The in vitro and in vivo antitumor effects of pepleomycin alone or in combination with radiation

Pepleomycin is a derivative of bleomycin, which is less toxic to the host but possesses greater antitumor activity. Experiments are described here in which Chinese hamster V-79 cells in culture and a murine squamous cell carcinoma in vivo have been used to obtain survival curves for pepleomycin alone or in combination with radiation. In both systems the survival curves for pepleomycin alone are biphasic, and this drug proved to be twice as effective as bleomycin. Further, in contrast to bleomycin, which showed simple additivity with radiation, pepleomycin potentiated radiation injury to the tumor cells both in vivo and in vitro. Therefore, pepleomycin may be superior to bleomycin not only in drug therapy but also in combined modality therapy.     

Estimating microbial inactivation parameters from survival curves obtained under varying conditions--the linear case

When the isothermal semi-logarithmic survival curves of heat inactivated microbial cells or spores are known to be linear it is possible to calculate their survival parameters from curves obtained under nonisothermal conditions, provided that the temperature history (’profile’) satisfies certain simple mathematical requirements. These requirements have been identified. The concept was tested by retrieving the survival parameters of a Listeria-like organism from generated survival curves for linear and nonlinear heating profiles on which noise had been superimposed. The availability of such a procedure eliminates the need to determine the survival parameters under perfect isothermal conditions, which are difficult to create for technical reasons. It will also enable determination of the survival parameters in the actual medium of interest, which may contain particles or may be too viscous to be treated in a capillary or narrow tube as is currently done. The method can also be used to assess survival parameters in nonthermal inactivation. A treatment with a dissipating chemical agent or anti-microbial is an example. In principle, the concept can be extended to the more general situation where the isothermal or iso-concentration semi-logarithmic survival curves are clearly nonlinear, but this will require a modification of the model and a different numerical calculation procedure.     

Generating microbial survival curves during thermal processing in real time

AIMS: To develop a method to calculate and record theoretical microbial survival curves during thermal processing of foods and pharmaceutical products simultaneously with the changing temperature. Moreover, to demonstrate that the method can be used to calculate nonisothermal survival curves, with widely available software such as Microsoft Excel. METHODS AND RESULTS: It has been assumed that the targeted organism's isothermal survival curves are not log linear and hence, the inactivation rate in nonisothermal processes is a function of the momentary temperature and the corresponding survival ratio. This could be expressed by a difference equation, which is an approximation to the continuous rate model. The concept was tested with the isothermal survival parameters of Clostridium botulinum and Bacillus sporothermodurans spores, and Salmonella enteritidis cells, using different kinds of survival models and under temperature profiles resembling those of commercial processes. As expected, there was an excellent agreement between the curves produced by solving the differential equation of the continuous model and by the incremental method, which has been posted on the web as freeware. CONCLUSIONS: It is possible to calculate nonisothermal survival curves, in real time, with an algorithm that can be written in the language of general purpose software, to follow the inactivation of one or more targeted organisms simultaneously and to simulate microbial survival patterns under existing or planned industrial thermal processes. SIGNIFICANCE AND IMPACT OF THE STUDY: Replacement of the traditional 'F0-value', which requires the log linearity of the organism's isothermal survival curves, by the more realistic theoretical survival ratio estimate as a measure of the thermal process efficacy.     

Bayesian monitoring of clinical trials with failure-time endpoints

This article presents an aid for monitoring clinical trials with failure-time endpoints based on the Bayesian nonparametric analyses of the data. The posterior distribution is a mixture of Dirichlet processes in the presence of censoring if one assumes a Dirichlet process prior for the survival distribution. Using Gibbs sampling, one can generate random samples from the posterior distribution. With samples from the posterior distributions of treatment-specific survival curves, one can evaluate the current evidence in favor of stopping or continuing the trial based on summary statistics of these survival curves. Because the method is nonparametric, it can easily be used, for example, in situations where hazards cross or are suspected to cross and where relevant clinical decisions might be based on estimating when the integral between the curves might be expected to become positive and in favor of the new but toxic therapy. An example based on an actual trial illustrates the method.     

Radiosensitivity of normal human epidermal cells in culture

Using an in vitro culture system we have derived radiation survival curves for the clonogenic cells of normal human epidermis. The culture system used allows the epidermal cells to stratify and form a multi-layered sheet of keratinizing cells. The cultures appear to be a very good model for epidermis in vivo. The survival curves show a population which is apparently more sensitive than murine epidermis in vivo. It remains unclear whether this is an intrinsic difference between the species or is a consequence of the in vitro cultivation of the human cells.     

Monitoring response, prediction methodology, staging, and imaging in prostate cancer

The predictive values of various tests and examinations are assessed as they relate to prostate cancer progression and treatment. The usefulness of post-treatment biopsy specimens is greatest 2 years after radiation therapy completion. Gleason grading is not reliable in the setting of hormonal ablation therapy. For patients with extracapsular extension, the survival curves separate depending on whether positive or negative surgical margins are obtained. Prostate-specific antigen doubling time is increasingly used as an indicator of disease recurrence after local therapy and prostate cancer-specific survival.     

The level of induced DNA double-strand breaks does not correlate with cell killing in X-irradiated mitotic and G1-phase CHO cells

The neutral (pH 9.6) filter elution technique was used to evaluate DNA damage induced in CHO cells irradiated at mitosis or in G1-phase under various incubation and postirradiation treatment conditions. Mitotic and G1/S border cells were more sensitive to radiation than G1 cells with respect to cell killing, but showed similar (G1/S) or lower (M) DNA elution dose--response curves. Similar cell survival and DNA/elution dose--response curves were obtained with plateau-phase cultures containing mainly G1-cells, as well as with G1 cells obtained after division of mitotic cells in either fresh or conditioned medium. However, survival of plateau-phase cells could be modified substantially by delayed-plating or postirradiation treatment with araA. These results, together with previously published observations, indicate that induction of DNA dsb cannot be invoked as an explanation for the variations in radiosensitivity observed through the cycle, or as an explanation for the formation of the survival curve shoulder. It is proposed that repair and fixation of radiation-induced DNA damage, expressed at the cell survival level as repair and fixation of alpha-PLD, are responsible for these effects.     

Inference for the proportional hazards model with misclassified discrete-valued covariates

We consider the Cox proportional hazards model with discrete-valued covariates subject to misclassification. We present a simple estimator of the regression parameter vector for this model. The estimator is based on a weighted least squares analysis of weighted-averaged transformed Kaplan-Meier curves for the different possible configurations of the observed covariate vector. Optimal weighting of the transformed Kaplan-Meier curves is described. The method is designed for the case in which the misclassification rates are known or are estimated from an external validation study. A hybrid estimator for situations with an internal validation study is also described. When there is no misclassification, the regression coefficient vector is small in magnitude, and the censoring distribution does not depend on the covariates, our estimator has the same asymptotic covariance matrix as the Cox partial likelihood estimator. We present results of a finite-sample simulation study under Weibull survival in the setting of a single binary covariate with known misclassification rates. In this simulation study, our estimator performed as well as or, in a few cases, better than the full Weibull maximum likelihood estimator. We illustrate the method on data from a study of the relationship between trans-unsaturated dietary fat consumption and cardiovascular disease incidence.     

Human mammary cell survival following ionizing radiation

We have developed and evaluated methods of culturing defined stromal and epithelial populations of normal human breast cells. These cell populations were used to generate radiation dose/survival curves. The epithelial cell population required specific hormones, growth factors, and conditioned media, as well as fibroblast feeder layers for clonal growth. Stromal cells grew well in a less complex medium. The stromal and parenchymal cell populations of the normal human breast were characterized by light and electron microscopy, immunohistochemical human fibronectin staining, gamma glutamyltranspeptidase histochemical staining, and cell sizing. Survival curves were generated using cells from four donors. The average D0 for epithelial cells was 122 cGy, with an average n value of 2.4. The average D0 and n values for stromal cells were 114 cGy and 2.0. The survival of human breast epithelial cells is compared to that of the cells of the rat mammary gland. The D0 values of both species are essentially the same, while the n value for human epithelial cells is lower. This difference in the n value may be a species specific response to radiation, or may merely reflect a difference in the two assay systems used to generate the survival curves.     

Theory of survival of bacteria exposed to ionizing radiation. I. X and gamma rays

A new model for the survival of bacteria exposed to ionizing radiation is constructed in the framework of a target theory based on microdosimetric concepts, where single- and double-strand breaks of DNA and their repair in vivo can be described consistently in terms of the microdosimetric quantity j (number of effective primary events per track per target). In this model, the ability of cells to repair DNA damage is taken into consideration in terms of the repair capacities for single- and double-strand breaks of DNA, xi 1 and xi 2 (0 less than or equal to xi 1, xi 2 less than or equal to 1). To apply this model to Escherichia coli K-12 strains with different repair abilities, values of the repair capacity for single-strand breaks, xi 1, were derived from experimental survival curves. The theoretical survival curves for 60Co gamma rays were found to be effectively insensitive to the value of xi 2. Experimental survival curves for the wild-type, uvr, and rec strains of E. coli K-12 were well reproduced in this model. From these results, it is concluded that the theoretical formulation for the survival fraction of bacteria can afford a quantitative method for analysis of the repair process for radiation-induced single-strand breaks in DNA in vivo.     

Exponential or shouldered survival curves result from repair of DNA double-strand breaks depending on postirradiation conditions

The yeast mutant rad54-3 is temperature conditional for the rejoining of DNA double-strand breaks, but cells do proliferate at both the restrictive and permissive temperatures. Thus, after irradiation with 30 MeV electrons, survival curves can be obtained which may or may not involve double-strand break rejoining under certain experimental conditions. Because of this special property of rad54-3 cells, it was possible to demonstrate that rejoining of radiation-induced double-strand breaks under nongrowth conditions yields exponential survival curves the slopes of which decrease as a function of the rejoining time. These survival data suggest that, under nongrowth conditions, the rejoining of double-strand breaks is an unsaturated process and lacks binary misrepair. In contrast, whenever rejoining of double-strand breaks occurs under growth conditions, shouldered survival curves are observed. This is true for immediate plating as well as for delayed plating survival curves. It is proposed that it is the unsaturated rejoining of double-strand breaks under nongrowth conditions, lacking binary misrepair, which is responsible for potentially lethal damage repair.     

Variance Calculations for Direct Adjusted Survival Curves,with Applications to Testing for No Treatment Effect

Several investigators have recently constructed survival curves adjusted for imbalances in prognostic factors by a method which we call direct adjustment. We present methods for calculating variances of these direct adjusted survival curves and their differences. Estimates of the adjusted curves, their variances, and the variances of their differences are compared for non-parametric (Kaplan-Meier), semi-parametric (Cox) and parametric (Weibull) models applied to censored exponential data. Semi-parametric proportional hazards models were nearly fully efficient for estimating differences in adjusted curves, but parametric estimates of individual adjusted curves may be substantially more precise. Standardized differences between direct adjusted survival curves may be used to test the null hypothesis of no treatment effect. This procedure may prove especially useful when the proportional hazards assumption is questionable.     

Is the mean inactivation dose a good measure of cell radiosensitivity?

The relationship between the mean inactivation dose, D, and cell radiosensitivity, particularly in the low-dose region, was investigated for linear-quadratic (LQ) and two-component (TC) cell survival curves. Although D is approximately equal to D40 for most survival curves, the results showed that there is no consistent relationship, in theory, between D and the extent of cell killing in the low-dose region. Curves with equal D values could potentially represent cell populations with markedly different levels of survival after 200 rad, and conversely, curves with equal surviving fractions at 200 rad could have D values that differ by a factor of two or more. The variability in replicate estimates of D was also investigated and compared with that of SF200, the surviving fraction at 200 rad. The results of a simulation experiment suggest that estimates of D are somewhat more variable than those of SF200 and are more dependent on the choice of cell survival model and the range of available data used for the estimation. It is concluded that SF200 is a more reliable and numerically stable parameter of cell radiosensitivity than D.     

Graphical Methods for Analysing Combination Effects in Dose-Response Experiments

An organbath experiment with bovine tracheal muscle strips with cumulative increases in concentrations of a substance A in the absence and presence of a fixed concentration of a second substance B is considered as an example for demonstrating graphical methods to analyse drug combination effects. The response of each strip is individually described and estimated by a nonlinear dose response curve. From the curves of the combined action theoretical curves of substance A are derived, which were expected if the combination effect was simple similar or independent, respectively. The first graphical method consists in comparing the derived curves for substance A with the curves for substance A directly fitted. It is cheeked by eye if the group of derived curves can clearly be distinguished from the group of directly fitted curves. The second graphical method differs from the first method in so far, as not the curves are visualized but the parameter vectors corresponding to them. In contrast to widely used analytical methods the proposed graphical methods allow to treat individual instead of averaged dose response relationships. The methods can help to decide if the combination effect may be considered as independent, simple similar or none of both.     

Intestinal crypt clonogens: a new interpretation of radiation survival curve shape and clonogenic cell number

Estimates of the clonogen content (number of microcolony-forming cells) of murine intestinal crypts using microcolony assays show an apparent dependence on the radiation dose used in the assay of clonogen content. Crypt radiation survival curves often show increased curvature beyond that expected on the basis of the conventional linear-quadratic model. A novel form of crypt survival curve shape is proposed based on two contributory mechanisms of crypt killing. Six previously published sets of microcolony data were re-analysed using a dual-kill model, where target cells are killed by two contributory mechanisms, each described by a linear-quadratic function of dose. The data were analysed as two series--high-dose rate and low-dose rate irradiation. The data were fitted to the models using direct maximization of a quasi-likelihood, explicitly allowing for overdispersion. The dual-kill model can reproduce both the apparent dose-dependence of the clonogen estimates and the high-dose curvature of the dose-response curves. For both series of data the model was a significantly better fit to the data than the standard linear-quadratic model, with no evidence of any systematic lack of fit. The parameters of the clonogenic cell component of the model are consistent with other studies that suggest a low clonogen number (somewhat less than five) per crypt. The model implies that there is a secondary mechanism decreasing clonogen survival, and hence increasing clonogen number estimates, at high doses. The mechanisms underlying the modification of the dose-response are unclear, and the implied mechanisms of, for example, slow growth, induced either directly in the surviving cells or indirectly through stromal injury or bystander effects are only speculative. Nevertheless, the model fits the data well, demonstrating that there is greater kill at high doses in these experimental series than would be expected from the conventional linear-quadratic model. This alternative model, or another model with similar behaviour, needs to be considered when analysing in detail and interpreting microcolony data as a function of dose. The implied low number of < or = 5 of these regenerative and relatively radioresistant clonogenic cells is distinct from a similar number of much more radiosensitive precursor stem cells which undergo early apoptosis after doses around 1 Gy.     

The effectiveness of short-term versus long-term exposures to Photofrin II in killing light-activated tumor cells.

Asynchronous populations of mouse EMT-6 tumor cells were exposed to various doses of 630-nm light in slowly stirred aerobic suspensions after both short-term and long-term exposures to Photofrin II. All survival curves are characterized by a "threshold" light dose below which no cell inactivation occurs followed by a steep light-dose response. Both the shoulder widths and the inactivation curve slopes are functions of Photofrin II concentration. After high doses of light where survival levels are 0.003 and lower, "resistant tails" are observed on some survival curves. Light doses required to inactivate 50% of tumor cell populations were obtained from whole survival curves and their reciprocals (1/D50% survival) used as inactivation "rates". The amount of Photofrin II within cells was measured by a fluorescence assay. Per unit of fluorescence, this photosensitizer is at least 10 times more effective after long-term than after short-term exposures. After long-term exposures, both fluorescence activity and photosensitizing effectiveness are retained in washed cells for several hours. After short-term exposures, a majority of both the fluorescence and photosensitizing activity is lost by multiple washings or stirring in tissue culture medium without drug. These data suggest that the cellular compartments associated with photosensitization after short-term exposures to Photofrin II are probably different from the cellular compartments associated with photosensitization after long-term exposures to the drug. The data are consistent with known properties of the monomeric and oligomeric components of Photofrin II.     

Predictive accuracy and explained variation in Cox regression

We suggest a new measure of the proportion of the variation of possibly censored survival times explained by a given proportional hazards model. The proposed measure, termed V, shares several favorable properties with an earlier V1 but also improves the handling of censoring. The statistic contrasts distance measures between individual 1/0 survival processes and fitted survival curves with and without covariate information. These distance measures, Dx and D, respectively, are themselves informative as summaries of absolute rather than relative predictive accuracy. We recommend graphical comparisons of survival curves for prognostic index groups to improve the understanding of obtained values for V, Dx, and D. Their use and interpretation is exemplified for a Yorkshire lung cancer study on survival. From this and an overview for several well-known clinical data sets, we show that the likely amount of relative or absolute predictive accuracy is often low even if there are highly significant and relatively strong prognostic factors.     

Finding Risk Groups by Optimizing Artificial Neural Networks on the Area under the Survival Curve Using Genetic Algorithms

We investigate a new method to place patients into risk groups in censored survival data. Properties such as median survival time, and end survival rate, are implicitly improved by optimizing the area under the survival curve. Artificial neural networks (ANN) are trained to either maximize or minimize this area using a genetic algorithm, and combined into an ensemble to predict one of low, intermediate, or high risk groups. Estimated patient risk can influence treatment choices, and is important for study stratification. A common approach is to sort the patients according to a prognostic index and then group them along the quartile limits. The Cox proportional hazards model (Cox) is one example of this approach. Another method of doing risk grouping is recursive partitioning (Rpart), which constructs a decision tree where each branch point maximizes the statistical separation between the groups. ANN, Cox, and Rpart are compared on five publicly available data sets with varying properties. Cross-validation, as well as separate test sets, are used to validate the models. Results on the test sets show comparable performance, except for the smallest data set where Rpart’s predicted risk groups turn out to be inverted, an example of crossing survival curves. Cross-validation shows that all three models exhibit crossing of some survival curves on this small data set but that the ANN model manages the best separation of groups in terms of median survival time before such crossings. The conclusion is that optimizing the area under the survival curve is a viable approach to identify risk groups. Training ANNs to optimize this area combines two key strengths from both prognostic indices and Rpart. First, a desired minimum group size can be specified, as for a prognostic index. Second, the ability to utilize non-linear effects among the covariates, which Rpart is also able to do.     

Absence of contact effects in irradiated EMT6-Rw tumors

Some cells have been reported to show greater resistance to drugs or radiation when growing with close intercellular contacts in spheroids or in solid tumors than when growing with few intercellular contacts in sparse cultures. In some cases this increased resistance reflects an increased capacity of cells in close contact to repair cytotoxic damage. However, not all tumors show contact effects, and in some tumors and spheroids the increased resistance appears to be produced by environmental factors, such as hypoxia, rather than by changes in the repair capacity of the cells. To assess whether EMT6-Rw cells showed increased intrinsic radioresistance when grown as solid tumors, we compared survival curves for cells in exponentially growing monolayers and in solid tumors in BALB/c mice. To avoid complications arising from regional heterogeneity in oxygenation within solid tumors, these irradiations were performed under conditions of uniform, maximal hypoxia. The two survival curves were indistinguishable. Moreover, survival curves for cells suspended from solid tumors, plated at low densities and irradiated immediately, after 5 h of incubation or after 24 h of incubation, were indistinguishable from one another and were indistinguishable from survival curves for cells suspended from exponentially growing monolayers and irradiated immediately using an identical protocol. It therefore appears that contact effects are insignificant for irradiated EMT6-Rw tumors and that the intrinsic radiosensitivity of these cells is similar in culture and in solid tumors.