Vol.:(0123456789)
Vi chows A chi
h ps://doi.o g/10.1007/s00428-025-04026-4
ORIGINAL ARTICLE
Immunohis ochemical analysis o 147 cases o low-g ade endome ial
s omal sa coma: e ining heimmunohis ochemical p o ile o LG-ESS
onala ge, molecula ly con i med se ies
Mi osla aFlíd o á1· Pa elDund 1· RomanaV ánko á1· K is ýnaNěmejco á1· Da idCibula2· Rena aPonco á2·
K ě osla aMichalo á3,4· JiříBouda5,6· JanLaco7· MunachisoNdukwe8· JanuszRyś9· Ma iuszKsiążek10,19·
Albe oBe jon11· IgnacioZapa diel12· I anF anin13· An onelaNja o14· Ji kaHausne o á15· Pe aB e o á16·
Vladimí Židlík17· Ja osla Klá 18· Zoa dTibo K asznai19· Robe Poka19· Na aliyaVolodko20· I ynaYezho a20·
Rado anPilka21,22· RadimMa ek21· Geo ginaKolniko a23· MilanK koška24· MichaelHalaška25· JanaD ozeno á26·
Dagma Dolinská27· Vladimí Kalis 28· Ma cinBobiński29· Ma aOs owska‑Leśko30· MagdalenaBizoń31,32·
Włodzimie zSawicki32· MaciejS ukan33,34· Ka olinaG abowska33· Ma cinJęd yka35· Tymo euszPop awski36·
SimonaS olnicu37· MihaiEmilCăpîlna38· ZuzanaŠpů ko á39· MichalZikán40· F ancescaCicca one41·
Gio anniScambia41,42· A chilSha ashenidze43· Mi andaGudadze44· Te ianaPia ny ska45· Iho Va chak45·
MichaelaKendallBá ů1
Recei ed: 2 Decembe 2024 / Re ised: 23 Decembe 2024 / Accep ed: 6 Janua y 2025
© The Au ho (s) 2025
Abs ac
Low-g ade endome ial s omal sa coma (LG-ESS) can p esen diagnos ic challenges, due o i s o e lapping mo phological
ea u es wi h o he u e ine mesenchymal umo s. Misdiagnosis a es emain signi ican , and immunohis ochemical da a o
LG-ESS a e limi ed o small se ies and inconsis en an ibody panels. This s udy aimed o e ine he IHC p o ile o LG-ESS
by analyzing a la ge, molecula ly con i med se ies o 147 cases using a panel o 24 an ibodies, including newe ma ke s
like ansgelin and smoo helin. CD10 and IFITM1, key endome ial s omal ma ke s, we e exp essed in 86% (92% o hose
ex ensi ely) and 69% (60% o hose ex ensi ely) o cases, wi h usion-posi i e umo s showing signi ican ly highe exp es-
sion. Smoo h muscle ma ke s (α-SMA, desmin, h-caldesmon, calponin, ansgelin) we e a iably exp essed, p edominan ly
in ocal o low-in ensi y pa e ns, wi h α-SMA eaching he highes equency o exp ession (44%). Howe e , he in ensi y
o smoo h muscle ma ke exp ession was usually e y low. Smoo helin was a ely exp essed. Ho mone ecep o s we e e-
quen ly posi i e, wi h PR showing a highe equency (92% s. 83%) and in ensi y han ER. Ma ke s like S-100, HMB45,
and CD117 we e la gely nega i e; all umo s we e p53 wild- ype, wi h p ese ed SMARCB1/SMARCA4 exp ession and
ALK and ROS1 nega i i y. This wo k ep esen s he la ges molecula ly alida ed IHC s udy on LG-ESS, p o iding a obus
diagnos ic p o ile o ou ine pa hology. By add essing key diagnos ic limi a ions and examining newe ma ke s, ou s udy
suppo s a mo e s anda dized app oach o diagnosing LG-ESS and unde sco es he alue o immunohis ochemical panels,
pa icula ly in usion-nega i e umo s whe e diagnosis elies on mo phological and immunohis ochemical in e p e a ion.
These indings con ibu e c i ical da a o imp o ing diagnos ic accu acy.
Keywo ds Low-g ade endome ial s omal sa coma· LG-ESS· Immunohis ochemis y· Endome ial s omal ma ke s·
Smoo helin
In oduc ion
Low g ade endome ial s omal sa coma (LG-ESS) is a
a e malignan mesenchymal umo which mo phologi-
cally esembles p oli e a i e phase endome ial s oma and
exhibi s an in il a i e g ow h pa e n [1]. Endome ial s o-
mal sa comas cons i u e up o 1% o all u e ine cance s and
Ex ended au ho in o ma ion a ailable on he las page o he a icle
Vi chows A chi
6–20% o all u e ine sa comas, wi h LG-ESS ep esen ing
he second mos common u e ine sa coma a e leiomyo-
sa coma [2, 3]. While mos cases a ec he u e ine body,
p ima y ex au e ine umo s can also occu , o en associa ed
wi h endome iosis, making diagnosis mo e challenging [4,
5].
In mos cases, LG-ESS can be eliably diagnosed based
solely on mo phology due o i s dis inc appea ance. How-
e e , in some cases, LG-ESS can exhibi signi ican mo -
phological he e ogenei y wi h a wide ange o his ological
pa e ns, which can also be e lec ed in hei immunohis o-
chemical p o ile [6]. Some less common, a ian mo pholo-
gies include smoo h muscle-like di e en ia ion, ib oblas ic
and/o myxoid ea u es, sex co d-like s uc u es, pseudopap-
illa y o ma ions, clea cell change, skele al muscle di e -
en ia ion, adipocy ic me aplasia, angioma ous pa e n, and
cells wi h habdoid appea ance o biza e nuclei [7–11].
The e o e, he di e en ial diagnosis o LG-ESS p ima ily
includes cellula leiomyoma (and po en ially o he smoo h
muscle umo s), in lamma o y myo ib oblas ic umo (IMT),
and u e ine umo s esembling sex co d-s omal umo s
(UTROSCT) [6, 8, 9, 12–15]. Some cases wi h smoo h mus-
cle di e en ia ion exhibi a cha ac e is ic “s a bu s ” pa e n,
wi h collagen ibe s adia ing h oughou he umo , while
o he s may show dispe sed collagen plaques o myxoid
changes, which can mimic he eg essi e changes ound in
leiomyoma o a myxoid leiomyosa coma [7]. The p esence
o epi helioid/ ound cell di e en ia ion may aise suspicion
o high g ade endome ial s omal sa coma (HG-ESS) o
pe i ascula epi helioid cell umo (PEComa). In some
cases, soli a y ib ous umo (SFT), gas oin es inal s omal
umo (GIST), o adenosa coma can also s and in he di -
e en ial diagnosis. Mo eo e , ecen ly desc ibed sa comas
wi h he KAT6B/A::KANSL1 usion usually ha e o e lap-
ping ea u es be ween LG-ESS/endome ial s omal nodule
(ESN) and smoo h muscle umo s [16–19].
Wi h he inc easing accessibili y o molecula gene ic
me hods, especially nex gene a ion sequencing (NGS), and
he expanding knowledge o ecu en gene ic al e a ions in
u e ine mesenchymal neoplasms, iden i ying cha ac e is ic
usions has g ea ly enhanced diagnos ic accu acy. Up o 75%
o LG-ESS ha bo ecu en usions, wi h JAZF1::SUZ12
being he mos common, ollowed by JAZF1::PHF1,
EPC1::PHF1, and MEAF6::PHF1 [6, 8, 20]. Howe e , ha
lea es almos a qua e o LG-ESS which a e no cu en ly
associa ed wi h any known ecu en gene ic al e a ion. In a
ou ine diagnos ic se ing, access o NGS is s ill a he lim-
i ed and o en a ies be ween coun ies, wi h he added ques-
ion o inancial a ailabili y. As such, immunohis ochemis y
as a me hod which is as ly mo e accessible s ill emains a
i al ool in eaching he co ec diagnosis.
We conduc ed an ex ensi e immunohis ochemical analy-
sis on a la ge, molecula ly examined coho o 147 LG-ESS,
all subjec ed o igo ous cen al e iew. Ou s udy es ed a
wide panel o immunohis ochemical ma ke s including sex
co d-s omal ma ke s, smoo h muscle ma ke s, ho mone
ecep o s, and selec ed no el an ibodies, some o which
ha e no been assessed in LG-ESS coho s o his size. Ou
goal was o de ine he immunohis ochemical p o ile o he
la ges coho o LG-ESS o da e, hus p o iding a p ac i-
cal diagnos ic guide o pa hologis s, especially in se ings
whe e molecula es ing is una ailable.
Ma e ials andme hods
Samples
The samples ep esen a pa o he coho assembled unde
he Ra e Gynecological Sa coma (REGYS) s udy, which
is an in e na ional p ojec in ol ing 23 pa icipa ing ins i-
u ions om 10 coun ies, consis ing mos ly o membe s
o he Cen al and Eas e n Gynecology Oncology G oup
(CEEGOG). The p ojec included a de ailed assessmen o
he mo phological ea u es, immunohis ochemical analysis,
and DNA and RNA NGS analysis. A de ailed desc ip ion o
he p ojec , he o e all coho assembled, and he molecu-
la esul s a e p o ided in a di e en s udy (cu en ly unde
e iew) and a e no desc ibed he e. B ie ly, a cen al e iew
was pe o med on all hema oxylin and eosin (HE)-s ained
slides a ailable o indi idual cases by wo expe ienced
gynecopa hologis specialis s (PD, MKB). The equi ed
numbe o ep esen a i e slides om he co-ope a ing ins i-
u ions o each case was 2 (n = 100). In hose cases which
came om he a chi es o ou depa men (as well as cases
which we e sen o us as consul a ions, n = 47), he e alu-
a ion was pe o med on all slides a ailable om he en i e
biopsy/ esec ed specimen. The mo phological e alua ion
was combined wi h he esul s o he immunohis ochemical
analysis and molecula es ing o each he co ec diagnosis.
An immunohis ochemical (IHC) examina ion wi h a
b oad panel o 24 selec ed an ibodies was pe o med o
each umo . Only cases unequi ocally diagnosed as LG-ESS
a e he cen al e iew we e included in he s udy, esul ing
in a inal sample se o 147 cases. O hese 147 cases, 133
had RNA NGS esul s a ailable which showed ha 101 cases
(75.9%) ha bo ed a ecu en usion. The mos common
al e a ion was he JAZF1::SUZ12 usion ound in 67 cases
(66.3% o all usions), ollowed mainly by JAZF1::PHF1
(n = 9), MEAF6::PHF1 (n = 8), and EPC1::PHF1 (n = 4).
The s udy was pe o med on o malin- ixed, pa a in
embedded (FFPE) issue blocks. Hema oxylin and eosin-
s ained slides om he FFPE issue blocks we e e iewed,
and issue mic oa ays (TMAs) we e cons uc ed om sui -
able umo a eas. Two issue co es o a 2.00 mm diame e
we e ex ac ed om each FFPE issue block using he TMA
Vi chows A chi
ins umen TMA Mas e (3DHISTECH L d., Budapes ,
Hunga y).
Immunohis ochemical analysis
Immunohis ochemis y was pe o med on 4 µm hick sec ion
using he TMAs. In cases whe e he TMA app oach was no
possible, pa icula ly due o small umo size o echnical
di icul ies du ing sample p ocessing, whole- issue sec ions
we e used whe e possible. The IHC was e alua ed indepen-
den ly by wo pa hologis s (MKB, MF). The an ibodies used
o IHC examina ion we e selec ed based on hei diagnos-
ic u ili y, wi h an emphasis on uling ou he o he en i ies
mos commonly in ol ed in he di e en ial diagnosis. The
whole panel was comp ised o es ogen ecep o (ER), p o-
ges e one ecep o (PR), al a-smoo h muscle ac in (α-SMA),
desmin, h-caldesmon, calponin, CD10, IFITM1, ansgelin,
BCOR, BCORL1, NTRK, S-100, HMB45, CD117, WT1,
SMARCA4 (BRG1), SMARCB1 (INI1), SMARCA2, ALK,
ROS1, p53, smoo helin, and cyclin D1. The comple e lis o
an ibodies used, including hei clones, dilu ion, and manu-
ac u e s is p o ided in Table1.
The immunohis ochemical esul s we e e alua ed based
on he o e all pe cen age o posi i i y (0–100%). Cases we e
classi ied as nega i e (comple e absence o s aining o < 1%
o posi i e umo cells), 1 + (1–25% o posi i e umo cells),
2 + (26–50% o posi i e umo cells), o 3 + (> 50% o posi-
i e umo cells). Fo he an ibodies NTRK and smoo helin,
he s aining was e alua ed independen ly as bo h nuclea and
cy oplasmic, o WT1 only he nuclea s aining was e alu-
a ed. The immunohis ochemical esul s we e also assessed
using he H-sco e me hod p e iously desc ibed by o he s
[21]. This me hod inco po a es bo h he pe cen age o posi-
i e cells and he s aining in ensi y (1 + o weak in ensi y,
2 + o mode a e, and 3 + o s ong). The inal H-sco e is
calcula ed by adding he mul iplica ion o he di e en s ain-
ing in ensi ies acco ding o he ollowing o mula: [1 x (%
o cells 1 +)] + [2x (% o cells 2 +)] + [3x (% o cells 3 +)],
esul ing in an H-sco e alue o 0–300. To cha ac e ize
exp ession in e ms o posi i e and nega i e cases, a cu -
o alue o 1% was used (posi i e: ≥ 1% o cells showing
exp ession). Fo compa ing he IHC exp ession wi hin he
LG-ESS coho based on he p esence o usion, he cu -o
alue was modi ied o 5% o accoun o mino non-speci ic
s aining a ia ions.
S a is ical analyses
All s a is ical analyses we e conduc ed using R so wa e,
e sion 4.3.3 (2024–02–29). S anda d desc ip i e s a is ics
we e applied o summa ize he da ase : Ca ego ical a iables
we e epo ed as equencies and pe cen ages, and con inu-
ous a iables we e desc ibed using means wi h s anda d
de ia ion (SD) o medians wi h in e qua ile ange. Co ela-
ions be ween he exp ession o IHC ma ke s (ca ego ized as
posi i e s. nega i e) and usion s a us (p esence s. absence
o usion) we e assessed using Pea son’s chi-squa ed es o
Fishe ´s Exac es based on expec ed alues. All es s we e
wo-sided, and a p alue < 0.05 was conside ed s a is ically
signi ican .
Resul s
A de ailed o e iew o he IHC esul s is p o ided in
Table2. The IHC es ing o all an ibodies was no possible
in all o he cases due o limi ed ma e ial. Rep esen a i e
images o selec ed IHC ma ke s a e p o ided in Figs.1, 2, 3,
and 4. The coho consis ed o 122 cases (83%) o LG-ESS
wi h he usual mo phological pa e n, 5 cases (3%) wi h pu e
ib oblas ic pa e n, and a single case each (0.6%) displaying
p edominan smoo h muscle-like mo phology o a glandu-
la pa e n. Fou een umo s showed a mixed pa e n—mos
commonly usual + ib oblas ic (n = 7.5%) and usual + smoo h
muscle-like (n = 2.1%), while 6 cases (4%) showed a sex
co d-s omal componen o a iable ex en . Two cases (1%)
we e made up o a mix u e o ib oblas ic and myxoid pa -
e ns, and one case displayed a combina ion o usual, ib o-
blas ic, and myxoid mo phology.
As expec ed, he umo s showed high le els o exp es-
sion o he endome ial s omal ma ke s CD10 and
IFITM1. CD10 exp ession was seen in 86% o cases, wi h
a majo i y o he posi i e cases (92%) exhibi ing di use,
ex ensi e s aining o a high in ensi y (median H-sco e
196). The o he ma ke o endome ial s omal di e en-
ia ion IFITM1 was posi i e in a lowe numbe o cases
(69%) wi h a he e ogenous ex en o exp ession, which
was mos ly di use bu o a ying in ensi y. The exp ession
o he s anda d smoo h muscle ma ke s (α-SMA, desmin,
h-caldesmon, calponin, and ansgelin) was p esen o a
a iable deg ee in all o he examined umo s, wi h α-SMA
eaching he highes equency o posi i e cases (44%).
The ex en o exp ession o hese ma ke s was ypically
on he opposi e ends o he spec um—ei he only a e and
ocal (wi h posi i ely s aining a eas comp ising less han
25% o he umo issue) o ex ensi e and di use (wi h
mo e han 50% o he umo showing exp ession). How-
e e , he in ensi y o he smoo h muscle ma ke exp ession
was usually e y low, wi h he highes a e age H-sco e
o 46 (obse ed o α-SMA) and he median H-sco e 0
o all examined ma ke s. Smoo helin showed exp ession
in only wo cases o LG-ESS, which was only ocal and
o a weak in ensi y. Ho mone ecep o s we e exp essed
in a high p opo ion o he umo s, wi h PR eaching a
sligh ly highe equency (92% s. 83%) and also in ensi y
Vi chows A chi
Table 1 Lis o an ibodies used o immunohis ochemical analysis
An ibody Clone Dilu ion P oduce Pla o m De ec ion E alua ed exp ession
ALK D5F3 1:100 Cell Signaling Tech-
nology, Dan e s,
Massachuse s, USA
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Cy oplasmic
BCOR C-10 1:50 San a C uz Bio ech-
nology, Dallas,
Texas, USA
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Nuclea
BCORL 1 Polyclonal abbi 1:200 A las an ibodies,
B omma, Sweden
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Dako) + Linke
Nuclea
SMARCA4 (BRG1) EPNCIR 111A 1:200 Abcam, Camb idge,
UK
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Dako)
Nuclea
H-caldesmon h-CALD 1:800 San a C uz Bio ech-
nology, Dallas,
Texas, USA
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Dako)
Cy oplasmic
Calponin CALP 1:400 Dako, Glos up,
Denma k
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Dako)
Cy oplasmic
CD10 56C6 1:50 No ocas a, Leica
Biosys ems, We -
zla , Ge many
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Cy oplasmic
CD117 c-ki 1:200 Dako, Glos up,
Denma k
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Ul aView Cy oplasmic
Cyclin D1 EP 12 RTU Dako, Glos up,
Denma k
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Dako)
Nuclea , cy oplasmic
Desmin D33 1:200 Dako, Glos up,
Denma k
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Cy oplasmic
ER SP1 1:200 Zy omed Sys ems
GmbH, Be lin,
Ge many
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Nuclea
HMB45 HMB 45 1:50 Dako, Glos up,
Denma k
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Dako)
Cy oplasmic
IFITM1 Polyclonal abbi 1:400 Abcam, Camb idge,
UK
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Agilen )
Cy oplasmic
SMARCB1 (INI1) MRQ-27 RTU Cell Ma que, Rocklin,
CA, USA
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Nuclea
p53 DO-7 1:400 Dako, Glos up,
Denma k
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Nuclea , cy oplasmic
PR clone 16 1:100 No ocas a, Leica
Biosys ems, We -
zla , Ge many
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Nuclea
ROS1 D4D6 1:100 Cell Signaling Tech-
nology, Dan e s,
Massachuse s, USA
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Cy oplasmic
S-100 4C4.9 1:400 DCS, Hambu g,
Ge many
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Agilen )
Cy oplasmic
α- SMA 1A4 1:1600 Dako, Glos up,
Denma k
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Agilen )
Cy oplasmic
Vi chows A chi
RTU eady o use
Table 1 (con inued)
An ibody Clone Dilu ion P oduce Pla o m De ec ion E alua ed exp ession
SMARCA2 1:800 A las an ibodies,
B omma, Sweden
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA)
EnVision FLEX
(Dako) + Linke
Nuclea
Smoo helin R4A 1:50 Ze a Co po a ion,
Mon o ia, CA, USA
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Cy oplasmic, nuclea
NTRK EPR17341 1:100 Abcam, Camb idge,
UK
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Nuclea , cy oplasmic
T ansgelin 2A10C2 1:300 Cell Ma que, Rocklin,
CA, USA
Ven ana BenchMa k
ULTRA (Roche,
Basel, Swi ze land)
Op iView Cy oplasmic
WT1 6F-H2 1:400 BioSB, San a Ba -
ba a, CA, USA
Dako Omnis, (Agi-
len , San a Cla a,
CA, USA
EnVision FLEX
(Agilen )
Nuclea , cy oplasmic
Table 2 De ailed o e iew o IHC exp ession o he an ibodies mos commonly used in he di e en ial diagnosis o LG-ESS
SD s anda d de ia ion, cy opl. cy oplasmic. The numbe o analyzed cases o each s ain di e s due o he amoun o a ailable issue. Pe cen -
ages a e ounded up/down. Pe cen ages o 1 + /2 + /3 + posi i i y a e coun ed only om he posi i e cases
IHC ma ke Nega i e (0–1%)
n (%)
Posi i e (≥ 1%)
n (%)
Posi i i y
1 + (1–25%)
n (%)
Posi i i y
2 + (26–
50%)
n (%)
Posi i i y
3 + (> 50%)
n (%)
H-sco e mean
(SD)
H-sco e median
( ange)
ER (n = 143) 25 (17%) 118 (83%) 6 (5%) 8 (7%) 104 (88%) 118 (87.3) 100 (0–300)
PR (n = 144) 11 (8%) 133 (92%) 8 (6%) 5 (4%) 120 (90%) 201 (106.5) 223 (0–300)
α-SMA (n = 142) 80 (56%) 62 (44%) 20 (32%) 7 (11%) 35 (56%) 46 (75.4) 0 (0–300)
Desmin (n = 141) 106 (75%) 35 (25%) 15 (43%) 6 (17%) 14 (40%) 24 (66.8) 0 (0–300)
H-caldesmon
(n = 141)
121 (86%) 20 (14%) 12 (60%) 3 (15%) 5 (25%) 7 (28.0) 0 (0–200)
Calponin (n = 143) 94 (66%) 49 (34%) 22 (45%) 4 (8%) 23 (47%) 31 (72.7) 0 (0–300)
CD10 (n = 141) 20 (14%) 121 (86%) 6 (5%) 4 (3%) 111 (92%) 171 (110.9) 196 (0–300)
IFITM1 (n = 143) 45 (31%) 98 (69%) 28 (29%) 11 (11%) 59 (60%) 65 (77.2) 20 (0–300)
T ansgelin
(n = 141)
110 (78%) 31 (22%) 13 (42%) 6 (19%) 12 (39%) 16 (49.7) 0 (0–300)
BCOR (n = 142) 135 (95%) 7 (5%) 2 (29%) 1 (14%) 4 (57%) 3 (19.3) 0 (0–170)
BCORL1
(n = 141)
107 (76%) 34 (24%) 31 (91%) 2 (6%) 1 (3%) 2 (7.9) 0 (0–70)
NTRK nuclea
(n = 145)
137 (94%) 8 (6%) 5 (63%) 3 (38%) 0 (0%) 2 (11.9) 0 (0–100)
NTRK cy opl.
(n = 145)
141 (97%) 4 (3%) 2 (50%) 1 (25%) 1 (25%) 1 (8.8) 0 (0–100)
S-100 (n = 143) 143 (100%) 0 (0%) 0 (-) 0 (-) 0 (-) 0 (-) 0 (-)
HMB45 (n = 143) 140 (98%) 3 (2%) 3 (100%) 0 (0%) 0 (0%) 0 (1.8) 0 (0–20)
CD117 (n = 146) 145 (99%) 1 (1%) 1 (100%) 0 (0%) 0 (0%) 0 (1.7) 0 (0–20)
WT1 (n = 140) 84 (60%) 56 (40%) 18 (32%) 4 (7%) 34 (61%) 25 (43.3) 0 (0–200)
Smoo helin
nuclea (n = 142)
141 (99%) 1 (1%) 1 (100%) 0 (0%) 0 (0%) 0 (0.1) 0 (0–1)
Smoo helin
cy opl. (n = 142)
141 (99%) 1 (1%) 1 (100%) 0 (0%) 0 (0%) 0 (0.1) 0 (0–1)
Cyclin D1
(n = 141)
62 (44%) 79 (56%) 39 (49%) 26 (33%) 14 (18%) 25 (48.6) 2 (0–300)

Vi chows A chi
Vi chows A chi
o s aining, compa ed wi h ER. Rega ding he an ibodies
use ul in di e en ial diagnosis such as S-100, HMB45, and
CD117, hese showed almos comple e nega i i y.
O hose ma ke s no included in Table2, all cases
we e p53 wild ype, and ALK and ROS1 nega i e. The
exp ession o SMARCB1 (INI1) and SMARCA4 (BRG1)
was p ese ed, while se en cases showed an isola ed
loss o SMARCA2 (BRM). As he molecula p o ile o
hese umo s was known, none o hem ha bo ed BRM
al e a ions.
The di e ences in exp ession o all he s udied ma ke s
be ween he umo s wi h a con i med usion (n = 101) and
hose wi hou a ecu en usion (n = 32) we e also ana-
lyzed (Supplemen a y Table1). The esul s showed ha
only he endome ial s omal ma ke s CD10 and IFITM1
di e ed signi ican ly be ween he usion posi i e and nega-
i e g oups, wi h usion posi i e umo s showing a mo e
equen posi i e exp ession o bo h CD10 (p < 0.001) and
IFITM1 (p = 0.007).
Due o he imbalanced numbe o cases showing a ian
mo phologies compa ed wi h he usual pa e n, i was no
possible o s a is ically analyze he di e ences in exp ession
be ween hese g oups. Tumo s wi h a leas a componen o
ib oblas ic mo phology (n = 13) showed he exp ession o
α-SMA, desmin, h-caldesmon, calponin, and ansgelin in
5/13 (38%), 4/13 (31%), 3/13 (23%), 5/10 (50%), and 3/13
(23%) cases espec i ely. CD10 was posi i e in 11/13 (85%)
cases and IFITM1 in 7/13 (54%) cases. Tumo s wi h a sex
co d-s omal componen (n = 7) showed he exp ession o
α-SMA, desmin, h-caldesmon, calponin, and ansgelin in
5/7 (71%), 3/7 (43%), 2/7 (29%), 3/7 (43%), and 2/7 (29%)
cases, espec i ely, while CD10 was posi i e in 6/7 (86%)
cases and IFITM1 in 5/7 (71%) cases. The h ee umo s wi h
smoo h muscle mo phology we e all nega i e in desmin
and h-caldesmon, and he exp ession o α-SMA, calponin,
and ansgelin was seen in 2/3 cases. All cases we e CD10
posi i e, wi h IFITM1 exp ession in 2/3 cases.
Discussion
Low g ade endome ial s omal sa coma p esen s a diag-
nos ic challenge due o i s signi ican mo phological o e -
lap wi h o he spindle cell mesenchymal umo s o he
u e us. Recen s udies e-e alua ing he pa hological diag-
noses o u e ine mesenchymal umo s ha e epo ed ha
LG-ESS can be misdiagnosed in up o 20% o cases [22].
Cu en knowledge abou he immunohis ochemical p o ile
o LG-ESS is de i ed om s udies wi h a ela i ely low
numbe o examined cases, wi h he la ges immunohis o-
chemically analyzed se ies including ewe han 50 cases
[23]. Much o he published da a comes om single-case
s udies and small an ibody panels, limi ing he a ailabili y
o obus da a. Addi ionally, LG-ESS is bu dened wi h a
ela i ely high in e obse e a iabili y, as some cases can
only be eliably diagnosed using molecula es ing. This
in oduces bias in o olde da a, pa icula ly due o e ol ing
e minology and de ini ions, as some olde s udies only
use he e m “endome ial s omal sa coma,” which makes
he compa ison o hese esul s p oblema ic. Gi en hese
limi a ions, ou aim was o p o ide eliable da a om a
ca e ully selec ed, molecula ly examined se ies o LG-
ESS, he la ges se ies desc ibed in he li e a u e o IHC
p o iling o da e. Reliable da a on he immunop o ile o
LG-ESS a e especially c i ical in ou ine p ac ice, pa icu-
la ly o cases wi h equi ocal ea u es o cases whe e only
limi ed umo issue is a ailable (e.g., small diagnos ic
biopsies o u-cu biopsies).
The main IHC ma ke s o diagnosing endome ial s o-
mal umo s a e CD10 and he mo e ecen ly in oduced
ma ke IFITM1 [24, 25]. The exp ession o CD10 has long
been es ablished as one o he key diagnos ic ma ke s, as
i is sensi i e (75–100%) and has been epo ed in 258/294
(88%) o all cases o LG-ESS wi h he a ailable IHC
esul s o da e. Howe e , CD10 is no highly speci ic o
LG-ESS, as a small po ion o LG-ESS can be nega i e,
especially in cases wi h poo ixa ion [26]. CD10 exp es-
sion is also epo ed in up o 30% o smoo h muscle umo s
(including cellula leiomyoma), as well as in PEComa and
he sa coma ous componen o adenosa coma [10, 27–30].
Cu en ly, he e is limi ed da a on he exp ession o
IFITM1 in LG-ESS. Two s udies epo IFITM1 exp es-
sion in 83% (10/12) and 100% (16/16) o LG-ESS cases,
compa ed wi h 30% (6/20) and 40% (12/30) o smoo h
muscle umo s [25, 31]. Bo h s udies concluded ha
IFITM1 speci ici y (70% and 86,7%) su passes CD10,
al hough each s udy examined a ela i ely small se ies o
cases. IFITM1 exp ession, while no en i ely es ic ed o
LG-ESS, ends o be weak and ocal in smoo h muscle
umo s. Zhao e al. also emphasized IFITM1’s supe io
sensi i i y and speci ici y o e CD10, al hough also based
Fig. 1 Examples o CD10 and IFITM1 exp ession in di e en mo -
phological a ian s o LG-ESS. All mic opho og aphs aken a
100 × magni ica ion. A Di use, s ong exp ession o CD10 in a usual
LG-ESS (no usion de ec ed). B Focal exp ession o CD10 o a a i-
able in ensi y in a case wi h p edominan smoo h-muscle mo phology
(JAZF1::SUZ12 usion). C Mode a e o s ong CD10 exp ession in
an LG-ESS wi h a sex co d s omal-like mo phology (JAZF1::SUZ12
usion). D Di use, s ong exp ession o CD10 in a myxoid LG-ESS
(JAZF1::SUZ12 usion). E Di use and s ong exp ession o IFITM1
in LG-ESS wi h a usual mo phology (SVIL::EPC1 usion). F Com-
ple e IFITM1 nega i i y in a case wi h p edominan ly smoo h muscle
mo phology (JAZF1::SUZ12 usion). G Dispe se IFITM1 exp es-
sion o a a iable in ensi y in an LG-ESS wi h sex co d s omal-like
mo phology. Same case as depic ed in 1C (JAZF1::SUZ12 usion). H
Focal, occasional g anula exp ession o IFITM1 in a case wi h myx-
oid ea u es. Same case as depic ed in 1D (JAZF1::SUZ12 usion)
◂
Vi chows A chi
on a small se ies [32]. Ou s udy ound IFITM1 exp ession
in 98/143 cases (69%), mos ly ex ensi e o di use wi h
a iable in ensi y. IFITM1 seems o ou pe o m CD10 in
dis inguishing LG-ESS om smoo h muscle umo s; how-
e e , la ge s udies a e s ill needed o con i m i s u ili y. A
use ul app oach would he e o e include using bo h CD10
and IFITM1 as a pa o he IHC panel. In e es ingly, ou
s udy ound ha bo h CD10 and IFITM1 we e signi ican ly
mo e commonly exp essed in LG-ESS cases wi h a ecu -
en usion, which unde sco es he diagnos ic challenges o
Fig. 2 Examples o he exp ession o selec ed smoo h muscle ma k-
e s in di e en mo phological a ian s o LG-ESS. A α-SMA in
a usual LG-ESS, 100 × magni ica ion (JAZF1::SUZ12 usion). B
Focal i egula exp ession o α-SMA in LG-ESS wi h p edominan ly
smoo h-muscle mo phology, 200 × magni ica ion (JAZF1::SUZ12
usion). C Occasional a e exp ession o α-SMA in indi idual cells
in a myxoid a ian o LG-ESS (wi h posi i ely s aining essels in
he ield), 100 × magni ica ion (JAZF1::SUZ12 usion). D Almos
comple e nega i i y o ansgelin in a usual LG-ESS (wi h posi-
i ely s aining admixed myome ium), 100 × magni ica ion (no usion
de ec ed). E Focal exp ession o ansgelin smoo h muscle a ian o
LG-ESS, 200 × magni ica ion (JAZF1::SUZ12 usion, same case as
2B). F T ansgelin in an LG-ESS wi h sex co d s omal-like mo phol-
ogy showing i egula bu ex ensi e weak o mode a e exp ession,
100 × magni ica ion (JAZF1::SUZ12 usion)
Vi chows A chi
usion-nega i e umo s gi en he lack o ecu en gene ic
al e a ions and po en ially mo e equi ocal IHC esul s.
Ho mone ecep o s, speci ically ER (ERα) and PR, a e
ano he c ucial pa o he diagnos ic panel. The a ailable
li e a u e indica es ha ER is exp essed in 84,5% o LG-
ESS, wi h PR exp ession eaching 87% [15, 23, 33–38].
Mos s udies epo mo e ex ensi e, s onge PR exp ession
han ER, which is consis en wi h ou indings. The exp es-
sion o and ogen ecep o (AR) has also been desc ibed in a
high pe cen age o LG-ESS, hough he numbe o s udies
is limi ed [35]. While ER and PR a e ypically less use ul
o dis inguishing smoo h muscle neoplasms, hey can aid
in uling ou o he non-Mülle ian o igin umo s, pa icula ly
o ex au e ine LG-ESS.
Ho mone ma ke exp ession (especially ER and PR) may
also ha e p edic i e signi icance, wi h hei high exp ession
a es in LG-ESS suppo ing ho mone he apy such as high-
dose p oges ins, a oma ase inhibi o s, o GnRH analogues
as a iable ea men op ion [39–41]. A ecen me a-analysis
ound ha adju an ho mone he apy can educe ecu ence
isk in pa ien s wi h FIGO s ages I–II disease, bu wi h no
bene i conce ning o e all su i al [42]. Ano he s udy sug-
ges ed ha ho mone he apy can also educe ecu ence isk
e en in s ages II–IV LG-ESS, ecommending 12 mon hs o
high-dose p oges in ea men pos -su ge y [43]. While he
e ec i eness o ho mone he apy is deba ed, epo ing umo
ho mone s a us emains essen ial as i can help guide ea -
men decisions. Con lic ing e idence also exis s ega ding
ER/PR exp ession as a p ognos ic ac o , equi ing alida-
ion h ough a molecula ly con i med se ies [35, 43, 44].
In ou ine p ac ice, one o he common diagnos ic pi -
alls lies in di e en ia ing cellula leiomyoma (CL) om
LG-ESS. The co ec diagnosis o CL e sus LG-ESS is
o ex eme clinical impo ance, gi en he di e en biologi-
cal na u es and beha io o hese umo s. In which case,
a combina ion o smoo h muscle ma ke s such as α-SMA,
desmin, h-caldesmon, calponin, ansgelin, and smoo he-
lin, and endome ial s omal ma ke s, such as CD10 and
IFITM1, is essen ial. Smoo h muscle ma ke s a e equen ly
posi i e in LG-ESS, especially in cases wi h smoo h muscle
Fig. 3 Examples o he exp ession o selec ed smoo h muscle ma k-
e s in di e en mo phological a ian s o LG-ESS, con inued. A
Nega i i y o desmin in a usual a ian o LG-ESS, 100 × magni ica-
ion (no usion de ec ed, same case as 2D). B Dispe se s ong exp es-
sion o desmin in a myxoid a ian o LG-ESS, 100 × magni ica ion
(JAZF1::SUZ12 usion, same case as 2C). C) Focal weak o mode a e
exp ession o h-caldesmon in an LG-ESS wi h smoo h muscle mo -
phology, 200 × magni ica ion (JAZF1::SUZ12 usion). D Smoo helin
in a usual LG-ESS, 200 × magni ica ion (no usion de ec ed, same
case as 2D and 2G)
Vi chows A chi
32 Chai andDepa men o Obs e ics, Gynaecology
andGynaecological Oncology, Medical Uni e i y o Wa saw,
Wa saw, Poland
33 Depa men o Gynecological Oncology, Pome anian
Hospi als, Gdynia, Poland
34 Su gical Oncology Clinic, Medical Uni e si y inGdansk,
Gdansk, Poland
35 Depa men o Gynecological Oncology, W oclaw Medical
Uni e si y, W oclaw, Poland
36 Depa men o Oncological Gynecology, Lowe Silesian
Oncology, Pulmonology andHema ology Cen e , W oclaw,
Poland
37 Depa men o Pa hology, Uni e si y o Medicine, Pha macy,
Sciences andTechnology GE Palade, Ta guMu es, Romania
38 Depa men o Gynecology, Uni e si y o Medicine,
Pha macy, Sciences andTechnology GE Palade,
Ta guMu es, Romania
39 Depa men o Pa hology, Uni e si y Hospi al Bulo ka,
P ague, CzechRepublic
40 Depa men o Gynecology andObs e ics, Cha les
Uni e si y - Fi s Facul y o Medicine andUni e si y
Hospi al Bulo ka, P ague, CzechRepublic
41 Gynecologic Oncology Uni , Depa men o Woman
andChild Heal h andPublic Heal h, Fondazione Policlinico
Uni e si a io ’A. Gemelli’ IRCCS, Rome, I aly
42 Sec ion o Obs e ics andGynecology, Uni e si y
Depa men o Li e Sciences andPublic Heal h, Uni e si à
Ca olica del Sac o Cuo e, Rome, I aly
43 Caucasus Medical Cen e, Onco-gynecological Depa men ,
Tbilisi, Geo gia
44 Caucasus Medical Cen e, Megalab,Tbilisi, Geo gia
45 Depa men o Gynecologic Oncology, Khmelny skyi
egional an i umo cen e , Khmelny skyi, Uk aine