Transcript

MIC CAVAZZINI:                Welcome to Pomegranate Health. I’m Mic Cavazzini for the Royal Australasian College of Physicians. In the words of the Beatles, I get by with a little help from my friends. My guest host today is one of the most recently appointed podcast reviewers, Dr Rahul Barmanray. Welcome, Rahul.

RAHUL BARMANRAY:     Thanks a lot.

MIC CAVAZZINI:                The most attentive listeners might remember you from an IMJ On-Air episode back in January 2023, presenting an observational study you conducted at Royal Melbourne Hospital where, am I right, you’re a consultant physician endocrinologist there?

RAHUL BARMANRAY:     That’s right.

MIC CAVAZZINI:                Tell us tell us what else we should know about you and your work.

RAHUL BARMANRAY:     My clinical roles at present are working in endocrinology and general medicine Royal Melbourne, working in endocrinology at Western health, as well as just completing a PhD now through Melbourne University, looking at the effect of proactive glycaemic control in patients admitted to hospital. I also have the opportunity to advise on a non-invasive glucose monitoring technology startup by the name of Opuz, and have really big interest in prevention of complications in diabetes, vascular complications in particular.

MIC CAVAZZINI:                Thanks for volunteering a paper for this new, sort-of, journal club format. So the article we’re looking at today was published December 7 in the New England Journal of Medicine. It’s titled “Baricitinib and β-Cell Function in Patients with New-Onset Type 1 Diabetes” and it was authored by a cast of clinicians and lab researchers from Melbourne’s top institutions. Before we introduce a couple of those authors to talk about the paper, give me your elevator pitch on why this was important.

RAHUL BARMANRAY:                    Yeah, so while we've been able to change the course of some autoimmune diseases, of which type 1 diabetes is very much one, it's still predominantly treated with hormone replacement therapy following the occurrence of irreversible and organ damage. But as most of our listeners will be familiar, insulin replacement has a really high treatment burden associated with it. And it's not just insulin, there’s glucose checking and all the rest of it and even then, it doesn't abolish the vascular complications, the risk of premature death associated with poor glycaemic control and affect that disease process itself. So in the last few years research and animal models has given us some hope that the residual pancreatic β-cell function which is present at the time of diagnosis of clinical diabetes can be preserved. And in so doing, we can reduce the dependence on insulin therapy, maybe not lifelong, but at least for a time.

So quick picture of the immunology behind type 1 diabetes; β-cells, those cells in the pancreas that produce insulin, they present auto antigens on surface receptors called MHC class 1. As an aside, hopefully, all Australians know about Peter Doherty, the Australian Nobel laureate who actually discovered these, so it's very much a close to home topic. And these are recognized by the CD8 positive T cells as targets for apoptosis, for cellular distraction. But in preclinical studies, the surface expression of these MHC class 1 molecules can be reduced by blocking these intracellular effectors called Janus kinases, specifically JAK1 and JAK2.

So if you can stop the formation of these immune synapses between the β-cells and those CDA positive T cells, then you should be able to prevent the death of the β-cells. And then if you can preserve the function of β-cells, you should see improvements in glycaemic control because you have further endogenous production of insulin or at least more than you would have otherwise had. And so that's what was demonstrated in the trial that we're going to talk about today.

MIC CAVAZZINI:               Perfect, thanks. Okay, let's get our guest authors on the line.

MIC CAVAZZINI:               Baricitinib was already approved by the Australian Registry of Therapeutic Goods in 2018 under the brand name Olumiant. It’s listed as a second line therapy for adult patients with moderate to severe active rheumatoid arthritis and severe alopecia areata. It’s also indicated for the treatment of moderate to severe atopic dermatitis in adults who are candidates for systemic therapy.

The trial we’re talking about today in children and adults newly diagnosed with Type 1 Diabetes was nicknamed BANDIT. The study had 40 researchers behind it and I won’t name them all here. The first two authors of this paper are Michaela Waibel PhD of St. Vincent's Institute of Medical Research and Associate Professor John Wentworth of the Walter and Eliza Hall Institute of Medical Research. We’ve had him on the podcast before in his role as IMJ section editor for endocrinology. There are other authors of this paper affiliated with Royal Melbourne Hospital, Royal Children’s Hospital, Melbourne Uni, the Murdoch Children's Research Institute. Speaking with us today are Professor Jennifer Couper,

JENNY COUPER:                I’m a paediatric endocrinologist Women's and Children's Hospital and the University of Adelaide.

MIC CAVAZZINI: and Dr Michelle So.

MICHELLE SO:    I'm an adult endocrinologist. So I work at Royal Melbourne and the Northern Hospital, and then do a bit of clinical research at St. Vincent's institute.

RAHUL BARMANRAY:                    Michelle, Jenny, thanks so much for joining us. So first of all, in this trial that you conducted called BANDIT, you tried baricitinib out for the first time in patients with new onset symptomatic type 1 diabetes. Many listeners might not be very familiar with the staging system of type one diabetes. Could you, first of all, briefly summarize this for those who aren't familiar?

JENNY COUPER:                So the first stages are the presymptomatic stages. And stage 1 is when you have multiple islet antibodies, but your blood glucose levels are normal. And stage 2 is when you have multiple islet antibodies and your blood glucose is what we call a bit disglycaemic—it goes up after food, but then comes down again, but you still, importantly, have no symptoms. And then stage 3 is when you have diagnostic type 1 diabetes, with raised blood glucose levels, but symptoms are starting to appear at that time as well. And that’s the stage we where clinical type 1 diabetes management is begun.

RAHUL BARMANRAY:                    Thanks. And most studies like this have quite a long journey that's behind them. Could you describe to us how this study was conceived?

MICHELLE SO:    Yep. So there's been decades of work done on JAK inhibition in preclinical models, it has been shown that it can preserve β-cells in vitro, as well as in the NOD [Non Obese Diabetes] mouse model—so the mouse model that most reflects human type 1 diabetes

RAHUL BARMANRAY:                    That’s the Non Obese Diabetes mouse model is it not?

MICHELLE SO:    Yes that’s the one. And that showed not only that it could preserve insulin production in the mouse, but even potentially reverse type 1 diabetes. Now, of course, in human type 1 diabetes, you want to ultimately prevent the need for insulin. But we would usually, in immunotherapy trials in type 1 diabetes start at recent onset. They're easier to detect, and when you get it very early in the diagnosis, you've got enough β-cell function to be able to see the amount of C peptide that they have and prevent the decline. So that was the rationale behind that. And baricitinib was already available for use in other disease states and so we didn't need to go to long standing type 1 diabetes as a safety precaution, which is often what is done for drugs that are completely experimental.

JENNY COUPER:                It’s just a lovely Australian story because Professor Tom Kay and Professor Helen Thomas have been working together at St Vincent's for 30 years. There are many people that should be mentioned but I guess they were the ones that underpinned the study for so many years. The last 10 to 12 years of which was work underpinning this very successful trial, So a completely Australian discovery right from the bench right through to the bedside.

MIC CAVAZZINI:               And I’ll skim over some of the study design so we can cut to the chase. The trial was a phase two double-blind, placebo controlled RCT is conduct across multiple Victorian centres over two years, starting November 2020. You recruited patients between 10 and 30 years of age diagnosed with type 1 diabetes no more than 100 days before starting the trial. And you naturally excluded patients with coexisting conditions that could confound your interpretation like cytopenias, known cancer or anyone receiving other immunomodulatory treatments. So this left you with 60 patients assigned to receive baricitinib and 31 to receive placebo. All received standard care for diabetes with a targeted glycated haemoglobin level of less than 7 percent in adults and less than 7.5 percent in children. Have I missed anything critical?

MICHELLE SO:    I think that that sounds like a good summary.

MIC CAVAZZINI:                Back to you Rahul.

RAHUL BARMANRAY:                     So, the primary outcome of this trial was the mean C peptide level during a two hour mixed meal tolerance test after 48 weeks of daily treatment with baricitinib. And in the week 48 assessment, you found a significantly higher meal-stimulated C peptide in the baricitinib group than in the placebo group. A median of 0.65 nmol per liter per minute vs 0.43 nmol per liter per minute, so quite a difference. So what does this tell us about insulin secretion in these patients?

MICHELLE SO:    So just taking a step back, the mixed meal tolerance test is the current gold standard for measuring endogenous insulin secretion in type one diabetes. So C peptide is used as a surrogate of insulin, and it's measured every 30 minutes for two hours after a standardized drink. So it's got carbs, lipids and protein and then we derive the area under the curve. So typically, with the onset of type 1 diabetes over the first year you would expect a gradual decline of endogenous insulin secretion. So this was observed as expected in the group of individuals receiving placebo. But in contrast, we found in the individuals who took baricitinib for 48 weeks the mean C peptide levels were relatively preserved. C peptide only dropped by 4%from baseline as opposed to 30 percent in the placebo group. So that indicates that baricitinib was able to protect the individual’s β-cells from immune destruction and maintain their endogenous insulin production.

MIC CAVAZZINI:                And for comparison, this effect of baricitinib on C-peptide levels was similar in size to that of interventions such as teplizumab, low-dose antithymocyte globulin, and golimumab. But these all require intravenous infusion or subcutaneous injection. Is adherence a big problem with those other therapies?

JENNY COUPER:                Well, I guess, I guess as a paediatrician— I mean, no, in as much as if you give an infusion, adherence really isn't the issue, the issue is the burden for the individual. And, for example, teplizumab—which is a good comparison, I think, because the increments were not dissimilar in C peptide with baricitinib, it requires 12 to 14 infusions over that many days. And so you can imagine for the younger child in particular, that requires long lines and an anaesthetic for the long line, so the intensity of the care is a much greater burden and for most families would be a substantial put off at this stage, even if it's safe in the overall scheme of things. So it's an enormous advantage, certainly for younger children and families for an oral treatment.

RAHUL BARMANRAY:                     So considering the outcome of this study, patients treated with baricitinib needed less insulin at 48 weeks than those receiving placebo, but they still needed insulin treatment. So how would you describe the benefit of this reduction of exogenous insulin, this partial preservation, at least, of β-cell function?

JENNY COUPER:                These were people with so-called stage three type 1 diabetes, so we weren't anticipating that they wouldn't be able to come off insulin, because much of the insulin-producing cells that already been destroyed, so it's really late in the progress. But by reducing exogenous insulin requirements, we're basically saying we've increased endogenous insulin production. And it's been shown in longer term studies, the benefit of even preserving small amounts of your own insulin, at really substantially even lower levels than what was shown in this study, has quite significant, meaningful reductions in vascular complications of type 1 diabetes and in severe hypoglycaemia. And there are other benefits, which I think were only more recently able to pick up such as less variation in blood glucose levels with more of your own insulin, which also vary a bit how the patient feels. So even preserving some and still requiring exogenous insulin, is still of benefit to the patients.

RAHUL BARMANRAY:     So, to paraphrase the pancreas is still the best determinant of insulin requirements than a human brain or even a pump brain?

JENNY COUPER:                And I guess another thing we didn't completely look at in this study, but you could argue for future studies, that if you could just reduce a patient's insulin schedule to e.g. one injection a day, rather than four or continuous insulin pump delivery, that would also be an enormous reduced burden of care. So the complete reduction of any need for insulin would be the ideal, but even these lesser changes still greatly help the patient.

MIC CAVAZZINI:                The maximum difference between groups in C peptide levels and insulin doses was already seen by 12 weeks, and the gap didn’t really increase over 24 or 48 weeks. So, does this tell you something about the way the drug is impacting β-cell function or the way that the disease is progressing in the placebo group?

JENNY COUPER:                Well, I think if there was to be an effect it would be seen quite soon and it would have some ceiling, as it were, because we've got a limited number of insulin-producing cells remaining, we can't precisely define that. And indeed, that is the way other effective immunomodulatory agents have worked; relatively quickly and then with some ceiling on that benefit. So I think that was anticipated just looking at other studies and the stage of disease which we are at in these sorts of studies.

MIC CAVAZZINI:                And secondary outcomes included the glycated haemoglobin level, the total daily insulin dose, and glycaemic control as assessed with the use of a continuous glucose monitoring device. So at week 48 of the study, continuous glucose monitoring showed there was less variation in interstitial glucose level in the treatment group, and measures of “time within range” and “time above range” also favoured the baricitinib treatment. Can you explain for a generalist how each of these different measures might fill in a different part of the picture you're looking at?

MICHELLE SO:    Yeah, sure. So, as Jenny mentioned, it's about preservation of endogenous insulin secretion. And if you imagine you're preserving the endogenous insulin secretion, you would expect a clinically significant impact on the amount of exogenous insulin required to maintain glycaemic control. And so sure enough, we found that that mean daily insulin dose was significantly lower at 48 weeks and they still managed to maintain a similar level of glucose control. So that baricitinib having a mean HbA1c of 7 percent at 48 Weeks was comparable to the 7 and a half percent in placebo, they weren't statistically different.

The other key secondary outcome was that continuous glucose monitoring. So, every participant wore a continuous glucose monitor which measures the interstitial glucose levels every few minutes. And that was worn for two weeks at very specific time points; So, baseline, week 12, week 24, week 48. And it can give a more granular picture of glucose control compared to a three month average, which is what HbA1c does. So, the most significant improvement was the glucose variability. We measured that with the coefficient of variation and that was significantly lower in the baricitinib group.

So that wasn't surprising in a way because where you've got endogenous insulin secretion preserved, your own native way of being able to manage meals is better. And that finding was consistent with continuous glucose monitoring studies in preclinical diabetes, where they found that those people more likely to progress to clinical diabetes had a rise in this glucose variability. And “time in range” would be similarly self-explanatory. You have people managing to stay within the target range more often where their own pancreas producing insulin.

MIC CAVAZZINI:                So tell me if I’m going off on a branch here. It looked like there was a glitch in one of those trends. The curves for “time within range” and “time above range” were significantly different between the cohorts at weeks 12 and 24 but they dipped closer together at week 48 such that the error bars actually overlapped. And this seemed to be the placebo group improving a little rather than the baricitinib group getting worse. What do you think could be going on there?

JENNY COUPER:                Yes, I think we've got to realize that while it was really nice to see these changes to “time in range”— but trials looking at commercially-available automatic delivery pumps that are now available that can improve “time in range” quite independent of improving C peptide. So, I think the answer really is that “time in range” is affected by many things. One of them is C peptide, but there's also effects of exactly how the patient's able to negotiate the care, and the fact of the type 1 diabetes, exactly what their method of insulin delivery, is exactly how well tuned they are in relation to this system of mentioned, delivery, et cetera. There's so many variables affecting “time in range”, that it's not surprising that there wasn't the sort of perfect relationship at all.

RAHUL BARMANRAY:     Now, with any trial products we're, of course, always very interested in the side effect profile, I suppose baricitinib itself is not a trial product, it is approved just not in this in this context. It's reassuring that the frequency or severity of events, adverse events, was no different between the groups. No events were attributed to either baricitinib or placebo. But this is, of course an immunomodulatory treatment, so clinicians might rightly be worried about adverse effects related to immunosuppression. How concerned should general clinicians be about immunosuppression-related adverse effects with the use of baricitinib specifically in patients with preclinical or clinical type 1 diabetes.

JENNY COUPER:                I think the term that’s being used a lot now is that we’re sort of tweaking the immune system. This is not what we would generally understand as immune suppression, it is immune modulation and there's obviously enormous difference. I mean there is a history also—a good history—of JAK inhibitor safety in this age group. And really, under the 65 age group there is very good safety data. And, of course, a short trial like this can't address this question but longer-term trials, in children down to two years have not really shown severe adverse effects due to immunomodulation. Similarly, agents such as the teplizumab and Abatacept thus far have seemed to be very well tolerated.

MIC CAVAZZINI:                And just to fill in for our listeners the wider spread use we’re talking about; baricitinib is already widely available to treat rheumatoid arthritis, alopecia areata, and severe coronavirus disease. And other JAK inhibitors are used to treat juvenile arthritis and inflammatory bowel disease. So is that the pool of long-term data you're talking about?

JENNY COUPER:                I guess it's also used to treat severe atopic dermatitis. And it's also used to treat severe ulcerative colitis and a, sort of, rare form that affects quite young children, as well.

RAHUL BARMANRAY:     And of course, the BANDIT study isn't occurring in isolation. In 2019 the results of teplizumab delaying the onset of clinical type 1 diabetes was presented at the American Diabetes Association conference. I was lucky enough to be there in the audience and it received a standing ovation. It created quite a stir of the diabetes community and lay media alike, I think the company that made teplizumab their share price increased by 200 percent overnight or something, something ridiculous. The BANDIT trial is different, targeting a different population for one thing. How else does BANDIT complement, build upon, or differ from previous evidence on teplizumab?

MICHELLE SO:    I guess, in a way, every immune therapy trial performed in type 1 diabetes builds on the other. So if you picture in your mind a schematic of a β-cells and the different cells of the immune system that are involved in the pathophysiology of type 1 diabetes. So you have that CD8 T effector cell, instigating the attack; you have antigen-presenting cells stimulating the T cells and the β-cells also activating the T cells; and then you've got the cytokines that make up the inflammatory milieu; each immunomodulating agent has added to our understanding of what part of that schematic is critical to the ultimate destruction of the β-cells, and at what stage of the disease we can actually modulate the disease trajectory.

So templizumab is an anti-CD3 antibody. So for the listeners, CD3 is a protein complex present on all T cells. And it's believed teplizumab was effective by impairing T cell function by inducing partial CD8 T cell exhaustion. So it's been shown to be effective now in stage 2 and stage 3 type 1 diabetes. On the other hand, baricitinib blocks the formation of immune synapses between β-cells in CD8 T cells, and that prevents the death of β-cells, or is believed to. So both demonstrate the critical role of the T cells in type 1 diabetes pathophysiology, but work via quite different mechanisms. So, teplizumab is cell surface, baricitnib is intracellular signaling. But similar to teplizumab, the next step for baricitinib would be to determine the efficacy in stage 2 type 1 diabetes. So to delay the need for exogenous insulin, and that will be critical to understanding the role of JAK1, 2 or JAK inhibition in general, in the disease.

MIC CAVAZZINI:                And I think you've alluded to this already that many patients diagnosed with stage three, type 1 diabetes have too much irreversible damage to their β-cell mass for cessation of insulin therapy to be a realistic goal. And you speculate that initiating baricitinib earlier might produce greater benefits, but how would you identify patients with presymptomatic type 1 diabetes in that sort of strategy?

JENNY COUPER:                Well, stage 2 in particular, is not such an easy stage to pick up. And there's enormous difference between a stage 1 which is progressing very slowly or a stage 1 that's just about to progress through to stage 2 in terms of its likely response. So the only way to find these people is to screen. And there are screening programs in Australia, but I guess none yet have been ready to completely take this quite enormous task on. There's a type 1 screening program looking at first degree relatives and screening people who are at increased risk—because they're 15 times the risk of someone without a first degree relative. And there's been a pilot also in this country of general population screening to see the health economics, and also the benefits of potentially having a general population screening program. If we were though, to move towards, for example, a stage two trial with baricitinib it seems almost very unlikely that Australia alone could do such a trial. And it would almost certainly need to be a multinational study.

MIC CAVAZZINI:                Okay, but on paper, on the back of an envelope the costs of population screening aren't an immediate nonstarter.

JENNY COUPER:                No. And globally, Italy has just passed a law whereby public health units are required to implement general population screening. And but they're also good examples; the first in Bavaria, and also in Colorado and other parts of USA, where general population screening is occurring over discrete areas. So this is certainly happening internationally, primarily, to prevent the serious complication of ketoacidosis when people present really late with type 1 diabetes and require intensive care. But also to monitor people, so they have a more gentle entry into requiring insulin and have good education prior to starting that.

MIC CAVAZZINI:                Right, so yeah, you're offsetting those very big costs of downstream treatment. And yeah, I don't know if you want to go into this. You also collected some mechanistic outcomes to explore a downstream stage in the signalling pathway involving transcription factors STAT1 and STAT3 (for signal transducer and activator of transcription). Did you find some support for the model?

MICHELLE SO:    Yes, we did. We were expecting baricitinib to inhibit the JAK-STAT pathway signalling, as measured by the phosphorylation of STAT1 and STAT3, as you mentioned in our assays. And we did find this to be the case. And so it is a good measure of pharmacodynamics of baricitinib activity in our participants.

RAHUL BARMANRAY:     And following on from that, from your preclinical studies, you speculate that there's a signalling role for interferon gamma signalling at these immune synapses between the CDI positive T cells and the β-cells. So how might this possibility inform the design of future trials?

MICHELLE SO:    So actually, the model is that interferon gamma induces upregulation of HLA class one on the β-cells, which increases the ability of CD8 T cells to recognize β-cells. So this might not happen, actually, in the immune synapse. The interferon gamma might be made by other cell types like CD4 or CD4 T cells or macrophages, for example. So we can't actually measure this in our participants because we don't have access to their islets or β-cells, but we have shown it to be the case in mouse studies. So this might inform future trials with different, more selective JAK inhibitors. For example, one that just blocks JAK1, which—blocking JAK1 should still inhibit interferon gamma signalling, so would be appropriate to test according to this model.

RAHUL BARMANRAY:     So in those NOD mice who've experienced remission of type 1 diabetes with this JAK inhibitor therapy, there's been this gradual relapse after cessation of treatment. So similarly, in human patients with autoimmune alopecia, or juvenile idiopathic arthritis, recurrence of disease has been observed on discontinuation of JAK inhibitor treatment. So what are the implications of this for future research with regards to JAK inhibition and management of type 1 diabetes with such therapies?

JENNY COUPER:                I mean, the second year of the study, participants off any treatment has not yet been analysed but we would anticipate that the benefit will wane relatively quickly when it’s stopped, according to those studies. And indeed, particularly according to the clinical data which you present in people with alopecia who get a response and then stop, how quickly that can return, which is obviously a very visible clear loss of benefit.

RAHUL BARMANRAY:     And following on from some other things that you've mentioned, Michelle, you mentioned when you were describing the various cellular targets there’s JAK inhibition, teplizumab targeting T cells that you can target various elements, imagining a model like you described—Is there any possibility in the future that will see combinations of these elements? For example, baricitinib and teplizumab, perhaps?

MICHELLE SO:    Yes, that's a great question and definitely baricitinib would be a great drug to trial in combination therapy. There's already a trial currently open in Melbourne through Trialnet, and I believe other states in Australia using Rituximab and Abatacept. And also there's one looking at intranasal insulin and Abatacept. So there are combination therapies being trialled already. And yeah, it will only get better because of the number of immune-modulating therapies showing efficacy in type 1 diabetes.

RAHUL BARMANRAY:     And I'm glad you mentioned Trialnet, because that really is my next question. Let's say you're an endocrinologist, or even a GP and you find a patient with early stage diabetes. How can a practitioner, perhaps, refer patients to be involved in these studies? Obviously, the studies moving in a very quick pace, and you know, the studies that are available and so on, is going to be changing very quickly. Is there any ability for doctors to refer patients to your trial program?

JENNY COUPER:                Well, there's ATIC, the Australian Trials Immunotherapy Collaborative links sites around the country in each state which have been involved in these studies, but also has information very clearly about the current studies that are. So I think the first step really would be to go to that website.

MIC CAVAZZINI:                And how long might you expect it to take to progress successfully through Phase 3 and changed indications at the TGA?

JENNY COUPER:                You know, I guess, a Phase 3 study is in different hands. And I think that is big question, a very important one. And certainly, discussions are occurring.

MIC CAVAZZINI:                Just to follow up from that. I'd like to convey to listeners that this was an investigator-initiated trial. The questions weren't being asked by Big Pharma. But what declarations of interest would you like to share?

JENNY COUPER:                I don't have any declarations of interest.

MICHELLE SO:    No conflict of interest.

MIC CAVAZZINI:                But, I think there was a list in in the paper, wasn't there, apart from donations of the drug?

MICHELLE SO:    Yep, the drug was provided by Lilly. But they had no involvement in the analysis, the write up or any subsequent part.

RAHUL BARMANRAY:     So thanks so much for answering our questions thus far, is there anything we haven't talked about that you'd like to convey to the listeners about this study, the wider program.

JENNY COUPER:                I think something that struck me, was just the enthusiasm of families to come in. I mean, I'm a paediatrician. We did actually look after children at my centre, its was the only centre and South Australia, up until 30. But most of them were younger families were in the middle of COVID. People weren't, you wouldn't have thought so keen to come to the hospital. Melbourne was locked down a lot of the time of the study. And yet despite that, there really was an excellent recruitment rate that actually exceeded even the projected required recruitment rate. I think that was just an interesting kind of statement on the enthusiasm—how the families feel about this work very much. Do you want to enter that perhaps Michelle?

MICHELLE SO:    Yeah, just to follow on from what you said, Jenny. It was during the pandemic, and we didn't show any significant differences in COVID infections between the two groups. So it just shows again, baricitinib was extremely well tolerated and safe. And generally all the immune therapy trials that have shown efficacy, haven't shown any severe adverse events. And so, I think as generalists and even general endocrinologist can sometimes feel a little bit nervous about immune therapy in type 1 diabetes and/or think that it's technology versus immune therapy. But I think immune therapy has an increasing place in the management of type 1 diabetes, and it's going to be part of the equation to improve the quality of lives of people with type 1 diabetes.

MIC CAVAZZINI:                As I was editing this podcast, I was asked by a couple of reviewers about the economic calculus for such therapy. The answer turned out to much messier that I expected. Well, the first part is easy enough; Baricitinib is subsidised by Australia’s Pharmaceutical Benefits Schedule to the tune of $1270 for 28 pills, which adds up to $16,555 for a whole year’s supply.

Not cheap, but nor is the cost of the type 1 diabetes itself. In a 2020 paper from the Juvenile Diabetes Research Foundation and accounting firm Accenture, analysts took a top-down approach to this problem. Medication, blood tests, consults and hospital care added to about 0.7 billion dollars a year in direct costs across Australia and there was another 0.4 billion dollars in unemployment benefits, disability support and formal care. Dividing by the approximately 127,000 Australians living with Type 1 diabetes gets you a figure of $8,600 per patient per year, of which about a quarter is shouldered by private payers.

By contrast, in a 2009 analysis for Diabetes Australia called DiabCo$t (see also) researchers came to an average of about $3,800 a year in direct costs for each person with type 1 diabetes. This study went bottom up, relying on surveys with 2,200 randomly selected patients or their families, and they included in these costs expenses on transport to medical appointments and diabetic-appropriate food. Patients with multiple complications could have annual costs almost four times higher than the mean, and these include neuropathy, cardiovascular disease, blindness and kidney disease.

To get some clarity on this question I spoke to Professor Stephen Colagiuri of the University of Sydney who led the DiabCo$t study. He said that the difference of almost $5,000 between the two cost estimates could in part be explained by the rollout of insulin pumps and continuous glucose monitoring devices in the decade between the publication of those studies. These high-tech interventions can each cost a few thousand dollars a year per person and the Australian government subsidises this cost for about 58,000 people under the age of 21 or with specific indications.

But the two research papers differed much more when it came to the indirect costs of type 1 diabetes; the lost productivity and diminished wellbeing of patients or their carers. Where the DiabCo$t group arrived at an average of $800 per patient based on their surveys, the accountants behind the JDRF modelling produced a figure of almost $15,000. They don’t display a lot of working publicly, but say that this latter estimate was derived from the health economist’s nominal value of a statistical life year, $213,000, then scaled by the years and quality of life lost to death and disability. The accountants also used national wage data and tax data to put a number of lost productivity and even leaned on some of the findings from the prior DiabCo$t study.

For example, from respondents who were of working age about a third were not in employment at the time at the time of the survey and a third of those said they weren’t able to work at all because of their disease. Of the carers who responded to the survey, 19 percent reported that they weren’t working in order to care for the person with diabetes, and most of those who working had to give up a almost a full day of work each month. On top of these economic impacts, both studies also considered less easily quantifiable ones like quality of life and the mental health impacts of chronic disease.

From the information at hand, I can’t say whether the indirect costs of type 1 diabetes were under-reported in the surveys or weighted too highly in the JDRF’s modelling. I suspect a bit of both. But whether the total economic burden in a year of type 1 diabetes closer to $4,700 per patient or $22,000 seems important when deciding whether to fund a drug that costs $16,500.

Or is it? As Professor Stephen Colagiuri said to me, as much as we’d like the numbers to give a black and white answer, there are judgement calls right through the process, not least the subjective determination of what a statistical life year is worth. Professor Thomas Kay, who led the BANDIT trial, said that new treatments are often more expensive than old ones when they’re introduced, but we back them because they have advantages for the target population. He added that we can expect baricitinib to get cheaper once other competitors and generics make it to market.

Both Professors noted that we already choose to fund antivirals for Hep C that cost over a hundred thousand dollars per patient. Or immunotherapies for cancer that sometimes only grant an extra year or two of life. At the end of the day there’s always a moral variable in these funding decisions too, and we naturally weight quality of life in children very high. Even if groundbreaking therapies like teplizumab and baricitinib only delay type 1 diabetes, rather than preventing it outright, they can help kids reach their full potential during the formative years.

As mentioned earlier, though, the catch is in identifying the right candidates for early intervention. Every year in Australia there are some 3000 new diagnoses of type 1 diabetes from an estimated pool of about 25,000 people with preclinical disease. Screening for islet autoantibodies has shown pretty good sensitivity in cohorts with genetic risk factors for type 1 diabetes, but not so much in undifferentiated populations. The mechanics and mathematics behind screening programs are a fascinating puzzle in their own right, and one we’ll have to come back to in another episode.

For now, I want to thank Michelle So and Professor Jenny Couper for contributing their expertise to Pomegranate Health.  Professors Stephen Colagiuri and Thomas Kay were also very patient with all my questions. And many thanks to Rahul Barmanray for volunteering to host this journal club episode. If you’ve come across some hot research conducted by RACP members, please share it with me and consider whether you’d like to act as a guest interviewer. You can write to the address podcast@racp.edu.au.

I’d love to help share more member-generated content like this, particularly from specialty groups who can contribute semi-regularly. Perhaps you have a colleague who’s just given a great lecture. If you sketch out some question to tease out that material, I can record the conversation and share it with the broad Pomegranate audience.

Did you know that every episode gets downloaded over 6000 times over the first 12 months from publication and on those numbers we’re probably among the top 7 percent of podcasts on the internet. Interestingly, a fifth of listeners are located outside of Australia and Aotearoa-New Zealand, so it’s one way to get your hard work out to the world.

And did you know that every month there are several webinars hosted by College and specialty societies? Members should already receive an email titled RACP Events Digest and some recordings can be found later under “past event presentations” from the RACP homepage, at the College Youtube Channel or the Medflix video library located at elearning.racp.edu.au. There really is something for everyone.

This podcast was produced by the waters of the Gadigal people. I pay respect for their custodianship of this country over tens of thousands of years. I’m Mic Cavazzini. Thanks for listening.