The Impact of Cobalt-60 Source Age on
Biologically Effective Dose in Gamma Knife
Thalamotomy
BH Kann, JB Yu, J Bond, C Loiselle, VL Chiang, RS Bindra, JL Gerrard, DJ Carlson
Leksell Gamma Knife Society Meeting 2016
May 17th, 2016
Presented by
Benjamin H. Kann, MD
Department of Therapeutic Radiology
Yale University School of Medicine
New Haven, CT, USA
SLIDE 0
Disclosures
• I have no conflicts of interest to disclose
SLIDE 1
Background
• Tremor greatly reduces QoL
– Parkinsonian
– Multiple Sclerosis
– Essential
• Incidence ~616 per 100,000 people
• Treatment options
–
–
–
–
–
Medical management
Surgical thalamotomy
RF ablation
DBS
Gamma Knife (GK) radiosurgery
J. Benito-León, et al. Incidence of essential tremor in three elderly populations of central Spain Neurology, 64 (2005), pp. 1721–1725
SLIDE 2
GK Thalamotomy
• Single 4mm shot to the thalamic ventralis intermedius (VIM)
nucleus to ~130 Gy*
Kooshkabadi, et al. Gamma knife thalamotomy for tremor in the magnetic resonance imaging era. J Neurosurg, 118 (2013), pp. 713–718
Yamada, et al. MR Imaging of Ventral Thalamic Nuclei. AJNR 2010 31: 732-735
SLIDE 3
GK Thalamotomy
• Promising efficacy
– >80% tremor improvement in most series
• Toxicity rare (<10%), but can be severe
– Hemiparesis, paresthesia, dysarthria, dysphagia, gait disturbance, death due to
hemorrhage/necrosis
– Multifactorial
Anecdotal reports: increased toxicity with Cobalt-60 source replacement
Campbell, et al. Gamma knife stereotactic radiosurgical thalamotomy for intractable tremor: A systematic review of the literature, Radiotherapy and Oncology. March 2015,
R.F. Young, et al. Gamma knife thalamotomy for treatment of tremor: long-term results. J Neurosurg, 93 (2000), pp. 128–135
Lee et al. Higher dose rate Gamma Knife radiosurgery may provide earlier and longer-lasting pain relief for patients with trigeminal neuralgia. Journal of Neurosurgery. Oct 2015,
Vol. 123, No. 4, Pages 961-968
SLIDE 4
GK Thalamotomy Source Age / Dose Rate
Cobalt-60 lifespan (half-life ~5.25 years)
New source
Replacement
Old source
Source Activity ↓
Dose Rate ↓
Treatment Time ↑
↑ toxicity?
1)
↓ efficacy?
Fowler et al. Loss of biological effect in prolonged fraction delivery. May 2004, Pages 242–249
SLIDE 5
Hypotheses
• Given high Rx dose of GK thalamotomy, biologically effective dose
(BED) of treatment may vary significantly with Cobalt-60 source
age
• This variation may substantial enough to expect effect on clinical
outcome
• A BED model based on the linear-quadratic (LQ) model with
correction for intra-fraction DNA repair could estimate BED
variation with source age
SLIDE 6
Methods - BED: the LQ Model
 G [λ , T ] D 
− ln S ( D)
=+
D 1
BED ≡

α
α /β 

“G” = protraction factor
(~intra-fraction repair)
T = irradiation time (~Co-60 source age)
τ= DNA repair half-time
(~0.1 to 10 hr)
λ=
ln 2
τ
Adapted from Carlson et al.
SLIDE 7
GK Thalamotomy BED Model Construction
Model Parameters:
D, T, α/β, τ
–
–
–
–
Dose
Treatment time
α/β Alpha / Beta
τ DNA repair half-time
λ=
ln 2
τ
SLIDE 8
GK Thalamotomy BED Model Construction
Model Parameters:
– Dose = 130 Gy
λ=
ln 2
τ
– Treatment time ~ Co-60 Source Age
• New Source: Dose Rate(t=0) = 2.99 Gy/min*
• Exponential activity decay (T1/2 = 5.25 years = 63 months)
A =Ao × exp [ − ln(2) / 63 × t ]
t = Source Age (mo)
Dose Rate(t) = Dose Rate(0) x A
Treatment time (T) = Rx Dose / Dose Rate (t)
*derived from departmental/vendor data
SLIDE 9
GK Thalamotomy BED Model Construction
Model Assumptions:
• α/β
- Empirical data suggests α/β ~ 2 for CNS tissue
λ=
ln 2
τ
• τ
– Bi-exponential DNA repair half-times (“fast” and “slow” components)
• Model range: τ(fast) = 0.1 to 0.7 h; τ(slow) = 2 to 6 h
• “Most plausible” scenario: τ(fast) = .38 h; τ(slow) = 4.1 h
K.K Ang, Radiotherapy and Oncology, 1992. Pop et al, Radiotherapy and Oncology, 2000. Landuyt et al, Radiotherapy and Oncology, 1997.
S L I D E 10
Results
Change in Treatment Time (T) with Co-60 Source Age
200
180
Tx time (minutes)
160
140
Treatment Time
Old source: 88 min
120
100
(1.48 Gy/ min)
80
60
New source: 43 min
40
20
0
(2.99 Gy/min)
TxTime
0
12
24
36
48
60
72
84
96
108
120
132
144
Age of source (months)
S L I D E 11
Results
With new source
replacement:**
% Decrease in BED with Cobalt-60 source age for 130 Gy
thalamotomy
0%
15.6%
relative
increase in
BED
(range: 1120%)
-5%
-10%
BED
-15%
-20%
-25%
-30%
τf = .7, τs = 6
-35%
τf = .38, τs = 4.1 *
-40%
τf = .1, τs = 2
0
12
24
36
48
60
72
84
96 108
Cobalt-60 Source Age (months)
120
132
144
*Using “most plausible” Tau assumptions. **with replacement after 63 months
S L I D E 12
Conclusions
• Using biologically “most plausible” assumptions, Cobalt-60 source
replacement causes a relative BED increase of ~16% for GK
Thalamotomy, that then declines by ~3% per year over Cobalt60 lifespan
– Magnitude of BED change dependent on DNA-repair kinetics, timing
of source replacement, initial source activity
• Co-60 source age and dose rate should be considered in the study
and use of GK thalamotomy and other high-dose GK treatments
S L I D E 13
Limitations / Next Steps
• Clinically correlated data needed
– Toxicity vs. efficacy
– Radiographic changes
• Practice changes?
– Prescription dose adjustment with Co-60 source age
– Prolong treatment time with selective collimation
S L I D E 14
Thank You
• Acknowledgements
David Carlson, PhD
James Yu, MD
Department of Therapeutic Radiology, Yale School of Medicine
Department of Neurosurgery, Yale School of Medicine
S L I D E 15
Appendix
Factors affecting toxicity:
-Target placement
-Patient sensitivity
-Lesion size result
S L I D E 16
Results
Decrease in BED with increased treatment times 130 Gy thalamotomy
8000
Increase Rx dose
 Increase %BED
change
7000
6000
Increase starting
dose rate,
DECREASE %BED
change
BED
5000
τf = .7, τs = 6
4000
3000
Increasing “Tau”
 DECREASE
%BED change
τf = .38, τs = 4.1 *
2000
1000
0
τf = .1, τs = 2
40
50
60
70
80
90
100
110
120
130
140
Treatment Time (min)
*Using “most plausible” Tau assumptions
S L I D E 17
GK Thalamotomy BED Model Construction
Model Assumptions:
• α/β
λ=
- Empirical data suggests α/β ~ 2 for CNS tissue
ln 2
τ
• τ(DNA repair half-time)
- Empirical evidence suggest bi-exponential DNA repair half-times (“fast”
and “slow” components)
Rat Spinal Cord Study
K.K Ang, Radiotherapy and Oncology, 1992
Pop et al, Radiotherapy and Oncology, 2000
Landuyt et al, Radiotherapy and Oncology, 1997
Mean
Fast (hr)
0.7
0.19
0.25
0.38
Slow (hr) % fast repair a/b (gy)
3.8
38%
2
2.16
51%
2.47
6.4
50%
2
4.1
•
Model range: τ(fast) = 0.1 to 0.7 h; τ(slow) = 2 to 6 h
•
“Most plausible” scenario: τ(fast) = .38 h; τ(slow) = 4.1 h
– With assumed equal contribution (50:50) in DNA repair
S L I D E 18
Dose rate effects and DNA damage repair
• Cell survival increases with
decreasing dose rate
Surviving Fraction
100
0.12 Gy h-1
• If dose protraction effects
included in the model,
a unique set of parameters
can predict the entire data
set:
10-1
0.5 Gy h-1
10-2
– α = 0.04 Gy-1
– β = 0.02 Gy-2
– τ = 6.4 h
45 Gy h-1
10-3
10-4
0
5
10
15
20
25
30
35
Absorbed Dose (Gy)
Chinese Hamster Ovary
(CHO)
Measured data from Stackhouse M.A. and
Bedford J.S. Radiat. Res. 136, 250-254 (1993) and Wells R.L. and Bedford J.S. Radiat. Res. 94(1), 105-134
S L I D E 19
Biologically Effective Dose (BED)
 BED is an LQ based estimate of the effective biological dose that
accounts for delivered total dose, the dose fractionation, and
the radiosensitivity of tissue
 Commonly used for isoeffect calculations
Recall
S ( D) = exp  −α D − β GD 2 + γ T 
Take the negative logarithm of S and divide by α :
BED ≡
− ln S ( D)
α

GD  γ T
=
−
D 1 +

 α /β  α
Physical dose
Relative effectiveness
“Lost” dose due to
repopulation effect
S L I D E 20
BED in the literature
This derived expression is more general than the commonly
used BED formalism
1/n
0

GD  γ T
− ln S ( D ) / α =
−
BED =
D 1 +

 α /β  α
In the literature, repopulation effects are usually
neglected and G is set equal to 1/n (appropriate for
daily fractions), where n = number of fractions.

d 
BED
= D 1 +

α
/
β


where d = dose per fraction (Gy) and D = total
treatment dose (= nd).
HR Withers, HD Thames, LJ Peters. A new isoeffect curve for change in dose per fraction. Radiother. Oncol. 1, 187-191 (1983).
S L I D E 21
Absolute Decrease in BED with Cobalt-60 source age for 130 Gy
thalamotomy
8000
7000
6000
BED
5000
4000
3000
2000
1000
0
0
10
20
30
40
50
60
70
Cobalt-60 Source Age (months)
BED2
BED3
BED4
S L I D E 22
Results
With new source
replacement:
% Decrease in BED with Cobalt-60 source age for 130 Gy
thalamotomy
0,0%
-2,0%
-4,0%
BED
-6,0%
-8,0%
-9.5%
-10,0%
-12,0%
τf = .38, τs = 4.1 *
-16,0%
-18,0%
-13.2%
τf = .7, τs = 6
-14,0%
-16.5%
τf = .1, τs = 2
0
10
20
15.6%
relative
increase
in BED
30
40
50
60
70
Cobalt-60 Source Age (months)
*Using “most plausible” Tau assumptions
S L I D E 23