In: Handbook of Metal-Ligand Interactions in Biological Fluids - Bioinorganic Medicine, Volume 2, Ed: G. Berthon, Marcel Dekker, Inc., New York, pages 1253 - 1265.
Raymond P. Warrell, Jr., M.D.
Memorial Sloan-Kettering Cancer Center,
Cornell University Medical College,
1275 York Avenue,
New York, NY 10021
Elemental gallium is a potent inhibitor of bone resorption that acts to maintain or restore bone mass. By virtue of these biologic effects, gallium compounds are potentially useful treatments for a variety of diseases that are characterized by accelerated bone loss, including cancer-related hypercalcemia, bone metastases, Paget's disease, and post-menopausal osteoporosis. This chapter provides an overview of the laboratory and clinical studies that underlie these observations.
After the anticancer activity of platinum coordination compounds was recognized in the late 1960's (Rosenberget al., 1969), a variety of elements were screened by the U.S. National Cancer Institute for potential antineoplastic uses. All of the group IIIa elements, including boron, aluminum, gallium, and thallium, exhibited cytotoxic effects; however, gallium was the most active and least toxic of these agents when tested against experimental animal tumors (Hart etal., 1971; Hart and Adamson, 1971). Having satisfactorily completed animal toxicologic testing, in 1976 gallium nitrate entered clinical studies in human subjects as a method of cancer treatment. By 1992, more than 1,200 patients with advanced cancer had been treated with relatively high doses of this drug in numerous investigations. Prominent side-effects from these high doses included nausea, vomiting, and kidney damage. However, with the exceptions of certain patients with malignant lymphoma (Warrell et al.,1983) and bladder cancer (Foster et al., 1986), the drug has not exhibited significant anticancer activity against human tumors and the agent is currently not being used for this purpose.
In these early studies,several investigators reported that patients frequently experienced a reductionin their serum calcium concentration (hypocalcemia) subsequent to the administration of gallium nitrate (Bedikian et al.,1978; Krakoff et al., 1979; Warrell et al., 1983). One report indicated that gallium treatment increased urinary excretion of calcium in rodents (Newman et al., 1979). Subsequent studies by our group, however, showed that hypocalcemia was in fact associated with a marked reduction in urinary calcium excretion (Warrell et al., 1985). This discrepancy suggested to us that in fact the major effect of gallium might be inhibition of calcium release from the skeleton, and further studies have confirmed this speculation.
III. PRECLINICAL STUDIES
Gallium has certain physicochemical properties that are relevant to bone physiology. Elemental gallium precipitates in the presence of phosphate at pH ^ 5.0, forming various metal/cation complexes of gallium-phosphate. Gallium also readily adsorbs to hydroxyapatite, the predominant calcium-phosphate species in mineralized bone. Early studies indicated that radiolabeled gallium was rapidly incorporated into osteogenic foci (Dudley and Maddox, 1949) and within cells in metabolically active tissues (Anghileri, 1971). As discussed below, a variety of experiments have been performed to further clarify the various effects of gallium on bone.
A. Localization in bone
Therapeutic doses of gallium result in trace levels (i.e. parts per million) accumulating in bone (Bockman et al., 1986; Repo etal., 1988). The quantification and precise localization of gallium in bone tissue has posed significant technical challenges; however, satisfactory spatial resolution of gallium has been achieved using synchrotron-generated X-ray microscopy. This method entails the use of a densely collimated beam of high energy X-rays that enables detection of various elements by X-ray fluorescence analysis. In one experiment, rats were treated with the equivalent of 7 mg elemental gallium over 14 days, and contour maps of gallium distribution in long bones were generated, as illustrated in Figure 1. From such maps, the highest gallium content was measured in the metaphysis, notably in the epiphyseal region (Bockmanet al., 1990). By contrast, extremely low levels of gallium were measured in mid-cortical regions which is only slowly remodeled with a low rate of calcium turnover. The skeletal localization of gallium was associated with important effects on both the mineral and protein matrix components of bone tissue.
B. Effects on bone-mineral
Measurements of bone powder from gallium-treated rats using X-ray diffraction analysis showed narrowing of the absorption peak on the long axis of the hydroxyapatitecrystals compared to controls. These changes are characteristic of an increase in the size or crystalline perfection of hydroxyapatite mineral. Infrared spectroscopy further suggested that bone carbonate content was decreased (Bockman et al., 1986). The physicochemical consequence of these alterations was a significant decrease in the solubility of gallium-treated bone mineral in acidic buffer solutions (Repo et al., 1988).
In later studies, bone particles taken from the metaphyseal regions of gallium-treated rats showed a shift to increased density, along with an increase in calcium and phosphorus content (Repo et al., 1988). Moreover, injection of radiolabeled calcium into gallium-treated rats demonstrated significantly higher accretion of new calcium into bone relative to control animals. These findings strongly indicate that gallium treatment enhanced mineralization of newly forming bone rather than simply acting to decrease physiologic resorption.
C. Effects on bone cells
A number of studies using bone explant models have clearly documented the capacity of gallium to prevent osteolysis. Fetal rat bones labeled in utero with 45Ca displayed a dose-dependent inhibition of calcium release over a broad range of gallium concentrations (10-200 microM). These effects were independent of the anion to which elemental gallium was attached (i.e. citrate, nitrate, and acetate salts were equally effective). Gallium was also effective in blocking bone resorption induced by parathyroid hormone (Warrell et al., 1984) as well as certain bone-resorbing cytokines such as tumor necrosis factor (Bockman et al., 1987a).
In other studies, bone particles from rats treated in vivo with gallium nitrate were then transplanted subcutaneously into other animals. Over a three-week period, bone-resorbing mononuclear cells were attracted to the bone particles and eventually effected their dissolution. A significant delay in resorption was seen when bone taken from gallium-treated rats was compared to bone from untreated controls. Histological sections of bone explants from gallium-treated rats have shown normal-appearing osteoclasts that are normally apposed to bone (Fig. 2A). This result is in striking contrast to the cytotoxic effects observed with certain other agents such as platinum and mithramycin (plicamycin) (Fig. 2B) (Bockman, 1991). Over the therapeutic range of concentrations previously noted, gallium did not alter DNA or general protein synthesis in bone explants or osteoblast cell lines. However, gallium significantly enhanced the synthesis and content of Type-I collagen (Bockman et al., 1987b), the major protein found in skeletal matrix.
One of the most important effects of gallium nitrate was recently found by Blair and associates (1992). Acidification of mineralized matrix is a principal mechanism whereby osteoclasts effect dissolution of bone. Using an inverted osteoclast membrane model, gallium nitrate was found to significantly impair ATPase-dependent hydrogen ion transport, thereby inhibiting bone resorption.
Since subtle changes in mineral or matrix could result in decreased tensile strength, standardized biomechanical tests of bone strength were performed on bone from gallium-treated rodents who received 27-33 mg of elemental gallium for periods ranging from 9 to 11 weeks. In these tests, no decrease in bone strength was observed compared to untreated control animals (Adelman et al., 1989).
To summarize these preclinical studies, elemental gallium preferentially accumulates in metabolically active regions of bone that undergo rapid resorption and formation. Gallium induces changes in the physical properties of newly synthesized bone mineral that renders the mineral less susceptible to dissolution. Gallium also appears to increase new calcium accretion into bone and to increase bone collagen synthesis. The drug is a potent inhibitor of cell-mediated bone resorption, probably by interfering with the acidification process of osteoclasts. Importantly, these effects occur at concentrations that are not cytotoxic to normal bone cells. These dual effects -to decrease resorption and stimulate new bone formation- probably account for the very high potency of this agent that has been observed in clinical studies.
IV. CLINICAL STUDIES
A. Human Pharmacokinetics
Following parenteral injection, gallium binds to a variety of plasma proteins, particularly the iron-bonding protein transferrin (Clausen et al., 1974). Considerable variability in plasma half-life after injection has been reported depending on the duration of the initial injection (Krakoff et al., 1979; Hall et al., 1979; Kelsen et al., 1980). The initial half-life in plasma (T1/2alpha) is approximately 1 hour and the T1/2beta is approximately 24 hours. The terminal half-life is highly dependent on the method of administration. When administered by prolonged intravenous infusion or repeated subcutaneous injection, the plasma half-life increases significantly (Warrell et al., 1987a). This change undoubtedly reflects binding to and slow release from a deep tissue compartment (presumably bone). The calculated volume of distribution (VDSS) is also quite large for similar reasons. Skeletal tissue has been shown to reversibly accumulate elemental gallium in a dose-dependent manner (Anghileri, 1971; Bockman et al., 1986). Release of bone-bound gallium occurs quite slowly and measurable concentrations of gallium have been found in human plasma several months following cessation of therapy.
The major route of excretion is via the kidney and like other metal-based pharmaceuticals, this organ is a principal site of toxicity after treatment with high doses. However, the mechanism of nephrotoxicity is substantially different. For example, platinum is directly cytotoxic to brush border cells of the renal tubule; renal damage from platinum compounds tends to be cumulative and irreversible. By contrast, kidney damage from gallium is associated with intraluminal precipitation of gallium-calcium-phosphate complexes within the renal tubule - an effect that is generally preventable and reversible by increasing urinary flow to avoid solute precipitation. Renal clearance of the drug is not affected by hydration and diuresis (Krakoff et al., 1979; Newman et al., 1979; Foster et al., 1986).
B. Cancer-related hypercalcemia
The clinical condition that is associated with the most rapid loss of calcium from bone tissue is cancer-related hypercalcemia. In this disorder, factors released from cancer cells cause a marked increase in osteoclastic bone resorption and calcium exits bone so rapidly that it overwhelms renal excretory capacity. Serum calcium concentration may then increase to levels that become acutely life-threatening. In the initial study (Warrell et al., 1984), gallium nitrate was given to 10 hypercalcemic patients with various tumor types. The drug was administered as a continuous intravenous infusion at a daily dose of 200 mg/m2 for periods that ranged from 5-7 days. All patients responded with a reduction in serum calcium to normal or sub-normal concentrations. These results were confirmed in a larger dose-ranging study in which 36 infusions were administered to 31 patients. A clear dose-response correlation was observed in that patients treated with the higher dose (200 mg/m2/day) achieved a superior rate of normocalcemic control compared to the lower dose (100 mg/m2/day), and the response was sustained for a substantially longer period of time (Warrell et al., 1986). Notably, the effective dose for the treatment of hypercalcemia is less than 40% of the usual antitumor dose.
In 1985, the first randomized double-blind study to compare active methods of treatment in patients with cancer-related hypercalcemia was initiated. In that study, hospitalized patients resistant to 2 days of parenteral hydration were randomized to receive either gallium nitrate or high doses of calcitonin, a hormone used for treatment of bone-resorptive disorders. The effects of both drugs on mean daily serum calcium measurements are depicted in Figure 3A. Overall, 75% of patients who received gallium nitrate achieved normocalcemia compared to only 27% of patients who received calcitonin (P < 0.0006) (Warrell et al., 1988). It should be noted that response criteria in this study were quite stringent (i.e. intent-to-treat analysis, no exclusions, and adjustments of all serum calcium levels for serum albumin concentrations). If conventional, less conservative criteria were employed such as have been used in other studies (including no adjustment of serum calcium and exclusion of early deaths), the normocalcemic response to gallium nitrate would have equaled 100% compared to 69% of patients treated with calcitonin (p < 0.001).
Duration of normocalcemia is difficult to assess given multiple confounding factors in critically ill patients, especially the subsequent use of cytotoxic or other potentially hypocalcemic drugs. If time to recurrence is censored at the time the serum calcium was first above the normal range or other treatment was administered (a very conservative method of analysis), the mean duration of normocalcemia was 6 days for patients treated with gallium nitrate and 1 day for patients treated with calcitonin ( < 0.001). If these effects are not censored from the analysis, the duration of normocalcemia was 13+ days for patients treated with gallium nitrate and 2 days for patients treated with calcitonin.
Recent studies have shown that epidermoid (squamous) carcinomas are particularly associated with elevated plasma levels of a protein with biochemical activities identical to parathyroid hormone (PTH) -the so-called "PTH-related protein" (Budayr et al., 1989; Burtis et al., 1990). In a previous study, gallium nitrate was shown to he highly effective for treatment of hypercalcemia in patients with parathyroid carcinoma, a virulent hypercalcemic syndrome caused by grossly elevated serum levels of FTH (Warrell et al., 1987a). Prior to initiation of the gallium/calcitonin study, patients were stratified by tumor histology since patients with hypercalcemia due to epidermoid carcinomas were believed more likely to have a resistant, "humorally-mediated" hypercalcemia. Patients with epidermoid carcinoma who received calcitonin in that study fared especially poorly, and only 1 of 10 such patients achieved a normal serum calcium However, the response to gallium nitrate was independent of histology, and equal proportions (75%) of patients with both epidermoid and non-epidermoid tumor types responded. These studies have particular importance since the current generation of bisphosphonates appear to be significantly less effective for hypercalcemic syndromes mediated by the PTH-related protein (Thiebaud et al., 1990).
A recent study has shown exceptional therapeutic superiority of gallium nitrate compared to etidronate (Warrell et al., 1991), one of a group of drugs known as bisphosphonates. The relative effects of these two drugs on mean daily serum calcium measurements in this study are shown in Figure 3B. Overall, 85% of patients who received gallium responded compared to 42% of patients who received etidronate. The duration of the effect was again significantly longer for gallium-treated patients. A follow-up study that compares the effectiveness of gallium to yet another bisphosphonate (pamidronate; APD) is underway in an international collaborative trial. Additional clinical studies are evaluating dosing regimens of shorter duration and low-dose schedules of single-day subcutaneous injections that can be administered chronically to prevent recurrence of hypercalcemia (Warrell, 1991). These clinical studies have confirmed in vitro observations that gallium antagonizes bone resorption irrespective of mechanism or cancer type. At present, it appears that gallium nitrate is the most effective drug in clinical use for the treatment of cancer-related hypercalcemia.
C. Studies in bone metastases
The concept of using certain medical therapies as adjuncts to traditional anticancer treatment in order to strengthen bone tissue against erosion from metastases has gamed increased credence. Ideally, such therapy should not only minimize further bone loss but also restore bone that has been previously eroded. Other drugs, such as bisphosphonates, fluorides and calcitonin, have previously been proposed for this use. However, the dual actions of gallium to both decrease bone resorption and enhance new bone formation has suggested that this agent might be exceptionally useful for preservation or restoration of bone that has been destroyed due to cancer.
Accelerated bone turnover in patients with bone metastases is commonly associated with increased urinary excretion of calcium and hydroxyproline. In a preliminary study, gallium nitrate was administered by a continuous intravenous infusion to 22 patients with lytic bone metastases. Urinary calcium excretion was significantly reduced in 21 of 22 patients, and urinary hydroxyproline excretion was significantly lowered in patients with high basal levels of excretion (Warrell et al., 1987b). Similar findings have been observed in patients with prostatic cancer with sclerotic (osteoblastic) bone metastases (Scher et al., 1987). These results document that therapy with gallium acutely lowers biochemical parameters indicative of accelerated bone turnover. Several key issues are yet unresolved including the optimal effective dose and schedule for long-term treatment and whether extended treatment will substantially ameliorate the morbid effects of bone metastases, such as pain, pathologic fracture, and hypercalcemia. Methodologic difficulties inherent in performing such studies have been discussed elsewhere (Warrell and Bockman, 1989).
In the United States, research has focused on multiple myeloma and breast cancer -the former since it is the prototypic osteolytic disease in cancer, and the latter since it is the most prevalent osteolytic condition. In the myeloma study, patients who had achieved a "plateau" in their disease with chemotherapy were randomized to receive gallium nitrate or no additional treatment for 6 months along with their chemotherapy. Patients randomized to no gallium were crossed over after 6 months. The principal outcome determinant was the measurement of total-body calcium content using a highly sensitive technique (neutron activation analysis) (Lovett et al., 1989). Patients who received gallium nitrate showed a significant increase in total-body calcium relative to untreated controls. The increase was usually (though not uniformly) observed after 6 months of treatment. Most patients also experienced substantial relief of bone pain while receiving gallium. Clinically apparent pathologic fractures, along with one episode of hypercalcemia, were observed only in patients who were initially randomized not to receive gallium treatment (Warrell et al., 1989). Some of these patients with myeloma have remained on intermittent therapy with gallium nitrate for periods in excess of 4 years. These results, though still preliminary, suggest that prolonged treatment with gallium nitrate can be effective in reducing the morbidity associated with bone metastases.
A follow-up study has been initiated in patients with osteolytic metastases from breast cancer whereby patients receiving standard systemic treatment are randomized to several different dose-schedules of gallium nitrate (administered as subcutaneous injections) or to no additional therapy. In addition to conventional measures of pain and fracture, 3-dimensional quantitative computed tomography is being evaluated in that study as a method of quantifying focal changes in lytic bone destruction.
D. Studies in metabolic bone diseases
Paget's disease afflicts up to 5% of the adult population older than age 40. Although the disease may remain quiescent and require no specific treatment in many individuals, a substantial proportion of patients experience bone pain, skeletal deformation, fractures, and impaired mobility. Similar to hypercalcemia, conventional medical treatment has consisted of calcitonin injections or bisphosphonates. In view of the superiority of gallium nitrate relative to both of these agents, an initial study was undertaken in patients with advanced Paget's disease that was resistant to conventional treatment. Similar to the clinical trial in bone metastases, the goal of this study was to evaluate whether a brief course of treatment with gallium nitrate could reduce biochemical parameters of accelerated bone turnover. Five patients were entered into each of three dose-schedules: 2.5 mg/kg/day (100 mg/m2/d) by continuous intravenous infusion for 7 days; 0.5 mg/kg/day for 14 days by subcutaneous injection; and 0.25 mg/kg/day for 14 days by subcutaneous injection. Reductions in serum alkaline phosphatase and urinary hydroxyproline excretion were observed after treatment with each dose-schedule; the median maximum decreases in serum alkaline phosphatase activity were 49%, 39%, and 18%, respectively. A typical response pattern is shown in Figure 4. The median maximum decreases in urinary hydroxyproline excretion were 50%, 52%, and 16%. The maximum decrease in urinary hydroxyproline excretion occurred within a median of 2 weeks from the start of treatment, whereas the maximum decrease in serum alkaline phosphatase activity occurred substantially later at a median time of 6 weeks. Response duration ranged from 6-42 weeks (Warrell et al., 1990). These results suggest significant activity for this drug in Paget's disease and further studies have been initiated to test prolonged treatment as a means of inducing extended complete remission. The favorable responses observed in Paget's disease, as well as other disorders characterized by rapid bone loss, suggest that gallium-containing compounds will be effective treatment for less extreme disorders, particularly osteoporosis - a condition in which loss of bone mass extends over many years.
Gallium nitrate appears to be a uniquely acting agent for treatment of bone-resorptive diseases. Although this activity was found incidentally to its original therapeutic purpose, extended study using considerably lower doses indicates that this agent is a safe and highly effective method of reducing accelerated bone loss in patients with cancer and metabolic bone disease.
Adelman, R.J., Jones, K.W., Donnelly, R., Wright, T.M., and Bockman, R.S. (1989). Trans. Orthoped. Res. Soc., 15: 415.
Anghileri, L. (1971). Strahlentherapie, 142: 456.
Bedikian, A.Y., Valdivieso, M., Bodey, G.P., Burgess, M.A., Benjamin, R.S., Hall, S., and Freireich, E.J. (1978). Cancer Treat. Rep., 62: 1449.
Blair, H.C., Teitelbaum, S.L., Tan, H.L., and Schlesinger, P.H. (1992). J. Cell. Biochem., 48: 401.
Bockman, R.S. (1991). Semin. Oncol., 18 (Suppl. 5): 21.
Bockman, R.S., Boskey, A.L., Blumenthal, N.C., Alcock, N.W., and Warrell, R.P., Jr. (1986). Calcif. Tissue Int., 39: 376.
Bockman, R.S., Repo, M.A., Warrell, R.P., Jr., Israel, R., and Gabrilove, J. (1987a). Proc. Amer. Assoc. Cancer Res., 28: 449.
Bockman, R.S., Israel, R., Alcock, N.W., Ferguson, R., and Warrell, R.P., Jr. (1987b). Clin. Res., 35: 620A.
Bockman, R.S., Repo, M.A., Warrell, R.P., Jr., Pounds, J.G., Schidlovsky, G., Gordon, B.M., and Jones, K.W. (1990). Proc. Nat. Acad. Sci. USA, 87: 4149.
Budayr, A.A., Nissenson, R.A., Klein, R.F., Pun, K.K., Clark, O.H., Diep, D., Arnaud, C.D., and Strewler, G.J. (1989). Ann. Int. Med., Ill: 807.
Burtis, W.J., Brady, T.G., Orloff, J.J., Ersbak, J.B., Warrell, R.P., Jr., Olson, B.R., Wu, T.L., Mitnick, M.A.E., Broadus, A.E., and Stewart, A.F. (1990). N. Engl. J. Med., 322: 1106.
Clausen, J., Edeling, C.J., and Fogh, I. (1974). Cancer Res., 34: 1931. Dudley, H.C., and Maddox, G.E. (1949). J. Pharmacol. Exp. Therap., 96: 224.
Foster, B.J., Clagett-Carr, K., Hoth, D., and Leyland-Jones, B. (1986). Cancer Treat. Rep., 70: 1311.
Hall, S.W., Yeung, K., Benjamin, R.S., Stewart, D., Valdivieso, M., Bedikian, A.Y., and Loo, T.L. (1979). Clin. Pharmacol. Therap., 25: 82.
Hart, M.M., and Adamson, R.H. (1971). Proc. Nat. Acad. Sci. USA, 68: 1623.
Hart, M.M., Smith, C.F., Yancey, S.T., and Adamson, R.H. (1971). J. Nat. Cancer Inst., 47: 1121.
Kelsen, D.P., Alcock, N., Yeh, S., Brown, J., and Young, C.W. (1980), Cancer, 46: 2009.
Krakoff, I.H., Newman, R.A., and Goldberg, R.S. (1979). Cancer, 44: 1722.
Lovett, D., Dilmanian, F.A., Heelan, R., Gnecco, C., and Warrell, R.P., Jr. (1989). Blood, suppl. 74: 377a.
Matkovic, V., Apseloff, G., Shepard, D.R., and Gerber, N. (1990). Lancet, 335: 12.
Newman, R.A., Brody, A.R., and Krakoff, I.H. (1979). Cancer, 44: 1728.
Repo, M.A., Bockman, R.S., Betts, F., Boskey, A.L., and Warrell, R.P., Jr. (1988). Calcif. Tissue Int., 43: 300.
Rosenberg, B., Vancarnp, L., and Trosko, J.E. (1969). Nature, 222: 385.
Schenk, R., Merz, W.A., Muhlbauer, R., Russell, R.G.E., and Fleisch, H. (1973). Calcif. Tissue Res., 11: 196.
Scher, H.I., Curley, T., Geller, N., Dershaw, D., Chan, E., Nisselbaum, J., Alcock, N., Hollander, P., and Yagoda, A. (1987). Cancer Treat. Rep., 71: 887.
Thiebaud, D., Jaeger, P., and Burckhardt, P. (1990). J. Bone Mineral Res., 5: 221. Warrell, R.P. (1991). Semin Oncol., 18 (suppl. 5): 26.
Warrell, R.P., and Bockman, R.S. (1989). Important Advances in Oncology (V.T. DeVita, Jr., S. Hellman and S.A. Rosenberg, eds.), J.B. Lippincott, New York, p. 205.
Warrell, R.P., Coonley, C.J., Straus, D.J., and Young, C.W. (1983). Cancer (Phila.), 51: 1982.
Warrell, R.P., Bockman, R.S., Coonley, C.J., Isaacs, M., and Staszewski, H. (1984). J. Clin. Invest., 73: 1487.
Warrell, R.P., Isaacs, M., Coonley, C.J., Alcock, N.W., and Bockman, R.S. (1985). Cancer Treat. Rep., 69: 653.
Warrell, R.P., Skelos, A., Alcock, N.W., and Bockman, R.S. (1986). Cancer Res., 46:
Warrell, R.P., Isaacs, M., Alcock, N.W., and Bockman, R.S. (1987a). Ann. Int. Med., 107: 683.
Warrell, R.P., Alcock, N.W., and Bockman, R.S. (1987b). J. Clin. Oncol., 5: 292.