docs/Statistical Review - COMIRNATY.txt

Application Type
STN
CBER Received Date
PDUFA Goal Date
Division/ Office
Committee Chair
Clinical Reviewers)
Project Manager
Priority Review
Reviewer Names)
Review Completion Date /
Stamped Date
Supervisory Concurrence
Applicant
Established Name
(Proposed) Trade Name
Pharmacologic Class
Formulation(s), including
Adjuvants, etc
Dosage Form(s) and Route(s) of
Administration
Dosing Regimen
Indication(s) and Intended
Population(s)
Statistical Review
STN: 125742/0
BLA, Original Application
125742/0
May 18, 2021
January 16, 2022
DVRPA /OVRR
Ramachandra Naik
Ann Schwartz; Susan Wollersheim Mike Smith; Laura Gottschalk Yes Lei Huang
Tsai-Lien Lin, Branch Chief, VEB, DB,
John A. Scott, Director, DB, OBE
BioNTech Manufacturing GmbH (in partnership with Pfizer, Inc.)
COVID-19 Vaccine, mRNA
COMIRNATY
Vaccine
After preparation, each 0.3 mL dose contains 30ug modified mRNA encoding
SARS-CoV-2 spike glycoprotein
Injectable Suspension, Intramuscular
Two 0.3 mL doses, 3 weeks apart Active immunization to prevent coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in individuals 16 years of age and older.
Page i

Statistical Review
STN: 125742/0
Table of Contents
Glossary....
1. Executive Summary..
2. Clinical and Regulatory Background
3. Submission Ouality and Good Clinical Practices ...............
3.1 Submission Quality and Completeness........
3.2 Compliance With Good Clinical Practices And Data Integrity.
4. Significant Efficacy/Safety Issues Related to Other Review Disciplines............
5
5. Sources of Clinical Data and Other Information Considered in the Review ........... 5
5.1 Review Strategy......
5.2 BLA/IND Documents That Serve as the Basis for the Statistical Review.........
5.3 Table of Studies/Clinical Trials...
6. Discussion of Individual Studies/Clinical Trials.
6.1 Study C4591001.
6.1.1 Objectives.
6.1.2 Design Overview.
6.1.3 Population
6.1.4 Study Treatments or Agents Mandated by the Protocol
6.1.6 Sites and Centers
6.1.7 Surveillance/Monitoring.
6.1.8 Endpoints and Criteria for Study Success
6.1.9 Statistical Considerations & Statistical Analysis Plan
6.1.10 Study Population and Disposition
6.1.11 Efficacy Analyses.....
6.1.12 Safety Analyses.
7. Integrated Overview of Efficacy...........
10
10
12
16
. 25
• 25
8. Integrated Overview of Safety.......
........ 25
9. Additional Statistical Issues...........
........... 25
10. Conclusions..........
10.1 Statistical Issues and Collective Evidence .........
10.2 Conclusions and Recommendations..
............. 25
......... 25
....... 26
Page ii

Statistical Review
STN: 125742/0
GLOSSARY
BIMO
BNT16262
CDC
CI
COVID-19
EUA
HHS
HIV
IM
IR
LNP modRNA
NAAT pV
RT-PCR
SARS-CoV-2
VE
VRBPAC
WHO
Bioresearch Monitoring
PfizerBioNTech COVID-19 Vaccine
Centers for Disease Control and Prevention
Confidence interval coronavirus disease 2019
Emergency Use Authorization
Health and Human Services human immunodeficiency virus intramuscular
Information request lipid nanoparticle nucleoside-modified messenger RNA nucleic acid amplification-based test person-years
reverse transcription-polymerase chain reaction severe acute respiratory syndrome coronavirus 2 vaccine efficacy
Vaccines and Related Biological Products Advisory Committee
World Health Organization
1. EXECUTIVE SUMMARY
Pfizer submitted a Biologics License Application (BLA 125742/0) on May 18, 2021 to seek licensure of the Pfizer-BioNTech COVID-19 Vaccine (BNT16262) for active immunization to prevent Coronavirus Disease 2019 (COVID-19) caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) in individuals 16 years of age and older. The BLA is supported by safety, efficacy, and immunogenicity data from two ongoing studies (C4591001 and BNT-162-01). This statistical review focuses on the analyses of efficacy data collected during the blinded, placebo-controlled follow-up in the Phase 2/3 portion of Study C4591001.
Study C4591001 is an ongoing, randomized, placebo-controlled, observer-blinded Phase 1/2/3 study being conducted in the United States, Argentina, Brazil, Germany, South Africa, and Turkey. In the Phase 2/3 portion of the study, 44,165 subjects aged 16 and above were randomized 1:1 to receive two doses of BNT162b2 or placebo 21 days apart.
Randomization was stratified by age group. Starting December 14, 2020, following issuance of an Emergency Use Authorization (EUA), participants 16 years of age and older were systematically unblinded when eligible per local recommendations and offered BNT1622 vaccination if they had been randomized to placebo.
In the updated efficacy analvsis for cases accrued during blinded placebo-controlled follow-up (cutoff date: March 13, 2021) of Study C4591001 in participants 16 years of age and older, the estimated vaccine efficacy (VE) against confirmed COVID-19
Page 3

Statistical Review
STN: 125742/0
occurring at least 7 days after Dose 2 was 91.1% (95% CI: 88.8%, 93.1%), with 77 COVID-19 cases in the BNT1622 group compared to 833 cases in the placebo group among participants without evidence of SARS-CoV-2 infection before and during the vaccination regimen; the estimated vaccine efficacy (VE) against confirmed COVID-19
occurring at least 7 days after Dose 2 was 90.9% (95% CI: 88.5%, 92.8%), with 81 COVID-19 cases in the BNT1622 group compared to 854 cases in the placebo group among participants with or without evidence of SARS-CoV-2 infection before and during the vaccination regimen.
With respect to efficacy against severe COVID-19 cases occurring at least 7 days after Dose 2, the estimated VE was 95.3% (95% CI: 71.0%, 99.9%), with 1 and 21 cases in the BNT1622 and placebo groups, respectively, among participants without evidence of SARS-CoV-2 infection; the VE result was the same among participants with or without evidence of SARS-CoV-2 infection.
Overall, the updated efficacy analysis results show that BNT1622 provided high VE in preventing symptomatic COVID-19 and severe COVID-19 cases.
2. CLINICAL AND REGULATORY BACKGROUND
Coronavirus disease 2019 (COVID-19) is an infectious disease caused by SARS-CoV-2, a novel coronavirus that emerged in late 2019 in patients with pneumonia of unknown cause. On January 31, 2020, the United States Secretary of Health and Human Services (HHS) made the declaration that COVID-19 constitutes a nationwide public health emergency. On March 11, 2020, the World Health Organization (WHO) declared the COVID-19 outbreak a pandemic.
The BNT16262 vaccine, developed by BioNTech Manufacturing GmbH in partnership with Pfizer, Inc., was granted Fast Track Designation on July 7, 2020 for individuals ≥18 years of age. An Emergency Use Authorization (EUA) was granted in the U.S. on December 11, 2020 for individuals >16 years of age (EUA 27034). An amendment to the EUA was submitted on May 10, 2021 to support emergency use in participants 12 to 15 years of age.
3. SUBMISSION QUALITY AND GOOD CLINICAL PRACTICES
3.1 Submission Quality and Completeness
The submission was adequately organized and integrated to accommodate the conduct of a complete statistical review.
3.2 Compliance With Good Clinical Practices And Data Integrity
Please refer to Haecin Chun's Bioresearch Monitoring (BIMO) review memo.
Page 4

Statistical Review
STN: 125742/0
4. SIGNIFICANT EFFICACY/SAFETY ISSUES RELATED TO OTHER REVIEW
DISCIPLINES
Please refer to other review disciplines' memos.
5. SOURCES OF CLINICAL DATA AND OTHER INFORMATION CONSIDERED IN THE REVIEW
5.1 Review Strategy
This memo focuses on the statistical review of clinical efficacy data. Please refer to Dr.
Ye Yang's memo for the statistical review of clinical safety data, and to Dr. Xinyu Tang's memo for the statistical review of non-clinical data.
To demonstrate efficacy of BNT162b2, the applicant provided the efficacy results from the interim analysis (cutoff date: November 4, 2020), the final analysis (cutoff date:
November 14, 2020), and an updated analysis for cases accrued during blinded placebo-controlled follow-up (cutoff date: March 13, 2021) for Study C4591001. As the efficacy results from the interim and final analyses supported the issuance of an EU and have been reviewed under EUA 27034, this statistical review primarily focuses on the updated efficacy results.
5.2 BLA/IND Documents That Serve as the Basis for the Statistical Review
The following documents submitted to the BLA are reviewed:
125742/0 (submitted on 5/6/2021)
Module 2. Common Technical Document Summaries
• Clinical Overview
• Summary of Clinical Efficacy
Module 5. Clinical Study Reports
• C4591001 Statistical Analysis Plan
C4591001 Interim 6-Month Report
125742/0.3 (submitted on 5/19/2021)
Module 1.11.3 Clinical Information Amendment
• Response to FDA 18 May 2021 IR
125742/0.17 (submitted on 7/26/2021)
Module 1.11.3 Clinical Information Amendment
• Response to CBER Clinical 22 July 2021 Info Request
125742/0.18 (submitted on 7/28/2021)
Module 1.11.3 Clinical Information Amendment
• Response to CBER 22 July 2021 Info Request
125742/0.27 (submitted on 8/2/2021)
Page 5

Statistical Review
STN: 125742/0
Module 1.14.1 Draft Labeling
125742/0.28 (submitted on 8/02/2021)
Module 1.11.3 Clinical Information Amendment
• Response to CBER Clinical 22 July 2021 Information Request
125742/0.32 (submitted on 8/05/2021)
Module 1.11.3 Clinical Information Amendment
• Response 22 Jul 2021 - Follow-up #3
Module 5 Clinical Study Reports
• C4591001 - 508 Efficacy Tables
125742/0.38 (submitted on 8/09/2021)
Module 1.14.1 Draft Labeling
Module 5 Clinical Study Reports
• C4591001 - Source Vaccine Efficacy Tables
125742/0.49 (submitted on 8/16/2021)
Module 1.14.1 Draft Labeling
Module 5 Clinical Study Reports
• C4591001 - Follow Up Table (with and without evidence of infection)
5.3 Table of Studies/Clinical Trials
Data from two ongoing clinical studies were submitted to support the licensing application for BNT16262 and are summarized in Table 1 below. The pivotal data are derived from a single study, C4591001, which is a multi-center, Phase 1/2/3, randomized, double-blinded, placebo-controlled safety, immunogenicity, and efficacy study; the second study, BNT162-01, is a Phase 1 safety and immunogenicity study evaluating various vaccine candidates and dose levels.
Page 6

Statistical Review
STN: 125742/0
Table 1. Clinical Trials Supporting Licensure of the Pfizer-BioNTech COVID-19 Vaccine
Study Number/ Description
BNT162b2 (30 jug)*
Placebo
Country
participants
participants
(N)
(N)
C4591001
Phase 1/2/3 randomized, placebo-
Phase 1ª: 24
Phase 1a. 6
Argentina, Brazil, controlled, observer-blind; to
Phase 2/3b: 22085
Phase 2/3b: 22080
Germany, S.
evaluate safety, immunogenicity
Africa, Turkey, and efficacy of COVID-19 US.A.
vaccine
BNT162-01
Phase 1/2 randomized, open-label; 24
0
Germany
to evaluate safety and
immunogenicity, dose escalation
N= total number of randomized participants 16 years of age and older, as of March 13, 2021 Placebo:
Study
Status
Ongoing
Ongoing
saline.
- Studies C4591001 and BNT162-01 started in April 2020 (first participant, first visit).
* Phase 1 studies included additional participants vaccinated with other dose levels and other mRNA vaccine candidates.
a Phase 1: enrolled individuals 18-85 years of age
b Phase 2/3: Phase 2: enrolled individuals >18 years of age (stratified as 18 to 55 years and 56 to 85 years); Phase 3: enrolled individuals ≥16 years of age (stratified as 16-55 years and >55 years of age).
Source: Summarized by reviewer based on information provided in Module 2 - Clinical Overview.
6. DISCUSSION OF INDIVIDUAL STUDIES/CLINICAL TRIALS
6.1 Study C4591001
Title: Phase 1/2/3, Placebo-Controlled, Randomized, Observer-Blind, Dose-Finding Study to Evaluate the Safety, Tolerability, Immunogenicity, and Efficacy of SARS-CoV-
2 RNA Vaccine Candidates Against COVID-19 in Healthy Individuals
First Subject First Visit: April 29, 2020
Data Cut-off: Mach 13, 2021
6.1.1 Objectives
The objectives and endpoints are presented below are for the Phase 2/3 portion of the study. The objectives for the Phase 1 portion are described in Section 6.1.2 (Design Overview).
Primary efficacy obiectives
1. To evaluate the efficacy of BNT1622 against confirmed COVID-19 occurring from 7 days after Dose 2 in participants without evidence of SARS-CoV-2 infection before vaccination.
Endpoint: COVID-19 disease based on laboratory-confirmed nuclei acid amplification-based test (NAAT) in participants with no serological or virological evidence (up to 7 davs after Dose 2) of past SARS-CoV-2 infection.
Page 7

Statistical Review
STN: 125742/0
2. To evaluate the efficacy of BNT16262 against confirmed COVID-19 occurring from 7 days after Dose 2 in participants with and without evidence of SARS-CoV-2 infection before vaccination.
Endpoint: COVID-19 disease based on laboratory-confirmed NAAT
Secondary efficacy objectives
• To evaluate the efficacy of BNT1622 against confirmed COVID-19 occurring from 14 days after Dose 2 in
o participants without evidence of SARS-CoV-2 infection before vaccination
(Dose 1)
o participants with and without evidence of SARS-CoV-2 infection before
vaccination (Dose 1)
Endpoint: COVID-19 disease based on laboratory-confirmed NAAT
• To evaluate the efficacy of BNT1622 against severe COVID-19 occurring from 7 days and from 14 days after Dose 2 in
• participants without evidence of SARS-CoV-2 infection before vaccination o participants with and without evidence of SARS-CoV-2 infection before
vaccination
Endpoint: Severe COVID-19 disease
• To describe the efficacy of BNT162b2 against confirmed COVID-19 (CDC-defined symptoms) occurring from 7 days and from 14 days after Dose 2 in o participants without evidence of SARS-CoV-2 infection before vaccination o participants with and without evidence of infection before vaccination
Endpoint: COVID-19 disease (CDC-defined symptoms) based on laboratory-confirmed NAAT
6.1.2 Design Overview
Study C4591001 is an ongoing, randomized, placebo-controlled, observer-blinded Phase 1/2/3 study being conducted in the U.S., Argentina, Brazil, Germany, South Africa and Turkey. Initially the study was designed as a Phase 1/2 study in healthy adults in the U.S. for vaccine candidate and dosage selection, as well as evaluation of immunogenicity and preliminary efficacy. The protocol was expanded to include a Phase 2/3 portion of the study to evaluate clinical disease efficacy endpoint in individuals 12 years of age and older in the U.S. and additional sites outside of the U.S. This review will focus on data collected from participants 16 years of age and older.
The Phase 1 portion of the study was designed to identify a preferred vaccine candidate(s) and vaccine dose level(s) for further development based on safety, tolerability, and immunogenicity. To this end, two age groups were evaluated in separate
Page 8

Statistical Review
STN: 125742/0
cohorts: younger adults 18 through 55 years of age (N=45) and older adults 65 through 85 years of age (N=45). The study population included healthy men and women and excluded participants at high risk of SARS-CoV-2 infection or with serological evidence of prior or current SARS-CoV-2 infection. Two different vaccine candidates were evaluated, and younger participants received escalating dose levels Evaluation of escalating dose levels in the older age group (65 through 85 years), were based on recommendations from an internal review committee that reviewed safety and immunogenicity data. For each vaccine candidate and dose level, participants were randomized 4:1, such that 12 participants received the vaccine candidate and 3 participants received placebo. Review of the safety and immunogenicity from Phase 1, in combination with data from Study BNT162-01 (see Section 6.2 of this review), supported the final vaccine candidate and dose level (BNT1622 at 30 ug, given 21 days apart) to proceed into Phase 2/3.
In Phase 2/3, participants were initially enrolled with stratification by age (younger adults: 18 through 55 years of age; older adults: over 55 years of age) and a goal of 40% enrollment in the older adult age group. Adolescents 16-17 years of age (and subsequently 12-15 years of age) were added to the protocol later, based on review of safety data in younger adults enrolled in the ongoing study. The study population for Phase 2/3 includes participants at higher risk for acquiring COVID-19 and at higher risk of severe COVID-19 disease, such as participants working in the healthcare field, participants with autoimmune disease, and participants with chronic but stable medical conditions such as hypertension, asthma, diabetes, and infection with HIV, hepatitis B or hepatitis C. Participants were randomized 1:1 to receive 2 doses of either BNT16262 or placebo, 21 days apart. The Phase 2 portion of the study evaluated reactogenicity and immunogenicity for 360 participants enrolled early, and these participants also contribute to the overall efficacy and safety data in the Phase 3 portion.
The ongoing Phase 3 portion of the study is evaluating the safety and efficacy of
BNT1622 for the prevention of COVID-19 disease occurring at least 7 days after the second dose of vaccine. Efficacy is being assessed throughout a participant's follow-up in the study through surveillance for potential cases of COVID-19. If, at any time, a participant develops acute respiratory illness, an illness visit occurs. Assessments for illness visits include a nasal (midturbinate) swab, which is tested at a central laboratory using a reverse transcription-polymerase chain reaction (RT-PCR) test (e.g., Cepheid;
FDA authorized under EUA), or other sufficiently validated NAAT, to detect SARS-CoV-2. The central laboratory NAAT result is used for the case definition, unless it is not possible to test the sample at the central laboratory. In that case, the following NAAT results are acceptable: Cepheid Xpert Xpress SARS-CoV-2, Roche cobas SARS-CoV-2 real-time RT-PCR test (EUA200009/A001), and Abbott Molecular/RealTime SARS-
CoV-2 assay (EUA200023/A001).
The study design included a planned interim analysis of the first primary efficacy endpoint at pre-specified numbers of COVID-19 cases (at least 62, 92, and 120 cases), and all primary and secondary efficacy endpoints were analyzed in the final efficacy analysis after at least 164 COVID-19 cases were accrued (see Statistical Analysis section,
Page 9

Statistical Review
STN: 125742/0
below). Participants are expected to participate for a maximum of approximately 26 months.
Starting December 14, 2020, following issuance of the Emergency Use Authorization for the Pfizer-BioNTech COVID-19 Vaccine, study participants 16 years of age and older have been unblinded to their treatment assignment when eligible per local recommendations, and offered BNT1622 vaccination if they had been randomized to placebo.
The study was unblinded in stages as each participant was either individually unblinded upon eligibility for vaccination outside the study or had concluded their 6-month post-Dose 2 study visit. Every participant 16 years of age and older who participated in the Phase 2/3 study was given the opportunity to receive BNT16262 no later than the 6-month timepoint after the second study vaccination. Participants who originally received placebo but then went on to receive BNT1622 were moved to a new visit schedule to receive both doses of BNT162b2, 3 weeks apart.
6.1.3 Population
comorbidities.
Individuals 12 years of age and older including those with stable infections and common
6.1.4 Study Treatments or Agents Mandated by the Protocol
Study C4591001 (Phase 1) evaluated a 2-dose series of investigational vaccine or placebo (0.9% normal saline) administered at a 21-day interval. Subiects were randomized to receive one of three levels of investigational RNA vaccine candidates (or placebo) for active immunization against COVID-19. The investigational RNA vaccine candidates included:
• BNT1621 (BNT162 RNA-LNP vaccine utilizing modRNA and encoding the
RBD): dose levels 10 Mg, 20 Mg, 30 Mg, 100 kg
• BNT16262 (BNT162 RNA-LNP vaccine utilizing modRNA and encoding the P2
S): dose levels 10 ug, 20 ug, 30 Mg
Based upon the preliminary results, the vaccine candidate selected for further evaluation in the Phase 2/3 studies was BNT16262 BNT162 RNA-LN vaccine utilizing modRNA
(6) (4) mcg/0.5 mL] at a dose of 30 Mg.
6.1.6 Sites and Centers
The study was conducted in a total of 153 sites: 131 in the U.S., 9 in Turkey, 6 in Germany, 4 in South Africa, 2 in Brazil, and 1 in Argentina.
6.1.7 Surveillance/Monitoring
Please refer to Drs. Susan Wollersheim and Ann Schwartz's clinical review memo.
Page 10

Statistical Review
STN: 125742/0
6.1.8 Endpoints and Criteria for Study Success
Please refer to Section 6.1.1 for efficacy endpoints.
Study success criteria:
In Phase 2/3, the assessment of VE is based on posterior probability of VE >30% and VEr>30%, where VEi represents VE for prophylactic BNT16262 against confirmed COVID-19 in participants without evidence of infection before vaccination, and VE2 represents VE for prophylactic BNT162b2 against confirmed COVID-19 in all participants after vaccination. Only the first primary endpoint was analyzed at interim analyses. The criteria for success at an interim analysis are based on the posterior probability, i.e. Pr(VE>30%|data) at the current number of cases. Efficacy will be declared if the posterior probability is higher than the success threshold, where the success threshold for each interim analysis was calibrated to maintain a familywise type I error rate of 2.5%. If the first primary objective is met, the second primary objective will be evaluated at the final analysis.
6.1.9 Statistical Considerations & Statistical Analysis Plan
The statistical analyses for the Phase 1 portion were descriptive.
For Phase 2/3, the evaluable efficacy population, which included all randomized participants who received all study interventions as randomized within the predefined window and had no other important protocol deviations as determined by the clinicians, was the primary analysis population for all efficacy analyses. Additional analyses based on the all-available efficacy population, which included all randomized subjects who received either at least 1 dose of vaccine or placebo (Dose 1 all-available set) or 2 doses (Dose 2 all-available set), were also performed.
The VE is defined as VE = 100 × (1 - IRR), where IRR is calculated as the ratio of the confirmed COVID-19 illness rate in the vaccine group to the corresponding illness rate in the placebo group. Assuming a true VE of 60%, 164 COVID-19 cases would provide 90% power to conclude true VE >30%. Because the analyses are based on the number of cases rather than the number of participants, the total number of participants enrolled in Phase 2/3 would vary depending on the incidence of COVID-19 at the time of enrollment, the true underlying VE, and a potential early stop for efficacy or futility. Four interim analyses were planned to be performed after accrual of at least 32, 62, 92, and 120 cases. However, for operational reasons, the first IA was not performed until 94 cases were accrued, followed by the final analysis with 170 cases.
VE was evaluated using a beta-binomial model and the posterior probability of VE being
>30% was assessed. A minimally informative beta prior, beta (0.700102, 1), was
proposed for 0 = r(1-VE)/(1+r(1-VE)), where r is the ratio of surveillance time in the
BNT162b2 group over that in the placebo group. For participants with multiple confirmed cases, only the first case contributed to the VE calculation. The two primary efficacy endpoints were evaluated sequentially to control the familywise type I error rate at 2.5% (one-sided). For the primary endpoint analysis, missing efficacy data were not imputed; only participants with known disease status were included. A sensitivity
Page 11

Statistical Review
STN: 125742/0
analysis was performed by imputing missing values with the assumption of missing at random (MAR). Secondary endpoints were evaluated similarly to the primary endpoints.
After the final efficacy analyses at 170 cases, updated efficacy analyses on primary and secondary efficacy endpoints were performed with additional data accrued. The point estimate of VE in the blinded follow-up period and associated 2-sided 95% CI were derived using the Clopper-Pearson method, adjusting for surveillance time. The posterior probability, Pr(VE>30%|data), was also provided.
6.1.10 Study Population and Disposition
6.1.10.1 Populations Enrolled/Analyzed
Participants 18 through 55 years of age and 56 years of age and older began enrollment into Phase 2/3 from July 27, 2020 and participants 16 through 17 years of age began enrollment from September 16, 2020.
6.1.10.1.1 Demographics
The population for the updated analysis of vaccine efficacy endpoint (March 2021 data cutoff) included 42,436 participants 16 years of age and older (21,136 in the BNT16262 group and 21,300 in the placebo group), with or without evidence of prior infection with SARS-CoV-2 through 7 days after the second dose. Table 2 presents the specific demographic characteristics in the studied population.
The evaluable efficacy population who received BNT1622 included 48.6% females, 81.9% White, 9.5% African American, 4.4% Asian, and <3% from other racial groups;
25.6% of participants were Hispanic/Latino. The median age was 51 years. One or more comorbidities that increase the risk of severe COVID-19 disease were present among 46% of participants. Evidence of prior SARS-CoV-2 infection was observed in 3% of participants. Geographically, <2% of participants lived in Germany, Turkey and South Africa, 6.8% lived in Brazil, 12.7% lived in Argentina, and 76.4% of participants lived in the U.S.
Page 12

Statistical Review
STN: 125742/0
Table 2. Demographics and Other Baseline Characteristics, Participants 16 Years of Age and Older, With or Without Evidence of Infection Prior to 7 Days After Dose 2, Evaluable Efficacy Population (Data Cutoff March 13, 2021)
BNT162b2 (30 ug)
(Na=21136
Characteristic
nb (%)
Placebo
(N°=21300)
nb (%)
Total
(Na=42436)
nb (%)
Sex: Female
Sex: Male
Age at Vaccination: Mean years (SD)
Age at Vaccination: Median (years)
Age at Vaccination: Min, max (years)
Age Group: 16 to <18 years
Age Group: 18 to 55 years
Age Group: >55 years
Age Group: ≥65 years
Race: American Indian or Alaska Native
Race: Asian
Race: Black or African American
Race: Native Hawaiian or Other Pacific Islander
Race: White
Race: Multiracial
Race: Not reported
Ethnicity: Hispanic or Latino
Ethnicity: Not Hispanic or Latino
Ethnicity: Not reported
Obesity: Yes
Obesity: No
Comorbidities: Yesd
Comorbidities: No
Baseline evidence of prior SARS-CoV-2 infection:
Negative™
Positive®
Missing
Country: Argentina
Country: Brazil
Country: Germany
Country: South Africa
Country: Turkey
Baseline evidence of prior SARS-CoV-2 infection:
Baseline evidence of prior SARS-CoV-2 infection:
10280 (48.6)
10856 (51.4)
49.8 (16.0)
51.0
(16, 89)
370 (1.8)
12120 (57.3)
8646 (40.9)
4407 (20.9)
204 (1.0)
929 (4.4)
2009 (9 5)
56 (0.3)
17304 (81.9)
545 (2.6)
89 (0.4)
5403 (25 6)
15628 (73.9)
105 (0.5)
7239 (34.2)
13897 (65.8)
9712 (46.0)
11424 (54.0)
20365 (96.4)
627 (3.0)
144 (0.7)
2686 (12.7)
1437 (6.8)
240 (1.1)
391 (1.8)
241 (1.1)
10579 (49.7) 20859 (49.2)
10721 (50.3) 21577 (50.8)
49.7 (16.0)
49.7 (16.0)
51.0
51.0
(16, 91)
(16, 91)
362 (1.7)
732 (1.7)
12252 (57.5) 24372 (57.4)
8686 (40.8) 17332 (40.8)
4429 (20.8)
8836 (20.8)
190 (0.9)
394 (0.9)
924 (4.3)
1853 (4.4)
2036 (9.6)
4045 (9.5)
32 (0.2)
88 (0.2)
17487 (82.1) 34791 (82.0)
519 (2.4)
1064 (2.5)
112 (0.5)
201 (0.5)
5409 (25.4) 10812 (25.5)
15778 (74.1) 31406 (74.0)
113 (0.5)
218 (0.5)
7386 (34 7) 14625 (34 5)
13914 (65.3) 27811 (65.5)
9736 (45.7) 19448 (45.8)
11564 (54.3) 22988 (54.2)
20511 (96.3) 40876 (96.3)
669 (3.1) 1296 (3.1)
120 (0.6)
264 (0.6)
2710 (12.7)
1432 (6.7)
243 (1.1)
392 (1.8)
238 (1.1)
5396 (12.7)
2869 (6.8)
483 (1.1)
783 (1.8)
479 (1.1)
Page 13

BNT162b2 (30 Mg)
(Na=21136)
nb (%)
Placebo
(N°=21300)
n' (%)
Statistical Review
STN: 125742/0
Total
(N°=42436)
nb (%)
Characteristic
Country: United States of America
16141 (76.4)
16285 (76.5) 32426 (76.4)
Abbreviation: SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2.
Note: HIV-positive subjects are included in this summary but not included in the analyses of the overall study obiectives.
a. N= number of subjects in the specified group, or the total sample. This value is the denominator for the percentage calculations.
Note: The analysis was based on treatment group as randomized. b.
n = Number of subjects with the specified characteristic.
C.
Subiects who had BMI >30 kg/m?.
Number of subiects who have 1 or more comorbidities that increase the risk of severe COVID-19 disease: defined as subjects who had at least one Charlson comorbidity index category or BMI >30 kg/m?.
e. Positive N-binding antibody result at Visit 1, positive NAAT result at Visit 1, or medical history of COVID-19.
f. Negative N-binding antibody result and negative NAAT result at Visit 1 and no medical history of COVID-19.
Source: Table F of C4591001-508-efficacy-tables submitted to ST 125742/0.32.
6.1.10.1.2 Medical/Behavioral Characterization of the Enrolled Population
Please refer to Drs. Susan Wollersheim and Ann Schwartz's clinical review memo.
6.1.10.1.3 Subject Disposition
The disposition of all Phase 2/3 participants 16 years of age and older is presented in Table 3. During the blinded placebo-controlled follow-up period, most participants randomized received Dose 1 (99.7%) and Dose 2 (98.0%).
Table 3. Disposition of Participants 16 Years of Age and Older, Phase 2/3 Subjects, Efficacy Population (Data Cutoff March 13, 2021)
BNT162b2
(30 ug)
n° (%)
Randomized
Dose 1 all-available efficacy population
Subiects without evidence of infection before Dose 1
Subjects excluded from Dose 1 all-available efficacy population
Reason for exclusion°
Did not receive at least 1 vaccination
Data considered potentially unreliable due to lack of PI oversight identified as significant quality event
22085 (100.0)
22009 (99.7)
21172 (95.9)
76 (0.3)
Placebo na (%)
22080
(100.0)
Total n° (%)
44165
(100.0)
22008
44017
(99.7)
(99.7)
21168
42340
(95.9)
(95.9)
72 (0.3) 148 (0.3)
55 (0.2)
21 (0.1)
50 (0.2) 105 (0.2)
22 (0.1) 43 (0.1)
Page 14

Statistical Review
STN: 125742/0
BNT16262
(30 ug)
nª (%)
Placebo na (%)
Total na (%)
Dose 2 all-available efficacy population
21648 (98.0)
21624
43272
(97.9)
(98.0)
Subjects without evidence of infection prior to 7 days after Dose 20536 (93.0)
20487
41023
2
(92.8)
(92.9)
Subjects excluded from Dose 2 all-available efficacy population
437 (2.0)
456 (2.1) 893 (2.0)
Reason for exclusion°
Did not receive 2 vaccinations
374 (1.7)
Data considered potentially unreliable due to lack of PI oversight
21 (0.1)
430 (1.9) 804 (1.8)
22 (0.1) 43 (0.1)
identified as significant quality event
Unblinded prior to 7 days after Dose 2
44 (0.2）
11 (0.0) 55 (0.1)
Evaluable efficacy (7 days) population
21136 (95.7)
Subiects without evidence of infection prior to 7 days after Dose 20064 (90.8)
2
Subjects excluded from evaluable efficacy (7 days) population
Reason for exclusion
Randomized but did not meet all eligibility criteria
Data considered potentially unreliable due to lack of PI oversight identified as significant quality event
Did not receive all vaccinations as randomized or did not receive
Dose 2 within the predefined window (19-42 days after Dose 1)
Unblinded prior to 7 days after Dose 2
Had other important protocol deviations on or prior to 7 days after Dose 2
949 (4.3)
21300
42436
(96.5)
(96.1)
20197
40261
(91.5)
(91.2)
780 (3.5) 1729 (3.9)
32 (0.1)
21 (0.1)
30 (0.1) 62 (0.1)
22 (0.1) 43 (0.1)
718 (3.3) 729 (3.3) 1447 (3.3)
44 (0.2)
240 (1.1)
11 (0.0) 55 (0.1)
58 (0.3) 298 (0.7)
Note: HIV-positive subjects are included in this summary but not included in the analyses of the overall study obiectives.
Note: The analysis was based on treatment group as randomized.
а.
n = Number of subjects with the specified characteristic.
b. These values are the denominators for the percentage calculations.
c. Subjects may have been excluded for more than 1 reason.
Source: Table D of C4591001-508-efficacy-tables submitted to ST 125742/0.32.
Reviewer Comment
1. There were more protocol deviations leading to exclusion from analvses in the BNT162b2 group than in the placebo group. The majority of protocol deviations were in the category of investigational products, including dosing/administration error and investigational product deemed not suitable for use. Protocol deviations in other categories appeared balanced across the two treatment groups. The additional analysis on the all-available efficacy population may be regarded as a sensitivity analysis and showed very similar efficacy results.
2. The Dose I all-available efficacy population excluded 43 subjects (21 in the BNT162b2 group and 22 in the placebo group) due to a specific protocol
Page 15

Statistical Review
STN: 125742/0
deviation, i.e. data considered potentially unreliable due to lack of PI oversight identified as a significant quality event, while the Dose I all-available set is defined as all randomized participants who received at least I vaccination in the SAP. I conducted a sensitivity analysis without excluding these 43 subjects for efficacy analyses using the Dose I all-available population, when applicable, and it showed minimal impact on VE results.
6.1.11 Efficacy Analyses
6.1.11.1 Analyses of Primary Endpoints
The Interim and Final Analyses
At the interim analysis, there were 4 confirmed COVID-19 cases in the BNT1622 group and 90 confirmed cases in the placebo group among subjects without evidence of prior SARS-CoV-2 infection prior to 7 days after Dose 2, resulting in a VE point estimate of 95.5% (95% credible interval: 88.8%, 98.4%) and a 99.99% posterior probability for the true VE being >30%, which met the prespecified success criterion of posterior probability
>99.5%. The median follow-up duration for subjects included in the first interim efficacy analysis was slightly less than the planned 2 months. In the final analysis, the case split between the BNT162b2 and placebo groups was 8:162 (VE: 95.0%; 95% credible interval: 90.3%, 97.6%) among subjects without evidence of prior SARS-CoV-2 infection prior to 7 days after Dose 2, and 9:169 (VE: 94.6%; 95% credible interval:
89.9%, 97.3%) among subjects with and without evidence of prior SARS-CoV-2 infection prior to 7 days after Dose 2. The final analysis extended the median follow-up for these subjects to greater than 2 months, and the results indicate that the conclusions from the first interim efficacy analysis would not change when including additional follow-up to November 14, 2020. This pre-specified primary efficacy analysis was the basis for issuance of the Emergency Use Authorization (EUA) for the Pfizer-BioNTech COVID-19 Vaccine on December 10, 2020.
Reviewer Comment
1. The efficacy results presented above included 88 subjects 12-15 years of age (46 in the BNI162b2 group and 42 in the placebo group). Since none of these 12-15 years old subjects developed protocol defined cases and the number of subjects is small relative to the evaluable population, the efficacy results excluding these subjects are very similar to the results including them. Based on my calculation, VE for 16 years and older subjects is 94.6% (95% credible interval: 90.3%, 97.6%).
2. The interim and final analyses were reviewed under EUA 27034, and hence the review is not replicated for this BLA submission.
Updated Efficacy Analyses
Updated efficacy analyses were performed with additional confirmed COVID-19 cases accrued during blinded placebo-controlled follow-up through March 13, 2021,
Page 16

Statistical Review
STN: 125742/0
representing up to 6 months of follow-up after Dose 2 for participants in the efficacy population.
For participants without evidence of SARS-CoV-2 infection prior to 7 days after Dose 2, the updated VE against confirmed COVID-19 occurring at least 7 days after Dose 2 was 91.1%. The case split was 77 COVID-19 cases in the BNT16262 group compared to 833
COVID-19 cases in the placebo group (Table 4).
Table 4. Updated Efficacy of BNT162b2 Against Confirmed COVID-19 From 7 Days
After Dose 2 in Participants Without Evidence of Prior SARS-CoV-2 Infection
- Evaluable Efficacy Population, 16 Years and Older (Data Cutoff March 13, 2021)
BNT16262
Placebo
(Na=19993)
(Na=20118)
Cases nib
Cases nib
Surveillance Time°
Surveillance Time
Pre-specified Age Group
(n2d)
(n2d)
Vaccine Efficacy %
(95% CHe
All participants
77
833
6.092
5.857
91.1
(88.8. 93.1)
(19711)
(19741)
16 to 55 years
52
568
3.593
3.439
91.2
(88.3, 93.5)
(11517)
(11533)
>55 years and older
25
265
2.499
2.417
90 9
(86.2. 94.2)
(8194)
(8208)
Abbreviations: N-binding = SARS-CoV-2 nucleoprotein-binding; NAAT = nucleic acid amplification test;
SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; VE = vaccine efficacy.
Note: Subiects who had no serological or virological evidence (prior to 7 davs after receipt of the last dose) of past SARS-CoV-2 infection (i.e., N-binding antibody [serum] negative at Visit 1 and SARS-CoV-2 not detected by NAAT (nasal swab] at Visits 1 and 2), and had negative NAAT (nasal swab) at any unscheduled visit prior to 7 days after Dose 2 were included in the analysis.
a. N= number of subjects in the specified group. b.
n = Number of subjects meeting the endpoint definition.
c. Total surveillance time in 1000 person-years for the given endpoint across all subjects within each group at risk for the endpoint. Time period for COVID-19 case accrual is from 7 days after Dose 2 to the end of the surveillance period.
d. n2 = Number of subjects at risk for the endpoint.
e. Confidence interval (CI) for VE is derived based on the Clopper and Pearson method adjusted for surveillance time.
Source: Table H of C4591001-508-efficacv-tables submitted to STN 125742/0.32.
Reviewer Comment
1. One subject (C4591001 (b) (6)
) reported "covid-19 antibody test
positive" in medical history but was included in the VE analvis in participants without evidence of prior infection in Table 4. An information request (IR) was sent on July 22, 2021. In the IR response submitted on July 26, 2021, the applicant clarified that "without evidence of prior infection" was based only on the NAAT tests at Visits I and 2 and the N-binding assay results due to the
Page 17

Statistical Review
STN: 125742/0
potential uncertainty of a medical history entry without knowledge of circumstances, assay performed, etc. Because this subject received placebo and was not a case, inclusion of subject would result in no real change to the VE estimate.
2. A total of 9 participants in the placebo group with COVID-19 symptoms starting on the same day of unblinding with PCR confirmation either on the same day or a few days after, were included in these analyses as positive cases.
3. Initially, there was one additional case reported in the placebo group, for Subject
C4591001 (b) (6)
This subject reported three COVID symptom
episodes: from October 8, 2020 to October 16, 2020, November 2, 2020 to December 11, 2020, and December 17, 2020 to January 16, 2021 (referred to as Episodes A, B and C, respectively). The PCR tests were negative for the first two episodes and positive for Episode C. Since the three episodes were more than 4 days apart, they should be treated as separate episodes per the statistical analysis plan (SAP). Hence, this subject should be considered to be a case with an onset on December 17, 2020, one day after the unblinding on December 16, 2020, and should be excluded from the analysis. In the IR response submitted on July 26, 2021, the applicant explained that Episodes B and C were merged into one episode as this subject was hospitalized from (b) (6)
to (b) (6)
, connecting Episodes B and C. We did not agree with the merging of the two episodes, because hospitalization is not a symptom or criterion pre-specified in the protocol for COVID-19 definition and there were no other data that could corroborate that this hospitalization was due to COVID-19. The applicant agreed to remove this case and updated efficacy tables were submitted on August 5, 2021.
4. The set of subjects used for efficacy analyses excluded those who had reported
COVID symptoms but had missing or unknown PC results at any time. It may be reasonable to exclude subjects who had reported COVID symptoms but had missing/unknown PCR results prior to 7 days after Dose 2 for efficacy analyses in participants without evidence of prior infection. However, subjects who reported symptoms and had missing/unknown PCR results after 7 days post Dose 2 were also excluded from the risk set, while they were at risk for the efficacy endpoint (lab-confirmed COVID-19 starting from 7 days post Dose 2). An IR was sent to the applicant on July 22, 2021. In the IR response submitted on July 26, 2021, the applicant explained that subjects who reported symptoms and had missing/unknown PC results do not have a chance to be counted in the numerator and inclusion of these subjects may result in an underestimation of the incidence rate. Since the percentages of such subjects were small and slightly higher in the placebo group, excluding them from the analyses likely had minimal impact on VE results. Per our request, the applicant also provided a sensitivity analysis under the missing at random (MAR) assumption, where missing efficacy endpoints were imputed based on predicted probability from logistic regression model using the fully conditional specification method for a total of 648 subjects (279 in BNT162b2 group and 369 in placebo group) in the evaluable population who reported COVID-19 symptoms from 7 days post Dose 2 but had missing/unknown PCR results. As a supplementary sensitivity analysis, the
Page 18

Statistical Review
STN: 125742/0
applicant also applied a conservative approach to the model by assuming a higher than the observed case rate when imputing missing efficacy endpoints from participants in the BNTI62b2 group only, to reflect potentially unknowable missing not at random effects that are unfavorable for efficacy result of the study.
As shown in Table 5, the average VE after imputation was 90.76% under the MAR assumption, which is consistent with the efficacy results reported in Table 4. The sensitivity analyses under the missing-not-at-random assumptions show that the efficacy results are robust, e.g. at least a 16-fold increase of positivity rate in the BNTI62b2 group is required for the average VE to fall below 70%, which we do not consider to be a plausible scenario.
Table 5. Sensitivity and Robustness Analysis of Missing Laboratory Results for Vaccine Efficacy - First COVID-19 Occurrence From 7 Days After Dose 2 - Subjects Without
Evidence of Infection Prior to 7 Days After Dose 2 - Evaluable Efficacy (7 Days)
Population
Assumed
Missing Data
Mechanism
Infection Rates
Average Positive Based on Existing Median Median of
Rate (%) Across and Imputed
Posterior
Lower
all Imputations
Values
Probability Limit of
(BNT162b2:
(BNT162b2:
of
95% CI for Median
Average
Placebo)a
Placebo)b
VE>30%
VE
VE (%) VE (%)
MAR
MNARI
MNAR2
MNAR3
MNAR4
MNAR5
4.0:28.5
10.1:28 5
23.3:285
45.3:28.5
69.1:28.5
85.9:28.5
4.21:45.31
5.01:45.31
6.76:45.31
9.69:45.31
100.00
100.00
100.00
100.00
12.85:45.31
15.08:45.31
100.00
100.00
88.56
86.55
82.30
75.36
67.71
62.36
90 78
88.97
85.12
78.79
71.78
66.81
90.76
88.98
85.14
78.69
71.75
66.85
Abbreviations: MAR = missing at random; MNAR = missing not at random; VE = vaccine efficacy.
Note: Each row of this table represents summary results from 500 imputations that were generated using
SAS PROC MI Fully Conditional Specification (FCS) method. Each imputation filled in the missing laboratory results based on a logistic regression model at the subject level, under the assumed missing data mechanism.
a. Average positive rate for each vaccine group was calculated as the mean of positive rates across all imputations among subjects with missing data after each imputation. Under the MAR assumption, the imputation model assumes the probability of positive cases for each vaccine group to be the same as observed from subjects with no missing data in that group. Under each MNAR assumption, while keeping the imputation model for placebo group unchanged, an increase in the positive rate for the BTN16262 group was assumed to reflect a potential conservative and unknowable MNAR scenario for efficacy results of the study.
b. Infection rate in each vaccine group was the number of cases divided by a total number of subjects in that vaccine group times 1000.
Source: Adapted from Table I of response-22jul2021-followup submitted to STN 125742/0.28.
For participants with and without evidence of SARS-CoV-2 infection before and during vaccination regimen, the updated VE against confirmed COVID- 19 occurring at least 7 days after Dose 2 was 90.9%, with 81 and 854 cases in the BNT1622 and placebo groups, respectively (Table 6).
Page 19

Statistical Review
STN: 125742/0
Table 6. Updated Efficacy of BNT16262 Against Confirmed COVID-19 From 7 Days
After Dose 2 in Participants With or Without Evidence of Prior SARS-CoV-2 Infection
- Evaluable Efficacy Population, 16 Years and Older (Data Cutoff March 13, 2021)
BNT16262
Placebo
(Nª=21047)
(NA=21210)
Cases
Cases
nib
nib
Surveillance Time° Surveillance Time Vaccine Efficacy %
Pre-specified Age Group
(n24)
(n20)
(95% CI)°
All participants
81
854
6.340
6.110
90.9
(88.5, 92.8)
(20533)
(20595)
16 to 55 years
56
584
3.766
3.619
90.8
(87.9, 93.1)
(12088)
(12142)
>55 years and older
25
270
2.573
2.492
91.0
(86.5, 94.3)
(8445)
(8453)
Abbreviations: SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; VE = vaccine efficacy.
N = number of subjects in the specified group.
b. n1 = Number of subjects meeting the endpoint definition.
c. Total surveillance time in 1000 person-years for the given endpoint across all subjects within each
group at risk for the endpoint. Time period for COVID-19 case accrual is from 7 days after Dose 2 o the
end of the surveillance period.
d. n2 = Number of subiects at risk for the endpoint.
e. Confidence interval (CI) for VE is derived based on the Clopper and Pearson method adiusted for
surveillance time.
Source: Table I of C4591001-508-efficacy-tables submitted to STN 125742/0.32.
VE in participants in the all-available efficacy population was similar to results in the evaluable efficacy population (Table 7). The VE for the prevention of COVID-19 disease after Dose 1 is 87.6%, in the all-available efficacy population. Based on the number of cases accumulated after Dose 1 and before Dose 2, there seems to be some protection against COVID- 19 disease following one dose (VE=56.4%); however, these data do not provide information about longer term protection beyond 21 days after a single dose.
Page 20

Statistical Review
STN: 125742/0
Table 7. Primary Efficacy Endpoint - Participants 16 Years of Age and Older - Dose 1 All-Available Efficacy Population (Data Cutoff March 13, 2021)
BNT162b2
Placebo
(N°=21909)
(Na=21908)
Cases
Cases
nib
nib
Surveillance Time Surveillance Time Vaccine Efficacy %
Efficacy Endpoint Subgroup
(n29)
(n29)
(95% CI)e
First COVID-19 occurrence after Dose 1
After Dose 1 to before Dose 2
Dose 2 to 7 days after Dose 2
≥7 Days after Dose 2
128
8.155
(21385)
43
1.273
(21385)
3
0 403
(21049)
82
6 479
(21019)
998
7.874
(21315)
98
1.266
(21315)
30
0.401
(20952)
870
6.207
(20901)
87.6
(85.1, 89.8)
56.4
(37.0, 70.3)
90.0
(68.0, 98.1)
91.0
(88.7, 92.9)
Abbreviation: VE = vaccine efficacy.
a. N= number of subjects in the specified group. b.
n1 = Number of subjects meeting the endpoint definition.
c. Total surveillance time in 1000 person-years for the given endpoint across all subjects within each group at risk for the endpoint. Time period for COVID-19 case accrual is from Dose 1 to the end of the surveillance period. d.
n2 = Number of subiects at risk for the endpoint.
e.
Confidence interval (CI) for VE is derived based on the Clopper and Pearson method adiusted for surveillance time.
Source: Table O of C4591001-508-efficacy-tables submitted to ST 125742/0.32.
Reviewer Comment
As mentioned, the Dose I all-available efficacy population excluded 43 subjects with a protocol deviation of data being considered potentially unreliable due to lack of PI oversight identified as a significant quality event. In my additional analysis with these subjects included, the case split for the first COVID-19 occurrence after dose I is
129:1003, resulting in an estimated VE of 87.6% (95% CI: 85.1%, 89.7%). Hence, the exclusion of these subjects likely had minimal impact on the VE results.
6.1.11.2 Analyses of Secondary Endpoints
Protocol-Defined Severe cases
Undated efficacy analvses of the secondary efficacy endpoint for the use of BNT16262 for the prevention of severe COVID-19 were also evaluated. Vaccine efficacy against severe COVID-19 is presented in Table 8 for participants without prior SARS-CoV-2 infection. In the updated analysis, among participants without evidence of prior infection, the estimated VE against severe COVID-19 disease occurring at least 7 days after Dose 2 was 95.3% (71.0%, 99.9%), with one subject who received BNT16262 and 21
Page 21

Statistical Review
STN: 125742/0
participants who received placebo experiencing severe disease. The same number of severe cases were reported among participants with or without evidence of prior infection and the estimated VE was the same (95.3%). These updated analyses of the secondary vaccine efficacy with a larger number of severe cases now shows more definitive evidence of protection against severe COVID-19 disease offered by BNT16262 (the data from the November 14, 2020 cut-off were limited to 4 total severe cases).
Table 8. First Severe COVID-19 Occurrence From 7 Days After Dose 2 - Subiects
Without Evidence of Infection Prior to 7 Days After Dose 2 - Participants 16 Years of Age and Older - Evaluable Efficacy Population (Data Cutoff March 13, 2021)
BNT16262
Placebo
(Na=19993)
(N°=20118)
Cases
Cases
nib
nib
Surveillance
Surveillance
Time©
Time©
Secondary Efficacy Endpoint
(n2d)
(n2d)
Vaccine
Efficacy %
(95% CI)e
First severe COVID-19 occurrence from 7 days after Dose 2 in participants without evidence of prior SARS-CoV-2 infection
1
6.103
(19711)
21
5.971
(19741)
95.3
(71.0, 99.9)
Abbreviations: N-binding = SARS-CoV-2 nucleoprotein-binding; NAAT = nucleic acid amplification test;
SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; VE = vaccine efficacy.
Note: Subiects who had no serological or virological evidence (prior to 7 days after receipt of the last dose) of past SARS-CoV-2 infection (ie, N-binding antibody [serum] negative at Visit 1 and SARS-CoV-2 not detected by NAAT [nasal swab] at Visits 1 and 2), and had negative NAAT (nasal swab) at any unscheduled visit prior to 7 days after Dose 2 were included in the analysis.
a. N= number of subjects in the specified group. b.
n1 = Number of subjects meeting the endpoint definition.
C.
Total surveillance time in 1000 person-years for the given endpoint across all subjects within each group at risk for the endpoint. Time period for COVID-19 case accrual is from 7 days after Dose 2 to the end of the surveillance period.
d. n2 = Number of subiects at risk for the endpoint.
e.
Confidence interval (CI) for VE is derived based on the Clopper and Pearson method adjusted for surveillance time.
Source: Table M of C4591001-508-efficacy-tables submitted to ST 125742/0.32.
In the all-available efficacy population, 31 participants had severe COVID-19 disease after Dose 1 (one subject who received BNT16262 and 30 participants who received placebo) as summarized in Table 9.
Page 22

Statistical Review
STN: 125742/0
Table 9. First Severe COVID-19 Occurrence After Dose 1 - Participants 16 Years of Age and Older - Dose 1 All-Available Efficacy Population (Data Cutoff March 13, 2021)
BNT162b2
Placebo
(Na=21909)
(N°=21908)
Cases
Cases
nib
nib
Surveillance Time Surveillance Time Vaccine Efficacy %
Secondary Efficacy Endpoint
(n2d)
(n2)
(95% CIe
First severe case occurrence after Dose 1
1
8.181
(21385)
After Dose 1 to before Dose 2
1.285
(21385)
Dose 2 to 7 days after Dose 2
>7 Days after Dose 2
0.403
(21056)
1
6 493
(21029)
30
8.032
(21316)
6
1.293
(21316)
1
0 402
(20962)
23
6.337
(20940)
96.7
(80.3, 99.9)
100.0
(14.6, 100.0)
100.0
NA
95.8
(73.9, 99.9)
Abbreviation: VE = vaccine efficacy.
a. N= number of subjects in the specified group. b.
n1 = Number of subjects meeting the endpoint definition.
c. Total surveillance time in 1000 person-years for the given endpoint across all subjects within each group at risk for the endpoint. Time period for COVID-19 case accrual is from Dose 1 to the end of the surveillance period d.
n2 = Number of subiects at risk for the endpoint.
e.
Confidence interval (CI) for VE is derived based on the Clopper and Pearson method adiusted for surveillance time.
Source: Table N of C4591001-508-efficacy-tables submitted to STN 125742/0.32.
Severe Case Based on CDC-Definition
Vaccine efficacy against severe COVID-19 based on the CDC definition is presented for participants with or without prior SARS-CoV-2 infection (Table 10) as the COVID-19 case counts in participants without prior SARS-CoV-2 infection were the same as those in participants with or without prior SARSCoV2 infection in both the vaccine and placebo groups.
Page 23

Statistical Review
STN: 125742/0
Table 10. First Severe COVID-19 Occurrence Based on CDC-Definition From 7 Days
After Dose 2 - Subiects With or Without Evidence of Infection Prior to 7 Days After Dose 2 - Participants 16 Years of Age and Older - Evaluable Efficacy Population (Data
Cutoff March 13, 2021)
Efficacy Endpoint
BNT162b2
(N°=21047)
Cases nib
Surveillance
Time (n20)
Placebo
(N°=21210)
Cases nib
Surveillance
Time© (n29
Vaccine
Efficacy %
(95% CIe
First severe COVID-19 occurrence based on CDC-definition from 7 days after Dose 2
0
6.345
(20513)
31
6.225
(20593)
100.0
(87.6, 100.0)
Abbreviations: VE = vaccine efficacy.
N = number of subjects in the specified group.
b. n1 = Number of subjects meeting the endpoint definition.
c. Total surveillance time in 1000 person-years for the given endpoint across all subjects within each group at risk for the endpoint. Time period for COVID-19 case accrual is from 7 days after Dose 2 to the end of the surveillance period.
d. n2 = Number of subjects at risk for the endpoint.
e. Confidence interval (CI) for VE is derived based on the Clopper and Pearson method adjusted for surveillance time.
Source: Adapted from Table ADCI9EF_VE_SEV_7PD2_CDC_EVALofC4591001-ve-tables submitted to
STN 125742/0.38.
6.1.11.3 Subpopulation Analyses
VE point estimates for the primary endpoint in participants without evidence of prior infection were comparable across sex, age groups (16 to 55 years and >55 years), race, ethnicity, and country, excluding categories with too few cases to analyze. Additional subgroup analyses were performed for the second vaccine efficacy endpoint (i.e.
COVID-19 for participants with and without evidence of infection prior to vaccination) because this endpoint may generalize better to the population who may receive the vaccine, as baseline evidence of prior infection may not be known by all people who might receive the vaccine. VE point estimates were generally high (>84%) across the subgroups examined (i.e. sex, age, race, ethnicity, comorbidity, baseline SARS-CoV-2 status, and country with the exception of participants identified as positive or unknown for baseline SARS-CoV-2 status and with un-reported ethnicity, for which there were too few COVID-19 cases to interpret efficacy data for these subgroups.
6.1.11.4 Dropouts and/or Discontinuations
Dropouts and discontinuations are generally balanced across the groups. There were 352 (1.6%) participants in the BNT162b2 group and 528 (2.4%) participants in the placebo group who discontinued from the vaccination period (Dose 1 to 1 month after Dose 2).
Most participants completed the visit at 1 month post-Dose 2 (>96.4%). Few participants in the BNT1622 and placebo groups were withdrawn from the study (1.6% and 2.2%,
Page 24

Statistical Review
STN: 125742/0
respectively), and most were due to withdrawals by the participant, or they were lost to follow-up without other cause given.
Starting December 14, 2020, following issuance of the Emergency Use Authorization for the Pfizer-BioNTech COVID-19 Vaccine, study participants 16 years of age and older have been unblinded to their treatment assignment when eligible per local recommendations, and offered BNT16262 vaccination if they had been randomized to placebo. The length of blinded follow-up appears to be balanced between the BNT16262 and placebo groups. During the blinded placebo-controlled follow-up period, 52.4% of participants in the BNT1622 group and 52.6% of participants in the placebo group in the evaluable efficacy population with or without evidence of infection prior to 7 days after dose 2 had follow-up time between ≥4 months to <6 months after Dose 2, and 8.4% in the BNT16262 group and 6.1% in the placebo group had follow up ≥6 months.
6.1.11.5 Exploratory and Post Hoc Analyses
Not Applicable.
6.1.12 Safety Analyses
Please refer to Dr. Ye Yang's memo for the statistical review of the clinical safety data of
Study C4591001.
7. INTEGRATED OVERVIEW OF EFFICACY
Data supporting the effectiveness of the vaccine were primarily generated in Study
C4591001. Consequently, no pooled efficacy analyses were performed.
8. INTEGRATED OVERVIEW OF SAFETY
Please refer to Dr. Ye Yang's memo for the statistical review of the clinical safety data.
9. ADDITIONAL STATISTICAL ISSUES
Not Applicable.
10. CONCLUSIONS
10.1 Statistical Issues and Collective Evidence
In the updated efficacy analysis for cases accrued during blinded placebo-controlled follow-up (cutoff date: March 13, 2021) of Study C4591001 in participants 16 years of age and older, the estimated vaccine efficacy (VE) against confirmed COVID-19
occurring at least 7 days after Dose 2 was 91.1% (95% CI: 88.8%, 93.1%), with 77 COVID-19 cases in the BNT162b2 group compared to 833 cases in the placebo group among participants without evidence of SARS-CoV-2 infection before and during the vaccination regimen; the estimated vaccine efficacy (VE) against confirmed COVID-19
occurring at least 7 days after Dose 2 was 90.9% (95% CI: 88.5%, 92.8%), with 81 COVID-19 cases in the BNT1622 group compared to 854 cases in the placebo group
Page 25

Statistical Review
STN: 125742/0
among participants with or without evidence of SARS-CoV-2 infection before and during the vaccination regimen.
With respect to efficacy against severe COVID-19 cases occurring at least 7 days after Dose 2, the estimated VE was 95.3% (95% CI: 71.0%, 99.9%), with 1 and 21 cases in the BNT1622 and placebo groups, respectively, among participants without evidence of SARS-CoV-2 infection; the VE result was the same among participants with or without evidence of SARS-CoV-2 infection.
10.2 Conclusions and Recommendations
Overall, the updated efficacy analysis results show that BNT1622 provided high VE in preventing symptomatic COVID-19 and severe COVID-19 cases that is consistent with the VE results reported in the interim and final analyses.
Page 26