Characterization of 2009 Pandemic Influenza A (H1N1) Virus Specimens with a Positive Hemagglutinin 1 Signal in the Luminex xTAG Respiratory Viral Panel Assay

Abstract

This retrospective review analyzed Luminex xTAG respiratory viral panel (RVP) results for 2009 pandemic influenza A (H1N1) virus specimens. Comparing median fluorescence intensity (MFI) signals for the influenza A virus and hemagglutinin 1 (H1) reactions for specimens with very low positive (MFI < 1,000) or “no-call” H1 results reliably distinguished 2009 H1N1 from seasonal virus.

A novel strain of influenza A virus, 2009 pandemic influenza A (H1N1) virus (2009 H1N1), emerged in the spring of 2009 and became the dominant strain during the 2009-2010 respiratory virus season (1). On 9 September 2009, our laboratory implemented the Luminex xTAG respiratory viral panel (RVP) assay (Luminex Corporation, Austin, TX) for the diagnosis of viral respiratory infections. The RVP is a solution-based microarray assay that detects influenza viruses A and B, respiratory syncytial viruses A and B, parainfluenza viruses 1, 2, and 3, metapneumovirus, rhinovirus, and adenovirus. In addition, this assay allows for concurrent detection and subtyping of influenza A virus based on three primer pairs specific for (i) the matrix gene common to all subtypes of influenza A virus (FLUA), (ii) the hemagglutinin 1 (H1) gene, and (iii) the hemagglutinin 3 (H3) gene. The qualitative results for each target are generated based on the RVP software's analysis of the median fluorescence intensity (MFI) of each specific bead population in the solution microarray. Results for viral targets are categorized as positive (MFI ≥ 300), negative (MFI < 150), or “no call.” Assuming adequate performance of the internal control, a no-call result for a viral target is likely due to an MFI value in the equivocal zone (150 ≤ MFI < 300) (5–7). Although the RVP detects 2009 H1N1 as influenza A virus, the H1 gene of the 2009 pandemic strain is not reliably detected. Therefore, given the epidemiologic environment of the 2009-2010 respiratory virus season, specimens testing as FLUA positive but H1 and H3 negative could be presumptively identified as 2009 H1N1 virus (3, 4). Our goal was to evaluate the data characteristics of confirmed positive 2009 H1N1 clinical specimens that generated no-call or positive H1 results in the RVP assay.

We performed a retrospective review of RVP results over a 7-month period from 10 September 2009 through 22 April 2010. All specimens positive for influenza A virus by RVP (n = 338) were included in the study. Data from date-matched specimens that were negative for any respiratory virus were collected for the negative-control group (n = 76). No presumptive seasonal H1N1 influenza virus was encountered during the study period at our institution, so commercially prepared material (influenza A, NATRVP1, NATtrol respiratory validation panel, Zeptometrix Corp., Buffalo, NY), a reference strain isolate (A/PR/9/34), and 13 archived seasonal influenza A virus specimens from the 2007-2008 respiratory virus season were tested to generate data for seasonal H1N1 virus (n = 15). As part of the clinical testing process, specimens yielding FLUA-positive and H1-positive or no-call results were sent to the Georgia Public Health Laboratory (GPHL) for subtyping based on their acceptance criteria. Any H1-positive or no-call specimens not analyzed by the GPHL were subtyped on a research basis at our institution using the CDC protocol of real-time reverse transcription (RT)-PCR for influenza A (H1N1) virus for the purposes of this study (2). Data collected for all specimens included the MFI values for both FLUA and H1. For data analysis, specimens were grouped as follows: FLUA positive/H1 negative (H1NEG), FLUA positive/H1 no call (H1NC), FLUA positive/H1 positive (H1POS), seasonal influenza A (H1SEAS), and respiratory virus negative (RVPNEG). The ratio of the H1 MFI to FLUA MFI was calculated for each specimen and expressed as a percent (H1/A%). The average, standard deviation, and range were calculated for the FLUA MFI values, H1 MFI values, and H1/A% for each group. Two-tailed Student t tests assuming unequal variance were performed to compare relatedness between groups.

A total of 414 clinical specimens and 15 control, reference, and historical specimens were included in the study. Of the 338 clinical specimens positive for influenza A virus, 17 yielded positive (n = 6) or no-call (n = 11) results for H1 in the RVP assay, accounting for 5% of all influenza A virus-positive specimens. The remaining 321 influenza A virus-positive specimens yielded negative results for H1 in the RVP assay and were presumed to be 2009 pandemic influenza A (H1N1) virus. The average, standard deviation, range, and H1/A% for each group are presented in Table 1.

Table 1.

Averages, standard deviations, and ranges of FLUA MFI values, H1 MFI values, and H1/A% by specimen group^a

Specimen group	FLUA MFI			H1 MFI			H1/A%
	Avg	SD	Range	Avg	SD	Range	Avg	SD	Range
RVPNEG (n = 76)	31.4	20.0	0.0–98.0	35.4	19.2	0.0–78.5	NA	NA	NA
H1NEG (n = 321)	6,562.1	3,314.8	304.0–11,261.5	42.5	31.3	0.0–148.0	NA	NA	NA
H1NC (n = 11)	9,045.6	540.6	8,295.0–10,032.5	189.0	37.9	152.5–281.0	2.1	0.5	1.6–3.3
H1POS (n = 6)	9,784.7	708.7	9,188.0–11,063.0	426.2	115.9	331.0–603.0	4.4	1.2	3.4–6.6
H1SEAS (n = 15)	7,156.7	1,296.0	3,294.0–8,231.0	4,435.3	952.8	1,956.5–5,335.0	61.7	4.5	52.9–68.7

^aRVPNEG, negative for all respiratory viruses in the RVP assay; H1NEG, FLUA positive and H1 negative in the RVP assay [presumed 2009 pandemic influenza A (H1N1) virus]; H1NC, FLUA positive and H1 “no call” in the RVP assay [confirmed 2009 pandemic influenza A (H1N1) virus]; H1POS, FLUA positive and H1 positive in the RVP assay [confirmed 2009 pandemic influenza A (H1N1) virus]; H1SEAS, FLUA positive and H1 positive in the RVP assay [seasonal influenza A (H1N1) from control material or archived patient specimens from the 2007-2008 respiratory virus season]; NA, not applicable.

Comparison of FLUA MFI values demonstrated that the H1NC and H1POS groups had higher FLUA MFI values than the H1NEG group (P < 0.01 and P < 0.01, respectively). The H1NC and H1POS groups also had higher FLUA MFI values than the H1SEAS group (P < 0.01 and P < 0.01, respectively). FLUA MFI values were not significantly different between the H1NEG and H1SEAS groups (P = 0.13). Comparison of H1 MFI values demonstrated that H1POS specimens had significantly lower H1 MFI values in the positive range than the H1SEAS group (P < 0.01). Comparison of the calculated H1/A% demonstrated that the H1SEAS group had significantly higher H1/A% values than both the H1NC and H1POS groups (P < 0.01 and P < 0.01, respectively). Most importantly, there was no overlap in the ranges of H1/A% between the H1SEAS and H1NC groups or between the H1SEAS and H1POS groups.

We found that a small but clinically significant group (5%) of patients with 2009 H1N1 virus had positive or no-call results for the H1 component of the RVP assay as opposed to a negative result for H1 as had previously been reported (3, 4). Given the differences in susceptibility between 2009 H1N1 and the most recent seasonal influenza A (H1N1) virus strains, relying on the RVP typing results to distinguish 2009 H1N1 and seasonal viruses may lead to mischaracterization of the virus and suboptimal treatment. However, we found that comparing the raw MFI signals for the FLUA and H1 reactions could reliably distinguish between 2009 H1N1 and seasonal viruses. All H1/A% values for 2009 H1N1 virus were <10%, while all H1/A% values for seasonal (H1N1) virus were >50%. Given the time-sensitive nature of the initiation of treatment, this observation could help direct therapeutic decision-making pending definitive viral subtyping.

One limitation of our study was the small number of seasonal influenza A (H1N1) virus specimens available for testing. Although we had multiple seasonal specimens in the middle to high FLUA MFI range compared to that of 2009 H1N1 virus, we had no seasonal specimens at the lower end of the positive ranges for FLUA and H1 and no seasonal specimens with no-call values. The validity of the H1/A% comparison would need to be confirmed in the low MFI value ranges for FLUA and H1 in seasonal specimens. Also, future cocirculation of seasonal (H1N1) and 2009 H1N1 viruses with possible genetic recombination of hemagglutinin genes between subtypes may impact the H1-to-FLUA MFI relationship observed in this study.