Meta-analysis confirms benefits and safety of precision medicine
Prof Razelle Kurzrock - University of California, San Diego, USA
First of all, let me explain what a meta-analysis is: it’s an analysis of existing data. It takes data that has been published in the literature in individual units of 50-100 patients, sometimes even less, and it puts all that data together to see if we can get better and bigger information. Although it was 50,000 patients on the title, it ended up being 70,000 patients so that’s really a lot of data. What we were trying to do is put together all the clinical trials that are phase II and phase III trials, so one of the analyses was phase II and the other was phase III trials, and look at it and see if precision medicine embedded in those trials made a difference. So some of the trials had a biomarker based approach or precision medicine, also known as personalised medicine, approach and a lot of other trials did not. We want to see what is the difference.
What was precision medicine compared against in the meta-analysis?
The comparison group was the same trials done during the same time period that did not have a biomarker based approach.
What were the main results of the meta-analysis?
It’s quite remarkable, in every way that we looked at it using personalised medicine is associated with a better outcome. It’s associated with higher response rates, it’s associated with longer time until the disease progresses and it’s associated with better overall survival. Really, of all the features we looked at, that was the most important feature for predicting a better outcome.
Did you look at the tolerability of precision medicine over standard therapy?
Tolerability was hard to look at from the data in the trials but it’s a very important question. We looked at safety, which is a measure ultimately of tolerability, and in particular we looked at the percentage of patients that died as a result of the drug that had toxicity. In one of the studies there was no difference between the patients that had traditional therapy and personalised therapy and this was very important because some people have claimed that the personalisation might be dangerous so that showed that it was safe. In the second study the patients with personalised therapy actually had a much lower rate of death that was due to drug so it was actually much safer than traditional therapy as well.
What were some of the other findings and how can they be interpreted?
I think that there are three main things, two of them that we were looking for and then one that was unexpected. So the first is do patients that have personalised therapy or precision medicine therapy, whichever term you want to use, do better? In every outcome we looked at they did better. The second is what I just mentioned, was this safe? Absolutely it was safe and, indeed, in one of the meta-analyses the rate of death due to toxicity or side effects from the drug was lower than in traditional therapy, so it was even safer than the traditional therapy. And those were two things that we set out to look for but sometimes when you’re doing a study you also find something that you didn’t expect that is really informative and that happened in this study. One of the ways that people think about targeted therapy is that targeted therapy in of itself is a great idea and that it’s better than chemotherapy because chemotherapy is harsh. What we found was that that’s not really true. Targeted therapy that is personalised, that is matched to the patient is the best thing that you can do for the patient but targeted therapy that is given to patients that are not selected in any way, is just given randomly to people, is actually the worst thing you can do to patients; it’s almost worthless when you give it to unselected patients. Indeed, we were wrong, it’s not better than chemotherapy, it’s worse than chemotherapy. So if we go from best to worst, best is personalised targeted therapy, second best is chemotherapy and the worst is targeted therapy that is given to unselected patients.
