Genotype GI.6 Norovirus, United States, 2010–2012

We report an increase in the proportion of genotype GI.6 norovirus outbreaks in the United States from 1.4% in 2010 to 7.7% in 2012 (p<0.001). Compared with non-GI.6 outbreaks, GI.6 outbreaks were characterized by summer seasonality, foodborne transmission, and non–health care settings.


The Study
Since 2009, the Centers for Disease Control and Prevention has operated 2 surveillance systems for norovirus outbreaks in the United States: CaliciNet and the National Outbreak Reporting System (NORS). CaliciNet is an electronic laboratory surveillance network that collects information on genetic sequences of noroviruses implicated in outbreaks (5). As of 2011, public health laboratories in all 50 states are either certified members of CaliciNet or submit norovirus-positive specimens to 1 of 5 regional Ca-liciNet Outbreak Support Centers. NORS is an electronic surveillance system for reporting all enteric disease outbreaks, regardless of etiology or mode of transmission (3). Data reported in NORS include outbreak characteristics, demographics, symptoms, implicated exposures, clinical outcomes, and etiologies We identified GI.6 outbreaks with a first illness onset date during January 1, 2010-December 31, 2012, from CaliciNet and linked them to NORS by using unique outbreak identification numbers. Supplemental information derived from NORS included mode of transmission, outbreak setting, and patients' demographic features and clinical outcomes. State health departments were queried about outbreaks that could not be linked to NORS and requested to provide such supplemental information directly. Annual variation in GI.6 outbreaks was assessed by χ 2 test for trend, and GI.6 seasonality was identified on the basis of visual examination of trends over time and compared with non-GI.6 seasonality by using Mid-P exact test. Noroviruspositive specimens were typed by using region D sequence analysis (5) (Figure 1 The most commonly identified mode of transmission was person-to-person, which occurred in 81 (57.4%) GI.6 outbreaks ( Table 1). Foodborne transmission was more frequent among GI.6 than among non-GI.6 outbreaks reported to CaliciNet during the same period (rate ratio [RR] 1.77, 95% CI 1.25-2.51). Waterborne transmission also was more common in GI.6 outbreaks; the 2 waterborne GI.6 outbreaks occurred during June and July.
GI.6 outbreaks accounted for 4,375 reported illnesses, with a median of 22 (range 2-178) reported illnesses per outbreak. Supplementary demographic and clinical outcome information was available for 66 (46.8%) outbreaks, comprising 2,220 reported illnesses. Data on hospitalization and death were provided for most (>79.0%) of these illnesses; other patient characteristics were reported less frequently ( Molecular typing data demonstrated that GI.6 viruses can be grouped into 2 clusters (Figure 1), with earlier outbreaks occurring deeper in the tree. One of the earliest occurring outbreaks in cluster A (2011-OB-274) occurred in Tennessee in February 2011 and involved a conference with 8,000 attendees and 143 reported cases in persons from 12 states.

Conclusions
We detected an increase in GI.6 outbreaks in the United States since 2010, with peak activity during summer 2012. Summer seasonality, foodborne transmission, and non-health care settings characterized GI.6 outbreaks, compared with non-GI.6 outbreaks reported through Cali-ciNet. Noroviruses are the most common cause of gastroenteritis outbreaks, and although GI.6 noroviruses remain responsible for a relatively small proportion of all reported norovirus outbreaks, they have unique characteristics and public health implications that differ from those of more common genotypes.  were downloaded, trimmed, and analyzed as described (5). In brief, a 3-parameter model, TPM1, with equal frequencies and invariable sites was run in PhyML 3.0 (www.atgc-montpellier.fr/phyml/ binaries.php) as determined by jModel test by using the corrected Akaike information criterion. The best of 5 random trees was used to start the analysis, and the approximate likelihood ratio test was used for branch support. GI.6 reference sequences (GenBank accession nos. GQ856463| GI.6 Beijing and AJ277615| GI.6 Sindlesham) were included. Two clusters of genetically related outbreaks (cluster A and cluster B) are marked by brackets. *The distance of GI.1 Norwalk to the nearest GI.6 cluster is 2.29 substitutions per site. Scale bar indicates nucleotide substitutions per site for the phylogenetic tree. the unique characteristics of GI.6 outbreaks described here primarily reflect differences between GI.6 and GII.4 noroviruses.
The absolute number of outbreaks and the proportion of outbreaks caused by GI.6 noroviruses peaked during April-July. This summer seasonal pattern contrasts with the overall winter seasonality of norovirus outbreaks driven primarily by winter surges in GII.4 norovirus activity (3,(6)(7)(8). A study of GI norovirus outbreaks in Australia demonstrated peak outbreak activity during their summer months, compared with a late winter peak for GII norovirus outbreaks (9). In a previous study in the United States, the highest number of GI outbreaks occurred during April-May, but no apparent seasonality was noted (7).
Region D typing data presented in this study indicates 2 clusters of GI.6 noroviruses in the United States. In February 2011, an outbreak among persons from multiple states occurred at a conference in Tennessee; this outbreak might have been a dissemination event for GI.6 activity because outbreaks of genetically related GI.6 noroviruses belonging to the same cluster occurred later in several of the states in which the conference attendees resided. However, more sequence information from the complete open reading frame (ORF2) or the hypervariable region of the protruding domain (P2) is needed to confirm possible links among the outbreaks (5).
Our study has several limitations. These include incomplete linkage of outbreaks reported in CaliciNet to outbreak reports in NORS and the resulting gaps in data on transmission mode and setting, as well as low rates of reporting of demographic characteristics, symptoms, and clinical outcomes. These limitations preclude direct comparison of GI.6 outbreak characteristics with characteristics of outbreaks linked to other genotypes. Efforts to improve reporting rates and integration between CaliciNet and NORS are under way (8).
Noroviruses are a diverse group of pathogens with varied characteristics. Continued surveillance for norovirus outbreaks through CaliciNet and NORS will enable further assessment of the public health implications and significance of emergence of relatively rare noroviruses, such as GI.6. Proper hand hygiene, environmental disinfection, and isolation of ill persons remain the mainstays of norovirus prevention and control (11).