PEREGRINE is a three-dimensional Monte Carlo dose calculation system written specifically for radiotherapy. This paper describes the implementation and overall dosimetric accuracy of PEREGRINE physics algorithms, beam model, and beam commissioning procedure. Particle-interaction data, tracking geometries, scoring, variance reduction, and statistical analysis are described. The BEAM code system is used to model the treatment-independent accelerator head, resulting in the identification of primary and scattered photon sources and an electron contaminant source. The magnitude of the electron source is increased to improve agreement with measurements in the buildup region in the largest fields. Published measurements provide an estimate of backscatter on monitor chamber response. Commissioning consists of selecting the electron beam energy, determining the scale factor that defines dose per monitor unit, and describing treatment-dependent beam modifiers. We compare calculations with measurements in a water phantom for open fields, wedges, blocks, and a multileaf collimator for 6 and 18 MV Varian Clinac 2100C photon beams. All calculations are reported as dose per monitor unit. Aside from backscatter estimates, no additional, field-specific normalization is included in comparisons with measurements. Maximum discrepancies were less than either 2% of the maximum dose or 1.2 mm in isodose position for all field sizes and beam modifiers.

No abstract

Four programs have been written to enable radiobiologists to build a computer data base of cellular dose-survival data, calculate cell survival with a correction for cell multiplicity at the time of irradiation, fit various survival models to the data by iteratively weighted least squares, and calculate the ratio of survival levels corresponding to specified doses or the ratio of doses that produce specified survival levels (e.g., oxygen enhancement ratio or relative biological effectiveness). The programs make plots of survival curves and data, and they calculate standard errors and confidence intervals of the fitted survival curve parameters and ratios. The programs calculate survival curves for the linear-quadratic, repair-saturation, single-hit multitarget, linear-multitarget, and repair-misrepair models of cell survival and have been designed to accommodate the addition of other survival models in the future. The programs can be used to compare the accuracy with which different models fit the data, determine if a difference in fit is statistically significant, and show how the estimated value of a survival curve parameter, such as the extrapolation number or the final slope, varies with the survival model. The repair of radiation-induced damage is analyzed in a novel way using these programs.

Computerized video time lapse (CVTL) microscopy was used to observe cellular events induced by ionizing radiation (10-12 Gy) in nonclonogenic cells of the wild-type HCT116 colorectal carcinoma cell line and its three isogenic derivative lines in which p21 (CDKN1A), 14-3-3sigma or both checkpoint genes (double-knockout) had been knocked out. Cells that fused after mitosis or failed to complete mitosis were classified together as cells that underwent mitotic catastrophe. Seventeen percent of the wild-type cells and 34-47% of the knockout cells underwent mitotic catastrophe to enter generation 1 with a 4N content of DNA, i.e., the same DNA content as irradiated cells arrested in G(2) at the end of generation 0. Radiation caused a transient division delay in generation 0 before the cells divided or underwent mitotic catastrophe. Compared with the division delay for wild-type cells that express CDKN1A and 14-3-3sigma, knocking out CDKN1A reduced the delay the most for cells irradiated in G(1) (from approximately 15 h to approximately 3- 5 h), while knocking out 14-3-3sigma reduced the delay the most for cells irradiated in late S and G(2) (from approximately 18 h to approximately 3-4 h). However, 27% of wild-type cells and 17% of 14-3-3sigma(-/-) cells were arrested at 96 h in generation 0 compared with less than 1% for CDKN1A(-/-) and double-knockout cells. Thus expression of CDKN1A is necessary for the prolonged delay or arrest in generation 0. Furthermore, CDKN1A plays a crucial role in generation 1, greatly inhibiting progression into subsequent generations of both diploid cells and polyploid cells produced by mitotic catastrophe. Thus, in CDKN1A-deficient cell lines, a series of mitotic catastrophe events occurred to produce highly polyploid progeny during generations 3 and 4. Most importantly, the polyploid progeny produced by mitotic catastrophe events did not die sooner than the progeny of dividing cells. Death was identified as loss of cell movement, i.e. metabolic activity. Thus mitotic catastrophe itself is not a direct mode of death. Instead, apoptosis during interphase of both uninucleated and polyploid cells was the primary mode of death observed in the four cell types. Knocking out either CDKN1A or 14-3-3sigma increased the amount of cell death at 96 h, from 52% to approximately 70%, with an even greater increase to 90% when both genes were knocked out. Thus, in addition to effects of CDKN1A and 14-3-3sigma expression on transient cell cycle delay, CDKN1A has both an anti-proliferative and anti-apoptosis function, while 14-3-3sigma has only an anti-apoptosis function. Finally, the large alterations in the amounts of cell death did not correlate overall with the small alterations in clonogenic survival (dose-modifying ratios of 1.05-1.13); however, knocking out CDKN1A resulted in a decrease in arrested cells and an increase in survival, while knocking out 14-3-3sigma resulted in an increase in apoptosis and a decrease in survival.

Different methods were used for evaluating data for DNA double-strand breaks (DSBs), as obtained by pulsed-field gel electrophoresis (PFGE) after X irradiation of Chinese hamster ovary cells. A total of 60 data points in the dose range of 0 to 116 Gy, along with repair data for 30 and 60 Gy, were analyzed by four methods: (1) percentage of DNA released from the plug, (2) specific size markers (percentage of DNA less than specific sizes, (3) fragment size distributions and (4) shape of the molecular weight (M) distributions. With the last method, both the slope and the intercept of the logarithm of the amount of radioactive DNA/delta M/M plotted as a function of M were used for calculating DSBs/100 Mbp. The slope and the intercept analyses differ in that the former is relatively independent of DNA trapped in the agarose plugs, i.e. cannot be released by doses of 100-150 Gy, whereas the intercept is dependent on the percentage of DNA trapped. Also, calculations of DSBs/100 Mbp for methods 1, 2 and 3 depend on the amount of DNA trapped in the plug. However, the slope method is unreliable for doses below about 20 Gy, and the scatter of data points is much greater than that obtained by the intercept method and by methods 1, 2 and 3. Therefore, the fragment size distribution and the specific size marker methods give the most consistent results, with 0.49 +/- 0.03 (95% CI) (DSBs/100 Mbp)/Gy. With the specific size marker method, however, care must be taken in selection of size markers in relation to the levels of DSBs of interest. Assuming randomly distributed DSBs, all four methods gave essentially the same results; i.e., the dose response was linear with a calculated level of 0.5-0.6 (DSBs/100 Mbp)/Gy, which is the same as 0.47-0.62 determined previously by calibrating with 125IdU.