Abstract
Background and aims
The majority of influenza infections during the 2012/2013 influenza season in Scotland have been due to influenza A H3N2. We report an outbreak of influenza A H3N2 in a vaccinated population of adults in the Regional Virology Laboratory in Glasgow. This investigation was carried out to confirm the epidemiological link between cases.
Methods and results
Staff with clinical symptoms of influenza-like illness were included. Samples were tested by real-time polymerase chain reaction and sequencing. Staff were interviewed to obtain information regarding symptom onset and vaccination status. Eight confirmed cases and six clinically diagnosed cases were reported, which all occurred within 4 days of a lunchtime Christmas quiz. The eight samples subtyped as H3 virus. The haemagglutinin gene in the confirmed cases was sequenced and shown to be identical. Most of the attendees had been immunised against influenza with the same vaccine batch at least 6 weeks earlier.
Conclusion
This outbreak appears to have been an isolated incident, which arose due to a social event that provided the ideal conditions for transmission of a respiratory disease. It may have been compounded by low-vaccine effectiveness this season. Sequence data supported the epidemiological link.
Introduction
The number of GP consultations for influenza-like illness (ILI) in Scotland and the rest of the United Kingdom peaked across all age groups between the end of December 2012 and start of January 2013.1,2 The majority of influenza infections in Scotland during the 2012/2013 influenza season were due to influenza A (subtype H3). 1 Most influenza H3 strain sequences released by the National Institute for Medical Research (NIMR), Mill Hill, for the 2012/2013 influenza season are highly similar to the vaccine strain (A/Victoria/361/2011 (H3N2)-like virus). 3 Vaccine uptake in Scotland for those over 65 years of age and those under 65 years of age, in the “at risk” groups, was similar to the previous year. 1 Despite relatively good vaccine uptake and sequence similarity, interim vaccine effectiveness (VE) reports suggest that vaccine efficacy is reduced (49% in UK). 4 Lower VE could explain the increase in influenza outbreaks seen between weeks 49 (2012) and 3 (2013). Four acute respiratory disease outbreaks were reported to Health Protection Scotland between weeks 49 and 51 (one due to influenza A, one due to influenza B, one due to rhinovirus and one due to both influenza A and B). 5 A further three influenza A outbreaks were reported between week 52 (2012) and week 03 (2013), all of which occurred in closed settings (one in a nursing home, one in a community hospital and the present outbreak, which occurred in a work setting). 5
On 18 December 2012, the virology laboratory (WoSSVC) in Glasgow held a lunchtime Christmas quiz. During the quiz, one member of staff reported feeling unwell with ‘flu-like’ symptoms. Over the next 3 days, a number of staff members also reported feeling unwell with ILI. Gargle samples were taken from the majority of staff members reporting ILI and screened for influenza and other respiratory infections by in-house multiplex real-time polymerase chain reaction (PCR).6,7 Eight staff members tested positive for influenza A, which were then subtyped as H3. We suspected an influenza A outbreak; therefore, an investigation was initiated.
Patients and methods
Identification of patients and clinical data
As not all symptomatic staff members were tested, we categorised our cases into confirmed cases and clinically diagnosed cases. Confirmed cases were cases of influenza A (H3) confirmed by a real-time PCR of a gargle sample taken when the patient was symptomatic with ILI. Clinical cases were those staff members who reported ILI but did not provide samples for testing (due to absence from work). One staff member, also considered a clinical case, provided a sample for testing (which tested negative) when their symptoms had subsided. Staff members were interviewed to gather information on symptoms, location during the quiz, treatment and vaccination history. The occupational health department was contacted to investigate whether there were vaccine batch differences between those infected and those who remained uninfected. The investigation was conducted by members of the WoSSVC staff.
Laboratory diagnosis
Viral nucleic acid was extracted using the Qiagen MDx Biorobot automated extractor with the QIAamp Virus MDx Kit according to the manufacturer’s instructions (Qiagen). Samples were tested for the following viruses: influenza A, B, C; parainfluenza 1, 2, 3 and 4; adenovirus; respiratory syncytial viruses A and B; coronaviruses 229E; OC43; NL63; rhinovirus; human metapneumovirus; and Mycoplasma pneumoniae, using five multiplex real-time PCR assays developed in the laboratory and performed routinely for diagnostic purposes. 6 The influenza A multiplex assay simultaneously detects influenza A (H1N1)pdm09 virus if present. 7 Non-H1 samples are then subtyped using a monoplex H3 subtyping real-time PCR assay developed by HPA, Colindale (HPA SOP). For all assays, one-step rtRT–PCR was performed on 6 µl of RNA extract with the Platinum One-step qRT-PCR kit (Invitrogen) on an ABI Prism 7500 SDS real-time platform (Applied Biosystems). The following thermal profile was used: a single cycle of reverse transcription for 30 min at 50℃, 15 min at 95℃ for reverse transcriptase inactivation, followed by 40 amplification cycles of 30 s at 95℃, 30 s at 58℃ and 1 min 10 s at 72℃ each (annealing-extension step); 72℃ 10 min; 4℃ ∞. Thresholds were set manually for analysis.
Sequencing and phylogenetic analysis
PCR and sequencing primers for H3 HA.
First round primers were used for one-step RT-PCR, and second round primers were used for nested PCR using Taq PCR mastermix. Purified PCR products were then sequenced using the sequencing primers.
Antigenic determination
The antigenic properties of the isolates were investigated following culture of five of the eight staff isolates in MDCK-SIAT1 cells and performing haemagglutination inhibition (HI) assays. The assays were carried out using 1% guinea pig red blood cells in the presence of 20 nM oseltamivir carboxylate to circumvent agglutination of the red blood cells through the virus NA. 8 The panel of post-infection ferret antisera included antisera raised against both exclusively egg-propagated A/Victoria/361/2011 and exclusively cell-propagated A/Victoria/361/2011, the strain included in the influenza vaccine for the Northern Hemisphere winter 2012/2013. 9
Results
Setting
The outbreak occurred in a large diagnostic laboratory in Glasgow with 58 members of staff. Thirty-three members of staff attended the lunchtime quiz, which involved spending 1 h, in a relatively enclosed space. Members of the same quiz team were seated in close proximity to each other, and finger-food and non-alcoholic drinks were served (Figure 1). During the quiz, one staff member reported feeling unwell with ILI (indicated with an I in Figure 1). A gargle sample was taken, screened for a number of respiratory infections and tested positive for influenza A. It then subtyped as H3 virus. Over the course of the next 3 days, a number of other staff members reported feeling unwell with ILI (see Figures 1 and 2). Gargle samples were taken and screened for a number of respiratory infections, and seven tested positive for influenza A. One sample (taken when the staff member’s symptoms had subsided) tested negative. The seven additional samples also subtyped as H3 virus.
A plan of the room in which the quiz took place (not to scale), which shows the spatial closeness of symptomatic (confirmed and clinical cases) and non-symptomatic individuals. Confirmed cases are shaded red, and clinical cases are shaded amber. The index case is highlighted with an I. Finger food and non-alcoholic drinks were served on two tables at the top of the room. The door is shaded brown. Timeline showing all confirmed and clinical cases detected during the outbreak. The date highlighted is the date of onset of symptoms. All confirmed cases tested positive for influenza A and subtyped as H3. ‘Staff 7’ did not attend the quiz but spent a considerable amount of time with another confirmed case (‘Staff 8’). All cases occurred within 3 days of the quiz (which took place on December 18).

Findings
As can be seen in the room plan (Figure 1), which details the location of staff during the lunchtime quiz, there were a total of eight confirmed cases and six clinical cases. Confirmed cases are shaded red, and clinical cases are shaded amber. There appears to be a cluster of confirmed cases near the door, which includes the suspected index case (I). As can be seen from the timeline in Figure 2, the suspected index case (I) reported symptoms during the quiz and was the first confirmed case, testing positive on December 18. The incubation period for influenza A is between 2 and 7 days (mean of 3 days), which corresponds to December 21, the date on which all other confirmed cases tested positive. One confirmed clinical case did not attend the quiz but is known to have spent a considerable amount of time in close proximity to another confirmed cases suggesting secondary transmission. All confirmed cases were interviewed to determine whether additional secondary transmission had occurred, which did not appear to be the case.
As can be seen from Figure 3, all staff HA H3 sequences fall into clade 3C and were virtually identical (99.5–100% amino acid and nucleic acid sequence identity), thus supporting the epidemiological link. They were also highly similar to other sequences circulating in Scotland (96.8–100% nucleic acid identity and 94.2–100% amino acid sequence identity) during the 2012–2013 influenza season and the vaccine strain (96.2–96.6% nucleic acid identity and 97.9–99.8% amino acid identity).
A distance-based neighbour-joining phylogenetic tree with bootstrap analysis (1000 replicates) was generated using MEGA 5.0.5 software. Reference influenza A (H3) strains, including the influenza A (H3N2) vaccine component for the northern hemisphere 2012/2013 influenza season (A/Vic/208), one Danish strain, A/Denmark/74
10
and five sample sequences (samples 1–5) tested by WoSSVC during the 2012/2013 influenza season were included for comparison. Outbreak sequences are Staffs 1–3 and 5–8 (index case is Staff 2).
All outbreak sequences contained two mutations in the HA antigenic regions, T128A and R142G, which are absent in the vaccine strain, but present in Danish strains including A/Denmark/65 (Figures 3 and 4). The NA gene of all four of the samples had identical sequences and fell in clade 3C (data not shown).
A portion of the amino acid sequence alignment of the haemagglutinin gene in seven outbreak sequences and two reference sequences (including the 2012/2013 Northern Hemisphere vaccine component, A/Vic/208 and one Danish strain, A/Denmark/74 from Bragstad et al., 2013
10
) within the antigenic region, containing two amino acid substitutions. The T128A and R142G substitutions are highlighted in yellow, with the original sequence highlighted in blue.
Differences in inhibition by postinfection ferret antisera of the agglutination of guinea pig red blood cells by MDCK cell-grown viruses.
E: egg passage; MK: monkey kidney cells; NVD: suspension MDCK cells from Novartis Vaccines and Diagnostics; SIAT: MDCK cells expressing α-2,6-sialyltransferase 10 ; C2, unspecified tissue culture cells. The number refers to the number of cell passages each virus underwent in the originating laboratories and in the laboratory carry.
Northern Hemisphere Vaccine component.
Homologous titres are displayed in boldface.
Note: Clade 3C viruses are A/Vic/361/2011, A/Berlin/93/2011, A/Texas/50/12 and all test viruses, A/Stockholm/18/2011 is a clade 3A virus, A/Athens/112/2012 is a clade 3B virus, A/Alabama/5/2010 is a clade 5 virus and A/Iowa/19/2010 is a clade 6 virus. A/Perth/16/2009 and A/Victoria/208/2009 are the prototype viruses of their respective clades.
Potential source of exposure
The index case spent a considerable amount of time in the company of someone who had ILI (unconfirmed, probable case) on December 15 and suspects that this was the source of her infection.
Discussion
The WHO recommended that the 2012/2013 influenza vaccine for the northern hemisphere should contain the following: an A/California/7/2009 (H1N1) pdm09-like virus; an A/Victoria/361/2011 (H3N2)-like virus; and a B/Wisconsin/1/2010-like virus. 9 This study describes an influenza A (H3) outbreak, which occurred in a group of vaccinated adults in the virus laboratory in Glasgow in December 2012. The eight confirmed cases subtyped as H3 and sequencing revealed them to be highly similar to the ‘A/Victoria/361/2011 (H3N2)-like virus’ found in the northern hemisphere vaccine. All eight of the confirmed cases and five of the six clinical cases had been immunised with the same vaccine batch (J8292-1). However, 15 of the 18 uninfected had also been vaccinated with the same vaccine batch; therefore, vaccine batch failure is unlikely. A recent report indicated that interim UK trivalent influenza vaccine (TIV) VE for influenza A for 2012/2013 influenza season was 49%. 4 This is lower than expected, as VE to seasonal influenza vaccine is expected to range from 50 to 70%.11,12 Although, it is in line with reports of VE elsewhere (USA 62% and Canada 45%).13,14 As 13 out of 31 immunised staff members became infected (∼42%), this appears to represent slightly lower than expected VE.
Recently, Bragstad et al. 10 reported a new H3 subclade of the vaccine strain in Denmark that contains a number of substitutions in the HA gene, including two within the antibody-binding regions: T128A and R142G. 10 The T128A substitution lies within antigenic site B and causes the loss of an N-linked glycosylation site, which could affect antibody binding.10,15,16 This new subclade has been detected in USA, Canada and Europe.3,10 All seven staff H3 sequences contained the T128A and R142G mutations (Figure 4), as have all H3 samples sequenced to date (with two exceptions) by the West of Scotland Specialist Virology Centre (2012/2013 influenza season, started October 2012). Bragstad et al. 10 suggest that these mutations could be due to antigenic drift. This is considered unlikely as antigenic drift in H3N2 has been muted in recent years. 8 Indeed, the results from HI analysis of the viruses from the outbreak (Table 2) are consistent with the results from HI and virus neutralisation results of other strains that carry the same substitutions and indicate that there is no evidence that substitutions at residues 128 and 142 have a marked impact on the antigenic properties of the virus when compared with other viruses representative of those in circulation in 2010, 2011 and 2012. 3
However, the isolates were antigenically distinguishable by HI from egg-propagated viruses derived over the same time scale. As egg-propagated viruses are used as vaccine seed viruses, it has been considered necessary to change the vaccine recommendation to an egg-grown virus that is a closer antigenic match to the cell-grown A/Victoria/361/2011, namely A/Texas/50/2012. 3
Conclusion
This outbreak appears to have been an isolated incident brought about by a social setting, which provided the ideal opportunity for transmission of a respiratory virus such as influenza. This may have been compounded by lower than expected VE this year. It is certainly illustrative of vaccine failure, as the majority of staff members had been vaccinated and would have been expected to be protected. The index case had clear contact with a probable case and subsequent contact with other confirmed and clinical cases within the laboratory. Fortunately, no staff members needed antiviral therapy, and there was no secondary transmission out with the laboratory.
Footnotes
Acknowledgements
We would like to thank Dr John McCauley (WHO Collaborating Centre for Reference and Research on Influenza) for reviewing and commenting on this manuscript. We would also like to thank NHS Education Scotland for supporting the work of the corresponding author.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
