while the concept that bisphosphonate could potentially ... docs... · web viewtitle: while the...
TRANSCRIPT
CONFIDENTIAL
IN PRESS IN ANTI-CANCER AGENTS IN MEDICINAL CHEMISTRY. NOT FOR
PUBLICATION OR CITATION
Bisphosphonates and Breast Cancer Prevention
Rowan T. Chlebowski, MD, PhD1; Nananda Col, MD, MPP, MPH, FACP2
1Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, Torrance, CA; 2Maine Medical
Center, Portland, ME
Target Journal: Anti-cancer Agents in Medicinal Chemistry
Previous Journal Submission: None
Address correspondence and reprint requests to
Rowan T. Chlebowski, MD, PhD
Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center
1124 W. Carson St.
Torrance, CA 90502
Phone: 310-222-2219
Facsimile: 310-320-2564
E-mail: [email protected]
Source(s) of support (in the form of grants, equipment, drugs, or all of these):
Running Title (limit: none): Bisphosphonates in Preventing Breast Cancer
Word Counts:Manuscript (limit 27,000 words) = 3,254Abstract (limit: 250 words) = 219
Number of Figures (limit: none) = 1
Number of Tables (limit: none) = 3
Number of References (limit: none) = 57
BP Anti-cancer RA DRAFT Page 2NPC ZOM 6148 2/23/11
ABSTRACT
Bisphosphonates are commonly used in patients with breast cancer to reduce skeletal-related events in
metastatic disease and to mitigate bone loss associated with cancer therapy in early stage disease. In addition,
adjuvant breast cancer trials evaluating the oral bisphosphonate clodronate suggested a reduction in cancer
recurrence, but the findings were mixed, with 2 positive and 1 negative report. In the Austrian Breast and Colorectal
Cancer Study Group (ABCSG) 12 study, adding the intravenous bisphosphonate zoledronic acid to endocrine
therapy in premenopausal breast cancer patients significantly prolonged disease-free survival versus endocrine
therapy alone (hazard ratio = 0.68; p = 0.008) at 62 months, and reduced local, regional, and distant recurrences.
Clinical trial findings from other adjuvant trials (Z-FAST, ZO-FAST), neoadjuvant studies, and studies involving
disseminated tumor cells (DTCs) are generally supportive of the ABCSG-12 conclusion, and recent data from
AZURE suggest the importance of menopausal status. Preclinical studies provide data on the mechanisms of action
that could mediate bisphosphonate direct and indirect anti-cancer effects. Recently, several observational studies (2
cohort studies and 2 case-control analyses) have associated oral bisphosphonate use with a lower breast cancer
incidence. Such reports require cautious interpretation because confounding by indication is an issue:
bisphosphonates are prescribed for women with low bone mineral density, and women with low bone density are at
decreased breast cancer risk.
Key words: adjuvant treatment, anti-cancer, bisphosphonates, early breast cancer, prevention, zoledronic acid
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 3NPC ZOM 6148 2/23/11
INTRODUCTION
Bisphosphonates have come into increasingly common use for preventing and treating osteoporosis [1].
Although the concept that bisphosphonates could also potentially influence breast cancer outcome is actually several
decades old [2], several recent clinical reports have prompted renewed interest in the question of bisphosphonate
impact on breast cancer prevalence in healthy women [3-5].
The potential for bisphosphonates to influence breast cancer outcome has preclinical support (as reviewed
by Winter [6]), with an increasing body of evidence supporting anti-cancer properties for bisphosphonates (as
reviewed by Green and Lipton [7]). These properties include cell apoptosis and proliferation, reduction of
angiogenesis, inhibition of tumor-cell invasion, activation of the immune system against cancer cells, synergy with
anti-cancer agents, as well as bone-mediated effects inhibiting osteoclast activity and preventing the release of tumor
growth factors [8-10]. As a result, adjuvant trials involving zoledronic acid (ZOL) were initiated that target not only
bone mineral density (BMD) maintenance but also breast cancer recurrence and overall survival.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 4NPC ZOM 6148 2/23/11
BISPHOSPHONATES AND CANCER TREATMENT-INDUCED BONE LOSS
Interest in the potential of bisphosphonates to reduce not only breast cancer recurrence but also breast
cancer incidence was kindled by results from the more recently reported trials evaluating adjuvant bisphosphonate
treatment in breast cancer. Previously, bisphosphonates such as pamidronate, clodronate, ibandronate, and ZOL
were shown to reduce the risk of skeletal-related events (SREs) in patients with bone metastases. Also, they have
been shown to increase BMD in women with bone loss [11-14]. More recently, aromatase inhibitors (AIs), which
substantially reduce circulating estrogen levels, have been associated with increased fracture risk, and were found to
be superior as adjuvant treatment compared with tamoxifen in postmenopausal women with early stage, hormone-
receptor–positive breast cancer [15]. However, the use of AIs is associated with decreases in BMD and increased
risk of fractures [15]. As a result, interest in mitigating the aromatase inhibitor-associated bone loss (AIBL) with
bisphosphonates emerged. In a series of trials with ibandronate [16,17], risedronate [18,19], and ZOL [20-22], and
the receptor activator of nuclear factor kappa B ligand (RANKL) inhibitor denosumab [23], all demonstrated an
ability to prevent and/or reverse AIBL.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 5NPC ZOM 6148 2/23/11
BISPHOSPHONATES AND BREAST CANCER RECURRENCES: CLODRONATE TRIALS
Initial clinical studies exploring the potential influence of bisphosphonates on breast cancer outcome were
directed at recurrence and progression. In 1987, Elomaa and colleagues described fewer new metastases after
administration of clodronate in a controlled but not randomized study in patients with breast cancer who had
established bone metastases [2]. The initial study of bisphosphonates in the adjuvant setting was reported by Diel
and colleagues in 1998 [24]. This study randomized 302 women with resected early breast cancer, to the oral
bisphosphonate clodronate (1,600 mg/day) or no bisphosphonate for 2 years. All patients received standard systemic
adjuvant therapy based on country guidelines, which could have included tamoxifen, CMF chemotherapy
(cyclophosphamide, methotrexate, and 5-fluorouracil), and/or goserelin. After 36 months of follow-up, the
clodronate group had a significant reduction in the number of bone metastases (p = 0.003) and distant metastases (p
= 0.003) compared with the control arm. Additionally, there was an overall survival benefit with clodronate (p <
0.001). Long-term follow-up (median, 103 months) demonstrated that only 20% of patients in the clodronate group
had died compared with 41% in the control group (p = 0.04) [25].
In contrast, a second similarly-sized study evaluating adjuvant clodronate reported substantially different
results. Saarto and colleagues examined the effects of clodronate (1,600 mg/day) versus no bisphosphonate for 3
years in 299 women with node-positive, early stage, resected breast cancer [26]. In this study, CMF chemotherapy
was administered to all premenopausal patients and tamoxifen to all postmenopausal patients. No differences in
bone metastasis frequency were reported between the treatment arms. Additionally, overall survival was lower in the
clodronate group (70% vs 83%, respectively; p = 0.009). This negative result is at least partially explained by an
imbalance in treatment patterns as well as estrogen receptor status between the groups (negative receptor status, 35%
vs 23%, respectively, for the clodronate and control groups). As a result of this, at least 12% more women in the
clodronate group received no effective adjuvant therapy, because their cancers most likely would not have
responded to endocrine therapy compared with women in the no-bisphosphonate group, potentially confounding the
study results.
In a third and the largest of the adjuvant clodronate trials, Powles and colleagues randomized 1,079 patients
with early stage breast cancer to standard adjuvant therapy and placebo or clodronate (1,600 mg/day) for 2 years
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 6NPC ZOM 6148 2/23/11
[27]. They reported a decrease in the incidence of bone metastases during the initial 2 years (p = 0.01) and increased
survival in the clodronate arm after a median follow-up of 5.6 years versus placebo (HR = 0.77, p = 0.048) [28].
Given these divergent clodronate results, the National Surgical Adjuvant Breast and Bowel Project (NSABP) is
currently conducting a fourth trial in which 3,200 early stage breast cancer patients receiving standard adjuvant
therapy have been randomized to placebo or clodronate (1,600 mg/day) for 3 years [29]. Accrual was completed in
March 2004, and results are awaited.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 7NPC ZOM 6148 2/23/11
BISPHOSPHONATES AND BREAST CANCER RECURRENCE: ZOLEDRONIC ACID TRIALS
Zoledronic acid has several studies that evaluated disease-free survival (DFS) and breast cancer recurrence
(as reviewed further by Gnant on page X). The ABCSG-12 trial randomized 1,803 premenopausal patients with
early stage breast cancer who all received goserelin 3.6 mg every 28 days plus tamoxifen (20 mg/day) or anastrozole
(1 mg/day), with or without ZOL (4 mg intravenously [IV] every 6 months) for 3 years [4]. Disease-free survival
was the primary endpoint. There was no DFS difference between anastrozole and tamoxifen. However, the women
receiving ZOL experienced significantly improved DFS (HR = 0.64; 95% confidence interval [CI] = 0.46, 0.91; p =
0.01) at 48 months’ median follow-up, and there was also a trend for increased overall survival with ZOL
administration (HR = 0.60; p =.011) [4]. There were also fewer bone metastases, distant metastases, local or regional
recurrences, and contralateral breast cancers with ZOL versus no ZOL.
Other recently reported clinical study results are generally supportive of a favorable association between
ZOL and breast cancer outcome. Three similarly designed randomized trials (Zometa-Femara Adjuvant Synergy
Trials: Z-FAST, ZO-FAST, and E-ZO-FAST) are evaluating upfront versus delayed use of ZOL in postmenopausal
patients with hormone-receptor–positive, early stage breast cancer [30]. These studies randomized patients to the AI
letrozole (2.5 mg/day) together with upfront ZOL (4 mg IV every 6 months) or letrozole together with delayed ZOL
(administered if BMD declined or fracture occurred) for 3 years. In Z-FAST, ZO-FAST, and E-ZO-FAST, upfront
ZOL mitigated loss of BMD associated with letrozole use [20,31]. In an early combined analysis of the Z-FAST and
ZO-FAST trials, upfront ZOL also had a favorable effect on DFS (defined as time to first appearance of breast
cancer recurrence or death from any cause) [3]. Fewer patients receiving upfront ZOL experienced disease
recurrence (7 patients [0.84%]) compared with patients receiving delayed ZOL (17 patients [1.9%]; p = 0.04) [3].
Eidtmann and colleagues have reported similar BMD benefits and a significant improvement in DFS for upfront
versus delayed ZOL (HR = 0.59; p = 0.031 at 36 months) [31]. However, when Coleman and colleagues reported a
combined analysis that included updated results and the similar E-ZO-FAST study [32], a Gail-Simon test was
statistically significant for heterogeneity in treatment effects between studies. Therefore, combining Z-FAST,
ZO-FAST, and E-ZO-FAST study results was felt not to be statistically appropriate [33]. In contrast, Mauri and
colleagues, using a somewhat earlier dataset and incorporating results from ABCSG-12, Z-FAST, ZO-FAST, E-ZO-
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 8NPC ZOM 6148 2/23/11
FAST, and including an additional small randomized trial, found a significant reduction in breast cancer recurrence
with ZOL use (odds ratio = 0.68; 95% CI = 0.48, 0.95; p = 0.025) [34,35]. As an exploratory analysis for this
commentary, we used updated results from the previously identified ZOL trials to generate a new meta-analysis
(Figure 1) [5,20,31-33,35]. We used MetaAnalyst software (Tufts-New England Medical Center, Boston, MA) to
pool data based on the random effects model of DerSimonian and Laird [36], and found no heterogeneity by
calculating the Q-statistic. Although the overall odds ratio was 0.79 for the risk of DFS events, favoring ZOL, these
pooled analyses did not reach statistical significance. However, duration and type of follow-up varied between these
datasets. It is recognized that 3 trials (Z-FAST, ZO-FAST, E-ZO-FAST) [33] are evaluating early versus delayed
ZOL, and all but 1 dataset [4,5] represent early reports.
Adjuvant treatment with ZOL in patients with breast cancer received additional evaluation in the AZURE
trial (BIG 01/04), and initial results were recently presented at the 2010 San Antonio Breast Cancer Symposium
[37]. The AZURE trial enrolled 3,360 patients with stage II-III invasive breast cancer who had no evidence of
metastases and who had not received prior bisphosphonate therapy during the last year. Participants were
randomized to standard therapy including chemotherapy and or hormone therapy at investigator discretion, or
standard therapy plus ZOL 4 mg at a schedule of every 3 to 4 weeks × 6 doses, then every 3 months × 8 doses, then
every 6 months × 5 doses, thus completing 5 years of therapy. The prospectively defined primary endpoint was
DFS, and secondary endpoints included OS. In the overall population, baseline characteristics, including distribution
of therapy, were well balanced between groups. Notably, more than 95% received adjuvant chemotherapy. At a
median follow-up of approximately 59 months, there was no significant difference in DFS in the overall population
(HR = 0.98; 95% CI = 0.85, 1.13; p = 0.79). However, there was a trend toward improved OS with ZOL versus no
ZOL (HR = 0.85; 95% CI = 0.72, 1.01; p = 0.07).
In pre-planned analyses, there was heterogeneity of ZOL effect on DFS by menopausal status
(heterogeneity p = 0.02) [37]. Among patients who were postmenopausal for > 5 years, ZOL significantly reduced
the risk of DFS events compared with control (HR = 0.76; 95% CI = 0.60, 0.98), and this benefit was not seen in the
remaining participants who were premenopausal, < 5 years postmenopausal, or of unknown status (HR = 1.13; 95%
CI = 0.95, 1.35). Similarly, ZOL significantly reduced the risk of death compared with control among patients who
were postmenopausal for > 5 years or > 60 years of age (HR = 0.71; 95% CI = 0.54, 0.94; p = 0.02). Overall, the
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 9NPC ZOM 6148 2/23/11
combination of adjuvant therapy plus ZOL was generally well tolerated; however, 17 confirmed cases of
osteonecrosis of the jaw were reported in the ZOL group (1.16%).
The AZURE investigators concluded that adjuvant use of ZOL at this dose and schedule did not improve
DFS in breast cancer patients treated with adjuvant chemotherapy [37]. However, a trend favoring ZOL use for OS
emerged. In addition, heterogeneity of effect by menopausal status suggested that ZOL significantly improved DFS
and OS in those > 5 years from their last menstrual period. These findings, in conjunction with those of ZO-FAST
and ABCSG-12, support the hypothesis that ZOL’s beneficial effect on breast cancer outcomes may be dependent
on a low estrogen concentration in the bone microenvironment.
Other smaller randomized trials have reported on the influence of ZOL on breast cancer. In a neoadjuvant
substudy (n = 205) of the larger Adjuvant Zoledronic Acid to Reduce Recurrence (AZURE) clinical trial (N =
3,360), patients with breast cancer randomized to ZOL (4 mg IV with each chemotherapy cycle) had smaller tumors
at resection and increased frequency of pathologic complete response compared with patients who received
chemotherapy alone [38]. In 3 translational studies, ZOL reduced residual cancer burden by reducing disseminated
tumor cells [39-41].
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 10NPC ZOM 6148 2/23/11
BISPHOSPHONATES AND BREAST CANCER INCIDENCE
Against the background of information such as the ABCSG-12 results presented in 2008, 4 separate groups
decided to conduct analyses of the association between bisphosphonate use in the osteoporosis setting and breast
cancer incidence in observational studies. However, as bisphosphonates are prescribed for women with low BMD, a
potential confounding problem arises. Perhaps because of the correlation between estrogen levels and BMD, women
with bone loss are at substantially lower breast cancer risk [42,43], and bisphosphonate use would be expected to be
associated with lower breast cancer risk because of this correlation. Thus, a means of adjusting for differences in
BMD between bisphosphonate use and non-use is needed, if reliable risk estimates are to be established.
Chlebowski and colleagues were able to address this issue in their analysis in the Women’s Health Initiative
(WHI) cohort of 154,768 postmenopausal women [44]. Of these participants, 10,418 women had a BMD assessment
at study entry as part of a substudy conducted at 4 of the 40 WHI clinical centers. The WHI investigators then
compared a hip fracture prediction score calculated from a validated analytic model, which incorporated clinical
findings but not overall BMD to total-hip BMD, in participants who had both determinations [45]. A strong
statistically significant (p < 0.001) correlation was seen between hip fracture prediction score and BMD. Thus, in a
multivariate analysis, the hip fracture prediction score, which was available in the entire WHI cohort, was used to
adjust for potential BMD differences between bisphosphonate use and non-use. After 7.8 mean years of follow-up,
invasive breast cancer incidence was 32% lower in participants who received bisphosphonates after adjusting for hip
fracture prediction score, menopausal hormone therapy use, and breast cancer risk factors (p < 0.01). Interestingly, a
similar lower incidence of both estrogen-receptor–positive and estrogen-receptor–negative breast cancers was
observed in participants who received bisphosphonates, although the latter association was not statistically
significant. In contrast, the incidence of ductal carcinoma in situ was significantly greater in participants who
received bisphosphonates. The authors concluded that oral bisphosphonate use was associated with significantly
lower invasive breast cancer incidence, suggesting that bisphosphonates may have inhibiting effects on breast cancer
[44].
Two case-control studies examining the association between bisphosphonate use and breast cancer
incidence were also recently reported [46,47]. Newcomb and colleagues, in their analysis of women with breast
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 11NPC ZOM 6148 2/23/11
cancer and age-matched controls, made adjustments for body mass index and postmenopausal hormone therapy use
[46]. To address potential BMD differences, adjustments were made for adult height loss and reported physician-
diagnosed osteoporosis. These variables would likely allow incomplete adjustment for BMD differences because
height loss is a late finding in the course of bone loss, and physician-reported osteoporosis would not account for
women not seeking assessment of their BMD. Nevertheless, the incidence of breast cancer was decreased with
bisphosphonate use versus no bisphosphonate use (4.4% vs 6.2%, respectively; odds ratio = 0.67; 95% CI = 0.51,
0.89) [46]. In the case control study of Rennert and colleagues, body mass index and breast cancer risk factors were
incorporated in the analyses, but no variables allowing for more direct adjustment of BMD differences were
available [47]. This study also reported fewer breast cancer cases in patients who received bisphosphonates (odds
ratio = 0.72; 95% CI = 0.57, 0.90).
An additional cohort study provides somewhat mixed results. From Danish registers, Vestergaard et al
collected records for 87,104 female bisphosphonate recipients; each patient in this cohort was compared with 3 age-
matched controls from the general population who were not receiving bisphosphonates (n = 261,322) [48].Reduced
risk of breast cancer was observed after patients initiated alendronate (HR = 0.53; 95% CI = 0.38, 0.73) or etidronate
(HR = 0.80; 95% CI = 0.73, 0.89). However, as no dose-response relationship was seen with the 2 agents, the
authors concluded that “no causal relationship seemed to be present” [48]. No adjustment for potential BMD
difference was made in this analysis. The comparable findings of the 3 observational studies and the 1 study with
somewhat mixed results are outlined in Table 1 [44,46-48].
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 12NPC ZOM 6148 2/23/11
ADJUVANT BISPHOSPHONATE CLINICAL TRIALS ANALYZING FOR BREAST CANCER
PREVENTION
Despite the promising results cited above, it will likely be difficult to conduct a full-scale bisphosphonate
primary prevention study given the current broad use of bisphosphonates for bone health indications, especially in
the United States. However, the ongoing randomized clinical studies using bisphosphonates in the breast cancer
adjuvant therapy setting provide an opportunity to evaluate bisphosphonate influence on contralateral breast cancer
as a surrogate for primary breast cancer risk reduction. In addition to the previously reported ABCSG-12 trial, 4
additional randomized trials are evaluating ZOL in a variety of schedules for patients with early stage resected breast
cancer (AZURE, SUCCESS, SWOG 0307, and NATAN) [5,49,50]. These adjuvant trials will enroll more than
14,000 breast cancer patients and provide ample opportunity to assess clodronate, ibandronate, and ZOL influence
on contralateral breast cancer development (Table 2) [5,49,50].
There are also 3 phase III trials evaluating other antiabsorptives versus placebo in early breast cancer
(Table 3) [5,49,50]. The NSABP B-34 study randomized 3,323 patients with early stage, resected breast cancer
receiving standard adjuvant therapy to placebo or oral clodronate (1,000 mg/day) [29]. The German Breast Group
(GBG) trial 33 (GAIN) will randomize 3,000 patients with node-positive stage II-III breast cancer receiving
standard adjuvant therapy to either oral ibandronate or placebo [50]. ABCSG-18 (NCT00556374) is randomizing
3,400 postmenopausal women with endocrine-responsive breast cancer and adjuvant aromatase inhibitor treatment
to receive either placebo or the RANKL inhibitor denosumab. Finally, the D-CARE trial is in the process of
randomizing 4,500 patients with early stage breast cancer receiving standard adjuvant therapy to either placebo or
denosumab [51].
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 13NPC ZOM 6148 2/23/11
CONSIDERATIONS WITH BISPHOSPHONATE USE IN BREAST CANCER PREVENTION
Although preclinical and clinical results suggest that several bisphosphonates may influence breast cancer
outcome, a question can be raised regarding whether a bisphosphonate class effect will emerge. Only 1 of the
previously presented studies involves an active comparison of different bisphosphonates (SWOG 0307, comparing
ZOL, clodronate and ibandronate). However, in a phase III trial in multiple myeloma, patients with stage I, II, or III
disease were randomized to either ZOL (4 mg every 3 to 4 weeks) or daily oral clodronate. Patients who received
ZOL had significantly longer (p = 0.01) survival than patients who received clodronate, suggesting that differences
may also emerge in the breast cancer setting [52].
Despite the range of potential mediating mechanisms identified in preclinical studies, the well-described
influence on bone turnover could be a major factor in the observed bisphosphonate influences on breast cancer
incidence and outcome. In this regard, it is noteworthy that the 2 agents approved for breast cancer risk reduction in
the United States, tamoxifen and raloxifene, both result in reduced bone turnover [53-55]. As bisphosphonates and
selective estrogen receptor modulators share this common bone influence, their relative influence on breast cancer
outcomes will be addressed in ongoing trials. In ABCSG-12, early results suggest that ZOL may add to tamoxifen’s
favorable influence on breast cancer DFS [5].
Finally, bisphosphonates are not free of risk, with the potential for renal toxicity [13], osteonecrosis of the
jaw [56], and, rarely, atypical femoral fractures [57]. Oral bisphosphonate are commonly associated with
gastrointestinal problems [13]. The ongoing, full-scale, randomized, controlled bisphosphonate adjuvant trials will
provide important information regarding the risk-to-benefit ratio of moderate-term bisphosphonate use to influence
breast cancer outcome.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 14NPC ZOM 6148 2/23/11
CONCLUSIONS
In randomized adjuvant breast cancer trials, ZOL likely reduces breast cancer recurrence risk, suggesting
breast tumor inhibitory effects. Preclinical studies provide several potential mechanism of action for the direct breast
tumor inhibitory effects of bisphosphonates (as reviewed by Clezardin on page X). Three recent observational
studies provide consistent reports associating bisphosphonate use with lower breast cancer incidence, but 1 cohort
study reported somewhat inconsistent findings. Ongoing randomized clinical trials evaluating bisphosphonates and
other antiresorptive agents may provide more definitive evidence regarding bisphosphonate use and breast cancer
incidence based on contralateral breast cancer findings.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 15NPC ZOM 6148 2/23/11
LIST OF ABBREVIATIONS
ABCSG = Austrian Breast and Colorectal Cancer Study Group
AI = aromatase inhibitor
AIBL = aromatase inhibitor-associated bone loss
AZURE = Adjuvant Zoledronic Acid to Reduce Recurrence
BMD = bone mineral density
CI = confidence interval
CLO = clodronate
CMF = cyclophosphamide, methotrexate and 5-fluorouracil
DFS = disease-free survival
DOC = docetaxel
FEC = 5-fluorouracil, epirubicin, cyclophosphamide
GBG = German Breast Group
GEM = gemcitabine
IBAN = ibandronate
IV = intravenous(ly)
NSABP = National Surgical Adjuvant Breast and Bowel Project
OR = odds ratio
RANKL = receptor activator of nuclear factor kappa B ligand
SWOG = Southwest Oncology Group
WHI = Women’s Health Initiative
Z-FAST, ZO-FAST, E-ZO-FAST = Zometa-Femara Adjuvant Synergy Trials
ZOL = zoledronic acid
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 16NPC ZOM 6148 2/23/11
CONFLICT OF INTEREST
Rowan Chlebowski is a consultant for Novartis and on their speakers bureau. He also is a consultant for Amgen and
has grant funding support from that company.
ACKNOWLEDGMENTS
Financial support for formatting assistance was provided by Novartis Pharmaceuticals. Dr. Chlebowski takes
responsibility for the content of the manuscript. He provided the initial draft of the manuscript and provided critical
revisions of the manuscript for important intellectual content. We thank ProEd Communications, Inc., for their
assistance with formatting and submitting this manuscript.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 18NPC ZOM 6148 2/23/11
FIGURE LEGEND
Figure 1. Risk of disease-free survival (DFS) events between zoledronic acid and control groups in recent trials
[5,20,31-33,35]. Disease-free survival was defined as breast cancer recurrence or death from any cause. Squares
represent odds ratio with the size proportional to the trial size. The 95% confidence intervals (CIs) are represented
by horizontal bars. Odds ratios < 1 favor zoledronic acid. Abbreviations: ABCSG, Austrian Breast and Colorectal
Cancer Study Group; Z-FAST, ZO-FAST, E-ZO-FAST, Zometa-Femara Adjuvant Synergy Trials.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 19NPC ZOM 6148 2/23/11
TABLES
Table 1. Association Between Bisphosphonate Use and Breast Cancer Incidence in Observational Studies
Study Study Design Breast Cancer Association (95% CI)
Rennert G, et al [47] Case-control OR = 0.72 (0.57, 0.90)
Newcomb PA, et al [46] Case-control OR = 0.67 (0.51, 0.89)
Chlebowski RT, et al [44] Cohort HR = 0.68 (0.52,0.88)
Vestergaard P, et al [48] Cohort HR = 0.53 (0.38, 0.73), alendronate; HR = 0.80 (0.73, 0.89), etidronate; but no dose-response relationship seen
Abbreviations: CI, confidence interval; HR, hazard ratio; OR, odds ratio.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 20NPC ZOM 6148 2/23/11
Table 2. Randomized Clinical Trials Evaluating Zoledronic Acid (ZOL) in Women With Early Stage Breast Cancer
Trial Patients, n Treatment Arms Primary Endpoint
ABCSG-12 [5] 1,803 (stage I, II)
TAM; ANA; TAM + ZOL; ANA + ZOL
DFS at 5 yr
AZURE [49] 3,360 (stage II, III)
Standard therapy ± ZOL DFS at 5 yr
SUCCESS [50] 3,754 (stage I, II, III)
FEC + DOC then endocrine therapy + ZOL; FEC + DOC + GEM then endocrine therapy + ZOL
DFS at 5 yr
SWOG-0307 [50] 4,500 (stage I, II, III)
ZOL; CLO; IBAN DFS at 3 yr
NATAN [50] 654 (stage II, III)
Standard therapy ± ZOL EFS at 5 yr
Abbreviations: ABCSG, Austrian Breast and Colorectal Cancer Study Group; ANA, anastrozole; AZURE, Adjuvant Zoledronic Acid to Reduce Recurrence; CLO, clodronate; DFS, disease-free survival; DOC, docetaxel; FEC, 5-fluorouracil, epirubicin, cyclophosphamide; GEM, gemcitabine; IBAN, ibandronate; SWOG, Southwest Oncology Group; TAM, tamoxifen.
CONFIDENTIAL — Working Draft
BP Anti-cancer RA DRAFT Page 21NPC ZOM 6148 2/23/11
Table 3. Randomized Clinical Trials Evaluating Clodronate, Ibandronate, and Denosumab in Women With Early Breast Cancer
Trial Patients, n Treatment Arms Primary Endpoint Status
NSABP-B34 [29] 3323 (stage I, II)
Clodronate vs placebo
DFS Active, not recruiting
D-CARE [51] 4500 (stage II, III)
Denosumab vs placebo
BMFS Recruiting
GAIN GBG33 [50] 3000 (stage II, III)
Ibandronate vs placebo
DFS Recruiting
Abbreviations: BMFS, bone-metastases–free survival; DFS, disease-free survival; GBG, German Breast Group; NSABP, National Surgical Adjuvant Breast and Bowel Project.
CONFIDENTIAL — Working Draft