10 questions for Peter Sterk, Scientific Adviser

Aug 31, 2020

Peter Sterk has launched exhaled breath analysis by eNose in various clinical settings. Peter is a biomedical scientist (MD with a PhD in Physiology) trying to incorporate novel biological information into the diagnosis and management of patients with respiratory diseases.

What motivates you to do what you do?

The usage of modern molecular technologies has revolutionized our view on role of biological networks in health and disease. This has its roots in physics and chemistry, where the behavior of complex systems (yes, we are…!) has been captured by pattern recognition rather than trying to understand its individual parts. This is a real step-change in 21st century biology and thereby in medicine.

Which strategies did you use to be successful in your career?

Whether successful or not, I enjoyed and cherished interdisciplinary contacts, which led to real cross-fertilization. What seems to be novel to us, is often already common grounds and more elaborate for others in- and outside medicine. My most successful projects are all linked to such exchange of ideas.

What do you consider to be one of your greatest achievements?

Personally, I am pleased to having trained so many young scientists, who went into several directions. From basic science to solid clinical practice. As a team we achieved that quantitative assessments at all levels of disease (from molecules to symptoms) has led to better disease outcomes for patients here and there. In my case this relates to capturing small airways disease, bronchial hyperresponsiveness, airways inflammation and molecular fingerprints in asthma and COPD.

How did your interest in breath analysis emerged?

In fact, already back in 1976 I used a mass-spectrometer for exhaled breath analysis. At that time for measuring uneven ventilation as a marker of lung disease. For us, assessing exhaled volatile organic compounds (VOCs) started around 2003, again primed by others (colleagues in the US) who tried to do this for identifying lung cancer. To me the turning point was to move from mass-spectrometry as a laboratory technique to electronic noses (eNoses) as instrumentation specifically developed for recognition of composite VOC mixtures. I believe that the future of medicine will not be represented by ever more expensive technology. In contrast, most progress in healthcare can be expected to come from cheap, simple and preferably portable systems. eNoses are suitable for clinical application at point of care, whereas gas-chromatography and mass-spectrometry (GC-MS) and related technologies are not. It appeared that the eNoses were already applied for safety and security and for agricultural purposes. This led to our studies with eNoses in asthma, COPD and lung cancer, first in Leiden and later in Amsterdam.

Why is early diagnosis so important for these lung diseases?

Good question: early diagnosis is only relevant when it has consequences for treatment and prognosis. In all these diseases it is not only the mere diagnosis that is important for clinical management, but the more so the individual phenotype (profile) of the patient. Detailed phenotyping of patients is required for the selection of today’s therapeutic options. It appears that eNoses are pretty good in doing just that.

”For the patients (and care takers) the eNose technology appears to be highly appealing, because it is non-invasive, easy and quick.”

What are your thoughts about breath analysis being a part of the new trend of personalized medicine?

I don’t think personalized medicine is a trend, it is a necessity these days. Currently, still, most diagnoses are based on 19th-20th century clinical pattern recognition, whereas incorporating composite cellular and molecular parameters leads to stratification of patients along operational biological networks. Modern drugs are targeting specific networks, thereby requiring the assessment of those networks. The latter can best be done by comprehensive molecular measurements (omics, such as transcriptomics, proteomics, metabolomics), allowing a broad picture of the patient’s biological condition. During the past decade I coordinated a big EU project granted by IMI (U-BIOPRED) to explore the potential of omics technologies in severe asthma. Exhaled breath analysis by eNose also represents such technology (breathomics), being very rapid and non-invasive. It turns out from validated studies that it can recognize inflammatory profiles in asthma and COPD that are relevant for therapeutic choices, whilst it can also identify responders to new biological treatment of lung cancer.

What are the biggest challenges for bringing breath analysis to clinical care?

I believe there are three modest hurdles to be taken. First, it requires recognition amongst professionals that fingerprints as provided by pattern recognition are better markers of individual disease than traditional biomarkers in medicine. This is a real step-change, coming from modern molecular biology that is learning to value the concept of ‘competence without full comprehension’. The latter is very topical and timely in many fundamentally complex areas in science and society, where artificial intelligence and machine learning are driving previously unanticipated progress. Second, the results of breath analysis are being presented as probabilities. These represent the numerical chance of a particular diagnosis or phenotype in an individual patient. Probabilities are not black or white (positive or negative), but represent on outcome along a continuum. This is entirely realistic in medicine, but may still need some endorsement in day to day practice. And third, it is a matter of practicalities: how can such breath analysis be performed within minutes at point of care, when the patient is still in the doctor’s office? This requires a rapid and simple measurement and powerful (cloud based) computing and feedback. I am very pleased it is operational now, albeit it can always be further improved.

What do you think, is the first application of breath analysis?

The scope and potential are broad, but based on the current scientific publications with experiments that have been validated in multiple patient populations, I expect that diagnosing asthma and COPD is a first hit. However, as said, the technology is equally important for selecting treatment options by sub-classification of patients (phenotyping). Notably, selection of responders to immunotherapy amongst patients with lung cancer has recently been validated by breath analysis. Screening for lung cancer would be nice as there are good preliminary data, but this still requires dissemination studies, particularly in series with CT-scanning. And finally, based on new data and the growing scientific literature there will be other areas in lung diseases, such as cystic fibrosis (CF), interstitial lung diseases (ILD), pneumonia, COVID-19, and beyond, which are currently being tested.

What are the biggest advantages for patients to bring breath analysis to clinical practice?

The nice aspect of this question is that the answer is quite simple. For the patients (and care takers) this technology appears to be highly appealing, because it is non-invasive (does not hurt), easy (not much effort) and quick (direct answer). Still, it is highly informative as it integrates information from hundreds to thousands of molecules and their fragments. The technology takes advantage of big data whilst the computations can be done anonymously, thereby meeting today’s cautiousness regarding the security and privacy of medical data.

And last but not least: what is your life motto?

“It is good to be here, it is good to be anywhere” (Keith Richards)