A multicenter, randomized clinical trial comparing zoledronic acid versus observation in patients with asymptomatic myeloma

Authors


Abstract

BACKGROUND.

Bisphosphonates (BPs) are effective in the prevention and treatment of skeletal-related events (SREs) in patients with symptomatic myeloma who are receiving chemotherapy. Recent data also suggest a possible antineoplastic activity of BPs. Few studies published to date have explored the role of BPs in patients with untreated, asymptomatic myeloma (AM). No data are available on the efficacy of zoledronic acid in these patients.

METHODS.

The authors conducted a prospective, multicenter, open-label, phase 3, randomized trial comparing the administration of zoledronic acid versus simple observation in patients with AM. One-hundred sixty-three patients were enrolled and randomized (1:1) to receive zoledronic acid (n = 81 patients) or not to receive zoledronic acid (n = 82 patients) for 1 year at a dose of 4 mg monthly as a single, 15-minute, intravenous infusion.

RESULTS.

After a median follow-up of 64.7 person-months, 44.4% of patients in the zoledronic acid group and 45.1% of the control group progressed to ‘symptomatic’ myeloma requiring chemotherapy (P = .9307). The median time to progression was 67 months and 59 months for the treatment and control groups, respectively (P = .8312). At progression, SREs were significantly lower in the zoledronic acid-treated group (55.5%) than in the control group (78.3%; P = .041), whereas anemia, renal failure, and extramedullary disease were not statistically different. More frequent adverse events observed in the zoledronic acid-treated group were asymptomatic hypocalcemia and fever. One patient developed reversible osteonecrosis of the jaw. No renal failure caused by zoledronic acid was reported.

CONCLUSIONS.

The monthly use of zoledronic acid for 1 year in patients with AM reduced the development of SREs at progression but did not influence the natural history of the disease. Cancer 2008. © 2008 American Cancer Society.

Asymptomatic (also defined as ‘inactive’ or ‘smoldering’) myeloma (AM) accounts for approximately 20% of patients with a diagnosis of multiple myeloma (MM).1, 2 AM is characterized by the absence of end-organ damage or tissue involvement, such as anemia, bone lesions, hypercalcemia, and renal failure, or by other relevant clinical conditions, such as hyperviscosity, amyloidosis, and recurrent infections.2 AM differentiates from monoclonal gammopathy of undetermined significance (MGUS) on the basis of higher levels of circulating M-component (>3 g/dL) and/or a percentage of bone marrow plasma cells (BMPCs) >10%.2 The annual progression rate of AM into ‘symptomatic’ or ‘active’ MM is approximately 10% during the first 5 years.1 This favorable clinical outcome and the occurrence of relevant side effects of chemotherapies suggest the use of simple observation in these patients,3 whereas conventional chemotherapy4–6 and even the use of new drugs, such as thalidomide,7, 8 generally are not recommended until end-organ damage develops.

Bisphosphonates (BPs) have entered clinical practice as a useful adjunct for the treatment of symptomatic MM, providing, in controlled trials, effective results in terms of reduction of skeletal-related events (SREs) (bone lesions, pathologic fractures, spinal compression requiring radiotherapy, and hypercalcemia) and pain control.9–13 However, in vitro, BPs also have demonstrated some direct or indirect antineoplastic activities.14, 15 These effects have been investigated particularly on both tumor and bone marrow stromal cells of MM.16–20 Furthermore, evidence of a possible antimyeloma effect of BPs has been reported occasionally in some patients.21, 22 It is noteworthy that some subcategories of patients with MM who received BPs along with chemotherapy had improved survival with respect to controls who received the same chemotherapy but without BPs.11, 13 Therefore, theoretically, there may be a rationale for using BPs in patients with AM if a role is hypothesized for these agents in interfering with the transforming process from indolent to active disease, which is mediated at least in part by the strict interactions that exist between clonal tumor cells and the bone marrow microenvironment.23, 24 In this setting, pamidronate, a second-generation BP, was able to reduce bone turnover and to decrease the development of SREs at progression, albeit without evidence of significant antitumor effects in terms of the rate of transformation or the time to disease evolution into ‘symptomatic,’ overt MM.25–27

Zoledronic acid (zoledronate) is a third-generation, nitrogen-containing BP with higher activity and better in vitro-defined antitumor effects with respect to other BPs.28 Zoledronic acid has demonstrated significant efficacy on bone disease in symptomatic MM29–31 and is used largely in clinical practice. Thus, zoledronic acid may represent a reasonable agent to test in the early stages of disease, such as AM. Accordingly, we conducted a clinical trial to verify the effects of zoledronic acid in patients with AM.

MATERIALS AND METHODS

Patients

This was a prospective, multicenter, open-label, phase 3, randomized clinical trial comparing administration of zoledronic acid versus simple observation in patients with AM who did not require further treatments. The study started in June 2001 and was closed in June 2004. Nine Italian Centers enrolled evaluable patients.

Inclusion criteria for participation were as follows: immunoglobulin G or immunoglobulin A M-component >2 g/dL; BMPCs >10%; hemoglobin level >11.5 g/dL; no evidence of bone disease; normal serum levels of creatinin (<1.2 mg/dL), calcium (<10 mg/dL), β2 microglobulin (<3 mg/dL), albumin (>3.5 g/dL), C-reactive protein (<5 mg/L), and lactate dehydrogenase (<480 IU/L); no bone pain; a World Health Organization performance status 0 or 1; no evidence of amyloidosis; and no concomitant or previous active treatments for myeloma. In doubtful cases, the absence of skeletal lesions was confirmed by magnetic resonance imaging (MRI) or positron emission tomography (PET), as appropriate. These criteria, which were defined arbitrarily in 2001, corresponded in all patients to the current definition of AM more recently suggested by the International Myeloma Working Group2 and also identified only stage I patients according to the International Staging System,32 both of which were published after the start of our study. One difference was the lower level of M-component (2 g/dL instead of 3 g/dL) that we chose to have the possibility of including a larger number of patients; however, we excluded patients with ‘true’ MGUS, because only patients with >10% BMPCs were enrolled. Exclusion criteria were another cancer, psychiatric disease, and severe cardiac, respiratory or liver dysfunction

The study was approved by the institutional review board at each participating center. All patients provided written informed consent before entering the study, in compliance with the ethical principles derived from the Declaration of Helsinki, from good clinical practice guidelines, and from applicable legislation.

Study Design and Procedures

Randomization (1:1) to receive or not receive zoledronic acid was performed at the individual center level by using a method of sequence generation based on a local random-number table. In patients who were randomized to receive the drug, zoledronic acid (Zometa; Novartis Pharmaceuticals, Origgio, Italy) was administered for 1 year on an outpatient basis, at a dose of 4 mg as a single, monthly, 15- minute intravenously infusion. Controls simply were observed. We chose to administer zoledronic acid for 1 year only after a wide panel discussion in which the majority of participating centers concluded that this period was long enough to provide clinical and safety indications and to be well accepted by patients, reducing the risk of possible side effects and of an excessive dropout rate in a population without need of other specific treatments.

Patients were monitored by evaluating hemoglobin levels, white blood cell and platelet counts, liver function and clotting tests, total protein levels, serum protein electrophoresis, urine analysis, C-reactive protein, lactate dehydrogenase and serum β2 microglobulin levels every 3 months in both groups of patients. Electrolytes and renal function also were determined every month in patients who were receiving zoledronic acid to evaluate its safety. Bone marrow examination, x-ray skeletal survey (and MRI or PET studies, if necessary) also were repeated every year or when clinically indicated. These examinations were not centralized.

The primary endpoint of the study was 5-year progression-free survival (PFS), which was measured from the time of the first visit of the study. Secondary endpoints were the time to progression (TTP), the percentage of progressed patients with SREs, and the safety/tolerability of zoledronic acid in this setting of patients.

To define progression into symptomatic MM, the appearance of 1 or more of the following criteria (related to myeloma and excluding any other possible cause) was required: anemia (<11 g/dL hemoglobin), abnormal creatinine (both with at least 2 consecutive measurements), SREs (single or multiple osteolytic lesions, pathologic fractures, hypercalcemia), or extramedullary tumors. In all such patients, an active treatment (chemotherapy and/or radiotherapy) was started. The simple increase of M-component and/or BMPCs generally was not considered disease progression, with the exception of 2 patients (1 in the zoledronic acid group and 1 in the control group), who received therapy because these parameters increased rapidly, although in the absence of other signs or symptoms. However, in nearly all patients, progressive disease (as defined above) was accompanied by at least a 25% increase in M-component and/or BMPCs.

Statistical Analysis

On the basis of 85% power, a sample size of 208 patients (104 patients per arm) was required initially to detect a 20% difference (from an expected 50% to 30% difference) in the proportion of zoledronic acid-treated patients experiencing progressive disease at 5 years with respect to the control group (constant hazard ratio, 1.737; α error, .05; P = .05; 2-sided; nQuery Advisor, version 7.0; Statistical Solutions Ltd., Dublin, Ireland). To compensate for nonevaluable patients, we planned to enroll 115 patients per group.

An interim analysis was done for safety monitoring in May 2004 because of the description of the first cases of osteonecrosis of the jaw (ONJ) in patients with myeloma who were receiving BP.33, 34 At that time, the steering committee determined that this was not an acceptable risk for patients who otherwise did not require specific treatments (for further details, see Discussion, below) and decided to prematurely stop the trial in June 2004 after 163 patients had been randomized (70.8% of the planned sample size). The final analysis was performed as of December, 31, 2007.

Central and dispersion statistics for continuous variables and frequency and percentages for categorical variables were used (SAS version 9.1 for Windows XP). In particular, univariate logistic regression (Proc. Logistic) was used to investigate the relations between the development of SREs and exposure to zoledronic acid. PFS and TTP analysis methodology (product limit survival) was applied to estimate the median time free from disease progression with log-rank tests and the Kaplan-Meier method for comparisons between groups (Proc Life Test) and with log-time exponential regression to model the right-censored data and to estimate the slope and 95% confidence interval (CI) (Proc Life Reg). The Fisher exact test was used to evaluate the statistical differences in progression and survival rates (Stat Exact, version 7.0).

RESULTS

Enrolment and Follow-up

Overall, 198 patients were evaluated for participating this study (Fig. 1). Thirty-five patients were excluded (n = 20 patients who did not meet criteria, n = 12 patients who refused to participate, and n = 3 patients who were excluded for other reasons). Therefore, 163 patients with AM were enrolled and randomized. Table 1 shows their main clinical and laboratory characteristics at baseline. The 2 groups were comparable in terms of age, sex, time from diagnosis, levels and type of M-component, and BMPCs.

Figure 1.

The Consolidated Standards Reporting of Trials (CONSORT) trial profile.

Table 1. Baseline Clinical and Laboratory Characteristics
CharacteristicZoledronic Acid GroupControl Group
  1. IgG indicates immunoglobulin G; IgA, immunoglobulin A.

No. of patients8182
Median age [range], y66 [41-82]67 [42-84]
No. of men/women43/3847/35
Median M-component [range], g/dL2.5 [2.1-4.7]2.3 [2.1-4.9]
Median bone marrow plasma cells [range], %21 [11-55]19 [11-52]
No. with IgG/IgA58/2361/21
No. with κ/λ59/2257/25
Median mo from diagnosis [range]16 [0-39]14 [0-42]

Seven patients (3 patients in the zoledronic acid group and 4 patients in the control group) died for reasons unrelated to AM or treatment between 6 months and 49 months (Fig. 1). Seven additional patients (3 patients in the zoledronic acid-treated group and 4 patients in the control group) were lost to follow-up after 6 months to 26 months. However, all 163 patients were analyzed by group, including 81 patients who received zoledronic acid and 82 nontreated control patients (Fig. 1). Excluding the patients who died or were lost at follow-up, all other patients had at least 36 months of complete observation.

Safety

Among the zoledronic acid-treated patients, 15 patients experienced asymptomatic hypocalcemia (without need to stop treatment and promptly corrected by oral substitutive therapy), and 7 patients developed fever after the initial infusions (1 patient interrupted early drug administration). One patient developed a reversible ONJ at the end of the treatment that remitted completely after ozone and antibiotic therapy.35 We also were informed of a second case of ONJ that developed in a patient who continued on therapy with zoledronic acid beyond the 12 planned cycles (protocol violation). No patient had renal failure attributable to zoledronic acid treatment.

Efficacy

No significant reduction of M-component (>25%) or BMPCs was observed throughout the study in either group. After 9491 cumulative person-months of follow-up (median, 64.7 person-months; range, 36-72 person-months), there were 36 progressions (44.4%) to ‘symptomatic’ MM in the zoledronic acid arm and 37 progressions (45.1%) in the control group (odds ratio, 1.03; 95% CI, 0.55-1.91; P = .9307). The percentage of patients who progressed at 5 years (primary endpoint) was 42% versus 42.7% of patients who did not progress (odds ratio, 1.03; 95% CI, 0.55-1.92; P = .9272) (Table 2). The median TTP (Fig. 2) was 67 months in the zoledronic acid group and 59 months in the control group (P = .8312). PFS was almost identical in both groups (P = .7901).

Figure 2.

The time to progression is illustrated for 163 patients with asymptomatic myeloma treated or not with zoledronic acid (log-rank test, 0.0454; P = .8312). Because of the substantial amount of right-censored time data, log-time exponential regression also was used to estimate the regressor with 95% confidence limits (CL): estimate, 0.0568; standard error, 0.2341; 95% CL, −0.35,146 to 0.4021; P = .8084. Trt indicates treatment.

Table 2. Absolute Numbers and Percentages of Patients (Treated or Not Treated With Zoledronic Acid) Who Remained Stable or Developed Progressive Disease at 5 Years
Patient GroupNo. of Patients (%)
Stable Disease*Progressive Disease*
  • *

    P=.9272.

Treated (n=81)47 (58)34 (42)
Not treated (n=82)47 (57.3)35 (42.7)
Total (n=163)94 (57.7)69 (42.3)

At the time of progression, SREs were significantly lower in the zoledronic acid group (n = 20 patients; 55.5%) than in the control group (n = 29 patients; 78.3%; odds ratio, 2.90; 95% CI, 1.04-8.06; P = .041) whereas anemia, renal failure, and extramedullary disease were not statistically different between the 2 groups (Table 3). Similarly, the levels of M-component (median: 3.8 g/dL [range, 2.8-6.1 g/dL] in the zoledronic acid group vs 3.7 g/dL [range, 3-5.7 g/dL] in the control group) and BMPCs (median: 35% range, 22%-95%] in zoledronic acid group vs 36% [range, 24%-100%] in the control group) also were similar at progression (P value not significant). Finally, there were no significant differences in outcome between patients with BMPCs <20% versus >20% or between patients with or without osteoporosis at baseline (data not shown).

Table 3. Causes (Absolute Numbers and Percentages) That Determined Progressive Disease in Patients Treated or Not Treated With Zoledronic Acid*
Patient GroupNo. of Patients (%)
AnemiaRenal FailureExtramedullary TumorSRE
  • SRE indicates skeletal-related events.

  • *

    More than 1 cause occurred in several patients.

  • P value not significant.

  • P=.041.

  • §

    There were 16 osteolytic lesions, 9 hypercalcemias, and 3 pathologic fractures.

  • ||

    There were 22 osteolytic lesions, 10 hypercalcemias, and 6 pathologic fractures.

Treated (n=36)25 (69.4)8 (22.2)5 (13.8)20 (55.5)§
Not treated (n=37)27 (72.9)6 (16.2)4 (10.8)29 (78.3)||

Treatment After Evolution Into Symptomatic Myeloma and Survival

All patients received active treatments after progression to symptomatic MM. Overall, these treatments were quite heterogeneous and included autologous stem cell transplantation in approximately 40% of patients in both groups. The overall response to first-line therapy after progression did not differ, with approximately two-thirds of patients responding in both groups (data not shown). On December, 31, 2007, 14 of 36 patients (38.8%) in the zoledronic acid group who progressed and 15 of 37 patients (40.5%) in the control group had died from MM (odds ratio, 1.07; 95% CI, 0.38-3.04; P = .8381).

DISCUSSION

The use of BPs has not been investigated extensively in patients with AM. One report described 12 patients with ‘smoldering’ or ‘indolent’ myeloma who received 12 pamidronate courses.25 One minor response, 8 patients with stable disease, and 3 patients with disease progression were observed. Circulating levels of some cytokines potentially involved in bone disease (interleukin-6, interleukin-1β, and oncostatin-M) remained stable, whereas tumor necrosis factor α levels increased. Bone density also was augmented, but markers of bone resorption and bone formation decreased, suggesting a reduction of bone turnover in the absence of significant antitumor effects.

In our 2 previous studies, ‘prophylactic’ administration of pamidronate was evaluated in patients with untreated stage I/IIA or ‘smoldering myeloma.’26, 27 Pamidronate did not reduce the rate of or the time required to develop symptomatic disease. In addition, no significant effect on the M-component level was observed, but the development of SREs decreased significantly.26, 27 Accordingly, the recently updated American Society of Clinical Oncology clinical practice guidelines on the role of BP in MM are not changed from the initial (2002) statement, confirming that starting BP is not recommended for patients with solitary plasmocytoma or smoldering (asymptomatic) or indolent myeloma.36

To the best of our knowledge, this is the first study exploring the role of zoledronic acid in patients with AM. Zoledronic acid is a molecule approximately 1000-fold more potent than pamidronate that has documented in vitro antiproliferative, proapoptotic, and cytotoxic effects against myeloma cells.37, 38 Zoledronic acid also exerts in vitro antiangiogenesis and immunomodulating (on γ-δ lymphocytes and dendritic cells) activities as well as modulation of adhesion molecules and interference with mevalonate metabolic pathway.39–44 It is noteworthy that Berenson et al recently reported a better survival rate with zoledronic acid than with pamidronate among chemotherapy-treated patients with MM who had high baseline bone alkaline phosphatase levels.45 Similarly, Aviles et al confirmed a possible favorable impact of zoledronic acid on the response rate, event-free survival, and overall survival of patients with MM who received standard chemotherapy.46

Despite these favorable reports, we could not demonstrate a significant benefit from zoledronic acid in AM patients, at least in terms of the percentage of transformation into overt myeloma and PFS/TTP duration. By contrast, a reduction in SREs at the time of disease progression was observed. This finding is similar to our previous reports on the use of pamidronate in patients with AM.26, 27 In the current study, however, a trend in favor of zoledronic acid was evident during the first 2 years of observation (Fig. 2). This raises the question of whether a longer period of zoledronic acid administration could be more effective in this setting.

BPs generally are tolerated well; however, since 2004, the use of BPs has been associated with a possible increased risk of ONJ.33 The patients with the greatest risk of developing ONJ seem to be those with MM who have received zoledronic acid for more than 24 months.47–49 By using adequate prophylactic measures, such as a careful baseline dental assessment, oral hygiene, and no invasive local procedures during the treatment with BP, the risk of ONJ decreases significantly.50, 51 However, in 2004, this information was not available, and the exact relation between zoledronic acid and ONJ was not clearly established. At that time, we decided to interrupt our study as a prudential measure, to decrease the risk of ONJ in otherwise asymptomatic patients. This choice, taken for ethical considerations, undoubtedly reduced the statistical power of the current trial and may have limited the validity of our conclusions, which will need confirmation in subsequent studies.

In conclusion, ‘prophylactic’ administration of zoledronic acid for 1 year in patients with untreated, early-stage AM was feasible, had predictable and manageable side effects, and decreased the development of SREs at the time of disease progression; however, it did not produce any clinical antitumor effects and did not reduce the risk of transformation into overt, ‘symptomatic’ MM. A recently started randomized trial by the GIMEMA (Italian Group for Adult Hematologic Diseases) that is characterized by a larger number of patients with AM, a longer treatment with zoledronate (2 years), a careful ONJ prophylaxis, and multiple endpoints, including instrumental (PET, MRI) and biologic (bone turnover markers) parameters, will provide further information regarding zoledronic acid treatment in this field.

Ancillary