American Journal of Gastroenterology ISSN 0002-9270


C 2006 by Am. Coll. of Gastroenterology doi: 10.1111/j.1572-0241.2006.00375.x
Published by Blackwell Publishing

CME

Familial Adenomatous Polyposis

Polymnia Galiatsatos, M.D., F.R.C.P.(C),1 and William D. Foulkes, M.B., Ph.D.2
1
Division of Gastroenterology, Department of Medicine, The Sir Mortimer B. Davis Jewish General Hospital,
McGill University, Montreal, Quebec, Canada, and 2 Program in Cancer Genetics, Departments of Oncology
and Human Genetics, McGill University, Montreal, Quebec, Canada

Familial adenomatous polyposis (FAP) is an autosomal-dominant colorectal cancer syndrome, caused by a

germline mutation in the adenomatous polyposis coli (APC) gene, on chromosome 5q21. It is characterized by
hundreds of adenomatous colorectal polyps, with an almost inevitable progression to colorectal cancer at an
average age of 35 to 40 yr. Associated features include upper gastrointestinal tract polyps, congenital hypertrophy
of the retinal pigment epithelium, desmoid tumors, and other extracolonic malignancies. Gardner syndrome is
more of a historical subdivision of FAP, characterized by osteomas, dental anomalies, epidermal cysts, and soft
tissue tumors. Other specified variants include Turcot syndrome (associated with central nervous system
malignancies) and hereditary desmoid disease. Several genotypephenotype correlations have been observed.
Attenuated FAP is a phenotypically distinct entity, presenting with fewer than 100 adenomas. Multiple colorectal
adenomas can also be caused by mutations in the human MutY homologue (MYH) gene, in an autosomal
recessive condition referred to as MYH associated polyposis (MAP). Endoscopic screening of FAP probands and
relatives is advocated as early as the ages of 1012 yr, with the objective of reducing the occurrence of colorectal
cancer. Colectomy remains the optimal prophylactic treatment, while the choice of procedure (subtotal vs
proctocolectomy) is still controversial. Along with identifying better chemopreventive agents, optimizing screening
of extracolonic cancers and applying new radiological and endoscopic technology to the diagnosis and
management of extracolonic features are the major challenges for the future.
(Am J Gastroenterol 2006;101:385398)

INTRODUCTION tion in the adenomatous polyposis coli (APC) gene, located

Familial adenomatous polyposis (FAP) is an inherited col- on chromosome 5q21. APC is a tumor suppressor gene, first
orectal cancer syndrome characterized by the early onset localized in 1987, and cloned in 1991 following mutation
of hundreds to thousands of adenomas throughout the large analyses in unrelated families with FAP (35). It has an 8,538
bowel. Left untreated, there is a nearly 100% progression to bp open reading frame, and consists of 15 transcribed exons
colorectal cancer (CRC) by the age of 3540 yr (1, 2), as well (68) (Fig. 1). This gene is expressed in a variety of fetal
as a heightened risk of various other malignancies. CRC can and adult tissues, including mammary and colorectal epithe-
be prevented by the timely implementation of rigid screening lium (6). It encodes for a 312-kDa protein, 2,843 amino acids
programs, and certain medico-surgical interventions. long (6).
Inactivation of the APC gene product constitutes the initial
METHODS step in the development of CRC in FAP. APCs major func-
tion is that of a scaffolding protein, affecting cell adhesion
A literature search was performed from 1980 to August and migration. It is part of a protein complex, modulated by
2005, using the computerized PubMed database, looking the Wnt signaling pathway, which regulates the phosphory-
for English publications regarding familial adenomatous lation and degradation of -catenin (9, 10). -catenin is an
polyposis, attenuated familial adenomatous polyposis, intracellular protein that binds to the cell adhesion molecule
and MYH associated polyposis. Original articles and case E-cadherin and links E-cadherin to the actin cytoskeleton (6).
reports discussing genetics, clinical features, genotype The phosphorylation of -catenin attracts ubiquitin ligases,
phenotype correlations, screening, and prophylaxis were re- leading to its destruction at the proteasome (9, 10). When
viewed. Relevant articles (before or after 1980) were also ext- APC is mutated, -catenin accumulates in the cytoplasm and
racted manually from the references of retrieved publications. binds to the Tcf family of transcription factors, altering the
GENETICS expression of various genes affecting the proliferation, dif-
ferentiation, migration, and apoptosis of cells, namely those
Familial adenomatous polyposis (FAP) is a highly penetrant encoding cyclin D1, the proto-oncogene c-myc, the metal-
autosomal-dominant disorder, caused by a germline muta- loproteinase matrilysin, as well as ephrins and capsases (9,
To access a continuing medical education exam for this article, please visit www.acg.
10). APC also plays a role in controlling the cell cycle, by
gi.org/journalcme. inhibiting the progression of cells from the G0 /G1 to the S

385
386 Galiatsatos and Foulkes

Figure 1. APC cDNA (below) and important protein motifs (above). Adapted from Van Es et al. (6), Fearnhead et al. (7), and Foulkes (8).

phase, helping to suppress tumorigenesis (10). Furthermore, match repair genes that have been difficult to detect by other
APC stabilizes microtubules, thus promoting chromosomal means (15). Multiplex ligation-dependent probe amplifica-
stability (9). Inactivation of APC can lead to defects in mitotic tion (MLPA) is yet another test that quantifies all APC exon
spindles and chromosomal missegregation, with the resulting copy numbers. It has been useful in identifying a deletion of
aneuploidy leading to cancer (9). the entire APC gene in a patient with classical FAP, as well
Over 700 different disease-causing APC mutations have as large deletions involving several exons of the gene, often
been reported to date, but the most common germline mu- missed by conventional tests (16).
tation involves the introduction of a premature stop codon, The APC missense polymorphism I1307K, resulting from
either by a nonsense mutation (30%), frameshift mutation a T-to-A transversion, has also been indirectly linked to col-
(68%), or large deletion (2%), leading to truncation of the orectal adenoma and carcinoma, by rendering the gene sus-
protein product in the C-terminal region (11). Most of these ceptible to somatic mutations. This variant allele has been
germline mutations are clustered at the 5 end of exon 15, identified in 6% of Ashkenazi Jewish controls and 10% of
otherwise referred to as the mutation cluster region (12). The Ashkenazi CRC patients (17). Apart from a modest increase
two most frequently mutated codons are at positions 1061 and in risk of CRC, I1307K has also been linked to a heightened
1309 (11). Correlations have been observed between sites of risk of breast cancer among Ashkenazi Jews (18), although
mutations and variations in the phenotype, as will be dis- the latter association remains controversial. Another gene that
cussed later. An updated database of APC gene mutations is is possibly linked with APC mutation-negative FAP in Ashke-
available online at http://www.cancer-genetics.org. nazi Jewish families is CRAC1 (colorectal adenoma and car-
Despite genetic testing, 2030% of classical FAP patients cinoma gene), located on chromosome 15q14-q22 (19). This
have no detectable APC germline mutation by routine screen- gene has also recently been implicated in the hereditary mixed
ing methods. One possibility is the presence of a nontruncat- polyposis syndrome in a large Ashkenazi family, character-
ing missense mutation that could be missed by protein trun- ized by the development of various different colorectal tu-
cation testing (PTT), previously used as a first-line screening mors (including juvenile, hyperplastic, adenomatous polyps,
tool. Heinimann et al. found that 12.9% of APC mutation as well as CRC). It is inherited as an autosomal-dominant
carriers are missed by standard PTT (13). Using SNP as- trait (20).
say and direct DNA sequencing, they identified four possibly Although microsatellite instability (MSI) is found in about
pathogenic germline missense mutations: R99W and E1317Q 6% of tissue specimens from APC mutation-negative polypo-
in the coding region, A290T within the APC promoter, and sis patients, the inconsistency of MSI in different tumors from
A8822G in the 3 UTR end of the gene (13). However, valida- the same patient suggests a somatic inactivation of hMLH1
tion of these candidate disease-associated mutations was not by promoter hypermethylation, rather than a germline muta-
performed. Nowadays, PTT has been largely replaced as the tion (13). Hence, mismatch repair (MMR) gene deficiency is
first-line genetic test by other mutation-finding techniques, unlikely to be a cause of APC mutation-negative polyposis.
particularly DNA sequencing (currently the standard screen- Recently, biallelic mutations in MYH have been deemed re-
ing tool in most North American centers), and newer diag- sponsible for up to 7.5% of APC-negative classic FAP (21),
nostic methods. Monoallelic mutation analysis (MAMA) is a although MYH mutations are more frequently associated with
more sensitive second-line technique, whereby the two APC a milder phenotype, as will be discussed later.
alleles can be examined independently. Using this method,
Laken et al. demonstrated significantly reduced expression
in one APC allele 7 of 9 FAP patients with no identified trun-
EPIDEMIOLOGY
cating APC mutation (14). The authors concluded that APC
mutations can be identified in >95% of FAP patients when The birth frequency of FAP in Northern European popula-
MAMA is combined with standard genetic testing (14). A tions is estimated at roughly 1 in 13,000 to 1 in 18,000 live
modified version of this technique is now called conversion births (1, 22), and is responsible for less than 1% of all CRC
analysis, and has also been used to identify mutations in mis- cases (23). In the Swedish Polyposis Registry, the median age
 Familial Adenomatous Polyposis 387

of probands with CRC at diagnosis was 42 yr, compared to 2661% (27, 3336), compared with a 0.81.9% incidence
34 yr for those without CRC, while asymptomatic relatives in the general population (37, 38). FGPs are indeed the most
were diagnosed at a median age of 22 yr (22). In the Poly- common type of gastric polyp to occur in FAP patients. In
posis Registry of Japan, the mean age of diagnosis of FAP contrast to sporadic FGPs, FAP-related FGPs are more nu-
in patients without CRC was 28 yr, compared to 33 yr for merous, tend to occur at a younger age, with more equal
those with early cancer (in situ or submucosal) and 40 yr for gender distribution (39). Helicobacter pylori and its associ-
advanced cancer (24). The cumulative risk of CRC exceeded ated atrophic gastritis seem to have the same protective effect
50% by the age of 42 yr in women, and 44 yr in men (24). In against the development of FAP-related FGPs as they do in
the Danish and Finnish registries, 6669% of symptomatic sporadic FGPs (40). Although FGPs in the general popula-
probands, versus only 27% of call-up patients (relatives re- tion are typically benign lesions, up to 25% of those in FAP
cruited from pedigrees) had CRC at the time of diagnosis of patients show foveolar dysplasia (41), and cases of gastric car-
FAP (25, 26). cinoma associated with diffuse FGP have also been reported
(4245). In a study of 41 FGPs from 17 FAP patients, 51%
of polyps demonstrated an inactivating somatic APC gene
CLINICAL PRESENTATION alteration, whereas there were no such APC gene mutations
in 13 sporadic FGPs used for comparison (46). There was
Lower Gastrointestinal Tract Polyps and Cancer no significant difference in mutation rate among FGPs with
The hallmark of FAP is the development of hundreds of ade- or without dysplasia (46). These second-hit somatic APC
nomatous polyps in the colon and rectum usually in adoles- gene alterations, superimposed on germline APC mutations,
cence, with an almost inevitable progression to CRC by the could account for the neoplastic potential of FGPs in FAP,
age of 3540 yr, significantly younger than sporadic cancers. which appear to be not only clinically, but also pathologi-
A total of 7080% of tumors occur on the left side of the cally distinct from sporadic lesions. In fact, sporadic FGPs
colon (22). have recently been shown to harbor a high frequency of so-
matic mutations in exon 3 of the -catenin gene, not identi-
Upper Gastrointestinal Tract Polyps and Cancer fied in FAP-associated FGPs (47), reinforcing the difference
Upper gastrointestinal polyps (gastric and duodenal adeno- in initial mutational events (despite similarity in the altered
mas) are present in nearly 90% of FAP patients by the age pathways) causing sporadic versus syndromic FGPs.
of 70 yr, with a median age of diagnosis of 38 yr, based on
a prospective study from Nordic and Dutch polyposis reg- Congenital Hypertrophy of the Retinal Pigment
istries (27). Interestingly, 12% of duodenal polyps discov- Epithelium
ered during the initial upper endoscopies in this study were Congenital hypertrophy of the retinal pigment epithelium
microadenomas, diagnosed from random biopsies without (CHRPE) refers to the presence of characteristic pigmented
visible lesions (27). Roughly two-thirds of duodenal adeno- fundus lesions that are thought to occur in roughly 7080% of
mas occur in the papilla or periampullary region (28). Ad- patients with FAP (4850). These ophthalmic manifestations
vanced duodenal adenomas confer an increased risk of small are usually present at birth, largely preceding the develop-
bowel cancer, which is the third leading cause of death in ment of intestinal polyposis, and are asymptomatic with no
FAP patients (8.2%), apart from metastatic CRC (58.2%) and malignant potential. They are specific to FAP, as opposed to
desmoid tumors (10.9%) (29). The cumulative risk of devel- other hereditary or sporadic colonic cancers (51, 52). The
oping advanced (Spigelmans stage IV) duodenal polyposis is diagnostic criteria with the highest specificity/sensitivity for
estimated to be 43% at the age of 60 yr and 50% at the age of CHRPE include the detection of four small pigmented le-
70 yr, using side-viewing and forward upper endoscopy, with sions, or two lesions of which one is large (>25% of disc
systematic biopsies of the duodenal papilla (30). The mean surface), using bilateral lens fundoscopic examination (53).
age of duodenal cancer diagnosis ranges between 47 and 51 The presence of multiple bilateral lesions appears to be a
yr according to Dutch and Danish polyposis registries, with highly specific marker for FAP (95100% specificity) (54).
a cumulative risk of 34% by the age of 70 yr (31). In the CHRPE positivity has been associated with increased severity
prospective Nordic study mentioned above, the cumulative of FAP in probands (namely earlier age of polyposis develop-
incidence rate was as high as 4.5% at the age of 57 yr (27). ment and upper gastrointestinal involvement), and correlates
One retrospective study of 180 Swedish FAP patients found with DNA test positivity in undiagnosed kindred belonging
an unusually elevated cumulative risk of periampullary ade- to FAP families that are CHRPE-positive (48). This makes
nocarcinoma of 10% by the age of 60 yr (32). This stresses ophthalmological examination an attractive noninvasive and
the importance of routine upper endoscopic surveillance, and early diagnostic test for at-risk family members, aside from
the added value of random biopsies even in the absence of genetic analysis. CHRPE lesions can also help predict the
visible lesions. mutation site, since they are restricted to a specific mutation
FAP patients are also at an increased risk for fundic gland subgroup along the APC gene (55) (see section genotype
polyps (FGPs) in the stomach, with an estimated incidence of phenotype correlations for more details).
388 Galiatsatos and Foulkes

Desmoid Tumors Table 1. Extracolonic Cancer Risks in FAP

Desmoid tumors are rare, locally invasive fibromatoses that Relative Absolute Lifetime
are a major cause of morbidity, and the second leading cause Malignancy Risk Risk (%)
of death in FAP patients (29). They occur rarely in the general Desmoid 852.0 15.0
population, where they have been linked to somatic mutations Duodenum 330.8 3.05.0
of -catenin (56), or associated with estrogens. The overall Thyroid 7.6 2.0
prevalence of desmoid disease in FAP is 15% (57), with a rel- Brain 7.0 2.0
ative risk of 850 times that of the general population (58). Ampullary 123.7 1.7
Pancreas 4.5 1.7
The majority of tumors occurs within the abdomen (50%), Hepatoblastoma 847.0 1.6
usually involving small bowel mesentery, or in the abdomi- Gastric 0.6
nal wall (48%), with few arising in the trunk or limbs (59).
Note. Adapted from Giardiello et al. (78), Jagelman et al. (76), Sturt et al. (57), Lynch
Most tumors are solitary (58%), although the number of tu- et al. (58), Bulow et al. (27).

mors per individual can range from 1 to 10 (59). Although The Leeds Castle Polyposis Group.

most abdominal wall tumors are asymptomatic, intraabdom-

inal tumors can cause abdominal pain, or can be complicated
by bowel obstruction or perforation, ureteric obstruction, in-
testinal hemorrhage, even enterocutaneous fistula (59). The toma among children of FAP patients is 1 in 235, compared to
average age of diagnosis is 32 yr (59). Traditionally, desmoids 1 in 100,000 in the general population (75). As with sporadic
have been linked to trauma, particularly abdominal surgery hepatoblastomas, boys within an FAP kindred are particularly
such as prophylactic colectomy. In a Canadian retrospective at risk (72).
study, 80% of desmoids developed postcolectomy, after an
average of 4.6 yr (60). Females have twice the odds of devel- Other Extracolonic Malignancies
oping desmoids compared to males (57, 61). Family history, Although rare, five cases of jejunal and one case of ileal ade-
presence of osteomas, and germline mutations after codon nocarcinoma were reported by Jagelman et al. among 1,255
1399 have also been identified as independent risk factors for patients with FAP (76). Other extraintestinal cancers associ-
desmoid occurrence (57, 61). To this day, treatment remains ated with FAP include adrenal, pancreatic, and biliary tract
a challenge. Surgical excision carries risks of bleeding and malignancies. Patients from the Johns Hopkins Registry had
short bowel syndrome, with recurrence rates as high as 45% a relative risk of pancreatic adenocarcinoma of 4.46 com-
(62). Nonsteroidal antiinflammatories (usually sulindac) and pared to the general population, with an absolute risk of 21.4
antiestrogens have been used, but less than a third of the tu- cases per 100,000 person-years (77). Table 1 summarizes the
mors stabilize or regress (59, 62). Results from the use of different extracolonic malignancies with their relative and
cytotoxic chemotherapy and radiotherapy have also been dis- lifetime risks (27, 57, 58, 76, 78).
appointing.

Thyroid Cancer
SPECIFIED VARIANTS OF FAP
Another malignancy associated with FAP is thyroid cancer,
with an estimated incidence of 12% (63, 64). The average Gardner Syndrome
age of diagnosis is 25 to 33 yr, with an overwhelming pre- Gardner syndrome is more of a historically coined variant
dominance of females (17:1) (6365). The most common his- of FAP rather than a truly distinct subtype of the disease. It
tological type of thyroid cancer in these patients is papillary is characterized by the association of gastrointestinal poly-
(>75%), with an unusual cribriform pattern (6367). Most posis with osteomas, as well as multiple skin and soft tissue
tumors are multicentric, unilateral (6466), with one North tumors (7981), including desmoids and thyroid tumors. Al-
American series reporting a strong predilection for the left though most FAP patients can be found to have at least sub-
lobe (64). They tend to be well circumscribed, nonaggressive, tle findings of Gardner syndrome on thorough investigation,
with a low metastatic potential and 10-yr mortality (6365, the term is usually used by health professionals to refer to
67). Aside from recommended regular physical examinations patients and families where the aforementioned extraintesti-
of the thyroid gland, there is ongoing debate about the need nal features are especially prominent. Osteomas typically oc-
for additional radiological screening, due to the rarity and cur in the mandible, but can also present in the skull and
excellent long-term prognosis of these tumors. long bones (82). Benign and painless, they usually precede a
clinical or radiological diagnosis of intestinal polyposis (82).
Hepatoblastomas Epidermal cysts are the most common skin manifestation of
Hepatoblastomas are rapidly progressive embryonal liver tu- Gardner syndrome, typically occurring at an earlier age and
mors, usually affecting children under the age of 2.5 yr, with at multiple sites, including the face, scalp, and extremities
a male:female ratio of 2.3:1 (68). Several reports, since 1983, (82). Other cutaneous features of the syndrome include lipo-
have made the link between these lethal tumors and a family mas, fibromas, and leiomyomas. Dental abnormalities, such
history of FAP (6974). Indeed, the incidence of hepatoblas- as supernumerary and impacted teeth, are seen in 2230%
 Familial Adenomatous Polyposis 389

of FAP patients on panoramic radiographs, and constitute yet with 100% penetrance. All affected family members were
another feature of Gardner syndrome (8385). found to have truncating frameshift mutations at codon 1924
of the APC gene, located in the 3 half of exon 15 (88).
Turcot Syndrome
Turcot syndrome, formally described in 1959, refers to the oc-
GENOTYPEPHENOTYPE CORRELATIONS
currence of a primary central nervous system (CNS) tumor, in
conjunction with colorectal polyposis (86). Turcot syndrome Several genotypephenotype correlations have been consis-
has been linked to FAP, as well as the hereditary nonpolyposis tently observed (Fig. 2). As regards aggressivity of the dis-
colorectal cancer syndrome (HNPCC), attributable to muta- ease, mutations at codon 1309 have been typically associated
tions in mismatch repair genes. In families with germline with a more severe clinical phenotype. Patients with muta-
APC mutations, the most common CNS tumor is medul- tions at this site tend to develop bowel symptoms more than
loblastoma, although anaplastic astrocytomas and ependy- 10 yr earlier (mean, 19.8 yr) than those with mutations at other
momas have also been described (87). This contrasts with sites (89), and have significantly more colorectal polyps (ap-
HNPCC, where the major associated CNS tumor seems to proximately 4,000) at the time of colectomy compared with
be glioblastoma (87). Strict neurological evaluation has been matched FAP controls (90). Also, mutations at codon 1309
recommended for FAP families with a member affected by a are associated with an earlier age of CRC development (mean,
CNS tumor, due to evidence of familial clustering (87). No 35 yr) (84).
guidelines exist, however, and there are no studies so far to As regards extracolonic manifestations, mutations from
determine whether such an intervention could improve sur- codons 976 to 1067 are associated with a three- to four-
vival. fold increased risk for developing duodenal adenomas, while
those spanning between codons 543 and 1309 are associated
Hereditary Desmoid Disease with a high risk of CHRPE (84). In fact, CHRPE lesions are
In 1996, Eccles et al. proposed the existence of yet another hardly ever present with mutations before exon 9, but are sys-
variant of FAP, termed hereditary desmoid disease (HDD). tematically present with mutations past this exon (55). Mu-
The authors described a family with multiple desmoid tumors tations beyond codon 1309 are linked to a six-fold increased
inherited across three generations, but occurring at sites un- risk of desmoid tumors (84), with the majority of those muta-
usual for FAP-related desmoid disease (paraspinal muscles, tions concentrated between codons 1445 and 1580 (89, 91).
breast, occiput, arms, and lower limbs) (88). Affected kindred Patients with papillary thyroid cancer often have mutations
also lacked the colonic features of FAP, except for one patient between codons 140 and 1309, the majority of which are
who had a palpable rectal mass, and another who had <50 concentrated in the CHRPE-associated area on exon 15 (92).
adenomatous polyps documented by colonoscopy (88). Like Mutations beyond codon 1444 are associated with a two-fold
FAP, HDD was inherited in an autosomal-dominant fashion, increased risk of osteomas (84, 93).

Figure 2. APC cDNA (below) and extracolonic genotypephenotype correlations (above). Except for CHRPE (congenital hypertrophy of
the retinal pigment epithelium), most lesions can occur with mutations anywhere along the APC gene, but are more likely in the locations
illustrated. AFAP = attenuated FAP. Adapted from Fearnhead et al. (7), Foulkes (8), Bertario et al. (84), and Cetta et al. (92).
390 Galiatsatos and Foulkes

Apart from mutation site, variations in phenotype have orectal adenomas and controls (108110). This variant was
also been potentially attributed to environmental factors, as also absent in 194 Swedish CRC patients, with sporadic or
well as interdependence of first and second hits, consistent familial tumors (111).
with Knudsons two-hit hypothesis. Moreover, the discovery
of a modifier locus (Mom1) on chromosome 4 of the mouse MYH Associated Polyposis
polyposis model (94) has led to the identification of a pos- More recently, an autosomal recessive type of oligopolypo-
sible modifier gene on human chromosome 1p3536, which sis has also been recognized, involving the human MutY
may also be implicated in the clinical heterogeneity of FAP homologue (MYH, or more accurately MUTYH) gene, re-
(95,96). Variations in the N-acetyltransferase loci NAT1 and ferred to as MYH associated polyposis (MAP). MYH, lo-
NAT2, located on chromosome 8p22, have also been shown cated on the short arm of chromosome 1, is a base excision
to affect the severity of disease (97). These modifier genes repair gene preventing mutations from products of oxidative
are not in clinical use at this point in time. In most cases, the damage, particularly the oxidized guanine lesion 8-oxodG
family history is the best guide as to the likely phenotypic (112). Biallelic mutations in MYH were first associated with
expression of APC mutations. polyposis after the study of a family with multiple colorec-
tal adenomas/carcinomas, who lacked inherited mutations in
APC (113). Since then, other patients with multiple adenomas
MULTIPLE COLORECTAL ADENOMA SYNDROMES have also been found to be homozygous or compound het-
erozygous carriers of MYH mutations. Y165C and G382D
Attenuated FAP missense mutations account for the majority (>80%) of
Attenuated familial adenomatous polyposis (AFAP) is a phe- disease-causing alleles in Caucasians, whereas E466X non-
notypically distinct variant of FAP, characterized by the pres- sense mutation has been identified in Indian families, and
ence of fewer than 100 adenomas, a more proximal colonic Y90X in Pakistani families (114). A mutation in exon 14 of
location of polyps, and delayed age of CRC onset (15 yr later the MYH gene, 1395delGGA, has also recently been identi-
than patients with classic FAP). The cumulative risk of CRC fied in three Italian patients with colorectal polyposis (115).
by the age of 80 yr is estimated to be 69%, and 75% of tumors In a British study of multiple colorectal adenoma patients,
occur in the proximal colon (98). Patients often have no fam- 29% of those with 15100 adenomas had biallelic pathogenic
ily history of polyps or CRC, and lack extracolonic features, MYH mutations, in comparison to 7.5% of patients with APC
apart from FGPs which are quite common, and duodenal ade- mutation-negative classic polyposis (>100 adenomas) (116).
nomas (99). Other studies have noted similar frequencies of MYH alter-
AFAP arises from mutations in the extreme proximal ations in multiple colorectal adenoma patients, ranging from
or distal portions of the APC gene, specifically truncating 23% to 36% (114, 117). No unaffected carrier of biallelic
frameshift mutations at the 3 end of the gene (100104), and MYH mutations has been identified to date, suggesting a high
nonsense/frameshift mutations at exons 3, 4, and 5 (101, 105). penetrance for this condition (118). In the largest prospective
Also reported as a cause of AFAP are nonsense mutations at cohort study to date, including 2,239 CRC cases and 1,845
exon 9 (100, 101). controls from across Scotland, G382D/G382D homozygotes
and Y165C/G382D compound heterozygotes had a 93-fold
APC Gene Polymorphisms excess risk of CRC (95% CI 42213) compared to wild-type
Only a minority of patients with multiple colorectal ade- individuals, while all G382D/G382D homozygote carriers
nomas (usually defined as the presence of 3100 colonic had developed CRC by the age of 65 years (119). The impli-
adenomas) harbor an identifiable APC germline mutation, cations of a single MYH-mutated allele remain unclear, but
which has made genetic diagnosis challenging. In a British the risk for CRC is unlikely to be more than 50% increased.
study of 164 unrelated patients with multiple colorectal ade- In Siebers study of multiple adenoma patients, 6 patients
nomas, only 8.5% carried germline APC variants, with pos- (3.8%) were heterozygotes for an MYH mutation, and had
sible pathogenic effects (100). The most common variant was 312 (median of 4) adenomas in the colon, as opposed to
the missense polymorphism E1317Q, carried by 4.3% of pa- 18100 adenomas in homozygous carriers (116). In a Cana-
tients (relative risk 11.17, p < 0.001), while 1.8% of the dian case-control study comparing 1,238 CRC patients and
patients (all Ashkenazi) carried the I1307K APC variant, dis- 1,255 healthy controls, 2.34% of case patients versus 1.67%
cussed earlier (100). Reports concerning the pathogenicity of of control subjects were heterozygous for either the Y165C
the E1317Q variant are so far contradictory and controver- or G382D mutation, suggesting a possible weakly penetrant
sial. Frayling et al. found an E1317Q variant to be present in autosomal-dominant inheritance pattern of increased CRC
2 of 134 multiple adenoma patients, 2 of 30 CRC patients, risk associated with monoallelic germline MYH mutations
but in none of 80 controls (106). In another study, the odds (120). Carriers of either single mutation had a combined OR
ratio of E1317Q in multiple adenoma patients versus controls of 1.4 for CRC, although this did not reach statistical signif-
was 2.0, but did not reach statistical significance (p = 0.4) icance (95% CI 0.82.5) (120). In the Scottish cohort study
(107). More recent studies have shown no difference in the mentioned above, there was a 1.68-fold excess risk of CRC
prevalence of this variant between patients with multiple col- (95% CI 1.072.95) for heterozygote carriers aged >55 yr,
 Familial Adenomatous Polyposis 391

while the risk for heterozygotes of all ages did not reach sta- form of subtotal colectomy should continue close endoscopic
tistical significance (119). Further studies are needed before surveillance of the remaining rectum approximately every 6
more accurate estimates of risk can be quoted. months, for recurrent adenomas or cancer (126).
Proctocolectomy with an ileal pouchanal anastomosis
(IPAA) has emerged as the surgical treatment of choice, al-
SCREENING lowing for complete resection of vulnerable colorectal mu-
cosa, while preserving transanal defecation. Although IPAA
Screening of patients and family members, with timely treat-
has been associated with a higher rate of postoperative com-
ment of affected individuals, has led to a 55% reduction in
plications, the functional results of IRA and IPAA appear to
the occurrence of CRC at diagnosis of FAP, and an improve-
be similar, as far as the frequency of bowel movements and
ment in cumulative survival for all FAP patients (121, 122).
daytime soiling are concerned (129). Incontinence occurs in
The American Gastroenterological Association recommends
roughly 5.9% of FAP patients with an IPAA, and the average
an annual sigmoidoscopy, beginning at the age of 1012 yr,
number of bowel movements is 56 per 24-h period (129,
for patients with a genetic diagnosis of FAP, or at-risk fam-
130). Pouch failure can occur in 7.7% of FAP patients over
ily members who have not undergone genetic testing (123).
a 210-yr follow-up period, mostly due to ischemia and late-
Most authors also recommend front and/or side-viewing en-
onset pelvic sepsis (131). Pouchitis occurs in only 11% of
doscopies of the stomach, duodenum, and periampullary re-
FAP patients, compared to 53.8% in patients with underly-
gion, every 6 months to 4 yr depending on the polyp burden
ing ulcerative colitis (130). Concerns have been raised about
(27, 35, 36, 124). Some even advocate the systematic use of
a marked reduction in female fertility following IPAA, as
0.5% indigo carmine dye, and routine biopsy of the duode-
noted in ulcerative colitis patients (132, 133). Pelvic adhe-
nal papilla, even in the absence of macroscopic lesions (30).
sions, disrupting the normal anatomic relationships between
As far as thyroid cancer is concerned, most would agree that
the fallopian tubes and ovaries, may be the cause. In a recent
it is reasonable to include a simple thyroid palpation in the
study by Olsen et al., the fecundity of women with FAP after
routine physical exam (63), while others would go as far as
IPAA dropped to 46% compared to the preoperative level (p
recommending routine thyroid ultrasonography (77). Some
= 0.001), while there was no observed change in fecundity
experts recommend screening for hepatoblastomas in chil-
before and after IRA (133). Still, the fertility rate of women
dren of FAP parents by use of routine alpha-fetoprotein levels
after IPAA was greater for those with FAP compared to ul-
and imaging of the liver (72, 125), but no standard guidelines
cerative colitis (134).
exist. Proposed algorithms for screening probands and unaf-
IPAA patients remain at risk for ileal polyps. The risk of
fected first-degree relatives are presented in Figure 3A and
developing adenomas within the ileal pouch 5, 10, and 15 yr
3B, respectively.
after proctocolectomy is roughly 7%, 35%, and 75%, respec-
tively (135). Others have estimated an incidence of 53% to
as high as 83%, 1020 yr postsurgery (136, 137). The risk
PROPHYLAXIS
of developing anastomotic adenomas is significantly lower in
Colectomy is the recommended treatment to reduce the risk patients undergoing a mucosectomy with hand-sewn anasto-
of colorectal cancer in FAP patients with adenomatosis. In mosis, compared with the simpler, more traditional stapled
children and adolescents, surgery can usually be safely post- anastomosis (138). Although IPAA significantly decreases
poned for several years, while continuing with annual colono- the residual risk of rectal cancer that accompanies an IRA,
scopies, until an appropriate psychological age is reached there have been four cases of invasive adenocarcinoma at
where colectomy can be accepted (usually late teens to early the ileoanal anastomosis (139142), one case of adenocar-
twenties). Surgical options include a subtotal colectomy with cinoma within the ileal pouch (143), and two cases of ade-
ileorectal anastomosis, a total proctocolectomy with a conti- nocarcinoma within the anal transitional zone (144) reported
nent ileostomy, or a proctocolectomy with ileoanal pouch. in the literature. Continued endoscopic surveillance of the
Subtotal colectomy with ileorectal anastomosis (IRA), al- ileoanal anastomosis is probably warranted, although no for-
though simpler and traditionally associated with less periop- mal guidelines exist.
erative complications and better functional results, has be- Medical interventions for CRC prevention have also been
come a less attractive option due to an ongoing CRC risk proposed, although at this point in time they are not effective
associated with residual rectal mucosa. The estimated cumu- enough to be considered a reasonable alternative to surgery.
lative risk of rectal cancer with this limited procedure is 10% Sulindac, a nonsteroidal antiinflammatory drug, has been
at the age of 50 yr, reaching up to 29% by the age of 60 proven to cause regression of colorectal adenomas in FAP
yr (126). Meanwhile, others have argued that the risk of dy- patients, by the induction of apoptosis (145). Most studies,
ing from rectal cancer after an IRA is only 2% after a 15-yr however, have shown incomplete polyp regression, and over
follow-up, making it an acceptable primary treatment option short follow-up periods (1 yr) (146150). Long-term ben-
for FAP patients (127). Laparoscopic colectomy with IRA efits of sulindac therapy for FAP patients having had IRA
has also proven to be a safe and minimally invasive treatment are inconsistent, ranging from no difference (151), to a 72%
option for selected FAP patients (128). Those undergoing any decrease in baseline polyp number (152). COX-2 inhibitors
392 Galiatsatos and Foulkes

A Proband with clinical diagnosis of FAP

Offer first-line genetic testing: full DNA

sequencing of APC gene

Mutation identified No mutation identified

Offer first-degree
relatives genetic testing
Second-line genetic testing:
Refer for appropriately for known mutation
MAMA, MLPA. Consider
timed colectomy MYH.

Subtotal Total colectomy forward + side-viewing upper

colectomy with with IPAA. Mutation identified Mutation not identified
endoscopy every 1-3yrs
IRA. Flexible Periodic flexible (depending on polyp burden) *
sigmoidoscopy sigmoidoscopy for
every 6 months.
annual thyroid exam by
pouch
palpation (consider U/S)
Consider sulindac polyps/cancers
or other NSAID to consider periodic abdominal Screen first-degree
Refer for relatives with flexible
decrease U/S for pancreatic cancer +
appropriately timed sigmoidoscopy as of age
incidence of rectal desmoid tumor screening
colectomy 10-12. Consider eye
polyps. consider routine alpha-
exam for CHRPE. If
fetoprotein and liver U/S for
proband has CHRPE,
children of proband until age 5.
offer eye exam to first-
degree relatives as
screening test

B
Unaffected first-degree relative of FAP proband

No mutation identified in Test for mutation identified in Proband not available for
proband, or mutation proband as of age 10-12 genetic testing
identified but relative refuses
or has no access to genetic
testing.
Offer relative genetic testing
(as in Fig 3a)
Mutation absent Mutation present
Flexible sigmoidoscopy
annually starting at age 10-12,
biennially from 26-35, every Mutation absent
third year from 36-50. Average population Colonoscopy annually
Consider eye exam for CHRPE. colon cancer screening from age 10-12
guidelines
Follow algorithm as for
families where no mutation is
identified in proband

If no polyps If found Refer for appropriately Onset of polyps

by age 50, to have timed colectomy
follow polyps
general forward + side-viewing upper endoscopy
population every 1-3yrs (depending on polyp
colon cancer Subtotal colectomy with IRA. burden)
Total colectomy with
screening
guidelines
Flexible sigmoidoscopy every 6 IPAA. Periodic flexible annual thyroid exam by palpation
months. Consider sulindac or other sigmoidoscopy for pouch (consider U/S)
NSAID to decrease incidence of rectal polyps/cancers consider periodic abdominal U/S for
polyps.
pancreatic cancer + desmoid tumor
screening

Figure 3. (A) Algorithm for proband with clinical diagnosis of FAP. Extracolonic cancer screening is not yet established for MYH mutation
carriers. (B) Algorithm for first-degree relatives of FAP proband. CHRPE = congenital hypertrophy of the retinal pigment epithelium, IPAA
= ileal pouch-anal anastomosis, IRA = ileorectal anastomosis, MAMA = monoallelic mutation analysis, MLPA = multiplex ligation-
dependant probe amplification, NSAID = nonsteroidal antiinflammatory drug, U/S = ultrasound.

have also been investigated. Celecoxib was proven to signif- concern (155, 156), and make the use of these drugs for this
icantly decrease duodenal polyposis after 6 months of high- indication much less appealing.
dose treatment (400 mg twice daily) (153). Rofecoxib was Although surgical prophylactic measures have favor-
also shown to reduce the rate of colorectal polyp formation ably changed the natural history of FAP with regards to
in eight patients with FAP (154). Recent reports, however, of CRC risk, management of duodenal adenomatosis remains
increased cardiovascular and thrombotic events with COX-2 a challenge. There are several endoscopic options avail-
inhibitors in adenoma chemoprevention trials are cause for able, including snare polypectomy, thermal ablation (using
 Familial Adenomatous Polyposis 393

monopolar/bipolar cautery or argon plasma), and laser coag- that genetic testing only be done in the setting of pre- and
ulation. Unfortunately, the multiplicity of lesions, their of- posttest counseling, to address the clinical, psychological,
ten sessile or flat configuration, and the risk of scarring and and ethical issues that are raised during the process (164). In
stricturing of the ampulla as a result of repeated excisions a nationwide study of 177 American patients being tested for
and diathermy limit their usefulness. Photodynamic therapy APC gene mutations, only 18.6% received pretest counsel-
(PDT) is a method used to induce localized necrosis using an ing, while only 16.9% provided informed consent (165). The
endoscopic light after the administration of a photosensitiz- authors found that nearly 20% of tests were ordered for indi-
ing agent. There is still very little experience in using PDT to cations considered unconventional, resulting in a low rate of
treat duodenal polyps in FAP, and one pilot study has shown positive results (2.3%) in this subgroup of individuals (165).
only limited responses, with reduction in adenoma size but Also, it was estimated that as many a third of patients tested
no complete eradication (157). A more radical approach for would have received misleading answers, owing to the inabil-
isolated extraductal ampullary lesions involves endoscopic ity of many physicians interviewed to correctly interpret the
snare papillectomy with or without pancreatic stent place- test results, particularly where false negatives were concerned
ment. This technique seems to be well tolerated, with an 8 (165). The use of consistent pre- and post-genetic counseling
15% complication rate (including pancreatitis, bleeding, and would likely have avoided many of these problems. Ideally,
perforation); however, adenomas have been shown to recur genetic counseling sessions should be face-to-face, with a
(158, 159). professional who could collect the necessary data to con-
The high prevalence of FAP-associated duodenal adeno- struct a three-generation pedigree, educate the patient and
mas, the difficulty in early detection of duodenal cancer, the family as to the medical aspects of the disease, the inher-
limitations of local ablative techniques, and the grim progno- itance pattern, and the recommended screening guidelines,
sis of invasive tumors raise the question of preventive surgery explore the psychosocial aspects of testing, obtain informed
for severe or progressive duodenal adenomatosis. However, consent, disclose the results and address the risks and man-
the optimal timing and technique remain unclear, and con- agement, as well as be available to answer further questions
clusive evidence of improved prognosis with early surgical and assure follow-up when this is required (166, 167).
treatment is still lacking.
Most experts agree that prophylactic surgery should
be considered for Spigelman stage IIIIV polyps (villous
SUMMARY
changes, severe dysplasia), rapidly growing lesions, peri-
ampullary adenomas in patients over 3540 yr of age, par- FAP is an autosomal-dominant syndrome, most commonly
ticularly if there is a family history of duodenal cancer caused by a truncating mutation in the APC gene at chromo-
(160). Surgical options include pylorus-preserving pancreati- some 5q21. It is characterized by the early onset of numerous
coduodenectomy (PPPDR), pancreas-sparing duodenectomy, colonic adenomas, with an almost inevitable progression to
duodenotomy with surgical polypectomy, and ampullectomy CRC. Other features include gastroduodenal polyps, desmoid
(161). Duodenotomy with polypectomy is the least preferred, tumors, and extraintestinal manifestations including CHRPE,
as it has been associated with up to 100% recurrence of adeno- osteomas, and other malignancies. Genotypephenotype cor-
mas within 636 months (162). There are few follow-up data relations have been observed. An attenuated form of FAP
on the use of pancreas-sparing duodenectomy, and concern exists, characterized by the development of <100 colorectal
that cancer may develop in the area of mucosa surrounding adenomas, and a delayed CRC onset. MYH-associated poly-
the ampulla that is left behind. PPPDR is the preferred pro- posis is an distinct autosomal recessive condition, caused by
cedure in most centers. Notwithstanding, there is a reported mutations in the MYH gene, which should figure in the dif-
morbidity and mortality rate in the range of 40% and 4.5%, ferential diagnosis of anyone with multiple colorectal ade-
respectively (163). Outcomes of PPPDR may be worse in FAP nomas, particularly in the absence of an identifiable APC
patients due to adhesions and desmoplastic changes related mutation. Strict endoscopic surveillance is recommended for
to previous surgery, notably colectomy. all FAP patients and at-risk family members. The optimal
treatment remains prophylactic colectomy, while continued
surveillance of the rectal remnant or ileoanal anastomosis
seems warranted because of ongoing risks of adenomas and
GENETIC COUNSELING
carcinomas within residual mucosa.
Genetic counseling is essential in the management of FAP Although heightened awareness, endoscopic surveillance,
patients and families, and in most centers constitutes a pre- and the establishment of polyposis registries have success-
requisite for genetic testing. Not only do individuals need fully decreased the incidence and mortality from CRC, the
to understand the clinical aspects and implications of FAP, challenge now lies in determining the optimal screening and
they must be made aware of the risks, benefits, and limita- therapeutic modalities for associated extracolonic malignan-
tions of genetic testing in order to make an informed deci- cies that are consequently becoming more prominent. The
sion, and be prepared to cope with the eventual results. The emergence of capsule endoscopy raises the question of its
American Society of Clinical Oncology (ASCO) advocates utility in detecting small bowel polyps and cancers in the
394 Galiatsatos and Foulkes

context of FAP, as in other hereditary polyposis syndromes. 16. Meuller J, Kanter-Smoler G, Nguyen AO, et al. Identifi-
With the progress of endoscopy, techniques such as ampullec- cation of genomic deletions of the APC gene in familial
tomy with or without thermal ablation are being evaluated as adenomatous polyposis by two independent quantitative
techniques. Genet Test 2004;8:24856.
an alternative to surgery for the management of periampullary 17. Laken S, Petersen G, Gruber S, et al. Familial colorectal
adenomas or malignancies confined to the papilla. Ultimately, cancer in Ashkenazim due to a hypermutable tract in APC.
the aim is to decrease morbidity, and strive for longevity and Nat Genet 1997;17:7983.
an acceptable quality of life for those affected. 18. Woodage T, King SM, Wacholder S, et al. The APCI1307K
allele and cancer risk in a community-based study of
Ashkenazi Jews. Nat Genet 1998;20:625.
19. Tomlinson I, Rahman N, Frayling I, et al. Inherited suscep-
ACKNOWLEDGMENTS tibility to colorectal adenomas and carcinomas: Evidence
for a new predisposition gene on 15q14-q22. Gastroen-
We wish to thank the two anonymous referees for their helpful terology 1999;116:78995.
comments. 20. Jaeger EE, Woodford-Richens KL, Lockett M, et al. An
ancestral Ashkenazi haplotype at the HMPS/CRAC1 locus
Reprint requests and correspondence: William D. Foulkes, M.B., on 15q1314 is associated with hereditary mixed polyposis
Ph.D., Department of Medical Genetics, Room A801, Sir Mortimer syndrome. Am J Hum Genet 2003;72:12617.
B. Davis Jewish General Hospital, 3755 Cote Ste-Catherine, Mon- 21. Sieber OM, Lipton L, Crabtree M, et al. Multiple colorec-
treal, QC, H3T 1E2, Canada. tal adenomas, classic adenomatous polyposis, and germ-
Received July 11, 2005; accepted September 12, 2005. line mutations in MYH. N Engl J Med 2003;348:791
9.
22. Bjork J, Akerbrant H, Iselius L, et al. Epidemiology of
REFERENCES familial adenomatous polyposis in Sweden: Changes over
time and differences in phenotype between males and fe-
1. Bisgaard ML, Fenger K, Bulow S, et al. Familial adenoma- males. Scand J Gastroenterol 1999;34:12305.
tous polyposis (FAP): Frequency, penetrance, and mutation 23. Burt RW, Bishop DT, Lynch HT, et al. Risk and surveil-
rate. Hum Mutat 1994;3:1215. lance of individuals with heritable factors for colorectal
2. Bussey HJR. Familial polyposis coli; family studies, cancer. WHO Collaborating Centre for the Prevention of
histopathology, differential diagnosis, and results of treat- Colorectal Cancer. Bull World Health Organ 1990;68:655
ment. Baltimore: John Hopkins University Press, 1975. 65.
3. Bodmer WF, Bailey CJ, Bodmer J, et al. Localization of the 24. Iwama T, Tamura K, Morita T, et al. A clinical overview of
gene for familial adenomatous polyposis on chromosome familial adenomatous polyposis derived from the database
5. Nature 1987;328:6146. of the Polyposis Registry of Japan. Int J Clin Oncol
4. Kinzler KW, Nilbert MC, Su LK, et al. Identification 2004;9:30816.
of FAP locus genes from chromosome 5q21. Science 25. Bulow S, Bulow C, Nielsen TF, et al. Centralized reg-
1991;253:6615. istration, prophylactic examination, and treatment results
5. Groden J, Thliveris A, Samowitz W, et al. Identification in improved prognosis in familial adenomatous polyposis.
and characterization of the familial adenomatous polyposis Results from the Danish Polyposis Register. Scand J Gas-
coli gene. Cell 1991;66:589600. troenterol 1995;30:98993.
6. Van Es JH, Giles RH, Clevers HC. The many faces of the 26. Jarvinen HJ. Epidemiology of familial adenomatous poly-
tumor suppressor gene APC. Exp Cell Res 2001;264:126 posis in Finland: Impact of family screening on the
34. colorectal cancer rate and survival. Gut 1992;33:357
7. Fearnhead NS, Britton MP, Bodmer WF. The ABC of APC. 60.
Hum Mol Genet 2001;10:72133. 27. Bulow S, Bjork J, Christensen IJ, et al. Duodenal ade-
8. Foulkes WD. A tale of four syndromes: Familial adeno- nomatosis in familial adenomatous polyposis. The DAF
matous polyposis, Gardner syndrome, attenuated APC and Study Group. Gut 2004;53:3816.
Turcot syndrome. Q J Med 1995;88:85363. 28. Bertoni G, Sassatelli R, Nigrisoli E, et al. High preva-
9. Nathke I. APC at a glance. J Cell Sci 2004;117:48735. lence of adenomas and microadenomas of the duodenal
10. Goss KH, Groden J. Biology of the adenomatous polyposis papilla and periampullary region in patients with famil-
coli tumor suppressor. J Clin Oncol 2000;18:196779. ial adenomatous polyposis. Eur J Gastroenterol Hepatol
11. Beroud C, Soussi T. APC gene: Database of germline and 1996;8:12016.
somatic mutations in human tumors and cell lines. Nucl 29. Arvanitis ML, Jagelman DG, Fazio VW, et al. Mortality in
Acids Res 1996;24:1214. patients with familial adenomatous polyposis. Dis Colon
12. Nagase H, Nakamura Y. Mutation of the APC (adenoma- Rectum 1990;33:63942.
tous polyposis coli) gene. Hum Mutat 1993;2:42534. 30. Saurin JC, Gutknecht C, Napoleon B, et al. Surveillance
13. Heinimann K, Thompson A, Locher A, et al. Nontruncat- of duodenal adenomas in familial adenomatous polyposis
ing APC germ-line mutations and mismatch repair defi- reveals high cumulative risk of advanced disease. J Clin
ciency play a minor role in APC mutation-negative poly- Oncol 2004;22:4938.
posis. Cancer Res 2001;61:761622. 31. Vasen HFA, Bulow S, Myrhj T, et al. Decision analysis
14. Laken SJ, Papadopoulos N, Petersen GM, et al. Analysis in the management of duodenal adenomatosis in familial
of masked mutations in familial adenomatous polyposis. adenomatous polyposis. Gut 1997;40:7169.
Proc Natl Acad Sci USA 1999;96:23226. 32. Bjork J, Akerbrant H, Iselius L, et al. Periampullary adeno-
15. Casey G, Lindor NM, Papadopoulos N, et al. Conver- mas and adenocarcinomas in familial adenomatous poly-
sion analysis for mutation detection in MLH1 and MSH2 posis: Cumulative risks and APC gene mutations. Gas-
in patients with colorectal cancer. J Am Med Assoc troenterology 2001;121:112735.
2005;293:799809. 33. Watanabe H, Enjoji M, Ohsato K. Gastric lesions in
 Familial Adenomatous Polyposis 395

familial adenomatosis coli: Their incidence and histologic ocular fundus lesions in the inherited gastrointestinal poly-
analysis. Hum Pathol 1978;9:26983. posis syndromes and in hereditary nonpolyposis colorectal
34. Iida M, Yao T, Itoh H, et al. Natural history of cancer. Ophthalmology 1988;95:9649.
fundic gland polyposis in patients with familial ade- 53. Tiret A, Taiel-Sartral M, Tiret E, et al. Diagnostic value of
nomatosis coli/Gardners syndrome. Gastroenterology fundus examination in familial adenomatous polyposis. Br
1985;89:10215. J Ophthalmol 1998;81:7558.
35. Sarre RG, Frost AG, Jagelman DG, et al. Gastric and 54. Morton DG, Gibson J, Macdonald F, et al. Role of con-
duodenal polyps in familial adenomatous polyposis: A genital hypertrophy of the retinal pigment epithelium in
prospective study of the nature and prevalence of upper the predicitve diagnosis of familial adenomatous polypo-
gastrointestinal polyps. Gut 1987;28:30614. sis. Br J Surg 1992;79:68993.
36. Church JM, McGannon E, Hull-Boiner S, et al. Gastroduo- 55. Olschwang S, Tiret A, Laurent-Puig P, et al. Restriction
denal polyps in patients with familial adenomatous poly- of ocular fundus lesions to a specific subgroup of APC
posis. Dis Colon Rectum 1992;35:11703. mutations in adenomatous polyposis coli patients. Cell
37. Kinoshita Y, Tojo M, Yano T, et al. Incidence of fundic 1993;75:95968.
gland polyps without familial adenomatous polyposis. 56. Miyoshi Y, Iwao K, Nawa G, et al. Frequent mutations
Gastrointest Endosc 1993;39:1613. in the beta-catenin gene in desmoid tumors from pa-
38. Marcial MA, Villafana M, Hernandez-Denton J, et al. tients without familial adenomatous polyposis. Oncol Res
Fundic gland polyps: Prevalence and clinicopathologic fea- 1998;10:5914.
tures. Am J Gastroenterol 1993;88:17113. 57. Sturt NJH, Gallagher MC, Bassett P, et al. Evidence for
39. Odze RD, Marcial MA, Antonioli D. Gastric fundic gland genetic predisposition to desmoid tumours in familial ade-
polyps: A morphological study including mucin histo- nomatous polyposis independant of the germline APC mu-
chemistry, stereometry, and MIB-1 immunohistochem- tation. Gut 2004;53:18326.
istry. Hum Pathol 1996;27:896903. 58. Lynch HT, Fitzgibbons R Jr. Surgery, desmoid tumors, and
40. Nakamura S, Matsumoto T, Kobori Y, et al. Impact of He- familial adenomatous polyposis: Case report and literature
licobacter pylori infection and mucosal atrophy on gastric review. Am J Gastroenterol 1996;91:2598601.
lesions in patients with familial adenomatous polyposis. 59. Clark SK, Neale KF, Landgrebe JC, et al. Desmoid tumours
Gut 2002;51:4859. complicating familial adenomatous polyposis. Br J Surg
41. Wu TT, Kornacki S, Rashid A, et al. Dysplasia and dys- 1999;86:11859.
regulation of proliferation in foveolar and surface epithelia 60. Soravia C, Berk T, McLeod RS, et al. Desmoid disease in
of fundic gland polyps from patients with familial adeno- patients with familial adenomatous polyposis. Dis Colon
matous polyposis. Am J Surg Pathol 1998;22:2938. Rectum 2000;43:3639.
42. Coffey RJ, Knight CD, Van Heerden JA, et al. Gastric ade- 61. Bertario L, Russo A, Sala P, et al. Genotype and phe-
nocarcinoma complicating Gardners syndrome in a North notype factors as determinants of desmoid tumors in pa-
American woman. Gastroenterology 1985;88:12636. tients with familial adenomatous polyposis. Int J Cancer
43. Goodman AJ, Dundas SA, Scholefield JH, et al. Gastric 2001;95:1027.
carcinoma and familial adenomatous polyposis (FAP). Int 62. Heiskanen I, Jarvinen HJ. Occurence of desmoid tumours
J Colorectal Dis 1988;3:2013. in familial adenomatous polyposis and results of treatment.
44. Zwick A, Munir M, Ryan CK, et al. Gastric adenocarci- Int J Colorect Dis 1996;11:15762.
noma and dysplasia in fundic gland polyps of a patient with 63. Bulow C, Bulow S, Leeds Castle Polyposis Group. Is
attenuated adenomatous polyposis coli. Gastroenterology screening for thyroid carcinoma indicated in familial ade-
1997;113:65963. nomatous polyposis? Int J Colorectal Dis 1997;12:240
45. Hofgartner WT, Thorp M, Ramus MW, et al. Gastric ade- 2.
nocarcinoma associated with fundic gland polyps in a pa- 64. Truta B, Allen BA, Conrad PG, et al. Genotype and pheno-
tient with attenuated familial adenomatous polyposis. Am type of patients with both familial adenomatous polyposis
J Gastroenterol 1999;94:227681. and thyroid carcinoma. Fam Cancer 2003;2:959.
46. Abraham SC, Nobukawa B, Giardiello FM, et al. Fundic 65. Perrier ND, Van Heerden JA, Goellner JR, et al. Thyroid
gland polyps in familial adenomatous polyposis. Am J cancer in patients with familial adenomatous polyposis.
Pathol 2000;157:74754. World J Surg 1998;22:73842.
47. Abraham SC, Nobukawa B, Giardiello FM, et al. Spo- 66. Harach HR, Williams GT, Williams ED. Familial adeno-
radic fundic gland polyps: Common gastric polyps arising matous polyposis associated thyroid carcinoma: A dis-
through activating mutations in the -catenin gene. Am J tinct type of follicular cell neoplasm. Histopathology
Pathol 2001;158:100510. 1994;25:54961.
48. Ruhswurm I, Zehetmayer M, Dejaco C, et al. Ophthalmic 67. Cameselle-Teijeiro J, Chan JK. Cribriform-morular vari-
and genetic screening in pedigrees with familial adenoma- ant of papillary carcinoma: A distinctive variant repre-
tous polyposis. Am J Ophthalmol 1998;125:6806. senting the sporadic counterpart of familial adenoma-
49. Romania A, Zakov ZN, McGannon E, et al. Congenital tous polyposis-associated thyroid carcinoma? Mod Pathol
hypertrophy of the retinal pigment epithelium in famil- 1999;12:40011.
ial adenomatous polyposis. Ophthalmology 1989;96:879 68. Hartley AL, Birch JM, Kelsey AM, et al. Epidemiological
88. and familial aspects of hepatoblastoma. Med Pediatr Oncol
50. Traboulsi EI, Krush AJ, Gardner EJ, et al. Prevalence and 1990;18:1039.
importance of pigmented ocular fundus lesions in Gard- 69. Kingston JE, Herbert A, Draper GJ, et al. Association be-
ners syndrome. N Engl J Med 1987;316:6617. tween hepatoblastoma and polyposis coli. Arch Dis Child
51. Hartvigsen A, Myrhoj T, Bulow S, et al. Ophthalmoscopy 1983;58:95962.
for congenital hypertrophy of the retinal pigment epithe- 70. Garber JE, Li FP, Kingston JE, et al. Hepatoblastoma
lium (CHRPE) in patients with sporadic colorectal carci- and familial adenomatous polyposis. J Natl Cancer Inst
noma. Int J Colorect Dis 1995;10:1389. 1988;80:16268.
52. Traboulsi EI, Maumenee IH, Krush AJ, et al. Pigmented 71. Bernstein IT, Bulow S, Mauritzen K. Hepatoblastoma in
396 Galiatsatos and Foulkes

two cousins in a family with adenomatous polyposis. Re- ilies with familial adenomatous polyposis coli. Ann Surg
port of two cases. Dis Colon Rectum 1992;35:3734. 1999;229:35061.
72. Giardiello FM, Petersen GM, Brensinger JD, et al. 92. Cetta F, Montalto G, Gori M, et al. Germline mutations
Hepatoblastoma and APC gene mutation in familial ade- of the APC gene in patients with familial adenomatous
nomatous polyposis. Gut 1996;39:8679. polyposis-associated thyroid carcinoma: Results from a
73. Thomas D, Pritchard J, Davidson R, et al. Familial hep- European Cooperative Study. J Clin Endocrinol Metab
atoblastoma and APC gene mutations: Renewed call for 2000;85:28692.
molecular research. Eur J Cancer 2003;39:22004. 93. Davies DR, Armstrong JG, Thakker N, et al. Severe Gard-
74. Inukai T, Furuuchi K, Sugita K, et al. Nuclear accumula- ner syndrome in families with mutations restricted to
tion of beta-catenin without an additional somatic muta- a specific region of the APC gene. Am J Hum Genet
tion in coding region of the APC gene in hepatoblastoma 1995;57:11518.
from a familial adenomatous polyposis patient. Oncol Rep 94. Dietrich WF, Lander ES, Smith JS, et al. Genetic iden-
2004;11:1216. tification of Mom-1, a major modifier locus affect-
75. Hughes LJ, Michels VV. Risk of hepatoblastoma in ing Min-induced intestinal neoplasia in the mouse. Cell
familial adenomatous polyposis. Am J Med Genet 1993;75:6319.
1992;43:10235. 95. Tomlinson IP, Neale K, Talbot IC, et al. A modifying lo-
76. Jagelman DG, DeCosse JJ, Bussey HJR. Upper gastroin- cus for familial adenomatous polyposis may be present on
testinal cancer in familial adenomatous polyposis. Lancet chromosome 1p35-p36. J Med Genet 1996;33:26873.
1988;1:114951. 96. Dobbie Z, Heinimann K, Bishop DT, et al. Identification of
77. Giardiello FM, Offerhaus GJ, Lee DH, et al. Increased risk a modifier gene locus on chromosome 1p3536 in familial
of thyroid and pancreatic carcinoma in familial adenoma- adenomatous polyposis. Hum Genet 1997;99:6537.
tous polyposis. Gut 1993;34:13946. 97. Crabtree MD, Fletcher C, Churchman M, et al. Analysis of
78. Giardiello FM, Offerhaus JGA. Phenotype and can- candidate modifier loci for the severity of colonic familial
cer risk of various polyposis syndromes. Eur J Cancer adenomatous polyposis, with evidence for the importance
1995;31A:10857. of the N-acetyl transferases. Gut 2004;53:2716.
79. Gardner EJ, Plenk HP. Hereditary pattern for multiple os- 98. Burt RW, Leppert MF, Slattery ML, et al. Genetic testing
teomas in a family group. Am J Hum Genet 1952;4:316. and phenotype in a large kindred with attenuated familial
80. Gardner EJ, Richards RC. Multiple cutaneous and subcu- adenomatous polyposis. Gastroenterology 2004;127:444
taneous lesions occuring simultaneously with hereditary 51.
polyposis and osteomatosis. Am J Hum Genet 1953;5:139 99. Lynch HT, Smyrk T, McGinn T, et al. Attenuated fa-
47. milial adenomatous polyposis (AFAP). A phenotypically
81. Gardner EJ. Follow-up study of a family group exhibiting and genotypically distinctive variant of FAP. Cancer
dominant inheritance for a syndrome including intestinal 1995;76:242733.
polyposis, osteomas, fibromas and epidermal cysts. Am J 100. Lamlum H, Al Tassan N, Jaeger E, et al. Germline APC
Hum Genet 1962;14:37690. variants in patients with multiple colorectal adenomas, with
82. Bilkay U, Erdem O, Ozek C, et al. Benign osteoma with evidence for the particular importance of E1317Q. Hum
Gardner syndrome: Review of the literature and report of Mol Genet 2000;9:221521.
a case. J Craniofac Surg 2004;15:5069. 101. Soravia C, Berk T, Madlensky L, et al. Genotype-
83. Oku T, Takayama T, Sato Y, et al. A case of Gardner syn- phenotype correlations in attenuated adenomatous poly-
drome with a mutation at codon 1556 of APC: A suggested posis coli. Am J Hum Genet 1998;62:1290301.
case of genotype-phenotype correlation in dental abnor- 102. Friedl W, Meuschel S, Caspari R, et al. Attenuated familial
mality. Eur J Gastroenterol Hepatol 2004;16:1015. adenomatous polyposis due to a mutation in the 3 part of
84. Bertario L, Russo A, Sala P, et al. Multiple approach to the the APC gene. A clue for understanding the function of the
exploration of genotype-phenotype correlations in famil- APC protein. Hum Genet 1996;97:57984.
ial adenomatous polyposis. J Clin Oncol 2003;21:1698 103. Van der Luijt RB, Khan PM, Vasen HFA, et al. Germline
707. mutations in the 3 part of APC exon 15 do not result in
85. Wolf J, Jarvinen HJ, Hietanen J. Gardners dento-maxillary truncated proteins and are associated with attenuated ade-
stigmas in patients with familial adenomatosis coli. Br J nomatous polyposis coli. Hum Genet 1996;98:72734.
Oral Maxillofac Surg 1986;24:4106. 104. Brensinger JD, Laken SJ, Luce MC, et al. Variable pheno-
86. Turcot J, Despres J-P, St Pierre F. Malignant tumors of type of familial adenomatous polyposis in pedigrees with
the central nervous system associated with familial poly- 3 mutation in the APC gene. Gut 1998;43:54852.
posis of the colon: Report of two cases. Dis Colon Rectum 105. Spirio L, Olschwang S, Groden J, et al. Alleles of the
1959;2:4658. APC gene: An attenuated form of familial polyposis. Cell
87. Hamilton SR, Liu B, Parsons RE, et al. The molecular basis 1993;75:9517.
of Turcots syndrome. N Engl J Med 1995;332:83947. 106. Frayling IM, Beck NE, Ilyas M, et al. The APC variants
88. Eccles DM, Van der Luijt R, Breukel C, et al. Hereditary I1307K and E1317Q are associated with colorectal tumors,
desmoid disease due to a frameshift mutation at codon 1924 but not always with a family history. Proc Natl Acad Sci
of the APC gene. Am J Hum Genet 1996;59:1193201. USA 1998;95:107227.
89. Friedl W, Caspari R, Sengteller M, et al. Can APC mutation 107. Fearnhead NS, Wilding JL, Winney B, et al. Multiple rare
analysis contribute to therapeutic decisions in familial ade- variants in different genes account for mutlifactorial inher-
nomatous polyposis? Experience from 680 FAP families. ited susceptibility to colorectal adenomas. Proc Natl Acad
Gut 2001;48:51521. Sci USA 2004;101:159927.
90. Nugent KP, Phillips RK, Hodgson SV, et al. Phenotypic 108. Figer A, Irmin L, Geva R, et al. Genetic analysis of the
expression in familial adenomatous polyposis: Partial pre- APC gene regions involved in attenuated APC phenotype
diction by mutation analysis. Gut 1994;35:16223. in Israeli patients with early onset and familial colorectal
91. Gebert JF, Dupon C, Kadmon M, et al. Combined molec- cancer. Br J Cancer 2001;85:5236.
ular and clinical approaches for the identification of fam- 109. Gismondi V, Bonelli L, Sciallero S, et al. Prevalence of the
 Familial Adenomatous Polyposis 397

E1317Q variant of the APC gene in Italian patients with for familial adenomatous polyposis. Dis Colon Rectum
colorectal adenomas. Genet Test 2002;6:3137. 1997;40:6758.
110. Hahnloser D, Petersen GM, Rabe K, et al. The APC 129. Ambroze WL Jr, Dozois RR, Pemberton JH, et al. Famil-
E1317Q variant in adenomatous polyps and colorectal can- ial adenomatous polyposis: Results following ileal pouch-
cers. Cancer Epidemiol Biomarkers Prev 2003;12:10238. anal anastomosis and ileorectostomy. Dis Colon Rectum
111. Evertsson S, Lindblom A, Sun XF. APC I1307K and 1992;35:125.
E1317Q variants are rare or do not occur in Swedish col- 130. Barton JG, Paden MA, Lane M, et al. Comparison of post-
orectal cancer patients. Eur J Cancer 2001;37:499502. operative outcomes in ulcerative colitis and familial poly-
112. Slupska MM, Baikalov C, Luther WM, et al. Cloning and posis patients after ileoanal pouch operations. Am J Surg
sequencing a human homolog (hMYH) of the Escherichia 2001;182:61620.
coli mutY gene whose function is required for the repair of 131. Korsgen S, Keighley MRB. Causes of failure and life
oxidative DNA damage. J Bacteriol 1996;178:388592. expectancy of the ileoanal pouch. Int J Colorect Dis
113. Al-Tassan N, Chmiel NH, Maynard J, et al. Inherited vari- 1997;12:48.
ants of MYH associated with somatic G:CA mutations 132. Olsen KO, Joelsson M, Laurberg S, Oresland T. Fertility
in colorectal tumors. Nat Genet 2002;30:22732. after ileal pouch-anal anastomosis in women with ulcera-
114. Sampson JR, Dolwani S, Jones S, et al. Autosomal re- tive colitis. Br J Surg 1999;86:4935.
cessive colorectal adenomatous polyposis due to inherited 133. Olsen KO, Juul S, Berndtsson I, et al. Ulcerative colitis:
mutations of MYH. Lancet 2003;362:3941. Female fecundity before diagnosis, during disease, and af-
115. Gismondi V, Meta M, Bonelli L, et al. Prevalence of ter surgery compared with a population sample. Gastroen-
the Y165C, G382D and 1395delGGA germline muta- terology 2002;122:159.
tions of the MYH gene in Italian patients with adenoma- 134. Olsen KO, Juul S, Bulow S, et al. Female fecundity before
tous polyposis coli and colorectal adenomas. Int J Cancer and after operation for familial adenomatous polyposis. Br
2004;109:6804. J Surg 2003;90:227231.
116. Sieber OM, Lipton L, Crabtree M, et al. Multiple colorectal 135. Parc YR, Olschwang S, Desaint B, et al. Familial ade-
adenomas, classic adenomatous polyposis, and germ-line nomatous polyposis: Prevalence of adenomas in the
mutations in MYH. N Engl J Med 2003;348:7919. ileal pouch after restorative proctocolectomy. Ann Surg
117. Venesio T, Molatore S, Cattaneo F, et al. High fre- 2001;233:3604.
quency of MYH gene mutations in a subset of patients 136. Parc Y, Piquard A, Dozois RR, et al. Long-term outcome
with familial adenomatous polyposis. Gastroenterology of familial adenomatous polyposis patients after restorative
2004;126:16815. coloproctectomy. Ann Surg 2004;239:37882
118. Lipton L, Tomlinson I. The multiple colorectal adenoma 137. Iida M, Itoh H, Matsui T, et al. Ileal adenomas in post-
phenotype and MYH, a base excision repair gene. Clin colectomy patients with familial adenomatosis coli/ Gard-
Gastroenterol Hepatol 2004;2:6338. ners syndrome. Incidence and endoscopic appearance. Dis
119. Farrington SM, Tenesa A, Bernetson R, et al. Germline sus- Colon Rectum 1989;32:10348.
ceptibility to colorectal cancer due to base-excision repair 138. Van Duijvendijk P, Vasen HFA, Bertario L, et al. Cumu-
gene defects. Am J Hum Genet 2005;77:1129. lative risk of developing polyps or malignancy at the ileal
120. Croitoru M, Cleary SP, Di Nicola N, et al. Associa- pouch-anal anastomosis in patients with familial adenoma-
tion between biallelic and monoallelic germline MYH tous polyposis. J Gastrointest Surg 1999;3:32530.
mutations and colorectal cancer risk. J Natl Cancer Inst 139. Hoehner JC, Metcalf AM. Development of invasive adeno-
2004;96:16314. carcinoma following colectomy with ileoanal anastomosis
121. Bulow S. Results of national registration of familial ade- for familial polyposis coli. Report of a case. Dis Colon
nomatous polyposis. Gut 2003;52:7426. Rectum 1994;37:8248.
122. Heiskanen I, Luostarinen T, Jarvinen HJ. Impact of screen- 140. Von Herbay A, Stern J, Herfarth C. Pouch-anal cancer after
ing examinations on survival in familial adenomatous poly- proctocolectomy for familial adenomatous polyposis. Am
posis. Scand J Gastroenterol 2000;35:12847. J Surg Pathol 1996;20:9959.
123. Winawer S, Fletcher R, Rex D, et al. (U.S. Multisoci- 141. Vuilleumier H, Halkic N, Ksontini R, et al. Columnar cuff
ety Task Force on Colorectal Cancer). Colorectal can- cancer after restorative proctocolectomy for familial ade-
cer screening and surveillance: Clinical guidelines and nomatous polyposis. Gut 2000;47:7324.
rationaleUpdate based on new evidence. Gastroenterol- 142. Remzi FH, Church JM, Bast J, et al. Mucosectomy vs sta-
ogy 2003;124:54460. pled ileal pouch-anal anastomosis in patients with familial
124. Morpurgo E, Vitale G, Galandiuk S, et al. Clinical adenomatous polyposis: Functional outcome and neoplasia
characteristics of familial adenomatous polyposis and control. Dis Colon Rectum 2001;44:15906.
management of duodenal adenomas. J Gastrointest Surg 143. Palkar VM, deSouza LJ, Jagannath P, et al. Adenocarci-
2004;8:55964. noma arising in J pouch after proctocolectomy for famil-
125. Giardiello FM, Offerhaus GJA, Krush AJ, et al. The risk ial polyposis coli. Indian J Cancer 1997;34:169.
of hepatoblastoma in familial adenomatous polyposis. J 144. Ooi BS, Remzi FH, Gramlich T, et al. Anal transitional
Pediatr 1991;119:7668. zone cancer after restorative proctocolectomy and ileoanal
126. Nugent KP, Phillips RK. Rectal cancer risk in older pa- anastomosis in familial adenomatous polyposis. Report of
tients with familial adenomatous polyposis and an ile- two cases. Dis Colon Rectum 2003;46:141823.
orectal anastomosis: A cause for concern. Br J Surg 145. Pasricha PJ, Bedi A, OConnor K, et al. The effects of sulin-
1992;79:12046. dac on colorectal proliferation and apoptosis in familial
127. De Cosse JJ, Bulow S, Neale K, et al. Rectal cancer risk in adenomatous polyposis. Gastroenterology 1995;109:994
patients treated for familial adenomatous polyposis. The 8.
Leeds Castle Polyposis Group. Br J Surg 1992;79:1372 146. Waddell WR, Loughry RW. Sulindac for polyposis of the
5. colon. J Surg Oncol 1983;24:837.
128. Milsom JW, Ludwig KA, Church JM, et al. Laparoscopic 147. Waddell WR, Ganser GF, Cerise EJ, et al. Sulindac for
total abdominal colectomy with ileorectal anastomosis polyposis of the colon. Am J Surg 1989;157:1759.
398 Galiatsatos and Foulkes

148. Labayle D, Fischer D, Vielh P, et al. Sulindac causes re- papillectomy for tumors of the duodenal papillae. Gastroin-
gression of rectal polyps in familial adenomatous polypo- test Endosc 2004;60:75764.
sis. Gastroenterology 1991;101:6359. 159. Norton ID, Geller A, Petersen BT, et al. Endoscopic
149. Giardiello FM, Hamilton SR, Krush AJ, et al. Treatment surveillance and ablative therapy for periampullary ade-
of colonic and rectal adenomas with sulindac in familial nomas. Am J Gastroenterol 2001;96:1016.
adenomatous polyposis. N Engl J Med 1993;328:1316. 160. Kadmon M, Tandara A, Herfarth C. Duodenal adenomato-
150. Giardiello FM, Offerhaus JA, Tersmette AC, et al. Sulin- sis in familial adenomatous polyposis coli: A review of the
dac induced regression of colorectal adenomas in familial literature and results from the Heidelberg Polyposis Reg-
adenomatous polyposis: Evaluation of predictive factors. ister. Int J Colorectal Dis 2001;16:6375.
Gut 1996;38:57881. 161. De Vos tot Nederveen Cappel WH, Jarvinen HJ, Bjork
151. Tonelli F, Valanzano R, Messerini L, et al. Long-term treat- J, et al. Worldwide survey among polyposis registries of
ment with sulindac in familial adenomatous polyposis: Is surgical management of severe duodenal adenomatosis in
there an actual efficacy in prevention of rectal cancer? J familial adenomatous polyposis. Br J Surg 2003;90:705
Surg Oncol 2000;74:1520. 10.
152. Cruz-Correa M, Hylind LM, Romans KE, et al. Long-term 162. Penna C, Bataille N, Balladur P, et al. Surgical treatment of
treatment with sulindac in familial adenomatous polyposis: severe duodenal polyposis in familial adenomatous poly-
A prospective study. Gastroenterology 2002;112:6415. posis. Br J Surg 1998;85:6658.
153. Phillips RK, Wallace MH, Lynch PM, et al. A randomised, 163. Gallagher MC, Shankar A, Groves CJ, et al. Pylorus-
double blind, placebo controlled study of celecoxib, a se- sparing pancreaticoduodenectomy for advanced duode-
lective cyclooxygenase 2 inhibitor, on duodenal polyposis nal disease in familial adenomatous polyposis. Br J Surg
in familial adenomatous polyposis. Gut 2002;50:85760. 2004;91:115764.
154. Hallak A, Alon-Baron L, Shamir R, et al. Rofecoxib re- 164. American Society of Clinical Oncology (ASCO) policy
duces polyp recurrence in familial polyposis. Dig Dis Sci statement update: Genetic testing for cancer susceptibility.
2003;48:19982002. J Clin Oncol 2003;21:23973406.
155. Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular 165. Giardiello FM, Brensinger JD, Petersen GM, et al. The
events associated with rofecoxib in a colorectal adenoma use and interpretation of commercial APC gene test-
chemoprevention trial. N Engl J Med 2005;352:1092102. ing for familial adenomatous polyposis. N Engl J Med
156. Solomon SD, McMurray JJ, Pfeffer MA, et al. Car- 1997;336:8237.
diovascular risk associated with celecoxib in a clinical 166. Trimbath JD, Giardiello FM. Review article: Genetic test-
trial for colorectal adenoma prevention. N Engl J Med ing and counselling for hereditary colorectal cancer. Ali-
2005;352:107180. ment Pharmacol Ther 2002;16:184357.
157. Mlkvy P, Messmann H, Debinski H, et al. Photodynamic 167. Wong N, Lasko D, Rabelo R, et al. Genetic counseling
therapy in familial adenomatous polyposisA pilot study. and interpretation of genetic tests in familial adenomatous
Eur J Cancer 1995;31A:11605. polyposis and hereditary nonpolyposis colorectal cancer.
158. Cheng CL, Sherman S, Fogel EL, et al. Endoscopic snare Dis Colon Rectum 2001;44:2719.