Cell Survival vs Irradiation dose

Transcription

1 What have we learned from Radiation Treatment, and what I think will be the next generation ion therapy accelerators? S.Y. Lee, Department of Physics, Indiana University Cell Survival vs Irradiation dose Colonies counted Survival fraction (SF) is defined as: Surviving fraction cells seededpe/1 Plate Efficiency (PE) is the percentage of cells that grow into colony! The SF vs radiation dose has a characteristic curve: For more than 1 years of applying radiation to medicine, we gain some controls on cancer. These experiences provide our knowledge on radiobiology and modality of radiotherapy, which is an evolving field or research and development. I will briefly discuss the transformation of IU Cyclotron Facility to Med-West Proton Research Institute (MPRI). Since then, there are new facilities in Illinois, Minnesota, Pennsylvania, etc. An outstanding medical center must also function as an advanced research center. Accelerator design to fulfill these research goals will be of great interest in our efforts. Dr. Chee-Wai Cheng had visited Chang-Gung in Statistical nature of survival curve of cells 1. Single-Hit survival curve Determining D, D q and n Consider N-biological cells. Given a small dose of radiation dd, a small number dn will be inactivated by radiation is 1 dn NdD, D N N exp( D ) D The mean lethal dose D is the dose that reduces the population by 37%. 2. Multi-target single hit survival curve Suppose the cell has n targets, each of which must be in-activated to cause cell death. The probability for a single cell not being hit is exp( D/D ), and the probability that any individual site being hit is [1-exp( D/D )]. The probability that all n-sites being hit is [1- exp( D/D )] n. Thus the probability of the survival of a cell is N N n 1 exp( D/ D ) N nexp( D/ D ) 1 DD The quasi-threshold dose is defined as 1 nexp( Dq / D ); Dq D lnn Characteristics of 28 kvp X-ray is D q ~2.4 Gy; D ~1.45 Gy; n~exp(2.4/1.45)~ Noted that the single-hit-multi-site model has zero-slope at small dose! 4 Dose (Gy) Dose (Gy) Number of surviving cells per cm 2 of skin D Number of surviving cells per cm 2 of skin D q n = exp[d q / D ] D q D

2 3. Single-hit + multi-targets models Contrary to most observations, the single-hit-multi-site model has zero-slope at small dose! A model of cell survival with an initial slope is N N exp( D/ DS ) 1 1exp( D/ D ) The parameter D S defines the initial slope of the survival curve. a. If D S >>D, then D define the final slope. b. If D S is the same order as D, then the final slope is determined by D D S /(D + D S ). 4. Linear-quadratic (LQ) model: The biological effect depends also on dual radiation effects 2 N N exp[ D D ] N (SF) Linear(SF) quadratic Parameters α and β can be inferred through multifraction experiments: 2 SF exp[ d d ] lnsf lnsf d nd D Plotting ln(sf)/d vs d (dose per fraction), parameters α and β can be obtained. These parameters are important in radiation therapy! n Early (X-ray) Late α/β (Gy) α/β (Gy) Skin 9-12 Spinal cord Jejunum 6-1 Kidney Colon 1-11 Lung Testis Bladder Callus 9-1 n Effective single dose survival curve reconstructed from muftifraction experiments for colonogenic cells of the jejunal crypts of mice. The numbers on the curve refer to the number of fractions used to re-construct that part of the curve. 1. The initial and the final slopes are 3.57 and 1.43 Gy respectively. 2. The quasi-threshold is Dq=4.3 Gy. The data can be equally well fitted by the linear-quadratic formula (dashed line). [Thames HD, et al, Int. J. Rad. Oncol. Bio. Phys. 7, 1591 (1981).] 1. Hadron beams have the advantage of dosage (energy deposit) in a localized solid tumor (Bragg peak). This property will be less damage to the noncancerous part of the body. 2. The heavy-ion beam has also the advantage of high LET or higher RBE, and less dependence on OER in a cancerous region, 3. Charged particle beam is easier in control, manipulation, and measurement of its dosage : R. Wilson propose to use protons in radiotherapy. R.R. Wilson, Radiology 47, (1946) source-to-surface distance (SSD). Second malignancy rates (MGH patients statistics) Chung et al. ASTRO % of proton patients (32 patients) 12.8% of photon patients (23 patients) Even with the advancement of IMRT/IGRT, the hadron beam holds advantage of treatment local control. Particularly, Hadron Pencil Beam Scanning (PBS) treatment plans. The lateral beam profile (left) and the depth physical dose distribution (right) of the typical spot beam. The 3D dose distribution measured by the multilayer ionization chamber

3 Straggling from Multiple Particle Paths 1 MeV electrons 5 histories 8 MeV protons 5 histories 15 MeV/n carbon ions 5 histories 4. Because of multiple-scattering, micro-beam probe is not as practical as one wishes for proton therapy, however, the micro-beam may be a useful feature for heavy ion beams. Central-axis depth dose distributions for the pencil beam algorithm compared with measurements using a diode. The beam is collimated by a 2.4 mm radius aperture. The air gap between the aperture and the water tank is 1 or 6 cm. 1. Proton beams Multiple Coulomb scattering produces beam penumbra rms spread! 2. Heavy ion nuclear breakup interaction produces distal penumbra! Linda Hong et al., Phys. Med. Biol. 41 (1996) Multiple non-overlapping pencil beams can produce Bragg-peak, and SOBP. Spread-out-Bragg-Peak 1. Energy degrader or 2. Changing energy in synchrotrons 1. The entrance dose increases with the size of SOBP. Even so, the entrance dose is still much smaller than that of IMRT 2. The entrance dose may be reduced by reducing the SOBP size in creative treatment plan, yet to be developed! Therapeutic Ratio In early days of radiotherapy, it was usually assumed that normal cells were less sensitive to radiation than the tumor cells, but vast majority of radiological studies indicate that D values for different mammalian cells irradiated with X or γ rays cluster quite closely around a value of 1.3 Gy. There seems to be a wider variation in Dq or n. In the LQ-model, they have similar α/β! There is no evidence that malignant cells have different sensitivity to radiation than the normal cells. Most tumor and critical normal tissue combinations, the curves A and B are much closer, even reversed. Radiobiology is to find ways of improving the therapeutic ratio, e.g. IMRT, IMPT, multi-fields treatment plan, etc. Hadron beams hold advantage over photon beams over the reduction of energy deposit to normal tissue by at least a factor of 2! Careful treatment plan and precision will be important to take this advantage to the limit!

4 RADIOBIOLOGY -- LET / RBE / OER The linear energy transfer (LET) represents the energy absorbed per unit length in a medium. It is imperative that the cell killing effectiveness depends on LET! The unit commonly used for the LET is measured in kev/μm, which is equivalent to ev/nm. RBE: Radiation Biological Effectiveness RBE D D ion In dosimetry, the RBE is by the radiation weighting factor, (W R ), or formerly, the quality factor (Q). These weighting factors, arrived at by consensus of governments, industry, and regulators, convert absorbed dose (measured in units of grays or rads) into formal biological equivalent dose for radiation exposure (n sieverts or rem). RBE depends on many factors: tissue type and biological endpoint; radiation type; biological system; choice of radiation standard ; radiation dose and dose rate; number of dose fractions (& dose per fraction). Cell Survival and LET Survival Fraction RBE=3. Optimum LET and Double Hits 1 2 RBE=1.5 RBE=2.6 Dose (Gy) Dose (Gy)

5 Carbon The oxygen effect The biological effects of the radiation are greater in the presence of oxygen than in its absence. The Oxygen enhancement ratio (OER) is defined as the ratio of doses for isoresponse on hypoxic cell to that of cells at 1 atm of air pressure. Studies show that OER is about 1. for LET values greater than 2 kev/μm. OER and RBE vs LET Left: Cell survival curves for four representative human tumour cell lines irradiated at high dose rate. HXl 42 neuroblastoma; HX58 pancreatic; HXl 56 cervix; RTl 12 bladder carcinoma. (From Steel, G. G., Radiother. Oncol., 2, 71-83, Bottom: Using the data of LQ-fitted model, one study the Contributions to the surviving fraction at 2 Gy from the linear and quadratic components of radiation cell killing for 17 human tumour cell lines. α component = exp( αd) β component = exp( βd 2 ) RBE and OER are two un-certainty factor in the proton therapy. The dotted line indicates equality of the two components. (Steel, G. G. and Peacock, J. H., Radiother. Oneal., 15, 63-72, [(1) Lymphoma, myeloma, neuroblastoma; (2) medulloblastoma, small-cell lung cancer; (3) breast, bladder, cervic carcinoma; (4) pancrea, colo-rectal, squamous lung cancer; (5) melanoma, osteosacoma, glioblastoma.] Cell killing is dominated by the linear component of the survival curve!

6 Fractionation For over 1 century, radiation has been used to treat patients with malignant diseases. Fractionating the radiation treatments have results in a better therapeutic ratio for most tumors than giving the treatment in a single dose. In s, experiments in Paris showed that Rabbits and Rams could be sterilized with fractionations of X-ray dosages without damage to the skin (Claudius Regaud). In 1922, Henri Coutard showed evidence that the laryngeal cancer can be treated by the fractionated radiation without disastrous effects. In 1934, Coutard established fractionated radiation treatment plan, which is still being used today. In radiation therapy, the total dose is fractionated: 1. Fractionation allows normal cells time to recover, while tumor cells are generally less efficient in repair between fractions. 2. Fractionation also allows tumor cells that were in a relatively radio-resistant phase of the cell cycle during one treatment to cycle into a sensitive phase of the cycle before the next fraction is given. Tumor cells that were chronically or acutely hypoxic (and therefore more radioresistant) may reoxygenate between fractions, improving the tumor cell kill. When the radiation treatment is fractionated, it is found that a much greater total dose is required to achieve a given level of biological damage than that of a single dose. The recovery is a complex process, involving 4Rs: Repair of sublethal dosage; Redistribution (reassortment) of cells within the cell cycle; Repopulation, Reoxygenation. Fractionated Radiation treatment At a sublethal dose, surviving cells may Repair, Repopulation, Redistribution, and Reoxygenation (4Rs). Repopulation: Assuming exponential growth, the survival curve needs to multiply an exponential growth factor: exp( T I / T ) Here T I is the interval of time, T R is the growth time. Typical T R =7 days for the Human skin cells. In the model, each small jump is exp(1/7); while the Monday jump is exp(2/7). Survival of a population of 1 6 cells as a function of dose for a variety of dose fractionations. 1. Curve OAB: fraction of 12. Gy; 2. Curve OACDEFG 6 fractions each of 2. Gy; 3. Curve OA C'D'E'F'H 6 fractions each of 2. Gy assuming regrowth; 4. Curve OK, 12 fractions each of 1. Gy with regrowth. Treatment started on Monday. The extra regrowth before Monday's treatment is due to regrowth on the weekend. R Characteristics of 6Co photon: D =1.3Gy, D s =2.4 Gy, n=3, T R =7. d Isoeffect curves and Nominal Standard Dose 1944: Strandqvist found that if the dose lay above this curve, overdose effect or necrosis of the normal skin occurred, while if the dose lay below, under-doses effect or tumor recurrence appeared. The isoeffect curve follows the scaling of T.33. Skin necrosis isoeffect His results imply that 2 Gy in 1 day is equivalent to 3 Gy in 4 days, 4 in 11 days, 5 in 25 days, and 6 in 45 days. Similar results of Paterson and Friedman are also shown. This is called the isoeffect curves. Effects can be reduced by IMRT, IGRT to reduce dose in normal tissues. Skin erythema isoeffect Nominal standard dose (NSD) When the radiation treatment is fractionated, it is found that a much greater total dose is required to achieve a given level of biological damage than that of a single dose. The recovery is a complex process, involving 4Rs: Repair, Reoxygenation, Redistribution, and Repopulation. 1944: Strandquist produced the first clinical isoeffect curve, calculating the relations between time, dose, cure, and skin reactions in the treatment of skin cancers. 1967: Ellis et al. recognized that the time factor was actually dependent on both the overall treatment time and the number of fractions. He devised the nominal standard dose (NSD) formula tolerable to connective tissue: D = NSD N.24 T.11 where D = dose, N = number of fractions, and T = overall time (days). The NSD is a constant. When multiplied by time and number of fractions to the appropriate power yield the total dose in Gy. (may be called Gray Equivalent Therapy (GET)). To use NSD equation, one must determine the numerical value of the NSD. Difficulties: 1. The system is based on skin reaction data, it does not predict LATE EFFECTS. 2. Effects of the Proliferation of cells have not been taken into account! LQ approach to dose fractionation seems to be much more reliable than that of NSD

7 Proliferation Proliferation of tumor cells and fractionation early-responding tissues (e.g. skin, intestinal epithelium) late-responding tissues (e.g. lung, spinal cord) Extra dose required to compensate for tumor proliferation, yet control normaltissue damage, is sigmoidal over time: in short time frames, none needed; at some point, large amt. needed Extending radiotherapy time has little sparing effect on late reactions, but a large sparing effect on early reactions Extra dose required to counteract proliferation in the skin of mice as a function of time after starting daily irradiation with 3 Gy per fraction. The proliferation does not become significant until ~2 weeks after the start of daily fraction. A delay followed a rapid rise is typical in time. In mouse skin, the delay is about 2 weeks. In human it is about 4 weeks. Prolonging overall time within the normal radiotherapy range has little sparing effect on late reactions but a large sparing effect on Dose-Response : Early and Late Responding Tissues The dose-response relationship for late-responding tissues is more curved than for early responding tissues. late-responding tissues are more sensitive to changes in fractionation patterns. The / dose value for early responding tissues is larger than that of the late responding tissues. Linear-quadratic model: N N exp[ D D early reactions ] / ~ 3 Gy / ~ 1 Gy Early (X-ray) Late α/β (Gy) α/β (Gy) Skin 9-12 Spinal cord Jejunum 6-1 Kidney Colon 1-11 Lung Testis Bladder Callus 9-1 Lung Fraction size is the dominant determining late effects, while overall treatment has little influence. 2. By contrast, fraction size and overall treatment time determine the response of acutely responding tissues. Isoeffect curve vs dose per fraction. Isoeffect curves are steeper for late effects Late and Early Response In practice, treatment protocols requiring a few large fractions (to produce equal early effects) result in more severe late effects Isoeffect curves are steeper for late effects High-LET particles produce more severe late effects, Dose conformity and dose distribution are critical in its treatment plan As dose per fraction increases, late responding tissues can take less dose Fraction Size and Treatment Time Early effects: both fraction size and overall treatment time are important Treatment with any cytotoxic agent, including radiation, can trigger surviving cells (clongens) in a tumor to divide faster than before, known as accelerated repopulation. Late effects: number of fractions dominates overall treatment time has little influence 28

8 Withers and his colleagues estimated the dose to achieve 5% of tumor local control of the squamous cell tumors head and neck cancers TCD 5 as a function of the overall duration. The analysis suggests that the clonogen repopulation in human cancer accelerates at about 28 days after the initiation of radiotherapy! Instead of the Dashed line, based on 2-months clonogen doubling rate! A dose increment of.6 Gy/day is needed to compensate this repopulation! Accelerated repopulation The Dose-Rate Effect Chinese hamster cells and x rays broad shoulder dramatic dose-rate effect significant differences in biological effect survival decreases as dose rate increases! HBO: High-pressure oxygen trial; Miso: Misonidazole. 4 different human cell lines irradiated at a HDR and then a LDR Multiple Fractions Per Day Treatment prolongation Advantages: spare early reactions and allow reoxygenation in tumors Disadvantages: early reactions decrease, but no sparing of late injury; allows for surviving tumor cells to proliferate during treatment Multiple fractions per day minimizes treatment prolongation (treatment to be completed as soon as practical); Two distinctly different multiple-fraction/day strategies: hyperfractionation and accelerated treatment Hyperfractionation reduce late effects with equivalent tumor control (increases early effects) Overall treatment time remains conventional at 6 to 8 weeks but since two fractions are given daily, the total number of fractions is doubled (6-8) Further fractionation works as long as dose fraction is on curved portion of survival curve flexure dose (~.1 /) is where significant bending starts Accelerated Treatment reduce re-population in rapidly proliferating tumors when patients have fast growing tumors Involves a conventional total dose with a conventional fraction number but since two fractions a day are given, the overall time is nearly halved little or no change in late effects since number of fractions and dose/fraction are unaltered early response becomes limiting may need rest period during treatment 31 Fractionation Regimes Must ensure that the two fractions are separated by adequate time for repair of sublethal damage Current convention dictates that the period between doses should be at least 6 hours. Evidence comes from the Radiation Therapy Oncology Group. For a given total dose, the incident of late effect was worse for interfraction intervals less than 4 hrs compared with interfraction intervals longer than 6 hrs. There is a trade-off between relative benefits of: hyperfractionation to reduce late-effects accelerated treatment to improve tumor control T pot and Accelerated Treatment Classified by Fast (Tpot < 4 days) and slow (Tpot > 4 days) growing tumors; A study by the European Corporative Radiotherapy Group (EORTC) For slow growing tumors, no apparent difference between results of conventional and accelerated treatments For fast growing tumors, accelerated treatment results in substantially better local control; comparable to those obtained for slow-growing tumors 32

9 Calculating Equivalent Doses Biological-effect: For a single acute dose, D, the biological effect is E=D+D 2 For n well-separated fractions of dose d, the effect is E=n(d+d 2 )= nd (+d / [) = total dose) (relative effectiveness The total dose for n fractions of dose d is D=nd BIOLOGICAL EFFECTIVE DOSE: E/ = D (1 + d/[/]) Biologically effective dose = E/ Calculations are not to be Relative effectiveness = 1 + d/[/] considered a substitute for clinical Late-responding tissues, / ~ 3 Gy judgment and experience Early-responding tissues, / ~ 1 Gy Equivalency should not be compared between early and late effects For a given tissue, if the dose per fraction is changed, how must the total dose be adjusted? We are interested in a constant effective dose E for a total dose D and dose per fraction d in comparison with a reference total dose D ref and dose per fraction d ref. d dref D dref / D1 Dref 1, or / / Dref d / 33 Departures from the LQ-Model for doses below 1 Gy There is now a large body of experimental evidence that for many mammalian cell lines, the surviving fraction at doses below around 1. Gy is not well predicted by the LQ model. The Figure below shows the survival of asynchronous T98G human glioma cells between Gy and 6. Gy measured by Short et al. (1999). The cells exhibit increased radiosensitivity at doses up to around 1. Gy, and this phenomenon is known as low-dose hyper-radiosensitivity (HRS). This important behavior has only come to light relatively recently because of the difficulty of making measurements at doses less than 1 Gy; effective techniques involve either a so-called fluorescenceactivated cell sorter (FACS) to plate a predetermined number of cells, or microscopic scanning to identify an exact number of cells after plating. We know little about low dose effects: inducible vs protective mechanisms, threshold, priming, dose rate effects, LET within one system. Technological advances may permit much needed studies at low doses in the areas of both treatment and protection. HRS is very important in radiation oncology, e.g. effect of penumbra on normal cells, effects of IRR for fractionated doses, OER, brachytherapy, etc How about effects of LET, dose rate, threshold, radiation sensitizer drugs, OER.? HRS: The survival deficit from the LQmodel up to 2 Gy is plotted vs the SF2 (survival fraction at 2 Gy) for various cell lines including human cell lines in the inlet. There is much unknown about radiobiology of human cell lines. Research in this field will be very important in radiation oncology and radiation protection!

10 Expectation in 23 Patient count in PTCOG 12 Capacity and Demand Summary YR 1 YR 2 YR 3 YR 4 YR # of Treatment Rooms # of Fractions Delivered 1,958 7,316 12,857 16,395 19,556 # of Patients Treated Capacity Utilization 7% 7% 8% 8% 85% Loma Linda HIMAC MGH Jacksonville U Penn 4 # of Cancer Cases in Market Area 262,55 262, , , ,29 # Cases Suitable for Protons 39,587 39,654 39,721 39,789 39,856 # of Patients Treated Market Penetration.2%.8% 1.4% 1.8% 2.1% 2 Actual number of patients: 257 in Patients per year Cost and efficiency Loma Linda Nice Orsay HIMAC LBL BevaLAC 433 (Ne) 254 (α) (1957;74-92) GSI 44 patients ( ) HIBMC, Hyogo MGH MD Anderson UFPTI, Jacksonville RPTC, Munich Hadron-Therapy Centers HIT, Heidelberg HIT, Heidelberg UPenn, Philadelphia GHMC Gumma IMP, Lanzhou CDH Proton Center, Warrenville Medipolis MRI, Ibusuki CNAO, Pavia 42: PTC, Czech 43: SCCA, PTC WA Conclusion: 1. Chang-Gung University Hospital has played a leader role in providing hadron beam therapy. This effort needs vision and courage! CGU will continue to carry out advanced medical service and research. 2. Radiation therapy is an indispensable option in cancer therapy. Radiobiology plays a key role in advancing this therapy. For a century, radiation biology research has made much progresses, but there are still much to learn in human radiobiology. CGU s effort can provide excellent example for other medical facilities. 3. Looking forward to your successes! In Delivering High-Quality Cancer Care - Charting a New Course for a System in Crisis, (December 213, National Academic Press 18359): The National Comprehensive Cancer Care Network, the American Society of Clinical Oncology, and the American Society of Radiation Oncology have worked with clinical experts to develop guidelines for more than 135 cancers or processes of care. A quote: Knowing is not enough; we must apply. Willing is not enough; we must do. -- Goethe Many thanks for your attention.