- Email: [email protected]
Effects of carbetimer, a new antineoplastic drug, on bone metabolism
Bone, 12, 139-142 (1991) Printed in the USA. All rights reserved.
Copyright
87%3282/91 $3.00 + .OO 0 1991 Pergamon Press plc
Effects of Carbetimer, a New Antineoplastic Drug, on Bone Metabolism J. J. BODY,
S. NEJAI,
G. FERNANDEZ,
F. GLIBERT,
and G. O’BRYAN-TEAR’
Department of Medicine and Laboratory of Endocrinology and Breast Cancer Research (J.C. Heuson Laboratoq), Institut J. Border, Brussels, Belgium ‘G.D. Searle & Co., Chicago, USA Address for correspondence and reprints: J.J. Body, M.D.,
Dept. of Medicine,
Abstract
factor-Interferon
resorption-Tumor
1 rue HCger-Bordet,
1000 Bruxelles,
Belgium
immunomodulation, as suggested by the induction of cytotoxic T-cells in mice and the discordant in vitro and in vivo activities against the Lewis lung carcinoma model (Searle & Co. 1988). Carbetimer does not have the usual side effects of chemotherapeutic agents. Bone marrow depression, mucositis or alopecia are minimal or absent, whereas neurotoxicity and hypercalcemia constitute the main toxic effects. Hypercalcemia has been the dose-limiting toxicity in phase I trials and appeared to be dose- and treatment duration-dependent (Grunberg et al. 1987; Dodion et al. 1989). During the phase I trial performed in our Institution, 11 out of 26 patients developed hypercalcemia (Dodion et al. 1989). The recommended dosage from this trial was 6.5 g/m’/day using a daily x 5 schedule every three to four weeks. These long-term effects of Carbetimer on calcium metabolism sharply differ from its short-term or acute effects. We indeed demonstrated that Carbetimer infusions caused an acute decrease in ionized calcium levels and an increase in urinary calcium excretion; these findings were best explained by druginduced calcium chelation with resultant hypercalciuria (Body et al. 1989). The acute and transient hypocalcemia induced by each Carbetimer infusion could evidently not explain the longterm hypercalcemic effects of the drug, and we undertook in vitro studies to unravel the effects of Carbetimer on bone metabolism.
Carbetimer is a new antineoplastic agent whose main side effects consist of neurotoxicity and long-term dose-dependent hypercalcemia. We previously showed that Carbetlmer is a potent calcium chelator responsible for an acute decrease in ionized Ca levels observed in vivo. However, the mechanism of the progressive increase in serum Ca remains unknown. We have evaluated the bone-resorbing effects of Carbetimer on 45Ca-prelabelled neonatal mouse calvariae. Carbetimer induced a dose-dependent increase in 45Ca release which started at a concentration of 1 mg/ml and reached a mean of 3.3 times the control values at 10 mg/ml. This marked increase in 45Ca release was similar on previously killed bones and could not be inhibited by calcitonin. Such concentrations are probably therapeutically relevant given the known affinity of Carbetimer for bone and the large daily doses administered to cancer patients (l&15 g). Since Carbetimer could exert its antineoplastic action through immunomodulation, we also studied its effects on the production of TNF-a and IFN-y which are also known to affect bone metabolism. Carbetimer did not stimulate TNF-a release from isolated normal human monocytes or lymphocytes, but it markedly inhibited T-lymphocyte production of IFN-y, which became undetectable at a concentration of 1 mg of Carbetimer/ml. In summary, Carbetimer-induced hypercalcemia appears to be due to a direct stimulation of osteolysis, but possibly also to an inhibition of IFN-y production. Key Words: Carbetimer-Bone
Instutut J. Bordet,
Bone Metabolism Unit;
Methods Reagents
Carbetimer powder was provided by G.D. Searle & Co. (Chicago, USA), dissolved in water, and further diluted in the appropriate culture medium. Salmon calcitonin (sCT) was kindly given by Sandoz Co. (Base], Switzerland). Pregnant mice (EF-1) were bought from Iffa-Credo (Lyon, France), hPTH l-34 from Peninsula (Belmont, California, USA), Lymphoprep from Nycomed (Oslo, Norway), and all culture reagents from Gibco (Gent, Belgium). We measured the production of “tumor necrosis factor-a” (TNF-a) and “interferon-y” (IFN-7) by using the radioimmunoassays of Medgenix Diagnostics (Flerus, Belgium) (Franchimont et al. 1987; Girardin et al. 1988).
necrosis
y-Calcitonin.
Introduction Carbetimer is a new antineoplastic agent with an unusual spectrum of preclinical antineoplastic activity against tumor cell lines and animal tumor models. Carbetimer is a low molecular weight (1590) polymer derived from ethylene and maleic anhydride and has already demonstrated interesting antitumor properties in humans, particularly against malignant melanoma (Grunberg et al. 1987; Searle & Co. 1988; Dodion et al. 1989). However, its mechanism of action remains unclear. Besides direct antiproliferative effects, Carbetimer could act through
Bone resorption
assay
We used a bone resorption published methods (Reynolds 139
assay adapted from previously et al. 1970; Stem and Krieger
J. J. Body et al.: Carbetimer
140
1983). The assay is based on the measurement of 45Ca release by half-calvariae of neonatal mice (three to four days old) whose mothers had been injected with 0.2 mCi 45Ca two or three days before delivery. Tubes (12 x 75 mm) were placed on a roller tube apparatus (Analys, Namur, Belgium) in an incubator at 37°C. After a 24-hour preincubation in RPMI-1640 supplemented with 2% L-Glutamine (Gln) 200 mM, 10,000 U Penicillin, 10 mg streptomycin/ml, and 20% heat-inactivated serum (15% horse serum, 5% fetal calf serum), bones were incubated in fresh complete culture medium containing various concentrations of Carbetimer. Bone resorption was evaluated after 48 hours and expressed as treated/control ratios (T/C) for pairs of half-calvariae. In some experiments, effects of Carbetimer were evaluated on bones previously killed by repeated cycles of freeze-thawing.
Dead bones
Live bones 1412. IO-
.
L . .
TIC
8. 64-
01sPTH 2-
.
. ;.
.
’
i
.
t .’
_: _*_-o_,_-;. .
L
0
1
5
10
:
-s-
;
25
-
0
10 25
Carbetimer , mg / ml
Isolation and culture of immune cells We derived our isolation procedure of circulating lymphocytes and monocytes from the reports of Freundlich and Avdalovic (1983) and of Horton et al. (1974). Cells were obtained from healthy individuals who had given their written informed consent for blood withdrawal. Heparinized blood was diluted I: I with phosphate-buffered saline (PBS) and layered onto Lymphoprep@. The tubes (30 ml of diluted blood per 15 ml of Lymphoprep@) were then centrifuged at a speed progressively reaching 1000 x g for 20 min. at 18°C. Plasma was recovered, centrifuged (750 x g for 10 min at 15°C). and layered over Petri flasks 3002 F previously coated with gelatin (0.2%), 4 ml of plasma being deposited in each flask for 1 h at 37°C. During this time, the buffy coats collected at the interfaces were centrifuged at 750 x g for 10 min. at 15°C; the cells were then put together, washed with PBS, centrifuged at 150 x g for 1.5 min. to discard remaining platelets, and counted in a Neubauer cell. The plasma-treated flasks were rinsed with calcium- and magnesium-free PBS before receiving the cells. We incubated 6 x lo6 cells per flask for 1 h at 37°C in 3 ml complete medium which consisted of RPM1 1640 supplemented as above. After incubation, cells in the supematants - mainly lymphocytes - were collected and thereafter mixed with the cells obtained by two subsequent washes of the flasks. The cell viability routinely exceeded 97% (as determined by the exclusion of Trypan Blue) and the purity 95% (as determined by light microscopy and esterase staining, and confirmed by cytofluorometry). Monocytes were recovered by treating the flasks with 2mM EDTA in complete medium at 37°C (3 ml per flask for 10 min.) with regular mild shaking of the flasks. The monocyte-containing supematants were collected with a Pasteur pipet, the flasks being washed once with RPM1 1640 containing 10% FCS at 37°C. Monocytes were then put together and similarly diluted in complete medium; their viability exceeded 95% and their purity 90%. Lymphocyte and monocyte preparations were contaminated by less than 0.2% of red blood cells and less than 0.1% of platelets and polymorphonuclear cells. In some experiments, we further separated T-lympohocytes from B-lymphocytes using the method of Mage and McHugh (1985) which is based on the adsorption of B-lymphocytes on flasks (Falcon 1029 F) previously coated with immunoglobulins (goat Ig antihuman IgG and human IgG). Nonadherent T-lymphocytes were further depleted of B-lymphocytes by two consecutive incubations on other IgG-coated flasks. Purity of T-lymphocytes was between 85% and 95% (as determined by cytofluorometry using a panel of fluorescent monoclonal antibodies). Isolated immune cells were cultured in the same complete medium. Carbetimer was diluted in culture medium at various
and bone metabolism
Fig. 1. Effects of Carbetimer on ?a release from prelabelled neonatal mouse calvariae. Bone-resorbing activity was determined on live or dead bones and compared to the effects of PTH lo-‘M. Individual results are shown and are expressed as ratio of treated/control (TIC) pairs of half-calvariae. Bars represent the mean values.
concentrations and spontaneous production of TNF-LXand IFN-y was evaluated after 24 hours and 120 hours of culture, respectively. Because of large interindividual variations in spontaneous cytokine production, all results were expressed as percentages of respective control cultures without Carbetimer. Statistical methods Data are reported as means ? 1 SEM. Statistical was evaluated by two-tailed paired t-testing.
significance
Results Efects
of Carbetimer on bone resorption
Carbetimer induced a dose-dependent and significant QKO.01) increase in 45Ca release from prelabelled neonatal mouse calvariae which started at I mg/ml and reached 3.26 * 0.24 times the control values at 10 mg/ml. The effects of Carbetimer were considerably larger than the effects of PTH 10e8 M (P
J. J. Body et al.: Carbetimer
and bone metabolism
.._.._ ........-.....-. -1’
% ot control
\
,’ oioo Concentration
Fig. 2. Effects
of
Carbettmer , )g / ml
of Carbetitner on spontaneous interferon-y human T-lymphocytes (see Methods).
production
by
The indicated values represent the mean ? SE of 4 to 12 experiments for each concentration. Data are expressed as percentages of the mean control value (22.2 U/ml/2 X 10’ cells). isolated
normal
= 14: NS) and 81 2 6% (n = 38; P
141
its antineoplastic activity through immunomddulation (Searle & Co. 1988), we indeed postulated that it could stimulate TNF-a release which would be responsible for its cytostatic activity (Ruggiero et al. 1987) and, as a side effect, increase bone resorption and contribute to hypercalcemia (Konig et al. 1988). Our experiments clearly show that this was not the case. On the contrary, Carbetimer slightly inhibited TNF-a release from lymphocytes but had no evident effect on TNF-a secretion by monocytes. It is now well demonstrated that TNF-o can indeed be secreted by lymphocytes independently of TNF-a release by monocytes (Cuturi et al. 1987). We have confiied this selective effect of Carbetimer in other experiments by measuring TNF-a with an immunoradiometric assay (data not shown), but further work would be necessary to delineate the implications of these findings. These effects were not due to nonspecific cytotoxicity, and the marked inhibition of IFN-y release by relatively low concentrations of Carbetimer suggests that this drug can profoundly affect the function of T-lymphocytes. EN-y can inhibit bone resorption stimulated by various cytokines and is also able to reduce osteoclast formation (Gowen and Mundy 1986; Takahashi et al. 1986). The marked inhibition of spontaneous IFN-?/ release occurred at lower concentrations of Carbetimer than the effects on bone resorption, and it is tempting to speculate that this inhibition contributes to the long-term hypercalcemic effects of the drug. In summary, the effects of Carbetimer on calcium and bone metabolism in humans are quite complex. Acute effects consist mainly of drug-induced calcium chelation with resultant hypercalciuria and stimulation of parathyroid function (Body et al. 1989) whereas long-term effects consist of hypercalcemia that appears to be due to a direct stimulation of osteolysis and perhaps to an inhibition of EN-~ production. Acknowledgments: This work has been supported by grants from Fondation Lef&vre (Brussels, Belgium), Caisse Getterale D’Epargne et de Retraite (Belgium, Contract No. CA870102). and G.D. Searle & Co. (Chicago, USA). The authors gratefully acknowledge N. Raymakers for dedicated technical assistance and A. Collet for typing the manuscript.
Discussion These data suggest two possible explanations for the long-term hypercalcemic effects of Carbetimer in humans. It is known from animal experiments that Carbetimer binds to bone (Searle & Co. 1988) and we demonstrate here that it is a very powerful bone-resorbing agent. Bone resorption, as evaluated by 45Ca release from prelabelled mouse calvariae, was of the same magnitude on dead as on live bones and was not affected by pharmacological doses of calcitonin. These findings indicate that Carbetimer-induced bone resorption is not mediated by increased osteoclastic activity, but by direct effects of the drug on the bone matrix. The bone-resorbing effects of Carbetimer were dose-dependent and much larger than the effects of pharmacological concentrations of PTH. The lack of pharmacokinetic data on Carbetimer in humans makes it hazardous to draw precise correlations between active in vitro concentrations on bone and achieved in vivo serum or bone concentrations of the drug. Nevertheless, it is probably correct to assume that concentrations of l-10 mg/ml are therapeutically relevant given the large doses (10-15 g) administered at each Carbetimer infusion (Grunberg et al. 1987; Body et al. 1989). Although the long-term hypercalcemia induced by Carbetimer could be attributed to direct bone lysis, we have also examined another potential mechanism, namely an effect on the release of osteotropic cytokines. Since Carbetimer could exert
Body, J. J.: Magritte, A.: Cleeren. A.; Borkowski, A.: Dodion, P. Short-term effects of Carbetimer on calcium and bone metabolism in man. Eur. J. Cancer Clin. Oncol. 251831-1835; 1989. Cuturi. M. C.; Murphy, M.; Costa-Giomi, M. P.; Weinmann, R.; Perussia. B.; Trinchieri, G. Independent regulation of tumor necrosis factor and lymphotoxin production by human peripheral blood lymphocytes. J. Exp. Med. 165:1581-1594; 1987. Dodion, P.; De Valeriola. D.; Body, 1. J.; et al. Phase I clinical trial with Carbetimer. Eur. J. Cancer Clin. Oncol. 25~219-286; 1989. Franchimont, P.: Reuter, A.; Gysen, P.; Bemier, 1.; Vrindt-Gevaen, Y. Tumor Necrosis Factor-a : un m6diateur de la fonction macrophagique. Actions biologiques et dosage radio-immunologique. Med. Hyg. 452160-2168: 1987. Freundlich, B.; Avdalovic, N. Use of gelatin/plasma coated flasks for isolating human peripheral blood monocytes. J. Immunol. Method. 62:31-37; 1983. Girardin. E.; Grau. G. E.; Dayer, J. M.; Roux-Lombard, P.; Lambert, P-H.; the J5 Study Group. Tumor necrosis factor and interleukin-1 in the serum of children with severe infectious purpura. N. En@. J. Med. 319:397a; 1988. Gowen, M.; Mundy, G. R. Action of recombinant interleukin-1, interleukin-2, and interferon-y on bone resorption in vitro. J. Immunol. 136:2478-2482; 1986. Grunberg, S. M.; Ehler, E.; Mitchell, M. S. Administration of a 5-day course of Carbetimer. Pi-m. Am. Assoc. Cancer Res. 28:224; 1987. Horton, J. E.; Oppenheim, J. J.; Mergenhagen. S. E.; Raisz, L. G. Macrophagelymphocyte synergy in the production of osteoclast activating factor. J. Immunol. 113:1278-1287; 1974. Kanig, A.; Miihlbauer. R. C.; Fleisch, H. Tumor necrosis factor u and interleukin-l stimulate bone resorption in viva as measured by urinary [‘HI tetracy-
142
cline excretion from prelabeled mice. J. Bone Miner. Res. 3:621-627; 1988. Mage, M. G.; McHugh, L. L. Separation of lymphocytes on antibody-coated plates. Meth. Enzymol. 10&l 18-124; 1985. Reynolds, J. J.; Dingle. J. T. A sensitive in vitro method for studying the induction and inhibition of bone resorption. C&if. Tiss. Res. 4:339-349; 1970. Ruggiero, V.; Latham, K.; Baglioni. G. Cytostatic and cytotoxic activity of tumor necrosis factor on human cancer cells. 1. Immunol. KU:271 l-2717: 1987. Searle and Co., Skokie, Illinois. Investigational brochure on Carbetimer: An antineoplastic agent; 1988. Stem. P. H.; Krieger, N. S. Comparison of fetal rat limb bones and neonatal
J. J. Body et al.: Carbetimer and bone metabolism mouse calvaria: Effects of parathyroid hormone and 1,25-dihydroxyvitamin D,. C&if. Tissue ht. 35172-176: 1983. Takahashi. N.: Mundy, G. R; Roodman, G. D. Recombinant human interferon-y inhibits formation of human osteoclast-like cells. J. Immunol. 137:35+3549: 1986.
Dare Received:
July 24,
1990
November 13. 1990 Date Accepted: November 26, 1990 Dare Revised:
Copyright
87%3282/91 $3.00 + .OO 0 1991 Pergamon Press plc
Effects of Carbetimer, a New Antineoplastic Drug, on Bone Metabolism J. J. BODY,
S. NEJAI,
G. FERNANDEZ,
F. GLIBERT,
and G. O’BRYAN-TEAR’
Department of Medicine and Laboratory of Endocrinology and Breast Cancer Research (J.C. Heuson Laboratoq), Institut J. Border, Brussels, Belgium ‘G.D. Searle & Co., Chicago, USA Address for correspondence and reprints: J.J. Body, M.D.,
Dept. of Medicine,
Abstract
factor-Interferon
resorption-Tumor
1 rue HCger-Bordet,
1000 Bruxelles,
Belgium
immunomodulation, as suggested by the induction of cytotoxic T-cells in mice and the discordant in vitro and in vivo activities against the Lewis lung carcinoma model (Searle & Co. 1988). Carbetimer does not have the usual side effects of chemotherapeutic agents. Bone marrow depression, mucositis or alopecia are minimal or absent, whereas neurotoxicity and hypercalcemia constitute the main toxic effects. Hypercalcemia has been the dose-limiting toxicity in phase I trials and appeared to be dose- and treatment duration-dependent (Grunberg et al. 1987; Dodion et al. 1989). During the phase I trial performed in our Institution, 11 out of 26 patients developed hypercalcemia (Dodion et al. 1989). The recommended dosage from this trial was 6.5 g/m’/day using a daily x 5 schedule every three to four weeks. These long-term effects of Carbetimer on calcium metabolism sharply differ from its short-term or acute effects. We indeed demonstrated that Carbetimer infusions caused an acute decrease in ionized calcium levels and an increase in urinary calcium excretion; these findings were best explained by druginduced calcium chelation with resultant hypercalciuria (Body et al. 1989). The acute and transient hypocalcemia induced by each Carbetimer infusion could evidently not explain the longterm hypercalcemic effects of the drug, and we undertook in vitro studies to unravel the effects of Carbetimer on bone metabolism.
Carbetimer is a new antineoplastic agent whose main side effects consist of neurotoxicity and long-term dose-dependent hypercalcemia. We previously showed that Carbetlmer is a potent calcium chelator responsible for an acute decrease in ionized Ca levels observed in vivo. However, the mechanism of the progressive increase in serum Ca remains unknown. We have evaluated the bone-resorbing effects of Carbetimer on 45Ca-prelabelled neonatal mouse calvariae. Carbetimer induced a dose-dependent increase in 45Ca release which started at a concentration of 1 mg/ml and reached a mean of 3.3 times the control values at 10 mg/ml. This marked increase in 45Ca release was similar on previously killed bones and could not be inhibited by calcitonin. Such concentrations are probably therapeutically relevant given the known affinity of Carbetimer for bone and the large daily doses administered to cancer patients (l&15 g). Since Carbetimer could exert its antineoplastic action through immunomodulation, we also studied its effects on the production of TNF-a and IFN-y which are also known to affect bone metabolism. Carbetimer did not stimulate TNF-a release from isolated normal human monocytes or lymphocytes, but it markedly inhibited T-lymphocyte production of IFN-y, which became undetectable at a concentration of 1 mg of Carbetimer/ml. In summary, Carbetimer-induced hypercalcemia appears to be due to a direct stimulation of osteolysis, but possibly also to an inhibition of IFN-y production. Key Words: Carbetimer-Bone
Instutut J. Bordet,
Bone Metabolism Unit;
Methods Reagents
Carbetimer powder was provided by G.D. Searle & Co. (Chicago, USA), dissolved in water, and further diluted in the appropriate culture medium. Salmon calcitonin (sCT) was kindly given by Sandoz Co. (Base], Switzerland). Pregnant mice (EF-1) were bought from Iffa-Credo (Lyon, France), hPTH l-34 from Peninsula (Belmont, California, USA), Lymphoprep from Nycomed (Oslo, Norway), and all culture reagents from Gibco (Gent, Belgium). We measured the production of “tumor necrosis factor-a” (TNF-a) and “interferon-y” (IFN-7) by using the radioimmunoassays of Medgenix Diagnostics (Flerus, Belgium) (Franchimont et al. 1987; Girardin et al. 1988).
necrosis
y-Calcitonin.
Introduction Carbetimer is a new antineoplastic agent with an unusual spectrum of preclinical antineoplastic activity against tumor cell lines and animal tumor models. Carbetimer is a low molecular weight (1590) polymer derived from ethylene and maleic anhydride and has already demonstrated interesting antitumor properties in humans, particularly against malignant melanoma (Grunberg et al. 1987; Searle & Co. 1988; Dodion et al. 1989). However, its mechanism of action remains unclear. Besides direct antiproliferative effects, Carbetimer could act through
Bone resorption
assay
We used a bone resorption published methods (Reynolds 139
assay adapted from previously et al. 1970; Stem and Krieger
J. J. Body et al.: Carbetimer
140
1983). The assay is based on the measurement of 45Ca release by half-calvariae of neonatal mice (three to four days old) whose mothers had been injected with 0.2 mCi 45Ca two or three days before delivery. Tubes (12 x 75 mm) were placed on a roller tube apparatus (Analys, Namur, Belgium) in an incubator at 37°C. After a 24-hour preincubation in RPMI-1640 supplemented with 2% L-Glutamine (Gln) 200 mM, 10,000 U Penicillin, 10 mg streptomycin/ml, and 20% heat-inactivated serum (15% horse serum, 5% fetal calf serum), bones were incubated in fresh complete culture medium containing various concentrations of Carbetimer. Bone resorption was evaluated after 48 hours and expressed as treated/control ratios (T/C) for pairs of half-calvariae. In some experiments, effects of Carbetimer were evaluated on bones previously killed by repeated cycles of freeze-thawing.
Dead bones
Live bones 1412. IO-
.
L . .
TIC
8. 64-
01sPTH 2-
.
. ;.
.
’
i
.
t .’
_: _*_-o_,_-;. .
L
0
1
5
10
:
-s-
;
25
-
0
10 25
Carbetimer , mg / ml
Isolation and culture of immune cells We derived our isolation procedure of circulating lymphocytes and monocytes from the reports of Freundlich and Avdalovic (1983) and of Horton et al. (1974). Cells were obtained from healthy individuals who had given their written informed consent for blood withdrawal. Heparinized blood was diluted I: I with phosphate-buffered saline (PBS) and layered onto Lymphoprep@. The tubes (30 ml of diluted blood per 15 ml of Lymphoprep@) were then centrifuged at a speed progressively reaching 1000 x g for 20 min. at 18°C. Plasma was recovered, centrifuged (750 x g for 10 min at 15°C). and layered over Petri flasks 3002 F previously coated with gelatin (0.2%), 4 ml of plasma being deposited in each flask for 1 h at 37°C. During this time, the buffy coats collected at the interfaces were centrifuged at 750 x g for 10 min. at 15°C; the cells were then put together, washed with PBS, centrifuged at 150 x g for 1.5 min. to discard remaining platelets, and counted in a Neubauer cell. The plasma-treated flasks were rinsed with calcium- and magnesium-free PBS before receiving the cells. We incubated 6 x lo6 cells per flask for 1 h at 37°C in 3 ml complete medium which consisted of RPM1 1640 supplemented as above. After incubation, cells in the supematants - mainly lymphocytes - were collected and thereafter mixed with the cells obtained by two subsequent washes of the flasks. The cell viability routinely exceeded 97% (as determined by the exclusion of Trypan Blue) and the purity 95% (as determined by light microscopy and esterase staining, and confirmed by cytofluorometry). Monocytes were recovered by treating the flasks with 2mM EDTA in complete medium at 37°C (3 ml per flask for 10 min.) with regular mild shaking of the flasks. The monocyte-containing supematants were collected with a Pasteur pipet, the flasks being washed once with RPM1 1640 containing 10% FCS at 37°C. Monocytes were then put together and similarly diluted in complete medium; their viability exceeded 95% and their purity 90%. Lymphocyte and monocyte preparations were contaminated by less than 0.2% of red blood cells and less than 0.1% of platelets and polymorphonuclear cells. In some experiments, we further separated T-lympohocytes from B-lymphocytes using the method of Mage and McHugh (1985) which is based on the adsorption of B-lymphocytes on flasks (Falcon 1029 F) previously coated with immunoglobulins (goat Ig antihuman IgG and human IgG). Nonadherent T-lymphocytes were further depleted of B-lymphocytes by two consecutive incubations on other IgG-coated flasks. Purity of T-lymphocytes was between 85% and 95% (as determined by cytofluorometry using a panel of fluorescent monoclonal antibodies). Isolated immune cells were cultured in the same complete medium. Carbetimer was diluted in culture medium at various
and bone metabolism
Fig. 1. Effects of Carbetimer on ?a release from prelabelled neonatal mouse calvariae. Bone-resorbing activity was determined on live or dead bones and compared to the effects of PTH lo-‘M. Individual results are shown and are expressed as ratio of treated/control (TIC) pairs of half-calvariae. Bars represent the mean values.
concentrations and spontaneous production of TNF-LXand IFN-y was evaluated after 24 hours and 120 hours of culture, respectively. Because of large interindividual variations in spontaneous cytokine production, all results were expressed as percentages of respective control cultures without Carbetimer. Statistical methods Data are reported as means ? 1 SEM. Statistical was evaluated by two-tailed paired t-testing.
significance
Results Efects
of Carbetimer on bone resorption
Carbetimer induced a dose-dependent and significant QKO.01) increase in 45Ca release from prelabelled neonatal mouse calvariae which started at I mg/ml and reached 3.26 * 0.24 times the control values at 10 mg/ml. The effects of Carbetimer were considerably larger than the effects of PTH 10e8 M (P
J. J. Body et al.: Carbetimer
and bone metabolism
.._.._ ........-.....-. -1’
% ot control
\
,’ oioo Concentration
Fig. 2. Effects
of
Carbettmer , )g / ml
of Carbetitner on spontaneous interferon-y human T-lymphocytes (see Methods).
production
by
The indicated values represent the mean ? SE of 4 to 12 experiments for each concentration. Data are expressed as percentages of the mean control value (22.2 U/ml/2 X 10’ cells). isolated
normal
= 14: NS) and 81 2 6% (n = 38; P
141
its antineoplastic activity through immunomddulation (Searle & Co. 1988), we indeed postulated that it could stimulate TNF-a release which would be responsible for its cytostatic activity (Ruggiero et al. 1987) and, as a side effect, increase bone resorption and contribute to hypercalcemia (Konig et al. 1988). Our experiments clearly show that this was not the case. On the contrary, Carbetimer slightly inhibited TNF-a release from lymphocytes but had no evident effect on TNF-a secretion by monocytes. It is now well demonstrated that TNF-o can indeed be secreted by lymphocytes independently of TNF-a release by monocytes (Cuturi et al. 1987). We have confiied this selective effect of Carbetimer in other experiments by measuring TNF-a with an immunoradiometric assay (data not shown), but further work would be necessary to delineate the implications of these findings. These effects were not due to nonspecific cytotoxicity, and the marked inhibition of IFN-y release by relatively low concentrations of Carbetimer suggests that this drug can profoundly affect the function of T-lymphocytes. EN-y can inhibit bone resorption stimulated by various cytokines and is also able to reduce osteoclast formation (Gowen and Mundy 1986; Takahashi et al. 1986). The marked inhibition of spontaneous IFN-?/ release occurred at lower concentrations of Carbetimer than the effects on bone resorption, and it is tempting to speculate that this inhibition contributes to the long-term hypercalcemic effects of the drug. In summary, the effects of Carbetimer on calcium and bone metabolism in humans are quite complex. Acute effects consist mainly of drug-induced calcium chelation with resultant hypercalciuria and stimulation of parathyroid function (Body et al. 1989) whereas long-term effects consist of hypercalcemia that appears to be due to a direct stimulation of osteolysis and perhaps to an inhibition of EN-~ production. Acknowledgments: This work has been supported by grants from Fondation Lef&vre (Brussels, Belgium), Caisse Getterale D’Epargne et de Retraite (Belgium, Contract No. CA870102). and G.D. Searle & Co. (Chicago, USA). The authors gratefully acknowledge N. Raymakers for dedicated technical assistance and A. Collet for typing the manuscript.
Discussion These data suggest two possible explanations for the long-term hypercalcemic effects of Carbetimer in humans. It is known from animal experiments that Carbetimer binds to bone (Searle & Co. 1988) and we demonstrate here that it is a very powerful bone-resorbing agent. Bone resorption, as evaluated by 45Ca release from prelabelled mouse calvariae, was of the same magnitude on dead as on live bones and was not affected by pharmacological doses of calcitonin. These findings indicate that Carbetimer-induced bone resorption is not mediated by increased osteoclastic activity, but by direct effects of the drug on the bone matrix. The bone-resorbing effects of Carbetimer were dose-dependent and much larger than the effects of pharmacological concentrations of PTH. The lack of pharmacokinetic data on Carbetimer in humans makes it hazardous to draw precise correlations between active in vitro concentrations on bone and achieved in vivo serum or bone concentrations of the drug. Nevertheless, it is probably correct to assume that concentrations of l-10 mg/ml are therapeutically relevant given the large doses (10-15 g) administered at each Carbetimer infusion (Grunberg et al. 1987; Body et al. 1989). Although the long-term hypercalcemia induced by Carbetimer could be attributed to direct bone lysis, we have also examined another potential mechanism, namely an effect on the release of osteotropic cytokines. Since Carbetimer could exert
Body, J. J.: Magritte, A.: Cleeren. A.; Borkowski, A.: Dodion, P. Short-term effects of Carbetimer on calcium and bone metabolism in man. Eur. J. Cancer Clin. Oncol. 251831-1835; 1989. Cuturi. M. C.; Murphy, M.; Costa-Giomi, M. P.; Weinmann, R.; Perussia. B.; Trinchieri, G. Independent regulation of tumor necrosis factor and lymphotoxin production by human peripheral blood lymphocytes. J. Exp. Med. 165:1581-1594; 1987. Dodion, P.; De Valeriola. D.; Body, 1. J.; et al. Phase I clinical trial with Carbetimer. Eur. J. Cancer Clin. Oncol. 25~219-286; 1989. Franchimont, P.: Reuter, A.; Gysen, P.; Bemier, 1.; Vrindt-Gevaen, Y. Tumor Necrosis Factor-a : un m6diateur de la fonction macrophagique. Actions biologiques et dosage radio-immunologique. Med. Hyg. 452160-2168: 1987. Freundlich, B.; Avdalovic, N. Use of gelatin/plasma coated flasks for isolating human peripheral blood monocytes. J. Immunol. Method. 62:31-37; 1983. Girardin. E.; Grau. G. E.; Dayer, J. M.; Roux-Lombard, P.; Lambert, P-H.; the J5 Study Group. Tumor necrosis factor and interleukin-1 in the serum of children with severe infectious purpura. N. En@. J. Med. 319:397a; 1988. Gowen, M.; Mundy, G. R. Action of recombinant interleukin-1, interleukin-2, and interferon-y on bone resorption in vitro. J. Immunol. 136:2478-2482; 1986. Grunberg, S. M.; Ehler, E.; Mitchell, M. S. Administration of a 5-day course of Carbetimer. Pi-m. Am. Assoc. Cancer Res. 28:224; 1987. Horton, J. E.; Oppenheim, J. J.; Mergenhagen. S. E.; Raisz, L. G. Macrophagelymphocyte synergy in the production of osteoclast activating factor. J. Immunol. 113:1278-1287; 1974. Kanig, A.; Miihlbauer. R. C.; Fleisch, H. Tumor necrosis factor u and interleukin-l stimulate bone resorption in viva as measured by urinary [‘HI tetracy-
142
cline excretion from prelabeled mice. J. Bone Miner. Res. 3:621-627; 1988. Mage, M. G.; McHugh, L. L. Separation of lymphocytes on antibody-coated plates. Meth. Enzymol. 10&l 18-124; 1985. Reynolds, J. J.; Dingle. J. T. A sensitive in vitro method for studying the induction and inhibition of bone resorption. C&if. Tiss. Res. 4:339-349; 1970. Ruggiero, V.; Latham, K.; Baglioni. G. Cytostatic and cytotoxic activity of tumor necrosis factor on human cancer cells. 1. Immunol. KU:271 l-2717: 1987. Searle and Co., Skokie, Illinois. Investigational brochure on Carbetimer: An antineoplastic agent; 1988. Stem. P. H.; Krieger, N. S. Comparison of fetal rat limb bones and neonatal
J. J. Body et al.: Carbetimer and bone metabolism mouse calvaria: Effects of parathyroid hormone and 1,25-dihydroxyvitamin D,. C&if. Tissue ht. 35172-176: 1983. Takahashi. N.: Mundy, G. R; Roodman, G. D. Recombinant human interferon-y inhibits formation of human osteoclast-like cells. J. Immunol. 137:35+3549: 1986.
Dare Received:
July 24,
1990
November 13. 1990 Date Accepted: November 26, 1990 Dare Revised: