Transcript

MIC CAVAZZINI:               Welcome to Pomegranate Health, a podcast about the culture of medicine. I’m Mic Cavazzini, for the Royal Australasian College of Physicians.

Australia and New Zealand-Aotearoa have the dubious honour of ranking 1st and 2nd in the world for melanoma incidence. The incidence in Australia has actually doubled since the 1980s to around 54 cases per 100,000 population. Melanoma isn’t the most common cancer but it is the most deadly as measured by loss of disease‐adjusted life years because it affects so many people in their prime.

According to US data aggregated between 2005 and 2011 16% of melanomas are diagnosed at the metastatic stage, and five year survival of these patients is under 17%. This includes those with proximal lymph node metastases that could be removed. But for those with unresectable metastases of stage IIID or stage IV disease the bleak choice had for a long time been between the systemic chaemotherapy Dacarbazine or palliative care.

That all changed with a Phase 3 trial of the immunotherapy ipilimumab published in 2011. Three years from diagnosis, the survival rate for the group receiving Dacarbazine alone was 12% while it was almost 21% for those also receiving ipilimumab. This was the first improvement in overall survival of such cases in three decades, and most remarkable was the virtual plateau of survival curves at later follow ups.

Within a few years of this landmark, results were published from trials of two more immunotherapeutics called pembrolizumab and nivolumab. The advantage of nivolumab over chaemotherapy was so marked, that the 2014 Checkmate-066 trial was stopped early and control subjects were allowed to cross over to the novel treatment. That was the last time Dacarbazine was used as a comparator. In the next offering from the Checkmate group 44% of patients receiving nivolumab survived to five years, and in combination with ipilimumab it was an astounding 52%. That’s almost five times as many survivors as would have been predicted just a few years earlier.

We’ll get into the action of immune-checkpoint inhibitors later and the sometimes serious toxicities that come as a trade-off. For oncologists today, there are profound challenges in managing expectations of patients about survival and quality of life. This is even more true as these therapies are rolled out to prevent recurrence of resected cancers, though government-funding might not always be available to cover the very high costs. To guide me through all this I spoke to Associate Professor Matt Carlino and I started by asking him whether there was any precedent in oncology for the impact that immune checkpoint inhibitors had made.

MATT CARLINO:               Hi, my name is Matt Carlino, I'm a medical oncologist. I work at Westmead and Blacktown hospital and I'm a faculty member of the Melanoma Institute Australia. The development of checkpoint inhibitors, led in melanoma but now in multiple cancers, has really been a large improvement in care. And there are few other examples where the care of a disease has changed so rapidly. Both in terms of, yes, these numbers and improvement in survival. But also the mindset of the treaters and the patient. So those patients we saw were being treated with Dacarbazine were getting palliative treatment. They'd see us, we'd say, “You've unfortunately got incurable cancer. We have a chaemotherapy, that's not very active, and unfortunately, even if the chaemotherapy does work, its chance of working for more than a few months is very slim.”

Now, if you look at that, Checkmate-067 study which was a landmark study, about 50% of patients are alive at five years. And I've got patients treated with checkpoint inhibitors now 12 years out from therapy. So I truly believe a proportion of patients with advanced melanoma are cured with immunotherapy. And so I see this treatment as having the potential for cure—unfortunately, not the guarantee of cure—I think, clearly 50% is not good enough—it's better than 5% which is what the historical number would but there's still room to move.

MIC CAVAZZINI:               You've answered my next question. Most, most of the attrition in responders now is from causes other than melanoma. And if you filter those out that the survival curves have basically plateaued. Do you tell patients that if they've made it to three years or five years that they are essentially cured?

MATT CARLINO:               Yeah, so I tell them that that as we get further away from those timelines—and I think three and five years are really useful landmarks—that their chance of their melanoma coming back is incredibly slim. And I tell them I want them to think about their life with that risk being low; you know, approaching zero. And that's important for decisions you make. So as an example, I make sure that all these patients are seeing a dermatologist and getting skin checks. Because just because we've controlled the metastatic melanoma, there is a chance of a new primary melanoma. If this patient is a woman in her sixties or seventies she should be getting mammograms. If that person has diabetes, I want them managing their diabetes effectively because I don't want them dying of a heart attack prematurely, and so those survivorship issues come to the fore. Whereas the honest truth is, if I met someone with melanoma ten years ago, and they were starting Dacarbazine, I wasn't telling them to get mammograms to exclude subsequent breast cancer because their prognosis was so poor, that that the melanoma dictated their survival.

MIC CAVAZZINI:               Let’s explain how these immune checkpoint inhibitors actually work. Melanomas are vulnerable to such drugs because they’re highly immunogenic to begin with; at least 2000 melanoma-specific antibodies have been identified in patient serum. It all starts with dendritic cells that pick up tumour epitopes and migrate to the lymph nodes where they present them to T lymphocytes. Activated T-cells go on to proliferate and attack tumour cells directly or mount a larger adaptive immune response. But they also have a hand-brake in the form of a membrane receptor called CTLA-4. This immune-checkpoint stops the cytotoxic response from getting overheated, and it’s also found on regulatory T-cells which release immune suppressive cytokines.

Ipilimumab is a monoclonal antibody to CTLA-4 that blocks its activation by dendritic cells and releases the handbrake on the immune response. But there is a second immune checkpoint further downstream, at the point where cytotoxic T-cells are locking in on a tumour. To avoid becoming collateral damage, healthy tissues bind the programmed cell death receptor PD-1 expressed on the T cell surface. Melanomas have figured this out too and they express a PD-1 ligand that gives them some degree of immune privileged. Nivolumab and pembrolizumab are two inhibitors of PD-1 that counteract melanoma’s subterfuge. An unfortunate side-effect of all the therapies described is that they expose native tissues to auto-immune attack. I asked Matt Carlino to paint a picture of what the immediate physiological response to immune checkpoint inhibitors looks like, as well as the related toxicities.

MATT CARLINO:               So the question becomes, at a tumour level, what does an immune response look like? As part of Melanoma Institute, and other investigators around the world, we've had patients who have kindly agreed to have biopsies before starting treatment and biopsies after starting treatment, and then even when their tumours grow. So one of the things you see in a responding patient very early in treatment is an immune infiltrate. So tumours often have T-cells in them, right, but you get an expansion of these T-cells. When you get more sensitive assays, so looking at the T-cell receptor on these T-cells in the blood and in the tumour, these T-cells are specific to the tumour. So you're getting this expansion of an immune response against the tumour. There are subtle differences in how this looks from ipilimumab to PD-1 alone and the combination. One practical thing is ipilimumab more likely expands the repertoire of T-cells, you know, the diversity of T-cells against that tumour, targeting multiple more antigens than PD-1. And that to me kind of makes sense both in how the drug works

MIC CAVAZZINI:               It’s further up the pathway isn’t it?

MATT CARLINO:               But it also makes sense why ipi has more side effects because you're broadening your repertoire of good, but you're also broadening your repertoire of bad in terms of tox. So the key is that expansion of T-cells. One of the things I explain to patients is this phenomenon called pseudoprogression. That occurs in a few percent of patients radiologically. You actually see patients who have lumps or subcutaneous spots, who actually can say, “After I started therapy, my lesions felt a bit angrier, you know, I thought they were growing,” and certainly patients get quite frightful if they've got lumps under their skin that suddenly get bigger when you start a drug. And the way I describe it is I say, “Look, imagine you've got a cup of melanoma, and it's full. If you pour T-cells into that cup, the cup overflows. And so that's why you get this very rapid change.”

MIC CAVAZZINI:               But unfortunately, as well as disabling melanomas defences, the monoclonal therapies also make healthy tissues vulnerable to attack from self-reactive T-cells. Around 90% of patients undergoing ICI therapy experience an immune-related adverse event of any severity. These might be itchy rashes and blisters, loss of appetite and diarrhea, and fatigue associated with hypothyroidism and adrenal insufficiency. Can you briefly describe the typical pattern of emergence and why some tissues show up before others.

MATT CARLINO:               So, you've got to be a little bit cautious when you interpret numbers; 90%, 60%, whatever. I always find it amusing when we look at a placebo-controlled trial and you see placebo has a 30% toxicity rate, even grade 3 and 4 toxicities in placebo. And so some of our nonspecific side effects are actually related to the cancer or life. You know, if I was on a trial, even if I'm on the placebo arm, and I get infective gastroenteritis, that gets coded as diarrhea. So the numbers need to be interpreted with caution.

But you're right, checkpoint inhibitors have this group of toxicities which are termed immune-related adverse events or immune-driven toxicities. How I describe them is, “toxicity is us boosting your immune system and attacking normal tissue versus just the cancer.” There are toxicities that are common to both ipi and the PD-1 agents. There are toxicities that are more common with the PD-1 agents alone. So thyroid toxicity is really common with the anti PD-1 agents. Ten percent of patients will develop hypothyroidism. If you include the patients that get transient hyperthyroidism or subclinical hyperthyroidism, often due to a thyroiditis, that number is higher. Hypophysitis—so inflammation of the pituitary gland leading to cortisol deficiency—is more common with ipi, so clearly there must be a predilection for how these drugs work for different side effects.

MIC CAVAZZINI:               But it's not as simple as saying the homeostatic balance of the skin tissue is more sensitive than the gut and then then you get…

MATT CARLINO:               Yeah look, it could be. I think there's a mix of things. So there must be an importance of PD-1 to the thyroid. Some of the toxicities are clearly cross-reactive antigens. So the best example of that is vitiligo. You don't see vitiligo in patients with lung cancer, or kidney cancer getting PD-1. But in melanoma, it's really common; if you come to our melanoma clinic, you will see five patients in a clinic who have vitiligo. And that makes sense. What's vitiligo? It's an immune reaction against melanocytes. Well, melanoma comes from melanocytes, so it's clearly a cross-reactive antigen. Interestingly, vitiligo is a very strong prognostic feature. If a patient develops vitiligo, it is associated with a good outcome. And that also makes sense to me.

MIC CAVAZZINI:               You can manage more severe symptoms, immune-related adverse events, with immunosuppressive drugs. Should we talk about some of the broad-brush options and some of the more targeted ones?

MATT CARLINO:               Yeah, let's go there. So, so managing immune toxicities is really a large part of that a medical oncologist’s job now. The first step in managing an immune toxicity is identification. So step one is you need to know about it. I don't know if they've got colitis, unless I know they've got diarrhea. The patient who has diarrhea at home, they need to know to communicate, be that calling me, calling a nurse I work with et cetera. The next question is, is this toxicity immune-driven, or is it unrelated? So using diarrhea as an example again, if that patient tells me when on the phone that both of their preschool aged children also have diarrhea, the fair bet is they've got viral gastro, right? And so that that is a key part of identification and management; Iis this immune-driven?

The next step in the management of immune-driven toxicities for most of them is immunosuppression, usually using steroids. So we use prednisone to, if you like, dampen down that immune response. And the vast majority of immune-related toxicities are managed with prednisone. Relatively high doses of prednisone compared to what many other doctors are used to using. So even a relatively mild colitis, someone who I'm managing as an outpatient with diarrhea—once I've confirmed or think they're very likely to have immune-driven colitis—we would start them on one milligram per kilogram of prednisone. You know, a standard size man or woman that's 75, maybe 100 milligrams a day—that's a pretty solid dose of pred. People who are sicker, you know, in hospital may even get a higher dose of intravenous steroids. So that's our first step of managing immune-driven toxicities is steroids. That works for the vast majority of patients. We then slowly wean them off the prednisone. So you don't get them better with prednisone in two days and stop it. You often have to wean them off prednisone over three or four weeks, maybe even longer.

MIC CAVAZZINI:               And are you consulting a lot with the endocrinologist and dermatologists and other specialists, or it's pretty standard practice now?

MATT CARLINO:               Yes and no. So I can't do a colonoscopy, so if I need a colonoscopy as part of the management I need to a gastroenterologist. So use your friends, but sometimes your friends will equate this to the disease it mimics. This is not the same as ulcerative colitis or Crohn's, it is different. This hopefully is a one-episode thing we get over. So I have a few go-to friends in each specialty, but don't handoff your toxicity to someone else to manage. I think the medical oncologist needs to drive the management of the toxicity.

MIC CAVAZZINI:               Yeah, and what's the distribution in patient responses? So we're talking 50% of patients respond to the checkpoint inhibitors. How many of them have tolerable effects? How many need immunosuppressants? And how many have to cease therapy because of intolerable toxicity?

MATT CARLINO:               Yeah, so, so best to start thinking about the single agent, anti PD-1, which is the most-tolerated of the immune therapies, and also the most active as a single agent. And so when I tell patients I say, look, a group of patients will get no side effects. You know, let's say that somewhere in the order of 50%, acknowledging that there's nonspecific side effects where you're not sure what's causing them. There's another group of patients, they're going to get minor side effects. You know, so the person who gets a rash that I give them a topical steroid and it goes away. You know, no big issue, you get an itch, I give you an antihistamine, it goes away. The person who has transient subclinical hyperthyroidism that goes away. That's another group of patients, let's say that's 20-30%. And I put the hypothyroidism in that minor group, even though it does need an intervention. But if you've got advanced cancer and your cancer gets controlled, and you need thyroxin, that's not a bad deal. It's enough to justify as a minor toxicity.

There's probably 15% of patients who get a toxicity that I need to actively manage with immunosuppression and so that that's a hodgepodge of things, be it joints, be it colitis, be it hypophysitis, and all those other toxicities. Of those, some need an inpatient stay to manage their toxicity, and most of that 15% probably need to stop treatment. The important thing to say is, unlike chaemotherapy, you stop for a toxicity, doesn't mean your cancer will come back or progress.

So if you look at those trials you mentioned Checkmate-067, and the pembrolizumab trials, Keynote-001,006, the vast majority of those patients are off treatment. In fact at Westmead there are many, many patients alive from those trials, none of them on therapy. When we first ran the trials, it was ongoing treatment. The wording was, “until disease progression, or unacceptable toxicity,” which is a long time. We then started doing trials to two years, but even two years may be too long. And so there are trials going on now of early cessation of immunotherapy. and I have patients who had one dose of therapy, got side effects and are fine years later.

MIC CAVAZZINI:               For listeners interested in more of the tit-for-tat subtleties of managing immune-related adverse events, there are a couple of detailed case studies in a review for the Journal of the Advanced Practitioner in Oncology from the Sloane Kettering Cancer Centre. They show, for example, how you might be able to reduce the burden of side effects by withholding treatment for a bit, or alternating delivery of two different agents.

Some melanomas can be also be controlled with drugs targeted at an enzyme called B-RAF kinase. About 40% of melanomas express a mutant that leads to uncontrolled cell proliferation. A 2010 publication by researchers at the Peter MacCallum Cancer Institute contained a breathtaking demonstration of targeted inhibitors using whole body FDG-PET scans. While the baseline scan was riddled with metabolically-active metastases, 15 days after the B-RAF kinase blocker was initiated there was barely a signal to be seen.

A similar compound called dabrafenib, registered in 2013, is more effective still. In combination with another blocker of the same pathway called trametinib, one year survival rates of patients with non-resectable BRAF-mutant melanoma have been put at 71%. Unfortunately, there’s a rapid drop off and in five year data published in 2019 only 34% of the original cohort were still alive.

Immune checkpoint inhibitors don’t have quite the same problem of acquired resistance, but unfortunately there are no known genetic or molecular markers that would predict suitability for treatment. Such markers could help improve the effectiveness of ICIs moving forward, but they’d also identify non-responders before they were subjected to the side-effects and false hope of immunotherapy. Focusing on chaemo or palliative care would be better for these patients and also spare healthcare resources for others more likely to benefit.

While there aren’t yet any genetic flags for response to ICI therapy, there are other ways to help individualize prognoses for patients with advanced melanoma. Those who start out with a robust infiltrate of T-cells in tumour biopsies are more likely to have a strong immune response to checkpoint inhibitors. Density of PD-1 ligands in the tumour and interferon-gamma in serum have also been associated with a better response rate. Conversely, high serum levels of lactate dehydrogenase and poor ECOG performance status are correlated with a much worse response to both ICIs and targeted therapies. I asked Associate Professor Carlino which of these were most useful for guiding prognosis.

MATT CARLINO:               So what I use most in practice is the clinical ones, right? Good ECOG. You know, what does that mean? The person that you see with asymptomatic cancer that was picked up incidentally on a scan; I factor that into my clinical decision-making. M1B disease, melanoma that's limited to the lung; much higher chance of response. You know, desmoplastic primaries—the old Australian bloke who had a melanoma that started and his scalp—more likely to respond than a melanoma that hadn't had a huge amount of sun exposure. Clearly, the person who's ECOG2 doesn't probably have that 40% response rate. Those clinical biomarkers of response are used daily, consciously and subconsciously. They're the things I'm using more than asking the lab to do PD-L1 or—we can't get an interferon gamma signature in a clinically meaningful time, and my sense is, even if I had that, I'd still be using those clinical factors. Interferon gamma signatures do not trump ECOG.

MIC CAVAZZINI:               So I'm interested to hear how you discuss the risks and benefits given this sort of hodgepodge of reactions. Annie Wong is an oncologist at Peter MaCallum Cancer Centre who wrote a paper titled “Balancing the Hype with Reality.” And there she concludes that, “Only a minority of patients expressed a good understanding of the likely efficacy and potential treatment side effects.” Palliative care dual specialist Molly Williams made similar comments in a in a panel discussion with you. She said that trial participants always think they are going to have the miracle response, so it then comes a terrible shock if their condition starts to deteriorate. How do you map out the possible trajectories and expectations for patients that are starting immunotherapy?

MATT CARLINO:               Yeah, so the first thing is you've got to in your head, get a feel for the likelihood of benefit and the likelihood of risk. So when you've got patients with advanced disease, they are essentially going to die of their disease unless this immune therapy works. I'm then asking myself the question, what's their likelihood of response to PD-1 alone, and also to ipi-nivo. Right, and I'm doing this in my head. So I think medical oncologists with a subspecialty interest in melanoma can do those that pretty individualized. So let's say based on everything I know about a patient, their ECOG, their burden of disease, I predict their chance of response to ipi-nivo is 20%. But 20% response to ipi-nivo, or to pembro, is not the same as 20% response to chaemo. Response means they have a chance of a durable response, and I’d argue that many patients would want an immune therapy and the benefit of that immune therapy response with a chance of 20%. Now, it's hard to give an exact number and actually, exact numbers are less important for an individual. For the individual, you don't have a 20% response; you respond or you don't.

The other side of the equation is, is your risk of tox or the potential sequelae of toxicity, higher or lower than normal? So, I've recently treated a woman who's got a kidney transplant with immune therapy. Her risk of loss of that kidney graft—is significant, perhaps 50%. The risk-benefit for that patient is very different to a standard patient, so you can't use a one size fits all on the chance of benefit; chance of harm. Another example, you see a patient who's in their eighties, lives at home, but needing some assistance from their family. If that patient gets a bad immune tox, in hospital for two weeks, deconditions, can't get home and be independent. So you've got to individualize the benefit side of the equation in my mind, and the risk side of the equation. Reporting numbers from trials, I think, doesn't serve patients as individuals well. We need to interpret that data and individualize it to the patient.

MIC CAVAZZINI:               There was even as a 2013 paper from the Tufts Center for the Evaluation of Value and Risk in Health. It showed that patients were prepared to pay significantly more for a hypothetical cancer drug if it was presented as landmark survival compared to the same stats presented as median survival. So you have to really make sure, yeah, you're communicating what you're trying to….

MATT CARLINO:               Yeah so, I don't like medians, particularly for drugs with plateaus of the curve. So medians work well, for that normal distribution of everyone unfortunately dies and we can—but the survival curves for the immunotherapies are not like that. I actually think the best way to think about immunotherapy, and the way I talk about it, is landmarks. And patients intuitively can think of a landmark; your chance of being progression-free or chance of being alive at one year, two years, three years, five years. You know, and when I say to a patient, “What’s your goal?” People’s goals are, “I want to be alive for my child to start school. I want to be alive for my child to finish school. They are survival events, that’s what patients are after.

Probably one of the hardest things in immune therapy is, you get a group of patients who do incredibly well. But you don't know who they are at the beginning. And so often I say to people, “Look, we hope you're going to be one of the people who's going to do incredibly well with these drugs. But we don't know now.” The best time to revisit that is that that first restaging. So you've given someone three months of drug they've come in and this scan shows a really deep response. You can alter your discussion. “Remember when I said the chance of you responding was 40%? The overall survival at five years was 50%. That’s when people enter the trial. We can we can now individualize that information. I can say, “Based on your scan, showing your complete response, I think your chances of being alive at five years are x.” I think it's important to embrace that uncertainty and say this is an ongoing conversation.

The other thing that is really hard, is every oncologist who does this a bit has a group of patients who are the exceptions to the rule. A couple stick in my head; one is a lovely bloke, he got one dose of ipi-nivo, came in with a bowel obstruction from malignant disease, so his abdomen was full of cancer. We spoke to the surgeons and it was clear that an operation wouldn't help him. You know, if they'd opened him up it would have been an open-shut operation. We had discussions about how awful life was, you know, his prognosis we felt was measured in weeks. His family and everyone started to come to terms with this. He actually went to the palliative care unit. You know—he's alive. He had a delayed response to that immune therapy as alive years later. I have other patients who had brain metastases and were rapidly progressing and dropped their ECOG to 3, touching 4s, who are alive.

Those exceptions—I have a handful of them—I can picture all their faces and I still see these people. Unfortunately, I've had many, many more patients die of melanoma. So you can't say that there's no chance. But you do a patient a disservice if you say, “You're rapidly progressing, your ECOG is falling. This drug’s got a chance of turning you around.” You need to be honest and say, “The most likely outcome is poor. But there is a slim chance,” and that is really hard to discuss. You know, patients can get angry because you know, “You gave up on me and sent me to pall care, and now I'm fine.”

And so a few things I've altered in my practice around that is —I link people to pall care, I engage pall care, we transfer patients to pall care if that's needed for their care. I don't discharge patients from my clinic. So I often arrange a follow up phone call. So yesterday I saw a gentleman who was in a state like that; is rapidly progressing; we’ve stopped treatment, he's elected to stop treatment. My follow up plan for him is, “I'm going to call you in two weeks, just to check in.” If I ring his wife and she says, “Look, Joe Bloggs is feeling better, he's looking better, he's starting to potter around the house” –he's the one in fifty or one in a hundred who have a delayed response—we regroup. Now, I'd love that to be true. I don't think it'll be true, but there's that slim chance. And so I think if I had just discharged him and say, “Go to pall care,” I think our the therapeutic relationship doesn't work as well for me and I think for the patient. So, it's a little bit of pray or hope, depending on your belief structure, for the one in the hundreds but acknowledge they're rare.

MIC CAVAZZINI:               I have to confess that I was spared lot of the difficult reading on this subject through an incredible resource called the Melanoma Education Portal. It’s a well-curated archive of over 100 recorded lectures, clinical tools and staging cards. There are different collections for physicians, GPs, nurses and patients too. The web address is melanomaeducation.org.au or you can follow the links from the home page of the Melanoma Institute of Australia.

Just to give you a sense of how far the use of immune checkpoint inhibitors has spread, in the US, the list of cancer types for which these therapies have been approved has now reached at least 17. These include certain lung cancers, head and neck squamous cell carcinoma, urothelial carcinoma and solid tumours with microsatellite instability or mismatch repair deficiency. But many other cancers are just too mutationally quiet to provoke a serious immune response, even with the help of ICI therapy. In 2019 nivolumab and pembrolizumab were added to the World Health Organisation’s List of Essential Medicines. Because they’ve had such an impact on survival rates they’ve been deemed worth investing in for health systems around the world. And the playing field continues to be populated with monoclonals that block different immune checkpoints.

Until now we’ve been talking about patients with unresectable metastatic cancer, people with little other hope of treatment. The risk-benefit discussion becomes much more nuanced in patients with stage III melanoma who can undergo surgery, and in more than half of cases go on to survive long-term. Australian researchers and patients have been at the forefront of exploring the benefits of immunotherapy in this setting at preventing recurrence of melanoma. There’s a review in the Annals of Oncology led by Alex Menzies and Georgina Long that I highly recommend, but here’s the brief version.  

Until recently, the only approved adjuvant therapy for patients with resected melanoma had been interferon-gamma. A Cochrane analysis of 18 trials showed that this drug reduced the risk of recurrence by a fraction, but it didn’t really improve survival rates beyond a couple of percentage points. Adjuvant radiotherapy following surgery also provides limited gains for considerable toxicity. Since 2015, the immune checkpoint inhibitors were tested in stage III patients who’d had primary tumours and bulky lymph nodes resected. The first ipilimumab trial showed a recurrence-free survival of 47% at 3 years compared to 35% in the group receiving surgery and placebo. In a 2017 publication from the Checkmate group nivolumab was shown to provide a further survival benefit with less toxicity. I asked Matt Carlino to summarise what the consensus is for treating these patients and how conversations about risk versus benefit change compared to the settings we’ve heard about already. There are also questions about how public funding for novel indications keeps up with the research findings.

MATT CARLINO:               Adjuvant therapy is approved in Australia and funded so we're really fortunate that both nivolumab and pembrolizumab are government-funded drugs for a resected stage III melanoma. My main comment was about the risk-benefit discussion about an adjuvant treatment generally, and a treatment that has data about improvement in what we term relapse-free survival, but not yet data showing an overall survival benefit. When you give someone adjuvant therapy, or when you design an adjuvant therapy trial, if the risk of recurrence is 40% for a treatment, you don't know which 40% will recur. You don't know which patients need to be treated.

So I said earlier, if a patient has metastatic melanoma and they have a response to immunotherapy and the only “cost” is hypothyroidism requiring thyroxin, I would say that is a very fair deal and most patients would take that deal. The adjuvant patient who ends up with hypothyroidism, whose melanoma wasn't going to recur, irrespective, is now on thyroxin for the rest of their life. Okay, thyroxin, but what about if they develop type one diabetes? So that, to me, is the important thing and adjuvant, it is that understanding and the nuance of what you're consenting to. It is very different and my discussions, both about the benefit of adjuvant and the toxicity are different in adjuvant versus metastatic.

The other question about immune therapies because we can “cure” advanced disease, there's a question of “treat now versus treat later”. You give adjuvant therapy to everyone or do you give treatment only when they've developed metastatic disease. So you need first to discuss that uncertainty and those pros and cons in the adjuvant setting.

MIC CAVAZZINI:               I’m going to rudely skim over a great seminar on the Melanoma Education Portal from Alex Menzies. He talked about how early trials suggest that immune checkpoint inhibitors may even benefit Stage IIIA patients identified by microscopic metastases in the sentinel lymph nodes rather than palpable ones. That raises a question as to whether it’s even worth doing a biopsy rather than going straight to ICI. And there’s even a small pool of data in favour of neo-adjuvant therapy; that’s to say checkpoint inhibitors prior to surgery. Dr Menzies believes that in some stage III patients, immunotherapy will even obviate the need for lymph node resection entirely. Are you similarly optimistic that the use of checkpoint inhibitors will have such a broad application?

MATT CARLINO:               Yeah, there are two reasons to do neoadjuvant therapy before instead of adjuvant. There is preclinical data to suggest that if you give the same drugs neoadjuvantly versus adjuvantly, you get a better outcome with immune therapy. To me, the way I described it to patients is well, “You're training your immune system to kill the cancer, if the cancer is there, it knows what it's fighting against.” And there is nice data from patient data to say that the expansion of T-cell receptors, so the expansion of the immune response, is greater in neoadjuvant therapy versus adjuvant. You've got this preclinical and clinical data suggesting the immune response may, I stress may, be better with neoadjuvant versus adjuvants. So that's the first thing

The second thing is the prognostication. So you give the drug for six weeks, you do the surgery. The neoadjuvant studies have shown really nicely, if the surgery shows a response, you can tell that patient their risk of relapse is much lower. In adjuvant therapy, you don't have that feedback. There is no surgical specimen removed. So patients like that feedback. Taken to its extreme, if you've got a BRAF-mutant patient, you give them neoadjuvant therapy, you do the surgery, if they've had a complete response to the therapy, you can confidently tell them their chance of relapse is very low. If they had no response in the surgical therapy, you can actually start them on adjuvant targeted therapy; get a second bite of the cherry. So that's the second reason for neodjuvant. Importantly, the benchmark for proving a treatment is better is a randomized phase three trial and one such trial is being run, led by in Australia by Georgina Long with Melanoma Institute. That trial will take time and we're currently recruiting patients to that trial all across the country.

MIC CAVAZZINI:               In your recent editorial for the Australia and New Zealand Journal of Surgery you noted that in fact patients with stage III melanoma receiving immune checkpoint inhibitors now do better than those with stage IIB and IIC disease who currently can only get surgery. The five-year survival rates are in the 90s rather than the 80s. The Keynote-716 publication from two years ago has shown that ICIs would provide a survival benefit even in those stage II patients, but with the potential toxicity this therapy has not been approved in Australia or New Zealand. Is that the right level of caution?

MATT CARLINO:               Stage II patients are difficult. Based on the stage III data, we ran a trial of stage IIB and C, and that's the 716 trial you mentioned. That trial has very preliminary data showing an improvement in relapse-free survival of pembrolizumab over placebo in resected stage IIB/IIC. I acknowledged this data is very immature, so I don't think it's unreasonable yet that the Australian regulators have not yet approved on what is immature data. The drug company, as far as I'm aware haven't even got to that point.

MIC CAVAZZINI:               Well, let's talk about the dollar costs. For some perspective, a full course of anti-PD1 monoclonals costs about $40 to $60,000 and anti PD-1 ligand drugs almost three times as much. For a listed indication, the PBS pays most of that and the patients has a gap of less than $100. There was an article in the Sydney Morning Herald just a couple of weeks ago about a melanoma survivor who had paid this sum out of pocket to prevent recurrence of melanoma. At the time this was still an off-licence indication. The Cancer Council made submissions to PBAC to broaden the listing as it led to “inequity for those individuals and families who cannot pay the private prescription costs.” Are many patients falling through that gap waiting for proven treatments to get funded?

MATT CARLINO:               I remember reading that story, and unfortunately when every drug comes along, there is a window of patients who are unlucky to present between the evidence being presented, (A), and (B) the drug getting approved. So no one likes that gap. The regulator, particularly the PBS, has a really important job of making drugs affordable for the system. And we don't know what the Australian Government pays for anti PD-1 therapy. They go into very confidential discussions with pharmaceutical companies, and I hope the Australian Government drives a hard bargain, because the cheaper we, as a society, pay for those drugs, the more money in the budget that can be used for other things. So I think that that ticket price is probably not what the Australian Government pays for that drug. I think the Australian system in general terms is excellent, the thing that obviously, we can improve on is the time between that drug being having evidence and getting funded. But the solution I don't think is just pay more as an Australian government because there is only presumably a certain amount of money in the pot. So I think the PBS has a really difficult job. And there will be treatments that presumably the cost-benefit don't stack up.

MIC CAVAZZINI:               Well, there was an interesting incident around 2000 [sic, actually 2001] where, Trastuzumab, for metastatic breast cancer had not met cost-effectiveness criteria to be listed on the PBS. But advocacy groups and some in the media presented it as this narrative of patients denied a life-saving drug by a mean-spirited government. The result being that in the run up to a federal election, both major parties made commitments to subsidise the drug and after the election an entirely unprecedented special-access program was set up. So we have to be careful don’t we about the spin and the…

MATT CARLINO:               Yeah and we often do advocacy work around the PBS submission for drugs and we've got a drug currently going through a PBS submission for another indication where I can tell you these patients will benefit—I can interpret the clinical trial data for you. I don't know the other side of the puzzle. What does that course of anti-PD1 one being $40,000, $30,000 mean to the budget? What's the opportunity cost as a society? I think the system of having that driven by health economists and bureaucrats is probably better than politicians.

MIC CAVAZZINI:               And we trust that threshold of around $50,000 per QALY gained?

MATT CARLINO:               Yeah, and I acknowledge I'm not an expert in that. And one of the truisms of medicine, is there are population-based decisions, and there are the individuals sitting in front of you. And perhaps as the person who sees the individual, I have some insights, but there are other insights that a health economist who is hands off, may be better suited than me—I acknowledge that difficulty.

MIC CAVAZZINI:               We've gotten to the last question, which explores that a little bit. So there was an interesting survey published in 2017; Deme Karikios and Martin Stockler of the NH&MRC Clinical Trials Centre. They showed that almost half of oncotherapies given in Australia contain a drug that is not PBS-listed. Nearly every oncologist surveyed had prescribed an unfunded cancer drug, but many told researchers it was a harder conversation to have with patients than talking about end-of-life issues. In an older paper, around a third of Australian oncologists surveyed said they avoided mentioning expensive and unsubsidised drugs if they thought the patient wouldn’t be able to afford them because they didn’t want to cause additional distress. Meanwhile most of the patients surveyed did want to know about expensive drugs even if they weren’t prepared to pay. I think paternalism is too unkind a word but how do you go about these conversations?

MATT CARLINO:               Yeah, so I think you've got to be careful—first thing is you’ve got to check your own biases. And so the line where, “What I would do if this was me,” is actually not the most important question. It's what you would do, and my job is to give you guidance and advice. You know, I think if I had a cancer where immunotherapy wasn't funded but there is some activity, would I redraw my mortgage? Would I check in my super? I actually don't know. I think when a patient is paying it does make that cost-benefit discussion a bit more front of mind. If it’s a treatment that has no plateau, so a last line therapy which improves survival by a small amount with toxicities, that costs $30,000—if that was me, am I better served taking that $30,000 and organizing a party for all my mates? So you don't want to project but you do discuss it slightly differently.

The other thing we spend a lot of time doing is using clinical trials as a workaround. There is a group of indications where immunotherapy is an important treatment, but it's not yet PBS-funded. Just recently, we got funding for microsatellite-unstable colorectal cancer to get immunotherapy, so those patients don't need to pay. But if you've got a gastric cancer or endometrial cancer that's microsatellite-unstable—if that was me, I would pay for the drug, you know, acknowledging I am a professional who probably could find that money. But I've had many patients who have accessed immunotherapy on clinical trials with those indications. So that’s one way that I think oncologists are quite effective at saying, “Yes, you need immunotherapy,” or, “we'd like you to have immunotherapy, I prefer you not to pay $60,000. There is a clinical trial running down the street, or even one hour down the road that is giving an immunotherapeutic—often it's PD-1 plus a novel immunotherapy—my advice to you is to go on that clinical trial to try to access PD-1 for in ‘free’.” So while those discussions are difficult, clinical trials often act as a very good alternative.

MIC CAVAZZINI:               Thanks to Professor Matt Carlino for lending his expertise to this episode of Pomegranate Health. There’s a complete transcript with linked references at our website and don’t forget to checkout the melanoma education portal at melanomaeducation.org.au. The Melanoma Institute’s “State of the Nation” report launched in February aspires to bring melanoma deaths to zero by 2030. Apart from the amazing developments in research we’ve talked about, there’s a big role for GPs and generalists in prevention and early detection.

I want to thank the members of the podcast editorial group who helped me polish this story up for you. They, and the music composers featured are credited at racp.edu.au/podcast. If you click on the episode link you’ll also find a shortcut for logging your CPD credits. There’s a comments section too where you can leave some thoughts on the podcast, or you’re welcome to drop me a line via the address podcast@racp.edu.au. I’m Mic Cavazzini, I hope to hear from you.