Order out of Chaos

How we finally learned to understand the chaos in cancer

Have you ever wondered why cancer is such a difficult disease to treat? Why some cancers are more aggressive than others? And do you ask yourself why some patients will not respond to their chemotherapies?

We, a team of 14 scientists from the Cancer Research UK Cambridge Institute, CNIO in Madrid and The Francis Crick Institute in London, have found answers to those questions and recently published our findings in Nature. As spearhead of our team, it is my pleasure to introduce you to our research on chromosomal instability (CIN), a major and - up until now - poorly understood hallmark of cancer affecting 4 out of 5 patients.

What is chromosomal instability or CIN?

Let’s start with our genetic code: It is distributed across 46 chromosomes, very much like 46 books of an Encyclopaedia that make up you. 23 books or chromosomes come from our mothers and 23 from our fathers. Everything is present in two copies. This fundamental fact underpins all our work. However, this state of 46 chromosomes is almost always disturbed in cancer, with 4 out of 5 patients affected. During the development of cancer, chromosomes get lost or duplicated, chromosomes break and get stitched together randomly, parts of the chromosomes get deleted or multiplied. Think of it as a ripping out pages out the 46 books, copying pages or sentences, gluing pages or whole chapters together all across the other books. You get the picture: the genetic code of every single cancer is uniquely chaotic (see Figure 1 below).

Normally, our genetic code is extremely closely protected, monitored and repaired. Human cells do not tolerate changes to their chromosomes. But in cancer all bets are off. Suddenly cells can accumulate seemingly endless changes and produce surviving, thriving and resilient cells. This phenomenon is perplexing to say the least and has scientists left baffled for over 130 years.

Figure 1: Chromosomal instability transforms normal cells with 46 chromosomes into tumour cells with no fixed structure

In fact, this chaos is a fundamental characteristic of cancer and was the first molecular feature to be discovered under the microscope by David von Hansemann and Theodor Boveri in the late 19th century. Though CIN was one of the first characteristics to be discovered, it turned out to be one of the hardest to be understood. We simply know too little about our genetic code to understand or predict what it means to lose a copy of chromosome or to delete a few bits here or multiply a few bits there. This is why there are few effective treatments and therapies available for cancer patients with high levels of chaos in their cancers (see Figure 2). This is why cancers such as Ovarian, Oesophageal or Lung cancers have some of the lowest rates of survival of all cancer types. This is why we care and invest our careers into studying CIN.

Figure 2: A limited understanding of chromosomal instability leads to lower 5-year survival rates

??

17 signatures of complex patterns bring order into the chaos?

In short, we were able to identify patterns in the chaos. We spent years studying the genetic code of almost 8,000 cancer patients representing 33 tissue types. Over time, we started to see patterns. Using computational algorithms, we identified combinations of complex patterns which we called “signatures”. Using the Encyclopaedia analogy introduced earlier, a signature could be the deletion of chapters of a certain length or the amplification of short sentences. In total, we identified 17 signatures of CIN which allow us to dissect the chaos into its fundamental building blocks.

What makes our study so strong is that we not only identified 17 signatures but we also linked them to faulty mechanisms that drive cancer. With high certainty, we can now look at the genetic code of a patient’s cancer and say which mechanisms have been acting with which activity. This is huge. We showed that there is much more order in the chaos than ever thought or shown before. Figure 1 does not show a chaotic cancer genome anymore, it shows the result of the activity of a distinct number of faulty mechanisms.

Now that we have gained these unique and novel insights into CIN and cancer, what can we do to help cancer patients?

Once you know the mechanism, you can predict the treatment outcome

You probably have heard of platinum-based chemotherapies or platins for short. These are the backbone of today’s chemotherapies with roughly half of patients receiving platins. But platins have a long list of side effects and don’t work in all patients. We know since the late 90s that platins interfere with the repair of our genetic code and that some cancers cannot handle this interference very well and die. Now, because we identified 4 signatures of faulty repair mechanisms, we hypothesised that only a subset of patients with a particular signature will react to platins. In other words, we developed a clinical predictor for platinum response.

We studied this hypothesis on Ovarian cancer patients which were regularly treated with platins and found that only 27% of patients had the right combination of signatures to potentially benefit from their therapies. Those 27% of predicted responders survived almost 70% longer compared to the 73% predicted non-responders: from an average 3,5 years to almost 6 years (see Figure 3). It is extremely promising to predict this large survival benefit for a subset of patients simply by studying patterns of their genomic changes without any additional information such as driver genes. We have further confirmed the ability to predict response on other cancer types such as oesophageal and breast cancer.

Bringing this classifier into the clinics could mean a great step forward to saving patients from the ordeal of painful and non-effective chemotherapies.

Figure 3: Our signatures allowed us to identify patients which had a substantial survival benefit by receiving platinum-based chemotherapies

Can we develop new therapies now that we know all this about cancer?

We certainly hope so. Our signatures also allowed us to identify completely new links in the complex biology of cancer. This means, in the near future new therapies might be developed for 49 new drug targets (see Figure 4).

In fact, three of our senior authors have founded a UK-based precision medicine company called Tailor Bio to do exactly this! They are working tirelessly to translate our findings into the clinics.?

Figure 4: The 17 signatures predicted response to 44 known drugs and identified 49 druggable targets for which novel therapies could be developed

In summary, our work enables us to better understand what happens in many cancer patients. It allows us to predict response to certain chemotherapies and also to guide the development of new therapies. Now that we found order in the chaos, we finally start to understand why cancers look the way they do. I hope this is very much in the spirit of David von Hansemann and Theodor Boveri who discovered and described CIN over 130 years ago.

Figure 5: The cover of this month’s print issue of Nature

Link to our article: https://www.nature.com/articles/s41586-022-04789-9

Link to the pdf: https://rdcu.be/cPIKf

Are you interested in a more scientific introduction to our work? Head over to our Tweetorial on Twitter: https://twitter.com/gjmacintyre/status/1537089902210428928

#17DeadlyCINs

More about Tailor Bio: https://www.dhirubhai.net/pulse/tailor-bio-17-deadly-cins-tailor-bio/

Contributions – It takes a village to raise a publication?

This long and challenging project started out as my PhD thesis but then became the large-scale publication which we now can share with the world. Only the efforts of a strong and dedicated team made my initial work and our novel ideas into the publication that it is today. Thank you everyone!

Thank you Geoff Macintyre and Florian Markowetz!

Thank you Barbara Hernando, Maxime Tarabichi, Kerstin Haase, Tom Lesluyes, Philip S. Smith, Lena Morrill Gavarró, Dominique-Laurent Couturier, Lydia Liu, Michael Schneider, James D. Brenton and Peter Van Loo!

The article was written by Ruben M Drews who is a Visiting Scholar at the Cancer Research UK Cambridge Institute and an employee of GSK Plc. This article as well as all the research mentioned were written and performed fully outside of his employment at GSK Plc and have no connection to the company.