Rapid Selection of Oseltamivirand Peramivir-Resistant Pandemic H1N1 Virus during Therapy in 2 Immunocompromised Hosts

The 2009 H1N1 influenza virus pandemic has highlighted the limited armamentarium of antivirals available to treat influenza A, especially for those at high risk of severe disease and complications of influenza. Oseltamivir, one of the neuraminidase inhibitors (NAIs), has been the drug of choice for treatment [1]. Early in the pandemic, reports of resistance to oseltamivir were limited, but recently a number of isolates have been detected that contain the H275Y mutation in the neuraminidase gene (NA), which confers a significant reduction in sensitivity to oseltamivir [1–4].

In addition to oseltamivir, 2 other antivirals are available: zanamavir, which is a US Food and Drug Administration-approved inhaled NAI, and peramivir, which is an intravenous NAI released under an emergency use authorization [5]. In vitro laboratory testing of seasonal H1N1 isolates containing the H275Y mutation have shown an increase in 50% inhibitory concentration (IC50) to peramivir [6, 7], but the demonstration of clinically significant resistance to peramivir in an individual patient has not yet been described.

Selection for resistance mutations in immunocompromised individuals infected with influenza A viruses has recently been reported [2, 8–10]. In this report, we describe 2 cases of influenza in immunocompromised hosts infected with the 2009 pandemic H1N1 virus who were treated with an extended course of NAIs.

Case reports. Patient A was a transfusion-dependant 24-year-old woman who was treated for myelodysplastic syndrome with a matched unrelated peripheral blood stem cell transplantation in December 2008, which was complicated by post-transplantation lymphoproliferative disorder and graft versus host disease. She recently received a stem cell boost and courses of corticosteroids, rituximab, and tacrolimus during the 2 months before infection with influenza A virus.

The patient presented on 24 October 2009 with fever (temperature, 38.5°C), coryza, myalgias, productive cough with clear sputum, and loose stool. She had course breath sounds bilaterally, her heart rate was 120 beats per min, and she was tachypnic with an oxygen saturation of 89%. A chest radiograph demonstrated a right-sided infiltrate, and a complete blood cell count showed slight lymphopenia. The patient was treated with oxygen, empirical oseltamivir (administered at a dosage of 75 mg twice daily), and piperacillin-tazobactam. Rapid test results were positive for influenza A virus and negative for all bacterial cultures. The patient required oxygen for ∼8 days, and chest computed tomography (CT), performed on day 6, showed bilateral patchy infiltrates. The patient was treated for 30 days continuously with oseltamivir and remained symptomatic, but respiratory symptoms began to improve after day 21 of therapy. All additional nasopharyngeal wash samples and a bronchioalveolar lavage sample remained positive for influenza A virus by viral culture until 44 days after the initial diagnosis was made (Table 1).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Samples and Isolates Collected from Patients A and B

Patient B was a 49-year-old man who underwent matched unrelated donor peripheral blood stem cell transplantation in March 2009 to treat recurrence of diffuse large B cell lymphoma. The patient had recently been treated for reactivation of cytomegalovirus infection and graft versus host disease, and he was receiving medications that included sirolimus and prednisone. On 22 October 2009, this patient presented with mild symptoms of upper respiratory tract infection. He was afebrile and breathing comfortably. Chest radiographs showed no evidence of disease, and complete blood cell count revealed lymphopenia. The respiratory virus culture was positive for influenza A virus at 24 h, and the patient was subsequently treated with 75 mg of oseltamivir administered twice daily.

After 14 days of oseltamivir therapy, the patient was admitted to the hospital with worsening fatigue, cough, sinus pressure, and significant lower extremity edema. Radiographs and chest CT demonstrated extensive bilateral patchy infiltrates. Cultures of samples from both a nasal wash and bronchoscopy were positive solely for influenza A virus. Oseltamivir therapy was continued, and levofloxacin was added empirically.

After 24 days of continuous oseltamivir therapy, the patient developed respiratory distress and was admitted to the intensive care unit for treatment with noninvasive positive-pressure ventilation. A second bronchoscopy was performed that yielded specimens that were culture-positive for influenza virus. Further empirical antibiotics were added to the patient's treatment regimen, and 10 days of peramivir were administered intravenously. Thirty-one days after the diagnosis was made, the patient's severe symptoms had stabilized, but the patient remained symptomatic, and nasopharyngeal wash samples remained positive for influenza A virus on day 32 and day 40. The patient then received 10 days of inhaled zanamivir therapy; by day 46, a nasopharyngeal wash specimen was obtained that was negative for influenza A virus, and the patient experienced an overall improvement in symptoms.

Methods. Samples were collected with the consent of participants under an National Institute of Allergy and Infectious Diseases Institutional Review Board-approved protocol, “Influenza in the Normal and Immunocompromised Host” (institutional review board number 07-I-0229). Viral isolates were obtained using shell vial culture technique. Viruses were passed 1–2 times in Madin-Darby canine kidney cells, as previously described [11]. Reverse-transcription polymerase chain reaction, and sequencing was performed on the primary viral isolates, as previously described [12]. NA sequences determined have been deposited in GenBank (GenBank accession numbers GU571152-GU571156). Antiviral susceptibility and NA activity were measured using methyl-umbelliferyl-N-acetyl neuraminic acid substrate, as previously described [10, 13], and IC50 was determined by regression analysis (Prism; GraphPad Software). A/California/07/2009 (H1N1) was used as control.

Results. Viral isolates collected from Patient A on day 0 and day 5 contained the wild-type histidine at position 275 of the NA gene. All other isolates after day 9 contained the H275Y mutation (Table 1). Viral isolates collected on day 0 from Patient B contained the wild-type NA, but all additional isolates contained the H275Y mutation. No other changes were observed in the NA gene (Table 1).

IC50 values for viral isolates collected containing the wildtype NA and H275Y mutation are shown in Table 2. The H275Y-containing isolates from both patients show a >200-fold increase in IC50 to oseltamivir and 50-fold increase to peramivir, indicating a significant loss in susceptibility to both antivirals. No significant change in IC50 to zanamivir was observed.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2

Neuraminidase Activity and Neuraminidase Gene Inhibition Assay

Discussion. Selection for the H275Y mutation during therapy for pandemic H1N1 infections of 2 immunocompromised hosts with prolonged illness was previously demonstrated early in the pandemic, but only after >24 days of continuous therapy [2]. The 2 cases described here are the first to demonstrate a rapid selection of the H275Y mutation, in <9 and 14 days of oseltamivir therapy for Patient A and Patient B, respectively. Oseltamivir treatment failed clinically in both cases, and peramivir failed to reduce shedding in 1 of these individuals. This is, to our knowledge, the first described case of clinically significant resistance to peramivir correlating to the measured IC50 in vitro.

The IC50 that signifies NAI resistance may differ between viruses and drugs tested, but a change in IC50 of 10-fold or greater between a single virus before and after treatment is commonly considered to be the hallmark of resistance [14]. Isolates collected from these patients that contained the H275Y mutation in the NA gene clearly meet this definition and demonstrate a significant increase in IC50 to both oseltamivir and peramivir

The clinical failure of therapy in both Patient A and Patient B further strengthens the importance of this change in IC50. Patient A showed signs of illness and viral shedding for 45 days, and Patient B progressed to a more severe respiratory illness with signs of viral pneumonia, despite continuous oseltamivir therapy for the first 30 and 24 days respectively. Patient B then received a 10-day course of intravenous peramivir that failed to reduce viral shedding.

In both cases, these individuals were significantly immunocompromised at the time of infection, which likely contributed to the failure of the antivirals. These 2 cases were observed during a 2-month period in which just 7 other cases of pandemic influenza in immunocompromised individuals were enrolled at the National Institutes of Health Clinical Center. This high frequency of resistance development is concerning, because it might suggest that the selection for multidrug-resistant viruses in immunocompromised hosts may be more common than previously believed.

It has become fairly common for immunocompromised patients at higher risk of severe complications of influenza infection to receive a longer course of oseltamivir therapy than the recommended 5 days if they remain ill or continue to shed. These 2 cases, together with the reports of similar resistance selection in immunocompromised patients [2, 8–10], suggest that we may need to reevaluate our usage of NAIs. Demonstration of rapid selection of resistance during therapy, in as fast as 9–14 days in these cases, or as fast as 5 days in seasonal influenza [10], suggests that prolonged therapy with a single agent may provide conditions that are optimal for the development of drug resistance mutations. Therefore, discontinuation of oseltamivir therapy or switching to zanamivir if possible after the recommended 5 days may be an appropriate strategy to prevent these mutations.

The rapid selection of the H275Y mutation leading to clinical failure of peramivir to reduce shedding in one of these cases is a significant finding and suggests that we need to investigate further the effectiveness of peramivir as a treatment for patients who experience failure of oseltamivir therapy. The rapid emergence of oseltamivir and peramivir resistance in already amantadine- resistant pandemic H1N1 viruses highlights the difficulties faced in treating influenza [15], especially in those patients who are susceptible to prolonged infection, such as immunocompromised individuals. These results indicate that additional investigation of drug resistance and the development of new classes of antivirals is imperative to reduce the adverse impact that current and future influenza pandemics can have on human health.

• Received December 29, 2009.
• Accepted January 29, 2010.