hpv genotyping trial design

With the results of the GBI clinical trial due this month I took the time to do some research into the regulatory body requirements and clinical trial design considerations relevant to HPV typing.

A research group out of the US National Cancer Institute have published an academic paper discussing the clinical utility of HPV genotyping that can be accessed at this URL:

http://www.eurogin.com/MaJ/2006/ConsensusReport/HPV%20Genotyping.pdf

This paper details that

"Careful validation is often expensive, but there are no short cuts. Simple correlation studies of one assay versus another validated assay using convenience specimens do not equate to clinical performance. The FDA licensure typically requires demonstrated performance for the detection of CIN2+. For a test that is predicted to be 90% sensitive for CIN2+, a sample size of 134 cases of CIN2+ is needed to achieve a 95% confidence interval of 85%–95%; in a high-risk population with a 2% risk of CIN2+, this translates to a population sample size of 6700. For reproducibility, a few repeated tests within a batch cannot determine if there is assay drift over time, which can only be assessed with careful quality control measures. Without validation and the precautions of quality control and assurance, genotyping has the capability of causing more harm than benefit."

Given this, it will be interesting to see if the GBI trial includes the sample size of 6,700 this group describe as being required to establish the efficacy of a HPV typing test. Numbers less than 6,700 (or at least comparable) will most likely dent the confidence of potential partners performing due diligence on PapType.