The Prostate 21:287-295 (1992)
Radiation Sensitization by Estramustine Studies on Cultured Human Prostatic Cancer Cells Solveig Eklov, Magnus Essand, Jorgen Carlsson, and Sten Nilsson Depaflment on Oncology, University Hospital (S.E., S.N.); Department of Radiation Sciences, Uppsala University (M.E., J.C.),Uppsala, Sweden In low-stage prostatic carcinoma, local cure can be obtained with radiation therapy alone, while in locally advanced disease the chances for cure are less. In this study, we have addressed the question of whether estramustine (EM), the main cytostatic metabolite of estramustine phosphate (Estracyt), may act as a radiosensitizing agent. This drug accumulates in prostatic cancer and has also been shown to arrest cancer cells at metaphase both in vitro and in vivo. The human prostatic cancer cell line DU 145 was grown as cultures monolayer and incubated with EM in concentrations varying from 1 to 20 kg/ml. External beam irradiation was performed with doses ranging from 0 to 8 Gy using gamma rays from a %o source. Clonogenic cell survival (CS) was used to analyse the radiation sensitizing effect of EM. The radiation dose modifying factor (DMF) at the survival level 0.1 was found to be 0.77 in the presence of EM (5 kgiml), i.e., 2 3 8 sensitization was obtained. When irradiating cells at the standard fraction dose of 2 Gy in the absence of EM, 22% of the cells lost their clonogenic ability. In presence of EM ( 5 kgiml), 2 Gy caused 40% of the cells to lose their clonogenic ability. Thus a radiation sensitizing effect of EM was established in the CS assay. It was also of interest to determine if the radiosensitizing effect of EM could be confirmed in a rapid assay. The rapid fluorescence assay was used to observe early damage of the cells. Results showed that by 2 days after exposure to irradiation a weak tendency towards sensitization was seen, while a clear sensitization was obtained after 4 days. This indicates that the rapid assay might be developed to a predictive assay for detection of the responae of primary prostate tumor cells to the radiation sensi tizing effect of EM. Q 1992 wilcy-Li\\. h c .
Key words: DU 145, prostate cancer, radiation effects
INTRODUCTION
Curative treatment for low-stage prostatic carcinoma can be obtained with radiotherapy or radical surgery. The long-time survival rates are comparable for these two techniques [l]. A clear relationship exists between initial tumor size and the possibility of obtaining local cure with radiation [ 2 ] .While Stage A patients have a local recurrence of about 3 percent 10 years after radiotherapy, the corresponding value for Stage C patients is 2 1 % 131. Since doses above 70 Gy to the pelvis increase
Received for publication February 24, 1992; accepted June 16, 1992. Address reprint requests to Solveig Eklov, Department of Oncology, University Hospital. S-75 1 85, Uppsala, Sweden.
0 1992 Wiley-Liss, Inc.
288
Eklov et al.
the risk for severe complications, there is a need for improved radiation therapy techniques, and one strategy that could be further pursued is to develop radiation sensitizing agents. One drug of interest in this respect is estramustine (EM), the main cytostatic metabolite of Estracyt, introduced into clinical use in 1969 [4] and, since then, used in the treatment of advanced prostatic carcinoma. Estracyt is rapidly desphosphorylated in the gastrointestinal tract, producing estramustine [5] and its oxidized analogue estramustine components [6]. EM accumulates in the prostatic gland and in prostatic cancer [7]. It has been shown to arrest prostatic cancer cells at atypical metaphase in vitro [8] and in vivo [9]. It also inhibits clonogenic cell survival of human prostatic cancer cells in vitro [ 101. The techniques used in this study were clonogenic cell survival (CS) and a rapid fluorescence assay (RF). By the CS assay, it is possible to distinquish cells with a seemingly unlimited proliferation capacity from cells that have lost their reproductive capacity. Human prostatic cancer cells DU 145 were preincubated with EM in varying concentrations prior to treatment with ionizing radiation to analyse whether or not the drug was able to modify the radiosensitivity of the cells. The RF technique was tested both as an adjunct to the CS technique and as a potential predictive test for radiation sensitization.
MATERIALS AND METHODS Cell Culture
The human prostatic cancer cell line, DU 145, was originally isolated and characterized by Stone [ 111 and Mickey [ 121 from a brain metastasis of a patient with metastatic prostatic carcinoma. The cells are hormone-independent both in vitro [ 101 and in vivo [13]. Monolayers of DU 145 cells were cultured in tissue culture flasks (J Bibby Science Products Ltd., Stone, Staffordshire, England), and the cells were kept in an incubator (ASSAB Medicin AB, Sweden) with 20% 02, 5% CO,, and high air moisture at a temperature of 37°C. The culture medium was changed 3 times a week, and the cultures were trypsinized and replated once a week. In the CS assay Ham’s F10 medium was used, while RPMI 1640 medium was used in RF assay. Then 10% foetal calf serum, penicillin (200 IU/ml), L-glutamine (2 mM), and streptomycin (200 pg/ml) were added to the medium. In the CS assay fungizone (2.5 pg/ml) was added to avoid fungi contamination. All chemicals were obtained from Biological Industries, Kibbutz Beth Haemek, Israel. The discrepancy in the choice of culture medium was dependent on different routines at the laboratories where the assays CS 2141 and RF [15] were run. No medium-type-dependent differences in the growth pattern of the DU 145 cells were observed. Chemicals Estramustine (EM) was obtained from Kabi-Pharmacia AB, Helsingborg, Sweden. The substance was dissolved in ethanol or dimethylsulphoxide (DMSO) at a concentration of 10 pg/pl. This solution was then diluted in culture medium to an EM concentration of 20 pg/ml. The final concentration of ethanol or DMSO was then 0.02 ppm, which did not disturb the growth of the DU 145 cells and did not have any radiosensitizing effect.
Radiation Sensitization by Estramustine
289
External Beam Irradiation DU 145 monolayer cultures were preincubated for 23 hours with 0, 1 , 5 , 10, and 20 p g EM per milliliter of culture medium. Then the cultures, still remaining in medium containing EM, were irradiated under standard incubation conditions by gamma rays from a 6oCo source. Single doses were given in an interval ranging from 0 to 8 Gy. The dose rate, when irradiating cell cultures for the CS assay, was 0.94 2 0.02 Gy/min, while for the RF assay it was 1.35 r+ 0.04 Gy/min. Clonogenic Survival Assay
One hour after exposure to irradiation, the cultures were trypsinized and suspended in culture medium to a single-cell-suspension. Cells were plated for clonogenic growth in tissue culture dishes (Bibby, 10 cm) under standard culture conditions. Untreated cells were plated as controls. Fresh culture medium was added once during the period of clonogenic growth. After 13 days of incubation, the formed cell colonies were washed in PBS and fixed in 10% neutral formalin solution (Sigma Diagnostic, St. Louis, MO). The cell colonies were then stained with haematoxylin. Colonies consisting of more than 50 cells were counted by an image analyser (Analytical Measuring Systems, Analytical Instruments Ltd, Sherill, Shaffron Walden, Essex, England). For each of the different concentrations of EM, the clonogenic cell survival was plotted as a function of the radiation dose. Computer programs, similar to those previously described by Albright [ 161, were used in the analyses of clonogenic cell survival data. The programs fitted the single-hit multi-target and the linear-quadratic models to the data by the iterative, weighted, least square method. The models have been described by, for example, Steel et al. 1171 and Tubiana et al. [18]:
s
= 1
-
(1
-
e--D/D~)n
Single-hit multitarget model
s
=
e-(aD
+ PDz)
Linear-quadratic model.
Parameters evaluated for the single-hit multitarget were Do (the dose reducing survival by 37% at the exponential part of the survival curve) and n (the extrapolation number, formally equivalent to the number of targets in a cell that each have to be inactivated by a single hit). In the linear-quadratic model, values of a (the linear component, representing lethal events) and p (the quadratic component, representing sublethal events) were evaluated. Rapid Fluorescence Assay
The prinicipal steps of the rapid fluorescence assay, the RF assay, have previously been described by Larsson and Nygren [ 151. Four thousand (4,000) cells in culture medium were seeded into the wells of a flat-bottomed 96-well microtiter plate (NUNC, Roskilde, Denmark). After 3 hours, EM was added and 23 hours thereafter the cells were irradiated. One hour after irradiation, the EM was removed by gentle aspiration and new culture medium without EM was added. For each concentration of EM, 6 wells were analyzed. The viability of the cells was counted 1, 2, 3, 4, and 7 days after irradiation.
290
Eklov et al. TABLE I. Clonogenic Survival of DU 145 Cells After Exposure for 24 Hours to Varying Concentrations of EM EM conc. (pgiml) 0 1 5 10 20
Survival fraction
Standard deviation
1 .oo
0.00 0.05 0.05 0.04 0.03
1.04 0.71 0.46
0.22
Mean values are given from the evaluation of 6 culture dishes in 2 independent experiments.
The 96-well microtiter plates were centrifuged (200 g, 5 min) and the medium was removed by flicking the plates. The wells were washed once in phosphate buffered saline (PBS pH 7.4). The fluorogenic substance fluorescein diacetate, FDA (Sigma Chemical, St. Louis, MO) was dissolved in DMSO and kept frozen at -20°C as a stock solution (10 mg/ml). It has previously been shown that FDA is clearly dependent on the fraction of intact and viable cells [ 151. FDA was diluted in PBS buffer to a concentration of 10 pg/ml and added column-wise at 200 pl per well. The plates were incubated for 60 min in 37°C. A 96-well scanning fluorometer (Fluoroscan 2) was used to count the emitted fluorescence, and the results were transferred to a Macintosh SE personal computer. The plates were kept under standard culture conditions during the whole experiment. Cell-suspension, culture medium, drugs, and buffers were added to the wells using 8-channel multitip pipettes. The RF assay is a novel approach to radiosensitivity testing of malignant cells. It is of interest to study whether or not this technique can demonstrate the same effects as seen with the more laborious clonogenic assay. RESULTS
Table I shows the clonogenic cell survival fraction of the human prostatic cancer cells DU 145 after exposure to varying concentrations of EM. Figure 1 shows the clonogenic cell survival of the human prostatic cancer cells DU 145 as a function of the radiation dose. The cell cultures were preincubated for 23 hours with varying concentrations of EM, and EM was kept present during the irradiations. The values of the survival fraction due to exposure to EM (see Table I) were in Fig. 1, normalized to survival = 1. The deviations between the curves for different concentrations of EM were then interpreted as the radiation sensitization effect. Two mathematical models were used to describe the actual cell survival data in Fig. 1: the single-hit multi target model and the linear-quadratic model. Parameter values evaluated from the computer program are presented in Table 11. The single-hit multitarget model adapted best to the experimental data. The dose modifying factor, DMF, was defined as the ratio of radiation doses to give a fixed level of survival, with/without preincubating the cell cultures with EM. At the survival level 0.1, the EM concentration 5 pg/ml gave DMF = 0.77 (5.4 Gy/7.O Gy), while the EM concentration 20 pg/ml gave DMF = 0.70 (4.9 Gy/7.O Gy). Thus EM at a concentration of 5 pg/ml culture medium radiosensitized the DU
Radiation Sensitization by Estramustine
1.00
291
~"'""1"""""'1
0.10
0.01
I
'
0
'
I
2
I
'
"
4
'
'
'
I
6 Dose (Gy)
"
"
8
"
'
' 10
Fig. 1 . Clonogenic cell survival curves for DU 145 at varying concentrations of EM, after external beam irradiation with gamma rays from a "Co source. The concentrations of EM were 0 = 0, 0 = 1, A = 5 , = 10, and = 20 kg/ml. Each point corresponds to the evaluation of 6 culture dishes in 2 independent experiments. Mean values and standard deviations are given.
+
145 cultures with 23% (1-0.77). At the EM concentration of 20 pg/ml, the radiosensitizing effect was 30% (1-0.70). In Fig. 2, the clonogenic cell survival at the standard fraction dose, 2 Gy, is given for DU 145 cells kept in culture medium at varying concentrations of EM. The survival values presented were all normalized to the survival of 0 Gy for each specific concentration of EM. It is seen that 2 Gy of gamma rays in the absence of EM killed 22% of the cells (surviving fraction = 0.78), while 2 Gy in the presence of EM (5 pg/ml) killed 40% of the cells (surviving fraction = 0.60). It is noteworthy that the sensitizing effect did not dramatically increase when the EM concentration was further increased. The relative fluorescence of DU 145 cells from the RF assay is shown in Table 111. In Table IIIa, the results 2 days after exposure to irradiation and EM can be seen, while Table IIIb shows the results after 4 days. Similar measurements were also made 1 , 3, and 7 days after the treatments. It turned out that 7 days was too long an evaluation time for the RF assay. The fluorescence values were normalized to the fluorescence of the untreated control cells. Both irradiation and EM had an effect on the relative fluorescence but, as seen especially in Table IIIb, the combined effect was higher than the effect obtained when simply adding the effects from radiation and the effects of EM. Thus, the rapid fluorescence assay as well as the clonogenic cell survival assay indicated radiation sensitization by EM. The radiosensitizing effect was, in the RF assay, most obvious 4 days after exposure to irradiation. Figure 3 shows the relative fluorescence signal 4 days after irradiation and exposure to EM. All fluorescence values were normalized to the fluorescence signal of the nonirradi-
292
Eklov et al. TABLE 11. Clonogenic Survival Parameters For DU 145 Cells Exposed to Irradiation and Varying Concentrations of EM Single-hit multitarget
EM conc. (Fgiml) 0 1 5 10 20
Linear-quadratic
n
DO
a
P
3.13 3.35 2.02 1.60 1.91
2.05 2.03 1.80 2.07 1.35
0.12 0.10 0.30 0.30 0.40
0.029 0.030 0.020 0.015 0.0087
The parameter values were evaluated from the single-hit multitarget and the linearquadratic models [ 181.
1
-A
3
0.1
0
5
10
EM conc. (pg/ml)
15
20
Fig. 2. Clonogenic cell survival as a function of the concentration of EM at an external beam irradiation of 2 Gy with gamma rays from a 6oCo source. Each point corresponds to evaluation of 6 dishes in 2 independent experiments. Mean values and standard deviations are given.
ated cultures at each concentration of EM. A radiation sensitizing effect was obtained at 5 , 10, and 20 pg EM per milliliter of culture medium. DISCUSSION
Some chemotherapeutic agents are likely to interact with radiation and to improve the tumor response to radiotherapy. The results in this study show that estramustine gives sensitization to radiation in the human prostatic cancer cell line DU 145. In the CS assay, the dose modifying factor was 0.77 at the survival level 0.1 in the presence of EM at a concentration of 5 pg/ml culture medium, i.e., the radiation sensitization was 23%. Doses of 5-20 pg/ml EM are similar to the concentrations that can be achieved locally in prostatic carcinoma in a patient [7]. This indicates that EM could be useful in combination with radiotherapy in the treatment of prostatic cancer patients.
Radiation Sensitization by Estramustine
293
TABLE 111. Effects of Irradiation and EM on DU 145 Cells in the RF Assay 2 Days and 4 Days After Exposure to Irradiation and EM Relative fluorescence EM conc. ( p g h l ) a) 2 days after irradiation 0
I .O 5.0 10.0 20.0
b) 4 days after irradiation 0 1.o 5.0 10.0 20.0
0 GY
2 GY
4 GY
8 GY
100% 92.9 51.7 40.3 33.7
97.6 92.0 53.1 40.3 37.4
90.5 85.9 54.8 35.5 31.7
79.5 74.4 41.4 27.4 25.3
100% 93.3 72.1 62.6 65.4
95.5 98.3 69.4 56.1 53.7
95.8 89.0 54.3 46.3 48.3
70.4 74.8 44.0 29.3 31.0
The fluorescence values were normalized to the fluorescence of the cells that were neither treated with irradiation nor EM. Mean values from evaluation of at least 6 wells are given.
6!
100
8 c; 8
E
s
G:
z
*
-4 (d
H
d
10 0
2
4
6 Dose (Gy)
8
10
Fig. 3 Relative fluorescence 4 days after irradiating DU 145 cells at varying concentrations of EM. The concentrations were 0 = 0, = 1, A = 5, = 10, and = 20 kg/ml. Mean values are given from the evaluation of 6 wells. The errors are at all points less than 10%.
+
The surviving fraction obtained in the CS assay after the standard dose of 2 Gy was 0.78. According to the classification by Deacon et al. [19], the cell line DU 145 must then be considered radioresistant. When irradiating cells in presence of EM (5 pg/ml), the radiation dose of 2 Gy gave a surviving fraction of 0.60. Thus, a significant sensitizing effect of EM was seen at the low radiation dose of 2 Gy.
294
Eklov et al.
Two standard models, the single-hit multitarget and the linear-quadratic models [ 181, were used to mathematically describe the clonogenic cell survival data from the
survival curve presented in Fig. 1. The dose Do is defined as the dose that reduces the surviving fraction in the exponential phase of the curve to 37%. The expression n is equivalent to the value where the asymptotic line crosses the y-axis. Both the value of Do and n decreased when the DU 145 prostatic carcinoma cells were exposed to estramustine during exposure to irradiation. This means that the radiation dose needed to reduce the surviving fraction to 37% was less, when estramustine was present during exposure to irradiation. The (Y parameter describes the dose-effect relation in the shoulder or low-dose region, while the p parameter describes the dose-effect relation in the high-dose region of the survival curve. The ratio a/@is an index of their relative importance: It is high when the survival curve is almost exponential and low when the shoulder is very wide. It was found that a@ increased when the cells were exposed to higher concentrations of estramustine during exposure to irradiation. This means that the shoulder region of the survival curve was less pronounced, i.e., low doses have higher efficiency in inactivating cells when estramustine is present during exposure to irradiation. This indicates that EM is a good radiation sensitizing agent at the low radiation doses which are preferred as daily doses in a fractionation schedule in the treatment of cancer patients. Many different drugs have been found to be radiosensitizing and thereby potentially valuable for clinical purposes, but most of them have caused a high degree of toxicity. The results obtained in this study show that EM potentiates the effect of ionizing radiation in the prostatic cancer cells studied. This finding is of interest since it is known that this substance accumulates in high amounts in prostatic cancer [7] and EM may therefore be used as a radiosensitizing agent. Estramustine-binding protein (EMBP) is a binding protein for EM; it binds EM with high affinity and capacity [20]. This protein was first discovered in prostatic tissue [21] and is believed to be the reason for the accumulation of EM in cancer cells. EMBP is expressed in different types of malignant tumours, e.g., some lung cancer cells and glioblastomas. We therefore do not believe that the radiosensitizing effect of EM is specific only for prostatic cancer cells. Further studies, however, have to be performed to show a potential clinical use in this respect. In vivo experiments are currently being performed in our laboratory in which heterotransplants of the human prostatic cancer cell line DU 145 in nude mice treated with radiation and EM are being used. The trend in modern radiation therapy is to individualize treatment. Patients undergo explorative laparotomy for staging, and the treatment volume can be effectively selected using CT-based dose planning systems. There is also an urgent need for predictive tests with respect to the radiation sensitivity of the tumor cells. The RF assay could easily be used on tumor biopsies from patients and therefore could become a clinicai aid in this respect, since in the experimental setup it visualized the radiation-sensitizing characteristics of EM within a few days. ACKNOWLEDGMENTS
The authors wish to thank Erianne Christina Carlsson at the Department of Radiation Sciences and Lena Lennartsson at the Department of Oncology for valuable
Radiation Sensitization by Estramustine
295
technical assistance with cell cultures. This work was financially supported by the Swedish Cancer Society, grants 1176-B91-05XAB and 2503-B91-6XAC; the Swedish National Board for Technical Development 88-03 129P; the Swedish National Board of Laboratory Animals 89-40; and the Swedish Lions’ Cancer Foundation in the Uppsala region. REFERENCES 1. Wittes RE (ed): “NCI Monograph 7.” Bethesda, MD: National Cancer Institute, 1988. 2. Hanks GE, Leibel S , Krall JM, Krame S: Dose response observations for local control of adenocarcinoma of the prostate. Int J Radial Oncol Biol Phys 11:153-157, 1985. 3. Hanks GE: In Moss WT, Cox, JT (eds): “Radiation Oncology.” St. Louis: CV Mosby Co, 1989, pp 487-5 11, 4. Alfthan DS, Rusk J: Estracyt in advanced prostatic carcinoma. Ann Chir Gynecologine Fennial 58:234, 1969. 5. Anderson SB, Gunnarsson PO, Nilsson T, Plym-Forshell G: Metbolism of estramustine phosphate (Estracyt) in patients with prostatic carcinoma.oEur J Metab Pharmacokinet -: 149-154, 198 1. 6. Gunnarsson PO, Plym-Forshell G, Fritjofsson A, Norlen BJ: Plasma concentrations of estramustine phosphate and its major metabolites in patients with prostatic carcinoma treated with different doses of estramustine phosphate (Estracyt). Scand J Urol Nephrol 15:2016206, 198 1. 7. Norltn BJ, Anderson S-B, Bjork P, Gunnarsson PO, Fritjofsson A: Uptake of estramustine phosphate (Estracyt) metabolites in prostatic cancer. J Urol 140:1058-1062, 1988. 8. Hartley-Asp B: Estramustine-induced mitotic arrest in two human prostatic carcinoma cell lines DU 145 and PC-3. Prostate 5:93-100, 1984. 9. Eklov S , Nilsson S , Larson A, Bjork, P, Hartley-Asp B: Evidence for a nonestrogenic cytostatic effect of estramustine on human prostatic carcinoma cells in vivo. Prostate 20:43-50, 1992. 10. Harley-Asp B, Gunnarsson PO: Growth and cell survival following treatment with estramustine, nor-nitrogen mustard, estradiol and testosterone of a human prostatic cancer cell line (DU 145). J Urol 1275318-822, 1982. 11. Stone KR, Mickey DD, Wunderli H, Mickey GH, Paulson DF: Isolation of a human prostate carcinoma cell line (DU 145). Int J Cancer 21:274-281, 1978. 12. Mickey DD, Stone KR, Wunderli H, Mickey GH, Vollmer RT, Paulson DP: Heterotransplantation of a human prostatic adenocarcinoma cell line in nude mice. Cancer Res 37:4049-4058, 1977. 13. Eklov S , Larson A, Bjork P, Nilsson S: The expression of estramustine-binding protein in the human prostatic cancer cell line DU 145 is not androgen dependent. Scand J Urol Nephrol 26: 119-125, 1992. 14. Capala J , Prihl M, Scott-Robson S , Ponten J, Westermark B, Carlsson J: Effects of ‘3’I-EGF on cultured human glioma cells, J Neuro-Oncolo 9:201-210, 1990. 15. Larsson R, Nygren P: A rapid fluorometic method for semiautomated determination of cytotoxicity and cellular proliferation of human tumor cell lines in microculture. Anticancer Res 9: 1 1 11-1 120, 1989. 16. Albright N: Computer programs for the analysis of cellular survival data. Radia Res 112:331-340. 1987. 17. Steel GG, A d a m GE, Horwich A: Survival of clonogenic cells: Cell-survival curves. In “Biological Basis of Radiotherapy, 2nd ed.” Amsterdam: Elsevier, 1989, pp 45-63. 18. Tubiana M, Dutreix J, Wamberise A, (Translated into English by DR Bewley): “Introduction to Radiobiology.” London: Taylor & Francis, 1990, pp 97-1 10. 19. Deacon J, Peckham MJ, Steel GG: The radioresponsiveness of human tumors and the initial slope of the cell survival curve. Radiother Oncol 2:317-323, 1984. 20. Forsgren B, Gustafsson J-A, Pousette A, Hogberg B: Binding characteristics of a major protein in rat ventral prostate cytosol that interacts with estramustine: A nitrogen mustard derivative of 17pestradiol. Cancer Res 39:5155-51?, 1979. 21. Bjork P, Forsgren B, Gustafsson J-A, Pousette A, Hogberg B: Partial characterization and “quantitation” of a human prostatic estramustine-binding protein. Cancer Res 42:1935-1942, 1982.
Radiation Sensitization by Estramustine Studies on Cultured Human Prostatic Cancer Cells Solveig Eklov, Magnus Essand, Jorgen Carlsson, and Sten Nilsson Depaflment on Oncology, University Hospital (S.E., S.N.); Department of Radiation Sciences, Uppsala University (M.E., J.C.),Uppsala, Sweden In low-stage prostatic carcinoma, local cure can be obtained with radiation therapy alone, while in locally advanced disease the chances for cure are less. In this study, we have addressed the question of whether estramustine (EM), the main cytostatic metabolite of estramustine phosphate (Estracyt), may act as a radiosensitizing agent. This drug accumulates in prostatic cancer and has also been shown to arrest cancer cells at metaphase both in vitro and in vivo. The human prostatic cancer cell line DU 145 was grown as cultures monolayer and incubated with EM in concentrations varying from 1 to 20 kg/ml. External beam irradiation was performed with doses ranging from 0 to 8 Gy using gamma rays from a %o source. Clonogenic cell survival (CS) was used to analyse the radiation sensitizing effect of EM. The radiation dose modifying factor (DMF) at the survival level 0.1 was found to be 0.77 in the presence of EM (5 kgiml), i.e., 2 3 8 sensitization was obtained. When irradiating cells at the standard fraction dose of 2 Gy in the absence of EM, 22% of the cells lost their clonogenic ability. In presence of EM ( 5 kgiml), 2 Gy caused 40% of the cells to lose their clonogenic ability. Thus a radiation sensitizing effect of EM was established in the CS assay. It was also of interest to determine if the radiosensitizing effect of EM could be confirmed in a rapid assay. The rapid fluorescence assay was used to observe early damage of the cells. Results showed that by 2 days after exposure to irradiation a weak tendency towards sensitization was seen, while a clear sensitization was obtained after 4 days. This indicates that the rapid assay might be developed to a predictive assay for detection of the responae of primary prostate tumor cells to the radiation sensi tizing effect of EM. Q 1992 wilcy-Li\\. h c .
Key words: DU 145, prostate cancer, radiation effects
INTRODUCTION
Curative treatment for low-stage prostatic carcinoma can be obtained with radiotherapy or radical surgery. The long-time survival rates are comparable for these two techniques [l]. A clear relationship exists between initial tumor size and the possibility of obtaining local cure with radiation [ 2 ] .While Stage A patients have a local recurrence of about 3 percent 10 years after radiotherapy, the corresponding value for Stage C patients is 2 1 % 131. Since doses above 70 Gy to the pelvis increase
Received for publication February 24, 1992; accepted June 16, 1992. Address reprint requests to Solveig Eklov, Department of Oncology, University Hospital. S-75 1 85, Uppsala, Sweden.
0 1992 Wiley-Liss, Inc.
288
Eklov et al.
the risk for severe complications, there is a need for improved radiation therapy techniques, and one strategy that could be further pursued is to develop radiation sensitizing agents. One drug of interest in this respect is estramustine (EM), the main cytostatic metabolite of Estracyt, introduced into clinical use in 1969 [4] and, since then, used in the treatment of advanced prostatic carcinoma. Estracyt is rapidly desphosphorylated in the gastrointestinal tract, producing estramustine [5] and its oxidized analogue estramustine components [6]. EM accumulates in the prostatic gland and in prostatic cancer [7]. It has been shown to arrest prostatic cancer cells at atypical metaphase in vitro [8] and in vivo [9]. It also inhibits clonogenic cell survival of human prostatic cancer cells in vitro [ 101. The techniques used in this study were clonogenic cell survival (CS) and a rapid fluorescence assay (RF). By the CS assay, it is possible to distinquish cells with a seemingly unlimited proliferation capacity from cells that have lost their reproductive capacity. Human prostatic cancer cells DU 145 were preincubated with EM in varying concentrations prior to treatment with ionizing radiation to analyse whether or not the drug was able to modify the radiosensitivity of the cells. The RF technique was tested both as an adjunct to the CS technique and as a potential predictive test for radiation sensitization.
MATERIALS AND METHODS Cell Culture
The human prostatic cancer cell line, DU 145, was originally isolated and characterized by Stone [ 111 and Mickey [ 121 from a brain metastasis of a patient with metastatic prostatic carcinoma. The cells are hormone-independent both in vitro [ 101 and in vivo [13]. Monolayers of DU 145 cells were cultured in tissue culture flasks (J Bibby Science Products Ltd., Stone, Staffordshire, England), and the cells were kept in an incubator (ASSAB Medicin AB, Sweden) with 20% 02, 5% CO,, and high air moisture at a temperature of 37°C. The culture medium was changed 3 times a week, and the cultures were trypsinized and replated once a week. In the CS assay Ham’s F10 medium was used, while RPMI 1640 medium was used in RF assay. Then 10% foetal calf serum, penicillin (200 IU/ml), L-glutamine (2 mM), and streptomycin (200 pg/ml) were added to the medium. In the CS assay fungizone (2.5 pg/ml) was added to avoid fungi contamination. All chemicals were obtained from Biological Industries, Kibbutz Beth Haemek, Israel. The discrepancy in the choice of culture medium was dependent on different routines at the laboratories where the assays CS 2141 and RF [15] were run. No medium-type-dependent differences in the growth pattern of the DU 145 cells were observed. Chemicals Estramustine (EM) was obtained from Kabi-Pharmacia AB, Helsingborg, Sweden. The substance was dissolved in ethanol or dimethylsulphoxide (DMSO) at a concentration of 10 pg/pl. This solution was then diluted in culture medium to an EM concentration of 20 pg/ml. The final concentration of ethanol or DMSO was then 0.02 ppm, which did not disturb the growth of the DU 145 cells and did not have any radiosensitizing effect.
Radiation Sensitization by Estramustine
289
External Beam Irradiation DU 145 monolayer cultures were preincubated for 23 hours with 0, 1 , 5 , 10, and 20 p g EM per milliliter of culture medium. Then the cultures, still remaining in medium containing EM, were irradiated under standard incubation conditions by gamma rays from a 6oCo source. Single doses were given in an interval ranging from 0 to 8 Gy. The dose rate, when irradiating cell cultures for the CS assay, was 0.94 2 0.02 Gy/min, while for the RF assay it was 1.35 r+ 0.04 Gy/min. Clonogenic Survival Assay
One hour after exposure to irradiation, the cultures were trypsinized and suspended in culture medium to a single-cell-suspension. Cells were plated for clonogenic growth in tissue culture dishes (Bibby, 10 cm) under standard culture conditions. Untreated cells were plated as controls. Fresh culture medium was added once during the period of clonogenic growth. After 13 days of incubation, the formed cell colonies were washed in PBS and fixed in 10% neutral formalin solution (Sigma Diagnostic, St. Louis, MO). The cell colonies were then stained with haematoxylin. Colonies consisting of more than 50 cells were counted by an image analyser (Analytical Measuring Systems, Analytical Instruments Ltd, Sherill, Shaffron Walden, Essex, England). For each of the different concentrations of EM, the clonogenic cell survival was plotted as a function of the radiation dose. Computer programs, similar to those previously described by Albright [ 161, were used in the analyses of clonogenic cell survival data. The programs fitted the single-hit multi-target and the linear-quadratic models to the data by the iterative, weighted, least square method. The models have been described by, for example, Steel et al. 1171 and Tubiana et al. [18]:
s
= 1
-
(1
-
e--D/D~)n
Single-hit multitarget model
s
=
e-(aD
+ PDz)
Linear-quadratic model.
Parameters evaluated for the single-hit multitarget were Do (the dose reducing survival by 37% at the exponential part of the survival curve) and n (the extrapolation number, formally equivalent to the number of targets in a cell that each have to be inactivated by a single hit). In the linear-quadratic model, values of a (the linear component, representing lethal events) and p (the quadratic component, representing sublethal events) were evaluated. Rapid Fluorescence Assay
The prinicipal steps of the rapid fluorescence assay, the RF assay, have previously been described by Larsson and Nygren [ 151. Four thousand (4,000) cells in culture medium were seeded into the wells of a flat-bottomed 96-well microtiter plate (NUNC, Roskilde, Denmark). After 3 hours, EM was added and 23 hours thereafter the cells were irradiated. One hour after irradiation, the EM was removed by gentle aspiration and new culture medium without EM was added. For each concentration of EM, 6 wells were analyzed. The viability of the cells was counted 1, 2, 3, 4, and 7 days after irradiation.
290
Eklov et al. TABLE I. Clonogenic Survival of DU 145 Cells After Exposure for 24 Hours to Varying Concentrations of EM EM conc. (pgiml) 0 1 5 10 20
Survival fraction
Standard deviation
1 .oo
0.00 0.05 0.05 0.04 0.03
1.04 0.71 0.46
0.22
Mean values are given from the evaluation of 6 culture dishes in 2 independent experiments.
The 96-well microtiter plates were centrifuged (200 g, 5 min) and the medium was removed by flicking the plates. The wells were washed once in phosphate buffered saline (PBS pH 7.4). The fluorogenic substance fluorescein diacetate, FDA (Sigma Chemical, St. Louis, MO) was dissolved in DMSO and kept frozen at -20°C as a stock solution (10 mg/ml). It has previously been shown that FDA is clearly dependent on the fraction of intact and viable cells [ 151. FDA was diluted in PBS buffer to a concentration of 10 pg/ml and added column-wise at 200 pl per well. The plates were incubated for 60 min in 37°C. A 96-well scanning fluorometer (Fluoroscan 2) was used to count the emitted fluorescence, and the results were transferred to a Macintosh SE personal computer. The plates were kept under standard culture conditions during the whole experiment. Cell-suspension, culture medium, drugs, and buffers were added to the wells using 8-channel multitip pipettes. The RF assay is a novel approach to radiosensitivity testing of malignant cells. It is of interest to study whether or not this technique can demonstrate the same effects as seen with the more laborious clonogenic assay. RESULTS
Table I shows the clonogenic cell survival fraction of the human prostatic cancer cells DU 145 after exposure to varying concentrations of EM. Figure 1 shows the clonogenic cell survival of the human prostatic cancer cells DU 145 as a function of the radiation dose. The cell cultures were preincubated for 23 hours with varying concentrations of EM, and EM was kept present during the irradiations. The values of the survival fraction due to exposure to EM (see Table I) were in Fig. 1, normalized to survival = 1. The deviations between the curves for different concentrations of EM were then interpreted as the radiation sensitization effect. Two mathematical models were used to describe the actual cell survival data in Fig. 1: the single-hit multi target model and the linear-quadratic model. Parameter values evaluated from the computer program are presented in Table 11. The single-hit multitarget model adapted best to the experimental data. The dose modifying factor, DMF, was defined as the ratio of radiation doses to give a fixed level of survival, with/without preincubating the cell cultures with EM. At the survival level 0.1, the EM concentration 5 pg/ml gave DMF = 0.77 (5.4 Gy/7.O Gy), while the EM concentration 20 pg/ml gave DMF = 0.70 (4.9 Gy/7.O Gy). Thus EM at a concentration of 5 pg/ml culture medium radiosensitized the DU
Radiation Sensitization by Estramustine
1.00
291
~"'""1"""""'1
0.10
0.01
I
'
0
'
I
2
I
'
"
4
'
'
'
I
6 Dose (Gy)
"
"
8
"
'
' 10
Fig. 1 . Clonogenic cell survival curves for DU 145 at varying concentrations of EM, after external beam irradiation with gamma rays from a "Co source. The concentrations of EM were 0 = 0, 0 = 1, A = 5 , = 10, and = 20 kg/ml. Each point corresponds to the evaluation of 6 culture dishes in 2 independent experiments. Mean values and standard deviations are given.
+
145 cultures with 23% (1-0.77). At the EM concentration of 20 pg/ml, the radiosensitizing effect was 30% (1-0.70). In Fig. 2, the clonogenic cell survival at the standard fraction dose, 2 Gy, is given for DU 145 cells kept in culture medium at varying concentrations of EM. The survival values presented were all normalized to the survival of 0 Gy for each specific concentration of EM. It is seen that 2 Gy of gamma rays in the absence of EM killed 22% of the cells (surviving fraction = 0.78), while 2 Gy in the presence of EM (5 pg/ml) killed 40% of the cells (surviving fraction = 0.60). It is noteworthy that the sensitizing effect did not dramatically increase when the EM concentration was further increased. The relative fluorescence of DU 145 cells from the RF assay is shown in Table 111. In Table IIIa, the results 2 days after exposure to irradiation and EM can be seen, while Table IIIb shows the results after 4 days. Similar measurements were also made 1 , 3, and 7 days after the treatments. It turned out that 7 days was too long an evaluation time for the RF assay. The fluorescence values were normalized to the fluorescence of the untreated control cells. Both irradiation and EM had an effect on the relative fluorescence but, as seen especially in Table IIIb, the combined effect was higher than the effect obtained when simply adding the effects from radiation and the effects of EM. Thus, the rapid fluorescence assay as well as the clonogenic cell survival assay indicated radiation sensitization by EM. The radiosensitizing effect was, in the RF assay, most obvious 4 days after exposure to irradiation. Figure 3 shows the relative fluorescence signal 4 days after irradiation and exposure to EM. All fluorescence values were normalized to the fluorescence signal of the nonirradi-
292
Eklov et al. TABLE 11. Clonogenic Survival Parameters For DU 145 Cells Exposed to Irradiation and Varying Concentrations of EM Single-hit multitarget
EM conc. (Fgiml) 0 1 5 10 20
Linear-quadratic
n
DO
a
P
3.13 3.35 2.02 1.60 1.91
2.05 2.03 1.80 2.07 1.35
0.12 0.10 0.30 0.30 0.40
0.029 0.030 0.020 0.015 0.0087
The parameter values were evaluated from the single-hit multitarget and the linearquadratic models [ 181.
1
-A
3
0.1
0
5
10
EM conc. (pg/ml)
15
20
Fig. 2. Clonogenic cell survival as a function of the concentration of EM at an external beam irradiation of 2 Gy with gamma rays from a 6oCo source. Each point corresponds to evaluation of 6 dishes in 2 independent experiments. Mean values and standard deviations are given.
ated cultures at each concentration of EM. A radiation sensitizing effect was obtained at 5 , 10, and 20 pg EM per milliliter of culture medium. DISCUSSION
Some chemotherapeutic agents are likely to interact with radiation and to improve the tumor response to radiotherapy. The results in this study show that estramustine gives sensitization to radiation in the human prostatic cancer cell line DU 145. In the CS assay, the dose modifying factor was 0.77 at the survival level 0.1 in the presence of EM at a concentration of 5 pg/ml culture medium, i.e., the radiation sensitization was 23%. Doses of 5-20 pg/ml EM are similar to the concentrations that can be achieved locally in prostatic carcinoma in a patient [7]. This indicates that EM could be useful in combination with radiotherapy in the treatment of prostatic cancer patients.
Radiation Sensitization by Estramustine
293
TABLE 111. Effects of Irradiation and EM on DU 145 Cells in the RF Assay 2 Days and 4 Days After Exposure to Irradiation and EM Relative fluorescence EM conc. ( p g h l ) a) 2 days after irradiation 0
I .O 5.0 10.0 20.0
b) 4 days after irradiation 0 1.o 5.0 10.0 20.0
0 GY
2 GY
4 GY
8 GY
100% 92.9 51.7 40.3 33.7
97.6 92.0 53.1 40.3 37.4
90.5 85.9 54.8 35.5 31.7
79.5 74.4 41.4 27.4 25.3
100% 93.3 72.1 62.6 65.4
95.5 98.3 69.4 56.1 53.7
95.8 89.0 54.3 46.3 48.3
70.4 74.8 44.0 29.3 31.0
The fluorescence values were normalized to the fluorescence of the cells that were neither treated with irradiation nor EM. Mean values from evaluation of at least 6 wells are given.
6!
100
8 c; 8
E
s
G:
z
*
-4 (d
H
d
10 0
2
4
6 Dose (Gy)
8
10
Fig. 3 Relative fluorescence 4 days after irradiating DU 145 cells at varying concentrations of EM. The concentrations were 0 = 0, = 1, A = 5, = 10, and = 20 kg/ml. Mean values are given from the evaluation of 6 wells. The errors are at all points less than 10%.
+
The surviving fraction obtained in the CS assay after the standard dose of 2 Gy was 0.78. According to the classification by Deacon et al. [19], the cell line DU 145 must then be considered radioresistant. When irradiating cells in presence of EM (5 pg/ml), the radiation dose of 2 Gy gave a surviving fraction of 0.60. Thus, a significant sensitizing effect of EM was seen at the low radiation dose of 2 Gy.
294
Eklov et al.
Two standard models, the single-hit multitarget and the linear-quadratic models [ 181, were used to mathematically describe the clonogenic cell survival data from the
survival curve presented in Fig. 1. The dose Do is defined as the dose that reduces the surviving fraction in the exponential phase of the curve to 37%. The expression n is equivalent to the value where the asymptotic line crosses the y-axis. Both the value of Do and n decreased when the DU 145 prostatic carcinoma cells were exposed to estramustine during exposure to irradiation. This means that the radiation dose needed to reduce the surviving fraction to 37% was less, when estramustine was present during exposure to irradiation. The (Y parameter describes the dose-effect relation in the shoulder or low-dose region, while the p parameter describes the dose-effect relation in the high-dose region of the survival curve. The ratio a/@is an index of their relative importance: It is high when the survival curve is almost exponential and low when the shoulder is very wide. It was found that a@ increased when the cells were exposed to higher concentrations of estramustine during exposure to irradiation. This means that the shoulder region of the survival curve was less pronounced, i.e., low doses have higher efficiency in inactivating cells when estramustine is present during exposure to irradiation. This indicates that EM is a good radiation sensitizing agent at the low radiation doses which are preferred as daily doses in a fractionation schedule in the treatment of cancer patients. Many different drugs have been found to be radiosensitizing and thereby potentially valuable for clinical purposes, but most of them have caused a high degree of toxicity. The results obtained in this study show that EM potentiates the effect of ionizing radiation in the prostatic cancer cells studied. This finding is of interest since it is known that this substance accumulates in high amounts in prostatic cancer [7] and EM may therefore be used as a radiosensitizing agent. Estramustine-binding protein (EMBP) is a binding protein for EM; it binds EM with high affinity and capacity [20]. This protein was first discovered in prostatic tissue [21] and is believed to be the reason for the accumulation of EM in cancer cells. EMBP is expressed in different types of malignant tumours, e.g., some lung cancer cells and glioblastomas. We therefore do not believe that the radiosensitizing effect of EM is specific only for prostatic cancer cells. Further studies, however, have to be performed to show a potential clinical use in this respect. In vivo experiments are currently being performed in our laboratory in which heterotransplants of the human prostatic cancer cell line DU 145 in nude mice treated with radiation and EM are being used. The trend in modern radiation therapy is to individualize treatment. Patients undergo explorative laparotomy for staging, and the treatment volume can be effectively selected using CT-based dose planning systems. There is also an urgent need for predictive tests with respect to the radiation sensitivity of the tumor cells. The RF assay could easily be used on tumor biopsies from patients and therefore could become a clinicai aid in this respect, since in the experimental setup it visualized the radiation-sensitizing characteristics of EM within a few days. ACKNOWLEDGMENTS
The authors wish to thank Erianne Christina Carlsson at the Department of Radiation Sciences and Lena Lennartsson at the Department of Oncology for valuable
Radiation Sensitization by Estramustine
295
technical assistance with cell cultures. This work was financially supported by the Swedish Cancer Society, grants 1176-B91-05XAB and 2503-B91-6XAC; the Swedish National Board for Technical Development 88-03 129P; the Swedish National Board of Laboratory Animals 89-40; and the Swedish Lions’ Cancer Foundation in the Uppsala region. REFERENCES 1. Wittes RE (ed): “NCI Monograph 7.” Bethesda, MD: National Cancer Institute, 1988. 2. Hanks GE, Leibel S , Krall JM, Krame S: Dose response observations for local control of adenocarcinoma of the prostate. Int J Radial Oncol Biol Phys 11:153-157, 1985. 3. Hanks GE: In Moss WT, Cox, JT (eds): “Radiation Oncology.” St. Louis: CV Mosby Co, 1989, pp 487-5 11, 4. Alfthan DS, Rusk J: Estracyt in advanced prostatic carcinoma. Ann Chir Gynecologine Fennial 58:234, 1969. 5. Anderson SB, Gunnarsson PO, Nilsson T, Plym-Forshell G: Metbolism of estramustine phosphate (Estracyt) in patients with prostatic carcinoma.oEur J Metab Pharmacokinet -: 149-154, 198 1. 6. Gunnarsson PO, Plym-Forshell G, Fritjofsson A, Norlen BJ: Plasma concentrations of estramustine phosphate and its major metabolites in patients with prostatic carcinoma treated with different doses of estramustine phosphate (Estracyt). Scand J Urol Nephrol 15:2016206, 198 1. 7. Norltn BJ, Anderson S-B, Bjork P, Gunnarsson PO, Fritjofsson A: Uptake of estramustine phosphate (Estracyt) metabolites in prostatic cancer. J Urol 140:1058-1062, 1988. 8. Hartley-Asp B: Estramustine-induced mitotic arrest in two human prostatic carcinoma cell lines DU 145 and PC-3. Prostate 5:93-100, 1984. 9. Eklov S , Nilsson S , Larson A, Bjork, P, Hartley-Asp B: Evidence for a nonestrogenic cytostatic effect of estramustine on human prostatic carcinoma cells in vivo. Prostate 20:43-50, 1992. 10. Harley-Asp B, Gunnarsson PO: Growth and cell survival following treatment with estramustine, nor-nitrogen mustard, estradiol and testosterone of a human prostatic cancer cell line (DU 145). J Urol 1275318-822, 1982. 11. Stone KR, Mickey DD, Wunderli H, Mickey GH, Paulson DF: Isolation of a human prostate carcinoma cell line (DU 145). Int J Cancer 21:274-281, 1978. 12. Mickey DD, Stone KR, Wunderli H, Mickey GH, Vollmer RT, Paulson DP: Heterotransplantation of a human prostatic adenocarcinoma cell line in nude mice. Cancer Res 37:4049-4058, 1977. 13. Eklov S , Larson A, Bjork P, Nilsson S: The expression of estramustine-binding protein in the human prostatic cancer cell line DU 145 is not androgen dependent. Scand J Urol Nephrol 26: 119-125, 1992. 14. Capala J , Prihl M, Scott-Robson S , Ponten J, Westermark B, Carlsson J: Effects of ‘3’I-EGF on cultured human glioma cells, J Neuro-Oncolo 9:201-210, 1990. 15. Larsson R, Nygren P: A rapid fluorometic method for semiautomated determination of cytotoxicity and cellular proliferation of human tumor cell lines in microculture. Anticancer Res 9: 1 1 11-1 120, 1989. 16. Albright N: Computer programs for the analysis of cellular survival data. Radia Res 112:331-340. 1987. 17. Steel GG, A d a m GE, Horwich A: Survival of clonogenic cells: Cell-survival curves. In “Biological Basis of Radiotherapy, 2nd ed.” Amsterdam: Elsevier, 1989, pp 45-63. 18. Tubiana M, Dutreix J, Wamberise A, (Translated into English by DR Bewley): “Introduction to Radiobiology.” London: Taylor & Francis, 1990, pp 97-1 10. 19. Deacon J, Peckham MJ, Steel GG: The radioresponsiveness of human tumors and the initial slope of the cell survival curve. Radiother Oncol 2:317-323, 1984. 20. Forsgren B, Gustafsson J-A, Pousette A, Hogberg B: Binding characteristics of a major protein in rat ventral prostate cytosol that interacts with estramustine: A nitrogen mustard derivative of 17pestradiol. Cancer Res 39:5155-51?, 1979. 21. Bjork P, Forsgren B, Gustafsson J-A, Pousette A, Hogberg B: Partial characterization and “quantitation” of a human prostatic estramustine-binding protein. Cancer Res 42:1935-1942, 1982.
