NeuroOncology Branch, The National Institute of Health NIH

1
Neuro-Oncology Branch, The National Institute of
Health (NIH)

• A collaborative, joint effort by the National
  Cancer Institute (NCI) and the National Institute
  Neurologic Disease and Strokes (NINDS) to advance
  basic science discovery of brain tumors and
  develop improved, novel therapeutics for children
  and adults with brain and spinal cord tumors.

2
Tumors of The Central Nervous System

• Primary CNS Tumors
• Bimodal incidence 0-4 y.o. and then begins to
  rise in late adolescence
• 2nd and 4th leading cause of cancer- related
  mortality n patients lt18 y.o.and lt34 y.o,
  respectively.
• Secondary CNS Tumors
• 25 of all system cancer patients will have
  clinically apparent CNS disease 40-60 will have
  pathologic (autopsy) evidence of CNS metastases.

3
Neuro-Oncology Topics for Discussion

• Primary CNS Tumors
• Astrocytomas
• Oligodendrogliomas
• Primary CNS Lymphoma
• Secondary CNS Tumors
• Brain metastases
• Neoplastic Meningitis
• Epidural Spinal Cord Compression

4
Astrocytoma

• Most common primary brain tumor of adults.
• Tumors derived from astrocytic lineage
• Genetics
• Low grade tumors
• p53 mutations
• PDGF (ab) and PDGFR (alpha and beta)
  over-expression
• High grade tumors
• EGFR mutation/amplification/over-expression
• PTEN deletions
• RB deregulation (p16 deletion, cdk4
  amplification, cyclin D1 over-expression,
  transcriptionally silent rb)

5
Astrocytoma Predisposing Factors

• Cranial radiation
• Vinyl Chloride (? polycyclic aromatic
  hydrocarbons)
• NCI Case-Control Study
• NOT cell phones!!
• Polymophisms in certain glutathione-S-transferase
  and cytochrome P450 isoenzymes
• Inverse relationship to allergic and autoimmune
  disease

6
Astrocytomas Grading

• Divided into low and high grade tumors.
• High grade tumors include anaplastic astrocytoma
  (AA) and glioblastoma multiforme (GBM).
• Multiple grading schemas
• Kernohan
• Three Tier System
• Daumas - Duport
• WHO

7
Astrocytoma Daumas-Duport Classification
Endothelial proliferation, nuclear pleomorphism,
mitoses, necrosis
_at_ Daumas-Duport, 1989
8
Data Analysis of the First 87 Gliomas
--Log-ratios for each microarray was normalized
by the locally robust smoother lowess --Hierarchic
al clustering was used to analysis gene
expression profiles --2-sided t-tests were used
to identify the differentially expressed genes
with plt0.001 --A univariate Cox model was used to
test for an association between gene expression
and survival --Both unsupervised and supervised
methods were used for predicting survival
9
Multidimensional Scaling
GBM Oligo Astro Mixed
10
216 Genes Associated With Survivalof Glioma
Patients at 0.001 Level
Graymissing data Purple on bar graphpatients
still alive Length of barnumber of days
survived
-2 -1 0 1 2 (log
base 2)
11
Patterns of Tumor Growth Within the CNS
Type III
Type I
Type II
Solid tumor with infiltrating tumor cells
Infiltrating tumor cells
Solid tumor
From Dumas-Duport
12
Astrocytoma Natural History
Low grade Tumors

• Presentation Seizures, headaches, subtle
  neurological signs
• Radiographic
• CT Scan-Low attenuation, poorly defined,
  non-contrast enhancing.
• MRI-Increased T2 signal, non-gadolinium enhancing
• Natural History Infiltrative, slowly
  progressive 60 of tumors transform to higher
  grade gliomas

13
Astrocytoma Treatment
Low Grade
Grade I
Grade II
Shaw, 1988
Since 1988
New grading systems few grade 1 tumors
Better intra-operative guidance more
extensive surgery
14
Astrocytoma Treatment
Low Grade

• Radiotherpy

Grade I
Grade II
Shaw, 1988
Dose and timing of XRT remain uncertain
Long term neuro-cognitive toxicity remains a
concern
Highly conformal XRT becoming the standard
15
Astrocytoma Natural History
High Grade

• Presentation Headache, seizures, acute global
  and/or focal neurologialc signs
• Radiographic
• CT Scan low attenuation, contrast enhancing
• MRI increased T2 signal representing both tumor
  and edema gadolinium enhancing.
• Rapidly growing, infiltrative, destructive. Death
  from local cerebal destruction, edema, and
  increased intracranial pressure leading to
  herniation.

16
Why did we not find a difference between GBMs and
astrocytomas Could GBMs be a heterogeneous
group
Two Molecular Pathways to GBM
Cells of origin
-EGFR amplification/mutation
(7p12) -INK 4A loss (9q26) -LOH 10 -PTEN
mutation (10q24) -RB alterations (13q13)
Astrocytoma Anaplastic Astrocytoma
-PDGF, FGF2 overexpression -P53 mutation -CDK4
amplification(12q13) -RB loss -LOH 19q -LOH
10q -PTEN mutation
Primary GBM
Secondary GBM
Do they have distinct gene expression profiles?
17
Mean89.26
(n8)
(n8)
Mean4.35
Mean4.81
Mean0.42
Mean4.9
Mean5.69
(n24)
(n27)
(n5)
(n16)
(n6)
18
63 Genes (gt 2 x) Differentially Expressed Between
High/Low EGFR Expression GBMs at 0.001 Level
High
Low
Graymissing data
-2 -1 0 1 2 (log
base 2)
19
MRI of a Glioblastoma
20
Infiltrative Nature of Malignant Gliomas Whole
Mount
21
Infiltrative Nature of Malignant Gliomas
Microscopic
22
Astrocytoma Surgical Treatment
Advantages of Large Resection

• Accurate histological grade
• Palliation of mass effect
• ? Improved quality of life
• Decreased dexamethasone requirement
• ? Impact on overall survival
• Retrospective studies say yes
• No prospective, randomized trials

23
Astrocytoma Radiation Therapy

• A series of randomized trials (BTSG, RTOG, EORTC)
  in the 1980s demonstrated clear survival
  advantage
• 5940 cGY in 33 fxs (or equivalent) considered
  standard
• Involved field XRT (rather then whole brain), as
  defined by MRI T2 signal abnormality considered
  standard.

24
Astrocytoma High-Dose Focal Radiation
Phase II studies suggest survival benefit when
used as an adjunct to standard fractionated
XRT.
Randomized RTOG study GBM Fx XRT /-
SRS Results No difference in survival
25
Astrocytoma Treatment
Radiotherapy

• Dose-Toxicity

Derived from Marks and Adler
26
Astrocytoma Adjuvant Chemotherapy

• gt 20 randomized trials
• Results mixed some trials showing benefits
  others not
• Trials included very heterogeneous patient
  population
• Different tumor histologies
• Different grade tumors
• Patient with different prognoses
• Relatively small trials
• Three meta-analyses have evaluated the data.

27
Meta-Analysis Results
Analysis includes 17 trials and gt 3000 patients
p0.002
(DSL DerSimonian Laird)
28
(No Transcript)
29
Meta-analysis Conclusions

• Chemotherapy following external beam radiation is
  advantageous in adults with anaplastic gliomas.
  Much less benefit is seen in patients with
  glioblastoma.
• The few glioblastoma patients who benefit tend to
  be those that live the longest, suggesting that
  treatment preferentially benefits patients with
  favorable prognostic factors (i.e. young age,
  good P.S., minimal post-op, residual tumor).
• 2 recent meta-analyses (2000, 2002) have
  confirmed findings of benefit of chemotherapy.

30
High-Grade Astrocytomas Adjuvant Chemotherapy
Two recent retrospective studies question
validity BCNU as good as PCV
31
Chemotherapy for Recurrent High Grade Astrocytomas

• Response rates highly variable in literature In
  reality objective response rates 5-30
• Few active agents
• Carboplatin
• Tamoxifen
• CPT-11No!
• Short-lived responses (short TTP)
• Responders may live longer no clear overall
  survival prolongation.
• Few quality of life studies however, progressive
  tumor growth results in rapid loss of QOL. Tumor
  reduction or stabilization may maintain QOL for a
  period of time.

Incudes stable disease
32
Gliadel BCNU Wafers

• Pros
• No systemic toxicity
• ? Survival advantage
• 4-6 weeks for recurrent GBMs
• 4-6 weeks for newly diagnosed GBMs..maybe?????
• Cons
• Needs surgery
• Very expensive
• Increased incidence of wound healing
• May cause necrosis
• ? Optimal pharmacology

33
Temozolomide

• Active metabolite of DTIC
• Crosses the blood-brain barrier
• Oral dosing 200mg or 150mg/m2/x5d q4 weeks
• Well tolerated nausea/vomiting, myelosuppression
• Idiosyncratic myelosuppression

34
Temozolomide

• Two large phase II trials
• Active in anaplastic gliomas (first recurrence)
• 35 response rate
• 5.4 month PFS
• 13.4 month median survival
• Improved HAQOL
• Marginal activity in glioblastoma
• lt10 response rate
• Standard of care for recurrent anaplastic
  gliomas.
• Role in initial therapy uncertain
• Probably equivalent to nitrosourea or PCV for
  high grade gliomas
• Role as a radiosensitizer highly questionable

35
Oligodendroglioma

• 1/10th as common as astrocytic tumors.
• Peak incidence occurs in young to middle aged
  adults accounting for 6 of all intra-cranial
  neoplasms in this age group. Rare in the young,
  or elderly.
• Presentation Seizures
• Natural history highly variable depending on grade

36
Oligodendroglioma Treatment

• Complete surgical resection optimal
• Radiation therapy of questionable benefit
• Anaplastic oligodendrogliomas are chemosensitive
  tumors Active regimens include
• Single agent nitrosoureas
• PCV
• Temozolomide
• Other alkylating agents
• Chromosomal karyotypes may help predict
  therapeutic and clinical outcomes.

37
Anaplastic Oligodendrogliomas
1p loss
1p intact
Group 1 Combined, isolated 1p
19q loss
Group 2 1p loss other
Group 3 P53 mutation
Group 4 No p53 mutation
Response rate 100
100 33
18 Duration response gt31ms
11ms 7ms
5ms Survival from dx gt123ms
71ms 71ms
16ms
Do they have distinct gene expression profiles?
38
1p/19q status of Oligo Patients
LOH
Maintained
nnon-informative lLOH mmaintained ndno data
39
97 Genes Differentially Expressed Between LOH
and Maintained 1p/19q OLIG Patients at 0.001 Level
LOH Maintained
Graymissing data
-2 -1 0 1 2 (log
base 2)
40
Anaplastic Oligodengroglioma Chemotherapy
NCIC Prospective Trial
I-PCV CCNU135mg/m2 PCB 75mg/m2 x 14d VCR
1.4 mg/m2
9 Ineligible Pts
7 wrong path 2 other
Cairncross G. et al 1992
41
Oligodendrogliomas and Mixed Gliomas

• Anaplastic oligos very chemosensitive
• Anapl. Mixed gallons chemosensitive
• Low grade oligo/mixed ? chemosensitive.

Questions

• When to use chemoRx Pre/post XRT, recurrence
• Opitmal Regmen PCV vs IPCV
• Length of therapy

42
Neuro-Oncology Pharmacology

• Hepatic cytochrome P450 isoenzymes
• Detoxification and chemical modification.
• Involved in chemotherapy metabolism.
• Induced by many anti-epileptic drugs (e.g.
  dilantin, tegretol, phenobarbitol
• Patients on enzyme-inducing anti-epileptics have
  significantly altered drug metabolism (e.g. MTD
  of Taxol and CTPT-11 gt 3 fold higher).

43
Primary Central Nervous System Lymphoma (PCNSL).

• Recognized entity for gt 80 years
• Nature of tumor initially unclear
• Perithelial Sarcoma
• Reticulum Cell Sarcoma
• Microglioma

44
PCNSL Epidemiology

• Neurologic Involvement of the CNS in systemic
  NHL 5-29
• PCNSL accounts for 1-2 of NHL
• Large increase in incidence for both population
  at risk
• Immunocompetent
• Immunocompromised
• Aids
• Organ allograft recipients
• Congenital

45
PCNSL Pathophysiology

• Focal Lesion most common presentation others
  include diffuse, uveal, leptomeningeal, and
  intramedullary.
• Infiltrates normal brain diffusely.
• Spreads along CSF pathways.
• Rarely spreads outside the CNS.

46
(No Transcript)
47
(No Transcript)
48
PCNSL Patient Characteristics
Data from Cohort of 1107 Patients
All cases reported in the literature between
1980-1993 (72 published series)
49
PCNSL Clinical Presentation
50
PCNSL Radiographic Presentation PCNSL Diagnosis
No data
51
(No Transcript)
52
(No Transcript)
53
PCNSL Diagnosis

• Lumbar Puncture (if safe)
• 80 of patients have a CSF pleocytosis
• lt 10 have obvious lymphomatous involvement of
  CSF.
• Surgery
• Systemic staging
• CNS Axis staging
• Spinal evaluation (MRI)
• Ophthalamic Slit Lamp Examination
• 10-20 of patients develop lymphomatous uveitis.

54
(No Transcript)
55
(No Transcript)
56
(No Transcript)
57
PCNSL Histology
Relevance of histological subtypes unclear
58
PCNSL Treatment
Surgery

• For Diagnostic purposes only extensive resection
  contraindicated.
• Histologic diagnosis may be difficult from a
  stereotactic biopsy.

59
PCNSL Radiotherapy

• Historically standard Tx
• 80 radiographic CR
• Recurrence local /- CSA
• 14-18 month median survival

60
PCNSL Treatment
PCNSL recurrent after radiation is often
chemotherapy sensitive, although few (if any)
are cured.
Drugs should possess anti-lymphoma activity and
have at least some ability to traverse a
partially disrupted, if not intact BBB
Multiple drugs shown to have anti-PCNSL activity

• Methotrexate (high-dose)
• Procarbazine, CCNU, Vincristine
• BCNU

• Ara-C
• Cyclophosphamide, Adriamycin,Vincristine,
  Prednisone (CHOP).
• Decadron

61
PCNSL Survival
62
Pre-Radiation Chemotherapy
RMH
MGH
OHSU
MSKCC
RTOG (88-06)
Blay
63
Pre-Radiation Chemotherapy2003

• MSKCC regimen hdMTX/PCB/VCR/itMTX XRT
  Hi-araC.
• High response rate (90)
• 2040 long term disease-free survival
• High and increasing incidence of neuro-toxicity
• NABTT Regimen HD-MTX (8gms/m2)
• High response rate (90)
• 20-40 long term disease-free survival
• 50 of CRs have not required XRT
• Little neuro-toxicity
• Eye penetration
• Few patients treated
• Shorter follow-up.

64
PCNSL Current Treatment Recommendations

• Most patients should receive pre-radiation
  chemotherapy with a regimen that at least
  contains high-dose methotrexate.
• Radiation therapy following chemotherapy may
  improve survival further although at the risk of
  significant long term neurotoxicity.
• Strongly encourage enrollment on a clinical trial.

65
CNS Involvement by Systemic Cancer

• Brain Metastasis
• Spinal Cord Compression
• Carcinomatous Meningitis

66
Brain Metastases What do we Know?

• 170,000 new cases/yr
• 20-40 of all cancer patients
• 50-60 lung CA 15 breast CA, 11 unknown
• 50 solitary by CT scan 25 by MRI scan

67
Brain Metastases Diagnosis

• Differential Diagnosis of Intracranial
• Space-occupying Lesion
• Infection
• bacterial abscesses, toxoplasmosis, nocardia
• Vascular
• infarction, hemorrhage, A-V malformations
• Primary Brain Tumors
• astocytoma, meningioma, etc.
• Idiopathic
• multiple sclerosis, radiation, necrosis, PML

68
Brain Metastases

• Median survival 4 months but survival is highly
  variable.
• Prognostic factors influence outcome more than
  therapy and thus must guide therapeutic options
• Single vs. solitary cerebral metastases
• Number of cerebral metastases
• Extent/activity or systemic disease
• Age
• Performance status
• Histology
• Neurological functions

69
Brain Metastases Treatment

• Solitary Metastases
• Goals
• 1. Palliate neurologic symptoms
• 2. Decrease future neurologic morbitiy
• 3. Prolong Survival
• 4. Potentially cure selected patients
• Treatment
• 1. Surgery if medically and anatomically
  indicated.

70
Solitary Metastases

• Med
• Study Pt Surv. Ind CNN Local
• (wk) Surv. Death Recur
• Patchell et.al 25 40 38 29 20
• Noordijk et al 32 43 33 35 nd
• Patchell et al 23 15 8 50 52
• Noordijk et al 31 26 15 33 nd
• Phase II Trials 122 56 44 25 14

Surgery XRT
Surgery
SRS XRT
71
Brain Metastases Treatment

• Multiple Metastases
• Goals
• 1. Palliate neurological symptoms
• 2 Decrease future neurological morbidity
• 3. Prolong survival in selected patients
• Treatment
• 1. Whole brain XRT
• 2. Chemotherapy for chemosensitive tumors (i.e.
  breast cancer, small cell lung carcinoma)
• Little intact BBB

72
Leptomeningeal Carcinomatosis

• Occurs in 5-8 of patients with solid tumors
• Most common tumors to metastasize to the
  leptomeninges are small cell carcinoma (9-18),
  and NHL (5-29)
• Other tumors include non-small cell lung, breast,
  melanoma, and G.U.

73
Leptomeningeal Carcinomatosis

• Pathogenesis
• Hematogenous seeding of the arachnoid
• 1/3 of patients will have concurrent epidural
  spinal, or brain metastasis.
• Invasion along Virchow-Robin spaces, and direct
  invasion of cranial and spinal nerves mediate
  neurologic damage.

74
Leptomeningeal Carcinomatosis Clinical
Presentation

• SYMPTOMS SIGNS
• Headaches 33 Reflex assymetry 71
• Lower Motor Weakness 38 Weakness 60
• Parasthesias 34 Mental Status Change 31
• Back/Neck Pain 26 Sensory Loss 27
• Radicular Pain 33 CN III Paresis 20
• Diplopia 20 CN VII Paresis 17
• Mental Status change 17
• from Wasserstrum, 1982

75
Carcinomatous Meningitis

• Diagnosis
• MRI
• Cerebral Spinal Fluid
• Rarely myelography

76
Leptomeningeal Carcinomatosis

• Goals of Treatment
• Destroy circulating and perineural neoplastic
  cells to stabilize or improve neurologic
  dysfunction.
• Prolong survival in specific tumor types
  (lymphoma, leukemia, breast carcinoma).

77
Leptomeningeal Carcinomatosis
Treatment

• Radiation
• Involved field for acute cranial or spinal nerve
  involvement
• No cranial-spinal
• Chemotherapy
• Intra-CSF Intrventricular vs. Intrathecal (MTX,
  ara-C, Depocyte, Thiotepa)
• Ease of administration
• Patient Comfort
• Optimal CSF drug distribution
• Systemic (?)

78
Epidural Metastases and Spinal Cord Compression

• Symptomatic epidural metastases occur in 5 of
  all patients with disseminated cancer.
• Frequency of asymptomatic metastases much higher.
  
• Most common cancers to cause cord compression are
  breast, lung, prostate, lymphoma, myeloma, and
  kidney.

79
Spinal Cord Compression
Symtoms

• Back and/or radicular pain occurs in almost all
  patients at presentation.
• 75 of patients present with motor weakness
• 50 of patients have sensory or autonomic loss at
  the time of diagnosis.

80
Spinal Cord Compression
Diagnosis

• Plain Radiographs of spine
• High Predictive value (gt80)
• Myelography/CT scan with metrizamide
• Historical gold standard
• MRI with gadolinium
• Noninvasive
• Intramedullary tumors
• Visualize entire vertebral column
• New gold standard

81
Spinal Cord Compression
Treatment

• Dexamethasone
• Low dose vs high dose vs no dose
• Surgical Decompression
• Laminectomy
• Anterior resection
• Radiotherapy
• Historically considered the treatment of choice.

82
Spinal Cord Compression

• Indications for Surgical Interventions
• Chance of restoring/preserving neurological
  function (anterior resection)
• Initial presentation from an unknown neoplasm.
• Previously irradiated field.
• Vertebral instability.
• Progressive neuroloic deteriation during and
  following XRT.
• Intractable pain despite adequate XRT, steroids,
  and analgesics.

The decision to surgically intervene ultimately
depends on an estimate of the morbidity of
surgery versus the potential fpr significant
quality survival.