Perspectives on the Current Utility and Future Potential of Biomarkers in Idiopathic Pulmonary Fibrosis
An Interview with Naftali Kaminski, MD
An interview was conducted with Naftali Kaminski, MD, on the topic of biomarkers for idiopathic pulmonary fibrosis (IPF). Dr. Kaminski is currently Professor of Medicine and Pharmacology, and Chief of Pulmonary, Critical Care, and Sleep Medicine at the Yale School of Medicine. Dr. Kaminski’s main research interests involve the application of genomic approaches to diagnose and monitor pulmonary conditions and to develop novel therapeutic interventions. His research focuses on IPF but also involves other advanced lung diseases such as chronic obstructive pulmonary disease, sarcoidosis, and severe asthma. Below are the highlights from the interview, and reflect Dr. Kaminski’s experience and opinions.
The Need for Biomarkers
What types of biomarkers are currently being investigated for IPF?
Dr Kaminski: Before talking about specific biomarkers in IPF, I think it would be useful to give a brief synopsis of biomarkers in general. In the simplest sense, a biomarker is anything that describes the biological process of the disease. As a result, there are a plethora of types which can be broadly classified on the basis of their intended purpose. These include, but are not limited to: diagnostic, disease susceptibility, prognostic, disease activity, and drug efficacy. All of the above subtypes are currently being investigated for IPF. If you would like further information on these subtypes of biomarkers in IPF, which goes beyond the discussion today, there is a paper that reviews all of the above and which you may find useful.1 In short, diagnostic biomarkers inform on the presence or absence of disease while susceptibility markers identify at risk populations such as those harboring genetic mutations. Prognostic biomarkers can be differentiated from activity markers because the former are predictive of disease course while the latter give a snapshot of the disease at any given time. Although the two may overlap they are not equivocal. Drug efficacy biomarkers, as the name suggests, inform on treatment response based on a compound’s mechanism of action. Again, this is not an exhaustive listing, however, it does highlight some of the more common and very useful biomarkers utilized in IPF. Beyond this there are a variety of more complex applications. For example, in drug development there are biomarkers that actually tell you whether the drug is engaged as well as biomarkers for drug toxicity.
What characteristics make for a good biomarker with regards to IPF?
Dr Kaminski: With respect to what is a good biomarker, this qualification is not dependent on a particular disease state, such as IPF, but is more dependent on what type of information the clinician wants to obtain in a particular setting, how reliable that information is, and what value the information will provide. One common feature, however, of what I consider to be good biomarkers is that they provide actionable information. Actionable information leads to more optimal management, treatment, and counseling of individual patients in the clinic and provides valuable insight into the pathophysiology of the disease to advance the science at the clinical trial level. With respect to IPF in particular, even the results of the spirometry that we all do in our offices or in the pulmonary function lab can be considered biomarkers. What has emerged in the last few years is the notion that pulmonary function tests, clinical assessments and imaging are very good for some things, but they're not good for other things, and that’s why there has been an emergence of molecular biomarkers. Using some kind of bodily fluid or sample to identify and detect molecules that are indicative of disease presence, activity, early detection, prognosis and therapeutic response fills in the gaps to provide optimal care.
Which classes of biomarkers are the most developed or reliable with regard to IPF?
Dr Kaminski: IPF is interesting because there's actually a wealth of information on one or two types of biomarkers but a paucity in others. I would have to say prognostic and susceptibility biomarkers are the most developed for IPF. So much so that an individual patient’s biomarker profile contributes much more valuable information than can be obtained from the clinical presentation. Matrix metalloprotease 7 [MMP-7] levels for example, although stable with disease progression, are predictive of mortality and have been shown to be elevated in those with positive CT findings who are free of symptoms. In addition to MMP-7, some more reliable biomarkers are surfactant proteins A and D [SP-A and SP-D]. These markers have been validated in multiple cohorts.2
The Utility of Biomarkers for Early Detection of Disease
Please discuss the role that biomarkers may play in the predisposition, early detection, and diagnosis of IPF.
Dr Kaminski: This is an important topic and more studies are needed to determine whether early treatment can modify the course of disease and have a significant effect on outcomes. Drugs have been shown to slow the decline in pulmonary function as measured by forced vital capacity, or FVC3, so it makes sense that biomarkers can help identify patients who would benefit from early intervention. With respect to predisposition, there are a number of genetic abnormalities which can identify patients at risk such as mutations in surfactant proteins and telomere shortening. These would fall under the category of susceptibility markers discussed previously. Let's say, for instance, a patient is found to have a disease causing a genetic variant. The family members of the genetically susceptible patient could be first genetically screened, and if they carry the same genetic variant, undergo follow-up testing for the proteins mentioned as biomarkers. Some data suggest that markers such as MMP-7 are increased when disease is subclinical. What is particularly promising is that these are all simple ELISA tests [enzyme-linked immunosorbent assay] that are extremely fast, simple and cost-effective. The information to support this approach is out there. The biggest challenge is to standardize and commercialize the tests so they could be used routinely.
Are there any confounding factors that may affect the sensitivity and specificity of IPF biomarkers?
Dr Kaminski: This is currently under investigation but there is a definite clear signal for disease presence in IPF. If you look at proteins in the bloodstream, a patient with IPF is very different from the healthy controls but they are also very different from individuals who are smokers and those with COPD. In addition, IPF cases seem to be discernable from other interstitial lung diseases. This distinction, however, is a bit more complicated.
The Use of Biomarkers to Predict Mortality in IPF
How can biomarkers measure IPF disease activity?
Dr Kaminski: Disease activity markers are not as developed as susceptibility and prognostic biomarkers in IPF. In fact, there is currently a lack of consensus on what defines disease activity in IPF. However, one could posit that SP-A and MMP-7 are markers of alveolar epithelial injury while CCL-18 [chemokine ligand 18] is a marker of macrophage activation. At this time, we do not have markers for some of the key events that underlie the pathophysiology of IPF such as the formation of new collagen and more studies need to be done to elucidate these mechanisms. On a more promising note, several groups, including my group in collaboration with some in Germany, have been able to obtain over 100 lavage samples from patients with IPF done at diagnosis in Europe at three centers in Italy, Belgium, and Germany. Although we are just now working on the results, there appears to be a strong signal for both outcome and disease activity.
Do you believe that biomarker examination could eventually replace forced vital capacity lung function testing as the primary measure of IPF disease progression?
Dr Kaminski: Not in the immediate future. Again, the most reproducible and reliable biomarkers to date are MMP-7, SP-A and SP-D, and CCL-18. There are others but the interesting thing is most of these markers don’t really predict disease progression the way we define it by decline in FVC. They do, however, perform very well in predicting mortality. Additionally, they do so relatively early in the course of the disease and this has been reproduced many times. Most recently we took a threshold for MMP-7 derived in Korea and applied it to our Yale population – levels above this threshold were significantly predictive of mortality even after adjustment to clinical variables.4 My group is also delving into whether any of the drugs that we give now affect the expression of particular biomarkers, as we think that this could be very informative. I'll give you an example from some work that is currently ongoing. MMP-7 levels do not fluctuate during the course of the disease. If you're high, you're high, if you're low, you're low, but it's predictive of outcome. So let's say now that a drug actually changes the levels, reduces them. Although we don’t know for sure, we can now follow up and that’s going to be our clinical study design to see whether this patient is actually a responder to the drug. And if we had that information, then we would have another initiative or motivation to actually use the drugs and the target as well.
Are there any clinical events (such as acute exacerbations, rapid respiratory declines, etc) that have their own set of unique biomarkers?
Dr Kaminski: This is another area of disease activity my group has looked at. In the setting of acute exacerbation in IPF [AE-IPF], KL-6 is the biomarker which has been most widely studied. It seems that AE-IPF is associated with increases in blood KL-6 but the mechanisms are not clear. We compared the gene expression in the lungs of patients with AE-IPF lungs to stable IPF and identified 579 differentially expressed genes, independent of infectious or inflammatory cause. We also noticed an increase in α-defensins, a group of innate antimicrobial peptides, in the plasma protein level of AE-IPF patients. Because of this we feel that genes should be evaluated as potential biomarkers5.
Methods for Obtaining Samples to Examine Biomarkers
Please discuss the manner in which samples are obtained to evaluate biomarkers.
Dr Kaminski: Ideally, we would like to sample the lungs. That’s where the disease is. The problem is that it is dangerous and not indicated in many of the patients, so for these reasons we use surrogate tissues such as blood. In the clinical setting, surgical biopsy has largely been replaced by high resolution CT [HRCT] scanning because you can make a definitive diagnosis of IPF through a UIP pattern in conjunction with a multidisciplinary assessment. Therefore, and I may be a little bit pessimistic in this, but I don’t see us getting routine lung tissues for patients with IPF anytime in the near future.
Do bronchoalveolar lavage samples provide a better sample than lung biopsy or non-invasive peripheral blood samples?
Dr Kaminski: Bronchoalveolar lavage [BAL] is primarily used in patients with pulmonary fibrosis to rule out things, so it's not a part of the routine evaluation of patients with IPF in the US. In Europe, and actually in the majority of the world, BAL is done quite routinely in patients with IPF but it’s clinical value is debatable. However, in the near future it may provide key information with respect to a patient’s microbiome. For years, people thought that the lung was sterile. In the past decade we have learned that even the healthy lung, while not thriving with bacteria, does have some resident microbial flora. Two groups, one from Imperial College in London and one from Michigan, have shown that the microbiome changes significantly in IPF and these changes may be predictive of different outcomes. So unlike changes in proteins, changes in the microbiome are actionable since you may decide to give the patient antibiotics. Thankfully, there are studies underway which will likely provide evidence into which patients may benefit. Dr. Fernando Martinez from Cornell, for example, is leading a multicenter study looking at whether antibiotic therapy would have any effect on the outcome of patients with pulmonary fibrosis.
The Role of Genetics in IPF
Can the exploration of genetic biomarkers help screen at-risk patients with a familial history who do not otherwise exhibit signs or symptoms of IPF?
Dr Kaminski: In biology there is a hierarchy. You could look at DNA, which tells you about the genetic predisposition of a patient, or the epigenome to analyze the expression of these genetics. Methylation is a common point of focus, as is RNA, proteins and even metabolites. The point is that there have been an amazing number of breakthroughs in discovering genetic mechanisms associated with IPF over the past few years, which are benefiting patients as we speak. This includes both familial and sporadic forms. In the rarer familial form, mutations in SP-A and telomere shortening have been shown to be predictive where in the sporadic form the most important is clearly the muc-5b mucin gene. One of the most important things to remember, and I don’t think that pulmonologists or the rest of the scientific community appreciates this enough, is the extremely strong genetic association in IPF. The genetic association in IPF is far stronger than in other multi-factorial chronic diseases including diabetes, asthma, COPD and coronary heart disease. This has been replicated in an unprecedented way. In particular, the hazard ratio of the muc-5b genetic polymorphism has been shown to be up to 20 fold in multiple cohorts. This mutation alone can explain almost one third of the cases and is so strong in these instances, I advocate that we drop the “I” in “IPF” and refer to it as “muc-5b associated pulmonary fibrosis”. There is also evidence that these patients behave differently than the non muc-5b group and therefore, identification serves a practical purpose. For instance, although they develop the disease more frequently, their disease is actually milder, so it has prognostic implications.6 On the other hand, let's say you have a patient with familial disease. You definitely want to screen them and other family members because this may have important implications even when they are not exhibiting symptoms. They could be included in studies for early detection of disease and potentially, hopefully, there will be interventions for early disease in the near future.
Taking this a bit further raises the question: should we genotype patients routinely? The truth is we don’t really know. We think, at least I think, we should because gaining this information can have both prognostic [in the case of MUC-5B] and diagnostic [in familial] value.
Does the Yale School of Medicine have any programs in place that screen for genetic markers of IPF?
Dr Kaminski: Yes, we offer comprehensive programs for our patients in collaboration with the leading laboratories in the field. Although usually we won't do the lab testing at Yale, we routinely work very closely with eminent leaders in the genetics of IPF at other institutions. In fact, because of the positive impact of these programs, we are now looking at extending our services to better meet patients’ unmet needs.
The Future of Biomarkers in IPF
What do you envision for the future of IPF with regards to biomarkers?
Dr Kaminski: As I mentioned earlier, we have an intuition that since pharmacotherapy reduces the decline in FVC, earlier intervention will have a significant impact on modifying the course of disease. Ideally, the patient never gets to a stage of disability but we just don’t know yet. Once the trials provide evidence that current treatments affect early disease, then there will be a huge motivation to both screen for disease predisposition as well as evidence of activity. You don’t want to do CT scans on everybody, so effective serum biomarkers, possibly MMP-7, could be very helpful.
Will the implementation of genetic biomarkers contribute to the development of “personalized medicine” in IPF?
Dr Kaminski: Personalized medicine and individualized therapy are crucial to optimal patient management in most chronic diseases and IPF is no exception. Advancements made in the field of oncology and other diseases with contributions from genetic biomarkers will benefit IPF and vice versa. Medical geneticists should share their knowledge and resources so that all can benefit. In fact, I am glad to say that Yale New Haven Hospital and Yale School of Medicine (which has one of the top departments of genetics in the country) are considering an initiative in genomic medicine which will have a great impact. Our section of Clinical Oncology at the Yale Cancer Center is led by one of the world’s leaders in lung cancer checkpoint inhibitors and precision medicine, Dr. Roy Herbts, so this collaboration holds a great deal of promise. In my section we have opened a new center called the Center for Precision Pulmonary Medicine (P2MED). It is a research center, focusing on implementing approaches in precision medicine for advanced lung diseases. We have all the equipment for biomarker discovery, a sample repository, a computer cluster for parallel computing and 3 faculty that are experts in data analysis, bioinformatics and computational biology. The reason we located the center in the section is because we want this center focused on implementation - moving fast from discovery to the bedside.
Is there anything else you would like to add to our discussion today?
Dr Kaminski: So, I think there are a few additional points. One is that this is not a complete discussion of biomarkers. There's much more to the story. The second is that the biggest revolution that happened in IPF occurred because we aggressively recruited patients into research. It's critically important to refer patients to centers that offer clinical trials because that’s what drives the field forward. I also recommend that physicians in the community, patients, and caregivers ensure they have access to the most current and accurate information available. In my opinion, one of the most helpful resources for both patients and caregivers is the Pulmonary Fibrosis Foundation website. It provides a wealth of valuable information and their data is vetted by an outstanding group of people. Finally, I recommend that all IPF patients are seen at least once annually in a center with a multidisciplinary team of IPF experts and, as recommended by the guidelines, I think the diagnosis should be made by a multidisciplinary team. Preferably, this will take place at a top academic center with access to medical geneticists and counseling because different mutations often have different implications. Such a center will be staffed by those who know how to communicate and advise through very complex issues. Such a setting also improves access to transplant, palliative care and may also assist in getting insurance to cover the drugs, so there are multiple layers of complexity which justify it. I’m not saying that the burden of care needs to be exclusively at the center, but if the pulmonologist in the community doesn’t feel comfortable, then they should refer to the most well-equipped institution possible.
- Zhang Y, Kaminski N. Biomarkers in idiopathic pulmonary fibrosis. Curr Opin Pulm Med. 2012;18(5):441-446.
- Ley B, Brown KK, Collard HR. Molecular biomarkers in idiopathic pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol. 2014;307(9):681-691.
- Raghu G, Rochwerg B, Zhang Y, et al. An official ATS/ERS/JRS/ALAT clinical practice guide: treatment of idiopathic pulmonary fibrosis: an update of the 2011 clinical practice guideline. Am J Respir Crit Care Med. 2015;192(2):e3-e19.
- Tzouvelekis A, Herazo-Maya JD, Slade M, et al. Validation of the prognostic value of MMP-7 in idiopathic pulmonary fibrosis. Respirology. 2016; 1-8.
- Konishi K, Gibson KF, Lindell KO, et al. Gene expression profiles of acute exacerbations of idiopathic pulmonary fibrosis. Am J Respir Crit Care Med. 2009;180(2):167-175.
- Tzouvelekis A, Herazo-Maya J, Sakamoto K, et al. Biomarkers in the evaluation and management of idiopathic pulmonary fibrosis. Curr Top Med Chem. 2016;16(14):1587-1598.
Naftali Kaminski MD
Professor of Medicine and Pharmacology
Chief of Pulmonary, Critical Care and Sleep Medicine
Yale School of Medicine
Naftali Kaminski, M.D. is the Boehringer Ingelheim Endowed Professor of Medicine with secondary appointment in pharmacology and Chief of Pulmonary, Critical Care and Sleep Medicine at Yale (http://medicine.yale.edu/intmed/pulmonary/people/naftali_kaminski.profile). Born and trained in Israel, Kaminski completed his research training at the University of California, San Francisco, and after heading the Simmons Center for ILD in Pittsburgh was recruited to Yale School of Medicine in 2013. Kaminski is known for his significant contributions to research and respiratory medicine. He is considered a pioneer in the application of high throughput methods and systems biology to implement precision medicine approaches to pulmonary fibrosis and advanced lung disease research. Among his discoveries are the role of MMP7 in pulmonary fibrosis, the altered repertoire of microRNAs in idiopathic pulmonary fibrosis (IPF) lung, and their potential as therapeutics, and the discovery of highly reproducible peripheral blood biomarkers in IPF. Since completing his clinical training, Dr. Kaminski authored more than 200 research papers, review articles and book chapters and has given numerous invited talks at national and International conferences. Dr. Kaminski has been consistently funded by NIH and is the PI of multiple NIH grants.
Dr. Kaminski’s achievements and contributions have been widely recognized in recent years, Dr. Kaminski was a recipient of the Marvin I. Schwarz Award for contributions to patient care and research in pulmonary fibrosis from the Coalition for Pulmonary Fibrosis in 2010, the University of Pittsburgh Innovator Award in 2012, the American Thoracic Society Recognition of Scientific Achievements award and the Helmholtz Institute International Fellow award in 2013. In 2015 he was elected to the Association of American Physicians. In 2016 Kaminski was elected as Fellow of the European Respiratory Society (ERS) and was awarded the ERS Gold medal for Interstitial Lung Disease. the Kaminski serves currently as deputy editor for Thorax BMJ and has been recently been elected as president-elect of the Association of the Directors of Pulmonary, Critical Care, and Sleep Medicine Divisions.
In the last three years, as the chief of Pulmonary, Critical Care and Sleep Medicine at Yale, he led the section through an unprecedented period of growth including junior faculty retention and promotion, significant growth of all clinical programs, near doubling of research funding, and creation and reorganization of novel centers of excellence (for more detail visit http://pccsm.yale.edu/).