publication date: Jun. 12, 2020
Conversation with The Cancer Letter
BTK inhibitors improve COVID-19 outcomes by targeting upstream switch for inflammation, early data suggest
Wyndham H. Wilson, MD, PhD,
Head, Lymphoma Therapeutics Section;
Senior investigator, Lymphoid Malignancies Branch;
Center for Cancer Research, National Cancer Institute
Louis M. Staudt, MD, PhD
Chief, Lymphoid Malignancies Branch;
NIH Distinguished Investigator;
Director, Center for Cancer Genomics,
Center for Cancer Research, National Cancer Institute
This story is part of The Cancer Letter’s ongoing coverage of COVID-19’s impact on oncology. A full list of our coverage, as well as the latest meeting cancellations, is available here.
Pondering hyperinflammation related to SARS-CoV-2, two NCI scientists had a hunch:
“It just dawned upon us, ‘Hey, we know how to block inflammation,’” recalls Louis Staudt, chief of NCI’s Lymphoid Malignancies Branch and director of the Center for Cancer Genomics.
Staudt and colleague Wyndham Wilson, head of the Lymphoma Therapeutics Section in NCI’s Lymphoid Malignancies Branch, realized that they were staring at a potential role for Bruton tyrosine kinase inhibitors as a treatment for COVID-19-related cytokine release syndrome.
Staudt and Wilson have been working on BTK inhibitors, a class of lymphoma drugs, for over a decade before the COVID-19 pandemic hit.
The investigators immediately got in touch with AstraZeneca, the sponsor of the BTK inhibitor acalabrutinib (Calquence), and set out to assemble a multidisciplinary team of researchers to test the lymphoma drug in patients with severe manifestations of COVID-19.
Their findings were published June 5 in Science Immunology.
In a small study that enrolled 19 patients, the team of researchers found that the drug significantly improved oxygenation over a 10 to 14-day treatment course:
Eight out of 11 patients (72.7%) that required supplemental oxygen had been discharged on room air,
Four out of 8 patients (50%) no longer needed mechanical ventilation, with two out of the 8 patients (25%) discharged on room air.
Oxygenation in these patients often improved within one to three days, and no discernable toxicity was observed. Measures of inflammation—C-reactive protein and IL-6—normalized quickly in most patients, as did lymphopenia, in correlation with improved oxygenation.
“This was definitely a situation in which chance favored the prepared mind, as Louis Pasteur said. Wyndham and I had been just chatting, as we do, about medical things, and, obviously, we were chatting about COVID-19 and the inflammation that was getting people in serious medical trouble,” Staudt, co-senior author of the study said to The Cancer Letter.
“We were struck by the biphasic clinical course in which some people infected with SARS-CoV-2 get fever and then get better, but other people, after initially getting better for a while, get this very severe lung inflammation,” Staudt said to The Cancer Letter. “It was our prior experience with these drugs in cancer that led us to jump to realization that they might be effective in COVID-19.”
Acalabrutinib resolves pulmonary inflammation in severe COVID-19. – Source: NCI
AstraZeneca has initiated a randomized phase II/III study in 50 to 100 sites around the world. Another company, BeiGene is testing zanabrutinib, its version of a BTK inhibitor, in smaller studies.
“Despite this being a small study, the rapid and consistent improvement in these parameters in many patients and temporally associated with starting acalabrutinib appears quite convincing. Of course, our findings require confirmation in a randomized trial, which has been launched,” Wilson, co-senior author of the study, said to The Cancer Letter.
“Once the safety component is completed—and we are hoping it will be completed in the next several weeks—[the AstraZeneca study] will then be expanded to a phase III double-blind randomized study involving around 500 cases,” Wilson said. “Endpoints will be looking for improvement in oxygenation and clinical outcomes and, of course, inflammatory proteins. That will be a multi-national study. This is an international study.”
Wilson and Staudt’s research on BTK inhibitors grew out of their work on cancer patients that were getting fungal infections.
“This turned out to be quite specific to patients who had received high dose steroids for prolonged periods of time in our study of primary CNS lymphoma,” Wilson said. “So, that observation led us to look at the effects of BTK inhibition in the innate immune system, as that is primarily responsible for control of fungal infections. That is neutrophils and macrophages.”
Based on these early data, BTK inhibitors are expected to reduce or cut off the downstream production of inflammatory cytokines—e.g. IL-6 and IL-1—that play a major role in causing Acute Respiratory Distress Syndrome, the hyperinflammatory disease that makes COVID-19 especially deadly.
This effect will be studied in the late-stage trials that will enroll patients with severe COVID-19.
“Just metaphorically, treatment with tocilizumab and anakinra is like cutting the limbs off a tree, but treatment with acalabrutinib is like cutting the tree down from the trunk,” Staudt said. “By analyzing blood samples from patients with COVID-19, we showed that the target of acalabrutinib, BTK, is active in the monocyte-macrophage lineage, which are innate immune cells that are abundantly present in the lungs of these patients.
“We also showed that in the blood, it is specifically the monocytes and not the other immune cells that have activated BTK and are making IL-6, an important cytokine in the cytokine storm that contributes to hyperinflammation in COVID-19.”
Large phase III trials of IL-6 inhibitors—tocilizumab, sarilumab, and siltuximab—are underway. Under an NCI protocol, tocilizumab is available to cancer patients at institutions that are not participating in Genentech’s phase III trial of the drug. The institute is also creating a registry of cancer patients with COVID-19 at over 1,000 sites in the U.S. (The Cancer Letter, March 27, April 10, April 17, 2020).
Patients with comorbid conditions appear to develop severe COVID-19, because their immune systems are chronically activated, leading to a rapid and overwhelming inflammatory response to SARS-CoV-2 infection, the researchers say.
“One area we have not discussed is the link between macrophages and the risk factors associated with severe COVID-19,” Wilson said. “We have all seen the data that shows that patients at highest risk of severe symptoms are enriched by those with obesity, hypertension, and diabetes.”
The phenomenon, in which the immune system is “marked” by previous exposures to foreign pathogens or medical conditions, is fairly well understood, Staudt said.
“It’s not a genetic change in immune cells, but rather an epigenetic change. This phenomenon has been called ‘trained immunity’. The immune cells are ‘trained’ by their previous exposures,” Staudt said. “In the setting of COVID-19, the immune system may be trained by various comorbid medical conditions that chronically stimulate macrophages to change their epigenetic state. So, these trained immune cells may be on a hair trigger when something like SARS-CoV-2 comes around.
“That’s a hypothesis. It’s going to be fascinating to carefully study the function of macrophages in patients that have, or do not have, these comorbidities, and that’s the science we’ll be doing in the course of the recently launched randomized phase II/III CALAVI clinical trial.”
Staudt and Wilson spoke with Matthew Ong, associate editor of The Cancer Letter.
At first glance, your study comes across as very small, with only preliminary data. Even so, it seems to suggest significant improvement in outcomes. Could you describe your findings and what you’ve learned about using BTK inhibitors in patients with COVID-19?
Let me start and say that it is a small series, but I think what is remarkable about this—particularly given that 18 of the 19 patients actually had worsening oxygen requirements when they started acalabrutinib—is that within several days of treatment with acalabrutinib, most patients showed a reduction in the C-reactive protein, improvement often to normal [levels] in the absolute lymphocyte count, and improvement in oxygenation.
Most of these patients were also receiving relatively high oxygen support, including high-flow oxygen or mechanical ventilation.
By the end of the study, 8 of 11 patients who were on supplemental oxygen no longer required oxygen and were discharged from the hospital. We also looked at the outcome after four more weeks of observation to assess if there were any recurrences, which we did not see.
Now, the group on ventilators was quite ill, in general. Some had been on a ventilator for extended periods and had multi-organ failure. But even within that group, there were four patients that showed a similar pattern of improved CRP, absolute lymphocyte count, and in fact, coming off the vent.
So, despite this being a small study, the rapid and consistent improvement in these parameters in many patients and temporally associated with starting acalabrutinib appears quite convincing. Of course, our findings require confirmation in a randomized trial, which has been launched.
When you see such a trial that is small and not controlled, you have to say, “Why do I believe this?”—and rightly so.
As Wyndham just said, the rapid effect of giving acalabrutinib on oxygenation and inflammation is a key finding, possibly suggesting causation. However, the additional and important concept is that we have a hypothesis for the action of acalabrutinib in COVID-19 and correlative evidence supporting the hypothesis.
By analyzing blood samples from patients with COVID-19, we showed that the target of acalabrutinib, BTK, is active in the monocyte-macrophage lineage, which are innate immune cells that are abundantly present in the lungs of these patients.
We also showed that in the blood, it is specifically the monocytes and not the other immune cells that have activated BTK and are making IL-6, an important cytokine in the cytokine storm that contributes to hyperinflammation in COVID-19.
So, when you put our clinical findings together with a mechanism of action that’s plausible, I think and absolutely hope that when we complete the right study from a medical research point of view—a randomized controlled trial—we will find that acalabrutinib does indeed ameliorate severe COVID-19
I think what makes our study stand apart from other studies is that we are targeting what we believe is the upstream instigator of inflammation. Studies of monoclonal antibodies that are directed at single cytokines, such as the IL-6 receptor and IL-1beta, are only targeting one of many inflammatory mediators. For example, in the case of tocilizumab, a monoclonal antibody against the IL-6 receptor, the CRP falls and oxygenation improves somewhat.
However, the blood concentrations of IL-6 actually rise, indicating that the source of the cytokine has not been quieted. It is likely the IL-6 rises due to a feedback mechanism, and this feedback may actually lead to increases in other cytokines that could obviate some of the benefits of blocking IL-6 action.
Of course, this needs to be explored, but does illustrate potential issues with targeting single cytokines among the many that are increased in severe COVID-19.
Notably, inhibition of BTK also targets neutrophils as well. Indeed, these cells also produce cytokines and are replete in pathological specimens from the lungs of patients with COVID-19.
Just metaphorically, treatment with tocilizumab and anakinra is like cutting the limbs off a tree, but treatment with acalabrutinib is like cutting the tree down from the trunk.
The word “upstream” did come to mind, and I was just thinking of asking you about that.
It’s very, very similar to the way we think about cancer. In cancer, we don’t want to chase every small subclonal genetic mutation in a tumor. We want to get at the trunk of the tree—the oncogenic mechanisms that are engaged at the beginning of the cancer process.
The therapies that target those primary mechanisms are more likely to be successful than those that target secondary oncogenic processes. So, I think this is quite related to how one thinks about cancer therapy.
I think that from a practical point of view, we should recognize that acalabrutinib and other BTK inhibitors are small molecules that are oral and relatively inexpensive compared to monoclonal antibodies.
Generally, BTK inhibitors also have a favorable safety profile, which is particularly true for acalabrutinib given its specificity for BTK and relatively low “off-target” effects. Indeed, we didn’t see any safety issues, which was not unexpected given we only administered it for 10 to 14 days. You may know in cancer that this drug was developed for B-cell tumors, where it’s given for months to years and generally is well tolerated.
Could you explain the mechanistic relationship between acalabrutinib, the BTK inhibitor, and the reduction in production of IL-6 and other cytokines?
Matt, we spent a lot of time on Figure 1 of the paper. It’s a drawing of a macrophage. What it shows is the presumed mechanism, namely that single strand viral RNA activates Toll-like receptors within the macrophage. These receptors activate the BTK kinase that then signals down to NF-kappa B, a transcription factor that turns on many of the cytokines, including IL-6, IL-12 and many others.
Separately, a fascinating new insight is that BTK activates the inflammasome, which is a molecular machine that is necessary to produce active interleukin-1 beta, another critical cytokine in the COVID-19 cytokine storm. BTK is directly associated with one of the subunits of the inflammasome, NLRP3, and induces that subunit to organize into large inflammasome structures that process and release interleukin-1 beta.
Our model is that BTK is ideally positioned to block all the signaling mechanisms that are contributing to the increased transcription of cytokines, or in the case of IL-1 beta, its processing into its secreted form.
That was a long explanation, but the picture in Figure 1 is worth a thousand words.
Model of macrophage activation in severe COVID-19 leading to a cytokine storm. – Source: NCI
I was following. This might be apples and oranges, but how do BTK inhibitors compare with some of the broad immunosuppressive agents that have been tested in patients with COVID-19?
I think the key here is to examine the targets of these other drugs and to consider what is known about cytokine release syndromes in COVID-19. In many ways, the hyperinflammatory response looks quite similar to what is known as macrophage activation syndrome.
Indeed, we hypothesized some years ago that BTK inhibitors might be useful in such syndromes, and even treated a patient with HLH and observed a rapid reversal of the inflammation.
Apart from the downstream targeting of individual cytokines I discussed earlier, others have proposed drugs that target JAK/STAT signaling that is important for cytokine production.
Baricitinib targets JAK1 and JAK2 and is under investigation. While it is likely to quell some of the cytokine storm, it also inhibits T cell activation and is associated with viral reactivation and clots. It also has a complex and dual role in macrophage polarization to M1 (activate-inflammation) and M2 (inhibitory). Thus, in the setting of COVID-19, its effect on macrophages will be difficult to predict without further study.
We believe we have reasonable evidence for our model, which hypothesizes that the hyperinflammatory milieu of COVID-19 is dependent on macrophage-monocytes, based on two lines of evidence.
First, we show that inhibition of BTK is associated with a rapid decline in inflammatory proteins, including CRP and IL-6, and second, we show activation of BTK in blood monocytes, validating the activation of our target.
JAK, as in Janus kinase?
Yes, Janus kinase. As I discussed earlier, JAK-STAT signaling is critical to many immune cells, including T cells and macrophages, and inhibition may have unpredictable effects.
And, of course, disabling T cell function could have adverse effects on viral control as observed when used in rheumatoid arthritis. And, of course, this all raises the question of whether the cytokine storm in COVID-19 is driven by T cells.
Matt, using the previous metaphor, JAK inhibitors also cut off limbs of the tree. They would block IL-6 signaling, and signaling by other JAK-dependent cytokine receptors, so they would act downstream of the cytokine storm.
They would have no demonstrable effect on the macrophage, at least in the model that we have involving the Toll-like receptors and the inflammasome.
Just to say the obvious, steroids have not proven effective in previous coronavirus infections like SARS, and the World Health Organization does not recommend the use of steroids in SARS-CoV-2 infections.
We all know that the recent studies suggest that hydroxychloroquine is certainly not beneficial and may get some people in trouble.
I think the beauty of BTK inhibitors, which we have been studying because of their specificity for B-cell receptor signaling, is really that they only target an activated state of certain immune cells but don’t hit more basic cellular functions.
I would say that the clinical experience with the safety of BTK inhibitors in thousands and thousands of patients suggests that they do not impair most normal immune responses, since treated patients only rarely get opportunistic infections. Rather, pathological situations involving intense activation of macrophages may be the setting in which BTK inhibitors could be helpful.
One area we have not discussed is the link between macrophages and the risk factors associated with severe COVID-19. We have all seen the data that shows that patients at highest risk of severe symptoms are enriched by those with obesity, hypertension, and diabetes.
It is remarkable that these conditions are associated with macrophages where it has been shown they have activation of the NLRP3 inflammasone and have higher blood levels of CRP and IL-6. It is also known that fatty acids, via toll-like receptors, activate the inflammasone and likely play a role in the heightened inflammation found in these clinical conditions.
We have hypothesized that these clinical risk factors lead to a higher set point, if you will, in these patients and thus their macrophages are more sensitive to viral triggers via TLR receptors and development of a pathological hyperinflammatory state.
So, basically their immune response is activated in a way that predisposes these high-risk patients to developing severe COVID with ARDS?
That’s what we hypothesize. It’s actually a fairly well-understood phenomenon in which the immune system is “marked” by its previous exposures to either infectious agents or to other medical conditions. It’s not a genetic change in immune cells, but rather an epigenetic change.
This phenomenon has been called “trained immunity.” The immune cells are “trained” by their previous exposures. In the setting of COVID-19, the immune system may be trained by various comorbid medical conditions that chronically stimulate macrophages to change their epigenetic state. So, these trained immune cells may be on a hair trigger when something like SARS-CoV-2 comes around.
That’s a hypothesis. It’s going to be fascinating to carefully study the function of macrophages in patients that have, or do not have, these comorbidities, and that’s the science we’ll be doing in the course of the recently launched randomized phase II/III CALAVI clinical trial.
Speaking of which, what further studies are underway, and what are the next steps for studying BTK inhibitors in COVID-19 patients?
We engaged AstraZeneca, the manufacturer of this drug, very early on. They have now launched a randomized phase II/III study. The first component is a randomized phase II of 60 patients, which is an open label study in severe COVID-19 for safety purposes.
Once the safety component is completed—and we are hoping it will be completed in the next several weeks—it will then be expanded to a phase III double-blind randomized study involving around 500 cases.
Endpoints will be looking for improvement in oxygenation and clinical outcomes and, of course, inflammatory proteins. That will be a multi-national study. This is an international study.
Now, there are other smaller studies of BTK inhibitors.
BeiGene has zanabrutinib as well.
Are you looking at a minimum of about two months from now before preliminary results can be analyzed?
I would say two months would be a reasonable timeline. Correct.
AstraZeneca is doing a full-court press on this and they’re going to open the trial in 50 to 100 sites around the world.
How do you plan to test the hypothesis we discussed earlier, to determine whether patient populations with hair-trigger immune systems would have a significant response to BTK inhibitors?
Well, there is a sub-study that’s being run out of our group by Mark Roschewski and Mihalis Lionakis from NIAID where multiple translational endpoints will be explored.
Where possible, blood cells will be collected and analyzed for monocyte activation via phospho-BTK, IL-6 and response to TLR stimulation in multiple settings including patients who developed COVID-19 with few symptoms and those with severe symptoms and during and after recovery from the infection. We will also be looking at a host of cytokines and also gene expression profiling. Lou, do you want to comment?
Yes, I think the truly big picture here is that, unfortunately, we have a situation where the entire human population is at risk of being exposed to a single infectious agent. Therefore, we have an opportunity to observe heterogeneity among human beings with respect to how their immune systems respond to the very same foreign pathogen.
So, we can directly get at the question of whether obesity affects the type of immune response that you mount. And the same question can be addressed in the setting of other COVID-19 comorbid conditions, such as diabetes, hypertension, and atherosclerosis. We will do those studies with our colleagues in NIAID, especially Mihalis Lionakis.
We will go well beyond just measuring cytokine levels. There will be an elaborate single-cell RNA-seq component to look at all the immune cells in the peripheral blood. There’s going to be multiplex proteomic methods used as well to phenotype these immune cells. This will be a nice joint NIH effort, as has already been the case during our exploratory clinical study.
That’s a great segue into NCI’s role in viral pandemics. How have decades of federally-funded work in cancer research led us to a point at which we can reliably use what we’ve learned to inform research into COVID-19?
You bring up a very good point. Lou and I have been working on BTK inhibitors for over a decade. During our work, we noticed that some patients were getting fungal infections. This turned out to be quite specific to patients who had received high dose steroids for prolonged periods of time in our study of primary CNS lymphoma.
So, that observation led us to look at the effects of BTK inhibition in the innate immune system, as that is primarily responsible for control of fungal infections. That is neutrophils and macrophages.
While that was not the primary focus of our work, obviously, we collaborated with Mihalis Lionakis in NIAID where he showed that inhibition of BTK in a murine model promoted aspergillus infections and it was due to inhibition of the innate immune system.
Thus, it was our understanding of BTK’s normal role in immunity that led us to our hypothesis in COVID-19 inflammation.
I would just add that this was definitely a situation in which chance favored the prepared mind, as Louis Pasteur said. Wyndham and I had been just chatting, as we do, about medical things, and, obviously, we were chatting about COVID-19 and the inflammation that was getting people in serious medical trouble.
We were struck by the biphasic clinical course in which some people infected with SARS-CoV-2 get fever and then get better, but other people, after initially getting better for a while, get this very severe lung inflammation.
It finally dawned on us that we know how to block inflammation with the BTK inhibitors that we have studied for over 10 years in lymphoma. So, it was our prior experience with these drugs in cancer that led us to jump to realization that they might be effective in COVID-19, which is supported by our small clinical study and we hope and expect will be confirmed in the ongoing randomized trial.
The other thing I would like to say, especially for The Cancer Letter, is that cancer is a significant risk factor for getting bad COVID. Though more work is needed, a paper from Chinese researchers reported that patients with blood cancers and lung cancers had the highest relative risk for developing severe COVID.
That really resonated with me, because the blood cancers are obviously derived from the immune system, and they consequently can secrete cytokines and interact with normal immune cells in ways that may train the immune system, as we were discussing earlier. That may be why SARS-CoV-2 infection triggers a hyperinflammatory response in these patients. This needs to be studied fully.
The National Cancer Institute itself, as you may know, is going to start a registry of patients with cancer who get COVID-19, and follow them over several years to see, number one, how they do with the COVID-19 and number two, see how they do with their cancer, having had COVID-19.
I think that’s an interesting open question.