intestinal cancer development. Nature
30 Prokhortchouk A, Sansom O, Selfridge J, et al.
Kaiso-deficient mice show resistance to intestinal
cancer. Mol Cell Biol 2006;26:199–208.
31 Perreault N, Sackett SD, Katz JP, et al. Foxl1 is a
mesenchymal modifier of Min in carcinogenesis of
stomach and colon. Genes Dev 2005;19:311–5.
32 Oster H, Leitges M. Protein kinase C alpha but not
PKCzeta suppresses intestinal tumor formation in
ApcMin/+ mice. Cancer Res 2006;66:6955–63.
33 Wilson CL, Heppner KJ, Labosky PA, et al. Intestinal
tumorigenesis is suppressed in mice lacking the
metalloproteinase matrilysin. Proc Natl Acad
Sci U S A 1997;94:1402–7.
34 Sansom OJ, Mansergh FC, Evans MJ, et al.
Deficiency of SPARC suppresses intestinal
tumourigenesis in APCMin/+mice. Gut
35 Hanahan D, Weinberg RA. The hallmarks of
cancer. Cell 2000;100:57–70.
36 Framson PE, Sage EH. SPARC and tumor growth:
where the seed meets the soil? J Cell Biochem
37 Murphy-Ullrich JE. The de-adhesive activity of
matricellular proteins: is intermediate cell adhesion
38 Bradshaw AD, Sage EH. SPARC, a matricellular
protein that functions in cellular differentiation and
tissue response to injury. J Clin Invest
39 Brekken RA, Puolakkainen P, Graves DC, et al.
Enhanced growth of tumors in SPARC null mice is
associated with changes in the ECM. J Clin Invest
40 Sangaletti S, Stoppacciaro A, Guiducci C, et al.
Leukocyte, rather than tumor-produced SPARC,
determines stroma and collagen type IV deposition
in mammary carcinoma. J Exp Med
... ............ ............. ............ ............. ............ ............. ............
What is the role of iFOBT in screening
for colorectal cancer?
David F Ransohoff
Policy makers will need to consider if it has one, not only as an
adjunct to gFOBT screening, but also as a primary screening test
randomised controlled trials (RCTs) to
reduce CRC mortality.1–3gFOBT testing is
endorsed as an option for CRC screening
in the United States4–7and is being
implemented in the United Kingdom.
People with a positive gFOBT receive
colonoscopy to detect early cancers and
advanced adenomas that, if untreated,
might cause CRC mortality. Because
gFOBT has a high rate of false positive
results, however, gFOBT screening can
incur substantial cost and use of colono-
scopy resources. A method that could
determine which people with a positive
gFOBT have false positive results—and do
notneed colonoscopy—would make
gFOBT screening more practical.
A study in this issue of Gut8Fraser et al
(see p 1415) shows that doing iFOBT
(human haemoglobin immunochemical
based FOBT) in people with a positive
gFOBT will detect almost all clinically
individuals, while reducing false positive
results and the need for colonoscopy. In
this study, both iFOBT and colonoscopy
were done in those with a positive gFOBT,
defined as one to four positive gFOBT
ovals (people with five or six positive
ovals were automatically referred for
colonoscopy). The sensitivity of iFOBT
among people with positive gFOBT was
95.9%, while specificity was 59.2%, result-
ing in a 30% reduction of colonoscopy
use. The mechanism by which iFOBT
achieves higher specificity than gFOBT
creening for colorectal cancer (CRC)
using gFOBT (guaiac based faecal
occult blood test) has been shown in
presumably is elimination of false posi-
tives that occur in guaiac based testing
from sources other than human haemo-
globin, like diet.
Learning that iFOBT may have a role in
a programme of gFOBT screening is
practical and important, but it raises a
larger question that will need to be sorted
out over time. Should iFOBT replace
gFOBT as the primary test in a pro-
gramme of FOBT screening? The authors
note that the unit cost for iFOBT is higher
than for gFOBT, but that is only one
increased benefit that might be achieved,
additional costs (including false posi-
tives), and the public’s willingness to
pay for greater benefit.
The potential usefulness of iFOBT vs
gFOBT screening depends largely on the
‘‘absolute sensitivity’’ of each FOBT, a
question not addressed in this study that
assessed iFOBT sensitivity only among
those who already had a positive gFOBT.
Used in this way, iFOBT can never do
better than gFOBT.
What is the absolute sensitivity for
each FOBT? Obtaining absolute sensitiv-
ity for any FOBT is logistically difficult,
requiring administration of both the
FOBT and a ‘‘gold standard’’ exam (like
colonoscopy). In asymptomatic indivi-
duals the prevalence of CRC is so low,
roughly 1–3 per 1000, that thousands of
people need to be studied. In one report,
gFOBT sensitivity for CRC was about
30%9; in another it was 13%.10
different results might be explained by
use of centralised FOBT processing in the
former study (as is also done in the
current study, in RCTs, and in the UK
screening programme being implemen-
ted), while in the latter study processing
was done in each physician’s office.
While these are low values for absolute
enough, as shown in RCTs, to reduce
CRC mortality by 33% in a US1trial and
by about 16% in two European trials.2 3
FOBT in a programme of screening.
gFOBT screening is considered not only
an ‘‘option’’ for CRC screening, but when
combined with sigmoidoscopy, it may be
competitive, in terms of effectiveness,
with a programme of colonoscopy screen-
ing.11 12This seeming paradox—that a less
sensitive test like FOBT might be as
effective or more effective than a ‘‘gold
standard’’ test like colonoscopy—can be
explained by considering how a ‘‘pro-
gramme’’ of screening works. If a test
with lower sensitivity at any one applica-
tion is applied repeatedly over time in a
programme of screening, it may result in
a higher ‘‘programme’’ sensitivity than
when a more sensitive test (like colono-
scopy) is applied less frequently, because
a more frequently applied test can detect
fast growing CRC that would be missed
by a less frequently applied test.11–13
If gFOBT is not simply ‘‘acceptable’’ but
perhaps is even competitive with colono-
scopy in some settings, then what would
be the impact of a better FOBT? The
absolute sensitivity of iFOBT in recent
studies has been shown to be roughly
60% for CRC,14–17much higher than for
gFOBT. Even without new RCT data to
assess CRC mortality reduction, policy
decisions may be based on data about
the sensitivity and specificity of newer
tests, existing RCT results, and model-
ling.7Policy makers will need to consider
whether iFOBT has a role not only as an
adjunct to gFOBT screening, but also as a
primary screening test.
Correspondence to: David F Ransohoff, CB7080,
University of North Carolina at Chapel Hill,