Bowel Screening in Scotland – Current Challenges and Possible Solutions

Bowel Screening in Scotland – Current

Challenges and Possible Solutions

Prof. Bob Steele

Ninewells Hospital, University of Dundee

Proving Screening Works

Population-based randomised trials in which the whole group offered screening (including refusers and interval cancers) is compared with the control group

Disease-Specific Mortality in gFOBT Randomised Trials

(Relative Risks)

• Minnesota– Annual 0.67 (CI 0.51-0.83)– Biennial 0.79 (CI 0.62 - 0.97)

• Nottingham– Biennial 0.85 (CI 0.74 - 0.98)

• Funen– Biennial 0.82 (CI 0.68 - 0.99)

• Göteborg– Biennial 0.84 (CI 0.71-0.99)

National UK Colorectal Cancer Screening Pilot

Aim: to test the feasibility of introducing gFOBT screeing into the NHS

Single Centre

Investigation and treatment devolvedto health boards (n=14)

Age range 50 - 74

Organisation of the bowel cancer screening programme - Scotland

0

1

2

3

4

5

CR

C m

orta

lity/

10

00 p

ers

ons

0 1 2 3 4 5 6 7 8 9 10Years since screening/matched date

Invited for screening Controls

Rate and 95% CI (Nelson-Aalen estimates)Cumulative Mortality from Colorectal Cancer

Rate ratio of Colorectal Cancer invited vs controls

Overall0.90 (0.830 – 0.989)

Relative reduction in CRC mortality 10%

Participants only 0.73 (0.653 – 0.824)

Relative reduction in CRC mortality 27%

Positive Predictive Value of Screening Colonoscopy

CarcinomaCarcinoma

14.6%14.6%

AdenomaAdenoma

35.9%35.9%No NeoplasiaNo Neoplasia

49.5%49.5%

Uptake- Gender and Deprivation

%

SIMD

Round 1 Round 2 Round 3

Screen -detected 351 (56.6%)

208 (46.5%)

139 (35.7%)

True interval 193 (31.2%)

213 (47.7%)

229 (58.9%)

Missed 2 (0.3%)

4 (0.9%)

2 (0.5%)

Miscellaneous 66 (10.7%)

22 (4.9%)

19(4.9%)

Not on Socrates 6(1%)

0 0

Cancers Diagnosed in the Screened Population

Gender distribution - all rounds

%

Site distribution - all rounds

%

Issues to address

• Interval Cancers

• Gender inequality

• Rectal and right-sided cancers

• Uptake

“Blood in stool” tests

Flexible Sigmoidoscopy

Colonoscopy

Colonoscopy

• No RCT results

• Case control studies only

• But – highly sensitive and 100% specific

If an insensitive test with imperfect specificity reduces

mortality…..

Test Accuracy Acceptability Risk

FOBT + +++ -

Flex-sig ++ ++ +

Colonosc +++ + +++£££££

£

ICRF/MRC Study(Oct 1996 – March 1999)

• Single flexible sigmoidoscopy with removal of adenomas– 55-64 years

• High risk colonoscopy– adenoma > 1cm– 3+ adenomas– tubulovillous or villous histology– 20+ hyperplastic polyps above distal rectum– cancer

ICRF/MRC Study

Total no: 354262

Interested : 194726 (55%)

Randomised: 170432

Control: 113178Invited for screening: 57254

Attended: 40674 (71%)

Mortality from CRC

Incidence of CRC

Incidence of L-sided CRC

Incidence of R-sided CRC

Potential Advantages of FS

• Disease prevention– Enhanced detection of left-sided

adenomas

• Detection of rectal cancer

• Unlikely to be a gender difference

Potential Problems with FS

• Uptake– Unlikely to be >30%– Possibility of exaggerated deprivation

gradient

• Effect on right-sided cancers

Alternative Strategy

Increasing sensitivity of FOBT?

gFOBT vs FIT

• gFOBT– Based on Guaiac reaction– Not specific for haemoglobin– Messy to do

• FIT– Immunological– Specific for human haemoglobin– Easy to do– QUANTITATIVE

Normal DD HP LRA HRA Cancer0

100

200

300

400

500

600

700

800

Fae

cal

hae

mo

glo

bin

(n

g/m

l)Quanitative FIT and Disease

n=20358 n=17783

“FIT 400”

n=20358 n=17783

“FIT 50”

Sensitivity Specificity

FIT 50

• 10% positivity rate

• 90% sensitive for cancer• 40% sensitive for adenoma• Lower detection limit may be more sensitive

But…

Question

• Sigmoidoscopy for all

• FIT 50 and colonoscopy for ~ 10%

or ?

Potential Pros

• Sigmoidoscopy– Detection of left-sided adenomas and protection from left sided cancer

– Detection of rectal cancer

• FIT 50– Uptake

– Detection of right-sided cancer