The Microbiome’s Diagnostic Potential in Infectious Disease Point-of-Care Assays

Kaitlin Searfoss:
Hi everyone. Welcome to this podcast from Cambridge Healthtech Institute for the Advances in Microbiome Diagnostic Symposium taking place August 18th as part of the Next Generation Dx Summit. I'm Kaitlin Searfoss, conference producer.

We have with us today one of our speakers, Dr. Georgios Kitsios, an Instructor of Medicine and a Post-doctoral scholar in the Pulmonary, Allergy, and Critical Care Medicine Division at the University of Pittsburgh Medical Center. Thank you for joining us, Dr. Kitsios.

Georgios Kitsios:
Thank you for having me, Kaitlin.

Kaitlin Searfoss:
Your current research involves rapid microbiome-based diagnostics of infectious disease, particularly pneumonia. Can you comment on how you see the microbiome filling a need in the diagnostic community to develop faster and more accurate point of care assays for infectious disease?

Georgios Kitsios:
Oh, absolutely. Our research focus is on severe pneumonia in the intensive care unit. That is often complicated by the serious syndromes of sepsis and the acute respiratory distress syndrome. In this population it is a true emergency that we provide the right antibiotic prescription as early as possible.

Diagnosing pneumonia is challenging though because in these patients the conventional diagnostic criteria of fever, leukocytosis, purulent sputum, and abnormalities of chest x-ray are not particularly useful as they are typically abnormal. And we have to rely heavily on microbiologic cultures of respiratory specimens, such as sputum or bronchoalveolar lavage fluid in order to identify a causative pathogen. However, it is important to emphasize that microbial growth on a Petri test for pathogen identification is a 19th century technology and despite being our current gold standard it is by no means golden.

Cultures can be terribly insensitive. In some recent studies up to 60% of the time cultures are negative and even when positive it typically takes three days for pathogen growth, identification and susceptibility testing to obtain actionable results. So even in the minority of patients in which microbiology cultures can tell us what the pathogen is, because of such delays we miss the critical initial period for decision making on antibiotic prescriptions. This is why we think that next generation sequencing offers an unprecedented opportunity for break-throughs in modern diagnostics.

Without need for ex-vivo growth but directly from patient samples sequencing of microbial genomes, we can reliably obtain a comprehensive picture of each bacteria, viruses or Fungi are present in our patients sample and at what relative proportions of what the actual pathogen is, if any. This approach would allow us deliver early, appropriate, targeted antibiotic prescriptions in severe pneumonia and represents a novel path for accomplishing personalized medicine in the intensive care unit.

Kaitlin Searfoss:
One of the drivers behind your research seems to be the epidemic of the antibiotic resistance. How has this phenomenon shaped your research in how you approach critical care medicine? How does the microbiome play into this and where do you see it moving in the future?

Georgios Kitsios:
So I completely agree with you, Kaitlin, that antibiotic resistance is a true epidemic and in fact one of the top priorities in the World Health Organization agenda.

So how is it shaping our research? We are looking at it from both patient and ICU community perspective. For individual patients infected with a multi-drug resistant organism, we are able to pull one of the lab resort antibiotics after the patient has failed typical regimens. If culture confirm resistance, with the delays we discussed, then the critical initial period for intervention is lost. There are now a couple studies in the laboratory that show proof-of-concept ability to predict susceptibility based on sequencing of antibiotic resistant genes. So one of our immediate recent priorities is to examine the predictive ability of point of care sequencing for antibiotic resistance in real world samples in the ICU.

Now from the global perspective of antibiotic resistance spreading in intensive care units, we ought to recognize that the major driver of resistance is iatrogenesis due to selective pressure of broad spectrum antibiotic regimens. However, our preliminary results indicate that especially in culture-negative patients start antibiotics are most of the time disproportionately intense or entirely unnecessary, especially when the pathogen is viral and there are no bacteria causing the infection. Consequently, real time next generation sequencing offers the great potential for accomplishing antibiotic stewardship in the intensive care unit and thus eliminate as much as possible, the empiric administration of big gun antibiotics that drive resistance.

Kaitlin Searfoss:
What are the biggest challenges you and other scientists face in microbiome research in general, and for diagnostic potential in particular?

Georgios Kitsios:
Well first of all, lung microbiome research is a nascent field and in fact a few years ago there was even debate about whether or not a lung microbiome exists or is it upper airway contamination that we detect. Now that more than 35 studies have unequivocally demonstrated a low biomass microbiome in the lungs of normal subjects, we can better understand the microbiome of those with infections when we have a normal reference standard for comparisons. Nonetheless, methodology is still evolving in sample acquisition, processing, sequencing protocols and analytic methods, and the field is still maturing.

Now from a diagnostic perspective in particular, I will point out that we face a common challenge with technological innovations. And the question is how can we assess diagnostic performance of a new test when there is no gold standard? Or rather, our current standard is not golden. So conventional approaches of sensitivity and specificity comparisons simply will not work because cultures are already not sensitive or specific. That is when we need statistical and computational biology methods to create a construct gold standard of severe pneumonia with incorporation of culture data, clinical data and host biomarkers suggestive of infection, such as procalcitonin levels to compare with our sequencing results. In this way, we'll be able to derive a sequencing based definition of pneumonia, that will be far more sensitive and specific of cultures. Nonetheless, the final test of truth will be demonstration of clinical utility of sequencing definition on antibiotic prescriptions in a clinical trial. And as one can imagine, this is time consuming, labor- and research-intensive. So we are few steps away before we are able to implement sequencing in our clinical practice.

Kaitlin Searfoss:
What are you most excited about seeing at the Next Generation DX Summit?

Georgios Kitsios:
So I'm especially interested in technological innovation in rapid sample processing and sequencing methods, and in the evolving analytical approaches to rapidly and reliably process the vast amounts of genomic big data in decision algorithms. I do believe there is truly an ongoing revolution in microbiome-based diagnostics of infection and I really excited to hear about the major breakthroughs in the field at the Next Generation Diagnostic Summit.

Kaitlin Searfoss:
Thank you for your time today, Dr. Kitsios.

Georgios Kitsios:
Thank you for having me, Kaitlin, and I look forward to the conference.

Kaitlin Searfoss:
That was Dr. Georgios Kitsios from the University of Pittsburgh Medical Center. He'll be speaking at the Advances in Microbiome Diagnostic Symposium on August 18th at the Next Generation DX Summit. I am Kaitlin Searfoss, thank you for listening.