Abstract
Background
Pandemic influenza virus strains originate in avian species. We examined veterinarians in the United States for evidence of previous avian influenza virus infection.
Methods
We performed a controlled, cross-sectional seroprevalence study among 42 veterinarians and 66 healthy control subjects using serum samples collected from 2002 through 2004. Serum samples were tested using a microneutralization assay against 9 influenza A virus strains.
Results
Using multivariable logistic regression modeling, veterinarians exposed to birds demonstrated statistically significant elevated titers against the H5, H6, and H7 avian influenza virus isolates, compared with control subjects.
Conclusions
These data suggest that occupational exposure to avian species may increase veterinarians’ risk of avian influenza virus infection. Veterinarians should be considered for priority access to vaccines and antiviral drugs in pandemic planning.
Influenza A viruses are known to infect a wide variety of animals, including humans, pigs, birds, horses, and sea mammals. The primary reservoir of influenza A virus is aquatic waterfowl, and birds are the source of all influenza viruses in other species [1]. It was previously believed that adaptation in an intermediate host was necessary for avian influenza strains to acquire the ability to infect humans. However, recent experience with the H5N1 virus has clearly demonstrated that the virus may jump directly from birds to humans without an intermediate host.
Since 1997, there have been 285 cases and 170 fatalities in humans caused by H5N1 virus [2]. Most infected individuals have had direct contact with sick poultry through exposures such as butchering or culling infected poultry [3] or through preparing poultry for consumption [4]. More recently, cases of human illness due to avian H9N2 [5], H7N2 [6], H7N7 [7], and H7N3 viruses [8] have been reported.
A serosurvey conducted in China in 1992 suggests frequent human infection from avian influenza strains [9]. Little is known about the seroprevalence of avian influenza among humans in the United States. In this pilot study, we sought to estimate the seroprevalence of antibodies against avian influenza viruses in veterinarians with exposure to birds and to determine risk factors for infection.
SUBJECTS AND METHODS
Study subjects
Veterinarians attending a conference of the Iowa Veterinary Medical Association in the spring of 2004 were invited to enroll in the study. Control subjects were comprised of volunteers associated with the University of Iowa (Iowa City) who were enrolled in the study during the spring of 2006. Study participants completed occupational risk factor questionnaires and were only permitted to participate if they had no immunocompromising conditions, were >18 years of age, and were not pregnant. The study was conducted after institutional review board approval and with signed informed consent.
Laboratory methods
Virus stocks were grown in the allantoic cavities of 10-day old embryonated hens’ eggs. Infected eggs were incubated for 3 days at 37°C and were chilled at 4°C overnight. Allantoic and amniotic fluids were harvested and clarified by centrifugation (600 g for 10 min). A hemagglutination assay was performed on the fluids from individual eggs; fluids with positive results were then pooled, aliquotted, and frozen at −80°C. Serum specimens were collected in serum separator tubes, allowed to clot at room temperature, centrifuged for 10 min at 1100–1300 g, aliquotted, and frozen at −80°C on the same day. Madin-Darby canine kidney cells used for the microneutralization assays were derived from the London lineage obtained from the Centers for Disease Control and Prevention and maintained in 5% Dulbecco minimum essential media containing 5% fetal bovine serum (Invitrogen/Gibco).
A microneutralization assay, adapted from that of Rowe et al. [10], was used to detect antibodies to avian strains H4–H12 (table 1). Because prevalence was expected to be low in the control group, serum samples were first screened at a dilution of 1:10, with 2-fold serial dilutions from 1:10 to 1:1280 run on all positive samples and tested in duplicate. Anticipating more reactivity, 2-fold serial dilutions from 1:10 to 1:1280 in duplicate were run on all veterinarian serum samples. Serial one-half–log dilutions of virus (1:100) in Modified Eagle’s Medium with Earle’s salts and L-glutamine, 7.5% bovine serum albumin, Hepes, and penicillin-streptomycin solution were performed in 96-well, flat-bottom cell culture plates (Falcon 35–3072). Madin-Darby canine kidney cells in log phase growth were adjusted to 2.0 × 105 cells/mL with diluent. One hundred microliters of cells were added to each well, and the plate was incubated at 37°C with 5% CO2 for 24 h. Plates were washed twice with PBS, fixed with cold 80% acetone, and incubated at room temperature for 10 min. ELISA was performed to detect the viral nucleoprotein. The end point titer was expressed as the reciprocal of the highest dilution of serum with optical density (OD) less than X, where X = [(average OD of virus control wells) + (average OD of cell control wells)]/2. Test cells with an OD >2 times the cell control OD mean were considered to be positive for virus growth. Virus suspension was calculated by the Reed Muench method to determine the TCID50/100 µL. The back titer was run in duplicate and was only accepted when both replicates had matching results.
Table 1.
Influenza viruses and antisera used for serological assays.
| Influenza virus, subtype | Antigen | Antisera |
|---|---|---|
| Avian | ||
| H4 | A/Duck/Czechoslovakia/1/56 (H4N8) | A/Duck/Shantou/461/00 |
| H5 | A/Chucker/Minnesota/14591-7/98 (H5N2) | A/Goose/Hong Kong/437/99 and A/Tern/South Africa/61 |
| H6 | A/Turkey/Massachusetts/65 (H6N2) | A/Turkey/Massachussets/65 |
| H7 | A/Turkey/Virginia/4529/02 (H7N2) | A/Fowl Plague Virus/Rostock/34 |
| H8 | A/Turkey/Ontario/68 (H8N5) | A/Turkey/Ontario/68 |
| H9 | A/Turkey/Minnesota/38391-6/95 (H9N2) | A/Turkey/Minnesota/38391-6/95 |
| H10 | A/Chicken/Germany/49 (H10N7) | A/Chicken/Germany/49 |
| H11 | A/Duck/Memphis/546/76 (H11N9) | A/Duck/Hong Kong/M603/98 |
| H12 | A/Duck/Alberta/60/76 (H12N5) | A/Duck/Alberta/60/76 |
| Human | ||
| H1 | A/New Caledonia/20/99 (H1N1) | … |
| H3 | A/Panama/2007/99 (H3N2) and A/Nanchang/933/95 (H3N2) | … |
Serum samples were also tested using a hemagglutination inhibition (HI) assay against 3 isolates of recently circulating human influenza A virus (table 1). The Centers for Disease Control and Prevention HI serological protocol was followed. Antigen strains were grown in embryonated chicken eggs. Serum samples were pretreated with receptor-destroying enzyme (1 part serum to 3 parts enzyme) from Vibrio cholerae overnight and then were hemadsorbed with guinea pig blood. HI titer results are reported as the reciprocal of the highest dilution of serum that inhibited virus-induced hemagglutination of a 0.5% solution of guinea pig erythrocytes.
HI assays with horse erythrocytes were performed on all serum samples with positive microneutralization assay results and on a subset of serum samples with negative results for subtypes H5, H6, and H7 [11]. Serum samples were treated with receptor-destroying enzyme, heat inactivated at 56°C for 30 min, and hemadsorbed with horse erythrocytes. Two-fold serial dilution series of serum were incubated with virus at 8 hemagglutinin U per 50 µL with 1% horse erythrocytes in 0.5% bovine serum albumin in phosphate buffered saline for 1 h at room temperature in V-bottom plates.
Statistical methods
Geometric mean HI titers were calculated for each virus strain and were compared by risk factor using the Wilcoxon rank-sum test with normal approximation. Only strains that demonstrated a significant difference between veterinarians and control subjects were further analyzed.
Exact logistic regression was then used to screen risk factors for their association with the outcome, with HI titers ≥1:10 considered to be positive. Covariates with bivariate P values <.25 were considered for inclusion in the multivariable logistic regression models. Final multivariable models included screened risk factors using a saturated model with manual backwards elimination.
RESULTS
We enrolled a total of 75 veterinarians and 77 control subjects. Veterinarians who had no exposure to birds were excluded from the analysis. Of the 42 remaining veterinarians, 32 worked with live chickens, 21 worked with live ducks, 18 worked with live turkeys, 12 worked with live geese, and 7 worked with live quail. Eleven of the control subjects reported exposure to birds and were excluded from the analysis, leaving 66 control subjects. We had insufficient serum samples for 1 veterinarian and were not able to include samples from that veterinarian in all analyses.
Demographic characteristics of participants are presented in table 2. Veterinarians were more likely to be male, but there were no other significant differences between the groups. Geometric mean antibody titers were elevated for the avian strains H5, H6, H7, and H9 in veterinarians and differed from the geometric mean antibody titers in control subjects by the Wilcoxon rank-sum test with normal approximation (table 3).
Table 2.
Demographic characteristics of the study population.
| Control subjects | Veterinarians | |
|---|---|---|
| Variable | (n = 66) | (n = 42) |
| Age | ||
| 18–41 years | 31 (47.0) | 13 (31) |
| 42–51 years | 17 (25.8) | 15 (35.7) |
| 52–77 years | 18 (27.3) | 14 (33.3) |
| Median years | 44 | 48 |
| Sex | ||
| Female | 45 (68.2) | 14 (33.3) |
| Male | 21 (31.8) | 28 (66.7) |
| Race | ||
| Asian | 2 (3.0) | 0 (0) |
| White | 64 (97.0) | 42 (100) |
| Ethnicity | ||
| Not Hispanic | 60 (98.4) | 42 (100) |
| Hispanic | 1 (1.6) | 0 (0) |
| Smoked a total of ≥5 packs of tobacco products in the past year | ||
| No | 62 (93.9) | 42 (100) |
| Yes | 4 (6.1) | 0 (0) |
NOTE. Data are no. (%) of subjects, unless otherwise indicated.
Table 3.
Geometric mean and distribution of antibody titers against avian influenza viruses.
| Control subjects | Veterinarians | |
|---|---|---|
| Influenza virus, titer | (n = 66) | (n = 42) |
| Avian H4 | ||
| <1:10 | 66 (100) | 40 (97.6) |
| 1:10 | 0 (0) | 1 (2.4) |
| Geometric mean titer | 5.0 | 5.1 |
| Avian H5a | ||
| <1:10 | 66 (100) | 36 (87.8) |
| 1:10 | 0 (0) | 4 (9.8) |
| 1:20 | 0 (0) | 1 (2.4) |
| Geometric mean titer | 5.0 | 5.5 |
| Avian H6a | ||
| <1:10 | 64 (97) | 32 (76.2) |
| 1:10 | 0 (0) | 9 (21.4) |
| 1:20 | 2 (3) | 1 (2.4) |
| Geometric mean titer | 5.2 | 6.0 |
| Avian H7a | ||
| <1:10 | 66 (100) | 35 (85.4) |
| 1:10 | 0 (0) | 6 (14.6) |
| Geometric mean titer | 5.0 | 5.5 |
| Avian H8 | ||
| <1:10 | 66 (100) | 42 (100) |
| Geometric mean titer | 5.0 | 5.0 |
| Avian H9a | ||
| <1:10 | 65 (98.5) | 34 (89.5) |
| 1:10 | 1 (1.5) | 3 (7.9) |
| 1:20 | 0 (0) | 1 (2.6) |
| Geometric mean titer | 5.1 | 5.5 |
| Avian H10 | ||
| <1:10 | 66 (100) | 41 (97.6) |
| 1:10 | 0 (0) | 1 (2.4) |
| Geometric mean titer | 5.0 | 5.1 |
| Avian H11 | ||
| <1:10 | 66 (100) | 40 (97.6) |
| 1:10 | 0 (0) | 0 (0) |
| 1:20 | 0 (0) | 0 (0) |
| 1:40 | 0 (0) | 1 (2.4) |
| Geometric mean titer | 5.0 | 5.3 |
| Avian H12 | ||
| <1:10 | 65 (98.5) | 39 (97.5) |
| 1:10 | 0 (0) | 0 (0) |
| 1:20 | 1 (1.5) | 0 (0) |
| 1:40 | 0 (0) | 0 (0) |
| 1:80 | 0 (0) | 1 (2.5) |
| Geometric mean titer | 5.1 | 5.4 |
NOTE. Data are no. (%) of subjects, unless otherwise specified.
Differed from that of control subjects by the Wilcoxon rank sum analysis.
We examined a number of possible risk factors, including age, chronic medical conditions, race and ethnicity, medication use, military service, children in the home, and smoking (data not shown). Potential occupational risk factors included working with birds known to be infected with influenza, number of years of exposure to birds, and the use of protective personal equipment, such as gloves and masks.
Possible confounding caused by serologic cross-reactivity between avian and human viral strains was assessed by evaluating associations with receiving human influenza vaccine in the past 3 years, receiving the swine influenza vaccine in 1976, or having titers ≥40 to human antibody strains H1N1 and H3N2. None of these possible confounders were found to be significantly associated with elevated titers against any of the viral strains. A power analysis (exact method) of the exposure variable data indicated that we had >90% probability to detect an important association (OR, ≥10) with avian strains H5, H6, or H7 serologic assays.
After adjusting for possible confounders, multivariable logistic regression revealed that veterinarians had elevated ORs for 3 of the avian influenza strains (table 4). Although the geometric mean antibody titer for avian H9 was significantly higher in veterinarians, there was no significant difference in seropositivity between veterinarians and control subjects with logistic regression. Veterinarians had much greater adjusted ORs than did control subjects for being seropositive for avian H5 (adjusted OR, 16.7; 95% CI, 2.1–∞), avian H6 (adjusted OR, 12.2; 95% CI, 2.0–138.2), and avian H7 (adjusted OR, 17.7; 95% CI, 2.3–∞). Veterinarians who reported having examined birds known to be infected with influenza presented an increasing trend of being seropositive, compared with veterinarians without this exposure and with control subjects. No other risk factors showed a statistically significant association with elevated antibody titers. Because data were sparse for certain potential risk factors (e.g., few subjects answered “yes” to chronic medical conditions and smoking), the power to detect statistically significant associations of these variables with the avian influenza virus serologic assays was <.8. In general, horse erythrocyte HI titers were higher than those for microneutralization assay, with agreement (±1 titer) of 65% for H5, 85% for H6, and 100% for H7.
Table 4.
ORs for a positive serological response against avian influenza virus by microneutralization assay using exact logistic regression.
| Avian H5 |
Avian H6 |
Avian H7 |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable | No. of subjects | No. (%) of subjects with titers ≥ 1:10 | Unadjusted OR (95% Cl) | Adjusted OR (95% Cl) | No. of subjects | No. (%) of subjects with titers ≥ 1:10 | Unadjusted OR (95% Cl) | Adjusted OR (95% Cl) | No. of subjects | No. (%) of subjects with titers ≥ 1:10 | Unadjusted OR (95% Cl) | Adjusted OR (95% Cl) |
| Age, continuous | 107 | 5 (4.7) | 1 (0.9–1.1) | … | 108 | 12 (11.1) | 1 (0.9–1) | 0.96 (0.9–1.0) | 107 | 6 (5.6) | 1 (0.9–1.1) | … |
| Exposure | ||||||||||||
| Avian-exposed veterinarians | 41 | 5 (12.2) | 11.9 (1.6–∞) | 16.7 (2.1–∞) | 42 | 10 (23.8) | 9.8 (1.9–97.1) | 12.2 (2–138.2) | 41 | 6 (14.6) | 14.8 (2.1-∞) | 17.7 (2.3-∞) |
| Control subjects | 66 | 0 (0.0) | Reference | Reference | 66 | 2 (3) | Reference | Reference | 66 | 0 (0.0) | Reference | Reference |
| Reported having received a flu shot within the last 3 years | ||||||||||||
| Yes | 76 | 4 (5.3) | 1.7 (0.2–84.8) | 4.3 (0.4–231.5) | 77 | 6 (7.8) | 0.4 (0.1–1.5) | 1.1 (0.2–5.3) | 76 | 4 (5.3) | 0.8 (0.1–9.4) | 2.1 (0.3–25.8) |
| No | 31 | 1 (3.2) | Reference | Reference | 31 | 6 (19.4) | Reference | Reference | 31 | 2 (6.5) | Reference | Reference |
NOTE. Titers >1:10 were considered to be positive. Final multivariable models were designed after screening risk factors for associations with the outcomes (P<.25), using a saturated model that included all potential risk factors of important bivariate predictors and manual backwards elimination.
DISCUSSION
This study provides evidence that veterinarians are at increased risk for infection due to avian influenza virus. To our knowledge, this is the first investigation of seroprevalence of a wide variety of avian influenza subtypes in occupationally exposed veterinarians in the United States.
Distribution of hemagglutinin subtypes in birds varies geographically, temporally, and by species. However, the H6 subtype appears to be commonly found in birds, whereas H5 and H7 are less common [12–14]. Why these veterinarians have elevated risk for antibodies to subtypes not frequently found in birds is unknown. Different subtypes are likely to vary in their ability to infect and produce an antibody response in humans. In addition, surveillance data for avian species are limited, and it is possible that outbreaks of these subtypes have occurred but have gone unnoticed.
The possibility of cross-reaction between subtypes must also be considered. However, the lack of an association between seropositivity and having received a human influenza vaccine suggests that there is no cross-reactivity of the avian H5, H6, and H7 serotypes against human H1 and H3 subtypes. We were unable to assess the degree of cross-reactivity of avian subtypes against each other in this small study.
Pandemic influenza viruses must have the ability to spread efficiently from person to person. When different influenza subtypes coexist in a single host, recombination of genetic material can occur, creating a novel virus with new characteristics that may increase infectivity and transmissibility [15]. Increasing vaccination rates among veterinarians with human influenza vaccine would decrease opportunities for reassortment events and should be considered.
There are several limitations of this study. The study design did not allow for the evaluation of an association of seropositivity with clinical symptoms. The nature, frequency, and severity of clinical illness caused by infection due to avian influenza remains unknown. The use of gloves and masks in this population was inconsistent and infrequent, making it unlikely that we would be able to detect any protective effect of personal protective equipment, even if it were present.
A better understanding of interspecies transmission of avian influenza is a crucial component in efforts to minimize the effects of the next pandemic. Humans with frequent and close contact to infected birds may be among the first to be infected and, by spreading the illness to their families and communities, may serve as a bridging population to the general population [16]. Early detection and intervention will be an important component of pandemic preparedness. Knowledge of prevalence rates and risk factors for zoonotic influenza transmission are fundamental in developing pandemic plans. These study data suggest that avian-exposed veterinarians may be at increased risk of zoonotic influenza infection. We posit that they should be considered for inclusion on priority access lists for pandemic vaccines and antiviral drugs.
Acknowledgments
We thank Debbie Strauss and Ghazi Khayali (University of Iowa Center for Emerging Infectious Diseases, Iowa City), for providing serologic assay study support; the Iowa Veterinary Medical Association (Ankeny), for helping us to engage their members; Drs. Jackie Katz and Alexander Klimov (Centers for Disease Control and Prevention, Atlanta, Georgia), Dr. Richard Webby (St. Jude Children’s Research Hospital, Memphis, Tennessee), and Dennis Senne (Diagnostic Virology Laboratory at the US Department of Agriculture’s National Veterinary Services Laboratories, Ames, Iowa), for kindly providing avian influenza viruses and antisera, and for their advice and assistance with laboratory reagents and the adaptation of serologic procedures.
Financial support. National Institute of Allergy and Infectious Diseases (R21 AI059214-01 to G.C.G.).
Footnotes
Potential conflicts of interest. All authors: no conflicts.
References
- 1.Webster RG, Bean WJ, Gorman OT, Chambers TM, Kawaoka Y. Evolution and ecology of influenza A viruses. Microbiol Rev. 1992;56:152–179. doi: 10.1128/mr.56.1.152-179.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.World Health Organization (WHO) [Accessed 17 may 2007];Cumulative number of confirmed human cases of avian influenza A/(H5N1) reported to the WHO. Available at: http://www.who.int/csr/disease/avian_influenza/country/cases_table_2007_03_29/en/index.html.
- 3.Bridges CB, Lim W, Hu-Primmer J, et al. Risk of influenza A (H5N1) infection among poultry workers, Hong Kong, 1997–1998. J Infect Dis. 2002;185:1005–1010. doi: 10.1086/340044. [DOI] [PubMed] [Google Scholar]
- 4.Dinh PN, Long HT, Tien NTK, et al. Risk factors for human infection with avian influenza A H5N1, Vietnam, 2004. Emerg Infect Dis. 2006;12:1841–1847. doi: 10.3201/eid1212.060829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Peiris M, Yuen KY, Leung CW, et al. Human infection with influenza H9N2. Lancet. 1999;354:916–917. doi: 10.1016/s0140-6736(99)03311-5. [DOI] [PubMed] [Google Scholar]
- 6.Wentworth DE, McGregor MW, Macklin MD, Neumann V, Hinshaw VS. Transmission of swine influenza virus to humans after exposure to experimentally infected pigs. J Infect Dis. 1997;175:7–15. doi: 10.1093/infdis/175.1.7. [DOI] [PubMed] [Google Scholar]
- 7.Fouchier RA, Schneeberger PM, Rozendaal FW, et al. Avian influenza A virus (H7N7) associated with human conjunctivitis and a fatal case of acute respiratory distress syndrome. Proc Natl Acad Sci U S A. 2004;101:1356–1361. doi: 10.1073/pnas.0308352100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tweed SA, Skowronski DM, David ST, et al. Human illness from avian influenza H7N3, British Columbia. Emerg Infect Dis. 2004;10:2196–2199. doi: 10.3201/eid1012.040961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shortridge KF. Pandemic influenza: a zoonosis? Semin Respir Infect. 1992;7:11–25. [PubMed] [Google Scholar]
- 10.Rowe T, Abernathy RA, Hu-Primmer J, et al. Detection of antibody to avian influenza A (H5N1) virus in human serum by using a combination of serologic assays. J Clin Microbiol. 1999;37:937–943. doi: 10.1128/jcm.37.4.937-943.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Meijer A, Bosman A, van de Kamp EE, Wilbrink B, van Beest Holle Mdu R, Koopmans M. Measurement of antibodies to avian influenza virus A (H7N7) in humans by hemagglutination inhibition test. J Virol Methods. 2006;132:113–120. doi: 10.1016/j.jviromet.2005.10.001. [DOI] [PubMed] [Google Scholar]
- 12.Hanson BA, Stallknecht DE, Swayne DE, Lewis LA, Senne DA. Avian influenza viruses in Minnesota ducks during 1998–2000. Avian Dis. 2003;47:867–871. doi: 10.1637/0005-2086-47.s3.867. [DOI] [PubMed] [Google Scholar]
- 13.Krauss S, Walker D, Pryor SP, et al. Influenza A viruses of migrating wild aquatic birds in North America. Vector Borne Zoonotic Dis. 2004;4:177–189. doi: 10.1089/vbz.2004.4.177. [DOI] [PubMed] [Google Scholar]
- 14.Slemons RD, Hansen WR, Converse KA, Senne DA. Type A influenza virus surveillance in free-flying, nonmigratory ducks residing on the eastern shore of Maryland. Avian Dis. 2003;47:1107–1110. doi: 10.1637/0005-2086-47.s3.1107. [DOI] [PubMed] [Google Scholar]
- 15.Klempner MS, Shapiro DS. Crossing the species barrier—one small step to man, one giant leap to mankind. N Engl J Med. 2004;350:1171–1172. doi: 10.1056/NEJMp048039. [DOI] [PubMed] [Google Scholar]
- 16.Saenz RA, Hethcote HW, Gray GC. Confined animal feeding operations as amplifiers of influenza. Vector Borne Zoonotic Dis. 2006;6:338–346. doi: 10.1089/vbz.2006.6.338. [DOI] [PMC free article] [PubMed] [Google Scholar]