Gabriel K. Y. Lam

Explanation of Dose-Rate and Split-Dose Effects on Mouse Foot Reactions Using the Same Time Factor'

The effect of dose rate on X-ray-induced skin reactions was determined for 1, 2, and 10 fractions using the mouse foot system. Data for 1.6, 0.15, and 0.06 Gy/min were obtained, and the time-course-dependent effects of these experiments were analyzed using an extension of the formulation by L. G. Lajtha and R. Oliver (Br. J. Radiol. 34, 252-257, 1961). r, the time for decay of the effective dose to 1/e of its initial value, was determined to be 114 + 28 min. On the basis of this analysis, predictions were made for split-dose and pulsed irradiations. These were consistent with experimentally measured results obtained using such irradiation protocols. Thus it has been shown that dose fractionation, low-dose-rate, split-dose, and pulsed irradiation effects can be handled conceptually as different particular examples of variable time-course irradiations. In addition, this analysis has been used to generate guidelines for minimum acceptable dose rates consistent with acute exposure and minimum fraction intervals consistent with complete repair, to assist in the design of future multifraction experiments using the mouse foot system.

The Survival Response of a Biological System to Mixed Radiations

The concept of additive radiation action is applied to the process of merging of the intermediate lesions at a common stage in the radiation inactivation pathways for lesions produced by different radiations. This gives rise to a natural nonindependent effect for combined irradiation. Even though the exact nature of this common intermediate lesion is unknown, the effect of this lesion additivity can still be formulated into a mathematical model using the assumptions: (1) there exists a stage in the chain of radiation inactivation events where different types of lesion precursors, produced by different types of radiations in a mixture, inflict lesions which become functionally indistinguishable and hence additive thereafter, to produce the same end point observed; (2) all precursors of all types are simultaneously competing for the opportunity to inflict lesions at the stage indicated in assumption 1, and each precursor has equal opportunity regardless of its origin; (3) if the radiations are delivered sequentially within a sufficiently short time, the lesion precursors of both radiations arrive at the above stage at about the same time and hence inflict lesions which are additive as described in assumptions 1 and 2. The model is quantitative but contains no free-fitting parameters. It is shown to be capable of explaining a large variety of apparently unrelated published experimental results observed for mixtures of high- and low-LET radiations.

Practical and theoretical considerations in the use of the mouse foot system to derive epithelial stem cell survival parameters

A series of experiments investigating the effects of 1, 2, 4, 10, and 20 dose fractions of x rays on mouse foot skin are reported. These include data for levels of acute gross reactions from dry to full moist desquamation. We have shown the following for the irradiated epidermal stem cell population of the mouse foot. (1) Detectable proliferation was not found for regimes employing 10 daily dose fractions. (2) Significant interexperimental variation occurred which did not result from changes in cellular radiosensitivity. (3) The cell-survival equation obtained from data limited to reactions below the point of moist desquamation was not different from that obtained from higher reaction data. (4) The absolute target cell-survival levels corresponding to our mean reaction scale were established using published skin clone data and extended down to 4 x 10/sup -8/, and the clonogenic population was estimated as 0.6 to 3.0 x 10/sup 7/. (5) The relative radiobiological usefulness of the various levels of acute reactions has been characterized. (6) Skin-reaction scores obtained by averaging reactions over 7 days encompassing the peak response gave the same results with fewer readings than the conventional 22-day averaging period.

Pions and pig skin: Preclinical evaluation of RBE for early and late damage

The skin of 50 pigs has been irradiated with negative pi mesons and with X rays in order to determine the RBE for early epidermal and later dermal damage. Late fibrosis was not studied. Four, 7, 9 and 10 fractions were used. An estimate of the RBE was made from the reactions on each pig for both early and late damage so that interanimal variability would be avoided. The data were also averaged to obtain mean dose response curves. There was no tendency for higher RBEs for late than for early skin damage. These pig studies have demonstrated an RBE of about 1.5 for early epidermal reactions and a slightly lower RBE (approximately 1.4) for later dermal damage in the same animals. This indicates that at doses of about 2.0 to 3.5 Gy pions, the medium wave skin damage is unlikely to be more severe than would be predicted from the early skin reactions and the accumulated clinical experience with X rays. However, if the trend to a steeper slope for the RBE versus dose per fraction for late injury is correct, as indicated by other published studies a relative increase in the late injury might be expected if much lower doses per fraction are used. The present clinical studies at Vancouver using 15 X 2.1 Gy pions indicate that an RBE of 1.5 is appropriate for epithelia, brain and colorectum.

THE USE OF POSITRON EMISSION TOMOGRAPHY IN PION RADIOTHERAPY

The radioactive debris produced by pion radiotherapy can be imaged by the technique of Positron Emission Tomography (PET) as a method of non-invasive in situ verification of the pion treatment. This paper presents the first visualization of the pion stopping distribution within a tumor in a human brain using PET. Together with the tissue functional information provided by the standard PET scans using radiopharmaceuticals, the combination of pion with PET technique can provide a much better form of radiotherapy than the use of conventional radiation in both treatment planning and verification.

Radiobiology of pions at TRIUMF

Radiobiological studies at TRIUMF of the effects of pion beams have been carried out using cultured cells, mice and pigs. CHO cells in gel/medium were used for RBE determinations throughout the dose distributions. The RBE was shown to increase with depth in the stopping region, while the average RBE value decreased with increasing width of the stopping peak and also increased with increasing field size. The results suggest that the effects of peak width and field size will approximately cancel out when treatment volumes are changed, provided that all three dimensions are changed more or less proportionally. The peak center RBE values ranged from 1.2-1.3 for single doses. The oxygen enhancement ratio (OER) for pions was 2.2 compared to 2.8 for X rays. Split dose recovery was found to be somewhat reduced in the peak region. In vivo studies with mouse and pig skin have been done for 1 to 20 fractions. For mouse skin, pions showed an RBE of 1.5-1.6 for fraction numbers greater than 10. The RBE for pig skin was 1.4 for early reactions, but it may be higher for medium term reactions.

Method of analysis to derive cell survival from observation of tissue damage following fractionated irradiation.

A general analysis technique is introduced for extracting in vivo cell survival equation parameters from observation of tissue damage following fractionated doses of radiation. This technique has been applied to a new set of mouse foot skin-reaction data and it has been possible to demonstrate that the parameters for an assumed survival equation can be determined independently of the arbitrary definition of the average skin reaction. Since all data points are used, no graphical data interpolation is required and the use of an arbitrary level of damage for isoeffect comparison is not necessary.

Range modulation in proton therapy - an optimization technique for clinical and experimental applications

A fast optimization algorithm for range modulation in clinical and experimental applications of proton therapy is described. The method is versatile towards the number of parameters provided for range modulation, i.e. the trade-off between accuracy and simplicity of the latter can be chosen freely. The approach is, therefore, adaptable to most operating proton therapy facilities. It requires only a few basic measurements as input data and results in a depth dose uniformity of better than 2%. A typical calculation takes less than 90 s on a DEC VAXstation 3100, FORTRAN. The method has been extensively tested at the TRIUMF Proton Therapy Facility in Vancouver, BC.

Detection of pion‐induced radioactivity by autoradiography and positron emission tomography (PET)

An autoradiographic technique incorporating a new imaging system was used to detect pion-induced radioactivity in Plexiglass and the results were compared with aluminium activation and PET imaging. The activity distribution in the region of the pion-stopping peak was similar in all three cases. Another large signal in the entrance region due to in-flight interactions [12C(pi-, pi- n) 11C] was detected by autoradiography and by PET but was not reflected in the aluminium activation measurements. This new technique is capable of defining the stopping region in phantoms with a better resolution than PET scanning and is useful as a complementary technique to other methods of pion dosimetry.

Analysis of interaction for mixtures of agents using the linear isobole

The linear isobole that is commonly used as a reference for the study of interaction is derived from the interaction of an agent with itself. It is shown that the general use of the linear isobole in the study of the combined effects of mixtures of agents implies interaction between the agents whether the dose-effect curves of the agents are the same or not. It is difficult to generalize the interaction between two doses of the same agent to the interaction between two doses of different agents with different action mechanisms without the use of a mechanistic model. Predictions using non-interaction defined as independent action are generally different from those using linear isobole. A simple mechanistic framework based on the concept of common intermediate lesions is introduced in this paper to relate these two methods used for the analysis of synergism and antagonism. In this framework of lesion development, two agents that have no common intermediate lesion in their action will be non-interactive (referred to as independent action). When the two agents share a common intermediate, it is shown that the combined effect will follow the linear isobole (referred to as common action). This simple framework of analysis is applicable to the general study of interaction between two agents with different types of dose-effect curves.