During April 2009–June 2010, the Centers for Disease Control and Prevention (CDC) requested state public health laboratories to submit specimens for antiviral susceptibility testing by 2 routes. In the first route, the first 5 influenza specimens of any type or subtype collected every 2 weeks from each laboratory underwent virus isolation for comprehensive antiviral testing, including testing by neuraminidase inhibition (NI) assay, sequencing viruses with elevated 50% inhibitory concentration (IC₅₀) values, and pyrosequencing for adamantine resistance–conferring M2 mutations. In the second route, the first 5 additional clinical specimens from pandemic (H1N1) 2009 virus–infected patients that were collected each week by these laboratories were submitted and screened for the oseltamivir-resistant conferring neuraminidase H275Y mutation by using pyrosequencing. Patients with oseltamivir-resistant pandemic (H1N1) 2009 infection had demographic and clinical information collected by using a standard form.

Oseltamivir resistance was determined by either NI or pyrosequencing for the H275Y mutation. NI was performed on virus isolates with a chemiluminescent substrate; viruses with elevated IC₅₀ values for oseltamivir were identified as resistant, based on previously set criteria (1,2). All oseltamivir-resistant viruses had H275Yconfirmed by pyrosequencing (1). Original clinical specimens collected from surveillance were screened by pyrosequencing for H275Y, without NI. NI testing was performed at CDC, and pyrosequencing for H275Y was performed at CDC and state laboratories in Wisconsin, New York, and California. All oseltamivir-resistant viruses referenced here were reported on FluView (3). Four patients, identified in June and August 2009, were reported previously (4,5).

A comparison group of hospitalized patients infected with oseltamivir-susceptible pandemic (H1N1) 2009 was identified from the Influenza Hospitalization Network (FluSurv-NET). FluSurv-NET includes 10 states that participate in the Emerging Infections Program, a population-based surveillance for hospitalized patients with influenza infection (California, Colorado, Connecticut, Georgia, Maryland, Minnesota, New Mexico, New York, Oregon, Tennessee), plus 6 states (Iowa, Idaho, Michigan, North Dakota, Oklahoma, South Dakota) added in response to the 2009 pandemic, as previously described (3,6). The counties within FluSurv-NET represent 26 million persons (8.5% of the US population) (6). The 16 states participating in FluSurv-NET collected demographic and clinical information for all hospitalized patients with laboratory-confirmed influenza infection within their catchment counties (6). We identified patients hospitalized in FluSurv-NET who had specimens submitted to national antiviral resistance surveillance by using Link Plus software to link antiviral resistance surveillance and FluSurv-NET data by patient county of residence, age, and sex and specimen collection date.

We considered identical matches on all 4 variables as a high probability match, e.g., a patient from FluSurv-NET who had a pandemic (H1N1) 2009 virus specimen submitted to national antiviral resistance surveillance who had an oseltamivir-susceptible pandemic (H1N1) 2009 virus infection. We validated our linking methods with Oregon data (n = 41); all 4 patients identified as high probability matches were true matches. For validation purposes, we identified 4 specimens that were matched on county, age, and sex but not on specimen collection date up to 7 days, e.g., moderate probability matches; 1 patient was hospitalized, 2 were outpatients, and 1 specimen was from a medical examiner (patient not hospitalized). The Oregon surveillance specimens that were neither high nor moderate probability matches were surveillance specimens from outpatients and cluster investigations (M. Vandermeer, pers. comm.).

Overall, 6,740 virus isolates and specimens were submitted to surveillance systems; 37 (0.5%) viruses were oseltamivir resistant (3); 18 were identified by NI, contained the H275Y mutation, and were susceptible to zanamivir and resistant to adamantanes; the 19 remaining viruses were detected by pyrosequencing for H275Y. Oseltamivir-susceptible viruses exhibited IC₅₀ values ranging from 0.05 to 1.44 nmol/L. Oseltamivir-resistant viruses exhibited a median IC₅₀ value of 80.08 nmol/L (range 6.24–116.48 nmol/L).

Most patients infected with oseltamivir-resistant pandemic (H1N1) 2009 viruses were hospitalized (81%), had a severe immunocompromising condition (76%), and had been exposed to oseltamivir before collection of the specimen tested for antiviral resistance (89%) (Table); 9 (30%) had received oseltamivir as chemoprophylaxis, and 21 (70%) had received oseltamivir as treatment. Four patients with oseltamivir-resistant pandemic (H1N1) 2009 virus infection had no documented exposure to oseltamivir before collection of the specimen for testing, including exposure to family members receiving oseltamivir. No epidemiologic links were found between the 4 patients.

Among the 28 patients infected with oseltamivir-resistant pandemic (H1N1) 2009 virus, with a severe immunocompromising condition and a complete case form, 24 (86%) had a malignancy reported, 23 had a hematologic malignancy and were receiving chemo- or immunosuppressive therapy at the time of their infection, and 10 (38%) were recipients of a hematopoietic stem cell transplant (SCT). One patient had AIDS and a lymphoma of the central nervous system. Among the 3 immunosuppressed patients without a malignancy, 2 were recipients of solid organ (renal) transplants, and another had received SCT <6 months before influenza illness (reason for SCT could not be confirmed).

Among 1,982 national surveillance oseltamivir-susceptible specimens from the 16 FluSurv-NET states, 1,607 (81%) had county information; among these, 401 (25%) specimens were from FluSurv-NET counties, and 65 patients from FluSurv-NET were high probability matches to patients identified in antiviral resistance surveillance data (Table). Compared with patients with oseltamivir-resistant pandemic (H1N1) 2009 infections identified in national surveillance, few (11%) FluSurv-NET patients with an oseltamivir-susceptible pandemic (H1N1) 2009 virus infection had severely immunosuppressive conditions, and few (14%) had oseltamivir exposure before collection of the specimen for testing, none were reported to have received oseltamivir as chemoprophylaxis. Among all 8,740 FluSurv-NET hospitalized patients with pandemic (H1N1) 2009 during this period, 10% had an immunosuppressive condition. Patients with oseltamivir-resistant infections had specimens for testing collected a median of 11 (range 2–37) days after oseltamivir initiation, and results may reflect testing due to clinical suspicion of resistance. Among the 6 FluSurv-NET patients with specimens collected after oseltamivir was begun, the median time between oseltamivir initiation and specimen collection was shorter.

Infections with oseltamivir-resistant pandemic (H1N1) 2009 viruses were rare in the United States during April 2009–June 2010. Few patients had no oseltamivir exposure before resistant virus was detected, and none had epidemiologic links to another patient. Thus, evidence for community transmission of oseltamivir-resistant pandemic (H1N1) 2009 viruses was rare (7). Patients with severe immunocompromising conditions with prior exposure to oseltamivir were most likely to have an oseltamivir-resistant infection. Infections were most frequently reported in patients with hematologic cancers who were undergoing immunosuppressive treatment, chemotherapy, or SCT. Other studies have also reported a high frequency of patients with hematologic malignancies or SCT and oseltamivir exposure among patients with oseltamivir-resistant pandemic (H1N1) 2009 virus infections (8,9). Oseltamivir resistance should be considered among patients with severe immunocompromising conditions and pandemic (H1N1) 2009 in the setting of oseltamivir treatment or chemoprophylaxis failure.

Although the number of patients with oseltamivir-resistant pandemic (H1N1) 2009 virus infections was small in the United States during this period, this is the largest case series published and confirms findings from reports with smaller samples (8–10). Although all patients in our comparison group of patients with oseltamivir-susceptible pandemic (H1N1) 2009 were hospitalized, most patients in the oseltamivir-resistant group were also hospitalized. Finally, we do not have a comparison group of patients with immunocompromising conditions and oseltamivir-susceptible pandemic (H1N1) 2009 virus infections; thus, risk factors for infection with oseltamivir-resistant infection among patients with immunocompromising conditions cannot be determined. The finding of oseltamivir-resistant pandemic (H1N1) 2009 viruses associated with oseltamivir treatment highlights the need for new antiviral agents and new treatment strategies.