Advancing infectious disease diagnostics from HIV to AI and metagenomics

Infectious disease diagnostics are evolving at a rapid pace. In a post-COVID landscape, clinical labs are under pressure to deliver faster results, handle increasingly complex cases, and integrate tools that go far beyond conventional PCRs or cultures. Technologies like shotgun metagenomics and artificial intelligence (AI) are reshaping how clinicians detect and respond to infections. Meanwhile, long-standing threats like HIV, hepatitis C and antimicrobial resistance continue to demand new strategies, including long-acting prevention and even the possibility of functional cures.

Stacey A. Rizza, M.D. executive medical director for International Practice and Asia Pacific at Mayo Clinic, has worked across infectious disease disciplines, from public health to diagnostics and research. She also leads efforts to build long-term collaborations with health systems globally, advancing knowledge-sharing and enhancing patient care.

In an interview with SelectScience^®, Rizza discusses key developments in clinical microbiology, including how HIV causes damage beyond immune cells, the shift toward patient-collected samples, and how labs are applying machine learning to predict outbreaks and resistance. Rizza’s insights reflect a growing need for labs to adopt new diagnostic models while remaining grounded in public health outcomes. 

Shotgun metagenomics

Shotgun metagenomics is increasingly recognized in diagnostic microbiology as a valuable approach for broad-spectrum pathogen detection. Unlike methods dependent on targeted primers or panels, this technique analyses all non-human genetic material present in a sample, thereby providing an extensive overview of infectious agents.

Shotgun metagenomics involves taking a sample and removing most of the cellular material, which is primarily human. Whatever is left that is not human is therefore your infection, says Rizza.

Rizza explains that this method allows for unbiased detection across viruses, bacteria, fungi and parasites in a single process. As opposed to PCR testing, which targets a narrow range of organisms, metagenomics can reveal unexpected co-infections or pathogens that standard tests miss. Metagenomic testing is also faster and more specific, to the point that it can actually predict resistance patterns.

The resistance prediction component is key. With antibiotic resistance rising globally, labs are under pressure to determine not just what pathogen is present but also whether it can be treated effectively. Shotgun sequencing allows for resistance gene identification without the need for culture, which is especially useful in fast-paced hospital settings.

HIV and hepatitis insights

Rizza has performed extensive research on HIV’s role in end-organ damage, even in patients with controlled viral loads. She explains that HIV proteins can cause harm in tissues like the brain, liver and kidney, even when the virus itself is not actively infecting those cells.

“HIV can cause apoptosis in end organs like the kidney, the brain, the liver without HIV infecting the cell.”

This understanding has clinical implications, especially for patients co-infected with hepatitis C. The overlap between viral mechanisms can worsen inflammation and speed up tissue damage. Early testing and treatment remain essential.

“We’ve known for a long time that having both HIV and hepatitis C accelerates liver disease, kidney disease, and brain damage. But now we’re starting to understand that HIV alone can do this too, even when the viral load is undetectable.”

Long-acting drugs and 'kick and kill'

The landscape of HIV treatment has shifted in recent years with the development of long-acting antiretroviral drugs. These allow for monthly or bi-monthly injections rather than daily pills. Rizza notes that long-acting agents like cabotegravir and the newly approved lenacapavir are making it easier for patients to stay on treatment or prevent infection altogether.

“We put a particular amount of emphasis on something called PrEP or pre-exposure prophylaxis. PrEP has evolved a lot. Nowadays, we know that you can use long-acting drugs, so you get a shot every two months.”

“Another class of drugs called lenacapavir was just approved for PrEP.That’s two shots a year and you won’t get HIV.”

Rizza also discusses ongoing research into HIV cure strategies, particularly those aiming to eliminate latent viral reservoirs.

The ‘kick and kill’ approach, being studied as a potential HIV cure, aims to activate hidden viruses in cells so that treatment can eliminate them. Since HIV can remain dormant and undetectable, this method seeks to expose and destroy infected cells.

This approach is still in the research phase, but labs involved in HIV studies should prepare to support clinical trials with immune monitoring, reservoir quantification, and long-term follow-up testing.

Expanding access to infectious disease diagnostics

Rizza highlights the growing role of patient-collected samples, especially in the context of HIV and STI screening. Self-collection can increase testing uptake, reduce stigma and reach populations who might not otherwise attend clinics.

“People actually can screen for HIV by themselves at home or they can screen for chlamydia or gonorrhea either orally, anally or vaginally. They send it in and get a result so we can treat sexually transmitted infections and HIV earlier.”

For labs, this model requires validated collection devices, stable transport media and streamlined reporting systems that maintain quality while scaling access.

The role of AI in infectious disease diagnostics

Artificial intelligence (AI) is becoming an integral part of diagnostic workflows. Rizza describes its use during the COVID-19 pandemic to predict viral spread and assist public health planning.

“During the COVID-19 pandemic, we applied AI algorithms to forecast which cities were likely to be impacted next. This enabled us to track the virus’ progression and anticipate emerging hotspots with greater precision.”

Rizza also mentionsthe utility of AI in streamlining laboratory analysis and speeding up interpretation.

“I think we will see AI being used more and more in interpretation. Machine learning can already help with analyzing large data sets.”

The role of AI is not limited to diagnosis. It is also accelerating drug discovery, trial design and pattern recognition in disease outbreaks.

“Without a doubt AI is going to help in drug development. Now we can get drugs from conception to in patients within months to a year.”

Clinical laboratories can expect to see more integration of AI tools in result triage, anomaly detection and decision support systems, especially where data volumes exceed human capacity.

Addressing antimicrobial resistance through early detection

Resistance remains a major challenge for infectious disease labs. Rizza emphasizes the importance of identifying resistance patterns at the genomic level before initiating treatment.

“Genomic testing enables the prediction of potential resistance to specific antibiotics, allowing for more targeted and effective treatment strategies.”

However, Rizza cautions against over-reliance on positive findings without clinical context.

“An essential consideration is whether every finding warrants treatment. In many cases, the answer is likely no.”

This nuance is central to effective stewardship. Labs need to equip clinicians with not just a list of detected organisms, but meaningful insights into which infections require intervention.

Looking ahead in infectious disease testing

As technologies mature, the role of the clinical lab is no longer just about detection. It is about interpretation, stewardship and public health impact. From shotgun metagenomics to long-acting HIV prevention, and from self-collection kits to AI-powered workflows, the diagnostic lab is becoming a hub for real-time, patient-centered action.

“The key to the kingdom is to find an infection early, to know exactly what it is and to know exactly how to treat it, and to do all of this in a fast and accurate way.”

Rizza’s perspective reflects a broader shift in lab medicine; one where precision, prevention and responsiveness matter as much as the test result itself.