Establishing Potency Specifications for Antigen Vaccines Clinical Validation of Statistically Derived Release and Stability Specifications

LEVEL: ADVANCED G ardasil (a registered trademark of Merck and Co., Inc., www. merck.com) is the first vaccine approved for women aged 9–26 years old in prevention of cervical cancer and genital warts as well as vulvar and vaginal precancerous legions. The vaccine contains noninfectious virus-like particles (VLPs) corresponding to HPV types 6, 11, 16, and 18. It is produced by recombinantly expressing the major HPV capsid protein, L1, for each type in yeast (1, 2). The L1 monomers self-assemble to produce icosahedral VLPs that are structurally similar to the native virion in size, assembly, and immunological properties (3–5). The VLPs are purified chromatographically, diluted, and then adsorbed onto Merck’s aluminum adjuvant to produce monovalent bulks (1, 2). A single dose contains 20, 40, 40, and 20 μg of VLP types 6, 11, 16, and 18, respectively, formulated on 225 μg of the aluminum adjuvant and administered in a final volume of 0.5 mL. Each bulk lot is formulated based on protein concentration. Four monovalent bulks, one per HPV type, are subsequently diluted and blended to produce the final container material. The dose is fixed for each type and based on mass; thus, there is minimal lot-to-lot variation in protein concentration. Because the vaccine’s potency depends on the specific activities (specific antigenicities) of each type-specific VLP source bulk, some variation in potency among final containers is anticipated. An enzyme-linked immunoassay, referred to as the in vitro relative potency assay (IVRP), is used as the potency test for the Gardasil vaccine (6). It measures the amount of antibodies bound to neutralizing epitopes for each HPV type (7–10). Results are reported relative to a Gardasil lot that was used in a phase 3 clinical trial. The assay, therefore, provides a direct comparison of the antigen content of each VLP type in a given test sample and the corresponding antigen content of a lot that has been shown to be efficacious in humans. IVRP results correlate with immunogenicity results obtained using a traditional mouse potency test and are considered predictive of immunogenicity in humans (6). At present, there is no immune correlate of protection for HPV. In the phase 2/3 clinical trial program, prophylactic vaccination was highly efficacious in preventing infection and disease caused by HPV types 6, 11, 16, and 18 for at least five years (11). No breakthrough cases due to waning immunity have been described (12–14). Because there is no immune correlate, and only a limited number of final container lots were manufactured before licensure, a novel approach for establishing potency specifications was developed and applied to the Gardasil vaccine. Preliminary specifications were derived using a propagation-of-error calculation starting from the IVRP values of bulk production lots. The statistically derived specifications were Three-dimensional reconstruction of a human papillomavirus (L1 type 11) virus-like particle, a component of Gardasil®.