establishing a preclinical multidisciplinary board for brain tumors · minimum time–concentration...

Cancer Therapy: Preclinical

Establishing a Preclinical Multidisciplinary Boardfor Brain TumorsBirgit V. Nimmervoll1, Nidal Boulos2, Brandon Bianski3, Jason Dapper4,Michael DeCuypere5, Anang Shelat6, Sabrina Terranova1, Hope E. Terhune4,Amar Gajjar7, Yogesh T. Patel8, Burgess B. Freeman9, Arzu Onar-Thomas10,Clinton F. Stewart11, Martine F. Roussel12, R. Kipling Guy6,13, Thomas E. Merchant3,Christopher Calabrese14, Karen D.Wright15, and Richard J. Gilbertson1

Abstract

Purpose: Curing all children with brain tumors will require anunderstanding of how each subtype responds to conventionaltreatments and how best to combine existing and novel therapies.It is extremely challenging to acquire this knowledge in the clinicalone, especially among patients with rare tumors. Therefore, wedeveloped a preclinical brain tumor platform to test combina-tions of conventional and novel therapies in amanner that closelyrecapitulates clinic trials.

ExperimentalDesign:Amultidisciplinary teamwasestablishedto design and conduct neurosurgical, fractionated radiotherapyand chemotherapy studies, alone or in combination, in accuratemousemodelsof supratentorial ependymoma(SEP) subtypes andchoroid plexus carcinoma (CPC). Extensive drug repurposingscreens, pharmacokinetic, pharmacodynamic, and efficacy studieswere used to triage active compounds for combination preclinicaltrials with "standard-of-care" surgery and radiotherapy.

Results:Mouse models displayed distinct patterns of responseto surgery, irradiation, and chemotherapy that varied with tumorsubtype. Repurposing screens identified 3-hour infusions ofgemcitabine as a relatively nontoxic and efficacious treatment ofSEP and CPC. Combination neurosurgery, fractionated irradia-tion, and gemcitabine proved significantly more effective thansurgery and irradiation alone, curing one half of all animals withaggressive forms of SEP.

Conclusions: We report a comprehensive preclinical trialplatform to assess the therapeutic activity of conventionaland novel treatments among rare brain tumor subtypes. Italso enables the development of complex, combinationtreatment regimens that should deliver optimal trial designsfor clinical testing. Postirradiation gemcitabine infusionshould be tested as new treatments of SEP and CPC. ClinCancer Res; 24(7); 1654–66. �2018 AACR.

IntroductionDespite decades of research, the treatment of brain tumors

has remained largely unchanged. These cancers are treated withan aggressive combination of neurosurgery, radiotherapy, andchemotherapy that frequently fails to cure but inflicts signifi-cant side effects (1–4). This limited progress has occurreddespite an active clinical trials effort: more than 2,580 braintumor trials are currently registered with clinicaltrials.gov, butonly six drugs are approved for treatment of brain tumors, ofwhich only two—Everolimus, an inhibitor of the mTOR (5),and Bevacizumab, an inhibitor of VEGFA (1)—are molecular-targeted treatments.

So why have we failed to identify effective new brain tumortherapies? One possibility is that the preclinical systems used toselect drugs for clinical trial do not predict therapeutic activity inpatients (6). This explanation is plausiblewhenone considers thatmost preclinical studies are conducted in mice harboring subcu-taneous brain tumor xenografts that cannot recapitulate accurate-ly the pharmacology or biology of brain tumor treatment. Fur-thermore, although brain tumor patients receive complex, multi-modality therapy, mice in preclinical studies usually receive drugsasmonotherapies. Such studies are unlikely to predict the survivalbenefit of a new treatment above that afforded by standard of care.Prioritizing treatments with the greatest potential for clinicalefficacy is especially important for rare tumors that have limitedpatient populations available for clinical trial.

1Cancer Research UK Cambridge Institute and Department of Oncology, Uni-versity of Cambridge, Cambridge, England, United Kingdom. 2Department ofHematology, St. Jude Children's Research Hospital, Memphis, Tennessee.3Department of Radiological Sciences, St. Jude Children's Research Hospital,Memphis, Tennessee. 4Department of Developmental Neurobiology, St. JudeChildren's Research Hospital, Memphis, Tennessee. 5Department of Surgery, St.Jude Children's Research Hospital, Memphis, Tennessee. 6Department of Chem-ical Biology and Therapeutics, St. Jude Children's Research Hospital, Memphis,Tennessee. 7Department of Oncology, St. Jude Children's Research Hospital,Memphis, Tennessee. 8Department of Pharmaceutical Sciences, St. Jude Chil-dren's Research Hospital, Memphis, Tennessee. 9Preclinical PharmacokineticsCore, St. Jude Children's Research Hospital, Memphis, Tennessee. 10Departmentof Biostatistics, St. Jude Children's Research Hospital, Memphis, Tennessee.11Department of Pharmaceutical Sciences, St. Jude Children's Research Hospital,Memphis, Tennessee. 12Department of Tumor Cell Biology, St. Jude Children'sResearch Hospital, Memphis, Tennessee. 13University of Kentucky College ofPharmacy, Lexington, Kentucky. 14Preclinical Imaging Core, St. Jude Children'sResearch Hospital, Memphis, Tennessee. 15Boston Children's Hospital and Har-vard Medical School, Boston, Massachusetts.

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

B.V. Nimmervoll and N. Boulos contributed equally to this article.

Corresponding Authors: Richard J. Gilbertson, Cambridge University, Cam-bridge CB2 0RE, England, United Kingdom. Phone: 01223769590; E-mail:[email protected]; and Karen D. Wright, Boston Children'sHospital and Harvard Medical School, 450 Brookline Avenue, Boston, MA02215. E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-17-2168

�2018 American Association for Cancer Research.

ClinicalCancerResearch

Clin Cancer Res; 24(7) April 1, 20181654

on February 8, 2020. © 2018 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst January 4, 2018; DOI: 10.1158/1078-0432.CCR-17-2168

http://crossmark.crossref.org/dialog/?doi=10.1158/1078-0432.CCR-17-2168&domain=pdf&date_stamp=2018-3-5

http://clincancerres.aacrjournals.org/

Modifying long-established treatment regimens that haveevolved empirically over many years is also challenging. Forexample, the treatment of supratentorial ependymoma (SEP) andchoroid plexus carcinoma (CPC)—two rare pediatric braintumors—has evolved over decades to include maximum surgicalresection and postoperative cranial irradiation (7–14). Thesetreatments are effective, but evidence suggests that this efficacyvaries with tumor subtype. For example, although most SEPscontaining the C11ORF95-RELA translocation (hereon, SEP-CR[þ]) resist combination surgery and irradiation, the majorityof SEP-CR(–) tumors are cured with this therapy (11, 15, 16).Despite these differences in treatment sensitivity, ongoing clinicaltrials are testing whether classic histology SEPs, regardless ofmolecular tumor subtype, can be cured with total tumor resectionalone (NCT01096368). Thus, there is a pressing need to deter-mine if SEP-CR(þ) resists surgery, radiotherapy, or both. Suchknowledge is also important if we are to combine conventionaland novel therapies to better treat these tumors; however, thisknowledge is unlikely to be acquired solely in the clinic, especiallygiven the rarity of disease variants. Therefore, to better understandthe response of SEP and CPC subtypes to surgery and radiother-apy, and to design clinical trials that integrate conventional andnew treatments, we established a preclinical multidisciplinaryteam (pMDT) with the capacity to conduct randomized, multi-modality trials in mice harboring accurate models of SEP or CPC.

Materials and MethodsTumor cells and implants

The isolation, culture, and orthotopic implantation of allmouse and human tumor cells was described previously(15–18). Thenomenclature, species, tumor type, driver oncogene,and implanted cell number of each xenograft and allograft areprovided in Supplementary Table S1. All cells were maintained asin vivo grafts and confirmed by ELISA as mycoplasma negativeprior to and following in vitro studies. All animal studies wereapproved by the Animal Care and Usage Committees at St JudeChildren's ResearchHospital and theUniversity of Cambridge. As

discussed in detail in the Supplementary Methods, host mice forall allografts and xenografts were CD-1 nude mice (strain code:086; Charles River). All preclinical surgery, radiotherapy, andchemotherapy studies were performed among randomizedcohorts of mice harboring tumors with �1e107 photons/secbioluminescence (16). Tumor progression and treatmentresponse were assessed clinically and by weekly bioluminescence(16). Mice displaying signs of excessive clinical morbidity (�20%weight loss and/or neurological impairment) were euthanized.

Preclinical neurosurgery, radiotherapy, and chemotherapyFollowing baseline bioluminescence imaging, mice were

appropriately anaesthetized, a craniotomy fashioned over thesite of maximum bioluminescence, and tumors resected usinga small suction tip. Postoperative hemostasis was achievedwith thrombin-soaked gel foam prior to skin closure. Micewere reimaged in the immediate postoperative period, mon-itored on heating pads, and treated for 3 days with ibuprofen-supplemented drinking water, dexamethasone (0.6 mg/kg/6 hours), and mannitol (100 mg/kg/6 hours). Note that 54Gy of radiotherapy was delivered to appropriately anesthetizedmice as 2 Gy/day fractions via an orthovoltage irradiator orimage-guided rodent irradiator (SARRP, Xstrahl). Drugs weredelivered via tail vein bolus injections or using Alzetpumps (2001D, mean pumping rate � 8.0 mL/h; loaded with150 mg/mL gemcitabine solution prepared in 50:50 PEG300:propylene glycol; Supplementary Methods; SupplementaryTable S2). For combination surgery–radiotherapy or sur-gery–radiotherapy–chemotherapy studies, mice were rested for72 hours in between therapeutic modalities.

Pharmacokinetic and pharmacodynamics studiesPharmacokinetic studies are described in detail in the Supple-

mentary Methods. Briefly, blood samples were collected fromeuthanized mice via cardiac stick into tubes containing tetrahy-drouridine (THU, final concentration 150 mg/mL). Plasma wasseparated and samples were stored at �80�C until analysis.Intracranial microdialysis studies were performed as describedpreviously (19). A guide cannula (MD-2255, BASi) and allo-grafted tumor cells were implanted stereotactically in the cortexof immunocompromised mice. Once tumors formed, a precali-brated microdialysis probe (MD-2211, BASi; 38 KDa MWCOmembrane) was implanted through the microdialysis guide can-nula and perfused with artificial cerebrospinal fluid (0.5 mL/min).Mice were dosed with gemcitabine and plasma samples collectedvia retro-orbital bleeds. Drug levels were measured using a val-idated high-performance liquid chromatography–mass spec-trometry method. Tumor cell proliferation and apoptosis wereassessed by immunohistochemical quantification of Ki67 andCaspase 3, respectively (Supplementary Methods).

In vitro drug testingHigh-throughput screens were performed by seeding tumor

cells in 384-well plates as described in the SupplementaryMethods and previously (19). Each plate included dilution seriesof test compounds (8.3 mmol/L to 0.5 nmol/L), DMSO-onlynegative controls and cycloheximide or bortezomib single point(0.5 mmol/L) and dose-response (0.5 mmol/L to 0.01 nmol/L)–positive controls. Cell number was determined in each well usingthe Cell Titer Glo reagent (Promega). All assays were conducted intriplicate.Wash-out studies were similarly performed to assess the

Translational Relevance

Existing drug development pipelines have failed to bringnew treatments to children with brain tumors. A lack offaithful preclinical models has prevented the discovery andprioritization of potential new therapies, and the rarity of thesediseases presents an insurmountable hurdle for drug devel-opment through clinical trial alone. Therefore, we establisheda preclinical multidisciplinary tumor board comprising biol-ogists, statisticians, pharmacologists, and clinicians to conductpreclinical studies thatmimic the clinic.Mousemodels includ-ed those of specific ependymoma and choroid plexus carci-noma subtypes—two rare pediatric brain tumors. In contrastwith previous brain tumor preclinical platforms, our approachenables the testing of potential new treatments of very raretumors, in the context of "standard-of-care" neurosurgery andfractionated irradiation. This approach enables assessment ofthe potential therapeutic "value added" of candidate treat-ments and thereby prioritizes novel treatment combinationsfor clinical trial.

A Preclinical Multidisciplinary Brain Tumor Board

www.aacrjournals.org Clin Cancer Res; 24(7) April 1, 2018 1655




minimum time–concentration exposure required to inhibit cellgrowth by 50% by replacing drug-containing medium with freshmedium 1, 3, 6, 10, 24, or 72 hours after dosing. Tumor cellapoptosis was assessed by fluorescence-activated cell sorting todetect Annexin V staining (apoptosis) and DAPI staining (DNAintegrity).

ResultsPreclinical multidisciplinary brain tumor board

We recruited from our clinical MDT, a pMDT comprisingstatisticians, biologists, chemists, pharmacologists, and clinicians.The pMDT met weekly to design, conduct, and review preclinicalstudies that closely recapitulate multimodality clinical trials(Fig. 1). Trial statisticians ensured appropriate randomization oftumor-bearing animals and statistical powering of study arms;neurosurgeons performed all mouse neurosurgery; radiationoncologists prescribed and delivered fractionated radiotherapyto mice; clinical pharmacologists and oncologists guided trialdrug doses and schedules; and radiologists and small animalimaging specialists evaluated treatment response. The pMDTadhered to strict, pre-agreed, standard operating procedures thatdictated the progress of therapies through the preclinical pipeline(Fig. 1). Preclinical trial datawere accessible to all pMDTmembersin real time via a centralized electronic mouse medical record.

Using cross-species functional genetic screens, we previouslygenerated a series of orthotopic, genetic mouse (m), and humanxenograft (x) models that recapitulate the histology, transcrip-tome, and growth of SEP-CR(þ), SEP-CR(–), and CPC tumors(mSEP-CR[þ], xSEP-CR[þ],mSEP-CR[-]RTBDN,mSEP-CR[-]EPHB2,mCPC; Supplementary Table S1; refs. 15–18). Because clinicaltrials frequently employMRI to assess treatment response, we firstconfirmed that bioluminescence (our preferred method of imag-ing) and MRI provide equivalent measures of tumor volume inour mouse models (R2¼ 0.96, P < 0.0001; Fig. 2A and B). Armedwith these data and the survival rates of 294 tumor-bearing mice,pMDT statisticians then employed the Wilcoxon rank-sum testand Noether's power formula to design studies with a >83%power to detect a significant survival difference between animalsreceiving test or control treatment.

Preclinical neurosurgeryTo test the therapeutic value of surgery in our models, we es-

tablished cohorts of mice harboring mSEP-CR(þ), xSEP-CR(þ),mSEP-CR(–)RTBDNa, mSEP-CR(–)RTBDNb, or mCPC as describedpreviously (15, 16, 18). Mice bearing equivalent sized tumorswere then randomized to undergo microscope-guided tumorresection by neurosurgeons or anesthesia alone (Fig. 2C). Grosstotal resection (�10% residual postoperative bioluminescence)was achieved in 100%(n¼14/14), 64% (n¼ 9/14), 51% (n¼ 24/47), 71% (n ¼ 15/21), and 43% (n ¼ 13/30) of mice harboringmSEP-CR(þ), xSEP-CR(þ), mSEP-CR(–)RTBDNa, mSEP-CR(–)RTBDNb, or mCPC, respectively, recapitulating the total resec-tion rates of these tumors in children (Fig. 2D–H; refs. 11–14).Surgical resection of mSEP-CR(þ) and xSEP-CR(þ) producedonly transient, significant reductions in tumor volume, and thesetumors regrew rapidly following total resection, resulting in nooverall survival advantage (Fig. 2D and E). In contrast, totalresection produced sustained, significant reductions in the vol-ume of mSEP-CR(–)RTBDNa and mSEP-CR(–)RTBDNb andincreased the survival of mice harboring these tumors, curing

some animals (Fig. 2F and G). Thus, the relatively poor prognosisof patients with SEP-CR(þ) may in part reflect the failure ofsurgery to control these tumors (7). Gross total resection isgenerally regarded as optimal therapy of CPC, although this hasnot been demonstrated definitively because the disease is so rare(12, 13). In support of this notion, total resection significantlyreduced tumor burden for around 1 week and marginally, butsignificantly, extended the survival of mice with mCPC (Fig. 2H).

Preclinical fractionated radiotherapyPostoperative cranial irradiation has been a mainstay of

ependymoma therapy for decades and is used to treat somepatients with CPC (7–9, 11–13). To test the efficacy of radio-therapy in our mouse models, we randomized mice withequally sized mSEP-CR(þ), xSEP-CR(þ), mSEP-CR(–)RTBDNa,mSEP-CR(–)RTBDNb, or mCPC to receive 27 daily fractions of2 Gy cranial irradiation (mimicking that given to patients) ormock treatment (Fig. 3A). In stark contrast with surgery, radio-therapy significantly impaired the growth of all SEP-CR(þ) andSEP-CR(–) models relative to controls for between 2 and 10weeks, resulting in a significant survival advantage for treatedmice (Fig. 3B–E). Notably, regrowth of mSEP-CR(þ) andmSEP-CR(–)RTBDN was observed before the end of radiother-apy, suggesting the emergence of resistant clones, potentiallyexplaining why this treatment ultimately failed. Conversely,and in agreement with the limited radiosensitivity of infantCPC, radiotherapy only transiently impaired mCPC growth andhad no therapeutic efficacy against this tumor (Fig. 3F).

Having evaluated the efficacy of surgery and radiotherapyindependently, we conducted a series of combination studies todetermine the benefit of combining these modalities (Fig. 3G).Although surgery alone did not benefit mice harboring mSEP-CR(þ), postoperative irradiation significantly prolonged tumorcontrol in these animals relative to radiotherapy alone, resultingin cures for almost half of all treated mice (Fig. 3H). In contrast,surgical resection of xSEP-CR(þ) did not prolong the survival ofmice with this tumor relative to those treated with irradiationalone, possibly reflecting the rapid regrowth of these tumorsfollowing surgical debulking, resulting in a shorter periodof tumor control overall (Fig. 3I). However, combinationsurgery and irradiation significantly impaired the growth ofmSEP-CR(–)RTBDNa relative to surgery alone and extended thesurvival of mice with these tumors beyond that achieved witheither therapy alone (Fig. 3J). Combination surgery and radio-therapy were not attempted in mCPC because this tumor resistedboth treatments.

Repurposing of chemotherapyHaving established the value of surgery and radiotherapy

among our models of SEP and CPC, and shown that the patternof response to these treatments approximates that observedin patients, we looked to see if our models might be usefulfor developing chemotherapies. Using an integrated in vitroand in vivo screen thatwedeployedpreviously to identify potentialbrain tumor treatments for clinical trial, we screened 114 drugsthat are FDA approved or currently in clinical trial (19, 20).mSEP-CR(–)RTBDNb and mCPC cells were chosen for these studiesbecause they represent relatively responsive and resistant tumortypes, respectively. In line with their relative resistance to treat-ments, 40 drugs inhibited the proliferation ofmSEP-CR(–)RTBDNb

cells by 50% (IC50) at concentrations �1 mmol/L after 72 hours

Nimmervoll et al.

Clin Cancer Res; 24(7) April 1, 2018 Clinical Cancer Research1656




in vitro compared with only 26 drugs against mCPC cells: 22of these drugs had IC50 � 1 mmol/L against both cell types (P <0.0001, Fisher exact for overlap; Fig. 4A).

Thirteen drugs with IC50 values � 1 mmol/L at 72 hours werethen subjected to "washout" studies to determine the minimumconcentration–time exposure required to inhibit cell proliferation

(Fig. 4B). Exposure to < 1 mmol/L of cabazitaxel, pralatrexed,gemcitabine, panobinostat, carfilzombib, or vosaroxin for just1 hour inhibited the proliferation of both mSEP-CR(–)RTBDNb

and mCPC cells by >50%; chaetocin was similarly active againstmCPC, whereas the IC50 of acivicin after 1-hour exposurealmost achieved 50% inhibition in the �1 mmol/L range.

Electronic medical record of real-time resultsfeasibility, efficacy, toxicity

Preclinical brain tumor board

Mousemodelsbiologist

Chemistclinical

pharmacologist

Neurosurgeon Radiationoncologist

Clinicaloncologist

Pathologist Statistician Radiologistsmall animal

imaging

Drug selection 1high through-put drug screen; predicted to be BBB-penetrant; cancer active

Drug selection 2in vitro wash-out studies; BBB penetration in vivo pharmacokinetics

Drug selection 3modelled ‘optimal’ dose in vivo efficacy studies

Preclinical trial designmultimodality trial relative to conventional treatment

R

V V V V V

Figure 1.

Composition of the preclinical Multi-DisciplinaryTumor Board and the multistep approach taken todevelop new treatment approaches. BBB, blood–brainbarrier; R, randomization.






A

MR

IB

iolu

min

esce

nce

7 14 21Day:

105

106

107

108

109

1010

0.1

1

10

100

1,000

7 14 21 7 14 21Day post allograft

Bio

lum

ines

cenc

e (p

hoto

ns/s

ec)

MR

I volume (m

m3)

Bioluminescence MRI volume

B

R2 = 0.96P < 0.0001

0 1 7 14 21Time [days]:C

R

=Allograft =Surgical resection=BioluminescenceR =Randomised

Vehicle (n = 6)>90% surgical resection (n = 9); P = ns

xSEP-CR(+)

<90% surgical resection (n = 5); P = ns

D

0

100

50

25

75

0 20 40 60

Per

cent

sur

vivi

ng

Survival [days]

Vehicle (n = 5)>90% resection (n = 14); P = ns

mSEP-CR(+)

Time post implant [days]0 5 10 15 20 25 30 35 40 45 50

105

106

107

108

109

1010

Control>90% surgical resection (d7−12)

Vehicle (n = 18)>90% surgical resection (n = 13); P = 0.03

mCPC


0 20 40 600

100

50

25

75

Per

cent

sur

vivi

ng

Survival [days]15 20 25 30 35 40

106

107

108

109

1010

Control<90% surgical resection (ns)>90% surgical resection (d18−23)

Time post implant [days]

0 10 20 30 40 500

25

50

75

100

Survival [days]

Per

cent

sur

vivi

ng

Vehicle (n = 15)>90% surgical resection (n = 15); P = 0.0007

mSEP-CR(-)RTBDNb

<90% surgical resection (n = 6); P = 0.008

0 10 20 30 40 50106

107

108

109

1010


Control<90% surgical resection (d6)>90% surgical resection (d6−10)

Lum

ines

cenc

e (p

/s ±

SE

)

Time post implant [days]0 20 40 60 80 100

105

106

107

108

Control<90% surgical resection (ns)>90% surgical resection (d22)

E

F

G

H

mSEP-CR(+)

xSEP-CR(+)

mSEP-CR(-)RTBDNb

mCPC

0 50 100 1500

25

50

75

100

Per

cent

sur

vivi

ng

0

100

50

25

75

Per

cent

sur

vivi

ng

0 100 200 300

Vehicle (n = 29)>90% surgical resection (n = 24); P < 0.0001

mSEP-CR(-)RTBDNa


mSEP-CR(-)RTBDNa

400Survival [days]

0 20 40 60 80 100 120 140105

106

107

108

109

1010


Control<90% surgical resection (ns)>90% surgical resection (d16−58)

Survival [days]

Lum

ines

cenc

e (p

/s ±

SE

)Lu

min

esce

nce

(p/s

± S

E)

Lum

ines

cenc

e (p

/s ±

SE

)Lu

min

esce

nce

(p/s

± S

E)

Figure 2.

Preclinical brain imaging and neurosurgery.A, Concurrent MRI (top) andbioluminescence (bottom) imaging ofthe same mouse with a SEP-CR(þ) tumor.B, Correlation of MRI and bioluminescenceimaging of the same cohort of 5 mice withSEP-CR(þ) tumors. C, Preclinical mouseneurosurgical protocol. D–H (left),Bioluminescence measurement of tumorgrowth. In parentheses are the days (d)when treated tumor volumewas significantly(P < 0.05, nonparametric) less than control.(Right) survival curves ofmicewith indicatedanimal numbers and tumor types treatedwith surgery or control. P value ¼ log-rankrelative to control.

Nimmervoll et al.





Published pharmacokinetic data indicated that cabazitaxel, gem-citabine, pralatrexate, and pemetrexed would penetrate the cen-tral nervous system (CNS) and provided appropriate doses and

scheduling for vosaroxin, chaetocin, and acivicin (20–32). There-fore, to select which of these drugs might be suitable for furtherpreclinical development, the pMDT designed and conducted a

GTime [Days]: 3 14 21 28 35

I I I I I I I I I I I I I I I I I I I I I I I I I I I 0 1 42 49

R

=2 x 5 Gy cranial irradiationI I I I I=Surgical resection=Bioluminescence=Allograft R =Randomized

H

I0 50 100 150 200

0

25

50

75

100

Time [Days]

Per

cent

sur

vivi

ng

0 10 20 30 40 60 80 100105

106

107

108

109

1010

1011

Time [Days]

ControlCranial radiation (d7−50)>90% resection (d7−10)>90% resection + radiation (d7−50)

mSEP-CR(+)

Vehicle (n = 16)>90% resection (n = 14)Cranial irradiation (n = 10)>90% resection + cranial irradiation (n = 7)

***

****

mSEP-CR(+)

Time [days]0 100 200 300 400

0

25

50

75

100

Per

cent

sur

vivi

ng

0 10 20 30 40 50 100 150

105

106

107

108

109

1010

1011

104

Time [days]

ControlCranial radiation (d16−60)>90% resection (d10−30)>90% resection + radiation (d10−37)

mSEP-CR(-)RTBDNa


**

***

mSEP-CR(-)RTBDNa

JTime [days]

Per

cent

sur

vivi

ng

105

106

107

108

109

104

Time [days]

ControlCranial radiation (dy32−85)>90% resection>90% resection + radiation (dy57−85)

xSEP-CR(+)


n.s.

***

xSEP-CR(+)

0 50 100 1500 50 100 150 200 250

0

25

50

75

100

Lum

ines

cenc

e [p

/s+S

E]

Lum

ines

cenc

e [p

/s+S

E]

Lum

ines

cenc

e [p

/s+S

E]

A0 1 7 14 21Time [days]: 28 35

I I I I I I I I I I I I I I I I I I I I I I I I I I I R

=Allograft =BioluminescenceR =Randomized =2 x 5 Gy cranial irradiationI I I I I

B C

D E

0 25 50 75 1000

100

50

25

75

Per

cent

sur

vivi

ngTime [days]

Per

cent

sur

vivi

ng

0 20 40 60 800

100

50

25

75

Time [days]

1010

1011

0 20 40 60 80106

107

108

109

Time [days]

mSEP-CR(+)

mSEP-CR(-)RTBDNa

0 50 100 150106

107

108

109

1010

Time [days]0 10 20 30 40

0

25

50

75

100

Per

cent

sur

vivi

ng

Time [days]

0 50 100 150 200 2500

25

50

75

100

Per

cent

sur

vivi

ng

Time [days]0 50 100 150

104

105

106

107

108

109

Time [days]

0 10 20 30105

106

107

108

109

Time [days]

xSEP-CR(+)

mSEP-CR(-)RTBDNb

P < 0.001

mSEP-CR(+) xSEP-CR(+)

mSEP-CR(-)RTBDNamSEP-CR(-)RTBDNb

Lum

ines

cenc

e (p

/s ±

SE

)Lu

min

esce

nce

(p/s

± S

E)

Lum

ines

cenc

e (p

/s ±

SE

)Lu

min

esce

nce

(p/s

± S

E)

(d13−28)

n = 9n = 10

(d10−85)

P = 0.03

n = 5n = 5

(d16-60)(d7-22)

P = 0.0002

n = 10n = 9

P < 0.0001

n = 9n = 15

mCPC Per

cent

sur

vivi

ng

0 10 20 30 400

25

50

75

100

Time [days]

0 10 20 30 40105

106

107

108

109

1010

Time [days]

F

mCPC

Lum

ines

cenc

e (p

/s ±

SE

)

ControlRadiation

Panels A−F

(d17)

P = ns

n = 8

n = 8

Figure 3.

Preclinical fractionated surgery andirradiation. A, Preclinical mouseradiation protocol. B–F (left),Bioluminescence measurement oftumor growth. In parentheses are thedays (d) when treated tumor volumewas significantly (P < 0.05,nonparametric) less than control.(Right) survival curves of mice withindicated tumor types treated withfractionated irradiation or control.G, Preclinical combination mousesurgery and radiotherapy protocol.H–J, Bioluminescence measures oftumor growth (left) and survivalcurves (right) of mice with theindicated tumor types treated withsurgery with or without fractionatedirradiation (P values are log-rankrelative to control). Comparisonsbetween treatments are log-ranknot significant (ns), P < 0.005(��), P < 0.0005 (��), andP < 0.00005 (��).






A

mSEP-CR(-)

RTBDNb

mCPC

0-

500-

1,00

0-

1,50

0-

Active/completedclinical trialsFD

A ap

prov

edPRALATREXATE

LUMINESPIBCABAZITAXELGANETESPIBDELANZOMIB

OLTIPRAZDOXORUBICINIXABEPILONEGEMCITABINEVINORELBINE

HOMOHARRINGTONINEOPROZOMIB

ACIVICINUCN-01

CARFILZOMIBCHAETOCINDACTOLISIB

CERIVASTATINDINACICLIB

PHENOXODIOLSEPANTRONIUM

NVP-BGT226SELINEXOR

RIGOSERTIBTALAZOPARIB

IXAZOMIBQUISINOSTAT

ETOPOSIDECUDC-907

VOSAROXINTANESPIMYCIN

CLADRIBINEROMIDEPSIN

BMS-754807PANOBINOSTAT

AMUVATINIBLESTAURTINIB

GMX-1778AZD7762

LY2801653BMS-911543

MK-8776AXITINIB

PALIFOSFAMIDEGSK-923295

RABUSERTIBMK-2206

VOLASERTIBMK-1775

PF562271AZD1480

DOCETAXELTOFACITINIBEVODIAMINEDANUSERTIB

OMIPALISIBENMD-2076

BARDOXOLONEGANDOTINIB

CUDC-101TERAMEPROCOL

VORINOSTATBARICITINIB

GEDATOLISIBTEMSIROLIMUS

TIVANTINIBPACRITINIB

ERIBULINTAFENOQUINE

CRIZOTINIBNAVITOCLAXARTESUNATE

FEDRATINIBLCL161

MEFLOQUINEPF-477736

NIRAPARIBCOBIMETINIB

GSK461364ALISERTIB

REGORAFENIBGSK-2636771

TH-302ERASTIN

TRAMETINIBMLN 4924

LINSITINIBRUXOLITINIB

GDC-0032RG7112

DIMETHYL FUMARATE SULFORAPHANE

AZD5363BORTEZOMIB

OLAPARIBARRY-520

RIBAVIRINSELUMETINIB

OLMESARTAN MEDOXOMILSIROLIMUSVELIPARIB

VEMURAFENIBLOMEGUATRIB

PX-478ABT-199

BARASERTIBTAK-960

AT-406BIRINAPANT

FG-4592PS-1145

LY2157299VISTUSERTIB

PEMETREXED

IC50<0.1 1 >10

I I I

72 hr IC50 (µmol/L)

5I

B

<0.1 0.5 >1,0001I I I

Washout IC50 (µmol/L)

I

1 3 6 10 24 72time (hours)

GMX-1778

1 3 6 10 24 72time (hours)

mSEP-CR(-)RTBDNb

Cabazitaxel

mCPC

Vosaroxin

Selinexor

Rigosertib

Gemcitabine

Pralatrexate

Carfilzomib

Panobinostat

Pemetrexed

Acivicin

Chaetocin

Cladribine

0.03 0.09

0.16 0.08

0.25 0.16

0.19 0.39

0.11 0.83

0.56 0.62

5.70 1.22

2.00 10.0

10.2 11.1

9.07 0.11

1.32 1.89

C

D Time post implant [days]

Per

cent

sur

vivi

ng

0 10 20 30 400

25

50

75

100

Pemetrexed (n = 9) P = nsCabazitaxel (n = 9) P = ns Pralatrexate (n = 10) P = ns Vosaroxin (n = 10) P = 0.01 Gemcitabine (n = 8) P = 0.002

Vehicle (n = 36)

mSEP-CR(-)RTBDNb

0 20 40 60 80 1000

25

50

75

100


Per

cent

sur

vivi

ng

Pemetrexed (n = 9) P = nsCladribine (n = 13) P = nsPralatrexate (n = 8) P = ns Chaetocin (n = 6) P = ns

Gemcitabine (n = 15) P < 0.0001

Vehicle (n = 60)

Acivivin (n = 8) P = ns

mCPC

0 10 20 30 40105

106

107

108

109

1010

Pemetrexed (400 mg/kg; ns)Cabazitaxel (ns)Pralatrexate (15 mg/kg; ns) Vosaroxin (20 mg/kg; ns)Gemcitabine (120 mg/kg; d11−18)

Vehicle

Time post implant [Days]

mSEP-CR(-)RTBDNb

Lum

ines

cenc

e [p

/s ±

SE

]

0 10 20 30 40105

106

107

108

109

1010


mCPC

Lum

ines

cenc

e (p

/s ±

SE

)

Pemetrexed (400 mg/kg; ns)Cladribine (75 mg/kg; ns)Pralatrexate (15 mg/kg; ns)Chaetocin (0.2 mg/kg; ns)

Gemcitabine (120 mg/kg; ns)

Vehicle

Acivicin (0.4 mg/kg; ns)

Figure 4.

Preclinical repurposing of chemotherapies.A, Repurposing screen of 114 FDA-approvedand/or clinical trial drugs. Heatmap (left)reports 72-hour IC50 against the indicated celltype; graybars (middle) indicate FDA-approveddrugs; graph (right) reports number ofcompleted trials of each drug. Arrows denotedrugs selected for further study in B. B,"Washout studies": heatmaps report IC50 valuesafter timed exposures of the indicated cell typesto drug. C–D (top), Bioluminescencemeasurement of tumor growth. In parenthesesare the days (d) when treated tumor volumewas significantly (P < 0.05, nonparametric) lessthan control. (Bottom) survival curves of micewith indicated tumor types treated with theindicated drug monotherapy. Comparisonsbetween treatments are log-rank not significant(ns) and P < 0.0005 (��).

Nimmervoll et al.





series of monotherapy preclinical trials in mice harboring mSEP-CR(–)RTBDNb or mCPC. The goal of these studies was to look forany evidence of antitumor activity (growth and/or survival).Doses and schedules of each drug were designed to mimic thoseachievable in patients. We also conducted a monotherapystudy of cladribine as a "negative control" compound becausethis drug was relatively inactive in vitro and was predicted not topenetrate the CNS. Of all agents tested, only gemcitabine (120mg/kg intravenous bolus) displayed significant activity againstboth mSEP-CR(–)RTBDNb and mCPC: this treatment was the onlymonotherapy to significantly impair the growth of mSEP-CR(–)RTBDNb and to prolong the survival of mice with these tumors;therefore, this drug was selected for further repurposing studies(Fig. 4C andD). Vosaroxin—a topoisomerase II inhibitor causingsite-selective DNA damage—also produced a modest but signif-icant survival advantage for mice harboring mSEP-CR(-)RTBDNb

tumors (Fig. 4C). These data underscore that drugs with relativelypotent activity in vitromay lack efficacy in vivowhen administeredat clinically relevant doses. In light of the considerableactivity of gemcitabine,we selected this drug for further preclinicaldevelopment.

Optimization of gemcitabine therapyGemcitabine can be administered as an intravenous bolus or

infusion, resulting in very different pharmacokinetic profiles(33). Therefore, we treated mice bearing mSEP-CR(–)RTBDNb ormCPC with various gemcitabine regimens and simultaneouslymeasured concentrations of the drug in plasma and braintumor extracellular fluid (tECF) using intratumoral microdia-lysis. Mice were treated initially with two clinically relevantgemcitabine regimens: 60 mg/kg i.v. bolus that is active againsta mouse model of group 3 medulloblastoma; or continuous 3-hour infusion via subcutaneous Alzet pumps (19, 20). Notethat 60 mg/kg i.v. bolus gemcitabine produced a plasmaAUC0–6hr of 25.9 mmol/L�hr that is equivalent to that observedin children treated with 1,200 mg/m2 (Fig. 5A; ref. 34). Thetumor-to-plasma partition coefficient for unbound gemcita-bine (Kp,uu) at this dose was 0.51 and 0.18 for mice bearingmSEP-CR(–)RTBDNb and mCPC tumors, respectively. The gem-citabine concentration in tECF produced by this regimen onlyremained above the in vitro washout IC50 of each tumor typefor less than 3 hours (compare Figs. 4B and 5A). In contrast,3-hour infusions of gemcitabine produced plasma exposuresof 95.1 � 24.1 mmol/L�hr—equivalent to treating childrenwith 2,000 mg/m2

—and in both models maintained a tECFconcentration above the IC50 in washout studies for �7 hours(compare Figs. 4B and 5B). To determine if these in vivoexposures produce the antitumor cell effects predicted in vitro,we harvested tumors from mice at 3, 8, 24, and 48 hoursfollowing initiation of gemcitabine therapy and estimatedlevels of tumor cell proliferation and apoptosis. Three-hourinfusions of gemcitabine induced significantly greater andmore sustained levels of tumor cell apoptosis in mSEP-CR(–)RTBDNb and mCPC than did 60 mg/kg i.v. bolus treatment,and 3-hour infusions produced a more significant andsustained reduction in tumor cell proliferation, although thiswas only observed in mSEP-CR(–)RTBDNb (Fig. 5C; Supple-mentary Fig. S1).

As a final step to select the optimal dose and schedule ofgemcitabine for preclinical assessment, we further expanded therepertoire of gemcitabine regimens to assess the relative activity of

200 mg/kg bolus and 6-hour infusions (n � 10 mice percohort; Fig. 5D–F).Of all regimens tested, 3- and 6-hour infusionsof gemcitabine were most efficacious, producing similar degreesof tumor growth suppression and enhanced overall survival;however, 3-hour infusions proved the least toxic. Additional 3-hour gemcitabine infusion monotherapy trials identified signif-icant active against mSEP-CR(þ), xSEP-CR(þ), and a mSEP-CR(–) model driven by EPHB2 (Fig. 5G–I; ref. 17). Three-hourgemcitabine infusions were more efficacious than combinationcisplatin/cyclophosphamide or cisplatin/etoposide/vincristinethat approximates "standard of care" chemotherapy regimensthat have been tested against ependymomaandCPC, respectively,in the clinic (Fig. 5J and K; refs. 7, 35). Therefore, we selected 3-hour infusions of gemcitabine for our final phase of preclinicalrepurposing.

Combining gemcitabine and conventional therapyThe efficacy of gemcitabine monotherapy in our model

systems suggests it may have value as an adjuvant therapy inthe clinic. Therefore, the pMDT designed a combinationstudy aimed at testing the value of adding 3-hour gemcitabineinfusions to "standard-of-care" surgery and fractionated radio-therapy (Fig. 6A). With regard to ependymoma, we focused onSEP-CR(þ) disease because this tumor type is the most aggres-sive form of SEP. The mSEP-CR(þ) rather than xSEP-CR(þ)model was chosen for these studies because these modelsdisplayed similar responses to surgery, radiotherapy, and gem-citabine as monotherapies, but the more rapid growth profile ofmSEP-CR(þ) enabled completion of these large combinationstudies in a timely manner. Mice were treated with GTR fol-lowed by 54 Gy fractionated irradiation and then 3 weeks ofconsecutive 3-hour gemcitabine infusions. This treatment wastolerated remarkably well. Although the average tumor burdenof mice receiving gemcitabine was lower than that of animalstreated with surgery and irradiation alone, this differencewas not significant (Fig. 6A); however, the addition of gemci-tabine doubled the median survival (183 days) of mice relativeto those treated with surgery and irradiation alone (96 days),and cured 50% of animals (P < 0.00001; Fig. 6B). These dataunderscore the need to assess both tumor volume and animalsurvival as response metrics to preclinical therapy becausetumor imaging in small animals may operate at the limitsof resolution. We next assessed the value of adding serialpostoperative, 3-hour gemcitabine infusions to the treatmentof mCPC (Fig. 6C and D). As observed previously, gross totaltumor resection alone produced a modest but significant sur-vival advantage for mice harboring mCPC (median survivalsurgery ¼ 34 days vs. control ¼ 22 days; P < 0.003; Fig. 6D);and gemcitabine therapy alone markedly extended the survivalof mice with these tumors (median survival gemcitabine ¼ 42days vs. control ¼ 22 days; P < 0.0001; Fig. 6D). Notably, nosignificant difference in survival was observed between miceundergoing surgery or gemcitabine therapy alone; however,surgical resection followed by gemcitabine significantly extend-ed survival above that of animals receiving surgery alone(median survival surgery alone ¼ 44 vs. surgeryþgemcitabine¼ 46.5 days; P < 0.0001; Fig. 6D). Together, these data suggestthat 3-hour infusions of gemcitabine may add therapeutic valueto "standard-of-care" surgery and radiation in the treatment ofSEP and may improve the results of surgical resection of CPC.We propose that these regimens should be tested in the clinic.






A

0 2 4 60.01

0.1

1

10

100

1,000

Time after dosing (hr)0 2 4 6

0.01

0.1

1

10

100

1,000

Time after dosing (hr)0 2 4 60.01

0.1

1

10

100

1,000

Time after dosing (hr)0 2 4 6

0.01

0.1

1

10

100

1,000

Time after dosing (hr)

0 2 4 6 80.01

0.1

1

10

100


B

0 2 4 6 80.01

0.1

1

10

100


Gemcitabine (3 hr inf.)mCPC

Gemcitabine (3 hr inf.)mSEP-CR(-)RTBDNb

0 2 4 6 80.01

0.1

1

10

100


Gemcitabine (3 hr inf.)mCPC

Time after dosing (hr)0 2 4 6 8

0.01

0.1

1

10

100

Gem

cita

bine

(µm

ol/L

)

Gem

cita

bine

(µm

ol/L

)

Gem

cita

bine

(µm

ol/L

)

Gem

cita

bine

(µm

ol/L

)

Gem

cita

bine

(µm

ol/L

)

Gem

cita

bine

(µm

ol/L

)

Gem

cita

bine

(µm

ol/L

)

Gem

cita

bine

(µm

ol/L

)

Tumor ECF (±95CI)Tumor ECF (±95CI) Plasma (±95CI)Plasma (±95CI)

10

20

30

03 hours 8 hours 24 hours 48 hours

Time post onset of treatment

*

*

% C

aspa

se 3

+ cel

ls (±

SE

)

% C

aspa

se 3

+ cel

ls (±

SE

)

10

30

50

20

40

3 hours 8 hours 24 hours 48 hoursTime post onset of treatment

****

***

0

mCPCmSEP-CR(-)RTBDNb

C D

-20

-10

0

10

20

% W

eigh

t cha

nge

(min

-max

)

1 day 4 days 7 days 8 daysTime post onset of treatment

** ****

*

EGemcitabine 60 mg/kg ivGemcitabine 3 hour pump

Vehicle control

6 hour pumps3 hour pumpsVehicle pumps

120 mg/kg iv200 mg/kg iv

Vehicle iv

F

Gemcitabine (60 mg/kg)mSEP-CR(-)RTBDNb

Gemcitabine (60 mg/kg)mCPC

Gemcitabine (60mg/kg)mCPC

Plasma Plasma PlasmaTumor ECF

Gemcitabine (60 mg/kg)

PlasmaTumor ECF

mSEP-CR(-)RTBDNb

Gemcitabine (3 hr inf.)mSEP-CR(-)RTBDNb

0 10 20 30 40 50 60 70

106

107

108

109

Time (days)

mCPC

60 mg/kg (ns)200 mg/kg (ns)6 hr infus. (d8−48)3 hr infus. (d10−50)

Vehicle

0 20 40 60 80 1000

25

50

75

100

Time (days)

mCPC

60 mg/kg; n = 16, P = ns200 mg/kg; n = 8, P = 0.026 hr inf; T = 10, P = 0.00053 hr inf; n = 17, P < 0.0001

Vehicle (n = 65)

0 10 20 30 40 50105

106

107

108

109

1010

Time (days)

60 mg/kg (ns)200 mg/kg (ns)6 hr inf (d6−20)3 hr inf (d12−20)

VehiclemSEP-CR(-)RTBDNb

0 10 20 300

25

50

75

100

40

Time (days)

60 mg/kg; n = 9, P = ns200 mg/kg; n = 9, P = 0.026 hr inf; n = 10, P = 0.0063 hr infus; n = 9, P = 0.02

Vehicle (n = 43)mSEP-CR(-)RTBDNb

Time (days) Time (days) Time (days) Time (days)0 10 20 30

105

106

107

108

109

1010

1011

mSEP-CR(-)EPHB2

3 hr infus. (d6−20)Vehicle

0 10 20 30 400

25

50

75

100

Per

cent

sur

vivi

ng

3 hr infus; n = 9P = 0.02

Vehicle (n = 10)mSEP-CR(-)EPHB2

0 50 100 150

107

108

109

1010

3 hr infus. (d78-88)Vehicle

xSEP-CR(+)

0 100 200 3000

25

50

75

100

3 hr infus; n = 6P = 0.0008

Vehicle (n = 6)xSEP-CR(+)

0 20 40 60107

108

109

1010

Time (days) Time (days) Time (days) Time (days)

mSEP-CR(+)

3 hr infus. (d7−24)Vehicle

0 20 40 60 800

25

50

75

100

Per

cent

sur

vivi

ngP

erce

nt s

urvi

ving

Per

cent

sur

vivi

ngP

erce

nt s

urvi

ving3 hr infus; n = 20

P < 0.0001

Vehicle (n = 18)mSEP-CR(+)

G H

I

0

25

50

75

100

Per

cent

sur

vivi

ng

0 20 40 60 80

CDDP/CP (n = 9)P < 0.001

Vehicle (n = 18)mSEP-CR(+)

0 20 40 60

107

108

109

1010

mSEP-CR(+)

CDDP/CP (d11−41)Vehicle

J

0

25

50

75

100

Per

cent

sur

vivi

ng

0 20 40 60 80

CDDP/VP16/VCR (n = 16), P = nsVehicle (n = 14)

mCPC

Time (days) Time (days)

K

mCPC

CDDP/VP16/VCR (ns)Vehicle

0 5 10 15 20 25 30106

107

108

109

1010

1011

Lum

ines

cenc

e (p

/s ±

SE)

Lum

ines

cenc

e (p

/s ±

SE)

Lum

ines

cenc

e (p

/s ±

SE)

Lum

ines

cenc

e (p

/s ±

SE)

Lum

ines

cenc

e (p

/s ±

SE)

Lum

ines

cenc

e (p

/s ±

SE)

Lum

ines

cenc

e (p

/s ±

SE)

Figure 5.

Pharmacokinetic, toxicity, and efficacy studies of gemcitabine. Plasma and tECF concentration–time plots in the indicated tumor types treated with 60 mg/kgbolus (A) or 3-hour infusion (B) gemcitabine. C, Graphs showing induction of tumor cell apoptosis measured by cleaved Caspase 3 immunohistochemistryin mice with the indicated brain tumors treated with 60 mg/kg or 3-hour infusion of gemcitabine. D, Toxicity determined by weight loss in mice treatedwith the indicated doses and regimens of gemcitabine (in C and D: � , P < 0.05; �� , P < 0.005; �� , P < 0.0005, Mann–Whitney). E–I, Bioluminescence measurementof tumor growth in mice treated with indicated dose and schedule of gemcitabine. In parentheses are the days (d) when treated tumor volume was significantly(P < 0.05, nonparametric) less than control. (Right) survival curves of the same mice shown left. P value ¼ log-rank relative to control. Similar growth andsurvival curves are shown in J and K for mice treated with cisplatin/cyclophosphamide or cisplatin/etoposide/vincristine, respectively.

Nimmervoll et al.





DiscussionThe past decade has witnessed a revolution in our understand-

ing of human cancer. The integration of genomic and develop-mental biology has shown that morphologically similar cancerscomprise subtypes, driven by different genetic alterations, whichlikely arise within distinct cell lineages (36). These data helpexplain why cancers once regarded histologically as homoge-neous diseases display discrepant behaviors. For example,medul-loblastoma and ependymoma are now known to include sub-types with extraordinarily good (e.g., WNT-medulloblastomaand SEP-CR[-]) or bad (Group 3-medulloblastoma with MYCamplification and SEP-CR[þ]) prognosis (11, 37). This knowl-edge could pinpoint patients who might be cured with lesstoxic therapy, as well as poor prognosis patients who need newtreatments. Indeed, clinical trials of decreased radiotherapyare ongoing among patients with WNT-medulloblastoma(NCT01878617). But integrating understanding of tumor biologyinto established clinical practice is enormously challenging andrequires a number of assumptions that are often made withoutknowledge of subtype-specific treatment efficacy. For example,reducing radiotherapy for children with WNT-medulloblastomaassumes that this therapy, rather than surgery or chemotherapy,is relatively redundant. And trials of new treatments for 'poorprognosis' tumors often assume that relatively ineffective con-ventional therapies should be retained; this approach runs the riskof increasing toxicity unnecessarily.

So how can we integrate new understanding of cancer biologyand therapy into empirical treatment regimens that havedeveloped over decades? It is unlikely that we will achieve this

through clinical trials alone, especially among patients with raredisease variants: the small populations of patients with thesetumors limit the number of drugs and regimens that can be testedin a timely manner. The preclinical platform described hereprovides an evidence-based approach to guide clinical trials forrare brain tumor subtypes. It is important to note that thisplatform is not designed to replace or reduce the clinical trialplatform; but rather to better triage drugs so that clinicians canfocus on novel regimens with the greatest potential to cure. Keyfeatures of our approach include the use of accurate mousemodels of human brain tumors and the coordinated engagementof clinical and research professionals in regular pMDT discus-sions, greatly facilitating the codevelopment of clinically relevantpreclinical trials.

Maximal surgical resection of ependymoma followed by irra-diation is consistently associated with a better patient outcomeregardless of primary tumor site (8, 10, 11, 14). This observationhas led to thewidespread notion that SEPs have a high probabilityof cure with surgery alone, and underpins Arm 1 of an ongoingChildren's Oncology Group study in which children achievinga gross total resection of classic histology SEP receive no furthertreatment (NCT01096368; ACNS0831). But in our studies,surgery alone had no therapeutic value in the treatment ofmSEP-CR(þ) or xSEP-CR(þ), only benefiting mice with mSEP-CR(–)RTBDNa. However, total resection of mSEP-CR(þ) didmarkedly improve the efficacy of irradiation, curing a signifi-cant number of animals and also improved the survival of micewith mSEP-CR(–)RTBDNa. These data underscore the importantpoint that treatments can display surprising interactions,

0 50 100 150 200 2500

25

50

75

100

Per

cent

sur

viva

l

mSEP-CR(+)

Surgery/radiation (n = 7), median survival 96 days P < 0.0001Vehicle (n = 16), median survival 31 days

Surgery/radiation/gemcitabine (n = 9), median survival 190 days P < 0.0001

BA

0 10 20 30 40104

105

106

107

108

109

1010

1011

60 80 100Time (days)

Time (days) Time (days)

Time (days)

ControlSurgery/radiation (d17−50) Surgery/radiation/gemcitabine (d6−50)

mSEP-CR(+)

0 25 50 75 1000

25

50

75

100

Per

cent

sur

viva

l

D

15 20 25 30 35 40 45 50106

107

810

109

1010C

Control

Gemcitabine (d17−21)Surgery/gemcitabine (d17−24)

mCPC

Surgery (d17−23)

mCPC

Surgery (n = 44), med survival 34 days P < 0.003Vehicle (n = 25), med survival 22 days

Surgery/gemcitabine (n = 14), med survival 46.5 days P < 0.0001Gemcitabine (n = 17), med survival 42 days P < 0.0001

*** n.s.n.s.

Lum

ines

cenc

e (p

/s ±

SE

)Lu

min

esce

nce

(p/s

± S

E)

Figure 6.

Combination surgery, fractionated irradiation, and postirradiation gemcitabine therapy. Bioluminescence measures of tumor growth (A) and survival curves(B) of mice with mSEP-CR(þ) treated with surgery and radiotherapy alone or surgery, radiation, and 3-hour infusions of gemcitabine. Bioluminescencemeasures of tumor growth (C) and survival curves (D) ofmicewithmCPC treatedwith surgery and 3-hour infusions of gemcitabine aloneor surgery andgemcitabine.Figures in parentheses in bioluminescence plots are the days (d) when treated tumor volume was significantly (P < 0.05, nonparametric) less than control.






producing high cure rates when used in combinations that notare not apparent when the treatments are used individually.Our data also support the notion that irradiation is a highlyeffective treatment of SEP and suggest that avoiding radiother-apy for all patients with totally resected SEP, regardless ofsubtype, may be inappropriate. Rather, as a minimum, werecommend the prospective evaluation of SEP molecular sub-type in ACNS0831 to ensure that SEP-CR(þ) patients are notundertreated. Reverse translation of these clinical data will becritical to validate the predictions made by our preclinicalsystem. This later point is particularly important because ourmodel system is likely to be predictive but not infallible.Indeed, in contrast with mSEP-CR(þ), total resection ofxSEP-CR(þ) did not improve the efficacy of radiotherapy,indicating that further iteration between preclinical and clinicalwork will be required to understand the ependymoma-sub-type–specific relevance of combination treatment.

The effectiveness of preirradiation surgery in our models alsosupports the widely held notion that cytoreductive surgeryincreases radiosensitivity and chemotherapy by removing thera-py-resistant, hypoxic, and highly-proliferative tumor cores (38).These data might also explain why resection and irradiation aremore effective than partial resection and irradiation amongpatients with posterior fossa subtype-A ependymoma—anotheraggressive disease subtype (14). Thus, our models provide anopportunity to explore the biological basis of cytoreductive sur-gical efficacy. Our models may also facilitate the identification,isolation, and study of radiation resistant tumor clones, becauseour imaging studies revealed regrowth of mSEP-CR(þ), xSEP-CR(þ), and mSEP-CR(-) prior to completion of radiotherapy.

In contrast with our SEP models, radiotherapy proved ineffec-tive against mCPC. The radioresistance of mCPC may reflect theTp53-null status of these tumors, because this gene mediates celldeath mechanisms in irradiated cells (39). Notably, 60% ofhuman CPCs contain mutant Tp53: these tumors also tend tobe radioresistant, clinically aggressive, and to develop in infants(40, 41). It is interesting that our mCPC model is initiated inembryonic choroid plexus; therefore, these tumors likely modelradioresistant, aggressive, and TP53-mutant infant CPC (18).

Although chemotherapy has been evaluated in ependymomaand CPC, its role remains controversial with only limited benefitreported (7, 13). These data are in keeping our observations thatmost drugs displaying potent activity against our models in vitrofailed to produce therapeutic benefit in vivo. This notion isalso supported by our observation that mouse models of SHH-medulloblastoma—a more chemosensitive disease—respondedto treatments that were ineffective against mSEP and mCPC, e.g.,pemetrexed and 60 mg/kg bolus gemcitabine (20, 42). Ourpreclinical in vitro and in vivo pipeline did identify 3-hour infu-sions of gemcitabine as a potential new treatment of SEP andCPC.This regimen generated tECF concentrations above the in vitro IC50

for�7 hours and provedmore effective against mSEP-CR(þ) andmCPC than combination conventional chemotherapy regimenswith reported activity in patients (7, 35). Thus, we suggest that3-hour infusions of gemcitabine should be tested in patientswith SEP and CPC.

Fewer than 150 adults and children with all variants of SEP andCPC are available for enrollment on clinical trials in the UnitedStates each year, severely limiting studies of new treatments (43).Indeed, it is widely agreed that the rarity of CPC poses an almostinsurmountable hurdle to the efficient development of new

treatments through clinical trial (13). For example, the onlymulticenter CPC clinical trial conducted to date was initiated in2000 (CTP-SIOP-2000), but 17 years later, the results of this trialare yet to be published. Our preclinical system provides analternative, evidence-based approach to prioritize combinationregimens for the clinic, potentially avoiding years of trials ofineffective therapies. Of particular note, by recapitulating surgery,irradiation, and chemotherapy, our approach allows for preclin-ical trials of multiple doses, delivery routes, and schedules ofnovel chemotherapies in the context of standard-of-care treat-ment. In this regard, sequential total tumor resection, fractionatedradiotherapy, and 3-hour gemcitabine infusions doubled themedian survival of mice with mSEP-CR(þ) relative to surgeryand irradiation alone, curing half of all animals. Our studies alsoprovide evidence that combination surgery and gemcitabineinfusion therapy may benefit the treatment of CPC. These dataare in keeping with the activity of gemcitabine in other chemore-sistant cancers including pancreatic cancer (44, 45). We thereforerecommend that gemcitabine infusions might prove effective aspostsurgery and irradiation chemotherapy. Furthermore, becausegemcitabine may also serve as a radiosensitizer, we are currentlyexploring the timing of gemcitabine treatment relative to irradi-ation and whether gemcitabine may be added to conventionaltreatment regimens in younger patients.

Although our model system provides a promising tool toprioritize complex combination treatment regimens for clinicaltrial, the accuracy of thesepredictions remains tobe assessed. It is ahard reality that most cancer treatments that are effective inanimal models fail in patients (46, 47). Indeed, although ourmodels closely replicate themorphology and transcriptomeof thecorresponding human tumors, they are maintained in immuno-compromised hosts and therefore cannot account for contribu-tions of the host immune system to tumor biology and treatment.Thus, preclinical platforms such as the one presented here requirecareful iterative study with clinical translation to be validated andrefined. This important ongoing process further underscores thevalue of convening pMDT teams comprising laboratory andclinical oncology professionals.

Disclosure of Potential Conflicts of InterestB.B. Freeman is an employee of and has ownership interests (including

patents) at KinDynaMet LLC. M.F. Roussel is a consultant/advisory boardmember for Cold SpringHarbor Laboratories and theNational Cancer Institute,and reports receiving commercial research support from Eli Lilly. No potentialconflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: B.V. Nimmervoll, N. Boulos, A. Shelat, A. Gajjar,Y.T. Patel, R.K. Guy, T.E. Merchant, K.D. Wright, R.J. GilbertsonDevelopment of methodology: B.V. Nimmervoll, N. Boulos, B. Bianski,J. Dapper, A. Shelat, S. Terranova, A. Gajjar, Y.T. Patel, B.B. Freeman,A. Onar-Thomas, C.F. Stewart, R.K. Guy, T.E. Merchant, C. Calabrese,K.D. Wright, R.J. GilbertsonAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): B.V. Nimmervoll, N. Boulos, B. Bianski, J. Dapper,M. DeCuypere, Y.T. Patel, B.B. Freeman, C.F. Stewart, T.E. Merchant,C. Calabrese, K.D. Wright, R.J. GilbertsonAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): B.V. Nimmervoll, N. Boulos, J. Dapper, A. Shelat,Y.T. Patel, B.B. Freeman, C.F. Stewart, R.K. Guy, T.E. Merchant, C. Calabrese,K.D. Wright, R.J. GilbertsonWriting, review, and/or revision of the manuscript: B.V. Nimmervoll,N. Boulos, A. Shelat, A. Gajjar, A. Onar-Thomas, C.F. Stewart, M.F. Roussel,T.E. Merchant, K.D. Wright, R.J. Gilbertson

Nimmervoll et al.





Administrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): B.V. Nimmervoll, N. Boulos, A. Gajjar,B.B. Freeman, K.D. Wright, R.J. GilbertsonStudy supervision: B.V. Nimmervoll, N. Boulos, A. Gajjar, K.D. Wright,R.J. GilbertsonOther (wrote the manuscript): R.J. Gilbertson

AcknowledgmentsThis work was supported by grants from the NIH, P01CA96832

(R.J. Gilbertson, B.V. Nimmervoll, N. Boulos, A. Gajjar, C.F. Stewart, andM.F. Roussel) and R0CA1129541 (R.J. Gilbertson andN. Boulos); the AmericanLebanese Syrian Associated Charities (R.J. Gilbertson, B.V. Nimmervoll,N. Boulos, A. Gajjar, C.F. Stewart, M.F. Roussel, B. Bianski, J. Dapper, A. Shelat,

Y.T. Patel, B.B Freeman, A. Onar-Thomas, R.K. Guy, T.E. Merchant,C. Calabrese, and K.D. Wright); Cancer Research UK (R.J. Gilbertson,B.V. Nimmervoll, and S. Terranova); the Mathile Family Foundation(R.J. Gilbertson, B.V. Nimmervoll, and S. Terranova); and Cure Search(R.J. Gilbertson, B.V. Nimmervoll, and S. Terranova).

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received July 26, 2017; revised November 21, 2017; accepted December 21,2017; published OnlineFirst January 4, 2018.

References1. Gilbert MR, Sulman EP, Mehta MP. Bevacizumab for newly diagnosed

glioblastoma. N Engl J Med 2014;370:2048–9.2. Gajjar A, Bowers DC, Karajannis MA, Leary S, Witt H, Gottardo NG.

Pediatric brain tumors: innovative genomic information is transform-ing the diagnostic and clinical landscape. J Clin Oncol 2015;33:2986–98.

3. Brinkman TM, Krasin MJ, Liu W, Armstrong GT, Ojha RP, Sadighi ZS, et al.Long-term neurocognitive functioning and social attainment in adultsurvivors of pediatric CNS tumors: results from the St Jude lifetime cohortstudy. J Clin Oncol 2016;34:1358–67.

4. Chemaitilly W, Li Z, Huang S, Ness KK, Clark KL, Green DM, et al. Anteriorhypopituitarism in adult survivors of childhood cancers treated withcranial radiotherapy: a report from the St jude lifetime cohort study.J Clin Oncol 2015;33:492–500.

5. Krueger DA, CareMM,Holland K, Agricola K, Tudor C,Mangeshkar P, et al.Everolimus for subependymal giant-cell astrocytomas in tuberous sclero-sis. N Engl J Med 2010;363:1801–11.

6. Kamb A. What's wrong with our cancer models? Nat Rev Drug Discov2005;4:161–5.

7. Khatua S, Ramaswamy V, Bouffet E. Current therapy and the evolvingmolecular landscape of paediatric ependymoma. Eur J Cancer 2017;70:34–41.

8. Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford RA. Conformalradiotherapy after surgery for paediatric ependymoma: a prospective study.Lancet Oncol 2009;10:258–66.

9. Merchant TE, Mulhern RK, Krasin MJ, Kun LE, Williams T, Li C, et al.Preliminary results from a phase II trial of conformal radiation therapy andevaluation of radiation-related CNS effects for pediatric patients withlocalized ependymoma. J Clin Oncol 2004;22:3156–62.

10. Pajtler KW, Mack SC, Ramaswamy V, Smith CA, Witt H, Smith A, et al. Thecurrent consensus on the clinical management of intracranial ependy-moma and its distinct molecular variants. Acta Neuropathol 2017;133:5–12.

11. Pajtler KW, Witt H, Sill M, Jones DT, Hovestadt V, Kratochwil F, et al.Molecular classification of ependymal tumors across All CNS compart-ments, histopathological grades, and age groups. Cancer Cell 2015;27:728–43.

12. Lafay-Cousin L, Keene D, Carret AS, Fryer C, Brossard J, Crooks B, et al.Choroid plexus tumors in children less than 36 months: the Canadianpediatric brain tumor consortium (CPBTC) experience. Childs Nerv Syst2011;27:259–64.

13. Sun MZ, Oh MC, Ivan ME, Kaur G, Safaee M, Kim JM, et al. Currentmanagement of choroid plexus carcinomas. Neurosurg Rev 2014;37:179–92; discussion 92.

14. Ramaswamy V, Hielscher T, Mack SC, Lassaletta A, Lin T, Pajtler KW, et al.Therapeutic impact of cytoreductive surgery and irradiation of posteriorfossa ependymoma in the molecular era: a retrospective multicohortanalysis. J Clin Oncol 2016;34:2468–77.

15. Parker M, Mohankumar KM, Punchihewa C, Weinlich R, Dalton JD, Li Y,et al. C11orf95-RELA fusions drive oncogenic NF-kappaB signalling inependymoma. Nature 2014;506:451–5.

16. Mohankumar KM, Currle DS, White E, Boulos N, Dapper J, Eden C, et al.An in vivo screen identifies ependymoma oncogenes and tumor-suppres-sor genes. Nat Genet 2015;47:878–87.

17. Johnson RA, Wright KD, Poppleton H, Mohankumar KM, Finkelstein D,Pounds SB, et al. Cross-species genomicsmatches drivermutations and cellcompartments to model ependymoma. Nature 2010;466:632–6.

18. Tong Y, Merino D, Nimmervoll B, Gupta K, Wang YD, Finkelstein D, et al.Cross-species genomics identifies TAF12, NFYC, and RAD54L as choroidplexus carcinoma oncogenes. Cancer Cell 2015;27:712–27.

19. Atkinson JM, Shelat AA, Carcaboso AM, Kranenburg TA, Arnold LA, BoulosN, et al. An integrated in vitro and in vivo high-throughput screen identifiestreatment leads for ependymoma. Cancer Cell 2011;20:384–99.

20. Morfouace M, Shelat A, Jacus M, Freeman BB 3rd, Turner D, Robinson S,et al. Pemetrexed and gemcitabine as combination therapy for the treat-ment of Group3 medulloblastoma. Cancer Cell 2014;25:516–29.

21. Dieras V, Lortholary A, Laurence V, Delva R, Girre V, Livartowski A, et al.Cabazitaxel in patients with advanced solid tumours: results of a phase Iand pharmacokinetic study. Eur J Cancer 2013;49:25–34.

22. FerronGM,Dai Y, SemiondD. Population pharmacokinetics of cabazitaxelin patients with advanced solid tumors. Cancer Chemother Pharmacol2013;71:681–92.

23. VrignaudP, SemiondD, Lejeune P, BouchardH,Calvet L, CombeauC, et al.Preclinical antitumor activity of cabazitaxel, a semisynthetic taxane activein taxane-resistant tumors. Clin Cancer Res 2013;19:2973–83.

24. Taylor S, Belt RJ, Joseph U, Haas CD, Hoogstraten B. Phase I evaluation ofAT-125 single dose every three weeks. Invest New Drugs 1984;2:311–4.

25. Weiss GR,McGovren JP, Schade D, Kufe DW. Phase I and pharmacologicalstudy of acivicin by 24-hour continuous infusion. Cancer Res 1982;42:3892–5.

26. Baruchel S, Bernstein M, Whitehead VM, Devine S, Bell B, Dubowy R, et al.A phase I study of acivicin in refractory pediatric solid tumors. A pediatriconcology group study. Invest New Drugs 1995;13:211–6.

27. Mould DR, Sweeney K, Duffull SB, Neylon E, Hamlin P, Horwitz S, et al.A population pharmacokinetic and pharmacodynamic evaluation ofpralatrexate in patients with relapsed or refractory non-Hodgkin's orHodgkin's lymphoma. Clin Pharmacol Ther 2009;86:190–6.

28. O'Connor OA. Pralatrexate: an emerging new agent with activity in T-celllymphomas. Curr Opin Oncol 2006;18:591–7.

29. Reichelova V, Juliusson G, Spasokoukotskaja T, Eriksson S, Liliemark J.Interspecies differences in the kinetic properties of deoxycytidine kinaseelucidate the poor utility of a phase I pharmacologically directed dose-escalation concept for 2-chloro-20-deoxyadenosine. Cancer ChemotherPharmacol 1995;36:524–9.

30. Kearns CM, Blakley RL, Santana VM, Crom WR. Pharmacokinetics ofcladribine (2-chlorodeoxyadenosine) in children with acute leukemia.Cancer Res 1994;54:1235–9.

31. Lee YM, Lim JH, Yoon H, Chun YS, Park JW. Antihepatoma activity ofchaetocin due to deregulated splicing of hypoxia-inducible factor 1alphapre-mRNA in mice and in vitro. Hepatology 2011;53:171–80.

32. Lancet JE, Ravandi F, Ricklis RM, Cripe LD, Kantarjian HM, Giles FJ, et al.A phase Ib study of vosaroxin, an anticancer quinolone derivative, inpatients with relapsed or refractory acute leukemia. Leukemia2011;25:1808–14.

33. Ciccolini J, Serdjebi C, Peters GJ, Giovannetti E. Pharmacokinetics andpharmacogenetics of gemcitabine as a mainstay in adult and pediatriconcology: an EORTC-PAMM perspective. Cancer Chemother Pharmacol2016;78:1–12.






34. Reid JM, QuW, Safgren SL, Ames MM, Krailo MD, Seibel NL, et al. Phase Itrial and pharmacokinetics of gemcitabine in children with advanced solidtumors. J Clin Oncol 2004;22:2445–51.

35. Grundy RG, Wilne SA, Weston CL, Robinson K, Lashford LS, Ironside J,et al. Primary postoperative chemotherapy without radiotherapy for intra-cranial ependymoma in children: the UKCCSG/SIOP prospective study.Lancet Oncol 2007;8:696–705.

36. Gilbertson RJ. Mapping cancer origins. Cell 2011;145:25–9.37. Northcott PA, JonesDT, KoolM,RobinsonGW,GilbertsonRJ, ChoYJ, et al.

Medulloblastomics: the end of the beginning. Nat Rev Cancer 2012;12:818–34.

38. Stummer W, KampMA. The importance of surgical resection in malignantglioma. Curr Opin Neurol 2009;22:645–9.

39. Fei P, El-Deiry WS. P53 and radiation responses. Oncogene 2003;22:5774–83.

40. Merino DM, Shlien A, Villani A, Pienkowska M, Mack S, Ramaswamy V,et al. Molecular characterization of choroid plexus tumors reveals novelclinically relevant subgroups. Clin Cancer Res 2015;21:184–92.

41. Wolff JE, Sajedi M, Coppes MJ, Anderson RA, Egeler RM. Radiationtherapy and survival in choroid plexus carcinoma. Lancet 1999;353:2126.

42. Phoenix TN, Patmore DM, Boop S, Boulos N, Jacus MO, Patel YT, et al.Medulloblastoma genotype dictates blood brain barrier phenotype. Can-cer Cell 2016;29:508–22.

43. CBTRUS. Statistical Report: primary brain tumors in the United States,1995–1999. Hinsdale, IL:Central Brain Tumor Registry of the UnitedStates; 2006.

44. Pauwels B, Korst AEC, Lardon F, Vermorken JB. Combined modalitytherapy of gemcitabine and radiation. Oncologist 2005;10:34–51.

45. Weinberg BA, Philip PA, Salem ME. Evolving standards of carefor resected pancreatic cancer. Clin Adv Hematol Oncol 2017;15:141–50.

46. Perrin S. Preclinical research: make mouse studies work. Nature 2014;507:423–5.

47. Begley CG, Ellis LM. Drug development: raise standards for preclinicalcancer research. Nature 2012;483:531–3.


Nimmervoll et al.




2018;24:1654-1666. Published OnlineFirst January 4, 2018.Clin Cancer Res Birgit V. Nimmervoll, Nidal Boulos, Brandon Bianski, et al. Establishing a Preclinical Multidisciplinary Board for Brain Tumors

Updated version

10.1158/1078-0432.CCR-17-2168doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2018/01/04/1078-0432.CCR-17-2168.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/24/7/1654.full#ref-list-1

This article cites 46 articles, 11 of which you can access for free at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/24/7/1654To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-17-2168

http://clincancerres.aacrjournals.org/content/suppl/2018/01/04/1078-0432.CCR-17-2168.DC1

http://clincancerres.aacrjournals.org/content/24/7/1654.full#ref-list-1

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/24/7/1654


establishing a preclinical multidisciplinary board for brain tumors · minimum time–concentration...

Documents