Copyright
U360-3016/90 $3.00 + .oO ‘ci 1990 Pergamon Press plc
0 Original Contribution
RADIORESPONSE OF HUMAN ASTROCYTIC TUMORS ACROSS GRADE AS A FUNCTION OF ACUTE AND CHRONIC IRRADIATION
CHRISTOPHER
J. SCHULTZ,
M.D. AND CHARLES R. GEARD,
PH.D.
Department of Radiation Oncology, Center for Radiological Research. College of Physicians and Surgeons of Columbia University, New York, NY
Astrocytomas make up the largest group of primary brain tumors of glial origin. Long term survival is rare with high grade tumors (grades 3 and 4), which recur despite subtotal resection, chemotherapy, and aggressive postoperative radiation therapy. In contrast, the 5-year survival for low grade astrocytomas (grades 1 and 2) following subtotal resection and postoperative radiotherapy approaches 50%. Variable sensitivity across grade may contribute to the difference in the behavior of these tumors. To investigate this possibility, the radioresponse of human glial tumors across grade as a function of the dose rate of irradiation was studied. Cell lines derived from a low grade astrocytoma (grade 1) and two high grade astrocytomas (grades 3 and 4) were established in culture. Clonal survival was determined following irradiation of the three cell lines with Cesium 137 gamma rays at high dose rate, 78 Gy/ hr, and at low dose rate, range 14 cCy to 79 cGy/hr. The low grade astrocytoma was found to be more radiosensitive than either of the high grade tumors. The alpha/beta (Gy-‘/Gym*) values (linear quadratic model) were 0.35/0.082 for the grade 1 line and 0.20/0.036 and 0.30/0.045 for the grade 3 and 4, respectively. D0 (cGy) values (single-hit multi-target model) were 99, 144, and 117 for grades 1, 3, and 4, respectively. A dose rate effect was present for all three tumor lines irradiated from 14 cGy/hr to 78 Gy/hr. An inverse dose rate effect was also noted at 37 cGy/ hr for each of the astrocytic lines. These findings may be useful in the development of strategies to treat astrocytic brain tumors which use high and/or low dose rate irradiation. Radiosensitivity, Human glial tumors, Dose rate effects.
INTRODUCTION
(grades 1 and 2) following resection and postoperative radiotherapy is much better than that reported for high grade tumors. The efficacy of postoperative radiotherapy for low grade astrocytomas is supported by several independent retrospective series (1, 3. 6). Five-year survival in these studies for patients receiving postoperative ra-
Astrocytomas constitute the largest group of primary brain tumors of glial origin ( 19). According to the NC1 SEER program, it is estimated that 10,000 new primary brain tumors will be diagnosed in 1989 in the USA. Approximately 50% of these tumors will be of astrocytic origin. Sixty to 80% will be high grade, with the remaining 2040% low grade. Treatment of these astrocytic tumors remains difficult and frustrating. Long term survival is rare with high grade tumors (grades 3 and 4), which recur despite subtotal resection, chemotherapy, and aggressive postoperative radiotherapy (2, 14). Of the available modalities, external beam radiation therapy has been the only intervention that has significantly affected survival, the overall survival time directly correlating with increasing dose delivered (16). Survival for low grade astrocytomas
diotherapy approaches 50%. significantly higher than for patients treated with surgery alone. Several prospective trials are currently testing these findings. Variable radiosensitivity across grade of astrocytoma may contribute to the improved control of low grade astrocytomas treated with postoperative radiation therapy compared to high grade tumors treated in a similar fashion. Several in vitro studies have shown high grade astrocytomas to be radioresistant relative to other human tumor lines when irradiated acutely at high dose rates (4. 11. 17). In contrast,
Presented at the 3 1st Annual Meeting of the American Society for Therapeutic Radiology and Oncology, San Francisco, California, October 1989. Reprint requests to: Christopher J. Schultz, M.D., The Medical College of Wisconsin, Department of Radiation Oncology, 8700 W. Wisconsin Ave., Milwaukee, WI 53226. Acknowledgments-The authors wish to extend their appreci-
ation to Dr. Eric Hall, Dr. Jack Fowler. and Dr. Timothy Kin&la for their assistance in preparing this manuscript. Supported by ASTRO Research Fellowship, American Cancer Society Clinical Oncology Fellowship, NCI-CA 24232 (C.J.S.) and NCI-CA 12536 (C.R.G.) from the U.S. Department of Health and Human Services. Accepted for publication 2 1 June 1990. I397
1398
1. J. Radiation Oncology 0 Biology 0 Physics
little information is available on the in vitro radiosensitivity of low grade astrocytomas irradiated acutely. Recognizing the demonstrated efficacy of external beam radiation therapy in the treatment of high and low grade astrocytic tumors, strategies to escalate tumor dose safely have been explored. This is a difficult problem, as normal brain tolerance is felt to be approximately 60 Gy using standard fractionation. One attractive approach to dose escalation is brachytherapy. Astrocytic tumors are most often unifocal, of limited size, often fairly well circumscribed, and tend to recur locally. These characteristics make them amenable to the use of interstitial brachytherapy. A limited implant volume combined with the properties of low dose rate irradiation may permit localized dose escalation with minimal increases in associated normal tissue toxicity. Little in vitro laboratory information is available on the radiosensitivity of astrocytic tumor lines of high or low grade chronically irradiated at low dose rate. Studies exploring the radiosensitivity of astrocytic tumors across grade, as a function of dose rate, may be pertinent to the future application of external beam and interstitial irradiation in the treatment of these tumors. Such studies have been performed in our laboratory and are reported here.
METHODS
AND
MATERIALS
Origin and maintenance of cell line, growth rate, chromosomal analysis Cell lines derived from a low grade astrocytoma (Kernohan, grade 1). and two high grade astrocytomas (Kernohan, grades 3 and 4) were established in culture from patients undergoing resection at The Neurological Institute of New York, College of Physicians and Surgeons of Columbia University, Columbia Presbyterian Medical Center, New York, NY. The cells were grown in minimum essential medium (Eagle MEM) supplemented with: 10% fetal calf serum containing serum extender,* essential amino acids, nonessential amino acids, L-glutamine, multivitamins, and gentamycin. Stock cultures were grown in unsealed plastic flasks.+ The cultures were split weekly during the course of the experiments so as to maintain an exponentially growing monolayer. The cultures were kept in an environment of 5% CO* and air saturated with water vapor at 37°C. Cell passage levels 10 through 20 were used for experiments with the grade 1 line. For the grade 3 and grade 4 lines, cell passage levels 77 to 87 and 3 1 to 4 1, respectively, were used. Population doubling times were determined by counting* the number of trypsinized cells in a single cell suspension obtained at various times intervals after initial plating. To measure the grade 1 astrocytoma doubling * SerXtend’“. HANA Biologic, Inc. Alameda, CA. + Corning Laboratory Products, Corning, NY. t Coulter Counter, Hialeah, FL.
December 1990, Volume 19. Number 6
times at low density, 3-4 X IO3 cells/cc’ were initially plated in 65 mm dishes.$ For the grade 3 and grade 4 doubling times at low density, 7 X lo2 cells/cc’ were initially plated in 100 mm dishes. For doubling times at high density, 2-4 X IO4 cells/cc’ were initially plated in 35 mm dishes+ for all three 3 astrocytic lines. Chromosome counts were performed on each cell line from chromosome spreads obtained in the standard fashion. The glial properties of the cell lines were conhrmed by light microscopy and GFAP staining.
.4cute high dose rate assay method Exponentially growing cells near confluence were lethally irradiated with a single 30 Gy fraction using 13’Cs gamma rays.** The cells were trypsinized to a single cell suspension, counted, and plated as a feeder layer at a concentration of between 4 and 5 X lo4 cells/dish in 100 mm plastic dishes** containing 10 cc of medium/dish. The feeder layer was allowed to attach overnight. Subsequently, an exponentially growing monolayer of the same cell type was trypsinized to a single cell suspension and counted. Known cell numbers were plated into the dishes containing the feeder layer. Six dishes were plated per dose point. Following cell attachment, the dishes were irradiated acutely in single 2.08 Gy fractions at 1.3 Gy/min using 13’Cs gamma rays to doses ranging from O-12.48 Gy. Three weeks following irradiation, the dishes were fixed with 25% acetic acid in methanol and stained with crystal violet. The number of colonies was determined by counting cell aggregates containing more than 50 cells. A minimum of three experiments was performed for each cell line.
Chronic 10~1dose rate assay method Exponentially dishes containing
growing cells were plated in 35 mm 3-3.5 cc medium/dish at a concentra-
tion of 2-3 X 10’ cells/dish. Approximately 24 hr later, the dishes were placed in a low dose rate 13’Cs incubator/ irradiator capable of delivering the following dose 14 cGy/hr, 26 cGy/hr. 37 cGy/hr. and 79 cGy/hr. dose rates were achieved by placing the dishes to radiated on shelves positioned at different distances
rates: These be irfrom
the r3’Cs source. Four dishes at each dose rate were irradiated. At different time intervals from the initiation of irradiation (range 4 hr to 120 hr) the dishes were removed from the incubator/irradiator so as to deliver four similar total doses at each dose rate (ranges: 208-336 cGy, 592672 cGy, 888- 1264 cGy, 1680- 1896 cGy). Upon removal from the incubator/irradiator, the cells were trypsinized and counted. Known numbers of cells were plated in 100 mm dishes containing a feeder layer of lethally irradiated cells that had previously been prepared as in the high dose d Falcon Plastics, Rutherford, NJ. ** Atomic Energy of Canada Ltd., Gamma Cell 40.
1399
Radioresponse of human glial tumors across grade 0 C. J. SCHULTZ AND C. R. GEARD Astrocytoma Grade 1
rate assay. At each removal time, control plates were likewise removed from a standard incubator and processed in the same fashion as the irradiated cells. Five dishes were plated per dose point. Three weeks following irradiation the dishes were fixed with 25% acetic acid in methanol and stained with crystal violet. The number of colonies was determined by counting cell aggregates containing more than 50 cells. A minimum of three experiments were performed for each cell line.
100 ,
I
70 60 50 40 10
0
30
20
Astrocytoma Grade 3 100 ~
g
90
g
%
80 70
g L
60 50
6
40 Astrocytoma Grade 4 90
80 70 60 50 40 0
10
20
30
40
50
Spread# Fig. 2. Chromosome counts per cell (one chromosome per cell) for the human astrocytic tumor lines.
spread
Data analysis The mean surviving fraction and its standard error were calculated from replicate experiments for each cell line. Survival curves for the acute high dose rate studies were obtained using a linear quadratic model. High dose rate survival parameters were calculated using the Berkeley Cell Survival Parameters Program. Chronic low dose rate survival curves were fitted well using a linear model. RESULTS
Fig. I. Normarski Phase contrast micrographs of human astrocytic tumors. Gr 1 = grade 1 astrocytoma: Gr 3 = grade 3 astrocytoma: Gr 4 = grade 4 astrocytoma.
Cellular morphology. growth rate. chromosomal analysis Normarski phase contrast micrographs of the three cell lines are shown in Figure I. Doubling times for the low grade astrocytoma plated at a density of 3-4 X IO3 cells/cc2 was greater than 100 hr. For both high grade lines plated at a density of 7 X 10’ cells/cc’ the doubling times were 34 hr. At higher density, 2-4 X 1O4cells/cc*, the doubling times were approximately 34 hr for all three cell lines. Chromosomal counts of the three cell lines reveal all the lines to be aneuploid (Fig. 2). The modal chromosome
1400
1. J. Radiation Oncology 0 Biology 0 Physics
December 1990. Volume 19. Number 6
number for the low grade line was 58, whereas it was 62.5 and 72 for the grade 3 and grade 4 lines, respectively.
HDR Survival Across Grade ? ? Grade 1
Acute high dose rate survival The three cell lines were evaluated for their ability to form colonies from single cells following irradiation. Feeder layers of lethally irradiated cells of the line being studied were used in all assays, resulting in the following plating efficiencies: grade 1, 9-16%; grade 3, 46-55%; grade 4, 7-25%. The response of these cell lines to acute high dose rate irradiation are shown in Figure 3. Acute high dose rate survival curve parameters are listed in Table 1 for all 3 cell lines. 0
Chronic low dose rate survival Growth curves for the cell lines irradiated at low dose rate are shown in Figure 4. The response of these cell lines to low dose rate irradiation are given in Figure 5. Acute high dose rate survival is also included in Figure 5 to illustrate further the dose rate effect. DISCUSSION
Maintenance of cell line, growth rate. chromosomal analpis The studies performed on three human astrocytic lines confirm the growth requirements necessary to use a clonogenic assay successfully to determine in vitro radiosensitivity. These requirements are, namely, the use of feeder layers as well as medium supplemented with fetal calf serum (4, 13). The 10% fetal calf serum supplemented with serum extender* used in our studies produced similar if not superior plating efficiencies to those obtained with 30% fetal calf serum in the above reports. The growth rate studies revealed doubling times that are similar to those reported by Gerweck et al. (4). The
2
4
6
8
10
12
14
Dose Gy
Fig. 3. Survival curves for astrocytic tumor lines: Grade I, Grade 3, and Grade 4 irradiated in single 2.08 Gy fractions using Cesium I37 gamma rays at high dose rate, 78 Gy/hr (I .3 Gy/min). Mean surviving fraction for three experiments per cell line. Standard error of the mean is plotted when larger than the data point. Data fit to a linear quadratic model. Note grade I astrocytoma is more radiosensitive than the high grade lines, grades 3 and 4.
low grade doubling time was ~100 hr whereas the high grade lines doubling times were approximately 35 hr. These initial determinations were at low density as described above and are in good agreement with results obtained by Gerweck et al. At higher density, the doubling time for the low grade line decreased to 34 hr. identical to that for the high grade lines. Doubling times for the high grade lines did not change at higher cell density. The initial cell density is not stated in Gerweck’s report so no further comparison can be made. The reason for the decreased doubling time at higher cell density for the low grade line is not clear and was not investigated further.
Table I. Acute HDR survival parameters s2 Cell type
DO ~GY
N
~GY
Alpha Gy-’
DQ
Beta Gy-’
Obs
LQcalc
99 (6.4)
3.3 (0.84)
II8 (19)
0.35 (0.07)
0.082 (0.014)
0.34
0.36
I
Astro Grade 3
144 (4.6)
4.5 (0.65)
217 (15)
0.20 (0.02)
0.036 (0.003)
0.58
0.58
Astro Grade 4
II7 (7.4)
190 (25)
0.30 (0.07)
0.045 (0.01)
0.58
0.46
,:::,
Glioblastoma (Malaise et al.)
144
I2
325
0.24
0.029
0.58
0.56
Astro Grade
(Figures in brackets are standard errors of the mean.) Note: Acute high dose rate survival parameters for Grade I, Grade 3. and Grade 4 astrocytomas from current experiments. Glioblastoma from literature (7). S2 observed: surviving fraction after 2 Gy fraction observed experimentally. S2 LQcalc; surviving fraction after 2 Gy calculated from LQ model. DO, N, Do; terminal slope, extrapolation number, and quasi-threshold dose from single-hit multi-target model. Note the low grade astrocytoma is clearly more sensitive than the higher grade lines.
Radioresponse of human glial tumors acrossgrade 0 C. J. SCHULTZ AND
2e+6
1e+6
I
Oe+O 4
LDR Growth Curve Grade 3 14 cGylhr d-
Oe+O
-
26cGyihr
--t
37cGyihr
--(t
79cGyihr
-
Control
4
2e+6
C. R.
1401
GEARD
it may in part explain the superior clinical radioresponse seen with low grade tumors in comparison to high grade tumors. Further studies of other low grade lines are necessary to test this finding. A range of sensitivities for high grade astrocytomas exists. In general, they fall in the radioresistant range of human tumor lines (4,7.8). The published survival curves have the characteristic broad shoulder felt to be a result of accumulation of sublethal damage. The survival curves obtained for the grade 3 and grade 4 lines in our studies demonstrate this characteristic shape and are within the range of radiosensitivities of high grade glial tumors reported elsewhere (Fig. 6) (4. 1 I. 17). There appears to be no direct correlation between grade of tumor and radiosensitivity within the high grade tumors. that is. grades 3 and 4. In the current study. the grade 4 line was found to be somewhat more sensitive than the grade 3 line. Similar results have been reported by Gerweck t’f al. (4). The AZ, grade 4 line in Gerweck’s report is also much more sen-
LDR Growth Curve Grade 4 10
Grade
1
1 e+6
* Oe+O
/l-
0
24
48
72
96
120
144
Time (hrs) Fig. 4. Growth curves for astrocytic tumor lines irradiated with Cesium 137 gamma rays at low dose rate: range 14 cGy/hr to 79 cGy/hr. Data is from one of three experiments per cell line. Proliferation continues below 26 cGy/hr.
Studies investigating the density dependent growth control of human malignant glioma cells have been reported elsewhere ( 18). All three lines were found to be aneuploid. The association between ploidy and radiosensitivity remains controversial (12). The modal chromosome number for the grade 1. grade 3, and grade 4 lines was 58, 62.5, and 72. respectively. The order of radiosensitivity from most radiosensitive to least radiosensitive in these experiments is grade 1. grade 4. and grade 3. No clear trend between ploidy and radiosensitivity was apparent in our studies.
10
Grade 4
0001
The survival curves for the three astrocytic human tumor lines evaluated in this study following acute high dose rate irradiation are given in Figure 3. These survival curves demonstrate that the low grade astrocytoma (grade 1) is more radiosensitive than either of the high grade lines (grades 3 and 4). The low grade line also appears to be more sensitive than other high grade lines reported in the literature Figure 6 (4, 11, 17). If this finding is true,
0
2
4
6
8
Dose
10
12
14
16
18
20
Gy
Fig. 5. Survival curves for astrocytic tumor lines irradiated with Cesium 137 gamma rays at high dose rate (78 Gy/hr) and low dose rate (14 cGy/hr to 79 cGy/hr). Mean surviving fraction for three experiments per cell line. Standard error of the mean is plotted when larger than the data point. High dose rate data fit to a linear quadratic model. Low dose rate data fit to a linear model. Note dose rate effect and inverse dose rate effect.
1402
I. J. Radiation Oncology 0 Biology
0 Physics
Acute HDR Survival Astrocytoma 1
??
Weichselbaum
-4 0
4
8
12
16
20
24
Dose Gy Fig. 6. Survival curves for human astrocytic tumors across grade of tumor irradiated in single fractions at high dose rate. Grade
1, Grade 3, Grade 4 curves from current study. A2 (Grade 4) A3 (Grade 4), A7 (Grade 3). Weichselbaum (Glioblastoma). Nilsson (Grade 3/4). Surviving fraction for experimental data and data from the literature fit to a linear quadratic model (4, II, 17).
sitive than the A7, grade 3 cell line. However, the second grade 4 line, A3, was not different from the A7 grade 3 cell line. Variable radiosensitivity exists for the high grade tumors reported in our series and those thus far reported by others. Acute HDR survival parameters for the three astrocytic tumor lines studied are listed in Table 1. Note that S2 (observed), the surviving fraction following 2 Gy, for our grade 3 and 4 lines as well as the S2 reported for glioblastoma multiforme by Malaise et al. are identical, all three being relatively resistant (7). The S2 (observed) for the grade 1 astrocytoma is significantly lower than that for the high grade lines, again demonstrating that the low grade tumor is more radiosensitive. It is the S2 parameter together with the mean inactivation dose and alpha term from the LQ model that is reported to correlate best with clinical radioresponsiveness (7). Again, perhaps these findings in part explain the superior results of irradiation for low grade astrocytomas.
December 1990, Volume
19. Number
6
rate effect, that is, the increase in surviving fraction per unit dose as the dose rate decreases, results from the repair of sublethal damage as the dose rate is decreased. at least until the “inverse dose rate effect” appears (see below). Proliferation also contributes to the dose rate effect below a critical dose rate where all sublethal damage is repaired. These effects are illustrated by a decreasing slope and loss of, or reduction in size of, the shoulder on the low dose rate survival curves. The only large difference in magnitude of dose rate effect was at 14 cGy/hr where proliferation is the major factor affecting response (Fig. 7). There is a dramatic difference in the efficacy of irradiation with increasing dose rate for the grade 4 line. This difference is less pronounced although still present for the grade 3 and grade 1 lines. This different response to chronic low dose rate irradiation reflects intrinsic cellular differences that are not completely predictable from the high dose rate single fraction studies. Predictions based on high dose rate single fraction experiments do not take into account the dynamic interrelationship of repair, proliferation, and redistribution. For example the grade 3 line has the broadest shoulder and is the least radioresponsive in the high dose rate studies, yet the grade 4 line has the widest response to low dose rate irradiation and is the least sensitive at the lowest dose rate, 14 cGy/hr. Identification of a characteristic response to low dose rate irradiation for individual tumors or even for a group of similar types of tumors may be useful in planning a course of fractionated high dose rate irradiation or low dose rate brachytherapy. Studies exploring the dose rate effect for a variety of different human tumor lines have recently been reported ( 15). In addition to the dose rate effect, an inverse dose rate effect is present for all three of the astrocytic lines studied. This improved efficacy of irradiation, within a limited range of decreasing dose rate, results from redistribution of cells into, and subsequent delay in, GzM, a more sen-
1% Survival Across Dose Rate
Chronic low dose rate survival The chronic low dose rate studies demonstrate that cellular proliferation continues at dose rates from 14 cGy/ hr to 26 cGy/hr up to total doses of 1680 cGy and 1872 cGy, respectively (Fig. 4). At the higher dose rates, proliferation is initially apparent; however, the rate of proliferation decreases with increasing dose rate. Cell cycle times were not specifically determined; therefore, no comment can be made regarding radioresponse in relation to “dose per cell cycle”. The low dose rate survival curves shown in Figure 5 demonstrate that a dose rate effect exists from 14 cGy/hr to 78 Gy/hr for all three of the cell lines studied. The dose
Dose Rate cGy/min Fig. 7. Dose to achieve 1% surviving fraction across dose rate for the human astrocytic tumor lines. Mean surviving fraction for three experiments per cell line. Note wide difference in dose rate effect in general and inverse dose rate effect at 37 cGy/hr.
Radioresponse of human glial tumors across grade 0 C. J. SCHULTZ AND C. R. GEARD
sitive phase in the cell cycle (9, 10). In our lines this inverse dose rate effect occurs at 37 cGy/hr. At dose rates less than this, cellular proliferation continues, whereas at rates greater than 37 cGy/hr, cells are slowed or stop proliferating completely. At 37 cGy/hr, cells are able to cycle but at a slower rate, so that they spend more time in G2M or are actually blocked in this sensitive phase of the cell cycle. Coincidentally, a similar inverse dose rate effect has been described at 37 cGy/hr for S3 HeLa cells chronically irradiated ( 10). Others have not seen significant dose rate effects nor inverse dose rate effects with the human and rodent cell lines they tested (5). Note that many of these studies did not include dose rates below 60 cGy/hr so that dose rate and inverse dose rate effects that exist below this dose rate could have been missed. Furthermore. in many of the experiments, plateau or “reduced growth phase” cells were studied which may not allow for the expression of survival differences which depend on cell cycle effects. The dose rate and inverse dose rate effects seen in our studies for the astrocytic lines tested may become important if normal brain is found to have a different response to similar dose rates. Wide differences in responses to low dose rate irradiation among different mammalian cell lines and human tumor lines have been reported (9, 15).
1403
The low dose rate studies underscore the importance of repair, redistribution, and repopulation which occur in tissues responding to chronic low dose rate irradiation. The magnitude and significance of these responses varies considerably over a relatively small range of dose rates for the three astrocytic tumor lines studied. CONCLUSION A difference in radiosensitivity exists between the high and low grade astrocytoma cell lines studied, the low grade line being more sensitive than the high grade lines. Cellular proliferation continues at dose rates less than 26 cGy/hr. A dose rate effect exists for exponentially growing cells irradiated at dose rates from 14 cGy/hr to 78 Gy/hr. An inverse dose rate effect exists at 37 cGy/hr due to differences in progression through the cell cycle. An effective range of dose rates has been identified for these astrocytic cell lines, 37 cGy/hr to 79 cGy/hr, which may be applicable to clinical applications of low dose rate irradiation. These findings may prove useful in developing future dose escalation strategies in the treatment of low and high grade astrocytic brain tumors.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
9.
Bouchard, J. Effects of irradiation in treatment of intercranial gliomas-treatment results by histologic groups (Chapter 5). In: Bouchard. J., ed. Radiation therapy of tumors and diseases of the nervous system. Philadelphia, PA: Lea and Febiger; 1966:78-I 18. Davis, L. W. Presidential address: malignant glioma-A nemesis which requires clinical and basic investigation in radiation oncology. Int. J. Radiat. Oncol. Biol. Phys. 16: 13551365; 1989. Garcia, D. M.: Fulling, K. H.: Marks, J. E. The value of radiation therapy in addition to surgery for astrocytomas of the adult cerebrum. Cancer 55:919-927; 1985. Gerweck, L. E.; Kornblith, P. L.: Burlett, P.; Wang, J.; Sweigert, S. Radiation sensitivity of cultured human glioblastoma cells. Radiology 125:23 l-234; 1977. Hall. E. J.; Marchese. M. J.; Astor, M. B.: Morse, T. Response of cells of human origin, normal and malignant, to acute and low dose rate irradiation. Int. J. Radiat. Oncol. Biol. Phys. l2:655-659; 1986. Laws, E. R.; Taylor, W. F.: Clifton, M. B.: Okazaki. H. Neurosurgical management of low grade astrocytomas of the cerebral hemispheres. J. Neurosurg. 6 1:665-673; 1984. Malaise. E. P.; Fertil, B.: Chavaudra, N.; Guichard, M. Distribution of radiation sensitivities for human tumor cells of specific histological types: comparison of in vitro and in vivo data. Int. J. Radiat. Oncol. Biol. Phys. 12:617-624; 1986. Malaise. E. P.; Fertil, B.; Deschavanne, P. J.: Chavaudra, N.; Brock. W. A. Initial slope of radiation survival curves is characteristic of the origin of primary and established cultures of human tumor cells and fibroblasts. Rad. Res. 111:319-333; 1987. Mitchell, J. B.; Bedford, J. S.: Bailey, S. M. Dose rate effects in mammalian cells in culture. III. Comparison of cell killing
and cell proliferation during continuous irradiation for six different cell lines. Rad. Res. 79537-55 I; 1979. 10. Mitchell, J. B.; Bedford, J. S.; Bailey, S. M. Dose-rate effects on the cell cycle and survival of S3 HeLa and V79 cells. Rad. Res. 79:520-536; 1979. I I. Nilsson. S.; Carlsson, J.: Larrson, B. Survival of irradiated glia and glioma cells studied with a new cloning technique. Int. J. Radiat. Biol. 37:267-279: 1980. 12. Radford. I. R.; Hodgson, S. H. Effect of ploidy on the response of V79 cells to ionizing radiation. Int. J. Radiat. Biol. 5 1:765-778; 1986. 13. Rosenblum, M. L.; Vasquez, D. A.; Hoshino, T.; Wilson. C. B. Development a clonogenic cell assay for human brain tumors. Cancer 41:2305-2314: 1978. 14. Shapiro, W. R. Therapy of adult malignant brain tumors. What have the clinical trials taught us? Sem. Oncol. 13:3845; 1986. 15. Steel, G. G.: Deacon, J. M.; Duchesne, G. M.; Horwich. A.; Kelland, L. R.; Peacock, J. H. The dose rate effect in human tumour cells. Radiother. Oncol. 9:299-310: 1987. 16. Walker. M. D.: Strike, T. A.; Sheline, G. A. An analysis of dose-effect relationship in the radiotherapy of malignant brain tumors. Int. J. Radiat. Oncol. Biol. Phys. 5:17251731; 1979. 17. Weichselbaum, R. R.: Epstein, J.; Little, J. B.; Kornblith. P. L. In vitro cellular radiosensitivity of human malignant tumors. Eur. J. Cancer 12:47-5 1: 1976. 18. Westermark. B. The deficient density-dependent growth control of human malignant glioma cells and virus-transformed glia-like cells in culture. Int. J. Cancer 12:438-45 1; 1973. 19. Zulch, K. J. Brain tumors. Their biology and pathology, 3rd edition. Berlin-Heidelberg: Springer-Verlag: 1986.
U360-3016/90 $3.00 + .oO ‘ci 1990 Pergamon Press plc
0 Original Contribution
RADIORESPONSE OF HUMAN ASTROCYTIC TUMORS ACROSS GRADE AS A FUNCTION OF ACUTE AND CHRONIC IRRADIATION
CHRISTOPHER
J. SCHULTZ,
M.D. AND CHARLES R. GEARD,
PH.D.
Department of Radiation Oncology, Center for Radiological Research. College of Physicians and Surgeons of Columbia University, New York, NY
Astrocytomas make up the largest group of primary brain tumors of glial origin. Long term survival is rare with high grade tumors (grades 3 and 4), which recur despite subtotal resection, chemotherapy, and aggressive postoperative radiation therapy. In contrast, the 5-year survival for low grade astrocytomas (grades 1 and 2) following subtotal resection and postoperative radiotherapy approaches 50%. Variable sensitivity across grade may contribute to the difference in the behavior of these tumors. To investigate this possibility, the radioresponse of human glial tumors across grade as a function of the dose rate of irradiation was studied. Cell lines derived from a low grade astrocytoma (grade 1) and two high grade astrocytomas (grades 3 and 4) were established in culture. Clonal survival was determined following irradiation of the three cell lines with Cesium 137 gamma rays at high dose rate, 78 Gy/ hr, and at low dose rate, range 14 cCy to 79 cGy/hr. The low grade astrocytoma was found to be more radiosensitive than either of the high grade tumors. The alpha/beta (Gy-‘/Gym*) values (linear quadratic model) were 0.35/0.082 for the grade 1 line and 0.20/0.036 and 0.30/0.045 for the grade 3 and 4, respectively. D0 (cGy) values (single-hit multi-target model) were 99, 144, and 117 for grades 1, 3, and 4, respectively. A dose rate effect was present for all three tumor lines irradiated from 14 cGy/hr to 78 Gy/hr. An inverse dose rate effect was also noted at 37 cGy/ hr for each of the astrocytic lines. These findings may be useful in the development of strategies to treat astrocytic brain tumors which use high and/or low dose rate irradiation. Radiosensitivity, Human glial tumors, Dose rate effects.
INTRODUCTION
(grades 1 and 2) following resection and postoperative radiotherapy is much better than that reported for high grade tumors. The efficacy of postoperative radiotherapy for low grade astrocytomas is supported by several independent retrospective series (1, 3. 6). Five-year survival in these studies for patients receiving postoperative ra-
Astrocytomas constitute the largest group of primary brain tumors of glial origin ( 19). According to the NC1 SEER program, it is estimated that 10,000 new primary brain tumors will be diagnosed in 1989 in the USA. Approximately 50% of these tumors will be of astrocytic origin. Sixty to 80% will be high grade, with the remaining 2040% low grade. Treatment of these astrocytic tumors remains difficult and frustrating. Long term survival is rare with high grade tumors (grades 3 and 4), which recur despite subtotal resection, chemotherapy, and aggressive postoperative radiotherapy (2, 14). Of the available modalities, external beam radiation therapy has been the only intervention that has significantly affected survival, the overall survival time directly correlating with increasing dose delivered (16). Survival for low grade astrocytomas
diotherapy approaches 50%. significantly higher than for patients treated with surgery alone. Several prospective trials are currently testing these findings. Variable radiosensitivity across grade of astrocytoma may contribute to the improved control of low grade astrocytomas treated with postoperative radiation therapy compared to high grade tumors treated in a similar fashion. Several in vitro studies have shown high grade astrocytomas to be radioresistant relative to other human tumor lines when irradiated acutely at high dose rates (4. 11. 17). In contrast,
Presented at the 3 1st Annual Meeting of the American Society for Therapeutic Radiology and Oncology, San Francisco, California, October 1989. Reprint requests to: Christopher J. Schultz, M.D., The Medical College of Wisconsin, Department of Radiation Oncology, 8700 W. Wisconsin Ave., Milwaukee, WI 53226. Acknowledgments-The authors wish to extend their appreci-
ation to Dr. Eric Hall, Dr. Jack Fowler. and Dr. Timothy Kin&la for their assistance in preparing this manuscript. Supported by ASTRO Research Fellowship, American Cancer Society Clinical Oncology Fellowship, NCI-CA 24232 (C.J.S.) and NCI-CA 12536 (C.R.G.) from the U.S. Department of Health and Human Services. Accepted for publication 2 1 June 1990. I397
1398
1. J. Radiation Oncology 0 Biology 0 Physics
little information is available on the in vitro radiosensitivity of low grade astrocytomas irradiated acutely. Recognizing the demonstrated efficacy of external beam radiation therapy in the treatment of high and low grade astrocytic tumors, strategies to escalate tumor dose safely have been explored. This is a difficult problem, as normal brain tolerance is felt to be approximately 60 Gy using standard fractionation. One attractive approach to dose escalation is brachytherapy. Astrocytic tumors are most often unifocal, of limited size, often fairly well circumscribed, and tend to recur locally. These characteristics make them amenable to the use of interstitial brachytherapy. A limited implant volume combined with the properties of low dose rate irradiation may permit localized dose escalation with minimal increases in associated normal tissue toxicity. Little in vitro laboratory information is available on the radiosensitivity of astrocytic tumor lines of high or low grade chronically irradiated at low dose rate. Studies exploring the radiosensitivity of astrocytic tumors across grade, as a function of dose rate, may be pertinent to the future application of external beam and interstitial irradiation in the treatment of these tumors. Such studies have been performed in our laboratory and are reported here.
METHODS
AND
MATERIALS
Origin and maintenance of cell line, growth rate, chromosomal analysis Cell lines derived from a low grade astrocytoma (Kernohan, grade 1). and two high grade astrocytomas (Kernohan, grades 3 and 4) were established in culture from patients undergoing resection at The Neurological Institute of New York, College of Physicians and Surgeons of Columbia University, Columbia Presbyterian Medical Center, New York, NY. The cells were grown in minimum essential medium (Eagle MEM) supplemented with: 10% fetal calf serum containing serum extender,* essential amino acids, nonessential amino acids, L-glutamine, multivitamins, and gentamycin. Stock cultures were grown in unsealed plastic flasks.+ The cultures were split weekly during the course of the experiments so as to maintain an exponentially growing monolayer. The cultures were kept in an environment of 5% CO* and air saturated with water vapor at 37°C. Cell passage levels 10 through 20 were used for experiments with the grade 1 line. For the grade 3 and grade 4 lines, cell passage levels 77 to 87 and 3 1 to 4 1, respectively, were used. Population doubling times were determined by counting* the number of trypsinized cells in a single cell suspension obtained at various times intervals after initial plating. To measure the grade 1 astrocytoma doubling * SerXtend’“. HANA Biologic, Inc. Alameda, CA. + Corning Laboratory Products, Corning, NY. t Coulter Counter, Hialeah, FL.
December 1990, Volume 19. Number 6
times at low density, 3-4 X IO3 cells/cc’ were initially plated in 65 mm dishes.$ For the grade 3 and grade 4 doubling times at low density, 7 X lo2 cells/cc’ were initially plated in 100 mm dishes. For doubling times at high density, 2-4 X IO4 cells/cc’ were initially plated in 35 mm dishes+ for all three 3 astrocytic lines. Chromosome counts were performed on each cell line from chromosome spreads obtained in the standard fashion. The glial properties of the cell lines were conhrmed by light microscopy and GFAP staining.
.4cute high dose rate assay method Exponentially growing cells near confluence were lethally irradiated with a single 30 Gy fraction using 13’Cs gamma rays.** The cells were trypsinized to a single cell suspension, counted, and plated as a feeder layer at a concentration of between 4 and 5 X lo4 cells/dish in 100 mm plastic dishes** containing 10 cc of medium/dish. The feeder layer was allowed to attach overnight. Subsequently, an exponentially growing monolayer of the same cell type was trypsinized to a single cell suspension and counted. Known cell numbers were plated into the dishes containing the feeder layer. Six dishes were plated per dose point. Following cell attachment, the dishes were irradiated acutely in single 2.08 Gy fractions at 1.3 Gy/min using 13’Cs gamma rays to doses ranging from O-12.48 Gy. Three weeks following irradiation, the dishes were fixed with 25% acetic acid in methanol and stained with crystal violet. The number of colonies was determined by counting cell aggregates containing more than 50 cells. A minimum of three experiments was performed for each cell line.
Chronic 10~1dose rate assay method Exponentially dishes containing
growing cells were plated in 35 mm 3-3.5 cc medium/dish at a concentra-
tion of 2-3 X 10’ cells/dish. Approximately 24 hr later, the dishes were placed in a low dose rate 13’Cs incubator/ irradiator capable of delivering the following dose 14 cGy/hr, 26 cGy/hr. 37 cGy/hr. and 79 cGy/hr. dose rates were achieved by placing the dishes to radiated on shelves positioned at different distances
rates: These be irfrom
the r3’Cs source. Four dishes at each dose rate were irradiated. At different time intervals from the initiation of irradiation (range 4 hr to 120 hr) the dishes were removed from the incubator/irradiator so as to deliver four similar total doses at each dose rate (ranges: 208-336 cGy, 592672 cGy, 888- 1264 cGy, 1680- 1896 cGy). Upon removal from the incubator/irradiator, the cells were trypsinized and counted. Known numbers of cells were plated in 100 mm dishes containing a feeder layer of lethally irradiated cells that had previously been prepared as in the high dose d Falcon Plastics, Rutherford, NJ. ** Atomic Energy of Canada Ltd., Gamma Cell 40.
1399
Radioresponse of human glial tumors across grade 0 C. J. SCHULTZ AND C. R. GEARD Astrocytoma Grade 1
rate assay. At each removal time, control plates were likewise removed from a standard incubator and processed in the same fashion as the irradiated cells. Five dishes were plated per dose point. Three weeks following irradiation the dishes were fixed with 25% acetic acid in methanol and stained with crystal violet. The number of colonies was determined by counting cell aggregates containing more than 50 cells. A minimum of three experiments were performed for each cell line.
100 ,
I
70 60 50 40 10
0
30
20
Astrocytoma Grade 3 100 ~
g
90
g
%
80 70
g L
60 50
6
40 Astrocytoma Grade 4 90
80 70 60 50 40 0
10
20
30
40
50
Spread# Fig. 2. Chromosome counts per cell (one chromosome per cell) for the human astrocytic tumor lines.
spread
Data analysis The mean surviving fraction and its standard error were calculated from replicate experiments for each cell line. Survival curves for the acute high dose rate studies were obtained using a linear quadratic model. High dose rate survival parameters were calculated using the Berkeley Cell Survival Parameters Program. Chronic low dose rate survival curves were fitted well using a linear model. RESULTS
Fig. I. Normarski Phase contrast micrographs of human astrocytic tumors. Gr 1 = grade 1 astrocytoma: Gr 3 = grade 3 astrocytoma: Gr 4 = grade 4 astrocytoma.
Cellular morphology. growth rate. chromosomal analysis Normarski phase contrast micrographs of the three cell lines are shown in Figure I. Doubling times for the low grade astrocytoma plated at a density of 3-4 X IO3 cells/cc2 was greater than 100 hr. For both high grade lines plated at a density of 7 X 10’ cells/cc’ the doubling times were 34 hr. At higher density, 2-4 X 1O4cells/cc*, the doubling times were approximately 34 hr for all three cell lines. Chromosomal counts of the three cell lines reveal all the lines to be aneuploid (Fig. 2). The modal chromosome
1400
1. J. Radiation Oncology 0 Biology 0 Physics
December 1990. Volume 19. Number 6
number for the low grade line was 58, whereas it was 62.5 and 72 for the grade 3 and grade 4 lines, respectively.
HDR Survival Across Grade ? ? Grade 1
Acute high dose rate survival The three cell lines were evaluated for their ability to form colonies from single cells following irradiation. Feeder layers of lethally irradiated cells of the line being studied were used in all assays, resulting in the following plating efficiencies: grade 1, 9-16%; grade 3, 46-55%; grade 4, 7-25%. The response of these cell lines to acute high dose rate irradiation are shown in Figure 3. Acute high dose rate survival curve parameters are listed in Table 1 for all 3 cell lines. 0
Chronic low dose rate survival Growth curves for the cell lines irradiated at low dose rate are shown in Figure 4. The response of these cell lines to low dose rate irradiation are given in Figure 5. Acute high dose rate survival is also included in Figure 5 to illustrate further the dose rate effect. DISCUSSION
Maintenance of cell line, growth rate. chromosomal analpis The studies performed on three human astrocytic lines confirm the growth requirements necessary to use a clonogenic assay successfully to determine in vitro radiosensitivity. These requirements are, namely, the use of feeder layers as well as medium supplemented with fetal calf serum (4, 13). The 10% fetal calf serum supplemented with serum extender* used in our studies produced similar if not superior plating efficiencies to those obtained with 30% fetal calf serum in the above reports. The growth rate studies revealed doubling times that are similar to those reported by Gerweck et al. (4). The
2
4
6
8
10
12
14
Dose Gy
Fig. 3. Survival curves for astrocytic tumor lines: Grade I, Grade 3, and Grade 4 irradiated in single 2.08 Gy fractions using Cesium I37 gamma rays at high dose rate, 78 Gy/hr (I .3 Gy/min). Mean surviving fraction for three experiments per cell line. Standard error of the mean is plotted when larger than the data point. Data fit to a linear quadratic model. Note grade I astrocytoma is more radiosensitive than the high grade lines, grades 3 and 4.
low grade doubling time was ~100 hr whereas the high grade lines doubling times were approximately 35 hr. These initial determinations were at low density as described above and are in good agreement with results obtained by Gerweck et al. At higher density, the doubling time for the low grade line decreased to 34 hr. identical to that for the high grade lines. Doubling times for the high grade lines did not change at higher cell density. The initial cell density is not stated in Gerweck’s report so no further comparison can be made. The reason for the decreased doubling time at higher cell density for the low grade line is not clear and was not investigated further.
Table I. Acute HDR survival parameters s2 Cell type
DO ~GY
N
~GY
Alpha Gy-’
DQ
Beta Gy-’
Obs
LQcalc
99 (6.4)
3.3 (0.84)
II8 (19)
0.35 (0.07)
0.082 (0.014)
0.34
0.36
I
Astro Grade 3
144 (4.6)
4.5 (0.65)
217 (15)
0.20 (0.02)
0.036 (0.003)
0.58
0.58
Astro Grade 4
II7 (7.4)
190 (25)
0.30 (0.07)
0.045 (0.01)
0.58
0.46
,:::,
Glioblastoma (Malaise et al.)
144
I2
325
0.24
0.029
0.58
0.56
Astro Grade
(Figures in brackets are standard errors of the mean.) Note: Acute high dose rate survival parameters for Grade I, Grade 3. and Grade 4 astrocytomas from current experiments. Glioblastoma from literature (7). S2 observed: surviving fraction after 2 Gy fraction observed experimentally. S2 LQcalc; surviving fraction after 2 Gy calculated from LQ model. DO, N, Do; terminal slope, extrapolation number, and quasi-threshold dose from single-hit multi-target model. Note the low grade astrocytoma is clearly more sensitive than the higher grade lines.
Radioresponse of human glial tumors acrossgrade 0 C. J. SCHULTZ AND
2e+6
1e+6
I
Oe+O 4
LDR Growth Curve Grade 3 14 cGylhr d-
Oe+O
-
26cGyihr
--t
37cGyihr
--(t
79cGyihr
-
Control
4
2e+6
C. R.
1401
GEARD
it may in part explain the superior clinical radioresponse seen with low grade tumors in comparison to high grade tumors. Further studies of other low grade lines are necessary to test this finding. A range of sensitivities for high grade astrocytomas exists. In general, they fall in the radioresistant range of human tumor lines (4,7.8). The published survival curves have the characteristic broad shoulder felt to be a result of accumulation of sublethal damage. The survival curves obtained for the grade 3 and grade 4 lines in our studies demonstrate this characteristic shape and are within the range of radiosensitivities of high grade glial tumors reported elsewhere (Fig. 6) (4. 1 I. 17). There appears to be no direct correlation between grade of tumor and radiosensitivity within the high grade tumors. that is. grades 3 and 4. In the current study. the grade 4 line was found to be somewhat more sensitive than the grade 3 line. Similar results have been reported by Gerweck t’f al. (4). The AZ, grade 4 line in Gerweck’s report is also much more sen-
LDR Growth Curve Grade 4 10
Grade
1
1 e+6
* Oe+O
/l-
0
24
48
72
96
120
144
Time (hrs) Fig. 4. Growth curves for astrocytic tumor lines irradiated with Cesium 137 gamma rays at low dose rate: range 14 cGy/hr to 79 cGy/hr. Data is from one of three experiments per cell line. Proliferation continues below 26 cGy/hr.
Studies investigating the density dependent growth control of human malignant glioma cells have been reported elsewhere ( 18). All three lines were found to be aneuploid. The association between ploidy and radiosensitivity remains controversial (12). The modal chromosome number for the grade 1. grade 3, and grade 4 lines was 58, 62.5, and 72. respectively. The order of radiosensitivity from most radiosensitive to least radiosensitive in these experiments is grade 1. grade 4. and grade 3. No clear trend between ploidy and radiosensitivity was apparent in our studies.
10
Grade 4
0001
The survival curves for the three astrocytic human tumor lines evaluated in this study following acute high dose rate irradiation are given in Figure 3. These survival curves demonstrate that the low grade astrocytoma (grade 1) is more radiosensitive than either of the high grade lines (grades 3 and 4). The low grade line also appears to be more sensitive than other high grade lines reported in the literature Figure 6 (4, 11, 17). If this finding is true,
0
2
4
6
8
Dose
10
12
14
16
18
20
Gy
Fig. 5. Survival curves for astrocytic tumor lines irradiated with Cesium 137 gamma rays at high dose rate (78 Gy/hr) and low dose rate (14 cGy/hr to 79 cGy/hr). Mean surviving fraction for three experiments per cell line. Standard error of the mean is plotted when larger than the data point. High dose rate data fit to a linear quadratic model. Low dose rate data fit to a linear model. Note dose rate effect and inverse dose rate effect.
1402
I. J. Radiation Oncology 0 Biology
0 Physics
Acute HDR Survival Astrocytoma 1
??
Weichselbaum
-4 0
4
8
12
16
20
24
Dose Gy Fig. 6. Survival curves for human astrocytic tumors across grade of tumor irradiated in single fractions at high dose rate. Grade
1, Grade 3, Grade 4 curves from current study. A2 (Grade 4) A3 (Grade 4), A7 (Grade 3). Weichselbaum (Glioblastoma). Nilsson (Grade 3/4). Surviving fraction for experimental data and data from the literature fit to a linear quadratic model (4, II, 17).
sitive than the A7, grade 3 cell line. However, the second grade 4 line, A3, was not different from the A7 grade 3 cell line. Variable radiosensitivity exists for the high grade tumors reported in our series and those thus far reported by others. Acute HDR survival parameters for the three astrocytic tumor lines studied are listed in Table 1. Note that S2 (observed), the surviving fraction following 2 Gy, for our grade 3 and 4 lines as well as the S2 reported for glioblastoma multiforme by Malaise et al. are identical, all three being relatively resistant (7). The S2 (observed) for the grade 1 astrocytoma is significantly lower than that for the high grade lines, again demonstrating that the low grade tumor is more radiosensitive. It is the S2 parameter together with the mean inactivation dose and alpha term from the LQ model that is reported to correlate best with clinical radioresponsiveness (7). Again, perhaps these findings in part explain the superior results of irradiation for low grade astrocytomas.
December 1990, Volume
19. Number
6
rate effect, that is, the increase in surviving fraction per unit dose as the dose rate decreases, results from the repair of sublethal damage as the dose rate is decreased. at least until the “inverse dose rate effect” appears (see below). Proliferation also contributes to the dose rate effect below a critical dose rate where all sublethal damage is repaired. These effects are illustrated by a decreasing slope and loss of, or reduction in size of, the shoulder on the low dose rate survival curves. The only large difference in magnitude of dose rate effect was at 14 cGy/hr where proliferation is the major factor affecting response (Fig. 7). There is a dramatic difference in the efficacy of irradiation with increasing dose rate for the grade 4 line. This difference is less pronounced although still present for the grade 3 and grade 1 lines. This different response to chronic low dose rate irradiation reflects intrinsic cellular differences that are not completely predictable from the high dose rate single fraction studies. Predictions based on high dose rate single fraction experiments do not take into account the dynamic interrelationship of repair, proliferation, and redistribution. For example the grade 3 line has the broadest shoulder and is the least radioresponsive in the high dose rate studies, yet the grade 4 line has the widest response to low dose rate irradiation and is the least sensitive at the lowest dose rate, 14 cGy/hr. Identification of a characteristic response to low dose rate irradiation for individual tumors or even for a group of similar types of tumors may be useful in planning a course of fractionated high dose rate irradiation or low dose rate brachytherapy. Studies exploring the dose rate effect for a variety of different human tumor lines have recently been reported ( 15). In addition to the dose rate effect, an inverse dose rate effect is present for all three of the astrocytic lines studied. This improved efficacy of irradiation, within a limited range of decreasing dose rate, results from redistribution of cells into, and subsequent delay in, GzM, a more sen-
1% Survival Across Dose Rate
Chronic low dose rate survival The chronic low dose rate studies demonstrate that cellular proliferation continues at dose rates from 14 cGy/ hr to 26 cGy/hr up to total doses of 1680 cGy and 1872 cGy, respectively (Fig. 4). At the higher dose rates, proliferation is initially apparent; however, the rate of proliferation decreases with increasing dose rate. Cell cycle times were not specifically determined; therefore, no comment can be made regarding radioresponse in relation to “dose per cell cycle”. The low dose rate survival curves shown in Figure 5 demonstrate that a dose rate effect exists from 14 cGy/hr to 78 Gy/hr for all three of the cell lines studied. The dose
Dose Rate cGy/min Fig. 7. Dose to achieve 1% surviving fraction across dose rate for the human astrocytic tumor lines. Mean surviving fraction for three experiments per cell line. Note wide difference in dose rate effect in general and inverse dose rate effect at 37 cGy/hr.
Radioresponse of human glial tumors across grade 0 C. J. SCHULTZ AND C. R. GEARD
sitive phase in the cell cycle (9, 10). In our lines this inverse dose rate effect occurs at 37 cGy/hr. At dose rates less than this, cellular proliferation continues, whereas at rates greater than 37 cGy/hr, cells are slowed or stop proliferating completely. At 37 cGy/hr, cells are able to cycle but at a slower rate, so that they spend more time in G2M or are actually blocked in this sensitive phase of the cell cycle. Coincidentally, a similar inverse dose rate effect has been described at 37 cGy/hr for S3 HeLa cells chronically irradiated ( 10). Others have not seen significant dose rate effects nor inverse dose rate effects with the human and rodent cell lines they tested (5). Note that many of these studies did not include dose rates below 60 cGy/hr so that dose rate and inverse dose rate effects that exist below this dose rate could have been missed. Furthermore. in many of the experiments, plateau or “reduced growth phase” cells were studied which may not allow for the expression of survival differences which depend on cell cycle effects. The dose rate and inverse dose rate effects seen in our studies for the astrocytic lines tested may become important if normal brain is found to have a different response to similar dose rates. Wide differences in responses to low dose rate irradiation among different mammalian cell lines and human tumor lines have been reported (9, 15).
1403
The low dose rate studies underscore the importance of repair, redistribution, and repopulation which occur in tissues responding to chronic low dose rate irradiation. The magnitude and significance of these responses varies considerably over a relatively small range of dose rates for the three astrocytic tumor lines studied. CONCLUSION A difference in radiosensitivity exists between the high and low grade astrocytoma cell lines studied, the low grade line being more sensitive than the high grade lines. Cellular proliferation continues at dose rates less than 26 cGy/hr. A dose rate effect exists for exponentially growing cells irradiated at dose rates from 14 cGy/hr to 78 Gy/hr. An inverse dose rate effect exists at 37 cGy/hr due to differences in progression through the cell cycle. An effective range of dose rates has been identified for these astrocytic cell lines, 37 cGy/hr to 79 cGy/hr, which may be applicable to clinical applications of low dose rate irradiation. These findings may prove useful in developing future dose escalation strategies in the treatment of low and high grade astrocytic brain tumors.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
8.
9.
Bouchard, J. Effects of irradiation in treatment of intercranial gliomas-treatment results by histologic groups (Chapter 5). In: Bouchard. J., ed. Radiation therapy of tumors and diseases of the nervous system. Philadelphia, PA: Lea and Febiger; 1966:78-I 18. Davis, L. W. Presidential address: malignant glioma-A nemesis which requires clinical and basic investigation in radiation oncology. Int. J. Radiat. Oncol. Biol. Phys. 16: 13551365; 1989. Garcia, D. M.: Fulling, K. H.: Marks, J. E. The value of radiation therapy in addition to surgery for astrocytomas of the adult cerebrum. Cancer 55:919-927; 1985. Gerweck, L. E.; Kornblith, P. L.: Burlett, P.; Wang, J.; Sweigert, S. Radiation sensitivity of cultured human glioblastoma cells. Radiology 125:23 l-234; 1977. Hall. E. J.; Marchese. M. J.; Astor, M. B.: Morse, T. Response of cells of human origin, normal and malignant, to acute and low dose rate irradiation. Int. J. Radiat. Oncol. Biol. Phys. l2:655-659; 1986. Laws, E. R.; Taylor, W. F.: Clifton, M. B.: Okazaki. H. Neurosurgical management of low grade astrocytomas of the cerebral hemispheres. J. Neurosurg. 6 1:665-673; 1984. Malaise. E. P.; Fertil, B.: Chavaudra, N.; Guichard, M. Distribution of radiation sensitivities for human tumor cells of specific histological types: comparison of in vitro and in vivo data. Int. J. Radiat. Oncol. Biol. Phys. 12:617-624; 1986. Malaise. E. P.; Fertil, B.; Deschavanne, P. J.: Chavaudra, N.; Brock. W. A. Initial slope of radiation survival curves is characteristic of the origin of primary and established cultures of human tumor cells and fibroblasts. Rad. Res. 111:319-333; 1987. Mitchell, J. B.; Bedford, J. S.: Bailey, S. M. Dose rate effects in mammalian cells in culture. III. Comparison of cell killing
and cell proliferation during continuous irradiation for six different cell lines. Rad. Res. 79537-55 I; 1979. 10. Mitchell, J. B.; Bedford, J. S.; Bailey, S. M. Dose-rate effects on the cell cycle and survival of S3 HeLa and V79 cells. Rad. Res. 79:520-536; 1979. I I. Nilsson. S.; Carlsson, J.: Larrson, B. Survival of irradiated glia and glioma cells studied with a new cloning technique. Int. J. Radiat. Biol. 37:267-279: 1980. 12. Radford. I. R.; Hodgson, S. H. Effect of ploidy on the response of V79 cells to ionizing radiation. Int. J. Radiat. Biol. 5 1:765-778; 1986. 13. Rosenblum, M. L.; Vasquez, D. A.; Hoshino, T.; Wilson. C. B. Development a clonogenic cell assay for human brain tumors. Cancer 41:2305-2314: 1978. 14. Shapiro, W. R. Therapy of adult malignant brain tumors. What have the clinical trials taught us? Sem. Oncol. 13:3845; 1986. 15. Steel, G. G.: Deacon, J. M.; Duchesne, G. M.; Horwich. A.; Kelland, L. R.; Peacock, J. H. The dose rate effect in human tumour cells. Radiother. Oncol. 9:299-310: 1987. 16. Walker. M. D.: Strike, T. A.; Sheline, G. A. An analysis of dose-effect relationship in the radiotherapy of malignant brain tumors. Int. J. Radiat. Oncol. Biol. Phys. 5:17251731; 1979. 17. Weichselbaum, R. R.: Epstein, J.; Little, J. B.; Kornblith. P. L. In vitro cellular radiosensitivity of human malignant tumors. Eur. J. Cancer 12:47-5 1: 1976. 18. Westermark. B. The deficient density-dependent growth control of human malignant glioma cells and virus-transformed glia-like cells in culture. Int. J. Cancer 12:438-45 1; 1973. 19. Zulch, K. J. Brain tumors. Their biology and pathology, 3rd edition. Berlin-Heidelberg: Springer-Verlag: 1986.
