榴莲视频

For patients with metastatic renal cell carcinoma (RCC), there are as yet no reliable biomarkers to predict response to therapy, but investigators are pursuing many avenues of research.

"Although tumor-intrinsic features such as pathological characteristics, genomic alterations, and transcriptional signatures have been extensively investigated, they have yet to provide definitive, robust predictive biomarkers," David Braun, MD, PhD, of Yale Cancer Center in New Haven, Connecticut, and colleagues wrote in the

Emerging techniques such as single-cell analysis, circulating tumor DNA, and study of the microbiome show promise in improving the understanding of RCC and uncovering biomarkers, Braun and colleagues said.

Braun, a medical oncologist who is also a member of Yale's Center of Molecular and Cellular Oncology, discussed the state of emerging techniques in the following interview.

Why has discovering predictive biomarkers been especially challenging in RCC?

Braun: I think there are a few factors at play here. First, current immunotherapies have complex mechanisms of action that often involve the interplay of many factors, and so biomarker discovery for immunotherapies is a challenge in general.

For RCC specifically, it has been particularly difficult in part because kidney cancer does not always follow the conventional paradigms for solid tumor immunology. For most cancer types, having a high number of somatic mutations -- which can translate into a large number of neoantigens -- has been associated with better response to immunotherapy. However, this is not the case in kidney cancer. This is just one of many examples of how RCC is immunobiologically distinct from many other solid tumors, and so it becomes more difficult to extrapolate biomarkers from other cancer types and have a high confidence that they will be applicable to RCC.

To put it briefly, I think that to develop robust biomarkers in RCC, we will have to (a) better understand the immunobiology of the disease; and (b) study and test putative biomarkers specifically in kidney cancer.

How has single-cell RNA sequencing improved the understanding of RCC?

Braun: Single-cell RNA-sequencing (RNA-seq) represents a true technological advance that furthers our understanding of kidney cancer. Whereas traditional RNA-sequencing and other "bulk" methods for measuring gene expression look at the average across all cells in a given tumor sample, single-cell methods allow us to understand the cellular composition, transcriptional phenotype, and even cellular interactions within the tumor microenvironment.

I frequently use the analogy of the fruit salad representing the heterogeneous nature of a tumor microenvironment, with each type of fruit representing a different type of cell. Conventional RNA-seq puts the fruit salad into a blender and makes a fruit smoothie. It's really hard to look at a fruit smoothie and understand exactly which type of fruit is most important for its color and taste. Similarly, it's very difficult to look at bulk RNA-seq and understand which cell type is contributing to therapeutic response or resistance.

Single-cell methods are analogous to being able to look at the original fruit salad, one piece of fruit at a time, and understand how it might differ in various situations. Similarly, single-cell RNA-seq allows us to look at each individual cell and understand differences in early versus advanced disease or tumors that are responsive or resistant to a given therapy.

There are lots of examples of how single-cell RNA-seq is improving our understanding of the disease. In one example from our group, we used this technique to understand how the tumor microenvironment changes as you move from normal kidney to early-stage RCC all the way to metastatic RCC. We found that the CD8 T cells become more exhausted (dysfunctional) in advanced disease, but also the myeloid component transitions to a more immunosuppressive phenotype.

Importantly, these changes do not appear to be independent, but rather the T cells and myeloid cells are talking to each other -- the myeloid cells produce ligands for inhibitor receptors on T cells that drive exhaustion, and those T cells produce factors that support this adverse myeloid state.

Overall, these cells form a bi-directional "immune dysfunction circuit," and the hope is that, after identifying this circuit through single-cell RNA-seq, we might now be able to test which "wire we can cut" (i.e., which cell interaction to inhibit) in order to restore effective anti-tumor immunity.

What is the status of research into circulating tumor DNA for predictive biomarkers in RCC?

Braun: There has been tremendous excitement about circulating tumor DNA as an emerging method for early diagnosis, detection of minimal residual disease (impacting who might receive adjuvant therapy), identifying early recurrences, and even tracking response to systemic therapy.

There have been substantial challenges, however, in applying these methods to RCC. While some excellent efforts have been made, the amount of DNA that RCC "sheds" into the blood is very low compared with most other solid tumors (and is probably most comparable in this regard to a brain tumor like glioblastoma multiforme).

While "off-the-shelf" and bespoke assays have made some progress, it has been difficult to get to the point of having a clinically actionable test in RCC.

One emerging approach is to look at patterns of cell-free methylated DNA. In work from Drs. Nuzzo, Berchuk, Korthauer, Choueiri, Freedman, and colleagues in , investigators used cell-free methylated DNA immunoprecipitation to identify DNA that was differentially methylated in RCC tumors compared with healthy controls. Rather than looking for just one or perhaps a few tumor-specific mutations (a needle in the haystack), this method allowed investigators to look at hundreds of differentially methylated regions.

The early results were promising, with highly accurate classification of plasma as coming from RCC patients or healthy controls. This was an initial proof of concept, and now needs further validation to move it forward into the clinic.

One potentially promising therapy developed from knowledge of the gut biome is the live bacterial product CBM588. What can you tell us about this?

Braun: There is now a robust body of scientific literature to support the impact of the gut microbiome on anti-tumor immunity, and specifically the efficacy of immune checkpoint inhibitors. With this background, Dr. Pal and colleagues have examined the live bacterial product CBM588 in combination with anti-PD-1 -- based treatment regimens in metastatic clear cell RCC.

For CBM588, the non-pathogenic spore-forming bacterium Clostridium butyricum, it was postulated that this bacteria may positively impact anti-tumor immunity both by producing butyrate (which, like other short-chain fatty acids, can augment T cell responses) and by supporting the growth of bifidobacteria (which had previously been shown by Dr. Gajewski and colleagues to promote effective anti-tumor immunity in the context of PD-1 axis blockade).

To test this hypothesis, Drs. Dizman, Pal and colleagues conducted a small prospective clinical trial of a standard first-line regimen, nivolumab+ipilimumab, with or without CBM588. The results, reported in , were impressive: even in this small study, there was a notable improvement in response and progression-free survival among patients receiving the CBM588.

In a second trial using a different standard-of-care regimen, Drs. Ebrahimi, Pal, and colleagues investigated nivolumab+cabozantinib with or without CBM588. Again in this small study, presented at the 2023 , the addition of CBM588 improved clinical outcomes.

While these are very promising early results, it is important to note that they are both small studies, and so need large-scale validation. Further, there did not appear to be a significant change in bifidobacterium species (one of the putative mechanisms of action), highlighting the need to better understand the biology of this approach.

How is artificial intelligence helping to identify biomarkers in RCC?

Braun: This is a big topic, and I think it is safe to say that AI [artificial intelligence] approaches are finding their way into just about every aspect of RCC, from radiomics to pathology to clinical outcomes based on large clinical datasets.

In my view, we have a number of very exciting biomarker approaches that are emerging, including understanding tumor-intrinsic biology (e.g., somatic alterations), to dissecting the tumor microenvironment (with single-cell RNA-seq), to understanding host immunity (e.g., HLA carrier status and diversity), to dynamic measurements with molecular imaging, and all the way to functional modeling using ex-vivo tumor samples.

Each of these approaches may provide some insight, but ultimately, integration of these high-dimensional, multimodal data will be necessary. This is one area where AI can play an enormous role.

Read the review here and expert commentary about it here.