Please introduce yourself – what brings you to AACR this year?

My name is Anand Jeyasekharan. I'm a medical oncologist, and I treat cancer patients with lymphoma and sarcoma. I have a research lab at the National University of Singapore where we study why some of these cancers are cured by chemotherapy and others are not. A lot of that has to do with the other cells in the tumor that constitute what's called the microenvironment. We study the interactions between tumor cells and the tumor microenvironment in these cancers and try and relate them to chemotherapy responses – I'm presenting some of our work at AACR 2025. It’s such a broad conference with so many emerging themes, and I'm specifically interested in the immune system and what it's doing in cancer.

Tell me about your NextGen Star award

The NextGen Star program is for assistant professors and younger scientists. You submit an abstract and if chosen by AACR on this prestigious program, you can present your work as part of the scientific program at the annual meeting. The work we’re presenting is about why some people with lymphoma relapse after chemotherapy. Chemotherapy is the main form of treatment of these cancers, and in most cases, the tumor shrinks down nicely, but in some cases it comes back, and we've been trying to understand why this happens. We found that certain types of cells contain different oncogenes within the same cell, and that creates a “bad cell group”.

"For the first time in this work, we demonstrated that it's not only the quantity of these cells that matters – it’s also their spatial distribution, which appears to influence the surrounding environment."

How are you using spatial biology in your translational research?

Exactly as we were discussing – the number of cells is important, but their distribution within the tumor is equally crucial. This is a new area, and we've been drawing on principles from other fields like ecology and meteorology, where spatial patterns are studied – for example, how certain plants grow together or where specific animals are found. These disciplines often take a two-dimensional perspective, much like meteorology. We've been leveraging models that have been used in these fields for years, applying them to cancer biology for the first time.

How did you make that connection?

We started by trying to understand the difference between those who relapsed and those who didn’t. As we examined various factors, we noticed differences in the numbers of certain cells. But beyond that, we also observed distinct patterns, which led us to explore ways to study them more effectively. To do so, we turned to other disciplines that have long been analyzing similar patterns, drawing on their methodologies to apply to our own work.

How do you see the field progressing in the next five to 10 years?

There are lots of good technologies emerging, but these technologies are only as good as the reagents that we have and how accurate they are – Abcam is important in that sense. To bring these advancements into clinical practice, extensive standardization and optimization are required, which presents a significant challenge. How do we translate these exciting discoveries into real-world treatments that can be used for patients every day? There’s still a great deal of work ahead to make that transition possible.

Another key challenge is how to integrate all this data effectively. With AI becoming increasingly prominent, there will be significant work in the area of integrative analysis. The sheer volume of data we’re generating is so vast that manually identifying all the patterns would be impossible. AI will undoubtedly play a crucial role in making sense of it all.

What excites you about the field?

Everything – it's a new way of looking at cancer.

"A few years ago, we didn’t have these tools – or rather, we did, but at a much lower plex, meaning we could only analyze few markers at a time. Back then, it was just called microscopy."

Now, with the more advanced capabilities we have, we’ve given it a fancy name: spatial biology. This allows us to examine multiple factors simultaneously, unlocking new possibilities in research. As a clinician scientist, this represents an exciting new level of understanding in the areas we care about, such as why some patients relapse or why a particular drug works in specific situations. I see it as a way to gain even more valuable insights into the questions that drive our research and practice.

Are there any emerging proteomics-based technologies that we should be keeping our eye on?

I think clearly the higher plex technologies, particularly those capable or analyzing 100 or maybe 200 proteins at single cell resolution will be interesting, as will looking at how they can be combined with RNA-based technologies. And finally, we’re certainly interested in protein- protein interactions, and how we can study these interactions in tissue samples – that's going to be pretty exciting as well. None of these things are going to be straightforward, especially when we do them at scale, but we're doing things that we didn't think we could do a couple of years ago. In terms of the goal, we want to be able to measure hundreds of proteins in every cell within a tumor – and do so consistently across samples. That would be ideal.

What opportunities and challenges are there?

I think the challenges present interesting opportunities. In a study where we’re evaluating 200 patients at once, it’s easier because all the samples are processed together, allowing for direct comparisons. But in clinical practice, you handle one sample at a time; as each patient comes in, you analyze their sample individually. That means the process needs to meet the highest standards of technical rigor and standardization. Only then can you confidently tell that particular patient whether their test is positive or negative. When you have 200 samples for reference, it’s much easier to make that call.

When you're working with just a single sample, the challenge becomes much greater – especially for many of these technologies. Unlike sequencing, where you can definitively determine whether a mutation is present or not, these methods often produce readouts on a gradient or distribution. That makes it much harder to define a clear cutoff between positive and negative results. I think that's going to be a significant challenge, but with the right controls and standardization, it's something that can absolutely be overcome, and needs to be to bring these approaches into the clinic.