Ileorectal anastomosis is appropriate for a subset of patients with familial adenomatous polyposis

GASTROENTEROLOGY 2000;119:1454 –1460

Ileorectal Anastomosis Is Appropriate for a Subset of Patients With Familial Adenomatous Polyposis ¨ LOW,* HANS VASEN,‡ HEIKKI JA¨RVINEN,§ JAN BJO ¨ RK,储 MARIE LUISE BISGAARD,*,¶ CHARLOTTE BU ¨ LOW* and STEFFEN BU *Danish Polyposis Register, Hvidovre University Hospital, Copenhagen, Denmark; ‡Dutch Polyposis Register, Foundation for the Detection of Hereditary Tumours, Leiden University Hospital, Leiden, The Netherlands; §Finnish Polyposis Register, Helsinki University Hospital, Helsinki, Finland; 储Swedish Polyposis Register, Department of Gastroenterology and Hepatology, Karolinska Hospital, Stockholm, Sweden; and ¶Department of Clinical Biochemistry, Rigshospitalet, Copenhagen, Denmark

Background & Aims: This study reevaluates the risk of rectal cancer and the frequency of subsequent proctectomy for nonmalignant causes in patients with familial adenomatous polyposis (FAP) who have undergone colectomy with ileorectal anastomosis (IRA). Potential risk factors for rectal cancer in this setting are also examined, and recommendations for the choice of surgical procedure are made. Methods: The national polyposis registries in Denmark, Finland, The Netherlands, and Sweden included 659 patients undergoing surgery with IRA in 1940 –1997. Kaplan–Meier analysis and Cox regression analysis were performed to evaluate cumulative risk, survival, and predictive risk factors. Results: Rectal carcinoma was diagnosed in 47 patients, with a cumulative 40-year risk of 0.32. The cumulative risk according to chronologic age was 0.30 at age 60, and higher in patients undergoing surgery above age 25 (P ⴝ 0.0016). Chronologic age was the only independent risk factor (P ⴝ 0.0016). The cumulative 5-year survival rate after rectal carcinoma was 0.60. The apc mutation was known in 167 patients, of whom 7 had rectal cancer. The cumulative 40-year risk of secondary proctectomy was 0.70, and higher in patients with a mutation in codon 1250 –1500 than outside this region (P ⴝ 0.005). However, all 7 rectal cancers were found in the latter group. None of the 18 patients with attenuated FAP (mutation in codon 0 –200 or >1500) had a secondary proctectomy. Conclusions: IRA is recommended in (1) young patients with few rectal adenomas and a family history of a mild phenotype and (2) patients with attenuated FAP (a mutation in codon 0 –200 or >1500), provided there is acceptance of life-long rectal surveillance. Patients with many rectal polyps and/or a family history of severe polyposis should be offered a restorative proctocolectomy with an ileal pouch–anal anastomosis.

urgery is the definitive treatment in familial adenomatous polyposis (FAP). Over the past decades it has been discussed whether the preferred surgical method

S

should be colectomy with ileorectal anastomosis (IRA) or a restorative proctocolectomy with ileal pouch–anal anastomosis (IPAA), and a recent survey showed the spectrum of opinion even among those responsible for management of the disease.1 Both methods have their advantages and disadvantages: IRA is a one-step procedure with few complications and good function, but implies life-long regular endoscopic follow-up and a 13%–25% cumulative risk of rectal cancer after 15–25 years despite surveillance.2– 6 A recent study indicated that the risk increases with chronologic age, rather than with time after the surgery, thereby suggesting that a conversion to a pouch should be considered in older patients.7 The crude 5-year survival rate after metachronous rectal cancer is 68%,4 which is better than in patients with sporadic rectal cancer, probably because of early tumor stage at diagnosis. Finally, there is a 44%–74% cumulative risk after 20 years of secondary proctectomy caused by increasing rectal polyposis.4,5,8 IPAA has an 18% cumulative risk of adenoma formation at the anastomosis or in the pouch after 7 years,9 but almost no risk of rectal cancer. The procedure most often implies a 2-stage procedure, including a temporary ileostomy. The results from comparative studies of complication rates and functional results after IRA and IPAA in FAP show similar results, but reveal interstudy differences.10,11 Until now, the choice of procedure has depended primarily on the phenotype of the patient (i.e., the number of rectal adenomas and the degree of dysplasia). A recent Dutch study has, however, indicated that genotype of the patient might influence future surgical Abbreviations used in this paper: FAP, familial adenomatous polyposis; IPAA, ileal pouch–anal anastomosis; IRA, ileorectal anastomosis. © 2000 by the American Gastroenterological Association 0016-5085/00/$10.00 doi:10.1053/gast.2000.20180

December 2000

ILEORECTAL ANASTOMOSIS

1455

Figure 1. Mutational spectrum in 167 Danish and Dutch patients. (■), Proctectomy because of rectal carcinoma; ( ) proctectomy because of uncontrollable polyposis.

decision making because IRA patients with a mutation after codon 1250 in the apc gene had a higher risk of secondary proctectomy than patients with a mutation before codon 1250.8 The aim of this study of patients with IRA was to reevaluate the cumulative risk of rectal cancer and of secondary proctectomy for nonmalignant causes and to identify possible risk factors, leading to recommendations for the choice of surgical procedure.

Materials and Methods The national polyposis registries of Denmark, Finland, Holland, and Sweden were searched for patients undergoing surgery from January 1, 1940, to December 31, 1997, with a colectomy and an ileorectal anastomosis as the primary procedure. Patients in regular follow-up with at least one annual proctoscopy were included in the study. The registered data included date of birth, sex, proband/ call-up case, familial/isolated case, location of mutation in the apc gene, date of IRA, colon cancer at IRA, date of exit of the study, status at exit (alive, dead, proctectomy, lost to followup), indication for proctectomy (rectal carcinoma, severe polyposis, functional problems, other causes), date of last proctoscopy before metachronous rectal carcinoma, date of death, and cause of death. Probands are polyposis patients diagnosed because of bowel symptoms and without any knowledge of the hereditary disease; call-up cases are first-degree relatives of probands diagnosed as a result of prophylactic examination. In the statistical analysis, the Kaplan–Meier estimate was chosen to calculate cumulative survival and incidence rates, differences being analyzed by the log rank test. The ␹2 test was used to compare proportions of events, and Cox regression analysis was used to identify independent predictive factors. A statistical significance level of 0.05 was chosen.

at IRA was 26 years (range, 7–75 years). Of them, 193 were probands (29%) and 418 (63%) were call-up patients, and the status proband/call-up case was unknown in 48 patients. Seventy-five patients (11%) were isolated cases with no family history of FAP. The specific apc gene mutation was known in 167 of 366 (46%) Dutch and Danish patients (Figure 1), whereas results of mutation analysis were not available in patients from Finland and Sweden. Forty-seven patients (7%) developed rectal carcinoma after a median follow-up of 11 years (range, 1–57 years) after IRA. The median age of development of rectal cancer was 42 years (range, 27– 62 years), and 16 (34%) had rectal cancer after the age of 50. The cumulative incidence rate by time was 0.04 after 10 years, increasing to 0.32 after 40 years (Figure 2). The cumulative risk of a secondary rectal cancer according to age showed a risk of 0.03 at age 20, increasing to 0.18 at age 50, and 0.30 at age 60 (Figure 3). The cumulative risk of rectal cancer after 30 years was 0.12 in patients undergoing surgery at an age ⬍25 years vs. 0.23 in patients undergoing surgery at an age ⬎25 years (P ⫽ 0.0016). The cumulative crude

Results The study consisted of 659 patients, 349 men (53%) and 310 women, from Denmark (126), Finland (105), Holland (240), and Sweden (188). The median age

Figure 2. Cumulative incidence rate of rectal cancer after IRA by time.

1456

¨ LOW ET AL. BU

Figure 3. Cumulative incidence rate of rectal cancer after IRA by chronologic age.

5-year survival rate after rectal carcinoma was 0.60 (Figure 4). The only significant and independent risk factor for development of metachronous rectal cancer was chronologic age (P ⫽ 0.0016). None of the following factors had any influence: years after colectomy, sex, proband/ call-up status, familial/isolated case, colon cancer at IRA, or location of mutation, but the mutation was only known in 7 of 47 patients with metachronous rectal carcinoma. A secondary proctectomy was performed for nonmalignant causes in 137 patients (21%): uncontrollable rectal polyposis in 109 (80%), functional problems in 6 (4%), patient wishes in 10 (7%), and unknown in 12 (9%). The rectum was left in situ in 475 patients. The risk of subsequent proctectomy caused by all causes (including rectal cancer) was 0.70 after 40 years. The risk

Figure 4. Survival rate after rectal cancer.

GASTROENTEROLOGY Vol. 119, No. 6

Figure 5. Cumulative risk of proctectomy caused by uncontrollable polyposis by time.

was higher in patients with a mutation in codon 1250 – 1500 than outside this region (P ⫽ 0.005). However, all 7 patients with a secondary rectal carcinoma and a known mutation were found in the latter group (Figure 1). The cumulative 40-year risk after secondary proctectomy caused by uncontrollable polyposis was 0.57 (Figure 5); this risk was also higher in patients with a mutation in codon 1250 –1500 than outside this region (P ⫽ 0.009). None of the 18 patients with attenuated FAP with a mutation either below codon 200 or above codon 1500 had a secondary proctectomy.

Discussion The choice of surgical method in FAP has remained controversial over the past 30 years, primarily because of different results on the long-term risk of secondary rectal cancer. The first surgical method was total proctocolectomy with ileostomy, which was gradually replaced by colectomy and IRA during the 1940s. IRA is a one-step procedure with almost no mortality, low postoperative morbidity, and good functional results, but the patients are left with their rectum in situ, and the risk of developing a secondary rectal carcinoma means that life-long regular rectal surveillance is mandatory. Over the following 20 years, IRA became a well-established method that was used in most patients without primary rectal cancer. In 1975, Bussey12 showed a low overall risk (2%) of secondary rectal carcinoma development in 89 patients at St. Mark’s Hospital, and 10 years later he showed a cumulative rectal cancer risk of 13% 25 years after IRA.3 In 1971, however, Moertel et al.13 published disturbing results in 143 IRA patients from the Mayo Clinic with a 59% cumulative risk of rectal cancer after 23 years. A subsequent report 10 years

December 2000

later on the same patient series confirmed these findings and showed a cumulative rectal cancer risk of 55% after 30 years.14 The Mayo Clinic study has been subjected to severe criticism concerning definition of FAP, surgical technique (ileosigmoidostomy rather than IRA in more than half of the patients), and insufficient follow-up.3 In the following years several reports showed less pessimistic results: the cumulative risk of rectal cancer after 20 –25 years was 12%–15% in studies from St. Mark’s Hospital (N ⫽ 174), the Cleveland Clinic (N ⫽ 133), and Scandinavia (N ⫽ 294).2,4,15 On the other hand, 2 studies from Japan (N ⫽ 320) and Finland (N ⫽ 100) revealed a 25%–37% cumulative rectal cancer risk after 20 years5,6; the patients of the latter study are included in the present patient series. The introduction in the beginning of the 1980s of the restorative proctocolectomy with an IPAA implied a total removal of the colorectal mucosa, which meant a minimal risk of secondary rectal cancer, and thereby a reduced need for close surveillance. These advantages compared with IRA must, however, be balanced against a 2-stage procedure, including a temporary ileostomy in most cases, higher risk of postoperative complications, longer accommodation period, and a potential difference in long-term functional results. Over the past 2 decades, IPAA has gradually gained increasing popularity. Recent studies generally show a somewhat higher rate of postoperative complications after IPAA than after IRA, but little difference in functional results.10,11,16 –19 Only 4 reports describe carcinoma formation after IPAA,20 –23 but a recent study showed a cumulative 18% risk of secondary adenoma formation in the pouch-anal anastomosis after 7 years.9 Also, there are reports indicating that the long-term risk of pouch adenomas is not negligible.24 –26 This means that patients with IPAA must be followed up postoperatively, and it is recommended that surveillance includes annual digital examination and endoscopy.9 The risk of pouchitis is much lower than in patients with colitis,27 and the risk of sexual dysfunction after IPAA (impotence or retrograde ejaculation), which is only 1%–2% in patients with ulcerative colitis,28 is probably even lower in FAP.10,11,18 IPAA can be performed as a secondary procedure after a previous IRA with similar results for morbidity and function as after a primary IPAA.29 IPAA is also an acceptable procedure in patients with a primary carcinoma in the colon or upper rectum, which can be resected for cure,30 or combined with a total mesorectal excision for a carcinoma in the lower two thirds of the rectum.31 IPAA can even be performed laparoscopically.32

ILEORECTAL ANASTOMOSIS

1457

Risk factors of future development of secondary rectal cancer are sparsely known. A recent study showed that a high number of colorectal adenomas and older age are associated with a higher risk of colorectal cancer in FAP in general.33 In a study of 224 patients with IRA, Nugent and Phillips7 showed that the chronologic age rather than the length of the postoperative follow-up is an important risk factor. The cumulative risk of rectal cancer was only 10% in patients younger than 50 years, but increasing steeply to 29% by age 60. The investigators concluded that either surveillance of the rectum in older patients must be improved or a secondary IPAA should be considered. In 1996, Vasen et al.8 published a study on 87 IRA patients in whom the pathogenic apc gene mutation had been identified. In patients with a mutation above codon 1250, the risk of secondary proctectomy was significantly higher than in patients with a mutation before this codon. The investigators stated that, in the future, molecular genetic testing may be used as a guide in the choice of surgical method, and they suggested that IRA should be the preferred treatment in patients with mutations before codon 1250, and IPAA in those with mutations after this codon.8 The choice of boundary at codon 1250 used by Vasen et al.8 was later criticized because of the inclusion of patients with a mutation after codon 1440 in the high-risk group of secondary proctectomy, because such patients have been reported to show a mild phenotype similar to patients with mutations at the extreme 5⬘-end of the gene.34 The present Scandinavian-Dutch series of 659 patients is based on 4 national polyposis registries established decades ago and with a high completeness of registration, which enables long-term results on an epidemiologically solid basis. In Table 1, our present findings are compared with the results of major studies in the literature. We show an overall frequency of secondary rectal cancer of 7%, which is similar to the results from St. Mark’s Hospital and the Cleveland Clinic,2,7,15 and only half of the Japanese figures.6 The median age at IRA was 26 years in our study, which is almost the same as in the Anglo-American studies (24 –29 years) but lower than in Japan. This difference may partly account for the difference in overall frequency of rectal cancer. The age at development of rectal cancer in this study was 42 years, compared with 44 years and 48 years in the St. Mark’s study and the Japanese series, respectively.6,7 The cumulative incidence rate of metachronous rectal cancer in the present series was 4%, 12%, 17%, and 32% after 10, 20, 30, and 40 years of follow-up, respectively. Bussey et al.2 found rates of 7%, 10%, and 13% after 10, 20, and 25 years, respectively, which correspond well to our find-

1458

¨ LOW ET AL. BU

GASTROENTEROLOGY Vol. 119, No. 6

Table 1. Major Series on IRA From the Literature Studies

Year No. of patients Age at IRA ( yr) Rectal cancer (%) Interval from IRA to rectal cancer ( yr) Dukes stage (A⫹B/all) Age at rectal cancer ( yr) Cumulative risk of rectal cancer by time (%)

Cumulative risk of rectal cancer by age (%) 5-yr cumulative survival after rectal cancer (%) Secondary proctectomy (%) Cumulative risk of proctectomy for polyposis by time (%)

Bu¨low3

Sarre et al.15

Bussey et al.2

Nugent et al.7

DeCosse et al.4

Iwama and Mishima6

Vasew et al.8

Heiskanen and Ja¨rvinen5

Present study

1984 58 25 5 5

1984 133 29 13 10

1985 174

1992 224 26 10 14

1992 297 27 4

1994 322 34 14 10

1996 225 28 7

1997 100 32 9 8

659 26 7 11

2/3 46 10 yr: 15

7/10 41 10 yr: 4 20 yr: 12

6/11 42 10 yr: 7 20 yr: 10 25 yr: 13 30 yr: 26

6

100

73

9

3

13/22 48 10 yr: 5 20 yr: 9 25 yr: 13 50 yr: 10 60 yr: 29 78

ings. On the other hand, Iwama and Mishima6 presented cumulative rates of 13% and 24% after 10 and 15 years, thus indicating a higher risk in Japan. In the present study, the cumulative incidence rate of rectal cancer showed a statistically significant increase with age, especially after the age of 50 years, which is similar to the findings in the St. Mark’s study.19 In the St. Mark’s study, age at IRA was not a statistically significant predictor of secondary rectal cancer, whereas we found a higher risk in patients undergoing surgery above the age of 25 compared with patients undergoing surgery at a younger age. We have confirmed the recent Dutch findings of a relationship between the site of the mutation and the risk of secondary proctectomy.8 Delineation of attenuated FAP has been suggested with some overlapping areas35–37; in our present calculation we have chosen codon 0 –200 and after codon 1500 as the compromise. We have compared the risk of secondary proctectomy in patients with a mutation in codon 1250 –1500 with those with a mutation outside this region, and show a statistically significant higher risk of proctectomy in the former group. Unfortunately, a potential relationship between location of the pathogenic mutation and subsequent rectal carcinoma could not be evaluated because of the low number of rectal cancer patients with a known mutation. Strangely enough, all 7 patients with a secondary rectal carcinoma were found in the low-risk group of proctectomy (Figure 1), whereas no case of rectal

8

5/9 43 10 yr: 6 15 yr: 8 20 yr: 25

44 10 yr: 13 15 yr: 24

50 yr: 5 60 yr: 14 69 11 10 yr: 9 15 yr: 14 20 yr: 17 25 yr: 31

50 yr: 13 60 yr: 26 60 20

12

27/47 42 10 yr: 4 20 yr: 12 30 yr: 17 40 yr: 32 50 yr: 18 60 yr: 30

21 40 yr: 57

cancer was found in patients with a mutation in the dense polyposis region. This does not, however, mean a low cancer risk in the latter mutation region, but is probably a result of a higher rate of proctectomy in such patients. In addition, it is remarkable that none of the patients with attenuated FAP had a secondary proctectomy. In the estimation of these relations to the mutation site it must be kept in mind, however, that the present patient series includes the patients from the previous Dutch series.8 In total, our present molecular data only justify recommendations about surgical procedures in patients with a mutation before codon 200 or after codon 1500 (i.e., in attenuated FAP). The cumulative risk of secondary proctectomy for uncontrollable rectal adenomatosis was 57% after 40 years. This is surprisingly high, but in accordance with a recent Finnish study.5 Over the past 20 years the decision to perform a secondary proctectomy has become less difficult because of the possibility of avoiding a permanent ileostomy and, instead, converting to an IPAA. This possibility may partly account for the increasing number of secondary proctectomies. Another possible explanation is that differences may exist in the spectrum of apc mutations between the participating countries, who have a uniform post-IRA surveillance strategy. What then are the major factors to consider in the choice between IRA and IPAA? (1) The most important is the risk of rectal cancer. Proven predictive factors of an increased rectal cancer risk are increasing number, size,

December 2000

and severity of dysplasia of rectal adenomas; remaining rectum ⬎12–15 cm; age of the patient; and noncompliance of the patient in postoperative surveillance, especially after IRA. Potential risk factors are a high number of colonic adenomas, presence of a colonic or even upperrectal carcinoma, or family history of a high number of colorectal adenomas.2,4 – 8,15,33 A mutation located before codon 200 or after codon 1500 probably means a low risk of rectal cancer. Finally, the survival rate after a rectal cancer is a major factor in the choice between an IRA and an IPAA.38 (2) The risk of postoperative complications, pelvic sepsis, and fistula formation is higher after IPAA than after IRA,10,11 and the former includes a 2-stage procedure with specific complications after ileostomy closure. (3) The risk after IRA of future proctectomy caused by increasing rectal adenomatosis is an argument in favor of IPAA2,3,5,33 and seems to be higher in patients with a mutation in codon 1250 –1500. (4) Postoperative pelvic nerve damage leading to sexual disturbance or voiding problems is very rare after IRA as well as after IPAA.10,11,18 (5) The postoperative long-term bowel function and continence are not significantly different after IRA and IPAA.10,17 (6) Pouchitis after IPAA is rare in polyposis patients.27 (7) The presence of a desmoid has been claimed to be an indication for an IRA10 to avoid the more extended dissection at IPAA, but this has never been proven, and the argument may be reversed because a secondary IPAA may turn out to be technically impossible after an IRA. Therefore, in our opinion, a desmoid is not a factor of major influence on the decision. Our proposal for recommendation of surgical procedure is as follows: 1. IRA is recommended for young patients with few rectal adenomas who have a family history of a mild phenotype and for patients with attenuated FAP (a mutation localized before codon 200 or after 1500). IRA candidates must accept life-long endoscopic surveillance of the rectum at intervals of no longer than 6 months. 2. IPAA is recommended for patients with many rectal polyps and/or a family history of severe polyposis. 3. Total proctocolectomy with Brooke ileostomy is only indicated in those few patients with cancer in the lower third of the rectum that cannot be removed for cure without sacrificing the anal sphincter. These guidelines for treatment are similar to those recently presented by Setti-Carraro and Nicholls,39 and they do not differ significantly from the opinion of most of the members of the Leeds Castle Polyposis Group.1

ILEORECTAL ANASTOMOSIS

1459

We foresee that the continuous and rapid development in the molecular genetic field may contribute substantially to a future prediction of the rectal cancer risk. However, in the meantime we need treatment guidelines on a clinical basis. Therefore, we propose that our strategy should be prospectively evaluated in a multicenter study to reach agreement on a future evidence-based gold standard for the surgical treatment of patients with FAP.

References 1. Vasen HFA, Bu¨low S, and the Leeds Castle Polyposis Group. Guidelines for the surveillance and management of familial adenomatous polyposis (FAP): a world wide survey among 41 registries. Colorectal Dis 1999;1:214 –221. 2. Bussey HJR, Eyers AA, Ritchie SM, Thomson JPS. The rectum in adenomatous polyposis: the St. Mark’s policy. Br J Surg 1985; (suppl):S29 –S35. 3. Bu¨low S. The risk of developing rectal cancer after colectomy and ileorectal anastomosis in Danish patients with polyposis coli. Dis Colon Rectum 1984;27:726 –729. 4. DeCosse JJ, Bu¨low S, Neale K, Ja¨rvinen H, Alm T, Hultcrantz R, Moesgaard F, Costello C, and the Leeds Castle Polyposis Group. Rectal cancer risk in patients treated for familial adenomatous polyposis. Br J Surg 1992;79:1372–1375. 5. Heiskanen I, Ja¨rvinen HJ. Fate of the rectal stump after colectomy and ileorectal anastomosis for familial adenomatous polyposis. Int J Colorectal Dis 1997;12:9 –13. 6. Iwama T, Mishima Y. Factors affecting the risk of rectal cancer following rectum-preserving surgery in patients with familial adenomatous polyposis. Dis Colon Rectum 1994;37:1024 –1026. 7. Nugent KP, Phillips RKS. Rectal cancer risk in older patients with familial adenomatous polyposis and an ileorectal anastomosis: a cause for concern. Br J Surg 1992;79:1204 –1206. 8. Vasen HF, van der Luijt RB, Slors JF, Buskens E, de Ruiter P, Baeten CG, Schouten WR, Oostvogel HJ, Kuijpers JH, Tops CM, Meera-Khan P. Molecular genetic tests as a guide to surgical management of familial adenomatous polyposis. Lancet 1996; 348:433– 435. 9. Duijvendijk P, Vasen HFA, Bertario L, Bu¨low S, Kuipers JHC, Schouten WR, Guillem JG, Taat CW, Slors JFM. Cumulative risk of developing polyps or malignancy at the ileal pouch–anal anastomosis in patients with familial adenomatous polyposis. J Gastrointest Surg 1999;3:325–330. 10. Ambroze WJ Jr, Dozois RR, Pemberton JH, Beart RW, Ilstrup DM. Familial adenomatous polyposis: results following ileal pouch– anal anastomosis and ileorectostomy. Dis Colon Rectum 1992,35:12–15. 11. Madden MV, Neale KF, Nicholls RJ, Landgrebe JC, Chapman PD, Bussey HJR, Thomson JPS. Comparison of morbidity and function after colectomy with ileorectal anastomosis or restorative proctocolectomy for familial adenomatous polyposis. Br J Surg 1991; 78:789 –792. 12. Bussey HJR. Familial polyposis coli. Baltimore, MD: The Johns Hopkins University Press, 1975. 13. Moertel CG, Hill JR, Adson MA. Management of multiple polyposis of the large bowel. Cancer 1971;28:160 –164. 14. Bess MA, Adson MA, Elveback LR, Moertel CG. Rectal cancer following colectomy for polyposis. Arch Surg 1980;115:460 – 467. 15. Sarre RG, Jagelman DG, Beck GJ, McGannon E, Fazio VW, Weakley FL, Lavery IC. Colectomy with ileorectal anastomosis for familial adenomatous polyposis: the risk of rectal cancer. Surgery 1987;101:20 –26. 16. Heiskanen I, Matikainen M, Hiltunen K-M, Laitinen S, Rintala R,

1460

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

¨ LOW ET AL. BU

Ja¨rvinen HJ. Colectomy and ileorectal anastomosis or restorative proctocolectomy for familial adenomatous polyposis. Colorectal Dis 1999;1:9 –14. Kartheuser AH, Parc R, Penna CP, Tiret E, Frileux P, Hannoun L, Nordlinger B, Loygue J. Ileal pouch–anal anastomosis as the first choice operation in patients with familial adenomatous polyposis: a ten-year experience. Surgery 1996;119:615– 623. Nyam DCNK, Brillant PT, Dozois RR, Kelly KA, Pemberton JH, Wolff BG. Ileal pouch–anal canal anastomosis for familial adenomatous polyposis. Ann Surg 1997;226:514 –521. Penna C, Tiret E, Kartheuser A, Hannoun L, Nordlinger B, Parc R. Function of ileal J pouch–anal anastomosis in patients with familial adenomatous polyposis. Br J Surg 1993;80:765–767. Wiltz O, Hashmi H, Schoetz DJ, Roberts PL, Murray JJ, Coller JA, Veidenheimer MC. Carcinoma and the ileal pouch–anal anastomosis. Dis Colon Rectum 1991;34:805– 809. Hoehner JC, Metcalf AM. Development of invasive adenocarcinoma following colectomy with ileoanal anastomosis for familial polyposis coli. Dis Colon Rectum 1994;37:824 – 828. Bassuini MMA, Billings PJ. Carcinoma in an ileoanal pouch after restorative proctocolectomy for familial adenomatous polyposis. Br J Surg 1996;83:506. Von Herbay A, Stern J, Herfarth C. Pouch-anal cancer after restorative proctocolectomy for familial adenomatous polyposis. Am J Surg Pathol 1996;20:995–999. Church JM, Oakley JR, Wu JS. Pouch polyposis after ileal pouch– anal anastomosis for familial adenomatous polyposis. Dis Colon Rectum 1996;39:584 –586. Nugent KP, Spigelman AD, Nicholls RJ, Talbot IC, Neale K, Phillips RKS. Pouch adenomas in patients with familial adenomatous polyposis. Br J Surg 1993;80:1620. Wu JS, McGannon E, Church JM. Incidence of neoplastic polyps in the ileal pouch of patients with familial adenomatous polyposis after restorative proctocolectomy. Dis Colon Rectum 1998;41: 525–527. Kmiot WA, Williams MR, Keighley MRB. Pouchitis following colectomy and ileal reservoir construction for familial adenomatous polyposis. Br J Surg 1990;77:1283. Meagher AP, Farouk R, Dozois RR, Kelly KA, Pemberton JH. J ileal pouch–anal anastomosis for chronic ulcerative colitis: complications and long-term outcome in 1310 patients. Br J Surg 1998; 85:800 – 803. Penna C, Kartheuser A, Parc R, Tiret E, Frileux P, Hannoun L, Nordlinger B. Secondary proctectomy and ileal pouch–anal anastomosis after ileorectal anastomosis for familial adenomatous polyposis. Br J Surg 1993;80:1612–1613. Ziv Y, Church JM, Oakley JR, McGannon E, Schroeder TK, Fazio

GASTROENTEROLOGY Vol. 119, No. 6

31.

32.

33.

34.

35.

36.

37.

38.

39.

VF. Results after restorative proctocolectomy and ileal pouch– anal anastomosis in patients with familial adenomatous polyposis and coexisting colorectal cancer. Br J Surg 1996;83:1578 – 1580. Panis Y, Bonhomme N, Hautefeuille P, Valleur P. Ileal pouch–anal anastomosis with mesorectal excision for rectal cancer complicating familial adenomatous polyposis. Eur J Surg 1996;162: 817– 821. Milsom J, Ludwig KA, Church JM, Garcia-Ruiz A. Laparoscopic total abdominal colectomy for familial adenomatous polyposis. Dis Colon Rectum 1997;40:675– 678. Debinski HS, Love S, Spigelman AD, Phillips RKS. Colorectal polyp counts and cancer risk in familial adenomatous polyposis. Gastroenterology 1996;110:1028 –1030. Evans DGR, Hill J, Dudding T, Burn J, Maher ER. Molecular tests in surgical management of familial adenomatous polyposis. Lancet 1997;350:1777. Nagase H, Miyoshi Y, Horii A, Aoki T, Ogawa M, Utsunomiya J, Baba S, Sasazuki T, Nakamura Y. Correlation between the location of germ-line mutations in the APC gene and the number of colorectal polyps in familial adenomatous polyposis patients. Cancer Res 1992;52:4055– 4057. Soravia C, Berk T, Madlensky L, Mitri A, Cheng H, Gallinger S, Cohen Z, Bapat B. Genotype-phenotype correlations in attenuated adenomatous polyposis coli. Am J Hum Genet 1998;62: 1290 –1301. Ficari F, Cama A, Valanzano R, Curia MC, Palmirotta R, Aceto G, Esposito DL, Crognale S, Lombardi A, Messerini L, Mariani-Constantini R, Tonelli F, Battista P. APC gene mutations and colorectal adenomatosis in familial adenomatous polyposis. Br J Cancer 2000;82:348 –353. Vasen HFA, van Duijvendijk P, Buskens E, Bu¨low C, Bjo¨rk J, Ja¨rvinen HJ, Bu¨low S. Decision analysis in the surgical management of patients with familial adenomatous polyposis. Gut (submitted). Setti-Carraro P, Nicholls RJ. Choice of prophylactic surgery for the large bowel component of familial adenomatous polyposis. Br J Surg 1996;83:885– 892.

Received March 30, 2000. Accepted July 26, 2000. Address requests for reprints to: Steffen Bu ¨low, M.D., D.M.Sc., The Danish Polyposis Register, Department of Surgical Gastroenterology 435, Hvidovre University Hospital, DK-2650 Hvidovre, Copenhagen, Denmark. e-mail: [email protected]; fax: (45) 3632-3200. The authors gratefully acknowledge the work of Dr. Carli Tops.