Dr. James V. Alvarez Video (Text Version)
Title: Pathways Regulating Tumor Recurrence
Investigator: James V. Alvarez, PhD, University of Pennsylvania
I received a Postdoctoral Fellowship to study the mechanisms by which breast cancers recur. By using inducible, conditional, transgenic models of breast cancer, we’ve identified a pathway involving the proapoptotic protein Par-4 that we think is a critical regulator of breast cancer recurrence.
Greater than 30% of women with breast cancer after their initial treatment, which is apparently successful, actually harbor within their bodies what’s referred to as minimal residual disease of minimal neoplastic cells. These cells are very, very difficult to study because they’re very, very difficult to identify, and there’s no in vitro approaches or approaches in test tubes if you will that allow you to study these approaches.
The real innovation in the approach that we’re taking is that in the mice we are able to isolate these neoplastic cells, these minimal residual disease, to try to characterize them, understand their properties so that we can hopefully one day eradicate them thereby presenting breast cancer recurrence.
To try to get at this, we’ve made a number of inducible mouse models of breast cancer where we can express different oncogenes under the control of the tetracycline-regulated promoter. And what this allows us to do is by crossing a driver mouse that has RTTA under the control of an MMTV mammary-specific promoter we can cross that to various responder strains that express different oncogenes under the control of the tetracycline operator.
So in these bi-transgenic mice and here I’m showing a mouse that drives the activated HER2/NEU gene under the TET operator, and these bi-transgenic mice in the absence of doxycycline the transgene is not expressed anywhere in the body. But when you add doxycycline, to the drinking water, the transgene gets expressed. And what we found is in almost all cases, the mice expressing these oncogenes develop invasive mammary adenocarcinomas that metastasize to the lungs with varying latencies between 3 and 6 months.
And so one key feature of these models is that they’re reversible so we can turn off the oncogene that we’ve expressed to mimic a targeted therapy and ask what happens to the cancer. And what is shown in B here is that when you do that, so if you add doxycycline at this point the tumors form and these are just tumor volumes shown here on the X-axis. And then if you remove doxycycline to turn off HER2/NEU, the tumors shrink very, very rapidly.
But what we found is if you just wait and monitor the tumors they ultimately recur in a stochastic process over the period of 3 to 6 months. And so what this tells us is that there must be in the mouse following tumor regression a focus or foci of latent probably dormant residual tumor cells that can ultimately reacquire the signals necessary for growth and lead to a recurrent tumor.
And C is just showing a Kaplan Meier plot indicating that almost 100% percent of the mice that previously harbored fully regressed tumors develop recurrent tumors within about a year. And D is just showing this schematically where this mouse here has tumors in almost all of its 10 mammary glands that completely regress but ultimately recur and when they recur they no longer express the initiating oncogene which is why the luciferase signal that’s present in the primary tumor is absent in the recurrent tumor.
And so we feel like these models are very important and they can give us some insight into identifying the processes that drive the formation of recurrent tumors and that drive the survival of residual neoplastic cells in the mice.
I think one of the important findings from my research is that this gene Par-4 and the process of multi-nucleation are really—seem to be commonly preserved. They’re—they’re common to several different mouse models whether they’re driven by the HER-2/NEU oncogene, the c-MYC oncogene, and the WNT oncogene, which are all oncogenes, that are themselves or whose pathways are implicated in a high proportion of women with breast cancer. So I think that the use of multiple mouse models really helps us identify things that are likely to be important for a number of women with breast cancer.
And so we—we finally just wanted to try to determine whether some of the findings that we’ve come across in mice also hold true in women with breast cancer. And so to do this, we reasoned that if a particular breast cancer has low Par-4 expression that following therapy and, in this case, it could be treatment with Herceptin or it could be traditional chemotherapy, cells with low Par-4 should survive better than cells with high Par-4 and that therefore tumors that have low Par-4 expression in their bulk tumor are likely to relapse more quickly than tumors with high Par-4. So to try to analyze this, we looked at a number of publicly available breast cancer gene expression data sets, and we divided the tumors into two groups—those with high Par-4 and those with low Par-4. And we asked whether the recurrence-free survival of those two groups was different. And what’s shown in figure six is Kaplan Meier recurrence-free survival curves indicate that was true in all of the data sets that we looked at. So in green is shown tumors with low Par-4 expression and in red is shown tumors with high Par-4 expression and you can see the tumors with low Par-4 expression recur more quickly in each of these six data sets suggesting that this pathway that we’ve identified in mice as playing a functional role in recurrence is likely to also hold true in women with breast cancer.
We always try to anchor our research that we do in mice to the situation as it exists in the clinic. While we believe that the mouse model is the best model for studying breast cancer breast cancer progression as it occurs in the body, ultimately we don’t want to cure breast cancer in mice. We want to cure breast cancer in women. And so the importance of taking these results which are I think very convincing and very compelling in mice and moving them to a point where they can be translated to actually you know affecting the lives of women with breast cancer can't be overstated. And so I think the next step with this project is really to try to identify proteins that lie in the same pathway that are attracted to drug targets, so we can hopefully intervene therapeutically in this pathway to prevent the emergence or the relapse of breast cancer.