Record of Telephone Conversation - Cervarix, June 7, 2007
Record of Telephone Conversation
STN Number: 125259/0
Manufacturer: GlaxoSmithKline Biologicals
Product: Human Papillomavirus Vaccine, ASO4 Adjuvant-Adsorbed
From: Helen S. Gemignani, Regulatory Project Manager
Subject: Discussion of –-b(4)--: Biopsy results data
Telecon Date: June 7, 2007 Telecon Time: 9:30 – 10:30 am
Telecon Initiated by: CBER
Contact Phone: 610 787-3726 Mr. Whitman
Summary of Telecon
CBER and GSK held a telecon to discuss responses to questions CBER had sent on 07-May-2007, regarding the --b(4)--: Biopsy results dataset. CBER had requested additional information regarding ascertainment of CIN 2 cases identified as endpoints in Protocol HPV-008. CBER was not able to reproduce the total number of CIN 2+ cases as noted in Table 31, p. 109, of Summary of Clinical Efficacy, nor the 23 cases identified in naïve subjects. These follow-up questions were sent to the sponsor on 05-June-2007. The GSK participants involved in the telecon were blinded as to PID number, but they did discuss the process.
In response to Item #3, Dr. Spiessens explained the process of identifying CIN 2+ cases. He explained that DIAGNOSI included the CIN 2+ cases as identified by the pathology panel, and if CIN 2+ was noted, the specimen was sent to -b(4)- for PCR testing. He further explained that cases that were read as CIN 1 by the pathology panel (identified by the DIAGNOSI variable), the specimen was referred to -b(4)- for further tissue cuts. If on further cuts of the tissue, CIN 2+ was identified, PCR testing was done at -b(4)-. The pathology slides, without PCR identification, were sent back to the pathology panel at -b(4)-for their diagnosis. If 2/3 agreed that the new cuts included CIN 2+, the case was added to the totals. This is the reason why we could not identify all cases represented in Table 31, p. 109, of the Summary of Clinical Efficacy. The EP_PROP variable identifies contains all CIN 2+ cases included in Table 31. The HR_PROP variable contains CIN 2+ cases with a HR HPV type (must have valid HPV test), and this specimen must have the PCR between the 2 slices with CIN 2+ as read by the pathology panel and -b(4)-. GSK also indicated that PCR was run on the deepest cut (S3 level). Baseline HPV-16 /-18 seropositivity also needs to be taken into account. Because of the complexity of this procedure, Dr. Lee requested that a written explanation be sent to CBER for further evaluation, but both Dr. Miller and Dr. Lee will re-review the datasets in order to see if they can reproduce the numbers with this knowledge.
Dr. Miller then asked about Item #4 (in which some CIN 2+ samples did not include PCR results). Of the 19 cases found on CBER’s review without PCR results, nine had not yet been run. CBER was concerned that these specimens may contain additional cases in naïve subjects due to HPV 16 or 18 that were not yet identified. Dr. Spiessens noted that the interim analysis was event driven, and they conducted their analysis at the time 23 cases were identified and had gone through the process to identify endpoints with full characterization. The nine cases noted by CBER represented cases that had not completed the full process because these were still in progress at the time the database was locked. Dr. Miller indicated CBER would not want to cause a problem with the final analysis, and will examine these cases further. Dr. Spiessens also answered a question as to the meaning of the “other slide” identified for four subjects. This designation includes S1 and S2 slides. S3 is the deepest cut and the one in which PCR was run.
Dr. Miller indicated that after learning the above information, she was going to recheck the other subject she had queried as perhaps being a CIN 2+ case with HPV 16 and 52 and appeared to be naïve for HPV 16 at baseline with a negative Pap test. Dr. Spiessens indicated that the unblinded statistician had provided them information that the CIN 2+ lesion only contained HPV 52, and not 16. That subject had been infected with HPV 52 at baseline and throughout, and this made sense. Dr. Miller and Dr. Lee thanked the sponsor for clarifying the process, and this made things clearer. CBER will go back and re-look at the datasets. Should they any more questions, there may be follow-up phone calls.
Dr. Raychaudhuri thanked the GSK staff for their assistance. She requested that their responses be sent in writing.
End of Teleconference
- Nancy Miller
- Martha Lee
- Helen Gemignani
- Gopa Raychaudhuri
- Vince Ahonkai
- Matt Whitman
- Bart Spiessens
- Toufik Zahaf