Personalised skin cancer vaccine

11 July 2017

Interview with

Catherine Wu, Harvard Medical School

In the last ten years or so the rates of malignant melanoma, which is a form of skin cancer, have doubled. But our ability to cure the disease has remained pretty dismal, with 5-year survival rates down at around 10% for people who present with a cancer that’s already spreading. This week there's some good news though: a pair of papers by researchers in Germany and the US have described techniques for producing personalised anticancer vaccines, which single out chemical differences between a patient's cancer cells and their healthy tissue... and triggers immune T cells to attack exclusively the cancer. It relies on the fact that cancers make a lot of genetic spelling mistakes - called mutations - when they copy their DNA, and this is what introduces the differences recognised by the vaccine. Cathy Wu leads the US team at the Dana Farber Cancer Institute. She explained to Chris Smith what her team have achieved...

Cathy - We were able to take tumour specimens, sequence them to understand what was the entire DNA content of these melanomas, and then using new computer tools predict which of those mutations could potentially be seen by the immune response as tags that mark the cancer as different.

Chris - So, essentially, you are comparing what's normal for that person - from healthy tissue - with what's happened to the genetics of their cancer to see if there are changes there that will be reflected in the way the cell appears on its surface markers that the immune system might be able to spot that will be specific to the cancer and will be absent from a healthy cell in any other part of the body?

Cathy - Exactly. What makes our vaccine different from what has been generated previously is it’s not a one-size-fits-all, but rather that it’s personalised and tailored to the individual cancer that we are studying.

Chris - Roughly, how many markers do you get for each patient?

Cathy - We get hundreds to thousands of mutations, but only a subset of them are going to be displayed in a format that the immune system can see. But even from those, we have several hundred from which we can prioritise and choose. So we selected the top 20 hits that we could predict and that’s what the vaccine is more composed of.

Chris - So having identified these DNA differences, the mutations that single out the tumour compared with a healthy tissue, you then work out what those genes would turn into on the cell surface in terms of these chemical markers and you make an artificial form of those chemical markers which is going to form the basis of your vaccine.

Cathy - Right. So that artificial form is what we call a peptide. It is a string of building blocks of proteins and they have that change generated by the mutation. So if you can imagine that we have up to 20 of these strings of amino acids and we put it into a syringe, and we give it together with an immune stimulant and we do give a series of injections – 5 over the first three weeks and then 2 boosters – the idea is that immune-stimulants sends out the signal for immune cells to come into the vaccine site, gather up those peptides, and then display it to the immune response where a cascade of events then lead to the calling of that army of T cells that might then recognise the patient’s tumour cells.

Chris - How long does it take to make one of these vaccines because one of the things about cancer is people don’t want to hang around, and we know that holding up treatment could actually lead to a disease acceleration or the disease progressing? So how long does it take you to get this ready to go and into a patient for each person?

Cathy - For our study, it required up to 3 months. However, there is no reason this could not be created within 4 to 6 weeks.

Chris - Critically, did it work, Cathy? Did you get T cells capable of combating the tumours in these people? Second question, did those cells actually combat tumours?

Cathy - We have to remember that this was a phase one study. Meaning that this was our first foot out the door and our goals were safety and feasibility. In that, we definitely succeeded. So we were able to show that this whole procedure caused very, very minimal side effects. The other goal that we had was asking: Could they generate a strong immune response? There, we were also very successful. We identified very, very strong immune responses that we could pick up in the blood. Finally, we could demonstrate that those immune responses could clearly not only circulate but also, recognise the patient’s own tumour. The best measure of whether or not we were successful is whether or not the patients did well in the long term. We observed our patients for 2 years or more and they have been without evidence of disease return.

Add a comment