Document text
1
Evaluation of Guillain -Barré Syndrome (GBS)
following Respiratory Syncytial Virus (RSV)
Vaccination Among Adults 65 Years and Older
Dr. Patricia Lloyd, ScM PhD
Health Statistician
Office of Biostatistics and Pharmacovigilance
Center for Biologics Evaluation and Research
U. S. Food & Drug Administration
MEETING OF THE ADVISORY COMMITTEE ON IMMUNIZATION PRACTICES (ACIP)
Respiratory Syncytial Virus (RSV) Vaccine, Adults
June 26 – 28, 2024
2
Outline
•Introduction
•Observed vs. Expected Analysis Summary
▪Study Methods and Results
•Self-Controlled Case Series (SCCS) Analysis
▪Study Methods and Results
•Discussion
•Conclusion
3
Introduction
•Two* RSV vaccines were approved for use in the U.S. in adults 60 years and
older
▪RSVPreF3+AS01 (GSK - AREXVY)
▪RSVPreF (PFIZER - ABRYSVO)
•An imbalance in the rates of Guillain -Barré syndrome (GBS) between vaccine
and placebo recipients was identified in clinical trials supporting licensure of
RSV vaccines 1, 2
•FDA is conducting a post -licensure RSV vaccine safety study using two designs:
▪Observed vs. Expected Analysis
▪Self-Controlled Case Series (SCCS) Analysis
*mRNA -1345 (Moderna - mRESVIA ®) was approved on May 31, 2024.
4
Observed vs. Expected Analysis Summary
Methods
•Evaluated risk of GBS following one dose of either RSVPreF3+AS01
or RSVPreF vaccines using a retrospective cohort design with the
2022 historical comparator
•Estimated the observed incidence rates (IRs) and compared to
historical comparator (expected) rates, to obtain incidence rate
ratios (IRRs) with 95% confidence intervals (CIs)
•Estimation of GBS positive -predictive value (PPV) -adjusted rates is
based on multiple imputed datasets
▪Chart review, PPV for GBS: 71% (95% CI: 63%, 79%) 3
5
Observed vs. Expected Analysis Summary
Results
RSVPreF3+AS01 RSVPreF
Inferential Analysis Results
Observed vs. Expected Analysis 2.76 (95% CI: 1.32, 5.07) 6.94 (95% CI: 3.70, 11.87)
PPV-Adjusted Analysis 2.75 (95% CI: 0.46, 5.04) 6.91 (95% CI: 1.85, 11.97)
GBS Cases per 1 million Doses 10.0 25.1
Descriptive Analysis Results
Total RSV Vaccine Doses 2,061,602
RSV Vaccine Doses 1,379,335 682,267
Observed GBS cases <11 13
Data Through Date: Dec 02, 2023•An elevated IRR was observed for GBS following RSV vaccination
•Only RSVPreF association was statistically significantly elevated in PPV -adjusted analysis
6
•The observed vs. expected analysis is a crude method with
limited adjustments for confounding, utilizing aggregate
historical incidence rates rather than individual historical persons
as comparators, increasing the potential for confounding and
bias
•The SCCS is a robust method that controls for time -invariant
confounding and does not rely on historical background
incidence ratesMotivation for SCCS Study
7
Self-Controlled Case Series (SCCS) Design
•Population Exclusion Criteria:
▪Beneficiaries with no incident
GBS outcome in the
observation period
OR
▪Beneficiaries who do not
meet criteria to identify an
incident outcome**
Days 43-90 Days 1 -42
365 Days Prior
Study Start Date
May 2023Study End Date
Apr 2024
RSV Vaccination (Day 0)= Risk Interval
= Control Interval
= GBS Outcome
*The clean window is relative to the outcome date; risk and control intervals are relative to the vaccination date
**Outcomes that are considered ‘incident’ after implementing the outcome -specific cleaning window are included,
and only first incident outcome in the observation period are retained•Population Inclusion Criteria:
▪Enrolled in Medicare Fee -
for-Service (FFS) during the
clean window
▪65 years of age or older on
RSV vaccination date
▪No GBS outcome during the
clean window= Clean Window (365 days)*
8
SCCS Analysis: Study Methods
Study Design Self-Controlled Case Series (SCCS) 4, 5
Data Sources and
Study PopulationCenters for Medicare & Medicaid Services (CMS) Medicare Beneficiaries ages 65 years
and older, enrolled in:
▪Medicare FFS (Parts A and B) and Part D on the date of RSV vaccination
▪Medicare FFS and in 1 -year prior
Study Period May 2023* – April 2024 ⬧**
GBS Outcome
Definition▪Risk Interval: 1 - 42 days
▪Control Interval: 43 - 90 days
▪Care Setting: inpatient – primary position only; ICD -10-CM DGN G61.0
Statistical
Analyses▪IRRs with 95% CIs
▪Absolute Risk: Attributable Risk (AR) with 95% CIs per 100,000 doses and 100,000
person -years (PY)
▪Adjustment for outcome -dependent observation time (Farrington) 6, seasonality, PPV
*FDA approval dates for RSVPreF3+AS01 and RSVPreF were May 3, 2023 and May 31, 2023, respectively
⬧Study end date for initial SCCS analysis was April 6, 2024
**RSV vaccinations prior to October 8, 2023 to have complete observation in 90 days post -vaccination and is expected to have 90% or greater data -completeness
9
SCCS Analysis: Vaccination Uptake Trends
020,00040,00060,00080,000100,000120,000140,000160,000180,000200,000Number of RSV Vaccinations
Calendar WeekWeekly Vaccination Uptake Trends in RSV Vaccines, By Vaccine Type
RSVPreF RSVPreF3+AS01
Vaccination Cut -off Date to ensure complete observation of 90 -day window
Data Through Date: Apr 06, 2024
10
Case Population Eligibility
CriteriaRSV Vaccinations
(n=1.33 M individuals; 1.33 M doses)*
RSVPreF3+AS01
(n= ~872k doses)*RSVPreF
(n= ~456k doses)*
Total GBS cases and total number
of days in study period160 cases [339 days] 92 cases [311 days]
GBS cases during 90 -day
observation period 105 74
Incident GBS cases after applying
clean window restriction55 36
GBS cases qualifying for SCCS
analysis
(vaccinated before Oct 8, 2023)11 17Case Counts for GBS following RSV vaccination by Vaccine Type
* n = Medicare Beneficiaries that received RSV vaccination and eligible for SCCS analysis are presentedSCCS Analysis: Descriptive Results
Data Through Date: Apr 06, 2024
11
Data Through Date: Apr 06, 2024SCCS Analysis: Results for GBS
Most Adjusted
Analysis
Least
Adjusted
Analysis
SCCS Analysis including most adjustments are highlighted in Blue
Farrington -Adjusted Analysis = Outcome -Dependent Observation Time AdjustmentIRR with 95% CI of GBS following RSV Vaccination Adjusted for Combinations of PPV,
Seasonality with Outcome -Dependent Observation Time
•An elevated IRR was observed for GBS following RSVPreF vaccination with two analyses that had the least adjustments
•Results additionally adjusted for PPV were no longer statistically significant
▪PPV, Seasonality with Farrington -Adjusted Analysis: 4.48 (95% CI: 0.88, 22.90) ▪PPV, with Farrington -Adjusted Analysis: 3.96 (95% CI: 0.77, 20.28)
Seasonality,
Farrington
FarringtonPPV, FarringtonPPV, Seasonality,
FarringtonAnalyses
12
SCCS Analysis: Results for GBS and
RSV vaccination
Inferential Analysis Results RSVPreF3+AS01 RSVPreF
Eligible Vaccinees 872,068 456,107
Cases in the Risk and Control
Intervals<11 12.1
IRR (95% CI) 2.30 (0.39, 13.72) 4.48 (0.88, 22.90)
AR per 100,000 Doses (95% CI) 0.32 ( -0.30, 0.95) 1.57 (0.30, 2.85)
AR Per 100,000 PY (95% CI) 2.81 ( -2.64, 8.26) 13.69 (2.59, 24.79)Incidence Rate Ratio (IRR) and Attributable Risk (AR) of GBS - Adjusted for PPV, Seasonality and
Outcome -Dependent Observation Time
Data Through Date: Apr 06, 2024
13
•SCCS study design provides
robust adjustment for potential
time-invariant confounding
•Large database facilitates
more precise evaluation of
health outcomes
•Study findings are
generalizable to U.S.
population 65 years and olderDiscussion
Strengths Limitations
•Potential outcome
misclassification
•Potential misspecification of
risk and control intervals
•Potential for residual
confounding
14
Discussion
Observed vs. Expected Analysis
•An elevated IRR was observed for GBS following RSV vaccination,
but only RSVPreF association was statistically significantly elevated
PPV (chart review) adjustment
•Crude method that utilized aggregate historical comparator rates
rather than individuals, increasing the potential for confounding
•Statistically significant results of GBS do not establish a causal
association between RSV vaccines and GBS
15
Discussion
SCCS Analysis
•Although, an elevated IRR was observed for GBS following
RSVPreF vaccination for two analyses that had fewer adjustments,
the results were not statistically significant when adjusted for PPV
•Only cases, i.e., persons with an incident outcome contribute to the
SCCS analysis
•Estimation of outcome risk occurs within, rather than between
individuals, adjusting for time -invariant confounding
16
Conclusion
•The results from two different types of analyses of potential GBS risk
following RSV vaccination are mixed and highly uncertain
•These analyses do not provide clear, conclusive evidence of an elevated
risk of GBS and an elevated risk cannot be ruled out at this time
•FDA is conducting medical chart review on GBS cases and will continue to
evaluate the safety of RSV vaccines as more data are available
•FDA maintains that the b enefits of RSV vaccination in preventing RSV
hospitalizations outweigh the potential risks associated with the vaccines
References
17
1.Respiratory Syncytial Virus Vaccine Recombinant, Adjuvanted ( Arexvy ). Vaccines and Related Biological
Products Advisory Committee Meeting. FDA Briefing Document. March 1, 2023.
https://www.fda.gov/media/165622/download
2.Respiratory Syncytial Virus Stabilized Bivalent Prefusion F Subunit Vaccine ( Abrysvo ). Vaccines and
Related Biological Products Advisory Committee Meeting. FDA Briefing Document. March 1, 2023.
https://www.fda.gov/media/165625/download
3.Arya,D.P ., et al. Surveillance for Guillain -Barré syndrome after 2015 -2016 and 2016 -2017 influenza
vaccination of Medicare beneficiaries. Vaccine, 2019. 37(43): p. 6543 -6549.
4.Petersen I, Douglas I, Whitaker H. Self controlled case series methods: an alternative to standard
epidemiological study designs. 2016;354:i4515.
5.Evaluation of Multiple Safety Outcomes following RSV Vaccination in Adults 60 Years and Older.
https://bestinitiative.org/wp -content/uploads/2024/01/BEST_RSV_Safety_Older_Adults_2023 -2024.pdf
6.Farrington, C. P., Anaya -Izquierdo , K., Whitaker, H. J., Hocine , M. N., Douglas, I., & Smeeth, L. (2011). Self -
Controlled Case Series Analysis With Event -Dependent Observation Periods. Journal of the American
Statistical Association, 106(494), 417 –426. https://doi.org/10.1198/jasa.2011.ap10108
Acknowledgements
18
Steven Anderson
Richard Forshee
Narayan Nair
Joann Gruber
Carla Zelaya
Tainya Clarke
FDA Partners:
Acumen, CMS
Questions?
19