o iginal epo s
Clinical Ou comes and Pa ien -Ma ched
Molecula Composi ion o Relapsed
Medulloblas oma
Rahul Kuma , PhD
1,2
; Kyle S. Smi h, PhD
1
; Maximilian Deng, MD
3
; Col Te hune, BA
4
; Giles W. Robinson, MD
4
; B en A. O , MD, PhD
6
;
An hony P. Y. Liu, MBBS
1,4
; Tong Lin, PhD
5
; Ca he ine A. Billups, MS
5
; Mu ali Chin agumpala, MD
7
; Daniel C. Bowe s, MD
8
;
Timo hy E. Hassall, MD
9
; Jo dan R. Hans o d, MD
10
; Dong Anh Khuong-Quang, PhD
10
; John R. C aw o d, MD
11
; Anne E. Bendel, MD
12
;
S idha an Gu u angan, MD
13
; K is in Sch oede , MD, MPH
13
; E ic Bou e , MD
14
; U e Ba els, MD
14
; Michael J. Fishe , MD
15
;
Richa d Cohn, MD
16
; Sonia Pa ap, MD
17
; S ewa J. Kellie, MD
18
; Geo ey McCowage, MD
18
; A nold C. Paulino, MD
19
;
S e an Ru kowski, MD
20
; Gud un Fleischhack, MD
21
; Gi ish Dhall, MD
22
; Lau a J. Klesse, MD, PhD
8
; Sa ah Lea y, MD
23
;
Ja ad Naza ian, PhD
24
; Ma cel Kool, PhD
25
; Pie e Wesseling, MD
26
; Ma ina Ryzho a, MD
27
; Olga Zheludko a, MD
28
;
And ey V. Golano , MD
29
; Roge E. McLendon, MD
30
; Roge J. Packe , MD
31
; Ch is ophe Dunham, MD
32
; Julie e Hukin, MB
33
;
Ma yam Fouladi, MD
34
; Claudia C. Fa ia, MD
35
; Jose Pimen el, MD
36
; And ew W. Wal e , MD
37
; Nada Jabado, MD, PhD
38
;
Yoon-Jae Cho, MD
39
; Sebas ien Pe eaul , MD
40
; Sidney E. C oul, MD
41
; Michal Zapo ocky, MD, PhD
42
; Cyn hia Hawkins, MD, PhD
43
;
U i Tabo i, MD
43
; Michael D. Taylo , MD, PhD
43
; S e an M. Pfis e , MD
27
; Paul Klimo J , MD
44
; F ede ick A. Boop, MD
44
;
Da id W. Ellison, MD, PhD
6
; Thomas E. Me chan , DO, PhD
45
; A zu Ona -Thomas, PhD
5
; And ey Ko shuno , MD
46
;
Da id T. W. Jones, PhD
3,49
; Ama Gajja , MD
4
; Vijay Ramaswamy, MD, PhD
43,48
; and Paul A. No hco , PhD
1
abs ac
PURPOSE We sough o in es iga e clinical ou comes o elapsed medulloblas oma and o compa e molecula
ea u es be ween pa ien -ma ched diagnos ic and elapsed umo s.
METHODS Child en and in an s en olled on ei he SJMB03 (NCT00085202) o SJYC07 (NCT00602667) ials
who expe ienced medulloblas oma elapse we e analyzed o clinical ou comes, including ana omic and
empo al pa e ns o elapse and pos elapse su i al. A la gely independen , pai ed molecula coho was
analyzed by DNA me hyla ion a ay and nex -gene a ion sequencing.
RESULTS A o al o 72 o 329 (22%) SJMB03 and 52 o 79 (66%) SJYC07 pa ien s expe ienced elapse wi h
significan ep esen a ion o G oup 3 and wingless umo s. Al hough mos pa ien s exhibi ed some dis al disease
(79%), 38% o pa ien s wi h sonic hedgehog umo s expe ienced isola ed local elapse. Time o elapse and
pos elapse su i al a ied by molecula subg oup wi h longe la encies o pa ien s wi h G oup 4 umo s.
Pos elapse adia ion he apy among p e iously noni adia ed SJYC07 pa ien s was associa ed wi h long- e m
su i al. Rei adia ion was only empo izing o SJMB03 pa ien s. Among 127 pa ien s wi h pa ien -ma ched
umo pai s, 9 (7%) expe ienced subsequen nonmedulloblas oma CNS malignancies. Subg oup (96%) and
sub ype (80%) s abili ies we e la gely main ained among he emainde . Ra e subg oup di e gence was ob-
se ed om G oup 4 o G oup 3 umo s, which is coinciden wi h gene ic al e a ions in ol ing MYC,MYCN, and
FBXW7. Subg oup-specific pa e ns o al e a ion we e iden ified o d i e genes and ch omosome a ms.
CONCLUSION Clinical beha io o elapsed medulloblas oma mus be con ex ualized in e ms o up- on
he apies and molecula classifica ions. G oup 4 umo s exhibi slowe biological p og ession. U ili y o adi-
a ion a elapse is dependen on pa ien age and p io ea men s. Deg ee and pa e ns o molecula con-
se a ion a elapse a y by subg oup. Relapse issue enables e ifica ion o molecula a ge s and iden ifica ion
o occul seconda y malignancies.
J Clin Oncol 00. © 2021 by Ame ican Socie y o Clinical Oncology
C ea i e Commons A ibu ion Non-Comme cial No De i a i es 4.0 License
INTRODUCTION
Medulloblas oma elapse ep esen s a key de e minan
o cance - ela ed mo ali y in he pedia ic popula ion.
1,2
Mul imodal he apy ha inco po a es maximal su gical
esec ion wi h c aniospinal i adia ion (CSI; child en
olde han 3 yea s) and chemo he apy has d i en
5-yea o e all su i al a es o 80%-85% o a e age-
isk disease and 60%-70% o high- isk disease.
3,4
Ne e heless, ea men ailu e and elapse occu in
up o one- hi d o pa ien s and con e s abysmal p og-
nosis, wi h only app oxima ely 10% o pa ien s su i -
ing beyond 5 yea s pos elapse.
5-7
Relapsed disease
hus emains ex emely e ac o y o exis ing he a-
pies, whe eas a e su i o s o en expe ience se ious
oxici y and de as a ing neu ocogni i e sequalae.
8,9
Al hough no s anda d app oach o ea ing elapsed
ASSOCIATED
CONTENT
Appendix
Da a Supplemen
Au ho a filia ions
and suppo
in o ma ion (i
applicable) appea
a he end o his
a icle.
Accep ed on
No embe 16, 2020
and published a
ascopubs.o g/jou nal/
jco on Janua y 27,
2021: DOI h ps://doi.
o g/10.1200/JCO.20.
01359
1
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
medulloblas oma exis s, sal age he apies, including
con en ional and a ge ed agen s, ha e la gely ailed o
con e du able su i al benefi .
10,11
A deepe clinical and
biological unde s anding o medulloblas oma ecu ence is
needed o he design and in e p e a ion o nex -gene a ion
clinical ials o elapsed disease.
Al hough ex ensi e genomic cha ac e iza ion o medullo-
blas oma has iden ified ou biologically and clinically dis inc
subg oups (wingless [WNT], sonic hedgehog [SHH], G oup 3,
and G oup 4), mos s udies ha e elied on diagnos ic samples
ha ha e no been exposed o umo -di ec ed he apies.
12
Compa a i e molecula s udies o diagnos ic e sus elapsed
medulloblas oma ha e sugges ed di e gen clonal selec ion,
leading o eme gence o specific molecula al e a ions a
elapse in he con ex o subg oup conse a ion.
13-16
Howe e ,
he gene alizabili y o such findings is diminished by ela i ely
modes coho sizes and employmen o di e en molecula
assays. Addi ionally, exclusion o seconda y CNS malignan-
cies subsequen o medulloblas oma he apy, pa icula ly
high-g ade gliomas, equi es obus classifica ion me hods o
p e en inad e en c oss-en i y compa isons and may ep-
esen a key limi a ion o p e ious compa a i e s udies.
17,18
Mul iple ecen s udies ha e le e aged DNA me hyla ion
p ofiling o desc ibe addi ional in e umo al he e ogenei y
wi hin he co e medulloblas oma subg oups.
19-21
These
sub ypes, p ima ily defined among SHH (a,b,g, and d)
and G oups 3 and 4 (I-VIII) umo s, ha e dis inc i e gene ic
and clinical ea u es.
22,23
Fu he mo e, sub ype-defined
isk pa adigms a e an ac i e a ea o in es iga ion. Howe e ,
con ex ualiza ion o hese sub ypes in elapsed disease
and assessmen o conse a ion a elapse a e lacking.
In his s udy, subg oup-specific clinical beha io o elapsed
medulloblas oma is desc ibed o pa ien s en olled on wo,
mul i-ins i u ional, isk-adap ed clinical ials. A la gely in-
dependen , pai ed molecula coho composed o pa ien -
ma ched diagnos ic and elapse umo s is in es iga ed o
de e mine he molecula ea u es o medulloblas oma e-
lapse using DNA me hyla ion p ofiling and nex -gene a ion
sequencing.
METHODS
Pa ien s and Samples
S udy popula ions a e summa ized in he Da a Supplemen
(online only). Eligible pa ien s wi h elapsed medulloblas-
oma om wo mul i-ins i u ional, isk-adap ed clinical i-
als, SJMB03 (NCT00085202; Gajja e al
24
and SJYC07
(NCT00602667),
23
we e included (Appendix Table A1,
online only; Da a Supplemen ). Pa ien s wi h a clinical
diagnosis o nonmedulloblas oma subsequen malignancy
we e excluded. Gi en he di e ing eligibili y c i e ia, isk
s a ifica ions, and ea men p o ocols be ween he ials,
pa ien s om each ial we e analyzed sepa a ely o clinical
ou comes based on molecula ea u es ga ne ed om p ima y
umo specimens. A la gely independen , pai ed molecula
coho o 127 pa ien s wi h o malin-fixed pa a fin-embedded
(FFPE) o ozen issue specimens a ailable om bo h hei
his opa hologically diagnosed p ima y medulloblas oma and
elapse o subsequen umo s was also assembled (Appendix
Table A2, online only; Da a Supplemen ). Compa a i e mo-
lecula analyses we e pe o med among he pai ed molecula
coho using pa ien -ma ched p ima y and elapsed umo
specimens.
Tumo Molecula P ofiling
Tumo specimens we e analyzed using Infinium Me hyla-
ion EPIC o 450K BeadChip a ays (Illumina, San Diego,
CA) om ei he eshly ozen o FFPE issue (Da a Sup-
plemen ). Medulloblas oma subg oup and sub ype p e-
dic ions we e de e mined using DNA me hyla ion–based
classifica ion o CNS umo s (Molecula Neu opa hology,
Heidelbe g, Ge many, e sion 11b4) and ained andom
o es p edic ions.
18
Genome-wide DNA copy numbe al-
e a ions we e in e ed om DNA me hyla ion a ays using
he Conumee R package.
Nex -gene a ion (whole-exome o a ge ed gene panel)
sequencing was pe o med on umo samples wi h su fi-
cien ma e ial a ailable. Addi ional de ails ega ding bio-
in o ma ic p ocessing a e gi en in he Appendix Me hods
(Da a Supplemen ). Genomic da ase s included in his
CONTEXT
Key Objec i e
To de e mine he clinical ou comes and molecula ea u es o elapsed medulloblas oma.
Knowledge Gene a ed
Time o elapse and pos elapse su i al a e associa ed wi h subg oup wi h G oup 4 umo s exhibi ing slowe biological
p og ession. U ili y o adia ion he apy a elapse depends on age and p e ious he apies. Mos elapses exhibi
conco dan subg oup classifica ions, excep in he cases o occul seconda y malignancies o a e di e gence om g oup
4 o g oup 3. D i e gene and ch omosome a m al e a ion pa e ns a y acco ding o molecula subg oup.
Rele ance
Fu u e ials o elapsed medulloblas oma mus be con ex ualized by molecula subg oup and up- on he apies. Relapse
issue should be acqui ed and used o confi ma ion o diagnosis and e ifica ion o molecula a ge s.
2© 2021 by Ame ican Socie y o Clinical Oncology
Kuma e al
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
s udy can be eely explo ed using he online S Jude Cloud
pedia ic genomic da a esou ce.
25
S a is ical Analysis
Time- o-e en analyses we e pe o med using Kaplan-
Meie me hods wi h log- ank es s. Haza d a ios (HRs)
wi h associa ed 95% CIs and P alues we e compu ed
using Cox eg ession. Dis ibu ions o ca ego ical a iables
we e compa ed using Fishe ’s exac o chi-squa e es .
Mul iple es ing co ec ion was pe o med using alse dis-
co e y a e. S a is ical analyses we e pe o med using R
e sion 3.5.1 and a e u he de ailed in he Appendix
Me hods (Da a Supplemen ).
RESULTS
Incidence o Relapsed Disease
An o e iew o inclusion c i e ia, isk s a ifica ions, and
ea men p o ocols o SJMB03 and SJYC07 is shown in
Figu e 1A.A o alo 72o 329(22%)pa ien s om
SJMB03and52o 79 omSJYC07(66%) elapsed(Fig
1B). Appendix Table A1 summa izes he demog aphic
and clinical cha ac e is ics o elapsed pa ien s om
each ial acco ding o molecula subg oup. No ably,
WNT subg oup elapsed medulloblas omas we e absen
in bo h ials. Dis ibu ions o subg oups ac oss clinical
ial isk g oups o elapsed pa ien s a e shown in
Figu e 1C.
Subg oup dis ibu ion be ween elapsed and non elapsed
pa ien s was no uni o m in ei he SJMB03 (P5.0019,
chi-squa e es ) o SJYC07 (P5.00083, chi-squa e es ;
Fig 1B). Compa ed wi h G oup 3 umo s, G oup 4 umo s
had a lowe elapse a e in SJMB03 (odds a io [OR], 0.31;
95% CI, 0.16 o 0.61), whe eas SHH umo s had a lowe
elapse a e in SJYC07 (OR, 0.08; 95% CI, 0.01 o 0.33).
The dis ibu ion o no el sub ypes
19,20,22,23
be ween e-
lapsed and non elapsed pa ien s was no uni o m o
G oups 3 and 4 sub ypes in SJMB03 (P,.0001, chi-
squa e es ) o SHH sub ypes in SJYC07 (P5.0058, chi-
squa e es ; Da a Supplemen ). P opo ions o elapsed
pa ien s be ween SJMB03 and SJYC07 di e ed o no el
G oups 3 and 4 sub ypes IV (P,.0001, Fishe ’s exac es )
and VII (P,.0001, Fishe ’s exac es ).
Time o Relapse
The median ime o elapse was 1.64 yea s (in e qua ile
ange [IQR], 0.91-3.03) o SJMB03 and 0.72 yea (IQR,
0.47-1.00) o SJYC07 (Da a Supplemen ). Fo SJMB03,
ime o elapse a ied by molecula subg oup (P5.00087,
log- ank es ) wi h a median ime o elapse o pa ien s wi h
G oup 4 umo o 2.79 yea s (IQR, 1.99-3.64) compa ed
wi h 0.91 yea (IQR, 0.68-1.37) o G oup 3 (HR, 3.01;
95% CI, 1.67 o 5.41) and 1.23 yea s (IQR, 0.96-2.34) o
pa ien s wi h SHH (HR, 2.41; 95% CI, 1.20 o 4.81; Fig 1D).
Fo SJYC07, ime o elapse also a ied by molecula
subg oup (P5.039, log- ank es ), la gely d i en by
di e ence be ween G oup 4 and G oup 3 umo s (HR,
2.74; 95% CI, 1.14 o 6.60; Fig 1E). Clinical isk g oup was
a significan co a ia e o ime o elapse only among pa-
ien s wi h G oup 4 umo in SJMB03 (Da a Supplemen ).
Time o elapse acco ding o no el SHH and G oups 3 and
4 sub ypes is shown in Da a Supplemen .
Ana omic Pa e ns o Relapse
Relapse occu ed wi h a dis an componen in 57 (79%)
SJMB03 pa ien s (Fig 1F) and 41 (79%) SJYC07 pa ien s
(Fig 1G). O pa ien s wi h SHH umo s, 43% in SJMB03
and 35% in SJYC07 p esen ed wi h isola ed local elapse
compa ed wi h only 14% (P5.028, Fishe ’s exac es ) and
10% (P5.067, Fishe ’s exac es ) o non-SHH umo s,
espec i ely (Figs 1F and 1G). Dis an elapses we e sig-
nifican ly associa ed wi h sho e ime o elapse only o
G oup 4 umo s in SJMB03 (HR, 5.08; 95% CI, 1.14 o
22.6) and SHH umo s in SJYC07 (HR, 3.48; 95% CI, 1.07
o 11.3; Da a Supplemen ).
Su i al A e Relapse
A da a cu o , se en elapsed pa ien s (10%) om SJMB03
and 24 (46%) om SJYC07 we e ali e. The median
pos elapse su i al (PRS) was 1.14 yea s (IQR, 0.30-2.36)
o SJMB03 and 2.12 (IQR, 0.39-NA) o SJYC07 (Da a
Supplemen ). Clinical ial isk g oup was no significan ly
associa ed wi h PRS o SJMB03 (P5.43, log- ank es ) o
SJYC07 (P5.099, log- ank es ; Da a Supplemen ).
Fo SJMB03, PRS ime a ied by molecula subg oup (P5
.016, log- ank) wi h he median PRS ime o pa ien s wi h
G oup 4 umo s o 2.27 yea s (IQR, 1.19-3.89) compa ed
wi h 0.44 yea (IQR, 0.29-1.29) o pa ien s wi h G oup 3
umo s (HR, 2.39; 95% CI, 1.29 o 4.42) and 1.06 yea s
(IQR, 0.27-1.88) o pa ien s wi h SHH umo s (HR, 1.90;
95% CI, 0.94 o 3.83; Fig 2A). Fo SJYC07, PRS did no a y
significan ly by molecula subg oup (P5.88, log- ank es ;
Fig 2B). Time o elapse was significan ly associa ed wi h
PRS o SJMB03 (HR, 0.65; 95% CI, 0.52 o 0.82) bu no
o SJYC07 (HR, 0.53; 95% CI, 0.23 o 1.23). No ably, he
associa ion o ime o elapse and PRS o SJMB03
emained significan wi h subg oup and clinical ial isk
g oup as addi ional co a ia es (HR, 0.67; 95% CI, 0.50 o
0.90). PRS acco ding o no el SHH and G oups 3 and 4
sub ypes is shown in he Da a Supplemen .
Twen y-fi e (35%) elapsed SJMB03 pa ien s (one pa ien
missing da a) and 30 (58%) SJYC07 pa ien s ecei ed
adia ion a e elapse. Fo SJMB03 pa ien s, a ansien
PRS benefi (P5.032, log- ank es ; Fig 2C) was obse ed
wi h addi ional adia ion a e elapse (HR, 0.56; 95% CI,
0.33 o 0.96). Fo SJYC07 pa ien s, pos elapse CSI (me-
dian, 36.0 Gy) was significan ly associa ed wi h long- e m
su i al (P,.0001, log- ank es ; Fig 2D), pa icula ly o
pa ien s wi h SHH (HR, 0.04; 95% CI, 0.01 o 0.37) and
G oup 3 umo s (HR, 0.27; 95% CI, 0.08 o 0.85) (Da a
Supplemen ).
Jou nal o Clinical Oncology 3
Clinico-Molecula Expe ience Wi h Medulloblas oma Relapse
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
C
10
(100%)
4 (18%)
11 (50%)
6 (27%)
1 (5%)
6 (30%)
11 (55%)
2 (10%)
1 (5%)
0
5
10
15
20
Low
(n = 10)
In e media e
(n = 22)
High
(n = 20)
SJYC07 Risk G oup
No. o Relapsed Pa ien s
SJYC07 (n = 52 elapsed pa ien s)
9 (24%)
11 (30%)
10 (27%)
7 (19%)
5 (14%)
15 (43%)
14 (40%)
1 (3%)
0
10
20
30
A e age
(n = 37)
High
(n = 35)
SJMB03 Risk G oup
No. o Relapsed Pa ien s
SJMB03 (n = 72 elapsed pa ien s)
MB Subg oup SHH G oup 3 G oup 4 Unclassi ied
A
Pos e io ossa lesion
Su gical esec ion
Medulloblas oma
70CYJS30BMJS
M0
GTR
M+
<GTR
M0
GTR
DNMB <GTR
non-DNMB
3-5 yo*
M+M0
HDMTX
VCR Cyclo Cispla in
Vinblas ineBoos 55-58 Gy
VCR Cyclo Cispla in Cyclo
Ca bopla in
E oposide
Focal RT
54 Gy
Topo ecan
Cyclo
CSI (> 3 yo)
CSI
23.4 Gy
CSI
36-39 Gy
ypa eh omehc noi cudnInoi aida i lanipsoina C
Cyclo Topo ecan
E lo onib
Main enance chemo he apy
High-dose chemo he apy Consolida ion he apy
< 3 yo* ≥ 3-21 yo
3-5 yo wi h M0, GTR, and
DNMB we e eligible o
inclusion on in e media e- isk
a m o SJYC07
Mus ha e all ea u es o inclusion
Any lis ed ea u e o inclusion
Risk s a i ica ion c i e ia:
A e age isk High isk Low isk In e media e isk High isk
*
Relapse No Yes
B
SJMB03 (n = 329 pa ien s)
Non-WNT Chi-squa e P = .0019
e .
n.s.
0%
40% 17%
29%
33%
WNT SHH G3 G4 UC
0
50
100
No. o Pa ien s
*
SJYC07 (n = 79 pa ien s)
Chi-squa e P = .00083
e .
n.s.
48% 92%
80% 67%
0
10
20
30
40
SHH G3 G4 UC
No. o Pa ien s
*
D
012345
Yea s Since Diagnosis
Cumula i e Relapses
SJMB03 Time o Relapse by MB Subg oup (n = 74 pa ien s)
Log- ank P = .00087
1495200
2694220
24 21 18 12 5 1
885210Unclassi ied
G oup 4
G oup 3
SHH
Numbe a Risk
Median TTR:
SHH: 1.23 y s
G oup 3: 0.91 y s
G oup 4: 2.79 y s
Unclassi ied: 2.07 y s
P < .05 **
0.25
0.50
0.75
1.00
E
0.25
0.50
0.75
1.00
01234
Yea s Since Diagnosis
Cumula i e Relapses
SJYC07 Time o Relapse by MB Subg oup (n = 52 pa ien s)
Log- ank P = .039
20 5 1 1 1
22 3 0 0 0
84110
21000Unclassi ied
G oup 4
G oup 3
SHH
Numbe a Risk
P < .05
MedianTTR:
SHH: 0.74
G oup 3: 0.55
G oup 4: 1.08
Unclassi ied: 0.78
*
FIG 1. Relapse pa e ns o SJMB03 and SJYC07 pa ien s wi h medulloblas oma. (A) O e iew o ial designs and isk s a ifica ions. (B) Appa en
elapse a es by medulloblas oma subg oup pe ial. (C) Medulloblas oma subg oup dis ibu ions pe ial isk s a ifica ions. Time o elapse by
subg oup o SJMB03 pa ien s (D) and SJYC07 pa ien s (E). Ana omic pa e ns o elapse pe subg oup o SJMB03 (F) and SJYC07 (G) pa ien s. CSI,
c aniospinal i adia ion; DNMB, desmoplas ic nodula medulloblas oma; GTR, g oss o al esec ion; HDMTX, high-dose me ho exa e; MB, medul-
loblas oma; RT, adia ion he apy; SHH, sonic hedgehog; TTR, ime o elapse; UC, unclassified; VCR, inc is ine; WNT, wingless.
4© 2021 by Ame ican Socie y o Clinical Oncology
Kuma e al
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
Pa ien -Ma ched Molecula Landscapes
Gi en he a i y o issue a ailabili y om elapsed medul-
loblas oma and wi h only se en pa ien s om each ial wi h
elapse issue a ailable, a la ge mul i-ins i u ional, pai ed
molecula coho o 127 pa ien -ma ched diagnos ic and e-
lapsed umo s was assembled (Fig 3A; Appendix Table A2;
Da a Supplemen ). Su ficien issue o DNA me hyla ion
p ofiling was equi ed o inclusion, and whole-exome o
a ge ed panel sequencing da a we e a ailable o bo h
pa ien -ma ched samples in 50% (64 o 127) o he pai ed
molecula coho (Fig 3B). DNA me hyla ion classifica ion
iden ified nine elapse cases (7%) as nonmedulloblas oma
subsequen umo s wi hin he pai ed molecula coho
(Fig 3C; Da a Supplemen ).
18
Molecula Subg oups and Sub ypes
Among 118 pa ien s wi h molecula ly confi med e-
lapsed medulloblas oma, diagnos ic umo s we e clas-
sifiedbyme hyla ionasWNTin1(1%),SHHin48
(41%),G oup3in22(19%),andG oup4in46(39%).
Medulloblas oma subg oup was conse ed a elapse in
113 pa ien s (96%) wi h di e gence obse ed in a o al o
fi e pa ien s be ween G oups 3 and 4 umo s. No el
molecula sub ypes wi hin SHH and G oups 3 and 4
umo s we e conse ed a elapse in 80% o pa ien s
(Figs 3D and 3E).
19,20
To assess he deg ee o molecula conse a ion be ween
pa ien -ma ched diagnos ic and elapsed umo s, mu a-
ions and copy numbe al e a ions o cu a ed medullo-
blas oma d i e genes and ch omosomal a ms we e
analyzed (Fig 4A; Da a Supplemen ).
19
Subg oup desig-
na ion o pai ed pa ien -ma ched cases was based on
subg oup a diagnosis.
D i e Gene Al e a ion Pa e ns
The dis ibu ion and pa e n o d i e gene al e a ions by
subg oup a e shown in Figu e 4B. Subg oup was associa ed
wi h a significan di e ence in conse a ion pa e n o d i e
gene al e a ions, la gely d i en by biases in SHH, which we e
p edominan ly conse ed (P,.0001, analysis o a iance).
The median numbe o disco dan d i e gene al e a ions
be ween pa ien -ma ched umo s was one wi h no di e -
ence be ween subg oups (P5.10, K uskal-Wallis es ;
Da a Supplemen ). A o al o 29% o cases had comple ely
conse ed d i e gene al e a ions wi h no s a is ical di e -
ence be ween subg oups (P5.29, chi-squa e es ). Wi hin
subg oups, he numbe o disco dan copy numbe a ia-
ions o mu a ions was no significan ly associa ed wi h age
a diagnosis, ime o elapse, o ecu ence pa e n (Da a
Supplemen ).
The incidence o sha ed d i e gene al e a ions was 52% in
SHH, 36% in G oup 3, and 37% in G oup 4 umo s (Figs 4A
F
Ana omic Pa e ns o Relapse o SJMB03
Local only e sus Dis an componen Chi-squa e P = .058
6 (43%) 4 (15%) 3 (12%) 2 (25%)
Local only
0
5
10
15
20
SHH
n = 14
G oup 3
n = 26
G oup 4
n = 24
Unclassi ied
n = 8
Subg oup
No. o Relapsed
Pa ien s
7 (50%)
21 (81%) 20 (83%)
6 (75%)
Dis an
0
5
10
15
20
SHH
n = 14
G oup 3
n = 26
G oup 4
n =24
Unclassi ied
n = 8
Subg oup
No. o Relapsed
Pa ien s
1 (7%) 1 (4%) 1 (4%)
Local + Dis an
0
5
10
15
20
SHH
n = 14
G oup 3
n = 26
G oup 4
n = 24
Unclassi ied
n = 8
Subg oup
No. o Relapsed
Pa ien s
G
Ana omic Pa e ns o Relapse o SJYC07
Local only e sus Dis an componen Chi-squa e P = .094
7 (35%) 2 (9%) 1 (12%) 1 (50%)
Local only
0
5
10
15
SHH
n = 20
G oup 3
n = 22
G oup 4
n = 8
Unclassi ied
n = 2
Subg oup
No. o Relapsed
Pa ien s
8 (40%)
16 (73%)
7 (88%)
Dis an
0
5
10
15
SHH
n = 20
G oup 3
n = 22
G oup 4
n = 8
Unclassi ied
n = 2
Subg oup
No. o Relapsed
Pa ien s
5 (25%) 4 (18%) 1 (50%)
Local + Dis an
SHH
n = 20
G oup 3
n = 22
G oup 4
n = 8
Unclassi ied
n = 2
0
5
10
15
Subg oup
No. o Relapsed
Pa ien s
FIG 1. (Con inued).
Jou nal o Clinical Oncology 5
Clinico-Molecula Expe ience Wi h Medulloblas oma Relapse
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.

and 4B). No able sha ed d i e gene al e a ions included
hose a ec ing PTCH1 (10 o 12, 88% sha ed) and DDX3X
(6 o 7, 86%) in SHH umo s. The incidence o elapse-
specific d i e gene al e a ions was 29% in SHH, 57%
in G oup 3, and 40% in G oup 4 umo s. Addi ionally,
nume ous low incidence elapse-specific al e a ions, pa -
icula ly among ch oma in modifie s (eg, CREBBP and
SMARCA4) and DNA epai machine y (eg, BRCA2), we e
obse ed ac oss subg oups. P ima y-specific d i e gene
al e a ions, such as ocal CDK6 amplifica ion in one p ima y
SHH umo and one p ima y G oup 4 umo , we e a e and
comp ised he mino i y in obse ed pa e ns o conse a-
ion and di e gence.
D i e genes wi h inc eased odds o al e a ion in elapsed
pa ien s we e iden ified using p ima y umo molecula da a
a ailable o he clinical ial coho s (Da a Supplemen ).
Conse a ion pa e n o such al e a ions, including TP53
mu a ions and MYC and MYCN amplifica ions, was hen
B
+
++
+
++ + + +
+
+
+
+
+++ +++
++
+
++
0.25
0.50
0.75
1.00
02468
Yea s Since Relapse
Pos elapse Su i al
SJYC07 PRS by MB subg oup
(n = 52 pa ien s)
20 (0) 9 (3) 3 (6) 1 (8) 0 (9)
22 (0) 9 (3) 7 (4) 3 (7) 0 (10)
8 (0) 2 (3) 1 (4) 1 (4) 0 (5)
2 (0) 0 (0) 0 (0) 0 (0) 0 (0)Unclassi ied
G oup 4
G oup 3
SHH
Numbe a Risk (numbe censo ed)
Log- ank P = .88
Median PRS:
SHH:
G oup 3:
G oup 4:
Unclassi ied:
2.26
2.24
1.61
0.26
SHH:
G oup 3:
G oup 4:
Unclassi ied:
3-yea OS (95% CI):
38 (20% o 71%)
46% (29% o 74%)
44% (17% o 100%)
..
1-yea OS (95% CI):
59% (40% o 85%)
57% (39% o 83%)
67% (38% o 100%)
..
D
Yea s Since Relapse
++
+
++
++
+++
++ + + ++++ +
+
+
+
+
+
0.25
0.50
0.75
1.00
02468
Pos elapse Su i al
SJYC07 PRS by RT
(n = 52 pa ien s)
30 (0) 17 (5) 10 (10) 5 (14) 0 (19)
22 (0) 3 (4) 1 (4) 0 (5) 0 (5)
Numbe a Risk (numbe censo ed)
No
Yes
Pos elapse
RT
Pos elapse RT: +Yes +No
Log- ank P = 4e-06
3-yea OS (95% CI):
62% (46% o 84%)
8% (1% o 49%)
1-yea OS (95% CI):
79% (65% o 50%)
25% (11% o 55%)
Pos elapse RT:
No pos elapse RT:
C
0.25
0.50
0.75
1.00
0 2.5 5 7.5 10
Yea s Since Relapse
Pos elapse Su i al
SJMB03 PRS by RT
(n = 71 pa ien s)
25 (0) 12 (0) 5 (1) 2 (1) 1 (2)
46 (0) 4 (2) 2 (3) 2 (3) 2 (3)
No
Yes
Pos elapse
RT
Numbe a Risk (numbe censo ed)
Pos elapse RT: +Yes +No
Log- ank P = .032
3-yea OS (95% CI):
39% (24% o 64%)
12% (5% o 27%)
1-yea OS (95% CI):
68% (52% o 89%)
50% (37% o 67%)
Pos elapse RT:
No pos elapse RT:
+
+
++
++
A
0.25
0.50
0.75
1.00
0 2.5 5 7.5 10
Yea s Since Relapse
Pos elapse Su i al
SJMB03 PRS by MB Subg oup
(n = 72 pa ien s)
14 (0) 2 (0) 1 (1) 1 (1) 1 (1)
26 (0) 1 (2) 1 (2) 1 (2) 1 (2)
24 (0) 11 (1) 4 (1) 2 (1) 1 (2)
8 (0) 2 (0) 1 (0) 0 (0) 0 (0)Unclassi ied
G oup 4
G oup 3
SHH
Numbe a Risk (numbe censo ed)
Log- ank P = .016
14% (4% o 52%)
10% (3% o 34%)
38% (22% o 63%)
25% (8% o 83%)
3-yea OS (95% CI):
1-yea OS (95% CI):
57% (36% o 90%)
35% (20% o 59%)
80% (65% o 97%)
63% (37% o 100%)
SHH:
G oup 3:
G oup 4:
Unclassi ied:
+
+++
+
+
Median PRS:
SHH: 1.06 y s
G oup 3: 0.44 y s
G oup 4: 2.27 y s
Unclassi ied: 1.07 y s
P < .05 *
FIG 2. Pos elapse ou comes o SJMB03 and SJYC07 pa ien s wi h medulloblas oma. PRS by subg oup o SJMB03 (A) and SJYC07 (B) pa ien s.
Pos elapse su i al by eceip o adia ion he apy a e elapse o SJMB03 (C) and SJYC07 (D) pa ien s. MB, medulloblas oma; OS, o e all su i al;
PRS, pos elapse su i al; RT, adia ion he apy; SHH, sonic hedgehog.
6© 2021 by Ame ican Socie y o Clinical Oncology
Kuma e al
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
BA
Inclusion C i e ia:
Pa ien s wi h his opa hologically diagnosed MB
Su icien issue a ailable o bo h p ima y and subequen umo s
131 pa ien s
271 umo s
1 p ima y umo unclassi iable
3 elapse umo s in no mal issue
Me hyla ion
p o iling
127 pa ien s
262 umo s
Pai ed molecula coho :
127 pa ien s
254 umo s
3 pa ien s wi h mul iple p ima y umo s
5 pa ien s mul iple elapse umo s
Relapsed MBs:
118 pa ien s
236 umo s
Subsequen umo s:
9 pa ien s
18 umo s
NGS
n = 7 pa ien s
NGS
n = 57 pa ien s
Me hyla ion
n = 118 pa ien s
Me hyla ion
n = 9 pa ien s
Relapsed Coho
Subsequen Tumo Coho
n = 1 pa ien
n = 5 pa ien s
Panel
WES
WES/Panel
n = 7 pa ien s
n = 8 pa ien s
n = 42 pa ien s
n = 1 pa ien
Panel
WES
WES/Panel
C
MB, SHH
MB, WNT
MB, G4
MB, G3
MB, SHH
MB, WNT
MB, G4
MB, G3
HGG
EWS Subsequen
umo s
Relapsed
MB
Pai ed Molecula Coho (n = 127 Pa ien s)
No. o Pa ien s
125
100
75
50
25
0
P ima y Relapse
D
G oup 3 and 4 Sub ypes (n = 69 pa ien s)
VIII
VII
VI
V
IV
III
II
I
VIII
VII
VI
V
IV
III
II
I
0
20
40
60
P ima y Relapse
No. o Pa ien s
E
SHH Sub ypes (n = 48 pa ien s)
γ
δ
β
(iSHH-I) (iSHH-I)
α
γ
δ
β
α
0
10
20
30
40
50
P ima y Relapse
No. o Pa ien s
(iSHH-II)
(iSHH-II)
FIG 3. O e iew o pa ien -ma ched molecula coho . (A) Flowcha desc ibing assembly o pai ed molecula coho composed o pa ien s wi h uly
elapsed medulloblas omas and hose wi h o he subsequen CNS malignancies. (B) Me hyla ion and nex -gene a ion sequencing da a a ailabili y
o each a m o he pai ed molecula coho . (C) Subg oup and en i y classifica ion o pa ien -ma ched issue om diagnosis and elapse. No el
G oups 3 and 4 (D) and SHH (E) sub ype classifica ions o pa ien -ma ched issue om diagnosis and elapse. EWS, Ewing sa coma; HGG,
high-g ade glioma; iSHH, in an SHH; MB, medulloblas oma; NGS, nex -gene a ion sequencing; SHH, sonic hedgehog; WES, whole-exome
sequencing; WNT, wingless.
Jou nal o Clinical Oncology 7
Clinico-Molecula Expe ience Wi h Medulloblas oma Relapse
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
Al e a ions
Age
Sex
Subg oup
Ins ance
AMB, WNT
(n = 1 pai )
MB, SHH
(n = 23 pai s)
MB, G3
(n = 10 pai s)
MB, G4
(n = 23 pai s)
0
5
Da a
Al e a ion
Pa e n
Da a
PTCH1
MYCN
DDX3X
OTX2
GLI2
KMT2D
MYC
PTEN
TP53
APC
BRCA2
KMT2C
TERT
CDK6
FBXW7
GSE1
SMO
FAT1
KDM6A
ARID1A
CREBBP
FMR1
IDH1
SMARCA4
TBR1
O he D i e
1q
2p
2q
3q
6p
6q
7p
9p
9q
10q
17p
17q
21q
05025
25
050
0
25
NA NA NA NA
No. o CNV
Al e a ions
No. o D i e
Al e a ions
Age
Sex
Subg oup
Ins ance
Sub ype
NA
SNV
Indel
Male
Female
MB, WNT
MB, SHH
MB, G3
MB, G4
P ima y
Relapse
Ampli ica ion
Dele ion
In an
Child
Adul
Panel
WES
Sha ed
Disco dan
B
No. o Cases (pai ed p ima y
and elapsed umo s)
No. o Cases (pai ed p ima y
and elapsed umo s)
No. o Cases (pai ed p ima y
and elapsed umo s)
O he d i e
TBR1
SMARCA4
IDH1
FMR1
CREBBP
ARID1A
KDM6A
FAT1
SMO
GSE1
FBXW7
CDK6
TERT
KMT2C
BRCA2
APC
TP53
PTEN
MYC
KMT2D
GLI2
OTX2
DDX3X
MYCN
PTCH1
MB, G4
(15/23 pai s)
50
O he d i e
TBR1
SMARCA4
IDH1
FMR1
CREBBP
ARID1A
KDM6A
FAT1
SMO
GSE1
FBXW7
CDK6
TERT
KMT2C
BRCA2
APC
TP53
PTEN
MYC
KMT2D
GLI2
OTX2
DDX3X
MYCN
PTCH1
MB, G3
(7/10 pai s)
40
MB, SHH
(23/23 pai s)
O he D i e
TBR1
SMARCA4
IDH1
FMR1
CREBBP
ARID1A
KDM6A
FAT1
SMO
GSE1
FBXW7
CDK6
TERT
KMT2C
BRCA2
APC
TP53
PTEN
MYC
KMT2D
GLI2
OTX2
DDX3X
MYCN
PTCH1
110
snoi a e la 03 = nsnoi a e la 41 = nsnoi a e la 86 = n
P opo ion o
D i e Al e a ions
%92%91 52% 57%7% %04%32%63 37%
Subg oup Ve sus Al e a ion Pa e n P < .001
Pa e n o al e a ion: Relapse onlyP ima y onlySha ed
FIG 4. Molecula landscape o elapsed medulloblas oma. (A) Oncop in depic ing pa ien cha ac e is ics, d i e gene al e a ions, and ch omosomal
copy numbe a ia ion o pa ien -ma ched umo pai s wi h a ailable nex -gene a ion sequencing (n 557 pa ien s). (B) Compa men -specific
pa e ns o d i e gene al e a ions by molecula subg oup. (C) Genomic ack o ch omosome 2 depic ing al e a ion pa e ns obse ed o MYCN.(D)
Compa men -specific pa e ns o ch omosome a m copy numbe a ia ion by molecula subg oup o pa ien -ma ched umo pai s (n 5107
pa ien s). Genomic acks depic ing al e a ion pa e ns o ch omosome 17 (E) and ch omosome 10 (F). CNV, copy numbe a ia ion; G3, G oup 3;
G4, G oup 4; MB, medulloblas oma; SHH, sonic hedgehog; SNV, single nucleo ide a ian ; WES, whole-exome sequencing; WNT, wingless.
8© 2021 by Ame ican Socie y o Clinical Oncology
Kuma e al
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
C
MYCN
MYCN
MYCN
MYCN
MYCN
MYCN
−1.2
−0.8
−0.4
0.0
0.4
0.8
1.2
MYCN
MYCN
MYCN
MYCN
MYCN
MYCN
Ins ance:
Pa ien :
Subg oup:
MB-7 MB-28 MB-121 MB-9 MB-70 MB-78
Log2 Ra io
Al e a ion Pa e n:
MYCN
Relapse OnlyP ima y OnlySha ed
ch 2 ch 2 ch 2 ch 2 ch 2 ch 2ch 2 ch 2 ch 2 ch 2 ch 2 ch 2
E
9
10
11
Ins ance
Subg oup
Pa ien
Ch omosome
F
16
17
18
Ins ance
Subg oup
Pa ien
Ch omosome
MB-128
MB-124
MB-78
MB-17
MB-7
MB-3
MB-37
MB-32
D
Pa e n o al e a ion: Relapse onlyP ima y onlySha ed
No. o Cases
(pai ed p ima y and
elapse umo s)
No. o Cases
(pai ed p ima y and
elapse umo s)
No. o Cases
(pai ed p ima y and
elapse umo s)
MB, G3
(21/21 pai s)
MB, G4
(45/46 pai s)
P opo ion o Al e ed
Ch omosomal A ms
MB, SHH
(33/40 pai s)
n = 294 al e a ionsn = 201 al e a ionsn = 222 al e a ions
10%
Subg oup Ve sus Al e a ion Pa e n P = .0047
Ch omosomal A ms
1p
1q
2p
2q
3p
3q
4p
4q
5p
5q
6p
6q
7p
7q
8p
8q
9p
9q
10p
10q
11p
11q
12p
12q
13q
14q
15q
16p
16q
17p
17q
18p
18q
19p
19q
20p
20q
21q
22q
0
10
20
30
**
***
*
*
Ch omosomal A ms
1p
1q
2p
2q
3p
3q
4p
4q
5p
5q
6p
6q
7p
7q
8p
8q
9p
9q
10p
10q
11p
11q
12p
12q
13q
14q
15q
16p
16q
17p
17q
18p
18q
19p
19q
20p
20q
21q
22q
0
5
10
15 *
*
*
*
*
*
Ch omosomal A ms
1p
1q
2p
2q
3p
3q
4p
4q
5p
5q
6p
6q
7p
7q
8p
8q
9p
9q
10p
10q
11p
11q
12p
12q
13q
14q
15q
16p
16q
17p
17q
18p
18q
19p
19q
20p
20q
21q
22q
0
5
10
15 *
61% 29%17% 42% %82%41%14 58%
FIG 4. (Con inued).
Jou nal o Clinical Oncology 9
Clinico-Molecula Expe ience Wi h Medulloblas oma Relapse
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
AUTHORS’DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
Clinical Ou comes and Pa ien -Ma ched Molecula Composi ion o Relapsed Medulloblas oma
The ollowing ep esen s disclosu e in o ma ion p o ided by au ho s o his manusc ip . All ela ionships a e conside ed compensa ed unless o he wise no ed.
Rela ionships a e sel -held unless no ed. I 5Immedia e Family Membe , Ins 5My Ins i u ion. Rela ionships may no ela e o he subjec ma e o his manusc ip .
Fo mo e in o ma ion abou ASCO’s conflic o in e es policy, please e e o www.asco.o g/ wc o ascopubs.o g/jco/au ho s/au ho -cen e .
Open Paymen s is a public da abase con aining in o ma ion epo ed by companies abou paymen s made o US-licensed physicians (Open Paymen s).
Giles W. Robinson
Consul ing o Ad iso y Role: Lilly, Roche/Genen ech
Resea ch Funding: No a is, Genen ech/Roche, No a is
Jo dan R. Hans o d
Consul ing o Ad iso y Role: Baye
John R. C aw o d
Hono a ia: Illumina
Speake s’Bu eau: As aZeneca
E ic Bou e
Consul ing o Ad iso y Role: No a is
Resea ch Funding: Roche, B is ol-Mye s Squibb
Michael J. Fishe
Hono a ia: As aZeneca
Resea ch Funding: As aZeneca, A ay BioPha ma, Exelixis
T a el, Accommoda ions, Expenses: As aZeneca, Sp ingWo ks
Sonia Pa ap
Consul ing o Ad iso y Role: Baye , Ge son Leh man G oup
Resea ch Funding: Abb ie
Geo ey McCowage
Consul ing o Ad iso y Role: Biogen
Resea ch Funding: No a is, Me ck
T a el, Accommoda ions, Expenses: BIOGEN
A nold C. Paulino
Employmen : MD Ande son Cance Cen e
Pa en s, Royal ies, O he In ellec ual P ope y: Royal y om Else ie Inc o
book on PET/CT in Radio he apy T ea men Planning
T a el, Accommoda ions, Expenses: Uni e si y o Sou he n Cali o nia
S e an Ru kowski
Consul ing o Ad iso y Role: B is ol-Mye s Squibb GmbH & Co KGaA,
Ge many, Celgene, Roche Pha ma AG, G enzach-Wyhlen
Resea ch Funding: Riemse Pha ma GmbH, G ei swald, Ge many
Lau a J. Klesse
Consul ing o Ad iso y Role: As aZeneca
T a el, Accommoda ions, Expenses: Sp ingwo ks, As aZeneca
Sa ah Lea y
Resea ch Funding: Blaze Bioscience
Roge E. McLendon
S ock and O he Owne ship In e es s: Gilead Sciences, Nan Kwes
Expe Tes imony: Johnson & Johnson
Roge J. Packe
Hono a ia: No a is
Consul ing o Ad iso y Role: No a is, As aZeneca
Julie e Hukin
S ock and O he Owne ship In e es s: Abb ie
Ma yam Fouladi
Resea ch Funding: PTC he apeu ics, Baye Sche ing Pha ma
Sebas ien Pe eaul
Leade ship: Baye
S ock and O he Owne ship In e es s: No ocu e
Hono a ia: Baye
Consul ing o Ad iso y Role: Baye
Speake s’Bu eau: Baye
Expe Tes imony: Baye
Michal Zapo ocky
Consul ing o Ad iso y Role: Baye
Pa en s, Royal ies, O he In ellec ual P ope y: IP o low g ade glioma and
sa coma usion panels as well as medulloblas oma subg ouping panel
S e an M. Pfis e
Resea ch Funding: Lilly, Baye , Roche, Pha maMa , Pfize
Pa en s, Royal ies, O he In ellec ual P ope y: Pa en on u ilizing DNA
me hyla ion p ofiling o umo classifica ion
F ede ick A. Boop
Employmen : Semmes Mu phey Clinic
Da id W. Ellison
Pa en s, Royal ies, O he In ellec ual P ope y: Sole in en o o US Pa en No.
9,005,907 issued Ap il 14, 2015 “Me hods and Composi ions o Typing
Molecula Subg oups o Medulloblas oma”, 62627291/S88435 1190US.P1
Feb ua y 2018 “Epigene ic His one Regula ion Media ed by CXo 67”published
Augus 2019 as WO 2019/155387
Thomas E. Me chan
T a el, Accommoda ions, Expenses: Philips Heal hca e
A zu Ona -Thomas
Consul ing o Ad iso y Role: Roche
Resea ch Funding: No a is, Apexigen, Pfize , Celgene, No a is, Me ck,
No ocu e
T a el, Accommoda ions, Expenses: Roche
Da id T. W. Jones
Pa en s, Royal ies, O he In ellec ual P ope y: Pa en WO 2013075237 A1,
i led “Mu a ions o his one p o eins associa ed wi h p oli e a i e diso de s”
Ama Gajja
Consul ing o Ad iso y Role: Roche/Genen ech
Resea ch Funding: Genen ech, Kazia Pha maceu ical
Vijay Ramaswamy
Hono a ia: As aZeneca
No o he po en ial conflic s o in e es we e epo ed.
© 2021 by Ame ican Socie y o Clinical Oncology
Kuma e al
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.

APPENDIX
TABLE A1. Demog aphics and Diagnos ic Clinical Cha ac e is ics o Relapsed Pa ien s
MB Subg oup
SJMB03 (n 572) SJYC07 (n 552)
SHH
n514
G oup 3
n526
G oup 4
n524
Unclassified
n58
SHH
n520
G oup 3
n522
G oup 4
n58
Unclassified
n52
Sex
Female 5 (36%) 8 (31%) 4 (17%) 1 (12%) 10 (50%) 12 (55%) 3 (38%) 1 (50%)
Male 9 (64%) 18 (69%) 20 (83%) 7 (88%) 10 (50%) 10 (45%) 5 (62%) 1 (50%)
Age (IQR) 8.9 (8.4-10.7) 6.4 (4.0-8.5) 7.8 (5.7-10.6) 9.1 (6.3-11.9) 2.0 (1.3-2.5) 2.6 (2.0-2.9) 3.71 (3.0-4.0) 1.1 (1.0-1.2)
M s age
M0 9 (64%) 12 (46%) 10 (42%) 7 (88%) 14 (70%) 11 (50%) 6 (75%) 1 (50%)
M1 0 (0%) 2 (8%) 0 (0%) 0 (0%) 0 (0%) 1 (5%) 0 (0%) 0 (0%)
M2 3 (21%) 0 (0%) 6 (25%) 0 (0%) 1 (5%) 1 (5%) 1 (12%) 0 (0%)
M3 2 (14%) 12 (46%) 8 (33%) 1 (12%) 5 (25%) 9 (41%) 1 (12%) 1 (50%)
His ology
Classic 4 (29%) 15 (58%) 21 (88%) 7 (88%) 2 (10%) 17 (77%) 7 (88%) 2 (100%)
DN 2 (14%) 0 (0%) 0 (0%) 0 (0%) 15 (75%) 0 (0%) 0 (0%) 0 (0%)
LCA 7 (50%) 11 (42%) 3 (12%) 1 (12%) 0 (0%) 5 (23%) 1 (12%) 0 (0%)
MBEN ——— —3 (15%) 0 (0%) 0 (0%) 0 (0%)
Medullomyoblas oma 1 (7%) 0 (0%) 0 (0%) 0 (0%) —— — —
Resec ion
GTR 10 (71%) 19 (73%) 16 (67%) 7 (88%) 15 (75%) 15 (68%) 7 (88%) 2 (100%)
NTR 2 (14%) 7 (27%) 7 (29%) 1 (12%) 0 (0%) 4 (18%) 0 (0%) 0 (0%)
STR 2 (14%) 0 (0%) 1 (4%) 0 (0%) 5 (25%) 3 (14%) 1 (12%) 0 (0%)
Pos elapse he apy
Radia ion
Yes 4 (29%) 4 (15%) 14 (58%) 3 (38%) 9 (45%) 15 (68%) 6 (75%) 0 (0%)
No 10 (71%) 21 (81%) 10 (42%) 5 (62%) 11 (55%) 7 (32%) 2 (25%) 2 (100%)
Unknown 0 (0%) 1 (4%) 0 (12%) 0 (12%) —— — —
Chemo he apy
Yes 11 (79%) 17 (65%) 21 (88%) 5 (62%) 16 (80%) 12 (55%) 3 (38%) 0 (0%)
No 1 (7%) 3 (12%) 1 (4%) 1 (12%) 4 (20%) 10 (45%) 5 (62%) 2 (100%)
Unknown 2 (14%) 6 (23%) 2 (8%) 2 (25%) —— — —
Abb e ia ions: DN, desmoplas ic nodula ; GTR, g oss o al esec ion; IQR, in e qua ile ange; LCA, la ge cell anaplas ic; MBEN, medulloblas oma wi h
ex ensi e nodula i y; NTR, nea o al esec ion; SHH, sonic hedgehog; STR, sub o al esec ion.
Jou nal o Clinical Oncology
Clinico-Molecula Expe ience Wi h Medulloblas oma Relapse
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.
TABLE A2. Demog aphic and Clinical Pa ame e s o Pa ien -Ma ched Molecula Coho
No. o Cases
Subg oup (Diagnosis)
Da a A ailable
N5127
MB, WNT
n52
MB, SHH
n552
MB, G3
n523
MB, G4
n550
Age, mean (SD) 25.0 (11.3) 15.9 (12.2) 5.87 (4.21) 8.75 (3.40) 124
Sex 127
Female 1 (50.0%) 21 (40.4%) 11 (47.8%) 19 (38.0%)
Male 1 (50.0%) 31 (59.6%) 12 (52.2%) 31 (62.0%)
His ology 127
Classic 1 (50.0%) 2 (3.85%) 3 (13.0%) 13 (26.0%)
DNMB 0 (0.00%) 16 (30.8%) 0 (0.00%) 1 (2.00%)
LCA 0 (0.00%) 8 (15.4%) 2 (8.70%) 0 (0.00%)
NOS 1 (50.0%) 26 (50.0%) 18 (78.3%) 36 (72.0%)
M S age 86
M11 (100%) 9 (26.5%) 8 (57.1%) 8 (21.6%)
M0 0 (0.00%) 25 (73.5%) 6 (42.9%) 29 (78.4%)
Resec ion 79
GTR o NTR 1 (100%) 21 (65.6%) 7 (50.0%) 26 (81.2%)
STR 0 (0.00%) 11 (34.4%) 7 (50.0%) 6 (18.8%)
Relapse pa e n 121
Dis an 1 (50.0%) 14 (28.6%) 17 (73.9%) 31 (66.0%)
Local 1 (50.0%) 35 (71.4%) 6 (26.1%) 16 (34.0%)
Subg oup conse a ion 127
Conse ed 1 (50.0%) 48 (92.3%) 21 (91.3%) 43 (86.0%)
Di e gen 0 (0.00%) 0 (0.00%) 1 (4.35%) 4 (8.00%)
Non-MB subsequen 1 (50.0%) 4 (7.69%) 1 (4.35%) 3 (6.00%)
The apy (diagnosis)
Chemo he apy 1 (50.0%) 31 (70.5%) 13 (86.7%) 27 (90.0%) 91
Radio he apy 2 (100%) 31 (70.5%) 7 (46.7%) 27 (90.0%) 91
The apy ( elapse)
Chemo he apy 0 (0.00%) 23 (82.1%) 12 (80.0%) 23 (92.0%) 69
Radio he apy 1 (100%) 14 (50.0%) 7 (46.7%) 10 (40.0%) 69
NGS da a 2 (100%) 26 (50.0%) 11 (47.8%) 25 (50.0%) 64
EFS (SD) 2.79 (1.12) 2.96 (3.49) 1.67 (1.04) 3.42 (2.54) 121
OS (SD) 9.88 (7.25) 5.29 (4.38) 4.08 (3.04) 5.94 (4.30) 97
Abb e ia ions: DNMB, desmoplas ic nodula medulloblas oma; EFS, e en - ee su i al; G3, G oup 3; G4, G oup 4; GTR, g oss o al esec ion; LCA, la ge
cell anaplas ic; MB, medulloblas oma; NGS, nex -gene a ion sequencing; NOS, no o he wise specified; NTR, nea o al esec ion; OS, o e all su i al; SHH,
sonic hedgehog; STR, sub o al esec ion; WNT, wingless.
© 2021 by Ame ican Socie y o Clinical Oncology
Kuma e al
Downloaded om ascopubs.o g by 79.168.162.59 on Feb ua y 5, 2021 om 079.168.162.059
Copy igh © 2021 Ame ican Socie y o Clinical Oncology. All igh s ese ed.