ORIGINAL RESEARCH Open Access
Neu on-speci ic enolase has po en ial alue
as a bioma ke o [
18
F]FDG/[
68
Ga]Ga-PSMA-
11 PET misma ch indings in ad anced
mCRPC pa ien s
Flo ian Rosa
1*
, Kalle Ribba
1
, Ma in Ries
1
, Johannes Linxweile
2
, Ma k Ba holomä
1
, S ephan Maus
1
,
Ma hias Sch eckenbe ge
3
, Same Ezziddin
1
and Fadi Kh eish
1
Abs ac
Backg ound: PSMA- a ge ed adioligand he apy (PSMA-RLT) yielded imp essi e esul s in he me as asized
cas a ion- esis an p os a e ca cinoma (mCRPC) se ing. High exp ession o PSMA is essen ial o success ul PSMA-
RLT. Howe e , some pa ien s de elop [
18
F]FDG-a id lesions wi h low o no PSMA exp ession ([
18
F]FDG/[
68
Ga]Ga-
PSMA-11 misma ch indings on PET/CT) in he cou se o ea men . Those lesions a e no a ec ed by PSMA-RLT and
a change in he apy managemen is needed. To enable ea ly misma ch de ec ion, possible blood pa ame e s as
indica o s o he occu ence o [
18
F]FDG/[
68
Ga]Ga-PSMA-11 misma ch indings on PET/CT we e e alua ed.
Me hods: Re ospec i e s udy o N= 66 ad anced mCRPC pa ien s wi h dual [
68
Ga]Ga-PSMA-11 and [
18
F]FDG PET/
CT imaging wi hin 4 weeks, who we e e e ed o o ecei ed [
177
Lu]Lu-PSMA-617 adioligand he apy. P os a e-
speci ic an igen (PSA), neu on-speci ic enolase (NSE), gamma-glu amyl ans e ase (GGT), and alkaline phospha ase
(ALP) we e es ed as indica o s o he occu ence o [
18
F]FDG/[
68
Ga]Ga-PSMA-11 misma ch indings. Addi ional o
absolu e alues, ela i e changes (ΔPSA, ΔNSE, ΔGGT, ΔALP) o e a pe iod o 4 ± 1 weeks p io o [
18
F]FDG PET/CT
we e analyzed.
Resul s: In o al, 41/66 (62%) pa ien s e ealed a leas one [
18
F]FDG/[
68
Ga]Ga-PSMA-11 misma ch inding on PET/
CT. These misma ch indings we e de ec ed in 13/41 (32%) pa ien s by sc eening o and in 28/41 (68%) pa ien s
du ing PSMA-RLT. NSE se um le el (55.4 ± 44.6 μg/l s.18.5 ± 8 μg/l, p< 0.001) and ΔNSE (93.8 ± 124.5% s.2.9 ±
39.5%, p< 0.001) we e signi ican ly highe in he misma ch g oup han in he non-misma ch g oup. No signi ican
di e ences we e ound o se um PSA (p= 0.424), ΔPSA (p= 0.417), se um ALP (p= 0.937), ΔALP (p= 0.611), se um
GGT (p= 0.773), and ΔGGT (p= 0.971). Fo NSE and ΔNSE, he maximum alue o he Youden index in ROC
analysis was a a cu -o le el o 26.8 μg/l (sensi i i y 78%, speci ici y 96%) and a + 13.9% (sensi i i y 84%, speci ici y
75%), espec i ely. An in oduced sco ing sys em o bo h pa ame e s achie ed a sensi i i y o 90% and a speci ici y
o 88% o he occu ence o [
18
F]FDG/[
68
Ga]Ga-PSMA-11 misma ch.
(Con inued on nex page)
© The Au ho (s). 2020 Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e
app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i
changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle's C ea i e Commons
licence, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle's C ea i e Commons
licence and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain
pe mission di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
* Co espondence: [email p o ec ed]
1
Depa men o Nuclea Medicine, Saa land Uni e si y—Medical Cen e ,
Ki be ge S . 100, Geb. 50, 66421 Hombu g, Ge many
Full lis o au ho in o ma ion is a ailable a he end o he a icle
Rosa e al. EJNMMI Resea ch (2020) 10:52
h ps://doi.o g/10.1186/s13550-020-00640-2
(Con inued om p e ious page)
Conclusion: We obse ed a signi ican ly highe absolu e se um concen a ion and a highe ela i e inc ease o NSE
in ad anced mCRPC pa ien s wi h [
18
F]FDG-a id and insu icien PSMA exp essing me as ases ([
18
F]FDG/[
68
Ga]Ga-
PSMA-11 misma ch indings on PET/CT) in ou coho . NSE migh be used as a po en ial labo a o y indica o o
[
18
F]FDG/[
68
Ga]Ga-PSMA-11 misma ch indings, i his obse a ion is con i med in u u e, ideally p ospec i e, s udies
in la ge pa ien coho s.
Keywo ds: P os a e cance , mCRPC, PSMA PET/CT, FDG PET/CT, Misma ch
In oduc ion
Wi h o e 1,000,000 new cases and app oxima ely 300,000
dea hs wo ldwide in 2012, p os a e ca cinoma is one o he
mos equen malignan diseases in men [1]. A signi ican
ac ion o p os a e ca cinoma pa ien s’p og esses o he le-
hal me as asized cas a ion- esis an p os a e ca cinoma
(mCRPC) se ing [2,3]. Du ing he las decade, howe e ,
he de elopmen o new ea men op ions o men wi h
mCRPC has led o an imp o ed su i al ime [4]. In
addi ion o chemo he apy wi h doce axel o cabazi axel [5,
6] and nex -gene a ion and ogen ecep o signaling inhib-
i ion wi h abi a e one o enzalu amide [7,8], adioligand
he apy a ge ing he p os a e-speci ic memb ane an igen
(PSMA) is a po en ial op ion in pallia i e se ings [9,10].
PSMA- a ge ed adioligand he apy (PSMA-RLT) wi h
[
177
Lu]Lu-PSMA-617 yielded imp essi e esul s in pallia i e
se ings while causing only mode a e side e ec s [11,12].
PSMA- a ge ed based posi on emission omog aphy
(PET)/compu e omog aphy (CT) wi h adiolabeled PSMA
ligands, such as [
68
Ga]Ga-PSMA-11, is equen ly used o
imaging o p os a e cance in clinical ou ine [13]. PSMA-
a ge ed PET/CT is no only used o he s aging o p os-
a e cance , bu i is also a use ul ool o he apy moni o -
ing [14,15] and indispensable o e i ying PSMA
exp ession p io o PSMA-RLT [16]. High exp ession o
PSMA is essen ial o success ul PSMA-RLT in pa ien s
wi h mCRPC. Howe e , some pa ien s de elop lesions wi h
low o no PSMA exp ession unde ongoing ea men [17].
Those lesions a e no a ec ed by PSMA-RLT and a change
in he apy managemen is needed [9,18]. [
18
F]FDG PET/
CT using
18
F-labeled luo odeoxyglucose ([
18
F]FDG) in
addi ion oaPSMA- a ge edPET/CTmaybeasui able
me hod o de ec ion o hose lesions. Mos ly, p os a e ca -
cinoma cells ha e a low glucose me abolism due o ene gy
gain by lipids and o he ene ge ic molecules bu in ad-
anced la e-s age disease he glucose me abolism is highly
inc eased by he Wa bu g e ec and shi ing o ae obic
glycolysis a e nume ous mu a ion e en s [19]. The e o e,
a combina ion o [
18
F]FDG and PSMA- a ge ed PET/CT
may be able o de ec clinically ele an glucose me abolic
iable umo lesions wi h low o no PSMA exp ession (mis-
ma ch lesions) in he ad anced mCRPC se ing. A ecen ly
p ospec i e phase-II ial o [
177
Lu]Lu-PSMA-617 RLT
excluded 16% o sc eened pa ien s due o missing o low
PSMA exp ession in [
18
F]FDG-a id me as ases [20]. Ea ly
de ec ion o hese misma ch indings is hus needed o p o-
ide pa ien s wi h al e na i e he apy op ions, addi ionally
o o ins ead o PSMA-RLT [18]. Howe e , equen ly pe -
o ming [
18
F]FDG PET/CT in addi ion o PSMA- a ge ed
PET/CT is e y cos in ensi e and associa ed wi h add-
i ional adia ion exposu e o he pa ien , which should be
a oided e en in a pallia i e se ing.
In his s udy, selec ed se um pa ame e s we e he e o e
es ed as indica o s o he occu ence o [
18
F]FDG/
[
68
Ga]Ga-PSMA-11 misma ch indings on PET/CT in his
s udy, including he p os a e-speci ic an igen (PSA) as he
ou ine con ol and esponse pa ame e [21], alkaline
phospha ase (ALP) as known o be ele a ed by bone me-
as ases [22], gamma-glu amyl ans e ase (GGT) as a pa -
ame e o li e unc ion a ec ed by li e me as ases [23],
and neu on-speci ic enolase (NSE) as a possible pa ame e
o ansdi e en ia ing o a neu oendoc ine ype o p os-
a e ca cinoma [24].
Ma e ials and me hods
S udy design
Re ospec i e monocen e s udy o mCRPC pa ien s
wi h dual [
68
Ga]Ga-PSMA-11 and [
18
F]FDG PET/CT
imaging wi hin 4 weeks, who we e e e ed o o e-
cei ed [
177
Lu]Lu-PSMA-617 adioligand he apy a he
clinic o nuclea medicine a Saa land Uni e si y Med-
ical Cen e om Oc obe 2015 ill Augus 2019. Pa ien s
wi h seconda y malignancies we e excluded o a oid po-
en ial in e e ence o image in e p e a ion.
Pa ien s and e hics
N= 66 o in o al 167 mCRPC pa ien s e e ed o o
ecei ed PSMA-RLT in ou cen e we e included in his
e ospec i e s udy. Two o he 167 pa ien s we e ex-
cluded because o incomple e blood examina ion, 3/167
because o seconda y malignancies and he emaining,
and 96/167 due o missing o un imely [
18
F]FDG PET/
CT. The pa ien s ecei ed a [
68
Ga]Ga-PSMA-11 PET/
CT and [
18
F]FDG PET/CT wi hin a sho ime pe iod
p io o in ended commencemen o PSMA-RLT (n=
14/66) o in he cou se o PSMA-RLT (n= 52/66). The
mean ime be ween bo h PET/CT scans was 7.3 ± 10.7
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 2 o 10
days (95% con idence in e al o he mean (CI) [4.6; 9.9]).
The mean age o he pa ien s was 69 yea s [ ange 45–89
yea s]. All pa ien s ecei ed se e al p e ea men s. De ailed
in o ma ion abou he p e ea men s and he pa ien cha -
ac e is ics is p esen ed in Table 1. And ogen dep i a ion
he apy (ADT) was con inued unchanged in all pa ien s o
a oid a ia ion o PSMA exp ession. [
68
Ga]Ga-PSMA-11
and [
18
F]FDG PET/CT we e pe o med on a compassiona e
use basis unde he Ge man Pha maceu ical Ac §13 (2b).
Pa ien s ga e w i en consen a e being ho oughly in-
o med abou he isks and po en ial side e ec s o his
in e en ion. Addi ionally, pa ien s consen ed o publica ion
o any esul ing da a in acco dance wi h he Decla a ion o
Helsinki. Re ospec i e analysis app o al was wai ed by he
local ins i u ional e iew boa d.
PET acquisi ion and analysis
Fo PET imaging, a mean ac i i y o 124.1 ± 14.4 MBq
[
68
Ga]Ga-PSMA-11 (CI [120.6; 127.6]) and 268.6 ± 28.7
MBq [
18
F]FDG (CI [261.6; 275.7]) was adminis e ed,
ollowed by a 500-ml in usion o NaCl 0.9%. Fas ing
mean blood glucose alue was 98.1 ± 17.3 mg/dl (CI
[93.8; 102.4]) be o e adminis a ion o [
18
F]FDG. The
mean up ake ime was app oxima ely 60 min (61.8 ± 6.6
min, CI [60.1; 63.4]) o [
68
Ga]Ga-PSMA-11 acco ding
o s anda d p ocedu es o p os a e cance imaging [25]
and 90 min (91.6 ± 8.7 min, CI [89.4; 93.7]) o
[
18
F]FDG, acco ding o he ou s anda d p ocedu e and
Ge man guideline o umo imaging [26]. Be o e da a
acquisi ion, all pa ien s we e ad ised o emp y hei
bladde . No diu e ics we e applied. All PET/CT scans
we e pe o med using a Biog aph 40 mCT PET/CT
scanne (Siemens Medical Solu ions, Knox ille, TN,
USA) wi h EANM Resea ch L d. acc edi a ion. The PET
acquisi ion was pe o med om e ex o mid- emu
wi h 3-min acquisi ion ime pe bed posi ion. A bed pos-
i ion co e s 21.4-cm ex ended ield-o - iew (T ueV).
The PET da ase s we e econs uc ed using an i e a i e
3-dimensional OSEM (o de ed-subse expec a ion
maximiza ion) algo i hm (3 i e a ions; 24 subse s) wi h
Gaussian il e ing and a slice hickness o 5 mm. Ran-
dom co ec ion, decay co ec ion, sca e a enua ion,
and a enua ion co ec ion we e applied. The CT was
pe o med in low-dose echnique using an X- ay ube
ol age o 120 keV and a modula ion o he ube cu en
by applying CARE Dose4D wi h a maximal ube cu en -
ime p oduc o 30 mAs. All PET/CT da a se s we e isu-
ally analyzed using ce i ied analysis so wa e (Sec a
PACS—Sec a Medical Sys ems GmbH, Cologne,
Ge many). [
18
F]FDG and [
68
Ga]Ga-PSMA-11 PET/CT
we e ead simul aneously by h ee expe ienced physicians
(a leas 5 yea s o expe ience in PET eading) sea ching
o misma ch indings. A misma ch inding was de ined as
me as asis wi h ema kable [
18
F]FDG up ake and no o
conside ably less conco dan [
68
Ga]Ga-PSMA-11 up ake
based on isual analysis. The decision o a misma ch ind-
ing was aken in consensus o all PET eade s.
Se um pa ame e s
Selec ed se um pa ame e s used in ou clinical p ac ice we e
es ed as indica o s o he occu ence o [
18
F]FDG/[
68
Ga]Ga-
PSMA-11 misma ch indings in PET/CT: p os a e-speci ic
an igen (PSA), neu on-speci ic enolase (NSE), gamma-
glu amyl ans e ase (GGT), and alkaline phospha ase (ALP).
Absolu e alues we e es ed in all 66 pa ien s. Addi ionally,
ela i e changes (ΔPSA, ΔNSE, ΔGGT, ΔALP) o e a pe iod
o 4 ± 1 weeks p io o [
18
F]FDG PET/CT we e analyzed in
55/66 pa ien s (83.3%). In 11/66 cases (16.7%) no labo a o y
da a om p e ious blood samples we e a ailable.
S a is ical analysis and sco ing sys em
Fo s a is ical analysis, Mann-Whi ney U es was ap-
plied using P ism 8 (G aphPad So wa e, San Diego,
Table 1 Pa ien cha ac e is ics
Cha ac e is ic Value
Age [yea s] 69 (45–89)
PSA [ng/dl] 70 (0.3–4742)
Time om ini ial diagnosis [yea s]
≤2 18 (27.3%)
>2 o≤5 20 (30.3%)
> 5 28 (42.4%)
P io he apy
P os a ec omy 29 (44%)
Radia ion 39 (59%)
ADT 66 (100%)
Enzalu amide 50 (76%)
Abi a e one 47 (71%)
Doce axel 37 (56%)
Cabazi axel 24 (36%)
Xo igo 7 (11%)
ECOG PS be o e i s cycle
0 18 (27%)
1 41 (62%)
2 5 (8%)
3 2 (3%)
Si e o me as asis
Bone 60 (91%)
Lymph node 49 (74%)
Li e 29 (44%)
Lung 10 (15%)
Da a a e p esen ed as n(%) o median ( ange)
Abb e ia ions:PSA p os a e-speci ic an igen, ECOG PS Eas e n Coope a i e
Oncology G oup Pe o mance S a us, ADT and ogen dep i a ion he apy
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 3 o 10
USA) o de e mine signi ican di e ences be ween he
g oups. A p alue < 0.05 was ega ded as s a is ically sig-
ni ican . Fo powe calcula ion o Mann-Whi ney U es ,
we calcula ed he e ec size; alues highe han 0.5 we e
conside ed as a s ong e ec size. Fo each pa ame e ha
u ned ou o be s a is ically signi ican , a ecei e ope a -
ing cha ac e is ic (ROC) analysis was pe o med and a
cu -o poin was de e mined using he maximal alue o
Youden index. In addi ion, a sco ing sys em combining all
signi ican pa ame e s was in oduced. This sco ing
sys em was cons uc ed o ease o clinical use and on he
basis o sensi i i y-speci ici y analysis o he s onges
pa ame e , assis ed by he o he signi ican pa ame e s,
enhancing he powe and accu acy o he sco e.
Resul s
In o al, 41/66 (62%) pa ien s included in his analysis e-
ealed a leas one [
18
F]FDG/[
68
Ga]Ga-PSMA-11 mis-
ma ch inding on PET/CT, and 25/66 (38%) had no
misma ch indings. O e all, 390 misma ch lesions we e
de ec ed: 211 in bone (in 24/41 pa ien s), 62 in lymph
nodes (in 21/41 pa ien s), 100 in he li e (in 20/41 pa-
ien s), 4 in he lung (in 1/41 pa ien s), and 13 in o he lo-
ca ions (in 9/41 pa ien s). Figu e 1p esen s an example o
a pa ien wi h mul iple misma ch indings in he li e . In
he misma ch g oup, 13/41 (32%) pa ien s had hei mis-
ma ch indings (121/390 misma ch lesions) de ec ed a
baseline imaging be o e in ended commencemen o
PSMA-RLT, whe eas in he emaining 28/41 (68%) pa-
ien s he misma ch (269/390 misma ch lesions) was diag-
nosed in he cou se o PRLT a e ha ing ecei ed a mean
o 4 ± 2 cycles o PSMA-RLT. When compa ing hese 269
misma ch lesions o he ini ial [
68
Ga]Ga-PSMA-11 PET/
CT be o e s a ing PSMA-RLT, 29/269 (11%) we e ini-
ially in ensely PSMA-posi i e, 52/269 (19%) we e mode -
a ely PSMA-posi i e, and he majo i y, 188/269 (70%),
we e no iden i iable on ini ial [
68
Ga]Ga-PSMA-11 PET/
CT. Figu e 2shows an example o wo misma ch lesions,
one in ensely PSMA-posi i e and one no iden i iable on
he ini ial [
68
Ga]Ga-PSMA-11 PET/CT.
NSE se um le el (55.4 ± 44.6 μg/l s.18.5 ± 8 μg/l, p<
0.001) and ΔNSE (93.8 ± 124.5% s.2.9 ± 39.5%, p<
0.001) we e signi ican ly highe in he misma ch g oup
han in he non-misma ch g oup. No signi ican di e -
ence was ound o se um PSA (p= 0.424), ΔPSA (p=
0.417), se um ALP (p= 0.937), and ΔALP (p= 0.611)
be ween he misma ch and he non-misma ch g oup, e-
spec i ely. Twen y-nine o 66 (44.0%) pa ien s had li e
me as ases, and 20/29 had a leas one misma ch inding
in he li e . In pa ien s wi h li e me as ases, GGT (p=
0.773) and ΔGGT (p= 0.971) also e ealed no signi i-
can di e ence be ween he misma ch and he non-
misma ch g oup. Fo NSE and ΔNSE, he e ec size was
0.70 and 0.59, which we e conside ed as s ong. Summa-
ized s a is ics o he da a a e p esen ed in Table 2. All
hese s a is ical compa isons a e illus a ed by box-plo
diag am o ma o each es ed pa ame e (Fig. 3). A
subg oup analysis compa ing he pa ien s diagnosed
wi h a misma ch by sc eening o PSMA-RLT wi h he
pa ien s diagnosed wi h a misma ch a e se e al cycles
PSMA-RLT e ealed no signi ican di e ence o se um
NSE alue (67.5 ± 51.7 μg/l s.49.9 ± 40.8 μg/l, p=
0.12). The p edic i e impac o NSE and ΔNSE ega ding
misma ch de ec ion is disce nible om he wa e all
plo s wi h highligh ing o misma ch and non-misma ch
indi iduals (Fig. 4). ROC analyses o NSE and ΔNSE
Fig. 1 Pa ien wi h hepa ic misma ch indings. Maximal in ensi y p ojec ion (MIP), PET/CT, and PET da a o [
68
Ga]Ga-PSMA-11 PET/CT (a) and o
[
18
F]FDG PET/CT (b)
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 4 o 10
e ealed an a ea unde he cu e (AUC) o 0.92 and
0.84, espec i ely (Fig. 5). Fo NSE, he maximum alue
o he Youden index (J= 0.74) was a a se um le el o
26.8 μg/l wi h a sensi i i y o 78% and a speci ici y o
96% o he occu ence o [
18
F]FDG/[
68
Ga]Ga-PSMA-11
misma ch lesions, whe eas he maximum alue o he
Youden Index o ΔNSE (J= 0.59) was a + 13.9% in-
c ease wi h a sensi i i y o 84% and a speci ici y o 75%.
Only one pa ien had misma ch indings wi h se um
NSE < 15 μg/l. On he o he hand, we obse ed se e al
misma ch (13/41) and non-misma ch (16/25) pa ien s
wi h NSE in he ange be ween 15 and 30 μg/l (Fig. 4a).
To dicho omize hese pa ien s and o imp o e he sensi-
i i y wi hou ma ked loss o speci ici y we in oduced a
sco ing sys em (Combined NSE Sco e) based on bo h
pa ame e s, which is p esen ed in Fig. 6a. The main pa
o he sco ing sys em was assigned o he absolu e alue
o NSE being he s onges pa ame e in ROC analysis.
Ze o poin s we e gi en o NSE in he physiological
ange o < 15 μg/l. Fo he ease o clinical use, NSE was
classi ied in he c i ical ange (15–30 μg/l) in s eps o
5μg/l, s a ing wi h 1 poin o 15–20 μg/l up o 3 poin s
o 25–30 μg/l co e ing he Youden’s index (26.8 ng/ml)
o he sensi i i y-speci ici y analysis. G ea e in e al
s eps we e chosen o NSE abo e 30 μg/l (4 poin s 30–
50 μg/l, 5 poin s 50–100 μg/l and 6 poin s > 100 μg/l).
ΔNSE was addi ionally included o enhancing he
powe and accu acy o he sco e. Addi ional poin s o
ΔNSE was se o −1 poin when NSE was dec easing
(mo e han −20 %), o 0 poin s wi h s able (−20 o + 20
Fig. 2 Example o wo misma ch lesions de ec ed a e 5 cycles o PSMA-RLT: one in ensely PSMA-posi i e (g een a ow) and one no iden i iable
(blue a ow) lesion on he ini ial [
68
Ga]Ga-PSMA-11 PET/CT a baseline p io o PSMA-RLT (a). [
68
Ga]Ga-PSMA-11 PET/CT (b) and [
18
F]FDG PET/CT
(c) a e 5 cycles o PSMA-RLT showing [
18
F]FDG/[
68
Ga]Ga-PSMA-11 misma ch in bo h lesions
Table 2 Desc ip i e s a is ics o se um pa ame e s
G oup nMedian (IQR) Mean (± SD) Minimum Maximum p alue
NSE [μg/l] Misma ch 41 35 (30.0–66.9) 55.4 (± 44.6) 15 188.6 < 0.001
Non-misma ch 25 16.9 (13.0–20.4) 18.5 (± 8.0) 10 50
PSA [ng/ml] Misma ch 41 168 (83.0–602.5) 367.6 (± 407.9) 0.3 1360 0.424
Non-misma ch 25 190 (116.8–743.0) 666.3 (± 1086.8) 3 4742
GGT [U/l] Misma ch 22 75 (52.5–184.5) 120.5 (± 104.5) 22 378 0.773
Non-misma ch 9 76 (30.0–163.0) 127 (± 149.9) 20 497
ALP [U/l] Misma ch 41 140 (86–242.5) 192.2 (± 163.5) 12 818 0.937
Non-misma ch 25 136 (83.5–306.5) 211.8 (± 209.6) 35 1042
ΔNSE [%] Misma ch 31 47 (18.6–137.3) 93.8 (± 124.5) −12 533 < 0.001
Non-misma ch 24 3 (−20.5 o 15.6) 2.9 (± 39.5) −62 140
ΔPSA [%] Misma ch 31 27 (−4.3 o 109.1) 254.7 (± 890.3) −60 4956 0.417
Non-misma ch 24 15 (−9.7 o 64.4) 63.8 (± 153.5) −89 631
ΔGGT [%] Misma ch 18 65.5 (30.3–143.8) 196.8 (± 464.9) −79 1986 0.971
Non-misma ch 9 70 (2.0–138.5) 205.3 (± 454.6) −31 1406
ΔALP [%] Misma ch 31 0 (−14.0 o 39.0) 35.3 (± 137.6) −33 753 0.611
Non-misma ch 24 12.5 (−13.8 o 40.0) 28.2 (± 76.4) −41 342
Abb e ia ions:NSE neu on-speci ic enolase, PSA p os a e-speci ic an igen, GGT gamma-glu amyl ans e ase, ALP alkaline phospha ase
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 5 o 10

%) o unknown NSE, and o 1 (20–50%) o o 2 (> 50%)
poin s o inc easing NSE. The ROC analysis o he de el-
oped sco ing sys em (Fig. 6b) e ealed an AUC o 0.91
and a maximum alue o he Youden index o 0.78 a 3
sco ing poin s wi h a sensi i i y o 90% and a speci ici y o
88% o he occu ence o [
18
F]FDG/[
68
Ga]Ga-PSMA-11
misma ch lesions. The gain o he Combined NSE Sco e is
seen a he sensi i i y le el o abou 90%, whe e he iso-
la ed use o NSE achie es only a speci ici y o 76%.
Discussion
We obse ed a signi ican ly highe absolu e se um con-
cen a ion and a highe ela i e inc ease o NSE in ad-
anced mCRPC pa ien s wi h [
18
F]FDG-a id and
insu icien PSMA exp essing me as ases ([
18
F]FDG/
[
68
Ga]Ga-PSMA-11 misma ch indings) in ou coho .
NSE is a highly speci ic ma ke o neu ons and pe iph-
e al neu oendoc ine cells, used as a bioma ke o agg es-
si e o ms o neu oendoc ine umo s and small cell
ca cinoma [27,28]. Ele a ed NSE se um le els ha e also
been obse ed in neu oendoc ine sub ypes o p os a e ca -
cinoma [29,30].Whileonlyasmallp opo iono p os a e
cance pa ien s ep esen neu oendoc ine ea u es a he
beginning o disease [31], mo e pa ien s de elop hese
unde he apy [24]. Se um NSE migh be inc eased in pa-
ien s wi h misma ch indings as a sign o dedi e en ia ion
o ans o ma ion o a neu oendoc ine sub ype. This could
be a mechanism o esis ance induced by selec i e ea -
men p essu e and was pa icula ly obse ed in pa ien s
unde going an i-and ogen he apy [32–34]. Neu oendo-
c ine di e en ia ion in p os a e ca cinoma is a pheno ypic
change by which p os a e cance cells ans-di e en ia e
in o neu oendoc ine-like cells. These neu oendoc ine-like
cells a e lacking exp ession o and ogen ecep o and
p os a e-speci ic an igen. Neu oendoc ine-like cells a e
known o p oduce pep ide ho mones and g ow h ac o s o
p omo e umo p og ession and a e apop osis- esis an
con ibu ing o ea men ailu e [35–37]. Some case e-
po s desc ibed ha PSMA- a ge ed PET/CT may be no
able o de ec neu oendoc ine p os a e ca cinoma [38–40].
Suppo ing hese obse a ions, Bakh e al. (2018) demon-
s a ed in i o on pa ien -de i ed xenog a (PDX) an in-
e se co ela ion be ween PSMA gene (FOLH1) and
neu oendoc ine bioma ke gene exp ession [41]. Pa icu-
la ly, supp ession o he PSMA gene was obse ed in 65%
o cases which o e exp essed he NSE gene (ENO2). They
also demons a ed ha he mos p og essions’pa hway o
neu oendoc ine ans-di e en ia ion unde ongoing
Fig. 3 Compa ison o absolu e se um alues o NSE (a), PSA (b) and ALP (c) be ween all misma ch- and non-misma ch pa ien s. Compa ison o
se um GGT (d) o pa ien s ha ing misma ch and non-misma ch li e me as ases. Rela i e change o each pa ame e : ΔNSE (e), ΔPSA ( ), ΔALP (g),
ΔGGT (h). Ex eme ou lie s a e no shown
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 6 o 10
Fig. 4 Wa e all plo o NSE (a) and ΔNSE (b) alues in descending o de and colo coding in o misma ch ( ed) and a non-misma ch (blue)
Fig. 5 ROC cu es o misma ch p edic ion by se um NSE (a) and ΔNSE (b) wi h maximum alue o he Youden Index (J)
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 7 o 10
ea men was associa ed subsequen ly wi h loss o he
PSMA exp ession. Thus, al e na i e molecula imaging
me hods a e equi ed o neu oendoc ine pheno ype.
The glucose anspo e GLUT1 has been ound o be
o e exp essed on cells o he ad anced la e-s age p os-
a e cance and i s exp ession was ela ed o umo ag-
g essi eness [42]. Meziou e al. (2020) assumed ha
GLUT1 exp ession may be inc eased in neu oendoc ine
p os a e ca cinoma sugges ing ha [
18
F]FDG PET/CT
migh be an impo an imaging ool o examine neu o-
endoc ine p os a e ca cinoma [43]. Few s udies demon-
s a ed he clinical u ili y o [
18
F]FDG PET/CT in
neu oendoc ine p os a e ca cinoma [44–46]. Sp a e al.
demons a ed in 23 pa ien s wi h neu oendoc ine p os-
a e cance ha [
18
F]FDG PET/CT has clinical bene i
and epo ed a high de ec ion a e o me as a ic disease,
especially o lymph node and isce al me as ases [46].
This hypo hesis o ans o ma ion o p os a e cance
wi h neu oendoc ine ea u es in ou misma ch coho
was con i med his opa hologically in one pa ien only.
Mo e e idence is equi ed o con i m his hypo hesis.
Thus, consis en his opa hological examina ion o mis-
ma ch me as ases should be pe o med in u u e s udies.
Pa ien s wi h mCRPC, p esen ing disco dan [
18
F]FDG-
a id lesions wi h low o no PSMA exp ession, ha e a poo
p ognosis wi h e y sho su i al ime [47]. This misma ch
phenomenon indica es a mo e agg essi e ype o mCRPC
in clinical ou ine and p ecludes pa ien s om PSMA-
di ec ed adioligand he apy [9]. Thus, ea ly diagnosis o
misma ch indings is e y impo an be o e and du ing
PSMA-RLT o p o ide hese pa ien s wi h al e na i e he -
apy op ions in addi ion o o ins ead o PSMA-RLT, includ-
ing chemo he apy, immuno he apy, bone-seeking
adiopha maceu icals, PARP inhibi ion, and o he no el
a ge ed ea men s [18,47].
Ou esul s ob ained om 66 mCRPC pa ien s sugges
ha se um NSE is a po en ial bioma ke o he exis -
ence o he desc ibed [
18
F]FDG/[
68
Ga]Ga-PSMA-11 mis-
ma ch. S a is ical analysis o ou da a p oposed an NSE
cu -o alue o a misma ch occu ence o 26.8 μg/l
(sensi i i y 78%, speci ici y 96%) and o + 13.9 % inc ease
o ΔNSE (sensi i i y 84%, speci ici y 75%). A combin-
a ion o bo h pa ame e s may imp o e he p edic i e
powe . We achie ed a sensi i i y and speci ici y o al-
mos 90% wi h ou in oduced Combined NSE Sco e
(cu -o : o al poin s ≥3, Fig. 6). The simplici y o he
sco e allows easy clinical applica ion as a guide o u -
he diagnos ic p ocedu es in he mCRPC se ing.
In con as o NSE, o he se um pa ame e s analyzed
in his s udy as PSA, ALP, and GGT we e no able o in-
dica e misma ch indings in ou coho . These bio-
ma ke s may only e lec o al umo bu den gi ing no
speci ic hin owa ds [
18
F]FDG-a id lesions wi h low o
missing PSMA exp ession.
We sugges including NSE assessmen s in o ou ine
blood es s o mCRPC pa ien s be o e and du ing PSMA-
RLT. Howe e , hese esul s should be no iced wi h cau-
ion conside ing a po en ial bias due o he e ospec i e
Fig. 6 Combined NSE Sco e (a) and ROC cu e o misma ch p edic ion by he sco e (b) wi h maximum alue o he Youden index (J)
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 8 o 10
s udy design and no - ep esen a i e pa ien coho . Qui e
o en, [
18
F]FDG PET/CT was pe o med when iable
PSMA-nega i e me as ases we e suspec ed be o e o du -
ing PSMA-RLT due o a wo sening cou se o disease o
by hin in o he imaging me hods (CT, MRI). Conse-
quen ly, mo e misma ch pa ien s (62% in ou coho ) han
non-misma ch pa ien s we e included in ou s udy, esul -
ing in a p eselec ed coho o mCRPC pa ien s no ep e-
sen ing he no mal incidence o misma ch in candida es
e e ed o PSMA-RLT. In a p ospec i e phase-II ial o
PSMA-RLT, Ho man e al. (2019) epo ed an incidence
o 16.3% (n=7/43) o [
18
F]FDG/[
68
Ga]Ga-PSMA-11 mis-
ma ch a ime o sc eening o PSMA-RLT [20]. This was
he i s p ospec i e s udy which excluded pa ien s wi h
misma ch lesions. A p ospec i e se ing including la ge
non-biased pa ien coho s is necessa y o con i m ou
indings. Fu he limi a ions a e he missing blood samples
in 11/66 (16.7%) pa ien s p io o sc eening. Those pa-
ien s we e included in he sco ing sys em as “unknown”
ΔNSE, which e lec s a equen clinical si ua ion in
mCRPC pa ien s e e ed o e alua ion o PSMA-RLT.
As ano he poin o c i icism, no quan i a i e SUV h esh-
old de ining low PSMA exp ession on [
68
Ga]Ga-PSMA-11
PET/CT was se , which is also a global p oblem in li e a-
u e su e ing om de ined cu -o s o adequa e PSMA
exp ession. Fu he mo e, addi ional se um pa ame e s,
which we e ecen ly epo ed as p ognos ic ma ke s in he
mCRPC se ing o PSMA-RLT, such as he neu oendo-
c ine ma ke ch omog anin A o he lac a e dehyd ogen-
ase (LDH) [48–50], we e no a ailable o es ing in his
s udy. Se um ch omog anin A alues migh also be highe
in pa ien s wi h misma ch indings simila o NSE. LDH,
which is a ma ke wi h s ong p ognos ic alue o esponse
p edic ion o PSMA-RLT [48], migh also be a po en ial
ma ke o misma ch. Bo h pa ame e s should hus also be
es ed in u u e s udies. I hose s udies would indica e a
misma ch p edic i e alue, hese pa ame e s could be in-
cluded o he sco ing sys em o inc ease i s sensi i i y and
speci ici y. Las ly, his opa hological con i ma ion o a ans-
o ma ion p ocess in misma ch lesions owa ds a neu oen-
doc ine pheno ype was pe o med only in one case.
His opa hological examina ion o misma ch me as ases
should be included in u u e p ospec i e s udies.
Conclusions
We obse ed a signi ican ly highe absolu e se um
concen a ion and a highe ela i e inc ease o NSE in ad-
anced mCRPC pa ien s wi h [
18
F]FDG-a id and insu i-
cien PSMA exp essing me as ases ([
18
F]FDG/[
68
Ga]Ga-
PSMA-11 misma ch indings on PET/CT) in ou coho .
NSE migh be used as a po en ial labo a o y indica o o
[
18
F]FDG/[
68
Ga]Ga-PSMA-11 misma ch indings, i his
obse a ion is con i med in u u e, ideally p ospec i e,
s udies in la ge pa ien coho s.
Abb e ia ions
ADT: And ogen dep i a ion he apy; ALP: Alkaline phospha ase; CI: 95%
con idence in e al o he mean; CT: Compu e omog aphy;
FDG: Fluo odeoxyglucose; GGT: Gamma-glu amyl ans e ase; LDH: Lac a e
dehyd ogenase; mCRPC: Me as asized cas a ion- esis an p os a e ca cinoma;
NSE: Neu on-speci ic enolase; PET: Posi on emission omog aphy;
PSA: P os a e-speci ic an igen; PSMA: P os a e-speci ic memb ane an igen;
PSMA-RLT: PSMA- a ge ed adioligand he apy; ROC: Recei e ope a ing
cha ac e is ic
Acknowledgemen s
No applicable
Au ho s’con ibu ions
Flo ian Rosa , Kalle Ribba , Ma hias Sch eckenbe ge , Same Ezziddin, and
Fadi Kh eish con ibu ed o he design o he s udy. Ma in Ries and Fadi
Kh eish pe o med da a acquisi ion wi h suppo om Ma k Ba holomä and
S ephan Maus. Flo ian Rosa , Kalle Ribba , and Johannes Linxweile analyzed
he da a. Flo ian Rosa and Kalle Ribba d a ed his pape , which was e ised
by Ma k Ba holomä, Same Ezziddin, and Fadi Kh eish. All au ho s app o ed
he inal manusc ip .
Funding
No hing o disclose
A ailabili y o da a and ma e ials
The da ase s used and analyzed du ing he cu en s udy a e a ailable om
he co esponding au ho on easonable eques .
Compe ing in e es
The au ho s decla e ha hey ha e no con lic o in e es .
E hics app o al and consen o pa icipa e
All p ocedu es pe o med in he pa ien s desc ibed he ein we e in
acco dance wi h he e hical s anda ds o he Ins i u ional and/o Na ional
Resea ch E hics Commi ees and wi h he 1964 Helsinki Decla a ion and i s
la e amendmen s, o wi h compa able e hical s anda ds. This epo does
no include any animal s udies. W i en in o med consen was ob ained om
all s udy pa icipan s
Consen o publica ion
All pa ien s ha e gi en w i en consen o publica ion.
Au ho de ails
1
Depa men o Nuclea Medicine, Saa land Uni e si y—Medical Cen e ,
Ki be ge S . 100, Geb. 50, 66421 Hombu g, Ge many.
2
Depa men o
U ology, Saa land Uni e si y—Medical Cen e , Hombu g, Ge many.
3
Depa men o Nuclea Medicine, Uni e si y o Mainz, Mainz, Ge many.
Recei ed: 21 Feb ua y 2020 Accep ed: 4 May 2020
Re e ences
1. B ay F, Fe lay J, Soe joma a am I, Siegel RL, To e LA, Jemal A. Global cance
s a is ics 2018: GLOBOCAN es ima es o incidence and mo ali y wo ldwide
o 36 cance s in 185 coun ies. CA Cance J Clin. 2018;68:394–424.
2. Ki by M, Hi s C, C aw o d ED. Cha ac e ising he cas a ion- esis an p os a e
cance popula ion: a sys ema ic e iew: The Epidemiology o CRPC. In J
Clin P ac . 2011;65:1180–92.
3. Wa sonPA,A o aVK,Sawye sCL.Eme gingmechanismso esis ance o
and ogen ecep o inhibi o s in p os a e cance . Na Re Cance . 2015;15:701–11.
4. Co n o d P, Bellmun J, Bolla M, B ie s E, De San is M, G oss T, e al. EAU-
ESTRO-SIOG Guidelines on P os a e Cance . Pa II: ea men o elapsing,
me as a ic, and cas a ion- esis an p os a e cance . Eu U ol. 2017;71:630–
42.
5. Be hold DR, Pond GR, Soban F, de Wi R, Eisenbe ge M, Tannock IF.
Doce axel plus p ednisone o mi oxan one plus p ednisone o ad anced
p os a e cance : Upda ed Su i al in he TAX 327 S udy. J Clin Oncol. 2008;
26:242–5.
Rosa e al. EJNMMI Resea ch (2020) 10:52 Page 9 o 10