TNR Grand Rounds Cytokine Storm Syndrome Save

Hi, everyone. Welcome to Tuesday night rheumatology, our grand round series. It is the 05/26/2020. Our program tonight features Doctor. Randy Kron from, Crohn, sorry, from UAB and Children's Hospital of Alabama.

Randy, welcome to Tuesday night rheumatology in our grand round series.

Thank you for having me.

Yeah. It's really great to have you because, you know, I was asked about four weeks ago to come up with a lecture on COVID and rheumatologic drugs and they wanted to know about cytokine storm and whatnot. You know, I knew a little bit, cytokine release syndrome with, you know, CAR T cell therapy and MAS and, but, you know, I had to do a cram course on all the information that's coming down. And Randy is the guy when it comes to the cytokine storm and understanding MAS. If you get into that subject, he's got a ton of publications.

So his being here tonight is really a treat for all of us, me included. It's gonna, I think help us as we try to help others with the syndrome. So a few things I wanna start out with and get Randy your take on, what do you think, what are you hearing around the country and within your academic center about the role of rheumatologists in managing COVID and managing cytokine storm. Are we integral? Are we marginalized?

I'm hearing sporadic reports. Some rheumatologists are really involved in planning and committees and whatnot. Others are just waiting for a consult. What's your experience so far?

Yeah, I find it fascinating. It seems to be one of two polar opposites. Having said that, I think we might be a little bit in between, but I can't remember if it was the ACR division directors group or the fellowship program director. I'm part of both of those. But anyway, one of them sent out a survey and the results were pretty fascinating.

Either people were completely just out of the loop and maybe ID and or the intensivists were running the show or it was like everybody, right? The way we would hope it would be that everyone's expertise has an input into what's going on. At UAB it's pretty much dominated by ID and the intensivists both on the adult and ped side. I've tried to push into both those spaces to a little bit of and over time actually I think they're starting to let us in a little bit but it's been a slow process.

Yeah, I think it's gonna be a little bit like what happens when an MAS happens to a systemic or a lupus patient. You'll get the consult, you get the consult late. You know, they expect you to show up with, you know, magic wands and whatnot. And whereas the, you know, the ID or the nephrologist are fighting with the hematologist about etoposide versus, you know, cyclosporine. And you come in and again, we want a rheumatologist to be comfortable with this really important disorder that's mediated by the inflammatory cytokines that we play with all the time.

Let's talk about a little bit about Plaquenil. A lot in the news, obviously in the last two months, it sort of had this rise in popularity based on nothing. And then it sort of, the supposed shortage never happened because we started getting these negative trials and a whole bunch of negative trials. And then warnings from the FDA and the EMA and Health Canada saying, don't take it for prevention. Don't take it if you're not in the hospital.

You're in the hospital trial, then go ahead and take it. And then, of course, Trump comes up and says he's taken it and now it's back up again. But then now you had the WHO today say they suspended their big chloroquine trial. So what's your take on all this at this point, as far as, if your rheumatologist were to ask you, should I be stopping it in my lupus patients? Should I be advocating it for everybody with COVID infection?

What do you think?

I think in general, whether it's Plaquenil or any of the meds that our patients are on, you should probably just stay on what you're on if you're doing well on it. If you can taper down and not flare great, but otherwise I would just stay on what you're on. I think that makes the most sense because it doesn't seem like our patients are at too much of an increased risk of having bad COVID compared to anyone else. So if you asked me personally, I didn't think Plaquenil was gonna be like this wonder drug. A lot of things that look good in vitro don't really ever pan out in vivo, but that's why the clinical trials are done.

There's also pretty good data that being on Plaquenil for whatever disease you have doesn't seem to keep you out of the hospital for COVID-nineteen. I mean, you're not gonna get it necessarily more than anyone else but it's not keeping you out of the hospital. So I'd be shocked actually personally to find it much of a home run-in this disease.

You know what I I would say that actually one of our listeners, Doctor. Fung down in Waco brought to my attention one of the early trials was that, all the negative stuff, either more hospitalizations or more cardiac events or arrhythmias or the things that everyone's worried about with hydroxychloroquine probably aren't from hydroxychloroquine. They're probably from the fact that the people that are getting hydroxychloroquine are sick. They're older. They have preexisting cardiovascular disease.

They're dying and having complications from their age and their comorbidities. And, you know, it's a bystander effect for hydroxychloroquine, including last week's big Lancet report where it was saying, you know, thirty seven percent more deaths and, you know, forty percent more ventricular arrhythmias. Again, I don't know that it's really related to that because in that study, it's not controlled. It's retrospective analysis of those on chloroquine, hydroxychloroquine with or without a macrolide antibiotic and who's gonna get those drugs other than people who are pretty sick.

I mean, I think adding the macrolide is something else too, right? We don't typically do that unless you've got whatever respiratory element you may have but that may make things worse but it certainly seems like the amount of side effects that are being reported and as scary at least as the press is making Plaquenil out to be was never really an issue for most of us particularly in pediatrics prior to COVID.

Yeah, all right. So I'm gonna hand this over to Doctor. Kron who's gonna talk about the cytokine kind storm syndrome and Randy take it over. Let you take it, let you drive the screen here.

All right so here we are. I like this kind of pointer so we'll stick with that if I can. These are my disclosures that are relevant to this talk.

Not seeing your slides. Oh sorry,

let me go back. How about now?

It looks like it's coming up. There we go.

All right. So these are my disclosures that are relevant to this talk. I do work pretty closely with Sobe and I'm a huge fan of Anna Kuner and most people know that about me. So my interest in this got really peaked when I was in Philadelphia. And I saw this young girl with Ed Barron's at the time who was one of my fellows at the time and who's now head of pediatric rheumatology at Children's Hospital of Philadelphia.

And Ed and I saw this 14 year old girl who came with a prior diagnosis of cytophagic histiocytic paniculitis. Something I had never heard of, hope to never see again, actually. And this poor young girl, she was 14, but she'd had fevers every day of her life since she was essentially born. I was diagnosed at like six months of age with this and had total body panniculitis and lipoatrophy to the point that in comparison to long distance runners, they seemed morbidly obese. Anyway, she presented to us with belly pain and she had a big liver on exam and seeing the word paniculitis, we gave her cyclosporine and corticosteroids.

And she had been off of traditional therapy for about a year and a half at this point. I think it was mostly on acupuncture, but she had been through the NIH and been on TNF inhibitors, a whole variety of things and just had never really done well. And she got a little bit better, but unfortunately within a month she was back. She said this time she was presented with seizures and belly pain and got admitted to the ICU and quickly deteriorated and had inflammation in her blood and her cerebral spinal fluid. She was coagulopathic, pancytopenic, hyperferritinemic.

And actually for me personally, probably the sickest kid I've ever seen or taken care of who ultimately came out of the ICU at the other end essentially unscathed. But she was on multiple inotropes, she had ARDS, high vent settings. She was in renal failure, liver failure, pancreatitis, all her lines were down bleeding from everywhere and her CNS was not well at all. And this was her marrow during that hospitalization and here is this activated macrophage or histiocyte engulfing other cell types or hemophagocytosis. And that's helpful, although it's only the present in about sixty percent of cases and it can be present in other conditions.

So it's neither highly sensitive nor specific, but it often is a criterion that people will use. This we got in retrospect, this was a skin biopsy from nine months of age showing these lymphocytes that should not be infiltrating her fat cells like they are here. There are these HLH four criteria been around for a long time and they were really developed for a familial or genetic form of this disease or cytokine storm HLH. And you can either have a molecular diagnosis by mutations in perforin or perforin pathway genes. Most infants don't show up with this unless they've had a sibling for example with this.

But they have these five out of eight criteria, three of which kind of E, F and G a lot of times you don't even order or get. So it's hard to meet five criteria and particularly for the secondary forms that we as rheumatologists tend to get consulted on. If you wait for that fifth criteria and your patient may be dead. However, this child was so sick even with the secondary form of this, that she met all these criteria in spades. And they had called the oncologist first because about this time since back at the 2004, the oncologist had been for a long time treating diseases like this MAS, HLH cytokine storm with the toposide and had a prospective study, not a randomized study but a prospective study to collect data on this.

And so people were comfortable going down this pathway. So she started down that pathway with the toposide, dexamethasone is what they tend to use for better CNS penetration. Although they only use about a 20% corticosteroid equivalent to our Solu Medrol doses for example. And cyclosporine and she was actually getting worse. And around this time, this was after Virginia Pasquale had kind of started exploring IL-one blockade in patients with systemic JA, which for both systemic JA and Still disease was revolutionary for those patients.

And it turns out that a lot of kids with systemic GA will develop a cytokine storm or macrophage activation syndrome. And so Ed and I reasoned this girl is gonna die. She was as close to death as I've seen anyone. And we might as well try this, this IL-one blockade with Anakinra, talk to the family, talk with the team, decided to do it. We use what we would now consider relatively low doses like around two milligrams per kilogram per day.

And within two days she woke from her coma within six days she was out of the ICU. Within two weeks she was out of the hospital. And for me it was really nothing short of miraculous. And I think for both Ed and I, Ed's now got a mouse model of MAS. And myself I tend to do more of the human studies.

It was kind of it really flipped our career paths, this one patient. And so what's the pathophysiology behind this? And so if we just take a very generic strength of the immune response on the y axis moving forward with time along the x, let's take a normal host and they're ticking along doing fine and they get some infection, let's say influenza and then your immune system is mounting a response to that infection. Hopefully it's starting to clear that infection and then your immune system is actually hardwired to shut itself down. Otherwise we'd be in one chronic state of inflammation or be a giant lymph node like an LPR mouse or an ALPS patient for example.

But what's not to occur in cytokine storm is this doesn't ramp itself down, it just keeps going. And one of the thoughts is, is that your cytotoxic CD8 T cells as well as your natural killer cells, when they recognize the infected antigen presenting cell have trouble killing it through the perforin pathway, for example, and that prolonged engagement between these two cells leads to the increased cytokine expression which leads to the multi organ failure. And that was modeled very nicely by Michael Jordan, not the basketball player, but the pediatric haemato oncologist when he was in Pippa Merrick's lab in Denver, Colorado as a fellow. And what Michael did was he took wild type or normal genetic mice and he took perforin deficient mice on both chromosomes, completely deficient in perforin. And these mice are actually perfectly fine in the mouse colony.

They're just hanging out, eating, growing, doing what they do until they get the wrong infection in this case a bug LCMV. And the wild type mice, the cells will get infected and present to the CDA T cells and they'll spew out gamma interferon and other cytokines. And through for and they'll lyse these cells and shut down the viral life cycle. And the mice do great. The perforin deficient mice are incapable of lysing the infected antigen presenting cell through the perforin pathway.

And they all die uniformly within ten to fourteen days of a cytokine storm. But what Michael did which and I think actually this paper should be required reading for every physician period at some point, particularly us in rheumatology. But when Michael did was he blocked either anti with anti CD8, the CD8 T cells. And then those mice under those conditions with this infection, they don't clear the virus as well as the wild type mice, but they don't die. And similarly, if you just block one of these pro inflammatory cytokines in this case, interferon gamma, they don't clear the virus quite as well, but they don't die.

So this is a really important concept that although the virus is the trigger, it's the immune system that's killing the host. And I think that translates to humans as well. And so perforin is one of these genes in this pathway, as well as a variety of other genes in this list keeps getting longer that if you're homozygous deficient in them, you will get a familial or infantile form of this disease. And this is modeled here. This could be either a cytotoxic CD8 T cell or a natural killer cell.

They share the same pathway. And when they engage their target cell, pre packaged cytolytic granules that contain proteins like perforin and granzymes. After the activation of the cell, they traffic along the actin cytoskeleton and the various proteins Rav27A and taxin 11, MONK13, four etc. Are all involved in this process, whether it's docking them or fusing them to the membrane. Perforin then gets released into the immunologic synapse and actually punches a hole or forms a channel or pore between the two cells.

Granzyme is delivered and triggers the cell to undergo cell death via apoptosis presuming all these proteins are intact. So when I got to Alabama, 2007, so a little while ago, we started getting consulted. I came with Tim Buchelman who was the other fellow with Ed Barron's at the time and I convinced him to come down with me. And anyway, we started getting consulted on a variety of patients, mostly in the ICU who were developing macrophage activation syndrome or cytokine storm. Some of them were our own patients like systemic JA patients, but many were not.

And I was surprised how many of them when we started looking at their genes had mutations in the same perforin pathway gene like way more than I was expecting. And so, excuse me, we hypothesize that although the familial form of this disease is from autosomal recessive deficiency in perforin, for example, or any of those other genes, even having a heterozygous mutation may be a risk, therefore contributing to the pathophysiology of the disease either through a dominant negative or partial dominant negative effect, or in terms of not having enough potentially a hypomorphic effect. And so we explored that. And this is work done almost entirely by Mingxin Zhang in the lab. And what Mingxin did was we took patient mutations that we identified in these proteins.

And we made cDNAs because the NK cell line that we use this this human natural killer cell line had all these genes intact. So we could make cDNA to reintroduce more of the wild type gene by a lentiviral approach or by site directed mutagenesis we could mutate the cDNA to match the patient's mutation. We could then introduce it into the NK cell line and compare it the wild type gene effect versus the mutant gene effect on a whole variety of assays. And so here's the lentivirus, it's got this immediate early CMV promoter, which is great at driving gene expression and lymphocytes including natural killer cells. And here's the copy DNA or cDNA for whatever gene we are interested in.

We give them back excess wild type gene which they already had, or we'd give them the patient mutation. It was linked so we could also express green fluorescent protein and either sort the cells or identify them by that. And the first one I'm gonna spend the bulk of my time on is this 18 year old patient who had this and most of these mutations were missense or amino acid changes, including her own in this protein called RAB27A. And this was a previously healthy 18 year old girl. In fact, at the time she was ranked number two in barrel racing the state of Alabama if you're familiar with barrel racing.

But she presented to the Children's Hospital with two weeks of fever and belly pain onto the GI service and sat there for a while, getting sicker, they switched to tendons the head of the GI service came on gave me a call said, Randy, I don't think this girl is doing well. I think you need to come see her. And I came to see her and they were kind of wheeling her into the ICU. She was in bad shape. She was still awake at that time.

You could talk to her. She didn't make sense. She had a fever and she had a big liver and spleen on exam. But really her labs were scary. She was pancytopenic.

Many of these patients will develop a hepatitis. She was coagulopathic, hyperferritinemic, elevate all of these features of even once again, she met the HLH criteria which are often way too restrictive for secondary forms, but she met them well. And it's important to point out the sedimentation rate of 10. So she was very inflamed and had a normal or low sedimentation rate. And that's because fibrinogen is one of the major drivers of your sedimentation rate.

So it sits on the outside of the cell and sinks them quickly. And it gets consumed in this DIC or coagulopathic process and though so your sedimentation may start up rate may start out high, but then if it drops, that's usually ominously a bad sign, particularly as the patient's not improving and other markers like the CRP are going up. And so by this time we had developed this cocktail of cyclosporine, solu medrol or dexamethylprednisolone, and the IL-one receptor antagonist anakinra. And clinically she got remarkably better very fast as did her platelet counts improve, her ferritin dropped within days and her liver enzymes came down and she did wonderfully. So we took her mutation back to the lab and introduced it into the NK cells along with the wild type version of that same gene.

And looked at, here's a single experiment where we looked at cell death. And so what we're doing is we're taking natural killer cells and putting them in with their natural target, which are K562 cells. And E to T is the number of effector to target cell ratio. So the more natural killer cells you have per target cell, the more killing you get. And that's shown here along the y axis.

And this is a summary of 12 experiments because the effects were subtle, but we wanted to make sure they were real. So if you just take the NK 92 cell line and introduce the empty lentiviral vector, they kill very nicely as you increase number of effectors to targets. If you express more wild type gene through the lentivirus for RAB27A, they kill even better. And if you put in the patient mutation, they kill slightly less well. And in the setting of IL-six, which is often elevated, everything the killing gets even worse because IL-six lowers things like perforin and granzymes.

And either even simpler assay is this degranulation assay. And all you really do is you take your natural killer cells and you mix them with their target cells that activates the NK cell. They move the cytolytic granules and at the same time they spit out this protein CD107A onto the surface of the cell and you can just measure it by flow. And so going down over time as we overexpress more of the RAB27A, you kind of get maximal CD107A expression at two hours. And at every time point you look, when you overexpress the patient's mutation, they're not as good.

And this is summarized here over multiple experiments, the degree of inhibition you get when you overexpress the patient mutation. So RAB27A the crystal structure had been solved for this protein. And so I turned to one of my colleagues, Mark Walter, who is very good at computer modeling of mutations and a variety of proteins by crystal structure, as well as some of the known binding partners in this case mimics of month thirteen four, which is also in that perforin pathway. And our patients mutation here in fuchsia or purple or pink or whatever color this is, was predicted to disrupt this interaction between RAB27A and its binding partner in the perforin pathway of MUC13.4. And so we first oops sorry, yikes.

See if I can get back to that. All right, so we did this by CoVNP precipitation or Western blot assay where we overexpressed wild type Monk thirteen-four that we epitope tagged with flag and wild type or patient mutation Rev27A that we tagged with another protein. These are all control lanes down here, and then looked for co immunoprecipitation. And at every time point upon activation of these NK cells with their target cells, there was a lot less binding of the patient RAB27A mutation to month thirteen, four, then there was the wild type gene at each of these time points. We also explore this in situ and this is a really cool asset I can happy to tell people about later, but for time sake, it allows you to look at where two proteins within a living cell are close enough to be detected by the assay I.

E. Right next to each other inside the cell. And they show up at these orange dots and this is over time. And so if you overexpress the wild type RAB27A, you get these nice little orange dots where RAB27A and MUNK13.4 are interacting. You also get orange dots if you overexpress the patient mutation, it's just on average about half as many.

So it's capable of getting close enough to bind, it just doesn't do it as well. So it makes it want to look by confocal microscopy at what was going on at the cellular level. And so in green or yellow here we're standing for grandsi and B something in these cytolytic granules. And so when these NK or effector cells bind their target cell, these grandsi containing cytolytic granules polarize or move along actincytic skeleton to this immunologic synapse. And you see that here, you don't see every single amount of granzyme but the bulk of it.

If you overexpress more of the RAB27A wild type gene, you get even better polarization. But if you overexpress the patient mutation, it's not so great. Now this is just one snapshot in time. So MiXi created a lot of work for himself. And so what he did was multiple experiments over a large time course using empty vector control overexpressing wild type or the patient mutation, and then counted 200 cells per condition per time point in a blinding fashion.

So he didn't know what he was counting. And you do this enough times, you come up with this, which is one graph but just an immense amount of work actually in this one graph. And what he showed and on the y axis here, we're just saying the percentage of these NK cells that polarized grandzyme B in this case, and time. And if you just put in the empty lentiviral vector, these NK cells will kind of max out their polarization around 6070% of the cells by about one to two hours at a population level. If you overexpress the wild type rap27A gene, it's a little bit better kind of comes back to the same baseline at two hours.

But the interesting thing was if you overexpress the patient mutation, the percentage of cells at a population level, not at an individual cell level, but at a population level, their ability to polarize is delayed by a few hours. And around this time, this really lovely paper from Misty Jenkins and colleagues in Melbourne, Australia came out. And this is what a lot of us were thinking, but this is kind of like the direct proof or evidence of what we were thinking. And what Misty and her colleagues did was she took human or mouse didn't matter NK cells or CD8 T cells didn't matter, perforin deficient or granzyme deficient didn't matter, but these were fully deficient. So homozygous deficient cells and put them together with their target cell and then kind of camera photography over time.

And what she found was if the NK cell or the CD8 T cell was incapable of killing its target cell because they were perfumed deficient or granzyme deficient, that they spent five times longer engaged with each other, which resulted in increased pro inflammatory cytokines, whether it's gamma interferon TNF or whatever she happened to look at. So it really kind of drove home this point. And so Mingxa saw this and did the same thing here we're doing it with the heterozygous because we have the wild type gene present, we're just overexpressing more of the wild type or the patient mutation and then looking at the percentage of cells that are expressing gamma interferon. And over time you get more gamma interferon with the NK cells binding their target. If you overexpress more of the wild type rAb27A, it doesn't really change anything.

But if you put in the patient mutation, which has this delayed degranulation process or cytolytic movement, you get more gamma interferon. So kind of basically the same thing MISTI was seeing. I had presented this data at a, I think at an ACR meeting and one of my colleagues Fabrizio DiBenedetti from Rome said, we have a 14 year old boy who has the exact same mutation that your 18 year old girl has. In fact, he's had cytokine storm as well. And both our patients' fathers had the same mutation and both our patients' fathers were fine.

Although the Roman patient's father has a chronic ferritin level around 800 for whatever that's worth. But the Roman patient and his father, whether we looked at degranulation or at specific silytic activity, either comparing it to the mother or sister who didn't have the RAB27A mutation or compared to healthy controls at baseline when they were both well, clinically well, their NK cell function was about half normal ish, which is fine most of the time. So that was a lot of work to show that one single mutation in one of these genes probably contributed to disease through a partial dominant negative fashion. This paper came out in 2010 and it caught my attention. It was put out by a bunch of pathologists at the University of Michigan.

And this is getting back to two thousand and nine H1N1 pandemic, which killed a lot of people, including people at the University of Michigan Hospital. And these were patients kind of 30, 40, 50 year old, not a lot of pre existing conditions. So these viruses come along, many of us get exposed, many of us get it. I got it in 2016, actually when I was in Australia because their summers and winters are flipped and I wasn't immunized. But anyway, I was sick as a dog, but I wasn't hospitalized and I didn't die.

So why is it that some people do so poorly? And so I contacted Paul and kind of brokered a deal with our colleagues in Cincinnati who were doing a lot of genetics at the time. And Alexei Grom and Grant Schuillard and colleagues sequenced fourteen of these patients who died from H1N1. And five of them had mutations in list, which is important for the cytolytic granule sorting. And two of those five also had a heterozygous mutation in perforin one.

So this number thirty six percent, you'll see over and over in the literature somewhere between thirty and forty percent of patients who have these secondary forms of HLH often have heterozygous mutations in these same familial HLH genes. In addition to looking at the exome, the bulk of our DNA 97% or something is non exomic and much of it is regulatory in nature. And Grant, Shuart and Cincinnati and colleagues had this systemic JIA patient who just repeatedly would get MAS over and over again. And they knew what genes to look for, but they could never find a mutation by whole exome sequencing. So they finally returned, they went to whole genome sequencing.

And they found this mutation in the first exon of this large protein, Monct thirteen, four, Unc 13 D. And it turns out that your lymphocytes actually express a version of Monct thirteen, four that excludes the first exon. So this actually serves as a promoter slash enhancer region. And I won't go into the details, but basically this patient expressed half as much of that version of this protein. And we showed in pretty good detail that that led to decreased expression through decreased transcription of an NF kappa B binding site.

And so hypomorphic mutations in some of these genes can also contribute to cytokine storm syndromes. This is a model that I think a lot of us favorites, the threshold model of disease. And so here's this threshold over which you can no longer kind of dampen your immune system to the point where you end up getting the cytokine storm out of control and the multi organ failure that follows. So if you are homozygous, let's say perforin deficient, you are already over that threshold. So these infants, it's rare to occur beyond age one, but they typically present within the first days, weeks or months of life with the cytokine storm syndrome or HLH because they're autosomal recessive deficient in this.

There are certain bugs or organisms, strains of influenza and other parts of the world, Leishmaniasis for example, that are very likely to trigger cytokine storms. And maybe you have a heterozygous mutation and you get one of these bugs and you're pushed over the top. Or what we see in rheumatology and other diseases, maybe you have lupus and you've got kind of a state of chronic inflammation, you're not completely controlled or most lupus patients actually present if they have cytokine storm, it's at onset. And maybe they have a heterozygous mutation in one of these perforin pathway genes, they get Epstein Barr or some infection and they get pushed over the edge. So it's like those RAB27A fathers, they're probably hanging out here all the time.

And they just been fortunate enough not to get the wrong organism to push them over that edge. And this is one of I think the silver linings of this pandemic is in addition to learning about cytokine storm syndrome as it relates to COVID-nineteen, patients all over the planet, pediatric and adult ICUs die every single day from cytokine storms from a whole variety of causes and triggers. And we're just, I think beginning to recognize the tip of the iceberg. Our intensivist colleagues have all these acronyms like multi organ dysfunction syndrome, systemic inflammatory response syndrome. My personal favorite is the culture negative sepsis.

They're on their tenth broad spectrum antibiotic. And nowadays where you can even diagnose most viral infections, and nothing grows or nothing is identified and they circle the drain because they're not getting treated for the cytokine storm. And so now we have a whole variety of conditions that can lead to a cytokine storm. So these are the ones we just talked about in the intensive care unit setting. This is what our patients with still disease and lupus and various vasculitides can get.

There are some rare genetic orders and metabolism that can lead to this. Cytokine release that most people now have heard about this. This is from, for example, CAR T cell therapy for treating refractory leukemias and lymphomas. That process itself can trigger a cytokine storm and up to twenty to thirty percent of those patients. Kids and some adults, but mostly kids with primary immune deficiencies, particularly those that make it difficult to clear viruses are at risk for this acquired immune deficiencies, either the virus itself or some of the secondary infections can trigger it that these patients get.

Certain hematologic malignancies such as leukemias and lymphomas are these patients are at high risk for this. Interestingly, cardiac bypass, that whole process of just going on the pump, even ECMO can trigger a cytokine storm. We now know some rare autosomal dominant auto inflammatory diseases like NLRC4 mutations can trigger this. These are these kids that we've been talking about with the genetic form of HLH. Hemorrhagic fever viruses, dengue puts about five hundred thousand people in the hospital every year and across the planet.

And a lot of those patients have a cytokine storm. Cremain Congo, maybe even Ebola. The most notorious bugs or organisms that trigger this are the herpes virus family members, namely Epstein Barr, but CMV, HHV6 and HSV, they can all do it. How do we diagnose it? There are these complicated H score, the HLH four, we develop various ones for specific diseases.

Those are all kind of troublesome and they're great for clinical trials, but practically they're tough. It turns out you may not need anything all that complicated. And this was some of the work that we did showing that just a ratio of the ferritin divided by the ESR could be helpful. Because as you get sicker, the ferritin goes up from the inflammation, the ESR comes down as we talked about before from the fibrinogen consumption. And in fact, you can get really reductionist and we set a specificity of ninety five percent and got an 89% sensitivity for ferritin alone over six twenty seven nanograms per ml.

And that's not that high. Many of these patients are in the 10s or 100s of 1000s. So how do we treat this? And I'm gonna focus on some of the stuff that's being used now to treat COVID-nineteen, particularly anakinra. So this is about a decade's worth of patients that we saw here in Alabama at the Children's Hospital who got an anakinra at some point during their cytokine storm.

And to be fair, we included patients who may have been sick for a month or so, but got anakinra the day or two before they died. And our mortality rate was not great, but it's certainly better than the HLH protocol, for example, the Etoposide protocol where their mortality is more like forty five percent. Anyway, most of the patients, if they died, they were on a variety of other immunosuppressive things. And a lot of this is probably secondary to corticosteroids. Phonchemia is often one of the kind of the end kind of bugs that you get, but you can just die from the shock or the multi organ failure as well.

And getting back to these mutations once again, this one here as TXBP2 in this pathway of these cytolytic granules it's involved in the fusion. And once again, we get this number, we tested thirty eight of these 44 genetically, this number of somewhere between thirty and forty percent, thirty four percent of them had at least one heterozygous several had more than one mutation in the perforin pathway. Unfortunately, if you had an STXBP2 heterozygous mutation that did not bode well for your survival. Other things at presentation, if your platelets are under a 100,000, that's not good either. You're not gonna do well likely.

An obvious one, the longer you're in the hospital, the less likely you're gonna come out alive at the end of this process. And if your serum ferritin doesn't drop substantially like 50% within the first two weeks of hospitalization, you're not gonna do well either. If at any point in time your platelets are low, that doesn't bode well or a couple of your cell lines are down, that's usually a bad prognosticator. Good news is the earlier you start therapy, it seems like a no brainer, it's nice to kind of demonstrate it. The earlier these patients started Anakin or the better their survival.

And for whatever reason, the kids with systemic JAA survived. And either that's because we were more attuned to them or aware of them earlier, or just because it's part of the pathophysiology of that disease that IL-one blockade is gonna be really helpful, we don't know. And so all the kids with systemic GAA survived. There were three kids with malignancy, two of whom actually survived their cytokine storm, but eventually went on to die from other causes. Leukemia or any of the cancer associated cytokine storms are notoriously hard to treat.

The lupus patients and MCTD patients were somewhere in between. Infection alone is not good, particularly if it's Epstein Barr virus, that's a really nasty bug for this condition. This is the way simple for me is good. So we have IL-one beta and we have IL-one alpha and anakinra just really blocks that pathway, right? So we have our own anakinra, right?

We generate recombinant human interleukin one receptor antagonist to mimic what our own body is doing. We're just not making enough of it. And so by giving anakinra, we're just balancing out excess interleukin-one. And I think there's a lot of reasons to particularly like this drug. Number one, it's a common human protein, not that you can give endless amounts of insulin to people, but really all you're doing here is balancing excess IL-one.

It tends to be fast acting. So if they're not getting better within forty eight hours clinically or by labs, you can just kind of rethink. It's got a really short half life. So even if it was harmful, which it tends not to be, it's out of the system before you know it. No one really knows for sure how important IL-one alpha is, IL-one beta is definitely important in this process, actions and has a ginormous therapeutic window.

There's not many drugs you can twenty times the dose and get more benefit without more side effects. In places like West Africa where seventy percent of the patients were dying from Ebola, a large part of that was they couldn't even get IV access in them. And yet you can give this drug subcutaneously. So there's a lot of reasons I like this particular drug. Bina Shakur when she was an adult rheumatology fellow here at UAB contacted Steven Opal and Charles Dinarello and got ahold of data.

So the intensivists they realized the cytokine storm was occurring in some of their sepsis patients. And so they did some pretty good trials, pretty large trials like 1,000 patient trials, and they kind of got to their penultimate trial. And here they're giving forty eight milligrams per kilogram per day of anakinra, it's a ton. And it didn't make anyone sicker in terms of mortality at least. But it was almost identical to placebo.

So thirty percent or so mortality whether you got anakinra or placebo. However, when BETA went back and reanalyze the data using the intensivist definition for multi organ failure, hepatobiliary dysfunction and DIC or coagulopathy, that subset of patients, their mortality was horrible. It was like sixty five percent if they got placebo. However, if they got an anakinra, it almost doubled their survival. So in subsets of patients who even have frank sepsis, if they have features of a cytokine storm, anakinra is likely going to be valuable for them.

Before COVID, Doctor. Chatham and I started a clinical trial to look at Anakinra specifically and secondary HLH or cytokine storms in both kids and adults. And the first five patients we enrolled, some of them had multiple heterozygous mutations in the same perforin pathway genes. It's five out of five, then we were up to nine and six of the nine, including this patient here had one as well as what we now think is a novel cytokine storm gene and dock eight. IL-six blockade, you hear a lot of this in tocilizumab and other agents being used to treat COVID-nineteen.

And once again, this was something that the oncologist learned really quickly, saved lives of the patients with refractory leukemia who are responding to their CAR T cell therapy, but at the same time developing a cytokine storm. IL-one blockade actually works very well too there if not better. Ruxolitinib and other JAK STAT inhibitors, you're gonna block multiple cytokines here, but they are another potential option for treating cytokine storms and are also being clinically trialed now. Emipalumab anti interferon gamma. This is the first FDA approved drug for the familial form of this disease.

So we may not need to use a toposide on all these patients, works pretty well. So in general, if you can get away with using corticosteroids, they may not be the first drug you go to, particularly if you can use something like anakinra, but they work well, they work fast. They've got all the side effects you know about. Cyclosporine, if your kidneys aren't completely shot, is pretty good at making your lymphocytes unhappy. And the CD8 T cells tend to be a problematic cell here.

I tend to avoid jumping right into cytotoxics. I mean, there's a place for them for sure. But not, we're getting a lot more better options now. And we have a whole variety of targeted cytokine approaches. So, and this is kind of the map of PubMed for these disorders, whether it's cytokine storm syndrome, MAS, cytokine release, HLH, not very many here.

This was the first kind of use of VP16 or toposide to treat the familial HLH. The first gene wasn't identified all that long ago, 1999, was perforin. That was the first one reported. This is that case report that I showed you at the very beginning of the talk. Anakinra's first use for MAS.

Now we have FDA approved tocilizumab for cytokine release, IL-six blockade and Emipalumab anti airfarin for primary agent. And now with COVID-nineteen, the number of publications is just astronomical. Many of us, when the reports were first coming out of China about the hospitalized patients with COVID-nineteen, the twenty percent who require hospitalization, many of their features were kind of reminding us of what a cytokine storm is. And so it wasn't rocket science to figure out that many of these patients were having cytokine storms who are hospitalized with COVID-nineteen. And so we have this umbrella term cytokine storm, which I like because it's good for the lay public, it's good for people who may not have heard of it before.

And under that we have various entities which are not all the same and some may be treated slightly differently than others. We have what we call MAS, the cancer associated, the cytokine release, the genetic or familial form and then this large Venn diagram circle of infection, particularly viral associated whether it's influenza, BV, and now we got COVID-nineteen. And ARDS can certainly occur in the setting of these cytokine storms, but it also occurs in the absence of them. So it doesn't completely get covered by the umbrella. And COVID-nineteen is really a unique beast in a lot of ways.

It's the hyperferritinemia and the IL-six elevation is not as extreme as we see in other conditions. It tends to go to the lung first. ARDS is often a late finding in cytokine storm or it's the first thing you see in these patients often. So there's a variety of differences here but I still think it falls under this umbrella. And some of the bad prognosticators, high white blood cell count particularly your neutrophils is bad.

Now in cytokine storm we usually see the opposite but with COVID high neutrophilia is bad for survival. Lymphopenia, these patients almost always are uniformly lymphopenic when they're hospitalized and the worse the lymphopenia, the worse the outcome. IL-six, once again, are not screaming high levels, they're elevated, but they're not that high. And the ferritin levels are not that high either. But C reactive protein can be very high and that's not good for survival.

Procalcitonin troponins are elevated. And then most of these patients have some kind of coagulopathy. And another difference with COVID-nineteen many of our cytokine storms, we have a lot of bleeding problems, can we get caught in too. But these patients really have a high tendency to clot. And then we get to treatments.

And so we're still waiting on the clinical trials to be published. There's some of the IL-six blockade trials have come out, but there's also these case series that just keep coming out because people are desperate. Whether it's Northern Italy or New York City or Philadelphia, you try to put everyone in a trial but you may not have enough trial space or the patient refuses, or you're just tired of watching people die. And so here's one hundred patients who got tocilizumab and this is on the y axis just a marker respiratory severity, whether they're in the ICU or on the general ward, they overall improved. This is twenty nine patients who got anakinra and they compared it to a historical control at the same site that was just treated the two weeks prior to them using anakinra.

And I could tell you there are other ones coming down the pipe because I reviewed them. And they're even more dramatic than this in terms of survival benefit. And so these papers will keep coming out while we're waiting in the clinical trials. And even corticosteroids, right? The World Health Organization has pretty much stamped that you should not use corticosteroids based on prior SARS and MERS and even influenza outbreaks.

But in the right population, given at the right dose at the right time, I. E. Not during the early viral phase, during the hyperinflammatory phase, cytokine storm can be very well treated COVID-nineteen irrespective with corticosteroids. And there aren't a whole lot of trials looking at that. There are some, but these case series come out.

I can tell you at Temple in Philadelphia in the adult side, they've treated fifteen hundred hospitalized patients with steroids the instant they come in the hospital. They all have respiratory stress, they all have fevers, many of them have laboratory features. It's like they all get moderate dose corticosteroids and they have a tough population. They have like seventy percent with multiple comorbid conditions. It's about fifty percent African American, thirty percent Hispanic.

People that you wouldn't expect necessarily do well, just based on others reporting and their mortality is under seven percent. So there's something to be said there, despite the fact that's not a trial. And I'm happy to take questions.

Okay, Randy, thank you very much for that. Let's see, I'm gonna get back on screen myself. There we go. So I wanna remind our audience, please use the Q and A button at the bottom if you have a question for Doctor. Cron.

Random, we got a bunch of questions. Given that, you you showed the plausibility of multiple interventions, how do you choose other than preference of using Anakin or an IL-one inhibitor versus an IL-six inhibitor versus Gamafant, which is emapalumab gamma ferrone inhibitor. Is it a matter of your experience? Like in the past it was the hem onc people using etoposide and everybody else using cyclosporine or is it a timing issue at the stage of the game or driven by a biomarker? What do you think?

Yeah, I think nobody knows. I mean, I would say one of the reasons I'll six got a lot of attention early is the Chinese don't have access to IL-one blockade. So they had access to IL-six blockade. The other thing is IL-six is pretty easy to measure in the blood. IL-one not so much.

So it's kind of a no brainer that IL-six got a lot of popularity. And then just my general take on reviewing manuscripts and reading what's out there is that IL-six blockade tends to work mostly. But I've been more impressed by the IL-one blockade. Now I have biases against that anyway. But I just, if I try to take my biases out of it, I think IL-one blockade will probably be better off for most people.

And there may be some patients who benefit from some and not others. And I don't think we have any idea at this point how to personalize that. The other thing is this Anakinra as I touted before is an incredibly safe drug. And so I tend to go to that first. Gamma Phant is probably gonna be a good drug.

Baricitinib, ruxolitinib, your JAX stat inhibitors are probably gonna be pretty good. The thing that you gotta be careful with blocking gamma whether it's through Emipalumab or through JAX stat is gamma is probably important for helping you get rid of the virus. Now, you got to weigh that against the patient dying from the cytokine storm. So I think my gut feeling is that all of these interventions are gonna be valuable. And in fact, way more valuable than getting out of the hospital four days sooner with remdesivir for example.

I mean, we should use antivirals, don't get me wrong, but those aren't gonna be the game changers. These drugs are gonna be the game changers.

Do you think these people at the time they get cytokine storm have generally gone through their infection or are they still at the height of their viremia? And this is now the downstream sequelae that's out of control.

I mean, ultimately, hopefully data will tell us that not a lot of people are actually culturing virus because you need a bill three facility or plus to do that. But yeah, I think by the time you reach the hospital and are sick enough to be hospitalized and I know that's a moving target just because some places are overwhelmed for example. But if you're sick enough to be hospitalized with respiratory distress and fever, you're probably already at that stage. And I'm not saying the virus is gone but I think the viral damage relative to the cytokine storm damages is less at that point. So yeah, I think by the time you walk through the door and that's the time you wanna get these patients.

So start getting your ferritins, your CRP, your IL-six whatever you're ordering LDH D dimers. The instant people walk through the door because if they've got respiratory distress and they've got elevated markers of inflammation and a fever and COVID-nineteen or ground glass opacities for the twenty percent who test COVID-nineteen negative who currently have COVID-nineteen, that's when you wanna intervene. That's not to say you should give up on people once they make it to the ICU or once they get invasive mechanical into ventilation or intubation. But your bang for your buck is gonna be, like I said earlier, the earlier you start, the better you're gonna do.

So with the tender trial in systemic JAA with tocilizumab and in the package insert, kids, you can use up to milligrams per kilogram to treat this and do it repeatedly. What's the dosing for anakinra? You know, it's a short half life drug, six hours. In Still's disease, I say take it at night because it's a circadian cytokine. But in these conditions, it's not a circadian cytokine.

It's all over the place all the time. So how do you choose the dose Anakinra here?

So the clinical trial that Doctor. Chapman and I started a while back pre COVID we kind of settled on a hundred milligrams Q6, some parts out of practicality, some parts around ten per kilo per day ish. I know people who go higher, I know people go less and of the at least the case series that I've read and or reviewed, people are all over the place, but they seem to work whether you're given two hundred milligrams a day and adult or four hundred milligrams a day. That's kind of the ballpark range that I see. Some people are just to save it because they can't get enough of it are using like a hundred milligrams a day.

If I had my druthers, I would probably start at a 100 Q six.

Is Castleman's disease a pre cytokine storm or a cytokine storm?

I think it's a cytokine storm at some point. Yeah, those patients get really sick and they benefit from this kind of therapy as well.

Any ideas as to why these patients get a coagulopathy?

I think there's the virus attacks the endothelium. I think that's a large part of it. I mean, they also have liver dysfunction and other features. But I think one of the reasons this virus is so different is it really goes after the endothelium. That's my best guess.

It's just hunch work, but I think that's it's probably involved.

So Doctor. Vu asked a question about measuring IL-six and IL-one levels to maybe guide you here. Is that worth doing? And then I'll ask you to follow-up to that. What is your go to lab, that you really hang your hat on here?

Is it, a lot of people hang your hat on ferritin. I like to hang my hat on Aldolase as a biomarker in this condition, but measuring those levels and what's your go to lab?

Yeah, so I don't measure cytokines acutely. I probably should. But as much as I'm a scientist as a clinician I'm always amazed by anecdote and we can think ourselves into corners sometimes and we measure things or oh, well then we should target that. But like I said, IL-one it's almost impossible to get a good measurement on. So if you're gonna base it on that you would never use an IL-one blocker for example.

IL-six I think is valuable track in these patients if you can get it because it does actually predict mortality in some of these studies. But it's not gonna change what I would do personally. I mean, I would just personally, people know my bias. Would give anakinra. But I also be fast to not to stop it and try something else if that's not working.

But the go to labs, I think, and this has worked from others that have just been reported. The LDH, the D dimer and the CRP are some of the ones that people really say wake you up to how sick these patients are. The ferritin tends to not be that high. It's some patients it's even normal and they have a cytokine storm. So it's valuable I think to track it but I wouldn't base my decisions on it.

But I think the LDH, the D dimer, the CRP is the ones that keep hearing repeatedly really predict how poorly these patients are doing. They don't tend and lymphopenia, they tend to be very lymphopenia, but that's gonna be all of them. But the thrombocytopenia actually is a pretty late finding too. So if they're thrombocytopenia upfront, I'd be very worried about them.

So I wanna underscore the ferritin. I think many rheumatologists have an overestimation of the importance of ferritin. And when you see 20,000 ferritin, yeah, it certainly helps, but you don't often see 20,000 ferritin even in these very sick patients. And I think I agree with you. We're working on those other working off those other labs maybe far better and lead to earlier recognition if you're waiting for the ferritin it could get in trouble.

Yeah. Mean, if it was my parent or me or my wife or whatever I mean, if they had a CRP and they had COVID and they're sick enough to be in the hospital, that's enough for me. Other people depending on what they're doing, if it's a clinical trial that takes everything out of the play because then you have actually hard criteria. But for patients who are not in clinical trials, it wouldn't take much for me to start treating their cytokine storm.

So when do you start to throw these drugs into the mix? At what point does it become a cytokine storm? And I'm thinking about, you know, patients who just get sick with sepsis and they get DIC and they start getting multi organ things. Are those people who are probably having a cytokine storm are just not labeled as such? Again, at what point do you now turn on the cytokine storm light and rush in with all these new drugs?

Is a well defined point here or do you gotta be thinking it to act on it early?

I mean, you just go for sepsis and start with that first. I think there is a significant subset of those patients. Certainly not all sepsis patients get a cytokine storm that could be amenable to these therapies. The other thing is that some of our smarter intensivists use is pheresis. I mean, does a lot.

It does a lot of things. It's removing all the evil humors. But I think for COVID, if you're I'll say it again. If you're sick enough to be hospitalized, you have a fever, you got respiratory distress and just a few lab markers of inflammation. Think that's the time you wanna attack.

And when you have to weigh benefit to side effect ratio. And for anakinra that ratio is huge. So I have very little qualms about starting anakinra even if it doesn't do anything it's unlikely to harm the patient. You mentioned I would say very soon after hospitalization, prior to intubation, prior to ICU, prior to shock would be I think your ideal timing.

You made the statement earlier, why do some patients and why does, do so poorly with this? You could say the same thing about COVID. Why do some COVID patients do so poorly and get this? If our patient, whoever that is, COVID or not gets cytokine storm, MAS, should we be testing for one of these perforin or HLH genes? And if you did find it, what would you do about it?

Yeah, so, I mean, that's being done. People are doing whole genome sequencing. We're doing it here. We got a small clinical trial of Anakinra that's getting ramped up and ready to go. And we'll do whole genome on them.

And people are doing it all over the place. The NIH everybody's doing whole genome. My guess is because it shares enough features with cytokine storm syndrome that we'll find some patients who have heterozygous mutations in these same genes. But I guess we're gonna my guess is we're also gonna find other genes we weren't even thinking about. And so what do you do with that knowledge?

It's not gonna it's probably in the short term, it's moot. But in the long term, and this is kind of pie in the sky. And I don't know if this will ever happen for a variety of reasons. But we have these one offs, right? We get Tay Sachs and sickle cell and all these things that we screen our infants for.

And that's really important. But it may be at some point we just get whole genome sequencing when you're born, right? And then if you end up having a perforin which and just be and I'll make this point too. Just because you have a mutation doesn't prove that it's doing anything. And so that's why we spent all that time doing it for that one single mutation.

Now you can have there are algorithms that might suggest it but really until you take it back to the lab, it's hard to know. But anyway, if I personally had a known mutation in one of these genes that I thought was disruptive, I would have a bracelet, a medical bracelet that said, I get a fever and I'm hospitalized, I don't even care if you know why I'm in the hospital. But if I got a fever and I'm hospitalized, start working me up and or start shoving Anakin or end of my.

Do you think somatic mutations or mosaicism are responsible for many of these?

Because the sooner you get to this, the better off you're gonna be. I missed the first part of that, Jack, sorry.

Do you think somatic mutations or mosaicism is responsible for some of these genetic problems?

I haven't seen that. I'm not saying it doesn't happen but I haven't seen reports along those lines.

Okay. If this was a lupus case of MAS would IL-one blockade be still preferable over IL-six or cyclosporine? I don't

know if it's preferable, once again, IL-one is safe. And Doctor. Chatham can give you a lot of anecdote about the benefits of IL-one blockade and lupus patients with cytokine storm. It's nice to have options. It's also confusing because we don't know what's best for what patients but once again, my own biases go with the safest and one that we have the most experience with.

So this pediatric syndrome that we're seeing this pediatric multi system inflammatory syndrome or PMIS or MSIC or something like that, is this part of cytokine storm or is this, an offshoot related to the DIC and the microthrombotic conditions that we're seeing?

Yeah, this virus just keeps throwing curve balls at us. And it's fascinating. These kids kind of showed up about four to six weeks after the adults in the same communities were in the hospital with florid COVID pneumonia. And most people at this point think that what the kids are getting and actually some adults get this too is a post infectious process. And it is a cytokine storm for many of these kids.

And some of them have Kawasaki disease like features which Kawasaki should probably be a syndrome anyway even before COVID but many of them don't. A lot of them go into shock. Fortunately, very few of these kids have died from this. Some have but very few in general have But it seems to be a cytokine storm that's a post infectious process. And so people are using IVIG because they're worried about coronary aneurysms and they're a subset of kids who get them but they tend to resolve actually with therapy.

But most of them are refractory to IVIG. So they end up getting corticosteroids and or anakinra.

Two more, two last quick questions. One, do we need to show the hemophagocytosis and bone marrow biopsy to make these diagnoses or not?

No, it's present sixty percent of the time it can be present in other conditions. I find it helpful if it's there, if it's not, it doesn't dissuade me.

And I got a case here of someone who had COVID has recovered and now has chronic urticaria that's new. Does that make any sense to you?

No, but this virus does a lot of stuff and give it also gives us like chill blames like feature to certain people. This virus does a lot of crazy stuff. Lot of patients actually get a vasculitis with this virus. So I'll leave it open to anything.

And then how do you know when to stop treatment for cytokine storm or MAS? My experience has been that the hard part is getting into the game with aggressive therapy. And as soon as they start to revert, now they're on the downward slope of getting better, and it's sort of an automatic thing. But how long do you usually have to continue therapy with either IL-one inhibition or etoposide or cyclosporine or even JAK inhibitors? Is

it that

18 what?

That 18 year old that I presented with the RAB27A mutation, I took a year and a half to taper off because I was really in my stage of taking care of these kids. So I got the first drug I got rid of in her and most of my patients were the steroids. Then if I can taper them off the cyclosporine if they're on that too, I get them off that. And the last drug I'm gonna taper is anti camera because there's other than the pain of the injection. We don't see a whole lot of downside to it.

But you can stop it much sooner than in a lot of these cases. But there's great things to follow both clinically and lab wise. It's pretty easy to track.

Well, Randy, this is a fabulous syndrome that you've shed a lot of light on that really helped us. You working both in the lab and in the clinic has been a blessing to us as clinicians. So we appreciate your insight and guidance here. Thanks so much for being here tonight and hopefully we can get you back again when we start the series again in the fall.

All right, well thanks again for having me. Always enjoy talking on this topic.

I wanna remind everyone next week is our final week before we hit the summer and that's gonna be John K talking about biosimilars and what's happening there. Have a good week. Take care of yourselves. Goodbye.