Use of serum sample banks to study

1
Use of serum sample banks to study causes of and
screening for cancer Matti Hakama NCU/ANCR
30.8.2005
2
Operational procedure Population based bank is
linked to cancer registry Specific
exposures and tumour markers are measured in
sera from cancer patients and (matched)
controls
3
Benefits for exposures Allows follow-up study
with reduced cost Nested case-co study
within a cohort Strong indication as to
direction of cause and effect
4
Benefits for tumour markers free of length
bias free of overdiagnosis allows estimation
of lead time But allows estimation of validity
only
5
Problems Differential handling light
temperature thawing Differential
misclassification Between cases and controls
6
Further reduction in cost by pooling Case
pool Age-, sex-, smoking-stratified
control pools Few individual measurements
for random variation in estimates Provides
indirectly means to evaluate bias
7
Example from Finland for etiology Hakama M,
Hakulinen T, Kenward MG, Aaran R-K, Aromaa A,
Knekt P, Nikkari T, Teppo L, Peto R. Blood
biochemistry and the risk of cancer. Effect of
sample pooling. Acta Oncologica 2004
43667674.
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
Stomach cancer ______________________________ ____
______________________________ Exposure Pooled
Individual ______________________________ ________
_______ ________________ Alpha-tocopherol
(µg/ml) 0.36 (0.41) 0.65 (0.40) Beta-carote
ne (µg/ml) 3.8 (14.7) 12.6 (13.9) Retinol
(µg/ml) 1.6 (21.8) 29.7 (21.8) Retinol-bindi
ng pr. (µg/ml) 1.42 (1.84) 3.05 (1.85) Cerulop
lasmin (mg/l) 26.4 (10.2) 11.2 (10.5) _______
________________________ _________________________
__________ Hakama et al. 2004
13
Lung cancer ______________________________ _______
___________________________ Exposure Pooled Ind
ividual ______________________________ ___________
____ ________________ Alpha-tocopherol
(µg/ml) 0.15 (0.34) 0.20 (0.23) Beta-carotene
(µg/ml) 2.5 (11.9) 11.7 (7.8) Retinol
(µg/ml) 59.9 (17.0) 36.6 (12.9) Retinol-bindin
g pr. (µg/ml) 0.19 (1.40) 2.88 (1.10) Ceruloplas
min (mg/l) 31.7 (7.9) 27.2 (6.1) ______________
_________________ ________________________________
___ Hakama et al. 2004
14
Examples from Finland for screening Hakama M,
Stenman U-H, Knekt P, Järvisalo J, Leino A,
Hakulinen T, Maatela J, Aromaa A. Tumor markers
and screening for gastrointestinal cancer a
follow up study in Finland. Journal of Medical
Screening 1994 16064. Hakama M, Stenman U-H,
Aromaa A, Leinonen J, Hakulinen T, Knekt
P. Validity of the prostate specific antigen test
for prostate cancer screening followup study
with a bank of 21,000 sera in Finland. The
Journal of Urology 2001 16621892192.
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
Pooling did not work pooling itself affects
the level of substance differentially We
assume that research at individual level
works differential handling
19
Biobanks are a valuable tool Good biobanking
practice is under development Present practice
vulnerable to cause bias both in etiological
and in screening research