From: Fiore, Cara
Sent: Wednesday, December 23, 2009 4:53 PM
Cc: Valenti, Elizabeth
Subject: Comments on PVP V59_34
Hi Chris, we have the following comments on the proposed pharmacovigilance plan (V59_34):
- Multiple EOIs originally proposed in the protocol synopsis have been dropped from the planned analysis (see Table 2 in Protocol V59_34 and the “Safety Objective” section of the Protocol Synopsis V59_34). Please provide your rationale for deleting the following EOIs that were present in your protocol synopsis: encephalitis, encephalopathy, brachial neuritis, optic neuritis, other neuropathy, transverse myelitis, ptosis, ataxia, acute disseminated encephalomyelitis, hemolytic anemia, myasthenia gravis, acute glomerulonephritis, nephrotic syndrome, and Graves’ disease.
- Please confirm that the chart reviewer who confirms the onset date of an EOI will be masked to exposure (i.e., Menveo receipt) status.
- In 4.1, we note that you will only include individuals in the analysis if they have been enrolled in the HMO for at least two years prior to the EOI. CBER disagrees with this approach and proposes that subjects should be included in the analysis as long as they have been enrolled in the HMO for six months prior to the EOI to minimize selection bias.
- In 7.7, you state that “Additional signals for testing could be identified for inclusion outside of the context of this study (e.g through VAERS).” Please confirm that signals identified by CBER or the sponsor through other means (e.g., VAERS) can prompt the same hypothesis testing analysis referenced in 7.5.1 that is undertaken for EOIs of borderline significance identified in your primary analysis of the HMO data. If so, the sentence should be revised to state, “If additional signals are identified outside the context of this study, they may be included in a hypothesis testing analysis according to the methodology specified for this study using an increased sample size focused on the new EOI.”
- As you note, the self controlled case series methodology proposed will not have sufficient power to detect a pre-specified EOI with a low annual incidence. CBER recommends including a secondary analysis, to be presented at the conclusion of the study, that will further characterize possible rare EOIs. EOIs included in this analyis will be those proposed in your protocol synopsis as well as any other EOIs identified outside the context of the protocol but included in hypothesis testing analysis. Subjects vaccinated during the study period (with and without EOIs) who were between age 11 and 19 at the time of vaccination and who were enrolled in the HMO for at least six months prior to vaccination will meet the simplified inclusion criteria for this analysis. The crude incidence of each EOI will be calculated from the total vaccinated population. Observed incidence for each EOI (within the exposure window) in this vaccinated cohort should then be divided by expected incidence in a comparable population based on estimates from the medical literature to yield another estimate of relative incidence.
- Please respond as soon as possible but no later than January 5, 2010 with specific responses to each numbered comment as well as a revised protocol with changes tracked.
Cara Fiore, Ph D
Division of Vaccines and Related Products Applications
Office of Vaccines Research & Review
Center for Biologics Evaluation & Research
U.S. Food & Drug Administration
WOC1 RM300N HFM-481
1401 Rockville Pike
Rockville, MD 20852