During 1998, severe outbreaks of influenza were observed in four swine herd
s in the United States. This event was unique because the causative agents,
H3N2 influenza viruses, are infrequently isolated from swine in North Amer
ica. Two antigenically distinct reassortant viruses (H3N2) were isolated fr
om infected animals: a double-reassortant virus containing genes similar to
those of human and swine viruses, and a triple-reassortant virus containin
g genes similar to those of human, swine, and avian influenza viruses (N. N
. Zhou, D. A. Senne, J. S. Landgraf, S. L. Swenson, G. Erickson, K. Rossow,
L. Liu, K.-J. Yoon, S. Krauss, and R. G. Webster, J. Virol. 73:8851-8856,
1999). Because the U.S. pig population was essentially naive in regard to H
3N2 viruses, it was important to determine the extent of viral spread. Hema
gglutination inhibition (HI) assays of 4,382 serum samples from swine in 23
states indicated that 28.3% of these animals had been exposed to classical
swine-like H1N1 viruses and 20.5% had been exposed to the triple-reassorta
nt-like H3N2 viruses. The HI data suggested that viruses antigenically rela
ted to the double-reassortant H3N2 virus have not become widespread in the
U.S. swine population. The seroreactivity levels in swine serum samples and
the nucleotide sequences of six additional 1999 isolates, all of which wer
e of the triple-reassortant genotype, suggested that H3N2 viruses containin
g avian PA and PB2 genes had spread throughout much of the country. These a
vian-like genes cluster with genes from North American avian viruses. The w
orldwide predominance of swine viruses containing an avian-like internal ge
ne component suggests that these genes may confer a selective advantage in
pigs. Analysis of the 1999 swine H3N2 isolates showed that the internal gen
e complex of the triple-reassortant viruses was associated with three recen
t phylogenetically distinct human-like hemagglutinin (HA) molecules. Acquis
ition of HU genes from the human virus reservoir will significantly affect
the efficacy of the current swine H3N2 vaccines. This finding supports cont
inued surveillance of U.S. swine populations for influenza virus activity.