Gast roent érologie Clinique et Biologique (2009) 33 , Suppl. 3, S190—S201
Colon cancer in inflammatory bowel disease:
recent trends, questions and answers
Cancer colique au cours des maladies inflammatoires
chroniques intestinales : actualité et perspectives
S. Viennot a, A. Deleporte a, D. Moussatab, S. Nanceyb, B. Flourié b,
J. -M. Reimunda, c, *
a Cent re Hospit al ier Universit aire de Caen, Service d’ Hépat o-Gast roent érol ogie et Nut rit ion, Pôl e Reins-Digest if -Nut rit ion,
Hôpi t al Côt e de Nacr e, B. P. 95182, 14033 Caen cedex 9, Fr ance
b Cent r e Hospi t al i er Lyon-Sud, Ser vi ce d’ Hépat o-Gast r oent ér ol ogi e, 69495 Pi er r e-Béni t e, Fr ance
c Uni ver si t é de Caen, IFR 146 ICORE, EA 3919 (Labor at oi r e de Bi ol ogi e Mol écul ai r e et Cel l ul ai r e de l a Si gnal i sat i on),
UFR de Médeci ne, Avenue de l a Côt e de Nacr e, 14032 Caen cedex, Fr ance
Summary
Pat ient s wit h chronic colit is (ulcerat ive colit is or colonic Crohn’s disease) have an
increased risk of colorect al cancer (CRC). Alt hough most of t he molecular alt erat ions
report ed in sporadic CRC have also been observed in colit is-associat ed CRC, t hey do not
occur at t he same t iming and f requency, indicat ing a dif f erent pat hophysiology. In part i-
cular, recent work highlight ed t he import ance of chronic mucosal in fl ammat ion as a key
f act or f avouring colorect al carcinogenesis in t hese pat ient s. This may also be one of t he
reasons explaining t he role of 5-aminosalicylat es as chemoprevent ive agent s f or CRC in
infl ammat ory bowel disease (IBD) pat ient s wit h colonic involvement . Beside chemopre-
vent ion, colonoscopic screening and surveillance have been shown t o be t he cornerst one
f or CRC prevent ion and early det ect ion in t his part icular pat ient s’ populat ion. Periodic
surveillance colonoscopy t o det ect dysplasia has been shown t o decrease t he mort alit y
at t ribut ed t o CRC. More recent ly, progress in imaging t echniques increased our abilit y
t o ident if y dysplasia, and should probably now be considered t o be an int egral part of
surveillance colonoscopy. In t he f ut ure, f urt her improvement of our knowledge of CRC
biology, refi nement of imaging t echniques, as well as molecular discovery (e. g. ident i fi -
cat ion of speci fi c mut at ions in st ool DNA ext ract s), might lead t o develop more accurat e
diagnost ic st rat egies t o reduce t he morbidit y and mort alit y relat ed t o CRC in pat ient s
wit h ulcerat ive colit is or colonic Crohn’s disease.
© 2009 Elsevier Masson SAS. All right s reserved.
* Corresponding aut hor:
E-mai l addr ess: reimund-j m@chu-caen. f r (J. -M. Reimund).
© 2009 Elsevier Masson SAS. Tous droit s réservés.
, Colon cancer in in fl ammat ory bowel disease S191
Résumé
Le cancer colorect al (CCR) est plus f réquent en cas de rect ocolit e hémorragique et de
maladie de Crohn colique. Si la plupart des alt érat ions moléculaires ident i fi ées dans le
CCR sporadique ont également ét é t rouvées dans les CCRs compliquant une maladie in-
fl ammat oire chronique int est inale (MICI) de localisat ion colique, il est bien démont ré que
leur f réquence dif f ère signi fi cat ivement dans ces deux sit uat ions et qu’ elles sur viennent
à des moment s dist inct s de la carcinogenèse colique, indiquant des mécanismes phy-
siopat hologiques dif f érent s. Des t ravaux récent s ont en part iculier souligné le rôle de
l’ infl ammat ion muqueuse comme f act eur clé dans la cancérogenèse au cours des MICI.
Celle-ci explique aussi, au moins part iellement , les ef f et s prot ect eurs des dérivés
5-aminosalicylés. À côt é de la chimioprévent ion par les 5-aminosalicylés, le dépist age
et la surveillance endoscopiques demeurent la pierre angulaire de la prévent ion du CCR
compliquant une MICI. Réalisée suivant des recommandat ions précises, cet t e surveil-
lance permet de diminuer la mort alit é par CCR au cours des MICI. Plus récemment ,
l’ améliorat ion des t echniques endoscopiques (avec en part iculier le développement des
colorat ions à l’ indigo carmin ou au bleu de mét hylène) a accru not re capacit é à ident i-
fi er les lésions dysplasiques. Ces t echniques sont donc désormais part ie int égrant e de
la st rat égie de surveillance. La meilleure compréhension de la physiopat hologie du CCR
compliquant les MICI, l’ améliorat ion des t echniques endoscopiques, et le développement
de nouveaux out ils moléculaires, conduiront cert ainement au cours des prochaines
années, à opt imiser nos capacit és diagnost iques, et ainsi, à réduire la morbi-mort alit é
liée au CCR compliquant les MICI.
© 2009 Elsevier Masson SAS. Tous droit s réservés.
Pat ient s present ing wit h in fl ammat ory bowel disease [ IBDs: 1. 91-3. 97] [ 2] . A lower risk has also been report ed in a small
ul cerat ive col it is (UC) and Crohn’ s disease (CD)] have an populat ion st udy f rom Olmst ed Count y in t he Unit ed St at es
i ncr eased r i sk of devel opi ng col or ect al cancer (CRC), (378 pat ient s wit h UC diagnosed bet ween 1940 and 2001):
est imat ed t o be gl obal l y 2 t o 5 t imes higher t han in t he only 6 CRCs were diagnosed, yielding a 30-year cumulat ive
general populat ion of t he same age group. Several st udies probabil it y of CRC of 2% in t his cohort , not st at ist ical l y
are endeavouring t o improve knowledge on pat hophysiology, di f f er ent f r om CRC r i sk i n t he non-IBD popul at i on [ 3] .
risk f act ors, screening and prevent ion of CRC. Surprisingly, no case of CRC was f ound in UC pat ient s whose
disease was diagnosed af t er 1980 [ 3] . This result was rein-
Colorectal cancer risk f orced by a larger st udy f rom Denmark (22, 290 person-years
compared t o 5, 567 person-years in Wint her et al. st udy [ 4] )
Epidemiology report ing a 30-year cumul at ive probabil it y of CRC in UC
pat ient s of 2. 1%, a risk not dif f erent t han in t he general
populat ion [ 4] . Finally, in a st udy perf ormed in a ref erence
Schemat ical l y, 1 t o 2% of CRCs observed in t he general
cent re (St . Mark’ s Hospit al , Unit ed Kingdom), Rut t er et
populat ion are a complicat ion of IBD. Nevert heless, t he real
al . report ing t heir 30 years experience on col onoscopic
magnit ude of t he risk of dysplasia and CRC remains dif fi cult
t o est imat e in UC and CD pat ient s, due t o t he import ant surveillance in UC, f ound a cumulat ive incidence of CRC of
var i at i on of t he dat a r epor t ed i n t he l i t er at ur e. These 2. 5% at 20 years, 7. 6% at 30 years, and 10. 8% af t er 40 years
discrepant resul t s are a consequence of several f act ors of disease [ 5] . This change in risk magnit ude has been prin-
such as het erogeneous st udy design, dif f erences in case cipally at t ribut ed t o a more widespread use of surveillance
defi nit ion, and ref erral cent re bias. In UC pat ient s, based colonoscopy, a more f requent use of chemoprevent ion, and
on a met a-analysis published in 2001, t he mean prevalence t he f act t hat surgery was more of t en used in UC t reat ment
of CRC risk has been est imat ed at 3. 7% independent ly of st rat egy. Nevert heless, t he in fl uence of ot her f act ors (i. e.
disease ext ent , rising t o 5. 4%in t he case of pancolit is [ 1] . By environment al f act ors) cannot be excl uded. At t he l ast
pooling t he result s of t he st udies report ing dat a on disease Unit ed European Gast roent erol ogy Week (UEGW) hel d in
durat ion by decade, t he aut hors cal cul at e a cumul at ive Oct ober 2008, Lut gens et al. [ 6] report ed result s of a met a-
probabilit y of CRC of 2% af t er 10 years of disease, 8% af t er analysis t hey perf ormed, including 48 publicat ions crit ically
20 years, and 18% af t er 30 years of disease [ 1] . However, appraised f or t ype of st udy populat ion, person years at risk,
st udies published more recent ly, especially t hose report ing disease localisat ion in CD pat ient s, and censoring f or colec-
general populat ion dat a, suggest t hat t his risk may probably t omy. Cumulat ive risk in all IBD pat ient s in populat ion-based
be lower or decreased over t ime. For example, in a large st udies were 1%, 2%, and 5% af t er 10, 20, and more t han
popul at ion-based st udy perf ormed by Bernst ein et al . in 30 years of disease respect ively, wit h a pooled st andardised
Manit oba (Canada), t he incidence rat e rat io f or developing morbidit y rat io (SMR) of 3. 6 (95% CI: 3. 1-4. 1), compared t o
CRC in UC pat ient s compared t o t he general popul at ion, 1%, 11%, and 43%, and a pooled SMR of 8. 8 (95%CI: 7. 3-11) in
was f ound t o be of 2. 75 [ 95% con fi dence int erval (95% CI): ref erral cent re st udies [ 6] . For UC pat ient s, ref erral cent re
Colon cancer in inflammatory bowel disease:
recent trends, questions and answers
Cancer colique au cours des maladies inflammatoires
chroniques intestinales : actualité et perspectives
S. Viennot a, A. Deleporte a, D. Moussatab, S. Nanceyb, B. Flourié b,
J. -M. Reimunda, c, *
a Cent re Hospit al ier Universit aire de Caen, Service d’ Hépat o-Gast roent érol ogie et Nut rit ion, Pôl e Reins-Digest if -Nut rit ion,
Hôpi t al Côt e de Nacr e, B. P. 95182, 14033 Caen cedex 9, Fr ance
b Cent r e Hospi t al i er Lyon-Sud, Ser vi ce d’ Hépat o-Gast r oent ér ol ogi e, 69495 Pi er r e-Béni t e, Fr ance
c Uni ver si t é de Caen, IFR 146 ICORE, EA 3919 (Labor at oi r e de Bi ol ogi e Mol écul ai r e et Cel l ul ai r e de l a Si gnal i sat i on),
UFR de Médeci ne, Avenue de l a Côt e de Nacr e, 14032 Caen cedex, Fr ance
Summary
Pat ient s wit h chronic colit is (ulcerat ive colit is or colonic Crohn’s disease) have an
increased risk of colorect al cancer (CRC). Alt hough most of t he molecular alt erat ions
report ed in sporadic CRC have also been observed in colit is-associat ed CRC, t hey do not
occur at t he same t iming and f requency, indicat ing a dif f erent pat hophysiology. In part i-
cular, recent work highlight ed t he import ance of chronic mucosal in fl ammat ion as a key
f act or f avouring colorect al carcinogenesis in t hese pat ient s. This may also be one of t he
reasons explaining t he role of 5-aminosalicylat es as chemoprevent ive agent s f or CRC in
infl ammat ory bowel disease (IBD) pat ient s wit h colonic involvement . Beside chemopre-
vent ion, colonoscopic screening and surveillance have been shown t o be t he cornerst one
f or CRC prevent ion and early det ect ion in t his part icular pat ient s’ populat ion. Periodic
surveillance colonoscopy t o det ect dysplasia has been shown t o decrease t he mort alit y
at t ribut ed t o CRC. More recent ly, progress in imaging t echniques increased our abilit y
t o ident if y dysplasia, and should probably now be considered t o be an int egral part of
surveillance colonoscopy. In t he f ut ure, f urt her improvement of our knowledge of CRC
biology, refi nement of imaging t echniques, as well as molecular discovery (e. g. ident i fi -
cat ion of speci fi c mut at ions in st ool DNA ext ract s), might lead t o develop more accurat e
diagnost ic st rat egies t o reduce t he morbidit y and mort alit y relat ed t o CRC in pat ient s
wit h ulcerat ive colit is or colonic Crohn’s disease.
© 2009 Elsevier Masson SAS. All right s reserved.
* Corresponding aut hor:
E-mai l addr ess: reimund-j m@chu-caen. f r (J. -M. Reimund).
© 2009 Elsevier Masson SAS. Tous droit s réservés.
, Colon cancer in in fl ammat ory bowel disease S191
Résumé
Le cancer colorect al (CCR) est plus f réquent en cas de rect ocolit e hémorragique et de
maladie de Crohn colique. Si la plupart des alt érat ions moléculaires ident i fi ées dans le
CCR sporadique ont également ét é t rouvées dans les CCRs compliquant une maladie in-
fl ammat oire chronique int est inale (MICI) de localisat ion colique, il est bien démont ré que
leur f réquence dif f ère signi fi cat ivement dans ces deux sit uat ions et qu’ elles sur viennent
à des moment s dist inct s de la carcinogenèse colique, indiquant des mécanismes phy-
siopat hologiques dif f érent s. Des t ravaux récent s ont en part iculier souligné le rôle de
l’ infl ammat ion muqueuse comme f act eur clé dans la cancérogenèse au cours des MICI.
Celle-ci explique aussi, au moins part iellement , les ef f et s prot ect eurs des dérivés
5-aminosalicylés. À côt é de la chimioprévent ion par les 5-aminosalicylés, le dépist age
et la surveillance endoscopiques demeurent la pierre angulaire de la prévent ion du CCR
compliquant une MICI. Réalisée suivant des recommandat ions précises, cet t e surveil-
lance permet de diminuer la mort alit é par CCR au cours des MICI. Plus récemment ,
l’ améliorat ion des t echniques endoscopiques (avec en part iculier le développement des
colorat ions à l’ indigo carmin ou au bleu de mét hylène) a accru not re capacit é à ident i-
fi er les lésions dysplasiques. Ces t echniques sont donc désormais part ie int égrant e de
la st rat égie de surveillance. La meilleure compréhension de la physiopat hologie du CCR
compliquant les MICI, l’ améliorat ion des t echniques endoscopiques, et le développement
de nouveaux out ils moléculaires, conduiront cert ainement au cours des prochaines
années, à opt imiser nos capacit és diagnost iques, et ainsi, à réduire la morbi-mort alit é
liée au CCR compliquant les MICI.
© 2009 Elsevier Masson SAS. Tous droit s réservés.
Pat ient s present ing wit h in fl ammat ory bowel disease [ IBDs: 1. 91-3. 97] [ 2] . A lower risk has also been report ed in a small
ul cerat ive col it is (UC) and Crohn’ s disease (CD)] have an populat ion st udy f rom Olmst ed Count y in t he Unit ed St at es
i ncr eased r i sk of devel opi ng col or ect al cancer (CRC), (378 pat ient s wit h UC diagnosed bet ween 1940 and 2001):
est imat ed t o be gl obal l y 2 t o 5 t imes higher t han in t he only 6 CRCs were diagnosed, yielding a 30-year cumulat ive
general populat ion of t he same age group. Several st udies probabil it y of CRC of 2% in t his cohort , not st at ist ical l y
are endeavouring t o improve knowledge on pat hophysiology, di f f er ent f r om CRC r i sk i n t he non-IBD popul at i on [ 3] .
risk f act ors, screening and prevent ion of CRC. Surprisingly, no case of CRC was f ound in UC pat ient s whose
disease was diagnosed af t er 1980 [ 3] . This result was rein-
Colorectal cancer risk f orced by a larger st udy f rom Denmark (22, 290 person-years
compared t o 5, 567 person-years in Wint her et al. st udy [ 4] )
Epidemiology report ing a 30-year cumul at ive probabil it y of CRC in UC
pat ient s of 2. 1%, a risk not dif f erent t han in t he general
populat ion [ 4] . Finally, in a st udy perf ormed in a ref erence
Schemat ical l y, 1 t o 2% of CRCs observed in t he general
cent re (St . Mark’ s Hospit al , Unit ed Kingdom), Rut t er et
populat ion are a complicat ion of IBD. Nevert heless, t he real
al . report ing t heir 30 years experience on col onoscopic
magnit ude of t he risk of dysplasia and CRC remains dif fi cult
t o est imat e in UC and CD pat ient s, due t o t he import ant surveillance in UC, f ound a cumulat ive incidence of CRC of
var i at i on of t he dat a r epor t ed i n t he l i t er at ur e. These 2. 5% at 20 years, 7. 6% at 30 years, and 10. 8% af t er 40 years
discrepant resul t s are a consequence of several f act ors of disease [ 5] . This change in risk magnit ude has been prin-
such as het erogeneous st udy design, dif f erences in case cipally at t ribut ed t o a more widespread use of surveillance
defi nit ion, and ref erral cent re bias. In UC pat ient s, based colonoscopy, a more f requent use of chemoprevent ion, and
on a met a-analysis published in 2001, t he mean prevalence t he f act t hat surgery was more of t en used in UC t reat ment
of CRC risk has been est imat ed at 3. 7% independent ly of st rat egy. Nevert heless, t he in fl uence of ot her f act ors (i. e.
disease ext ent , rising t o 5. 4%in t he case of pancolit is [ 1] . By environment al f act ors) cannot be excl uded. At t he l ast
pooling t he result s of t he st udies report ing dat a on disease Unit ed European Gast roent erol ogy Week (UEGW) hel d in
durat ion by decade, t he aut hors cal cul at e a cumul at ive Oct ober 2008, Lut gens et al. [ 6] report ed result s of a met a-
probabilit y of CRC of 2% af t er 10 years of disease, 8% af t er analysis t hey perf ormed, including 48 publicat ions crit ically
20 years, and 18% af t er 30 years of disease [ 1] . However, appraised f or t ype of st udy populat ion, person years at risk,
st udies published more recent ly, especially t hose report ing disease localisat ion in CD pat ient s, and censoring f or colec-
general populat ion dat a, suggest t hat t his risk may probably t omy. Cumulat ive risk in all IBD pat ient s in populat ion-based
be lower or decreased over t ime. For example, in a large st udies were 1%, 2%, and 5% af t er 10, 20, and more t han
popul at ion-based st udy perf ormed by Bernst ein et al . in 30 years of disease respect ively, wit h a pooled st andardised
Manit oba (Canada), t he incidence rat e rat io f or developing morbidit y rat io (SMR) of 3. 6 (95% CI: 3. 1-4. 1), compared t o
CRC in UC pat ient s compared t o t he general popul at ion, 1%, 11%, and 43%, and a pooled SMR of 8. 8 (95%CI: 7. 3-11) in
was f ound t o be of 2. 75 [ 95% con fi dence int erval (95% CI): ref erral cent re st udies [ 6] . For UC pat ient s, ref erral cent re
