Searching for New Biomarkers to Prevent Organ Transplant Rejection. Q&A with HOT Panel Grant Winner Dr. Evan Farkash

Categorized As:
Grants Transplants

Bob’s first question to me was: “How much do you want to learn about kidney pathology?” My response was that I wanted to learn as much as I could, and I don’t think I have stopped learning since then.” Dr. Evan Farkash, MD., PhD.

This Q&A zoom session with Dr. Evan Farkash, MD, PhD., was particularly pleasant: this brilliant and smiling scientist was in his office, where glass walls let you glimpse inside his lab and at the busy young researchers walking around their benches. He sounded very proud of this setting.

Dr. Farkash has gotten his degrees from University of Pennsylvania Medical School, trained as a Pathologist at Massachusetts General Hospital, and today he is Assistant Professor in the department of Pathology at the University of Michigan. His lab studies renal transplantation, long-term graft survival, and is searching for new biomarkers of organ rejection. They were the winners of the NanoString® Human Organ Transplant Panel Grant in November 2019.

Speaking with him, I had the impression that Dr. Farkash is not just an enthusiastic mentor to his students, but also a fun PI to work with.

NS: What led to your interest in renal pathology and transplantation?

EF: I would have to say influential mentors. At various points in my life, from undergraduate studies through graduate school, my residency and fellowship, key mentors have guided my path forward. Dr. Arlene Sharpe (Harvard Medical School) and Dr. Nina Luning Prak (University of Pennsylvania) are two pathologists that inspired me to consider pathology as a career. Dr. Robert Colvin was my mentor for renal pathology and transplantation pathology at Massachusetts General Hospital. Sitting across the scope on my first day of the rotation, Bob’s first question to me was, “How much do you want to learn about kidney pathology?” My response was that I wanted to learn as much as I could, and I don’t think I have stopped learning since then.

NS: Your second area of interest is the relationship between BK polyomavirus (BKV) and urologic cancer. Why did you expand your work into this area?

EF: I am broadly interested in long-term outcomes after kidney transplantation. We had a fortuitous observation during my fellowship—a transplant recipient whose kidney had failed in the first year after transplant due to BKV nephropathy then developed a urothelial carcinoma in the graft four years later. This is an unusual location to develop a tumor. We did some immunohistochemistry (IHC) stains to see if the two incidents were related and, in fact, there was diffuse BKV in every single cell of that tumor, suggesting a mechanistic correlation. This wasn’t the first instance of this; it had initially been observed by clinicians at Johns Hopkins back in in 2002. My curiosity about the linkage led me to pursue this new area of research.

NS: Is it common to see expression of the BKV antigen in transplant recipients?

EF: Unfortunately, BKV nephropathy is all too common. We probably diagnose it on biopsy in 10-15% of patients during their first year after transplant. This is part of the rationale for the two BKV genes on the Human Organ Transplant (HOT) panel. One of the things the panel can detect is the presence of active BKV infection. However, BKV-associated cancers have only been seen in organ transplant recipients, and even then are quite rare.

NS: What is the connection between the immunosuppressive drug tacrolimus and the BKV?

EF: Tacrolimus is a very effective immunosuppressant and the flip side of preventing rejection is the possibility of developing infection. In particular, renal transplant recipients are vulnerable to developing BKV nephropathy when they are more immunosuppressed.

NS: Are patients who receive a kidney transplant more susceptible to developing cancer after the transplant?

EF: Yes, in general transplant patients are known to be more susceptible to developing cancer. Specifically, they have a higher risk of developing virus-driven cancers and UV-related skin cancers. We’ve also have seen Merkel Cell Carcinoma in a small subset of our transplant population which is caused by a polyomavirus related to BK Virus.

NS: Why are you looking at an alternative to C4d as a biomarker for antibody-mediated transplant rejection?

EF: C4d is one component of a three-part test for antibody-mediated rejection using the Banff classification schema. We use it as a marker for donor-specific antibodies interacting with blood vessels in the transplant. C4d has good positive predictive value, but a number of studies in the early 2010s showed only moderate sensitivity. As pathologists, we were essentially missing cases of antibody-mediated rejection. A missed opportunity for diagnosis is a missed opportunity for treatment. Therefore, we need to look for alternative biomarkers.

NS: What characteristics make for an ideal biomarkers in organ rejection?

EF: Accessibility, affordability, and accuracy. Renal transplant biopsies are the gold standard but they are invasive, so an accessible sample is one that is easy to get from a patient, such as blood or urine. You want a test that’s affordable so that it can be used widely for screening a large population. Finally, a reliable biomarker has a pathologic range that is clearly distinguished from the normal range.

NS: Given these criteria, why did you choose to use NanoString’s nCounter® platform for your biomarkers discovery?

EF: I’ve been looking for an effective tool for expression analysis for my research for some time. As a pathologist, I have access to an archive of over 10,000 renal transplant biopsies from over the last 15 years at my institution. I’d like to be able to do more with the tissue we have, which are FFPE biopsy cores. We do have access to some frozen tissue and it’s always possible to gain patient consent to take extra cores in RNA-preserving solution up front. But we have this vast archive with extensive longitudinal follow-up that we would like to be able to access fully. This archive presents a wonderful opportunity for biomarker discovery as well as research into mechanisms of rejection.

NS: Your grant proposal aims to analyze kidney transplants biopsies using both the nCounter HOT panel and by single cell gene expression analysis. What do you expect to discover from this correlation?

EF: The HOT panel allows us to look at overall gene expression across all cells in the transplant biopsy and that will be affected by two things. First, the composition of the cells in the graft biopsy – are they normal kidney cells or is there an influx of other cell types like B cells, T cells, macrophages, NK cells, or other inflammatory cells that mediate rejection? Second, changes in expression in the normal cells – do the glomerular, tubular, and vascular cells show evidence of dysfunction, adaptation, or injury?  In contrast, single cell analysis allows us to measure changes in expression patterns of an individual cell. If we were forest rangers, the HOT panel would be like flying in a plane getting a broad view at the overall health of the forest, and single cell analysis would be like examining a representative set of trees in detail from the ground.  The HOT panel gives us a big picture of both what cells are present in the graft and whether they have changed their pattern of expression. We can then switch to single cell analysis to correlate those findings to changes in individual cell frequency and individual cell expression. For example, if the HOT panel identifies altered expression of genes associated with glomerular epithelial cells, using single cell sequencing we could tease out whether this is due to changes within a glomerular cell population versus a relative change in glomerular cell abundance due to an influx of inflammatory cells.  In addition, if we identify a trend in a gene or group of genes in the HOT panel, but it’s not statistically significant, we can use the single cell data to look in finer detail and see if that change is significant within a subpopulation of the patient’s cells.  Essentially, we are using the HOT panel as a tool for biomarker discovery, and the single cell analysis for validation and exploration of mechanism.

NS: Do you see the results of your research being used in the clinic in the relatively near future?

EF: We’re still early in the discovery phase at this point. There are a lot of potentially interesting questions in kidney transplantation outside of rejection or tolerance that haven’t received as much attention. For example, there is a lot still unknown about subpopulations of glomerular epithelial and endothelial cells and how they change over time in kidney transplants. These changes have important implications not just for acute events (like rejection) but for long term graft survival. If the kidney is going to survive for fifteen to twenty years in the recipient, then the endothelial cells and podocytes have to remain healthy for a long time, too.

NS: Your grant application described using GeoMx® Digital Spatial Profiling (DSP) and the nCounter System together in a “Forest and Tree” approach to biomarker discovery. How would you use these two technologies together?

EF: Pathologists rely on architecture significantly when making a diagnosis. The HOT panel has the ability to detect expression data from the paraffin-embedded tissue but at the cost of architectural information. GeoMx DSP and the nCounter platform is a nice technique for the best of both worlds to retain and pair architectural information and expression analysis.

NS: On a lighter note, we know that every lab has its own unique personality and history. Would you mind sharing a recent story about your lab with us?

EF: My lab is fairly small, comprised mainly of undergraduates and a technician. When we first started using the NanoString’s nCounter platform a year ago, I had a talented undergrad who did a brilliant job setting up the RNA extractions and doing the hybridization.  When it came time to loading the chip, I was excited to use the new technology so I pulled rank and insisted that I (as the PI) would load the samples onto the chip myself.  And wouldn’t you know it, I loaded half of the samples in completely the wrong order. The lab had to go back and reconstruct what I had done; fortunately we had replicates so they were easy to pair up. Which just goes to show you what happens when then PI steps out of the office and into the lab to try to seize control of the experiment.

Title provided for identification purposes only. The views and opinions expressed are those of the individual only and do not necessarily reflect the position of the University of Michigan

For Research Use Only.  Not for use in Diagnostic Procedures.

By Laura Tabellini Pierre
For research use only. Not for use in diagnostic procedures.