Report in English with a Dutch summary (KCE reports 45A)
Report in English with a Dutch summary (KCE reports 45A) Report in English with a Dutch summary (KCE reports 45A)
40 Screening for Colorectal Cancer KCE reports vol.45 A Swedish study focused on secondary cancers in 68.104 cases of CRC from the Swedish Family-Cancer Database 142. In 1.113 patients a secondary CRC was diagnosed; 25 of them had a family history of CRC. Cases of secondary CRC with a family history were diagnosed up to 10 years before sporadic cases. The RR of all secondary CRCs was 2,21 compared with the first CRC. Familial secondary CRCs had a 2-fold risk compared with the sporadic forms. Age of onset was the most important covariate of secondary CRCs; the relative risk at ages 15 - 39 years was 27 compared with the first CRC. Familial CRC was associated with a high risk of small-intestinal, endometrial, and gastric cancers apart from CRC, all typical of HNPCC. Among familial cases, 36% of secondary CRCs and 100% of endometrial cancers came from families that fulfilled the Bethesda criteria for HNPCC 145, 58. Only 12 families conformed to the Amsterdam criteria; in family members, the risk of secondary CRC was 127-fold and that of endometrial cancer 257-fold. Other sites that were in excess among all secondary cancers were many cancers linked to HNPCC and, additionally, breast, prostate, thyroid and other endocrine, skin, and genital cancers. The authors concluded that the high risk of secondary cancer after early-onset CRC calls for evaluation of family history and clinical surveillance. Adenoma prevalence and cancer risk in familial non-polyposis colorectal cancer was further studied in Sweden 123. Over a period of 10 years, 304 subjects at risk were included in ongoing surveillance with regular colonoscopies. To compile the medical findings and experience generated during this period, a retrospective cross sectional study was performed. Subjects were classified into three family groups: families with HNPCC (Lynch syndrome), families with hereditary colorectal cancer (HCC, non-Lynch syndrome) and a third group of families with only empirical risk estimates based on a family history of two close relatives (TCR) with CRC. The risk population was studied with regard to age at onset, prevalence, number, cancer risk, size, dysplasia, and distribution of adenomas. A comparison was made within the family groups and with a reference group representing the general population. In total, 195 adenomas and six cancers were detected among 85 individuals. The relative risk of having an adenoma in the whole risk population compared with the general population was 2,6. Subjects from TCR families had most adenomas and HNPCC subjects had the least. A shift from proximal adenomas to distal carcinomas in families with HCC and TCR suggested a higher cancer risk in distal adenomas in these syndromes. HNPCC families showed a younger age at onset and adenomas with a higher degree of dysplasia. In HNPCC there was a similar localisation of adenomas and carcinomas, suggesting a high risk of cancer in all adenomas. The study showed that there was clear overrepresentation of adenomas in all three family types compared with the reference population. In HNPCC there was earlier onset of adenomas and faster progression to cancer. Families with HCC, and even more so TCR subjects, had a later onset and lower risk of cancer from proximal adenomas. Based on these results, surveillance protocols in Sweden have been revised. A retrospective review of the French Calvados Cancer Registry 1993 - 1998, published in 2004 143, showed that colon cancer had a familial or genetic component but not rectal cancer: RR 1,47 (95% CI: 1,16-1,96; p value 0,004) vs. RR 0,98 (95% CI: 0,67-1,40). The familial/genetic component appeared stronger for proximal colon cancer than for distal colon cancer, but only among women: RR 2,24 (95% CI: 1,35-3,50) vs. RR 1,45 (95% CI: 0,83-2,36). Another French population study 140 aimed to estimate the lifetime risk (0 - 74 y) of CRC in the general population (males versus females) and in first degree relatives of patients with sporadic colorectal cancer or adenoma. The lifetime
KCE reports vol.45 Screening for Colorectal Cancer 41 risk of CRC was 1/23 in men and 1/40 in women. In males, 0,5% in the 55 - 59 age group and 4,5% in the 70 - 74 age group developped a CRC. The corresponding values in females were 0,4% and 2,5%. The cumulative risk at age 74 varied between 7,7% (one family member affected) and 25,6% (two affected) in males, and 4,3% and 14,3% respectively in females. The risk in the 40 - 44 year age group for individuals with one first degree relative affected before 45 years of age was 0,5%, similar to that of those aged 45 - 49 with one first degree relative affected with a colorectal cancer or a large adenoma (> 1 cm). The study results suggested that screening in the general population should start at 50 or 55. In individuals with one affected first degree relative before age 45, or with at least two affected first degree relatives, the lifetime risk appeared high enough (over 10%) to warrant colonoscopic screening and relatives of these patients should enter screening programs at age 40 to 44. Family history as a risk factor for colorectal cancer in Inflammatory Bowel Disease (IBD) was studied in a population-based cohort study of 19.876 individuals with ulcerative colitis or Crohn's disease born between 1941 and 1995 137. Familial CRC was associated with a more than 2-fold risk of CRC (RR = 2,5, 95% CI: 1,4 - 4,4) and an increase in absolute risk (AR) of CRC at 54 years from 3,8% to 6,9%. Patients with a first-degree relative diagnosed with CRC before 50 years of age had a higher RR (9,2, 95% CI: 3,7 - 23) and the highest AR (29%). No association with familial IBD was observed. Finally, a retrospective analysis of the Surveillance Epidemiology and End Results (SEER) program database for the period 1974 through 1995 identified 101.734 white and African-American women, age 25 yr, with prior cervical, endometrial, or ovarian cancer 141. Subsequent follow-up demonstrated no increased risk of colorectal cancer in women with cervical cancer. For endometrial cancer patients, increased risk of colorectal cancer was confined to women whose diagnosis of endometrial cancer was before age 50, but the increased risk was substantial in this group (RR 3,39; 95% CI: 2,73 - 4,17). For ovarian cancer patients, increased risk for colorectal cancer was substantial for those diagnosed with ovarian cancer before age 50 (RR 3,67; 95% CI: 2,74 - 4,80), and there was some increased risk for women diagnosed at ages 50 - 64 yr (RR 1,52; 95% CI: 1,25 - 1,83). 3.3.4 Systematic reviews with meta-analysis Some of the retrieved publications were based on systematic reviews of the literature with meta-analysis146, 147, 48, 114, 148, 47. The Web published NCI Colorectal Cancer (pDQ) Genetics update 47 recalls the estimated relative and absolute risks of developing CRC based on a systematic review and meta-analysis of familial colorectal cancer risk by Johns & Houlston 146, published in 2001 (Table 7).
- Page 1 and 2: Health Technology Assessment Colore
- Page 3 and 4: Health Technology Assessment Colore
- Page 5 and 6: KCE reports vol.45A Colorectale Kan
- Page 7 and 8: KCE reports vol.45A Colorectale Kan
- Page 9: KCE reports vol.45A Colorectale Kan
- Page 12 and 13: 2 Screening for Colorectal Cancer K
- Page 14 and 15: 4 Screening for Colorectal Cancer K
- Page 16 and 17: 6 Screening for Colorectal Cancer K
- Page 18 and 19: 8 Screening for Colorectal Cancer K
- Page 20 and 21: 10 Screening for Colorectal Cancer
- Page 22 and 23: 12 Screening for Colorectal Cancer
- Page 24 and 25: 14 Screening for Colorectal Cancer
- Page 26 and 27: 16 Screening for Colorectal Cancer
- Page 28 and 29: 18 Screening for Colorectal Cancer
- Page 30 and 31: 20 Screening for Colorectal Cancer
- Page 32 and 33: 22 Screening for Colorectal Cancer
- Page 34 and 35: 24 Screening for Colorectal Cancer
- Page 36 and 37: 26 Screening for Colorectal Cancer
- Page 38 and 39: 28 Screening for Colorectal Cancer
- Page 40 and 41: 30 Screening for Colorectal Cancer
- Page 42 and 43: 32 Screening for Colorectal Cancer
- Page 44 and 45: 34 Screening for Colorectal Cancer
- Page 46 and 47: 36 Screening for Colorectal Cancer
- Page 48 and 49: 38 Screening for Colorectal Cancer
- Page 52 and 53: 42 Screening for Colorectal Cancer
- Page 54 and 55: 44 Screening for Colorectal Cancer
- Page 56 and 57: 46 Screening for Colorectal Cancer
- Page 58 and 59: 48 Screening for Colorectal Cancer
- Page 60 and 61: 50 Screening for Colorectal Cancer
- Page 62 and 63: 52 Screening for Colorectal Cancer
- Page 64 and 65: 54 Screening for Colorectal Cancer
- Page 66 and 67: 56 Screening for Colorectal Cancer
- Page 68 and 69: 58 Screening for Colorectal Cancer
- Page 70 and 71: 60 Screening for Colorectal Cancer
- Page 72 and 73: 62 Screening for Colorectal Cancer
- Page 74 and 75: 64 Screening for Colorectal Cancer
- Page 76 and 77: 66 Screening for Colorectal Cancer
- Page 78 and 79: 68 Screening for Colorectal Cancer
- Page 80 and 81: 70 Screening for Colorectal Cancer
- Page 82 and 83: 72 Screening for Colorectal Cancer
- Page 84 and 85: 74 Screening for Colorectal Cancer
- Page 86 and 87: 76 Screening for Colorectal Cancer
- Page 88 and 89: 78 Screening for Colorectal Cancer
- Page 90 and 91: 80 Screening for Colorectal Cancer
- Page 92 and 93: 82 Screening for Colorectal Cancer
- Page 94 and 95: 84 Screening for Colorectal Cancer
- Page 96 and 97: 86 Screening for Colorectal Cancer
- Page 98 and 99: 88 Screening for Colorectal Cancer
40 Screen<strong>in</strong>g for Colorectal Cancer <strong>KCE</strong> <strong>reports</strong> vol.45<br />
A Swedish study focused on secondary cancers <strong>in</strong> 68.104 cases of CRC from<br />
the Swedish Family-Cancer Database 142. In 1.113 patients a secondary CRC was<br />
diagnosed; 25 of them had a family history of CRC. Cases of secondary CRC<br />
<strong>with</strong> a family history were diagnosed up to 10 years before sporadic cases. The<br />
RR of all secondary CRCs was 2,21 compared <strong>with</strong> the first CRC. Familial<br />
secondary CRCs had a 2-fold risk compared <strong>with</strong> the sporadic forms. Age of<br />
onset was the most important covariate of secondary CRCs; the relative risk at<br />
ages 15 - 39 years was 27 compared <strong>with</strong> the first CRC. Familial CRC was<br />
associated <strong>with</strong> a high risk of small-<strong>in</strong>test<strong>in</strong>al, endometrial, and gastric cancers<br />
apart from CRC, all typical of HNPCC. Among familial cases, 36% of secondary<br />
CRCs and 100% of endometrial cancers came from families that fulfilled the<br />
Bethesda criteria for HNPCC 145, 58. Only 12 families conformed to the<br />
Amsterdam criteria; <strong>in</strong> family members, the risk of secondary CRC was 127-fold<br />
and that of endometrial cancer 257-fold. Other sites that were <strong>in</strong> excess among<br />
all secondary cancers were many cancers l<strong>in</strong>ked to HNPCC and, additionally,<br />
breast, prostate, thyroid and other endocr<strong>in</strong>e, sk<strong>in</strong>, and genital cancers. The<br />
authors concluded that the high risk of secondary cancer after early-onset CRC<br />
calls for evaluation of family history and cl<strong>in</strong>ical surveillance.<br />
Adenoma prevalence and cancer risk <strong>in</strong> familial non-polyposis colorectal cancer<br />
was further studied <strong>in</strong> Sweden 123. Over a period of 10 years, 304 subjects at risk<br />
were <strong>in</strong>cluded <strong>in</strong> ongo<strong>in</strong>g surveillance <strong>with</strong> regular colonoscopies. To compile<br />
the medical f<strong>in</strong>d<strong>in</strong>gs and experience generated dur<strong>in</strong>g this period, a<br />
retrospective cross sectional study was performed. Subjects were classified <strong>in</strong>to<br />
three family groups: families <strong>with</strong> HNPCC (Lynch syndrome), families <strong>with</strong><br />
hereditary colorectal cancer (HCC, non-Lynch syndrome) and a third group of<br />
families <strong>with</strong> only empirical risk estimates based on a family history of two close<br />
relatives (TCR) <strong>with</strong> CRC. The risk population was studied <strong>with</strong> regard to age<br />
at onset, prevalence, number, cancer risk, size, dysplasia, and distribution of<br />
adenomas. A comparison was made <strong>with</strong><strong>in</strong> the family groups and <strong>with</strong> a<br />
reference group represent<strong>in</strong>g the general population. In total, 195 adenomas<br />
and six cancers were detected among 85 <strong>in</strong>dividuals. The relative risk of hav<strong>in</strong>g<br />
an adenoma <strong>in</strong> the whole risk population compared <strong>with</strong> the general population<br />
was 2,6. Subjects from TCR families had most adenomas and HNPCC subjects<br />
had the least. A shift from proximal adenomas to distal carc<strong>in</strong>omas <strong>in</strong> families<br />
<strong>with</strong> HCC and TCR suggested a higher cancer risk <strong>in</strong> distal adenomas <strong>in</strong> these<br />
syndromes. HNPCC families showed a younger age at onset and adenomas <strong>with</strong><br />
a higher degree of dysplasia. In HNPCC there was a similar localisation of<br />
adenomas and carc<strong>in</strong>omas, suggest<strong>in</strong>g a high risk of cancer <strong>in</strong> all adenomas. The<br />
study showed that there was clear overrepresentation of adenomas <strong>in</strong> all three<br />
family types compared <strong>with</strong> the reference population. In HNPCC there was<br />
earlier onset of adenomas and faster progression to cancer. Families <strong>with</strong> HCC,<br />
and even more so TCR subjects, had a later onset and lower risk of cancer<br />
from proximal adenomas. Based on these results, surveillance protocols <strong>in</strong><br />
Sweden have been revised.<br />
A retrospective review of the French Calvados Cancer Registry 1993 - 1998,<br />
published <strong>in</strong> 2004 143, showed that colon cancer had a familial or genetic<br />
component but not rectal cancer: RR 1,47 (95% CI: 1,16-1,96; p value 0,004) vs.<br />
RR 0,98 (95% CI: 0,67-1,40). The familial/genetic component appeared stronger<br />
for proximal colon cancer than for distal colon cancer, but only among women:<br />
RR 2,24 (95% CI: 1,35-3,50) vs. RR 1,45 (95% CI: 0,83-2,36).<br />
Another French population study 140 aimed to estimate the lifetime risk (0 - 74<br />
y) of CRC <strong>in</strong> the general population (males versus females) and <strong>in</strong> first degree<br />
relatives of patients <strong>with</strong> sporadic colorectal cancer or adenoma. The lifetime