**3. The genomic approach to identifying novel biomarkers of influenza vaccine safety**

The search for new biomarkers that can reflect the bioactivity assessed in the ATT and LTT was conducted by performing comprehensive gene expression analyses on major organs via the microarray technology [25]. Inactivated influenza vaccines have been widely used for preventing infections and the spread of infections; they can roughly be subdivided into two types: the HAV and WPV [30]. The HAV mainly contains hemagglutinin (HA). This type of vaccine has no strong bioactivity; it does not contain substances other than HA proteins that act as antigens, thereby leading to no adverse reactions. Nevertheless, their ability to induce antibody production is considered insufficient to prevent the progression of influenza virus infection [30]. Historically, however, vaccines have been more effective than the existing split vaccines. The WPV is considered effective against influenza virus infections. This type of vaccine contains the whole influenza virus particle, including lipid and single-stranded RNA, and therefore, drives various immune responses. On the other hand, various WPVinduced immune responses also cause adverse reactions in humans [29]. Therefore, although WPV is a highly effective vaccine, it has lately not been employed as a seasonal influenza vaccine and is only partially manufactured as a pandemic influenza vaccine. We have carried out the searches for safety assessment marker genes of the HAV using two types of vaccines: the WPV and split influenza vaccine (SV). The WPV has high reactogenicity (effectiveness and toxicity) and therefore serves as a toxicity reference. The SV has low reactogenicity and frequency of adverse reactions and is therefore employed as a safety control. As a result, the clearest clustering of gene expression patterns in the lungs of animals by different types of vaccines was obtained [25]. In particular, the gene expression patterns in the lungs differed between the SV-treated and WPV-treated animals. Furthermore, the gene expression levels, which showed large differences between the SV- and WPV-treated animals, were estimated. As a result, 18 genes expressed in the lungs were identified as biomarker genes (**Table 1**) [25]. Functionally, these biomarker genes tend to correlate with the white blood cell (WBC) count in the peripheral blood of animals (leukopenic toxicity) that were treated with the inactivated influenza vaccine [31, 32]. It has been known that intraperitoneal injection of the WPV causes leukopenic toxicity in mice, and this bioactivity has been used as an index for the safety evaluation of inactivated influenza vaccines in Japan, which is termed the LTT [19]. In addition, WPV-induced body weight loss may be reflected in biomarker gene expression levels [25]. Thus, WPV-like bioactivity assessed by the ATT could be predicted by the expression levels of biomarker genes. Furthermore, the biomarker genes can partially measure biological activities that cannot be quantitated by means of body weight changes. For example, the identified biomarker gene expression increases with leukopenia and weight loss; however, some genes show increased expression even in a state without body weight loss and without a decreased WBC count (without leukopenic toxicity) [33]. The SV hardly induces the elevated expression of biomarker genes. By contrast, if SVs produced at different manufacturing plants were evaluated using the biomarker gene expression levels, variations in their quality can be reflected in the expression levels [33]. This variation is not indicated by body weight changes or by the WBC count [33]. This case is an example where the gene expression level can detect biological reactions that cannot be detected by phenotypic changes. This is the advantage of the toxicogenomics technology. Thus, it is likely that the genomics technology is also useful for the safety assessment of vaccines. Furthermore, there is a possibility that the assay involving the newly identified biomarker genes can be widely adopted as an

Genomic Approaches Enable Evaluation of the Safety and Quality of Influenza Vaccines…

http://dx.doi.org/10.5772/intechopen.76361

117

alternative method to the currently popular methods: the ATT and LTT.

**gene expression**

produced in different batches.

**4. Safety evaluation of influenza vaccines on the basis of biomarker** 

The utility of the identified biomarker genes has been verified. For seasonal influenza vaccines, the ATT has been regarded as a test for safety and quality control. Therefore, a safety assessment of SVs manufactured at four manufacturing plants was conducted by means of biomarker gene expression and body weight as parameters (ATT) and by the LTT, with a WPV as a control [33]. With respect to phenotypic changes, body weight loss rates of all the SVs were found to be equivalent, and leukocyte number reduction was hardly observed for the HAVs from all the manufacturers. Nevertheless, in case of one manufacturer's HAV, analyses of the expression of 18 biomarker genes in lungs showed a significant difference in gene expression levels from other manufacturers' HAV [33]. This result suggests that the biomarker genes identified by the microarray analysis can capture biological changes that cannot be detected by body weight changes and leukocyte number reductions. This finding indicates that the analysis of expression of biomarker genes is a more sensitive assay than the conventional safety and quality control tests (ATT and LTT). This evaluation method can be applied not only to predict the toxicity but also to evaluate the homogeneity among vaccines

Subsequently, the safety assessment of trivalent virosome-type influenza vaccine (Inflexal Berna V) currently licensed in several European countries such as Switzerland and Italy was


**Table 1.** Marker genes for safety evaluation of influenza vaccines.

Functionally, these biomarker genes tend to correlate with the white blood cell (WBC) count in the peripheral blood of animals (leukopenic toxicity) that were treated with the inactivated influenza vaccine [31, 32]. It has been known that intraperitoneal injection of the WPV causes leukopenic toxicity in mice, and this bioactivity has been used as an index for the safety evaluation of inactivated influenza vaccines in Japan, which is termed the LTT [19]. In addition, WPV-induced body weight loss may be reflected in biomarker gene expression levels [25]. Thus, WPV-like bioactivity assessed by the ATT could be predicted by the expression levels of biomarker genes. Furthermore, the biomarker genes can partially measure biological activities that cannot be quantitated by means of body weight changes. For example, the identified biomarker gene expression increases with leukopenia and weight loss; however, some genes show increased expression even in a state without body weight loss and without a decreased WBC count (without leukopenic toxicity) [33]. The SV hardly induces the elevated expression of biomarker genes. By contrast, if SVs produced at different manufacturing plants were evaluated using the biomarker gene expression levels, variations in their quality can be reflected in the expression levels [33]. This variation is not indicated by body weight changes or by the WBC count [33]. This case is an example where the gene expression level can detect biological reactions that cannot be detected by phenotypic changes. This is the advantage of the toxicogenomics technology. Thus, it is likely that the genomics technology is also useful for the safety assessment of vaccines. Furthermore, there is a possibility that the assay involving the newly identified biomarker genes can be widely adopted as an alternative method to the currently popular methods: the ATT and LTT.
