Congress of the European Society of
Clinical Microbiology and Infectious Diseases
Each day, four key sessions will be covered and summarised here allowing you to catch up on some of the major development and discoveries that are being presented at ESCMID Global 2026. Check back each morning to see the highlights from the day before.
Laura Fuoli (Spain) began the session by highlighting how next-generation biomarkers have the potential to shift the diagnosis of invasive fungal disease (IFD) beyond culture-based methods. She explained that high mortality associated with fungal disease is due to non-specific symptoms, low clinical suspicion, delayed diagnosis and antifungal therapy. She then outlined the role of next-generation biomarkers in invasive fungal infections (IFI), including earlier detection, monitoring treatment response and predicting complications.
Fuoli outlined key biomarker types and explained the advantages of metagenomic next-generation sequencing (mNGS), including unbiased detection and resistance profiling, and highlighted promising mNGS sensitivity and specificity results from studies. She discussed how host biomarkers such as cytokines and chemokines should be incorporated into IFI diagnostic strategies. Fuoli concluded that next-generation biomarkers should include multi-biomarkers, multimodal pathogen-based diagnostic strategies and host biomarkers to transform IFD diagnosis.
Daniele Giacobbe (Italy) revisited the foundational question, “Can machines think?”, noting how the terminology has evolved, from “automata” in 1960s to present day “artificial intelligence” (AI). He explained that machine learning (ML) models learn by recognising patterns, such as pathogen characteristics. He further questioned how they differ from traditional models like logistic regression or total parenteral nutrition (TPN). Training ML models requires large datasets, yet a recent review showed that many perform similarly or worse than traditional approaches.
Giacobbe highlighted ongoing improvements in ML but stressed the challenge of explainability. He compared how the traditional “white box” models allow clinicians to understand risk factors, whereas ML often functions as a “black box”. In his closing remarks, he highlighted that AI has a potential role in fungal disease management, but future efforts require considerations of big data and better approaches to explainability.
Cecile Angebault (France) explained that while bacterial respiratory tract microbiota is well studied, fungal respiratory tract-mycobiota (RT-My) remain poorly understood. She discussed the origin of RT-My through micro-aspiration and inhalation and how fungal exposure varies with geographical location and climate. She further gave examples of the impact of climate change with an increased incidence of Aspergillus species and Coccidioidomycosis. She outlined the core RT-My, including Candida, Saccharomyces and Malassezia, and highlighted its high intra-individual variability and age-related changes. She concluded by highlighting that further research was warranted to better understand intra- and inter-kingdom microbial interactions, the role of fungi in lung carcinogenesis and new approaches to adapt RT-My.
In the last presentation, Andrés Henao (United States) discussed how corticosteroids increase vulnerability to IFI and what the optimal prophylactic strategies should be, particularly for non-human immunodeficiency virus (HIV) pneumocystis jirovecii pneumonia (PJP). He highlighted that corticosteroids independently increased the IFI risk (adjusted odds ratio 2.4), resulting in high mortality rates (p < 0.001).
Henao highlighted that pathogen-specific risk patterns and factors, including glucocorticoid type, duration and potency increase IFI risk. He proposed a PJP risk‑stratification and prophylaxis framework.
Key Takeaways:
Organised by ESCMID Fungal Infection Study Group (EFISG), European Confederation of Medical Mycology (ECMM), Portuguese Society of Infectious Diseases and Clinical Microbiology, Romanian Society for Medical Mycology and Mycotoxicology.
David Fisman (Canada) highlighted that the last year has seen an extraordinary disruption in public health governance and funding, against a background of creeping environmental change, conflicts and antimicrobial resistance.
In the most challenging contexts, the infectious disease threat environment remains severe, with tuberculosis, malaria and vector-borne diseases ranking among the greatest burdens – despite the availability of technical solutions. Modelling studies predict that mortality rates could nearly double under health divestment compared with sustained investment, reinforcing the effect of budgetary cuts – a conscious and avoidable choice. He noted that the silence of public health success is taken for granted.
In the best of times, rapid containment of new outbreaks (e.g. Marburg vs Ebola) exemplifies how more rapid deployment enables containment without further spread. In addition, the use of vaccine rollouts, novel developments in genomics to understand epidemics in real-time and vaccine research highlight the positive effects on outcomes and control, with appropriate investment and funding. He highlighted that some vaccines are emerging as tools for disease prevention beyond infections for use in dementia, oncology and vascular disease.
Kurt Hanevik (Norway) focused on key research studies from the last year. One global disease study that evaluated mortality, morbidity and risk factors reported a shift in mortality from infectious to non-communicable disease (e.g. cancer and vascular disease), primarily from conditions such as diabetes. He emphasised that a reduction in communicable diseases represents a major public health achievement.
Climate change research indicates that health risks are worsening over time, from heat-related mortality to increased infectious disease caused by improved suitability for transmission. Cholera represents an ongoing problem, particularly in Africa and Asia. Research on recent outbreaks in Sudan showed that vaccines were often deployed too late, highlighting the effect of war on health system disruption. He noted that other studies have reported microbial resistance patterns in war-torn countries such as Ukraine, stressing the importance of reducing unnecessary antibiotic prescription and antimicrobial stewardship.
Novel research has assessed high-dose influenza in the elderly and new vaccines, with a large number focused on technological advances based on outer membrane vesicles (OMV). With self-adjuvated properties and reported safety, OMV vaccines hold promise for future multivalent vaccine development.
Key Takeaways
Edward Holmes (Australia) drew on two decades of metagenomic research to demonstrate that the virosphere is far larger and more dynamic than previously understood, with major implications for how we detect and prevent future pandemics.
Large-scale sequencing of mammals has uncovered hundreds of novel viruses, demonstrating that virome composition is primarily shaped by host phylogeny. Rodents and carnivores harbour the greatest diversity, with rodents carrying many viruses capable of cross-species transmission. Crucially, some of these viruses can jump between hosts.
Using COVID-19 as a “poster child” for zoonotic emergence, Holmes presented evidence supporting spillover at the Huanan Seafood Market in Wuhan. Environmental samples clustered in areas where susceptible wildlife were sold, with metagenomic data confirming their presence. The closest known relative of SARS-CoV-2 remains a bat virus from Laos with approximately 96.8% similarity. Beyond the market, fur-farmed and other intensively farmed animals harbour diverse viruses with host-jumping potential, creating conditions for amplification and spillover.
Surveillance across bats, rodents, ticks, and other reservoirs continues to reveal an extensive pool of viruses with pandemic potential. Coronaviruses remain Holmes’ primary concern due to their ability to cross between mammalian hosts and cause respiratory disease. Furthermore, emerging threats are not geographically confined. Tick-borne viruses are also expanding in distribution, with modelling suggesting that large populations are at risk.
Holmes posed the question of how we can prevent the next pandemic, highlighting the need for a shift towards proactive prevention. This includes sustained surveillance at the human–animal interface, rapid and open data sharing, and stronger capacity in low- and middle-income countries. Artificial intelligence is expected to play an increasing role in zoonotic risk assessment, helping to prioritise which viruses pose the greatest threat.
He closed by highlighting efforts to build pre-emptive coronavirus vaccine libraries, targeting entire viral families with emergence potential rather than individual pathogens to enable faster responses to new threats. With current technology, pathogens can be identified within 24 hours and vaccines designed within 48 hours, making preparedness increasingly feasible.
Key Takeaways
In the first presentation, Carl Llor (Spain) reported the results from a study evaluating different antibiotic regimens in women with uncomplicated urinary tract infections (UTIs) in Spain.
Direct comparisons of current recommended first-line treatments are needed. This phase 4, pragmatic, multicentre, multi-arm, open label, randomised clinical trial (RCT) compared one or two fosfomycin doses with short-course nitrofurantoin and pivmecillinam regimens in 804 women with ≥1 UTI symptom.
Overall, nitrofurantoin was the most effective for treating UTIs while single-dose fosfomycin was the least effective, often requiring additional antibiotic treatment. Differences were greater among those with positive urine cultures. Adverse events were mild. Guidelines should be updated based on the available evidence.
Annalisa Paviglianiti (Spain) presented findings from a multicentre real-world retrospective study across Europe.
Maribavir (MBV) is approved for refractory/resistant cytomegalovirus (CMV) infections in adults after haematopoietic cell transplantation (HCT). Data outside the RCT setting are lacking. This registry study included 118 adult and paediatric HCT patients who received 126 courses.
MBV was given for CMV infection or disease in 51% of courses and due to toxicities in 49%. High response rates were achieved with MBV treatment (81% resolution) and with limited toxicity. Use beyond approved indications shows promising results and should be explored in further studies.
Rory De Vries (Netherlands) presented results from an observational, cross-sectional analysis in The Netherlands.
There is renewed interest in highly pathogenic avian influenza A(H5) viruses based on the potential threat of spillover events in mammals and humans.
Overall, 107 healthcare and animal workers provided blood samples for surveillance and epidemiological analysis. Low-level binding antibodies against the A(H5) haemagglutinin (HA) head were detected, but without HA inhibition activity. There was an abundance of functional antibodies targeting avian N1 and T-cells targeting avian H5 and N1. At the population level, there is partial cross-reactive population immunity against A(H5) influenza viruses, most likely through previous exposure to seasonal influenza variants.
Annelies Van Rie (Belgium) shared findings from SMARTT, to assess whether whole-genome sequencing (WGS)-guided therapy improves tuberculosis management and treatment in South Africa.
In this pragmatic, randomised, single-blind phase 4 medical device trial, 204 adults with pulmonary rifampicin-resistant (RR) tuberculosis received standard of care (SoC) or WGS-guided therapy generated by an artificial intelligence model.
In the WGS arm, the median treatment duration was reduced by 2.6 months. Unfavourable treatment outcomes (bacteriological failure, loss to follow-up or death) were less frequent with WGS versus SoC. Bacteriological response and serious adverse events were similar. WGS improved treatment outcomes and was accurate, safe and reduced costs.
Karl Hagman (Sweden) reported findings from an open-label, multicentre, randomised, controlled trial in Swedish hospitals.
Adjunctive corticosteroid treatment reduces mortality in adults with severe community-acquired pneumonia (CAP); however, the benefit in those with M. pneumoniae is currently unknown. The trial assessed the effect of 5 days of oral betamethasone (BMZ; 3 mg or 2 mg) versus SoC (without corticosteroids) in 70 hypoxemic adults (Sp02 <93%) with a verified diagnosis of M. pneumoniae CAP.
BMZ treatment significantly reduced the time to resolution of hypoxaemia and shortened the duration of hospitalisation versus SoC. Treatment was well tolerated.
Key Takeaways:
Maria Van Kerkhove (Switzerland) anchored her presentation to WHO’s mission, which remains unchanged since 1948: that the health of all people is fundamental to the attainment of peace and security, and dependent on the fullest co-operation of individuals and states. This commitment underpins the Triple Billion Targets defining WHO’s current mandate: 6 billion people living healthier lives, 5 billion accessing essential health services, and 7 billion better protected from health emergencies.
She described an increasingly complex threat landscape, where zoonotic spillover, antimicrobial resistance (AMR), conflict, climate change, demographic shifts and AI-generated pathogens represent a convergence of risks that could affect everyone, regardless of wealth or geography.
Van Kerkhove reminded the audience that COVID-19 has not gone away. As of February 2026, WHO has recorded over 779 million cases and 7.1 million reported deaths, although excess mortality estimates suggest the true toll is much higher. SARS-CoV-2 continues to evolve, with variant XFG now dominant. Beyond health, COVID-19 is estimated to have cost $16 trillion globally, disrupted education for 1.6 billion students, and pushed 135 million people into poverty. She noted that pandemics are increasing in both frequency and impact, and that COVID-19 will not be the last.
Van Kerkhove explained that WHO’s strategic response is built around the Health Emergency Preparedness and Response (HEPR) Framework, developed from over 300 recommendations and spanning three pillars: governance, financing and systems. Its legal backbone is the revised International Health Regulations (IHR) and the WHO Pandemic Agreement. The framework was adopted without a single Member State abstention, an unprecedented outcome in WHO’s history. It is supported by the Pandemic Fund and a network of over 730 international and 900 local partner organisations.
WHO processed approximately 4.5 million raw signals in 2025, identifying 472 new events with potential international health implications. This surveillance infrastructure, alongside spillover risk mapping, enables earlier detection and more targeted preparedness, helping ensure that laboratory systems and diagnostic capacity are in place where they are most needed.
Van Kerkhove closed by emphasising that rapid response is the result of sustained investment in prevention and preparedness. She stressed that science, trust, transparency, and collaboration are the foundations on which that investment must be built.
Key takeaways
In this Keynote Scientific Interview, Jon Friedland (United Kingdom) spoke with Senjuti Saha (Bangladesh) about how integrating technology and local capacity can strengthen disease surveillance and prevention in low-resource settings.
Saha began by reflecting on her early exposure to science. Growing up in a family of microbiologists, she spent time in her parents’ laboratories, looking through microscopes rather than playing with toys. From an early age, she knew she wanted to pursue a PhD, with access to science shaping her path in ways she later recognised as a privilege.
Saha described how her work has focused on building local capacity, particularly in genomics, as a foundation for effective surveillance. Without genomics, she noted, it is not possible to fully understand the pathogens circulating within a country.
Saha emphasised that sequencing is critical as disease patterns vary by region and will be further shaped by climate change, displacement and conflict. Future prevention strategies will need to be increasingly local and context-specific.
Saha described the COVID-19 pandemic as a turning point for science in Bangladesh. Within days of the first reported SARS-CoV-2 cases, her team adapted laboratory workflows, identified a positive case and generated one of the country’s first viral sequences. The work drew national attention and helped reshape perceptions of science and scientists in the country.
This shift extended beyond the laboratory. Saha began receiving messages from young people and families asking how to become scientists, prompting the creation of Building Scientists for Bangladesh. The initiative delivers hands-on science in rural communities, lab-based experiences and practical training for early-career scientists.
Saha also spoke candidly about the barriers researchers face in low-resource settings. She described them as being “scholarly poor”, constrained by journal paywalls, high publication costs, slow procurement and scientific isolation. She noted that digital tools, including large language models, can help address some of these barriers, particularly where language limits access.
Saha concluded by highlighting the broader challenge of trust in science, with misinformation and disinformation on social media continuing to undermine public confidence. For Saha, strengthening communication and engaging more effectively with policymakers and communities is essential to ensure that scientific evidence translates into real-world impact.
Key Takeaways
Christian Giske (Sweden) started the session by emphasising that while large language models (LLMs) are a form of machine learning (ML), not all ML are LLMs. He explained why LLM/ML approaches matter for antimicrobial susceptibility testing (AST), noting that interpretations are often overcomplicated by treating mechanisms as optional. He acknowledged the more recent view was that certain resistance mechanisms and model in context (MIC) modulation can influence predictions.
Giske highlighted how phenotypic AST is already complex, directing to examples of the European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines, breakpoint tables and expert rule-based interpretation. He noted ongoing work is improving how genomic data can support AST prediction and be a potential alternative to phenotypic AST. He discussed conceptual challenges of using ML for AST, such as the reference MIC used for comparison and strain diversity. MIC predictions do not add value when strains fall between susceptible and resistant categories. He presented published studies including receiver operating characteristic (ROC) curves showing variability in specificity and false susceptibility errors.
Giske then discussed recent progress with LLMs, highlighting a study that showed how commercial ChatGPT improved sensitivity and specificity for detecting antimicrobial resistance mechanisms. For genomic prediction, LLMs remain challenging but show promise in some areas (Staphylococcus aureus and Mycobacterium tuberculosis). Overall, LLMs can support rule-based interpretation and, in more advanced forms, contribute to susceptibility prediction from genomic data. He concluded by reiterating the need for high-quality reference genomic datasets from multiple geographical locations.
Andrea De Vito (Italy) followed by questioning whether LLMs truly improve clinical care, from diagnostics to patient management. He framed his discussion around four questions: how well LLMs perform, how much they can be trusted, whether their results are reproducible and what safe integration of LLMs into clinical practice might require.
De Vito presented findings from a prospective cohort study evaluating whether chatbot artificial intelligence (AI) could replace physicians in the management of bloodstream infections. The chatbot predicted correct diagnoses in 59% of cases, adequate diagnostic workup in 80%, and adequate empiric therapy in 64%, but optimal global management in only 2%. LLM failures included lack of diagnostic workup, missed contamination, inadequate source control and lack of optimal treatment plans.
De Vito then talked about LLM reproducibility concerns, using an example of a study comparing LLMs for antibiotic prescribing in different scenarios. The study found that some models performed better than others, and some newer versions of LLM underperformed earlier ones. Using a “car wash” analogy, he illustrated how LLMs can miss the real issue and vary in responses to identical prompts. He concluded that LLMs are not reliable enough to function independently and that their errors can have serious consequences for patients. Clinicians should challenge LLMs, rather than trust their first response.
Key Takeaways:
Marta Giovanetti (Italy) opened this session by highlighting the impact of climate change and mobility on the global epidemiological situation, which has seen the exacerbation of existing and the emergence of new arboviral pathogens.
Higher temperatures, increased flooding, storms and droughts, alongside increased global travel, shape transmission and drive introduction pathways towards more severe epidemics. Proactive mitigation strategies using integrated genomic surveillance, alongside climatic and environmental data, will be critical to improve early detection and preparedness, and to better understand these changes for future control.
Anna Papa-Konidari (Greece) reported that there are currently no approved treatments for West Nile virus. However, several candidates, including vaccines, are in preclinical or early clinical trials, and new diagnostic and surveillance tools are available.
Novel agents targeting different proteins (such as genomic polyproteins C, E, prM and NS proteins) disrupt the lifecycle of the virus. In addition, repurposing already approved medicines and combining therapies may improve synergy, but this is not without challenges, particularly in the elderly. She concluded that there is a need for validated animal models, animal‐to‐human translational studies, specific studies in aging populations, and testing of combination therapies to prevent drug resistance.
Felipe Naveca (Brazil) noted that although first discovered in the 1950s, the Oropouche virus (OROV) was neglected until acute febrile outbreaks were reported in recent decades in Brazil. It is now considered a priority emerging arbovirus in the Americas.
Gaps in surveillance delay detection and may amplify further outbreaks, underlining the importance of diagnostic decentralisation to enable timely intervention, along with regional coordination to prevent further spread. Dispersal dynamics have shown that OROV spread and infection coincide with extreme climactic events and is linked to agriculture in humid regions. He concurred that genomics will be vital to strengthen preparedness for growing OROV threats.
Gathsaurie Neelika Malavige (Sri Lanka) outlined that although the incidence of dengue has increased exponentially in recent decades, it is significantly under-reported due to differences in reporting practices.
Increased urbanisation and overcrowding, a reduction in the force of infection, co-circulation with other flaviviruses and arboviruses and comorbid conditions have led to increased infection and disease severity in susceptible individuals. Notably, diabetes, obesity and other metabolic conditions are associated with greater disease severity than secondary infections. Shifts from one to multiple serotypes have also occurred in recent years, coinciding with other viral infections. Additional research is needed to understand how this may affect vector competence and transmission.
Key Takeaways:
Co-organised with: SBI, Colombian Association of Infectious Diseases, Latin American Society for Travel Medicine, API, Dominican Society of ID, Ecuatorian Society for MM and Pediatric ID, ESCMID Emerging Infections Subcommittee, ESCMID Scientific Affairs Subcommittee.
Silvia Bertagnolio (Switzerland) presented results from the WHO pilot implementation of GLASS-AMR for fungal diseases, positioning invasive fungal disease (IFD) as a critical but under-recognised global health threat. She set the scene with estimates of 6.5 million infections and 3.8 million deaths annually, noting that these figures are modelled and carry uncertainty.
GLASS-Fungi was developed in response to the lack of standardised global surveillance data, enabling the systematic collection of antifungal resistance information. Building on the bacterial GLASS platform, now spanning over 100 countries, the fungal pilot included 3,447 patients across 301 sites in 13 countries, with 16 countries submitting data in 2024.
Results reveal that at least 13.6% of isolates were resistant to ≥1 antifungal agent. Resistance varied by species and geography, with higher fluconazole resistance in Candida parapsilosis and Nakaseomyces glabrata, particularly in low- and middle-income countries (LMICs). In contrast, echinocandin resistance remained generally low (<10%), and amphotericin B resistance was rare.
Bertagnolio highlighted a key data gap, with LMICs contributing less than one-third of isolates despite carrying the highest burden, reflecting limitations in diagnostic and surveillance capacity.
GLASS-Fungi is part of a broader WHO strategy, alongside the Fungal Priority Pathogens List (FPPL) and the fungal blueprint. These will soon be complemented by GLASS-EAR, a real-time early warning system for detecting high-risk fungal pathogens and outbreaks.
Nelesh Govender (South Africa) explored whether the WHO FPPL is translating into meaningful action at the country level. While the FPPL provides a framework to prioritise pathogens, guide research and align investment, he emphasised that prioritisation alone is insufficient without a clear pathway to implementation.
The WHO fungal blueprint was presented as a stepwise pathway from global priorities to national action, identifying key gaps across governance, workforce, diagnostics, surveillance and access to treatment. These gaps are underpinned by a lack of fungal disease visibility, with IFDs often being ignored, unseen and unheard.
Govender outlined the stepwise approach, with implementation progressing from inputs to impact, enabling countries to translate strategy into action. Interventions are organised across four domains: public health and health systems, innovation and therapeutics, laboratory systems and surveillance, and social and environmental drivers. Laboratory capacity and surveillance were highlighted as a key entry point, including integration into national AMR systems such as GLASS.
Using South Africa as an example, Govender highlighted the urgency of action. Candida auris fungaemia cases have more than doubled in recent years, with rising azole resistance and evidence of species replacement. He concluded that while global frameworks are now in place, impact will depend on translating these into coordinated, stepwise, and context-specific implementation.
Key takeaways
Susan Huang (United States) delivered a compelling overview of MRSA, highlighting how decades of progress have reshaped but not resolved the challenge of controlling this persistent pathogen.
Starting with the past, Huang described the rapid rise of MRSA in the 1990s, when it became a dominant healthcare-associated pathogen. Coordinated infection prevention measures reduced rates at scale, laying the foundation of modern infection control.
In the present day, MRSA remains a major cause of healthcare-associated infection (HAI). In the US alone, there are approximately 120,000 Staphylococcus aureus bloodstream infections annually, with around 20,000 associated deaths. It continues to rank among the leading pathogens across surgical site infections, pneumonia and bloodstream infections, and remains the top cause of paediatric HAIs. Even a single patient interaction can contaminate healthcare workers, regardless of role or setting.
According to Huang, patients can carry MRSA for months or years, facilitating spread across healthcare networks. In one study, around 10-14% of patients who acquired MRSA developed serious infection within a year of discharge.
Huang outlined three broad hospital strategies shaping current MRSA control. The first strategy, Relent, shifts focus from elimination to scalable, system-wide measures such as hand hygiene, antiseptic skin cleansing and device-related bundles. MRSA rates can remain stable when supported by strong surveillance, but control is partial and much of the burden lies beyond the hospital.
Target focuses on proactive prevention through surveillance. Identifying carriers enables earlier intervention, particularly decolonisation, which reduces patients’ own risk of infection. As most infections arise from a patient’s own nasal strain, targeted decolonisation can substantially reduce risk, including in surgical populations and after discharge. This approach delivers strong benefit but requires sustained investment.
The Expand strategy recognises that patients often carry multiple organisms, making pathogen-specific strategies less efficient. Universal, horizontal approaches such as antiseptic skin cleansing with or without nasal agents can reduce both MRSA and overall infection burden. In large trials, this approach delivered the greatest reductions in MRSA and bloodstream infections, while also protecting patients by reducing their own microbial burden.
Looking ahead, Huang pointed to a shift towards more precise and scalable prevention. Artificial intelligence could help identify which patients to screen, isolate or treat, enabling more targeted use of resources. Advances in decolonisation strategies, infection prevention tools and potentially vaccines also offer new opportunities.
Key Takeaways:
In the first presentation of articles with simultaneous publication, Timothy Savage (United States) reported the results from a study evaluating amoxicillin-clavulanate (AMC) versus amoxicillin (AMX) for acute sinusitis in adults.1
Savage observed that although the highest rate of antibiotic prescriptions is for infectious disease; there is no consensus on first-line treatment for acute sinusitis. This large, retrospective, US database study compared standard doses of AMC (875–125 mg BID) and AMX (875 mg BID/500 mg TID) in propensity score-matched adults <65 years with acute sinusitis.
Among >521,000 eligible patients, there was no difference in treatment failure. AMC was associated with a higher risk of adverse events, mostly driven by yeast and C. difficile infections. These findings support standard-dose AMX as a preferred first-line treatment for adults with acute sinusitis.
John Chandy (United States) presented findings from a study to investigate if severe malaria in early childhood is associated with worse cognition or academic achievement in later life.2
He highlighted that malaria continues to be a major cause of mortality, responsible for 75% of deaths in African children <5 years. Cerebral malaria (CM) and severe malarial anemia (SMA) are associated with cognitive impairment and impaired academic achievement after initial infection.
Children from Uganda who enrolled in previous studies were tested 4–15 years after severe malaria (mean: 8.4 years). CM and SMA were associated with cognitive impairment and lower academic achievement in maths after the initial episode. Attention and reading scores did not differ. Concomitant acute kidney injury, hyperuricaemia and elevated angiopoietin-2 levels were associated with worse long-term cognitive outcomes, reinforcing the need for early intervention in regions with a high burden of malarial disease.
Thomas Holland (United States) presented results from an exploratory study that assessed the pharmacokinetics (PK) of dalbavancin treatment for complicated Staphylococcus aureus bacteraemia, in a secondary analysis of the DOTS clinical trial.3
Holland recapped that in the original trial, patients with a clear bloodstream and who completed therapy with two doses of dalbavancin 1500 mg IV vs MRSA/MSSA standard of care, had a similar risk/benefit, meeting non-inferiority criteria.
This prespecified study evaluated the exposure-response using population PK modelling in 97 evaluable patients. Dalbavancin was found to be highly protein-bound (>99%). Higher total exposures were associated with greater clinical success without increased toxicity. Efficacy increased with exposure and was consistent across subgroups, supporting the use of the third dose in some patients. Exposure-guided dosing should be evaluated in controlled trials.
In the last presentation, Ritu Banerjee (United States) shared findings from the Fast Antimicrobial Susceptibility Testing (FAST) for Gram-negative bacteraemia (GNB) randomised clinical trial.
Banerjee noted that ineffective treatment for GNB bloodstream infections (BSIs) is associated with a >30% increased mortality rate. New and rapid diagnostic testing techniques reduce the time to optimal treatment but are more expensive and not yet validated in clinical settings.
This open-label trial compared the use of rapid antimicrobial susceptibility testing (AST) alongside standard AST in regions with high resistance in 899 hospitalised patients with GNB BSIs over 1.5 years. AST was not superior to standard testing by desirability of outcome ranking (primary outcome) but led to improvements in antibiotic modifications and targeted antibiotic therapy in some patient groups. These findings may help inform the use of FAST in clinical practice.
Key Takeaways:
References
Abdoulaye Djimdé (Mali) delivered a fascinating account of how molecular science has shaped malaria control across sub-Saharan Africa, while outlining the work still needed to translate evidence into impact.
Djimdé began by outlining the scale of the burden. Malaria remains high globally, with approximately 282 million cases and 610,000 deaths annually, the majority occurring in sub-Saharan Africa. In Mali, malaria accounts for 38% of hospital visits and 19% of deaths. Control relies on a range of interventions, including vector measures, rapid diagnostic testing, combination therapies, chemoprevention, and emerging vaccines and monoclonal antibodies. However, drug resistance remains a key challenge.
Djimdé explained that molecular surveillance has been central to tracking the emergence and spread of resistance. Resistance to artemisinin-based combination therapies is linked to mutations in the PfK13 gene, first identified in Africa as locally emerging rather than imported. Multiple mutations are now reported across regions, indicating ongoing evolution within the continent.
He also presented data from early genomic studies across 15 African countries, showing that parasite populations are highly diverse and structured by geography. This has supported a shift away from uniform strategies towards more locally tailored approaches.
Djimdé introduced the Pathogens Genomic Diversity Network Africa (PDNA), established to support coordinated surveillance, data sharing and capacity building across countries. The network has informed national policy and regional initiatives, including a West African surveillance network.
New data from Mali provide a largely reassuring picture. While Pfcrt mutations remain heterogeneous and drug pressure continues to shape Pfmdr1 selection, high-level resistance genotypes are rare. No PfK13 mutations associated with artemisinin partial resistance were identified, suggesting current frontline therapies remain effective.
Diagnostic resistance was also assessed. Among more than 8,000 positive samples, only four pfhrp2/3 deletions were identified (0.05%), well below the 5% threshold at which diagnostic strategies may need to change. This indicates that HRP2-based rapid diagnostic tests remain reliable in Mali.
Djimdé concluded that generating evidence is only part of the solution. Impact depends on engagement beyond the laboratory – collaboration with national malaria control programmes, communities and policymakers is essential to support uptake.
Key Takeaways:
Kimberly Fornace (Singapore) opened the session by discussing the relationship between climate change and infectious disease risks, with a particular focus on malaria and other vector-borne diseases. She highlighted that anthropogenic activities are driving ecological changes that influence both biodiversity and human health and wellbeing. Although climate change is global, its impact is highly regional, meaning the key drivers and health consequences vary by location. Fornace emphasised that there is robust evidence indicating that infectious diseases are aggravated by climate change.
The talk focused on vector ecology, highlighting how temperature and precipitation shape vector survival, behaviour and transmission potential. Climate change can alter vector distributions and expand climatic suitability for diseases. Fornace illustrated how optimal conditions for pathogens such as dengue differ across regions and how indirect effects such as climate-driven migration and displacement further increase disease risk by changing exposure patterns and straining health systems. She used the 2022 floods in Pakistan to demonstrate how severe climate disruption led to a malaria outbreak.
Fornace concluded by linking broader environmental change to infectious disease dynamics, noting that biodiversity loss and deforestation can reshape exposure pathways. She gave the example of monkey malaria (Plasmodium knowlesi) in Southeast Asia, where land-use change and forest-related exposure have increased human infection risks. She closed on an optimistic note, pointing to the wide range of monitoring tools now available and the development of novel climate‑ and health-adaptation frameworks across the Asia-Pacific region to strengthen infectious disease surveillance.
Carsten Rahbek (Denmark) expanded the discussion to global environmental change, stressing that humans are not isolated from the effects of their actions. Rahbek underscored that biodiversity loss and ecosystem disruption are linked to pandemic risk, as most emerging infectious diseases are zoonotic, and spillover becomes more likely when natural systems are stressed. With around 20% of the global species threatened with extinction and 70% of ecosystems degraded, human activity is increasing contact between wildlife, livestock and humans, creating new opportunities for pathogen transmission.
Rahbek explained that once spillover becomes possible, pathogen behaviour is shaped by host availability, changes in species communities, land-driven mixing of animals and host immune response. Disturbed ecosystems often favour generalist, disturbance-tolerant species that act as efficient reservoirs, while climate warming can shift ranges and potentially accelerate pathogen adaptation. He noted that intensive livestock can amplify transmission due to dense, genetically similar animal populations.
Rahbek concluded that pandemic risk is rising, but proactive action can reduce future spillover. This includes setting aside space for nature and maintaining self‑sustaining ecosystems.
Key Takeaways:
Kimberly Kline (Switzerland) opened the session by reminding the participants that most chronic infections are biofilm-related, and of these, many are polymicrobial. In the clinical setting, biofilms can disrupt antibiotic tolerance, further promoting polymicrobial infection, and affecting antibiotic susceptibility.
After infecting a host, the bacterial species (e.g. Enterococci spp) are faced with bacterial competition, fluid flow, nutrient availability ad pH and importantly, the host’s individual immune response, all of which are key to proliferation and suppressing immune activation. In vitro assays and animal models have shown that lactic acid drives immunosuppression via pH reduction, increased virulence and polymicrobial infection. Protease is a key factor for intracellular replication and primes the infecting bacterial species for reinfection.
She highlighted the use of new treatment strategies that include compounds already approved by the US Food and Drug Administration (FDA), which have enabling faster access to licensed treatments and shown efficacy in clearance of intracellular bacteria. Bosutinib (used for chronic myelogenous leukaemia) promotes actin remodelling leading to enhanced macrophage clearance, while methotrexate (used for cancer and multiple sclerosis) has a synergistic effect when combined with vancomycin, leading to increased clearance of antibiotic-resistant bacteria.
Joana Azeredo (Portugal) described how bacteriophages can be used to control and treat viral infections, with phage-associated enzymes, such as the endolysins and polymerases, also possessing antiviral properties.
Showing a development timeline, she emphasised that the concept of using bacteriophages for phage-therapy is not new, although it’s use was sidelined for some time after the development of antibiotics. There is renewed interest in phage therapies, particularly in the advent of antibiotic resistance, with recent clinical cases and studies underlining their efficacy and safety. Biofilms represent a challenge for phage therapy. In conditions such as cystic fibrosis (CF), where a heterogeneous bacterial population is present in the sputum, some strains can become resistant after treatment and proliferate further.
Key strategies to overcome antibacterial resistance include combination therapy with antibiotics and phage evolution where bacteriophages replicate repeatedly in the presence of new biofilms and show improved biofilm disruption. She concluded by noting that bacteriophage therapy is now approved in Belgium and Portugal, where magistral phage preparation enables a personalised approach to enable effective treatment under quality-controlled and validated conditions.
Tom Coenye (Belgium) outlined the differences between susceptibility, tolerance and persistence, which are contributing factors in the failure of antimicrobial therapy.
He described that bacterial strains above the minimum inhibitory concentration (MIC), are resistant, with persistence representing the most difficult to inhibit or kill. Susceptibility may be caused by genomic changes leading to increased efflux and target-site modifications, while tolerance and persistence are phenotypical and reversible, leading to impaired penetration and an adapted response. Bacteria are able to adapt their metabolic pathways in biofilm infections resulting in different outcomes for treatment. Reduced metabolic activity is a key driver of tolerance and persistence, while reduced susceptibility is often driven more by metabolic activity than by drug exclusion. A study of patients with prosthetic joint infections reported that the minimum biofilm inhibitory concentration (MBIC) was much higher than the MIC, highlighting the difficulty in preventing further biofilm development.
A number of approaches have been used to augment the use of standard antibiotic treatment. These include hyper-bariatric oxygen therapy where increased oxygen enhanced the bactericidal effect, adding carbon to activate metabolism, potentiate biofilm prevention and increase killing in P. aeruginosa. In most cases, antibiotics are able to penetrate the biofilm. Improved understanding of recused susceptibility in biofilms may facilitate new areas for treatment.
In the last presentation of this session, Valerie Waters (Canada) reviewed how biofilms in patients with antibiotic-resistant P. aeruginosa infections can be assessed in order to treat effectively.
She highlighted that the majority of infections (e.g. artificial joints/implants, burns, catheter infections, CF and other lung infections) are biofilm-based. In vitro models are important to aid understanding but are difficult to replicate, particularly when relating to polymicrobial interactions, growth environment, spatial structure (e.g. biogeography of biofilms) and host factors. Aggregates are also difficult to replicate in vitro.
Advances in imaging techniques (e.g. MiPACT) has enabled visualisation of spatial proximity and aggregate size. Studies in CF patients with P. aeruginosa infections showed that small aggregates were cleared with antibiotic therapy, while larger aggregates persisted allowing further replication and worsened symptoms. Low-dose nitric oxide has been used as targeted adjunctive therapy. In clinical trials, CF transmembrane conductance regulator (CFTR) modulator therapy disrupted biofilm-based infections but did not fully eradicate the infection. Trials of antibiotic compounds may be helpful to understand the effects of therapy and key markers of inflammation.
Key Takeaways:
Co-organised with: ESCMID Study Group for Biofilms (ESGB), ESCMID Study Group for Non-traditional Antibacterial Therapy (ESGNTA)
María Velasco (Spain) opened the session by discussing two key healthcare challenges for immunosuppressed patients: migration and travel.
Global migration has reached 300 million people, doubling since 1990, and represents a highly diverse population. At the same time, international travel has risen exponentially over the past decade. Together, these trends create a bidirectional impact on both patient outcomes and healthcare systems, making targeted screening and increased awareness essential. Even young, previously healthy migrants may develop illness due to adverse living conditions, malnutrition or exposure to infectious diseases.
Velasco highlighted that latent disease can be reactivated or become more acute, with immunosuppression (e.g. HIV, transplantation, immunosuppressant and chemotherapy use) and may increase susceptibility to some diseases. Screening, simple diagnostic testing and cost-effective, easily administered treatments are key to identifying and treating imported infections. Certain infections (e.g. T. cruzi, Strongyloides spp. and Shistosoma spp.) can modify disease, partially in those with HIV, and some guidelines are now recommending additional screening by geographical origin.
In the following presentation, Martin Peter Grobusch (Netherlands) focused on providing an overview of the key infections in travellers. The most common are diarrhoeal-disease, malaria and arbovirus infections.
Grobusch noted that other diseases may be altered in immunocompromised patients, including tuberculosis (primary infection/reactivation), strongyloidiasis (hyperinfection), leishmaniasis (visceral disease), microsporidial infections (HIV positive) and Dengue (more severe outcomes). Grobusch emphasised that pre-travel care is not complicated, but detailed (e.g. checking antigen titres).
Standard vaccines (Hepatitis A and Yellow Fever) show protection in almost all cases, with a small risk of breakthrough infection, although revaccination is advised in immunocompromised patients. Other vaccines, including against Rabies, are important to boost immunity.
Key Takeaways:
Co-organised with: Federation of European Societies for tropical medicine and international health (FESTMIH)
Esther Calbo (Spain) opened the session by reframing how success in antimicrobial therapy should be defined, distinguishing between the individual patient perspective and the population-level perspective. She emphasised that success is not binary and that beyond survival, patients experience a spectrum of outcomes. This highlights the need to move away from simply asking “what works?” (clinical outcomes) to also understanding “how care is delivered.”
To better capture this complexity, Calbo introduced two composite frameworks. The Desirability of Outcome Ranking (DOOR) integrates efficacy and safety into a single outcome, revealing trade-offs not captured by standard mortality endpoints. Response Adjusted for Duration of Antibiotic Risk (RADAR) incorporates antibiotic exposure, favouring shorter treatment when outcomes are comparable with longer strategies.
Overall, Calbo emphasised that endpoints must capture both what outcomes are achieved and how they are achieved, reflecting real-world patient experiences.
Antonios Katsounas (Germany) illustrated many of Calbo’s points via examples from ICU settings. He emphasised that mortality is a “blunt” endpoint that is infrequent, multifactorial and often underpowered. He advocated for multidimensional measures that better reflect patient trajectories. He also highlighted the importance of standardised stewardship metrics, spanning both process (e.g. days of therapy) and outcomes (e.g. adverse events).
Katsounas showed that multi-component stewardship interventions can reduce antibiotic use without compromising ICU outcomes. He also showed that cycling and mixing strategies have no significant impact on resistance acquisition, reinforcing the need for robust evaluation at both patient and population levels.
Katsounas closed by highlighting the growing importance of patient-centred outcomes, noting from COVID-19 that ICU survivors often experience long-term physical and cognitive impairments. These remain under-measured but are critical to understanding the true impact of antimicrobial strategies.
Key Takeaways:
Naomi Rupasinghe (United States) framed antimicrobial resistance (AMR) as both a health and economic crisis, citing World Bank estimates that high-resistance scenarios could reduce global GDP by up to 3.8% annually by 2050. Rupasinghe emphasised that antimicrobials are uniquely efficient, yet current payment systems fail to reward their value. A shift in policy is needed to align incentives, with the development of finance models and global partnerships highlighted as key levers.
Benedetta Allegranzi (Switzerland) took a data-driven approach, highlighting that approximately 136 million healthcare-associated infections (HAIs) occur annually, with prevalence ranging from around 7% in Europe to over 27% in Africa. Major gaps persist, with 1 in 4 facilities lacking water and 2 in 5 lacking adequate hand hygiene. With AMR associated with an estimated 4.95 million deaths globally, Allegranzi emphasised that infection prevention and control (IPC) interventions can significantly reduce infections and associated costs.
Padmini Srikantiah (United States) positioned vaccination as a central yet underutilised strategy to address AMR by preventing the bacterial infections that drive antibiotic use. Srikantiah highlighted that vaccines protect individuals, reduce complications and limit transmission, thereby decreasing both antibiotic demand and resistance spread. Using typhoid as a case study, she showed that typhoid conjugate vaccines are highly effective against both susceptible and resistant strains. Srikantiah believes that reframing AMR as a preventable burden of infection is critical to unlocking policy action.
Koen Pouwels (United Kingdom) used cross-country modelling to show that countries with similar overall antibiotic use may require different AWaRe* distributions when adjusted for disease burden. Analysis suggests that around 76% of antibiotic use should fall within Access antibiotics, broadly aligning with UN targets. Pouwels noted that antibiotics are among the top three drug expenditures in many LMICs, and that shifting from Watch to Access could generate substantial savings. However, Access is not always interchangeable with Watch, and patient-level factors, including out-of-pocket costs, continue to shape treatment decisions.
*A = Access; Wa = Watch; Re = Reserve
Javier Yugueros Marcos (France) discussed how antimicrobial use in animals has plateaued, with continued reliance on critically important antimicrobials. Resistance levels remain elevated, with around 1 in 6 isolates classified as multidrug-resistant. Marcos pointed to continued gaps in practices such as hand hygiene and emphasised that AMR economics is skewed toward high-income healthcare settings, with limited focus on animal health. He concluded that multisectoral analyses show strong returns on prevention, reinforcing WOAH’s priorities of prevention, surveillance, and sustainable funding.
Key Takeaways:
|
2026 |
|
|---|---|
|
ESCMID Global 2026: WHO-recommended antibiotics cover only 1 in 4 neonatal sepsis infections in low- and middle-income countries 21 April 2026 |
|
|
ESCMID Global 2026: Antibiotic resistance genes found in newborns within hours of birth, study shows 20 April 2026 |
|
|
ESCMID Global 2026: Improving oral care more than halves hospital-acquired pneumonia risk, major trial finds 20 April 2026 |
|
|
ESCMID Global 2026: HIV treatment reduces accelerated biological ageing by nearly four years, landmark study shows 20 April 2026 |
|
|
ESCMID Global 2026: Maternal RSV vaccination cuts infant hospitalisation risk by over 80%, major UKHSA study finds 18 April 2026 |
|
|
ESCMID Global 2026: Influenza frequently missed in winter deaths, new study finds 09 April 2026 |