To: Mr. Matt Whitman (GSK)
File: STN 125259/0
RE: Your May 29, 2007 Responses to CBER”s May 7, 2007 comments
Date: June 5, 2007
We have the following comments regarding your May 29, 2007, responses to our May 7, 2007, questions (Items 3 and 4) in regard to the “eppcon: Biopsy results” datasets for protocol 008 submitted to the BLA.
Item #3: Because of a discrepancy in the number of all CIN 2+ cases identified by CBER as compared to the total number of CIN 2+ (p. 126 Clinical Overview) identified in the TVC-1 population by GSK, CBER had requested clarification of the variable in the epccon dataset which identified study endpoint CIN 2+ cases as assessed by the study panel. The variable you identified as designating endpoint CIN 2+ cases as assessed by the study pathology panel was 'DIAGNOSI'. When we rechecked the additional cases we had not identified in our first search of the database, it was noted that except for subject # 8374, the additional cases you identified (PIDs 1148, 2229, 2652, 3955, 4249, 4451, 7850, 8131, 8326, 10800, 11723, 11942, 12390, 12553, 14542, 19613, 19620, 5780, 11871, 11985, 12106, 12611, 12646, 14099, and 70323) had diagnoses of CIN 1 by the DIAGNOSI variable. On further review of those subjects' biopsy results, a diagnosis of CIN 2 was found in either the HR_PROP or ER_PROP variable. Please discuss exactly how the total number of CIN 2 cases (120 in the HPV group and 112 in the HAV group) were identified and are represented in the datasets.
Item #4: CBER requested you provide PCR results for 19 subjects in which CIN 2+ was identified by the DIAGNOSI variable in the dataset epccon: Biopsy results but not found by CBER on their search, or alternative, an explanation as to why the result was not available. For 9 of the 19 subjects (1154, 1789, 2290, 85028, 21950, 73982, 12554, 11848, and 5985), the more recent date of the biopsy was the explanation for lack of results. CBER had concern that all subjects should have had testing for at least HPV 16 and 18 in the CIN 2 lesions identified, especially those who were naïve to HPV 16 and/or 18 [SN and PN] and were Pap negative at baseline (21950 and 73982 in the HPV group and 1154, 1789, and 2290 in the HAV group). Please discuss. Also in Item #4, please explain the designation "other slide" for subjects 1308, 7729, 13355, and 15894.
Subject 76478 (who received HPV vaccine) was noted to be naïve for HPV 16 and 18 at baseline with a negative Pap test, but HPV 16 and 52 were noted in a CIN 2 biopsy specimen in epccon dataset on my review. Please present your characterization for this subject (baseline Pap and PCR status for 16 and 18 and any other HPV types at baseline, and HPV types found in the CIN 2 lesion) and provide an explanation as to why this subject was not included as a CIN 2 case in the efficacy analysis.