In 1988, Lynch and colleagues [7] first described a cancer-prone family with many members who had fewer than 100 adenomas. The adenomas were flat in appearance and were located primarily in the right side of the colon. After molecular examinations of this family and six others with similar characteristics were done [8-10], a mutation was found in the adenomatous polyposis coli gene. This variant of familial adenomatous polyposis was named attenuated adenomatous polyposis coli by Spirio and colleagues [11]. The clinical characteristics of this attenuated variant include few adenomas, marked phenotypic variation within families, and onset of colorectal cancer occurring approximately 15 years later than in classic familial adenomatous polyposis but 10 years earlier than in sporadic colorectal cancer [10, 12]. When studying seven families that had an attenuated phenotype, Spirio and colleagues [11] found four different mutations of the adenomatous polyposis coli gene. These mutations were frame-shifts or changes to single base pairs, and they predicted truncated gene products (proteins) that are similar to the mutations that are found in classic familial adenomatous polyposis. However, these truncations were located very close to the 5' end of the gene in codons 85 to 157. With data from three other families that were studied by Olschwang and coworkers [13] and Fodde and associates [14], Spirio and colleagues concluded that mutations that were proximal to codon 158 predicted the attenuated phenotype whereas mutations that were distal to codon 167 were associated with classic familial adenomatous polyposis [10] (Figure 1). However, the clinical features of families that have mutations at various sites in the 5' region of the adenomatous polyposis coli gene have not been compared in detail.

View larger version (11K): [in this window] [in a new window]   Figure 1. The adenomatous polyposis coli gene. Codon location of mutations in proximal 5' and distal 5' families is shown.

In this study, we examined the phenotype of seven families with mutations in the adenomatous polyposis coli gene that occurred toward the 5' end, relative to codon 158 (proximal 5' families). These findings were compared with those of seven families with mutations that were immediately downstream from codon 158, in codons 179 to 625 (distal 5' families).

Methods

Top Methods Results Discussion Glossary Author & Article Info References

Patient Evaluation

At the time of this study, The Johns Hopkins Polyposis Registry contained information from 340 families with familial adenomatous polyposis that was initially gathered from a five-state region in the United States beginning in 1972. After informed consent was obtained, at least one member with familial adenomatous polyposis from each of 112 families was evaluated for a mutation of the adenomatous polyposis coli gene. These 112 families were the first to be consecutively enrolled in the Johns Hopkins Registry that had an affected member who was available for genetic testing. Genotype was analyzed in leukocyte DNA, RNA, or both from peripheral blood. The analysis was done by the in vitro synthesized protein (IVSP) assay, by cloning and sequencing the entire coding region of the gene, or by both methods as described elsewhere [15-17]. (The in vitro synthesized protein assay has replaced DNA sequencing for routine detection of adenomatous polyposis coli mutations and is the only test that is clinically available to detect mutations of the adenomatous polyposis coli gene.)

Medical and family information was obtained from patients through a standard registry questionnaire and a review of medical records. The clinical diagnosis of familial adenomatous polyposis in a family was verified by the clinical and pathologic criterion of 100 or more colorectal adenomas in a phenotypically affected member [5]. Phenotypically affected members of a family were assumed to have the same mutation as the affected member who was analyzed. All data were entered into computer data-bases using dBase (Borland, Scotts Valley, California) and pedigree analysis software (Cyrillic, Cherwell Scientific Publishing, Oxford, United Kingdom).

The following clinicopathologic variables were analyzed for each family member: sex; age at diagnosis of familial adenomatous polyposis; number of colorectal adenomas that were found at the first colorectal examination, recorded as fewer than 100 polyps or at least 100 polyps (polyposis); number of colorectal adenomas that were found at the first colorectal examination when the patient was 35 years of age or younger (95% of patients with familial adenomatous polyposis have more than 100 adenomas by 36 years of age [6]); predominant anatomical distribution of polyps, recorded as right-sided (proximal to splenic flexure), left-sided (distal to splenic flexure), or uniform; age at diagnosis of colorectal cancer; and anatomical location of colorectal cancer.

Statistical Analysis

Differences in the number of polyps found at first examination and the number of polyps found during examination when patients were younger than 36 years of age were evaluated by using the chi-square test. Differences in mean age were evaluated by using a t-test. Analysis for intrafamilial clustering of phenotype was done by using nested analysis of variance. The rate of development of colorectal cancer was compared between groups by Kaplan-Meier analysis with log-rank tests. A P value less than 0.05 was considered statistically significant. The True Epistat software package (Epistat Services, Richardson, Texas) was used for all analyses.

Results

Top Methods Results Discussion Glossary Author & Article Info References

Mutations in the Adenomatous Polyposis Coli Gene in 5' Families

Of 112 families with familial adenomatous polyposis and known mutations of the adenomatous polyposis coli gene, 7 (6%) were found to have a mutation toward the 5' end of the adenomatous polyposis coli gene relative to codon 158. Seven others had a mutation downstream from codon 158, in codons 179 to 625. The site of mutation in proximal 5' and distal 5' families is shown in Table 1 and Figure 1. Point and frameshift mutations were noted.

We evaluated the possibility of common ancestry by searching for surname, date of birth, and country of origin of the oldest members of all 14 families by using the extensive genealogy records maintained by the Church of Jesus Christ of Latter-Day Saints. No common ancestry could be established in the families described.

Clinical Features of Proximal 5' and Distal 5' Families

The clinical features of the families are described in Table 2. Demographic characteristics, mean age at diagnosis of familial adenomatous polyposis, age at first examination of the colon, and age at diagnosis of familial adenomatous polyposis in the proband of each family did not significantly differ between the proximal 5' and distal 5' families. Nested analysis of variance for age at diagnosis of familial adenomatous polyposis grouped by proximal 5' or distal 5' site and by family also showed no evidence of intrafamilial clustering (F = 1.25; P = 0.27). The intraclass coefficient (r = 0.049) reflects a modest contribution of intrafamilial variation to the overall variance. In contrast, the colonic phenotypes differed between proximal 5' and distal 5' families. Eight of 17 phenotypically affected patients (47%) in proximal 5' families were found to have fewer than 100 adenomas during their first colorectal examination; only 9 of 48 patients (19%) in distal 5' families had similar findings (P = 0.029). Seven of 12 patients in proximal 5' families who were 35 years of age or younger had fewer than 100 polyps compared with 7 of 39 (18%) patients in distal 5' families (P = 0.010). In family 3, two patients who had a mutation of the adenomatous polyposis coli gene at codon 140 had no adenomas found by colonoscopy done in the third decade of life (when the disease is usually evident). Nevertheless, in proximal 5' families, all but one patient who were older than 45 years of age had polyposis; in six of seven families, at least one member had the phenotype for classic familial adenomatous polyposis. Thus, the phenotype was heterogenous in the family rather than attenuated in all of its members.

In proximal 5' families, polyp distribution was predominantly right-sided in 4 patients (24%), uniform in 10 patients (59%), and left-sided in 3 patients (18%). In contrast, polyp distribution in distal 5' families was right-sided in 3 patients (6%), uniform in 44 patients (94%), and left-sided in no patients (P < 0.001) (Table 2).

Six of 20 (30%) phenotypically affected patients in four proximal 5' families had colorectal cancer compared with 18 of 53 patients (34%) in distal 5' families (Table 3). The mean age (±SD) at diagnosis of colorectal cancer in patients in proximal 5' families was 51 ± 16 years (range, 28 to 75 years). This value was somewhat greater than the mean age in patients in distal 5' families (39 ± 14 years [range, 23 to 71 years]), but the difference was not statistically significant. However, Kaplan-Meier analysis done with a log-rank test showed a statistically significant difference in the cumulative probability of survival without colorectal cancer between the two types of families (P = 0.041) (Figure 2). Patients in proximal 5' families had a higher rate of cancer-free survival. The occurrence of colorectal cancer was related to the number of polyps in patients in the proximal 5' families: Five of the six cases of cancer developed in patients who had the phenotype for classic familial adenomatous polyposis, whereas cancer developed in one patient who had only 20 polyps at 43 years of age.

View larger version (18K): [in this window] [in a new window]   Figure 2. Life table analysis of onset of colorectal cancer in proximal 5' and distal 5' families. Onset was delayed in patients in the proximal 5' families (P = 0.041 by log-rank test). The top row of numbers above the x-axis shows how many patients were being observed in the proximal 5' families, and the bottom row shows the corresponding numbers for the distal 5' families.

Discussion

Top Methods Results Discussion Glossary Author & Article Info References

Our study shows that patients who have a proximal 5' mutation of the adenomatous polyposis coli gene have fewer polyps found on the first colorectal examination than do patients from distal 5' families, particularly if the evaluation is done when the patient is 35 years of age or younger. Notably, two persons in family 3 who were genotypically affected with mutations at codon 140 (proximal 5' mutations) were phenotypically free of polyps when they were 29 and 31 years of age, respectively. Because the average age at diagnosis of familial adenomatous polyposis in our registry is 15 years [6], a delay in the development of polyposis in some of the members of 5' families is evident. However, most patients in proximal 5' families had polyposis at an older age, and some presented as young adults with the phenotype of classic familial adenomatous polyposis. These findings are similar to those first reported by Lynch and colleagues [7] and Spirio and colleagues [11] and emphasize that proximal 5' families represent an exaggerated heterogeneity of the colorectal phenotype that is seen in distal 5' families [18].

Spirio and coworkers [10] concluded that mutations that are 5' to codon 158 predict the attenuated phenotype, whereas mutations that are 3' to codon 167 are associated with classic familial adenomatous polyposis. In our study, seven families had mutations that were 5' to codon 158; six of these families had at least one member who had classic polyposis, and five of the seven had a member who had fewer than 100 adenomas. Family 8 had a mutation at codon 179, and phenotypically affected members had classic polyposis. The exact molecular site of demarcation and the mechanistic explanation for the attenuated phenotype awaits examination of additional families and clarification of the function of the adenomatous polyposis coli gene product and its domains.

The distribution of polyps in the colon was also affected by the site of the mutation. In most patients in the distal 5' families, polyps were evenly distributed throughout the colorectum, as is typically noted in familial adenomatous polyposis [19]. However, only 50% of phenotypically affected patients in the proximal 5' families had uniform polyp distribution. Two patients in families 4 and 5 were of particular concern because both had fewer than 100 right-sided polyps, thus precluding the clinical diagnosis of familial adenomatous polyposis by sigmoidoscopy (the method that is usually used for screening patients who are at risk). Our findings are consistent with preliminary results from Burt and colleagues' investigation of one large family with attenuated adenomatous polyposis [20].

The mean age at diagnosis of colorectal cancer for patients in our proximal 5' families was older (51 years) than that for patients in distal 5' families (39 years). Kaplan-Meier analysis also showed later development of colorectal cancer in proximal 5' families, which is consistent with the previous reports of Lynch and colleagues [7] and Spirio and colleagues [11]. However, colorectal cancer was diagnosed at young ages (28, 32, and 43 years) in several patients in our proximal 5' families. This further emphasizes the clinical heterogeneity in these families. The presence of many adenomas in members of attenuated families seemed to confer an increased risk for colorectal cancer. Nevertheless, in one patient in a proximal 5' family who had only 20 colorectal adenomas, colon cancer was diagnosed at 43 years of age in the absence of classic polyposis.

The phenotypic variability that we saw in families that had a proximal 5' mutation of the adenomatous polyposis coli gene carries several clinical implications, although several limitations of our study should be considered. First, the total number of patients that we investigated was small because of the infrequent occurrence of proximal 5' families (6% in our registry). Second, although the examinations did not show evidence of intrafamilial clustering of phenotype, the small number of cases hampered a full statistical analysis of this issue. Third, colonic surveillance of patients was not systematic or regular. Thus, the biology and clinical sequence of polyp development in proximal 5' families are unknown. Only sequential colonoscopy of genotypically positive patients for a period of several decades can fully define colorectal phenotypic expression in these patients.

Despite these limitations, our data argue for routine genotyping of one phenotypically affected member of a family with familial adenomatous polyposis to confirm the diagnosis of the disorder and to evaluate the site of the mutation of the adenomatous polyposis coli gene. The results of such evaluation may affect the timing and method of surveillance. Patients at risk who are in families that are found to have a mutation of the adenomatous polyposis coli gene 5' to codon 158 should have colonoscopy because sigmoidoscopy may not identify the manifestations of this syndrome. In contrast, patients in families that have distal 5' adenomatous polyposis coli mutation have uniform polyposis, which can generally be detected at lower cost with sigmoidoscopy.

Although the development of colorectal cancer is often delayed for patients in proximal 5' families, the lifetime risk seems to approach 100% and cancer can occur in the absence of typical polyposis. These findings argue for colectomy if multiple adenomas are present in members of these families. Finally, patients who have multiple adenomas on a sporadic basis may be phenotypically indistinguishable from patients who have the attenuated variant of familial adenomatous polyposis. Therefore, a patient who presents with multiple adenomas at a young age or who is a member of a family that has more than one person with multiple adenomas should be considered for genetic testing to evaluate the possibility of familial adenomatous polyposis.

Glossary

Top Methods Results Discussion Glossary Author & Article Info References

5'-3': The convention in which nucleic acid sequences are written in the 5' to 3' direction-that is, from the 5' terminus at the left to the 3' terminus at the right. At the 5' end, the terminal nucleotide has a free 5' group; the terminal nucleotide at the other end has a free 3' group.

Allele: One of several alternate forms of a gene that occupy a given locus on a chromosome.

Codon: A triplet of nucleotides that represents an amino acid or a termination signal to end the gene product.

Distal 5' families: Families in which the adenomatous polyposis coli gene mutation is in the 5' end of the gene but is downstream from codon 158.

Downstream: Sequences that proceed farther in the direction of gene reading or expression (in other words, sequences that move away from the 5' end of the gene).

Exon: Any segment of an interrupted gene that is represented in the mature RNA product and encodes the protein product.

Frameshift mutations: Deletions or insertions that are not a multiple of three base pairs and thus change the frame in which triplets are translated into protein.

Germline: The sperm and egg cells that have only a single set of chromosomes.

Genotype: The full genetic constitution of an intact organism. For any particular gene, the particular pair of alleles that a person possesses at a gene locus. One of these alleles is inherited from the mother, and the other is inherited from the father.

In vitro synthesized protein (IVSP) assay: A laboratory test for gene mutations in which DNA from peripheral blood leukocytes is transcribed into RNA and translated into protein. Results that show a shortened length of the adenomatous polyposis coli gene protein identify most mutations of the gene.

Mutation: A change in the nucleotide sequence of DNA.

Point mutations: Changes involving single base pairs.

Phenotype: The appearance or other characteristics of an organism that result from the expression of the genotype. The same genotype may be expressed differently from one person to the next as a result of differences at other gene loci or in the environment.

Proximal 5' families: Families in which the gene mutation is in the 5' end of the gene but is upstream from codon 158.

Sporadic adenomas: Dysplastic polyps that are not caused by any known germline genetic mutation and are therefore not inherited.

Truncated adenomatous polyposis gene product: The adenomatous polyposis coli protein of limited length that indicates that a mutation of the adenomatous polyposis gene is present. This result can be found when genomic DNA is analyzed for mutations of the adenomatous polyposis gene by the in vitro synthesized protein assay or when the DNA is sequenced and an altered codon is found within the exon.

Upstream: Sequences that proceed in the opposite direction from the gene reading or expression (in other words, sequences that move toward the 5' end of the gene).

From The Johns Hopkins Hospital, The Johns Hopkins Hereditary Colorectal Cancer Registry, The Johns Hopkins School of Public Health, and The Johns Hopkins University School of Medicine, Baltimore, Maryland; Myriad Genetics, Salt Lake City, Utah; LabCorp, Research Triangle Park, North Carolina; and The Elkhart Clinic, Elkhart, Indiana.

Ms. Brensinger, Ms. Krush, Ms. Bacon, and Ms. Booker: The Johns Hopkins Hereditary Colorectal Cancer Registry, 550 North Broadway, Suite 108, Baltimore, MD 21205.

Dr. Luce: Myriad Genetics, 421 Wakara Way, Salt Lake City, UT 84108.

Dr. Petersen: Department of Epidemiology, Room 6509, Johns Hopkins School of Public Health, 615 North Wolfe Street, Baltimore, MD 21205.

Mr. Cayouette: LabCorp, 1912 Alexander Drive, Research Triangle Park, NC 27709.

Dr. Bufill: The Elkhart Clinic, 303 South Nappanee Street, Elkhart, IN 46514.

Dr. Hamilton: Division of Gastrointestinal/Liver Pathology, Department of Pathology, Ross Research Building, Room 632, The Johns Hopkins University School of Medicine, 720 Rutland Avenue, Baltimore, MD 21205.