Optimization and qualification of a quantitative reversed-phase HPLC method for hemagglutinin in influenza preparations and its comparative evaluation with biochemical assays.

Abstract A previously described reversed-phase HPLC (RP-HPLC) method based on fast separations on a non-porous silica stationary phase [1] was optimized and qualified for the quantitative determination of hemagglutinin (HA) in influenza vaccine preparations. Optimization of the gradient elution conditions led to improved separation of the HA1 subunit from other vaccine constituents. The sensitivity of the method was significantly increased by using native fluorescence detection, resulting in an approximately 10-fold increase as compared to UV–vis detection. This enabled the elimination of the concentration step described in the original method and allowed direct analysis of vaccine preparations. The method was qualified for linearity, range, limit of detection, limit of quantitation and precision. Overall, it was found to be linear over the range of 2.5–100 μg HA/mL for all subtypes examined. This range covered 50–150% of the concentration found for individual strains in seasonal influenza vaccines and in the pandemic H1N1 vaccine. The limit of detection and limit of quantitation for each subtype were found to be suitable for the method's intended purpose and compared well to values found by the single radial immunodiffusion (SRID). The repeatability of the method gave RSD values below 5% for both retention time and peak areas. As expected for intermediate precision, larger RSD values for peak area were obtained but were below 10% and deemed acceptable. The RP-HPLC results were compared to Western blot analysis using a HA universal antibody for a set of 15 monovalent A/California H1N1 preparations and showed good correlation. Similarly, the quantitative nature of the RP-HPLC method was assessed in relation to the SRID assay currently used for the determination of the HA content in bulk antigen and final vaccine preparations. Thus, for a series of 23 monovalent A/Brisbane/59/2007 H1N1 bulks, ranging between 12.7 and 15.9 μg HA/mL by SRID, the RP-HPLC values were found to be in very good agreement, ranging between 11.9 and 14.1 μg HA/mL ( n  = 5) for five determinations carried out on 5 different days. During the 2009–10 H1N1 influenza pandemic the quantitative RP-HPLC method was used alongside several other test methods for the analysis of pandemic H1N1 vaccine preparations that included bulk antigen and final vaccines. The HA content of vaccines formulated at 15 or 30 μg/mL was measured by RP-HPLC and SRID and results showed that the HA content determined by RP-HPLC correlated well to that determined by SRID and to values determined by Western blot. Overall, the results provided further evidence of the usefulness of RP-HPLC for the detection and quantitation of the HA content once a reference standard has been established.