[Source: World Health Organization, Weekly Epidemiological Record, full PDF document: (LINK). Edited.]
Weekly epidemiological record / Relevé épidémiologique hebdomadaire
25 january 2013, 88th year / 25 janvier 2013, 88e année - No. 4, 2013, 88, 37–48 - http://www.who.int/wer
Detection of influenza virus subtype A by polymerase chain reaction: WHO external quality assessment programme summary analysis, 2012
Global influenza virological surveillance has been conducted through WHO’s Global Influenza Surveillance and Response System (GISRS) for over 50 years.(1) Currently 140 institutions in 110 WHO Member States are recognized by WHO as National Influenza Centres (NICs). The laboratory network also comprises 6 WHO Collaborating Centres, 4 essential regulatory laboratories and ad hoc groups established to address specific emerging issues.
With the continued evolution of influenza viruses, including both circulating human and non-human viruses causing occasional human infections with pandemic potential, timely surveillance and robust laboratory diagnostic capacity need to be maintained. With technological advances, the use of molecular methods for rapid and sensitive detection of influenza viruses has become more widely available, and has gained importance especially where viral culture facilities are not readily available. Recognizing the essential requirement to assure the quality of laboratory testing, the WHO external quality assessment programme (EQAP) for influenza virus subtype A by polymerase chain reaction (PCR) was initiated in 2007. It has become a resource for monitoring the quality and comparability of the performance of participating laboratories in routine diagnosis and surveillance of seasonal and other emerging influenza viruses, such as avian influenza A(H5N1) and the pandemic influenza A(H1N1)pdm09. Summaries of the performance of participating laboratories from panels 1 to 10 (2007 to 2011) have been reported in the Weekly Epidemiological Record.(2, 3, 4, 5)
In 2012, the project continued under the coordination of WHO’s Global Influenza Programme, implemented by the H5 Reference Laboratory and National Influenza Centre at the Centre for Health Protection, Hong Kong Special Administrative Region of China, and with support from WHO Collaborating Centres for Reference and Research on Influenza, other H5 reference laboratories and WHO regional offices.
The programme currently carries out testing once per year. This report summarizes the results of the assessment of virus panel 11, which was dispatched to participating laboratories between April 2012 and June 2012.
Preparation of panel
Vacuum-dried γ-ray inactivated influenza A and influenza B viruses were dispatched to participating laboratories. Viruses were grown in Madin-Darby canine kidney (MDCK) cells, inactivated by γ-ray irradiation, and then prepared as described previously.(2, 3, 4, 5)
Composition of panel
Panel 11 consisted of 10 coded samples, represented by different concentrations of influenza viruses, including influenza A(H5N1) genetic clade 18.104.22.168, A(H3N2), A(H1N1)pdm09, A(H9N2) and an influenza B virus. A sample that contained no virus was included in this panel. Details of the composition of the panel are shown in Table 1. Participants were instructed to reconstitute each sample with the provided reconstitution buffer prior to testing. A questionnaire on the methods of detection and gene targets used was also included.
Distribution of panel and response of participants
NICs and other national influenza laboratories were invited to participate before the panels were dispatched. Panel 11 was dispatched between April 2012 and June 2012, from the H5 Reference Laboratory and National Influenza Centre at the Centre of Health Protection in Hong Kong SAR at ambient temperature by courier service to 170 participating laboratories from 134 countries in the 6 WHO Regions as previously described.(2, 3, 4, 5)
The majority of laboratories received panel samples within 1 week of dispatch: 125/170 (73.5%) and 163/170 (95.9%) reported results before the closing date.
Results for the performance of laboratories
Based on the same assessment criteria for panels 1–10,2, 3, 4, 5 109/163 (66.9%) participating laboratories returned correct results for all of the 10 samples. Altogether 27 (16.6%) participants returned 1 incorrect result and 27 (16.6%) returned more than 1 incorrect results (Table 2). Among 159 participants reporting influenza A(H5) results, 131 (80.4%) correctly identified all 5 influenza A(H5) samples. The performance of participating laboratories by WHO Region is shown in Figure 1.
Seven participating laboratories reported positive results for the negative sample (V02-2012) and the false-positive rate was 4.3% (Table 1).
Two different influenza A(H5N1) viruses were included.
For the first virus sample (V04-2012) and a corresponding 10-fold diluted sample (V05-2012), the correct rates (...) were 154 (94.5%) and 140 (85.9 %) respectively. For the second virus sample (V07-2012) and 2 identical 10-fold diluted samples (V08-2012) and (V09-2012), the correct rates were 152 (93.3%), 149 (91.4%) and 150 (92%) respectively.
For the influenza A(H3N2) sample (V03-2012) and influenza A(H1N1)pdm09 sample (V10-2012), 155 (95.1%) and 156 (95.7%) laboratories reported correct results respectively. For the influenza B sample of Yamagata lineage (V06-2012), 159 (97.5%) participants reported correct results.
An influenza A(H9) sample (V01-2012) was introduced for the first time to tentatively evaluate the capacities of participating laboratories. Together with an influenza A(H5) sample at low concentration (V05-2012 with 6.352 copies/μl), these 2 challenging samples were excluded from scoring in panel 11. The performance of the participants returning all correct results among the 8 scored samples in panel 11 was 82.2% (134/163), while 87.7% (143/163) participants reported all correct results for the 4 scored influenza A(H5) samples. Comparison of laboratories’ performance for all panels is shown in Figure 2.
In this panel, 54/163 (33.1%) participants reported incorrect results. The details are presented in Table 3.
Methods of detection
There was considerable variation in testing strategies and PCR protocols used by participating laboratories to screen for influenza type A and B viruses and subtype A(H1)pdm09, A(H3) and A(H5) viruses as in previous panels.(2, 3, 4, 5) More than half of the participants used protocols from the United States Centers for Disease Control and Prevention. There was little difference in test results despite the use of different PCR protocols. Only 12.3% (20/163) participating laboratories performed influenza A(H9) subtyping. Of these 20 laboratories, 4 also performed N subtyping and correctly reported N2. Details on target genes, detection methods (...) and source of primers/probes and enzymes used were included in the summary report of performance that was distributed to all participants.
Factors affecting performance
Analysis of panel 11 showed that the use of real-time PCR was significantly associated (p<0.05) with the detection of influenza at low concentration (V05-2012).
Among the 159 participating laboratories which reported H5 subtyping results, 7 participants used conventional PCR alone for H5 subtyping, 5 of whom reported negative H5 results for the sample V05-2012.
This is in contrast to 9 out of 152 laboratories using real-time PCR and reporting false-negative results for this sample. All participants are strongly encouraged to review the testing procedures regularly to ensure good sensitivity and avoid mistaken-subtyping of influenza A viruses.
The analysis also showed that misinterpretation of results led to 9 participating laboratories reporting incorrect influenza A(H9) subtyping results. Participants are recommended to follow strictly the provided instructions for the adopted testing methods) during analysis of results. In addition, the associated Good Laboratory Practice (GLP) 2012 survey revealed that only 24.8% (42/148) of respondent laboratories performed influenza A(H9) detection, and this could explain why only 12.3% (20/163) correctly reported the influenza A(H9) sample.
The number of laboratories participating in the EQAP increased slightly from 158 to 163 since panel 8 in 2010. Viruses inactivated by γ-ray irradiation were introduced in panel 9 and completely replaced RNA samples in panel 11. With the additional extraction step, the allcorrect rate dropped slightly from 86% in panel 8 to 77% and 78% in panels 9 and 10 respectively. The increase to 82% in panel 11 may be a result of improvement in RNA extraction or an overall enhancement in testing capacity. For influenza A(H5) detection, the all-correct rate remained stable at around 88% in the 5 recent panels.
The scope of EQAP has been expanding. An influenza A(H9) sample was introduced for the first time in panel 11. Less commonly recognized subtypes of influenza A with pandemic potential will continue to be included in future panels. It is also planned to include evaluation of the ability to detect influenza antiviral resistance in the coming panel 12, as a trial to gauge the proficiency in the molecular methods used. NICs and other national influenza laboratories are encouraged to continue their participation in the EQAP to raise the quality and standard of the global capacity in diagnosis and surveillance of influenza viruses.
(2) See No. 45, 2008, pp. 401–412.
(3) See No. 48, 2009, pp. 493–504.
(4) See No. 3, 2011, pp. 17–24.
(5) See No. 3 2012, pp. 29–36.