Although the ferret model is an accepted model for studying influenza virus pathogenicity and vaccine efficacy 
, a statistical analysis comparing the changes of various biological parameters that can be used to define clinical profiles and predictors of survival/mortality as a result of infection, has not been published in the literature. We compared data from 269 male ferrets from 16 studies over a five-year period to evaluate changes in various biological parameters associated with influenza disease with the aim of describing a clinical profile and correlates of protection between three distinct influenza viruses. High mortality (approximately 93%) was associated with ferrets infected with HPAI H5N1 A/Vn/1203/04. Other studies performed in our laboratory have demonstrated various mortality rates associated with other HPAI viral strains, but the highest mortality rates have been associated with H5N1 A/Vn/1203/04 (data not shown). We did not observe mortality in ferrets challenged with various seasonal or swine influenza viruses. The difference in virulence may be associated with various viral or host system factors. For instance, different levels of viremia and the amount of viral replication within the specific host animal tissues may be a factor. A/Vn/1203/04 has been shown to have a lower infection 8 hours post-infection when compared to other influenza virus strains 
. However, H5N1 A/Vn/1203/04 replicated to titers similar to those of the other influenza virus strains 24 hours post-infection and induced higher levels of cell necrosis when compared to other influenza viruses 
. This published report also demonstrates a unique ability to recover H5N1 A/Vn/1203/04 from both the apical and basolateral surfaces of cells, which may be indicative of a virulence factor that is associated with this virus 
. Elevated mortality rates have also been correlated with elevated viral replication and viral load, resulting in extreme cytokine production 
. Moreover, H5N1 A/Vn/1203/04 has been shown to induce elevated levels of proinflammatory cytokines when compared to other influenza viruses 
. Additionally, the presence of the proper influenza virus receptors on host cells and the specificity of the various influenza viruses to these receptors may impact virulence 
. All of these factors likely contribute the virulence of A/Vn/1203/04 when compared to other viruses.
Data from our study suggest that statistical differences in several clinical parameters occur when comparing ferrets infected with HPAI, seasonal, or swine influenza viruses. For instance, HPAI-infected animals experience higher body temperatures, a long duration in fever, greater weight loss, less activity, and increased viral shedding. Additionally, this study demonstrates that temperature increases, thrombocytopenia, leukopenia, and lymphopenia are not only hallmarks of disease in humans 
, but may also be clinical profiles associated with disease in ferrets. Additionally, elevated AST, ALT, SDH, along with a decrease in eosinophil, monocyte, and neutrophil counts, as well as the observed thrombocytopenia, leukopenia, hypocalcemia, and lymphopenia can be used delineate the clinical profile for HPAI infection in ferrets. Changes in creatinine, total protein, BUN, chloride, and hypocalcemia suggest potential kidney damage in HPAI-infected ferrets. When comparing these levels to ferrets infected with seasonal or swine influenza, it appears that HPAI infection leads to more liver dysfunction. Thus, renal failure or decreased kidney function also appears to be biomarker of HPAI disease and may lead increased mortality when compared to the other influenza viruses tested. Additionally, changes observed to various clinical pathology profiles suggest HPAI infection in ferrets may lead to more severe anemia when compared to ferrets infected with seasonal or swine influenza viruses. Elevated RBC and HCT, hepatocyte damage, characterized by elevated AST, ALT, and SDH, a decrease in activity, fever, and weight loss are all clinical profiles associated with influenza disease. Thus, it appears that a clinical profile or disease signature is specific to HPAI, seasonal, or swine influenza infection. Additionally, evaluating such parameters on specific days post-infection, leads to a further characterized disease state in which HPAI, seasonal, and swine influenza disease appear to have a distinct disease signature. Furthermore, when evaluating mortality and survival associated with HPAI infection, specific clinical profile parameters can be used as correlates of protection (). A statistical analyses of HPAI-infected ferret data collected over a 5 year period demonstrates that increased thrombocytes and lymphocytes, three days post-challenge, may be used to predict whether an animal will survive. Thus, the data suggest that the total number of thrombocytes and lymphocytes influence survival when ferrets are infected with HPAI. Moreover, maintaining body weight and mean corpuscular volume [(MCV) average RBC volume], three days post-infection, may lead to a greater chance of survival (). HPAI-infected animals show signs of anorexia and begin to lose weight. However, the data suggest that maintaining body weight correlates with survival of the animal, which is a typical correlation associated with most disease states. Statistical analyses of these parameters show significant correlation with survival in the HPAI survival model since the positive slope estimate suggests that if there is a smaller decrease from baseline for a representative parameter, then the animal has a greater probability of surviving infection.
Correlates of Survival in HPAI-Infected Ferrets.
Severe human disease due to influenza virus infection results in lymphopenia, leucopenia, fever, anemia, and changes in clinical pathology. Complete analyses of clinical chemistry profiles in humans infected with swine influenza have been described 
and the observed changes in this current study add credence to the ferret model. Though many changes to the host system, as a result of influenza virus infection, have been observed, a statistical comparison using data from multiple studies on signs of disease and changes associated with body temperature, clinical pathology, and virus shedding has not been reported. Thus, this study aimed to describe the clinical profile associated with disease in ferrets caused by the various influenza viruses and to determine whether correlates of protection could be identified. The identification of such correlates of protection could shape the strategic targets of novel therapeutics and prophylaxis. For example, statistical analyses of the data presented here suggest that the survival of ferrets infected with various influenza viruses exhibited greater lymphocyte numbers, less change in MCV, higher platelet numbers, and overall higher body weights. Individually, these parameters may not affect the survival of a ferret infected with influenza viruses. However, collectively, these parameters have been statistically identified as correlates of protection and thus, may represent targets for novel therapeutics and prophylaxis.
In conclusion, establishing the clinical profile or disease signature associated with influenza disease is necessary to establish correlates of protection. Previous published work has suggested that virulence factors and mortality associated with various influenza viruses may correlate with several host and viral factors including: the presence of influenza receptors on cells; temperature changes; induction of cellular necrosis; viremia and viral titers in host tissues; and the induction of a severe immune response 
. Our study suggests that the severity and duration of febrile temperatures, overall lymphocyte and platelet numbers, changes in MCV, and overall body weight associated with the host animal after influenza infection may also have a role in the pathogenesis and disease state, which may serve as correlates of clinical disease in ferrts. Furthermore, delineation of the influenza clinical profile can be used to establish proper endpoints in studies designed to test vaccine or therapeutic/prophylaxis efficacy. In all, a statistical analysis of influenza data from multiple studies suggest clinical parameters that correlate to HPAI disease in ferrets and helps validate the use of ferrets as a model system to study influenza pathogenesis and evaluate product efficacy.