Sickle Cell Disease (SCD), iNKT Cells, and Regadenoson Infusion

Abstract

A humanized murine sickle cell–disease (SCD) model (NY1DD) has been used to study ischemia/reperfusion injury (IRI) in sickle cell anemia, and iNKT cells (a very small subset of murine and human T cells) have been found to instigate such injury in this model. Furthermore, levels of activated iNKT cells are generally elevated in the circulation of patients with SCD. Because activated iNKT cells are rich in adenosine A2A receptors which, when agonized, down-regulate the inflammatory cytokine expression that characterizes the cells, we have conducted a phase 1 trial of a constant infusion of low-dose regadenoson (an adenosine analogue with high A2A receptor specificity) to determine its safety and the capacity of a safe dose to down-regulate circulating iNKT cells in patients with SCD. We have found two dose rates that are both safe and effective and now plan a controlled Phase 2B clinical trial to determine whether our highest dose, administered as a 48-hour constant infusion, will induce faster remission in both painful vaso-occlusive crisis (pVOC) and acute chest syndrome (ACS).

INTRODUCTION

SCD is a pernicious illness characterized by chronic vasculopathy and multiple episodes of pVOC and (less frequent) ACS, all of which particularly damage the brain, lungs, bones, spleen, and kidneys. Although SCD, when seen only in the hospital, can appear to be very periodic because the frequency of hospital admission for either pVOC or ACS in any given patient may be as low as once a year (1, 2), most patients have much more disability. Smith et al reported that a high proportion of SCD patients have daily pain and disability (3).

Until the 1980s, the pathophysiology of SCD was thought to be due entirely to the distortion of the red cells by deoxyhemoglobin S and consequent mechanical regional obstruction of pre-capillary and capillary blood flow. The high viscosity of SCD blood induced by cells with abnormal flow characteristics also contributes to low oxygenation by reducing cardiac output and, therefore, oxygen delivery to the tissues. The severity of the various sickle syndromes depend almost entirely on the mixture of hemoglobins in the individual SCD erythrocytes (4). Those cells with increased amounts of fetal hemoglobin (HbF, an excellent anti-sickling agent) change their shape in hypoxic conditions far less readily than those that lack HbF. A milder course and lower mortality is usually observed if such cells, known as F-cells, constitute a high proportion of circulating erythrocytes (5). The protective role of HbF forms the basis of treatment of SCD with hydroxyurea, the only United States Food and Drug Administration (FDA)– approved drug approved for SCD management (6). When properly used, hydroxyurea is useful (7), but it is estimated that, in general, only approximately half of SCD patients are significantly benefitted and exhibit reduced hospitalizations for crises.

Hebbel et al added another important contribution to SCD pathophysiology when they noted that sickled SCD erythrocytes adhere tightly to human endothelial cells (8). This was confirmed by Narla et al (9). These latter observations have led to an additional understanding that SCD may actually cause IRI (10). The latter contributes to a vicious cycle because IRI (10) is associated with inflammation that causes a decrease in local pH, reduced cellular hydration, and increased local temperature, all of which enhance intracellular sickling.

It has been well established that iNKT cells may play an important role in the development of tissue damage and inflammation resulting from IRI (10). Large numbers of activated iNKT cells are found in the liver and kidney during an episode of IRI that may accompany organ transplantation (11–13). To investigate the potential role of iNKT cells in sickle cell disease, Linden et al have studied the NY1DD mouse, a murine model of relatively mild SCD. They have confirmed that the NY1DD mouse has chronic lung disease characterized by cellular invasion of the alveolar septae by erythrocytes and leukocytes: many of the latter are iNKT cells. The pulmonary pathology of these mice is markedly ameliorated by removal of iNKt cells from the murine T cell compartment either by antibodies or by genetic manipulation (14, 15). The pulmonary disease is also ameliorated by treatment of the mice with an adenosine analogue, an approach taken because activated iNKT cells express large numbers of adenosine A2A receptors on their surfaces. Linden et al have concluded that iNKT cells are the leading causes of tissue pathology in the NY1DD mouse and have also shown that circulating activated iNKT cells are markedly increased above normal in most patients with SCD (15).

Armed with this information, and concerned about a paucity of useful approaches to management of pVOC or ACS crises in SCD, our group has decided to mount a clinical trial to investigate the possible benefit of an adenosine as a receptor agonist administered intravenously during acute sickle cell crises in an in-patient setting.

MATERIALS AND METHODS

Clinical Trial

The first stage of the dose finding trial used a standard 3 × 3 design. A low dose (0.24 μg/kg/h of regadenoson [Lexiscan,]) was infused for 12 hours into patients with SCD who were considered “baseline” because they had not had a recent crisis requiring hospitalization. Serial measurement of pulse and blood pressure were obtained as indices of toxicity because the major toxic aspect of regadenoson is hypotension followed by reflex tachycardia. The low dose was followed in a second group of patients at 0.66 μg/kg/h and was increased in a third group to 1.44 μg/kg/h. In stage 2, the highest dose was used in a fourth group of patients who received 1.44 μg/kg/h for 24 hours. Thus far, a total of 21 patients have been studied in stages 1 and 2, and 1 patient is in stage 3 (a patient with pVOC).

Laboratory Measurements

Plasma regadenoson concentrations were measured by liquid chromatography/mass spectrometry using a deuterated regadenoson standard in all of the stage 1 patients.

Blood samples were obtained periodically throughout the infusions and maintained at 4 ° C before iNKT cell numbers and their state of activation were estimated by flow cytometry. The iNKT cells were separated from the leukocyte population and enumerated by means of their binding by flourescent MAb 6B11 (16), an antibody that reacts exclusively with the invariant T cell receptor on human iNKT cells. Their state of activation was initially measured by evaluating their surface expressions of CD69 and CXCR3 and intracellular interferon gamma. We then found that the most useful indices of activation were high expression of the Ser536-phosphorylated p65-subunit of activated NF-kB, low expression of the inhibitory subunit, IKBa, and high expression of adenosine A2A receptors. The latter measurements were used in stages 2 and 3.

RESULTS

All three dose levels and the two times of administration of the drug failed to induce any hypotension or tachycardia in stage 1 and stage 2 patients with baseline SCD or in the single patient studied thus far in PVOC. The drug was therefore judged to be safe at these ranges of administration in baseline patients 18 years of age and older. Follow-up examinations and several blood studies also revealed no other toxicities.

Plasma regadenoson levels achieved plateau after approximately 3 hours of constant intravenous infusion. The levels were consistent with the dose rates and achieved concentrations at the highest dose that had been previously shown to down-regulate iNKT cell lines in vitro. They persisted during the infusion and rapidly decreased after the infusion was stopped. Six hours after conclusion of the treatment, levels of regadenoson were barely detectable.

Flow cytometry revealed that iNKT cells were readily identified and were activated in the peripheral blood of nearly every patient. After 6 hours of regadenoson infusion at either 0.66 μg/kg/h or 1.44 μg/kg/h, the number of iNKT cells in the “activation gate” was reduced by between 30% and 50% in more than three quarters of the patients. Activation recovered in most patients after regadenoson administration was discontinued. We therefore concluded that regadenoson was both safe and biologically effective in baseline SCD patients at the doses specified.

DISCUSSION

This study was undertaken because there are no effective treatments for acute crisis in SCD other than fluids, narcotics, and blood transfusions. Although treatment of crisis is not a sufficient goal in SCD therapeutics because crisis management does little to reduce the ongoing chronic organ damage that characterizes the disease, faster relief of pain and acute disability in the crisis setting would be highly useful. Currently, hydroxyurea treatment to stimulate the number of F cells with large amounts of HbF within them is the only preventive treatment available to these unfortunate patients short of gene therapy or stem cell transplant. However, hydroxyurea is at best effective in 50% to 70% of patients depending on the determination of the physician to use it. The recent studies of BCL11A inhibition by Orkin and Sankaran and their coworkers offers great promise for a far better preventive therapy (17). But at this time, we need much better management of crises in this pernicious disease.

The potential value of adenosine A2A receptor agonist treatment in SCD crisis stems from the observations of Linden et al in NY1DD model mice transfected with human SCD (4). They clearly demonstrated that iNKT cells, a very small subset of hybrid T cells in mice and in humans that express large numbers of adenosine A2A receptors on their surfaces, are responsible for much of the organ damage in that model, and they showed that treatment with a potent adenosine A2A receptor agonist markedly reduces pulmonary pathology induced by exposure to a period of hypoxia.

Treatment of SCD with an adenosine receptor agonist is not, however, without risk. A standard pulse dose of regadenoson as used in cardiac imaging would not be safe because it would depress the blood pressure and cause reflex tachycardia in patients already anemic and in pain. We have avoided this risk by using a low-dose/constant infusion approach to the treatment. Furthermore, the selectivity of the adenosine may be very important. Xia et al have shown that adenosine, which binds to and agonizes all four adenosine receptors, may worsen hemolysis and anemia in the Berkeley mouse model of SCD by engaging adenosine A2B receptors on sickle erythrocytes or reticulocytes (14). Fortunately, regadenoson is highly selective if not completely selective for A2A receptors and does not appear to engage A2B receptors at the concentrations that we have used.

The data presented here, derived from our experience of 21 patients at baseline and 1 in PVOC, are not sufficient to permit us to begin a clinical trial immediately. We must still be certain that the drug is safe during a period of pVOC and we must examine safety in patients between the ages of 14 and 18 years old. But the data do reassure us that this adenosine A2A agonist is very likely to be safe and that it certainly reaches the target that we have chosen. The achievement of clinical effectiveness is our next pressing goal.

DISCUSSION

Schiffman, Providence: David, beautiful paper. What do you think the global effect will be by eliminating these cells? I was with you until you wanted to get rid of all of them. As Phil Mackowiak said to me, “they must be there for something.” Could we really be affecting the immune response?

Nathan, Boston: God gave us iNKT cells. Who is David Nathan to take them away? That's the question. Okay, here's the point about the antibody, which is extremely interesting. When you give an injection of this antibody, one shot, the iNKT cells disappear for months, but unlike other T-cell monoclonal antibodies, they come back. We do not get the precursor. We only get those completed cells, the mature cells. So they're gone. The precursor is still there so they can come back. But there is murine evidence of infection risk — particularly borrelia and strep pneumo. Yet, as far as anybody knows, iNKT cells have has no role in human infection. Now these are spleenless patients and they are fragile patients, so we are getting a lot of heat from the FDA and others about the antibody, but we're going to win through on that because the cells will come back. When we stop, they'll come back.

Spivak, Baltimore: That was really an exciting presentation, and one of the things I was thinking about it was: When we do bone marrow transplant to cure sickle cell anemia? Post-transplantation, as the patients recover, they have about as severe or even more severe pain crises in the sickle cell than they had before they had the transplant, and this can last up to a year and, of course, it's like the phantom limb syndrome, but nevertheless it's real, and I'm curious because then they are being reconstituted with a normal immune system, and I don't know if you've done any studies of these iNKT cells post bone marrow transplant because it seems that this would be a perfect mechanism for this so-called recrudescence of pain in the absence of sickle cells.

Nathan, Boston: I can think of a lot of reasons why that might happen, but that's a good idea, and I think we must look at that. There are other conditions, by the way, in which I would like to look at iNKT cells. One of them is something like Henoch-Schonlein purpura, and I know that sounds bizarre but I'm just a pediatrician and I'm interested in that disease. The facts are that in that disease, like sickle cell anemia, you get a temporary response to steroids and then it doesn't work anymore and they flare. That's because, I think, iNKT cells seem to be relatively insensitive to steroids. You can't get rid of them that way. You can shut down other inflammatory cells, but not iNKT cells. So, there are special circumstances. I will keep in mind that transplant issue.

Abboud, Iowa City: So, the iNKT cell appears to be proliferating or increasing only in sickle cell disease?

Nathan, Boston: Now, we haven't studied everybody, but we know that iNKT cells are increased by a factor of roughly 10 in the blood of patients with sickle cell disease. Now, sickle cell disease patients are spleenless, and we haven't controlled for splenectomy, so, I don't know whether this is because they're chronically inflamed all the time or because they are spleenless or both. So, we study thalassemia after splenectomy and see if iNKT cells are elevated in that group of patients. We haven't done that yet.

Abboud, Iowa City: So other innate immune cells should recognize normal cells and tissues, shouldn't they?

Nathan, Boston: Yes. Other immune cells should be involved in this, and I think they are but they're secondary. They're driven by the iNKT cell. I think, I can't really prove this yet, but the animal studies suggest that the disease starts via adhesion phenomenon, sickle cells sticking to each other and to endothelium. That sends signals that alert the iNKT cell, and those cells spew out all their cytokines and get all the other inflammatory cells going. That's the way I'm thinking about it at the moment.

Silverstein, Milwaukee: David, another disease where that happens is cerebral malaria. Are there any data in that group?

Nathan, Boston: Yes. I've been thinking a lot about that. You are right. That's a really fascinating inflammatory problem and kills because of the inflammation as well as adhesion.