Browsing by Subject "Influenza"

Background: Acute febrile respiratory illnesses, including influenza, account for a large proportion of ambulatory care visits worldwide. In the developed world, these encounters commonly result in unwarranted antibiotic prescriptions; data from more resource-limited settings are lacking. The purpose of this study was to describe the epidemiology of influenza among outpatients in southern Sri Lanka and to determine if access to rapid influenza test results was associated with decreased antibiotic prescriptions.

Methods: In this pretest- posttest study, consecutive patients presenting from March 2013- April 2014 to the Outpatient Department of the largest tertiary care hospital in southern Sri Lanka were surveyed for influenza-like illness (ILI). Patients meeting World Health Organization criteria for ILI-- acute onset of fever ≥38.0°C and cough in the prior 7 days--were enrolled. Consenting patients were administered a structured questionnaire, physical examination, and nasal/nasopharyngeal sampling. Rapid influenza A/B testing (Veritor System, Becton Dickinson) was performed on all patients, but test results were only released to patients and clinicians during the second phase of the study (December 2013- April 2014).

Results: We enrolled 397 patients with ILI, with 217 (54.7%) adults ≥12 years and 188 (47.4%) females. A total of 179 (45.8%) tested positive for influenza by rapid testing, with April- July 2013 and September- November 2013 being the periods with the highest proportion of ILI due to influenza. A total of 310 (78.1%) patients with ILI received a prescription for an antibiotic from their outpatient provider. The proportion of patients prescribed antibiotics decreased from 81.4% in the first phase to 66.3% in the second phase (p=.005); among rapid influenza-positive patients, antibiotic prescriptions decreased from 83.7% in the first phase to 56.3% in the second phase (p=.001). On multivariable analysis, having a positive rapid influenza test available to clinicians was associated with decreased antibiotic use (OR 0.20, 95% CI 0.05- 0.82).

Conclusions: Influenza virus accounted for almost 50% of acute febrile respiratory illness in this study, but most patients were prescribed antibiotics. Providing rapid influenza test results to clinicians was associated with fewer antibiotic prescriptions, but overall prescription of antibiotics remained high. In this developing country setting, a multi-faceted approach that includes improved access to rapid diagnostic tests may help decrease antibiotic use and combat antimicrobial resistance.

Influenza causes an estimated 3 to 5 million cases of severe illness annually, along with substantial morbidity and mortality, particularly among low and middle-income countries (LMICs). The most effective way to decrease the burden of influenza is vaccination. Currently, Sri Lanka has no influenza vaccination policies and does not offer vaccination within the public healthcare sector. Therefore, a cost-effectiveness analysis of influenza vaccine implementation for the Sri Lankan population was performed. A static Markov model that did not account for transmission dynamics was used for this study. The model followed a theoretical cohort of Sri Lankans from all ages through two potential scenarios: universal influenza vaccination and no influenza vaccination across twelve-monthly cycles. Cost-effectiveness was analyzed using a governmental perspective at the national level. Costs for the study were broken down into three categories: direct, indirect, and vaccine costs. All costs were identified from previous literature for the local context. Vaccine efficacy was expressed as a range (48% to 72%) identified from a previous meta-analysis investigating similar settings. One model arm was considered cost-effective if the ICER was below a three-fold gross domestic product (GDP) per capita per DALY averted limit and highly cost-effective if below a one-fold GDP per capita per DALY averted limit. Utilizing TreeAge Pro software, we conducted both probabilistic sensitivity analyses and one-way sensitivity analyses for all model variables. The vaccination model arm reduced all influenza outcomes by approximately 60% (170,283 episodes, 3,167 hospitalizations, and 152 deaths) compared to no vaccination. By implementing vaccination earlier in the year, this reduction in the influenza disease burden was maximized. Vaccination was considered cost-effective compared to no vaccination, with a base case incremental cost-effectiveness ratio (ICER) estimated at Rs. 968,071.45 /DALY (5,418.62 USD/DALY). Sensitivity analysis identified that the ICER was sensitive to implementation month, monthly probability of contracting influenza, cost of vaccination, and years of life disabled. Due to only considering a one-year period, the implementation month had the most substantial effect on the ICER because no potential rollover effects of vaccination could be seen for the later implementation months. Probabilistic sensitivity analyses were performed on all variables, and there was a 99% probability that vaccination was cost-effective below a WTP threshold of 1,157,047.92 Rs/DALY (6,476.38 USD/DALY). No value for a variable within our estimated ranges resulted in ICERs above the WTP threshold of Rs. 2,066,157 (USD 11,556) per DALY averted. In conclusion, vaccination was considered cost-effective when compared to the implementation of no vaccine. However, due to a lack of national data, large-scale national studies are needed to determine better the influenza disease burden, at-risk population, and implementation cost.

Peptide-based vaccines have received growing interest due to their specificity and ability to limit off-target effects, and they are currently being explored toward a variety of infectious diseases and therapeutic targets. However, the efficacy and applicability of such epitope-based vaccines are currently limited by difficulties in predicting immunogenic epitopes in outbred populations and a reliance on carrier proteins and adjuvants that can cause pain and swelling. Current vaccine platforms are further limited in their ability to combine multiple different epitopes, making it difficult to adjust humoral and cellular responses systematically. A vaccine platform containing broadly reactive T-cell epitopes that boosts responses to co-delivered antigens with minimal inflammation could address these limitations. To that end, the focus of this dissertation was to create peptide epitopes that can be incorporated within a supramolecular nanomaterial platform, together acting as a nano-adjuvant, a term that we will use here to describe materials whose adjuvanting properties depend on their nanoscale structure. To achieve this, we took inspiration from a class of materials termed glatiramoids, which promote anti-inflammatory and TH2 immune responses. We created an immunomodulatory supramolecular nanomaterial system inspired by the randomized nature of glatiramoids termed KEYA-Q11. By creating a glatiramoid-like peptide library integrated within self-assembling Q11 nanofibers, numerous epitopes can be presented simultaneously along the nanofibers for maximum antigen presenting cell uptake and activation. The first half of this document (Chapters 3 and 4) describes how this nanomaterial increased immunogenicity of co-assembled epitopes while also creating a KEYA-specific non-inflammatory response to the randomized component. Additionally, capitalizing on the potential for KEYA-Q11 to amplify immune responses to co-assembled epitopes, this technology is applied in the second half of this document (Chapters 5 and 6) to an epitope-based influenza vaccine. Initially we designed and synthesized a self-assembling nanomaterial inspired by glatiramoids and evaluated its TH2 T-cell polarizing properties (Chapter 3). Glatiramoids raise strong, protective immune responses in patients and have been examined in a variety of contexts from Multiple Sclerosis to HIV. However, due to their randomized polysequence structure, it remains challenging to incorporate glatiramoids into other materials and strategies to optimize them for specific therapeutics. Therefore, we designed a polysequence peptide sequence and synthesized it onto the chemically defined, supramolecular Q11 nanofiber platform to straightforwardly titrate it into other nanomaterial formulations. This polysequence nanomaterial was termed KEYA-Q11 for the four amino acids, lysine, glutamic acid, tyrosine, and alanine, that comprise its structure. Due to the extensive number of possible KEYA sequences, multiple batches of KEYA-Q11 were first examined with an array of biophysical characterization techniques to confirm reproducible synthesis and assembly. The optimal number of polysequence amino acid additions was determined to be 20 amino acids as (KEYA)20Q11 could reliably be synthesized and raise strong Type 2/TH2/IL-4 immune responses. Moreover, by modulating the concentration of KEYA-Q11 in a Q11 immunization, the strength of KEYA-specific B-cell responses were similarly altered. KEYA modifications dramatically improved uptake of peptide nanofibers in vitro by antigen presenting cells and served as strong B-cell and T-cell epitopes in vivo, inducing a KEYA-specific Type 2/TH2/IL-4 phenotype. KEYA modifications also increased IL-4 production by T cells, extended the residence time of nanofibers, and decreased overall T cell expansion compared to unmodified nanofibers, further suggesting a TH2 T-cell response with minimal inflammation. Subsequently, we exploited the modularity of the self-assembling system to maximize application of KEYA-Q11 as a nanoscale adjuvant without inflammation (Chapter 4). Adjuvants are commonly required to raise strong immune responses to peptide therapeutics, but often induce swelling and pain at the injection site and typically drive immune phenotype. Relative to common adjuvants, KEYA-Q11 had no detectable injection site swelling and was more effective at raising humoral responses despite a genetically diverse in vivo population. Furthermore, when combined with peptide epitopes KEYA-Q11 augmented antibody production against co-assembled B-cell epitopes for cytokine TNF, D-chiral MMP cross linker, and a conserved segment of the M2 influenza protein, and increased T-cell stimulation specific to co-assembled T-cell epitopes PADRE and a conserved segment of the nucleoprotein of influenza. Likewise, when combined with the influenza surface protein hemagglutinin, KEYA modifications strengthened the resulting influenza-specific cellular immune responses. Augmented immune responses typically followed native epitope polarization, as in a co-assembly of KEYA-Q11 and the nucleoprotein epitope raised Type 2/TH2/IL4 producing KEYA-specific responses and magnified the Type 1/IFN producing nucleoprotein-specific responses that epitope would produce without an adjuvant, and thus using KEYA-Q11 as the adjuvant allowed for finer control over immune phenotype. Building on the success of KEYA-Q11 as a nano-scale adjuvant without inflammation, we utilized these properties to decrease the severity of influenza infection and provide broad protection via immunization with peptide epitopes (Chapters 5 and 6). Much of the current focus on influenza vaccines revolves around partial or whole proteins to induce broadly protective antibodies, while other have demonstrated cross-reactive T-cell responses are vital for heterologous protection. Conserved peptide epitopes have been discovered but typically are included with larger proteins and adjuvants to increase immunogenicity. Supramolecular assemblies based on the Q11 peptide system containing KEYA, a B-cell epitope from a conserved surface protein on influenza, and CD4+ and CD8+ T-cell epitopes from influenza nucleoprotein and polymerase acidic protein, respectively, raised strong immune responses against all three epitopes. Inclusion of the KEYA component in prophylactic immunizations with these materials significantly improved protection following a lethal influenza challenge. It has been established that while peptide-based immunotherapies can have finely directed specificity for chosen epitopes, they generally lack sufficient immunogenicity to provoke suitable immune responses. This new strategy for augmenting immune responses to peptide-based therapeutics, especially those employing nanomaterials, and especially for applications where non-inflammatory responses are prioritized, can be employed for a variety of potential applications in vaccine development, towards infectious diseases and towards non-infectious applications such as inflammatory autoimmune diseases, wound healing, or graft rejection. KEYA-Q11 is a unique fusion of two materials, a highly ordered system with a highly disordered system, and examination of this nanomaterial has provided valuable insight into both randomly polymerized structures and non-inflammatory nano-scale adjuvants.

Despite high vaccination rates and effective adaptive immune responses from the part of infected individuals, influenza A viruses cause significant morbidity and mortality annually. This is due to influenza's rapid antigenic evolution, whereby continual mutations occurring in epitope regions of the virus's hemagglutinin protein result in the diminishment of long-term antibody recognition, in a process that has been termed `antigenic drift'. Although it is clear that antigenic drift enables previously infected individuals to become reinfected, the mechanism that is responsible for influenza's antigenic drift is still under debate. As recently as 2009, a new hypothesis of antigenic drift was put forward that argues that binding avidity changes in the viral hemagglutinin result in antigenic drift as a side effect. This hypothesis stands in contrast to the traditionally accepted hypothesis that mutations in epitope regions are positively selected for their ability to evade immune recognition. This thesis focuses on the use of epidemiological models and empirical data analysis to explore different hypotheses of antigenic drift.

In the first chapter, I am asking what effects on antigenic drift rate would be produced under the new hypothesis. I mathematically formulate the hypothesis that antigenic drift is simply a side effect of cellular receptor binding avidity changes that occur as the virus is transmitted between individuals of different immune status levels. I then use this formulation to explore how influenza's rate of antigenic drift depends on different epidemiological factors, including host contact rate, host lifespan, and the duration of infection. Finally, I use the model to assess alternative vaccination strategies by the impact they have on rates of antigenic drift and therewith rates of disease incidence/

In the second chapter, I critically evaluate the binding avidity hypothesis by comparing predictions of the hypothesis against empirical data. I first use a `phylodynamic' extension of the model presented in the first chapter to determine whether the hypothesis is consistent with the ladderlike phylogeny of influenza's hemagglutinin protein. I then use viral sequence data and metadata to determine whether older aged individuals (with a higher number of previous infections) harbor viruses with higher binding avidity than younger aged individuals (with a lower number of previous infections), a prediction made by the binding avidity hypothesis. Finally, I perform a phylogenetic analysis to determine how rapidly binding avidity changes occur. From these analyses, I conclude that the binding avidity hypothesis is not well supported by empirical data.

In the third chapter, I develop an integrated viral life cycle model, in which viral replication depends on three viral phenotypes: receptor binding avidity, neuraminidase activity, and antigenicity. This integrated model recognizes that receptor binding avidity changes will influence viral replication, but also allows for antigenic evolution to be brought about directly by epitope changes. I first use this model to show how the evolutionary dynamics of these phenotypes are dependent on one another and how antigenic drift can be interpreted within this framework. I then return to some of the questions addressed in the first chapter to ask how different epidemiological factors impact influenza's rate of antigenic drift.

Together, these three chapters highlight the importance of viral phenotypes other than antigenicity in contributing to influenza's antigenic evolution, and, more generally, the importance of computational and mathematical research in understanding constraints on viral adaptation.

The evolution of antigenically variable viruses cannot be understood without studying the interaction between viruses and the host immune system. Viral evolution is driven by their fast acquisition of genetic variation as well as by the strong selection imposed by the host immune response. Moreover, understanding viral evolution dynamics and its interplay on the host immune response can provide essential information for vaccine development. In this dissertation, I use an integrative approach to study various aspects of this interplay in two viral systems: influenza A (IAV) and cytomegalovirus (CMV), both ubiquitous in humans and significant public health threats.

Congenital cytomegalovirus infection is the leading infectious cause of congenital defects. As such, the study of viral dynamics is essential to develop better treatment and prevention procedures. In a monkey challenge study for congenital cytomegalovirus infection, I investigated viral transmission between maternal and fetal compartments. Using high-coverage sequencing data, I examined viral evolutionary dynamics in time and space. I found evidence of large transmission bottleneck sizes between maternal compartments and in congenital transmission. I also inquire about the role of preexistent CMV-specific antibodies in the virus population, finding no apparent effect in the viral genetic make-up but reduced viral load and also reduced congenital transmission.

One of the more promising vaccine formulation for CMV until now is the gB-MF59 vaccine, which is based on a soluble version of the immunodominant gB protein. To understand immune and viral factors contributing to vaccine efficacy for this formulation, I examined immunoglobulin G binding to a gB-specific peptide microarray from seropositive individuals and vaccinees prior to and after vaccination. The antibody profile observed from binding clustered by individual immune exposure history. While the antibody profile elicited by vaccination show high agreement with the one from seropositive individuals, I also identified regions in gB preferentially targeted in vaccinees. Moreover, I observe no difference between the antibody profiles of vaccinees with different clinical outcomes and instead found further evidence for reduced cross-immunity between divergent genotypes.

Original antigenic sin (OAS) refers to the tendency of the host immune system to focus on previously recognized viral epitopes during secondary challenges with related viral strains. This preference is sustained by antibody memory and can result in suboptimal immune protection. Mounting evidence highlights the importance of the initial viral strain encountered during childhood, which primes the antibody repertoire to contend against further infections. Here, the goal was to identify OAS-driven cohort effects in IAV driven by at least one antigenic mutation. Using sequence data and host's age information from individuals infected with H1N1 in the U.S. from 2009 to the present, we searched for potential signatures of birth year cohort changes driven by new variants with nonsynonymous mutations. We identified multiple variants with such properties and studied the age groups with differential abundance after such variant arose.