The Evolving History of Influenza Viruses and Vaccines
The Evolving History of Influenza Viruses and Vaccines
The characteristics of influenza virus B strains are slightly different from those of influenza A strains. It was generally believed, a few decades ago, that diseases arising from influenza B infections:
• Were less severe than those arising from influenza A infections;
occurred mainly as isolated cases or clusters, but not as epidemics;
• Were caused by viruses less likely to undergo antigenic variations than A viruses.
Can we say that this is still true today? Probably not, because there have been several reports of diseases caused by influenza B virus in all age groups that have been as severe as those caused by an influenza A virus and with similarly serious complications. In addition, although it is true that influenza B viruses have never been reported to be responsible for pandemics, they have been reported to be the cause of localized epidemics, particularly in closed communities and they have been the prevalent viruses during certain seasons. Thus, during the 2005–2006 winter in Europe, 59% of the isolated influenza viruses were B strains. Influenza B infections often occur later in the winter after the end of the annual epidemic of influenza A. Finally, the B virus does seem to be more stable than the A strain; the isolated B variants are not as different from the previous B strain as A subtypes from the original A strain.
However, since 1940, there has been a succession of drifted strains and the vaccine has often been modified to reflect these changes. Over the last 25 years, there have been two lineages of influenza virus B, which have been responsible for local epidemics (Figure 1). In 1987, B/Victoria/2/87 virus replaced the previous B virus throughout the world. This Victoria virus became the dominant B strain for several years before the emergence of B/Yamagata/16/88, which appeared in Japan in the 1990s. For nearly 10 years, the Yamagata strain was the dominant B strain throughout the world. In 2002, the Victoria strain reappeared and was in competition with the Yamagata strain. Unexpectedly, the new virus did not replace the previous one as has often happened with influenza A virus and the two strains started to cocirculate (Figure 1). Since then, the dominant strain varies according to the country and the season. These observations show that it is difficult to predict which strain will appear during the following season, in particular because the two strains can circulate simultaneously during the same season and even within the same country. Since the current vaccine contains only one influenza B strain, it is obvious that the mismatch between the vaccine and the circulating strain can occur often.
WHO recommendations have made attempts to provide guidance about what strain B should be selected, depending on the epidemiological situation. However, their choice has not always reflected the circulating strain the following season. Between 2000 and 2013, there were eight changes for the vaccine type B virus to be included in the vaccine compared with seven changes for the subtype A/H3N2 and four changes for A/H1N1 (Figure 3). In some regions, there was a perfect match, whereas in others, the match was not as good. Between 2002 and 2007, examples of mismatch were reported in various countries; the mismatched vaccine often provided only partial protection ( Table 1 ). For example, during the 2004–2005 season, when the WHO recommendations had proposed the inclusion of a B/Yamagata lineage virus in the vaccine, the B/Victoria was dominant in Italy and Taiwan (83% and 64% mismatch, respectively), whereas in the USA and Hong Kong, the circulating strains were related to the B/Yamagata strain (28% and 0% mismatch, respectively) ( Table 1 ). Data from the USA (CDC) show that the vaccine and the circulating strains were well matched for influenza virus B for only 5 years from 1999 to 2010.
This has led to proposals to systematically include two B strains in addition to the two A strains in the usual composition of the seasonal vaccines to overcome the unpredictable circulation of the B strain and to offer potentially a wider protection. A quadrivalent vaccine, currently licensed in the USA, could offer advantages from both the public health and economic perspectives. On 20 February 2013, WHO recommended the following strains for inclusion in the seasonal vaccine: A/California/7/2009, analogous to (H1N1)pdm09; an A(H3N2) A/Victoria/361/2011; and a B/Massachusetts/2/2012-like virus from the Yamagata lineage. The recommendations included a strain from the Victoria lineage (B/Brisbane/60/2008-like virus) in a quadrivalent vaccine, leaving it up to each country/state to decide to use either a trivalent or quadrivalent vaccine.
New Developments for Influenza Virus B Vaccines
The characteristics of influenza virus B strains are slightly different from those of influenza A strains. It was generally believed, a few decades ago, that diseases arising from influenza B infections:
• Were less severe than those arising from influenza A infections;
occurred mainly as isolated cases or clusters, but not as epidemics;
• Were caused by viruses less likely to undergo antigenic variations than A viruses.
Can we say that this is still true today? Probably not, because there have been several reports of diseases caused by influenza B virus in all age groups that have been as severe as those caused by an influenza A virus and with similarly serious complications. In addition, although it is true that influenza B viruses have never been reported to be responsible for pandemics, they have been reported to be the cause of localized epidemics, particularly in closed communities and they have been the prevalent viruses during certain seasons. Thus, during the 2005–2006 winter in Europe, 59% of the isolated influenza viruses were B strains. Influenza B infections often occur later in the winter after the end of the annual epidemic of influenza A. Finally, the B virus does seem to be more stable than the A strain; the isolated B variants are not as different from the previous B strain as A subtypes from the original A strain.
However, since 1940, there has been a succession of drifted strains and the vaccine has often been modified to reflect these changes. Over the last 25 years, there have been two lineages of influenza virus B, which have been responsible for local epidemics (Figure 1). In 1987, B/Victoria/2/87 virus replaced the previous B virus throughout the world. This Victoria virus became the dominant B strain for several years before the emergence of B/Yamagata/16/88, which appeared in Japan in the 1990s. For nearly 10 years, the Yamagata strain was the dominant B strain throughout the world. In 2002, the Victoria strain reappeared and was in competition with the Yamagata strain. Unexpectedly, the new virus did not replace the previous one as has often happened with influenza A virus and the two strains started to cocirculate (Figure 1). Since then, the dominant strain varies according to the country and the season. These observations show that it is difficult to predict which strain will appear during the following season, in particular because the two strains can circulate simultaneously during the same season and even within the same country. Since the current vaccine contains only one influenza B strain, it is obvious that the mismatch between the vaccine and the circulating strain can occur often.
WHO recommendations have made attempts to provide guidance about what strain B should be selected, depending on the epidemiological situation. However, their choice has not always reflected the circulating strain the following season. Between 2000 and 2013, there were eight changes for the vaccine type B virus to be included in the vaccine compared with seven changes for the subtype A/H3N2 and four changes for A/H1N1 (Figure 3). In some regions, there was a perfect match, whereas in others, the match was not as good. Between 2002 and 2007, examples of mismatch were reported in various countries; the mismatched vaccine often provided only partial protection ( Table 1 ). For example, during the 2004–2005 season, when the WHO recommendations had proposed the inclusion of a B/Yamagata lineage virus in the vaccine, the B/Victoria was dominant in Italy and Taiwan (83% and 64% mismatch, respectively), whereas in the USA and Hong Kong, the circulating strains were related to the B/Yamagata strain (28% and 0% mismatch, respectively) ( Table 1 ). Data from the USA (CDC) show that the vaccine and the circulating strains were well matched for influenza virus B for only 5 years from 1999 to 2010.
This has led to proposals to systematically include two B strains in addition to the two A strains in the usual composition of the seasonal vaccines to overcome the unpredictable circulation of the B strain and to offer potentially a wider protection. A quadrivalent vaccine, currently licensed in the USA, could offer advantages from both the public health and economic perspectives. On 20 February 2013, WHO recommended the following strains for inclusion in the seasonal vaccine: A/California/7/2009, analogous to (H1N1)pdm09; an A(H3N2) A/Victoria/361/2011; and a B/Massachusetts/2/2012-like virus from the Yamagata lineage. The recommendations included a strain from the Victoria lineage (B/Brisbane/60/2008-like virus) in a quadrivalent vaccine, leaving it up to each country/state to decide to use either a trivalent or quadrivalent vaccine.