Characteristics and mortality of severe influenza cases treated with parenteral aqueous zanamivir, United Kingdom, October 2009 to January 2011

Background Aqueous zanamivir solution, an investigational product, was provided by the manufacturer on compassionate grounds for parenteral administration to severe H1N1pdm09 influenza cases during the 2009 pandemic. Objective To describe characteristics and outcomes of UK patients receiving parenteral zanamivir therapy. Methods Collaborators at multiple hospital sites gathered retrospective data on patients receiving aqueous zanamivir therapy between Q4 2009 and Q1 2011. We present analysis of the demographics, clinical features, treatment and outcomes of this cohort. Results Data on 185 cases were obtained (response rate of 38%; median age 43 years; 62% male; 17% non‐Caucasian ethnic group). Most frequent co‐morbidities included cancer, immunosuppression and respiratory conditions. Most patients received intravenous zanamivir alone (90%), for durations of up to 21 days. 13% of cases had adverse effects related to zanamivir therapy. Thirty four percentage of cases died. No significant relationship was seen between mortality and timing or route of administration of aqueous zanamivir therapy. Conclusions The response rate of this observational study of the outcomes of treatment of severe influenza was low, allowing limited conclusions to be drawn. Some potential adverse effects were noted. Clinicians should carefully consider potential risks and benefits of use of this product. New treatment options are urgently required to improve outcomes for patients with severe influenza infections.

administered by powder inhaler. 3 Nebulised administration of zanamivir inhalation powder as a liquid formulation is not recommended. 4 Few alternative antiviral formulations were available for critically ill patients with oseltamivir-resistant infections, impaired gastric motility, malabsorption and/or gastrointestinal bleeding.
Zanamivir aqueous solution is an investigational product which may be administered via nebulisation or intravenously. At the onset of the 2009 pandemic, published evidence of the safety and effectiveness of intravenous or nebulised zanamivir aqueous solution was limited and it was not licensed for use in any country. 5   Follow-up information was available for 457 of the 485 cases for whom parenteral zanamivir was issued.
The recommended adult dosage of intravenous aqueous zanamivir was 600 mg twice daily for 5 days; adjustments were required for renal impairment. 7

| RE SULTS
Data were returned for 185 cases (out of 457 questionnaires sent; response rate 38%). Thirty-four sites each provided data for between 1 and 21 cases (median three cases). Data completeness was >95% for the majority of key variables.

| Treatment
Most cases (164 cases; 89%) received antiviral therapy prior to aqueous zanamivir therapy (where specified, oseltamivir was given for all but two cases, who received inhaled zanamivir therapy). Duration of prior antiviral therapy was reported for 73%: 72 cases (60%) received therapy for 5 days or less, 40 cases (33%) for 6-14 days and eight cases (7%) for greater than 14 days.

| Mortality
Outcomes were recorded for 175 cases (95%). Of these, 97 cases (55%) recovered, 18 cases (10%) recovered with permanent sequelae and 60 cases (34%) died. Death certificate data were available for all 60 deaths. Influenza was recorded as a primary cause of death for 37 cases (62%) and pneumonia, pneumonitis, bronchopneumonia or acute respiratory distress syndrome of unspecified cause for a further nine cases (15%). For one patient, influenza was a secondary cause of death. Influenza or compatible syndromes were not recorded among the causes of death for five cases (8%).
Age group (highest relative mortality in the 55-64 years age group), cardiovascular impairment, complications of pneumonia, renal replacement therapy, corticosteroid therapy and high SOFA scores were significantly associated with mortality ( Table 2). Nonserious adverse events temporally related to zanamivir were associated with significantly reduced mortality, and serious adverse events temporally related to zanamivir were not significantly associated with mortality.
Among those who died median time to zanamivir treatment was 13 days, compared with 10 days for those who survived (P = 0.22).   In the event of a future potential pandemic, information on novel treatments needs to be collected and analysed in real time in order to inform the response to a pandemic, using randomised, controlled trials where possible. Future evaluations of investigational products during influenza pandemics will require standardised collection of high-quality data to inform the evidence base, ideally planned in advance allowing for adequate staff training, and undertaken prospectively; a challenge that is currently being addressed by the PREPARE 24 and ISARIC 25 projects. New treatment options are urgently required to improve outcomes for patients with severe and life-threatening influenza infections.

ACK N OWLED G EM ENTS
The study was funded by an unrestricted grant from GlaxoSmithKline.
The funder received data and interim reports from Public Health England for information but did not influence analysis and reporting of the study.