Vaccines and Related Biological Products Advisory Committee

Meeting Date: December 15, 2005

FDA Clinical Briefing Document for

Merck & Co., Inc.

Zoster vaccine live (Oka/Merck)

Zostavax™

Patricia Rohan, M.D.

CBER/FDA

Page

1.0 General Information 3

PRODUCT NAME

PRODUCT COMPOSITION 3

PROPOSED INDICATION 3

DOSING REGIMEN AND ROUTE OF ADMINISTRATION 3

PRINICPAL STUDIES SUPPORTING LICENSURE 3

EXECUTIVE SUMMARY 4

2.0 Introduction and Background 7

2.1 EPIDEMIOLOGY OF HERPES ZOSTER 7

2.2 REGULATORY BACKGROUND 7

2.3 BASIS FOR LICENSURE 8

3.0 Clinical Development Overview 8

4.0 Pivotal Study Overview - Protocol 004 10

4.1 OBJECTIVES 10

4.2 DESIGN 10

4.3 POPULATION 10

4.4 ENROLLMENT CRITERIA 10

4.5 VACCINE ADMINISTRATION 11

4.6 OTHER PRODUCTS USED IN PROTOCOL 004 12

5.0 Endpoints – Protocol 004 13

6.0 Laboratory Methods and Surveillance - Protocol 004 14

7.0 Case Definitions – Protocol 004 18

8.0 Statistical Considerations – Protocol 004 18

9.0 Study Results – Protocol 004 19

9.1 POPULATION ENROLLED / DISPOSITION 19

9.2 STUDY FOLLOW-UP 20

9.3 DETERMINATION OF HERPES ZOSTER CASES 22

9.4 ENDPOINTS RESULTS 23

10.0 Additional Studies - Protocol 009 51

11.0 Additional Data Tables for Protocol 004 58

12.0 References 67

1.0 General Information

Product name

Proper name: Varicella Virus Vaccine Live (Oka/Merck)

Proposed trade name: ZOSTAVAX™

Product composition:

ZOSTAVAX, manufactured at Merck & Co., Inc. (Merck), West Point, Pennsylvania, is a sterile lyophilized product prepared by formulating the attenuated Oka/Merck VZV strain (a live attenuated virus) propagated in MRC-5 cell culture. The drug product is stored at –15 °C until use and is reconstituted with sterile diluent just prior to use.

Merck intends that the licensed product will be a 0.65 mL dose formulation which when reconstituted as directed and stored at room temperature for up to 30 minutes, each 0.65-mL dose contains a minimum of 19,400 PFU (plaque forming units) of Oka/Merck varicella-zoster virus; 31.16 mg of sucrose, 15.58 mg of hydrolyzed gelatin, 3.99 mg of sodium chloride, 0.62 mg of monosodium L-glutamate monohydrate, 0.57 mg of sodium phosphate dibasic, 0.10 mg of potassium phosphate monobasic, 0.10 mg of potassium chloride; residual components of MRC-5 cells including DNA and protein; and trace quantities of neomycin, and bovine calf serum. The product contains no preservatives.

The virus seeds, drug substance process and varicella vaccine bulk used for ZOSTAVAX™ are the same as for varicella component in Merck’s FDA licensed vaccines, VARIVAX™ and ProQuad™. The varicella component of these latter two vaccines is indicated for vaccination against varicella in individuals 12 months of age and older in the case of VARIVAX™ and in children 12 months to 12 years of age in the case of ProQuad®.

Products Used

Manufacturer Merck & Co., Inc.

Proposed indication Immunization of adults ≥50 yrs for prevention

of herpes zoster, postherpetic neuralgia and

reduction of acute chronic zoster associated

pain

Dosing regimen and

Route of administration Single dose, subcutaneously

PRINCIPAL CLINICAL STUDIES SUBMITTED IN SUPPORT OF LICENSURE AND LABELING

Protocol 004

This protocol was designed to demonstrate the safety and efficacy of a single 0.5 mL dose of ZOSTAVAX™ in persons aged 60 years and older to prevent herpes zoster (HZ), post-herpetic neuralgia (PHN) and the burden of illness due to HZ-associated pain as measured by the burden-of-illness (BOI) score.

Protocol 009

Protocol 009 was designed to support an indication in persons 50-59 years of age and use of a higher potency and volume (0.65 mL) ZOSTAVAX™ dose. Subjects were evaluated for safety.

Executive Summary:

This briefing document contains a summary of the safety, immunogenicity and efficacy data from Protocol 004 and safety data from Protocol 009 provided by Merck & Co., Inc. to support approval of their Varicella Virus Vaccine Live (Oka/Merck) ( Trade name: Zostavax™), for single dose immunization of adults 50 years of age and older.

Efficacy Data – Protocol 004

The submitted efficacy data from pivotal Study 004 demonstrate that in a group of relatively healthy adults, aged 60 years and older, a single dose of ZOSTAVAX™ (22,000 – 62,500 pfu/dose) reduced the incidence of herpes zoster (HZ) by 51.3% (95% CIs 44.2, 57.6), the incidence of PHN by 66.5% (95% CIs 47.5, 79.2) and reduced the burden-of-illness (BOI) score, a composite measure of HZ pain incidence, severity and duration, by 61.1% (95% CIs 51.1, 69.1) when comparing rates by treatment groups over the first ~3 years following vaccination.

Statistically significant differences were demonstrated in the duration of post-herpetic neuralgia (20 days in the vaccine group vs. 22 days in the placebo group (p-value <0.001 for MITT, p-value = 0.041 for evaluable HZ).

No statistically significant vaccine effects were seen on the rates of mortality, hospitalizations (overall and zoster-related), serious morbidity, use of pain medications and interference with activities of daily living (ADLI) over the course of the study.

The study report shows a trend of decreasing efficacy in all three major efficacy endpoints (HZ incidence, PHN incidence and BOI) over the first 3 years following vaccination. Interpretation of data in later years, i.e., more than 3 years postvaccination, is limited by the relatively small proportion of subjects with follow-up.

Age appears to be an important factor in some study measures of vaccine efficacy: the incidence of HZ was reduced by 63.9% in subjects 60-69 years old and by 37.6% in those 70 years of age and older; and the BOI score was reduced by 65.5% in subjects 60-69 years old and by 55.5% in those 70 years of age and older. A similar reduction in PHN was seen in both ZOSTAVAX™ and placebo recipients in both age groups.

In FDA analyses of Study 004 data, the incidence of post-herpetic neuralgia occurring or persisting at 90-days following HZ rash onset was reduced from 12.5% in placebo recipients who developed HZ to 8.6% in vaccine recipients who developed HZ (p-value [Fisher’s Exact = 0.08]. The median HZ BOI was 82.50 in vaccine recipients who developed HZ and 87.75 in placebo recipients who developed HZ (p-value (Wilcoxon) = 0.25). It appears that the major treatment effect in Protocol 004 is reflected by the decrease in the incidence of HZ (51.1%) and that the additional major efficacy endpoints, HZ BOI and incidence of PHN provide little additional information. Further information is contained in the FDA Statistical Reviewer’s Briefing Document.

FDA analysis of the data further reveals a consistent trend toward progressive loss of vaccine efficacy in prevention of HZ with increasing age, although the numbers of subjects and cases of HZ occurring at the oldest ages are relatively small.

Immunogenicity Data – Protocol 004

Blood samples were taken from subjects in the CMI substudy and from subjects developing HZ for gpELISA, responder cell frequency (RCF) and INF-γ ELISPOT assays.

Antibody data from gpELISA assay appears to be a measure of postvaccination immune response. Considering the fact that the vaccine is acting as a “booster” in subjects with previous primary varicella infection, the kinetics of the antibody response might be expected to exhibit higher and more discriminatory titers at earlier time points, e.g., 2-3 weeks postvaccination.

Measures of cellular immunity, RCF and IFN-γ ELISPOT data, are limited by lower the relatively lower magnitude of response and relatively higher variability in assay results as well as the technical complexity of handling the samples. The data derived from RCF and ELISPOT at this point do not appear to add significantly to the information derived from gpELISA data.

Naturally Occurring HZ – Protocol 004

Following naturally occurring HZ, both ZOSTAVAX™ and placebo recipients exhibited larger immune responses as measured by gp ELISA when compared to immune responses seen postvaccination with ZOSTAVAX™. 6 weeks following ZOSTAVAX™ vaccination, gpELISA geometric mean titer (GMT) was 575 (95% CIs: 442, 511) and geometric mean fold rise (GMFR) was 1.7 (95% CIs: 1.6, 1.8). In contrast, 6 weeks following onset of herpes zoster gpELISA GMT was 2042 (95% CIs: 1805, 2310) and GMFR was 3.2 (95% CI: 2.6, 3.9) in ZOSTAVAX™ recipients; and gpELISA GMT was 2260 (95% CI: 2070, 2467) and GMFR was 3.1 (95% CIs: 2.7, 3.5) in placebo recipients. The clinical relevance of this higher response with respect to risk and severity of subsequent HZ in naturally occurring HZ as compared to HZ occurring postvaccination with ZOSTAVAX™ remains to be elucidated.

Safety Data – Protocol 004

Overall, safety data from the pivotal clinical study demonstrate no particular pattern of postvaccination adverse events such as deaths, hospitalizations or serious adverse events.

In all cases in which VZV DNA was detected by PCR from suspected HZ lesions in Protocol 004, the strain identified was wild-type; the attenuated vaccine strain was not detected in any case. Protocol 001 did include one case of varicella-like rash with onset at Day 310 postvaccination which was positive for Oka/Merck varicella virus by PCR.

Reactogenicity following vaccination, as measured in the AE Monitoring Substudy, was higher in ZOSTAVAX™ recipients compared to placebo recipients for rate of injection-site adverse events (48% in of zoster vaccine recipients, 17% of the placebo recipients). This was primarily due to vaccine-associated increases in solicited adverse events, erythema, pain/tenderness and swelling. Several types of unsolicited injection-site reactions, including swelling, warmth and pruritus were reported at higher rates in ZOSTER™ recipients. There were no notable differences in the rates of systemic vaccine-associated AEs, including fever.

Local injection-site reactions were relatively higher in females (~5-10% in placebo recipients; ~40-50% in ZOSTAVAX™ recipients) versus males (~4-8% in placebo recipients; 15-25% in ZOSTAVAX™ recipients).

Local injection site reactions were also relatively higher in the younger 60-69 year old cohort (~5-10% in placebo recipients; ~30-43% in ZOSTAVAX™ recipients) versus ≥ 70 year old cohort (~4-7% in placebo recipients; 20-30% in ZOSTAVAX™ recipients).

Safety Follow-Up – Protocol 004

Day 0-42 postvaccination safety follow-up data was collected using the Automated Telephone Response System (ATRS). 55% of all study subjects responded to the ATRS around Day 42 and study investigators added information for an additional 11% subjects over a 4 year period postvaccination. There are 1240 additional reports for subjects already accounted for in this dataset – some of these additional reports are added several years after the initial entry.

No information on the reporting rates is available for monthly ATRS contacts used to identify potential cases of HZ in all subjects and to provide safety follow-up in the Adverse Event Monitoring Substudy (AE Substudy). Likewise, there is no information on reporting rates by month, by site, by baseline characteristics and by study outcomes.

Study termination procedures included directly querying each subject. The study reports final accounting (~2-5 years postvaccination) for 99% of subjects, consisting of those identified as directly contacted by study personal at the conclusion of study and those confirmed as deceased. Less than 0.5% of all study subjects were identified as drop-outs or those lost to follow-up. In 99% of those reported as having contact at study termination, there is no date given for last contact. The proportion of subject termination records resulting from direct subject contact and the proportion of subject records, if any, resulting from “pre-populated” data, i.e., derived from other data sources, is unclear.

No information is available regarding the identity of subjects who were enrolled in the pivotal study at VA Medical Center sites who were not eligible to receive VA Medical Center healthcare. Therefore it is impossible to determine whether differences in baseline subject characteristics or differential follow-up and access to healthcare information might impact on the reported safety and/or efficacy of ZOSTAVAX™.

Per the sponsor comment on page 309 of Protocol 004 Clinical Study Report: “Due to the passive and inconsistent nature of safety data collection in the Routine Safety Monitoring Cohort for adverse experiences occurring from Day 43 postvaccination through study end, caution should be exercised when interpreting these particular data.”

Limitations of Protocol 004

Although age is a primary determinant in the frequency and severity of HZ-related disease, there are relatively fewer subjects at the upper age range of the study population. Additionally, the study sought to enroll relatively healthy, older subjects, excluding those with common co-morbidities (e.g., regular use of inhaled corticosteroids; subject survival not estimated to be at least 5 years; subject homebound or not ambulatory; subject with cognitive impairment or severe hearing loss) any of which might have resulted directly or indirectly in lower efficacy estimates and higher reported rates of adverse events than those seen in Protocol 004. Additionally, this relatively healthy study population may not provide an adequate opportunity to evaluate whether or not the vaccine had any impact on zoster-related hospitalizations and other severe complications of HZ.

Since most study participants who developed HZ were treated with antivirals, the rate of HZ complications may have been decreased.

Given the lack of information on the proportion of subjects responding at the monthly ATRS contact and at study termination, and the documented deficiencies in subject reporting to the Day 42 safety follow-up, it is difficult to draw conclusions as to the relative safety of ZOSTAVAX™ at this time.

Protocol 009

This study evaluated safety of ZOSTAVAX™, comparing a higher potency dose (207,000 pfu) to a lower potency dose (58,000 pfu). The lower potency is in the range of the highest potency evaluated in the pivotal study, Protocol 004.

Subjects 50 years and older, stratified into two groups: 50-59 years old and ≥ 60 years old were enrolled in Protocol 009. Note: Protocol 004, the pivotal study, enrolled two age strata: 60-69 years old and ≥ 70 years old.

The two study endpoints for ZOSTAVAX™ were based upon the historical adverse event rate for PNEUMOVAX23™ and the assumption that no vaccine-related serious adverse events would occur, respectively. The clinical relevance of these two endpoints is unclear.

Subjects receiving the higher dose experienced higher rates of solicited and non-solicited vaccine-related injection site reactions, although few were characterized as severe. There did not appear to be a higher rate of systemic vaccine-related AEs.

The younger cohort (50-59 years old) experienced relatively higher rates of injection-site reactions as compared to the older cohort (≥ 60 years old). The younger cohort also characterized more specified systemic adverse events, e.g., headache, as severe.

No specific pattern of SAEs was observed, although 4 SAEs were seen in the higher dose group and 1 SAE in the lower dose group.

2.0 Introduction and Background

2.1 Epidemiology of Herpes Zoster

Varicella-zoster virus (VZV) is an alphaherpes virus and a member of the Varicellovirus genus in the Herpesviridae family of viruses. Infection in humans occurs via the respiratory tract and conjunctiva. As with other herpes viruses, VZV has the capacity to persist in the body after a primary infection, in the case of VZV in the sensory nerve ganglia.

Disease manifestions may include a mild, prodromal phase (malaise, fever) prior to the appearance of the typical pruritic, vesicular rash which can be distributed in a localized or diffuse pattern. Complications of primary varicella infection can include bacterial superinfection of involved skin and underlying soft tissues, and pneumonia particularly in adults. Infections may be severe and include septicemia, toxic shock syndrome, necrotizing fasciitis, osteomyelitis, bacterial pneumonia and septic arthritis. Less common complications can involve the central nervous system, e.g., cerebellar ataxia, encephalitis; liver and hematopoietic systems. Disseminated intravascular coagulation can occur in rare cases. Hospitalization rates of ~3/1,000 cases and death rates of ~1/60,000 cases have been reported following varicella infection. Adults, immunocompromised individuals and infants of mothers experiencing varicella rash in the perinatal period are all at increased risk of complications from varicella infection. Intrauterine infections, particularly early in gestation, can rarely result in congenital varicella syndrome.

Herpes zoster is due to reactivation of latent varicella zoster virus, usually many years following a primary varicella virus infection (chickenpox) and manifests as a unilateral, vesicular rash in a dermatomal distribution and systemic symptoms are unusual. Zoster-associated pain, paresthesias and pruritis usually resolve within a few weeks, but in some cases severe, debilitating pain and paresthesia may persist for a year or more and symptomatic treatment may be only partially effective. Other serious complications involve ocular and visceral organ involvement. While transmission of the virus to other individuals has been documented, it occurs rarely. Reactivation is associated with aging and immunosuppression. Individuals with intrauterine exposure or varicella at an early age (< 18 months) are also at higher risk of VZV reactivation.

VARIVAX™ (Varicella Virus Vaccine Live) a live, attenuated strain of the virus was licensed in the U.S. in 1995 for vaccination against varicella in individuals 12 months of age and older.

From the time of licensure to 2003, rates of primary varicella have decreased by approximately 85% as reported by the CDC Varicella Active Surveillance Project (VASP) with nationwide vaccine rate of 85% among the population for whom it is recommended. VASP is also monitoring rates of HZ, as circulating wild-type varicella has been suggested as a source of external boosting (i.e., exposure to varicella disease in the community) that prevents reactivation of VZV and subsequent HZ and its manifestations. Data from the Massachusetts Behavioral Risk Factor Surveillance System shows age-standardized rates of overall herpes zoster occurrence increasing from 2.77/1,000 to 5.25/1,000 (90%) in the period 1999–2003. This trend in both crude and adjusted rates was highly significant (p < 0.001), specifically in the 25–44 year and 65+ year age groups (Yih, 2005).

2.2 Regulatory Background

ZOSTAVAX™ has been formulated using the same varicella virus bulk product contained in the Merck VARIVAX™ and ProQuad™ vaccines. The varicella component contained in these latter two vaccines is indicated for vaccination against varicella in individuals 12 months of age and older.

ZOSTAVAX™ has not been licensed in any country to date.

The proposed trade name, ZOSTAVAX™, will be used throughout this document to indicate the formulations

of the Varicella Virus Vaccine Live (Oka/Merck) studied in Protocols 001, 002, 003, 004, 005, 007, and 009.

Only Protocols 004 and 009 are described in this briefing document.

2.3 Basis for Licensure

There is no current preventive treatment for herpes zoster.

The development strategy followed by Merck & Co., Inc. to support licensure of ZOSTAVAX™ in older adults was based upon the following:

· Demonstration of safety of a single dose of ZOSTAVAX™ in healthy adults 60 years and older

· Demonstration of efficacy of a single dose of ZOSTAVAX™ in healthy adults 60 years and older [decreases in the incidence of HZ, the incidence of post-herpetic neuralgia (PHN), the burden-of-illness (BOI) score, the duration of PHN and interference with activities of daily living (ADLI).]

· Demonstration of clinical lot-to-lot consistency by comparison of major efficacy endpoints (incidence of HZ, incidence of PHN and changes in BOI scores) in healthy adults aged 60 years and older by ZOSTAVAX™ lot administered

· Demonstration of safety (42 days postvaccination) of a higher potency ZOSTAVAX™ in adults aged 50-59 years.

3.0 OVERVIEW OF ZOSTER VACCINE CLINICAL STUDIES

The Biologics License Application (BLA) contains safety, immunogenicity and efficacy data from Protocol 004, and safety data with or without immunogenicity data from six supportive clinical studies in adults: Protocols 001, 002, 003, 005, 007 and 009. Reports from Protocol 012 and 049, used to support licensure of ProQuad™ and VARIVAX™, respectively are submitted. For each of these studies full study reports were submitted.

The overall safety database comprised approximately 21,000 subjects who received ZOSTAVAX™. 19,270 subjects received ZOSTAVAX™ in Protocol 004 and 698 subjects received ZOSTAVAX™ in Protocol 009.

4.0 PIVOTAL SAFETY AND IMMUNOGENICITY STUDY - PROTOCOL 004

Title: Trial of Varicella-Zoster Vaccine for the Prevention of Herpes Zoster and its Complications

Department of Veterans Affairs Cooperative Study #403

4.1 Objectives

Co-Primary Objectives

· To determine if immunization with zoster vaccine will reduce the incidence and/or severity of herpes zoster (HZ) and its complications, primarily post herpetic neuralgia (PHN), in persons 60 years of age and older.

· To determine if immunization with zoster vaccine will protect against PHN.

Secondary Objectives (originally designated as tertiary objectives)

· To determine if immunization with zoster vaccine will reduce the incidence of HZ.

· To determine if immunization with zoster vaccine will reduce the duration of HZ pain.

· To determine if immunization with zoster vaccine will reduce the Activities of Daily Living Interference (ADLI) in subjects who develop HZ.

Tertiary Objectives

· To assess the effect of the vaccine on the incidence of PHN using alternative definitions of pain that persists or appears more than 30, 60, 120, and 182 days after rash onset.

· To examine the vaccine efficacy and immunogenicity of three consistency lots of the vaccine.

· To examine the effect of a reduction in the plaque forming unit (PFU) content of the vaccine over time on its efficacy.

· To assess the varicella zoster virus (VZV) - specific immune response to the zoster vaccine.

· To collect specific data on the impact of HZ and its complications on the quality of life of older persons.

· To provide specific data on the natural history of HZ in older persons, during a time when most will be treated acutely with an antiviral drug (acyclovir, famciclovir, or valacyclovir).

· To assess the safety profile of the zoster vaccine in persons 60 years of age or older.

· To explore potential covariate effects of gender, age, antiviral and analgesic medications on the burden-of-illness (BOI) and vaccine efficacy in study subjects who develop HZ.

· To describe the HZ-related health care resource utilization (beyond study protocol mandated visits) and out-of- pocket expenses incurred by study subjects with HZ.

· To determine whether use of the zoster vaccine is associated with a significant reduction in the number of nonprotocol-mandated HZ-related health care contacts relative to placebo recipients.

4.2 Design

Protocol 004 is described as a randomized, double-blind, placebo-controlled, 22-center study of the safety, efficacy, immunogenicity and consistency of three manufacturing lots of ZOSTAVAX™ in relatively healthy adults (≥ 60 years of age).

4.3 Population

The study planned enrollment of approximately 37,200 adults 60 years of age and older, including a target enrollment of approximately 11,160 subjects in the 60 to 69 years of age stratum and approximately 7440 in the ≥ 70 years of age stratum in each vaccination group.

4.4 Enrollment Criteria

Inclusion Criteria

· History of varicella or long-term (≥30 years) residence in the continental USA.

· 60 years of age or older.

· Informed consent obtained from the subject.

Exclusion Criteria

· Immunosuppression resulting from disease (e.g., malignancy; human immunodeficiency virus [HIV] infection), corticosteroids (except intermittent topical or inhaled corticosteroid [<800 mcg/day beclomethasone dipropionate or equivalent]), or other immunosuppressive/cytotoxic therapy (cancer chemotherapy or organ transplantation).

· Active neoplastic disease (except local skin cancer or other malignancies [e.g., prostate cancer] that was stable in the absence of immunosuppressive/cytotoxic therapy).

· Prior HZ.

· Prior receipt of varicella vaccine.

· Allergic sensitivity to neomycin.

· History of anaphylactoid reaction to gelatin.

· Significant underlying illness that would be expected to prevent completion of the study (e.g., life-threatening disease likely to limit survival to less than 5 years).

· Not ambulatory (bed-ridden or homebound).

· Receipt of immune globulin or any other blood product within 3 months before or planned during the 3- to 5-year study period.

· Receipt of any other vaccines within 1 month before study vaccination (2 weeks in the case of inactivated influenza vaccines or other inactivated vaccines [e.g., diphtheria toxoid, tetanus toxoid, pneumoccocal vaccine, hepatitis A vaccine, hepatitis B vaccine]), or scheduled within 6 weeks after study vaccination.

· Receipt of antiviral therapy at the time of enrollment, in order to avoid potential confounding of vaccine effectiveness.

· Any other condition (e.g., extensive psoriasis, chronic pain syndrome, cognitive impairment, severe hearing loss) that, in the opinion of the site investigator, might have interfered with the evaluations required by the study.

· Intercurrent illness (e.g., urinary tract infection, influenza) that might have interfered with the interpretation of the study results.

· Female and premenopausal (women who entered the study had to be postmenopausal).

· Unlikely to adhere to protocol follow-up.

· Involved in a conflicting (vaccine or investigational drug) clinical trial.

· History of recurrent herpes simplex virus (HSV) and had more than 3 episodes per year for which the subject was treated with episodic antiviral therapy (e.g., 400 mg oral acyclovir 3 times daily for 5 days) or was on constant daily antiviral therapy.

· History of multiple sclerosis, in order to avoid potential confounding of vaccine effectiveness due to antiviral or immunosuppressive therapy.

4.5 Vaccine Administration

4.5.1 General Vaccine Information

All investigational vaccine was supplied lyophilized in 0.7-ml, single-dose vials. Vaccine was stored at ≤ -20ºC at Merck and at -15ºC during distribution and storage at the study sites. Sterile diluent (water without preservatives) was supplied in 0.7-ml vials. It could be stored refrigerated at 2-8ºC or at room temperature. The drug product is reconstituted with sterile diluent just prior to use.

The placebo contained the same stabilizers as the investigational vaccine, but did not contain the Oka/Merck varicella virus or neomycin. The placebo was visibly distinct from the investigational vaccine.

A single 0.5 mL dose was administered subcutaneously into the deltoid area of the non-dominant arm.

4.5.2 Protocol Specific Vaccine Information

Pivotal study, Protocol 004, employed 12 clinical lots of ZOSTAVAX™ administered in a volume of 0.5mL. 3 clinical lots were stored frozen until reconstitution and administration as described above, while an additional 3 clinical lots were derived from these lots and subjected to accelerated aging by storage at 2-8ºC for 3 months to intentionally decrease the potency as measured by plaque-forming units. The remaining 6 clinical lots were derived from additional “parental” lots (2 accelerated aged lots from each of 3 parental lots) and subjected to storage at 2-8ºC for 3 months to intentionally decrease the potency as measured in plaque-forming units.

4.5.3 Dose Selection

Most of the ZOSTAVAX™ clinical lots were “aged” by storage at higher than recommended temperature (2-8ºC) to evaluate the safety and efficacy at the anticipated expiry, approximately 20,000 plaque-forming units (pfu)/0.5 mL dose.

Table 4-1 ZOSTAVAX™ Lots Used in Protocol 004

“Parental” Lot*	Clinical Lot Number	Initial (Release) Potency PFU/0.5 mL Dose	Aged

Not applicable	1535W-E046	48,911	Not aged
Not applicable	1536W-E047	57,092	Not aged
Not applicable	1537W-E048	46,291	Not aged
A	1553W-E462	33,000	Aged
B	1554W-E463	39,600	Aged
C	1555W-E464	32,700	Aged
D	1562W-E471	27,300	Aged
E	1563W-E472	27,150	Aged
F	1564W-E473	33,500	Aged
D	1588W-G479	22,000	Aged
E	1589W-G480	26,850	Aged
F	1590W-G481	25,500	Aged
*For ease of review this table arbitrarily assigns parental lots a unique identifier.

(Source BLA 125123, Protocol 004 Table 5-1)

4.6 Other Products Used in Protocol 004

Antiviral Therapy

Famciclovir (Smith Kline Beecham Pharmaceuticals, UK, or Novartis Pharmaceuticals, USA)

500mg every 8 hours for 7 days

Subjects evaluated within 72 hours of onset of rash due to suspected herpes zoster were offered famciclovir. If subject were first seen beyond 72 hours after rash onset, famciclovir could be offered at the physician’s discretion.

Analgesia

Pain was managed at the discretion of the study site investigator and could include acetaminophen, non-steroidal anti-inflammatories, opiates, and topical anesthetics.

5.0 ENDPOINTS

5.1 Primary and Secondary Endpoints

Table 5-1 Summary of Primary and Secondary Endpoint Results and Analyses

		Endpoint/Analysis	Success Criteria
Co-Primary	1	Herpes Zoster Burden of Illness (HZ BOI) subjects ³ 60 years of age VE_BOI = HZ BOI_PLACEBO - HZ BOI_VACCINE	Point estimate > 47% LL 95% CI > 25%
Co-Primary	2	Incidence of Post-herpetic neuralgia (PHN) subjects ³60 years of age VE_PHN = PHN_PLACEBO– PHN_VACCINE/ PHN_PLACEBO	Point estimate > 62% LL 95% CI > 25%
Secondary*	1	Incidence of HZ in subjects ³60 years of age VE_HZ = HZ_PLACEBO – HZ_VACCINE / HZ_PLACEBO	LL 95% CI > 25%
	2	Duration of clinically significant pain vaccine and placebo, in subjects ³60 years of age Log-rank test	p-Value < 0.001
	3	Substantial Interference with Activities of Daily Living (SADLI)⁴ Above and Beyond VE_HZ subjects ³60 years of age SADLI = 1– Relative Risk SADLI / Relative Risk HZ	p-Value LL 95% CI > 0
LL = lower limit * Originally designated as tertiary endpoints, but during the study designated secondary endpoints.

(Source: STN 125123 Protocol 004 Section 7.1.2.2)

5.1.1 Co-primary Endpoints Defined

· Herpes Zoster Burden of Illness Score (HZ BOI)

A composite endpoint measuring incidence, severity and duration of pain:

Proportion of all subjects within treatment group who developed HZ

Mean severity-by-duration score of subjects who developed HZ

(weekly worst pain score multiplied by 7 days; scored on a 0-10 scale, 10 = worst pain)

Note: All pain scores reported for Days 0-30 after onset of HZ rash were included, but only pain scores ≥ 3 were included after Day 30 following onset of HZ rash.

· Incidence of Postherpetic Neuralgia (PHN) in each treatment group

HZ-associated pain scored ≥ 3 persisting or appearing at least 90 days after onset of HA rash. (Scored on 0-10 scale, 10 = worst pain). The cutoff for the time defining PHN was revised during the study from 30 days to 90 days.

5.1.2 Secondary Endpoints Defined

NOTE: During the course of the study these endpoints were “elevated” from tertiary to secondary endpoints

· Incidence of Herpes Zoster (HZ)

Number of evaluable HZ cases per 1000 person-years of follow-up per treatment group

· Duration of Clinically Significant HZ Pain

Number of days between the first day after rash onset when the subject had a “worst pain” score of ≥ 3 (as reported on either IZIG or ZBPI) and the first visit when the “worst pain” score reported on ZBPI was reported as <3 and remained <3 for the remainder of the follow-up period (up to 6 months after HZ rash onset) per treatment group.

· Substantial Activities of Daily Living Interference (ADLI or SADLI)

Combined ADLI score ≥ 2 for ≥ 7 days, beyond reduction in HZ in the 6-month period following HZ rash onset per treatment group.

Average of scores for seven pain interference items used a 0-10 scale, with 0 being no interference and 10 being maximum interference:

General activity
Mood
Walking ability
Normal work
Relations with others
Sleep
Enjoyment of life

5.1.3 Selected Tertiary Endpoints

· PHN Using Alternative Definitions

Pain that persists or appears more than 30, 60, 120 and 182 days after rash onset

· Efficacy and Immunogenicity of 3 Consistency Lots of the Vaccine

· Effect of a Reduction in Plaque-Forming Unit (pfu) Content Over Time on Vaccine Efficacy

· VZV-Specific Immune Responses

At baseline, Week 6, and Months 12, 24 and 26 post-vaccination

· Impact of HZ and Its Complications on Quality of Life of Older Persons

· Natural History of HZ in Older Persons

Most subjects were treated with an antiviral drug (acyclovir, famciclovir or valacyclovir)

· Safety Profile of Zoster Vaccine in Persons ≥ 60 Years Old

· Evaluation of Potential Covariates

· Age, gender, and antiviral and analgesic medication usage

6.0 LABORATORY METHODS, SURVEILLANCE AND MONITORED PARAMETERS

6.1 Laboratory Methods

Evaluation of Suspected HZ

Lesion specimens collected from all suspected HZ cases at 1 of 2 central laboratories were forwarded to Central Laboratory PCR for VZV & HSV. PCR testing utilized PCR primers to distinguish between HSV and VZV infections. Culture of VZV was performed when local laboratory facilities were available, thus, culture was not attempted at all study sites.

Blood Samples

Cell Mediated Immunity (CMI) Substudy: The ZOSTAVAX™/VZV immune response was evaluated in approximately 1200 subjects (~600 each from the San Diego and Denver study sites; ~ 600 subjects between 60 and 69 years old and ~ 600 subjects ≥ 70 years old). Enrollment and blood sampling began approximately 1 year after initiation of Protocol 004. Whole blood samples for VZV IFN-γ ELISPOT and RCF assays and serum samples for VZV-specific gpELISA antibody assays were collected just prior to vaccination, and at Week 6 and at months 12 ,24 and 36 postvaccination.

Naturally occurring HZ: Blood samples were obtained at Day 1, week 3 and week 6 following HZ rash onset.

Assay Methods

VZV PCR

DNA was extracted from specimen using ------------------------------------------------------------------------------------------ PCR utilizing virus-specific primers and fluorescent probes to detect and discriminate among: Wild-type/Oka-parent VZV (VZV-WT/VZV-P); Oka-type attenuated VZV (VAV-O), the vaccine strain; and HSV DNA (HSV types 1 and 2).

VZV Responder Cell Frequency (RCF)

Lymphocyte proliferation was measured by limiting dilution assay after a 10-day incubation of PBMC + VZV antigen, followed by 3H-thymidine pulse (6 hrs.). Counts per minute (CPM) / well determined with positive well indicated ≥ 1 responding cell. Results were calculated as number of VZV-specific responder cells /100,000 PBMCs.

VZV IFN-γ ELISPOT

PBMCs were stimulated by exposure to VZV antigen (prepared from a stock of UV-inactivated Oka/Merck vaccine virus). A mouse monoclonal antibody was used to coat the culture plates and bind IFN-γ produced upon incubation of PBMCs with VZV antigen. After PBMCs were washed away a second biotinylated monoclonal antibody was used to bind IFN-γ remaining on the plate (bound to the first antibody). Alkaline phosphatase-streptavidin binds the biotinylated antibody and reacts with a chromogenic substrate (NBT/BCIP) and the resulting dark blue spots are then counted as a measure of the number of cells that produced IFN-γ in response to VZV antigen. Results reported as frequency of spot forming cells (SFCs) / 10⁶ PBMCs. If results were negative, zero or <2, result reported as <0.5.

gpELISA

This method detects antibodies to VZV glycoprotein (gp), which have been purified from MRC-5 cells infected with the KMcC strain of VZV by lectin affinity chromatography. Serum sample titers as determined by gpELISA are shown to correlate with neutralizing antibody titers.

VZV glycoprotein (gp) antigen from VZV-infected MRC-5 cells or from uninfected MRC-5 cells (Tissue Culture Controls or TCC) was adsorbed to polystyrene microtiter wells and used as the solid phase antigen. Experimental, control, and standard curve sera were incubated in the VZV gp-coated and TCC-coated wells (2 wells for each antigen). For each serum sample, a delta optical density (DOD) is calculated as the difference between the average optical density (OD) of the 2 VZV antigen wells and the average OD of the 2 TCC wells. Quantitation was obtained by comparison of sample DOD with a standard curve. Results for the assay were reported as concentration of antibody in gpELISA units/mL.

The negative control used for this assay was an individual human serum at a dilution of 1:50, found to be negative for anti-VZV. The high-positive control was a VZV antibody-positive serum, diluted 1:500, which gave a response in the assay at the upper end of the standard curve. The low positive control was a VZV antibody-positive serum diluted 1:50, which gave a response in the assay at the lower end of the standard curve. A VZV antibody-positive individual human serum was used to generate a standard

curve.

6.2 Safety Surveillance and Monitoring

· Automated Telephone Reporting System (ATRS)

At approximately 42 days postvaccination, all subjects were instructed to call a toll-free ATRS number to report safety-related information. Subjects were asked a serious of programmed questions regarding occurrences of rash, unusual reactions, hospitalizations, disability, life-threatening events, new diagnosis of cancer, overdose of any medication.

Subjects were also instructed to call the ATRS on a designated day each month to answer a series of pre-programmed questions to ascertain whether or not the subject had, at that time or during the previous 30 days, signs and/or symptoms compatible with HZ.

If a subject failed to call the ATRS when scheduled, the ATRS initiated 4 telephone calls to the subject over 96 hours. If these telephone calls failed to establish contact, the ATRS faxed a notification of failed contact to the local Study Coordinator.

If any response was suggestive of HZ, the subject was instructed by the ATRS to immediately telephone his/her local Study Coordinator to be evaluated. Also, ATRS faxed a copy of the subject's responses to the local Study Coordinator, notifying the coordinator to immediately establish contact with the subject.

A voice message center allowed subjects to leave detailed messages for local Study Coordinator.

· Adverse Event Monitoring Substudy / Vaccine Report Cards (VRCs)

The Adverse Event (AE) Monitoring Substudy planned enrollment of ~6000 subjects, to include ~300 subjects from each of the participating study sites. Once each site had established routine enrollment into the main study, subjects were to be consecutively enrolled into the AE Monitoring Substudy at the time they were enrolled. Vaccine Report Cards (VRCs) specifically queried for specific solicited adverse events (AEs) as follows:

Days 0-4 postvaccination: Swelling, redness, pain, tenderness

Days 0-21 postvaccination: Temperature

Days 22-42 postvaccination: Temperature if felt by subject to be abnormal; unsolicited AEs

Days 0-42 postvaccination: Rash looking like chicken pox or shingles, other complaints or illnesses

Site personnel reviewed available medical records to identify any additional AEs occurring Days 0-42. Hospitalizations occurring after the 42-day postvaccination period were tracked using monthly subject reporting to the Automated Telephone Response System.

Subjects enrolled in the AE Monitoring Substudy were excluded from enrollment in the CMI Substudy.

· Routine Safety Monitoring Cohort

The cohort includes all subjects not enrolled in the Adverse Event Monitoring Substudy. Days 0-42 subjects were monitored as described in the ATRS monitoring section and after Day 42 safety follow-up data was passively collected. Site personnel were expected to review available subjects’ medical records to capture data relevant to adverse experiences or potential case of HZ.

· Serious Adverse Events (SAEs) and Deaths

Follow-up of a reported SAE was the same for the AE Monitoring Substudy and the Routine Safety Monitoring Cohort. Narratives are provided for SAEs occurring on study.

Narratives of serious adverse experiences with onset dates between Days 0 and 42 postvaccination that led to death are submitted in the study report.

Narratives are not provided for all deaths reported in the entire study period. Only the incidence of death was monitored throughout the entire study and used to compare the mortality rates between the 2 vaccination groups.

· Evaluation of Suspected Herpes Zoster

According to the protocol serial evaluation of by study personnel were to begin within 24 hours of first reported symptoms or as soon as possible thereafter and continuing over ≥ 182 days according to a protocol-specific schedule. Pain was evaluated using worst pain score (0-10 point scale, 10 = worst pain) on (time since rash onset) Days 1-3; 4 or 5; 6, 7, or 8; 9, 10 or 11; weekly on Weeks 2-8 and afterwards through Day 182 or until worst pain score has been below 3 for at least two consecutive weeks.

All pain scores reported for Days 0-30 were included, but only pain scores ≥3 were included after Day 30 in the calculation of BOI scores.

Pain scores of ≥ 3 are assumed to correlate with interference in activities of daily living.

Pain scores were recorded after first onset of rash and prior to the first visit using the Initial Zoster Impact Questionnaire (IZIQ) and at subsequent timepoints using the Zoster Brief Pain Inventory (ZBPI). The Brief Pain Inventory was originally designed to measure cancer pain (Cleeland 1994) and its evaluation of “worst pain in the last 24 hours” is reported to be a valid and responsive measure of HZ-related pain, correlating with changes in responses to health status questions as reported in results from a validation study conducted in 121 HZ subjects aged 60 years and older enrolled within 14 days of rash onset (Coplan 2004).

· Clinical Evaluation Committee (CEC)

The 5-member Clinical Evaluation Committee (CEC), blinded to treatment, clinical and laboratory information, evaluated all suspected HZ cases according to a Standard Operating Procedure (SOP). Blinded information including digital photos of rashes were evaluated individually by each CEC member. These results were used in determining evaluable cases of HZ when PCR or culture confirmation were not available. In addition, a comparison was made between those cases determined by clinical laboratory testing (PCR or culture) and the corresponding determination for the same cases made by CEC.

A hierarchical determination of HZ considered the results of PCR first, culture results next for those without PCR determination and CEC determination for those without either PCR or culture determination for each suspected HZ case. Suspected rashes determined not to be caused by HZ or considered indeterminable were classified as “non-evaluable.” Non-unanimous cases were reviewed at subsequent CEC meetings. Cases that remained “Indeterminate” after a second CEC meeting were classified as “Not a clinically diagnosed case of HZ”

· Data Safety Monitoring Board (DSMB)

A DSMB was responsible for reviewing the progress of the study, monitoring subject intake, outcomes, possible adverse experiences after vaccination, and various ethical issues and making recommendations as to whether the study should continue or terminate.

DSMB Interim Monitoring

Interim reports were provided to the DSMB by a designated unblinded biostatistician at the Veterans Affairs Cooperative Studies Program Coordination Center, who was not otherwise associated with this program.

Approximately two weeks prior to each of their meetings, the DSMB members were to receive an interim monitoring report including all analyses previously requested. Any unblinded tables that have been prepared by the third-party unblinded biostatistician used “X” or “Y” for treatment group labels. The actual identity of treatments was not revealed to the DSMB unless requested by the DSMB. No individuals other than the third-party unblinded biostatistician and the DSMB were allowed to see or know the content of the unblinded tables.

Interim Safety Monitoring:

The DSMB reviewed adverse experience data periodically and in the event that severe adverse reactions or increased incidence of HZ are noted to be excessive in the vaccine arm relative to the placebo arm, the DSMB may consider stopping recruitment into the study.

7.0 CASE DEFINITIONS

7.1 Suspected Herpes Zoster

A rash with unilateral dermatomal distribution and at least one of the following: vesicles in the area of the rash and/or pain in the area of the rash.

7.2 Evaluable Herpes Zoster (confirmed case)

Diagnosis of HZ was based upon a hierarchical approach: suspected cases were to be confirmed as VZV positive or negative by PCR whenever possible; suspected cases that could not be confirmed as positive or negative by PCR could be identified by a positive VZV viral culture; and lastly, cases that could not be confirmed as positive or negative by PCR or positive by viral culture were to be determined by a Clinical Evaluation Committee (CEC) which evaluated all suspected cases even those confirmed by PCR testing or viral culture. (CEC HZ determinations and hierarchical testing determinations of HZ cases were later compared in the efficacy analyses.)

7.3 Post-herpetic Neuralgia (PHN)

Using a pain scale of 0-10 with 0 = no pain & 10 = worst pain, PHN was defined as pain score ≥ 3 persisting or appearing more than 90 days after HZ rash onset (revised during the study from originally definition using a 30-day cutoff).

8.0 STATISTICAL CONSIDERATIONS

Loss to follow-up

Loss-to-follow-up was considered to have little impact on the planned power of the efficacy analyses. If a subject was lost to active follow-up during the trial but later contacted and no case of HZ had been missed, no information was considered lost with regard to the efficacy analyses.

All randomized subjects who were lost to follow-up before developing HZ during the study contributed follow-up information to the efficacy analyses until the date of the last completed contact. The analyses treated these subjects similar to those subjects who remained in contact with the study site throughout the study and did not develop HZ before the trial was terminated.

Although protocol-specified, the date of last contact is blank in the termination datasets for 99% of subjects not identified as deceased, lost to follow-up or dropouts.

8.1 Randomization and Stratification

Subjects were randomized in a 1:1 ratio to receive vacccine or placebo. Eligible subjects were sequentially assigned an allocation number (AN) in numerical order from the allocation schedules provided by the Cooperative Studies Program Coordination Center.

Randomization was stratified by site and by age group: 60-69 years and ≥ 70 years. Two (2) separate allocation schedules were generated; one for individuals 60 to 69 years of age and one for individuals ≥ 70 years of age. Randomization numbers were assigned sequentially within each age stratum at each participating site as subjects were enrolled. Target enrollment in each treatment group (vaccine or placebo) was ~11,160 in the 60-69 year age stratum and ~7440 in the ≥70 year age stratum.

8.2 Blinding

Placebo and vaccine were visually distinct. Therefore, an independent third party (vaccine technician) was responsible for labeling syringes and reconstituting and administering the vaccine/placebo. This unblinded person had no subsequent role in the assessment of subjects and did not maintain any separate record of study vaccine or placebo assignments. Subjects, site investigators and site personnel, and Veterans Affairs Cooperative Studies Program Coordination Center (CSPCC), Veterans Affairs Cooperative Studies Program Clinical Research Pharmacy Coordinating Center (CSPCRPCC) and Merck & Co., Inc. personnel, were all blinded.

8.3 Unblinding

A preliminary analysis of data from the CMI Substudy obtained at Day 0 and Day 42 postvaccination was performed. No clinical efficacy data were included in this preliminary analysis, which was performed by an independent, unblinded Veterans Affairs Cooperative Studies Program (CSP) Biostatistician who was not involved in the day-to-day operation of the study. The results were summarized by group and no individual identification of the subjects was disclosed. These results were made available to the management of the Clinical, Biostatistics, Regulatory, and Manufacturing departments at Merck & Co., Inc., and the Director of the West Haven Veterans Affairs Cooperative Studies Program Coordination Center (VA CSPCC). The DSMB was also provided with a copy of the preliminary analysis report. The conduct of the study was unaffected by this analysis.

8.4 Sample Size

The study used a conservative estimate of HZ incidence of 3 / 1000 person-years in individuals aged ≥ 60 years, and estimated 10% of subjects lost to follow-up annually, a total enrollment of 37,500 subjects randomized 1:1 to vaccine or placebo with approximately 4.5 years of follow-up per subject was planned. Fewer subjects aged ≥ 70 years were planned to be enrolled, as the incidence rates of HZ and PHN are relatively higher in this older age group.

8.5 Interim Analyses

Two interim analyses of efficacy endpoints were planned but later cancelled. No formal interim analyses were performed, however sample size re-estimation and other analyses were conducted.

8.6 Population for Analyses

The primary efficacy analysis population included all subjects ≥ 60 years old. This modified intent-to-treat (MITT) population excluded subjects with less than 30 days of follow up and those who developed HZ in the first 30 days following vaccination.

9.0 RESULTS

Study period: 06-Nov-1998 to 30-Apr-2004

Last HZ case accrued: 30-Sep-2003

Last subject terminated: 28-Apr-2004 (except one subject with termination date 20-Sep-2004)

9.1 Population Enrolled / Disposition

Age distribution in the treatment groups for MITT population, and CMI and AE Monitoring Substudies were all similar to that in the overall study population.

While a higher proportion of males (59%) was enrolled the study overall, a lower proportion of males were enrolled in the CMI and AE Monitoring Substudies (55.5% and 55.6% respectively).

Although the proportion of males in the CMI Substudy was 55.5%, a lower proportion was randomized to the zoster vaccine group within the CMI Substudy (50.4%).

Racial characteristics of the study populations were largely similar except that a lower proportion of Black subjects were enrolled in the CMI Substudy compared to the overall study (0.5% vs. 2.1%). The fact that 95% of all study participants were White makes it difficult to draw conclusions as to any differences in safety or efficacy based upon race.

The baseline characteristics of marital status, education, work status, health status by EuroQol, and functional/activity measures were similar in the AE Monitoring Substudy and the overall study population.

All baseline characteristics in the zoster vaccine and in the placebo groups were similar within each of the study subpopulations (MITT, CMI Substudy and AE Monitoring Substudy populations).

Table 9-1 Subject Disposition

Protocol 004	Zoster Vaccine (N=19270)		Placebo (N=19276)		Total (N=38546)
Protocol 004	n	(%)	n	(%)	n		(%)
Vaccinated	19270	( 100)	19276	( 100)	38546		( 100)
Completed	18359	(95.3)	18357	(95.2)	36716		(95.3)
Discontinued	911	( 4.7)	919	( 4.8)	1830		( 4.7)
Reasons for discontinuations:
Died	793	( 4.1)	792	( 4.1)	1585		( 4.1)
Withdrawn from the study	57	( 0.3)	75	( 0.4)	132		( 0.3)
Lost to follow-up	53	( 0.3)	40	( 0.2)	93		( 0.2)
Other†	8	( 0.0)	12	( 0.1)	20		( 0.1)
Reasons for withdrawal or lost to follow-up:
All	118	( 0.6)	127	( 0.7)	245	( 0.6)
Changed mind about being in study	31	( 0.2)	40	( 0.2)	71	( 0.2)
Moved away	13	( 0.1)	10	( 0.1)	23	( 0.1)
Discontinued due to adverse experience	11	( 0.1)	18	( 0.1)	29	( 0.1)
Other‡	62	( 0.3)	59	( 0.3)	121	( 0.3)
Missing	1	( 0.0)	0	( 0.0)	1	( 0.0)
† Reported "Other" as reason for study termination in Case Report Form. ‡ Reported "Other" as reason for withdrawal, or "Lost to follow-up" in Case Report Form. Completed = Completed end-of-study interview N = # subjects randomized n = # subjects in respective category.

(Source: STN 125123; Protocol 004, Tables 6-1, 6-2, 6-3)

Subject disposition in CMI Substudy and AE Monitoring Substudy were similar to that in Protocol 004.

9.2 Study Follow-Up

Herpes Zoster Case Follow-up

Subjects in the MITT population were followed for an average of 3.09 years (median: 3.10 years; range: 31 days to 4.90 years) postvaccination for the development of suspect HZ. According to the protocol, each suspected case of HZ was to be followed for HZ pain for 6 months after rash onset. Table 9-4 summarizes the length of follow-up for evaluable HZ pain by vaccination group. As shown in this table, more than 96% of evaluable HZ cases had follow-up (at least for 175 days). Among those subjects followed for less than 175 days, the majority reported a worst pain score ≤1 and rash healed at their last visit.

Table 9-2 Follow-Up of Evaluable HZ Pain After Rash Onset in MITT Population

	Placebo	Zoster	Total
# Evaluable HZ cases	642	315	957
Followed ≥ 182 days	586 (91.3 %)	287 (91.1%)	873
Followed ≥175, <182 days	38 (5.9%)	13 (4.1%)	51
Followed < 175 days	18 (2.8%)	15 (4.8% )	33
No Pain follow up	13 (2.0%)	8 (2.5%)	21

(Source: (STN 125123; Protocol 004 Table 11-10)

Clinical lots were introduced into the study in a dose de-escalation fashion, and equal numbers of subjects received each of the three clinical lots within each group. The administered dose and time of follow-up were both a function of the date the subject enrolled in the study. All immunogenicity data were obtained from Groups 3 and 4.

Table 9-3 Follow-Up by Vaccine Potency (ZOSTAVAX™ recipients)

	Group 1	Group 2	Group 3	Group 4
Vaccine recipients (N)	835	978	8720	8737
Dose (pfu/ 0.5ml dose)	50,000-62,000	34,000-42,000	26,000-33,000	21,000-26,000
Dates administered	11/98 - 11/99	04/99 – 11/99	07/99 – 12/00	07/00 – 09/01
Approx. Avg. F/U (days)	1400	1400	1200	900
*Clinical Lot denoted by last 3 digits of Clinical Lot number. See Table 4-1 for further information on Clinical Lots. ^#Accelerated aged lot groups

(Source: STN 125123, Protocol 004 DEMO datasets)

9.3 Determination of HZ Cases

The following table provides information on the proportion of HZ cases confirmed by method of ascertainment. Of suspected cases of HZ, 88.4% tested positive by PCR. Of suspected cases determined evaluable cases of HZ, 93.4% tested positive by PCR. In all cases in which VZV DNA was detected by PCR, the strain identified was wild-type, i.e., no Merck/Oka vaccine strain-induced HZ was found in this study.

9.4.1 Primary and Secondary Endpoints

The results of the primary and secondary analyses are presented in Table 9-6. Burden of Illness (BOI), incidence of PHN, and incidence of HZ met the predefined criteria for success. The duration of clinically significant pain (20 days in the ZOSTAVAX™ group and 22 days in the placebo group) met the predefined statistical criteria for success (p <0.05) however the clinical significance is unclear. The SIADL analysis did not meet the pre-defined criteria for success.

Table 9-6 Summary of Primary and Secondary Endpoint Results and Analyses

		Endpoint	Point Estimate (95% CI)	Additional related analyses and comments
Co-Primary	1	Herpes Zoster Burden of Illness (BOI) VE_BOI = HZ BOI_PLACEBO - HZ BOI_VACCINE	61.1% (51.1, 69.1)	Based upon pain scale (0-10, 10 = worst pain) Stratified by age group Using total follow-up times as weights All scores (0-10) Days 0-30; Scores ≥3 after Day 30 Treatment-by-age interaction: p=0.266
Co-Primary	2	Incidence of Post-herpetic neuralgia (PHN) VE_PHN = PHN_PLACEBO– PHN_VACCINE/ PHN_PLACEBO	66.5% (47.5, 79.2)	27 PHN cases in vaccine group 80 PHN cases in placebo group Treatment-by-age interaction: p-value > 0.999
Secondary	1	Incidence of Herpes Zoster VE_HZ = HZ_PLACEBO – HZ_VACCINE / HZ_PLACEBO	51.3% (44.2, 57.6)	315 evaluable¹ cases HZ in vaccine group 642 evaluable¹ cases HZ in placebo group VE_HZ in 60-69 yr. old strata = 63.9% VE_HZ in ≥ 70 yr. old strata = 37.6%
	2	Duration of clinically significant pain³		20 days in zoster vaccine group 22 days in placebo group p-value <0.001 (MITT) p-value = 0.041 (Evaluable HZ)
	3	Substantial Interference with Activities of Daily Living⁴ Above and Beyond VE_HZ SADLI = 1– Relative Risk SADLI / Relative Risk HZ		p-value = 0.341 Does not include vaccine effect on HZ incidence, unlike the other major efficacy endpoints Treatment-by-age interaction: p-value = 0.696
¹Not included in efficacy analyses: Evaluable HZ cases w/in 30 days post-vaccination (6 cases in vaccine recipients vs. 18 cases in placebo groups) ³ Clinically significant pain: Worst daily pain ≥3 after HZ rash onset until worst daily pain <3 for remainder of follow-up, up to 6 months following HZ rash onset. This was set at 0 for subjects who did not develop HZ. ⁴ ADLI score ≥ 2 (1-10 pt.-scale, 10 = max. score) for ≥ 7 days during 6-mo. follow-up after HZ rash onset: General Activity, Mood, Walking ability, Normal work, Relations w/ others, Sleep and Enjoyment of life categories.

(Source: STN 125123; Protocol 004 Tables 7-4, 7-5, 7-9; Section 7.1)

The following table shows the effect of ZOSTAVAX™ as compared to placebo on incidence of HZ, incidence of PHN and BOI during the 5 hears postvaccination. These analyses were not prespecified.

Table 9-7 Durability of ZOSTAVAX™ Effect on Major Efficacy Endpoints (MITT)

Years On Study	Zoster Vaccine (N = 19270)						Placebo (N = 19276)								Vaccine Efficacy: (95% CI)†
Annual Incidence of Evaluable HZ Cases¹
	n		m	Follow-Up Time (Person- Years)		HZ Incidence (Per 1000 Person-Years)	n	m	Follow-Up Time (Person- Years)			HZ Incidence (Per 1000 Person-Years)				HZ Incidence
1	76		19254	19132		3.972	201	19247	19081			10.534				0.623 (0.507, 0.714)
2	103		18994	18827		5.471	194	18915	18679			10.386				0.473 (0.328, 0.589)
3	98		18626	14505		6.756	171	18422	14327			11.936				0.434 (0.270, 0.563)
4	35		9942	5412		6.467	70	9806	5325			13.145				0.508 (0.252, 0.682)
5	3		1906	327		9.183	6	1856	324			18.500				0.504 (-1.324, 0.920)
Overall	315		19254	58203		5.412	642	19247	57736			11.120				0.513 (0.442, 0.576)
Annual Incidence of PHN¹
	n	m			Follow-Up Time (Person- Years)	HZ Incidence (Per 1000 Person-Years)	n	m			Follow-Up Time (Person- Years)			HZ Incidence (Per 1000 Person-Years)	PHN Incidence
1	5	19254			19132	0.261	33	19247			19081			1.729	0.849 (0.610, 0.954)
2	8	18994			18827	0.425	22	18915			18679			1.178	0.639 (0.159, 0.861)
3	10	18626			14505	0.689	17	18422			14327			1.187	0.419 (-0.344, 0.762)
4	3	9942			5412	0.554	7	9806			5325			1.315	0.578 (-0.847, 0.930)
5	1	1906			327	3.061	1	1856			324			3.083	0.007 (-76.930, 0.987)
Overall	27	19254			58203	0.464	80	19247			57736			1.384	0.665 (0.475, 0.792)
Estimated HZ BOI‡ Based on AUC Scale Over 6 Months of Follow-Up After HZ Rash Onset²§
	n	m			Total Follow-Up Time (Person- Years)	Estimated HZ Pain BOI‡	n	m		Total Follow-Up Time (Person- Years)			Estimated HZ Pain BOI‡		HZ Pain BOI
1	76	19254			19132	0.427	201	19247		19081			2.075		0.794 (0.682, 0.867)
2	103	18994			18827	0.801	194	18915		18679			1.661		0.518 (0.266, 0.683)
3	98	18626			14505	0.809	171	18422		14327			1.482		0.454 (0.193, 0.631)
4	35	9942			5412	0.367	70	9806		5325			1.007		0.635 (0.246, 0.824)
5	3	1906			327	0.094	6	1856		324			0.375		0.748 (0.190, 0.922)
Overall	315	19254			58203	2.208	642	19247		57736			5.682		0.611 (0.511 0.691)
¹Calculated: 1 - ratio of observed HZ incidence rates in zoster vaccine group & placebo group. CI based on exact conditional procedure. ² AUC in 6 mos. follow-up after HZ rash onset: (1) patient-reported data on HZ between rash onset & first patient interview collected on IZIQ; (2) excludes pain scores <3 that occur on ≥ 2 consecutive visits > 30 days after rash onset; and (3) included recurrent pain with score ≥3 beyond 30 days after HZ rash onset. ‡ Weighted average of observed HZ BOI stratified by age group (60 - 69 and ≥70 years) with weights proportional to the total follow-up time in each age group. § VE calculated as a weighted average of observed zoster vaccine efficacy stratified by age group with weights proportional to the total follow-up time in each age group. The CI is constructed based on the large sample approximation under the fixed-number-of-events design. MITT population: randomized, followed ≥ 30 days postvaccination, w/o evaluable HZ w/in days 0-30 postvaccination. N = # subjects randomized n = # evaluable HZ cases in time period m = # MITT subjects followed in time period

(Source: STN 125123, Protocol 004 Table 7-45)

Effect of Age on Vaccine Efficacy

As noted in Table 9.6, age is the most consistently and strongly associated factor in explaining vaccine response. Older subjects (≥70 years old) had lower vaccine efficacy in prevention of HZ, and higher rates of HZ in both vaccine and placebo groups as compared to the younger subjects (60-69 years old). The following tables (Table 9.8 and 9.9) show the effect of age on incidence of HZ.

Table 9-8 Incidence of evaluable HZ case (MITT population)

Age*	Placebo (N=19247)					Zoster Vaccine (N=19254)
Age*	# subjects	# subjects with HZ	Total follow-up time (yrs)	Incidence rate (1000 person yrs)	# subjects		# subjects with HZ	Total follow-up time (yrs)	Incidence rate (1000 person yrs)
59-64	5198	153	15384	9.945	5216		54	15693	3.441
65-69	5158	181	15569	11.626	5154		68	15630	4.351
70-74	4560	158	13814	11.438	4545		89	13830	6.435
75-79	2999	103	9105	11.312	3076		67	9329	7.182
80-84	1097	39	3189	12.230	1063		31	3172	9.773
85-89	210	7	605	11.570	181		5	498	10.040
90+	25	1	70	14.286	19		1	51	19.608
Total	19247	642	57736	11.120	19254		315	58203	5.412
*Age at randomization

(Source: FDA analysis of HZ datasets)

9.4.2 SELECTED TERTIARY EFFICACY ENDPOINTS

Tertiary Endpoint: PHN Using Alternative Definitions

Table 9-10 Incidence of PHN Using Alternative PHN Definition - MITT Population

PHN defined by Cutoff Day (After Rash Onset)	Zoster Vaccine (m = 19245) Follow-Up (person-yrs.) = 58203		Placebo (m = 19247) Follow-Up (person-yrs.) = 57736		Vaccine Efficacy with Respect to PHN Point Estimate (95% CI) ‡
PHN defined by Cutoff Day (After Rash Onset)	n	Incidence Rate of PHN† /1000 Person-Years	n	Incidence Rate of PHN† /1000 Person-Years
30	81	1.393	196	3.393	0.589 (0.466, 0.687)
60	45	0.774	113	1.956	0.604 (0.436, 0.726)
90	27	0.464	80	1.384	0.665 (0.475, 0.792)
120	17	0.292	54	0.934	0.687 (0.452, 0.830)
182	9	0.155	33	0.571	0.729 (0.421, 0.886)
†Weighted average of the observed incidence rate stratified by age group (60 to 69 and ≥ 70 years) with Mantel-Haenszel weights associated with the total follow-up time in each age group. ‡Calculated as 1 minus the ratio of the estimated incidence rates of PHN in the zoster vaccine and placebo groups. The CI was constructed based on the exact conditional procedure stratified by age group. PHN for the co-primary endpoint was defined as any HZ-associated pain rated ≥ 3 (on a 0 to 10 scale) persisting or appearing ≥ 90 days after the HZ rash onset. Alternative PHN definitions were pain ≥ 3 persisting or appearing more than 30, 60, 120, or 182 days after HZ rash onset. MITT: All randomized subjects who were followed ≥ 30 days postvaccination and did not develop evaluable cases of HZ (per the hierarchical algorithm specified in Protocol Amendment 6) within the first 30 days postvaccination. n = Number of PHN cases (defined as any HZ-associated pain ≥ 3 [on a 0 to 10 scale] persisting or appearing more than the respective cutoff days after the HZ rash onset) in the MITT population. m = Number of subjects in the MITT population. PHN = Postherpetic neuralgia. HZ = Herpes zoster. MITT = Modified intention-to-treat. CI = Confidence interval.

(Source: BLA 125123; 5.3.5.1.1.4 Table 7-15)

If the Co-Primary Endpoint in Protocol 004 (PHN) had not been redefined during the course of the study as reduction in the incidence of PHN occurring or persisting at Day 90, rather than at Day 30 as originally designated, following HZ rash onset, the study would have failed the pre-specified criterion for success. The pre-specified success criterion required that the point estimate for reduction in PHN incidence be > 62%. Based upon sponsor-conducted sensitivity analyses the study would have failed using either 30 or 60 days as the cutoff defining PHN persistence or occurrence.

Given that the majority of cases of PHN resolve completely within a few weeks after HZ rash onset, the use of a 90-day cutoff for evaluation of treatments for PHN appears useful. It is not clear that a 90-day cutoff is the most appropriate in a preventive study which seeks to evaluate the overall burden of illness due to PHN experienced in the study population. In the latter case it would seem that capturing the largest number of PHN cases would be informative as to the complete burden of illness due to PHN and also more sensitive to differences between treatment groups.

Tertiary Endpoint: Efficacy and Immunogenicity of 3 Consistency Lots of the Vaccine

Vaccine efficacy among 3 consistency lots was assessed by the estimation of pairwise ratios in the clinical endpoints of HZ pain BOI, the incidence of PHN, and the incidence of HZ among the 3 consistency lots using 90% CIs for lot-to-lot comparisons. No statistically significant differences were found in the analysis of these data. The immunogenicity results for each pair of consistency lots are shown in table 9-11.

Table 9-11 Fold Rise in gpELISA Titers 6 Weeks Postvaccination by Vaccine Lot

Consistency Lot	Vaccine Lot	Zoster Vaccine (N = 691)
Consistency Lot	Vaccine Lot	n	m	GMF	95% CI
1	1562W-E 471	73	78	1.7	(1.5, 2.0)
2	1563W-E 472	76	82	1.8	(1.5, 2.1)
3	1564W-E 473	93	97	1.9	(1.6, 2.3)
1	1588W-G 479	138	146	1.7	(1.5, 1.9)
2	1589W-G 480	141	147	1.6	(1.4, 1.8)
3	1590W-G 481	134	141	1.6	(1.4, 1.8)
N = Number of subjects vaccinated in the CMI Substudy. n = Number of subjects contributing to immunogenicity analysis. m= Number of subjects vaccinated in CMI Substudy in each lot category. gpELISA = Glycoprotein enzyme-linked immunosorbent assay. CMI = Cell mediated immunity. GMF = Geometric mean fold rise. CI = Confidence interval.

Source: STN 125123, Protocol 004, Table 11-118

Tertiary Endpoint: Effect of a Reduction in Plaque-Forming Unit (PFU) Content Over time on Vaccine Efficacy

The evaluation of the effects of potency on vaccine efficacy included all clinical lots. When used as a regression parameter, vaccine potency was found not significant in the analyses of each of the three efficacy endpoints (HZ incidence, PHN incidence and BOI scores). In fact the Cox Regression model indicated age as the most consistently significant explanatory variable for efficacy effect.

Comparison of efficacy by clinical lots is made difficult by the large number of clinical lots (twelve) and in many cases, the relatively small numbers of events in each clinical lot group. Small losses in efficacy measures for BOI scores and PHN incidence is seen in unaged clinical lots (046, 047, 048) as compared to the remaining aged lots, but this observation is of uncertain clinical significance.

Table 9-12 Summary of HZ BOI Based on the Protocol-Defined AUC Scale†

Over 6 Months of Follow-Up After HZ Rash Onset by Vaccine Lot (MITT Population)

Vaccine Lot	Dose Potency When Shipped (PFU/dose)	Zoster Vaccine (N=19270)
		n	m	Total Follow-Up (Person-Years)	Observed HZ BOI	95% CI




1535W-E 046	52018	6	278	1097	4.261	(2.239, 8.108)
1536W-E 047	61833	7	278	1119	6.943	(2.729, 17.660)
1537W-E 048	50063	8	279	1133	4.020	(1.697, 9.522)
1553W-E 462	37273	4	326	1277	1.712	(0.393, 7.458)
1534W-E 463	42403	8	326	1282	3.284	(2.078, 5.190)
1535W-E 464	34362	7	326	1266	2.160	(0.950, 4.910)
1562W-E 471	27633	56	2906	9764	3.170	(2.334, 4.305)
1563W-E 472	26371	63	2903	9766	2.168	(1.630, 2.882)
1564W-E 473	32588	53	2901	9745	2.658	(1.927, 3.665)
1588W-G 479	21480	35	2912	7271	1.543	(1.136, 2.095)
1589W-G 480	26218	38	2908	7238	1.954	(1.169, 3.266)
1590W-G 481	24931	30	2911	7247	2.193	(1.331, 3.614)
†Protocol-defined AUC: (1) incorporates patient-reported data on HZ between rash onset and the first patient interview collected on IZIQ; (2) excludes pain scores <3 that occur on 2 or more consecutive visits more than 30 days after rash onset; and (3) included recurrent pain with score ≥ 3 after 2 consecutive visits with worst pain scores <3 beyond 30 days after HZ rash onset. The MITT population included all subjects randomized in the study who were followed for at least 30 days postvaccination and did not develop evaluable cases of HZ (per the hierarchical algorithm specified in Protocol Amendment 6) within the first 30 days postvaccination. N = Number of subjects randomized in the vaccination group n = Number of evaluable HZ cases in the MITT population m = Number of subjects in the MITT population. HZ = Herpes zoster BOI = Burden of illness AUC = Area under the curve MITT = Modified intention-to-treat CI = Confidence interval IZIQ = Initial zoster impact questionnaire

(Source: STN 125123, Protocol 004 Table 11-71)

Table 9-13 Summary of Incidence of PHN by Vaccine Lot (MITT Population)

Vaccine Lot	Dose Potency When Shipped (PFU/dose)	Zoster Vaccine
		(N = 19270)
		n	M	Total Follow-Up (Person-Years)	Observed Incidence Rate of PHN (Per 1000 Person-Years)	95% CI
1535W-E 046	52018.5	2	278	1097	1.823	(0.221, 6.586)
1536W-E 047	61833.5	2	278	1119	1.787	(0.216, 6.457)
1537W-E 048	50063.0	0	279	1133	0.000	(0.000, 3.255)
1553W-E 462	37273.0	0	326	1277	0.000	(0.000, 2.889)
1534W-E 463	42403.5	2	326	1282	1.560	(0.189, 5.635)
1535W-E 464	34362.5	0	326	1266	0.000	(0.000, 2.914)
1562W-E 471	27633.5	5	2906	9764	0.512	(0.166, 1.195)
1563W-E 472	26371.5	2	2903	9766	0.205	(0.025, 0.740)
1564W-E 473	32588.0	4	2901	9745	0.410	(0.112, 1.051)
1588W-G 479	21480.5	3	2912	7271	0.413	(0.085, 1.206)
1589W-G 480	26218.0	3	2908	7238	0.414	(0.085, 1.211)
1590W-G 481	24931.0	4	2911	7247	0.552	(0.150, 1.413)
MITT: All randomized subjects who were followed ≥ 30 days postvaccination and did not develop evaluable cases of HZ (per the hierarchical algorithm, Protocol Amendment 6) within the first 30 days postvaccination. N = Number of subjects randomized. n = Number of PHN cases (defined as any HZ-associated pain ≥3 [on a 0 to 10 scale] persisting or appearing ≥90 days after the HZ rash onset) in the MITT population. m = Number of subjects in the MITT population HZ = Herpes zoster PHN = Postherpetic neuralgia MITT = Modified intention-to-treat CI = Confidence interval

(Source: STN 125123, Protocol 004, Table 11-72)

Table 9-14 Incidence of Evaluable HZ Cases by Vaccine Lot (MITT Population)

Vaccine Lot	Dose Potency When Shipped (PFU/dose)	Zoster Vaccine (N = 19270)
Vaccine Lot	Dose Potency When Shipped (PFU/dose)	n	m	Total Follow-Up (Person-Years)	Average Follow-Up (Days)	Observed Incidence Rate of HZ (Per 1000 Person-Years)	95% CI
1535W-E 046	52018.5	6	278	1097	1441	5.469	(2.007, 11.904)
1536W-E 047	61833.5	7	278	1119	1470	6.256	(2.515, 12.890)
1537W-E 048	50063.0	8	279	1133	1483	7.060	(3.048, 13.911)
1553W-E 462	37273.0	4	326	1277	1430	3.133	(0.854, 8.022)
1534W-E 463	42403.5	8	326	1282	1436	6.240	(2.694, 12.295)
1535W-E 464	34362.5	7	326	1266	1418	5.530	(2.223, 11.394)
1562W-E 471	27633.5	56	2906	9764	1227	5.736	(4.333, 7.448)
1563W-E 472	26371.5	63	2903	9766	1229	6.451	(4.957, 8.254)
1564W-E 473	32588.0	53	2901	9745	1227	5.439	(4.074, 7.114)
1588W-G 479	21480.5	35	2912	7271	912	4.814	(3.353, 6.695)
1589W-G 480	26218.0	38	2908	7238	909	5.250	(3.715, 7.206)
1590W-G 481	24931.0	30	2911	7247	909	4.140	(2.793, 5.910)
MITT: All randomized subjects who were followed ≥ 30 days postvaccination and did not develop evaluable cases of HZ (per the hierarchical algorithm, Protocol Amendment 6) within the first 30 days postvaccination. N = # subjects randomized n = # of evaluable HZ cases m = # subjects in the MITT population HZ = Herpes zoster MITT = Modified intention-to-treat CI = Confidence interval

(Source: STN 125123, Protocol 004, Table 11-73)

Tertiary Endpoint: VZV-Specific Immune Responses

Immunogenicity data (both humoral and cell-mediated) are from subjects enrolled in the Cell-mediated Immunity Substudy (CMI) and included results of gpELISA, Responder Cell Frequency (RCF) and IFN- γ ELISPOT assays. The gpELISA assay appears to be the most informative in terms of magnitude of response as well as differences between treatments groups.

Measures of cellular immunity, RCF and IFN-γ ELISPOT data, are limited by lower the relatively lower magnitude of response and relatively higher variability in assay results as well as the technical complexity of collecting and handling the samples. The data derived from RCF and ELISPOT at this point do not appear to add significantly to the information derived from gpELISA data and will not be discussed further.

Primary analyses are based upon a per-protocol population, excluding those with insufficient sample, sample not obtained in designated time window, high background, invalid sample, storage or handling problems and invalid assay results for the RCF and IFN- γ ELISPOT and gpELISA assays. There were no large notable differences between treatment groups in the proportion of data in each assay that was excluded from the per-protocol analysis. Antibodies to VZV were assessed 6 weeks following immunization with ZOSTAVAX™ or placebo using a gpELISA assay.

Table 9-15 gpELISA Titers in CMI Substudy Participants 6 Weeks Postvaccination

Endpoint	Zoster Vaccine		Placebo
	(N=691; n = 655-678)		(N=704; n = 673-691)
	Observed Response (95% CI)		Observed Response (95% CI)
Day 0 (gpELISA units/mL)
	n = 678		n = 691
% ≥200	60.5%	(56.7%, 64.2%)	64.4%	(60.7%, 68.0%)
GMT	278.8	(258.0, 301.4)	291.0	(269.7, 314.0)
6 Weeks Postvaccination (gpELISA units/mL)
	n = 667		n = 684
% ≥200	83.4%	(80.3%, 86.1%)	64.6%	(60.9%, 68.2%)
GMT	474.7	(441.5, 510.5)	291.4	(269.3, 315.3)
6 Weeks Postvaccination (Fold Rises from Day 0)
	n = 655		n = 673
% ≥2fold	35.7%	(32.1%, 39.5%)	7.1%	(5.3%, 9.3%)
% ≥3fold	16.9%	(14.2%, 20.0%)	2.5%	(1.5%, 4.0%)
% ≥4fold	10.7%	(8.4%, 13.3%)	1.0%	(0.4%, 2.1%)
% ≥5fold	7.0%	(5.2%, 9.3%)	0.4%	(0.1%, 1.3%)
Geometric Mean Fold Rise	1.7	(1.6, 1.8)	1.0	(1.0, 1.0)
90 Percentile of Fold Rise	4.2		1.7†
95 Percentile of Fold Rise	6.4		2.4‡
†Percent of subjects in vaccine group with a fold rise above this cutoff was 44.3% [95% CI= (40.4%, 48.2%)] ‡Percent of subjects in vaccine group w/ fold rise above this cutoff was 27.9% [95% CIs: 24.5%, 31.5%] N = # subjects vaccinated in CMI substudy n = # subjects contributing to immunogenicity analysis. gpELISA - Glycoprotein enzyme-linked immunosorbent assay GMT - Geometric mean titer CI - Confidence interval.

(Source: STN 125123, Protocol 004 Table 7-64)

The persistence of gpELISA titers from Day 0 to 3 years postvaccination is shown in the following table.

Table 9-16 Persistence of gpELISA Titers Among the CMI Substudy

		Zoster Vaccine			Placebo
		(N=691)			(N=704)
			Observed			Observed
Endpoint	Time Point	n	Responses	95% CI	n	Responses	95% CI
GMT	Day 0	678	278.8	258.0, 301.4	691	291.0	269.7, 314.0
	6 Weeks	667	474.7	441.5, 510.5	684	291.4	269.3, 315.3
	12 Months	649	353.7	328.1, 381.2	661	306.6	283.3, 331.9
	24 Months	636	329.5	304.5, 356.5	644	300.6	277.8, 325.3
	36 Months	625	331.6	305.1, 360.4	612	305.7	280.6, 333.2
Geometric	6 Weeks	655	1.7	1.6, 1.8	673	1.0	1.0, 1.0
Mean	12 Months	636	1.3	1.2, 1.3	650	1.1	1.0, 1.1
Fold Rises	24 Months	624	1.2	1.1, 1.2	633	1.1	1.0, 1.1
from Day 0	36 Months	612	1.2	1.1, 1.3	601	1.0	1.0, 1.1
N = Number of subjects vaccinated in the CMI Substudy. n = Number of subjects contributing to the immunogenicity analysis. gpELISA = Glycoprotein enzyme-linked immunosorbent assay. CMI = Cell mediated immunity. GMT = Geometric mean titer. CI = Confidence interval.

(Source: STN 125123, Protocol 004, Table 7-71)

Prevaccination titers, age and timing of the 6-week blood sample had statistically significant effects on the gpELISA responses at 6 weeks postvaccination.

Gender, study site and the vaccine potency subjects received did not have statistically significant effects.

The effect of age was more obvious in the lowest prevaccination titer category: the fold rises were higher in the younger age group than in the older age group when prevaccination titer was <100, but more similar when prevaccination titers were ≥ 100.

Overall, zoster vaccine recipients with relatively lower gpELISA titers at Day 0 appeared to have relatively lower titers but higher fold rises at 6 weeks postvaccination.

In the zoster vaccine group, the fold rises at 6 weeks postvaccination appeared to be comparable between males and females. In the placebo group, no increase of fold rises was observed in any of the subgroups.

Zoster vaccine recipients had comparable gpELISA titers at 6 weeks postvaccination, regardless of the vaccine lot they received.

Tertiary Endpoint: Immune Responses as Correlates of Protection

Subjects who did not develop HZ had higher gpELISA titers at 6 weeks postvaccination compared with the subjects who developed HZ. This difference held true for both the placebo as well as the zoster vaccine group, although the number of subjects who developed HZ was very small in the zoster vaccine group.

Table 9-17 Immune Responses Among CMI Substudy Participants by HZ Incidence Status

Endpoint

Zoster Vaccine

Placebo

Subject Cohort

(N=691)

(N=704)

Protocol #	001	002	003	004	005	007	009
Number of subjects	276	Dose 1: 398 Dose 2: 206	21	38,546	196	210	698
Number of ZOSTAVAX recipients	241	398	18	19,270	196	210	698
Population	Healthy adults Seropositive	Adults Healthy, DM or COPD Hx of varicella	Healthy adults Central & S. America & Philippines Low (≤5 gpELISA units/mL) or undetectable varicella-zoster virus antibody titer	Healthy Adults ≥ 60 yrs. old	Healthy adults Previous 1-2 doses zoster vaccine History of varicella	Healthy adults ≥ 60 yrs.	Healthy adults ≥ 50 yrs.
Strata Demographics	60-75 yrs: N = 144 ≥ 76 yrs: N = 132	60-75 yrs: ≥ 76 yrs	≥ 30 yrs: N = 21	60-69 yrs: ≥ 70 yrs:	61-89 yrs	None	50-59 yrs: ≥ 60 yrs:
# Doses	1	2	1	1	1	2	1
Schedule	0	0 & 18 mos.	0	0	0	0 & 42 days	0
Dose levels (pfu/dose)	Placebo 2,000 17,000 (aged) 19,000 34,000 67,000	Dose 1: Placebo 34,000 50,000 Dose 2: 50,000	50,000	Placebo 22,000 - 63,000 includes aged lots	50,000	25,550	58,000 207,000
Postvaccination Follow up	42 day safety, immunogenicity	42 day safety, immunogenicity	42 day safety, immunogenicity	42 day safety HZ ~2-4.5 yrs Immune ≤3 yrs.	42 day safety; 42 day-2 year immunogenicity	42 day safety, immunogenicity	42 day safety

Criteria					Zoster Vaccine				Placebo
					(N = 19270)				(N = 19276)
					(M = 467)				(M = 799)
PCR	Virus Culture		Clinical Adjudication		Evaluable		Non-Evaluable		Evaluable	Non-Evaluable
					n (%)		n (%)		n (%)	n (%)
All suspected HZ cases occurring ≥ 30 days postvaccination (MITT definition)
Total				316 ( 67.7)		151 ( 32.3)		644 ( 80.6)			155 ( 19.4)
VZV-positive†	Not considered	Not considered		295 ( 63.2)		0 ( 0.0)		602 ( 75.3)			0 ( 0.0)
VZV-negative†	Not considered	Not considered		0 ( 0.0)		88 ( 18.8)		0 ( 0.0)			91 ( 11.4)
HSV-positive‡	Not considered	Not considered		0 ( 0.0)		23 ( 4.9)		0 ( 0.0)			21 ( 2.6)
VZV-positive And HSV-positive	Not considered	HZ case by CEC		0 ( 0.0)		0 ( 0.0)		0 ( 0.0)			0 ( 0.0)
		Non-HZ case by CEC		0 ( 0.0)		0 ( 0.0)		0 ( 0.0)			1 ( 0.1)
Inadequate or missing specimen	Positive for VZV	Not considered		2 ( 0.4)		0 ( 0.0)		8 ( 1.0)			0 ( 0.0)
	Positive for HSV	Not considered		0 ( 0.0)		1 ( 0.2)		0 ( 0.0)			2 ( 0.3)
	No virus isolated or no viral culture	HZ case by CEC (seen early)		19 ( 4.1)		0 ( 0.0)		34 ( 4.3)			0 ( 0.0)
	No virus isolated or no viral culture	HZ case by CEC (seen late)		0 ( 0.0)		3 ( 0.6)		0 ( 0.0)			4 ( 0.5)
	No virus isolated or no viral culture	Non-HZ case by CEC		0 ( 0.0)		36 ( 7.7)		0 ( 0.0)			36 ( 4.5)
*Including multiple cases/subject †Samples were determined as HSV-negative or inadequate for HSV testing ‡These samples were determined as VZV-negative. Percentages calculated based on total number of suspected HZ cases per group (i.e., 100 x n/M) Seen early: Suspected HZ seen during the rash stage (crusted vesicles or earlier). Seen late: Suspected HZ first seen beyond crusting stage of rash. Evaluable: Suspected HZ, 1. Confirmed as VZV-positive by PCR or culture (if PCR result unavailable) 2. Adjudicated as HZ by CEC if not confirmed by laboratory results. N = Number of subjects randomized.				M = Number of suspected HZ cases. n = Number of cases in the respective category. PCR = Polymerase chain reaction. HZ = Herpes zoster. CEC = Clinical Evaluation Committee. VZV = Varicella-zoster virus. HSV = Herpes simplex virus. 2 placebo & 1 zoster vaccine recipients experienced a 2^nd evaluable HZ case – recurrent episode data not included in efficacy analyses

Zoster Vaccine (N = 19270)							Placebo (N = 19276)					Vaccine Efficacy for HZ BOI (95% CIs)
Effect of Age on BOI – AUC
	n		m	Follow-Up Time (Person-Years)	HZ BOI‡		n	M	Follow-Up Time (Person-Years)		HZ BOI‡
60-69 yrs.	122		10370	31323	1.495		334	10356	30953		4.334	0.655 (0.515, 0.755)
≥ 70 yrs.	193		8884	26881	3.471		308	8891	26783		7.781	0.554 (0.399, 0.669)
Effect of Age on PHN Incidence
		n	m	Follow- Up Time (Person-Years)	Incidence of PHN† (1000 Person-Yrs.)		n	M	Follow-Up Time (Person-Years)		Incidence of PHN† (1000 Person-Yrs.)	Vaccine Efficacy for PHN Incidence (95% CI)‡ ‡
60-69 yrs.		8	10370	31323	0.255		23	10356	30953		0.743	0.656 (0.204, 0.867)
≥ 70 yrs.		19	8884	26881	0.707		57	8891	26783		2.128	0.668 (0.433, 0.813)
Effect of Age on HZ Incidence
	n		m	Follow- Up Time (Person-Years)		Incidence of HZ (1000 Person-Yrs.)	n	M		Follow-Up Time (Person-Years)	Incidence of HZ (1000 Person-Years)	Vaccine Efficacy for HZ Incidence (95% CI)† †

60-69 yrs.	122		10370	31323		3.895	334	10356		30953	10.791	0.639 (0.555, 0.709)
≥70 yrs.	193		8884	26881		7.180	308	8891		26783	11.500	0.376 (0.250, 0.481)
†Protocol-defined AUC: (1) incorporates patient-reported data on HZ between rash onset and 1^st interview collected on IZIQ; (2) excludes pain scores < 3 that occur on 2 or more consecutive visits > 30 days after rash onset; and (3) included recurrent pain with score ≥ 3 after 2 consecutive visits with worst pain scores <3 beyond 30 days after HZ rash onset. ††1 minus the ratio of estimated incidence rates of HZ in zoster vaccine group and placebo group. CI was constructed based on the exact conditional procedure. ‡Weighted average of observed BOI stratified by age group (60-69 and ≥70 yrs) w/ weights proportional to total follow-up time in each age group. ‡‡ Calculated as 1 minus the ratio of estimated incidence rate of PHN between the zoster vaccine group and the placebo group. The CI was constructed based on the exact conditional procedure stratified by age group. §Calculated as a weighted average of the observed vaccine efficacy stratified by age group with weights proportional to the total follow-up time in each age group. The CI was constructed based on the large sample approximation under the fixed-number-of-events design. MITT-2: All randomized subjects followed ≥30 days postvaccination and did not develop (per clinical adjudication) w/in 30 days postvaccination. Subjects considered to be immunosuppressed if immunosuppressed at study entry (resulting from a disease), at rash onset, or at time of termination. N = Number of subjects randomized n = Number of evaluable HZ cases in specified population m = Number of subjects w/ active follow-up for HZ surveillance in ITT population AUC = Area under the curve HZ = Herpes zoster BOI = Burden of illness IZIQ = Initial zoster impact questionnaire ITT = Intention-to-treat CI = Confidence interval

	Description	N	% of study population	Placebo	Zoster vaccine	Placebo w/HZ	Zoster vaccine w/HZ	Total w/HZ
ITT	All randomized subjects - randomized just prior to vaccination	38,546	100%	19,276	19,270	660	321	981
	HZ by PCR	--	--	--	--	616	299	915
	HZ by virus culture	--	--	--	--	8	2	10
	HZ by CEC adjudication	--	--	--	--	36	20	56
MITT*	Followed ≥ 30 days postvaccination and did not develop evaluable cases of HZ w/in 30 days postvaccination	38,501	99.9%	19,247	19,254	642	315	957
	Excluded from MITT – HZ w/in 30 days	--	--	--	--	18	6	24
	Excluded from MITT: Discontinued w/in 30 days	--	--	11	10	--	--	--
MITT2	Followed ≥ 30 days postvaccination and did not develop HZ per Clinical Adjudication Committee (CEC) w/in 30 days postvaccination	38,504	99.9%	19,249	19,255	626	332	958
	Excluded from MITT2– HZ w/in 30 days	--	--	--	--	16	5	21
	Excluded from MITT2: Discontinued w/in 30days	--	--	11	10	--	--	--
* MITT is the primary efficacy analyses population †Evaluable cases of HZ were determined by protocol defined hierarchical algorithm (see section 6.1) except in MITT2 population. In MITT2, all evaluable cases of HZ cases were determined based upon clinical diagnosis by CEC irrespective of PCR or viral culture results.

Table of Contents

Page