The non Surgical Management of Paediatric

Low Grade Gliomas

Eric BouffetGarron Family Chair in Childhood Cancer Research

The Hospital for Sick ChildrenToronto

Low grade gliomas

• Most common paediatric brain tumour• Includes grade I and grade II astrocytic

tumours (aka: benign gliomas, low grade astrocytomas…)

• Can arise anywhere in the CNS

All LGG! Paediatric low grade glioma is not a single entity

Paediatric low grade glioma is not a single entity

Low Grade Glioma in children

0

5

10

15

20

1 3 5 7 9 11 13 15 17

Age

num

ber

OPG

PF

Pathology of Paediatric Low Grade Gliomas

• Grade 1 or Pilocytic• Grade 2

– Pilomyxoid– Fibrillary– Others

Biphasic Pattern

Smear

Slide

Rosenthal Fibers

Vessels in JPA

Predisposing conditions

• Neurofibromatosis type 1

JPA

• Tuberous sclerosis

SEGA

Behaviour of paediatric LGG

• Highly variable• Reasons still unclear • Can even disseminate• However, with some exceptions (thalamic

tumours), do not show malignant transformation (unlike adult LGG)

Treatment of paediatric low grade gliomas

• Complex equation!

Site * Operability * histology =(behaviour)4/3

Management of paediatric LGG

• Surgery +++• Surgery is curative for the surgical lesions

(cerebellar astrocytomas) • Surgery is still the mainstay of treatment for

most other LGGs

Low grade glioma, cerebellum

Long history (several months)

Mostly vomiting and headaches

Well circumscribed tumour

Rim enhancement

Neuro-navigation

Fonctionnal Mapping

Tectal Tumors in Children

• CSF diversion best accomplished by endoscopic third ventriculostomy

• Excellent results even in young children

• No other treatment

Surgery for low grade glioma in critical areas

• Optic pathways• Brainstem• Spinal cord• Thalamus • Various schools

– From observation to aggressive surgery– Be aware of the competition (shopping

around)

Blind, DI, swings with temperature, Hyperphagic,behaviouralchanges

Surgery for hypothalamic gliomas

Management of paediatric LGG

• Main issue:– Management of

unresectable low grade gliomas

– Management after biospy/incompleteresectionObservation (betting game)Treatment (Which one? When? Why?)

Post resection 3 months later2001

4 year-old, NF1

2002

No change in vision

No treatment

2003

Treatment or not?

2014

Treatment or not?

2006

6 year-old, neck plexiform neurofibroma

2007

6.5 year-old

Second case (NF1) Treatment or not?

2006

6 year-old

2007

6.5 year-old

August 2001 August 2011

Treatment or not?

August 2000 December 2001January 2000

Treatment or not?

Non surgical management of paediatric LGG

• Radiation and chemotherapy: an unsettled and absurd competition

• Traditionally, radiation is the standard treatment• But, who would systematically use up-front

radiation in a population of patients with a survival rate > 95% at 10 years and a close to normal life expectancy?– Risk of stroke (RR: 70 at 15 years, CCSS)– Risk of secondary brain tumour (RR: 6.7 at 5 years,

SEER)– Endocrine deficits– Decline in cognition (younger ++)

LGGTarget Volumes

Gross Tumour Volume (GTV)

Clinical Target Volume (CTV)

Planning Target Volume (PTV)

Outcome for All Clinically Relevant PLGG Patients by Upfront Radiation Therapy

Outcome for All Clinically Relevant PLGG Patients by Upfront Radiation Therapy

p<0.0001

36/717

46/154

Long Term OS by Tumour Location

p<0.0001

• Thalamic tumors demonstrate unique pattern of very late death (> 10 years of follow up)

Thalamic

Brainstem

Management of paediatric LGG

• Chemotherapy: – After failure of radiation (historically)– Then for young children only

• 5 year old (SIOP)• Then < 10 year old (North America)

At progression(after partial resection)

12 month later

12 month of vincristine-carboplatin

Aug 2001

Dec 2001

Feb 2003 (end of treatment)

Feb 2003:

start chemo

May 2003 blind

Jan 2005:

end chemo

2 ½ year old child, no NF1

Management of LGG: When should we treat?

• COG recommendations (COG 9952)

• Still unclear: – Who are the candidates for upfront treatment– What is a risk of neurologic impairment

Why?

• Aims of treatment – Response– Visual preservation (optic pathways)– Delay of radiation treatment– Cure/Survival– Minimize morbidityLGG: a chronic disease?

Which treatment?

Are there better recipes?Should we take into account other factors?

Chemotherapy for paediatric LGG

• First publication in 1976 (pre CT era): broad phase II study of vincristine inrecurrent brain tumours

• Then in 1988 Vincristine-actinomycin (Packer)

• In 1993: Vincristine-carboplatin (Packer)• In 1997: TPCV (Petronio-Prados)

1997: Update on Vincristine-carboplatin (Packer)

• 78 patients (15 NF1)• Mean age: 3 year• tumour site:

– Hypothalamic/chiasmatic: 58– Brainstem: 12

1997: Update on Vincristine-carboplatin (Packer)1997: Update on Vincristine-

carboplatin (Packer)

NF1

1997: Update on Vincristine-carboplatin (Packer)

Response to chemotherapy

Canadian review of allergic reactions to Carboplatin :

105 patients

Cumulativeincidence of allergicreactions: 41.9%

Lafay-Cousin,Cancer 2007

COG 9952 Carboplatin allergy

– CDDP 90 mg/m2 and etoposide 450 mg/m2 for10 cycles

– 34 patients (29 hypothalamic/chiasmatic – 8 NF1 – median age 45 months)

– 3 year-EFS: 78%, OS: 100%.– Prognostic factors:

• NF1 (no progression)• Age (< 1 year-old 33 vs 87% EFS, less than 5 year-

old 66 vs 100%)• Response not prognostic

– 28% hearing loss

2002: Cisplatin-etoposide (Massimino)

Italian Study Temozolomide(CCG phase II study)

• 5 day schedule (q 4 w)

• 21 patients, 20 evaluable• All recurrent (no information on prior

therapy)• Response:

– 1PR, 9 sustained SD, 10 PD

Temozolomide in Paediatric LGG

Gururangan 2007

SIOP-CCG: which one is the best?

LOW GRADE GLIOMAPATIENTS TREATED WITH CT

Progression Free Survival

N. pts N.failed % of survival at 3 years

155 43 48.7 (34.5 - 62.9)

3 y PFS : CCG (68 %)3 y PFS : SIOP (48 %)

Carboplatin : 560 mg/m2/4 weeks

Vincristine : 1.5 mg/m2/4 week (20)

More transfusion, as many allergic reactions

Carboplatin : 175 mg/m2/week

Vincristine : 1.5 mg/m2/week (46)

(COG 9952: 34 x Vcr)

COG 9952 TPCV versus Vincristine-carboplatin

A = vincristine-carboplatin B = TPCV

Good = NF1, small tumours Bad = young, large tumours

COG 9952 and other studies

Most children will require more than one line of treatment!

EFS is 38%at 5 years

New Recipes

• Vinblastine• Vinorelbine• Bevacizumab based regimens• mTOR inhibitors for SEGA

Vinblastine• Phase II study initiated in October 2000• Eligibility:

• relapse/progression after previous chemo or irradiation

• Schedule: 6 mg/m2/week• First assessment: between week 10 and 12• Duration: 52 weeks

October 2000 March 2001

Jan 2001

Jan 2002

Dec 2002 (end of VBL)

June 2005

Continued GH treatment during vinblastine!

March 2003June 2003 (71%)

Clinical progressionOff study?

Feb 2004 Feb 2005 (end of treatment)

Diagnosis7 year old

(largechiasmatic

glioma)

Progression 5 months after completion of

vincristine-carboplatin

May 2007

Phase II study of vinblastine

9952

First line38% at 5 years

Vinblastine

Second line43% at 5 years

Admission Vinorelbine – 18 Cycles Observation

Progression

Descontinue the drug

Initial

4 cycles

0

2

4

6

8

10

12

14

nov/09dec/10jan/10 feb/10mar/10

__ weight

4y boy, Diencephalic S.nystagmus, vomiting and

no weight gain

Initial

4 cycles

8 cycles

4m girl, diffuse brainstem tumor;swallowing difficulties and mechanical

ventilation dependent; biopsy

What can we learn from previous experiences

• Most children will need more than one treatment

• So, oncologists need to think ahead of time• Most unresectable paediatric low grade

gliomas cannot be cured with one shot• This is

A CHRONIC DISEASE

First and second line of chemo

• No clear evidence of acquired resistance to chemotherapy• Event free survival similar between first and second line chemotherapy• EFS is in the range of 40% at 5 years.

Scheinemann et al, PBC 2011

Additive/cumulative toxicity• Carboplatin/vincristine-carboplatin

– Allergic reactions• TPCV

– Hematological toxicity– Risk of secondary cancer– Infertility

• Etoposide-cisplatin– Hearing loss– Risk of leukemia

• Temozolomide– Still poorly known (second cancer?)

• Vinblastine, vinorelbine– No long term toxicity

• Bevacizumab-based– Hypertension, kidney problems, growth issues

Cost

• One year of

– Vincristine-carboplatin: $ 1650 (CCG schedule)– Temozolomide: $ 19320 (12 x 5 d cycles)– Vinblastine: $ 1090 (weekly)– Vinorelbine: $ 2300 (weekly)– Bevacizumab based: $ 60,000 (biweekly)

Fishing for the magic bullet

1 CR, 4 PR (573 patients) LGG: 35 patients, 2 PR

The BRAF-KIAA1549 (B-K) fusion gene

The BRAF-KIAA1549 (B-K) fusion gene

trametinib

trametinib

Avery et al, JAMA Ophthalmol 2014,

150 paediatric patients, to start in 2015

Phase I study

• 51 patients (2-21, median 10 year-old)• Dose levels: 15-116 mg/m2/day x 21 days• 2 objective responses (one thalamic astrocytoma,

one OPG)• 12 month PFS for patients with LGG was 67+13%

Rationalefor

ACNS1022

trametinib

nn Neurol 2006• Rapamycin causes regression of astrocytomas in TSC

patients• 4/5 paediatric patients

– 3, 5.5, 14.5, 15 years– Headaches, nystagmoid eye movements, change in mental

status– All patients responded to Rapamycin with decreased SEGA

sizes (3-5 months)– 23x20 18x13 mm– 10.2x12.7 7.3x9.9 mm– 13.7x23.4 8.1x13.7 mm– 66x50 58x42 mm

> 50%

mTOR pathway

Patient

2 y 9 y

Patient 1

3 months after Rapamycin

Patient

End of treatment 3 months later

Everolimus for SEGA in TSC

• Prospective, open-label phase I/II• Jan 2007 – Dec 2008• N=28 (17 male) • Median age 11 years (3-34 y)

• Dose 3 mg/m2 achieving trough levels 5-15 ng per milliliter

Darcy A. Krueger et al, NEJM 2010

EXIST-1 Study Accrual August 2009 -September 2010

EIAED = enzyme inducing antiepileptic drug.aEverolimus starting dose 4.5 mg/m2/day and adjusted to trough level of 5-15 ng/mL. Dose could be adjusted in cases of toxicity.ClinicalTrials.gov identifier NCT00789828.

2:1

Everolimus target trough 5-15 ng/mLa

n = 78

Placebon = 39

RANDOMIZE

Crossover at SEGA progression

Treatment until SEGA progression or unacceptable toxicity

Stratified by EIAED use

Progressive SEGAassociated with TSC (N =117)Definitive TSC per Gomezcriteria1 target SEGA lesion 1 cmin diameter

Serial SEGA growthconfirmed by imaging ornew/worseninghydrocephalus

SEGA Response Rate in Subgroups

Difference in SEGA response rate

Stratum

Sex

Age

All patients (N = 117)

with EIAED (N = 22)

without EIAED (N = 95)

Male (N = 67)

Female (N = 50)

<3 years (N = 20)

3 – <18 years (N = 81)

18 years (N = 16)

-20 0 20 40 60Placebo EverolimusIn favor of

EIAED = enzyme inducing antiepileptic drug.Exact 95% confidence interval obtained from the exact unconditional confidence limits.

27/78 (34.6)4/15 (26.7)23/63 (36.5)

12/49 (24.5)15/29 (51.7)

3/13 (23.1)

21/55 (38.2)

3/10 (30.0)

Everolimus0/39 (0.0)0/7 (0.0)0/32 (0.0)

0/18 (0.0)0/21 (0.0)

0/7 (0.0)

0/26 (0.0)

0/6 (0.0)

Placebo

Response n/N (%)

Franz DN, et al, Lancet, 2013

EXIST-1 Results

100%

86%

Weidman et al, PBC 2015

inefficient

Dosesbelow 2.5

mg/m2

inefficient

Everolimus in Paediatric LGG

• Kieran, SIOP 2013• 23 patients with recurrent LGG (3-17 year-

old)• 2.7 previous treatments• 4 PR, 13 SD• “Incorporation of everolimus into frontline

LGG therapy is being proposed.”

trametinib

BRAF Inhibitors

• Several BRAF inhibitors undergoing testing– GSK: Dabrafenib– Roche/Genentech: Vemurafenib– Genentech: GDC 0879– Novartis: Encorafenib (LGX818)– Cellagen: PLX 4720

BRAF Mutation in paediatric and adult gliomas

Rush S,JCO 2013

Vemurafenib in refractory ganglioglioma

Start of therapy(dabrafenib) 2 months later

Previous surgery, chemo and radiation

Start of Dabrafenib 2 months later April 2015 June 2015

Compassionate BRAF V600E inhibitor

trametinib

MEK Inhibitors

• Several MEK inhibitors undergoing testing– Astra-Zeneca: Selumetinib– GSK: Trametinib– Novartis: MEK162– Pfizer: PD0325901– Roche: GDC-0973

Selumetinib in LGG (ASCO 2014)

• 39 patients mean age at study entry 13.1 (5.6-20.8)• 56% JPA, 23% Gliomas, others.• Recommended dose: 25 mg/m2/dose

[TITLE]

Selumetinib in LGG• Phase II ongoing: 6 strata (PBTC 029B)

MEK inhibitor PD184352Microtublule destabilizing agent TZT 1027

Questions

• Assessment of response– No consensus– Urgent need for harmonization– Issue of “pseudoprogression”– Time of response

Start of therapy(dabrafenib) 2 months later

CompleteResponse

Stabledisease

No consensus: T1? T1 +C? FLAIR? T2?

Skrypek et alPBC 2014

Duration of treatment

• Vincristine-carboplatin (1993): 84 weeks• Reduced to 60 weeks in 9952• Currently patients on MEK or BRAF

inhibitors for more than 2 years• Long term toxicity of these treatments

unknown

Next steps

• Medical treatment of paediatric LGG is complex

• Still a role for chemotherapy• Targeted treatment will be the future for

most paediatric LGG• Perhaps more role for early biopsies when

targeted treatments are widely available