatous changes in the polyps, osteomas of facial and other bones, a periampullary carcinoma, transitional-cell carcinoma of the bladder, adrenal adenoma, and intraabdominal fibrous desmoid tumors with bowel obstruction.

Lower Gastrointestinal Tract

FAP is characterized by the development of hundreds of colorectal adenomas during adolescence. Colorectal cancer will develop in nearly all affected persons by the sixth decade of life if prophylactic colectomy is not performed (Giardiello et al., 2002).

Asman and Pierce (1970) reported a large kindred from Kentucky with familial multiple polyposis of the intestine. No extraintestinal features were found.

Shull and Fitts (1974) reported a family in which the father and 2 sons had both adenomatous and lymphoid polyps. Venkitachalam et al. (1978) pointed out that lymphoid polyposis had been reported several times in affected families.

Upper Gastrointestinal Tract

Schnur et al. (1973) reported the association of adenocarcinoma of the duodenum and Gardner syndrome. Erbe and Welch (1978) presented a patient with multiple polyps of the small bowel and 2 adenocarcinomas of the jejunum. Denzler et al. (1979) described 3 patients with FAP who also had adenomatous or hyperplastic polyps in the stomach and duodenum. The polyps were detected only by endoscopy or air-contrast radiographic examination. The findings suggested that gastric and duodenal polyps are more common in familial polyposis coli than previously recognized and should be considered an integral part of the syndrome.

Sugihara et al. (1982) reported a 48-year-old man with Gardner syndrome and rectal carcinoma who developed a well-differentiated adenocarcinoma of the duodenum. Histologic examination showed a large adenoma with focal carcinoma, 256 adenomas of the duodenum, and 91 adenomas of the gastric antrum. A review of the literature showed 29 cases of periampullary carcinoma and 12 cases of gastric carcinoma complicating FAP or Gardner syndrome.

Periampullary cancer is a well-recognized feature of FAP (Harned and Williams, 1982; Jones and Nance, 1977). The clustering of polyps around the ampulla of Vater implicates bile in the pathologic process (Pauli et al., 1980).

Burt et al. (1984) found that 6 of 11 patients of the original Utah kindred reported by Gardner (1951) had numerous small polyps of the gastric fundus and body. Another patient had a single antral adenoma. Eight patients exhibited small duodenal adenomas, and 6 had ileal adenomas. The results indicated that upper gastrointestinal polyps are a common pleiotropic manifestation of the genetic defect responsible for Gardner syndrome.

In a 26-year-old woman with Gardner syndrome, Walsh et al. (1987) found multifocal adenomatous change with severe dysplasia in the gallbladder. They referred to observations of others on bile duct cancer and carcinoma in situ of the gallbladder in patients with this form of hereditary polyposis.

Iida et al. (1988) reviewed the natural history of gastric adenomas in FAP. Thirteen of 26 FAP patients were found to have gastric adenomas; during a 6.8-year follow-up, 6 of the 13 patients developed additional gastric adenomas.

Offerhaus et al. (1992) commented on the fact that gastric cancer in Japan is more common than duodenal cancer in patients with FAP, and that gastric adenomas develop in 50% of Japanese patients with FAP. Jagelman et al. (1988) had observed that duodenal cancer was much more common than stomach cancer in Western APC gene carriers. Offerhaus et al. (1992) found that in the families in the Johns Hopkins Polyposis Registry, there was a greatly increased relative risk of duodenal adenocarcinoma and ampullary adenocarcinoma. No significant increased risk was found for gastric or nonduodenal small intestinal cancer.

Bapat et al. (1993) stated that 24 to 96% of FAP patients develop periampullary adenomas. The authors identified 2 distinct somatic mutations in the APC gene (611731.0019; 611731.0020) in 2 periampullary adenomas from an FAP patient. The findings were consistent with periampullary tumors being an extension of the same pathologic process.

Worthley et al. (2012) described 3 unrelated families segregating autosomal dominant proximal gastric polyposis, including a large 4-generation family from Australia and 2 smaller 3-generation families from the United States and Canada. In the Australian family, there were 20 affected family members, and fundic gland polyposis (FGP) were detected as early as 10 years of age. There were 2 family members with typical proximal polyposis of the stomach who died from intestinal-type gastric adenocarcinoma at ages 33 and 48 years; both developed liver metastases. Although all individuals represented in the 3 pedigrees were masked for sex, Worthley et al. (2012) stated that there was a 1:2.3 male:female ratio in the Australian family, with 4 instances of male-to-male transmission. In addition, 3 obligate carriers had normal upper gastrointestinal endoscopies at 68, 72, and 77 years of age, indicating incomplete penetrance. One affected individual underwent capsule endoscopy that excluded polyps of the small intestine. Colonoscopy results were available from 13 affected family members, none of whom had colorectal polyposis; the most advanced colorectal pathology involved 8 simple tubular adenomas removed over 4 colonoscopies, and there was no family history of colorectal cancer. In the other 2 families, 3 of 7 and 2 of 4 affected individuals, respectively, developed gastric cancer, the youngest at age 34 years. Colonoscopy was reported to be unremarkable in 3 affected family members from the 2 families. In the Canadian family, there were 4 family members with other cancers, including 1 case each of leukemia, brain, prostate, and lung cancer. Worthley et al. (2012) designated the phenotype 'gastric adenocarcinoma and proximal polyposis of the stomach (GAPPS).'

Congenital Hypertrophy of the Retinal Pigment Epithelium

Blair and Trempe (1980) observed that congenital hypertrophy of the retinal pigment epithelium (CHRPE) is a frequent finding in Gardner syndrome and can be a valuable clue to the presence of the gene in persons who have not yet developed other manifestations. The pigmented fundus lesion may be mistaken for malignant melanoma.

Lewis et al. (1984) described multiple and bilateral patches of CHRPE in affected members of 3 families with Gardner syndrome. Most CHRPE lesions were unilateral, solitary, nonfamilial, and not known to be associated with other ocular or systemic disorders. The patches were 1 or 2 disc diameters in size with a surrounding area of depigmentation, and have been referred to as 'pigmented scars.' The center of the lesion showed chorioretinal atrophy and the peripheral hyperpigmentation. In 4 other families, a total of 8 patients did not show CHRPE. Bull et al. (1985) also reported observations on CHRPE in the Gardner syndrome.

Traboulsi et al. (1987) examined 134 members of 16 families with Gardner syndrome for pigmented ocular fundus lesions. Of 41 patients with documented Gardner syndrome, 37 (90.2%) had such lesions. The lesions were bilateral in 32 of the patients and in 2 of 42 controls. Twenty (46.5%) of 43 first-degree relatives at 50% risk for Gardner syndrome had bilateral pigmented fundus lesions indicating that they probably had inherited the abnormal gene. The presence of bilateral lesions, multiple lesions (more than 4), or both appeared to be a specific (specificity = 0.952) and sensitive (sensitivity = 0.780) clinical marker for Gardner syndrome. Since the lesions were observed in a 3-month-old baby at risk, they were considered congenital.

Diaz-Llopis and Menezo (1988) suggested that CHRPE may be a useful marker to detect patients at risk for FAP. Combining eye examination for CHRPE with data on age of onset and linked DNA markers appeared to be highly effective in carrier exclusion. Lyons et al. (1988) concluded that the CHRPE phenotype is a more powerful marker than other phenotypic features of Gardner syndrome.

Baker et al. (1988) claimed that CHRPE is not as specific for Gardner syndrome compared to the presence of polyps. When ophthalmic examinations were performed on 56 at-risk patients, 8 patients were found to have the retinal lesions without any of the extracolonic features of Gardner syndrome. However, it was possible that the eye lesion may be the only extracolonic feature of Gardner syndrome.

Chapman et al. (1989) searched for CHRPE in 40 patients representing all 25 pedigrees with FAP identified in the northern region of the U.K. All had multiple lesions, ranging in number from 2 to more than 40. None of 35 controls had more than 2 lesions.

Houlston et al. (1992) suggested that CHRPE is not exclusively a manifestation of mutation at the APC locus. They described 3 patients who had 4 or more patches with no other extracolonic manifestations of FAP and all having fewer than 5 adenomatous polyps detected by colonoscopy. In the families of the 3 patients, a parent and the proband in each case had colorectal cancer. In 2 families, there was cancer of other types. Houlston et al. (1992) suggested that CHRPE can occur with cancer family syndromes. However, no search for mutations of the APC gene was made in these cases. Patients expressing CHRPE tend to cluster within specific polyposis families.

CHRPE is traditionally regarded as a benign stationary condition. However, in at least 5 cases, CHRPE has given rise to elevated solid tumors (Shields et al., 2000). Shields et al. (2001) reported the histopathology of a progressively enlarging peripheral fundus tumor that arose from a focus of classic CHRPE. After removal of the mass by local resection, histopathologic examination revealed a low-grade adenocarcinoma of the retinal pigment epithelium, apparently arising from CHRPE. The authors concluded that CHRPE should be observed periodically for the development of neoplasm.

Cutaneous and Skeletal Features

Fader et al. (1962) first reported dental anomalies in Gardner syndrome. These include impacted teeth, supernumerary teeth, congenitally missing teeth, and abnormally long and pointed roots on the posterior teeth (Carl and Herrera, 1987). Jarvinen et al. (1982) found dental anomalies in 18% of patients, but jaw osteomata were very frequent.

Hoffmann and Brooke (1970) described a family in which 6 persons in 3 generations had FAP and a mother and son had sarcoma of bone leading to death from metastases at 28 and 13 years of age, respectively. No evidence of polyposis was found in either but special studies including autopsies were not done.

Utsunomiya and Nakamura (1975) recorded jaw osteomata, which appear as radiopaque lesions without a translucent halo, in 95% of FAP patients, but interpretation of the orthopantomograms is difficult and limits this as a diagnostic investigation.

Greer et al. (1977) reported a patient with Gardner syndrome and chondrosarcoma of the hyoid bone.

Calin et al. (1999) described 2 unrelated patients with FAP with unusual extracolonic phenotypes, namely several abnormalities of mesodermal origin strongly resembling Marfan syndrome (MFS; 154700). One patient was a 28-year-old Romanian man who was unusually tall and thin, being 184 cm tall, compared to his father (165 cm tall), his mother (158 cm tall), and a brother and sister (168 and 161 cm tall, respectively). The patient's palate was narrow and high-arched with crowding of the teeth. There was moderate thoracic kyphoscoliosis, moderate hypermobility of all joints, and skin hyperextensibility. Moderate mental retardation was described. The second patient was a 38-year-old Romanian man who was 192 cm tall with arm span greater than height. An aortic diastolic murmur was heard. The diagnosis of FAP seemed well established in both patients; in the second patient the mother may have died at age 34 of FAP and a 36-year-old sister was found to have polyposis. Conventional cytogenetic and FISH analysis revealed no gross chromosomal rearrangement of 5q. In the second case, the FAP-causing mutation in the APC gene was found in the donor splice site of exon 4 and was shown to result in a frameshift and a premature termination codon. Calin et al. (1999) proposed that the connective tissue abnormalities resulted from germline APC mutations in combination with specific genetic and/or environmental modifying factors.

Desmoid Tumors

Simpson et al. (1964) reported the association of mesenteric fibromatosis in FAP and considered it to be a variant of Gardner syndrome. Mesenteric fibromatosis tended to develop after surgery. Also known as desmoid tumors, these slowly growing lesions were locally invasive and reached enormous proportions.

Fraumeni et al. (1968) described a family in which the father and a daughter had a malignant mesenchymal tumor, a son had polyposis coli, and another son had both polyposis coli and malignant mesenchymal tumor. The authors also suggested that it was a variant of FAP.

Klemmer et al. (1987) found an increased frequency of desmoids in patients with FAP. The crude frequency was about 6%, but the risk was dependent on age and sex. The lifetime risk was estimated to be 8% for males and 13% for females.

Clark et al. (1999) reviewed the occurrence of desmoid tumor in FAP patients ascertained through a polyposis registry. They identified 166 desmoids in 88 patients; 83 tumors (50%) were within the abdomen, and 80 (48%) were in the abdominal wall. All but 16 individuals (18%) had already undergone abdominal surgery. Intraabdominal desmoids caused small bowel and ureteric obstruction and resulted in 10 deaths; survival was significantly poorer than in patients with abdominal wall desmoid alone, and 8 of 22 patients who underwent resection of intraabdominal desmoid died in the perioperative period. Clark et al. (1999) concluded that abdominal wall desmoids caused no deaths or significant morbidity; although recurrence was common after excision, the treatment was safe. They concluded that intraabdominal desmoids can cause serious complications, and treatment is often unsuccessful; in particular, surgery for desmoids at this site is hazardous.

Hepatoblastoma

Heimann et al. (1987) described a male patient who presented at 25 months of age with precocious puberty and an abdominal mass that was found to be a virilizing hepatoblastoma. Shneider et al. (1992) reported that the patient remained disease-free for 53 months following liver transplantation, but was found to have multiple adenomatous polyps of the colon at age 8 years. There was a strong maternal family history of polyposis and colon cancer. Ophthalmologic examination revealed CHRPE. Total colectomy and ileoanal reconstruction was performed when he was 10 years of age.

Several groups noted the association of hepatoblastoma with polyposis coli (e.g., Kingston et al., 1982; Li et al., 1987; Krush et al., 1988). Li et al. (1987) observed hepatoblastoma in 4 unrelated children who had a family history of polyposis coli and found this association in 10 other kindreds in the literature. One child who survived hepatoblastoma showed multiple colonic adenomas at 7 years of age. She and 8 affected maternal relatives also had CHRPE. Krush et al. (1988) reported hepatoblastoma in 4 children from unrelated families. One child, 19 years old at the time of the report, survived after a resection of a hepatoblastoma in infancy and had recently been found to have Gardner syndrome. He, like many others in these 4 families, both affected and at risk, had osteomatous jaw lesions and pigmented ocular fundus lesions.

In a worldwide collaborative study, Garber et al. (1988) identified 11 children with hepatoblastoma and a family history of adenomatous polyposis; 14 additional instances of the association were collected from the literature. Among the 11 survivors of hepatoblastoma in the combined series, adenomatous lesions of the colon had been sought in 7 and detected in 6 patients at ages 7 to 25 years. Five of these patients also had CHRPE. Giardiello et al. (1991) studied the frequency of hepatoblastoma in the families registered in the familial polyposis registry maintained at Johns Hopkins since 1973. Seven members of these families had hepatoblastoma diagnosed at ages varying from 1 month to 4.5 years. Six of them were from Gardner syndrome families and 1 was from a polyposis family without extrabowel manifestations. Giardiello et al. (1991) calculated the relative risk of hepatoblastoma in persons with the APC gene from birth through age 4 as being 3.3 per 1,000 person/years.

In a retrospective review of their family history data, Hughes and Michels (1992) found that 2 of 470 (0.42%) children born to 241 patients with FAP had hepatoblastoma. This figure was significantly higher than the 1 in 100,000 incidence of hepatoblastoma in the general population. However, for genetic counseling purposes, an empiric risk of less than 1% for hepatoblastoma can be cited to persons with FAP for their children.

Brain Tumor-Polyposis Syndrome 2

Crail (1949) reported a 24-year-old man with adenomatous polyposis, colonic adenocarcinoma, brainstem medulloblastoma, and papillary adenocarcinoma of the thyroid.

Capps et al. (1968) described a family with 4 generations of polyposis and carcinoma of the colon. A brother of the proband died of brain tumor at age 9 years and had colonic polyposis. The proband, aged 14 years at first presentation, had carcinoma of the colon, ampulla of Vater, and urinary bladder.

Hamilton et al. (1995) identified APC mutations (see, e.g., 611731.0014 and 611731.0022) in 10 of 12 families with FAP in which at least 1 patient developed a central nervous system tumor, mainly medulloblastoma (79%), as an extracolonic manifestation of FAP. Since these index patients had both colonic polyposis and CNS tumors, they had originally been referred to as having Turcot syndrome (see 276300). However, Turcot syndrome is usually considered an autosomal recessive disorder resulting from biallelic mutations in mismatch repair (MMR) genes (see, e.g., MLH1, 120436); heterozygous mutations in MMR genes result in hereditary nonpolyposis colorectal cancer (HNPCC; see 120435). Hamilton et al. (1995) estimated that the relative risk of medulloblastoma in FAP patients was 92 times greater than that found in the general population. Several of the patients with APC mutations also had pigmented ocular fundus lesions, epidermal inclusion cysts, or osteosclerotic jaw lesions consistent with Gardner syndrome.

Paraf et al. (1997) proposed that Turcot syndrome, which they referred to as the 'brain tumor-polyposis (BTP) syndrome,' could be classified into 2 distinct entities. The authors referred to patients with mutations in mismatch repair genes as having 'BTP syndrome type 1' (BTPS1; 276300). Patients from FAP kindreds with germline APC mutations who develop CNS tumors were referred to as having 'BTP syndrome type 2' (BTPS2). Risk analysis showed an increased incidence of medulloblastoma in FAP patients. By contrast, APC mutations were not found in sporadic glioma or medulloblastoma.

In a review of reported FAP cases with medulloblastoma, Van Meir (1998) found that patients with medulloblastoma who also expressed the colorectal phenotype developed disease after age 17 years, whereas family members who did not express the colorectal phenotype had an age of brain tumor occurrence of less than 10 years. However, the authors noted that the young age of these patients may explain the absence of the colonic phenotype, which may have occurred at a later age. In a discussion of mechanism of inheritance, Van Meir (1998) suggested that the rarity of medulloblastoma in patients with FAP suggests the involvement of a second locus with a modifier gene or of environmental factors.

Endocrine Carcinoma

Camiel et al. (1968) described thyroid carcinoma in 2 sisters with Gardner syndrome, which was probably present in at least 3 generations of the family. Smith (1968) also described patients with the association of colonic polyps and papillary carcinoma of the thyroid. Herve et al. (1995) reported a case of papillary carcinoma in a 16-year-old girl with Gardner syndrome. They reviewed the literature and estimated that the incidence of thyroid carcinoma in patients with Gardner syndrome approached 100 times that of the general population. Cameselle-Teijeiro and Chan (1999) and Tomoda et al. (2004) noted that the papillary thyroid carcinoma most frequently associated with FAP is the distinctive cribriform-morular variant.

Marshall et al. (1967) described a case of Gardner syndrome with adrenal cortical carcinoma with Cushing syndrome.

In a member of the original Utah kindred with Gardner syndrome, Naylor and Gardner (1981) observed bilateral adrenal adenomas. They found reports of 6 cases of adrenal adenoma and 1 of primary adrenal carcinoma. They also reviewed 15 reported cases of thyroid tumors in Gardner syndrome. Bell and Mazzaferri (1993) reported what they alleged to represent the 37th report of the association of Gardner syndrome with papillary thyroid carcinoma. They pointed out that 94.3% of the patients have been women.

Chung et al. (2006) described a 19-year-old woman with the cribriform-morular variant of papillary thyroid carcinoma, which had been discovered 8 months before the discovery of polyposis of the colon, in whom they identified a de novo R302X mutation (611731.0006). The authors noted that a hereditary colonic syndrome can be associated initially with an extracolonic tumor.

Attenuated FAP

Hodgson et al. (1994) suggested that heterozygous deletion of the entire APC gene may be associated with a form of FAP characterized by more proximal distribution of adenomas than usual, of which some are sessile and some may be nonpolypoid or flat. They postulated that in the usual type of FAP where the mutation results in a truncated protein, this protein may interfere with the function of the protein product of the normal allele to cause a more severe disease than seen in their patients. They pointed to the large kindreds reported by Leppert et al. (1990) and Lynch et al. (1992) as possible examples of this particular phenotype. Samowitz et al. (1995) pointed out that this seemingly different phenotype was referred to by Lynch et al. (1992) as 'hereditary flat adenoma syndrome.' Later, when it was found that the family reported by Leppert et al. (1990) and the families of Lynch et al. (1992) had characteristic mutations in the 5-prime end of the APC gene, the syndrome was renamed 'attenuated adenomatous polyposis coli' (AAPC).

Attenuated adenomatous polyposis coli is characterized by the occurrence of less than 100 colonic adenomas and a later onset of colorectal cancer (age greater than 40 years) (Soravia et al., 1998).

Evans et al. (1993) reported families with an attenuated form of FAP. In 1 family, a 59-year-old patient showed no abnormality; late onset of polyps was discovered by endoscopy and biopsy in other members of that family and in 2 other families. Mutation analysis in these families was not reported.

Matsumoto et al. (2002) explored the possible association between serrated adenomas and FAP. Detailed colonoscopy and biopsy was undertaken in 11 individuals from 8 FAP families who had not undergone prophylactic colectomy. Serrated adenomas were detected in 3 individuals. Overall macroscopic polyp counts were less than 100 in these individuals. APC mutations were found in codons 161, 332, and 1556. These observations suggested that serrated adenomas may be an important feature of the attenuated form of FAP.

Diagnosis

Petersen et al. (1989) demonstrated how one could use linkage information to modify the genetic counseling recommendations for FAP. In the family of an affected 36-year-old man with a positive family history of FAP, there were 4 asymptomatic children under the age of 10 years. Before linkage analysis, all children had a 50% risk. The linkage information allowed a counselor to state to the family with 98% confidence that 3 of the children did not inherit the gene and that 1 child did. That child could be screened annually; the others could have screening every 3 years beginning at ages 12 or 13 and continuing until age 35.

Tops et al. (1989) identified 2 linked polymorphic DNA markers on either side of the FAP locus. They estimated that use of these markers could allow prenatal and presymptomatic diagnosis with more than 99.9% reliability in most families. Dunlop et al. (1990) described 6 DNA markers flanking the APC gene that were useful for presymptomatic diagnosis. Dunlop et al. (1991) performed presymptomatic analysis of DNA from 41 individuals at risk for FAP. Of these, 28 individuals were informative, and 14 whose probe-derived risk was greater than 0.93 were subsequently demonstrated to be affected by clinical screening. The authors suggested that an integrated risk analysis, including genotypic, colonic, and ophthalmologic evaluation for the presence of CHRPE, should be used in FAP screening programs. Cachon-Gonzalez et al. (1991) concluded on the basis of linkage studies using 4 DNA probes that presymptomatic diagnosis could be given with only 90% probability based on DNA typing alone.

Morton et al. (1992) demonstrated that DNA extracted from preserved tissue of dead relatives could be used to extend informativeness in FAP families.

Petersen et al. (1993) demonstrated the feasibility of presymptomatic direct detection of APC mutations in each of 4 families. Maher et al. (1993) concluded that intragenic and closely linked DNA markers were informative in most families at risk for FAP and that the reduction in screening for low-risk relatives rendered molecular genetic diagnosis a cost-effective procedure. In their population-based study, they estimated a minimum heterozygote prevalence of 1/26,000. Of 33 probands, 8 (24%) represented new mutations. Interfamilial variation in CHRPE expression was evident, with ophthalmologic assessment showing more than 3 CHRPEs in 27 of 43 (63%) affected patients and high-risk relatives, and none of 18 low-risk relatives.

Powell et al. (1993) developed a method based on the examination of APC proteins synthesized in vitro and study of endogenous APC transcripts, since most mutations in patients with FAP result in truncation of the APC gene product. In 62 unrelated patients from the Johns Hopkins Familial Adenomatous Polyposis Registry, primary screening identified a truncated protein in 51 of the 62 patients (82%). In 3 of the 11 remaining patients, allele-specific expression assay demonstrated significantly reduced expression of one allele of the APC gene. Use of the 2 assays in combination successfully identified germline APC mutations in 87% of the 62 patients. A so-called 'protein truncation test,' based on the in vitro transcription and translation of genomic PCR products, was developed also by van der Luijt et al. (1994).

Papadopoulos et al. (1995) reported the development of a sensitive and specific diagnostic strategy based on somatic cell hybridization termed monoallelic mutation analysis (MAMA). This simple and ingenious method involves the use of hamster/human somatic cell hybrids, which could be expected in many cases to have only 1 of the 2 alleles present. To show that single alleles were isolated in the clones, microsatellite markers proximal and distal to the gene of interest were assessed. Papadopoulos et al. (1995) demonstrated the utility of this strategy in FAP and in hereditary nonpolyposis cancer.

Thakker et al. (1995) presented a weighted scoring system for changes on dental panoramic radiographs, called the Dental Panoramic Radiographs Score (DPRS), as a diagnostic tool for FAP. The score took into account the nature, extent, and sight of osseous and dental changes, as well as the incidence of the anomaly in the general population. Using the highest threshold, a specificity of 100% and sensitivity of approximately 68% were obtained. If all positive findings were considered as significant, sensitivity was increased to approximately 82%, but the specificity was reduced to approximately 88%. Overall, approximately 68% of the affected subjects had significant changes, and approximately 18% had normal appearance on DPR, with the remainder having changes classified as minimal or equivocal.

The use of commercially available tests for genes linked to familial cancer is a source of concern about the possible adverse impact on patients. Giardiello et al. (1997) assessed indications for APC gene testing, through telephone interviews with physicians and genetic counselors in a nationwide sample of 177 patients from 125 families who underwent testing during 1995. Of the 177 patients tested, 83% had clinical features of FAP or were at risk for the disease. Only 18.6% (33 of 177) received genetic counseling before the tests, and only 16.9% (28 of 166) provided written informed consent. In 31.6% of the cases, the physicians misinterpreted the test results. Among the patients with unconventional indications for testing, the rate of positive results was only 2.3% (1 of 44). Giardiello et al. (1997) concluded that physicians should be prepared to offer genetic counseling if they order genetic tests.

Deuter and Muller (1998) described a highly sensitive and nonradioactive heteroduplex-PCR method (HD-PCR) for detecting APC mutations in stool DNA. Traverso et al. (2002) purified DNA from routinely collected stool samples and screened for APC mutations by a novel approach called digital protein truncation. Stool samples from 28 patients with nonmetastatic colorectal cancers, 18 patients with adenomas that were at least 1 cm in diameter, and 28 control patients without neoplastic disease were studied. APC mutations were identified in 26 of the 46 patients with neoplasia and in none of the 28 control patients. The authors emphasized, however, that their study had not established that the digital protein truncation test is a clinically useful screening procedure.

Clinical Management

Waddell and Loughry (1983) were the first to make a connection between nonsteroidal antiinflammatory drugs (NSAIDs) and colon cancer. The authors observed the disappearance of rectal polyps in a patient with Gardner syndrome and correctly attributed this disappearance to treatment with sulindac, an NSAID that was given for unrelated reasons.

In a random, double-blind, placebo-controlled study of 22 FAP patients, including 18 who had not undergone colectomy, Giardiello et al. (1993) found that oral sulindac reduced the number and size of colorectal adenomas. The effect was incomplete, however, leading the authors to conclude that it is unlikely to replace colectomy as primary therapy. Giovannucci et al. (1994) reported that patients who take aspirin or other NSAIDs on a regular basis have a 40 to 50% lower relative risk of colorectal cancer when compared with persons not regularly taking these medications. This effect has been confirmed in numerous clinical trials (Marcus, 1995), and the interpretation that NSAIDs promote regression of colon polyps by inhibiting prostaglandin synthesis has been supported by genetic studies in mice (Oshima et al., 1996).

Schnitzler et al. (1996) demonstrated that sulindac inhibited the growth of carcinoma cells in vitro and caused an increase in APC mRNA. They suggested that the effect of these agents on colonic carcinogenesis was not mediated entirely via inhibition of prostaglandin biosynthesis. Herrmann et al. (1998) found that sulindac sulfide, the metabolite of sulindac, strongly inhibited Ras-induced malignant transformation. Sulindac sulfide decreased the Ras-induced activation of its main effector, the c-raf-1 kinase.

Maule (1994) found that nurses could carry out screening by flexible sigmoidoscopy as accurately and safely as experienced gastroenterologists. However, Kassirer (1994), while accepting the usefulness of many more nurse practitioners in collaborative practice arrangements, doubted that their function as entirely independent practitioners of primary care would be either cost-effective or safe.

Steinbach et al. (2000) studied the effect of celecoxib, a selective COX2 inhibitor, on colorectal polyps in patients with FAP. In a double-blind, placebo-controlled study of 77 patients, they found that 6 months of twice-daily treatment with 400 mg of the agent significantly reduced the number of colorectal polyps.

Using immunohistochemistry with activation-specific antibodies, Hardwick et al. (2001) demonstrated expression of COX2 and NFKB (164011) in stromal macrophages in human colonic adenomas. In addition, active JNK (601158) was expressed in stromal and intraepithelial T lymphocytes and periendothelial cells of new blood vessels. Active p38 (MAPK14; 600289) was most highly expressed in stromal macrophages. Hardwick et al. (2001) concluded that active inflammatory signal transduction occurs predominantly in the stroma of colonic polyps. They suggested that NSAIDs may exert their chemopreventive effects in reducing colonic polyp size through effects on stromal rather than epithelial cells.

Giardiello et al. (2002) reported that standard doses of sulindac did not prevent the development of adenomas in FAP patients. During 4 years of treatment, adenomas developed in 9 of 21 subjects (43%) in the sulindac group and in 11 of 20 subjects in the placebo group (55%). There was no significant difference in the mean number or size of polyps between the groups. Prophylactic colectomy remained the treatment of choice to prevent colorectal cancer in patients with this disorder. Despite the relatively disappointing results with respect to the ability of inhibition of cyclooxygenase to prevent adenomatous polyps in patients with FAP, Chau and Cunningham (2002) suggested that NSAIDs and cyclooxygenase-2 inhibitors may be shown to have a role in the primary prevention or treatment of established colorectal cancer.

Martinez et al. (2003) studied the effect of an intronic polymorphism in the ornithine decarboxylase gene (ODC1; 165640.0001) on the risk of recurrence of colorectal adenoma. They concluded that the ODC polymorphism and aspirin act independently to reduce the risk of adenoma recurrence by suppressing synthesis and activating catabolism, respectively, of colonic mucosal polyamines. Individuals homozygous for the minor OCD A allele who reported using aspirin were only one-tenth as likely to have an adenoma recurrence as non-aspirin users homozygous for the major G allele.

Cytogenetics

Herrera et al. (1986) found a constitutional interstitial deletion of 5q15-q22 in a man with possible Gardner syndrome; the large bowel was 'carpeted' with more than 100 adenomatous polyps and contained a well-differentiated carcinoma of the rectum, a similar carcinoma of the ascending colon, and melanosis coli. A small mesenteric neurofibroma was also found. In addition, the patient had severe mental retardation, horseshoe kidney, absence of the left lobe of the liver, and agenesis of the gallbladder. This finding of a deletion prompted the search for linkage with RFLP markers on 5q; positive results indicated that the mutation determining Gardner syndrome is located on 5q, probably near bands 5q21-q22. Kobayashi et al. (1991) described an interstitial deletion of 5q in a boy with Gardner syndrome who had mental retardation and multiple minor anomalies. The deletion involved q22.1-q31.1.

Loss of constitutional heterozygosity (LOH) for markers on chromosome 5 is a frequent finding in colon cancers. Solomon et al. (1987) demonstrated that at least 20% of sporadic colorectal adenocarcinomas lose one of the alleles on chromosome 5q that is present in normal tissue of the host. Ashton-Rickardt et al. (1989) found that more than 50% of a large series of colorectal carcinomas had lost an allele near APC, suggesting that loss of the APC locus contributes to the malignant process.

Hockey et al. (1989) described an interstitial deletion of 5q15-q22 in 2 intellectually handicapped brothers with familial adenomatous polyposis. Their retarded mother died of an inoperable carcinoma of the colon with extensive polyposis.

Using probes linked to APC or chromosome 5, Okamoto et al. (1990) found a high incidence of allelic deletions among 51 colorectal tumors and 7 desmoids from 19 cases of FPC and 5 cases of Gardner syndrome, as well as 15 sporadic colon cancers. APC loss resulted primarily from interstitial deletion or mitotic recombination. Combined tumor and pedigree analysis in a Gardner syndrome family showed loss of normal 5q alleles in 3 tumors, including a desmoid tumor, which suggested the involvement of hemizygosity or homozygosity of the defective APC gene in colon carcinogenesis and possibly in extracolonic neoplasms.

Cross et al. (1992) reported a male patient and his maternal aunt with Gardner syndrome and mental handicap who had an interstitial deletion of 5q22-q23.2, deleting the APC gene. Two other normal family members had the underlying direct insertion of of chromosome 5 (dir ins(5)(q31.3q22q23.2)). Cross et al. (1992) suggested that familial direct insertions should be considered as a cause of recurrent microdeletion syndromes.

Hodgson et al. (1993) reported 2 cases of deletion of 5q associated with FAP. They noted that 5 such cases had previously been reported. In their 2 cases, as well as in one of those previously reported, macroscopic polyposis was confined to the proximal colon in patients in their thirties, although microscopic adenomatosis was shown in the more distal colon with occasional single polyps. Both of their subjects had dermoid cysts, and CHRPE was seen in one. The latter patient, who had the more extensive deletion, also showed Caroli disease (dilatation of distal intrahepatic bile ducts with intrahepatic stone formation); see 263200.

Barber et al. (1994) described an institutionalized adult woman who was referred for chromosome studies because of autistic behavior. A high risk of colorectal cancer was predicted when an interstitial deletion of 5q was found in lymphocytes and deletion of the MCC (159350) and APC genes confirmed by molecular analysis. Adenomatous polyposis coli and carcinoma of the rectum were subsequently diagnosed in the patient. The del(5)(q15q22.3) arose as a result of recombination within the small insertion loop formed at meiosis by the direct insertion found in the patient's mother.

In a Dutch FAP family, van der Luijt et al. (1995) detected a germline rearrangement in the form of a constitutional reciprocal translocation t(5;10)(q22;q25), resulting in the disruption of the APC gene and an apparently null allele. The patients exhibited atypical clinical features, namely a slightly delayed age of onset of colorectal cancer and a reduced number of colorectal polyps that were mainly sessile and located predominantly in the proximal colon. This was thought to be the first description of a reciprocal translocation disrupting the APC gene.

De Chadarevian et al. (2002) identified a constitutional inversion, inv(5)(q22-q31.3), associated with FAP in 3 generations of a Mexican family. The proband was a 16-month-old male with an 8-month history of liver masses, biopsies of which had been diagnosed as multiple adenomas. Liver transplantation was performed. His brother had died at almost 2 years of age following the resection of a tumor diagnosed as hepatoblastoma. Family history included a paternal grandfather who died of colon cancer. The child, his father, and a paternal uncle were found to carry the same constitutional inversion. Colonoscopy in the father demonstrated extensive colonic polyposis. Molecular analysis failed to demonstrate a truncated APC protein or an APC mutation, suggesting that the phenotype in this family may be the result of a position effect.

Mapping

By analysis of families with FAP, Bodmer et al. (1987) found linkage to marker C11p11 on chromosome 5q (maximum lod score of 3.26). The majority of the families studied were instances of Gardner syndrome, with extracolonic lesions such as epidermoid cysts, jaw osteomata, and fibrous desmoid tumors. The findings of Bodmer et al. (1987) suggested that Gardner syndrome and familial colorectal cancer are allelic disorders. In an accompanying paper, Solomon et al. (1987) demonstrated that at least 20% of sporadic colorectal adenocarcinomas lose one of the alleles on chromosome 5q that is present in normal tissue of the host. The findings suggested that a locus on 5q is critical for the progression of colorectal cancer.

Simultaneously and independently, Leppert et al. (1987) demonstrated linkage of FAP in 5 families to a marker in the region of 5q22 (maximum lod score of 5.0). In 2 of the families, affected individuals had lesions typical of Gardner syndrome; in 3 families, it was typical of familial polyposis coli. Although most of the linkage information was provided by the families with the FAP phenotype, the findings indicated that Gardner syndrome and FAP are allelic disorders.

Through studies in 6 families, Nakamura et al. (1988) refined the genetic localization of the polyposis locus to a position about 17 cM distal to the DNA probe C11p11 at 5q21-q22. Three of the families had the FAP phenotype and 3 had Gardner syndrome.

In Dutch FAP kindreds, Meera Khan et al. (1988) found that the RFLP used by Bodmer et al. (1987) and Leppert et al. (1987) was minimally informative because