The world of precision and individualized medicine is undergoing drastic changes with the application of new technologies. Among these, proteomic profiling and imaging is changing our understanding of cell functions, particularly in cancer biology.
Psomagen was excited to sit down with Dr. Manish Kohli to discuss two of his recent publications in the field of prostate cancer research. With decades of experience in clinical practice and research, Dr. Kohli now works as an oncologist at the Huntsman Cancer Institute in Salt Lake City, Utah.
The first paper, Circulatory prostate cancer proteome landscapes and prognostic biomarkers in metastatic castrate resistant prostate cancer, used Olink Explore 3072 at Psomagen to identify differentially expressed proteins (DEPs) in advanced stages of prostate cancer.
In the second paper, Plasma Cell-Free DNA Methylation-Based Prognosis in Metastatic Castrate-Resistant Prostate Cancer, researchers expanded on the first study by analyzing clinical biomarkers and additional patient samples to develop a nomogram quantifying the risk associated with the combined cfDNA methylation changes.
In the following article, learn more about his key takeaways from these recent research projects, and where these insights may lead for prognostic indicators or new clinical trials. Then, take a look at his thoughts on the future of precision medicine and the research disciplines that hold promise for disease treatment breakthroughs.
For Dr. Kolhi, insights brought forth by efforts like the Human Genome Project have long foreshadowed the future of precision medicine. “Since 2005, it’s become clear that the molecular world is going to come upon us to be used as a tool for patient care. And it’s even become more clear after maybe 2018, 2019, that omics platforms like epigenomes and proteomes are going to become more involved beyond circulating tumor DNA or beyond transcriptomics.”
Omics platforms have allowed for the generation of massive datasets across many disciplines of molecular biology. For over a decade, a major part of Dr. Kohli’s work has been developing and annotating real-world liquid biopsy biobank data.
Now, it’s becoming increasingly important to integrate multiple data sources in these datasets: “We try to use different omic platforms in these real-world biobanks to probe for questions which don’t have ready-made answers from clinical estimates — because clinical estimates is what we do at this point in time. Those clinical outcomes include prognosis, prediction, monitoring of disease using molecular tools. The idea is to derive signals, and then to test those signals and to validate those signals, and then to prospectively also take them into the research world.”
Dr. Kohli refers to his first paper using Explore 3072 as “a kind of alpha paper, or the canary in the shaft, using this technology.” The project made use of a well-annotated genitourinary liquid biopsy biobank developed by Dr. Kohli’s team the Huntsman Cancer Institute. They used Olink technology “to look at a temporal landscape of proteins specifically focusing on the 750 or 738 tumor-associated proteins in this panel.”
The project looked at two plates of samples for “patients in the continuum of prostate cancer, from local stage to a slightly more metastatic stage called metastatic hormone sensitive-stage, and then an even more stage of progression called metastatic castrate-resistant stage (mCRPC).” With disease outcome information, the team was able to compare protein expression across prostate cancer timepoints and prognoses. The goal was to create “a descriptive landscape of the protein repertoire on this panel across these different stages.”
The findings of the paper were significant. Of the 738 proteins targeted in the Explore 3072 panel, they found “about 26 proteins which are truly, truly statistically underexpressed or overexpressed in different stages.” This information allowed the team to then focus on a key goal of their research: they wanted to “look at the castrate-resistant stage of progression, look at the overexpressed proteins we can reach after a differential subtraction from the previous stages.”
“Then, after reaching those overexpressed proteins in the castrate-resistant stage, look at what are the candidates over there that are overexpressed in a stage-specific way, but also give us a signal that they may be used for prognosis in the patient population. In terms of short survival, is there overexpression linked with short survival? Is there overexpression linked with long survival?” These questions led to the generation of the heat map below (Fig. 1), which details the overexpressed and underexpressed DEPs found via this proteomic analysis.
When testing this technology and its utility for oncological research, it was critical to combine it with existing information. Dr. Kohli’s team wanted “to see whether our current knowledge is enhanced using this technique.” Their results proved that integrating proteomic data and targeted proteomic panels had success — of the 26 dysregulated proteins, two were strong prognostic indicators of short survival in castrate-resistant prostate cancer.
Identifying prognostic DEP biomarkers was an important step forward in advanced prostate cancer research. Dr. Kohli’s team turned their focus to increasing the scope and sample size of the next project. In their next study, Plasma Cell-Free DNA Methylation-Based Prognosis in Metastatic Castrate-Resistant Prostate Cancer, the team evaluated epigenomic features that predict clinical outcomes using methylation haplotype blocks specifically overexpressed in CRPC state.
An ongoing follow-up study now is evaluating an increased number of patient samples and DEPs of significance and integrating the DEPs with epigenomic signals.
“We’ve worked with the parent Olink company also to not double but a little more than double the total number of patients in the study…. We have gone back to look at the signals not just in 738 proteins but in 3,000 approximate proteins, all pathways, because we want some more robust data given that this approach seems to work.”
With data from over 350 patient samples, they found that “it’s not just two proteins, but there are at least 20 proteins overexpressed in a stage-specific way in castrate-resistant disease.” Current clinical tests rely on biomarkers that are non-specific to disease stage. However, by combining this new data with historic models, they’ve been able to create a plot showing the hazard ratio is for the clinical outcome of each DEP in metastatic castrate-resistant prostate cancer.
These results provide evidence of the biological processes at work. “It is clear that there are specific pathways in the gene ontology network which focus on oxidative reductase pathways. When proteins in those pathways are overexpressed, those patients do worse in this particular stage, which also opens up potentially the role for targeting those pathways in the future.”
Furthermore, this study was able to develop a cell-free DNA based prognostic model for metastatic castration-resistant prostate cancer, which Dr. Kohli considers a “first exercise” before moving on to predictions of drug efficacy.
The oxidative reductase pathway is the focus of other disease research and drug trials, so there are already possible therapeutics for these prognostic indicators. Dr. Kohli sees a clinical trial for mCRPC as “a very different exercise from this one, which is to say, that if we get a signature, we have to then look at what is the prevalence of that signature in the compilation to begin with? Because based on the prevalence, you will then be able to devise and design a prospective clinical precision medicine trial, or not. And so those are very different exercises in prospective clinical trial modeling compared to getting the signature.”
Having epigenomic and protein datasets is a strong first step, especially given the “mixed report card” Dr. Kohli has seen with DNA-based clinical trials. “If you look at big pharma’s definition of precision medicine at the current stage of evolution,” Dr. Kohli said, “it’s mostly based on DNA-based alterations. A 'Genotype Matched Treatment' (GMT) for drug-altered DNA approach has been the focus for precision medicine.”
However, this approach does not always work in reality. “For the most part, given the iMATCH trials from the NCI and so many trials from Europe, the DNA match has not been very successful, sadly. You may have the alteration, and you may have the drug, but it doesn’t always work…. Prospective clinical trials with drugs matching a DNA alteration have a response rate which is between 5 to 15% and not long-lasting. So that’s a disappointment in many ways for several patients whose biology may be defined better using molecular platforms beyond tumor biology-based altered DNA patterns.”
Where does that lead future precision medicine and clinical trial efforts? Dr. Kohli points to early successes in the transcriptome, epigenome, and proteome. First, RNA-based technologies have led to some oncology breakthroughs. “For example, a very large success is a companion diagnostic in the Oncotype DX breast cancer adjuvant trial situation, where you have a precision medicine instrument like the Oncotype DX RNA based signature, and you give a lady with breast cancer adjuvant chemotherapy or not based on that recurrence score.” Predictive scoring like this holds promise for improving treatment response rates.
The epigenome and proteome may also have a promising future in medical practice. “More than 95% of drugs are targeting proteins, they’re not targeting DNA alterations. I think this sets up stages to capture heterogeneity, which is actually expressed and may be associated with progressive tumor biology more closely than DNA based alterations, at least in prostate cancer.” Technologies that can capture this heterogeneity in tumor biology will lead to novel insights and potential treatment targets.
Dr. Kohli’s pioneering work demonstrates the use of proteomic profiling and multiomics in redefining how we understand and treat advanced prostate cancer. Through the use of cutting-edge tools like Olink Explore 3072 and large-scale biobank datasets, his research identifies biomarkers with prognostic value and points toward a future where data-driven treatment becomes the norm.
As precision medicine continues to evolve, efforts like these show the importance of integrated molecular insights and collaborative innovation in driving the next generation of cancer care.