Presentation on theme: "The Radiobiology Behind Dose Fractionation Bill McBride Dept"— Presentation transcript:
1 The Radiobiology Behind Dose Fractionation Bill McBride Dept The Radiobiology Behind Dose Fractionation Bill McBride Dept. Radiation Oncology David Geffen School Medicine UCLA, Los Angeles, Ca.Radiation Biology is study of the effects of radiation on living things. For the most part, this course deals with the effects of radiation doses of the magnitude of those used in radiation therapy.
2 Objectives To understand the mathematical bases behind survival curves Know the linear quadratic model formulationUnderstand how the isoeffect curves for fractionated radiation vary with tissue and how to use the LQ model to change dose with dose per fractionUnderstand the 4Rs of radiobiology as they relate to clinical fractionated regimens and the sources of heterogeneity that impact the concept of equal effect per fractionKnow the major clinical trials on altered fractionation and their outcomeRecognize the importance of dose heterogeneity in modern treatment planning
3 Relevance of Radiobiology to Clinical Fractionation Protocols Conventional treatment:Tumors are generally irradiated with 2Gy dose per fraction delivered daily to a more or less homogeneous field over a 6 week time period to a specified total doseThe purpose of convenntional dose fractionation is to increase dose to the tumor while PRESERVING NORMAL TISSUE FUNCTIONDeviating from conventional fractionation protocol impacts outcomeHow do you know what dose to give; for example if you want to change dose per fraction or time? Radiobiological modeling provide the guidelines. It usesRadiobiological principles derived from preclinical dataRadiobiological parameters derived from clinical altered fractionation protocolshyperfractionation, accelerated fractionation, some hypofractionation schedulesThe number of non-homogeneous treatment plans (IMRT) and extreme hypofractionated treatments are increasing. Do existing models cope?
4 In theory, knowing relevant radiobiological parameters one day may predict the response for Dose given in a single or a small number of fractionsSBRT, SRS, SRT, HDR or LDR brachytherapy, protons, cyberknife, gammaknifeNon-uniform dose distributions optimized by IMRTe.g. dose “painting” of radioresistant tumor subvolumesCombination therapies with chemo- or biological agentsDifferent RT options when tailored by molecular and imaging theragnosticsIf you know the molecular profile and tumor phenotype, can you predict the best delivery method?Biologically optimized treatment planning
5 The First Radiation Dosimeter prompted the use of dose fractionation
6 Modeling Radiation Responses Assumes that ionizing ‘hits’ are random events in spaceWhich are fitted by a Poisson DistributionP of x = e-m.mx/x!where m = mean # hits, x is a hitP survival(when x = 0)100 targets 100 hits m=1 e-1=0.368100 targets 200 hits m=2 e-2=0.137100 targets 300 hits m=3 e-3=0.05N.B. Lethal hits in DNA are not really randomly distributed, e.g. condensed chromatin is more sensitive, but it is a reasonable approximation
7 How many logs of cells would be killed by 23 Gy if D0 = 1 Gy? This Gives a Survival Curve Based on a Model where one hit will eliminate a single targetWhen there is single lethal hit per target S.F.= e-1 = 0.37This is the mean lethal dose D0D10 = 2.3 xD0In general, S.F. = e-D/D0or LnS.F. = -D/D0or S.F. = e-aD , i.e. D0 = 1/aWhere a is the slope of the curve and D0 the reciprocal of the slope1.00.10.010.0010.37S.F.The mathematical bent of early radiobiologists led them to describe survival curves by the mean lethal dose (D37 or D0), which is the dose required to cause on average one lethal hit per cell and result in 37% survival. In practice D10, the dose that would reduce survival to by one log10, which is 2.3x D0 is easier to use. The slope of the curve is given by , where D0 is 1/. Bacterial killing and protein inactivation follow this log-linear curve, although the D0 values are high compared with mammalian cells.D0How many logs of cells would be killed by 23 Gy if D0 = 1 Gy?D10DOSE Gy
8 Eukaryotic Survival Curves are Exponential, but have a ‘Shoulder’ Puck and Marcus, J.E.M.103, 563, 1956 First in vitro mammalian survival curveEukaryotic Survival Curves are Exponential, but have a ‘Shoulder’1.00.10.010.001Accumulation ofsub-lethaldamagesinglelethalhitsndoseTwo component modelIn 1956 Puck and Marcus published the first survival curve for mammalian cells and noted that the D0 was cGy. Furthermore, it had a shoulder region before the logarithmic decline. It is easiest to think of this as single-hit and multi-hit killing (another assumption!). At low doses, the rate of deposition of energy by a charged particle is inversely proportional to its energy, and as it loses energy through collisions and scattering the distribution of ionizing events become more dense and the probability of a lethal lesion being formed by a single track increases. At higher doses, accumulation of injury from other tracks (intertrack) becomes a more likely cause of a lethal lesion. Note that the nature of the chromosomal lesions will go from being predominantly deletions to more exchange-type (two-hit) lesions. Note that with doses of around 2Gy, the former will dominate.
9 Accumulation of sublethal Two Component ModelTwo Component Model(or single target, single hit + multi-target (n), single hit)S.F.=e-D/1D0[1-(1-e-D/nD0)n]singlelethalhitsn1.00.10.010.0011D0 =reciprocalinitial slopenD0 =final slopeS.F.ExtrapolationNumberSingle hitAccumulateddamageAccumulation of sublethaldamageDOSE Gy
10 Mean Inactivation Dose (Do) Virus D0 approx. = 1500 GyE. Coli D0 approx. = 100 GyMammalian bone marrow cells D0 = 1 GyGenerally, for mammalian cells D0 = GyWhy the differences?
11 FOOD TYPE DOSE (Gy) EFFECT Meat, Poultry, Fish,Shellfish, some vegetables20, ,000Spices, etc.8, ,000Meat, Poultry, Fish1, ,000Delays spoilage.Kills salmonella.Strawberries andsome other fruits1, ,000Delays mold growthGrain, Fruit, Vegetables,000Kills some insectsBananas and othernon-citrus fruitsDelays ripeningPorkInactivates trichinaeSterilization. Storageat room temperatureReduces micro-organismsand insectsPotatoes, Onions, etc.,.Inhibits sprouting
12 SBRT/SRS often aims at TISSUE ABLATION In general, history has shown repeatedly that single high doses of radiation do not allow a therapeutic differential between tumor and critical normal tissues. Dose fractionation does.SBRT/SRS often aims at TISSUE ABLATION
13 “Double Trouble” Does this Matter? Prescribed Dose: 25 fractions of 2Gy = 50GyHot spot: 110%Physical dose: 55GyBiological dose: 60.5Gy“Double Trouble”
14 EARLY MODELS OF THE EFFECT OF FRACTIONATION Strandquist ploteffect depended only on dose and timeD = const x T 1-pLinear on log/log plot1-p = slope = 0.22 from skin erythemaFowler 1963 in pig skin - Number of Fx importantEllis formula - nominal standard dose (NSD)Number of fx important based on pig skin expts.Dose = (NSD)T0.11.N0.24
15 Failed to account for differences between tissues.
16 Linear Quadratic Formula Biological effect is based on a linear term and a quadratic term Lea and Catchside 1942Radiation-induced chromosome aberrations in Tradescantia microsporesKellerer and Rossi, 1972Theory of dual radiation action based on microdosimetry
17 Linear Quadratic Model Cell kill is the result of single lethal hits plus accumulated damage from 2 independent sublethal eventsThe generalized formula is E = aD + bD2For a fractionated regimen E= nd(a + bd) = D (a + bd) Where d = dose per fraction and D = total dosea/b is dose at which death due to single lethal lesions = death due to accumulation of sublethal lesions i.e.aD = bD2 and D = a/b in Gy1.00.10.010.001aDS.F. = e-aDSingle lethal hitsbD2S.F.S.F. = e-(aD+bD2)Single lethal hits plus accumulated damagea/ in GyDOSE Gy
18 Over 90% of radiation oncologists use the LQ model: it is simple and has a microdosimetric underpinninga/b is large (> 6 Gy) when survival curve is almost exponential and small (1-4 Gy) when shoulder is widethe a/b value quantifies the sensitivity of a tissue/tumor to fractionated radiation.But:Both a and b vary with the cell cycle. At high doses, S phase and hypoxic cells become more important.The a/b ratio varies depending upon whether a cell is quiescent or proliferativeThe LQ model best describes data in the range of 1 - 6Gy and should not be used outside this range
19 The Linear Quadratic Formulation Does not work well at high dose/fxAssumes equal effect per fraction
20 N.B. Survival curves may deviate from L.Q. at low and high dose!!!! Certain cell lines, and tissues, are hypersensitive at low doses of Gy.The survival curve then plateaus over GyNot seen for all cell lines or tissues, but has been reported in skin, kidney and lungAt high dose, the model probably does not fit data well because D2 dominates the equationHT29 cellsAn additional complication has been reported by Joiner et al, who have shown that certain cell lines show a hypersensitivity zone at Gy that flattens out over Gy, before showing the normal shape of survival curve. The basis for this is not well established but hypersensitivity is thought to be associated with increased apoptosis and lack of G2 arrest.Lambin et al. Int J Radiat Biol 63:
21 The resultant slope is the effective D0 S.F. = e-D/eD0 2420161284.01.11Dose (Gy)S.F.Single doselimiting slope/low dose rate3 fractions5 fractionsMulti-fraction survival curves can be considered linear if sublethal damage is repaired between fractionsthey have an extrapolation number (n) = 1.0The resultant slope is the effective D0eD0 is often Gy and eD GyS.F. = e-D/eD0If S.F. after 2Gy = 0.5, eD0 = 2.9Gy; eD10 = 6.7Gy and 30 fractions of 2 Gy (60Gy) would reduce survival by (0.5)30 = almost 9 logs (or 60/6.7)If a 1cm tumor had 109 clonogenic cells, there would be an average of 1 clonogen per tumor and cure rate would be about 37%
22 Thames et al Int J Radiat Oncol Biol Phys 8: 219, 1982. The slope of an isoeffect curve changes with size of dose per fraction depending on tissue typeAcute responding tissues have flatter curves than do late responding tissues measures the sensitivity of tumor or tissue to fractionation i.e. it predicts how total dose for a given effect will change when you change the size of dose fractionReciprocaltotal dosefor an isoeffectSlope = Douglas and Fowler Rad Res 66:401, 1976Showed and easy way to arrive at an ratioIntercept = Dose per fraction
23 Response to Fractionation Varies With Tissue .01.111Acute RespondingTissues a/b = 10GyFractionatedLate EffectsS.F.S.F..1FractionatedAcute EffectsLate RespondingTissues - a/b = 2GySingle DoseLate Effectsa/b = 2GySingle DoseAcute Effectsa/b = 10Gya/b is high (>6Gy) when survival curve is almost exponential and low (1-4Gy) when shoulder is wide.0148121648121620Dose (Gy)Dose (Gy)Fractionation spares late responding tissues
24 20304050607080=3Gy; 1.5Gy/fx=30Gy; 1.5Gy/fx2.0Gy/fx=30Gy; 4Gy/fxD new=3Gy; 4Gy/fx80706050403020D oldNote how badly late responding tissues respond to increased dose/fraction
25 Sensitivity of Tissue to Dose Fractionation can be estimated by the ratio
26 What are a/ ratios for human cancers? In fact, for some tumors e.g. prostate, breast, melanoma, soft tissue sarcoma, and liposarcoma a/ ratios may be moderately lowProstateBrenner and Hall IJROBP 43:1095, 1999comparing implants with EBRTa/ ratio is 1.5 Gy [0.8, 2.2]Lukka JCO 23: 6132, 2005Phase III NCIC 66Gy 33F in 45days vs 52.5Gy 20F in 28 daysCompatible with a/ ratio of 1.12Gy ( )BreastOwen, J.R., et al. Lancet Oncol, 7: , 2006 and Dewar et al JCO, ASCO Proceedings Part I. Vol 25, No. 18S: LBA518, 2007.UK START Trial50Gy in 25Fx c.w. 39Gy in 13Fx; or 41.6Gy in 13Fx [or 40Gy in 15Fx (3 wks)]Breast Cancer a/ = 4.0Gy ( )Breast appearance a/ = 3.6Gy; induration a/ = 3.1GyAdenocarcinoma may be fractionation sensitive, like LRTIf fractionation sensitivity of a cancer is similar to dose-limiting healthy tissues, it may be possible to give fewer, larger fractions without compromising effectiveness or safety
27 What total dose (D) to give if the dose/fx (d) is changed New OldDnew (dnew + ) = Dold (dold +)So, for late responding tissue, what total dose in 1.5Gyfractions is equivalent to 66Gy in 2Gy fractions?Dnew (1.5+2) = 66 (2 + 2)Dnew = 75.4GyNB: Small differences in for late responding tissues can make a big difference in estimated D!
28 Biologically Effective Dose (BED) S.F. = e-E = e-(aD+bD2)E = nd(a + bd)E/a = nd(1+d/a/b)Biologically Effective DoseRelativeEffectivenessTotal doseFractionation alpha nd beta35 x 2Gy = B.E.D.of 84Gy10 and 117Gy3NOTE: 3 x 15Gy = B.E.D.of 113Gy10 and 270Gy3Normalized total dose2Gy= BED/RE= BED/1.2 for of 10Gy= BED/1.67 for of 3GyEquivalent to 162 Gy in 2Gy Fx -unrealistic!(Fowler et al IJROBP 60: 1241, 2004)
29 Isoeffect CurvesD= NSD x N0.24Withers et al Radiat Res 119:395, 1989
30 4Rs OF DOSE FRACTIONATION RedistributionRepairRepopulation700R1500RAssessed by varying the time between 2 or more doses of radiation
31 4Rs OF DOSE FRACTIONATION These are radiobiological mechanisms that impact the response to a fractionated course of radiation therapyRepair of sublethal damagespares late responding normal tissue preferentiallyRedistribution of cells in the cell cycleincreases acute and tumor damage, no effect on late responding normal tissueRepopulationspares acute responding normal tissue, no effect on late effects,danger of tumor repopulationReoxygenationincreases tumor damage, no effect in normal tissuesFractionation benefits
32 Repair“Repair” between fractions should be complete - N.B. we are dealing with tissue recovery rather than DNA repairCorrection for incomplete repair is possible (Thames)In general, time between fractions for most tissues should be >6 hoursSome tissues, such as CNS, recover slowly making b.i.d. treatment inadvisableBentzen - Radiother Oncol 53, 219, 1999CHART analysis HNC showed that late morbidity was less than would be expected assuming complete recovery between fractionsIs the T1/2 for recovery for late responding normal tissues hrs?Fractionation spare late tissues
33 Regeneration in Normal Tissues The lag time to regeneration varies with the tissueIn acute responding tissues,Regeneration has a considerable sparing effectIn human mucosa, regeneration starts days into a 2Gy Fx protocol and increases tissue tolerance by at least 1Gy/dyProlonging treatment time has a sparing effectAs treatment time is reduced, acute responding tissues become dose-limitingIn late responding tissues,Prolonging overall treatment time beyond 6wks has little effect, butprolonging time to retreatment may increase tissue tolerance
34 Repopulation in Tumor Tissue Rat rhabdosarcomaHuman SCC head and neckT2 T3705540local controlTotalDose(2 Gy equiv.)no local controlFractionation and time prolongationTreatment DurationHermens and Barendsen, EJC 5:173, 19694 weeks to start of accelerated repopulation.Thereafter T1/2 of 4 days = loss of 0.6Gy per dayWithers, H.R., Taylor, J.M.G., and Maciejewski, B.Acta Oncologica 27:131, 1988Treatment breaks are often “bad”
35 Regeneration assumed to be exponential S.F.regeneration = eT = e (ln2/Tp)TWhere T = overall treatment time; Tp = effective doubling timei.e. S.F. = e-(D+D2)+ln2/Tp(T-Tk)Where Tk is time of start of regeneration
36 Altered Fractionation or How to optimally distribute dose over time
37 Players Total dose (D) Dose per fraction (d) Interval between fractions (t)Overall treatment time (T)Tumor typeAcute reacting normal tissuesLate reacting normal tissues
38 TCP or NTC TCP or NTC Dose Tumor control Late responding tissue complicationsComplication-free cureTCPor NTCHyperfractionationAcceleratedFractionationTCPor NTCDose
39 Other Sources of Heterogeneity Biological DoseCell cycleHypoxia/reoxygenationClonogenic “stem cells” (G.F.)NumberIntrinsic radiosensitivityProliferative potentialDifferentiation statusPhysical DoseNeed to know more about the importance of dose-volume constraintsDoseoxichypoxicS.FPhillips, J Natl Cancer Inst 98:1777, 2006
40 TCP/NTCP and Heterogeneity 1001008080SF2= 0.5N=109TCP (%)N=109TCP (%)60SF2=0.760N=1010SF2=0.640SF2=0.540N=1011SF2=0.4Average20SF2=0.320102030405060708090100110120130140102030405060708090DOSE (Gy)DOSE (Gy)Rafi SuwinskiIn order to cure a tumor, the last surviving clonogen must be killed, which is a probability function of dose.TCP = e-(m. SF) or e-m.e-(ad+bD2)Where m is the initial number of clonogenic cellsTCP=37% when, on average, 1 cell survivesSlope of curve represents radiobiological heterogeneityAlternative or supplemental indicator of treatment outcome
41 Heterogeneity within and between between tumors in dose-response characteristics, often resulting in large error bars for valuesIn spite of this, the outcome of clinical studies of altered fractionation generally fit the models, within the constraints of the clinical doses used
42 Definitions Conventional fractionation Hyperfractionation Daily doses (d) of 1.8 to 2 GyDose per week of 9 to 10 GyTotal dose (D) of 40 to 70 GyHyperfractionationThe number of fractions (N) is increasedT is kept the sameDose per fraction (d) less than 1.8 GyTwo fractions per day (t)Rationale: Spares late responding tissuesConventional empirically developed FletcherRadiosensitive tumors can be controlled with low doses (seminoma and lymphoma), low incidence of normal tissue damageGBM very radioresistantMost tumors intermediate sensitivity SCC, adenocaTumor size also plays a roleConformal radiotherapy: dose escalation with sparing of normal tissues but when done in a conventional way, lengthening OTTHyperfractionation: escalate dose, improve tumor control without increasing risk of late complications.
43 Definitions Accelerated fractionation Hypofractionation Shorter overall treatment timeDose per fraction of 1.8 to 2 GyMore than 10 Gy per weekRationale: Overcome accelerated tumor repopulationHypofractionationDose per fraction (d) higher than 2.2 GyReduced total number of fractions (N)Rationale: Tumor has low a/b ratio and there is no therapeutic advantage to be gained with respect to late complicationsExceptions of tumors with low a/b: melanoma, prostate, liposarcomaApplied in the palliative setting, limited life expectancy, late side effects not an issueModerate hypofractionation used in some countries, total dose usually lower but OTT also shorter which may compensate for the expected reduction in local tumor controlA way to escalate dose in trials of CRT? SIBAccelerated fractionation:early normal tissue reactions are expected to increase. If interval between fractions is long enough late normal tissue side effects should be the same or less if fractionsize is lower than 1.8 or 2 Gy and/or total dose is decreased
44 Very accelerated with reduction of dose Conventional70 Gy - 35 fx - 7 wksHyperfractionated81.6 Gy - 68 fx - 7 wksVery accelerated with reduction of dose54 Gy - 36 fx - 12 daysModerately accelerated72 Gy - 42 fx - 6 wks
45 Hyperfractionated Barcelona (586), Brazil (112), RTOG 90-03 (1113), EORTC 22791 (356), Toronto (331) Very accelerated CHART (918), Vancouver (82), TROG (350),GORTEC (268)Moderately accelerated RTOG (1113), DAHANCA (1485), EORTC (512) CAIR (100), Warsaw (395)Other EORTC (348), RTOG (210)7623 patients in 18 randomized phase III trials !!HNSCC only will be discussed
46 Scatter plot of selected altered fractionation schedules tested in randomised controlled trials according to the dose per fraction employed and the rate of dose accumulation. The Manchester schedule is included for comparison. The trial codes and the corresponding literature references are: 22791: European Organization for Research and Treatment of Cancer (EORTC) trial, 22851: EORTC trial, CHART, DAHANCA, Gliwice I and II : CAIR with 2.0 and 1.8 Gy/F, respectively, GORTEC 9402, Pinto: Radiation Therapy Oncology Group (RTOG) RTOG (HF: hyperfractionation, CB: concomitant boost, SC: accelerated split-course.Bernier and Bentzen EJC 39:560, 2003
47 EORTC hyperfractionation trial in oropharynx cancer (N = 356) Oropharyngeal Ca T2-3, N0-1Horiot 199280.5 Gy - 70 fx - 7 wks control: 70 Gy fx wksSURVIVALp = 0.08LOCAL CONTROLp = 0.02Pooled grade 2 and 3 side effectsIncrease of about 19 %in long term local tumor controlInterfraction interval 4 to 6 hoursYearsYears
48 Very Accelerated: CHART (N = 918) Dische 199754 Gy - 36 fx - 12 days control: 66 Gy - 33 fx wksLoco-regional controlSurvival12 consecutive days, 3 fractions per day, interval 6 hours, 1.5 Gy, total dose 54 Gy, total dose is lower to remain within tolerance of acutely responding tissues918 patientsOTT reduced by 33 days, total dose is 12 Gy less but LC is the same.conventional CHARTconventional CHARTFavourable outcome with CHART: well differentiated tumors larynx carcinomas
49 CHART: MorbidityDische 199754 Gy - 36 fx - 12 days control: 66 Gy - 33 fx wksModerate/severe subcutaneous fibrosis and oedemaP = 0.04Mucosal ulceration and deep necrosisP = 0.003Moderate/severe dysphagiaP = 0.04Laryngeal oedemaP = 0.009Mucositis occured earlier but settled sooner as well, skin reactions were less severe.
50 Moderately Accelerated Overgaard 2000DAHANCA 6: only glottic, (N = 694) DAHANCA 7: all other sites, + nimorazole (N = 791)66-68 Gy fx - 6 wks control: Gy fx - 7 wksActuarial 5-year ratesLocal control DAHANCA 6 DAHANCA 7Nodal control DAHANCADisease-specific survival DAHANCA 6 + 7Overall survivalLate effects (edema, fibrosis)5 fx/wk 6 fx/wk73% 81% p=0.0456% 68% p=0.00987% 89% n.s.65% 72% p=0.04n.s.
51 Moderately Accelerated CAIR: 7-day-continuous accelerated irradiation (N = 100)Skladowski 200066-72 Gy fx - 5 wks control: Gy fx - 7 wks Gy fx wks control: Gy fx wksOVERALL SURVIVALCONTROLCAIRlog-rank p=Follow-up (months)Probability
52 Total dose (Gy) Treatment time (days) CAIR: two schedules of continuous 7 days / weekwith different dose per fractionTotal dose (Gy)Maciejewski 1996, Skladowski 200080excessive mucosal toxicity (2.0 Gy/day)70acceptable mucosal toxicity (1.8 Gy/day)605040conventional3020107142128354249Treatment time (days)Wygoda, A. IJROBP 2008
53 RTOG 90-03, Phase III comparison of fractionation schedules in Stage III and IV SCC of oral cavity, oropharynx, larynx, hypopharynx (N = 1113)Fu 2000Conventional70 Gy - 35 fx - 7 wksHyperfractionated81.6 Gy - 68 fx - 7 wksAccelerated with split67.2 Gy - 42 fx - 6 weeks (including 2-week split)Accelerated withConcomitant boost72 Gy - 42 fx - 6 wks
57 Acute effects in accelerated or hyperfractionated RT Toxicity of RT in HNSCCAcute effects in accelerated or hyperfractionated RTAuthor Regimen Grade 3-4 mucositisCont ExpHoriot (n=356) HF 49% 67%Horiot (n=512) Acc fx + split 50% 67%Dische (n=918) CHART 43% 73%Fu (n=536) Acc fx(CB) 25% 46%Fu (n=542) Acc fx + split 25% 41%Fu (n=507) HF 25% 42%Skladowski (n=99) Acc fx 26% 56%
58 Altered fractionation in head and neck cancer: meta-analysis Bourhis, Lancet 2006Randomized trials (no postop RT)15 trials included (6515 patients)Survival benefit: 3.4% (36% % at 5 years, p = 0.003) Loco-regional control benefit: 7% (46.5% % at 5 years, p < )
59 Conclusions for HNSCCHyperfractionation increases TCP and protects late responding tissuesAccelerated treatment increase TCP but also increases acute toxicityWhat should be considered standard for patients treated with radiation only?Hyperfractionated radiotherapyConcomitant boost accelerated radiotherapyFractions of 1.8 Gy once daily when given alone, cannot be considered as an acceptable standard of careTCP curves for SSC are frustratingly shallow … selection of tumors?
60 Conclusions for HNSCCThe benefit derived from altered fractionation is consistent with can be of benefit but should be used with careIn principle, tumors should be treated for an overall treatment time that is as short as possible consistent with acceptable acute morbidity, but with a dose per fraction that does not compromise late responding normal tissues, or total dose.Avoid treatment breaks and treatment prolongation wherever possible – and consider playing “catch-up” if there are anyStart treatment on a Monday and finish on a Friday, and consider working SaturdaysNever change a winning horse!
61 Other Major Considerations Not all tumors will respond to hyper or accelerated fractionation like HNSCC, especially if they have a low a/b ratio.High single doses or a small number of high dose per fractions, as are commonly used in SBRT or SRS generally aim at tissue ablation. Extrapolating based on a linear quadratic equation to total dose is fraught with danger.Addition of chemotherapy or biological therapies to RT always requires caution and preferably thoughtful pre-consideration!!!Don’t be scared to get away from the homogeneous field concept, but plan it if you intend to do so.
62 Questions: The Radiobiology Behind Dose Fractionation
63 Random events occurring in cell nuclei 109. A basic assumption in modeling of radiation responses is that lethal ionizing events areRandom events occurring in cell nucleiRandom events in space as defined by the Poisson distributionA Gaussian distributionLogarithmic dose response curves#2 – This mathematical assumption is the basis of the log linear survival curve
64 Is a measure of the shoulder of a survival curve 110. D0 isIs a measure of the shoulder of a survival curveIs the mean lethal dose for the linear portion of the dose-response curveRepresents the slope of the log linear survival curveIs constant at all levels of radiation effect#3 – It is the mean lethal dose which is also 1/slope.
65 The inverse of the terminal slope of the survival curve 111. Dq isThe inverse of the terminal slope of the survival curveA measure of the inverse of the initial slope of the survival curveA measure of the shoulder of the survival curveA measure of the intercept of the terminal portion of the survival curve on the y axis#3 – The terminal slope is extrapolated back to the x axis.
66 112. If Dq for a survival curve is 2Gy, what dose is equivalent to a single dose of 6Gy given in 2 fractions, assuming complete repair and no repopulation between fractions.4 Gy6 Gy8 Gy10 Gy#3 – When dose is fractionated Dq is repeated, so it is 6+2Gy.
67 113. If hematopoietic bome marrow stem cells have a Do of 1Gy, and there is no shoulder on the survival curve, what fraction will survival a lethal dose of 6.9Gy?0.00010.0010.010.37#2 – If Do is 1Gy, D10 is 2.3Gy i.e. 3xD10.
68 114. If 90% of a tumor is removed by surgery, what does this likely represent in term of radiation dose given in 2 Gy fractions?1-2 Gy3-4 Gy6-10 Gy10-20 Gy20-30 Gy#2 – The eDo for fractionated radiation is around Gy and the eD10 will be 2.3 times this.
69 115. What is true for the ratio It is unitless It is a measure of the shoulder of the survival curveIt measures the sensitivity of a tissue to changes in size of dose fractionsIt is the ratio where the number of non-repairable lesions equals that for repairable lesions#3 – Low ratios reflect the sensitivity of late responding tissues to fractionation and high ratios the lack of sensitivity of acute responding tissues.
70 Unrepairable DNA double strand breaks Lethal single track events 116. The alpha component in the linear quadratic formula for a survival curve can be thought of as representingUnrepairable DNA double strand breaksLethal single track eventsMultiply damaged sites in DNADamage that can not be altered by hypoxia#2 – The beta component may be thought of as representing intertrack accumulated damage
71 The extrapolation number 117. Which parameter contributes most to cell killing in standard clinical fractionated regimens in RTThe ratioDoAlphaBetaThe extrapolation number#3 – Single lethal hits predominate at low doses (2Gy).
72 118. If cells have a Do of 2 Gy, assuming no shoulder, what dose is required to kill 95% of the cells?6 Gy12 Gy18 Gy24 Gy30 Gy#1 – 3xDo or e-3 = 0.05
73 Dependent on the size of the dose per fraction 119. The extrapolation number N for a multi-fraction survival curve, allowing complete repair between fractions and no repopulation is1< 1>1Dependent on the size of the dose per fraction#1 – It is a straight log-linear curve with the slope extrapolating to the x/y intersect at 1.0.
74 Dependent on the size of the dose per fraction 120. The extrapolation number N for a single dose neutron survival curve is1< 1>1Dependent on the size of the dose per fraction#1 – It is a straight log-linear curve with the slope extrapolating to the x/y intersect at 1.0.
75 121. The extrapolation number N for a low dose rate survival curve is < 1>1Dependent on the size of the dose per fraction#1 – It is a straight log-linear curve with the slope extrapolating to the x/y intersect at 1.0.
76 122. The inverse of the slope of a multifraction survival curve (effDo) for x-rays is generally within the rangeGyGyGyGy#3 – This obviously has a lot of assumption, but is not a bad ‘ball-park’ figure to remember.
77 123. If the effDo for a multifraction survival curve is 3 123. If the effDo for a multifraction survival curve is 3.5 Gy, what dose would cure 37% of a series of 1cm diameter tumors (109 clonogens).56 Gy64 Gy72 Gy80 Gy#3 – The eD10 would be about 8Gy (2.3x3.5Gy), so 72Gy would reduce survival to on average 1 surviving cell or e-1 and would give 37% cure. Or TCP= e-m.SF
78 124. If the effDo for a multifraction survival curve is 3 124. If the effDo for a multifraction survival curve is 3.5 Gy, what dose would cure 87% of a series of 1cm diameter tumors (109 clonogens).56 Gy64 Gy72 Gy80 Gy#3 – 2 more eDo doses would reduce survival from 1 to e-2 or cells/tumor. TCP = e = 87%
79 125. If a tumor has an effective Do of 3. 5 Gy, what is the S. F 125. If a tumor has an effective Do of 3.5 Gy, what is the S.F. after 70 Gy?2 x 10-112 x 10-92 x 10-72 x 10-52 x 10-3#2 – The eD10 would be about 8Gy (2.3x3.5Gy), so 70Gy would reduce survival to about 2 x 10-9.
80 126. If 16 x 2 Gy fractions reduce survival by 10-4, what dose would be needed to reduce survival to 10-10?50 Gy60 Gy64 Gy70 Gy80 Gy#5 – x 2Gy = 80Gy
81 127. If 16 x 2 Gy fractions reduce survival by 10-4, what is the effective Do? #4 – The eD10 would be about 8Gy, so eDo would be 3.5Gy.
82 128. The ratio for mucosal tissues is closest to 1 Gy 3 Gy 5 Gy #4 – Acute responding tissues have a high ratio.
83 129. Which of the following human tumors has recently been thought to have an ratio of 1-2 Gy Oropharyngeal CaProstate CaGlioblastomaColorectal Ca#2 –Several studies have suggested this and therefore that hypofractionation may be of value.
84 130. If tissue tolerance is 60Gy at 2 Gy/fraction and 40 Gy at 4Gy/fraction, what is its a/b ratio? #2 – Dnew (dnew + ) = Dold (dold +)
85 131. It is decided to treat a patient with hypofractionation at 3 Gy/fraction instead of the conventional schedule of 60 Gy in 2 Gy fractions. What total dose should be delivered in order for the risk of late normal‑tissue damage to remain unchanged assuming an a/b for late damage of 3 Gy?40 Gy48 Gy50 Gy55.4 Gy75 Gy#3 – Dnew (3 + 3) = 60 (2 +3) = 50Gy
86 132. Hyperfractionation using a fraction size of 1 132. Hyperfractionation using a fraction size of 1.2 Gy is replacing a standard 70Gy in 2Gy fractions for HNSCC. Assume full repair of sublethal damage between fractions and an a/b of 3 Gy, what total dose should be used to maintain the same level of late complications?42 Gy58 Gy70 Gy83 Gy117 Gy#4 – Dnew ( ) = 70 (2 +3) = 83Gy
87 133. A standard treatment of 70 Gy in 2 Gy/fraction is changed to 83Gy in 1.2 Gy. Assuming no proliferation and complete repair between fractions, an a/b of 3 Gy for late responding tissue and 12 Gy for tumor, what would be the therapeutic gain.6%12%18%24%#2 – The response of the tumor is not going to change much, so you can guess 83/70 = 12%
88 134. Which of the following sites is the least suitable for b. I. d 134. Which of the following sites is the least suitable for b.I.d. treatmentHead and neckBrainLungProstate#2 – The brain does not respond well to b.i.d. treatment
89 135. The rationale behind accelerated fractionation is To spare late responding normal tissueTo combat encourage tumor reoxygenationTo exploit redistribution in tumorsTo combat accelerated repopulation in tumors#4 – The idea is to get the dose in during the lag time before accelerated repopulation starts.
90 136. The CHART regimen for HNSCC of 54Gy in 36 fractions over 12 days compared with 66 Gy in 33 fractions in 6.5 weeks, overall showedSuperior locoregional control, no increase in overall survival, increased late effectsSuperior locoregional control that translated into an increase in overall survival, no change in late effectsNo change in locoregional control and overall survival, decreased late effectsSuperior locoregional control, no increase in overall survival, increased acute effects#3 – The aim of this trial was not to increase response but to decrease normal tissue reactions, unlike a later NSCLC CHART trial
91 Was a hyperfractionation trial 137. DAHANCA 6 and 7 clinical trials with 66-68Gy given in 6 compared to 7 weeksWas a hyperfractionation trialInvolved treating patients 6 days a weekShowed no increase in local controlShowed no increase in disease-specific survival#2 – with better outcomes…
92 138. RTOG compared hyperfractionation, accelerated fractionation with a split, and accelerated fractionation with a boost. It showedHyperfractionation to be superior in terms of loco-regional control and late effectsAccelerated fractionation with a split to be equivalent to hyperfractionation in terms of loco-regional controlThere to be no advantage to altered fractionationAccelerated fractionation to be superior to hyperfractionation#1 – The lead investigator was K. Fu.