Oncology management of CNS tumours Neil Burnet

1
Oncology management of CNS tumoursNeil Burnet

• University of Cambridge Department of Oncology
  Oncology Centre, Addenbrookes Hospital

ECRIC CNS study day 7th April 2009
2
Introduction

• Treatment modalities for cancer
• What data do oncologists want?
• Examples of uses of Registry data

3
(No Transcript)
4
Cancer treatment modalities
5
Cancer treatment modalities

• Modalities
• (Surgery)
• Radiotherapy
• Chemotherapy
• Consider efficacy
• Consider costs

6
Oncology management
7
(No Transcript)
8
Radiotherapy

• Radiotherapy is an anatomical treatment
• Treats a specific area
• Localising the tumour target is crucial
• Imaging is key
• Better localisation better outcome
• Localising normal structures allows avoidance

9
CT the technology advance

• Late 1970s 1980s 2003

10
Glioblastoma imaging

• T1

• T1 Gd contrast

• T2

MR (magnetic resonance) imaging
11
Radiotherapy

• Immobilise the patient
• Relate today's patient position to tumour imaging

12
Radiotherapy

• High precision positioning
• Relocatable stereotactic frame

13
Radiotherapy
14
Radiotherapy imaging
CT MRI
15

• GBM planning
• Using CT MR together

MRI CT
16
Radiotherapy imaging

• Post-op planning CT

• Pre-op CT

17
Target volume delineation
18
Radiotherapy

• Planning and delivery technology now very
  different
• Old square planning
• Was conventional in 1960s 1990s
• Conformal (dose conforms to shape of target in
  3D)
• Ultra-conformal (includes concave shape)
• known as IMRT (intensity modulated radiotherapy)
• 21st century technology

19
Treatment volumes compared
Conformal
Ultra-conformal IMRT
Square plan
20

• Old square planning

21

• Some shielding with lead blocks

22
Treatment volumes compared
Conformal
Ultra-conformal IMRT
Square plan
23
Conformal RT plan
24
IMRT plan (TomoTherapy)

• Ca nasopharynx
• 68 Gy to primary (34)
• 60 Gy to nodes (34)
• Cord dose lt 45 Gy
• No field junctions
• No electrons

25
IMRT plan

• Skull base meningioma
• Shaping of dose around optic nerves and chiasm
• Tumour 60 Gy
• Optic chiasm 50 Gy

26
(No Transcript)
27
Radiotherapy dose

• Biological effect depends on
• Total dose
• Number of fractions
• (Dose per fraction)
• Overall treatment time

Complex relationship
28
Radiotherapy dose

• Single fraction
• Very destructive
• Known as radiosurgery
• Must physically avoid normal tissue
• Multiple fractions
• Spare normal tissue
• Enhances therapeutic radio
• Allows treatment including normal tissue

29
RT dose and fractions

• For a given dose, and overall time, biological
  effect depends on number of
• Actually depends on dose/

30
Chemotherapy

• Use in accordance with NICE Guidelines
• At first presentation, with (surgery ) RT
• Temozolomide
• Also at relapse
• PCV
• Monitor
• Blood count, nausea, liver function ( other s/e)
• Progression

31
Chemotherapy

• Most chemo for CNS tumours is oral
• Temozolomide
• Invented in UK
• Revolutionised treatment of GBM

32
RT TMZ for GBM
EORTC Randomised trial results
Plt0.001
33
Cancer cure and cost
34
Cancer cures by modality

• References
• SBU. The Swedish council on technology assessment
  in health care Radiotherapy for Cancer. 1996
• Cancer Services Collaborative 2002

35
The Cancer Reform Strategy
Prof. Mike Richards 2007
36
Effectiveness and cost

• cures of cancer Ratio
  care cost
• Radiotherapy 40 5 8.0
• Chemotherapy 11 18 0.6
• Surgery 49 22 2.2

37
What data do oncologists really want?
38
What data do oncologists really want?

• What data do oncologists really want or need?
• Types of CNS tumour
• Prognostic factors
• Treatment intent
• Treatment details
• Dates

39
Tumour types in oncology clinic

• Note 20 with benign tumours

40
CNS tumour types - 1

• Glial tumours
• Astrocytoma (inc Pilocytic Juvenile Pilocytic)
• Oligodendroglioma
• Oligo-astrocytoma
• Glioblastoma (GBM)
• Ependymoma ( subependymoma)
• Meningioma
• Pituitary adenoma Craniopharyngioma

41
CNS tumour types - 2

• Vestibular schwannoma (aka acoustic neuroma)
• Medulloblastoma
• Germinoma teratoma
• Lymphoma
• Neurocytoma Ganglioglioma
• Pineoblastoma
• Primitive neuro-ectodermal tumour (PNET)
• (Chordoma chondrosarcoma)
• (Metastases)

42
CNS tumour types - 3

• Many tumour types
• Prognosis varies enormously
• Survival from days to weeks to cure
• Affected by tumour type
• Grade (ie how malignant)
• Essential to know detail
• Detail must be collected

43
Grade affects prognosis

• High grade glioma
• Grade III
• Grade IV GBM
• - Surgery RT only
• - Radical treatment
• - Addenbrookes data

44
Grade affects prognosis

• Histology is not the only tumour feature which
  affects outcome

45
Radiotherapy Oncology 2007 85371-378

• Radiology adds to pathology grade
• Need to include information from imaging

46
What data do oncologists really want?

• Prognostic factors
• Age
• Performance status
• ? Size
• Extent of surgical resection (hard to evaluate)
• Treatment intent
• Radical
• Palliative

47
What data do oncologists really want?

• Treatment intent
• Might be clear from treatment
• GBM RT 60 Gy (30) radical
• 30 Gy (6) palliative
• Need to know if intent changes
• eg due to progression

48
Radiotherapy details

• Area treated
• Total dose
• Number of fractions
• Overall treatment time
• Dates
• Time (delay) to start RT
• Overall time (duration) of RT

49
Chemotherapy details

• Drug(s)
• Dose
• Number of cycles given
• Dates

50
(No Transcript)
51
Examples of Registry data use

• Measuring disease burden - AYLL
• GBM outcome
• Modelling chemotherapy use

52
Measuring disease burden
1

• Simple mortality figures do not tell the whole
  story
• Other measures show alternative aspects of
  mortality
• Burden on society
• Burden to the individual affected
• With particular thanks to Peter Treasure at ECRIC

53
Measuring disease burden

• Method
• Detail deaths from specific tumour type
• Compare to standardised matched population
• Sum the difference

54
Measuring disease burden

• CNS tumours
• 2 of cancer deaths simple mortality
• 3 of the years of life lost - YLL
• YLL shows the burden on society

55
Average Years of Life Lost

• Divide YLL by number of affected patients
• Average Years of Life Lost AYLL
• AYLL shows the burden to the affected person
• Easily understood measure, including by patients
• CNS tumours account for 20 years of lost life
• This is higher than any other adult tumour type

56
Average Years of Life Lost
57
Measuring disease burden

• CNS tumours
• 2 of cancer deaths
• 3 of the years of life lost YLL
• 20 years of lost life per individual - AYLL

58
Average Years of Life Lost

• In the 2007 Cancer Reform Strategy reference made
  to the poor overall outcome of brain CNS
  tumours in terms of AYLL
• Encouraging that alternative measures of
  mortality are being acknowledged by the
  government
• UK Government Department of Health (2007)
  http//www.dh.gov.uk/en/Publicationsandstatistics/
  Lettersandcirculars/Dearcolleagueletters/DH_080975

59
Measuring disease burden

• AYLL is an effective measure of disease burden to
  the affected person
• AYLL has other uses
• Compare disease burden with research spending
• AYLL does not match NCRI research spending
• The mis-match is most extreme for CNS tumours

60
Average Years of Life Lost per affected patient
versus NCRI spending
Burnet et al. Br J Cancer 2005 92(2) 241-5
61
GBM outcome
2
62
GBM outcome

• GBM traditionally terrible
  outloook
• Addition of temozolomide (TMZ) chemotherapy has
  transformed the outlook
• Can we reproduce trial results?

The scream Edvard Munck
63
TMZ RT for GBM
EORTC Randomised trial results
Plt0.001
64
TMZ RT for GBM
Addenbr RT alone
65
TMZ RT for GBM
Addenbr RT TMZ Addenbr RT alone
66
TMZ RT for GBM
Addenbr RTTMZ
Plt0.001
67
GBM outcome

• Our results match the international trial
• Endorsement of our treatment pathway
• Good news for patients !

Patient photo
68
Modelling chemotherapy use
3
69
Modelling chemotherapy use

• TMZ chemo combined with RT ( surgery) has
  revolutionised the outcome for patients with GBM
• TMZ is given in 2 parts
• Concurrent daily with RT
• Adjuvant for 6 cycles after RT
• Are both parts of value?

70
TMZ treatment schema

• Chemo-RT programme with temozolomide (TMZ)

RT
TMZ

• Component 2
• Adjuvant
• 5 days every 28, x 6 cycles

• Component 1
• Concurrent with RT
• Daily for 42 days

0 6 10 14 18
22 26 30 34
Week
71
Modelling chemotherapy use

• Build model of patient survival
• Allow treatment with RT and with chemo
• Fit model to Kaplan Meier survival curves to
  derive values for tumour growth and response to
  treatment
• Test
• TMZ RT concurrent
• RT followed by TMZ adjuvant

72
EORTC trial Model - RT concurrent TMZ
RT concurrent TMZ near perfect fit
73
Modelling chemotherapy use

• RT concurrent TMZ produces near perfect fit
• Suggests concurrent TMZ is the effective
  component
• Suggests adjuvant TMZ may not add anything
• Omitting 6 cycles of adjuvant TMZ would
• Spare toxicity
• Improve QoL (likely) - finish treatment 6/12
  earlier
• Save money

74
Modelling chemotherapy use

• Incidence of GBM
• 33 cases per million population per annum
• Cost of TMZ 1 course
• Concurrent 3900
• Adjuvant 7100
• With thanks to
• David Greenberg Peter Treasure,
• Eastern Cancer Registration Information Centre
  (ECRIC), Cambridge
• Brendan OSullivan,
• Chemotherapy Pharmacist, Addenbrookes Hospital

75
Modelling chemotherapy use

• UK
• Population 60 m
• GBM cases (33 x 60) 1,980 p.a.
• GBM patients treated radically 50
• Number requiring TMZ 990 p.a.

76
Modelling chemotherapy use

• UK
• Population 60 m
• GBM cases (33 x 60) 1,980 p.a.
• GBM patients treated radically 50
• Number requiring TMZ 990 p.a.
• Cost TMZ 11 m p.a.
• Saving by using only
  concurrent TMZ 7 m p.a.

77
Improving survivorship
Patient photo
Photo of patient and family

• AW on the beach

• AS at Christmas

78
Acknowledgements

• Colleagues
• Sarah Jefferies
• Raj Jena
• Fiona Harris
• Phil Jones

• Peter Treasure
• Norman Kirkby
• Lara Barazzuol
• EORTC

• National Institute for Health Research (NIHR)
  Cambridge Biomedical Research Centre
• RJ is supported by The Health Foundation, UK
• NFK was supported by an EPSRC discipline-hopping
  grant

79
(No Transcript)