BISPHOSPHONATES ARE POTENT INHIBITORS of normal and. A Dose-Finding Study of Zoledronate in Hypercalcemic Cancer Patients

Transcription

1 JOURNAL OF BONE AND MINERAL RESEARCH Volume 14, Number 9, 1999 Blackwell Science, Inc American Society for Bone and Mineral Research A Dose-Finding Study of Zoledronate in Hypercalcemic Cancer Patients J.J. BODY, 1 A. LORTHOLARY, 2 G. ROMIEU, 3 A.M. VIGNERON, 4 and J. FORD 4 ABSTRACT Zoledronate is a new heterocyclic imidazole bisphosphonate that is the most potent bisphosphonate administered in humans because it is times more potent than pamidronate, according to in vitro or animal models of bone resorption. We conducted an open-label, dose-finding, single-dose phase I study in tumor-induced hypercalcemia (TIH), which has been similarly used as a model to determine the active doses of other bisphosphonates. The primary objective was to determine, with a dose escalation schedule, two nontoxic dose levels of zoledronate able to induce normocalcemia in at least 80% of patients with TIH after rehydration (corrected Ca for albumin levels 2.75 mmol/l). Based on estimates of potency, the starting dose was mg/kg, and further tested doses were 0.005, 0.01, 0.02, and 0.04 mg/kg. To obtain a more precise estimate of the response rate, we treated 10 more patients at the highest of the two effective dose levels. The median infusion time of zoledronate was 30 minutes. Thirty out of the 33 treated patients were evaluable for efficacy. Thirty percent of the patients had breast cancer and 54% had metastatic bone involvement. For all groups combined, mean Ca levels at baseline was 3.0 mmol/l. The two effective dose levels were 0.02 mg/kg and 0.04 mg/kg. Five out of five patients became normocalcemic after 0.02 mg of zoledronate/kg and 14 out of 15 after 0.04 mg of zoledronate/kg. The success rate of the latter dose was thus 93% (95% confidence interval [CI] %). At this dose, the first day of normocalcemia was day 2 or 3 for all but one patient. The duration of normocalcemia for the two effective doses could be assessed in nine patients; seven patients remained normocalcemic throughout the trial (32 39 days). The fall in serum Ca was accompanied by a marked fall in fasting urinary Ca excretion. Zoledronate was well tolerated: 7 out of 33 patients developed transient hypophosphatemia, and 3 developed transient hypocalcemia. The only clinically detectable side effect was an increase in body temperature occurring in 10 (30%) patients. In summary, very low doses of zoledronate (0.02 mg/kg and 0.04 mg/kg, i.e., 1.2 mg and 2.4 mg for a 60-kg individual, respectively) administered by a short-time infusion effectively treated patients with TIH. The fall in serum Ca was rapid, and normocalcemia was often maintained for several weeks. Zoledronate was well tolerated. Future trials will determine whether prolonged treatment with this potent compound can have greater effects on the skeletal morbidity rate in patients with tumor bone disease than can be achieved with currently available bisphosphonates. (J Bone Miner Res 1999;14: ) INTRODUCTION BISPHOSPHONATES ARE POTENT INHIBITORS of normal and pathological bone resorption. They decrease osteoclast activity and/or number by various mechanisms, including changes in the cytoskeleton, disappearance of the ruffled border, apoptosis induction, decreased acid extrusion, and changes in enzyme activity. (1) Bisphosphonates have become standard therapy for tumor-induced hypercalcemia (TIH) and a new form of medical therapy for bone metastases. (2) The prolonged administration of oral or intravenous bisphosphonates can reduce the frequency of morbid skeletal events by one third to one half in patients with breast cancer metastatic to bone and in patients with mul- 1 Supportive Care Clinic & Bone Diseases Clinic, Institut J. Bordet, Université Libre de Bruxelles, Brussels, Belgium. 2 Centre Paul Papin, Angers, France. 3 Centre Val d Aurelle, Montpellier, France. 4 Novartis, Rueil-Malmaison, France and Basle, Switzerland. 1557

2 1558 tiple myeloma. (2 5) Newer more potent bisphosphonates may simplify current therapeutic schemes but could be especially important for improving these already impressive results. Zoledronate is a new heterocyclic imidazole bisphosphonate which is so far the most potent bisphosphonate administered in humans. In an in vitro bone resorption assay using mouse calvariae, zoledronate is at least 100 times more active than pamidronate. In the in vivo animal model of calcitriol-induced hypercalcemia in thyroparathyroidectomized rats, zoledronate is 850 times more active than pamidronate and more than 4 orders of magnitude more potent than clodronate. (6) Of all bisphosphonates clinically evaluated, zoledronate has the largest therapeutic ratio between the desired inhibition of bone resorption and the unwanted inhibition of bone mineralization. Various toxicology studies also showed that the compound is safe. These preclinical data formed the rationale to conduct a phase I trial in human subjects. Since there were no data on the active doses in humans, we conducted an open-label, dose-finding, single-dose phase I study in TIH, which has previously been used as a human model to determine the active doses of the bisphosphonate ibandronate. (7) MATERIALS AND METHODS Study design The trial was an open, multicenter, dose-finding phase I study. The trial protocol was approved by the local ethics committees of all participating institutions. The primary objective was to determine, with a dose escalation schedule, two nontoxic dose levels of zoledronate able to induce normocalcemia in four or five out of five patients with TIH. To obtain a more precise estimate of the response rate, we treated 10 more patients at the highest of the two effective dose levels. Each patient received a single infusion and was studied on only one occasion, i.e., no patient did receive different doses of zoledronate. Patients were followed daily for 1 week, then weekly for 4 more weeks. A clinical examination and laboratory tests were performed at each visit. Blood tests included serum calcium (Ca), phosphate, electrolytes, creatinine, albumin, complete blood count with differential count and liver function. Fasting urine albumin, blood, and casts were checked and the Ca/creatinine ratio (UCa) was measured in second-void specimens. Inclusion criteria included the persistence of TIH (defined as a Ca concentration 2.75 mmol/l or 11.0 mg/dl) after 24 h of intravenous rehydration, consisting of at least 2 3 l of saline. Serum Ca was corrected for albumin levels, according to the following formula: corrected Ca (mmol/l) measured Ca (mmol/l) albumin (g/l) We initially included only patients whose Ca levels were between 2.75 mmol/l and 3.25 mmol/l. However, when the first effective dose level was determined, the upper limit was increased to 3.75 mmol/l. Hypercalcemia was not a chronic condition in any patient. Exclusion criteria included serum creatinine >1.5 times the upper limit of the laboratory normal value after rehydration, previous treatment with bisphosphonates within 6 months, treatment with any drug known to influence Ca levels within 1 month preceding zoledronate therapy, hypercalcemia as a flare reaction within 2 weeks of starting endocrine therapy, and treatment with chemotherapy during the 7 days before zoledronate therapy. The treatment allocation procedure was centralized. Based on estimates of potency, the starting dose was mg/kg, and further tested doses were 0.005, 0.01, 0.02, and 0.04 mg/kg. Each dose level was first tested in two patients. If Ca levels increased or decreased by <10% compared with the postrehydration value after 72 h following treatment, the next higher dose level was tested, but if Ca decreased by 10% in one or two patients, a third patient was included at this dose level. If normocalcemia (defined as a corrected Ca level 2.60 mmol/l) was obtained by 7 days in only one out of three patients, this dose level was considered ineffective. If normocalcemia was obtained by 7 days in two or three out of three patients, two other patients received the same dose level. When normocalcemia was then obtained by 7 days in four or five out of five patients, that dose of zoledronate was considered effective, and the next higher dose level was tested in five additional patients. Patients were removed from the trial (failure of therapy) when normocalcemia was not obtained by 7 days. To be considered evaluable, patients could not receive chemotherapy before normocalcemia was reached; similarly, hormonotherapy could not be started or modified and any drug potentially able to modify Ca levels could not be started during the trial. Zoledronate was diluted in 5% dextrose and infused over minutes (median infusion time, 30 minutes). Statistics Analyses were carried out using SAS (Statistical Analysis System, Cary, NC, U.S.A.) version We used descriptive tests with no formal statistical comparisons between the dose groups. Results are presented as mean ± SEM or median and range when indicated. Geometric means were used for UCa. A 95% confidence interval (CI) was given for the observed proportion of successes based on serum Ca (corrected for protein) levels. RESULTS Patients characteristics BODY ET AL. Thirty out of the 33 treated patients were evaluable for efficacy. Three patients were indeed ineligible; one patient was not hypercalcemic according to the entry criteria for the trial and two were in the terminal phase of cancer and died 1 day after therapy. The characteristics of the 30 evaluable patients are summarized in Table 1. Thirty percent of the patients had breast cancer and, of the patients with solid tumors, 54% had metastatic bone involvement. For the five treatment groups, the mean corrected Ca levels at baseline were 2.97, 3.30, 3.49, 3.14, and 3.17 mmol/l, respectively. For all groups combined, the mean Ca level was 3.0 mmol/l

3 ZOLEDRONATE FOR CANCER HYPERCALCEMIA 1559 Characteristics TABLE 1. PATIENTS CHARACTERISTICS BY TREATMENT GROUP (n = 3)* (n =3) 0.01 (n =4) Dose in mg/kg 0.02 (n =5) 0.04 (n = 15) All doses (n = 30) Age (mean) (Range) (years) (40 56) (41 47) (44 82) (55 70) (32 85) (32 85) Gender female male Tumor types breast other gynecology digestive lung 3 3 hematological others Bone metastases * n, number of patients. Only considering patients with solid tumors (n 26). (median 3.16, range mmol/l). The mean weight of the patients was 60.3 kg. Efficacy Three out of 10 patients became transiently normocalcemic at the three lowest dose levels (one at each dose step). The effective dose levels were 0.02 mg/kg and 0.04 mg/kg, thus 1.2 mg and 2.4 mg total dose for the average 60 kg individual. Five out of five patients became normocalcemic after 0.02 mg of zoledronate/kg, and 14 out of 15 after 0.04 mg of zoledronate/kg. The success rate of the latter dose was thus 93%, with 95% CI of %. The fall in serum Ca concentrations (mean ± SEM) for the two effective dose levels is shown in the upper panel of Fig. 1. At the dose of 0.04 mg/kg, the first day of normocalcemia was day 2 or 3 for all but one patient. The duration of normocalcemia for the two effective doses could not be assessed in 10 of the 19 patients who reached a normal serum Ca level because of early death or loss to follow-up. Of the nine evaluable patients, only two relapses were documented, with seven patients remaining normocalcemic throughout the trial (32 39 days). As shown in the lower panel of Fig. 1, there was a marked fall in fasting urinary Ca excretion. At these two effective dose levels, mean UCa levels dropped dramatically within 2 or 3 days and then stabilized in the normal range. The evaluation has been stopped at day 7 for the dose of 0.02 mg/kg because there were only one or two samples at 14 days and 21 days. Tolerability Safety was evaluated in all 33 treated patients. Zoledronate was well tolerated. Seven patients developed transient hypophosphatemia and three developed transient hypocalcemia. This occurred mainly with the highest dose level of zoledronate; of the 15 patients treated by 0.04 mg of zoledronate/kg, 6 developed transient hypophosphatemia and 3 developed transient hypocalcemia. Hypomagnesemia was observed in only one case. The only clinically detectable side effect was an increase in body temperature, occurring in 10 patients (30%). Seven episodes were observed after 0.04 mg of zoledronate/kg, but fever was generally moderate and transient. One patient had an increase in liver function tests from day 3 to day 21 after 0.04 mg of zoledronate/kg. DISCUSSION This study demonstrates that very low doses of zoledronate (0.02 mg/kg and 0.04 mg/kg, i.e., 1.2 mg and 2.4 mg for a 60 kg individual) administered as a short infusion constitute an effective therapy for TIH. The success rate of the dose of 0.04 mg of zoledronate/kg was 93%, with 95% CI of %. The fall in serum Ca was rapid since the day of normalization of corrected serum Ca was day 2 or 3 for most patients. Moreover, normocalcemia was often maintained for several weeks, and the fact that seven patients at the two effective dose levels remained normocalcemic throughout the trial suggests that zoledronate could have quite prolonged effects. These first data should, however, be confirmed in a larger series of patients with more severe TIH before stating that zoledronate acts faster and has a longer duration of action than other bisphosphonates in this condition. However, parathyroid hormone-related protein or other bone-resorbing cytokines were not measured in this study and we cannot thus address their possible influence on the response to zoledronate therapy. Pamidronate at a dose of 90 mg corrects TIH in about 90% of unselected hypercalcemic cancer patients, and the median duration of normocalcemia is 3 4 weeks. (2,8,9) A dose escalation trial with ibandronate was conducted in 147 patients with Ca levels 3.0 mmol/l after rehydration, 125

4 1560 FIG. 1. Effects of intravenous zoledronate in 20 hypercalcemic cancer patients after saline rehydration, at a dose of 0.02 mg/kg (open circles; n 5) and 0.04 mg/kg (filled squares; n 15). (Upper panel) Effects on serum Ca levels corrected for albumin. (Lower panel) Effects on urinary fasting Ca excretion. The reference ranges are indicated by the hatched bars. Data are shown as the mean ± SEM. of whom were evaluable for response. The success rate was 50% in the 2 mg group, which was significantly lower than the responses in the 4 mg and 6 mg dose groups (76% and 77%, respectively), but the duration of response was not significantly influenced by dose. (7) Because patient characteristics vary between studies, prospective and randomized face-to-face comparisons between these compounds need to be performed before meaningful conclusions can be drawn. (2) Zoledronate was safely administered as a convenient 30- minute infusion, and the compound was well tolerated. Transient hypocalcemia and hypophosphatemia are expected adverse events of a potent bisphosphonate. As with aminobisphosphonates, fever was observed in 30% of all cases, but the cause of mild fever is often difficult to assess in patients with advanced cancer, and the true frequency of zoledronate-induced fever will have to be assessed in other patient populations. Alterations in liver function tests were observed in one patient, but caution is necessary before stating that zoledronate was the responsible factor. This dose response study should be especially interesting for the design of future trials in normocalcemic patients with bone metastases in whom it is more difficult to determine the adequate dose to be employed. Preliminary data indicate that intravenous doses of zoledronate of up to 8 mg are safe, inhibit bone resorption, and decrease bone pain. (10,11) However, dose response effects were not evident and, as suggested by our trial, it remains to be demonstrated that such high doses are more effective than lower doses. Bisphosphonates represent a major therapeutic advance in the management of the skeletal morbidity caused by metastatic breast cancer and multiple myeloma. (2) Intravenous bisphosphonates exert clinically relevant analgesic effects in patients with metastatic bone pain. Regular pamidronate infusions can also enhance the objective response rate to chemotherapy. In breast cancer, the prolonged administration of oral clodronate reduces the frequency of morbid skeletal events by more than one fourth, (12) whereas monthly pamidronate infusions for 1 or 2 years in addition to chemotherapy reduce by more than one third the frequency of all skeletal-related events. (3) Moreover, bisphosphonates in addition to chemotherapy are superior to chemotherapy alone in patients with stage III multiple myeloma and they can reduce the skeletal morbidity rate by about one half. (4,5) Last, recent data indicate that the prolonged administration of bisphosphonates after breast cancer surgery could reduce the incidence of bone metastases and prolong survival. (13) Future trials will have to determine if treatment with more potent bisphosphonates, such as zoledronate, can have more pronounced effects on the skeletal morbidity rate than can be currently achieved with available compounds in patients with cancer-related bone disease. ACKNOWLEDGMENTS We gratefully acknowledge the dedicated help of Mrs. M.A. Lumen for conducting the trial in Brussels and of Mrs. A. Lebec in France. This work was partly supported by Novartis and by grants from Fondation Médic and Fonds de la Recherche Scientifique Médicale (Convention No ). REFERENCES BODY ET AL. 1. Fleisch H 1997 Bisphosphonates in Bone Disease From the Laboratory to the Patient, 3rd Ed. The Parthenon Publishing Group, London, U.K. 2. Body JJ, Bartl R, Burckhardt P, Delmas PD, Diel IJ, Fleisch H, Kanis JA, Kyle RA, Mundy GR, Paterson AHG, Rubens RD, for the International Bone and Cancer Study Group 1998 Current use of bisphosphonates in oncology. J Clin Oncol 16: Hortobagyi GN, Theriault RL, Lipton A, Porter L, Blayney D, Sinoff C, Wheeler H, Simeone JF, Seaman JJ, Knight RD, Heffernan M, Mellars K, Reitsma DJ, for the Protocol 19 Are-

5 ZOLEDRONATE FOR CANCER HYPERCALCEMIA 1561 dia Breast Cancer Study Group 1998 Long-term prevention of skeletal complications of metastatic breast cancer with pamidronate. J Clin Oncol 16: Berenson JR, Lichtenstein A, Porter L, Dimopoulous A, Bordoni R, George S, Lipton A, Keller A, Ballester O, Kovacs M, Blacklock H, Bell R, Simeone JF, Reitsma DJ, Heffernan M, Seaman J, Knight RD, for the Myeloma Aredia Study Group 1998 Long-term pamidronate treatment of advanced multiple myeloma patients reduces skeletal events. J Clin Oncol 16: McCloskey EV, MacLennan ICM, Drayson M, Chapman C, Dunn J, Kanis JA 1998 A randomized trial of the effect of clodronate on skeletal morbidity in multiple myeloma. Br J Hematol 100: Green JR, Müller K, Jaeggi KA 1994 Preclinical pharmacology of CGP , a new, potent, heterocyclic bisphosphonate compound. J Bone Miner Res 9: Ralston SH, Thiébaud D, Herrmann Z, Steinhauer EU, Thürlimann B, Walls J, Lichinitser MR, Rizzoli R, Hagberg H, Huss HJ, Tubiana-Hulin M, Body JJ 1997 Dose response study of ibandronate in treatment of cancer-associated hypercalcaemia. Br J Cancer 75: Purohit OP, Radstone CR, Anthony C, Kanis JA, Coleman RE 1995 A randomised double-blind comparison of intravenous pamidronate and clodronate in the hypercalcaemia of malignancy. Br J Cancer 72: Body JJ, Dumon JC 1994 Treatment of tumor-induced hypercalcaemia with the bisphosphonate pamidronate: Dose response relationship and influence of the tumour type. Ann Oncol 5: Berenson JR, Lipton A, Rosen LS, et al Phase I clinical study of a new bisphosphonate, zoledronate (CGP-42446), in patients with osteolytic bone metastases. Blood 88 (Suppl 1):586a. 11. Body JJ 1997 Clinical research update: Zoledronate. Cancer 80: Paterson AHG, Powles TJ, Kanis JA, McCloskey E, Hanson J, Ashley S 1993 Double-blind controlled trial of oral clodronate in patients with bone metastases from breast cancer. J Clin Oncol 11: Diel IJ, Solomayer EF, Costa SD, Gollan C, Goerner R, Wallwiener D, Kaufmann M, Bastert G 1998 Reduction in new metastases in breast cancer with adjuvant clodronate treatment. N Engl J Med 339: Address reprint requests to: J.J. Body, M.D., Ph.D. Institut Jules Bordet Rue Héger-Bordet Brussels, Belgium Received in original form November 11, 1998; in revised form March 19, 1999; accepted April 6, 1999.