Vol. 80. No. 2 Journalof CbnicalEndocrinology and Pr’rrnrrd ,,, IJ S.A. CopyrIght0 1995by The EndocrineSociety

A Multicenter Trial of Low Dose in Patients with Advanced Paget’s Disease of

RICHARD S. BOCKMAN, FRANCOIS WILHELM, ETHEL SIRIS, FREDERICK SINGER, ARTHUR CHAUSMER*, RACHELLE BITTON, JON KOTLER, BARBARA J. BOSCO, DAVID R. EYRE, AND DAVID LEVENSON Department of Medicine, Hospital for Special Surgery and Cornell University Medical College (R.S.B., B.J.B., D.L.), New York, New York 10021; Research and Development, Fujisawa Pharmaceutical Co. (F.W.), Deer-field, Illinois 60015; Columbia University College of Physicians and Surgeons (E.S.), New York, New York 10032; St. John’s Hospital and Health Center, John Wayne Cancer Institute (F.S.1, Santa Monica, California 90404; the Division of Endocrinology, Long Island Jewish Hospital (R.B.1, New Hyde Park, New York 11042; Holy Cross Hospital (J.K.), Ft. Lauderdale, Florida 33008; and the Department of Orthopaedics, University of Washington (D.R.E.), Seattle, Washington 98195

ABSTRACT treated with the 0.5 mg/kg.day dose achieved a 50% or more reduction Gallium nitrate is a potent antiresorptive drug that has been ex- in enzyme activity. The nadir value in hydraxyproline excretion oc- tensively tested in patients with accelerated bone turnover. We have curred at 10 weeks, with mean changes of +9%, -lo%, and ~ 17% for evaluated the effects of this new agent in a pilot multicenter trial of the 0.05, 0.25, and 0.5 mg/kg.day doses, respectively; the difference 49 patients with advanced Paget’s disease of bone. Patients were was significant only at the 0.5 mg/kg.day level (P < 0.01). Urinary randomized to receive 0.05, 0.25, or 0.5 mg/kg.day gallium nitrate collagen cross-link excretion showed a significant decrease at the 0.25 administered by SCinjection in two 14-day cycles. Serum alkaline and 0.5 mg/kg.day doses. We also observed a definite, but nonsignif- phosphatase, fasting 2-h urinary hydroxyproline and N- telopeptide icant, trend for improved quality of life in patients treated at the collagen cross-links excretion, and quality of life were assessed every highest drug dose. Minor discomfort at the injection site was fre- 2 weeks for 12 weeks. The group mean alkaline phosphatase activity quently reported, but did not lead to interruption of therapy. Our at baseline was 854 2 100 (t SEM) IUL The mean changes from results in these patients who had received moderate to extensive prior baseline to week 12 in serum alkaline phosphatase were +0.5%‘, therapies with other drugs show that cyclical, low dose, SCadminis- ~24%, and -31%, respectively, for the three doses tested. The dif- tration of gallium nitrate is safe and effective for treating patients ferences for each of the higher dose levels (0.25 and 0.5 mg/kg.day) with advanced Paget’s disease of bone. (J Clin Endocrinol Metab 80: was statistically significant (P 5 0.05), and nearly half of the patients 595-602, 1995)

AGET’S disease of bone is characterized by abnormally osteoclastic activity have proven useful for the treatment of P accelerated bone remodeling that is driven by excessive active Paget’s disease, including antiresorptive agents, such osteoclastic resorption (1, 2). The affected are fre- as (7, 8, 15-19) and bisphosphonates (10-14, quently deformed and more susceptible to fracture due to 20-36), as well as agents that are cytotoxic for , their abnormal structure and architecture. The extent and such as plicamycin (mithramycin) (37). activity of Paget’s disease correlate with serum alkaline phos- Gallium nitrate is a potent new antiresorptive drug that phatase levels (a measure of osteoblastic activity) as well as has been approved for the treatment of cancer-related hy- urinary excretion of hydroxyproline (an index of bone matrix percalcemia (38). When administered iv at moderate doses resorption) (3,4). Serum alkaline phosphatase is considered (200 mg/m*.day or -5 mg/kg.day), several randomized a reliable biochemical index of disease activity, with a low double blind studies have shown highly significant superi- coefficient of variation within a single patient (5); however, ority of this drug relative to calcitonin (38) and bisphospho- changes in alkaline phosphatase often take weeks to occur nates (39,40). Additional studies have shown that the agent after effective therapy has been initiated (6-13). The coeffi- markedly reduces the biochemical parameters of accelerated cient of variation within patients for hydroxyproline is much bone turnover in patients with cancerous metastases (41) and greater than that for alkaline phosphatase, making hy- increases mineral content in patients with multiple myeloma droxyproline a less sensitive marker with which to assess (42). In vitro studies would suggest that gallium nitrate acts treatment response (5, 14). Although the etiology of this directly on osteoclasts to inhibit and that this disorder is unknown, drugs that are capable of suppressing action is not a consequence of a cytotoxic effect on bone cells (43, 44). Received May 20, 1994. Revision received October 7, 1994. Accepted Pilot studies have evaluated both moderate dose iv infu- October 17, 1994. sions (2.5 mg/kg.day) and low dose SCinjections (0.5 and 0.25 Address all correspondence and requests for reprints to: Richard S. mg/kg.day) of gallium nitrate in heavily pretreated and Bockman, M.D., Ph.D., Hospital for Special Surgery, 535 East 70th Street, drug-resistant patients with advanced Paget’s disease New York, New York 10021. * Current address: Veterans Administration Medical Center, 510 East (45, 46). These early studies showed a marked decrease in Stoner, Shreveport, Louisiana 71101. both serum alkaline phosphatase levels and urinary

595

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010 BOCKMAN ET AL. JCE & M . 1995 Vol80 . No 2

hydroxyproline excretion, with suggestive evidence of a was separated, frozen at -70 C, and sent to a central reference labora- dose-response relationship. The potent antiresorptive effects tory. For the urine specimens, the total volumes were measured, and an aliquot was frozen and sent to the reference laboratories. of this drug, its safety, and the favorable biochemical re- sponse to these reduced doses in pilot studies prompted us to conduct the first multicenter randomized trial of this agent Quality of life assessment in patients with advanced Paget’s disease of bone. At baseline, patients underwent a medical history and physical ex- amination. Each patient gave a subjective assessment of “pain experi- enced now,” “pain experienced during the last week,” physical activity, Subjects and Methods and quality of life. The first three parameters were measured by a loo-mm visual analog scale. Quality of life was assessed by completing Study centers a questionnaire regarding interference with seven specific activities (i.e. general activity, mood, walking ability, normal work, relationship to Between September 1, 1991, and April 20, 1992, 49 patients were others, sleep, enjoyment of life, and global quality of life). enrolled in a randomized prospective study at 6 sites. Patients with a known history of Paget’s disease underwent radionuclide bone scanning and radiographs of affected skeletal areas. Entry criteria included a Statistical analyses serum alkaline phosphatase level at least twice the upper limit of normal. In addition, patients could not have taken calcitonin or mithramycin For baseline characteristics, comparisons between treatment groups within the preceding 8 weeks, nor any bisphosphonate within the pre- were performed by analysis of variance (ANOVA) for parametric sta- ceding 6 months. None of the patients was taking other medications that tistics and 2 analyses for categorical data. For alkaline phosphatase and affected skeletal metabolism during the 12-week treatment phase. If a urinary hydroxyproline, the percent change from baseline at each mea- patient had been previously treated with any antiresorptive medication, sured time point was calculated for each patient. Comparisons were their serum alkaline phosphatase level had to have returned to their performed using ANOVA, with week number and treatment dose as the previous baseline value before entry. main variables. Intergroup and weekly comparisons were made using the Student-Newman-Keuls and pairwise t tests (least squared differ- ence) on the means, respectively. Additional analyses were performed, Baseline calibration studies with dose and site as independent variables. Changes in hemoglobin levels were analyzed by calculating the means and percentage of pa- Two-hour fasting urine specimens were collected daily for 5 days in tients with a decrease greater than 1.5 mg/dL. Statistical analyses were a subset of patients (12) from a single study site. This short 5-day interval performed with pairwise t and 2 tests, respectively. was selected to minimize the possible variability of disease activity. Bone pain and quality of life assessments were performed using Patients were instructed to fast after 2000 h and upon waking the next questionnaires developed for patients with rheumatological diseases. morning after a 12-h fast to discard the first voided urine specimen, then These methods of assessing subjective responses to disease state or to drink 24 oz water. Urine was collected over the first 2 h after waking. treatment efficacy have not been validated in patients with Paget’s Urinary volume, , and hydroxyproline were measured by the disease. Continuous variables (visual analog scales of bone pain and usual clinical laboratory procedures (47). Fasting 2-h samples show a physical activity) were compared using ANOVA, with week number high correlation with 24-h collections taken under strict dietary restric- and treatment dose as the main effects. Between-treatment groups and tions (48). Collagen cross-link determinations were made on the fasting between-week comparisons were made using the Student-Newman- urine samples by measuring N-telopeptides containing pyridinoline Keuls and pairwise t tests on the means, respectively. For categorical residues derived from bone collagen, assayed by enzyme-linked im- variables (general activity, mood, walking ability, normal work, rela- munosorbent assay using a mouse monoclonal antibody-based assay tions with other people, sleep, and enjoyment of life), contingency table (Ostex International, Seattle, WA) that had high affinity for an epitope analyses of pretreatment vs. each week’s data were performed. Descrip- specific for degraded human bone collagen (49). Baseline blood samples tive tabulations of each variable over time by treatment group were also were obtained at the beginning and end of the calibration studies. performed. All t tests were two-sided. The study protocol was approved by the institutional review boards Treatment schedule at the participating institutions. Written informed consent was obtained from all patients. This study was a randomized, open labeled, parallel group compar- ison of three different dosages of gallium nitrate. Although there was no placebo arm, we anticipated that the lowest dose (0.05 mg/kg.day) Results might be a “no effect” dose. Patients were randomized to receive gallium nitrate at 0.05 mg/kg.day (group A, 17 patients), 0.25 mg/kg.day (group Patient characteristics B, 17 patients), or 0.50 mg/kg.day (group C, 15 patients). Patients in- jected themselves SC, preferably in the inner thigh, once daily in the Characteristics for each group of patients are shown in morning. The medication was taken for 14 days, followed by 4 weeks off Table 1. Mean age, sex, race, height, and weight did not differ medication. This cycle was repeated once to complete a 12-week study significantly among the 3 groups. Two patients in group A, period. Follow-up measurements of efficacy and safety were obtained beyond the 12-week treatment phase only in the calibration subset of 7 in group B, and 9 in group C had received prior anti-Pagetic patients previously described. therapy; thus, with regard to prior therapy, group A (0.05 mg/kg.day) had significantly less prior treatment than Laboratory tests groups B and C (P < 0.025). Previously treated patients (37% of the total study group) had received calcitonin (14 patients), Serum alkaline phosphatase and 2-h morning fasting urine collections etidronate (6 patients), or risedronate (1 patient). The extent for hydroxyproline and creatinine determinations were obtained on two separate occasions at least 7 days apart. The calculated average of these and duration of disease among the three groups were similar. two baseline values served as the pretreatment value. A complete blood Baseline alkaline phosphatase, urinary hydroxyproline to count, serum , iron-binding capacity, transferrin, ferritin, biochem- creatinine ratio, and the subjective parameters of pain, phys- ical profile, electrolytes, and a 24-h urine collection for , hy- ical activity, and quality of life are shown in Table 2. There droxyproline, and creatinine were obtained every 2 weeks for 12 weeks. were no statistically significant differences among the groups The calibration subset of patients previously described had blood and urine sampling every 4 or 8 weeks for up to 52 weeks. For serum in any of the categories. The proportions of very severely specimens, blood was allowed to clot at room temperature; the serum affected patients (i.e. those with alkaline phosphatase levels

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010 GALLIUM NITRATE IN PAGET’S DISEASE 597

TABLE 1. Patient characteristics

Treatment group (mg gallium nitrate/kg. day) A B C 0.05 (n = 17) 0.25 (n = 17) 0.50 (n = 15)

Age [mean + SD (range)] 70.0 k 10.2 (44-83) 67.6 -c 7.4 (51-77) 68.4 2 7.0 (55-83) Sex (% male) 59% 53% 67% Race % Caucasian 82 94 60 % Black 18 6 33 % Asian 0 0 7 Ht (cm) 169 2 13.2 (147-188) 163 + 10.6 (145-178) 167 ? 9.0 (146-179) Wt (kg) 76.6 ? 20.1 (47.5-104) 74.6 k 18.7 (46.3-110) 74.6 -t 12.8 (54.7-96.7) Prior anti-Pagetic treatment (no.) 2 7 9 Monostotic disease (n) 3 1 Duration of disease (yr) 16 i- l&l-41) 18 ? 8 (2-33) 17 2 10 (4-46)

TABLE 2. Baseline biochemical values

A 0.05 B 0.25 c 0.50 Alkaline phosphatase in serum (III/L) 808.6 2 432.8(227-1625) 824.7 t 568.1(266-2202) 813 k 1037.1 (228-4445) No. >5x normal 9 8 8 Urinary hydroxyproline/creatinine 19.4 2 15.9 (6.9-66.7) 15.6 i- 9.8 (4.6-32.4) 19.1 ? 15.8 (5.4-56.7) Pain assessed now (o-100) 37.4 k 39.6 36.4 2 32.8 23.3 k 26.8 Pain last week (O-100) 36.4 -t 34.2 46.5 t 27.9 36.6 f 32.1 Physical activity (o-100) 24.2 t 26.0 26.9 t 24.8 34.6 -t 34.8 Quality of life (O-70) 21.1 + 15.1 24.9 t 18.8 21.9 -c 20.6

Values are the mean 2 SD; the range is in parentheses.

>5 times the upper limit of normal) were similar in the three The mean percent changes from baseline as a function of groups. ANOVAs with center and dose as the main variables time for serum alkaline phosphatase for the three dose showed no significant “center” effect for the data collected. groups are shown in Fig. 1. At the end of 12 weeks, group A had a mean increase of 0.5% (the change was not significant Biochemical parameters compared to baseline; P < 0.9), group B had a mean decrease of 24% (P < 0.05 compared to baseline and P < 0.02 compared A subgroup analysis of 12 patients allowed us to evaluate to group A), and group C had a mean decrease of 31% the reproducibility of the hydroxyproline measurements and (P < 0.05 compared to baseline and P < 0.001 compared to compare them to the newer collagen cross-link N-telopeptide group A). Log transformation of the data before statistical values. These patients had demographics similar to those of analyses produced no difference in results. the total patient population. Five consecutive daily 2-h fast- ing urine samples were collected from 12 patients (5 females and 7 males) before the initiation of drug therapy. The daily Serum Alkaline Phosphatase coefficient of variation for urinary hydroxyproline (milli- % Variation from Baseline grams per 2 h) measurements varied from lO.l-55.1%, with an average of 24.3%. When urinary hydroxyproline excretion % change from baseline Alk Phos was normalized for creatinine, the coefficient of variation for IOr urinary hydroxyproline/creatinine ratios varied from 2.8- 14.6%, with an average of 8.1%. The correlation between O3 serum alkaline phosphatase IU/L (average of 2 baseline mea- surements) and the mean value of the 5 consecutive mea- surements of hydroxyproline/creatinine ratios was highly significant (r = 0.89; P < 0.01). Within the calibration subgroup, the coefficient of varia- -40 tion for the collagen cross-linked N-telopeptides reported as 0 2 4 6 a 10 12 collagen type I equivalents (BCE) in picomoles per pmol Week Number creatinine excreted at baseline was 10.5%, similar to that for - 0.05 mglhglday 4- 0.25 mglhglday ++ 0.50 mg/kg/day the hydroxyproline/creatinine ratio. The correlation be- tween collagen cross-linked peptide levels in the urine and a-Pc0.05 b-Pc0.01 serum alkaline phosphatase was significant (r = 0.81; FIG. 1. Percent change in serum alkaline phosphatase activity dur- P < O.Ol), as was the correlation of collagen cross-linked ing the trial. The group receiving 0.05 mg/‘kg.day (n = 17) gallium peptides with urinary hydroxyproline/creatinine ratios nitrate showed no change. The groups receiving 0.25 mg/kg.day (n = 17) and 0.50 mg/kg.day (n = 15) showed significant reductions as (r = 0.97; P < 0.01). Based on the measured levels of hy- early as weeks 4 and 2, respectively. The time periods for the two droxyproline and bone cross-linked peptides, all patients courses of gallium nitrate treatment are indicated by the cross- demonstrated increased bone turnover. hatched bars along the x-axis. a, P < 0.05; b, P < 0.01.

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010 598 BOCKMAN ET AL. JCE & M . 1995 Vol 80 . No 2

Two-hour fasting urinary hydroxyproline/creatinine ra- tios for the total study groups are shown in Fig. 2A. At the Quality of Life Assessment end of 12 weeks, group A had a mean increase of 9%, and During the Last 2 Weeks group B had a mean decrease of 10% (neither change was Degree of interference (O-100) significant). Group C had a mean decrease of 17% (P < 0.1 4or compared to baseline and P < 0.05 compared to group A). Note that in group C, the hydroxyproline/creatinine ratios at weeks 2, 4, 6, 8, and 10 showed significant reductions compared to baseline (P < 0.05 for each). N-Telopeptide levels in the fasting 2-h urine samples for the calibration subset of patients are shown in Fig. 2B. At the 2-week sample point, the decreases in collagen cross-linked peptide levels were significant for groups B and C compared 4 6 6 10 12 Week Number

Urinary OH-Proline/Creat Ratio - 0.05 mg/kg/day - 0.25 mg/kg/day ++ 0.50 mg/kg/day % Variation from Baseline FIG. 3. Quality of life, as assessed by questionnaire. Seven areas of % change from baseline OH-ProXreat “interference with daily living” were explored. The scale is from O-100, and a lower score suggests less interference. The trend for improved quality of life is significant in the 0.50 mgikg,day group (P < 0.05).

to the 5-day average of the baseline values 154% (P < 0.001) and 48% (P < 0.031, respectively]. These differences in groups B and C compared to baseline remained significant through a b weeks 12 and 16 for groups B and C, respectively. No sig- nificant change from baseline was observed for group A. The

0 2 4 6 6 10 12 changes in urinary hydroxyproline/creatinine ratios corre- Week Number lated significantly with the changes in serum alkaline phos- phatase levels for groups B and C (r = 0.72 and 0.71, re- - 0.05 mglkgldsy - 0.25 mglkglday - 0.50 mglkglday spectively; P < 0.05 for each). Mean urinary hydroxyproline/ a-Pt0.05 b-PcO.01 A creatinine ratios returned to baseline by week 20 for the high and middle (group C and B) dose patients (data not shown). Collagen Cross-linked Peptide Levels % Variation from Baseline Quality of life

% change from baseline BCE/Creat Patient responses for intensity of current pain, pain over 20 r the last week, and ability to perform physical activity showed no statistically significant trends. Quality of life response data for the entire study population are presented in Fig. 3 and are reported as interference with daily activity (i.e. a lower score suggests less interference). Each data point re- flects the average over a 2-week period, and a data point was collected every 2 weeks. Thus, it is presented as a trend. The -60 trend of improvement in quality of life for group C reaches -60 statistical significance at week 12 (P < 0.05). 0 2 4 6 0 10 12 14 16 Week Number Adverse effects - 0.05 mglkglday -0.25 mglkglday - 0.50 mglkglday The mean hemoglobin values showed a trend for a reduc- a-PcO.05 b-PcO.01 B tion over time in groups B and C, with a nadir at week 10 of FIG. 2. A, Percent change in urinary hydroxyproline/creatinine ra- 1.2 g/dL below the baseline (P < 0.0001; Fig. 4). Zero percent tios during the trial. The groups receiving 0.05 mg/kg.day (n = 17) and of group A, 35% of group B, and 40% of group C had a fall 0.25 mg/kg.day (n = 17) gallium nitrate had no significant changes from baseline. The group receiving 0.50 mg/kg.day had significant in hemoglobin of greater than 1.5 g/dL at some time during reductions starting from week 2. a, P < 0.05; b, P < 0.01. B, Percent the 12-week study. However, only two patients (both in change in fasting urinary bone collagen cross-linked peptide levels group C) had a decrease in hemoglobin of greater than 2 normalized to creatinine in 12 patients during a treatment protocol g/dL. In all cases, the nadir in hemoglobin occurred during with gallium nitrate. Data are given as the mean change from base- line (100% value) for each of the dose groups studied [n = 4 for groups the eighth or tenth week, which was 2-4 weeks after the B and C (0.25 and 0.5 mg/kg.day respectively); n = 5 for group A (0.05 second administration of gallium. In all cases but one, the mg/kg.day)l. a, P < 0.05; b, P < 0.01. hemoglobin level was within 1.5 g/dL of baseline by the

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010 GALLIUM NITRATE IN PAGET’S DISEASE 599

treatment. This method of analysis would tend to bias the Hemoglobin Concentration data toward a less favorable outcome. Changes During Treatment

Hemoglobin (gram/d11 Discussion In this pilot randomized study, we again found that low dose gallium nitrate given by intermittent SC injections caused a significant reduction in biochemical parameters of Paget’s disease activity. More than half of the patients in this trial had baseline serum alkaline phosphatase levels that were greater than 5 times the upper limit of normal. Such patients seldom normalize their alkaline phosphatase levels 2 4 6 6 10 12 regardless of therapy (14, 19, 27, 31, 50, 51). Furthermore, Week Number many of the patients in this study had received prior therapy with other agents that might have rendered them more re- - 0.05 mglkglday 4- 0.25 mg/hglday - 0.50 mglkglday sistant to subsequent therapy (26,52). Despite the severity of FIG. 4. Mean hemoglobin concentrations (normal range, 11.5-16 their disease, the patients given the higher doses (0.25 and 0.5 g/dL in women and 13-17 g/dL in men). The nadir appeared during the eighth and tenth weeks, shortly after the second course of gallium. mg/kg.day; levels that are only 5-10% of the FDA-approved dose for cancer-related hypercalcemia) showed significant twelfth week of the study. No consistent changes in hemat- decreases in their biochemical parameters after two brief ocrit, blood cell counts, serum iron, ferritin, or transferrin courses of therapy. were seen. There was a significant reduction in the iron- The lowest dose of gallium nitrate that was tested (0.05 binding capacity at 2-8 weeks for the group C patients, mg/kg.day) had little or no effect on the measured biochem- which did not correlate with hemoglobin or changes in other ical or clinical parameters. For the other two doses tested (0.25 and 0.5 mg/kg.day), a rapid decline in bone resorptive blood parameters. Mean serum creatinine levels for the three parameters occurred shortly after initiating therapy, fol- dose schedules are shown in Fig. 5. Measurement of serum creatinine levels showed that 6% of group A patients, 12% of lowed by a slow, but persistent, fall in serum alkaline phos- group B, and 7% of group C patients had a rise in creatinine phatase activity. Similar uncoupling of osteoclastic/osteo- of greater than 0.2 mg/dL over baseline (data not shown). blastic activities had been seen in previous studies with gallium nitrate (45,46) as well as with other antiresorptives, The largest increment was 0.4 mg/dL, which occurred in two patients. In one of these patients, the peak value occurred such as alendronate (35), pamidronate (13, 27-29, 31, 32), etidronate (20,53), clodronate (11,26), and calcitonin (54,55). during the second week, but had decreased to 0.1 mg/dL below baseline by the end of the study. In the other patient, Isolated studies involving etidronate (56) and calcitonins (19, the increase in creatinine occurred at 12 weeks and was 57) failed to find this uncoupling. associated with urinary obstruction due to prostatic enlarge- The pattern of biochemical responses to treatment, char- ment. After relief of the obstruction, creatinine and creatinine acterized by a rapid fall in hydroxyproline and N-telopeptide clearance returned to baseline. Various subjective complaints levels, suggests that gallium nitrate works primarily by in- hibiting osteoclastic activity. This probable mechanism of were elicited as the patients returned for the follow-up ex- ams. Commonly reported complaints included slight pain at action is also suggested by the effectiveness of this drug in the injection site (45%), back pain (27%), headache (16%), and diarrhea (14%). The overall incidence of reporting any Serum Creatinine Levels adverse event was 77% in group A, 88% in group B, and 93% in group C patients. These events were mild to moderate in Changes During Treatment most instances, were not dose related, and were thought to be unrelated to the study medication. Three patients of the 49 discontinued the study. One pa- tient from group B had a severe stress fracture of the left femur resulting from trauma that required surgical inter- vention. Another patient from group B suffered a partial small bowel obstruction, which resolved with conservative therapy. One patient from group C developed a sensation of leg weakness, reported as paralysis, several hours after each of four administrations of the gallium. No objective evidence 4 6 6 10 12 of weakness was evident during an extensive neurological Week Number evaluation, and the symptoms abated after discontinuing the gallium. These patients are included in the analyses of base- - 0.05 mg/kg/day - 0.25 mg/kg/day - 0.50 mglkg/day line characteristics and for as long as their data were collected FIG. 5. Mean serum creatinine levels (normal, 0.4-1.2 mg/dL) during in an “intention to treat” fashion. Their baseline values were the trial. There was very little variation and no significant differences substituted for all weeks subsequent to their discontinuing among the groups.

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010 JCE & M . 1995 600 BOCKMAN ET AL. Vol 80 . No 2 the treatment of cancer-related hypercalcemia (38,58), a dis- Symptomatic relief of pain and an increased tolerance to order characterized by marked osteoclastic resorption. In physical activity were recorded in all symptomatic patients vitro studies have shown that gallium nitrate preferentially receiving the 0.25 and 0.5 mg/kg.day doses. Our study is accumulates in the metabolically active regions of bone similar to others that reported clinical improvement with where bone remodeling occurs. At these sites, gallium can be antiresorptive therapy (10,221; however, when we attempted measured at very low concentrations (in the parts per million to quantify these measures using a visual analog scale, we range) (59). Bone particles prepared from animals injected were unable to find significant differences. This may be due with gallium nitrate showed increased calcium and phos- to a shortcoming in the methods employed to quantify these phorus content, with larger and more mature hydroxyapatite subjective parameters. We used a modification of the crystals that were more resistant to cell-mediated dissolution Melszack-McGill quality of life assessment (65); unfortu- (39,59,60). Treatment of isolated osteoclasts with micromo- nately, we are not aware of any validated assessment of lar concentrations of gallium nitrate inhibited their ability to clinical improvement in patients with Paget’s disease. Fur- produce resorption pits in normal bone; resorption was sim- thermore, in placebo-controlled studies in Paget’s disease, ilarly inhibited if normal (untreated) osteoclasts were placed symptomatic improvement occurred in 35% of patients in the on pieces of intact bone that had been exposed to gallium placebo arm (10, 22). With such a high background rate of nitrate (43). Exposure of osteoclasts to concentrations of gal- improvement and a limited number of symptomatic patients lium nitrate as high as 300 pmol/L, levels more than 20 times the serum levels achieved in patients with cancer-related per group, there may not be the statistical power to detect hypercalcemia receiving 5 mg/kg.day as a continuous in- even moderate differences in clinical improvement. When fusion (58) and more than 200 times the concentrations quality of life was analyzed by our nonstandardized ques- achieved in the present study, produced no evidence of cy- tionnaire, a significant trend (P < 0.05) for improvement over totoxicity (43). Thus, the antiresorptive effects observed here time was found in the patients receiving the highest dose are clearly not a consequence of osteoclastic cytotoxicity. tested. More recent studies have shown that unlike the bisphos- In long term studies using antiresorptives for the treat- phonates and calcitonins, gallium nitrate can directly stim- ment of Paget’s disease, the return of bone resorptive pa- ulate osteoblasts to increase matrix gene expression. Treat- rameters to the high baseline state usually precedes and ment of explanted rat calvaria with gallium nitrate increased heralds a rise in the osteoblast marker alkaline phosphatase [“Hlhydroxyproline incorporation into collagenase-digest- (32). In patients with high baseline levels of alkaline phos- ible protein (40). Such treatment also increased procollagen phatase, prolonged or repetitive cycles of therapy with an- and fibronectin messenger ribonucleic acid in transformed tiresorptives are required for normalization of biochemical osteoblasts, primary osteoblasts, and human skin fibroblasts parameters (32). It is noteworthy in the current clinical trial (40, 61). In osteoblasts, gallium nitrate inhibited osteocalcin that the second course of gallium nitrate administration gene expression and decreased bone osteocalcin content after caused an additional 10% reduction in biochemical param- ill uivo treatment (61,62). As osteocalcin may serve as a signal eters. Thus, intermittent cyclical therapy with low dose gal- for osteoclastic resorption (63), gallium nitrate may reduce lium nitrate might be expected to normalize biochemical osteocalcin content in newly formed bone and thereby di- variables in most patients with advanced Paget’s disease of minish recruitment, which would protect the bone bone, and the extent of therapy would depend on the severity from resorption in the future. of disease. Overall, gallium nitrate given as low dose SCinjections was In the present study there was an upward trend of hy- well tolerated in this elderly group of patients. At the doses droxyproline (and N-telopeptides in the subgroup) after the used in this study, gallium nitrate compares favorably to the 2-4 weeks following cessation of each course of gallium current available therapies with calcitonins and bisphospho- nates. No clinically significant changes were measured in nitrate therapy. Despite this rebound in resorptive parame- serum creatinine, blood urea nitrogen, creatinine clearance, ters, gallium was able to continually suppress disease activ- blood cell counts, or hematocrits. Hemoglobin levels de- ity, as measured by the persistent linear decline in serum creased by slightly more than 2.0 g/dL in two patients re- alkaline phosphatase. In many other studies with calcitonins ceiving the highest dose tested, both patient’s hemoglobin and bisphosphonates, the nadir of alkaline phosphatase was levels returned to their baseline levels by the twelfth week of often not seen until after 3-6 months of therapy (12). How- study. No correlation was seen between the hemoglobin ever, we did not follow the entire patient group beyond 12 levels and any changes in other blood parameters. Interest- weeks. Gallium nitrate has a different mechanism of action ingly, the falls in hemoglobin levels seen in this study were and toxicity profile from the calcitonins or bisphosphonates, similar to those in a prior study (38), in which 6-12 times the future investigations might well explore combined dose of gallium nitrate was used. Gallium nitrate may inhibit treatments of gallium nitrate with these other agents. erythropoeisis or compete with iron in binding to transferrin (64); however, no decrease in reticulocyte counts was noted. Decreases in iron-binding capacity preceded the nadir in hemoglobin, but no coincident decreases in iron, transferrin, Acknowledgments or ferritin levels were seen. Thus, the effects of gallium nitrate We wish to thank Dr. I’. Schecter for his support, and especially Dr. on hemoglobin levels were not clearly related to effects on R. Warrell, Jr., for his expert guidance in the design of the study and for iron metabolism or transport. reviewing the manuscript.

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010 GALLIUM NITRATE IN PAGET’S DISEASE

References 24. Delmas I’D, Chapuy MC, Vignon E, et al. 1982 Long term effects of Dichloromethylene Diphosphonate in Paget’s disease of bone. 1. Kanis JA. 1991 Pathophysiology and treatment of Paget’s disease of J Clin Endocrinol Metab. 54:837-844. bone. Durham: Carolina Academic Press; 1. 25. Adami S, Guarrera G, Salvagno G, et al. 1984 Sequential treatment 2. Meunier PJ, Coindre J, Edouard CM, et al. 1980 Bone histomor- of Paget’s disease with human calcitonin and dichloromethylene phometry in Paget’s disease. Quantitative and dynamic analysis of diphosphonate (Cl, MDP). Metab Bone Dis. 5:265-267. Paget’s disease and nonpagetic bone tissue. Arthitis Rheum. 23: 26. Yates AJP, Gray RES, Urwin GH, et al. 1985 Intravenous clodronate 1095-1103. in the treatment and retreatment of Paget’s disease of bone. Lancet. 3. Franck WA, Bress NM, Singer FR, et al. 1974 Rheumatic manifes- 1:1474-1477. tation of Paget’s disease of bone. Am J Med. 56:592-603. 27. Frijlink WB, Velde JT, Bijvoet OLM, Heynen G. 1979 Treatment of 4. Walton RJ, Preston CJ, Bartlett M, et al. 1977 Biochemical mea- Paget’s disease with (3-amino-l-hydroxypropylidene)-l,l-bisphos- surements in Paget’s disease of bone. Distribution and correlations. phonate (A.P.D.). Lancet. 1:799-803. Eur J Clin Invest. 7:37-39. 28. Heynen G, Delwaide P, Bijvoet OLM, Franchimont P. 1982 Clinical 5. Morris HA, Wishart JM, Horowitz M, Need AG, Nordin BEC. 1990 and biological effects of low doses of (3 amino-l hydroxypropli- The reproducibility of bone-related biochemical variables in dene)-l,l-bisphosphonate (APD) in Paget’s disease of bone. Eur postmenopausal women. Ann Clin Biochem. 27562-568. J Clin Invest. 11:29-35. 6. Bordier P, Hioco D, Tun-Chot S. 1970 Calcitonin: acute effects upon 29. Fraser TRC, Ibbertson HK, Holdaway IM, et al. 1984 Effective oral serum calcium urinary hydroxyproline excretion, and osteoclasts in treatment of severe Paget’s disease of bone with APD (3.amino-l- man. In: Taylor S, Foster G, eds. Calcitonin 1969. Heinemann: hydroxypropylidene-l,l-bisphosphonate); a comparison with com- London; 339-347. bined calcitonin + EHDP (l-hydroxyethylidene-l,l-bisphospho- 7. Haddad JG, Birge S, Avioli L. 1970 Effects of prolonged thyrocal- nate). Aust NZ J Med. 14:811-818. citonin administration on Paget’s disease of bone. N Engl J Med. 30. Mautalen CA, Gonzalez D, Ghiringhelli G. 1985 Efficacy of the 283:549-555. bisphosphonate APD in the control of Paget’s bone disease. Bone. 8. Shai F, Baker RK, Wallach SL. 1971 The clinical and metabolic 6~429-432. effects of porcine calcitonin on Paget’s disease of bone. J Clin Invest. 31. Cantrill JA, Buckler HM, Anderson DC. 1986 Low dose intra- 50:1927-1940. venous 3-amino-l-hydroxypropylidene-l,l-bisphosphonate 9. Krane SM, Harris ED, Singer FR, et al. 1973 Acute effects of cal- (APD) for the treatment of Paget’s disease of bone. Ann Rheum citonin on bone formation in man. Metabolism. 22:51-58. Dis. 45:1012-1018. 10. Altman RD, Johnston CC, Khairi MRA, Wellman H, Serafini AN, 32. Fenton AJ, Gutteridge DH, Kent GN, et al. 1991 Intravenous ami- Sankey RR. 1973 Influence of disodium etidronate on clinical and nobisphosphonate in Paget’s disease: clinical, biochemical, histo- laboratory manifestations of Paget’s disease of bone (osteitis morphometric and radiological responses. Clin Endocrinol (Oxf). deformans). N Engl J Med. 289:1379-1384. 34:197-204. 11. Douglas DL, Duckworth T, Kanis JA, et al. 1980 Biochemical and 33. Price RI, Gutteridge DH, Stuckey BGA, et al. 1993 Rapid, divergent clinical responses to dichloromethylene bisphosphonate (Cl, MDP) changes in spinal and forearm bone density following short-term in Paget’s disease of bone. Arthritis Rheum. 23:1185-l 192. intravenous treatment of Paget’s disease with pamidronate diso- 12. Siris ES, Canfield RE, Jacobs TP, Stoddart KE, Spector PJ. 1981 dium. J Bone Miner Res. 8:269-217. Clinical and biochemical effects of EHDP in Paget’s disease of bone: 34. O’Dohertv DP. Gertz BI. Tindale W. Sciberras DG. Survill TT. patterns of response to initial treatment and to long-term therapy. , ,. Metab Bone Dis. 4&5:301-308. Kanis JA. 1992 Effects of five daily 1 h infusions of aiendronate in 13. Harinck HIJ, Papapoulos SE, Blanksma HJ, Moolenaar AJ, Paget’s disease of bone. J Bone Miner Res. 7:81-87. Vermeij I’, Bijvoet OLM. 1987 Paget’s disease of bone: early and late 35. Reginster JY, Colson F, Morlock G, Combe B, Ethgen D, Geusens responses to three different modes of treatment with aminohy- P. 1992 Evaluation of the efficacy and safety of oral tiludronate in droxypropylidene bisphosphonate (APD). Br Med J. 295:1301- 1305. Paget’s disease of bone. Arthitis Rheum. 35:967-974. 14. Gray RES, Yates AJP, Preston CJ, Smith R, Russell RGG, Kanis JA. 36. Brown JP, Kylstra JW, Bekker PJ, et al. 1994 Risedronate in Paget’s 1987 Duration of effect of oral diphosphonate therapy in Paget’s disease: preliminary results of a multicenter study. Semin Arthritis disease of bone. Q J Med. 245:755-767. Rheum. 23:272. 15. Bijvoet OLM, van der Sluys Veer JD, Jansen Al’. 1968 Effects of 37. Lebbin D, Ryan WG, Schwartz TB. 1974 Outpatient treatment of calcitonin on patients with Paget’s disease, thyrotoxicosis or hyper- Paget’s disease of bone with mithramycin. Ann Intern Med. 81: calcemia. Lancet. 1:876-881. 635- 637. 16. Greenberg PB, Doyle FH, Fisher MT, et al. 1974 Treatment of 38. Warrell RI’ Jr, Bockman RS, Coonley CJ, Isaacs M, Staszewski H. Paget’s disease of bone with synthetic human calcitonin. Biochem- 1984 Gallium nitrate inhibits calcium resorption from bone and is ical and roentgenologic changes. Am J Med. 56:867-870. effective treatment for cancer-related hypercalcemia. J Clin Invest. 17. DeRose J, Singer FR, Avramides A, et al. 1974 Response of Paget’s 73:1487-1490. disease to porcine and salmon calcitonins. Effects of long-term treat- 39. Bockman RS, Boskey AL, Blumenthal NC, Alcock NW, Warrell RP ment. Am J Med. 56:858-866. Jr. 1986 Gallium increases bone calcium, and crystallite perfection of 18. Grunstein HS, Clifton-Bligh I’, Posen S. 1981 Paget’s disease of hydroxyapatite. Calcif Tissue Int. 39:376-381. bone. Experiences with 100 patients treated with salmon calcitonin. 40. Bockman RS, Guidon PT, Pan LC, Salvatori R, Kawaguchi A. 1993 Med J Aust. 2:278-280. Gallium nitrate increases type I collagen and fibronectin mRNA and 19. Martin TJ, Jerums G, Melick RA, Xipell JM, Arnott R. 1977 Clinical, collagen protein levels in bone and fibroblast cells. J Cell Biochem. biochemical and histological observations on the effect of porcine 52:396-403. calcitonin in Paget’s disease of bone. Aust NZ J Med. 736-43. 41. Warrell Jr RI’, Alcock NW, Bockman RS. 1987 Gallium nitrate 20. Canfield R, Rosner W, Skinner J, et al. 1977 Diphosphonate therapy inhibits accelerated bone turnover in patients with bone metastases. of Paget’s disease of bone. J Clin Endocrinol Metab. 44:96-106. J Clin Oncol. 5:292-298. 21. Alexandre C, Meunier PJ, Edouard C, Khairi RA, Johnston CC. 42. Warrell RI’, Lovett D, Dilmanian A, Schneider R, Heelan RT. 1993 1981 Effects of ethane-1 hydroxy-l,l-diphosphonate (5 mg/kg/day Lo;v-dose gallium nitrate for prevention of osteolysis in myeloma: dose) on quantitative bone histology in Paget’s disease of bone. results of a uilot randomized studv. I Clin Oncol. 11:2443-2450. Metab Bone Dis. 3:309-315. 43. Hall TJ, Chambers TJ. 1990 Gallium’inhibits bone resorption by a 22. Khairi MRA, Altman RD, DeRosa GP, Zimmermann J, Schenk RK, direct effect on osteoclasts. Bone Miner. 8211-216. Johnston CC. 1977 Sodium etidronate in the treatment of Paget’s 44. Schlesinger PH, Teitelbaum SL, Blair HC. 1991 Osteoclast inhibi- disease of bone. Ann Intern Med. 87:656-663. tion by Ga’+ contrasts with bisphosphonate metabolic suppression. 23. Meunier PJ, Alexandre C, Edouard C, et al. 1979 Effects of disodium Competitive inhibition of H+ ATPase by bone-bound gallium. dichloromethylene diphosphonate on Paget’s disease of bone. J Bone Miner Res. 6(Suppl l):S127. Lancet. 2:489-492. 45. Matkovic V, Apseloff G, Shepard DR, Gerber N. 1990 Use of

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010 JCE & M . 1995 602 BOCKMAN ET AL. Vol 80. No 2

gallium to treat Paget’s disease of bone: a pilot study. Lancet. 1978 Treatment of Paget’s disease with combined calcitonin and 335(8681):72-75. diphosphonate (EHDP). Metab Bone Dis. 1:251-261. 46. Warrell RI’, Bosco B, Weinerman S, Levine B, Lane J, Bockman RS. 57. Hosking DJ, van Aken J, Bijvoet OLM, Will EJ. 1976 Paget’s bone 1991 Gallium nitrate for advanced Paget’s disease of bone: effec- disease treated with diphosphonate and calcitonin. Lancet. 1: tiveness and dose-response analysis. Ann Intern Med. 113:847-851. 615-617. 47. Stegeman H, Stalder K. 1967:Determination of hydroxyproline. 58. Warrell Jr RP, Israel R, Frisone M, Snyder T, Gaynor JJ, Bockman Clin Chim Acta. 18:267-273. RS. 1988 Gallium nitrate for acute treatment of cancer-related 48. Russell RGG, Beard DJ, Cameron EC, et al. 1981 Biochemical mark- hypercalcemia. Ann Intern Med. 108:669-674. ers of bone turnover in Paget’s disease. Metab Bone Dis. 4/5:255- 262. 59. Bockman RS, Repo MA, Warrell RI’, et al. 1990 Distribution of trace 49. Hanson DA, Weis M-AE, Bollen A-M, Maslan SL, Singer FR, Eyre levels of therapeutic gallium in bone as mapped by synchrotron DR. 1992 A specific immunoassay for monitoring human bone re- x-ray microscopy. Proc Nat1 Acad Sci USA. 87:4149-4153. sorption: quantitation of type I collagen crosslinked N-telopeptides 60. Repo MA, Bockman RS, Betts F, Boskey AL, Warrell Jr RP. 1988 in urine. J Bone Miner Res. 7:1251-1258. Effect of gallium on bone mineral properties. Calcif Tissue Int. 43: 50. Thiebaud D, Jaeger P, Gobelet C, Jacquet AF, Burkhardt I’. 1988 A 300-306. single infusion of the bisphosphonate AHPrBP (AI’D) as treatment 61. Jenis LG, Waud CE, Stein GS, Lian JB, Baran DT. 1993 Effect of of Paget’s disease of bone. Am J Med. 85:207-212. gallium nitrate in zitro and in normal rats. J Cell Biochem. 52: 51. Mautalen CA. 1983 Treatment of Paget’s bone disease with the 330-336. bisuhosohonate APD. Henrv Ford HOSD Med I. 31:244-248. 62. Guidon Jr PT, Salvatori R, Bockman RS. 1993 Gallium nitrate 52. Adami S, Salvagno G, Guarrera G, et ai. 1986 Treatment of Paget’s regulates rat osteoblast expression of osteocalcin protein and mRNA disease of bone with intravenous 4-amino-l-Hydroxybutylidene-1, levels. J Bone Miner Res. 8:101-110. I-Bisphosphonate. Calcif Tissue Int. 39:226-229. 63. Lian JB, Tassinari M, Glowacki J. 1984 Resorption of implanted 53. Dewis P, Prasad BK, Anderson DC, Willets S. 1985 Clinical expe- bone prepared from normal and warfarin-treated rats. J Clin Invest. rience with the use of two diphosphonates in the treatment of Paget’s 73:1223-1226. disease. Ann Rheum Dis. 44:34-38. 64. Chitambar CR, Seligman PA. 1986 Effects of different transferrin 54. Altman RD, Collins-Yudiskas B. 1987 Synthetic human calcitonin forms on transferrin receptor expression, iron uptake and cellular in refractory Paget’s disease of bone. Arch Intern Med. 147: proliferation of human leukemic HL60 cells: mechanisms respon- 1305-1308. sible for the specific cytotoxicity of transferrin-gallium. J Clin Invest. 55. Singer FR, Fredericks RS, Minkin C. 1980 Salmon calcitonin ther- 78:1538-1546. apy for Paget’s disease of bone. Arthitis Rheum. 23:1148-1153. 65. Melszack R. 1975 The McGill pain questionnaire: major properties 56. Bijvoet OLM, Hosking DJ, Frijlink WB, Velde JT, Vellenga CJLR. and scoring methods. Pain. 1:277-299.

Downloaded from jcem.endojournals.org at Stanford Univ. Medical Center Lane Medical Lib., Route 1 on February 8, 2010