Transcription

1 Biologically Guided Radiation Therapy (BGRT) Relative Biologically Effectiveness (RBE) and Oxygen Effects in Particle Therapy Rob Stewart, Ph.D. Associate Professor and Assistant Head, School of Health Sciences Director, Radiological Health Science Program School of Health Sciences Purdue University University of Washington Department of Radiation Oncology Friday April 23, 2010 School of Health Sciences

2 Slide 2/71 Physics Chemistry Biology Clinic Radiation Chemical 10-3 s 1 Gy ~ 1 in 10 Correct Repair 6 Repair Ionization Excitation 10-6 s 10 2 s 10 4 s Enzymatic Repair (BER, NER, NHEJ, ) to s DNA damage Angiogenesis and Early Effects Incorrect or 10 6 s Vasculogenesis 10 3 s (erythema, ) 10 5 s Incomplete Repair Loss of Function Inflamatory and Remodeling Responses Cell Death Late Effects (fibrosis, ) 10 8 s 10 4 s 10 5 s Cancer 10 8 s Clonal Expansion 10 7 s Slf Self renewal and Differentiation Heritable Effects Neoplastic Transformation 10 5 s Germline Somatic cells Small- and large-scale mutations (point mutations and chromosomal aberrations)

3 Slide 3/71 RBE Effects in the SOBP Delivered dose in the spread-out Bragg peak (SOBP) due to a mixture of low, intermediate and high energy protons Source: ICRU 78 (2007)

4 Slide 4/71 Effect of Radiation Quality (1 st level) 500 ev electron Simple DSB Opposed strand breaks within about 10 bp Complex DSB composed 2 strand breaks with collateral base damage in same or opposed strands Segment of a 4 MeV α particle ( 4 He He ) Complex SSB one or more strand breaks (same side) with collateral abasic sites and base damage in the opposed strand

5 Slide 5/71 Relative SSB and DSB induction electron 3.0 alpha lative SSB Yield Re alpha Rela ative DSB Yield electron Increasing LET Increasing LET Kinetic Energy (MeV) Kinetic Energy (MeV) Ratio of SSB per DSB decreases with increasing LET V.A. Semenenko and R.D. Stewart. Fast Monte Carlo simulation of DNA damage formed by electrons and light ions. Phys. Med. Biol. 51(7), (2006)

6 Slide 6/71 DSB complexity N umber of Lesions per Cluster Double-strand breaks electron 1 Increasing LET proton alpha Particle Energy (MeV) V.A. Semenenko and R.D. Stewart. Fast Monte Carlo simulation of DNA damage formed by electrons and light ions. Phys. Med. Biol. 51(7), (2006)

7 Slide 7/71 SSB and DSB yield ) SSB Yield (Gy -1 Gbp electron 200 total > 2 lesions > 4 lesions proton > 4 lesions > 2 lesions total 0 30 > 6 lesions DSB Yield (G Gy -1 Gbp -1 ) Total (2 or more) > 4 lesions > 6 lesions Total (2 or more) > 4 lesions Electron Energy (MeV) Proton Energy (MeV) V.A. Semenenko and R.D. Stewart. Fast Monte Carlo simulation of DNA damage formed by electrons and light ions. Phys. Med. Biol. 51(7), (2006)

8 Slide 8/71 DNA and Chromatin Induction of individual and clustered DNA lesions by ionizing radiation related stochastic ti events and processes on Spatial scale < 10 nm (100 to 1000 nucleotides) Time scales < 10-3 s nm 30 nm 11 nm Nucleosome ( bp)

9 Slide 9/71 Radiation Quality and Fragment Size Distributions (2 nd level) ~ 1000 bp (fiber) ~ 85 bp (nucleosome) ~ 85 bp (nucleosome) ~ 1000 bp (fiber) Holley WR, Chatterjee A. Clusters of DNA induced by ionizing radiation: formation of short DNA fragments. I. Theoretical modeling. Radiat Res. 145(2): (1996). Rydberg B. Clusters of DNA damage induced by ionizing radiation: formation of short DNA fragments. II. Experimental detection. Radiat Res. 145(2):200-9 (1996).

10 Slide 10/71 RMDS and DSB Proximity Significance? DSB in close spatial proximity are more likely to interact (less likely to be correctly repaired) ) than DSB separated by large distances High-LET radiation is more effective at producing spatially and temporally correlated DSB than low LET radiation On average, DSB formed by high-let radiation are more likely to cause harm than DSB from low-let radiation

11 Slide 11/71 Damage Distribution Among Cells (3 rd level) ency (DS SB -1 ) Frequ Gy absorbed dose 0.05 Low LET (1 kev/μm) DSB cell tracks DSB track z F = 2.04 mgy Number of DSB

12 Slide 12/71 DSB distribution (1 Gy, 2 MeV α particle) B -1 ) Freque ency (DS Not hit 1 Gy absorbed dose High LET (162.5 kev/μm) 1 hit 2 hit 3 hit DSB cell tracks 4 hit 46.3 DSB track Number of DSB z F = mgy Nucleus 5 μm in diameter

13 Slide 13/71 Effects of Radiation Quality on α and α/β z α θ 1 F Σ =Σ + θ / κ θ / κ α / β = 2 zf + Σ z F LET d Σ = number of DSB Gy - 1 cell - 1 (estimated using Monte Carlo simulations) θ and κ are dimensionless, cell-specific biological constants that are independent or a weak function of LET up to about kev/μm frequency-mean specific energy (Gy)

14 Slide 14/71 Proton RBE (experimental) In vitro In vivo 1.75 In vitro: (0.8, 1.5) roton RBE P 1.50 In vivo: (0.8, 1.3) ICRU 78 (2007) recommends an RBE = Fraction Size (Gy) P tti H Ni i k A A ki i M G k LE G it i M Paganetti H, Niemierko A, Ancukiewicz M, Gerweck LE, Goitein M, Loeffler JS, Suit HD. Relative biological effectiveness (RBE) values for proton beam therapy. Int J Radiat Oncol Biol Phys. 53(2), (2002)

15 Slide 15/71 RBE for Clonogenic Survival 10 0 ng Fraction Survivi Low LET 10-1 Higher LET (e -, γ-rays, y, )) protons 10-2 RBE Dγ 7.6 Gy 1.7 D 4.4 Gy = p Absorbed Dose (Gy)

16 Slide 16/71 Equivalent Dose of the Reference Radiation Outcome reference radiation = Outcome mixture of protons n = n = i= 1 SD ( ) SD ( ) SD ( ) SD ( ) SD ( ) SD ( ) γ 2 2 exp α D β G D = exp α D β GD γ γ γ γ γ i i i i i i Take logarithm, apply quadratic formula and rearrange terms i D γ ( α / β) γ 4Gγ GD i i = αidi 1+ 2 G γ α γ ( α / β ) γ i ( α / β ) i Formula has explicit corrections for total dose, fraction size, and dose rate effects Effect of radiation quality implicit in the biological parameters. α i /α γ, (α/β) i, (α/β) γ

17 Slide 17/71 RBE for a Mixture of Protons Define n D 1 f i αi / α γ, D, Gγ p i γ, n i= 1 γ and G i 1 n p RBE D γ nγ( α / β) γ 4 D i = fidi 1+ Dp 2 Dp nγ( α / β) γ i np( α / β) i physical parameters: n γ, n p, D p, D i biological parameters: f i, (α/β) γ, (α/β) i Similar RBE formula derived by Wilkens and Oelfke (2004) and references therein. Wilkens JJ, Oelfke U, A phenomenological model for the relative biological effectiveness in therapeutic proton beams. Phys Med Biol. 49(13), (2004).

18 Slide 18/71 Possible proton RBE range (α/β) γ = 1.5 Gy In vitro In vivo RBE 1.5 Proton RBE Fraction Size (Gy) 0.35 Gy -1 > α p > α γ 100 Gy > (α/β) α γ = 0.1 Gy -1 p > (α/β) γ Sampled using MC methods Fraction Size (Gy) Paganetti H, Niemierko i A, Ancukiewicz i M, Gerweck LE, Goitein i M, Loeffler JS, Suit HD. Relative biological effectiveness (RBE) values for proton beam therapy. Int J Radiat Oncol Biol Phys. 53(2), (2002)

19 Slide 19/71 Potential explanation for RBE < Su urviving Fract tion 10-1 proton RBE 1.5 photon Fraction Size (Gy) α γ = 0.1 Gy -1 (α/β) γ = 1.5 Gy Fraction Size (Gy) α p = 0.25 Gy -1 (α/β) p = 100 Gy

20 Slide 20/71 Is proton RBE constant? Any RBE in range from 0.8 to 1.5 consistent with available experimental data RBE = 1.1 unlikely to be useful (accurate) for Comparing the efficacy of photon vs proton treatments Guiding the selection of appropriate prescription dose and tolerance doses for normal tissues Constant RBE inconsistent with widely accepted biophysical mechanisms underpinning clonogenic survival, tumor control,

21 Slide 21/71 -lns vs Aberrations (= processed DSB) One to one correlation between lethal exchanges (dicentrics and centric rings) ) implies this form of damage is a major reason cells are killed by radiation Source: Cornforth and Bedford, Rad. Res., 111, p (1987). AC 1522 normal human fibroblasts irradiated by x-rays

22 Slide 22/71 Breakage and Reunion Theory DSB Chromosome Incorrect Rejoining Chromosome (= DNA molecule) R.K. Sachs and D.J. Brenner, Chromosome Aberrations Produced by Ionizing Radiation: Quantitative Studies ShowTOC&rid=mono_002

23 Slide 23/71 RBE for DSB induction DSB induction is closely linked to (but not identical to) ) a major cell killing mechanism for ionizing radiation RBE for DSB induction: dose of a reference radiation needed to produce the same # of DSB as another radiation ( isoeffect calculation ) RBE {dose of photons} D Σ = γ = {dose of proton} D Σ p p γ Σ = number of DSB Gy - 1 cell -1

24 Slide 24/71 Monte Carlo Damage Simulation (MCDS) MCDS successfully reproduces induction of DNA damage predicted by track structure simulations Electrons (> 80 ev), protons (> 105 kev), α (> 2 MeV) up to about 1 GeV (Semenenko and Stewart 2006) Photons from about 80 ev up to 1 GeV (Hsaio and Stewart 2008) Reasonable agreement with measured data Number of SSB and DSB per Gy per cell Computationally efficient and freely available Damage configurations for 100,000 cells exposed to 1 Gy can be simulated on 2.8 Ghz Pentium in about 1.5 minutes Y Hsiao, R.D. Stewart, Monte Carlo Simulation of DNA Damage Induction by X-rays and Selected Radioisotopes. Phys. Med. Biol. 53, (2008). V.A. Semenenko and R.D. Stewart. Fast Monte Carlo simulation of DNA damage formed by electrons and light ions. Phys. Med. Biol. 51(7), (2006)

25 Slide 25/71 DSB yields Track Structure and MCDS DSB Yiel ld (Gy -1 cell -1 ) Open symbols Nikjoo et al. (1994, 1997, 1999, 2001, 2002) Filled symbols Friedland et al. (2003, 2005) Solid Line Semenenko and Stewart (2004, 2006) Diamonds electrons Squares protons Triangles α particles Low LET (high h KE) High LET (low KE) Z eff 2 /β 2 Figure adapted from V.A. Semenenko and R.D. Stewart. Fast Monte Carlo simulation of DNA damage formed by electrons and light ions. Phys. Med. Biol. 51(7), (2006)

26 Slide 26/71 Fragmentation Analysis vs Theory (per track) -1 DSB trac ck -1 Gbp Tsao et al. (2007) Lobrich et al. (1996) Newman et al. (1997, 2000) 10 1 Hoglund et al. (2000) Belli et al. (2002) Rdb Rydberg et al. (2002) Radulescu et al. (2004) Meijer et al. (2005) 10 0 Esposito et al. (2005) Dini et al. (2005) 10-1 protons Lines denote damage yields for electrons (black black), protons (blue blue), and α particles (red)) predicted by MCDS LET (kev/μm)

27 Slide 27/71 Proton RBE for DSB induction μm (0.5 MeV) 3.0 Monte Carlo simulation (Semenenko k and Stewart 2006) LET (kev/μm m) μm (2 MeV) 1.2 mm (10 MeV) RBE for DSB indu uction ICRU 78 (2007) 0.5 MeV electron Kinetic Energy (MeV) Kinetic Energy (MeV) RBE for DSB induction computed using the MCDS

28 Slide 28/71 Model for α and α/β (conceptual basis) DSB Chromosome z α θ 1 F Σ =Σ + θ / κ θ / κ α / β = 2 zf + Σ z F LET d Incorrect Rejoining Σ = number of DSB Gy - 1 cell - 1 (estimated using Monte Carlo simulations) θ and κ are dimensionless, cell-specific biological constants that are independent or a weak function of LET up to about kev/μm

29 Slide 29/71 Radiosensitivity Parameters for Protons z F Σ α R, (α/β) R Other Radiation (same θ and θ/κ) Analyze clinical or laboratory z data for a reference radiation 1 F Σ α =Σ θ + θ / κ ( 60 Co γ-rays, 6 MV x-rays, ) θ / κ α / β = 2 zf + Σ LET d R, ( β) R LET z Σ F 2 d α R 2z F θ = 1 θ κ = Σ Σ [( α β) R 2zF] 2 Other Radiation ( αβ) R (e -, p, α, γ) ) Reference Radiation

30 Slide 30/71 Human Kidney T1 cells (in vitro) zfσ κ α =Σ θ 1 + β = Σ 0.25 θ / κ ) α (Gy ) β (Gy LET (kev/μm) LET (kev/μm) Radiosensitivity parameters for human kidney cells irradiated in vitro by x-rays, 2 H 1+ and 4 He 2+ ions. Filled symbols: estimates of α and β reported by Carlson et al. (2008) for x-rays (filled green triangles), 2 H + (filled red circles) and 4 He 2+ (filled blue squares). Error bars denote the 95% CI. Solid lines: predicted LQ parameter for protons with kinetic energies between 0.1 MeV to 1 GeV (θ = and θ/κ = 249.5). Estimate of θ and θ/κ computed from LQ parameters for x-rays with Σ = 50 DSB Gy -1 cell -1, α = Gy -1, α/β = 10 Gy. Carlson DJ, Stewart RD, Semenenko VA, Sandison GA, Combined use of Monte Carlo DNA damage simulations and deterministic repair models to examine putative mechanisms of cell killing. Rad. Res. 169, (2008)

31 Slide 31/71 Maximum (low dose) ) RBE When D << α/β, S(D) exp(-αd) and RBE reduces to nγ( α / β) γ 4 Di = fidi 1 2 D nγ( α / β) + γ i n ( α / β) RBE p p i 1 D = fidi where P i f i zf( LETi) Σi Σ i 1 α + θ / κ Σ z ( LET) Σ Σ = 1 α z + / 1 γ γ θ κ Σ Σ Σ Σ γ + θ / κ i i F i i i γ γ γ Σ i (protons) and Σ γ (reference radiation) determined using the MCDS (Semenenko k and Stewart t 2006, Hsaio and Stewart t 2008) Y Hsiao and R.D. Stewart, Monte Carlo Simulation of DNA Damage Induction by X-rays and Selected Radioisotopes. Phys. Med. Biol. 53, (2008). V.A. Semenenko and R.D. Stewart. Fast Monte Carlo simulation of DNA damage formed by electrons and light ions. Phys. Med. Biol. 51(7), (2006)

32 Slide 32/71 Maximum Proton RBE Max ximum RBE α/β =1.5 Gy α/β =3.1 Gy α/β =5 Gy α/β =10 Gy RBE for DSB induction Maximum RBE increases as α/β decreases RBE for cell survival is always greater than or equal to RBE for DSB induction to 3.00 (1 MeV) 1.13 to 1.18 (10 MeV) 1.05 to 1.10 (25 MeV) 2 1 ICRU 78 (2007) < 1.01 (> 100 MeV) Kinetic Energy (MeV) For comparison, ICRU 78 recommends an RBE = 1.1

33 Slide 33/71 Effective RBE in the SOBP Proton Dose Distribution EUD p Equivalent Photon Dose Distribution Niemierko, Med. Phys. 24(1), (1997) EUD γ RBE EUD γ EUD p MeV MeV protons incident on water phantom Pencil beam CSDA algorithm (NIST stopping power) Absorbed D ose (Gy) Entrance (5 mm skin ) Proximal (1 cm) Distal (1 cm) ) Depth into Water (cm)

34 Slide 34/71 RBE in the SOBP (4 cm target, 2 Gy fraction) Absor rbed Dose (Gy) Entrance (5 mm skin ) Proximal (1 cm) Distal (1 cm) Depth into Water (cm) α = Gy -1, α/β = 10 Gy Region EUD p EUD γ RBE Skin Proximal Middle Distal Target Avg α = Gy -1, α/β = 1.49 Gy Region EUD p EUD γ RBE Skin Proximal Middle Distal Target Avg RBE increases as photon α/β decreases

35 Slide 35/71 Effect of Target and Fraction Size 1.5 Photon: α = Gy -1, α/β = 1.49 Gy ctive RBE Effe Gy 2 Gy 0.01 Gy RBE increases as target size decreases RBE decreases as fraction size increases 1.1 ICRU 78 (2007) Width of SOBP (cm)

36 Slide 36/71 Summary and Conclusions Entrance RBE (e.g., skin) close to 1.0 RBE in distal edge of SOBP higher than proximal edge ( ) 1.8) Potential for biological hot and cold spots Average RBE in target (1.1 1 to 1.5) Increases as α/β decreases (tissue and tumor specific) Increases as target size decreases Increases as fraction size decreases For large fractions, RBE < 1 Range comparable to experimental observations (0.8 (08to 1.5) 15)

37 Monte Carlo Simulation of the Effects of Oxygen on Clustered DNA Lesions Formed by Ionizing Radiation Rob Stewart, Ph.D. Associate Professor and Assistant Head of Health Sciences Director, Radiological Health Science Program School of Health Sciences Purdue University Presented at 2008 RRS Annual Meeting (Sept 21-25, 2008) Boston, MA Stochastic Track Models Monday September 22, :30 to 4:30 pm School of Health Sciences

38 Slide 38/71 Effects of oxygen on DNA damage induction Long known that oxygen is a powerful modulator of DNA damage induction and closely related endpoints, such as clonogenic death and mutation Cells irradiated under reduced oxygen sustain less damage and are much more likely to survive (factor ~ 2-4) Models to predict the effects of oxygen on cluster induction could be used to help test hypotheses such as Reduced oxygen levels at the time of irradiation decreases cluster complexity (number of DNA lesions per cluster) and reduces the overall cluster yield per cell and per unit dose. The rate and fidelity (efficiency) of cluster repair increases as cluster complexity y( ff y) p p y decreases. Enhanced repair of clusters increases cell survival.

39 Slide 39/71 Chemical Basis of the Oxygen Effect Competition between oxygen fixation and chemical repair is the prevailing hypothesis (von Sonntag 2006) (1) DNA + ionizing radiation DNA lesion (biochemical repair required) (2) DNA + ionizing radiation DNA (various) (3) DNA + O 2 DNA-O 2 ( oxygen fixation biochemical repair required) (4) DNA + RSH DNA ( chemical repair restoration of the DNA * ) (5) DNA DNA lesion (biochemical repair required) * Von Sonntag notes that donation of a proton to a DNA radical may or may not restore the original chemical structure of the DNA. But, the chemical repair process evidently converts the DNA radical (or cluster of radicals?) ) into a form that is more amenable to biochemical repair Clemens von Sonntag, Free-Radical-Induced DNA Damage and its Repair A chemical perspective. Springer-Verlag, New York, NY (2006)

40 Slide 40/71 MCDS Modification Step 1 Use the original MCDS to simulate the location of DNA radicals At this stage of the simulation All lesions formed in the original MCDS treated as a radical that may undergo fixation or chemical repair No distinction made between radicals formed through direct and indirect mechanisms Preserves the ability of the MCDS to simulate lesion (radical) clustering y ( ) g effects without introducing any additional parameters into the modeling process.

41 Slide 41/71 MCDS Modification Step 2 (conceptual basis) Oxygen is uniformly distributed near the DNA and able to interact with all DNA radicals with equal probability O 2 O 2 O 2 O 2 O 2 O 2 O 2 O 2 O 2 If oxygen fixation occurs (DNA + O 2 DNA-O 2 ), radical converted to a strand break or damaged base as in the original MCDS cluster yields for normoxic conditions same as original MCDS. Fixation implicitly includes other processes that convert DNA radicals to a form requiring biochemical repair. DNA DNA lesion (biochemical repair) If chemical repair occurs (DNA + RSH DNA), DNA is restored to its original (undamaged) state lesion not created

42 Slide 42/71 Chemical repair or fixation? Define f as the fraction of the DNA radicals that undergo chemical repair Fraction (1-f f ) of the DNA radicals are fixed by oxygen O 2 O 2 O 2 O 2 O 2 O 2 O 2 O 2 O 2 f ([ O ]) = 2 1 [ O ] [ 2 O ] K + M K [O 2 ] = oxygen concentration at time of irradiation Formula is derived from (related to) ) oxygen-effect effect formula of Alper and Howard-Flanders (1956) Introduces two adjustable parameters (M and K) ) into the modeling process. Alper T, Howard-Flanders P. Role of oxygen in modifying the radiosensitivity of E. coli B, Nature (London) 178, (1956).

43 Slide 43/71 Effect of O 2 on fixation and chemical repair 1.0 Pro obability 0.8 [ f max = 1 1/M 1-f (fixation) chemical repair normoxic K = M = f at 1/M f O ] ( ) = 2 1 [ O ] [ 2 O ] 2 + K + M K M determines maximum fraction of radicals fixed at 0% O 2 K = oxygen concentration at which f equals 1/M K Oxygen Concentration (%)

44 Slide 44/71 OER for Damage Induction (low LET) Georgakalis ( unpublished) Frankenberg-Schwager et al. (1991) Nygren and Ahnstrom (1996) Hirayama et al. (2005) Whitaker McMillan (1992) Prise et al. (1992) Botchway et al. (1997) V79 cells CHO cells Human kidney T-1 cells OER MCDS (1 MeV e - ) OER Oxygen Concentration (%) 1.0 DSB CllS Cell Survival Oxygen Concentration (%) Trends and estimates of OER for DSB induction and cell survival comparable to values predicted by MCDS (K = , M = ). Carlson DJ, Stewart RD, Semenenko VA, Effects of oxygen on intrinsic radiation sensitivity: A test of the relationship between aerobic and hypoxic linear-quadratic (LQ) model parameters. Med Phys. 33(9), (2006).

45 Slide 45/71 Interplay between oxygen and LET (1) DNA + ionizing radiation DNA lesion (biochemical repair required) (2) DNA + ionizing radiation DNA (various) (3) DNA + O 2 DNA-O 2 ( oxygen fixation biochemical repair required) (4) DNA +RSH DNA ( chemical repair restoration of the DNA * ) (5) DNA DNA lesion (biochemical repair required) If ionization does not substantially alter the local cellular environment (e.g., oxygen-in-track hypothesis ), might expect reactions (3)-(5) ( ) to be same for low and high LET radiation. f [ O ] ( ) 2 1 = [ O2 ] [ O ] 2 + K Assume M and K are independent of radiation quality. + M K radiation quality.

46 Slide 46/71 OER OER with f independent of LET e - Cell survival (Carlson et al. 2006) Frankenberg-Schwager et al. (1991) Prise et al. (1989) Prise et al. (1990) Hirayama et al. (2005) LET (kev/μm) α Number of Lesions 2 MeV α (162.5 kev/μm) f = 39.6% DSB Yield (%) Anoxic Normoxic Model cannot easily explain the decrease in the OER as LET > increases Avg per DSB

47 Slide 47/71 Possible explanation for LET effect? (1) DNA + ionizing radiation DNA lesion (biochemical repair required) (2) DNA + ionizing radiation DNA (various) (3) DNA + O 2 DNA-O 2 ( oxygen fixation biochemical repair required) (4) DNA +RSH DNA ( chemical repair restoration of the DNA * ) (5) DNA DNA lesion (biochemical repair required) Ionization does not substantially alter the local cellular environment (reactions 3-5 same for low and high LET radiation), but reaction 1 is enhanced relative to reaction 2 f [ O ] ( ) 2 1 = [ O2 ] [ ] 2 + K O + M() l K K isindependent of radiation quality and M becomes function of radiation quality (convenient way to implement effect into model)

48 Slide 48/71 OER with f a function of radiation quality f ([ O2 ]) M() l = 1 where l [ O ] [ ] K O + M() l K ( M 11 ) 0 = M0 + 1 q/ l 2 Z eff β 2/3 ( β ) Z Z eff = 1 exp 125 Z β = ( 1 + T / mc 2 0 ) OER α Cell survival (Carlson et al. 2006) Frankenberg-Schwager g et al. (1991) 1.0 Prise et al. (1989) Prise et al. (1990) Hirayama et al. (2005) e - LET (kev/μm) M 0 = , q = , K =

49 Slide 49/71 Effect of Oxygen on Cluster Complexity Number of Lesions DSB Yield (%) Anoxic Normoxic > Avg per DSB OER MeV α (162.5 kev/μm) Number of Lesions DSB Yield (%) Anoxic Normoxic > Avg per DSB OER MeV e - (0.186 kev/μm)

50 Slide 50/71 OER for SSB and DSB induction MeV electron DSB 2.5 DSB OER OER 1.5 SSB 1.5 SSB Measured doer ssb ~ 2-4 4(l (low LET) Oxygen Concentration (%) LET (kev/μm) OER dsb (OER ssb ) 2 M(l) 2

51 Slide 51/71 Summary Predicted trends in the OER for DSB induction consistent with estimates derived from measured data for DSB induction and cell survival (OER ~ 2 to 3) Three new parameters introduced into the modeling process to account for oxygen concentration (M and K) and to correct for the effects of radiation quality (q) As LET increases, the probability chemical repair occurs per initial DNA radical may decrease (oxygen fixation increases and/or lesions are directly created) Predicted OER for SSB induction approximately equal to the square root of the OER for DSB induction (OER ~ 1.4 to 1.7) Most of the measured data suggest a higher OER for SSB induction Average cluster complexity increases as the oxygen concentration increases 1 MeV e - (2.5 lesions/dsb 0% O 2 and28lesions/dsb21%o) % 2 2 MeV α (6.8 lesions/dsb 0% O 2 and 8.0 lesions/dsb 21% O 2 )