Radiosurgery in brain tumours
Dr Debnarayan Dutta, MD Consultant Radiation Oncologist Apollo Speciality Hospital, Chennai duttadeb07@[Link]
CNS Tumours
Total number of tumours 132 Total number of malignant glial tumour ~ 20
WHO Classification. Louis D ; Acta Neuropathol 2007
Radiation therapy
Conventional RT: 1.8-2 Gy/# Majority of the tumours are treated with Conv RT Hypofractionated RT: >2 Gy/# Mainly for palliative treatment Radiosurgery: Single fraction high dose treatment Usually curative intent Fractionated Radiosurgery: Short course high dose treatment Usually curative
Radiosurgery: tools
Gamma-Knife LA based SRS Systems BrainLAB Novalis Trilogy Tomotherapy CyberKnife
Gamma knife
Gamma-knife: 201 Cobalt source
Only for intracranial lesions Rigid/ fixed Single
frame required
fraction treatment
Gamma-knife
Indications
- Small Meningiomas (<3 cm) - Small acuastic schwannoma (<3 cm) - Solitary / oligo brain metastasis with controlled primary (RPA Class I) - Small residual LGG
- AVMs (<3 cm)
- Trigeminal neuralgia (Functional disorder) More than 40 years experience / results with Gamma-Knife
CyberKnife: Unique properties
Highly precise treatment delivery
Motion management method Tumour tracking Dose painting
Excellent dose distribution
Fractionation schedule No rigid fixation
CyberKnife is an extension of Gamma-Knife
CK & GK: Similarity
- Principles of field arrangement - Dose distribution pattern - Multiple isocentre -Treatment principles - Treatment delivery accuracy similar - Delivered dose in single fractions - Intra-cranial indications
Hence, all the indications of GK are indications of CK also
Cyberknife
Indications for single fraction treatment as Gamma-Knife
- Small Meningiomas (<3 cm) - Small acuastic schwannoma (<3 cm) - Solitary / oligo brain metastasis with controlled primary - Small residual LGG
- AVMs (<3 cm)
- Trigeminal neuralgia
- Rec High grade glioma - Craniopharyngioma - Pituitary tumour
More than 40 years experience / results with Gamma-Knife
Cyberknife Vs Gamma-Knife: Dissimilarity
GK Immobilization device Rigid frame Orfit 6MV LA Inverse planning Complex Usually 80-95% May treat multiple fraction CK Comments CK has favorable orfit GK need to replace sources every 5/6 yrs Favorable dosimetry in CK Even neurosurgeons can plan in GK GK: more dose heterogeniety Radiobiology favorable in CK
RT source Co60 Planning No complex planning Planning method Simple Isodose prescription Usually 50% Fractions Single
Tumour size Only smaller lesions can be treated
Energy source Radiation Verification Not possible Indications Only brain lesions
Larger lesions also can be treated in fractionated schedule
Electricity Possible Both extra & intra cranial
Increased indications with CK
GK can work with less electricity Even Intra-fraction movement can be corrected CK more economical
Cyberknife Vs Gamma-Knife: Dissimilarity
Advantage of Inverse planning
GK planning
CK planning
Dose to mesial temporal lobe & Choclea is higher with GK Mean dose to mesial temporal lobe >6 Gy with SRS: IQ decline
Romanalli, Lancet 2009
% of patient with >10% drop in IQ
Left temporal lobe DVH
p=0.39
Volume (cc)
p=0.06
p=0.03 p=0.06
Jalali , Dutta et al IJROBP 2009
PTV margin in brain tumour
CTV-PTV Margin
Systemic Error () Random Error () ICRU 62 Strooms Van Herks
NR only Group:
Ant-Posterior Med-lateral Sup-Inferior 0.1 0.28 0.52 1.36 1.04 1.37 1.05 mm 1.01 mm 1.48 mm 1.15 mm 1.29 mm 2.0 mm 1.20 mm 1.43 mm 2.26 mm
NRF Group:
Ant-Posterior Med-lateral Sup-Inferior 2.24 0.78 0.94 1.28 1.41 1.39 3.14 mm 1.77 mm 1.91 mm 5.38 mm 2.55 mm 2.85 mm 6.50 mm 2.94 mm 3.32 mm
Prospective study
Two different head rest (NR & NRF) 220images (NR 100, NRF 120) Error estimation with 2D EPID
PTV margin: 3 mm.
Budrukkar , Dutta et al, JCRT 2008
Cyberknife Vs Gamma-Knife Vs X-Knife:
CK: Accuracy similar with Gamma-Knife
Treatment delivery accuracy: GK: ~1 mm CK : ~1 mm LA based SRS: 1-2 mm (iso-centric inacurracy; LUTZ test)
PTV margin: CK: <1 mm GK: <1 mm LA based SRS: 1-2 mm
GK/CK
LA based SRS
CK has the accuracy of GK and flexibility of LA based SRS
fSRS
Extended Indications for multiple fraction treatment
- Larger meningiomas (>3 cm) - Larger acuastic schwannoma (>3 cm) - Large solitary / oligo brain metastasis with controlled primary - Larger residual LGG
- AVMs (>3 cm)
- Chordomas - Rec HCC - Craniopharyngioma - Pituitary tumour
Short term data with robotic radiosurgery
New experiences with fSRS
Pre-Treatment
Post-Treatment
- More necrosis with CK than SRT (25Gy/5# Vs 54Gy/30#) - Difficult to have radiological interpretation - Require longer duration of steroid - Associated with more oedema
Outcome measures in benign/ low grade tumours
Radiological response may not be appreciable
Lack of progression is control in low grade/benign tumours Hence, function preservation is the mainstay of assessment of Rx outcome Function assessment:
Neuro-psychological assessment: IQ assessment Neuro-cognitive assessment: LOTCA Activities of daily living: Barthels , FIM FAM Quality of life
ADL in evaluation of efficacy in benign/low grade tumour
(n=38)
Dutta, Jalali JNO 2008
Response with fSRS in benign tumour
Conventional RT lack of progression is usual. In a few patients we have observed regression or complete response
New experiences with fSRS
Radiobiology & dose equivalent may be unpredictable with high dose/Fr Conventional BED calculation may not be appropriate Need to use different methodology for calculation of dose equivalence
60Gy @ 2Gy/Fr equivalence dose
New experiences with fSRS
Low dose region is less with CK compared with LA based SRS
Balaji, Dutta, Mahadev, AROI 2010 (Abstr)
Secondary malignancies: Impact of low dose region
Low dose region is less with CK compared with LA based SRS
Dose factors & sec malignancies
Sec malignancies high with higher 1-10 Gy volume
Coudi 2010
Experiences with SRS/ fSRS
Brain metastasis Acaustic schwannoma AVMs Meningiomas Pituitary tumour Craniopharyngioma Rec HGG New indications
Demography data: Brain tumours
Incidence*
All cases Benign SEX-Male Female Estimated new cases Paediatric All Male Female Lifetime Risk Male Female CBTRUS SEER 14.8 6.4 7.4 14.5 7.6 15.1 5.3 43,800 18,500 4.3 4.5 4.0 0.65% 0.50%
Prevalence#
All cases Benign Malignant Uncertain behaviour CBTRUS 130.8 97.5 29.5 3.8
# (per 1,00,000 population)
Brain metastasis
7-10 times of primary tumour
*(per 1,00,000 person-years)
Brain metastasis: SRS
Problem with Indian Subcontinent
Median age of presentation
Developed Countries* Metastatic brain Tumour Anaplastic astrocytoma Glioblastoma Oligodendroglioma Tata Hospital data**
61 yrs 49 yrs 62 yrs 41 yrs
49.4 yrs 36 yrs 50 yrs 37 yrs
Pituitary adenoma
Meningioma
39 yrs
55 yrs
41 yrs
46.5 yrs
Malignant Tumours: presentation one decade earlier in our data
* SEER and CBTRUS. **Tata Memorial Hospital NeuroOncology registry 2006 Jalali & Datta J Neurooncol (2008) 87:111114
Brain metastasis: WBRT alone
RPA class 1 2 3 Features KPS>70; Age<65; controlled primary; no extracranial disease KPS>70; Age>65; Uncontrolled primary; extracranial metastasis KPS<70 MS (mo) 7.1 4.2 2.3
Gasper et al; 1999
Brain metastasis: SRS/Sx
Prospective studies
MS (mo) Patchel WBRT+ Sx WBRT only 9.2 3.4 p-value 0.01
Vecht
Mintz Andrews Kondriolka
WBRT+ Sx
WBRT only WBRT+ Sx WBRT only WBRT+ Sx WBRT only WBRT+ Sx WBRT only
10
6 5.6 6.3 6.5 5.7 11 7.5
0.04
0.24 0.13 0.22
SRS: Brain metastasis Advantages
Surgery Lesion Effect Histopathology Larger (>4 cm), Non-eloquent area Rapid resolution of mass effect Tumour removed Confirmed Radiosurgery Small, deep lesions, eloquent area Minimally invasive Sterilized Not
Anesthesia
Steroid Follow up
Required
Tapped faster Less intensive
No
longer More
Suh J et al; NEJM 2010
SRS: Brain metastasis
Ideal lesions for SRS
Well defined on imaging (MRI & CT) Spherical or pseudospherical shape Most <4 cm in Max diameter Generally noninfiltrative
Located in grey-white junction
Suh J et al; NEJM 2010
Brain metastasis: fSRS
Prospective studies: Larger tumours
Study Alexender (1995) Median Vol (cm3) 3 KPS 80 Multiple lesions 31% MS (Mo) 9.4
Aucher (1996)
Breneman (1997) Shiou (1997) Shirato (1997) Pirzhall (1998) Kim (2000) Nishizaki (2006)
<4 cm 1.3 >2 cm:36% 2.1 7.2
90 90 60 80 90 80
0%
57% 46% 0% 26% 15% 45%
13
10 11 9 5.5 11 13
Nishizaki; Minim Invas Neurosurg 2006
AVMs
Epidemiology - Account for 10% SAH and 1% of strokes - Autopsy studies show 4-5% incidence in general population - Males: Female 2:1 Presentation - Hemorrhage (50%) usually during 2nd-4th decades - 10-20% risk of death if bleeds - 10-20% risk of long-term disability - Increased risk of re-bleed of 6% during first year after initial bleed - Seizures (25%) - HA (15%) migraine-type - Pulsatile tinnitus
Dose response curve: obliteration rate
3Yr obliteration 15-20 20-25 25-30 45% 55% 75%
5 year 85% 90% 75%
Obliteration after SRS depends upon marginal dose
Flickinger et al.. Rad Onc 2002; 63:347-354.
Complications : AVM Radiosurgery
Persistent neurological toxicity depends upon 12 Gy normal brain volume & location
Flickenger et al. IJROBP, 38(3):485-490,1997.
AVMs: SRS dosimetry
Dose prescription (Isocentre)
Marginal dose ( Gy)
12 Gy normal brain volume (cc)
Obliteration depends upon: marginal dose Complication depends upon: 12 Gy normal brain volume
Radiosurgery in AVMs
Gamma Knife
Accuracy PTV margin Isodose coverage Dose inhomgeniety Normal brain dose Complication probability Obliteration probability Sub-millimeter accuracy ~0-1 mm 50% high high high same
LA based SRS
not 1-2 mm 80-90% less less high same
Cyberknife
Sub-millimeter accuracy ~0-1 mm 80-90% less least Expected to be lower same
Cyberknife: sub-millimeter accuracy of gamma knife & higher dose homogeniety of LA based SRS
SRS in AVMs: Indian data (n=23)
Number of patient referred for SRS Number of patients planned for SRS Number of patients treated with SRS LFU status No deficits Neurological deficit persist Type of Imaging done for Assessment MRI and MRA done at 2 yrs FU DSA Imaging awaited on follow up Last Follow up status on Imaging MRA proven obliteration Obliteration confirmed on DSA No Obliteration on DSA Complication after SRS No complication Temporary worsening Persistent neurological deficit 87 23 21 22 01 15 12 06 15 11 01 18 02 01
Pre-SRS
Post-SRS 2 yr FU
Complete obliteration rate at 2 yrs DSA evaluation 92%
Jalali, Dutta et al. J Cancer Res Ther, 2009
Large AVMs
n Chang (2008) Pollock (2000) 55 10 (23)
Median FU (mo) 36 mo 12 mo
Results OR- 36% 12 Gy Vol dose acceptable
LTNS 15% -
Larger AVMs are treatable without increasing lat e neurological toxicity
Pollock IJROBP 2010
Meningiomas: SRS
- SRS is an option for small meningiomas (Incidental findings or symptomatic ) - Dose: 10-15 Gy; single Fr - Local control rate: 80-90% at 10 yrs - However, now emerging data, larger lesions (para-sagital) / Recurrent meningiomas may be treated with fractionated approach
CK Society website 2010
Atypical/ anaplastic meningiomas: SRS
Craniopharyngioma
Epithelial tumou rising from rathkes pouch remnants 2-5% of all primary intracranial tumours Common age of presentation <20 yrs 5-15% of primary tumour in children Two histopathological types: 1) Aadamantinomatous typemainly occurs in children 2) papillary type- occurs exclusively in adults.
Age & Sex distribution
Increasingly treated with conservative surgery + RT Good results with RT; 70-85% long term control Relatively high risk of treatment related effects
Review of 144 published data; Adamson & Yasargil 2008
Recurrence rate after only partial excision
Author Carbezudo Carmel Djordjevic Hoff Hoffman Lichter McMurrary Shapiro Stahnke Sweet Thomsett yr 1981 1982 1879 1972 1977 1977 1977 1979 1984 1976 1980 n 14 14 15 18 15 9 9 9 12 5 11 131 Recurrence 12 10 8 16 8 7 7 7 6 4 10 93 (71%) FU (yrs) 5-30 6.1 10 2-16 1-20 1-14 7.8 6.9 1-21 8.2
Recurrence rate 71% after only partial excision
Surgery alone vs Sur+ RT
Subtotal resection + RT: higher PFS (n=76)
Stripp et al IJROBP 2004
SRS/fSRS: Craniopharyngioma
Veeravagu et al, Neurosurg Focus 2010
Craniopharyngioma: SCRT- IQ assessment (n=18)
120 110 100 90 80 Mean IQ scores 70 60 50 40 30 20 10 0 Pre-RT 6 month 24 month 36 month VQ PQ FSIQ MQ 5 0 Pre-RT 6 month 24 month 36 month Mean Score
Mean IQ Scores
35 30 25 20 15 10
Mean Anxiety Score
Anxiety Trait (C1) Anxiety State (C2)
Mean IQ Scores are maintained at post-RT follow up. State anxiety had reduced after RT.
VQ: Verbal Quotient PQ: Performance Quotient MQ: Memory Quotient FSIQ: Full Scale IQ Dutta, Jalali et al WFNO 2009
Pituitary tumour: SRS
Problems with SRS: Pituitary tumour close to Optic pathway/ chiasm. Tumor close to chiasm may not be treated with surgery Also not possible to treat with single fraction SRS Constraint to chiasm: 10 Gy SRS dose required: 12 Gy fSRS is possible Higher dose can be delivered without increasing chiasm injury SRS/ fSRS increases early hormonal control without increasing toxicity (12 vs 40 mo)
Plowman Clinical Endocrinology 1999
Recurrent HGG: SRS studies
Romanelli, Neurosurg focus 2009
Recurrent GBM: SRS
SRS/fSRS MS (mo) 6-mo PFS (%) Radionecrosis Corticosteroid 6.5 20 60% SRS+TMZ 12 60 10% 80%
Conti 2010
Recurrent GBM: Survival function
Conti 2010
HGG: IMRT + CK boost Protocol
Eligibility Criteria: Histopathologically confirmed high grade gliomas (AA / GBM). Karnosky performance status >70. Willing for IMRT and Cyberknife treatment.
Methodology:
Conformal RT (50 Gy/25#/5 wks) CK 20Gy/5#
Conc TMZ (75mg/m2) x 6 wks Adj TMZ (200 mg/m2) x 6 cy
End point:
Survival function, Activities of daily livings QOL (ethical committee approved)
New Indications: SRS
-Temporal lobe epilepsy - Resistant seizure disorder - Behavioral disorders - Mood disorder - Obesity - Child hood attention deficit disorder / absence seizure - Skull base tumour
Quality of life is paramount important
EORTC QLQ C30 & BN20 Score in HGG (n=255)
TMH data EORTC QLQ-C- 30* Global score Emotional Cognitive Social Function Fatigue Pain BN-20** Future uncertainty Communication deficit Seizures Drowsiness 51.7 61.4 67.6 69.2 44.4 39.4 23.1 34.9 38.2 18.5 Taphoorn et al* 62.8 69.3 67.5 35.3 40.1 18.6 NA 26.4
Future uncertainty & communication deficits are different in our data & western data
Jalali, Buddrukar, Dutta JNO 2009
SRS in brain tumours Conclusions
- SRS is one of the standard of care is many small & benign brain tumours. - It seems, clinical outcome of robotic radiosurgery is similar to GK in these subset of pts - fSRS is an attractive option in larger benign/low grade and malignant tumours