Professor Georgina Long AO

So one really important point about science is the infrastructure or the way we think about answering questions, so we can divide that methodology into various ways. So, for example, I might have a question about why my patients on immunotherapy who respond but then get what we call immunosuppression for side effects, why do some of them not stay in response because of the immunosuppression, and some do? The first point is, what sort of methodology do I need? Is it a retrospective study where I look back and analyse the data, or is this something I can explore prospectively? So let me talk about the frameworks that you do need to understand within your curriculum.

A primary analysis – that means you are taking direct observations and analysing them. For example, I've observed in my clinic that maybe patients who have immunotherapy and get immunosuppression, some of them may progress later rather than staying cured. That's my observation. The primary analysis of that would be collecting data on a group of patients who have responded and dividing the group into those who got immunosuppression and those who didn't, and then how many did not stay in response, for example.

That's a primary analysis. That's looking at the data to try and understand the question I asked or my hypothesis.

Another type is secondary analysis. So now I've done my study, and I looked at my patients who'd responded. Some got immunosuppression, some didn't, and of the ones that got immunosuppression, a third progressed. Of the ones that didn't, no one progressed, let's say.

A secondary analysis would be someone saying, now I've seen Georgina and her team’s study, and then I've seen that the French have done one, and the Italians have done one, and the Americans have done one. What if I just grouped them all together so I'm not doing the initial work. I'm not raising the hypothesis. I'm looking at a group of studies that have tried to answer the one question to see if I can get more power, to get a more accurate answer to that question. So that's called a secondary analysis. Sometimes we do meta analyses, for example, in clinical trials where I might have a clinical trial of drug X.

Patients with cancer and they're randomised to the standard drug A or the new drug X. And then we've done a similar trial with similar types of drugs, with similar mechanisms in another point in time. A meta-analysis would be taking all of those patients together and trying to find out whether the new drug X actually improves survival in those patients. That's again, another example of a secondary analysis.

The other type of studies would be epidemiological studies. And that's where we are observing a health related problem. And trying to find the associations. Associations does not mean cause. That is really important in science. Just because I noticed that when COVID came along and everyone got vaccines, let's say, I noticed that there was some abnormality in brain function in a certain population. Let's say there was a health condition that happened, and so it's associated with vaccinating everybody. Does that mean that caused it? No, because there's another association – getting COVID itself. So how do you tease out whether you see this phenomenon, for example, let's call it brain fog or some other health related thing, how do we know it was a vaccination or actually getting COVID? These are called associations. They are not necessarily the cause. So epidemiological studies are just trying to make associations to try and understand that health related problem, but it does not necessarily make it the cause. Let me give you another example.

Let's say there was a family that had a genetic problem that was quite rare and they lived in a certain area where there were electric towers. And then over the course of a hundred years, the genetic problem got expanded and we started to see this disease and they all lived around the same area because they were a big family.

We could say, oh, this disease is due to the wires and the towers. What if it was actually genetic? And it was because of the family that hung around together and lived together. So these are associations that may not be causative. So epidemiological studies try to look at what's associated with the health condition to try to better understand that health condition.

[End of transcript]

Professor Georgina Long AO

So then we have clinical trials. That's another type of study. In fact, clinical trials are a type of primary analysis. And then a meta-analysis is where I get many clinical trials. So some of these things sit under each other, but a clinical trial is where we are testing a drug, for example, a method, a technique, an intervention of some kind.

And we try to have a control arm. Not all clinical trials have a control arm, but basically we are testing an intervention in humans. Within clinical trials, we have phase one, phase two, and phase three, and then we have post-marketing once we know something works. Phase one are first in humans trials. This is now talking particularly about drugs.

Phase two is where you might see a drug has a bit of a signal and it's tolerated, and we expand it in a single arm of a trial. So the drug X, it's tolerated. Phase two tells me whether it's got a bit of a signal of working in the cancer. Phase three is then a randomised trial where I test it against the current standard to see if it's something we should adopt.

So that's phase one, two, and three within clinical trials for drug therapies, you may have an intervention, let's say a psychological intervention. Let's say my melanoma patients are feeling distress about having melanoma and the fear of recurrence. Would a psychological intervention given every 3 months for the first 12 months help with that fear of recurrence?

So in that case, you would first do a pilot trial with the intervention to see if you see a change. Then you may do a randomised trial where you randomised people to the intervention, a psychological intervention versus not, and you then measure whether it's made any psychological difference to their quality of life.

So that's another type of clinical trial. And then lastly, we have blinded trials. So you can blind just the recipient, so the patient, for example, in a clinical trial. So they don't know what drug they're getting, they don't know if it's drug X or Y.

You can blind both the patient and the investigator. So me as the doctor who's treating a patient on a trial, I'm blind and they are blind. That's called double-blinded. They are common. I do a lot of double blinded studies. I don't know what treatment they're on, they don't know. It is ethical because we know they're at least getting the standard or a new drug, which has some good data behind it.

And then there's triple-blinded where the patient doesn't know what they're getting, drug X or Y, I don't know if they're getting drug X or Y, and the person analysing the data at the end of it does not know whether patient gets X or Y.

These terms epidemiology or primary analysis, secondary analysis, clinical trials, single-blinded, double-blinded, triple-blinded. They all sit within each other. You can have a trial that's a trial and a primary analysis and is double-blinded.

So using these terms, one study might come under several headings.

[End of transcript]

Professor Georgina Long AO

Now comes the question of, you've done your work as a researcher and you need to let the world know about it. You need to communicate your science results. How do we go about doing that?

So there's many different ways to do that. For example, you can present it at conferences where you write what we call an abstract, a summary. You submit it to the conference. If they accept it, you then go and present the full results. That's one way of getting your work out there. That's considered one level.

It's not peer reviewed, so it hasn't been reviewed yet, although you're showing it to the world and you'll get commentary about it, negative or positive that might help you then write the paper that you then publish. So where do we put it?

So original research papers would be that I am reporting the results from my primary analysis or a secondary analysis, and it is original work. So that would be published in a journal. It gets peer reviewed to poke holes in it, critically analysed, that is important in science that we allow our work to be critically analysed. So that's an original article. I might be an expert, so I'm an expert in immunotherapy and particularly melanoma. I've got two things I really want the world to know. It's from my experience as a researcher, having published a lot and from treating a lot of patients.

Number one, I want the world to know that you can reduce toxicity using immunotherapy. That's side effects. If you dose it in a different way to the way we think of chemotherapy, a different type of drug therapy. So I could write a commentary. So that's not me presenting my original research, that's me making a comment about my research I've done to date and my ideas as an expert, where the field needs to go.

That's a commentary, that's an opinion piece. It does go through peer review, but in a lesser way. It's more that it's sensible and not, not saying things that are going to take the scientific world off down a rabbit hole that's not appropriate. 'Treat everyone with cockatoo poo, it's going to work' like that's just not appropriate. It's not backed by science. So my opinion piece is based on my experience as a key opinion leader, and it's based on my research. And I have this opinion that we need to talk about so that we keep patients safe. So that's an opinion piece.

An editorial is where you're invited to make a comment about other people's work. Someone publishes a paper, and let's say it's on melanoma and immunotherapy, and I'm not involved in the research. They would ask me to do an editorial or if I'm an editor of a journal, and I am an editor of some journals, I might write a piece to contextualise that research, that original research that's different.

So commentary is about how I want the field to move forward, an editorial as an editor is to contextualise where this new data, this new original research fits in, in the world. So there are just some examples of research papers, but if you are presenting original results from original primary analyses or secondary analyses they will in general be an original paper.

[End of transcript]

Professor Georgina Long AO

Peer review is where other opinion leaders in your field make comment about the scientific method and the question that's being answered. Everything should be peer reviewed that is published in a scientific journal. We also use peer review for grant applications. So, for example, my program of research is underpinned by many grants, some of which are what we call NHMRC, the National Health and Medical Research Council, a very important body that funds a lot of medical research in this country.

And so when I put those grant applications in, they undergo peer review. It's a competitive process. In fact, this year for what we call NHMRC, investigator grants a very important grant. The rate of success was less than 20%. In fact, it was around the 13% mark. So every one of the a hundred percent that applied, only 13% got a grant. So you can see it's very competitive.

So it goes through peer review and the best grants get allotted funding. Is the question important? Is the method good? And will this person deliver? Do they have a track record of delivering? It's because when we make that investment in research, we want to see something for Australia.

[End of transcript]

Professor Georgina Long AO

Melanoma is my area of research. I'm a medical oncologist. I treat melanoma with drug therapies. When we are doing research, understanding the cause of a disease or a cancer can help us eradicate it. In Australia, the most common type of melanoma, is skin melanoma. There's increasing glamorisation of tanning even in China now in the younger population, so they're seeing a bit more of skin melanoma now, unfortunately.

But there are also some rare subtypes, mucosal, melanoma, melanoma that arises from the inside of your body, not due to the sun. We don't understand the cause of that, but they're very, very rare. We think there could be a viral element to the cause.

So back to Australia. Over 90% of melanomas which are on the skin are due to directly UV and sun exposure, tanning and sunburns. The lighter the skin you have, the higher the risk of getting melanoma is. It's multifactorial: many things play in your risk of melanoma, but fair skin. If you're dark skin, you still can get it though.

And we need to be sun safe.

[End of transcrpt]

Professor Georgina Long AO

The risk factors for melanoma, let's talk about Australia, are going to be UV exposure, particularly in your younger years. But the good news is that any age you start being sun safe, it will protect you from future melanoma. So even if you're 50 or 60 start being sun safe, it will reduce your chance of a melanoma down the track.

But our most important years are our younger years. Basically from zero to 30 we have to be really careful. But beyond that as well, What is sun safety? Sun safety is hats, sunglasses, clothing is really important. Clothing is your best protection. Shade, and then sunscreen for the bits of skin you can't cover up.

But if you can use clothing, you're better off. So what are the risk factors of melanoma besides UV? Fairer skin although when you're dark, you still can get it. Genetics. So if you have a strong family history of melanoma, so your parents or your siblings have had melanoma, the number of moles you have. So people who have more moles are at higher risk.

There's a syndrome called dysplastic nevi syndrome. Those people are at higher risk. And the older you are, the higher risk of melanoma because you've lived a longer time. But they are the main risk factors.

[End of transcript]

Professor Georgina Long AO

Immunotherapy. Where did this come from? How did we develop immunotherapy for cancer?

First things is: nothing happens like that [snaps fingers], it happens over decades, incremental improvements in understandings. Then you have your eureka, often not always a eureka moment, but it's actually built. It doesn't happen overnight.

So for example, in cancer, we've always known that cancer has a very unique relationship with the immune system, particularly melanoma. Long before we had any effective drug therapies for stage four melanoma. We did see the rare case of what we call spontaneous remission. I have a couple of patients where this happened.

It just went away on its own. That's the immune system eradicating it. One minute they had melanoma in the lung and their lymph nodes. Next thing, over the course of about three months, it just disappeared. So we knew that there was a relationship with the immune system and we just couldn't work out how to leverage it. In fact, in the 1800s we knew that certain infections, they would stir up some bacterial infections called Colley's toxins and inject it into patients and occasionally they'd have an anti-cancer effect and the tumours would shrink. So they knew there was something about the immune system.

We used to, and we still do, actually take immune cells from someone's body, so we'd cut out the cancer, extract the immune cells, and reinject them into the patient. And in that case, those immune cells, we'd see about 1 in 10 patients responding. So this is over decades and then around the late, naughties, let's say around the 2007, 8, 9 time, based on work that was done in the nineties.

So scientists had discovered that these immune cells called T cells had these checkpoints on them because you don't want your immune system to get overactive, otherwise it'll gobble up yourself and hurt you. So you need your T cells to fight the right enemies and not fight yourself.

[Graphic on screen: Suppression of anti-tumour activity

Shows a tumour cell with MHC, PD-L1 and PD-L2 receptors interacting with a T cell with PD-1 receptors.

PD-1 receptors are located on T cells. When ligands bind to these receptors, they trigger a process that leads to T cells undergoing a form of cell death called apoptosis, or programmed cell death, This helps to reduce immune responses, especially in cases of immunological tolerance.]

And to do that, they're very controlled these T cells, they have these checkpoints and these checkpoints on the outside mean that the T cells switch off. So if they come across your thyroid or they come across your lungs, they recognise lungs and thyroid tissue as self, these checkpoints are engaged and the T cells switch off, and that's what you want: a beautifully well controlled system.

Now, we do want those T cells not to be switched off by cancer.

[Graphic on screen: Suppression of anti-tumour activity

Shows a tumour cell with MHC, PD-L1 and PD-L2 receptors interacting with a T cell with PD-1 receptors.

This tumour cell has PD-LI and PD-L2 proteins on its surface. When these proteins bind to PD-1 receptors on T cells, the T cells die. This suppresses the immune response against the tumour.]

So what happens is cancer engages these checkpoints on T cells as a handshake and that switches T cells off, so they go away and don't kill the cancer. The immunotherapy that made a massive difference and started really impacting lots of patients were these checkpoint inhibitors, so they disrupted that handshake and allowed the T cells to then kill the cancer.

[Graphic on screen: Immunotherapy – tumour destruction

Shows a tumour cell with MHC, PD-L1 and PD-L2 receptors interacting with a T cell with PD-1 receptors and Pembrolizumab antibodies

Pembrolizumab is a small antibody that attaches to the PD-1 protein. This attachment does not lead to T cell death because it doesn't trigger the pathways that cause it. Instead. pembrolizumab prevents PD-L1 and PD-L2 from binding to PD-1.]

We have checkpoints that work at different points, so there's one checkpoint called PD-1, which works down at the cancer, where the T cells going in to kill the cancer comes up, stops the T cell, switches it off.

[Animation shows T cell with PD-1 interacting with and killing a tumour cell with PD-L1.]

We interrupted that, T cells kills the cancer: bingo. We also have checkpoints at the very beginning of the immune cycle, so this is when your immune cells, called antigen presenting cells, on your skin, for example, see a little bit of melanoma cancer, they gobble it up and they present the bit of melanoma to your T cells, and then the T cells go, oh, that's abnormal, that's abnormal.

[Graphic on screen: Suppression of anti-tumour activity

Shows a tumour cell with MHC, PD-L1 and PD-L2 receptors interacting with a T cell with PD-1 receptors.

This tumour cell has the MHC1 protein on its surface. which binds to a tumour-associated antigen. When this binding occurs, it activates the T cell. leading to the destruction of the tumour cell.]

They get activated, but then they get switched off. So we also have another checkpoint here where the T cells are going, Ooh, that's abnormal, that makes sure that the T cell says that is abnormal. So using these checkpoints together, we get this enhanced tumour kill.

When we say immunotherapy, it's a big term. There's lots of different types of drugs. I have only given you one set of immunotherapies and that's called checkpoint inhibitors. They have been the eureka moment for cancer.

But we are now building on that with many other immunotherapies. This whole area of science is built on decades of work that happened before checkpoints and drugs inhibiting checkpoints came about.

[End of transcript]

Professor Georgina Long AO

Another area of immunotherapy is CAR -T cells. So I've just finished saying there's immunotherapy in cancer. So immune stimulants, right? They stimulate the immune system to kill cancer, and checkpoint inhibitors are one type. There's also CAR -T cells. There's also genetically modified T cells. There's also cytokines, there's vaccines against cancer. All of this is immunotherapy.

Checkpoints have changed the way we treat cancer almost overnight, again, built on lots of science. CAR -T cells are where we take a patient's T cells from their blood, so they undergo something called plasmapheresis. We collect all their T cells, and then they are modified to have a receptor put into them that then brings them to the cancer.

[Animation showing a modified T cell interacting with a tumour cell. The tumour cell is destroyed.]

They have been successful in what we call cancers that have a single genetic or a single common thing we can tag the CAR -T cells to and attract to. Regrettably things like melanoma or breast cancer or colon cancer or lung cancer, they are so mixed up and heterogeneous that CAR -T cells can't kill all the cancer enough because they're so heterogeneous. You really need one framework for the CAR -T cell to connect to. However we're getting around that and so now we are looking at modifying CAR -T cells, sometimes put two or three receptors in them, or modify them genetically, even including using CRISPR techniques to then bring other T cells into the more heterogeneous cancers like lung and colon and melanoma.

So watch this space. CAR -T cells really started in single genetically modified cancers, which have one target, and we're now altering them to try and make them more appropriate for what we call solid tumours that have multiple different targets and are very heterogeneous and often find their way around CAR -T cells so it can develop resistance. CRISPR is just a technique to insert genes or to delete genes out of a cell.

[On screen: diagram showing the CRISPR-edited T cells process.

1. Remove blood from patient to get T cells.

2. Make CRIPSR-edited T cells in the lab

- insert gene for NY-ESO-1 receptor

- delete 3 genes with CRISPR (PDC1, TRAC and TRBC genes)

3. Grow millions of CRISPR-eidted T cells

4. Infuse CRISPR-edited T cells into patient

5. CRISPR-edited T cells bind to cancer cells and kill them.]

We use CRISPR techniques for all sorts of things in cancer research, particularly in our mouse models, to test mechanisms of resistance, but it's also used in genetic diseases, whether we can insert new genes or delete genes to eradicate a genetic disease.

[End of transcript]