- Email: [email protected]
Repopulation between radiation fractions in human melanoma xenografts
0360-3016/92 $5.00 + .UI Copyright 0 1992 Pergamon Press Ltd.
ht. J. Raduzrion Oncology Bid Phys.. Vol. 23, PD. 63-68 Printed in the U.S.A. All rights reserved.
0 Biology Original Contribution REPOPULATION BETWEEN RADIATION FRACTIONS IN HUMAN MELANOMA XENOGRAFTS EINAR K. ROFSTAD,
PH.D.
Institute for Cancer Research and The Norwegian Cancer Society, The Norwegian Radium Hospital, Montebello, 0310 Oslo 3, Norway Rate of tumor repopulation between radiation fractions was studied using two human melanoma xenograft lines (E.F. and V.N.). Tumors were given five radiation fractions under hypoxic conditions in viva and clonogenic cell survival was measured in vitro after the last radiation fraction. Dose-response curves were established for hrterfiaction times of 12,24,36, and 48 hr by varying dose per fraction from 4.2 to 11.2 Gy. Assuming an oxygen enhancement ratio of 2.8, these doses corresponded to doses of 1.5 to 4.0 Gy under aerobic conditions, that is, clinically relevant doses per fraction were used. The dose-response curves were nearly parallel and were shifted to the right with increasing interfraction time, demonstrating significant repopulation between the radiation fractions. Iso-effect analyses showed that additional radiation doses of 2.0 + 0.6 Gy/day (E.F.) and 2.2 + 0.6 Gy/day (V.N.), corresponding to doses of 0.7 + 0.2 Gy/day (E.F.) and 0.8 f 0.2 Gy/day under aerobic conditions, would be required to compensate for the repopulation. These doses were equivalent to the surviving clonogenic cells showing doubling times of 4050 hr (E.F.) and 30-40 hr (V.N.) during the treatment period. The radioresponsiveness of the two melanoma xenograft lines was also measured. Tumors in air-breathing mice were given from 5 to 15 daily fractions of 2.0 Gy and cell survival curves were established in vitro. Theoretical survival curves, calculated from SF2 in vitro and rate of repopulation during fractionated irradiation in vim, agreed fairly well with the measured survival curves. This suggested that the radioresponsiveness of the melanoma xenograft lines was governed by two main parameters: (a) cellular radiation sensitivity and repair capacity and (b) rate of repopulation between radiation fractions. Melanoma xenografts, Repopulation, SF2 in vitro, Prediction of radiation resistance.
INTRODUCTION
model system in our institute in attempts to develop predictive assays for clinical radiocurability ( 19). A significant correlation has been found between the radioresponsiveness of the melanomas in vivo (2.0 Gy fractions) and the initial slope of the cell survival curves in vitro (22). However, the melanomas were more resistant to radiation in vivo than predicted by SF2 in vitro alone, suggesting that other radiobiological parameters also were important. Previous studies have shown that the reoxygenation is rapid and extensive in the melanomas (20), suggesting that the radio-responsiveness was not significantly influenced by hypoxia. Moreover, none of the melanomas showed significant PLD-repair after fractionated irradiation in vivo (21), excluding also PLD-repair as a reason why the melanomas were more resistant in vivo than predicted by SF2 in vitro. The purpose of the study reported here was to investigate whether the radioresponsiveness of the melanomas is significantly influenced by repopu-
Experimental and clinical investigations suggest that in vitro assays for cellular radiation sensitivity may be useful in prediction of tumor radiocurability and development of individualized treatment strategies (9, 16). Thus, cell surviving fraction at 2.0 Gy (SF2) determined from survival curves of established human tumor cell lines has been found to correlate with clinical radiocurability of tumors of corresponding histology (5, 6). The radioresponsiveness of experimental tumors given fractionated irradiation in vivo has been shown to correlate with SF2 in vitro for cells isolated from the same tumors (2, 22). Moreover, preliminary clinical studies of squamous cell carcinoma of the head and neck (3) and of the uterine cervix (29) have indicated that SF2 in vitro may be higher for recurrent than for locally controlled tumors. Five human
melanoma
xenograft
lines are used as a
Reprint requests to: Einar K. Rofstad, Ph.D., Department of Biophysics, Institute for Cancer Research, The Norwegian Radium Hospital, Montebello, 03 10 Oslo 3, Norway. Acknowledgements-B. Mathiesen, K. Baekken, and H. Stageboe
Petersen are thanked for skillful technical assistance. Financial support was received from The Norwegian Cancer Society. Accepted for publication 30 September 199 1. 63
64
I. J. Radiation Oncology 0 Biology 0 Physics
lation
between
radiation
fractions.
(E.F.) and one rapidly-growing for this study. METHODS
AND
One slowly-growing
(V.N.) line were selected
MATERIALS
Mice and tumors Female BALB/c/nu/nu/BOM mice, bred at the animal department of our institute, were used. They were kept under specific pathogen-free conditions. The melanoma xenograft lines (E.F. and V.N.) were originally derived from lymph node metastases of patients admitted to The Norwegian Radium Hospital. Tumor tissue was transplanted directly into athymic mice without previous adaptation to in vitro culture conditions. Histologically the parent metastases were similar. They were composed of solid trabecules and nests of relatively large cells with hyperchromatic vesicular nuclei surrounded by partly abundant eosinophilic cytoplasm. Areas with more spindle-shaped cells were also seen. The cytoplasm contained little or no melanin. Numerous mitotic figures were found. The melanomas were grown serially in athymic mice by implanting tumor fragments, approximately 2 X 2 X 2 mm in size, subcutaneously into the flanks of recipient mice. Passages 35-60 of the melanomas were used in the present work. The melanomas were kinetically stable during the period the experiments were carried out, as ascertained by flow cytometric measurements of DNA histograms and measurements of volumetric growth rates. The volume-doubling times in the volume range 250500 mm3 were 18.2 + 2.1 days (E.F.) and 5.4 * 0.4 days (V.N.). Light and electron microscopic examinations showed that the histological appearance of the xenografts was similar to that of the metastases in the donor patients. The radiation biology of the melanomas has been studied in detail (17-23). The oxygen enhancement ratio (OER) of the melanoma cells has been determined to be 2.7-2.9 by irradiating cells from disaggregated tumors under aerobic and extremely hypoxic (~4 ppm 02) conditions in vitro (17). Irradiation An X ray unit,* operated at 220 kV, 19-20 mA, and with 0.5 mm Cu filtration, was used for irradiation. The tumors were irradiated in non-anesthetized mice at a dose rate of 5.1 Gy/min. Hypoxic conditions were obtained by applying a clamp occluding the tumor blood supply 5 min before irradiation. A 15 X 15 mm hole through a 2 cm thick lead block served as beam-defining aperture. During exposure the mice were kept in specially made, thin-walled polymethylmethacrylate tubes with a hole in
* Siemens Stabilipan, Erlangen, Germany. + Gibco Limited, Paisley, Scotland. * Lab-Blender 80, Seward Laboratory, London, UK.
Volume 23, Number 1, 1992
the cranial end through which they could breathe freely. A piston in the tail end positioned the mice firmly in the tubes. A hole was cut in each tube through which the tumors protruded. To ensure uniform doses throughout the tumor volumes, the tumors were exposed to radiation by two opposing treatment fields through each of which 50% of the dose was delivered. The tumors were irradiated when they attained a volume within the range 250-500 mm3. Callipers were used to measure tumor volumes. Two perpendicular diameters (length and width) were recorded, and the volumes were calculated as V = t - a - b2, where a and b are the longest and the shortest diameter, respectively. Colony assay The fraction of surviving cells in the tumors after irradiation in vivo was measured in vitro using a soft agar colony assay (4). Single cell suspensions were prepared from the tumors using a standardized mechanical procedure; the tumors were put into plastic bags with 20 ml culture medium (Ham’s F-12 medium+ with 20% fetal calf serum,+ penicillin+ (250 mg/l), and streptomycin+ (50 mg/l)) and disaggregated for 30 s with a stomacher.* The resulting suspensions were filtered through 30 pm nylon mesh. The cell concentrations were determined using a hemocytometer. The number of host cells in the tumors, especially macrophages, tended to increase with increasing time between the first and the last radiation exposure. The host cells could usually be distinguished easily from the melanoma cells on the basis of size. Melanoma cells having an intact and smooth outline with a bright halo were scored as morphologically intact and counted. The cell yield was calculated as the total number of morphologically intact melanoma cells divided by the tumor volume as measured immediately before the first radiation exposure. The soft agar was prepared from powdered agar§ and culture medium (see above). Erythrocytes from August rats and melanoma cells were added as described previously (17). Aliquots of 1 ml of soft agar with the appropriate number of melanoma cells were seeded in plastic tubes.** The cells were then incubated at 37°C for 3-4 weeks (E.F.) or 4-5 weeks (V.N.) in an atmosphere of 5% 02, 5% CO*, and 90% NZ. Culture medium (2 ml) was added on the top of the agar 5 days after seeding and then changed weekly. A stereomicroscope was used to count colonies. The dense colonies formed by the melanoma cells could be distinguished easily from the loose colonies formed by the macrophages. Melanoma cells giving rise to colonies larger than 50 cells were scored as surviving. The plating efficiency of the morphologically intact cells from unirradiated tumors was 30-60% (E.F.) or 15-35%
5 Bacto agar, Difco, Detroit, MI, USA. ** Falcon 2057 tubes, Becton Dickinson, Oxnard, CA, USA.
Repopulation during fractionated irradiation 0 E. K.
ROFSTAD
(V.N.). The fraction of surviving cells in an irradiated tumor was calculated from the mean number of colonies in four tubes with cells from that tumor and four tubes with cells from an unirradiated tumor, and the number of morphologically intact cells seeded and the cell yield for the two tumors, that is, the surviving fractions were measured relative to the number of clonogenic cells in the tumors immediately before the first radiation exposure and thus represent the fraction of clonogenic cells per tumor that survived the treatment.
65
Hypox~c
Conddmns
: 2.0 * 0.6
Gyldoy
Hypox~c
Conditmns:
2.2 f 0.6
Gy/doy
Aerobic
Conditions:
0.8 f 0.2
Gylday
Data analysis Survival curves were fitted to the data by least-squares linear regression analysis of log surviving fraction versus total radiation dose. The analysis was based on surviving fractions measured for individual tumors. RESULTS
Cell survival curves for tumors given five equal radiation fractions under clamped conditions are presented in Figure 1. Four different interfraction times, 12, 24, 36, and 48 hr, were used. The tumors were excised and assayed in vitro 12, 24, 36, or 48 hr after the last radiation exposure depending on the interfraction time, that is, the total treatment times were 2.5, 5.0, 7.5, and 10.0 days. The irradiation was performed under hypoxic conditions to avoid effects of reoxygenation. Total doses of 2 1 to 56 Gy were used. Assuming an OER of 2.8, these doses corresponded to doses per fraction of 1.5 to 4.0 Gy under aerobic conditions. The cross-hatched areas represent the ranges in which the survival curves would be expected to be positioned if the radiation response of the tumors were determined solely by the intrinsic radiation sensitivity of
J Fig. 1. Radiation survival curves, that is, fraction of clonogenic cells per tumor that survived the treatment versustotal radiation dose, for two human melanoma xenograft lines. The tumors were given five equal radiation fractions under clamped conditions. The interfraction times were 12 hr (0), 24 hr (a), 36 hr (A), or 48 hr (A). The points represent the means of five to eight individual tumors and the bars, standard errors. The crosshatched areas represent the ranges in which the survival curves would be expected to be positioned if the radiation response of the tumors were determined solely by the intrinsic radiation sensitivity of the tumor cells.
01 0
2.5
10.0
7.5
5.0 TREATMENT
TIME
12.5
(days)
Fig. 2. Total radiation dose necessary to reach a surviving fraction of 1 . lo-* wsus total treatment time for two human melanoma xenograft lines. The data were derived from the survival curves in Figure 1. The points and bars represent mean values and standard errors. The slopes of the curves, indicated in the figure, represent the radiation doses per day that would be required to compensate for tumor repopulation between radiation fractions.
the tumor cells and their repair capacity. These ranges
were calculated on the basis of survival curves for cells that were isolated from tumors of the two melanoma lines and irradiated in vitro under aerobic conditions (22). Thus, the cross-hatched areas were determined by raising to the fifth power the surviving fractions L SE (standard errors) for cells irradiated in vitro and assuming an OER of 2.8. The measured survival curves were positioned above the cross-hatched areas and the curves were gradually shifted to the right with increasing interfraction time, showing that tumor repopulation took place between the radiation fractions. The magnitude of the repopulation was quantified by performing iso-effect analyses of the data in Figure 1. Figure 2 shows total dose necessary to reach a surviving fraction of 1 - 10d2 as a function of total treatment time. The slopes of the curves represent the radiation doses per day that would be required to compensate for the increase in clonogenic cell number due to repopulation. These doses, determined by linear regression analyses, were found to be 2.0 k 0.6 Gy/day (E.F.) and 2.2 + 0.6 Gy/day (V.N.) for hypoxic tumors. Assuming an OER of 2.8, these doses correspond to doses of 0.7 f 0.2 Gy/day (E.F.) and 0.8 + 0.2 Gy/day (V.N.) under aerobic conditions. Similar analyses were also performed using surviving fractions of 1 * 10m3and 1 - 1O-4 as iso-effect levels. The doses quantifying the repopulation were found to be the same irre-
66
I. J. Radiation Oncology 0 Biology0 Physics
spective of iso-effect level (the survival curves in Figure 1
are nearly parallel). The corresponding doubling times of the surviving clonogenic tumor cells were calculated to be 40-50 hr (E.F.) and 30-40 hr (V.N.) by performing iso-dose analyses of the data in Figure 1; the increase in surviving fraction with increasing total treatment time was quantified at three different iso-dose levels (28, 35, and 42 Gy). To investigate whether the radioresponsiveness of the two melanoma lines (2.0 Gy fractions) could be explained from SF2 in vitro and rate of repopulation between radiation fractions without taking other radiobiological parameters into consideration, experiments involving fractionated irradiation of tumors in air-breathing mice were. performed. Tumors were irradiated with daily fractions of 2.0 Gy (24 hr between each fraction) and cell surviving fractions were measured in vitro 24 hr after the last radiation exposure. Total doses ranging from 10 to 30 Gy were used. The survival curves are presented in Figure 3. The cross-hatched and the shaded areas represent the ranges in which the survival curves would be expected to be positioned if the radioresponsiveness of the tumors were determined solely by SF2 in vitro (cross-hatched areas) or by SF2 in vitro and rate of repopulation between radiation fractions in vivo (shaded areas). Thus, the crosshatched areas were defined by (SF2 f SE)N, where N represents the number of 2.0 Gy fractions. The shaded areas were derived from the cross-hatched areas assuming that doses of 0.7 Gy/day (E.F.) and 0.8 Gy/day (V.N.) would be required to compensate for the repopulation. There was a fairly good agreement between the measured survival curves and the shaded areas, although the measured
I lo
20
30
I o-0
DOSE IGyl
I 10
20
30
Fig. 3. Radiation survival curves, that is, fraction of clonogenic cells per tumor that survived the treatment versustotal radiation dose, for two human melanoma xenograft lines. The tumors were irradiated in air-breathing mice with daily fractions of 2.0 Gy (24 hr between each fraction). The points represent the means of five to eight individual tumors and the bars, standard errors. The cross-hatched and the shaded areas represent the ranges in which the survival curves would be expected to be positioned if the radiation response of the tumors were determined solely by the intrinsic radiation sensitivity of the tumor cells (cross-hatched areas) or by the intrinsic radiation sensitivity of the tumor cells and the repopulation between radiation fractions (shaded areas).
Volume 23, Number 1, 1992
cell surviving fractions were in the upper part of or slightly above these areas (Fig. 3). DISCUSSION
Repopulation during treatment Two human melanoma xenograft lines were used to study the influence of repopulation between radiation fractions on tumor radioresponsiveness. Tumors were given fractionated irradiation with clinically relevant doses per fraction. The magnitude of the repopulation was quantified by comparing the therapeutic effects of fractionation regimens that were identical in all respects except for overall treatment time. There are methodological problems involved in such studies (27). Firstly, possible effects of reoxygenation were avoided by irradiating the tumors under hypoxic conditions. However, irradiation under hypoxic conditions may possibly overestimate the rate of repopulation. Tumor clamping leads to a uniform distribution of cell inactivation throughout the tumor and the repopulation will therefore tend to occur from surviving cells in normally well-oxygenated regions, cells which would have been inactivated by the radiation under air-breathing conditions. An overestimation of the rate of repopulation is probably small, if present at all, in the experiments described here because a significant fraction of the hypoxic cells in the tumors under air-breathing conditions are acutely hypoxic (20). Secondly, interfraction times of 12 to 48 hr were chosen to eliminate, as far as possible, confounding effects of incomplete repair and redistribution between radiation fractions. However, the possibility that the results to some minor extent were influenced by these two radiobiological phenomena cannot be excluded. An iso-effect analysis such as that presented in Figure 2 is the only reliable method of measuring repopulation between radiation fractions and its influence on tumor radioresponsiveness. Other possible methods, for example, cell kinetic, histopathological, biochemical, or genetic methods, give only data describing the behavior of mostly radiation-inactivated, non-clonogenic tumor cells. The rate of repopulation during fractionated irradiation was significant in both melanoma lines. Compensation of the repopulation was found to require additional doses of 2.0 + 0.6 Gy/day (E.F.) and 2.2 k 0.6 Gy/day (V.N.) under hypoxic conditions and 0.7 t- 0.2 Gy/day (E.F.) and 0.8 * 0.2 Gy/day (V.N.) under aerobic conditions, corresponding to doubling times of 40-50 hr (E.F.) and 30-40 hr (V.N.) of the surviving clonogenic tumor cells. The median cell cycle time of untreated tumors of most human melanoma xenograft lines is 30-40 hr (23, 25) and this cell cycle time is usually shortened after irradiation (23, 24). The volume-doubling times of the lines studied here were 18.2 t 2.1 days (E.F.) and 5.4 + 0.4 days (V.N.) and the potential doubling times have been calculated to be 80-l 10 hr (E.F.) and 60-85 hr (V.N.)
Repopulation during fractionated irradiation 0 E. K. ROFSTAD
(18). Accordingly, the doubling times of the clonogenic tumor cells during the treatment period were somewhere between the cell cycle times and the potential doubling times. This observation indicates that it may be possible to estimate the rate of repopulation in tumors during fractionated irradiation from the cell cycle time and/or the potential doubling time before treatment (1, 8). Studies of transplantable rodent tumors have also indicated rapid repopulation following radiation treatment. Hermens and Barendsen (11) demonstrated a hundredfold increase in the clonogenic cell fraction in a rat tumor within few days after a large single radiation dose, corresponding to a doubling time of the clonogenic cells as short as the cell cycle time of less than 24 hr. Suit et al. (26) showed a large increase in TCDso with increasing overall treatment time in a mouse tumor given multifraction radiation treatment, corresponding to slightly more than one doubling of the number of clonogenic cells within 24 hr. The most comprehensive studies have been reported by Trott (27) and Trott and Kummermehr (28). They studied five mouse tumor lines that were treated with daily radiation fractions over three weeks, and found that the repopulation rates were not constant throughout the course of treatment, that is, they were different in the first, second, and third week. Moreover, the rate of repopulation as well as the changes with time depended on tumor line. The percentage of the total dose that was used to compensate for the repopulation ranged from 16% to 67%. Corresponding percentages for the two melanoma lines studied here were in the range from 20% to 53%, depending on the radiation dose per fraction. Analyses of clinical data have also given evidence that the repopulation may be rapid during the treatment period and significantly reduce local tumor control rates in protracted regimens (7, 8). Thus, trials using split-course irradiation to spare acute radiation damage have suggested that 2-3 extra fractions of 4-6 Gy are required to compensate for a 2-3 week gap in a 20-30 fraction regimen ( 12, 14, 15). Data allowing more rigid analyses are sparse. However, Trott and Kummermehr (28) have analyzed data from Friedman et al. (10) on local control of Hodgkin’s disease and found that an additional dose of 0.3 Gy/ day was needed to compensate for the repopulation when total treatment time exceeded three weeks. Similar analysis of local control rates in T3-4 squamous cell carcinoma of the larynx has suggested a repopulation rate equivalent to an additional dose of 0.5 Gy/day (13). These doses found for Hodgkin’s disease and squamous cell carcinoma are comparable to, but somewhat lower than, those measured for the melanoma lines studied here.
Tumor radioresponsiveness Previous work involving five human melanoma xenograft lines showed that the lines differed considerably in radioresponsiveness in vivo (2.0 Gy fractions) (22). SF2
61
in vitro also differed considerably among the lines and the radioresponsiveness in vivo showed a clear correlation to SF, in vitro (22). Consequently, the differences in radioresponsiveness among the lines were mainly due to differences in cellular radiation sensitivity and repair capacity. However, the absolute level of radioresponsiveness in vivo could not be explained from SF2 in vitro alone; the tumors were too resistant in vivo (22). The present work revealed that the repopulation between radiation fractions was significant in the tumors. Two lines, one having a short and the other a long volume-doubling time, were studied and the rate of repopulation, corresponding to doses of 0.7-0.8 Gy/day, was found to be approximately equal for the two lines. Consequently, repopulation between radiation fractions strongly reduced the absolute level of radioresponsiveness, but did not influence significantly the differences in radioresponsiveness among the lines. Theoretical values for tumor radioresponsiveness, calculated from SF;?in vitro and rate of repopulation during fractionated irradiation in vivo, agreed fairly well with the absolute level of radioresponsiveness measured after treatment with 2.0 Gy fractions in vivo (Fig. 3). This suggests that the absolute level of radioresponsiveness of the melanoma lines was governed by two main parameters: (a) cellular radiation sensitivity and repair capacity and (b) rate of repopulation between radiation fractions. However, the measured values for tumor radioresponsiveness were in the upper part of or slightly above the shaded areas in Figure 3, representing the theoretical values. This is an indication that also other radiobiological parameters may have given minor contributions to the absolute level of radioresponsiveness. A contribution from tumor hypoxia is one possibility. Although the tumors showed rapid and extensive reoxygenation after treatment with 2.0 Gy fractions, the fractions of hypoxic cells never decreased to levels below those in untreated tumors, that is, 22 f 8% in the E.F. line and 15 + 5% in the V.N. line (20). The human melanoma xenograft lines studied here showed short volume-doubling times compared with most tumors in patients. It is therefore possible that the consequences of repopulation between radiation fractions are less in clinical radiation therapy than predicted by the present experimental study. However, the radiation doses required to compensate for repopulation in the melanoma xenografts were only slightly larger than those estimated from clinical data. It is therefore highly probable that the clinical radioresponsiveness of tumors also is significantly influenced by repopulation, as is the radioresponsiveness of melanoma xenografts. Consequently, clinical investigations of the utilitarian value of possible predictive assays of tumor radioresponsiveness should include assays meaSUkXg cell repopulation rate as well as cellular radiation sensitivity.
68
I. J. Radiation Oncology 0 Biology 0 Physics
Volume 23, Number 1, 1992
REFERENCES 1. Begg, A. C.; Hofland, I.; Moonen, L.; Bartelink, H.; Schraub,
2.
3.
4.
5.
6.
I. 8.
9.
10.
11.
12.
13.
14.
S.; Bontemps, P.; Le Fur, R.; Van Den Bogaert, W.; Caspers, R.; Van Glabbeke, M.; Horiot, J. C. The predictive value of cell kinetic measurements in a European trial of accelerated fractionation in advanced head and neck tumors: an interim report. Int. J. Radiat. Oncol. Biol. Phys. 19: 14491453; 1990. Bristow, R. G.; Hill, R. P. Comparison between in vitro radiosensitivity and in vivo radioresponse in murine tumor cell lines II: in vivo radioresponse following fractionated treatment and in vitro/in vivo correlations. Int. J. Radiat. Oncol. Biol. Phys. 18: 33 l-345; 1990. Brock, W. A.; Baker, F. L.; Wike, J. L.; Sivon, S. L.; Peters, L. J. Cellular radiosensitivity of primary head and neck squamous cell carcinomas and local tumor control. Int. J. Radiat. Oncol. Biol. Phys. 18: 1283-1286; 1990. Courtenay, V. D.; Mills, J. An in vitro colony assay for human tumours grown in immune-suppressed mice and treated in vivo with cytotoxic agents. Br. J. Cancer 37: 26 l268; 1978. Deacon, J.; Peckham, M. J.; Steel, G. G. The radioresponsiveness of human tumours and the initial slope of the cell survival curve. Radiother. Oncol. 2: 3 17-323; 1984. Fertil, B.; Malaise, E. P. Inherent cellular radiosensitivity as a basic concept for human tumor radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 7: 62 l-629; 198 1. Fowler, J. F. Non-standard fractionation in radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 10: 755-759; 1984. Fowler, J. F. Potential for increasing the differential response between tumors and normal tissues: can proliferation rate be used? Int. J. Radiat. Oncol. Biol. Phys. 12: 641-645; 1986. Fowler, J. F. Scientific perspectives on recent progress in radiation oncology. Int. J. Radiat. Oncol. Biol. Phys. 12: 709-712; 1986. Friedman, M.; Pearlman, W. W.; Turgeon, L. Hodgkin’s disease, tumor lethal dose and iso-effect recovery curve. Am. J. Radiol. 99: 843-851; 1967. Hermens, A. F.; Barendsen, G. W. Changes of cell proliferation characteristics in a rat rhabdomyosarcoma before and after X-irradiation. Eur. J. Cancer 5: 173-189; 1989. Holsti, L. R. Clinical experience with split-course radiotherapy. A randomized clinical trial. Radiology 92: 591596; 1969. Maciejewski, B.; Preuss-Bayer, G.; Trott, K. R. The influence of the number of fractions and of overall treatment time on local control and late complication rate in squamous cell carcinoma of the larynx. Int. J. Radiat. Oncol. Biol. Phys. 9: 321-328; 1983. Overgaard, J.; Hjelm-Hansen, M.; Vendelbo Johansen, L.; Andersen, A. P. Comparison of conventional and splitcourse radiotherapy as primary treatment in carcinoma of the larynx. Acta Oncol. 27: 147-152; 1988.
15. Parsons, J.; Bova, F. J.; Million, R. R. A re-evaluation of split-course technique for squamous cell carcinoma of the head and neck. Int. J. Radiat. Oncol. Biol. Phys. 6: 16451652; 1980. 16. Peters, L. J. Inherent radiosensitivity of tumor and normal tissue cells as a predictor of human tumor response. Radiother. Oncol. 17: 177-190; 1990. 17. Rofstad, E. K. Radiation response of the cells of a human malignant melanoma xenograft. Effect of hypoxic cell radiosensitizers. Radiat. Res. 87: 670-683; 198 1. 18. Rofstad, E. K. Growth and vascular structure of human melanoma xenografts. Cell Tissue Kinet. 17: 9 l- 10 1; 1984. 19. Rofstad, E. K. Human tumor xenografts in development of predictive assays of tumor treatment response. In: Chapman, J. D., Peters, L. J., Withers, H. R., eds. Prediction of tumor treatment response. New York: Pergamon Press; 1989: 197-216. 20. Rofstad, E. K. Hypoxia and reoxygenation in human melanoma xenografts. Int. J. Radiat. Oncol. Biol. Phys. 17: 8 l89; 1989. 21. Rofstad, E. K. PLD-repair in human melanoma xenografts following single dose and fractionated irradiation. Br. J. Cancer 6 1: 856-860; 1990. 22. Rofstad, E. K.; Brustad, T. Radioresponsiveness of human melanoma xenografts given fractionated irradiation in vivorelationship to the initial slope of the cell survival curves in vitro. Radiother. Oncol. 9: 45-56; 1987. 23. Rofstad, E. K.; Lindmo, T.; Brustad, T. Effect of single dose irradiation on the proliferation kinetics in a human malignant melanoma in athymic nude mice. Acta Radiol. Oncol. 19: 261-269; 1980. 24. Spang-Thomsen, M.; Vindelev, L. L. Changes of proliferation kinetics after X-irradiation of a human malignant melanoma grown in nude mice. Eur. J. Cancer Clin. Oncol. 20: 849-855; 1984. 25. Spang-Thomsen, M.; Vindelov, L. L. Proliferation kinetics of a human malignant melanoma serially grown in nude mice. Cell Tissue Kinet. 17: 401-410; 1984. 26. Suit, H. D.; Howes, A. E.; Hunter, N. Dependence of response of a C3H mammary carcinoma to fractionated irradiation on fractionation number and intertreatment interval. Radiat. Res. 72: 440-454; 1977. 27. Trott, K. R. Cell proliferation and overall treatment time. Int. J. Radiat. Oncol. Biol. Phys. 19: 1071-1075; 1990. 28. Trot& K. R.; Kummermehr, J. What is known about tumour proliferation rates to choose between accelerated fractionation and hyperfractionation? Radiother. Oncol. 3: 1-9; 1985. 29. West, C. M. L.; Davidson, S. E.; Hunter, R. D. Evaluation of surviving fraction at 2 Gy as a potential prognostic factor for the radiotherapy of carcinoma of the cervix. Int. J. Radiat. Biol. 56: 761-765; 1989.
ht. J. Raduzrion Oncology Bid Phys.. Vol. 23, PD. 63-68 Printed in the U.S.A. All rights reserved.
0 Biology Original Contribution REPOPULATION BETWEEN RADIATION FRACTIONS IN HUMAN MELANOMA XENOGRAFTS EINAR K. ROFSTAD,
PH.D.
Institute for Cancer Research and The Norwegian Cancer Society, The Norwegian Radium Hospital, Montebello, 0310 Oslo 3, Norway Rate of tumor repopulation between radiation fractions was studied using two human melanoma xenograft lines (E.F. and V.N.). Tumors were given five radiation fractions under hypoxic conditions in viva and clonogenic cell survival was measured in vitro after the last radiation fraction. Dose-response curves were established for hrterfiaction times of 12,24,36, and 48 hr by varying dose per fraction from 4.2 to 11.2 Gy. Assuming an oxygen enhancement ratio of 2.8, these doses corresponded to doses of 1.5 to 4.0 Gy under aerobic conditions, that is, clinically relevant doses per fraction were used. The dose-response curves were nearly parallel and were shifted to the right with increasing interfraction time, demonstrating significant repopulation between the radiation fractions. Iso-effect analyses showed that additional radiation doses of 2.0 + 0.6 Gy/day (E.F.) and 2.2 + 0.6 Gy/day (V.N.), corresponding to doses of 0.7 + 0.2 Gy/day (E.F.) and 0.8 f 0.2 Gy/day under aerobic conditions, would be required to compensate for the repopulation. These doses were equivalent to the surviving clonogenic cells showing doubling times of 4050 hr (E.F.) and 30-40 hr (V.N.) during the treatment period. The radioresponsiveness of the two melanoma xenograft lines was also measured. Tumors in air-breathing mice were given from 5 to 15 daily fractions of 2.0 Gy and cell survival curves were established in vitro. Theoretical survival curves, calculated from SF2 in vitro and rate of repopulation during fractionated irradiation in vim, agreed fairly well with the measured survival curves. This suggested that the radioresponsiveness of the melanoma xenograft lines was governed by two main parameters: (a) cellular radiation sensitivity and repair capacity and (b) rate of repopulation between radiation fractions. Melanoma xenografts, Repopulation, SF2 in vitro, Prediction of radiation resistance.
INTRODUCTION
model system in our institute in attempts to develop predictive assays for clinical radiocurability ( 19). A significant correlation has been found between the radioresponsiveness of the melanomas in vivo (2.0 Gy fractions) and the initial slope of the cell survival curves in vitro (22). However, the melanomas were more resistant to radiation in vivo than predicted by SF2 in vitro alone, suggesting that other radiobiological parameters also were important. Previous studies have shown that the reoxygenation is rapid and extensive in the melanomas (20), suggesting that the radio-responsiveness was not significantly influenced by hypoxia. Moreover, none of the melanomas showed significant PLD-repair after fractionated irradiation in vivo (21), excluding also PLD-repair as a reason why the melanomas were more resistant in vivo than predicted by SF2 in vitro. The purpose of the study reported here was to investigate whether the radioresponsiveness of the melanomas is significantly influenced by repopu-
Experimental and clinical investigations suggest that in vitro assays for cellular radiation sensitivity may be useful in prediction of tumor radiocurability and development of individualized treatment strategies (9, 16). Thus, cell surviving fraction at 2.0 Gy (SF2) determined from survival curves of established human tumor cell lines has been found to correlate with clinical radiocurability of tumors of corresponding histology (5, 6). The radioresponsiveness of experimental tumors given fractionated irradiation in vivo has been shown to correlate with SF2 in vitro for cells isolated from the same tumors (2, 22). Moreover, preliminary clinical studies of squamous cell carcinoma of the head and neck (3) and of the uterine cervix (29) have indicated that SF2 in vitro may be higher for recurrent than for locally controlled tumors. Five human
melanoma
xenograft
lines are used as a
Reprint requests to: Einar K. Rofstad, Ph.D., Department of Biophysics, Institute for Cancer Research, The Norwegian Radium Hospital, Montebello, 03 10 Oslo 3, Norway. Acknowledgements-B. Mathiesen, K. Baekken, and H. Stageboe
Petersen are thanked for skillful technical assistance. Financial support was received from The Norwegian Cancer Society. Accepted for publication 30 September 199 1. 63
64
I. J. Radiation Oncology 0 Biology 0 Physics
lation
between
radiation
fractions.
(E.F.) and one rapidly-growing for this study. METHODS
AND
One slowly-growing
(V.N.) line were selected
MATERIALS
Mice and tumors Female BALB/c/nu/nu/BOM mice, bred at the animal department of our institute, were used. They were kept under specific pathogen-free conditions. The melanoma xenograft lines (E.F. and V.N.) were originally derived from lymph node metastases of patients admitted to The Norwegian Radium Hospital. Tumor tissue was transplanted directly into athymic mice without previous adaptation to in vitro culture conditions. Histologically the parent metastases were similar. They were composed of solid trabecules and nests of relatively large cells with hyperchromatic vesicular nuclei surrounded by partly abundant eosinophilic cytoplasm. Areas with more spindle-shaped cells were also seen. The cytoplasm contained little or no melanin. Numerous mitotic figures were found. The melanomas were grown serially in athymic mice by implanting tumor fragments, approximately 2 X 2 X 2 mm in size, subcutaneously into the flanks of recipient mice. Passages 35-60 of the melanomas were used in the present work. The melanomas were kinetically stable during the period the experiments were carried out, as ascertained by flow cytometric measurements of DNA histograms and measurements of volumetric growth rates. The volume-doubling times in the volume range 250500 mm3 were 18.2 + 2.1 days (E.F.) and 5.4 * 0.4 days (V.N.). Light and electron microscopic examinations showed that the histological appearance of the xenografts was similar to that of the metastases in the donor patients. The radiation biology of the melanomas has been studied in detail (17-23). The oxygen enhancement ratio (OER) of the melanoma cells has been determined to be 2.7-2.9 by irradiating cells from disaggregated tumors under aerobic and extremely hypoxic (~4 ppm 02) conditions in vitro (17). Irradiation An X ray unit,* operated at 220 kV, 19-20 mA, and with 0.5 mm Cu filtration, was used for irradiation. The tumors were irradiated in non-anesthetized mice at a dose rate of 5.1 Gy/min. Hypoxic conditions were obtained by applying a clamp occluding the tumor blood supply 5 min before irradiation. A 15 X 15 mm hole through a 2 cm thick lead block served as beam-defining aperture. During exposure the mice were kept in specially made, thin-walled polymethylmethacrylate tubes with a hole in
* Siemens Stabilipan, Erlangen, Germany. + Gibco Limited, Paisley, Scotland. * Lab-Blender 80, Seward Laboratory, London, UK.
Volume 23, Number 1, 1992
the cranial end through which they could breathe freely. A piston in the tail end positioned the mice firmly in the tubes. A hole was cut in each tube through which the tumors protruded. To ensure uniform doses throughout the tumor volumes, the tumors were exposed to radiation by two opposing treatment fields through each of which 50% of the dose was delivered. The tumors were irradiated when they attained a volume within the range 250-500 mm3. Callipers were used to measure tumor volumes. Two perpendicular diameters (length and width) were recorded, and the volumes were calculated as V = t - a - b2, where a and b are the longest and the shortest diameter, respectively. Colony assay The fraction of surviving cells in the tumors after irradiation in vivo was measured in vitro using a soft agar colony assay (4). Single cell suspensions were prepared from the tumors using a standardized mechanical procedure; the tumors were put into plastic bags with 20 ml culture medium (Ham’s F-12 medium+ with 20% fetal calf serum,+ penicillin+ (250 mg/l), and streptomycin+ (50 mg/l)) and disaggregated for 30 s with a stomacher.* The resulting suspensions were filtered through 30 pm nylon mesh. The cell concentrations were determined using a hemocytometer. The number of host cells in the tumors, especially macrophages, tended to increase with increasing time between the first and the last radiation exposure. The host cells could usually be distinguished easily from the melanoma cells on the basis of size. Melanoma cells having an intact and smooth outline with a bright halo were scored as morphologically intact and counted. The cell yield was calculated as the total number of morphologically intact melanoma cells divided by the tumor volume as measured immediately before the first radiation exposure. The soft agar was prepared from powdered agar§ and culture medium (see above). Erythrocytes from August rats and melanoma cells were added as described previously (17). Aliquots of 1 ml of soft agar with the appropriate number of melanoma cells were seeded in plastic tubes.** The cells were then incubated at 37°C for 3-4 weeks (E.F.) or 4-5 weeks (V.N.) in an atmosphere of 5% 02, 5% CO*, and 90% NZ. Culture medium (2 ml) was added on the top of the agar 5 days after seeding and then changed weekly. A stereomicroscope was used to count colonies. The dense colonies formed by the melanoma cells could be distinguished easily from the loose colonies formed by the macrophages. Melanoma cells giving rise to colonies larger than 50 cells were scored as surviving. The plating efficiency of the morphologically intact cells from unirradiated tumors was 30-60% (E.F.) or 15-35%
5 Bacto agar, Difco, Detroit, MI, USA. ** Falcon 2057 tubes, Becton Dickinson, Oxnard, CA, USA.
Repopulation during fractionated irradiation 0 E. K.
ROFSTAD
(V.N.). The fraction of surviving cells in an irradiated tumor was calculated from the mean number of colonies in four tubes with cells from that tumor and four tubes with cells from an unirradiated tumor, and the number of morphologically intact cells seeded and the cell yield for the two tumors, that is, the surviving fractions were measured relative to the number of clonogenic cells in the tumors immediately before the first radiation exposure and thus represent the fraction of clonogenic cells per tumor that survived the treatment.
65
Hypox~c
Conddmns
: 2.0 * 0.6
Gyldoy
Hypox~c
Conditmns:
2.2 f 0.6
Gy/doy
Aerobic
Conditions:
0.8 f 0.2
Gylday
Data analysis Survival curves were fitted to the data by least-squares linear regression analysis of log surviving fraction versus total radiation dose. The analysis was based on surviving fractions measured for individual tumors. RESULTS
Cell survival curves for tumors given five equal radiation fractions under clamped conditions are presented in Figure 1. Four different interfraction times, 12, 24, 36, and 48 hr, were used. The tumors were excised and assayed in vitro 12, 24, 36, or 48 hr after the last radiation exposure depending on the interfraction time, that is, the total treatment times were 2.5, 5.0, 7.5, and 10.0 days. The irradiation was performed under hypoxic conditions to avoid effects of reoxygenation. Total doses of 2 1 to 56 Gy were used. Assuming an OER of 2.8, these doses corresponded to doses per fraction of 1.5 to 4.0 Gy under aerobic conditions. The cross-hatched areas represent the ranges in which the survival curves would be expected to be positioned if the radiation response of the tumors were determined solely by the intrinsic radiation sensitivity of
J Fig. 1. Radiation survival curves, that is, fraction of clonogenic cells per tumor that survived the treatment versustotal radiation dose, for two human melanoma xenograft lines. The tumors were given five equal radiation fractions under clamped conditions. The interfraction times were 12 hr (0), 24 hr (a), 36 hr (A), or 48 hr (A). The points represent the means of five to eight individual tumors and the bars, standard errors. The crosshatched areas represent the ranges in which the survival curves would be expected to be positioned if the radiation response of the tumors were determined solely by the intrinsic radiation sensitivity of the tumor cells.
01 0
2.5
10.0
7.5
5.0 TREATMENT
TIME
12.5
(days)
Fig. 2. Total radiation dose necessary to reach a surviving fraction of 1 . lo-* wsus total treatment time for two human melanoma xenograft lines. The data were derived from the survival curves in Figure 1. The points and bars represent mean values and standard errors. The slopes of the curves, indicated in the figure, represent the radiation doses per day that would be required to compensate for tumor repopulation between radiation fractions.
the tumor cells and their repair capacity. These ranges
were calculated on the basis of survival curves for cells that were isolated from tumors of the two melanoma lines and irradiated in vitro under aerobic conditions (22). Thus, the cross-hatched areas were determined by raising to the fifth power the surviving fractions L SE (standard errors) for cells irradiated in vitro and assuming an OER of 2.8. The measured survival curves were positioned above the cross-hatched areas and the curves were gradually shifted to the right with increasing interfraction time, showing that tumor repopulation took place between the radiation fractions. The magnitude of the repopulation was quantified by performing iso-effect analyses of the data in Figure 1. Figure 2 shows total dose necessary to reach a surviving fraction of 1 - 10d2 as a function of total treatment time. The slopes of the curves represent the radiation doses per day that would be required to compensate for the increase in clonogenic cell number due to repopulation. These doses, determined by linear regression analyses, were found to be 2.0 k 0.6 Gy/day (E.F.) and 2.2 + 0.6 Gy/day (V.N.) for hypoxic tumors. Assuming an OER of 2.8, these doses correspond to doses of 0.7 f 0.2 Gy/day (E.F.) and 0.8 + 0.2 Gy/day (V.N.) under aerobic conditions. Similar analyses were also performed using surviving fractions of 1 * 10m3and 1 - 1O-4 as iso-effect levels. The doses quantifying the repopulation were found to be the same irre-
66
I. J. Radiation Oncology 0 Biology0 Physics
spective of iso-effect level (the survival curves in Figure 1
are nearly parallel). The corresponding doubling times of the surviving clonogenic tumor cells were calculated to be 40-50 hr (E.F.) and 30-40 hr (V.N.) by performing iso-dose analyses of the data in Figure 1; the increase in surviving fraction with increasing total treatment time was quantified at three different iso-dose levels (28, 35, and 42 Gy). To investigate whether the radioresponsiveness of the two melanoma lines (2.0 Gy fractions) could be explained from SF2 in vitro and rate of repopulation between radiation fractions without taking other radiobiological parameters into consideration, experiments involving fractionated irradiation of tumors in air-breathing mice were. performed. Tumors were irradiated with daily fractions of 2.0 Gy (24 hr between each fraction) and cell surviving fractions were measured in vitro 24 hr after the last radiation exposure. Total doses ranging from 10 to 30 Gy were used. The survival curves are presented in Figure 3. The cross-hatched and the shaded areas represent the ranges in which the survival curves would be expected to be positioned if the radioresponsiveness of the tumors were determined solely by SF2 in vitro (cross-hatched areas) or by SF2 in vitro and rate of repopulation between radiation fractions in vivo (shaded areas). Thus, the crosshatched areas were defined by (SF2 f SE)N, where N represents the number of 2.0 Gy fractions. The shaded areas were derived from the cross-hatched areas assuming that doses of 0.7 Gy/day (E.F.) and 0.8 Gy/day (V.N.) would be required to compensate for the repopulation. There was a fairly good agreement between the measured survival curves and the shaded areas, although the measured
I lo
20
30
I o-0
DOSE IGyl
I 10
20
30
Fig. 3. Radiation survival curves, that is, fraction of clonogenic cells per tumor that survived the treatment versustotal radiation dose, for two human melanoma xenograft lines. The tumors were irradiated in air-breathing mice with daily fractions of 2.0 Gy (24 hr between each fraction). The points represent the means of five to eight individual tumors and the bars, standard errors. The cross-hatched and the shaded areas represent the ranges in which the survival curves would be expected to be positioned if the radiation response of the tumors were determined solely by the intrinsic radiation sensitivity of the tumor cells (cross-hatched areas) or by the intrinsic radiation sensitivity of the tumor cells and the repopulation between radiation fractions (shaded areas).
Volume 23, Number 1, 1992
cell surviving fractions were in the upper part of or slightly above these areas (Fig. 3). DISCUSSION
Repopulation during treatment Two human melanoma xenograft lines were used to study the influence of repopulation between radiation fractions on tumor radioresponsiveness. Tumors were given fractionated irradiation with clinically relevant doses per fraction. The magnitude of the repopulation was quantified by comparing the therapeutic effects of fractionation regimens that were identical in all respects except for overall treatment time. There are methodological problems involved in such studies (27). Firstly, possible effects of reoxygenation were avoided by irradiating the tumors under hypoxic conditions. However, irradiation under hypoxic conditions may possibly overestimate the rate of repopulation. Tumor clamping leads to a uniform distribution of cell inactivation throughout the tumor and the repopulation will therefore tend to occur from surviving cells in normally well-oxygenated regions, cells which would have been inactivated by the radiation under air-breathing conditions. An overestimation of the rate of repopulation is probably small, if present at all, in the experiments described here because a significant fraction of the hypoxic cells in the tumors under air-breathing conditions are acutely hypoxic (20). Secondly, interfraction times of 12 to 48 hr were chosen to eliminate, as far as possible, confounding effects of incomplete repair and redistribution between radiation fractions. However, the possibility that the results to some minor extent were influenced by these two radiobiological phenomena cannot be excluded. An iso-effect analysis such as that presented in Figure 2 is the only reliable method of measuring repopulation between radiation fractions and its influence on tumor radioresponsiveness. Other possible methods, for example, cell kinetic, histopathological, biochemical, or genetic methods, give only data describing the behavior of mostly radiation-inactivated, non-clonogenic tumor cells. The rate of repopulation during fractionated irradiation was significant in both melanoma lines. Compensation of the repopulation was found to require additional doses of 2.0 + 0.6 Gy/day (E.F.) and 2.2 k 0.6 Gy/day (V.N.) under hypoxic conditions and 0.7 t- 0.2 Gy/day (E.F.) and 0.8 * 0.2 Gy/day (V.N.) under aerobic conditions, corresponding to doubling times of 40-50 hr (E.F.) and 30-40 hr (V.N.) of the surviving clonogenic tumor cells. The median cell cycle time of untreated tumors of most human melanoma xenograft lines is 30-40 hr (23, 25) and this cell cycle time is usually shortened after irradiation (23, 24). The volume-doubling times of the lines studied here were 18.2 t 2.1 days (E.F.) and 5.4 + 0.4 days (V.N.) and the potential doubling times have been calculated to be 80-l 10 hr (E.F.) and 60-85 hr (V.N.)
Repopulation during fractionated irradiation 0 E. K. ROFSTAD
(18). Accordingly, the doubling times of the clonogenic tumor cells during the treatment period were somewhere between the cell cycle times and the potential doubling times. This observation indicates that it may be possible to estimate the rate of repopulation in tumors during fractionated irradiation from the cell cycle time and/or the potential doubling time before treatment (1, 8). Studies of transplantable rodent tumors have also indicated rapid repopulation following radiation treatment. Hermens and Barendsen (11) demonstrated a hundredfold increase in the clonogenic cell fraction in a rat tumor within few days after a large single radiation dose, corresponding to a doubling time of the clonogenic cells as short as the cell cycle time of less than 24 hr. Suit et al. (26) showed a large increase in TCDso with increasing overall treatment time in a mouse tumor given multifraction radiation treatment, corresponding to slightly more than one doubling of the number of clonogenic cells within 24 hr. The most comprehensive studies have been reported by Trott (27) and Trott and Kummermehr (28). They studied five mouse tumor lines that were treated with daily radiation fractions over three weeks, and found that the repopulation rates were not constant throughout the course of treatment, that is, they were different in the first, second, and third week. Moreover, the rate of repopulation as well as the changes with time depended on tumor line. The percentage of the total dose that was used to compensate for the repopulation ranged from 16% to 67%. Corresponding percentages for the two melanoma lines studied here were in the range from 20% to 53%, depending on the radiation dose per fraction. Analyses of clinical data have also given evidence that the repopulation may be rapid during the treatment period and significantly reduce local tumor control rates in protracted regimens (7, 8). Thus, trials using split-course irradiation to spare acute radiation damage have suggested that 2-3 extra fractions of 4-6 Gy are required to compensate for a 2-3 week gap in a 20-30 fraction regimen ( 12, 14, 15). Data allowing more rigid analyses are sparse. However, Trott and Kummermehr (28) have analyzed data from Friedman et al. (10) on local control of Hodgkin’s disease and found that an additional dose of 0.3 Gy/ day was needed to compensate for the repopulation when total treatment time exceeded three weeks. Similar analysis of local control rates in T3-4 squamous cell carcinoma of the larynx has suggested a repopulation rate equivalent to an additional dose of 0.5 Gy/day (13). These doses found for Hodgkin’s disease and squamous cell carcinoma are comparable to, but somewhat lower than, those measured for the melanoma lines studied here.
Tumor radioresponsiveness Previous work involving five human melanoma xenograft lines showed that the lines differed considerably in radioresponsiveness in vivo (2.0 Gy fractions) (22). SF2
61
in vitro also differed considerably among the lines and the radioresponsiveness in vivo showed a clear correlation to SF, in vitro (22). Consequently, the differences in radioresponsiveness among the lines were mainly due to differences in cellular radiation sensitivity and repair capacity. However, the absolute level of radioresponsiveness in vivo could not be explained from SF2 in vitro alone; the tumors were too resistant in vivo (22). The present work revealed that the repopulation between radiation fractions was significant in the tumors. Two lines, one having a short and the other a long volume-doubling time, were studied and the rate of repopulation, corresponding to doses of 0.7-0.8 Gy/day, was found to be approximately equal for the two lines. Consequently, repopulation between radiation fractions strongly reduced the absolute level of radioresponsiveness, but did not influence significantly the differences in radioresponsiveness among the lines. Theoretical values for tumor radioresponsiveness, calculated from SF;?in vitro and rate of repopulation during fractionated irradiation in vivo, agreed fairly well with the absolute level of radioresponsiveness measured after treatment with 2.0 Gy fractions in vivo (Fig. 3). This suggests that the absolute level of radioresponsiveness of the melanoma lines was governed by two main parameters: (a) cellular radiation sensitivity and repair capacity and (b) rate of repopulation between radiation fractions. However, the measured values for tumor radioresponsiveness were in the upper part of or slightly above the shaded areas in Figure 3, representing the theoretical values. This is an indication that also other radiobiological parameters may have given minor contributions to the absolute level of radioresponsiveness. A contribution from tumor hypoxia is one possibility. Although the tumors showed rapid and extensive reoxygenation after treatment with 2.0 Gy fractions, the fractions of hypoxic cells never decreased to levels below those in untreated tumors, that is, 22 f 8% in the E.F. line and 15 + 5% in the V.N. line (20). The human melanoma xenograft lines studied here showed short volume-doubling times compared with most tumors in patients. It is therefore possible that the consequences of repopulation between radiation fractions are less in clinical radiation therapy than predicted by the present experimental study. However, the radiation doses required to compensate for repopulation in the melanoma xenografts were only slightly larger than those estimated from clinical data. It is therefore highly probable that the clinical radioresponsiveness of tumors also is significantly influenced by repopulation, as is the radioresponsiveness of melanoma xenografts. Consequently, clinical investigations of the utilitarian value of possible predictive assays of tumor radioresponsiveness should include assays meaSUkXg cell repopulation rate as well as cellular radiation sensitivity.
68
I. J. Radiation Oncology 0 Biology 0 Physics
Volume 23, Number 1, 1992
REFERENCES 1. Begg, A. C.; Hofland, I.; Moonen, L.; Bartelink, H.; Schraub,
2.
3.
4.
5.
6.
I. 8.
9.
10.
11.
12.
13.
14.
S.; Bontemps, P.; Le Fur, R.; Van Den Bogaert, W.; Caspers, R.; Van Glabbeke, M.; Horiot, J. C. The predictive value of cell kinetic measurements in a European trial of accelerated fractionation in advanced head and neck tumors: an interim report. Int. J. Radiat. Oncol. Biol. Phys. 19: 14491453; 1990. Bristow, R. G.; Hill, R. P. Comparison between in vitro radiosensitivity and in vivo radioresponse in murine tumor cell lines II: in vivo radioresponse following fractionated treatment and in vitro/in vivo correlations. Int. J. Radiat. Oncol. Biol. Phys. 18: 33 l-345; 1990. Brock, W. A.; Baker, F. L.; Wike, J. L.; Sivon, S. L.; Peters, L. J. Cellular radiosensitivity of primary head and neck squamous cell carcinomas and local tumor control. Int. J. Radiat. Oncol. Biol. Phys. 18: 1283-1286; 1990. Courtenay, V. D.; Mills, J. An in vitro colony assay for human tumours grown in immune-suppressed mice and treated in vivo with cytotoxic agents. Br. J. Cancer 37: 26 l268; 1978. Deacon, J.; Peckham, M. J.; Steel, G. G. The radioresponsiveness of human tumours and the initial slope of the cell survival curve. Radiother. Oncol. 2: 3 17-323; 1984. Fertil, B.; Malaise, E. P. Inherent cellular radiosensitivity as a basic concept for human tumor radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 7: 62 l-629; 198 1. Fowler, J. F. Non-standard fractionation in radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 10: 755-759; 1984. Fowler, J. F. Potential for increasing the differential response between tumors and normal tissues: can proliferation rate be used? Int. J. Radiat. Oncol. Biol. Phys. 12: 641-645; 1986. Fowler, J. F. Scientific perspectives on recent progress in radiation oncology. Int. J. Radiat. Oncol. Biol. Phys. 12: 709-712; 1986. Friedman, M.; Pearlman, W. W.; Turgeon, L. Hodgkin’s disease, tumor lethal dose and iso-effect recovery curve. Am. J. Radiol. 99: 843-851; 1967. Hermens, A. F.; Barendsen, G. W. Changes of cell proliferation characteristics in a rat rhabdomyosarcoma before and after X-irradiation. Eur. J. Cancer 5: 173-189; 1989. Holsti, L. R. Clinical experience with split-course radiotherapy. A randomized clinical trial. Radiology 92: 591596; 1969. Maciejewski, B.; Preuss-Bayer, G.; Trott, K. R. The influence of the number of fractions and of overall treatment time on local control and late complication rate in squamous cell carcinoma of the larynx. Int. J. Radiat. Oncol. Biol. Phys. 9: 321-328; 1983. Overgaard, J.; Hjelm-Hansen, M.; Vendelbo Johansen, L.; Andersen, A. P. Comparison of conventional and splitcourse radiotherapy as primary treatment in carcinoma of the larynx. Acta Oncol. 27: 147-152; 1988.
15. Parsons, J.; Bova, F. J.; Million, R. R. A re-evaluation of split-course technique for squamous cell carcinoma of the head and neck. Int. J. Radiat. Oncol. Biol. Phys. 6: 16451652; 1980. 16. Peters, L. J. Inherent radiosensitivity of tumor and normal tissue cells as a predictor of human tumor response. Radiother. Oncol. 17: 177-190; 1990. 17. Rofstad, E. K. Radiation response of the cells of a human malignant melanoma xenograft. Effect of hypoxic cell radiosensitizers. Radiat. Res. 87: 670-683; 198 1. 18. Rofstad, E. K. Growth and vascular structure of human melanoma xenografts. Cell Tissue Kinet. 17: 9 l- 10 1; 1984. 19. Rofstad, E. K. Human tumor xenografts in development of predictive assays of tumor treatment response. In: Chapman, J. D., Peters, L. J., Withers, H. R., eds. Prediction of tumor treatment response. New York: Pergamon Press; 1989: 197-216. 20. Rofstad, E. K. Hypoxia and reoxygenation in human melanoma xenografts. Int. J. Radiat. Oncol. Biol. Phys. 17: 8 l89; 1989. 21. Rofstad, E. K. PLD-repair in human melanoma xenografts following single dose and fractionated irradiation. Br. J. Cancer 6 1: 856-860; 1990. 22. Rofstad, E. K.; Brustad, T. Radioresponsiveness of human melanoma xenografts given fractionated irradiation in vivorelationship to the initial slope of the cell survival curves in vitro. Radiother. Oncol. 9: 45-56; 1987. 23. Rofstad, E. K.; Lindmo, T.; Brustad, T. Effect of single dose irradiation on the proliferation kinetics in a human malignant melanoma in athymic nude mice. Acta Radiol. Oncol. 19: 261-269; 1980. 24. Spang-Thomsen, M.; Vindelev, L. L. Changes of proliferation kinetics after X-irradiation of a human malignant melanoma grown in nude mice. Eur. J. Cancer Clin. Oncol. 20: 849-855; 1984. 25. Spang-Thomsen, M.; Vindelov, L. L. Proliferation kinetics of a human malignant melanoma serially grown in nude mice. Cell Tissue Kinet. 17: 401-410; 1984. 26. Suit, H. D.; Howes, A. E.; Hunter, N. Dependence of response of a C3H mammary carcinoma to fractionated irradiation on fractionation number and intertreatment interval. Radiat. Res. 72: 440-454; 1977. 27. Trott, K. R. Cell proliferation and overall treatment time. Int. J. Radiat. Oncol. Biol. Phys. 19: 1071-1075; 1990. 28. Trot& K. R.; Kummermehr, J. What is known about tumour proliferation rates to choose between accelerated fractionation and hyperfractionation? Radiother. Oncol. 3: 1-9; 1985. 29. West, C. M. L.; Davidson, S. E.; Hunter, R. D. Evaluation of surviving fraction at 2 Gy as a potential prognostic factor for the radiotherapy of carcinoma of the cervix. Int. J. Radiat. Biol. 56: 761-765; 1989.