AACR 2012 Annual Meeting, March 31st - April 4th, Chicago

Metformin to slow prostate cancer growth

Dr Anthony Joshua – Princess Margaret Hospital, Toronto, Canada

I will present a study, this is a collaboration between Princess Margaret Hospital and McGill with Dr Pollak also being instrumental in the study design and rationale. The study was based on this drug metformin which is derived from the French lilac plant which even in medieval times was noted to make the excessive urination of Frank diabetes better. It was approved in Europe in 1975 and in the United States in 1995 for the treatment of diabetes. Subsequently it has been explored as a treatment and/or a preventative treatment for cancer based on a conglomeration of laboratory, epidemiological and clinical data. The molecular data is summarised in that box which comes from a paper by Dr Pollak. On the left, briefly it may reduce insulin in the blood which may promote tumour growth; that insulin in prostate cancer in particular may make testosterone signalling more potent. In the middle it may stress cancer cells and in particular in lung cancer where they don’t have the ability to respond to that stress, they may then die. And it also, in the final little circle there, may inhibit one of the major growth pathways, the motor, if you like, of the cell which is known as the mTOR pathway and metformin may have both direct and indirect effects on the mTOR pathway which contribute to cancer growth.

So with that in mind we undertook a neoadjuvant study, this means that men came into our clinic with a diagnosis of prostate cancer done on a prostate cancer biopsy, which is illustrated here, and then they were offered to participate in the study to the point at which they were told their surgery date was going to be. So there was a window of opportunity for drugs to be used to test their effectiveness on the prostate tissue, when we compare the initial biopsy tissue to the final prostate which is taken out and that window, at least in Canada, is between four to twelve weeks. So the men were given metformin at a dose ultimately of 500mg three times a day, which is even slightly lower than the dose which is prescribed to diabetics.

The crunch of the study is that if we looked at the tumours that we were able to identify, comparing the biopsies to the final prostates that were taken out, it did seem to reduce the rate at which the cancers were growing. So it didn’t stop the cancer growing but this Ki-67, which is a measure of how quickly is the cancer growing, was reduced by an absolute value of 1.65%, which is a relative decrease of 32% compared to the original amount of proliferation. This was accompanied then by a decrease in the staining of a marker of the mTOR pathway, this is a protein known as phoso-4E-BP1 and this is a readout, if you like, of how quickly that pathway is sending signals into the cellular nucleus so that the cancer cells will grow and divide. So that marker is significantly reduced when we compare these samples after the drug to the samples before the drug.

It did have an effect on the PSA, I’m not confident that this means that the drug is acting on the PSA, some of these men undoubtedly may have other reasons why their PSA goes up and down and these values were not the high values you see in castration refractory disease that Dr Montgomery talked about, these are relatively low values. This waterfall plot is intriguing but it is not yet something that I am comfortable pointing out as an advantage of the drug, as opposed to the previous two slides.

So the conclusion of the study was that it certainly is promising; it adds to the rationale of metformin being used in early stage prostate cancer. There are some limitations, though, that both patients and doctors need to be aware of. Firstly, the study hasn’t finished; secondly, there is no control group, so some men may have, for example, become so nervous about their prostatectomy that other things may have happened in their body and these changes may have been seen anyway. We’re in the process of evaluating that in a matched cohort; there is certainly heterogeneity in the tissue which means that some parts of the tumour stain and some parts don’t. Most importantly, this marker of proliferation that we looked at on the prostate tissue is not a clinically relevant end-point; it doesn’t tell you whether the men are going to benefit from it or not, it is simply a marker for how quickly the cancer cells are growing. Whilst that may intuitively relate to benefit, the relationship between the decrease in the Ki-67 index and ultimate benefit to the men is still unclear.

But certainly these differences suggest that there is a direct effect of metformin in prostate cancer tissue and this does bode well for future trials.