Digital image analysis of the tissue surface areas of site-designated and bilaterally pooled prostate biopsies
Full text
Summa y. Ini ial epo s abou he leng h o bila e ally
pooled biopsies showed ala ming issue loss compa ed
o indi idual biopsies, bu he cu en unde s anding o
“noodle biopsies” and be e embedding echniques may
ha e imp o ed hei quali y. He e, we implemen ed
digi al image analysis o s udy he di e ences in issue
su ace a eas be ween indi idual and pooled co es.
P os a e biopsy epo s om 1242 consecu i e pa ien s
we e e iewed. U ologis -dependen bias on he biopsy
quali y was elimina ed by iden i ying ou u ologis s
who submi ed equally indi idual and bila e ally pooled
biopsies. Digi al image analysis was applied o he issue
su ace a eas o 936 i ual slides con aining 1440
biopsy co es (12 co es pe pa ien x 120 pa ien s) aken
by he ou u ologis s. The median ( ange) su ace a eas
we e 73.8 mm2(40.1-102.5) o he si e-designa ed
(n=57) and 77.1 mm2(49.5-119.2) o he bila e ally
pooled biopsies (n=63) (p=0.19). Fo h ee u ologis s,
he median su ace a eas we e 69.5 mm2(60.4-93.2),
75.5 mm2(48.2-98.7) and 78.2 mm2(47.1-92.7) o he
si e-designa ed and 79.2 mm2(49.5-116.4), 69.3 mm2
(49.6-119.2) and 79.2 mm2(55.1-96.7) o he pooled
biopsies, espec i ely (p=0.58-0.75). Fo one u ologis ,
he median su ace a ea was ma ginally highe o he
pooled biopsies, 68.1 mm2(40.1-102.5) s. 81.6 mm2
(62.7-108.8) (p=0.03). In conclusion, he his ological
yields o indi idual and pooled p os a e biopsies we e
p ac ically equal. The esul s should no be conside ed as
a ecommenda ion o inc easingly submi unspeci ied
bila e al co es bu o encou age pa hology labo a o ies o
embed and cu all ecei ed p os a e biopsies wi h special
a en ion, ega dless o submission ype.
Key wo ds: P os a e needle biopsy, Biopsy quali y,
Digi al image analysis, Guidelines
In oduc ion
Diagnosis o p os a e cance is based on he
his ological e alua ion o p os a e biopsies, and he
accu acy o he diagnosis depends on he quali y o
samples on he mic oscope slide. The commonly
accep ed biopsy scheme consis s o 10-12 sys ema ic
needle co es wi h addi ional co es om suspec a eas
(P es i, 2007; Ukimu a e al., 2013), bu submission and
embedding p o ocols a e a iables (Va ma e al., 2013).
Indi idually submi ed co es a e ecommended by he
in e na ional guidelines (Van de Kwas e al., 2003,
2013; P es i, 2007; Ukimu a e al., 2013), bu some
u ologis s p e e submi ing less labo -in ensi e bila e al
biopsies, depending on he clinical si ua ion. Acco ding
o a ecen su ey, app oxima ely hal o Eu opean
pa hology labo a o ies s ill ecei e bila e al biopsies
pooled in wo specimen con aine s, whe eas only 40 %
ecei ed all p os a e biopsy co es in sepa a e ials
(Va ma e al., 2013).
Because o high p e alence o p os a e cance and
he la ge numbe o co es pe biopsy se , pa hology
labo a o ies ha e become inc easingly awa e o he cos s
o p os a e biopsy handling. Pooling p os a e biopsy
Digi al image analysis o he issue su ace a eas o
si e-designa ed and bila e ally pooled p os a e biopsies
Lau a Koi usalo1,2, An i Kaipia1, Paula Kujala2, Jo ma Isola3and Teemu T. Tolonen2,3
1Depa men o Su ge y, Sa akun a Hospi al Dis ic , Po i, 2Depa men o Pa hology, Fimlab Labo a o ies,
Tampe e Uni e si y Hospi al and 3Depa men o Cance Biology, BioMediTech, Uni e si y o Tampe e, Tampe e, Finland
His ol His opa hol (2018) 33: 399-405
h p://www.hh.um.es
O p in eques s o: Teemu T. Tolonen, Depa men o Pa hology,
Fimlab Labo a o ies, Bioka u 4, 33520 Tampe e, Finland. e-mail:
[email p o ec ed]
DOI: 10.14670/HH-11-940
His ology and
His opa hology
F om Cell Biology o Tissue Enginee ing
co es may educe he cos s o biopsy p ocessing, bu i is
no encou aged by he in e na ional guidelines. One
majo conce n is ha he quali y ob ained om pooled
p os a e biopsy is no adequa e o p ope diagnosis,
because single co es a e easie o align in he pa a in
block o maximal su ace a ea ep esen a ion han
mul iple co es (Kao e al., 2002; Boccon-Gibod e al.,
2004; Gup a e al., 2004). Cu en guidelines by he
ERSPC (Eu opean Randomized S udy o Sc eening o
P os a e Cance ) commi ee allow he embedding o up
o h ee co es in he same pa a in block, p o ided ha
p ope p e-embedding me hods a e used o a oid cu ling
and loa ing o he biopsies (Roga sch e al., 2000;
Be accini e al., 2007; Van de Kwas e al., 2013).
F om a na ow quali y poin o iew, he maximum
numbe o biopsies in one block may be a ma e o
echnique as in ou p e ious s udy, he numbe o co es
(one o nine) embedded o a single pa a in block did no
a ec he leng h o he biopsies (Tolonen e al., 2015).
Digi al image analysis is inc easingly used in
pa hology due o ad ancing slide scanne and s o age
echnology and i is p esumably he mos accu a e
me hod o de e mining biopsy su ace a ea. Because in
ou p e ious wo k, he leng hs o bila e ally pooled
biopsies we e unexpec edly high (Tolonen e al., 2015),
we wan ed o u he e alua e whe he embedding six
biopsies o a single pa a in block would ha e an impac
on he ac ual his ological yields, and applied digi al
image analysis o he co e slice su ace a eas o he
whole slide images o he p os a e biopsies. A Tampe e
Uni e si y Hospi al (TAUH) dis ic , i is s ill mo e
common o sample bila e ally pooled biopsies han
indi idual co es, which may po en ially in oduce a bias
o he esul s. The e o e, we also elimina ed he
ope a o -dependen bias on biopsy quali y (Van de
Kwas e al., 2013) by iden i ying u ologis s who ook
bo h specimen ypes equally, and by analyzing hei
specimens indi idually acco ding o he u ologis .
Ma e ials and me hods
The s udy was conduc ed unde app o al o he
E hical Commi ee o Tampe e Uni e si y Hospi al,
e e ence numbe R03203.
We e ospec i ely sc eened 1242 consecu i e
p os a e needle biopsy cases e alua ed in he Depa men
o Pa hology a Fimlab Labo a o ies be ween Ma ch
2013 and Sep embe 2014 om he pa hology da abase
o Fimlab Labo a o ies. The biopsies had been aken
om pa ien s wi h ele a ed PSA- alues and/o suspec
indings in digi al ec al examina ion. All biopsies had
been aken unde ans ec al ul asound guidance using
an 18-gauge needle biopsy gun wi h an 18-mm sample
no ch (Ba d Pe iphe al Vascula , Temple, AZ, U.S.A.,
e no. MC 1825) and we e placed in o 10% bu e ed
o malin ials di ec ly om he biopsy needle. All
p os a e biopsies we e p ocessed a Fimlab Labo a o ies.
The con en s o each specimen ial we e s aigh ened in
he issue casse e be ween wo sponges wi h o ceps and
p ocessed be ween wo sponges p e- e ed wi h o malin.
The embedding p ocess was “as usual” and essen ially
he same o bo h indi idually submi ed and pooled
co es: he me allic issue mold was pa ly illed wi h ho
pa a in and placed on he hea ed pla e o he embedding
s a ion, and hen he s aigh ened biopsy co es we e
ans e ed o he bo om o he mold wi h o ceps and
cooled down on he cold pla e. No special equipmen o
embedding was used. Tissue ink was no u ilized o
iden i y biopsy loca ion. The con en s o one casse e
we e embedded in o one block, esul ing in 12 blocks o
indi idually submi ed co es and wo blocks o pooled
co es. The blocks we e sec ioned and s ained ou inely
wi h hema oxylin-eosin s ain (H&E).
The biopsy cases in he pa hology da abase we e
ca ego ized based on he submission and embedding
me hod (12 indi idual co es, 6+6 pooled co es, o o he ).
The cases we e ca ego ized based on bo h u ologis s’
e e als and pa hologis s’ epo s in he TAUH
pa hology da abase. Fou u ologis s who had submi ed
app oxima ely equal amoun s o bo h indi idual co es
and bila e ally pooled biopsy co es we e iden i ied, and
hei samples we e e ie ed om he sample a chi es.
The ype o biopsy was e i ied om he o iginal
mic oscope slides, and only cases con aining exac ly 12
co es we e included in he s udy. The inclusion scheme
and ypical se s o mic oscope slides o indi idually
submi ed and unspeci ied bila e al biopsies a e
illus a ed in (Fig. 1).
The mic oscope slides o indi idual co es con ained
usually wo o h ee, bu occasionally up o six slices pe
co e, always esul ing in one mic oscope slide. Pooled
co es we e consis en ly cu on ou le els, esul ing in
wo mic oscope slides. The mic oscope slides we e
scanned using an Ape io AT2 whole slide scanne (Leica
Biosys ems, Nussloch, Ge many) using 20x
magni ica ion wi h a esolu ion o 0.5 μm/pixel. The
scanned i ual slides we e con e ed in o JPEG2000-
iles using JP2 WSI Con e e so wa e, e sion 1.0.2
(BioMediTech, Uni e si y o Tampe e, Tampe e,
Finland). The JPEG2000- iles we e hen opened using
JVS iew i ual mic oscope so wa e, e sion 1.2
(Uni e si y o Tampe e, Tampe e, Finland), which
allowed expo a ion o he images o he image analysis
so wa e ImageJ, e sion 1.48 (Na ional Ins i u e o
Heal h, USA). Wi h ImageJ, he snapsho images om
i ual slides we e analyzed o de e mine he o al a ea
a ailable o his opa hological e alua ion.
Image analysis
To ans o m he images in o a o m ha could be
analyzed, we c ea ed a mac o sc ip o ImageJ. The
unc ion o he mac o is illus a ed s ep by s ep in Fig. 2,
and he ull mac o is p o ided in Table 1. Fi s , o
sepa a e he issue sample om he backg ound, a
h eshold was applied o he image. We used
expe imen ally de e mined h eshold alues o emo e
mos a i ac s and o ecognize he maximum amoun o
400
Digi al image analysis o he issue su ace a eas o p os a e biopsies
issue. Second, he pixels emaining ou side o he
sample a ea a e h esholding we e emo ed as ou lie s.
We de ined he maximum size o ou lie s o be 2 pixels.
Thi d, we used he “Fill holes” unc ion o ill he emp y
spaces in he sample a ea because he gland s uc u es o
p os a e issue a e pa o he co e su ace a ea.
The ac ual su ace a ea measu emen s we e
pe o med sepa a ely o each pa a in block slice image
u ilizing he “Analyze pa icles” unc ion o ImageJ. All
o he slice images ob ained om each block we e
analyzed. Because he numbe o slices om each block
a ied, we used he a e age o he wo bes (i.e., la ges
su ace a ea) slices om each o ob ain he leas biased
es ima e o compa ing he biopsy quali y o he wo
sample ypes. Fo indi idual co es, he a e age a ea
measu emen o each co e was summed o gi e he o al
su ace a ea o he biopsy. Fo he pooled co es, he
a e age combined a ea o he six co es om he le and
he igh we e summed.
S a is ical analysis
The s a is ical signi icance be ween he su ace a ea
measu emen s o indi idually submi ed and bila e ally
pooled co es was de e mined using he Wilcoxon-Mann-
Whi ney U- es a 95 % con idence le el (p=0.05). To
accoun o ope a o -dependen a ia ion, he
signi icance es ing was pe o med be ween each
u ologis ’s 12 indi idual and pooled samples. The
s a is ical analysis was pe o med wi h IBM SPSS
S a is ics, e sion 22.
Resul s
Based on he w i en epo s o he pa hology
da abase, 202 (16.3 %) o he o al 1242 cases we e
iden i ied as indi idually submi ed biopsies, and 613
(49.4 %) we e ca ego ized as pooled bila e al biopsies. A
o al o 427 (34.4 %) cases we e ca ego ized as “o he ”,
401
Digi al image analysis o he issue su ace a eas o p os a e biopsies
Fig. 1. A. Inclusion scheme o he cases o he s udy. B. An example o a ypical mic oscope slide a ays bila e ally pooled and si e-designa ed
p os a e biopsies.
including cases whe e he numbe and submission ype
o biopsy co es was una ailable (37 cases).
O e 20 di e en u ologis s om Tampe e
Uni e si y Hospi al dis ic con ibu ed o he 1242
samples submi ed o e alua ion a Fimlab Labo a o ies
du ing he pe iod o in es iga ion. The ou u ologis s
whose samples we e included in he s udy a e deno ed
by he numbe s 1-4. The ou u ologis s con ibu ed 16%
o all o he samples du ing his ime. U ologis s 1-4
we e all expe ienced in aking p os a e needle biopsies
wi h 10, 22, 16 and 6 yea s o expe ience, espec i ely.
The o e all median ( ange) su ace a eas we e
simila o bo h submission ypes, 73.8 mm2(40.1-
102.5) o he si e-designa ed (n=57) and 77.1 mm2
(49.5-119.2) o he bila e ally pooled biopsies (n=63)
(p=0.19). In addi ion, s a is ical signi icance was
calcula ed sepa a ely o each u ologis ’s samples (Fig.
3). The only s a is ically signi ican di e ence was ound
in u ologis 4’s biopsies, wi h he pooled co e biopsies
ha ing ma ginally la ge issue su ace a eas han he
indi idual co es (p=0.03).
Discussion
Si e-designa ed indi idual p os a ic needle biopsies
a e a o ed by all cu en in e na ional guidelines, bu
depending on he pa hology labo a o y and egion, i is
no uncommon o ecei e pooled biopsies submi ed by
he u ologis s (Bied zycki e al., 2003; Va ma e al.,
2013; Tolonen e al., 2015). Al hough poo mul i-
embedding echnique can lead o se e ely educed
biopsy leng h (Y an is e al., 2002), cu en awa eness o
he “noodle biopsies” and paying mo e a en ion o he
embedding p ocess should ha e imp o ed he quali y o
pooled biopsies. Ou aim was o compa e he issue
su ace a eas o indi idual s. pooled co es in a
pa hology labo a o y wi h expe ience in handling bo h
submission ypes.
Wi h he eme ging digi al pa hology, he e is an
ongoing pa adigm shi om a ule o high- esolu ion
image analysis o digi ized slides. In he p esen s udy,
we applied digi al image analysis o he issue su ace
a eas o 936 i ual slides con aining 684 indi idually
embedded and 756 bila e ally pooled biopsy co es sliced
in 2-4 planes. In addi ion, we ied o elimina e he
ope a o -dependen bias (Van de Kwas e al., 2013) by
selec ing u ologis s pe o ming equally bo h submission
ypes, and compa ing su ace a eas o sepa a ely
embedded and pooled biopsies indi idually acco ding o
he u ologis .
The o e all median issue su ace a eas we e simila
o 12 indi idually submi ed (73.8 mm2) and bila e ally
pooled 6+6 co es (77.1 mm2) (p=0.19), and he e we e
no signi ican di e ences be ween each u ologis s
indi idual and pooled biopsy a eas, excep o one
u ologis , who had ma ginally highe su ace a eas on
pooled biopsies (p=0.03). Al hough pooled biopsies
pe o med unexpec edly well, possibly due o
compensa o y slices acco ding o ou labo a o y
ins uc ions (2-3 slices om indi idual biopsies, ou
om pooled biopsies), he esul s a e conco dan wi h
ou p e ious wo k in which we did no ind signi ican
di e ences in he biopsy co e leng hs be ween he wo
submission ypes (Tolonen e al., 2015). In ou
labo a o y, he biopsies a e s aigh ened and p ocessed
be ween sponges ha ha e been p e-mois u ed wi h
o malin. The con en s o each biopsy con aine a e
p ocessed and embedded oge he , indi idual biopsies
leading o 12 pa a in blocks, and pooled biopsies
leading o wo blocks pe case. I is especially impo an
o embed biopsies ca e ully o he bo om o he me al
mold, and o no ole a e any o e lapping o he co es.
Slicing should a ge o he cen al axis o he biopsy (o
biopsies), which can a ec he slice su ace a eas
signi ican ly. In addi ion, some ma e ial should be
402
Digi al image analysis o he issue su ace a eas o p os a e biopsies
Table 1. ImageJ mac o o analyzing su ace a ea o a sample.
//Su ace a ea o p os a e samples
// un("Th eshold...");
min=newA ay(3);
max=newA ay(3);
il e =newA ay(3);
a=ge Ti le();
un("HSB S ack");
un("Con e S ack o Images");
selec Window("Hue");
ename("0");
selec Window("Sa u a ion");
ename("1");
selec Window("B igh ness");
ename("2");
min[0]=0;
max[0]=255;
il e [0]="pass";
min[1]=15;
max[1]=255;
il e [1]="pass";
min[2]=0;
max[2]=238;
il e [2]="pass";
o (i=0;i<3;i++){
selec Window(""+i);
se Th eshold(min[i], max[i]);
un("Con e o Mask");
i ( il e [i]=="s op") un("In e ");
}
imageCalcula o ("AND c ea e", "0","1");
imageCalcula o ("AND c ea e", "Resul o 0","2");
o (i=0;i<3;i++){
selec Window(""+i);
close();
}
selec Window("Resul o 0");
close();
selec Window("Resul o Resul o 0");
ename(a);
// Colou Th esholding-------------
makeLine(0, 0, 0, 0);
se Op ion("BlackBackg ound", alse);
un("Remo e Ou lie s...", " adius=2 h eshold=50 which=Da k");
un("Fill Holes");
spa ed o possible immunohis ochemical s ainings. All
his makes i qui e a di icul ask, and needs an
expe ienced echnician wi h good isuospa ial skills.
The e a e mul iple easons in a o o si e-
designa ed biopsies. The co es submi ed in sepa a e
ials a e less p one o agmen a ion and issue
en anglemen han pooled co es, which enables epo ing
he numbe o posi i e co es o be mo e accu a e
(Faja do and Eps ein, 2010). In addi ion, indi idual
biopsies a e easie o embed in o a single plane o
maximal issue ep esen a ion (Gup a e al., 2004),
which can in luence he de ec ion a e o cance and
small a ypical lesions (Iczkowski e al., 2002; Öbek e
al., 2012). Mos impo an ly, indi idually submi ed
biopsies spa e biopsy loca ion in o ma ion, which is
pi o al o he planning o he obo ic su ge y and ac as
a guide o e-biopsy si es in ac i e su eillance (Kao e
al., 2002; Van de Kwas e al., 2003, 2013; Boccon-
Gibod e al., 2004). On he o he hand, some au ho i ies
ha e poin ed ou ha a widesp ead use o indi idual co e
submission would inc ease he wo kload o al eady
s e ched pa hology labo a o ies. Fo ins ance, Bos wick
and Kahane ha e ecommended submi ing h ee biopsy
co es pe casse e o economic easons (Bos wick and
Kahane, 2013). Al hough he exac locus in o ma ion
will be los using sugges ed h ee-co es- h ee-slices
sys em, he au ho s did no ind e idence o lowe ed
diagnos ic a e o a ypical small acina p oli e a ion,
high g ade p os a ic in aepi helial neoplasia o
adenoca cinoma (Bos wick and Kahane, 2013).
Acco dingly, he esul s o he p esen s udy sugges ha
highe his ological yield alone seems o be a weak
a gumen o a o ing indi idual co es. The e o e, i
seems easonable ha pooled 6+6 biopsies migh be
u ilized o educe he pa hology labo a o y wo kload in
selec ed cases in eg. pa ien s ou o adical ea men due
o high age o PSA, o wi h ad anced disease. Fimlab
Labo a o ies ecei es bo h indi idual and pooled
p os a e biopsies om app oxima ely 1000 pa ien s
annually. Recei ing all biopsies sepa a ely would yield
12 000 pa a in blocks compa ed o 2000 blocks using
bila e al 6+6 biopsies. Al hough he mul i-embedding o
he la e may be mo e di icul and ime consuming, i
can aid educing he wo kload la e in he p ocess. We
es ima e ha cu ing 10 000 ex a blocks would ake
abou 100 labo days (o 5 mon hs) om an expe ienced
echnician.
Al hough ou aim was no o compa e ope a ing
u ologis s, he e was some ope a o -dependen a ia ion
in he esul s, as p oposed by Van de Kwas e al.
(2013). The o al su ace a ea measu emen s span la ge
anges, bu he e is no o e all end in he dispe sions o
he o al su ace a eas be ween pooled and indi idually
submi ed biopsies. The size o he ange o u ologis
1’s pooled biopsies (66.90 mm2) is much la ge han o
indi idual biopsies (32.85 mm2). Con e sely, o
403
Digi al image analysis o he issue su ace a eas o p os a e biopsies
Fig. 2. The unc ions
pe o med by he
ImageJ mac o on he
image o a pooled
p os a e biopsy
sample.
Howe e , we used he same image esolu ion h oughou
he s udy, so he esul an e o is sys ema ic and should
no a ec he compa abili y o he esul s. Second, he
“Fill holes” unc ion wo ks o enclosed egions,
whe eas open glands a e no aken in o accoun .
Meanwhile, i a co e has s e ched, lea ing emp y space
inside he a ea o a sample, i is illed and in luences he
a ea measu emen . This may cause andom e o in he
measu emen s.
Conclusions
Digi al image analysis o he issue su ace a ea o
p os a e biopsies is he ul ima e way o measu e hei
his ologic yield. Al hough a wide in a-ope a o
a ia ion was seen in image analysis o he issue su ace
a eas o bo h indi idual and pooled p os a e biopsies, he
median issue su ace a eas we e equal ega dless o he
submission ype. Ou esul s sugges ha he his ologic
yields o ca e ully embedded bila e ally pooled p os a e
biopsies can app oach he le el o indi idually
embedded biopsies; a leas hey seem o ha e done so a
ou ins i u e. To con i m hese esul s, a mul icen e
404
Digi al image analysis o he issue su ace a eas o p os a e biopsies
Fig. 3. Boxplo s showing
he dis ibu ions o he
su ace a eas o he
pooled and indi idually
submi ed p os a e
samples acco ding o he
ope a ing u ologis .
*signi icance a p=0.05.
u ologis 4, he ange o alues is la ge o indi idual
biopsies (62.40 mm2) han o pooled biopsies (46.13
mm2). Fo u ologis s 2 and 3, he sizes o he alue
anges a e app oxima ely equal o bo h ypes o
samples. The la ges o al su ace a ea alue is ound in
he pooled biopsies g oup o all u ologis s, which may
indica e ha he o al su ace a ea de e mina ion
echnique is biased owa ds pooled biopsies. I seems
likely ha measu ing he ac ual biopsy a ea in mm2
ins ead o jus measu ing hei leng h highligh s his
a ia ion. As men ioned ea lie , si e-designa ed biopsies
we e gene ally sliced on wo o h ee le els and pooled
biopsies a ou le els, which may in oduce he
a o emen ioned bias. Howe e , i seems o be a jus i ied
p ac ice as he his ological yields we e simila while a
he same ime he o al numbe o slides emain low
(Fig. 1).
The ImageJ mac o may ha e induced some e o in
he su ace a ea measu emen s. Fi s , he alue o he
measu ed a ea depends on he esolu ion (i.e., pixel size)
because we modi y he image on a pixel le el. The
e ec s o he “Remo e ou lie s” and “Fill holes”
unc ions especially depend on he image esolu ion.
s udy would be necessa y. Finally, he esul s should no
be conside ed as a ecommenda ion o inc easingly
submi unspeci ied bila e al co es bu o encou age
pa hology labo a o ies o embed and slice all ecei ed
p os a e biopsies wi h special a en ion, ega dless o
submission ype.
Acknowledgemen s: We would like o hank Tampe e Uni e si y
Hospi al u ologis s M. Aho, J. Koskimäki, A. Ko sa and T. Mu ola (in
alphabe ical o de ) o p o iding he samples used in his s udy and
Sa akun a and Pi kanmaa Hospi al Dis ic s o unding he esea ch.
We would also like o hank Ph.D. Pekka Ruusu uo i o his aluable
commen s. Finally, we would like o hank he dedica ed labo a o y
echnicians o Fimlab Labo a o ies.
Au ho s’ con ibu ion: L Koi usalo: Da a collec ion, image analysis
de elopmen , da a analysis, manusc ip w i ing; A Kaipia: P ojec
de elopmen , manusc ip edi ing; P Kujala: P ojec de elopmen ,
manusc ip edi ing; J Isola: Image analysis de elopmen , manusc ip
edi ing; TT Tolonen: P ojec de elopmen , manusc ip w i ing
This a icle does no con ain any s udies wi h human pa icipan s o
animals pe o med by any o he au ho s.
Con lic s o in e es : The au ho s decla e ha hey ha e no con lic s o
in e es .
Re e ences
Be accini A., Fandella A., P aye -Gale i T., Sca oni, V., Galosi, A. B.,
Fica a, V., T ombe a C., Gion M. and Ma o ana G. (2007).
Sys ema ic de elopmen o clinical p ac ice guidelines o p os a e
biopsies: A 3-yea i alian p ojec . An icance Res. 27, 659-666.
Bied zycki O., Va ma M. and Be ney D.M. (2003). Va ia ions in he
p ocessing o p os a ic needle co es in he UK; wha is sa e?. J.
Clin. Pa hol. 56, 341-343.
Boccon-Gibod L., an de Kwas T.H., Mon i oni R., Boccon-Gibod L.
and Bono A. (2004). Handling and pa hology epo ing o p os a e
biopsies. Eu . U ol. 46, 177-181.
Bos wick D.G. and Kahane H. (2013). Adequa e his ologic sec ioning o
p os a e needle biopsies. Ann. Diagn. Pa hol. 17, 357-360.
Faja do D.A. and Eps ein J.I. (2010). F agmen a ion o p os a ic needle
biopsy co es con aining adenoca cinoma: he ole o specimen
submission. BJU In . 105, 172-175.
Gup a C., Ren J.Z. and Wojno K.J. (2004). Indi idual submission and
embedding o p os a e biopsies dec eases a es o equi ocal
pa hology epo s. U ology 63, 83-86.
Iczkowski K.A., Casella G., Seppala R.J., Jones G.L., Mishle B.A., Qian
J. and Bos wick D.G. (2002). Needle co e leng h in sex an biopsy
in luences p os a e cance de ec ion a e. U ology 59, 698-703.
Kao J., Up on M., Zhang P. and Rosen S. (2002). Indi idual p os a e
biopsy co e embedding acili a es maximal issue ep esen a ion. J.
U ol. 168, 496-499.
Öbek C.I., Doğanca T., E dal S., E doğan S. and Du ak H. (2012). Co e
leng h in p os a e biopsy: size ma e s. J. U ol. 187, 2051-2055.
P es i J.C. (2007). P os a e biopsy: Cu en s a us and limi a ions. Re .
U ol. 9, 93-98.
Roga sch H., Mai inge T., Ho ninge W., Gschwend ne A., Ba sch G.
and Mikuz G. (2000). Op imized p eembedding me hod imp o es
he his ologic yield o p os a ic co e needle biopsies. P os a e 42,
124-129.
Tolonen T.T., Isola J., Kaipia A., Riikonen J., Koi usalo L., Huo inen S.,
Lau ila M., Po e S., Ti kkonen M. and Kujala P. (2015). Leng h o
p os a e biopsies is no necessa ily comp omised by pooling mul iple
co es in one pa a in block: an obse a ional s udy. BMC Clin.
Pa hol. 15, 4.
Ukimu a O., Coleman J.A., de la Taille A., Embe on M., Eps ein J.I.,
F eedland S.J., Gianna ini G., Kibel A.S., Mon i oni R., Ploussa d
G., Roobol M.J., Sca oni V. and Jones J.S. (2013). Con empo a y
ole o sys ema ic p os a e biopsies: Indica ions, echniques, and
implica ions o pa ien ca e. Eu . U ol. 63, 214-230.
Van de Kwas T.H., Lopes C., San onja C., Pihl C.G., Nee ens I.,
Ma ikainen P., Di Lollo S., Bubendo L. and Hoedemaeke R.F.
(2003). Guidelines o p ocessing and epo ing o p os a ic needle
biopsies. J. Clin. Pa hol. 56, 336-340.
Van de Kwas T., Bubendo L., Maze olles C., Raspollini M.R., Van
Leende s G.J., Pihl C.G. and Kujala P. (2013). Guidelines on
p ocessing and epo ing o p os a e biopsies: he 2013 upda e o
he pa hology commi ee o he Eu opean Randomized S udy o
Sc eening o P os a e Cance (ERSPC). Vi chows A ch. 463, 367-
377.
Va ma M., Be ney D.M., Algaba F., Campa o P., Compé a E., G i i hs
D.F., K is iansen G., Lopez-Bel an A., Mon i oni R. and Ege ad L.
(2013). P os a e needle biopsy p ocessing: a su ey o labo a o y
p ac ice ac oss Eu ope. J. Clin. Pa hol. 66, 120-123.
Y an is H.G., Lo e O.B. and Sil e be g S.G. (2002). P os a e co e
biopsies p ocessing: E alua ing cu en p ac ice. USCAP annual
mee ing, Chicago, p 347 A., Abs ac 1447.
Accep ed Oc obe 23, 2017
405
Digi al image analysis o he issue su ace a eas o p os a e biopsies