Annals of Oncology Advance Access originally published online on May 25, 2006
Annals of Oncology 2006 17(8):1328-1332; doi:10.1093/annonc/mdl101
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2006 European Society for Medical Oncology
Colorectal cancer (CRC) screening using sigmoidoscopy followed by colonoscopy: a feasibility and efficacy study on a cancer institute based population
1 Gastroenterology Department and 2 Epidemiology Clinic, Jules Bordet Institute, Brussels, Belgium
* Correspondence to: Prof. H. Bleiberg, Department of Gastroenterology, Institut Jules Bordet, Rue Héger-Bordet, 1, 1000 Brussels, Belgium. Tel: +32-2-541-32-07; Fax: +32-2-538-08-58; E-mail: harry.bleiberg{at}bordet.be
 |
Abstract |
|---|
 Top Abstract introductionpatients and methodsresultsdiscussionReferences |
|---|
Background: Advanced distal neoplasia found at sigmoidoscopy could be the marker for more proximal lesions.
Patients and methods: In the setting of a screening clinic, subjects underwent flexible sigmoidoscopy. If no significant lesion was found, sigmoidoscopy was planned after 5 years. If an advanced neoplasia was found, colonoscopy was performed just after the first sigmoidoscopy and at 1, 3 and 5 years. If a non-advanced neoplasia was found, sigmoidoscopy was performed at 1, 3 and 5 years and followed by colonoscopy if advanced lesion was found.
Results: At first screening 1704/1912 (88%) subjects had a negative sigmoidoscopy, 104 (5.4%) had an advanced neoplasia, 96 (6%) had a non-advanced neoplasia and eight (0.4%) had invasive colorectal cancer (CRC). At follow-up examinations at 1, 3 and 5 years, among 170 subjects with advanced and non-advanced neoplasia, one developed invasive CRC and 47 (31.6%), advanced neoplasia. At 5 years, among 718 first negative sigmoidoscopies, 572 (80%) were confirmed negative and 97 (14%) had advanced neoplasia. Colorectal cancer status at 5 years could be checked for interval cancers in 97% of subjects and no CRC was diagnosed in subjects who did not attend sigmoidoscopy at 5 years. Comparison of the incidence of invasive CRC to the data of registries of the Netherlands and Luxembourg suggested that the incidence of CRC was decreased by 36%–46%. Seven of the nine CRCs were Duke's A and the two others were Duke's B and C.
Conclusions: Screening with sigmoidoscopy followed by colonoscopy in case of positive sigmoidoscopy leads to substantial decreases in the incidence of CRC. Most CRCs found are at an early, curable stage of their development.
Key words: screening, colorectal cancer, sigmoidoscopy
|
introduction |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionReferences |
|---|
Worldwide, colorectal cancer (CRC) represents one of the most frequent cancer-related deaths in Western countries, with approximately 180 000 deaths from CRC in the 25 member states of the European Union in 2000 [1
].
Although the disease may sometimes be very aggressive, CRC diagnosed at an early stage, either as a result of alerted symptoms or, better, by screening, have more chance of being surgically cured and are associated with an improved prognosis. Moreover, most CRCs arise from adenomas. Since the adenoma-cancer sequence may take years to evolve, resection of premalignant adenomas would have a major impact on the incidence of CRC [2]. Therefore, screening procedures, such as endoscopy or fecal occult blood testing (FOBT), have a key role in the diagnosis of premalignant or early malignant lesions [5–9].
Common sense supports endoscopy to visualise the colon directly. Although highly sensitive and specific for lesions within the reach of the instrument, sigmoidoscopy could miss approximately 50% of proximal or right-sided neoplasms [10, 11]. If colonoscopy is initiated after identifying an adenoma in the distal colon, the sensitivity for the detection of advanced neoplasms (i.e. five or more adenomas of any size, at least one adenoma >1 cm, adenoma containing >25% villous structure and/or or high-grade dysplastic lesion) anywhere in the colon is 70% [4]. Addition of FOBT to the workup increases the sensitivity to only 76% [12].
Colonoscopy examines the entire colon and can remove adenomas and early cancers. For those with negative findings on colonoscopy examination, further screening should not be performed within 5–10 years. However, compared with sigmoidoscopy, colonoscopy has a higher rate of perforation (2 per 1000 versus 1 per 10 000 for sigmoidoscopy) [13], a higher rate of sedatives-associated complications and a higher cost. Therefore sigmoidoscopy appears as an attractive approach for CRC screening.
In 1993, the Screening Clinic of the Institut Jules Bordet started a study based on sigmoidoscopy followed by colonoscopy if advanced neoplasms were found in the distal colon. The aim of the study was to evaluate, by the compliance, the number of cancers and advanced neoplasia detected, the stage at surgery and the number of interval cancers at 5 years, the feasibility and efficacy of sigmoidoscopy followed by colonoscopy for screening CRC.
|
patients and methods |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionReferences |
|---|
study population and procedure
The Ethical Committee of the Université Libre de Bruxelles approved the study protocol (9 April 1991). The present cohort was enrolled from March 1993 to December 1994 at the Clinique de Dépistage de l'Université Libre de Bruxelles. At that time, no colorectal screening programme was offered to the Belgian population.
Healthy subjects, 50–75 years old, were approached for participation and the study was presented to them as a research study. The study was presented to asymptomatic subjects. They had to sign an informed consent and complete a questionnaire focusing on their personal and familial medical risk factors including a personal history of chronic inflammatory bowel disease, colorectal adenoma or cancer, or a first-degree relative (parent, sibling or offspring) who had had adenoma or cancer. Persons were excluded if they reported (i) symptoms of disease of the lower gastrointestinal tract, including rectal bleeding in the previous 6 months, that would normally require a medical evaluation, (ii) a serious medical conditions that could increase the risk associated with colonoscopy, (iii) a familial history of Hereditary Non Polyposis Colorectal Cancer (HNPCC) or Familial Adenomatous Polyposis (FAP) or (iv) a colonoscopy performed in the previous 5 years and a sigmoidoscopy in the previous 3 years.
Subjects with and without risk factors were submitted to sigmoidoscopy. The bowel preparation consisted of three fleet-enema. A first enema was administered the evening before the endoscopic procedure and the second and third, 2 and 1 hour before sigmoidoscopy. No premedication was given. The endoscope was moved forward in order to visualize the proximal sigmoid and descending colon. Sigmoidoscopy was considered complete if the distal colon was visualised. In case of negative findings no other sigmoidoscopy was performed before the planned 5-year follow-up examination.
If advanced neoplasia, including five or more adenomas of any size, at least one adenoma >1 cm, adenoma containing >25% villous structure and/or high-grade dysplastic lesion was detected, total colonoscopy was performed and all lesions were resected. Further colonoscopies were planned at 1, 3 and 5 years.
If non-advanced neoplasia, including less than five adenomas, all adenomas being less than 1 cm with low-grade dysplasia, was detected, following sigmoidoscopies were planned at 1, 3 and 5 years. If advanced neoplasia was found at subsequent sigmoidoscopy, total colonoscopy was performed and subjects followed accordingly.
Bowel preparation for colonoscopic consisted of polyethylene glycol (PEG) enema solution and a fleet-enema given before the examination. A coagulation workup was obtained before colonoscopy. A peripheral venous access was placed and sedation was achieved with 2–3 mg midazolam. Analgesia with 50 mg pethidine hydrochloride or 10 mg piritramide was given if required. If the caecum could not be reached, subjects underwent the colonoscopy under general anaesthesia.
Evaluation of bowel preparation, pain control and completeness of endoscopy was not planned in the study.
histological evaluation
All retrieved lesions were sent to the pathology laboratory for processing. Classification of tumours was based on the most advanced lesion. The criterion for cancer diagnosis was an invasion of malignant cells beyond the muscularis mucosa.
tracing of patients lost to follow-up
In order to identify interval cancers at the end of the study, we implemented a procedure for obtaining the medical status of the participant who attended the initial screening only and did not show up for the planned endoscopies. Participants' general practitioners were contacted and asked for the CRC status of their patient. If this was not possible we contacted the participant or his family directly. If a CRC was reported the histopathology records were verified.
impact of endoscopy on CRC incidence
For evaluating the impact of endoscopic screening on CRC incidence, we calculated the standard incidence ratio (SIR) by taking the number of CRC found in the cohort as ‘observed’ CRC and the number of CRC to be expected in unscreened populations. The observed number of CRC was calculated by computing the person-years contribution of each male or female subject included in the cohort, and person-years of each subject were distributed in successive 5-year age intervals, starting from 35 years old until 85 years old. The numbers of person-years by each subject in each 5-year interval were then summed up. This total was multiplied by the age–sex-specific incidence reported by cancer registries, which yielded the age-specific expected numbers of CRC. These age-specific numbers were then summed up, in order to provide the total expected number of CRC in the entire cohort. Exact Poisson 95% confidence intervals were calculated.
Because there is no complete cancer registry in Belgium, we took data from registries from the south of the Netherlands 1993–1997 [14] and from the Grand-Duché of Luxembourg [15]. In Luxembourg, data over 3 years were considered in order to compensate for the small population size of the Grand-Duché of Luxembourg.
|
results |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionReferences |
|---|
A total of 8718 persons where approached for participation in the study, of which 1912 (22%) agreed to participate and met the enrolment criteria. The mean age of participants was 61 years (range 27–84) with 84 (4%) participants less than 50 years old and 61 (3%) older than 75 years. The group of participants comprised 1268 women (66%) and 644 men (34%).
Family risk factors for CRC were reported in 343 (18%) patients. Two hundred and ninety-six (16%) participants had a first-degree relative with CCR and 47 (2%) had a first-degree relative with adenomas. One hundred and thirty-four participants (7%) had personal CRC risk factors, i.e. five had a history of neoplasia other than CRC (0.3%), 116 had a history of adenomas (6%) and 13 had a history of ulcerative colitis (0.7%).
During the first screening, sigmoidoscopy permitted sigmoid exploration in all subjects. Table 1 displays results of the screening process. At initial screening 1704/1912 (88%) had a normal sigmoidoscopy, 96 (6%) were found with a non-advanced neoplasia and 104 (5.4%) were found with an advanced neoplasia. Patients with advanced neoplasia underwent colonoscopy shortly after sigmoidoscopy and eight invasive CRC were found.
|
Eighty-two per cent of participants who had a bowel lesion (any type) at initial screening attended at least one further screening session during the 5-year follow-up. Only 718 of 1704 (42%) participants with a normal result at initial sigmoidoscopy attended the follow-up sigmoidoscopy at 5 years.
At follow-up examinations at 1, 3 and 5 years, per-protocol analysis was done. Of the 88 subjects with advanced neoplasia at initial screening, one (1%) developed invasive CRC and 28 (31.5%) developed advanced neoplasia. The invasive CRC was found at 3-year follow-up examination. Among 82 subjects with non-advanced neoplasia, no malignant lesion was found but 19 (23%) developed advanced neoplasia. At 5 years, 572 of the 718 (80%) participants with normal initial screening were confirmed negative and 97 (14%) had advanced neoplasia (Table 1).
Among the nine subjects diagnosed with invasive CRC, one was classified as pT1N0M0 (Duke's A), six as pT2N0M0 (Duke's A), one as pT3N0M0 (Duke's B) and one as pT3N3M0 (Duke's C).
At 5-year follow-up, the entire cohort of 1912 participants was checked for interval cancers. The CRC status was confirmed in 1860 subjects (97%). Twenty-five participants could not be retrieved, 27 left the country and 16 had died from causes other than CRC. Six subjects were suspected of having developed an interval CRC. After reviewing the hospital files, diagnosis of CRC was excluded in five cases. For the last suspected CRC case, no surgical or pathological report could be retrieved.
We examined the number of invasive CRC found during the follow-up with the number of CRC one could have expected in the general population. Men in the cohort contributed to 3001.1 person-years and women in the cohort contributed to 5989.5 person-years. Using age-specific incidence data from the south of the Netherlands, 14.6 invasive CCR were expected compared with nine observed, thus a SIR of 0.65 (95% CI 0.21–1.08). Using the Luxembourg age-specific incidence data, 16.7 invasive CCR were expected compared with nine observed, thus a SIR of 0.54 (95% CI 0.19–0.89).
|
discussion |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionReferences |
|---|
Sigmoidoscopic screening for CRC has been recommended by several organisations based on indirect evidence of effectiveness. One large randomised controlled trial conducted to demonstrate that one single sigmoidoscopy performed at the age of 60 years old could lower the incidence and mortality of CRC, is ongoing in the UK [13
]. This programme appears acceptable, feasible and safe but it will be several years before results on survival are obtained. Lieberman et al. [10] and Imperiale et al. [11] have shown that although sigmoidoscopy is highly sensitive and specific for lesions within the reach of the instrument, it misses approximately 50% of the neoplasms that are proximal or right-sided. In our study we were able to obtain information on the cancer status after a 5-year period of time for 97% of the 1912 subjects included in our cohort. CRC status of most subjects at the end of follow-up was reported by their general practitioner or was self-reported. There is thus the possibility of error in reporting the cause of death, or that some subjects could harbor a still undetected CRC at the study end. These reporting and detection biases were minimised by verification of the causes of death and the medical files when interval CRC was suspected. Total avoidance of these biases would have required a colonoscopy of all subjects at the study end, a procedure that would have been ethically unacceptable and difficult to implement. In fact, we are not aware of any prospective study on CRC screening (with FOBT or endoscopy) that performed colonoscopy of all subjects at the study end. In spite of uncertainties of possible missed CRC, our results suggest that, even if lesions were not identified in the right-side colon at 5-years follow-up, none of these subjects developed a right-sided colon cancer and, a fortiori, none died from a cancer arising in that location.
Comparison of our incidence of invasive CRC with the data from Netherlands and Luxembourg registries suggests that screening with sigmoidoscopy followed by colonoscopy in case of positive sigmoidoscopy decreases the incidence of CRC. This strategy based on the identification of high-risk subjects through sigmoidoscopy has the advantage of reducing risks and costs associated with colonoscopic examinations.
The majority (64%) of CRC were classified as Duke's A tumours. In trials that tested the value of screening with FOBT, the proportions of Duke's A screen-detected CRCs was 41% in the Nottingham trial [7] and 38% in the Funen trial [6]. These figures indicate that gains in mortality from CRC may be higher for screening with sigmoidoscopy, according to our scheme, than by screening with FOBT as diagnosis appears to be done at an earlier stage with substantial reductions in incidence related to the adenoma resection strategy.
The primary goal of cancer screening is the decrease of mortality due to cancer. This was achieved through reducing cancer incidence or through detecting at an earlier stage, when the cancer is curable, or through a combination of both modalities. Our study suggests that sigmoidoscopy followed by colonoscopy is able to reduce CRC incidence and to detect CRC at an earlier stage than with FOBT testing. Reduced incidence and earlier detection obtained with sigmoidoscopy according to our scheme will have additive effects on the reduction of CRC mortality.
It must be borne in mind that all invasive cancers (but one) were found at initial screening. Hence, as the cohort progresses in time, the observed to expected number of CRC will steadily decrease, mainly in subjects compliant with screening protocol. As a consequence, as the cohort of screened subjects progresses in time, the CRC mortality rate is likely to steadily decline.
Outside the screening context, 7%–9% of CRC patients reported having a first-degree relative who had a CRC [24, 25]. That proportion was 16% in our study. Family history of CRC screening encourages participation in CRC screening programmes [26, 27]. It is thus possible that in the absence of screening, our study group would have had a higher incidence of CRC than average, and actual numbers of CRC expected in our study population were probably higher than those calculated using CRC incidence rates in general populations. Hence, the reduction in CRC incidence induced by screening in our study group may have been underestimated.
Five per cent of the subjects were identified with advanced neoplasia at initial screening. This group was followed at 1, 3 and 5-year intervals by colonoscopy allowing the discovery of one cancer (1%) and 28 (31.5%) advanced neoplasia. These subjects identified by sigmoidoscopy at first screening truly represent a group at risk for CRC that justifies more frequent follow-up.
Although non-advanced neoplasia are generally not considered as a marker for the presence of advanced neoplasia elsewhere in the colon, there are data supporting the opposite point of view [11, 17, 18]. In our series none of the subjects with non-advanced neoplasia developed cancer or adenoma with high-grade dysplasia. However, 23% developed advanced neoplasia during the follow-up period of time, suggesting that non-advanced neoplasia might be, in the long run, a risk factor for CRC occurrence. We also believe that discovery of non-advanced neoplasia should be followed by colonoscopy with resection of all visible lesions, warranting more frequent surveillance of that group of subjects.
Numerous previous studies have estimated recurrence rates for colorectal adenomas. In most of these series, recurrence rate was 27% after 1 year [19], 37% after 2 years [20] and more than 50% after 5 years [21–23]. Although study populations and design for these studies were different from ours, our estimates (58% recurrence after 5 years and more) are consistent and within the ranges of these previous estimates.
The screening acceptance was low with only 22% of subjects agreeing to participate in the study. This could be explained by the attitude of the health and academic authorities in Belgium, which, until now, did not acknowledge the effectiveness of any CRC screening procedure. In that context one could conclude that sigmoidoscopy is not feasible in Belgium. However, any change in the attitude of the health authorities is likely to promote such a non-aggressive/efficient method. Moreover, our study shows that if subjects with a negative sigmoidoscopy do not require further surveillance before 5 years, once those at risk were identified, the compliance to further colonoscopy surveillance was about 80%, demonstrating, at that level, the feasibility of the approach.
One limitation of our study is the relatively small number of subjects, which might have an impact on our results. In our series, subjects with familial and personal risk factors did not disclose a higher incidence of advanced neoplasia nor CRC. As the incidence of CRC in the presence of familial risk factors could be less than twice the incidence in normal controls [28], our sample size is too small to identify such a tiny increase in incidence.
The frequency of distal lesions in our series is 11%, which is rather low compared with the 22%–49% observed in other series [10, 11, 13]. This difference cannot be explained by different methods of exploring the sigmoid, which are fundamentally similar, but reflects probable differences in patient selection and might also be a consequence of our small sample size.
In conclusion, our study demonstrates that screening sigmoidoscopy is feasible and well accepted. Advanced and non-advanced distal neoplasias appear to be relevant markers for advanced neoplasia in the rest of the colon. Follow-up of subjects with non-advanced and advanced neoplasias found at sigmoidoscopy with subsequent resection of all lesions in the entire colon, whether advanced or non-advanced, prevent the occurrence of invasive cancer and would thus reduce the mortality due to CRC.
Received for publication October 27, 2005. Revision received March 23, 2006. Accepted for publication March 30, 2006.
|
References |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionReferences |
|---|
1. Ferlay J, Bray F, Sankila R, Parkin DM (eds). Eucan—Cancer incidence, mortality and prevalence in the European Union. IARC Cancer base no. 4. Lyon: IARC 1999.
2. Winawer SJ, Zauber AG, Diaz B. The National Polyp Study: temporal sequence of evolving colorectal cancer from the normal colon. Gastrointestinal Endosc 1987; 203–207.
3. Dukes CE, Bussey HJR. The spread of rectal cancer and its effect on prognosis. Br J Cancer 1958; 12: 309.[Web of Science][Medline]
4. Winawer SJ, Fletcher RH, Miller L et al. Colorectal cancer screening: clinical guidelines and rationale. Gastroenterology 1997; 112: 594–642.[CrossRef][Web of Science][Medline]
5. Mandel JS, Bond JH, Church TR et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993; 329: 672.
6. Kronborg O, Fenger C, Olsen J et al. Randomised study of screening for colorectal cancer with faecal-occult-blood test. Lancet 1996; 38: 1467–1471.
7. Hardcastle JD, Chamberlain JO, Robinson MHE et al. Randomised controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet 1996; 348: 1472–1477.[CrossRef][Web of Science][Medline]
8. Ahlquist DA, Wieand HS, Moertel CG et al. Accuracy of fecal occult blood screening for colorectal neoplasia. JAMA 1993; 269: 1262–1267.
9. Gotzsche P. Screening for colorectal cancer. Lancet 1997; 349: 356.[Medline]
10. Lieberman DA, Weiss DG, Bond JH et al. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. N Engl J Med 2000; 343: 162–168.
11. Imperiale TF, Wagner DR, Lin CY et al. Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med 2000; 343: 169–174.
12. Lieberman DA, Weiss DG. One-tie screening for colorectal cancer with combined fecal occult-blood testing and examination of the distal colon N Engl J Med 2001; 345: 555–560.
13. UK Flexible Sigmoidoscopy Screening Trial Investigators. Single flexible sigmoidoscopy screening to prevent colorectal cancer: baseline findings of a UK multicentre randomised trial. Lancet 2002; 359: 1291–1300.[CrossRef][Web of Science][Medline]
14. Coebergh JW, Janssen M, Louwman M, Voogd A. Cancer, incidence, care and survival in the south of the Netherlands, 1995–1999. Eindhoven, The Netherlands: IKZ, Comprehensive Cancer Centre South 2001.
15. RMT. Registre Morphologique de Tumeurs. Nouveaux cas au Grand-Duché de Luxembourg, année 1995, 1996, 1997. Laboratoire National de la Sant. Luxembourg, 1995, 1996, 1997.
16. Schoenfeld P, Shod J, Ornseth E et al. Predictive value of diminutive colonic adenoma trial. The PREDICT trial. Clin Gastroenterol Hepatol 2003; 1: 195–201.[CrossRef][Medline]
17. Schoen RE, Corle D, Cranston L et al. Is colonoscopy needed for the non-advanced adenoma found on sigmoidoscopy. Gastroenterology 1998; 115: 533–541.[CrossRef][Web of Science][Medline]
18. Gondal J, Grotmal T, Hostad B et al. Grading of distal colorectal adenomas as predictors for proximal colonic neoplasia and choice of endoscope in population screening: experience from the Norwegian Colorectal Cancer Prevention Study (NORCCAP) Gut 2003; 52: 398–403.
19. Winawer SJ, Zauber AG, O'Brien MJ et al. Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps. N Engl J Med 1993; 328: 901–906.
20. Avidan B, Sonnenberg A, Schnell TG et al. New occurrence and recurrence of neoplasms within 5 years of a screening colonoscopy. Am J Gastroenterol 2002; 97: 1524–1529.[CrossRef][Web of Science][Medline]
21. Yood MU, Oliveria S, Boyer JG et al. Colon polyp recurrence in a managed care population. Arch Intern Med 2003; 163: 422–426.
22. Jorgensen OD, Kronborg O, Fenger C. A randomized surveillance study of patients with pedunculated and small sessile tubular and tubulovillous adenomas. The Funen Adenoma follow-up study. Scand J Gastroenterol 1995; 30: 686–692.[Web of Science][Medline]
23. Amonkar MM, Hunt TL, Zhou Z, Jin X. Surveillance patterns and polyp recurrence following diagnosis and excision of colorectal polyps in a medicare population. Cancer Epidemiol Biomarkers Prev 2005; 14: 417–421.
24. Mitchell RJ, Campbell H, Farrington SM et al. Prevalence of family history of colorectal cancer in the general population. Br J Surg 2005; 92: 1161–1164.[CrossRef][Web of Science][Medline]
25. Sandhu MS, Luben R, Khaw KT. Prevalence and family history of colorectal cancer: implications for screening. J Med Screen 2001; 8: 69–72.
26. Slattery ML, Kinney AY, Levin TR. Factors associated with colorectal cancer screening in a population-based study: the impact of gender, health care source, and time. Prev Med 2004; 38: 276–283.[CrossRef][Web of Science][Medline]
27. Ramji F, Cotterchio M, Manno M et al. Association between subject factors and colorectal cancer screening participation in Ontario, Canada. Cancer Detect Prev 2005; 29: 221–226.[CrossRef][Web of Science][Medline]
28. Bazzoli F, Fossi S, Sottili S et al. The risk of adenomatous polyps in asymptomatic first-degree relatives of persons with colon cancer. Gastronterology 1995; 109: 783–788.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F Ruthotto, F Papendorf, G Wegener, G Unger, B Dlugosch, F Korangy, M. Manns, and T. Greten Participation in screening colonoscopy in first-degree relatives from patients with colorectal cancer Ann. Onc., September 1, 2007; 18(9): 1518 - 1522. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||