The Intrinsic α/β Ratio for Human Tumour Cells: Is It a Constant?

The radiation response of 15 mammalian cell lines comprising 11 human tumour, two human fibroblast and two murine lymphoma cell lines, has been analysed using the linear-quadratic equation. As well as using conventional analysis of acute dose-survival curves to derive values for alpha and beta (termed alpha ac and beta ac), low dose-rate and split-dose experiments have been used to derive independent values of alpha and beta (alpha 1dr and beta RR), respectively. alpha 1dr provides a measure of irrecoverable damage, the magnitude of which agreed well with the initial slope of the acute survival curve for most cell lines. beta RR derived from split-dose experiments represents a unique measure of recovery for each cell line. Large differences were found between individual values of beta ac and beta RR, especially in the radiosensitive cell lines. Since beta RR is a functional measure of recovery we suggest that this is the more relevant parameter in studies of dose sparing. The most striking result of this analysis was found in considering the alpha/beta ratios. No relationship was observed between alpha ac and beta ac resulting in values of alpha ac/beta ac ranging from 1 to 175. In contrast a positive correlation was observed between alpha 1dr beta RR in the 11 tumour cell lines, giving an alpha/beta ratio of 9.4 +/- 1.8 Gy. This observation of the relative constancy of the ratio for human tumour cells leads to an hypothesis about the role of initial damage as a determinant of radiosensitivity.

Cellular Recovery in Two Sub-lines of the L5178Y Murine Leukaemic Lymphoblast Cell Line Differing in Their Sensitivity to Ionizing Radiation

Cellular recovery was assessed in two sublines of L5178Y murine lymphoma cells of differing radiosensitivity (LY-S and LY-A4) using low dose-rate irradiation and split-dose experiments. No increase in cell survival was observed in the LY-S cell line until the dose-rate was reduced to 2 cGy/min, whereas in the LY-A4 cell line 20 cGy/min was low enough to detect changes in survival. The extent of this change, as assessed by dose reduction factors at 2 logs of cell kill, was greater in the LY-A4 cell line. Fitting these data with the incomplete repair model of Thames led to anomalous values for the half-time of repair. In split-dose experiments the maximum observed recovery ratio increased as a function of dose in a manner that is consistent with the linear-quadratic equation. As was found previously with radiosensitive human tumour cells, the LY-S cell line showed more split-dose recovery at any given dose than the LY-A4 cell line.

The case against apoptosis

Apoptosis is not as dominant a process in cell loss from normal tissues and tumours as has sometimes been claimed. The term 'programmed cell death', which many authors regard as synonymous with apoptosis, is an unsatisfactory term that is best avoided. In studies on the response of tumours to drug and radiation treatment, the use of apoptosis assays concentrates attention on the first decade of cell killing (0-90%), whereas the outcome of treatment depends on multi-log cell kill; an assay of clonogenic cell survival is the appropriate method for this purpose. Loss of colony-forming ability is the key event in treated tumour cells, and the appearance of morphological and molecular evidence of apoptosis is probably downstream from this event. Published studies that have compared apoptosis and cell survival responses in tumour cells have generally failed to find a causal relationship.

Differences between a human bladder carcinoma cell line and its radiosensitive clone in the formation of radiation-induced DNA double-strand breaks in different chromatin substrates

It is well established that DNA-associated proteins, as well as soluble free-radical scavengers, can significantly influence the amount of damage inflicted in DNA by ionising radiation. It is not known, however, to what degree there is variation between cell lines in the effectiveness of these cellular components to protect DNA. In this study we have examined the level of strand break induction in a human bladder carcinoma cell line, MGH-U1, and its radiosensitive mutant, U1-S40b, when soluble scavengers and DNA-associated proteins were progressively removed. DNA double-strand breaks were measured using pulsed-field gel electrophoresis when cells were irradiated after lysis in solutions containing various salt concentrations. The two cell lines showed only a small, non-significant difference in damage induced in intact cells but isolated nuclei and chromatin devoid of non-histone proteins showed significantly more damage in the U1-S40b cells. Once the histone H1 was removed again there was no difference between the cell lines in the damage induced. We conclude that the different components of the cellular defences against free radical attack can have different influences in different cells. It is not clear whether this has an influence on the cellular sensitivity to the killing effects of radiation but it does suggest that artificial manipulation of the different components of the system may not affect overall damage induction to the same degree in all cells.

Comparison between pulsed-field gel electrophoresis and the comet assay as predictive assays for radiosensitivity in fibroblasts

The radiosensitivity of skin fibroblasts derived from patients as measured in vitro by a clonogenic survival assay appears to correlate with the risk of developing severe late reactions to radiation. Unfortunately, these assays are clinically impractical as a predictive test for radiosensitivity. The purpose of this study was to assess the utility of two possible surrogate assays for radiosensitivity, pulsed-field gel electrophoresis (PFGE) and single-cell gel electrophoresis (comet assay), both of which can be used to measure DNA double-strand breaks. Twenty-three nontransformed human fibroblast cell lines exhibiting a range of radiosensitivities were studied with both of these assays. The results were correlated with measurements of radiosensitivity obtained as part of a larger study examining the correlation between cellular radiosensitivity and clinical response. [2-(14)C]Thymidine-labeled confluent cultures were irradiated at 1.0 Gy/min with doses of 0 to 150 Gy. After allowing 4 h for repair at 37 degrees C, cells were trypsinized and aliquots were used for preparing slides for the comet assay. After neutral lysis and electrophoresis, the slides were stained with ethidium bromide and 50 comet moments were measured for each dose. The remainder of the cells were formed into agarose plugs and, after neutral lysis, were subjected to PFGE. The fraction of activity released (FAR) from the well was measured by scintillation counting of appropriate segments of each gel lane. Cellular radiosensitivity was measured with a standard clonogenic assay at a low dose rate of 1.2 cGy/min, and the dose that resulted in a surviving fraction of 0.01 (D0.01) was calculated. The slope of the plot of comet moment as a function of dose for each cell line did not correlate with D0.01 (R = 0.36, P > 0.1). In contrast, the slope of the FAR as a function of dose had a weak inverse correlation with D0.01 (R = 0.43 and P = 0.05) such that the more radiosensitive cell lines exhibited a steeper dose response for FAR. Although the correlation between the slope of the dose response for FAR and D0.01 was weak, refinement of the PFGE technique may provide a potentially useful predictive assay for radiosensitivity.

Treatment results and prognostic factors in 101 men treated for squamous carcinoma of the penis

PURPOSE

This retrospective study was performed to assess the treatment outcome and prognostic factors in 101 men with invasive squamous carcinoma of the penis treated at the Royal Marsden Hospital between 1960-1990.

METHODS AND MATERIALS

The tumor was confined to the glans penis (T1) in 79 patients, 82 were node negative (N0), and two patients had distant metastases at presentation. The histology was Grade 1 (G1) in 36, Grade 2 (G2) in 18, Grade 3 (G3) in 28, and unknown in 19 patients. Node-positive disease was commoner in patients with G3 (p = 0.02) or T2/3/4 tumors (p = 0.007). Treatment for the primary tumor was external beam radiotherapy (EBRT) in 59, interstitial brachytherapy in 13, and partial or total penectomy in 29 patients. The median dose, dose/fraction, and treatment time for EBRT was 60 Gy, 2 Gy/fraction, and 46 days, respectively. Eighty patients received no inguinal node treatment, 13 had EBRT (4 with chemotherapy), and 8 underwent groin dissection at presentation.

RESULTS

During a median follow-up of 5.2 years (2 months-22 years), 56 patients died (penile cancer 31, intercurrent illness 23 and unknown cause 2), giving 10 year overall and cause-specific survival (CSS) of 39 and 57%, respectively. Adverse prognostic factors for CSS on univariate analysis were G3, ulcerative/fungating or T2/3/ 4 tumors, node positive, Jackson's Stage 2/3/4, and surgical treatment for the primary. All but the last two were significant independent prognostic factors for CSS on multivariate analysis. Penile or perineal recurrence or residual disease after initial treatment was seen in 36 out of 98 evaluable patients, giving a 10-year local failure rate (LFR) of 45%. Local failure after initial treatment was successfully salvaged in the majority (26 out of 36) of patients with further surgery or radiotherapy, and local control was achieved ultimately in 74 out of 77 T1, 7 out of 12 T2; 3 out of 3 T3, and 3 out of 5 T4 tumors. In the 44 evaluable patients with T1 tumors treated by EBRT the only adverse RT parameter approaching prognostic significance (p = 0.052) was a BED value corrected for recovery of <60 Gy (alpha/beta 10, K = 0.5 Gy/day, mean = 21 days).

CONCLUSION

Invasive squamous carcinomas of the penis carry a significant risk of loco-regional recurrence after initial radiotherapy and this can be successfully salvaged in most patients with further treatment. This mandates close follow-up to detect loco regional recurrence early.

Hprt- mutation spectrum in a closely related pair of human bladder tumour cell lines after gamma-irradiation at different dose-rates

The spectrum of deletion sizes in mutants of two human bladder carcinoma cell lines has been examined. The cell lines were MGH-U1 and a radiation-sensitive subline (U1-S40b) that has been developed in this laboratory. Three groups, each of 20-30 mutants at the hprt locus were investigated: arising spontaneously, or induced after exposure to 10 Gy gamma-radiation either at high dose-rate (2 Gy/min) or low dose-rate (0.01 Gy/min). Data on the mutation frequency of the two cell lines at low dose-rate were obtained to supplement previously published data at high dose-rate. The mutation frequency was lower in U1-S40b than in MGH-U1 both for high and low dose-rate irradiation. The presence of intact copies of each of the nine hprt exons was examined using multiplex PCR, supplemented by single-exon PCR. The incidence of small hprt mutations (i.e. leading to no change in the size of the PCR products) was the same for spontaneous mutations in the two cell lines; for radiation-induced mutants it was higher in U1-S40b. The incidence of total deletions (i.e. no positive exon amplification) was lower in U1-S40b both for high and low dose-rate irradiation. The results are consistent with the hypothesis that large deletions tend to lead to the loss of adjacent essential genes and thereby to the death of potential mutants.

Cell-cycle variation in DNA migration in pulsed-field gel electrophoresis

Pulsed-field electrophoresis is being used extensively in the gene mapping studies and in the analysis of DNA strand breakage by ionizing radiation. We have evaluated the relationship between the fraction of S phase DNA in a cell population and its ability to modify the migration of DNA in pulsed-field gel electrophoresis. We have shown that increasing the proportion of S phase DNA reduced the effective rate of migration of MGH-U1 cellular DNA. This effect was observed after treatment with ionizing radiation or the restriction enzyme Not I. However, when radiation-induced damage was studied using intact cells, only the DNA with 70 percent S phase showed apparent differences in damage induction. These studies therefore provide data to indicate the percentage of S phase cells at which overall DNA migration might be affected significantly.

From targets to genes: a brief history of radiosensitivity

The biological work of Douglas Lea spanned the period from 1934 to his early death in 1947, and during this short period he made important contributions to the theory of radiation action. He interpreted experimental data relating to the effects of radiation on viruses, bacteria, bean roots, etc in terms of the inactivation of discrete targets, which he identified with cellular genes. He thus laid the foundation of much subsequent research. It is now well recognized that mammalian cells differ substantially in radiosensitivity, especially in the low-dose region of the survival curve. The dependence of radiosensitivity on dose rate has been widely studied; this has practical significance for clinical radiotherapy as well as mechanistic implications. Since Lea's time there have been a number of efforts to describe models that can relate cell killing to radiation dose, dose rate, and track structure. So far these have not led to a comprehensive and widely accepted picture. Microdosimetric considerations lead to the concept of differing severity of lesions induced in DNA. Much is known about the sequence of processes that subsequently lead to cell inactivation: this can be divided into phases of induction, processing, and manifestation. Chromosomal events are currently attracting much attention, as they did in Lea's time. Considerable progress has also been made in identifying genes that control the repair of radiation damage. It has been found that mutation is frequently associated with the loss of a large segment of the genome around the damage site and this will have important implications for interactive processes between particle tracks.

Glutathione manipulation and the radiosensitivity of human tumour and fibroblast cell lines

We have studied the role of glutathione (GSH) in determining radiation response in five human tumour and one human fibroblast cell line. GSH concentration was measured using the Tietze assay and compared with clonogenic survival following gamma-irradiation. No relationship between GSH concentration and aerobic radiosensitivity was observed. The addition of 10 mM extracellular cysteamine produced protection factors in all cell lines, ranging from 1.6 to 2.1, but had little influence on cellular GSH concentration. Depletion of GSH by buthionine sulphoximine (0.1 mM for 18 h) had negligible effect on cell survival, though moderate radiosensitization resulted from extreme GSH depletion after 30-min treatment with 1 mM dimethylfumarate. The degree of aerobic sensitization did not correlate with GSH levels. Irradiation under hypoxia produced oxygen enhancement ratios varying from 1.6 to 2.6, with no relationship to GSH content.

The comet moment as a measure of DNA damage in the comet assay

The development of rapid assays of radiation-induced DNA damage requires the definition of reliable parameters for the evaluation of dose-response relationships to compare with cellular endpoints. We have used the single-cell gel electrophoresis (SCGE) or 'comet' assay to measure DNA damage in individual cells after irradiation. Both the alkaline and neutral protocols were used. In both cases, DNA was stained with ethidium bromide and viewed using a fluorescence microscope at 516-560 nm. Images of comets were stored as 512 x 512 pixel images using OPTIMAS, an image analysis software package. Using this software we tested various parameters for measuring DNA damage. We have developed a method of analysis that rigorously conforms to the mathematical definition of the moment of inertia of a plane figure. This parameter does not require the identification of separate head and tail regions, but rather calculates a moment of the whole comet image. We have termed this parameter 'comet moment'. This method is simple to calculate and can be performed using most image analysis software packages that support macro facilities. In experiments on CHO-K1 cells, tail length was found to increase linearly with dose, but plateaued at higher doses. Comet moment also increased linearly with dose, but over a larger dose range than tail length and had no tendency to plateau.

DNA strand break rejoining defect in xrs-6 is complemented by transfection with the human Ku80 gene

The radiosensitive mutant xrs-6, derived from Chinese hamster ovary cell line CHO-K1, has been demonstrated to be defective in DNA double-strand break repair and also in its proficiency to undergo V(D)J recombination. Recent work has provided both genetic and biochemical evidence that the M(r) 80,000 subunit of the Ku protein is able to complement the radiosensitivity and the V(D)J recombination defect in the xrs-6 mutant. We demonstrate here that complementation of the radiosensitive phenotype in xrs-6 cells by the introduction of Ku80 cDNA is accompanied by the concomitant restoration of DNA double-strand break rejoining proficiency to almost that of the parental CHO-K1 cells, as measured both by neutral single-cell microgel electrophoresis (Comet) technique and by pulsed-field gel electrophoresis. These results provide further biochemical evidence for the involvement of the Ku protein in the repair of DNA double-strand breaks.

Non-surgical management of early breast cancer in the United Kingdom: radiotherapy fractionation practices. Clinical Audit Sub-committee of the Faculty of Clinical Oncology, Royal College of Radiologists, and the Joint Council for Clinical Oncology

A national survey of British radiotherapy schedules used in women with early breast cancer was undertaken to document variation in treatment practices and to consider its clinical significance. Although the variation is considerable, the analysis suggests that the majority of schedules in use are very similar in terms of treatment intensity when allowance is made for fraction size and overall time. Half the respondents used one of three dosage schedules, which probably differ very little in terms of late normal-tissue effects and tumour control from a conventional schedule giving 50 Gy in daily 2 Gy fractions. Eighty-two percent of respondents were using schedules that are equivalent to a dose of between 45 Gy and 50 Gy in 2 Gy fractions. The study suggests that the protocols in use by a minority of respondents may be unduly conservative or aggressive, and it leads to the proposal that oncologists should set up trials comparing commonly used schedules as a matter of urgency.

A comparison of methods for calculating DNA double-strand break induction frequency in mammalian cells by pulsed-field gel electrophoresis

Pulsed-field electrophoresis (PFGE) has become one of the most widely used methods for the evaluation of radiation-induced DNA double-strand breaks (dsb). In most studies a simple quantification of DNA migration from the well in the gel has been used as the correlate with dsb formation. Here we have compared such a method, as calibrated with 125I-labelled UdR, with two methods which involved the analysis of the distribution of sizes of DNA fragments migrating in the gel. We conclude that the three methods produce similar absolute values for dsb induction frequency. It is not clear which is the single method of choice but the comparison of the analyses increases the information which can be derived from PFGE experiments.

Dose-rate effect for DNA damage induced by ionizing radiation in human tumor cells

The effect of dose rate on clonogenic cell survival and DNA double-strand breaks (DSBs) has been examined in a human bladder carcinoma cell line, RT112, treated with ionizing radiation. Cell survival changed markedly over the range of dose rates used (0.01-1.28 Gy/min) with the curves becoming shallower and straighter as the dose rate was lowered. Similarly, the number of DSBs measured by pulsed-field gel electrophoresis (PFGE) immediately after irradiation varied with dose rate. Fewer DSBs were detectable after low-dose-rate irradiation. However, when a 4-h repair period was allowed after irradiation, cells treated at all dose rates exhibited approximately the same amount of damage. The final level of unrejoined DSBs, as detected by PFGE, therefore does not correlate with cell survival at different dose rates.

Initial radiation-induced DNA damage in human tumour cell lines: a correlation with intrinsic cellular radiosensitivity

The role of the initial DNA double-strand breaks (dsb) as a determinant of cellular radiosensitivity was studied in human breast and bladder cancer cell lines. Cell survival was measured by monolayer colony-forming assay as appropriate and differences in radiosensitivity were seen (alpha-values ranged from 0.12 to 0.54). After pulsed-field gel electrophoresis (PFGE) the initial slopes of dose-response curves were biphasic with a flattening of the curves above 30 Gy. When the frequency of DNA dsb induction was assessed using a mathematical model based on the DNA fragment size distribution into the gel lane, we found a statistically significant relationship between the number of DNA dsb induced and the corresponding alpha-values and fraction surviving after 2Gy (P = 0.0049 and P = 0.0031 respectively). These results support the view that initial damage is a major determinant of cell radiosensitivity.

Planned and unplanned gaps in radiotherapy: the importance of gap position and gap duration

Local tumour control in 971 patients with squamous carcinoma of the supraglottic larynx has been examined in relation to the occurrence of gaps in radiation therapy. The minimum follow-up time was 3 years. The reasons for a gap in radiotherapy fell into four categories: independent of the patient (national holidays, machine break-down, etc.), planned gaps (split-course therapy), severe normal-tissue reactions, and intercurrent disease. Only 11.7% of patients had no gap at all, 75.5% a single gap, and 2.1% had more than four gaps. The probability of tumour control increased with dose in all patient sub-groups; the average percentage increase for a 1% increase in dose was 4.3. The data were subjected to multivariate analysis, leading to the following conclusions. Patients in whom there was a single gap showed a remarkable trend of local control: if the gap began before day 19 after the start of therapy, the local tumour control was considerably below that in patients who did not suffer a gap in treatment. The local tumour control in patients whose gap began at day 20-29 was indistinguishable from that in patients who had no gap in treatment. A gap further towards the end of treatment was again associated with a severe drop in local control. This trend was independent of the recorded cause of the gap. The mechanism of this phenomenon is not clear. The effect of the timing of a treatment gap appears in this data set to have had a considerable impact on outcome and our observations should stimulate further study of this phenomenon in other clinical settings.

Prospects for proton-beam radiotherapy

Proton beams are already being employed for radiation therapy in 15 centres worldwide and over a dozen more are planned. Good clinical results have been reported in uveal melanomas and in sarcomas of the skull base. Calculated dose distributions for the treatment of brain, lung, head and neck and pelvic tumours predict an improvement relative to multiple-field or conformal photon radiotherapy. Protons may well provide high-precision radiotherapy that will lead to improved treatment of certain tumours in specific anatomical locations.

Cell-cycle progression during continuous low dose rate irradiation of a human bladder carcinoma cell line

At very low radiation dose rates, the proliferation of mammalian cells continues unaffected but as the dose rate is increased there comes a point at which it is interrupted. The dose rate at which this happens is often thought to be a significant factor in the effects of brachytherapy: it may determine the range from an implanted source at which cell-cycle redistribution and repopulation effects will occur. By means of mitotic counts and DNA flow cytometry, we have examined the dose rate effect in a human bladder carcinoma cell line (MGH-U1). Irradiation at dose rate 0.1 cGy/min had little or no effect on cell-cycle progression. Suppression of mitosis and arrest of cells in G2 was observed at 0.4 cGy/min and above. Surprisingly, the duration of mitotic arrest showed little dose rate dependence; it was followed by an overshoot of cells in mitosis after 24-39 h of irradiation. An even more pronounced overshoot of cells in G2 occurred and persisted throughout the irradiation period. The cell kinetic data indicate that after the temporary block in cell-cycle progression, cell proliferation continued at all dose rates up to 1.4 cGy/min. We have evaluated these results in the light of previous studies in this department of the dose rate effect for cell survival in the MGH-U1 cell line. After 24 h irradiation at 1.4 cGy/min the surviving fraction was below 10(-2), also after 30 h at 1.0 cGy/min. When cell-cycle blockade is considerable, so is the level of cell killing. Flow-cytometric data therefore are dominated by the properties of cells that are doomed to die.(ABSTRACT TRUNCATED AT 250 WORDS)

Mouse hemithoracic irradiation and its interaction with cytotoxic drugs

Irradiation of the mouse right or left hemithorax at 14 or 18 Gy produced a dose-dependent rise in breathing rate 16 weeks after irradiation without significant mortality. The measurement of breathing rate changes following right hemithoracic irradiation combined with the maximally tolerated dose of cytotoxic drugs was assessed as a method for qualitatively detecting drug-irradiation interactions which either exacerbate pneumonitis or alter its time course. Cyclophosphamide at 100 mg/kg accentuated and accelerated the rise in breathing rate, culminating in early mortality. BCNU 30 mg/kg delayed the appearance of the radiation response. Busulphan 30 mg/kg appeared to be radioprotective, but this was shown to be due to the DMSO-containing vehicle. Doxorubicin 6 mg/kg had no effect when combined with right or left hemithoracic irradiation. Carboplatin 100 mg/kg, vindesine 4 mg/kg and vinblastine 4 mg/kg had no substantial effect upon the changes in breathing rate.

Differences in the level of DNA double-strand breaks in human tumour cell lines following low dose-rate irradiation

It is now well accepted that differences exist in the intrinsic radiosensitivity of human tumour cells although the molecular basis of this is still unclear. Current evidence suggests that of the lesions induced in DNA by ionising radiation, double-strand breaks (DSB) are the most closely linked to cell death. In this study, levels of DSB were measured by neutral filter elution under conditions of both repair inhibition and maximum recovery and compared with clonogenic survival curves for high (HDR) and low dose-rate (LDR) irradiation in human carcinoma lines of differing radiosensitivity. Four human lung carcinoma lines were used, two small-cell (SCLC; HC12 and HX149) and two non-small cell lines (NSCLC; HX147A7 and HX148G7). Cell survival was measured by soft agar and monolayer colony-forming assays as appropriate and a large variation in sensitivity of the cell lines was seen (alpha values of 0.06 to 0.56 Gy-1). We have previously reported that the damage induced at high dose rate does vary in these cell lines but not in a way which correlates with their cell survival response [5]. Following irradiation to 15 Gy at low dose rate essentially no DSBs were detected in any of the four lines but at 70 Gy the more sensitive SCLC showed more residual damage than in the more radioresistant NSCLC lines. The prime determinant of the difference between the LDR and HDR damage curves is likely to be repair occurring during irradiation.(ABSTRACT TRUNCATED AT 250 WORDS)

Heterogeneity of radiosensitivity in a human glioma cell line

Sixteen clones were isolated from an early-passage human glioma cell line (IN859) and have been found to show variation in several biological characteristics including DNA content, modal chromosome number, and morphology. In addition, heterogeneity of radiosensitivity was detected: the doses that gave a surviving fraction of 0.01 varied by a factor of approximately 1.5. The most sensitive (clone 6) and the most resistant (clone 9) clones were selected for further study; their surviving fractions at 2Gy (SF2) were 0.37 and 0.64, respectively. When compared at a fixed radiation dose the sensitive clone surprisingly demonstrated greater split-dose recovery than the resistant clone; it also showed greater low dose-rate sparing.

Evaluation of cell attachment matrix (CAM) coated plates for primary culture of human tumour biopsies

Baker et al. [Cancer Res. 46: 1263-1274, 1986] developed an adhesive tumour cell culture system (ATCCS) using a culture surface coated with a cell attachment matrix (CAM). The ability of CAM-coated plates to support the growth of cells from human tumour biopsies has been evaluated. Successful growth was obtained in 9/22 samples (41%), but fibroblasts, rather than tumour cells, grew in the majority. Comparison of CAM with other surfaces showed that CAM was no better for establishing tumour cell growth than the presence of feeder cells or an alternative attachment factor vitronectin.

Cell proliferation in human tumour xenografts: measurement using antibody labelling against bromodeoxyuridine and Ki-67

Cell proliferation was investigated in human tumour xenografts using bromodeoxyuridine (BrdUrd) labelling, evaluated either by flow cytometry or in tissue sections, and also using the proliferation marker Ki-67. BrdUrd labelling was found to increase when cryostat tumour sections were digested with an enzymic solution. This yielded a labelling index up to four times higher than that obtained using the flow cytometer. Ki-67 indices were found to be higher than those reported for human tumour biopsies, as may be expected due to the enhanced growth rate of the xenografts. Significant heterogeneity was observed in the results for cervix, breast and bladder tumours, and the results of the three methods were poorly correlated. However, three of the four tumour types showed that the tumour with the lowest Ki-67 index also had the longest potential doubling time. Since the measurement of Ki-67 index was found technically easier to perform, and also adequately reflects relative tumour cell proliferation, it is preferred over the other techniques.

The ESTRO Breur lecture. Cellular sensitivity to low dose-rate irradiation focuses the problem of tumour radioresistance

This paper emphasises the radiobiology of human tumour cells irradiated at the relatively low dose rate of 1 Gy/h (i.e.1-2 cGy/min), described here as the "Regaud dose rate". Continuous irradiation at this dose rate is approximately isoeffective with fractionated radiotherapy using 2 Gy/fractions. At the Regaud dose rate, cell survival curves are approximately exponential and they appear to extrapolate the initial slope of the high dose rate survival curve. Little recovery occurs after such treatments since it has largely taken place during irradiation. At the Regaud dose rate human tumour cell lines show a wide range of radiosensitivities, differing by a factor of around 7. This may well be the most clinically-relevant way of describing the radiosensitivity of tumour cells. Current models of radiation cell killing envisage a component of damage that increases linearly with dose. It is this component that dominates the slope of the Regaud survival curve. It may be produced by DNA damage due to clusters of ionisation events, or perhaps by damage to hypersensitive parts of the genome. The steepness of this component of damage may be modified by exogenous inhibitors of DNA damage repair.

Growth and radiosensitivity testing of human tumour cells using the adhesive tumour cell culture system

The radiosensitivity of human tumour cell lines and cells cultured from xenografts or biopsy specimens was measured using the adhesive tumour cell culture system (ATCCS). For cell lines the derived surviving fractions at 2 Gy were in good agreement with values obtained by clonogenic assay. However, the assay tended to overestimate survival at higher radiation doses, and thus to give a false impression of radioresistance. When cells taken from xenografts or tumour biopsies were cultured there was no evidence for selective growth of tumour cells: fibroblast-like cells commonly grew. Immunohistochemical staining against the intermediate filament, vimentin, supported the mesenchymal origin of the fibroblast-like cells. In cultures of artificial mixtures of tumour cells and fibroblasts, low proportions of fibroblasts were not excluded by the assay and consequently modified the radiation response. The majority of cultures grown from bladder carcinoma biopsy specimens appeared fibroblast-like, although in some cases clearly distinguishable colonies of tumour cells were also grown. In such tumour types the reliable measurement of radiosensitivity in cells taken from biopsies will require further development of techniques that allow the selective growth of tumour cells.

The radiation dose-rate effect in two human neuroblastoma cell lines

The current use of targeted radiotherapy in the treatment of neuroblastoma has generated a requirement for further information on the radiobiology of these cells. Here we report on studies of the dose-rate effect in two human neuroblastoma cell lines (HX138 and HX142) and the recovery that they demonstrate in split-dose experiments. The sensitivity of the two cell lines to high dose-rate irradiation was confirmed. Surviving fractions at 2 Gy were 0.083 for HX138 and 0.11 for HX142. There was little evidence of a dose-rate effect above 2 cGy min-1 but significant sparing was seen at lower dose rates. Substantial recovery was seen in split-dose experiments on both cell lines, to an extent that was consistent with the linear quadratic equation. The data were used to derive values for the beta parameter of the linear-quadratic equation; the values for the neuroblastomas were higher than for any of the other human tumour cell lines that we have investigated to date. Thus, despite their high sensitivity to ionising radiation HX138 and HX142 do exhibit substantial levels of cellular recovery, suggesting that they may have a significant capacity for repair of radiation-induced lesions.

The inhibition of cellular recovery in human tumour cells by inhibitors of topoisomerase

A human bladder carcinoma cell line was irradiated at high and low dose rates and exposed to camptothecin and VP16, inhibitors of topoisomerase I and topoisomerase II respectively. Although camptothecin substantially modified the cytotoxic effects of high dose rate irradiation, abolished low dose rate sparing and inhibited the repair of sublethal and potentially lethal damage, VP16 had no effect on the survival curves even at highly cytotoxic doses. Thus, it is argued that there is a role for topoisomerase I but not topoisomerase II in the repair of DNA damage induced by ionising radiation.

Radiosensitivity, recovery and dose-rate effect in three human glioma cell lines

The results of radiotherapy in the treatment of high-grade gliomas are disappointing. In this study three recently established cell lines from high-grade human gliomas have been found to exhibit a sensitivity that is at the resistant end of the spectrum of radiosensitivities seen in human tumour cells generally. The results support the view that inherent cellular radioresistance may be an important cause of failure in this disease. All three cell lines showed an increase in survival when the radiation dose rate was reduced. In split-dose experiments, recovery was found to increase with dose in a manner consistent with the predictions of the linear-quadratic equation.

Override of the radiation-induced mitotic block in human tumour cells by methylxanthines and its relationship to the potentiation of cytotoxicity

The four methylated xanthine derivatives, caffeine, theophylline, theobromine and paraxanthine, were tested for their ability to override the mitotic block induced by ionizing radiation in the human bladder carcinoma cell line RT112. All four agents were found to partially override the block, ranking in terms of potency at a concentration of 1 mM in the order caffeine greater than theophylline greater than theobromine = paraxanthine. However, the effects of the four agents on the clonal survival of irradiated cells failed to correlate with the extent of override, both in terms of the relative effects of the four agents and the dose-response relationships; at a concentration of 1 mM only caffeine was found to potentiate cell killing as well as causing block override, whilst at higher concentrations all the agents had a significant effect on survival but little or no further influence on the degree of block override. It is therefore concluded that override of a mitotic block is not in itself sufficient to cause the increased killing seen when irradiated cells are incubated in the presence of caffeine, and that caffeine exerts its potentiating effect either by directly inhibiting repair of damage in DNA or by causing override of the radiation-induced inhibition of DNA synthesis.

The radiobiology of human cells and tissues. In vitro radiosensitivity. The picture has changed in the 1980s

Substantial developments have been made during the 1980s in the radiobiology of human tumours, in particular in studies of the radiosensitivity of human tumour cells. It is now clear that tumour cells differ considerably in radiosensitivity, to an extent that by itself is capable of explaining the clinical response of tumours to radiotherapy. There also is evidence that the radiosensitivity of human tumour cell lines to low radiation doses correlates with clinical experience. Irradiation at low dose rate amplifies the differences between cell lines. In conjunction with mathematical modelling, a study of the dose-rate effect also allows a distinction to be drawn between repairable and non-repairable damage. The differences seen between cell lines at low acute doses or low dose rates are associated with the non-repairable component. The most radiosensitive cell lines have a steep component of non-repairable damage and they give the impression of being recovery-deficient; this may, however, be incorrect for when evaluated at constant dose levels recovery is found to increase with increasing radiosensitivity. This leads to the view that recovery from radiation damage may reflect the amount of recoverable damage inflicted rather than the 'capacity' of the cells to recover.

A comparison of the in vivo and in vitro radiation response of three human cervix carcinomas

The radiation response of three carcinoma of the cervix tumours has been compared in vivo in xenografts using growth delay and in vitro by means of a monolayer-based clonogenic assay. Tumours have been irradiated with 60Co gamma-rays at both high (70-100 cGy/min) and continuous low dose rates (3-5 cGy/min) in order to determine the relative in vivo and in vitro sparing effects associated with lowering radiation dose rate. In vitro, two of the lines (HX155c and HX156c) showed significant low dose rate sparing when compared to the HX160c line (which showed little sparing) (p = 0.012). Despite greater scatter for the in vivo data, there was a general tendency for the in vivo results to follow that predicted from the in vitro experiments. In vivo, HX156 exhibited significant sparing (p = 0.011) whereas HX160 again exhibited no significant sparing (p = 0.15). Although the HX155 line did show some sparing in vivo this did not reach significance (p = 0.111). All three lines showed less actual specific growth delay in vivo than that predicted from in vitro data. These in vitro/in vivo findings lead to the conclusion that rapid predictive testing of radiosensitivity (ideally utilising low dose rate irradiation) would be beneficial in determining the choice of radiotherapy regimens for the treatment of this disease.

Why are some human tumours more radiosensitive than others?

There is now good evidence that the radiosensitivity of human tumour cells varies form one tumour type to another, and that the steepness of the initial part of the cell survival curve correlates with clinical radioresponsiveness. Studies at low dose rate allow differences between tumour cells to be seen more clearly. Current mathematical models of radiation cell killing include two components: a linear (i.e. exponential "alpha-component") and a bending component ("beta-component"). Repair of radiation damage affects only the beta-component. Among the 17 human tumour cell lines that we have studied, the average surviving fraction at 2 Gy due to the alpha-component is 0.44 and that due to the beta-component is 0.88. The beta-effect at 2 Gy appears to be similar in radiosensitive and radioresistant tumours; thus among radiosensitive tumours where the survival due to the alpha-component is below 0.3 the beta-effect makes a very small contribution to overall radiosensitivity in the low-dose region. Steep cell survival curves may appear straight but still be consistent with a modest beta-value. The overall effect of many small fractions will be to amplify the dominance of the alpha-component. The beta-effect is then unimportant because repair will be almost complete. Repair inhibitors may however change this situation, reducing recovery, increasing the beta-effect, and thereby increasing sensitivity. But in the absence of such inhibitors the differences between radiosensitive and radioresistant tumours must be looked for not in repair capacity but in the nature of the alpha-component. The most radiocurable tumours do not have low beta-values; they have higher alpha-values than the less curable tumours.

Recovery kinetics deduced from continuous low dose-rate experiments

Information on the rate of recovery from potentially lethal damage can be obtained from the analysis of the radiation dose-rate effect. Two mathematical models are useful in this respect: the Lethal-Potentially. Lethal (LPL) model of Curtis and the Incomplete Repair (IR) model of Thames. Full cell-survival curves are required at three or more dose rates in order to make a reliable analysis. Studies on human tumour cell lines yield values for recovery half-time in the range 0.1 to 0.9 h, and for each cell line the half-time found in this way is shorter than that obtained from a split-dose experiment. This result may imply that cellular recovery is a multi-component or a saturable process.

High intrinsic radiosensitivity of a newly established and characterised human embryonal rhabdomyosarcoma cell line

A new human rhabdomyosarcoma cell line (HX170c) has been established from a paratesticular embryonal tumour in a 5-year-old male. The cells grew as an adherent monolayer with a doubling time of 32 h and showed pleomorphic features. Intermediate filament analysis revealed the line to be mesenchymal in origin (reactivity to vimentin and desmin antibodies). The line was tumorigenic in nude mice, possessed elevated levels of creatine phosphokinase (mainly of the MM isoenzyme form) and had a near diploid mean chromosome number of 50. In vitro cell cloning determinations gave colony forming efficiencies of 0.01% in soft agar and 24% in a monolayer anchorage-dependent assay. Radiosensitivity determinations using a monolayer clonogenic assay with feeder layer support showed the cells to be among the more radiosensitive human tumour cell types (surviving fraction at 2 Gy of 0.26) that have been investigated. Furthermore, experiments utilising continuous low dose rate radiation at 3.2 cGy min-1, showed that, under these experimental conditions, the cells possessed only a very low capacity to recover from radiation-induced damage (dose reduction factor at 1% cell survival of 1.07 for 150 versus 3.2 cGy min-1). As other human tumour cells of an embryonal cell origin (e.g. neuroblastoma and germ cell tumours of the testis) have also been shown to be radiosensitive it appears that sensitivity to radiation may be a common property of this group of tumours.

Recovery from radiation damage in human squamous carcinoma of the cervix

The radiobiology of human carcinoma of the cervix has been investigated using five recently established cell lines. By means of an in vitro clonogenic assay, cell survival has been determined at dose-rates of 150, 3.2 and 1.6 cGy/min, and, in addition, split-dose experiments have been performed. Using approximately isoeffective doses (1 per cent cell survival), the amount of recovery observed using dose-rate and split-dose experiments has been correlated with the surviving fraction at 2 Gy (SF2 value) and the parameter alpha from the acute cell survival curves for the five lines. Two lines were radiosensitive, possessing significantly lower SF2 values (p = 0.03), and, at these isoeffect doses, showed significantly less sparing at low dose-rate (p = 0.012) and less split-dose recovery (p = 0.0096) than the three remaining cell lines. In general, the results indicate a good correlation between split-dose recovery-ratio, dose sparing achieved at low dose-rate compared to high, and SF2 value, and an inverse correlation with the alpha value derived from fitting dose rate dependence data to the incomplete-repair model of Thames. These observed differences in radiosensitivity within one human tumour type indicate the possible value of predictive testing of intrinsic radiosensitivity in the radiotherapy of carcinoma of the cervix.