ZOMETA^â (Zoledronic acid for injection)

 

 

 

 

 

ONCOLOGIC DRUGS ADVISORY COMMITTEE

BRIEFING DOCUMENT

 

 

 

 

Date of Meeting: January 31, 2002

 

 

 

 

NDA # 21-386

Treatment of bone metastases

 

 

 

 

 

 

 

Version: Final

 

Release date: January 3, 2002

 

 

AVAILABLE FOR PUBLIC DISCLOSURE WITHOUT REDACTION.

Summary

Many forms of cancer can originate in or metastasize to bone.  Cancers that involve bone can cause a variety of skeletal problems for patients, ranging from pain to fractures to spinal cord compression.  The majority of patients living with bone metastases have breast cancer or prostate cancer (since these cancers are common in our population, frequently metastasize to bone, and are often associated with a relatively longer survival than other cancers following metastasis to bone).  Other cancers  associated with frequent metastasis to bone include lung cancers, renal cancer, melanoma and thyroid cancer.  Bone complications are also common in patients with multiple myeloma, a hematologic malignancy that originates in bone.  In the US, it is thought that over 250,000 patients are living with bone metastasis at any time.

Damage to bone by cancer can result from direct effects of tumor growth (replacing/weakening the bone and/or compromising adjacent structures) and from indirect effects mediated by factors released by the tumor (which may promote osteoclast resorption of bone or abnormal bone growth/remodeling).  Effective systemic anticancer treatments can control tumor growth which, for a time, may stem the bony destruction and ameliorate or reverse damage to bone, with improvement in related signs and symptoms.  Unfortunately, many types of cancer that involve bone are poorly responsive to current antineoplastic treatments.  Patients with this diagnosis may often  suffer from morbid skeletal events, requiring palliative treatments such as pain medication, radiation therapy, or surgery.  These palliative treatments have their own limitations and potential for adverse consequences, and patients may be left with significant disability and loss of mobility and independence.

An alternative approach to control bone damage due to cancer has been to utilize intravenous bisphosphonate therapy to block tumor-stimulated osteoclast activity.  ArediaÒ (pamidronate disodium for injection) is marketed for treatment of patients with multiple myeloma or breast cancer metastatic to bone, where it has been shown to delay or reduce the occurrence of skeletal-related morbid events.  The utility of bisphosphonate treatment in reducing skeletal-related morbid events in patients with other forms of cancer has not previously been established, and there is no U.S.-approved bisphosphonate treatment for these patients.  In particular, prospective studies in prostate cancer failed to detect a beneficial effect of Aredia on skeletal-related events (1).  It is nonetheless anticipated that effective inhibition of osteoclast activity should have beneficial effects regardless of tumor type, since excessive osteoclast activity plays a central role in the pathophysiology of all (osteolytic, mixed, and osteoblastic) malignant bone lesions.  The magnitude of the observed benefit, however, may depend on many factors, including the potency of the bisphosphonate effect, the types and levels of osteoclast-activating factors elaborated by different tumor types, and the biology of the patient’s disease in other tissues (with competing causes of early morbidity/mortality in patients with rapid disease progression in other tissues).

Intravenous bBisphosphonates are generally well tolerated and can be administered safely to patients with a variety of advanced tumor types.  The most commonly noted side effect is a transient flu-like syndrome characterized by fever, arthralgias, myalgias, and chills.  Nausea, fatigue, and headache are among the other more common adverse events.  Bisphosphonate administration may also be associated with impairment of renal function, a complication that may be related to complexes between bisphosphonate and minerals in the kidneys.  Elevation of serum creatinine and (rarely) renal failure may occur in some patients.  Risk factors are believed to include insufficient hydration of the patient, rapid infusionof the bisphosphonate, and higher bisphosphonate drug dosages.

ZOMETA (Zometa for infusion, Novartis Pharmaceuticals Corporation, East Hanover, NJ) is a new-generation bisphosphonate that was selected for clinical development based on its extremely high pharmacological potency. as well as preclinical studies suggesting a potentially better margin of renal safety.  In previous clinical trials, Zometa has been shown to be superior to Aredia in the treatment of hypercalcemia of malignancy, providing a higher response rate and longer response duration.  Zometa is registered worldwide for this indication (2,3).

The Zometa study clinical development program is the largest and most comprehensive bisphosphonate clinical trial program conducted to date in patients with malignancies involving the skeletal system.. The study designs and clinical endpoints were discussed with the FDA prior to the start of the clinical trials   The study designs and clinical endpoints for these studies were revieweddiscussed with the FDA and agreed upon prior to the start of the clinical trials.  The clinical trials reported in this application evaluated Zometa versus Aredia in patients with multiple myeloma or breast cancer metastatic to bone (study 010).  and Zometa was compared with versus placebo in patients with prostate cancer (study 039) or and lung cancer or other solid tumors (study 011) metastatic to bone.  While variations were seen in the statistical significance of the treatment effects observed in the different studies and different analyses, Tthe study data are consistent in demonstrating a clinical benefit for Zometa in delaying and reducing the occurrence of skeletal related events (SREs) for the broad range of all tumor types studied.  The primary analyses compared the proportions of patients who experienced any SRE in the respective arms of each study.  Analyses of the time to first SRE and comparisons of the rate and recurrence rate of SREs over time on study (secondary analyses of this primary endpoint) provided further support for Zometa effectiveness across all tumor types studied.

Overall, no significant efficacy differences were observed between 4 mg and 8 mg Zometa doses in these studies, suggesting that the 4 mg dose is sufficient to provide maximal effectiveness.  AlsoRenal safety concerns resulted in a decision to increase the Zometa administration time from 5 minutes (in the original study protocols) to 15 minutes.  Over the course of these studies, the 8 mg dose was associated with an increased incidence of creatinine elevations, and in the later stages of this research all 8 mg Zometa patients were converted to the 4 mg dose.  Finally, renal safety concerns also resulted in a decision to increase the Zometa administration time from 5 minutes (in the original study protocols) to 15 minutes over the course of this research.  The safety profile of Zometa 4 mg administered over 15 minutes is comparable to that of other intravenous bisphosphonatesof Aredia.  The incidence of serum creatinine elevations in patients receiving Zometa 4 mg over 15 minutes was similar to that seen with administration of Aredia 90 mg over 2 hours in breast cancer and multiple myeloma patients and was similar to that seen with placebo in prostate cancer patients. It; and was only slightly greater than that seen with placebo in the other solid tumor patients.  A Finally a practical benefit associated with Zometa is the short 15 minute infusion time compared with the 2-4 hours required to administer Aredia to breast cancer or multiple myeloma patients.

In conclusion, the studies provided in this application demonstrate the safety and effectiveness of Zometa 4 mg infused over 15 minutes (repeated at 3-4 week intervals) to delay or reduce the occurrence of skeletal-related events in patients with cancer metastatic to bone or multiple myeloma. The results were consistent in that both Zometa treatment arms in all three protocols reduced the proportion of patients having at least one SRE. This reflects the benefit patients receive by having a longer initial period free of skeletal complications, and by having fewer additional complications over time. .

 

Table of Contents

1     Background......................................................................................................................... 10

1.1       Clinical consequence of skeletal related events.......................................................... 10

1.2       Pathophysiology...................................................................................................... 10

1.3       Overview of Zometa................................................................................................ 11

2     Clinical Pharmacokinetics and Metabolism........................................................................... 12

2.1.      Clinical pharmacology studies................................................................................... 12

2.2.      Metabolism and drug-drug interactions..................................................................... 14

2.3.      Pharmacokinetics in renal impairment....................................................................... 14

3     Use in Hypercalcemia of Malignancy.................................................................................... 15

3.1       Efficacy in HCM...................................................................................................... 15

3.2       Safety in HCM........................................................................................................ 15

4     Phase II Trial in treatment of bone metastases....................................................................... 15

4.1       Efficacy................................................................................................................... 16

4.2       Safety...................................................................................................................... 16

5     Phase III Trials.................................................................................................................... 16

5.1       Study Design........................................................................................................... 16

5.2       Breast cancer and multiple myeloma – Aredia-controlled trial (010).......................... 25

5.2.1 ....... Patient disposition................................................................................... 25

5.2.2 ....... Demographics and disease characteristics................................................ 25

5.2.3........ Primary efficacy analysis.......................................................................... 27

5.2.4........ Secondary skeletal endpoints analyses..................................................... 28

5.2.5........ Analysis of Other Endpoints.................................................................... 30

5.2.6........ Survival................................................................................................... 32

5.2.7........ Efficacy Summary (breast cancer / multiple myeloma)............................... 33

5.3       Prostate Cancer – Placebo-controlled trial (039)...................................................... 34

5.3.1........ Patient Disposition................................................................................... 34

5.3.2........ Demographic and Disease Characteristics................................................ 35

5.3.3........ Primary efficacy analysis.......................................................................... 36

5.3.4........ Secondary skeletal endpoint analyses....................................................... 37

5.3.5........ Analyses of Other Endpoints................................................................... 39

5.3.6........ Survival................................................................................................... 40

5.3.7........ Efficacy Summary (prostate study)........................................................... 41

5.4       Lung Cancer and Other Solid Tumors – Placebo-Controlled Trial (011)................... 43

5.4.1........ Patient Disposition................................................................................... 43

5.4.2........ Demographics and Disease Characteristics.............................................. 43

5.4.3........ Primary efficacy analysis.......................................................................... 45

5.4.4........ Secondary skeletal endpoint analyses....................................................... 46

5.4.5........ Analyses of Other Endpoints................................................................... 49

5.4.7    Efficacy Summary (other solid tumors)..................................................................... 51

6     Phase III Studies – Safety.................................................................................................... 53

6.1       Adverse Events....................................................................................................... 53

6.1.1........ Overall Incidence in Phase II/III bone metastases studies......................... 53

6.1.2........ Grade 3 and 4 Adverse Events................................................................ 54

6.1.3........ Serious Adverse Events........................................................................... 56

6.2       NCI Grade 3 and 4 Laboratory Abnormalities......................................................... 56

6.3       Renal Effects........................................................................................................... 58

6.3.1........ Renal Adverse Events in Phase II/III Bone Metastases Studies................. 58

6.3.4........ Grade 3 and 4 Creatinine Values............................................................. 60

6.3.5........ Renal Function Deterioration – Advisory Board Criteria........................... 61

6.3.6........ Time to Creatinine Increase..................................................................... 62

6.4       Deaths..................................................................................................................... 67

6.4.1........ Primary Cause of Death......................................................................... 67

7.    Summary of Safety.............................................................................................................. 68

8.    Summary of Effectiveness.................................................................................................... 68

9.    Benefit-Risk Assessment...................................................................................................... 71

10.  Overall Conclusion.............................................................................................................. 72

11.  References.......................................................................................................................... 73

ATTACHMENT 1: COMPOSITION OF DATA SAFETY MONITORING BOARD (DSMB)  75

ATTACHMENT 2: COMPOSITION OF RENAL ADVISORY BOARD (RAB).............. 76

ATTACHMENT 3:............................................................................................................. 77

PROPOSED ZOMETA LABELING (including HCM)....................................................... 77

 

List of tables

Table 2-1.                    Human pharmacokinetic studies............................................................... 12

Table 5-1.                    Overview of Phase III studies.................................................................. 17

Table 5-2.                    Efficacy parameters used in Phase III studies........................................... 20

Table 5-3.                    Patient disposition by treatment group, Aredia-controlled trial................... 25

Table 5-4.                    Baseline demographic characteristics – Aredia-controlled trial – ITT population  26

Table 5-5.                    Baseline disease characteristics of multiple myeloma patients.................... 26

Table 5-6.                    Baseline disease characteristics of breast cancer patients.......................... 27

Table 5-7.                    Proportion of patients with SREs, multiple myeloma and breast cancer..... 28

Table 5-8.                    Time to the first SRE up to Month 13 (Median and hazard ratio), multiple myeloma and breast cancer..................................................................................................... 29

Table 5-9.                    Multiple-event analysis of time to SRE up to Month 13, multiple myeloma and breast cancer (010)...................................................................................................... 30

Table 5-10.                  Mean/Median changes from baseline in quality of life scores at Month 13 by treatment group............................................................................................................... 31

Table 5-11.                  Time to progression of disease up to Month 13, multiple myeloma and breast cancer      31

Table 5-12.                  Best bone lesion response up to Month 13, multiple myeloma and breast cancer 32

Table 5-13.                  Summary of median survival (days) by stratum and treatment group.......... 32

Table 5-14.                  Summary of analysis of skeletal related events in Study 010 (Breast Cancer and Multiple Myeloma)............................................................................................... 34

Table 5-15.                  Patient disposition – prostate cancer study............................................... 35

Table 5-16                   Demographics in prostate cancer patients – ITT population...................... 35

Table 5-17.                  Baseline characteristics in prostate cancer patients (ITT population).......... 36

Table 5-18.                  Proportion of patients with prostate cancer with an SRE up to Month 15.. 37

Table 5-19.                  Time to the first SRE up to Month 15 in prostate cancer patients.............. 38

Table 5-20.                  Skeletal morbidity rate (risk set definition) of any SRE (-HCM) up to Month 15 38

Table 5-21.                  Multiple-event analysis of time to SRE up to Month 15 in prostate cancer patients (039) 39

Table 5-22.                  Summary of time to progression of disease at Month 15 in prostate cancer patients         40

Table 5-23.                  Frequency distribution of best response by treatment group...................... 40

Table 5-24.                  Summary of median survival (days) by stratum and treatment group.......... 41

Table 5-25.                  Summary of analysis of skeletal related events in Study 039 (Prostate cancer)     42

Table 5-26.                  Patient disposition, lung cancer and other solid tumors (011).................... 43

Table 5-27.                  Demographic characteristics by treatment group, Study 011 (ITT population)     44

Table 5-28.                  Baseline disease characteristics in patients with lung cancer or other solid tumors (011)   45

Table 5-29.                  Proportion of patients with lung cancer or other solid tumors having any SRE (011)       46

Table 5-30.                  Time to first SRE in patients with lung cancer and other solid tumors (011) 47

Table 5-31.                  Skeletal morbidity rate of any SRE in patients with lung cancer and other solid tumors  (011)............................................................................................................... 48

Table 5-32.                  Proportion of patients with lung cancer or other solid tumors having any SRE (+HCM)   48

Table 5-33.                  Multiple-event analysis of time to SRE up to Month 9 in patients with lung cancer and other solid tumors (011)................................................................................... 49

Table 5-34.                  Time to progression of disease in patients with lung cancer and other solid tumors (011) 50

Table 5-35.                  Frequency distribution of best bone lesion response (011)........................ 50

Table 5-36.                  Summary of median survival (days) by stratum and treatment group.......... 51

Table 5-37.                  Summary of analysis of skeletal related events in Study 011 (Lung cancer and other solid tumors)................................................................................................... 52

Table 6-1.                    No. (%) of patients with most frequent adverse events (³ 15%) – Phase II/III studies     54

Table 6-2.                    No. (%) of patients with grade 3 or 4 AEs (³ 5%) – Phase II/III studies.. 55

Table 6-3.                    No (%) of patients with SAEs (³ 5%) – Phase II/III Studies.................... 56

Table 6-4.                    No. (%) of patients with grade 3 or 4 serum chemistry values – Phase II/III studies        57

Table 6-5.                    No. (%) of patients with grade 3 or 4 hematology laboratory abnormalities – Phase II/III bone metastases studies................................................................................... 58

Table 6-6.                    No. (%) of patients with AEs (³ 1%) of the renal and urinary system – pre-15-minute infusion amendment............................................................................................. 59

Table 6-7.                    No. (%) of patients with AEs (³ 1%) of the renal and urinary system – post-15-minute infusion patients................................................................................................... 60

Table 6-8.                    No. (%) of patients with grade 3 or 4 serum creatinine values pre- and post-15-minute infusion amendment – Phase II/III bone metastases studies................................... 61

Table 6-9.                    No. (%) of patients with notable serum creatinine values pre- and post-15-minute infusion amendment – Phase II/III bone metastases studies................................... 62

Table 6-11.                  Primary causes of death (by body system) in Phase II/III studies.............. 67

Table 8-1.                    Proportion of patients with one or more SREs (Phase III trials, primary analyses)           69

Table 8-2.                    Median Time to first SRE (Phase III trials)............................................... 69

Table 8-3.                    Skeletal Morbidity Rate (rate of SREs over time, Phase III studies).......... 70

Table 8-4.                    Risk ratios compared to control, Multiple Event Analyses of SREs (Phase III studies)     70

 

List of figures

Figure 2-1.                   Mean Zometa plasma concentrations over 4 hours (left panel) and 24 hours (right panel)   13

Figure 2-2.                   Relationship between the renal clearance of Zometa (Clr) and creatinine clearance (CLcr) in 64 patients in 3 clinical pharmacology studies................................................ 14

Figure 5-1                    Individual SREs in breast/myeloma patients.............................................. 28

Figure 5.2                    Mean skeletal morbidity rate by treatment group...................................... 29

Figure 5-3.                   Kaplan-Meier curve of survival by treatment group.................................. 33

Figure 5-4.                   Individual SREs in patients with prostate cancer....................................... 37

Figure 5-5.                   Kaplan-Meier curve of survival by treatment group.................................. 41

Figure 5-6.                   Individual SREs in patients with lung cancer or other solid tumors............. 47

Figure 5-7.                   Kaplan-Meier curve for survival by treatment group................................. 51

Figure 6-1.                   Kaplan-Meier curve for time to the first renal function deterioration by treatment group for the pre 15-minute infusion Amendment patients (Protocol 010, Safety evaluable patients)     64

Figure 6-2.                   Kaplan-Meier curve for time to the first renal function deterioration by treatment group for the post 15-minute infusion Amendment patients (Protocol 010, Safety evaluable patients)   64

Figure 6-3.                   Kaplan-Meier curve for time to the first renal function deterioration by treatment group for the pre 15-minute infusion Amendment patients (Protocol 011, Safety evaluable patients)     65

Figure 6-4.                   Kaplan-Meier curve for time to the first renal function deterioration by treatment group for the post 15-minute infusion Amendment patients (Protocol 011, Safety evaluable patients)   65

Figure 6-5.                   Kaplan-Meier curve for time to the first renal function deterioration by treatment group for the pre 15-minute infusion Amendment patients (Protocol 039, Safety evaluable patients)     66

Figure 6-6.                   Kaplan-Meier curve for time to the first renal function deterioration by treatment group for the post 15-minute infusion Amendment patients (Protocol 039, Safety evaluable patients)   66

 

 

Glossary of Abbreviations and Definition of Terms

 

 

AE	Adverse Event
ANCOVA	Analysis of covariance
ANOVA	Analysis of variance
b.i.d.	Bis in diem/Twice daily
BMD	Bone mineral density
BP	Bisphosphonate
BPI	Brief Pain Inventory (composite score)
CMH	Cochran-Mantel Haenszel
CR	Complete response
CRF	Case report form
DSMB	Drug Safety Monitoring Board
ECOG	Eastern Cooperative Oncology Group
FACT-G	Functional Assessment of Cancer Therapy-General
GLM	General linear model
HCM	Hypercalcemia of malignancy (also known as TIH)
IEC	Independent Ethics Committee
IMN	International Medical Nomenclature
IRB	Institutional Review Board
ISE	Integrated Summary of Efficacy
ISS	Integrated Summary of Safety
ITT	Intent-to-treat (population)
i.v.	intravenous
KM	Kaplan-Meier
NDA	New Drug Application
NSLC	Non - small cell lung cancer
o.d.	Omni die/once per day
OST	Other solid tumor
pam	Pamidronate (Aredia)
p.o.	Per os/by mouth/orally
PSA	Prostate specific antigen
QOL	Quality of life
RAB	Renal Advisory Board
SAE	Serious Adverse Event
SMR	Skeletal-morbidity rate
SOP	Standard operating procedures
S.D.	Standard deviation
SRE	Skeletal-related event
SRE (- HCM)	Skeletal-related events excluding HCM
SRE (+ HCM)	Skeletal-related events including HCM
TIH	Tumor-induced hypercalcemia (also known as HCM)
WHO	World Health Organization
ZOL	Zometa (ZOMETA)

1           Background

1.1         Clinical consequence of skeletal related events

In patients with advanced cancer, the events defined as Skeletal Related Events (SREs) are the clinical complications of bone metastases.  The time that is necessary to treat these complications and the additional pain and suffering subtract from the time in which a patient can live free from symptoms and independent of additional medical care.

Many SREs are related to pathological fractures.  A pathological fracture can occur in bone that has been eroded by a bone metastasis, to the point that a slight trauma is sufficient to break the bone, cause pain, loss of mobility and the need for possible surgical intervention. If the metastasis has eroded into the structure of a vertebra, the vertebral body may collapse. The resulting tissue mass may bulge inward, into the spinal canal, and compress the spinal cord, leading to paralysis below the level of the fracture, a devastating complication.  Fractures are generally painful and require analgesic medication to treat pain.  Pathological fractures will not heal spontaneously when immobilized, unlike traumatic fractures.  Therefore, most patients will require additional procedures, such as orthopedic surgery, for curettage of the malignant tissue and to insert a metallic rod or plate to immobilize the bone and reduce pain.

A course of radiotherapy may be administered as part of the treatment of a pathological fracture.  Localized radiotherapy may eradicate the tumor tissue in a bone metastasis and is used to prevent pathological fractures, including the collapse of a vertebra, when a significant amount of bone has been destroyed by cancer.  Radiotherapy is also used to treat bone pain due to bone metastases, but frequently entails daily visits to a radiotherapy department for two to three weeks.  The need to undergo radiotherapy is therefore a significant burden on a patient. Radiotherapy itself may lead to osteopenia with the increased risk of new fractures.

Hypercalcemia of malignancy is often a late complication of cancer.  Symptoms include nausea, vomiting, dehydration, weakness and coma, culminating in death.  Emergency treatment at a clinic is required, which again subtracts from the time a patient has free from the need for additional medical care.  Finally, even though a change in antineoplastic therapy to treat bone pain appears to be of lesser significance compared to other SREs, it means to the patient that their cancer is advancing and may have a large psychological impact.

Each of these SREs is a serious debilitating event to anyone with cancer, threatening their independence and requiring additional medical treatment.  Psychologically, an SRE is another reminder to the patient of the fact that she or he must deal with cancer.  The ability to postpone any of these events leads to a longer time for the patient to be free from the complications of bone metastases, which together with the reduction in the rate of subsequent events, results in a lower proportion of patients suffering from these SREs.

1.2         Pathophysiology

[The osteoclast is thought to be the final common pathway in the pathophysiology of bone metastases. Osteoclasts are specialized bone cells which erode mineralized bone by secreting acids and lysosomal enzymes.  The lytic bone destruction associated with malignancy develops because tumor cells synthesize and release soluble factors that stimulate osteoclasts to resorb bone (4,5,6).  The osteoclastic activating factors released by tumor cells include parathyroid hormone-related peptide (PTHrP), growth factors, and cytokines (7-10).  The activation of osteoclasts is revealed by the increase in urinary NTX/creatinine ratio seen in patients with osteolytic bone metastases.  The same assay reveals that the highest levels of osteoclast activation occur in prostate cancer patients who have predominantly osteosclerotic osteoblastic bone metastases.(11).

The In patients with malignant bone lesions, the malignant activation of osteoclasts results in disruption of normal bone remodeling so that the equilibrium between bone resorption and bone formation is shifted towards increased bone resorption.  Thus, the predominant role of the osteoclast in the pathogenesis of bone destruction and the inhibitory effects of bisphosphonates on osteoclast function have formed the rationale for the use of bisphosphonates in the treatment of osteolytic bone metastases.  The common role, regardless of tumor type, of osteoclasts as mediators of bone destruction in metastatic skeletal disease is indicated by the effectiveness of bisphosphonates in the therapy of hypercalcemia of malignancy arising from any type of cancer (12-16).  Recent s

Bisphosphonates are effective inhibitors of bone resorption, and have the potential to delay or reduce the occurrence of SREs in patients with malignant bone lesions.  ArediaÒ (pamidronate), given in addition to standard anticancer therapy, is the only bisphosphonate currently approved in the United States (U.S.) for the treatment of patients with multiple myeloma or bone metastases from breast cancer.  Studies which established the current standard of care have shown that therapy with the bisphosphonate pamidronate (ArediaÒ) combined with anti-neoplastic therapy significantly reduces the proportion of patients having skeletal complications due to the lytic bone disease associated with multiple myeloma and breast cancer compared to anti-neoplastic therapy alone (17,18).  The new Zometa trials reported here have improved on the clinical and statistical methodologies that had been employed in the original phase III program demonstrating the efficacy of Aredia in breast cancer and multiple myeloma.  For example the methodology for  assessment of skeletal related events was more conservative than in the Aredia program, taking into account newer statistical insights which avoid overcounting events.

 Bisphosphonates differ from one another by substitution of side chains on their phosphorous - carbon - phosphorous structural backbone.  Nitrogen-containing bisphosphonates, including Zometa and Aredia, inhibit the biosynthetic mevalonate pathway required for protein prenylation in osteoclasts.  The established safety profile of intravenous bisphosphonates includes transient acute phase reactions (fever, arthralgias, myalgias, lymphopenia), injection site reactions (erythema, swelling, and/or induration), and renal insufficiency (19,20,21).

1.3         Overview of Zometa

Zometa is a third generation bisphosphonate, 2-(imidazol-l-yl-hydroxyethane-1, 1-bisphosphonic acid), characterized by a side chain consisting of an imidazole ring group.  Zometa, a more potent inhibitor of osteoclasts than earlier bisphosphonates, has the largest in vitro therapeutic ratio between the desired inhibition of bone resorption and the unwanted inhibition of normal mineralization of all the bisphosphonates.  It can be infused over a shorter time (15 minutes rather than 2 hours) than Aredia.  Zometa is approved worldwide for treatment of hypercalcemia of malignancy; in the U.S., the NDA for this use of Zometa was approved on August 20, 2001 [see Attachment 3, Proposed Product Labeling for ZometaÒ (including HCM)].

This briefing document presents the results of an extensive program of clinical pharmacology, phase II, and phase III trials, which have examined the use of Zometa as a treatment for patients with myeloma or with bone metastases due to a broad range of cancer types.  These data provide the basis of a supplemental NDA for the use of Zometa in a new indication, treatment of patients with myeloma or with cancer metastatic to bone.

2           Clinical Pharmacokinetics and Metabolism

Single 5-minute or 15-minute infusions of 2, 4, 8 or 16 mg Zometa^® (zoledronic acid for injection) were given to 32 patients with cancer and bone metastases.  The post-infusion decline of zoledronic acid concentrations in plasma was consistent with a triphasic process, showing population half-lives of t_1/2α 0.23 hours and t_1/2β 1.75 hours for the early distribution and elimination of the drug, with a terminal elimination half-life t_1/2γ of 167 hours describing the low concentrations in plasma observed up to 28 days post dose.

In a study in patients with cancer and bone metastases (n=32), 44 ± 18% of the administered Zometa dose was recovered in the urine within 24 hours.  The balance, representing drug presumably bound to bone, is slowly released back into the systemic circulation, giving rise to the observed 167-hour terminal half-life in plasma.  The area under the plasma concentration versus time curve of Zometa was linearly related to dose, and the cumulative percent of drug excreted in the urine 0-24 hours was independent of dose.  The 0-24 hour renal clearance of Zometa in these patients was 4.0 ± 2.3 L/h, and the plasma clearance, representing renal elimination plus uptake by bone, was 5.6 ± 2.5 L/h.

Zometa clearance was independent of dose, and was not affected by body weight, body mass index, or gender. In a study in patients with cancer, increasing the infusion time of a 4 mg dose of Zometa from 5 minutes (n=3) to 15 minutes (n=4) resulted in a 30% decrease in the Zometa concentration at the end of the infusion ([mean ± SD] 393 ± 100 ng/mL vs 267 ± 41 ng/mL) and a 21% increase in the total AUC (412 ± 107 ng/mL vs 496 ± 212 ng/mL).

In a pharmacokinetic study in patients with cancer and bone metastases (n=32) who had normal to moderately impaired renal function [mean baseline creatinine clearance of 81 ± 30 mL/min (4.9 ± 1.8 L/h)], the renal clearance of Zometa was found to closely correlate with creatinine clearance.  On average, Zometa clearance in these patients was 82 ± 35% of the creatinine clearance.  Pharmacokinetic data are not available for Zometa in patients with severe renal impairment.

2.1.        Clinical pharmacology studies

The pharmacokinetics of Zometa were derived from plasma and urine Zometa concentrations determined by specific radioimmunoassay in 64 cancer patients with bone metastases in three studies, J001, CZOL446D 0503, and CZOL446E 0506 (Table 2-1).

Table 2-1.          Human pharmacokinetic studies

Study	Study Objective	Zometa doses	Pharmacokinetic evaluations
J001 (Japan)	PK/PD in patients with bone metastases	2 mg,   5 min.   4 mg,   5 min.   8 mg,   5 min.	Plasma and quantitative urine sampling 0-48 h post dose on day 1 (1st infusion).
CZOL446D 0503 (USA)	PK/PD in patients with bone metastases	4 mg,   5 min.   4 mg, 15 min.   8 mg, 15 min. 16 mg, 15 min.	Plasma and quantitative urine sampling 0-24 h post dose on days 1, 29, and 57 (1st, 2nd, 3rd infusion); blood and spot urine days 8, 15, 29 post dose.
CZOL446E 0506 (USA and Canada)	PK/PD in patients with bone metastases and differing degrees of renal function ADME in patients with normal renal function	4 mg, 15 min.	Plasma and quantitative urine sampling 0-72 h post dose day 1 (1st inf.), and 0-24 h post dose days 29 and 57 (2nd and 3rd inf.); blood and spot urine days 8, 15, 36, 43, 64, 71, and 85.   14C-Zometa plasma, blood, and urine samples were obtained on days 1, 2,3, 8, 15, 29, 36, 43, 57, 64, 71, and 85.

 

The pharmacokinetics of Zometa were consistent across the three studies, demonstrating an overall pattern similar to that of other bisphosphonates.  After initiating the infusion of Zometa, the plasma concentrations of drug rapidly increased, achieving their peak at the end of the infusion period, followed by a rapid, multiphasic decline to <1% of peak after 24 hours post end infusion.  The rapid decline of Zometa plasma concentrations is illustrated in Figure 2-1, which shows the mean concentration data after a single 4 mg dose infused over 15 minutes, in three groups of cancer patients with normal, mild, and moderate to severe renal impairment (study CZOL446 0506).

 

Figure 2-1.        Mean Zometa plasma concentrations over 4 hours (left panel) and 24 hours (right panel)

Cartesian plot showing rapid decline of Zometa early post infusion	Semi-logarithmic plot showing prolonged low concentrations at later timepoints

After the steep, about 100-fold decline in the plasma concentration within the first 24 hours post-drug administration, which is characterized by population half-lives of t_½a = 0.24 hours and t_½b = 1.87 hours, there was a long period of detectable Zometa in plasma characterized by t_½g = 146 hours.  This pattern of early and rapid decline in plasma drug concentration, followed by a prolonged period of very low concentrations results from the injected drug either rapidly binding to osseous tissue or being removed by renal excretion.  The very low Zometa concentrations in plasma at later timepoints presumably represent the small amounts of drug continually released from bone into the systemic circulation during the remodeling process.

The pharmacokinetics of Zometa assessed by AUC_(0-24h) were dose proportional within the dose range of 2 to 16 mg.  The accumulation of Zometa following the repetitive 28-day dosing schedule was low.  There was no evidence of dose or cycle effect on accumulation properties of Zometa over the three administrations of Zometa studied.  Increasing the infusion period from 5 minutes to 15 minutes had no statistically significant effect on the drug exposure (area under the plasma concentration vs. time curve), but as expected lowered the peak Zometa concentration by about 30% (CZOL446D 0503).

Generally, between 10% to 65% of Zometa is excreted in urine within 24 hours post infusion. Based on the average (±s.d.) estimate from all 64 patients evaluated in the clinical pharmacology studies, a total of 39 ± 16% of the administered dose is excreted in the urine within 24 hours.  The remainder is retained in the body, subject to slow release from bone governed by the rate of bone remodeling.

The renal clearance of Zometa is significantly positively correlated and proportional to the creatinine clearance, as would be expected for this renally excreted drug (see Figure 2-2).

Figure 2-2.        Relationship between the renal clearance of Zometa (Clr) and creatinine clearance (CLcr) in 64 patients in 3 clinical pharmacology studies

As

2.2.        Metabolism and drug-drug interactions

Data from a study using ¹⁴C labeled Zometa (4 patients in CZOL446E 0506) showed that Zometa is not metabolized and is eliminated by renal excretion.

No specific clinical drug-drug interaction studies have been conducted with Zometa.  Since Zometa is not metabolized in humans and the drug was found to have little or no capacity as a direct acting and/or irreversible metabolism-dependent inhibitor of P450 enzymes in vitro, Zometa is unlikely to reduce the metabolic clearance of drugs of drugs which are metabolized via the cytochrome P-450 enzyme systems.

Zometa is not highly bound to human plasma proteins (56% bound) and binding is concentration independent.  Therefore, interactions resulting from displacement of highly protein-bound drugs are unlikely.

2.3.        Pharmacokinetics in renal impairment

There are no relevant differences in exposure to Zometa between renally impaired patients and patients with normal renal function (study CZOL446E 0506), evidenced by similar concentration versus time profiles in plasma (see Fig. 2-1), and no changes in the ratio of renal clearance to creatinine clearance.  In patients mild renal impairment, and moderate to severe renal impairment, the AUC_(0-24h) was increased by 26-37%, and 27-41%, respectively. Cmax was increased by 11-15% and 0-17%, respectively, compared to patients with normal renal function.  The increases in AUC and Cmax were the same after the first, or after subsequent doses (every 28 days) of Zometa.  The small increase in systemic exposure, and lack of accumulation of drug with multiple doses irrespective of renal function, suggest that dose adjustments of Zometa in mild (Cl_cr=50-80mL/min) and moderate renal impairment (Cl_cr=30-50mL/min) are not necessary.  As only limited data are available in severe renal impairment (creatinine clearance <30 mL/min), no dosing recommendations are possible for this population.

3           Use in Hypercalcemia of Malignancy

Two identical multicenter, randomized, double-blind, double-dummy studies of Zometa 4 mg or 8 mg given as a 5-minute intravenous infusion or pamidronate 90 mg given as a 2-hour intravenous infusion were conducted in 185 287 patients with hypercalcemia of malignancy (HCM).   This was the largest HCM clinical program ever done.  The mean age of the study population was 59 years; 81% were Caucasian, 15% were Black, and 4% were of other races.  Sixty percent of the patients were male.  The most common tumor types were lung, breast, head and neck, and renal.

In these studies, HCM was defined as a corrected serum calcium (CSC) concentration of ≥ 12.0 mg/dL (3.00 mmol/L).  The primary efficacy variable was the proportion of patients having a complete response, defined as the lowering of the CSC to ≤ 10.8 mg/dL (2.70 mmol/L) within 10 days after drug infusion.

3.1         Efficacy in HCM

To assess the effects of Zometa versus those of pamidronate, theThe two multicenter HCM studies were combined in a pre-planned analysis to assess the effects of Zometa versus those of pamidronate.  .  The results of the primary analysis revealed that tThe proportions of patients that had normalization of corrected serum calcium by Day 10 were 88% and 70% for Zometa 4 mg and pamidronate 90 mg, respectively (P=0.002).  Median time to relapse of HCM was significantly longer with Zometa 4 mg vs pamidronate 90 mg (30 vs 17 days P<0.001).  Thus Zometa is superior to pamidronate in the treatment of HCM, in the normalization of serum calcium and time to relapse.  These endpoints are clinically meaningful in that more patients will recover quickly and early relapse is less likely.  In these studies, no additional efficacy benefit was seen for Zometa 8 mg over Zometa 4 mg.

3.2         Safety in HCM

Adverse reactions to Zometa^® (Zoledronic acid for injection) in these trials were usually mild and transient and similar to those reported for other bisphosphonates.  Intravenous administration has been most commonly associated with fever.  Occasionally, patients experience a flu-like syndrome consisting of fever, chills, bone pain and/or arthralgias, and myalgias.  Gastrointestinal reactions such as nausea and vomiting have been reported following intravenous infusion of Zometa.  Local reactions at the infusion site, such as redness or swelling, were observed infrequently.  In most cases, no specific treatment was required and the symptoms subsided after 24-48 hours.

Grade 3 or 4 creatinine values were reported in 2 (2.3%), 5 (5.2%), and 4 (4.0%) of patients in the Zometa 4 mg, Zometa 8 mg, and Aredia 90 mg groups, respectively.

4           Phase II Trial in treatment of bone metastases

One A phase II dose ranging trial was performed.,  This study was a randomized, double-blind, double dummy, parallel group Phase 2 study trial in 280 patients aged at least 18 years with metastatic breast cancer or multiple myeloma with at least one osteolytic lesion (confirmed on plain film X-ray).  For patients with multiple myeloma, the study eligibility criteria specified that patients must have had a previous skeletal event or must have experienced failure of first-line chemotherapy.  All patients had to have ECOG performance scores of 0, 1 or 2.  Patients received either 0.4 mg, 2 mg, or 4 mg Zometa administered as 5-minute infusions, or Aredia 90 mg given as a 2-hour i.v. infusion every 3-4 weeks for 9 months.A double-dummy design was used to maintain blinding.

4.1         Efficacy

The primary efficacy endpoint was the proportion of patients receiving radiation to bone.  This endpoint was selected based on data from the pamidronate breast cancer and multiple myeloma studies.  Radiation therapy to bone was consistently affected in the pamidronate treatment arms in these studies.  This consistent efficacy effects was seen throughout the course of these trials.  The secondary parameters used in this trial were similar to the primary and secondary endpoints (proportion of patients with SREs, time to first SRE and rate of SREs/time).Unlike the later phase III studies, this trial used the proportion of patients receiving radiation to bone as the primary efficacy variable, but some secondary parameters were the same as secondary parameters used in the pivotal studies.  For the primary endpoint each treatment, except Zometa 0.4 mg, met the statistical criterion of effectiveness defined in the study objectives, i.e. that the proportion of patients having radiation to bone would be statistically less than 30% (p=0.104, 0.012, 0.033, and 0.003 for Zometa 0.4 mg, 2.0 mg, 4.0 mg and Aredia 90 mg, respectively   Results for the primary endpoint and for all SREs, with and without hypercalcemia, are shown below.This study is part of an overall Zometa clinical trial program culminating in larger Phase III trials in cancer patients with metastatic bone disease.  Of the three Zometa doses evaluated here, 0.4 mg was clearly ineffective compared with Zometa at doses of 2.0 mg to 4.0 mg and pamidronate at 90 mg.  Although results for the primary efficacy criterion were similar for the 2.0- and 4.0-mg doses of Zometa, several trends in the secondary efficacy parameters favor the 4.0-mg Zometa dose.  Skeletal events as a whole and pathologic fractures occurred slightly less frequently in patients treated with 4.0 mg than 2.0 mg Zometa.  The 4.0-mg Zometa group was the only one in which no patient developed hypercalcemia during the 10 months of the study.  Also, time to the first skeletal event was almost 2 months longer for patients in the 2.0-mg group.  Although this trial was not designed to detect statistically significant differences between treatment groups, these data suggest that, assuming similar safety profiles, the 4.0-mg dose was the better choice for confirmatory Phase III trials.

Table 1. Proportion of patients with SRE in Phase II bone metastases study

	n/N (%) patients
Type of SRE	Zometa 0.4 mg	Zometa 2.0 mg	Zometa 4.0 mg	Aredia 90 mg
Radiation therapy to bone	16/68 (24%)	14/72 (19%)	14/67 (21%)	13/73 (18%)
All excluding HCM (-HCM)	30/68 (44%)	23/72 (32%)	22/67 (33%)	22/73 (30%)
All including HCM (+HCM)	31/68 (46%)*	25/72 (35%)	22/67 (33%)	22/73 (30%)

4.2         Safety

The nature and frequency of commonly reported adverse events (³15% of patients) were similar for the Zometa and Aredia treatment groups.  The incidence of renal adverse events was low and was similar for the Zometa and Aredia treatment groups.  The percentage of patients experiencing injection site reactions with the 5-minute Zometa infusion was slightly lower than that with that 2-hour Aredia infusion.  The overall reporting of eye-related adverse events was low (12.5%) and was similar in all the treatment groups.  Conjunctivitis occurred more frequently in the Aredia and Zometa 0.4 mg groups than in the Zometa 2.0 and 4.0 mg groups.  The incidence of hypocalcemia and other metabolic or electrolyte abnormalities was low and similar in the treatment groups.

5           Phase III Trials

5.1        Study Design

Three pivotal phase III Zometa trials (2 placebo- and 1 pamidronate-controlled) were performed (Table 5-1).  The primary analysis for these trials was the proportion of patients having at least one skeletal-related event (SRE).  This trial design has been shown to be effective in the phase III Aredia development program.  Secondary analyses of the primary SRE endpoint included time to first SRE, skeletal morbidity rate (SMR, the number of events divided by the time on trial), and SRE multiple event analysis of time to each SRE (Anderson-Gill).  Both the proportion of patients having at least one SRE and the time to first SRE are conservative endpoints which only take into account the information of the first event.  The SMR and Anderson-Gill analyses also utilize information on subsequrentsubsequent events, in cases where patients had more than one event over the course of their study participation, and thus help to evaluate the longer-term effects of treatment with the study drug.

Per discussions with the FDA, the primary analyses specified in the study protocols excluded hypercalcemia of malignancy as an SRE.

Table 5-1.          Overview of Phase III studies

Study No.	No. of Patients	Treatment Duration	Zometa dose	Control	Patient population
010	1648	13 months	4 and 8* mg Q3 - 4weeks	Aredia 90 mg Q3 - 4 weeks	Multiple myeloma or metastatic breast cancer
039	643	15 months	4 and 8* mg Q3 weeks	Placebo	Metastatic prostate cancer
011	773	9 months	4 and 8* mg Q3 weeks	Placebo	Metastatic solid tumor other than breast or prostate cancer

*All patients on 8 mg Zometa were switched to 4 mg following a protocol amendment

The different study durations were based both on the skeletal related event analysis of previous pamidronate trials and the estimated survival of the cancer population being investigated.

Study Histories

In the original pivotal study protocols, patients were to be randomized to either of two Zometa doses, 4 mg or 8 mg, administered as 5-minute i.v. infusions.  Following a protocol amendment issued in June 1999 by Novartis, the infusion time was increased to 15 minutes, and the volume increased from 50 mL to 100 mL.  Following protocol amendments issued in June 2000, all patients on 8 mg were switched to 4 mg, and monitoring of serum creatinine was initiated.  These latter amendments were made on the suggestion of a Data Safety Monitoring Board (DSMB) and Renal Advisory Board (RAB) appointed to monitor overall and renal safety.  Throughout this document, any treatment group that was originally assigned to treatment with 8 mg Zometa but later was switched to 4 mg following the protocol amendments is referred to as an 8/4 mg group.

Due to the higher incidence of renal adverse events observed in the Zometa 8 mg group at the time of the amendment (reduce 8 mg to 4 mg) and the inhomogeneity of the treatment duration in the Zometa 8/4 mg group, Novartis made the decision that the Zometa 8/4 mg will not be part of the application for the indication as documented in the amendment.  However, analysis results of the Zometa 8/4 mg will be included in the presentation for the completeness of the study results.

Entry Criteria

Entry criteria were generally similar across the 3 trials, except for tumor-specific considerations.

Aredia- controlled trial

Study 010 was a randomized, double-blind, double-dummy, multicenter, parallel-group Phase III study in 1648 patients.  Patients had multiple myeloma with at least one osteolytic bone lesion or breast cancer with at least one bone metastasis, and were receiving anti-cancer therapy.  Breast cancer patients receiving hormonal therapy had to be using first- or second-line hormonal therapy.  Patients had an ECOG performance status of 0, 1, or 2, and no significant hepatic, renal or cardiac impairment, hypercalcemia, brain metastases, or lymphangitic lung metastases (breast cancer patients).

Placebo-controlled trials

Study 039 was a randomized, double-blind, multicenter, parallel-group, placebo-controlled, Phase III study in 643 patients with bone metastases due to prostate cancer.  Patients had biochemical evidence (rising serum PSA concentration) of disease progression despite hormonal therapy, and an ECOG performance status of 0, 1 or 2.  Patients who had received, or were receiving, cytotoxic chemotherapy, or who had had radiation therapy within 3 months prior to entry were excluded, as were those who had bone pain requiring strong narcotics.  Patients with hypercalcemia, significant renal or cardiac impairment, or a history of other cancers within the previous 5 years were also excluded.

Study 011 was a randomized, double-blind, multicenter, parallel-group, placebo-controlled, Phase III study in 773 patients with bone metastases from solid tumors other than breast or prostate cancers.  Patients had to have ECOG scores of 0, 1 or 2; those with ECOG scores of 2 had to have bone metastases diagnosed within 6 weeks of study entry.  Patients were excluded for significant hepatic, renal or cardiac impairment, hypercalcemia, or symptomatic brain metastases.

Randomization Procedures: treatment assignments and blinding

Randomization procedures were the same for all three double-blind trials.  The pharmacist, who was the only unblinded person during the studies, was provided with allocation cards linking the randomization number with the treatment group and was responsible for maintaining the blind at each center.  Emergency drug codes were also supplied to the investigator and kept on file at Novartis in Clinical Safety and EpidemologyEpidemiology (CS&E).  The randomization scheme was performed by Novartis Drug Supply management using a validated system that automates the random assignment of treatment groups to randomization numbers.  The randomization scheme was then reviewed by Quality Management Biostatistics and locked by them after approval.

In the placebo-controlled trials, patients were randomized to treatment with Zometa 4 mg, Zometa 8 mg, or placebo every 3 weeks for 15 months in the prostate cancer study and 9 months in the study of other solid tumors.  Following a protocol amendment (dated June 24 and 25, 1999 for 039 and 011, respectively), the infusion time was increased from 5 to 15 minutes and infusate volume increased from 50 to 100 mL.  In another protocol amendment (dated June 7, 2000), all patients on 8 mg Zometa were switched to 4 mg and creatinine monitoring was instituted.

In the prostate cancer study, patients were stratified into two groups: patients with no metastatic disease present at the time of the initial diagnosis of prostate cancer (TMN stage MX or MO) and patients with metastatic disease present at initial diagnosis (stage D2 disease or TMN Stage M1).  Per protocol, all patients had bone metastases at study entry.  In the study of other solid tumors, patients were also stratified into two groups: patients with lung cancer and patients with all other cancers.

In the study in patients with breast cancer or multiple myeloma, patients were randomized to Zometa 4 mg, Zometa 8 mg or Aredia 90 mg every 3-4 weeks for 12 months.  Following a protocol amendment dated June 25, 1999, the infusion time for Zometa or matching double-dummy placebo was increased from 5 to 15 minutes and infusate volume was increased from 50 to 100 mL.  To maintain blinding, Zometa was infused in a total volume of 100 mL, followed by 250 mL of normal saline given over 2 hours.  Aredia was administered as a 2-hour i.v. infusion in a total volume of 250 mL, preceded by a 100 mL 15-minute normal saline infusion.  Following a protocol amendment dated June 7, 2000, all patients on 8 mg Zometa were switched to 4 mg and creatinine monitoring was instituted.  Patients were stratified into three groups: Stage III multiple myeloma patients, Stage IV breast cancer patients receiving chemotherapy with or without combination hormonal therapy, and Stage IV breast cancer patients receiving first-line or second-line hormonal therapy for metastatic disease at the time of randomization.

Treatment Administration: Study Drug

Zometa was packaged in open-label fashion and shipped to the pharmacist at each center.  Medication labels complied with the legal requirements of each country and printed in the local language.  Zometa was supplied in 4mg lyophilized vials.  The pharmacist was responsible for the preparation of study drug.  Zometa 4 or 8 mg was given in 100 mg mL of normal saline and administered by intravenous infusion.

Efficacy Endpoints and Statistical Analysis

All efficacy analyses reported here were performed on intent-to-treat (ITT) populations.  The three pivotal studies utilized a common primary efficacy endpoint, the proportion of patients with at least one skeletal-related event (SRE), which had been demonstrated to be a relevant parameter in the phase III Aredia trials.  These studies also shared most secondary efficacy parameters.

Table 5-2.          Efficacy parameters used in Phase III studies

Primary	Major Secondary
Proportion of patients having at least one SRE (exclusive of HCM)	Skeletal: Proportion of patients having at least one SRE including HCM, time to first SRE, SMR, proportion of patients with each SRE, bone biochemical markers1, bone mineral density2 Quality of Life: Changes from baseline in pain score (BPI), analgesic score, ECOG performance score, quality of life (FACT-G and EURO QOL-5D3) Cancer:  Time to progression of bone lesions, time to progression of disease, best response of bone lesions
SRE: skeletal-related events; SMR: skeletal morbidity rate; BPI: basic pain inventory composite score 1 In breast cancer/multiple myeloma study, some biochemical markers were only evaluated in selected US/Canadian centers 2 In prostate cancer study and selected US/Canadian centers in breast cancer/multiple myeloma study only 3 In prostate cancer study only

 

Skeletal related events (SREs) were defined as follows:

Radiation therapy to bone including irradiation of bone to palliate painful lesions, to treat or prevent a pathologic fracture, to treat or prevent a spinal cord compression, and the use of i.v. strontium-89 (or other radioisotopes) to treat metastatic bone pain.

Surgery to bone, including surgical procedures performed to set or stabilize pathologic fractures or areas of spinal cord compression, or to prevent an imminent pathologic bone fracture or spinal cord compression.

Pathologic bone fractures were defined as bone fractures (vertebral and non-vertebral) which occurred spontaneously or resulted from trivial trauma and documented via radiographs.  Each pathologic fracture was to be documented by a plain X-ray film.  A new vertebral compression fracture was defined as a decrease in total, anterior or posterior vertebral height of ³ 25% from baseline.  Vertebral compression fractures and non-vertebral fractures were evaluated by a central radiologist who was not aware of the treatment regimens.

Spinal cord compression occurs due to impingement of tumor on the spinal cord, and is associated with neurologic impairment or back pain.  If spinal cord compression was associated with one or more vertebral compression fractures, each fracture was recorded as an SRE, in addition to the spinal cord compression itself.

Change of antineoplastic therapy to treat bone pain included any change of anticancer agents (to a different hormonal regimen or to a cytotoxic chemotherapy regimen) to palliate bone pain.  This was only included as an SRE in the study in prostate cancer patients where changes in antineoplastic therapy maybe more commonly used to achieve bone pain relief..  An alteration of analgesic medication for bone pain was captured as an analgesic score and was not recorded as an SRE in any study.

Hypercalcemia of malignancy (HCM) also termed tumor-induced hypercalcemia (TIH), was included as an SRE for some secondary efficacy analyses.  HCM/TIH was defined as a corrected serum calcium level of ³ 3.00 mmol/L (12.0 mg/dL) or a lower level of hypercalcemia which was symptomatic and required active treatment other than rehydration.

 

Other skeletal assessments (data not shown):

Biochemical markers of bone resorption and formation were measured by a central laboratory (Mayo Medical Laboratories) in the pivotal studies (in study 010, US/Canadian centers only):

·        Urinary N-telopeptide/creatinine ratio

·        Urinary pyridinoline/creatinine ratio

·        Urinary deoxypyridinoline/creatinine ratio

·        Serum bone alkaline phosphatase

·        Serum PTH

 

Bone mineral density (BMD) was determined in studies 010 (breast cancer and multiple myeloma patients) and 039 (prostate cancer patients) by dual-energy X-ray absorptiometry (DEXA).  Scanning sites were the AP lumbar spine (L2-L4), the proximal femur (neck, trochanteric region and Ward’s triangle) and the non-dominant forearm.  Total body bone mineral density was also calculated.

Analysis of primary endpoint and other skeletal endpoints

A placebo group was not utilized in the breast/multiple myeloma study, since Aredia has previously been shown to significantly delay or reduce the incidence of skeletal-related morbid events in these patients.  If Zometa and Aredia were equal in effectiveness, statistical power calculations indicated that the planned study enrollment (1600 patients) would have had power to exclude an 8% or greater inferiority for Zometa in the proportion of patients experiencing an SRE (the primary study endpoint), compared with anticipated beneficial treatment effect of 13.5% for Aredia (based on historical data).  The study results revealed a small difference in the proportion of patients experiencing an SRE, favoring Zometa 4 mg, and the lower confidence bound for this comparison with Aredia was 3.9% (indicating that Zometa was at least within 3.9% of the effectiveness seen with Aredia as measured by this statistic).  In the non-inferiority study versus Aredia 90 mg (breast cancer, multiple myeloma), the 95% confidence interval for the difference between treatment groups was provided, and non-inferiority of Zometa with respect to Aredia 90 mg was concluded if the upper bound of the 95% confidence interval was below 8%.  For the placebo-controlled studies (prostate cancer; solid tumors other than breast/prostate), the proportion of patients with at least one SRE was compared between groups using a Cochran-Mantel-Haenszel (CMH) test.

The proportion of patients having any SRE (including HCM) and the proportion of patients having each individual type of SRE were analyzed as described above for the primary variable.

Time to first SRE was compared between treatment groups using survival analysis, including Kaplan-Meier product-limit estimates of the ‘survival functions’, the log-rank test, and Cox regression stratified by the stratum.  The Anderson-Gill approach was used to analyze the time from randomization to each occurrence of the events (multiple event analysis).  The information of each patient from the study entry to the end of the follow-up was fully utilized in the Anderson-Gill approach to estimate the rate of recurrence over the course of the study and the test was aimed in detecting whether one group of study subjects had a higher recurrence rate than the other group.  The difference between the time to the first event analysis and the Anderson-Gill approach is that the former ignores the information after the occurrence of the first event (the number of events and the time when each of these recurrent events occurs) and the latter utilizes the information ignored by the former.

The SMR was defined as the number of SREs divided by the time at risk in years, counted from the randomization date with every counted event followed by a 20-day period during which no SRE experienced by the patient would be counted, nor the time be counted as at risk.  This was a more conservative approach than was used in the phase III Aredia program, where all events were counted. If a patient had no SRE, the whole study period was counted as at risk.  The SMRs were compared between treatment groups using a CMH test.

The same definition in counting the number of events and the time at risk as defined in SMR was used in the Anderson-Gill approach.  In the case where patients did not experience any SRE, the SMR analysis made no distinction for the length of follow up of each patient (i.e. SMR had a value of 0) while the Anderson-Gill approach takes this into account of the length of the follow-up.

Analysis of Quality of Life parameters

Pain was measured by the Brief Pain Inventory (BPI) composite score, which was derived from items 3 to 6 from the BPI.  Changes from baseline in BPI composite score were compared between treatment groups using analysis of covariance (ANCOVA) with baseline value as a covariate and stratum and treatment group as factors at 3, 6, 9, etc. months.

Analgesic score captured the use of pain medication in a five-point ordinal scale ranging from 0 (no analgesics used) to 4 (strong narcotics, e.g. morphine or hydromorphone, used).  Change from baseline in analgesic score was compared between treatment groups using the stratified Cochran-Mantel-Haenszel test with modified ridit scores at 3, 6, 9 etc., months.  Within-treatment differences from baseline were analyzed using the Wilcoxon signed-rank test.

ECOG performance scores measured performance on a five-point ordinal scale from 0 (fully active and able to carry out all pre-disease performance without restriction to 4 (completely disabled, cannot carry on any self care and totally confined to bed or chair).Changes from baseline in performance status were compared between treatment groups using the stratified Cochran-Mantel-Haenszel test with modified ridit scores at 3, 6, 9, etc., months.  Within-treatment differences from baseline were analyzed using the Wilcoxon signed-rank test.

Two quality of life instruments were used: FACT-G (in all three studies) and EURO QOL EQ-5D (in the prostate cancer study only).  FACT-G (version 4) uses a questionnaire with four subscales: physical well-being, social/family well-being, emotional well-being and functional well-being.  The sum of these provides the total score.  EURO QOL EQ-5D uses a questionnaire with 6 dimensions: mobility, self-care, usual activities, pain/discomfort, anxiety/depression, and overall health state.  Changes from baseline in FACT-G total scores and the four subscales were analyzed using ANCOVA with baseline value as covariate and disease as a factor.  Changes in EURO QOL scores were analyzed as for FACT-G.

Analysis of Other Disease Parameters

Progression of bone lesions was defined in the protocols as the appearance of new osteolytic or osteoblastic lesions, and/or a significant increase in the size of one or more existing osteolytic or osteoblastic lesions.  For ‘measurable’ lesions (those with greatest diameter >2 cm), progression was defined as an increase of >25% in the lytic or blastic area as measured by the product of the greatest diameter and the diameter perpendicular to the greatest diameter.  For ‘evaluable’ lesions (defined as those with <2 cm greatest diameter), progression was defined as a linear increase of 1 cm or more in the unidimensional measurement.  Time to progression of disease was compared between the treatment groups using survival analysis methods, including Kaplan-Meier product-limit estimates of the ‘survival functions’, the log-rank test, and Cox regression stratified by the stratum.

For each tumor type, criteria for progression of disease were defined in the study protocolsTime to progression of disease was analyzed using the same method used for time to progression of bone lesions.

Objective response of bone lesions, on the basis of radiologic evidence, was determined as follows:

·        Complete response, i.e. resolution of all osteoblastic lesions and complete recalcification of all osteolytic lesions.

·        Partial response, i.e. resolution of some but not all osteoblastic lesions or decreased size of one or more osteoblastic lesions (for ‘measurable’ lesions a decrease of ³ 50% in blastic area, for ‘evaluable’ lesions a decrease of ³ 30% of the unidimensional measurement) or at least partial recalcification of one or more osteolytic lesions and no appearance of new osteolytic or osteoblastic lesions or progression of existing osteolytic or osteoblastic lesions.

·        No change, i.e. no change in number or size of osteoblastic lesion, no evidence of recalcification of, or change in number or size of, osteolytic lesions.

·        Progressive disease, defined as above in the description of time to progression of bone lesions.

·        Not evaluable.

The best response of objective bone lesion response was summarized by treatment group.

 

All skeletal X-rays and bone scans from the patients in the three pivotal trials were read by a central radiologist who was blinded to the patient’s treatment assignment.

Sample Size and Power Considerations

Aredia-controlled trial

A placebo group was not utilized in this study, since Aredia has previously been shown to significantly delay or reduce the incidence of skeletal-related morbid events in these patients.  If Zometa and Aredia were equal in effectiveness, statistical power calculations indicated that the planned study enrollment (1600 patients) would have had power to exclude an 8% or greater inferiority for Zometa in the proportion of patients experiencing an SRE (the primary study endpoint), compared with anticipated beneficial treatment effect of 13.5% for Aredia (based on historical data).  The study results revealed a small difference in the proportion of patients experiencing an SRE, favoring Zometa 4 mg, and the lower confidence bound for this comparison with Aredia was 3.7% (indicating that Zometa was at least within 3.7% of the effectiveness seen with Aredia as measured by this statistic).

Historical data was used to estimate SRE rates.  Preceding studies comparing pamidronate to placebo (in multiple myeloma, breast cancer receiving chemotherapy or hormonal therapy), showed an average pamidronate - placebo difference of 13.4% after 9-12 months of treatment.

To demonstrate non-inferiority of either zoledronic acid dose to pamidronate 90-mg in the percentages of patients with SREs, the non-inferiority margin was set at 8%.  In order to maximize power at 80%, the percentage of patients with SREs was assumed to be 50% for each treatment group (maximizing the variance).  Using a two-group large-sample normal approximation test at a one-sided significance level of 0.05, the sample size was calculated (484 patient per arm) using the nQuery Advisor® (for equivalence of two proportions).  Based on previous dropout rates the decision was made to randomize approximately 1470 patients (490 per arm).  This sample size was increased to 1509 following discussions with the health authorities.

Placebo controlled trials

The studies in lung cancer or other solid tumors and in prostate cancer were designed to have 80% power which allowed to detect a 14% and a 16% difference in the proportion of patients reporting any “skeletal-related episode” (during the first nine and the first fifteen months of treatment) between the two Zometa dose levels (4-mg and 8-mg) and placebo, respectively.  The sample size for both protocols were calculated using the Bonferroni’s adjustment formula.

For the prostate cancer study, based on the Bonferroni’s adjustment, the sample size calculation assumed a 40% incidence rate on placebo treatment; a 24% incidence rate on either dose level of Zometa, with an overall Type I error rate of 0.05 (two-sided).  The total sample size was determined to be 519 patients (173 on each arm).  It was recommended that 550 patients be enrolled to allow for 5% noise due to the inclusion of the intent-to-treat patient population.

For the lung cancer and other solid tumors study, the sample size was calculated, assuming a 38% incidence rate on placebo; a 24% incidence rate on either dose level of Zometa, with an overall Type I error rate of 0.05 (two-sided).  The total sample size was determined to be 663 patients (221 on each arm).  Therefore, it was recommended that 700 patients be enrolled in order to allow for the 5% noise from the inclusion of the intent-to-treat patient population.

To determine the final patient accrual number, patient drop-out and SRE rates were monitored on an ongoing basis using blinded data.  In the study in lung cancer and other solid tumors, there was a higher than expected drop-out rate (40%) and a lower than expected SRE rate (<30%) in the first 400 randomized patients.  Therefore, there was a need to replace the drop-out patients increase the sample size to achieve the established 80% power.  Amendment 4, effective 16-Feb-2000, increased the sample size from 600 to 700 patients.  There was no need to increase the sample size in the prostate cancer study.

 

5.2     Breast cancer and multiple myeloma – Aredia-controlled trial (010)

5.2.1   Patient disposition

A total of 1648 patients were randomized between 16-October-1998 and 18-December-1999 at 248 centers in 20 countries including centers in North and South America, Europe, South Africa, the Middle East, and Australia and New Zealand.  The intent-to-treat (ITT) population excluded eight patients from one center because of GCP/ICH violations.

Table 5-3.          Patient disposition by treatment group, Aredia-controlled trial

Number of patients	Zometa 4 mg N = 564	Zometa 8/4 mg N = 526	Aredia 90 mg N = 558
Randomized Included in efficacy analysis (ITT)	564 561	526 524	558 555
Multiple myeloma	183 (32.4%)	160 (30.4%)	167 (29.9%)
Breast cancer with chemotherapy	178 (31.6%)	172 (32.7%)	181 (32.4%)
Breast cancer with hormonal therapy	200 (35.5%)	192 (36.5%)	207 (37.1%)
Completed 12 months	353 (62.9%)	313 (59.7%)	337 (60.7%)
Discontinued prematurely from ITT n (%)	208 (37.1%)	211 (40.3%)	218 (39.3%)
Adverse events	57 (10.2%)	70 (13.4%)	51 (9.2%)
Abnormal laboratory values	6 (1.1%)	3 (0.6%)	4 (0.7%)
Abnormal test procedure results	0 (0.0%)	2 (0.4%)	2 (0.4%)
Unsatisfactory therapeutic effect	18 (3.2%)	18 (3.4%)	22 (4.0%)
Condition no longer requires study drug	6 (1.1%)	7 (1.3%)	8 (1.4%)
Protocol violation	6 (1.1%)	4 (0.8%)	4 (0.7%)
Withdrew consent	47 (8.4%)	45 (8.6%)	56 (10.1%)
Lost to follow-up	3 (0.5%)	4 (0.8%)	3 (0.5%)
Administrative problems	5 (0.9%)	2 (0.4%)	4 (0.7%)
Death	60 (10.7%)	56 (10.7%)	64 (11.5%)

5.2.2        Demographics and disease characteristics

The majority of patients in all treatment groups in the breast cancer/multiple myeloma study were female and Caucasian, with a mean age of 59 to 60 years (see Table 5-4 below).

Table 5-4.          Baseline demographic characteristics – Aredia-controlled trial – ITT population

	Zometa 4 mg N=561	Zometa 8/4 mg N=524	Aredia 90 mg N=555
Age (years)
Mean ± SD	59.7 ± 12.00	58.9 ± 12.32	58.7 ± 12.66
Sex – n (%)
Male	104 (18.5)	96 (18.3)	92 (16.6)
Female	457 (81.5)	428 (81.7)	463 (83.4)
Race – n (%)
Caucasian	493 (87.9)	443 (84.5)	484 (87.2)
Black	34 (6.1)	42 (8.0)	43 (7.7)
Other	34 (6.1)	39 (7.4)	28 (5.0)
Weight (kg)
N	536	504	538
Mean ± SD	72.7 ± 16.43	72.9 ± 16.20	73.5 ± 16.43
Median	70.2	70.7	72.0
ECOG performance status - n (%)
ECOG 0-1	474 (84.5)	428 (81.7)	436 (78.6)
ECOG 2	86 (15.3)	95 (18.1)	116 (20.9)
Missing	1 (0.2)	1 (0.2)	3 (0.5)

 

Other baseline characteristics are presented by tumor type (multiple myeloma or breast cancer patients) in the tables below.

Table 5-5.          Baseline disease characteristics of multiple myeloma patients

	Zometa 4 mg	Zometa 8/4 mg	Aredia 90 mg
Disease characteristic	N=186	N=160	N=167
Previous SRE	150 (80.6%)	130 (81.3%)	136 (81.4%)
Yes	150 (80.6%)	130 (81.3%)	136 (81.4%)
Time from initial cancer diagnosis to Visit 2 (months, mean ± SD)* Visit 2 (months, median ± SD)*	18.3 ± 32.28	13.6 ± 22.30	17.3 ± 28.54
Baseline serum creatinine
Normal (<1.4 mg/dL)	147 (79.0%)	127 (79.4%)	145 (86.8%)
Abnormal (³1.4 mg/dL)	36 (19.4%)	32 (20.0%)	22 (13.2%)
Missing	3 (1.6%)	1 (0.6%)	0 (0.0%)

* 28 days in a month.

Table 5-6.          Baseline disease characteristics of breast cancer patients

	Zometa 4 mg	Zometa 8/4 mg	Aredia 90 mg
	N=377	N=364	N=389
First-line anti-neoplastic therapy	161 (42.7%)	180 (49.5%)	182 (46.8%)
Previous SRE	232 (61.5%)	207 (56.9%)	244 (62.7%)
Site of metastases:
Bone	377 (100%)	364 (100%)	389 (100%)
Liver	82 (21.8%)	69 (19.0%)	97 (24.9%)
Lung	69 (18.3%)	81 (22.3%)	80 (20.6%)
Brain	6 (1.6%)	5 (1.4%)	9 (2.3%)
`    Other	82 (21.8%)	76 (20.9%)	97 (24.9%)
Time from initial cancer diagnosis to Visit 2 (months, mean ± SD)*	78.6 ± 67.19	79.1± 74.89	71.9 ± 63.69
Time from initial cancer diagnosis to 1st metastasis (months, mean ±SD)**	57.0 ± 57.40	60.4± 65.83	54.4 ± 57.73
Time from 1st bone metastasis to Visit 2 (months, mean ±SD)*	17.5 ± 33.85	14.1± 22.87	12.6 ± 21.68
Baseline serum creatinine
Normal (<1.4 mg/dL)	364 (96.6%)	348 (95.6%)	369 (94.9%)
Abnormal (³1.4 mg/dL)	11 (2.9%)	11 (3.0%)	15(3.9%)
Missing	2 (0.5%)	5 (1.4%)	5 (1.3%)

* 28 days in a month
** Time from initial diagnosis of cancer to bone metastases or 1st metastatic disease is assigned to 0 when metastatic disease occurred before initial cancer diagnosis.

 

5.2.3       Primary efficacy analysis

The primary efficacy analysis in this study was aimed to demonstrate non-inferiority of the selected Zometa doses to 90 mg Aredia in the proportion of patients with SREs (where hypercalcemia would not be considered an SRE).  The non-inferiority margin was prospectively defined as 8%.   In order to show non-inferiority, the upper limit of the 95% confidence interval for the difference in the two groups would have to be less than 8%, indicating that the Zometa dose being compared was no more than 8% worse than pamidronate.  This criterion was met by both the Zometa groups (Table 5-7 below).  This table shows that the lower bound was 3.7% for the 4 mg group, and 5.8% for the 8/4 group, indicating that the predefined non-inferiority criteria were substantially exceeded. In the analysis of proportion of patients with SREs including hypercalcemia, a similar result was seen.

Table 5-7.          Proportion of patients with SREs, multiple myeloma and breast cancer

		95% C.I. and P-value* for the difference
	Proportion	Zometa 4 mg	Zometa 8/4 mg
Excluding TIH
Aredia 90 mg	257/555 (46%)	(-7.9%, 3.7%)p=0.461	(-6.1%, 5.8%) p=0.963
Zometa 4 mg	248/561 (44%)	-	(-3.9%, 7.9%) p=0.495
Zometa 8/4 mg	242/524 (46%)	-	-

5.2.4        Secondary skeletal endpoints analyses

Proportion of patients with each type of SRE

The proportion of patients receiving radiation therapy to bone was significantly lower for patients in the Zometa 4 mg group than for patients in the Aredia 90 mg group (p=0.031).

The proportion of patients with a pathological fracture, vertebral or non-vertebral, was comparable for all treatment groups.  Spinal cord compression, surgery to bone, and hypercalcemia (not shown) occurred for a small and comparable percentage of patients in each treatment group.

Figure 5-1         Individual SREs in breast/myeloma patients

Time to first skeletal-related event

The median time to the first occurrence of a SRE and the event rate at day 364 was similar for patients in all treatment groups (see Table 5-8 below).  The upper limit of the 95% CI for the hazard ratio for the 4 mg group is 1.090, with the hazard ratio of 0.915, indicates that Zometa 4 mg is likely to influence the time to first SRE at least as strongly as Aredia.

 

Table 5-8A.          Time to the first SRE up to Month 13 (Median and hazard ratio), multiple myeloma and breast cancer

			Zometa 4 mg v.s.		Zometa 8/4 mg v.s.
	N	NMedian (days)	Hazard ratio*	95% CI for Hazard ratio	Hazard ratio*	95% CI for Hazard ratio
Aredia 90 mg	555	363	0.915	(0.786, 1.090)	0.994	(0.834, 1.185)
Zometa 4 mg	561	373561	-	-	1.087	(0.910, 1.299)
Zometa 8/4 mg	524	353524	-	-	-	-

* Hazard ratio is the ratio of column versus row.  The hazard ratio and the 95% C.I. are from Cox-regression for the pairwise comparison of the time to the first event.

Table 9B.  Time to the first SRE up to Month 13, multiple myeloma and breast cancer

					P-values for the between treatment comparison
	N	Event rate at day 364	25% quartile (days)	Median (days)	Zometa 4 mg	Zometa 8/4 mg
Aredia 90 mg	555	50.9%	97.0	363.0	0.322	0.949
Zometa 4 mg	561	48.0%	103.0	373.0	-	0.355
Zometa 8/4 mg	524	51.5%	99.0	353.0	-	-

P-values are from Cox-regression with factor treatment stratified by the stratum.
Event rate, median and 25% quartile are derived from KM estimate.

Skeletal morbidity rate

The mean SMR of all SREs up to Month 13 was lower for patients in the Zometa 4 mg and Zometa 8/4 mg groups than in the Aredia 90 mg group, but the differences were not statistically significant.

Figure 5.2          Mean skeletal morbidity rate by treatment group 

Multiple event analysis of all SREs

The analysis results of multiple events analysis of all SRE (-HCM), using Anderson and Gill’s approach, showed that there was a trend toward statistical significance (p=0.155), favoring Zometa 4 mg over Aredia 90 mg

The difference seen between the SMR analysis and the multiple event analysis was partially due to patients experiencing no events during the study.  In the case of the SMR analysis, all patients contributed equal information regardless of a patient’s duration in the study, whereas the multiple event analysis took into account the length of stay in the study.

Table 5-9.          Multiple-event analysis of time to SRE up to Month 13, multiple myeloma and breast cancer (010)

	Zometa 4 mg v.s.				Zometa 8/4 mg v.s.
	Hazard ratio*	95% CI for Hazard ratio	P-value		Hazard ratio*	95% CI for Hazard ratio	P-value
Aredia 90 mg	0.885	(0.748, 1.047)	0.155		0.910	(0.768, 1.079)	0.277
Zometa 4 mg					1.025	(0.863, 1.218)	0.775

* Hazard ratio is the ratio of column versus row.  Hazard ratios and p-values are from Anderson-Gill approach for time to multiple events

P-value and the 95% C.I. are computed using the robust variance estimate

5.2.5        Analysis of Other Endpoints

Quality of Life Analyses

Pain

The change from baseline in mean BPI was compared between treatment groups at Months 3, 6, 9 and 13.  The composite pain score was decreased for all treatment groups at each analysis time point.  The mean change from baseline in BPI composite pain scores was comparable between the treatment groups at Month 13, regardless of whether the patient had pain at baseline or not.  For the patient population as a whole and for the patients with pain at baseline, the BPI score decreased slightly over the period of the trial, in spite of the fact that many patients experienced progressive disease.

Analgesics

Analgesic scores ranged from 0 to 4, higher scores indicate that stronger analgesics were used. The mean change from baseline in analgesic score decreased at Month 3 and remained lower up to Month 13 for the three treatment groups.  There was no statistically significant difference between any treatment groups in the change from baseline in analgesic score.

ECOG performance status

The mean change from baseline in ECOG score increased from baseline to Month 13 for all three treatment groups.  There was no statistically significant difference between any treatment groups in the change from baseline in ECOG score.  However, the within treatment change from baseline in ECOG score was statistically significant in all treatment groups.

Quality of Life (FACT-G)

The FACT-G total score is the sum of four subscales, with a maximum possible score of 108.  An increase in the score from baseline indicated improvement.  The mean increase from baseline in the FACT-G total score was statistically significant between Zometa 8/4 mg and Aredia 90 mg compared with Zometa 4 mg.  The mean change from baseline was similar for Zometa 4 mg and Aredia 90 mg in the physical and emotional subscales.  There was a significant difference in the mean change from baseline in the functional and social subscales, with Aredia 90 mg showing a greater increase than Zometa 4 mg.  These small changes however, was were not clinically meaningful.

Table 5-10.        Mean/Median changes from baseline in quality of life scores at Month 13 by treatment group

	Zometa 4 mg	Zometa 8/4 mg	Aredia 90 mg
	N=561	N=524	N=555
Change in FACT-G total score      Total no. of patients * Baseline       Mean ± SD Change from baseline      Mean ± SD      Median     P-value: vs Aredia 90 mg      P-value: vs Zol 4 mg	446 76.2 ± 16.3 0.5 ± 14.9 0.30.031    -	418 75.0 ± 16.9 3.1 ± 14.9 3.00.839 0.020	445 76.9 ± 15.8 2.1 ± 15.6 2.0   -    -

* Number of patients who had a non-missing score at both baseline and Month 13 with last observation carried forward.

Time to progression of disease

There was no pairwise statistically significant difference in time to progression of bone metastases or in time to overall disease progression between the two treatment groups.

Table 5-11.        Time to progression of disease up to Month 13, multiple myeloma and breast cancer

	Zometa 4 mg	Zometa 8/4 mg	Aredia 90 mg
	N=561	N=524	N=555
Time to progression of bone lesions      Event rate at day 364      Median (days)     P-value: vs Aredia 90 mg      P-value: vs Zol 4 mg	61.5% 179.00.499     -	62.9% 175.0 0.602  0.866	62.4% 171.0   -   -
Time to progression of disease      Event rate at day 364      Median (days)P-value: vs Aredia 90 mg      P-value: vs Zol 4 mg	72.0% 134.00.184     -	71.6% 125.0 0.083  0.702	74.9% 111.0-    -

P-values are from Cox-regression with factor treatment stratified by the stratum.
Event rate and median  are derived from KM estimate.
Source: Post-text tables 9.2-55 and 9.2-56.

Objective bone lesion response

The best bone lesion response rates were similar for all treatment groups (see Table 5-12 on next page).

Table 5-12.        Best bone lesion response up to Month 13, multiple myeloma and breast cancer

	Zometa 4 mg	Zometa 8/4 mg	Aredia 90 mg
	N=561	N=524	N=555
Best bone lesion response      Complete response      Partial response      Stable disease      Progression      Not evaluable	0 (0%)   98 (17%) 170 (30%) 222 (40%)   71 (13%)	0 (0%)   77 (15%) 164 (31%) 219 (42%)   64 (12%)	0 (0%) 102 (18%) 154 (28%) 238 (43%)   61 (11%)

Bone biochemical markers

The median urinary N-telopeptide /creatinine ratio was decreased from baseline for all treatment groups at each time point.  The percent change was slightly greater for patients in the Zometa 4 mg and Zometa 8/4 mg treatment groups than for patients in the Aredia 90 mg group.  At End of Study, these differences were statistically significant.

5.2.6        Survival

The following survival analysis has been updated to include all data available from the ongoing extensions up to October 26, 2001.  Median survival is provided in Table 5-13 and Kaplan-Meier curve of survival is provided in Figure 5-3 below.

In Study 010, median survival has been reached for all treatment groups in the strata of breast cancer with hormonal therapy of study entry.  Median survival has not been reached for any treatment group in the multiple myeloma strata.  There are no significant differences in survival between Zometa 4 mg and pamidronate in study 010, either in each individual stratum or the total.

Table 5-13.        Summary of median survival (days) by stratum and treatment group

	Zometa 4 mg	Zometa 8/4 mg	Aredia 90 mg
Stratum
Multiple myeloma	NR	NR	NR
Breast cancer chemotherapy	570.0	637.0	640.0
Breast cancer hormonal	806.0	910.0	868.0
Total	838.0	830.0	811.0

 

 

Figure 5-3.        Kaplan-Meier curve of survival by treatment group

5.2.7        Efficacy Summary (breast cancer / multiple myeloma)

In the Zometa clinical trials program, Study 010 was designed to evaluate whether Zometa provides the level of benefit seen with Aredia in patients with multiple myeloma or bone metastases from breast cancer.  A placebo group was not utilized in this study, since Aredia has previously been shown to significantly delay or reduce the incidence of skeletal-related morbid events in these patients.  If Zometa and Aredia were equal in effectiveness, statistical power calculations indicated that the planned study enrollment (1600 patients) would have had power to exclude an 8% or greater inferiority for Zometa  in the proportion of patients experiencing an SRE (the primary study endpoint), compared with an anticipated beneficial treatment effect of 13.5% for Aredia (based on historical data).  The study results revealed a small difference in the proportion of patients experiencing an SRE, favoring Zometa 4 mg, and the lower confidence bound for this comparison with Aredia was 3.7% (indicating that Zometa was at least within 3.7 % of the effectiveness seen with Aredia as measured by this statistic).

Other analyses of skeletal related events, as noted in the study protocol statistical plan, considered (1) the time to first SRE, (2) the rate of SREs over time (Skeletal Morbidity Rate, SMR), and (3) Multiple Event Analysis (Anderson-Gill methodology was used).  These analyses complement the analysis of proportion of patients experiencing an SRE, and provide useful further information regarding the comparative patient benefits of Zometa and Aredia (a drug that is known to provide substantial benefit in this population).  The results of these analyses are provided in table 5-14, below.

Hazard ratios show comparable efficacy for Zometa 4 mg and 8/4 mg to pamidronate.

Table 5-14.        Summary of analysis of skeletal related events in Study 010 (Breast Cancer and Multiple Myeloma)

	Proportion with SRE	Time to First SRE (Hazard ratio)	Mean Skeletal Morbidity Rate	Multiple Event analysis (Hazard ratio)
Zometa 4 mg	248/561 (44%)	0.915	1.13	0.885
Zometa 8/4 mg	242/524 (46%)	0.994	1.08	0.910
Aredia	257/555 (46%)	-	1.40	-

Hazard ratio were Zometa treatment groups versus Aredia treatment group.

While the results do not demonstrate any statistically significant differences among the treatment groups, it is interesting to note that all of these analyses of skeletal-related morbid events were consistent in showing similar to slightly better results for both Zometa study arms compared to the Aredia study arm.  Based on evaluations of the historical data from placebo-controlled studies of Aredia, and the features of the patients enrolled in study 010 (data not shown), it is anticipated that Aredia had substantial benefit for patients in study 010, and thus the similar findings with Zometa reliably establish the effectiveness of Zometa in this population.

 

 

5.3         Prostate Cancer – Placebo-controlled trial (039)

5.3.1        Patient Disposition

This trial was designed with a placebo control, because no bisphosphonate, including the clinically effective bisphosphonate Aredia, has demonstrated efficacy in clinical trials in this population. A total of 643 prostate cancer patients with a history of bone metastases were randomized: 214 patients in the Zometa 4 mg group, 221 patients in the Zometa 8/4 mg group, and 208 patients in the placebo group.  These patients were randomized between 22-June-1998 and 10-November-1999 at one of 136 centers in 17 countries including centers in North and South America, Europe, and Australia and New Zealand.

Table 5-15.        Patient disposition – prostate cancer study

Number of patients	Zometa 4 mg N=214	Zometa 8/4 mg N=221	Placebo NN=208
Randomized Included in efficacy analysis	214 214	221 221	208 208
Completed 15 months	81(37.9%)	62(28.1%)	65(31.3%)
Reason for discontinuation
Adverse events	38 (17.8%)	44 (19.9%)	29 (13.9%)
Abnormal laboratory values	3 (1.4%)	5 (2.3%)	2 (1.0%)
Abnormal test procedure results	1 (0.5%)	0 (0.0%)	0 (0.0%)
Unsatisfactory therapeutic effect	19 (8.9%)	17 (7.7%)	34 (16.3%)
Condition no longer required study drug	1 (0.5%)	3 (1.4%)	3 (1.4%)
Protocol violation	1 (0.5%)	0 (0.0%)	0 (0.0%)
Withdrew consent	41 (19.2%)	50 (22.6%)	35(16.8%)
Lost to follow-up	4 (1.9%)	0 (0.0%)	5 (2.4%)
Administrative problems	0 (0.0%)	0 (0.0%)	3 (1.4%)
Death	25 (11.7%)	40(18.1%)	32 (15.4%)

5.3.2        Demographic and Disease Characteristics

Baseline patient demographics were generally comparable for all treatment groups.  Most of the randomized patients were > 60 years of age.

Table 5-16         Demographics in prostate cancer patients – ITT population

	Zometa 4 mg	Zometa 8/4 mg	Placebo
Demographic variable	(N=214)	(N=221)	(N=208)
Age (years)
Mean ± S.D.	71.8 ± 7.92	71.2 ± 7.99	72.2 ± 7.89
Range	45 - 90	43 - 90	37 - 90
> 60	195 ( 91.1)	202 ( 91.4)	193 ( 92.8)
Race n (%) patients
Caucasian	178 (83.2%)	186 (84.2%)	172 (82.7%)
Black	24 (11.2%)	19 (8.6%)	19 (9.1%)
Other	12 (5.6%)	16 (7.2%)	17 (8.2%)
Weight (kg)
Mean ± S.D.	82.8 ± 14.16	82.1 ± 14.44	83.4 ± 16.08
Range	40 - 120	46 - 126	43 - 138

Treatment groups were also comparable with respect to baseline quality of life variables, and disease characteristics, with some minor exceptions.  Median baseline serum PSA was higher in both Zometa groups than in the placebo group.  Also, the BPI composite pain score and proportion of patients with pain at baseline were higher in the Zometa 8/4 mg group than in the other two groups.

Table 5-17.        Baseline characteristics in prostate cancer patients (ITT population)

	Zometa 4 mg	Zometa 8/4 mg	Placebo
Total number of patients	214	221	208
Previous SRE	66 (30.8)	71 (32.1)	78 (37.5)
Time since initial diagnosis of cancer (months, mean)	62.5	67.6	66.6
Time since bone metastasis (months, mean)	23.5	25.8	28.4
PSA (ng/dl)
mean ± S.D.	276.5 ± 737.1	350.9 ± 1148.9	211.1 ± 464.9
median	81.7	88.2	61.0
ECOG status n (%)
0-1	197 ( 92.1)	199 ( 91.3)	190 ( 91.3)
³2	17 ( 7.9)	19 ( 8.7)	18 ( 8.7)
BPI composite pain score
N	193	199	191
Mean ±SD	2.0 ± 1.98	2.5 ± 2.10	2.1 ± 2.04
Median	1.8	2.3	1.8
Analgesic score
0	94 ( 43.9)	73 ( 33.0)	77 ( 37.0)
1	69 ( 32.2)	84 ( 38.0)	77 ( 37.0)
2	9 ( 4.2)	11 ( 5.0)	9 ( 4.3)
3	40 ( 18.7)	48 ( 21.7)	41 ( 19.7)
4	2 ( 0.9)	3 ( 1.4)	3 ( 1.4)
FACT-G total score
N	193	193	187
Mean ±SD	81.0 ± 15.30	81.4 ± 13.76	82.2 ± 14.57
Median	82.5	82.2	82.8
Baseline creatinine
< 1.4 mg/dL	173 (80.8)	170 (76.9)	170 (81.7)
³ 1.4 mg/dL	41 (19.2)	48 (21.7)	33 (15.9)

5.3.3       Primary efficacy analysis

The proportion of patients having any SRE in the Zometa 4 mg group was significantly smaller than that in the placebo group (Table 5-18).  The Zometa 8/4 mg group also had a smaller proportion of patients with SREs than the placebo group, but the difference did not reach statistical significance.  There was also no statistically significant difference between the two Zometa groups for this comparison.  There were two patients who had HCM during the study, but each had other SREs and was thus counted in the SRE (-HCM) analysis.  Hence the proportion of patients with prostate cancer having any SRE (-HCM) is identical to the proportion of patients having any SRE (+HCM).

Table 5-18.        Proportion of patients with prostate cancer with an SRE up to Month 15

Proportion = (no. of patients with the event)/(total no. in the group) up to Month 15;
Confidence interval for the difference (treatment labeled in the column minus row) of percent of patients with events.
P-values are based on stratified Cochran-Mantel-Haenszel test for the proportion.

 

5.3.4        Secondary skeletal endpoint analyses

Proportion of patients with individual skeletal-related events

The proportion of patients having each type of SRE was lower in the Zometa 4 mg group and Zometa 8/4 mg group than in the placebo group except for the change of antineoplastic therapy (Figure 5-4).  The difference in the proportion of patients with a pathological fracture between the Zometa 4 mg group and the placebo group was statistically significant (p=0.015).

Figure 5-4.    Individual SREs in patients with prostate cancer

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Time to first skeletal-related event

The difference in time to the first occurrence of any SRE between the Zometa 4 mg group and the placebo group was statistically significant.  For Zometa 4 mg treatment group, the median time to the first occurrence of a SRE was not reached, but it was at least 420 days (the estimated event rate was less than 50% on day 420), while the median time in the placebo group was 321 days.  Hence the median time to the first occurrence of a SRE for the Zometa 4 mg group was increased by at least 99 days compared to the patients in the placebo group.

The times to the first SRE (-HCM) and the first SRE (+HCM) are identical in this study as hypercalcemia was not the first SRE for the two patients who had hypercalcemia.

 

Table 5-19.        Time to the first SRE up to Month 15 in prostate cancer patients

					P-values for the between treatment comparison
	N	Event rate at day 420	25% Quartile (days)	Median (days)	Zometa 4 mg	Zometa 8/4 mg
Placebo	208	57.19%	122	321	0.011	0.491
Zometa 4 mg	214	44.87%	182	Not Reached	-	0.059
Zometa 8/4 mg	221	53.17%	127	363	-	-

P-values are from the log-rank test for comparing the distribution of time to the event.
For stratum =total, the p-values are from Cox-regression with factor treatment, stratified by the strata. Event rate is the Kaplan Meier estimate of the event rate at day 420.
Event rate, Median and 25%quartile are derived from KM estimate.

Skeletal morbidity rate

The skeletal morbidity rate up to Month 15 was lower for patients in the Zometa 4 mg and Zometa 8/4 mg groups than in the placebo group (Table 5-20).  The difference in the SMR between the Zometa 4 mg group and the placebo group was statistically significant.

Table 5-20.        Skeletal morbidity rate (risk set definition) of any SRE (-HCM) up to Month 15

		Skeletal morbidity rates (no. of events per year)		p-values for the between treatment comparison
	N	Mean ± SD	Median	Zometa 4 mg	Zometa 8/4 mg
Placebo	208	1.49 ± 3.336	0.00	0.006	0.143
Zometa 4 mg	214	0.80 ± 1.703	0.00	-	0.191
Zometa 8/4 mg	221	1.06 ± 2.193	0.00	-	-

Skeletal morbidity rate (risk set definition) is the number of SREs divided by the time at risk in years, counted from the date of randomization, with every counted event followed by a 20 day period during which any SRE experienced by the patient would not be counted, nor would the patient be counted as at risk.  For patients with no SRE, the whole study period was counted as at risk.

Multiple event analysis of SREs

For SRE (-HCM), there was a statistically significant difference between the Zometa group and the placebo group in the time to multiple occurrence of the event in favor of the Zometa 4 mg group (p=0.004).  See Table 5-21 below.

 

Table 5-21.        Multiple-event analysis of time to SRE up to Month 15 in prostate cancer patients (039)

	Zometa 4 mg v.s.				Zometa 8/4 mg v.s.
	Hazard ratio*	95% CI for Hazard ratio	P-value		Hazard ratio*	95% CI for Hazard ratio	P-value
Placebo	0.643	(0.476, 0.870)	0.004		0.847	(0.640, 1.122)	0.247
Zometa 4 mg					1.318	(0.966, 1.798)	0.081

* Hazard ratio is the ratio of column versus row.  Hazard ratios and p-values are from Anderson-Gill approach for time to multiple events

P-value and the 95% C.I. are computed using the robust variance estimate

5.3.5        Analyses of Other Endpoints

Pain

Overall, the mean BPI composite pain score increased from baseline to Month 15 for all three treatment groups.  However, the difference in favor of Zometa between the BPI composite pain scores of both of the Zometa 4 mg and Zometa 8/4 mg groups and the placebo group at months 3 (p= 0.003 and p= 0.003, respectively) and 9 (p=0.030 and p=0.014, respectively) was statistically significant.  The difference in favor of Zometa was also statistically significant between the 8/4 mg group and the placebo group at several other time points.

Analgesics

The mean analgesic score increased from baseline to Month 15 for all three treatment groups. The difference in analgesic score from baseline for the total of all patients was either similar for all three treatment groups or the placebo group had a slightly higher score than the two Zometa groups at all timepoints.

Performance status (ECOG)

The mean ECOG score increased from baseline to Month 15 for all three treatment groups.  Overall, there was no statistically significant difference between any treatment groups in the change from baseline in ECOG score.

Quality of life (FACT-G)

The total score for FACT-G is the sum of the physical, functional, social and emotional subscales.  The EQ-D5 questionnaire consists of the EURO QOL-5D score and the EURO QOL-5D thermometer.  An increase in the score from baseline indicates improvement while a negative number indicates worsening.

At Month 15 there was some improvement in the FACT-G social subscale score and the EURO QOL-5D score for all treatment groups, but there were no statistically significant differences between treatment groups for any of the quality of life scores.

Time to Progression

There was no statistically significant difference between treatment groups in the distribution of time to progression of bone metastases or in the distribution of time to overall disease progression.

Table 5-22.        Summary of time to progression of disease at Month 15 in prostate cancer patients

	Zometa 4 mg	Zometa 8/4 mg	Placebo
	N= 214	N=  221	N= 208
Time to progression of bone lesions      Median (days)      P-value: vs Placebo      P-value: vs Zometa 4mg	92.0 0.275     -	89.0  0.710  0.342	87.0   -   -
Time to progression of disease      Median (days)      P-value: vs Placebo      P-value: vs Zometa 4mg	84.0 0.771     -	84.0 0.408 0.275	84.0   -   -

Note: Stratified log-rank test is used for the between treatment comparison. The median is derived from KM estimate.

Objective bone lesion response

No patients achieved complete response during the study. In both Zometa treatment groups, a higher percentage of patients had no change in bone lesions while a lower percentage had progression of bone lesions compared with patients in the placebo group.

Table 5-23.        Frequency distribution of best response by treatment group

	Zometa 4 mg	Zometa 8/4 mg	Placebo
	N=214	N=221	N=208
Best bone lesion response      Complete response      Partial response      No change      Progression      Not evaluable	0 (0%)  9 (4%)   47 (22%) 118 (55%) 40 (1 9%)	0 (0%)   6 (3%)   46 (21%) 123 (56%) 46 (21%)	0 (0%)  8 (4%) 35 (17%) 132 (63%) 33 (16%)

5.3.6        Survival

The following survival analysis has been updated to include all data available from the ongoing extensions up to October 26, 2001.  Median survival is provided in Table 5-24 and Kaplan-Meier curve of survival is provided in Figure 5-5 below.

 

 

 Table 5-24.       Summary of median survival (days) by stratum and treatment group

	Zometa 4 mg	Zometa 8/4 mg	Placebo
Stratum
No metastases at initial diagnosis	515.0	419.0	471.0
Metastases at initial diagnosis	646.0	377.0	466.0
Total	563.0	418.0	469.0

Figure 5-5.        Kaplan-Meier curve of survival by treatment group

 

5.3.7        Efficacy Summary (prostate study)

The primary efficacy variable was compared between the 4 mg Zometa treatment group and placebo using a stratified Cochran-Mantel-Haenszel test with modified ridit scores.

For time to the first SRE (-HCM) and each type of SRE, survival analysis methods, including Kaplan-Meier product-limit estimates of the “survival functions”, and the Cox regression stratified with the stratum were used to assess the effectiveness of Zometa 4 mg in prolonging the time to the first event.

For skeletal morbidity rate of SRE (-HCM), SRE (+HCM), and each type of SREs, stratified Cochran-Mantel-Haenzsel with modified ridit scores were used to assess the difference between the Zometa 4 mg and the placebo

Zometa 4 mg given as a 15-minute infusion every 3 weeks in addition to standard antineoplastic therapy was significantly more effective than placebo in treatment of bone metastases in patients with prostate cancer (see Table 5-25 below):

·        the proportion of patients having skeletal-related events was significantly lower with Zometa 4 mg than with placebo

·        the time to the first skeletal-related event was significantly longer with Zometa 4 mg than with placebo

·        the skeletal morbidity rate was significantly lower with Zometa 4 mg than with placebo

·        hazard ratio for SRE recurrent rates favored Zometa 4 mg over placebo

·        the proportion of patients with individual types of SREs and the SMR for each SRE was lower with Zometa compared with placebo

·        patients treated with Zometa had a smaller increase in bone pain than patients receiving placebo, with little between group difference in analgesic score, in spite of similar deterioration of performance status in the three treatment groups

Table 5-25.        Summary of analysis of skeletal related events in Study 039 (Prostate cancer)

	Proportion with SRE	Time to First SRE (Event Rate) Day 420	Skeletal Morbidity Rate	SRE Recurrence (Anderson-Gill) (Hazard ratio)
Zometa 4 mg	71/214 (33%)*	44.87%*	.80*	0.643*
Zometa 8/4 mg	85/221 (38%)	53.17%