The similarities between prostate and breast cancer

Mathematical biologists deal with spherical cows, that is, we look for patterns in nature. That means abstracting some details so that we can concentrate our attention on the key features that are relevant to our question.

That is why, as a mathematical modeller working on prostate cancer I was always interested in breast cancer.  The first is one of the most common cancers in men whereas the latter is one of the most common in women (although men can also have breast cancer). Both are cancers that initiate in the epithelial cells that form the glands that characterize prostate and breast. Both are usually quite treatable and most patients survive. When things take a wrong turn, both prostate and breast cancer usually metastasise to the bone. Surely there are substantial differences between the two cancers but there are also enough similarities that I was not too surprised with recent findings showing that genes that are usually mutated in breast cancer are found to be mutated in prostate cancer too (a fifth of prostate cancer patients have mutations in BRCA1 and 2). The study has been labeled the Rosetta stone of prostate cancer. One of the take-home messages is that 90% of metastatic tumours are dominated by cells with mutations for which we have treatment. This is a rather optimistic point of view though, as we know that these metastases are heterogeneous and that thus, emergence is likely to emerge.


One year ago: talk at Simons Institute

About a year ago I was in Berkeley, having accepted an invitation from Rick Durrett. At that stage we had recently published our research on an integrated computational model of the bone microenvironment and prostate cancer cell metastasis to the bone. I have described the paper before but here you can hear me mumble about it for about 45 minutes.

The t-shirt is quite cool and if you want one, well, the Pint of Science festival is just around the corner.

The complex route to metastasis

The latest issue of Nature contains this very interesting article describing two different pieces of research on the topic of prostate cancer to bone metastases. Unfortunately you need to have the right IP address to access these papers free of charge.

Having done some research on prostate cancer to bone metastases (thanks to the Lynch lab) our group has now a better understanding of this disease and about the genetic and cellular drivers that characterize it. We have been working on the role of heterogeneity in cancer before but the source of it gets confounded in metastases: does it generate as one single metastatic cell reaches the bone and the clones acquire mutations? To this now we can add: do heterogeneous metastases in the bone come from a heterogeneous group of  tumor cells traveling together from the prostate? do these metastases start in a homogeneous clonal fashion and become heterogeneous as prostate cancer cells from other metastatic sites go to the bone? A combination of all these possibilities is (wait for it) also a possibility.

A relatively complex model of prostate cancer to bone metastases


CancerEvo got some good news last week. The paper that we have been working together with the Lynch Lab has now been accepted for publication at Cancer Research. Of course being accepted and being published are two different things so it might take a few weeks before those of you with the right subscription get access to the research but we are working to get something in biorxiv in the next few days.

Expect to find a relatively complex model based on the hybrid discrete continuum CA paradigm. That is, an agent based model where cells are governed by rules and molecules like TGF-Beta and RANKL are described by partial differential equations (PDEs).

Did I mention that this model is relatively complex? Well, it kind of is and it is no accident. In general mathematical oncologists pride themselves in producing the simplest mathematical models that fit the problem: the simplest most elegant model is always the best. But I am not sure whether this approach works so well when studying things like homeostasis. Homeostasis in the bone emerges from the interactions of a variety of cells like osteoclasts, osteoblasts, MSCs, etc. Successful metastases to the bone will have to disrupt this homeostasis and co-opt these interactions for their own advantage.

Some of these are mediated by molecules like TGF-Beta and RANKL and that is what we have investigated in this work but I also wonder whether we are missing out by just restricting ourselves to the molecules and cell types that have been the focus of attention by experimentalists. Of course the advantage of doing that is that we have a solid understanding of the known biology thanks to Leah Cook and Conor Lynch.

Which leads me to another point: usually mathematical oncologist work with experimentalist at specific points of the development of a project. Hopefully at inception (not always), then later on when there’s some model results to discuss and finally during the writing of the manuscript. Conor and Leah were involved at every step of this project, which I think is one of the reasons why we could get away with a little bit of extra complexity in this case. It is unfortunate that the current academic publishing system still does not have a way to acknowledge authors in projects with more than one PI and more than one postdoc doing the bulk of the work. At least you, the reader of this post, knows.