Di e en ially De egula ed Mic oRNAs as No el
Bioma ke s o Neoplas ic P og ession in
Ulce a i e Coli is
Isabel Quin anilla, PhD
1
,
2
,
3
,
*, Ge ha d Jung, PhD
1
,
2
,
4
,
5
,
*, Mi eya Jimeno, PhD
1
,
6
, Juan Jos ´
e Lozano, PhD
2
,
7
, Julia Sido o a, PhD
2
,
7
,
Jo di Camps, PhD
1
,
2
, Sabela Ca ballal, PhD
1
,
2
,
4
,
5
, Luis Bujanda, PhD
2
,
8
, Ma ia Isabel Ve a, PhD
9
, En ique Quin e o, PhD
10
,
Ma a Ca illo-Palau, PhD
10
, Mi iam Cua ecasas, PhD
1
,
2
,
11
, An oni Cas ells, PhD
1
,
2
,
4
,
5
, Juli `
a Pan ´
es, PhD
1
,
2
,
4
,
5
, Elena Rica , PhD
1
,
2
,
4
,
5
,
Le icia Mo ei a, PhD
1
,
2
,
4
,
5
, F ancesc Balague , PhD
1
,
2
,
4
,
5
and Ma ia Pellis ´
e, PhD
1
,
2
,
4
,
5
INTRODUCTION: Colo ec al cance (CRC) is a po en ially li e- h ea ening complica ion o long-s anding ulce a i e coli is
(UC). Mic oRNAs (miRNA) a e epigene ic egula o s ha ha e been in ol ed in he de elopmen o UC-
associa ed CRC. Howe e , hei ole as po en ial mucosal bioma ke s o neoplas ic p og ession has no
been adequa ely s udied.
METHODS: In his s udy, we analyzed he exp ession o 96 p eselec ed miRNAs in human o malin- ixed and
pa a in-embedded issue o 52 case biopsies (20 no mal mucosa, 20 dysplasia, and 12 UC-associa ed
CRCs) and 50 con ol biopsies (10 no mal mucosa, 21 spo adic adenomas, and 19 spo adic CRCs) by
using Cus om TaqMan A ay Ca ds. Fo alida ion o de egula ed miRNAs, we pe o med indi idual
quan i a i e eal- ime polyme ase chain eac ion in an independen coho o 50 cases (13 no mal
mucosa, 25 dysplasia, and 12 UC-associa ed CRCs) and 46 con ols (7 no mal mucosa, 19 spo adic
adenomas, and 20 spo adic CRCs).
1
Ins i u d’In es igacions Biom `
ediques Augus Pi i Sunye (IDIBAPS), Ba celona, Spain;
2
Cen o de In es igaci ´
on Biom ´
edica en Red, En e medades Hep ´
a icas y
Diges i as (CIBERehd), Ba celona, Spain;
3
Na ional Cance Ins i u e, NIH, Be hesda, Ma yland, USA;
4
Depa men o Gas oen e ology, Hospi al Clinic, Ba celona,
Spain;
5
Facul y o Medicine, Uni e si y o Ba celona, Ba celona, Spain;
6
Depa amen o Pa hology, Hospi al Uni e si a i Ge mans T ias i Pujol, Badalona, Spain;
7
Bioin o ma ics Pla o m, CIBEREHD, Ba celona, Spain;
8
Depa men o Gas oen e ology, Biodonos ia Heal h Resea ch Ins i u e, Uni e sidad del País Vasco
(UPV/EHU), San Sebas i ´
an, Spain;
9
Depa men o Gas oen e ology, Uni e si y Hospi al Pue a de Hie o Majadahonda, Mad id, Spain;
10
Depa men o
Gas oen e ology, Uni e si y Hospi al o he Cana y Islands, San a C uz de Tene i e, Spain;
11
Depa men o Pa hology, Hospi al Clinic, Ba celona, Spain.
Co espondence: Ma ia Pellis ´
e, PhD. E-mail: [email p o ec ed].
*Isabel Quin anilla and Ge ha d Jung con ibu ed equally o his wo k.
Recei ed No embe 18, 2021; accep ed Ma ch 4, 2022; published online Ap il 8, 2022
© 2022 The Au ho (s). Published by Wol e s Kluwe Heal h, Inc. on behal o The Ame ican College o Gas oen e ology
Ame ican College o Gas oen e ology Clinical and T ansla ional Gas oen e ology
ARTICLE 1
INFLAMMATORY BOWEL DISEASE
RESULTS: Six y- ou miRNAs we e ound o be di e en ially de egula ed in he UC-associa ed CRC sequence.
Eigh o hese miRNAs we e chosen o u he alida ion. We con i med miR-31, -106a, and -135b o
be signi ican ly de egula ed be ween no mal mucosa and dysplasia, as well as ac oss he UC-associa ed
CRC sequence (all P<0.01). No ably, hese miRNAs also con i med o ha e a signi ican di e en ial
exp ession compa ed wi h spo adic CRC (all P<0.05).
DISCUSSION: UC-associa ed and spo adic CRCs ha e dis inc miRNA exp ession pa e ns, and some miRNAs
indica e ea ly neoplas ic p og ession.
SUPPLEMENTARY MATERIAL accompanies his pape a h p://links.lww.com/CTG/A803,h p://links.lww.com/CTG/A804,h p://links.lww.com/CTG/A805,h p://
links.lww.com/CTG/A806, and h p://links.lww.com/CTG/A807.
Clinical and T ansla ional Gas oen e ology 2022;13:e00489. h ps://doi.o g/10.14309/c g.0000000000000489
INTRODUCTION
Inflamma o y bowel disease (IBD) has been associa ed o a highe
isk o de eloping colo ec al cance (CRC) (1,2), and a ecen
la ge popula ion-based coho s udy concluded ha he isk o
diagnosis o CRC in pa ien s wi h IBD has no declined signifi-
can ly in he pas 35 yea s (3).
Cu en su eillance s a egies aim o iden i y dysplasia, which
is conside ed a p emalignan lesion and hus associa ed wi h a
high isk o de elop CRC. Al hough he e is only e idence om
case se ies and coho s udies, mos o he socie ies ecommend
1- o 5-yea ly colonoscopy o su eillance s a ing om 8 o 10
yea s o disease onse , unless specific isk ac o s o dysplasia,
such as p ima y scle osing cholangi is, a e also p esen (4,5).
Howe e , he e a e se e al sho comings when i comes o
de ec and diagnose dysplasia accu a ely. Fi s , hei low de-
ec ion a es du ing su eillance colonoscopy (6,7). In ac ,
in e al cance s a e significan ly mo e equen in IBD-
associa ed cases compa ed wi h spo adic cases and a e mos
likely due o no de ec ed o no comple ely esec ed dysplas ic
lesions (8). Second, he poo in e obse e ag eemen among
bo h endoscopis s and pa hologis s o dis inguish dysplas ic
om inflamma ion-associa ed mo phologic al e a ions. Thi d,
his poo in e obse e ag eemen is also applicable when di -
e en ia ing low-g ade dysplasia om high-g ade dysplasia (9).
And ou h, he difficul y in eliably dis inguishing coli is-
associa ed dysplasia om spo adic adenomas by his opa hol-
ogis s (10,11).
The difficul ies o accu a ely de ec and diagnose dysplasia as
he p emalignan lesion o coli is-associa ed cance ha e led o he
sea ch o mo e obus , objec i e, and minimally o nonin asi e
bioma ke s (12). Mic oRNAs (miRNAs) a e ideal candida es o
bioma ke s because o hei small size, s abili y in biological
samples, a ailabili y in blood samples (ci cula ing miRNAs),
abili y o egula e hund eds o mRNAs, and hei ela i ely small
o al numbe compa ed wi h mRNAs. Mo eo e , miRNAs ha e
been linked o pa hogenic p ocesses such as inflamma ion sig-
naling, endo helial-mesenchymal ansi ion, cance s em cells,
and me as a iza ion (13–15).
In UC-associa ed CRC (UC-CRC), he e idence abou he ole
o miRNAs in inflamma ion and ca cinogenesis is less abundan
compa ed wi h spo adic CRC (Sp-CRC), bu se e al s udies ha e
ound p omising p elimina y esul s (16–18). Al hough mos o
he p e ious s udies had ocused on disease-specific exp ession
pa e ns in bo h ulce a i e coli is (UC) and C ohn’s disease, some
ecen s udies aimed o iden i y diffe en ially exp essed miRNAs
in UC-CRC (19–22). Howe e , he e is only a hand ul o s udies
ha analyzed single miRNAs associa ed o UC-CRC, bu hei
sample sizes we e a he low, hey we e no alida ed wi h in-
dependen coho s, dysplasia was no conside ed, o specifici y
o UC-associa ed CRC o e spo adic CRC was no add essed.
Despi e he bu geoning knowledge o miRNAs as egula o s in
bo h main ained inflamma ion and inflamma ion-associa ed
ca cinogenesis, hese p ocesses a e no ully unde s ood, and
mo e esea ch is needed o find bioma ke s o clinical use. Ou
s udy aimed o find miRNAs ha a e diffe en ially and specifically
exp essed in he colo ec al ca cinogenesis associa ed wi h UCand
could se e as bioma ke s o imp o e he de ec ion o UC-
associa ed dysplasia (UC-Dys).
MATERIAL AND METHODS
Pa ien s and sample coho s
The s udy g oup included 103 issue specimens om 94 pa ien s
wi h UC who we e e e ed o dysplasia sc eening o diagnosed wi h
a CRC a ising on a coli ic a ea and ulfilling all he ollowing eligi-
bili y c i e ia: (i) endoscopically and pa hologically confi med UC
p oximal o he ec um, (ii) disease du a ion o a leas 8 yea s, and
(iii) absence o clinical ac i i y (i.e., nonin asi e 6-poin pa ial Mayo
Sco e (23) ,3), o a oid bias in oduced by inflamma o y p ocesses.
To be conside ed as a UC-associa ed lesion, dysplasias and CRCs
had obea isenina easp e iouslyaffec ed by ch onic in-
flamma ion. The con ol g oup included 96 issue specimens om
96 diffe en indi iduals om he FIT-based o ganized Ba celona-
Eixample-Esque a popula ion CRC Sc eening P og am, in which
all indi iduals aged 50–69 yea s a e in i ed o pa icipa e, and a FIT
cu offo $20 mg o hemoglobin/g o eces is used o indica e a
colonoscopy. Pe sonal his o y o CRC, adenoma, o inflamma o y
bowel disease, a amily his o y o CRC (defined as hose indi iduals
wi h 2 fi s -deg ee ela i es wi h CRC o 1 diagnosed be o e he age
o 60), known he edi a y CRC synd omes, se e e coexis ing illness,
colonoscopy pe o med wi hin he pas 5 yea s, p e ious colec omy,
and con aindica ion o colonoscopy a e conside ed defini i e o
empo a y exclusion c i e ia o sc eening.
The samples om he s udy g oup and he con ol g oup we e
each di ided o o m 2 independen coho s: (i) a disco e y coho
wi h 52 samples om he s udy g oup (20 UC no mal mucosa
biopsies [UC-NM], 20 UC-associa ed dysplas ic lesions [UC-
Dys], and 12 UC-associa ed colo ec al cance s [UC-CRC]) and
50 samples om he con ol g oup (10 no mal mucosa biopsies
[Sp-NM], 21 spo adic adenomas [Sp-Ad], and 19 spo adic CRC
[Sp-CRC]) (Figu e 1). All samples de i ed om diffe en unique
pa ien s excep o 3 pa ien s o he s udy g oup who p o ided 1
sample o no mal mucosa and 1 sample o dysplasia and 2
Clinical and T ansla ional Gas oen e ology VOLUME 13 | JULY 2022 www.clin anslgas o.com
INFLAMMATORY BOWEL DISEASE
Quin anilla e al.2
pa ien s who p o ided 2 samples o dysplasia. (ii) The alida ion
coho was composed o 51 samples om he s udy g oup: 14 UC-
NM, 25 UC-Dys, and 12 UC-CRC and 46 samples om he
con ol g oup: 7 Sp-NM, 19 Sp-Ad, and 20 Sp-CRC (Figu e 1). All
samples de i ed om diffe en pa ien s excep o 4 pa ien s who
p o ided 2 samples o dysplasia. Spo adic adenomas in pa ien s
wi h UC we e dis ega ded in his s udy.
Tissue samples o no mal mucosa, UC-associa ed dysplasia,
and spo adic adenoma we e ob ained by colonoscopy, whe eas
CRC specimens de i ed om su gical blocks. All samples we e
fixed by imme sion in buffe ed o malin immedia ely a e he
colonoscopy o su ge y and hen p ocessed and embedded in
pa affin (FFPE). All specimens we e e alua ed by pa hologis s a
each pa icipa ing ins i u ion, acco ding o he Vienna classifi-
ca ion and he se en h edi ion o he Ame ican Join Commi ee
on Cance umo , node, me as asis g ading sys em. Be o e RNA
ex ac ion, mic odissec ion o ele an issue was pe o med by
using a biopsy punch de ice.
RNA ex ac ion om FFPE specimens
Reco e All To al Nucleic Acid Isola ion Ki o FFPE ( e . AM1975,
In i ogen; The mo Fishe Scien ific, Wal ham, MA) was used o
ex ac o al RNA om issue co es, acco ding o he manu ac u e ’s
p o ocol. RNA concen a ion was de e mined wi h a NanoD op
1000 spec opho ome e (NanoD op, Wilming on, DE).
Quan i a i e e e se- ansc ip ion PCR assays
Fo he disco e y phase, 96 miRNAs we e selec ed a e a com-
p ehensi e li e a u e and da abase esea ch based on hei p e-
iously desc ibed unc ions in cell cycle con ol, signal ansduc ion,
cell-cell in e ac ion, inflamma ion, and umo igenesis, as well as
hei po en ial bioma ke ole in colo ec al cance o coli is-
associa ed cance . We used 96-well Cus om TaqMan A ay
Mic ofluidic Ca ds ( e . 4342261; Applied Biosys ems, The mo
Fishe Scien ific, Wal ham, MA) o he e alua ion o miRNA ex-
p ession le els in pa ien samples. All RNA samples we e fi s
e e se- ansc ibed o cDNA by using he TaqMan Mic oRNA
Re e se-T ansc ip ion Ki ( e . 4366596; Applied Biosys ems)
acco ding o he manu ac u e ’s ins uc ions. B iefly, 100 ng o
RNA was e e se- ansc ibed in a final olume o 15 mLpe eac ion
unde he ollowing condi ions: 30 minu es a 16°C, 30 minu es a
42°C, and 5 minu es a 85°C. Then, he cDNA was p eamplified o
14 cycles using he TaqMan P eAmp Mas e Mix ( e . 4488593;
Applied Biosys ems) combined wi h he co esponding Cus om
TaqMan P eAmp p ime pool. The p eamplified cDNA was ei he
di ec ly p ocessed o s o ed a 220°C bu o no longe han 64
hou s. Fo quan i a i e eal- ime polyme ase chain eac ion (qRT-
PCR), each o he 4 mic ofluidic ca d’spo swe efilled wi h 100 mL
o eac ion olume om a mas e mix consis ing o 225 mLTaqMan
Uni e sal Mas e Mix II No UNG ( e . 4440048; Applied Bio-
sys ems), 4.5 mLo p eamplified and 1:4 dilu ed cDNA (a a con-
cen a ion o 0.142 ng/mL), and nuclease- ee wa e up o 450 mL.
The qRT-PCR was un on a 7900HT Fas Real-Time PCR in-
s umen ( e . 4351405; Applied Biosys ems) using he ollowing
condi ions: hold o 10 minu es a 95°C, hen 15 seconds a 95°C,
and 1 minu e a 60°C o 40 cycles. All qRT-PCR eac ions we e
pe o med in duplica e o each sample and miRNA assay. C
alues we e calcula ed om au oma ic h eshold. RNU66, RNU48,
RNU44, and RNU6b we e ini ially included as endogenous con-
ols o no maliza ion, bu da a analysis showed ha miR-30a-5p,
miR-30e, and miR-28 we e mo e s ably exp essed ac oss all sam-
ples and we e he e o e used as be e con ols o ou s udy.
TaqMan Mic oRNA Assays ( e . 4449142; Applied Biosys ems)
we e used o alida e diffe en ial miRNA exp ession in 97 samples
by qRT-PCR. RNA was dilu ed o 4 ng/mL, and 10 ng was used as a
empla e o each e e se- ansc ip ion eac ion. All qRT-PCR
eac ions we e pe o med on a ViiA 7 Real-Time PCR ins umen
( e . 4453534; Applied Biosys ems) in iplica e o each sample
and miRNA. The mal condi ions we e as ollows: hold 10 minu es
a 95°C, hen 15 seconds a 95°C, and 1 minu e a 60°C o up o 50
cycles o de ec any exp ession possible. CT alues we e calcula ed
om au oma ic h eshold, and only hose below 40 we e ega ded
o final analysis. As in he disco e y phase, miR-30a-5p, miR-30e,
and miR-28 we e used as endogenous con ols.
S a is ical analysis
S a is ics o clinical and pa hological cha ac e is ics we e calcu-
la ed wi h G aphPad P ism ( e sion 9.1.2). The unpai ed S uden
es and Fishe exac es we e used whe e applicable o di -
e ences be ween s udy g oups. All s a is ics o expe imen al da a
including all plo s we e calcula ed and gene a ed wi h R so wa e.
We did no pe o m an in e pla e no maliza ion because he
p incipal componen analysis did no show a ba ch effec .
Quan i a i e RT-PCR esul s we e analyzed by s anda d 2-class
unpai ed Welch es and ANOVA whe e applicable. Finally, o
each selec ed miRNA, box and whiske plo s and ecei e ope -
a ing cha ac e is ic (ROC) cu es we e gene a ed, and he a ea
unde he cu e (AUC) was analyzed o es he pe o mance o
he diffe en models.
RESULTS
Cha ac e is ics o he s udy coho s
Pa ien s included in he disco e y and alida ion coho s p e-
sen ed simila dis ibu ion o sex and age a s udy inclusion and
diagnosis be ween all cases and con ols. Howe e , o he dis-
co e y phase, pa ien s o he s udy g oup we e sligh ly younge
(52 s 57 yea s, P50.02), and ad anced-s age umo s (III/IV)
we e wice as equen in he UC-associa ed CRC g oup com-
pa ed wi h he spo adic CRC g oup (41.6 s 21%, espec i ely),
bu which was no s a is ically significan (Table 1). Fo a ew
pa ien s in bo h s udy g oups, mo e han 1 sample was a ailable.
P ecisely, o 3 pa ien s, 1 sample o no mal mucosa and 1 sample
o dysplasia we e a ailable; o 6 pa ien s, 2 samples o dysplasia
we e a ailable. By con as , all samples o bo h con ol g oups
came om diffe en pa ien s, esul ing in a o al o 102 samples
om 97 pa ien s in he disco e y coho and 97 samples om 93
pa ien s in he alida ion coho . Rega ding he g oup o UC-
associa ed dysplasia, in he disco e y coho , 18 and 2 o 20 we e
low-g ade and high-g ade dysplasia, espec i ely, whe eas in he
alida ion coho , 100% we e low-g ade dysplasia. All spo adic
adenomas we e ubula adenomas wi h low-g ade dysplasia ex-
cep o a single one in he disco e y coho ha was a ubulo-
illous adenoma wi h high-g ade dysplasia (Table 1 and Figu e 1).
MiRNAs a e di e en ially exp essed ac oss he coli is-associa ed
cance sequence
We ound 64 miRNAs ha we e diffe en ially de egula ed ac oss
he UC-associa ed CRC sequence, whe eas 65 miRNAs we e al-
e ed in he spo adic CRC sequence (see Supplemen a y Table S1,
Supplemen a y Digi al Con en 1, h p://links.lww.com/CTG/A803).
We compa ed he 2 sequences o diffe en ially exp essed miRNAs
Ame ican College o Gas oen e ology Clinical and T ansla ional Gas oen e ology
INFLAMMATORY BOWEL DISEASE
Di e en ially De egula ed Mic oRNAs 3
(ANOVA P,0.05) and ound ha mos o he miRNAs we e
sha ed be ween he 2 sequences. Howe e , we also ound 9 miR-
NAs ha we e only diffe en ially egula ed in he UC-associa ed
sequence, whe eas 10 we e only significan o he spo adic se-
quence (see Supplemen a y Table S1, Supplemen a y Digi al
Con en 1, h p://links.lww.com/CTG/A803). We nex fil e ed he
significan ly de egula ed miRNAs by hose who showed he la ges
diffe ence in ela i e exp ession le els be ween no mal mucosa
(UC-NM) and UC-associa ed CRC (UC-CRC). Fo a diffe ence
g ea e han 61.0, we e ie ed 21 miRNAs ha we e up egula ed
and 11 ha we e down egula ed (see Supplemen a y Table S1,
Supplemen a y Digi al Con en 1, h p://links.lww.com/CTG/A803).
In e es ingly, wi hin each g oup (i.e., up egula ed and down-
egula ed), hey showed diffe en ypes o change in exp ession le el.
Fo ins ance, wi hin he up egula ed g oup, al hough some we e only
o e exp essed in he cance o in hedysplasiawi h espec ono mal
mucosa, o he s showed a consis en up egula ion ac oss he UC-
associa ed sequence (Figu e 2). We also compa ed he miRNA
exp ession p ofiles be ween UC-CRCs and Sp-CRCs and
e ie ed 47 miRNAs ha showed s a is ically significan diffe -
ences be ween bo h en i ies. Thi y o hese miRNAs showed a
mo e han 2- old diffe ence, 9 o which we e highe exp essed in
he UC-associa ed CRCs, whe eas 21 we e highe exp essed in he
spo adic CRCs (Figu e 3a; see Supplemen a y Table S2, Supple-
men a y Digi al Con en 2, h p://links.lww.com/CTG/A804).
miRNAs a e po en ial mucosal bioma ke s o dysplasia in
pa ien s wi h ulce a i e coli is
When compa ed wi h no mal colonic mucosa o pa ien s wi h
UC, dysplas ic lesions showed a dis inc miRNA exp ession
p ofile (Figu e 3b; see Supplemen a y Table S3, Supplemen a y
Digi al Con en 3, h p://links.lww.com/CTG/A805). In ac-
co dance wi h ou main objec i e o disco e ing issue bio-
ma ke s o iden ifica ion o dysplasia associa ed o UC and
which a he same ime we e highly specific o neoplas ic
p og ession associa ed o UC a he han he spo adic
adenoma-ca cinoma sequence, we used he ollowing c i e ia
o selec ing he miRNAs o he alida ion phase: (i) miRNAs
Figu e 1. S udy lowcha . AUROC, a ea unde he ecei e ope a ing cu e; Sp-Ad, spo adic adenoma; Sp-CRC, spo adic colo ec al cance ; Sp-NM,
spo adic no mal mucosa; UC, ulce a i e coli is; UC-CRC, ulce a i e coli is–associa ed colo ec al cance ; UC-Dys, ulce a i e coli is–associa ed dysplasia;
UC-NM, UC-associa ed no mal mucosa.
Clinical and T ansla ional Gas oen e ology VOLUME 13 | JULY 2022 www.clin anslgas o.com
INFLAMMATORY BOWEL DISEASE
Quin anilla e al.4
ha showed a significan de egula ion be ween no mal mucosa,
dysplasia, and cance , (ii) miRNAs ha showed a consis en de-
egula ion ac oss he UC-associa ed sequence meaning ha hose
miRNAs dys egula ed in he dysplasia cases ha did no show a
dys egula ion o highe magni ude in cance we e no ega ded as
meaning ul, (iii) miRNAs wi h a significan ly diffe en exp ession
le el in he coli is-associa ed cance s compa ed wi h he spo adic
cance s (Figu e 3A; see Supplemen a y Table S2, Supplemen a y
Digi al Con en 2, h p://links.lww.com/CTG/A804), and (i )
miRNAs ha showed a diffe en ial exp ession be ween no mal
mucosa o pa ien s wi h UC and UC-associa ed dysplasia
(Figu e 3B; see Supplemen a y Table S3, Supplemen a y Digi al
Table 1. Clinical cha ac e is ics o he pa ien s wi hin he di e en g oups
Disco e y (N 597) Valida ion (N 593)
S udy g oup Con ol g oup PS udy g oup Con ol g oup P
To al pa ien s, n 47 50 47 46
Sex
Female, n (%) 16 (34) 23 (46) 0.30 17 (36%) 23 (50) 0.21
Age
Mean age a inclusion, y (SD) 52 (13.8) 57 (3.1) 0.02 56 (12.2) 57 (3.5) 0.64
Mean age a diagnosis UC, y (SD) 36 (18.4) NA NA 38 (14.1) NA NA
UC-associa ed dysplasia, n 20 25
Low-g ade/high-g ade, n (%) 18/2 (90/10) 25/0 (100/0)
Spo adic adenoma, n 21 0.61 19 0.99
Low-g ade/high-g ade, n (%) 20/1 (95/5) 19/0 (100/0)
Sex
Female, n (%) 7 (35) 9 (43) 0.75 10 (40) 10 (53) 0.54
Age
Mean age a inclusion, y (SD) 53 (13.0) 58 (0.8) 0.08 56 (12.9) 58 (1.2) 0.51
Mean age a diagnosis UC, y (SD) 36 (14.6) NA NA 38 (14.7) NA NA
Ca cinomas, n 12 19 12 20
Sex
Female, n (%) 3 (25) 9 (47) 0.27 5 (42) 9 (45) 0.99
Age
Mean age a inclusion, y (SD) 56.9 (16.4) 54.4 (2.9) 0.52 60 (12.8) 55 (2.7) 0.09
Mean age a diagnosis CRC, y (SD) 57.8 (17.5) 54.4 (2.9) 0.42 58 (12.5) 55 (2.7) 0.32
D be ween diagnosis UC and CRC, y (SD) 13.5 (11.3) NA NA 14.9 (9.3) NA NA
TNM, n (%)
In si u 0 (0) 6 (31.6) 0 (0) 0 (0)
I 2 (16.7) 9 (47.4) 4 (33.3) 9 (45)
II 4 (33.3) 0 (0) 5 (41.7) 2 (10)
0.22 0.21
III 1 (8.3) 2 (10.5) 3 (25) 6 (30)
IV 4 (33.3) 2 (10.5) 0 (0) 1 (5)
Unknown 1 (8.3) 0 (0) 0 (0) 2 (10)
Loca ion
a
P oximal/dis al, n (%) 3/8 (25/67) 2/17 (11/89) 0.33 3/6 (25/50) 5/15 (25/75) 0.67
Unknown 1 (8) 0 (0) 3 (25) 0 (0)
Adju an ea men , n (%)
Yes/no 5/5 (42/42) 4/15 (21/79) 0.20 9/3 (75/25) 11/9 (55/45) 0.45
Unknown 2 (16.7) 0 (0) 0 (0) 0 (0)
CRC, colo ec al cance ; TNM, umo , node, me as asis; UC, ulce a i e coli is; NA, no applicable. P- alues ,0.05 a e highligh ed in bold.
a
P oximal and dis al o he splenic lexu e.
Ame ican College o Gas oen e ology Clinical and T ansla ional Gas oen e ology
INFLAMMATORY BOWEL DISEASE
Di e en ially De egula ed Mic oRNAs 5
Con en 3, h p://links.lww.com/CTG/A805). Acco dingly, we
only conside ed hose miRNAs o which he le els o specifici y
acco ding o hei ROC cu es we e a leas 0.7 o highe when
disc imina ing dysplasia om co esponding no mal mucosa.
We iden ified 6 miRNAs om he disco e y phase ha ul-
filled each o hese c i e ia: miR-20b, -24, -31, -106a, and -135b, all
o which we e up egula ed ac oss he sequence, and miR-le -7 ,
which was down egula ed (Table 2).
Success ul alida ion o he miRNA bioma ke s in he
alida ion coho
The 6 miRNAs ha ulfilled he abo e condi ions we e chosen o
u he alida ion. Mo eo e , we decided o include also miR-195
and miR-29a in o he analysis because hey also co ela ed well
wi h ca cinogenesis and we e accu a e disc imina o s o dys-
plasia, al hough ailed o show s a is ically significan diffe ences
be ween spo adic and UC-associa ed cance . As shown in Table 2
Figu e 2. Beha io o selec ed di e en ially de egula ed miRNAs ac oss he UC-associa ed CRC sequence in he disco e y phase. CRC, colo ec al cance ;
miRNA, mic oRNA; UC, ulce a i e coli is.
Clinical and T ansla ional Gas oen e ology VOLUME 13 | JULY 2022 www.clin anslgas o.com
INFLAMMATORY BOWEL DISEASE
Quin anilla e al.6
(and Supplemen a y Table S4, Supplemen a y Digi al Con en 4,
h p://links.lww.com/CTG/A806), we posi i ely alida ed he
exp ession o 3 o 8 miRNAs: miR-31, -106a, and -135b, all o
which we e up egula ed. Box plo s and ROC cu es o hese 3
miRNAs a e shown in Figu e 4.
Con a ily, miR-195 showed a s a is ically significan down-
egula ion in dysplasia (P50.0002), bu in he UC-CRC g oup, i
was exp essed a he same le el as in no mal mucosa (P50.65),
esul s no consis en wi h hose o he disco e y phase (see Sup-
plemen a y Figu es S5, Supplemen a y Digi al Con en 5,
h p://links.lww.com/CTG/A807). On he o he hand, miR-20b
showed a consis en and s a is ically significan de egula ion
ac oss he UC-associa ed and spo adic CRC sequences in bo h
he disco e y and alida ion coho , bu in he la e , i ailed o
show significan diffe ences be ween he UC-CRC and Sp-CRC
cases (P50.11) (Table 2). Finally, miR-24 did no show any
diffe ences ac oss he UC-associa ed CRC sequence, and miR-
29a and miR-le -7 did no show diffe ences be ween no mal
mucosa and dysplasia (see Supplemen a y Table S4, Supple-
men a y Digi al Con en 4, h p://links.lww.com/CTG/A806
and Supplemen a y Figu es S5, Supplemen a y Digi al Con en
5, h p://links.lww.com/CTG/A807).
DISCUSSION
CRC is s ill a se e e and li e- h ea ening complica ion o long-
s anding UC; hus, sc eening and ea ly ea men o p emalignan
lesions a e decisi e o dec ease mo bidi y and mo ali y. Cu en
su eillance s a egies aim o iden i y dysplasia by pe iodic
colonoscopies; howe e , he e a e impo an difficul ies o de ec
and diagnose dysplasia accu a ely, and colonoscopy is s ill an
in asi e p ocedu e, cos ly, and uncom o able o pa ien s.
The e o e, he e is a subs an ial need o find new bioma ke s ha
can p edic dysplasia o cance nonin asi ely.
miRNAs ep esen a c ucial pa in he complex ne wo k o
epigene ic egula ion and ha e become po en ial bioma ke s o
UC-CRC. Howe e , he e is s ill a need o imp o e he diagnos ic
pe o mance based on he de ec ion o miRNAs species and alida e
he findings. Indeed, he e ha e been p e ious a emp s o dis in-
guish IBD ela ed o spo adic colonic neoplasia wi h some success,
bu obse a ions a e no consis en . His ological diffe en ia ion o
coli is-associa ed CRC om spo adic CRC is no possible. Com-
monly, s udies ha e conside ed hose CRC a ising in an a ea wi h
known p e ious inflamma ion as coli is-associa ed CRC (24). Also,
coli is-associa ed CRC is o en labeled a biobanks as CRC wi hou
including any specifica ion o hei ela ionship wi h coli is, which
makes hei iden ifica ion and diffe en ia ion om spo adic CRC
difficul . We add essed his limi a ion in ou s udy by including as
UC-CRC only hose cases ha occu ed in pa ien s wi h known
longs anding UC and ha aised in a eas wi h p e ious coli is (i.e., a
p oximal CRC in a pa ien wi h dis al longs anding UC would no
ha e been included in he s udy). Mo eo e , o ensu e ha ou
esul s would no be biased owa d inflamma ion, we only included
dysplasias and CRCs om pa ien s wi h UC in clinical emission.
We canno exclude, howe e , ha he e was some deg ee o mi-
c oscopic inflamma ion in some o he samples, bu we belie e ha
he impac was minimal. On he o he hand, we ec ui ed as spo-
adic CRCs only hose cases diagnosed h ough a popula ion-based
CRC sc eening p og am in asymp oma ic indi iduals whe e pa-
ien s wi h IBD a e specifically excluded.
One majo s eng h o ou s udy is ha we mi o he pa h-
ophysiologic cascade o ca cinogenesis o bo h he UC-CRC and
Figu e 3. (a) Hea map o selec ed mic oRNAs di e en ia ing he exp ession pa e ns o UC-associa ed CRC (UC-CRC) in ligh blue and spo adic CRC
(Sp-CRC) in pink. (b) Hea map o selec ed mic oRNAs di e en ia ing he exp ession pa e ns o UC-associa ed dysplasia (UC-DYS) in ligh blue and no mal
colonic mucosa o pa ien s wi h UC (UC-NM) in pink.
Ame ican College o Gas oen e ology Clinical and T ansla ional Gas oen e ology
INFLAMMATORY BOWEL DISEASE
Di e en ially De egula ed Mic oRNAs 7
Sp-CRC sequence by analyzing samples o he in e media e s eps,
i.e., dysplasia o UC-CRC and spo adic adenoma o Sp-CRC.
By doing so, we ound six y-4 diffe en ially de egula ed miRNAs
o he UC-CRC sequence, which is compa able wi h p e ious
s udies (20–22,25). We obse ed ha mos miRNAs we e
equally de egula ed ac oss he UC-CRC and Sp-CRC sequence,
whe eas some miRNA species we e exclusi ely de egula ed
ac oss he UC-CRC sequence, o ins ance, down egula ed miR-
192 and up egula ed miR-126, indica ing ha al hough bo h
p ocesses sha e a wide ange o common ea u es, hey ha e
indeed dis inc miRNA exp ession pa e ns. We also ound
miRNAs ha we e associa ed wi h bo h UC-CRC and Sp-CRC
de elopmen bu on s a is ically significan diffe en le els (e.g.,
miR-31 being highe exp essed in he UC-CRC sequence),
which migh eflec a diffe en ele ance o he ca cinogenesis
such as as e p og ession o highe g ade o in asion o because
o he unde lying inflamma ion in UC-CRC. Hence, ano he
s eng h o ou s udy was ha we conside ed hese significan ly
diffe en le els as one o he ele an c i e ia o selec miRNAs
o u he alida ion.
The esul s o ou s udy also sugges ha miRNAs a e linked
o diffe en s ages o ca cinogenesis. Fo example, miR-126 and
-490 we e only up egula ed in cance and no in dysplasia,
sugges ing hey a e la e e en s, whe eas o he s (e.g., miR-20b,
-31, and -106a) seemed o be ea ly e en s because hey we e
al eady up egula ed in he dysplasia ye on a lowe le el han in
he cance . We hypo hesize ha hose miRNAs al e ed ea ly
du ing ca cinogenesis and ha main ain o u he enhance his
de egula ionincance a eo mos in e es as bioma ke s be-
cause hey could be an ea ly p edic o o a highe isk o de elop
cance . In his s udy, mo e han 90% o UC-associa ed dysplasia
we e classified as low g ade, whe eas 95% o adenomas we e
ubula adenomas wi h low-g ade dysplasia. Hence, he numbe
o samples o high-g ade dysplasia was oo low o pe o m a
subg oup analysis.
Ou main goal was o find miRNAs o be as specificaspos-
sible o coli is-associa ed dysplasia ha ha e a high po en ial
o mo ing on o cance . The e o e, we applied e y es ic i e
c i e ia o iden i y candida e mucosal miRNA bioma ke s ha
would no only accu a ely disc imina e dysplasia om no mal
mucosa bu also indica e a highe isk o CRC and a e diffe -
en ially exp essed in UC-CRC compa ed wi h Sp-CRC. Despi e
his na ow defini ion, we achie ed o iden i y 6 candida es, 3 o
which we e posi i ely alida ed in an independen coho . All 3
alida ed miRNAs (i.e., miR-31, -106a, and -135b) ha e been
p e iously desc ibed o a ious cance ypes including CRC
and ha e been associa ed wi h diffe en cance - ela ed pa hways
such as RAS signaling o miR-31 (26) and Wn /b-ca enin,
PI3K/AKT, TGFBR2, and PTEN o miR-135 (27–30). Mos
in iguingly, miR-31 has been linked o IBD-associa ed CRC
(21,31). On he o he hand, miR-106a, whose ole in cance cell
p oli e a ion, mig a ion, and in asion has been widely desc ibed
o CRC (32), has also been p o en o dis inguish be ween CD
and UC as well as o classi y inde e mina e IBD, bu ou findings
also sugges a no el ole o his miRNA in IBD-associa ed ca -
cinogenesis (33). MiR-135, which has been p oposed as a po-
en ial nonin asi e bioma ke in s ool o spo adic CRC and
ad anced adenoma, has no shown any diagnos ic alue in pa-
ien s wi h IBD so a , and he e o e, i s associa ion wi h UC-
CRC is a uly no el findingo ou s udy(34).
Despi e ou p omising esul s, we also acknowledge some
limi a ions o ou s udy. Fi s , he numbe o samples is s ill
limi ed; ne e heless, we we e able o iden i y nume ous highly
ele an bioma ke candida es and alida ed some o hem. Sec-
ond, he design o he s udy was e ospec i e, and he samples in
he UC-CRC sequence we e mos ly de i ed om diffe en pa-
ien s, which does no allow in e ing causali y. Thi d, we did no
co ela e he esul s o clinical da a, such as g ade o in-
flamma ion, umo s age, o disease- ee su i al, because hese
da a we e ei he no a ailable o he e ospec i e design did no
allow i s usage.
In ou opinion, he ideal u u e pe spec i e o miRNAs as
bioma ke s is hei u ili y as a nonminimally o minimally
in asi e sc eening me hod in blood, s ool, o ec al biopsy o
iden i y pa ien s wi h a high chance o ha ing p emalignan o
cance ous lesions. I es ed posi i e, in a nex s ep, pa ien s
would unde go colonoscopy-d i en biopsies o esec ions o
u he e alua e and s a i y acco ding o indi idual isk
p ofiles and be e a ionalize su eillance and ea men
s a egies.
Table 2. Resul s o disco e y phase and alida ion phase o 8 alida ed mic oRNAs (bold 5posi i ely alida ed)
miRNA
Disco e y Valida ion
UC-NM s
UC-Dys,
S uden P
UC-NM s
UC-Dys,
AUROC
UC-CRC
sequence,
ANOVA P
UC-CRC s
Sp-CRC,
S uden P
UC-NM s
UC-Dys,
S uden P
UC-NM s
UC-Dys,
AUROC
UC-CRC
sequence,
ANOVA P
UC-CRC s
Sp-CRC,
S uden P
le -7 0.003 0.720 0.012 5.20E-04 0.707 0.554 0.006 0.384
miR-20b 2.92E-04 0.820 1.58E-06 0.005 0.016 0.734 3.10E-06 0.110
miR-24 1.60E-06 0.900 2.22E-08 0.045 0.199 0.620 0.140 7.26E-04
miR-29a 3.90E-07 0.920 2.21E-09 0.062 0.645 0.517 0.003 0.021
miR-31 0.002 0.760 4.87E-06 0.020 2.64E-04 0.774 1.90E-06 0.015
miR-106a 6.16E-06 0.900 1.69E-06 0.001 0.002 0.777 0.003 0.012
miR-135b 4.57E-09 0.960 1.65E-13 0.010 5.12E-08 0.923 4.40E-05 0.040
miR-195 8.42E-04 0.800 4.33E-07 0.611 1.45E-04 0.831 1.40E-04 2.44E-04
AUROC, a ea unde he ecei e ope a ing cu e; Sp-CRC, spo adic colo ec al cance ; UC-CRC, ulce a i e coli is–associa ed colo ec al cance ; UC-Dys, ulce a i e
coli is–associa ed dysplasia; UC-NM, ulce a i e coli is–associa ed no mal mucosa.
Clinical and T ansla ional Gas oen e ology VOLUME 13 | JULY 2022 www.clin anslgas o.com
INFLAMMATORY BOWEL DISEASE
Quin anilla e al.8
Figu e 4. Le : Box and whiske plo s o ela i e exp ession o alida ed mic oRNAs compa ing he UC-associa ed (le ) and he spo adic sequence ( igh ):
miRNA-31 (a, b), miR-106a (c, d), and miR-135b (e, ). Righ : Sensi i i ies, speci ici ies, and ecei e ope a ing cu e o disc imina ing be ween no mal
mucosa and dysplasia o he ulce a i e coli is cases. Sp-Ad, spo adic adenoma; Sp-CRC, spo adic CRC; Sp-NM, spo adic no mal mucosa; UC-CRC, UC-
associa ed colo ec al cance ; UC-Dys, UC-associa ed dysplasia; UC-NM, ulce a i e coli is–associa ed no mal mucosa.
Ame ican College o Gas oen e ology Clinical and T ansla ional Gas oen e ology
INFLAMMATORY BOWEL DISEASE
Di e en ially De egula ed Mic oRNAs 9
