Nonmalignant Hematologic Disorders at ASH 2020: Our Choices for Most Notable Studies

In this commentary, Hanny Al-Samkari, MD; Sujit Sheth, MD; and Mark A. Schroeder, MD, highlight key studies in the management of immune thrombocytopenia (ITP), hemophilia B, sickle cell disease (SCD) and β-thalassemia, and graft-vs-host disease (GVHD) presented during the 2020 ASH annual meeting and share their perspectives on the clinical implications of these findings.

Hanny Al-Samkari, MD: ITP Studies

FLIGHT study
Bradbury and colleagues presented results from the phase III FLIGHT trial, which compared first‑line mycophenolate mofetil and corticosteroids vs standard steroid treatment in patients with newly diagnosed ITP (N = 120). This is a very important study in ITP in which patients who were at least 16 years of age with a platelet count < 30 x 10⁹/L and who required first‑line treatment received either mycophenolate mofetil 1 g twice daily orally plus a corticosteroid or to corticosteroids alone. The primary endpoint was the time to treatment failure (platelet count < 30 x 10⁹/L) as well as clinical need for second‑line treatment. In this study, frontline treatment with mycophenolate mofetil plus corticosteroids achieved a significantly lower rate of treatment failures vs the corticosteroid-alone arm (22% vs 44%, respectively; HR: 0.41; P = .0064). In addition, patients receiving mycophenolate mofetil were more likely to achieve platelet count recovery of > 30 x 10⁹/L following second-line treatment. For example, in patients with platelets > 100 x 10⁹/L after first-line treatment, the response rate was 91.5% with the addition of mycophenolate mofetil vs 63.9% with corticosteroids alone.

An important question was whether adding mycophenolate mofetil to corticosteroids would significantly increase toxicity for patients with ITP, and we saw no significant differences in adverse event (AE) rates between arms, which is reassuring. Approximately 25% of patients developed an infection or infestation, only 4 cases of neutropenia were seen (all in the corticosteroid-only arm), and diarrhea or other gastrointestinal events affected 25% of patients in the steroid-only arm and 34% in the combination arm.

Another important finding from the FLIGHT study was that patient-reported outcome measures were worse in the mycophenolate mofetil plus steroid group vs the steroid-alone group (SF-36: P = .012; FACIT-Fatigue: P = .05). However, quality-of-life differences were not statistically significant when a Bonferroni correction (P = .0033) was applied for multiple testing.

Results from FLIGHT suggest that mycophenolate mofetil plus corticosteroids appear safe and effective in patients with ITP. The combination was well tolerated with significantly lower rates of treatment failure than with corticosteroids alone. It is still unclear whether these findings will change the standard of care. It would be nice to see longer-term remission data to determine whether adding mycophenolate mofetil to frontline corticosteroids results in a higher rate of durable drug-free remission for several years. The fact that there were no excessive additional AEs with addition of mycophenolate mofetil is promising. It is unclear why there were lower quality-of-life scores for the combination group, and this is an important topic for additional study.

Rilzabrutinib trial
Kuter and colleagues presented long-term extension results from a dose-finding, open-label phase I/II clinical trial of the oral, reversible, BTK inhibitor rilzabrutinib. In the initial dose-escalation phase, patients received rilzabrutinib 200 or 400 mg once daily or 300 or 400 mg twice daily, with the optimal dose identified as 400 mg twice daily. The endpoint of the long-term extension study presented at ASH 2020 was safety and durable responses in the 13 patients who received rilzabrutinib 400 mg twice daily and had platelet counts ≥ 50 x 10⁹/L for at least one half of the final 8 weeks of treatment in the initial phase.

Patients in this study were heavily pretreated for ITP and had no other available treatment options. Patients in the dose-escalation cohort had received a median of 6 previous treatments (range: 1-53) and 5 in the long‑term extension cohort (range: 1-19). Approximately 25% of patients had undergone a splenectomy.

Despite being a heavily pretreated patient population, more than 42% of patients responded in the dose-escalation cohort and responses were durable; median treatment duration for patients who received rilzabrutinib 400 mg twice daily was 19.6 weeks. Median platelet count for patients who crossed over from the dose-escalation to the long‑term extension was 98 x 10⁹/L.

Based on data presented at ASH 2020, rilzabrutinib 400 mg twice daily appears safe overall, with no treatment‑related bleeding or thrombotic events, and no high‑grade treatment‑emergent AEs. Any-grade treatment‑emergent AEs judged to be related to rilzabrutinib occurred in 47% of patients. As expected with drugs in this class, treatment‑emergent AEs included diarrhea, nausea, fatigue, and other abdominal‑related symptoms, but all were low grade and transient.

Pending validation and confirmation in large randomized clinical trials, I think rilzabrutinib appears to be a safe and efficacious oral option for patients with highly refractory ITP who have no other available treatment options. I am eager to see data from the ongoing phase III LUNA3 trial (NCT04562766), which is comparing rilzabrutinib vs placebo in adult patients with ITP.

Sujit Sheth, MD: Hemophilia B, SCD, and β-Thalassemia Studies

Hemophilia B
Pipe and colleagues presented the primary analysis for an international, open-label, single-arm phase III HOPE-B study evaluating etranacogene dezaparvovec (AMT-061), an investigational gene therapy, consisting of an adeno-associated virus-5 vector (AAV5) with a naturally occurring hyperactive Padua factor IX (FIX) gene variant, in patients with hemophilia B. Patients were males 18 years of age or older with a FIX activity level ≤ 2% of normal and had been receiving continuous FIX prophylaxis for at least 2 months. The coprimary endpoints were FIX activity at 26 and 52 weeks. Secondary endpoints included safety and rates of total, spontaneous, traumatic, and treated and untreated bleeds. Other key endpoints included FIX consumption and response rate correlated with neutralizing AAV5 antibodies.

By Week 26, mean FIX activity levels were 37.20% or a +36.01% increase from baseline (P < .0001). For this patient population achieving and maintaining stable FIX activity in the 30% to 50% range would be protective of most bleeds and was observed across the entire population of 54 patients. Authors also reported a significant reduction in total bleeds (83% decrease) and treated bleeds (91% decrease) vs baseline. There was also a significant decrease in factor usage, decreasing from a mean of 290,769 units/year/patient at baseline to 12,537 units/year/patient at 6 months in all patients, and 52 (98%) patients were able to discontinue their previous prophylaxis indefinitely.

Of importance, approximately 43% of patients had detectable neutralizing AAV5 antibodies at baseline, with a maximum titer recorded over 3000. However, there was no correlation between response and preexisting neutralizing AAV5 antibodies titer, which is important because it was previously believed that patients with neutralizing antibodies would not be good candidates for this type of gene therapy. One patient with a very high antibody titer of 3212.3 did not respond to treatment, suggesting a threshold of antibody level at which a prospective patient might not be a good candidate.

I thought the safety profile of autologous etranacogene dezaparvovec gene therapy was quite good. Rates of treatment-related AEs were low and 76% were mild in severity. The most common AEs after infusion were influenza-like illness (13%), headache (13%), and liver enzyme elevation (alanine aminotransferase, 13%; aspartate aminotransferase, 9.3%), and 9 of 54 patients had elevation of liver enzymes that required intervention with steroids, but all had discontinued steroids before the 6-month analysis. Seven patients experienced infusion-related reactions, but the infusion was able to be completed in 6 of them; 1 patient received 10% of the dose and discontinued treatment thereafter. Of importance, there was no correlation between AAV5 capsid–neutralizing antibodies and safety, and no inhibitors to FIX were reported.

In summary, HOPE-B is the largest phase III study of gene therapy for hemophilia B to date. Data were reported for the primary analysis, which showed that in patients with moderately severe to severe hemophilia B, etranacogene dezaparvovec significantly increased mean FIX activity to near-normal levels by 6 months. Of importance, 98% of patients responded and were also able to discontinue previous FIX prophylaxis. There were no major treatment-related AEs reported. Thus, the study is very encouraging, and it will be interesting to see what longer‑term data show—as they want to collect data up to 60 months—and whether durability of FIX activity is extended.

CLIMB Studies
Frangoul and colleagues presented early results from 2 phase I/II studies evaluating CRISPR/Cas9 gene editing of BCL11A in stem and progenitor cells (CTX001) in order to allow continuing gamma globin expression and therefore fetal globin production: CLIMB THAL-111 in patients with transfusion-dependent β-thalassemia (TDT) and CLIMB SCD-121 in SCD. The primary endpoint of CLIMB THAL-111 was sustained reduction in transfusions of 50% for at least 6 months, beginning 3 months after infusion, and the primary endpoint of CLIMB SCD-121 was fetal hemoglobin level of ≥ 20% sustained for longer than 3 months, beginning 6 months after infusion.

In both CLIMB THAL-111 and CLIMB SCD-121, most AEs occurred within the first 60 days of the infusion, and all were consistent with complications expected and related to the myeloablation and the autologous transplant. In all patients with at least 3 months of follow-up (n = 7), only 1 patient reported a serious AE related or possibly related to study drug.

Outcomes were very promising in both patients with TDT and those with SCD. In patients with TDT patients with at least 3 months of follow-up, total hemoglobin levels reached near normal levels (range: 9.7-14.1 g/dL), and this was enough to achieve transfusion independence in all evaluable patients. The shortest and longest on-study time for patients with TDT without the need for red blood cell transfusions were 1.8-20.5 months. In patients with SCD, hemoglobin responses were quite robust with all patients achieving hemoglobin levels of ≥ 10 g/dL by 3 months. Of importance, the percentage of sickle hemoglobin in patients with SCD decreased by approximately 50%, and all 3 patients, albeit at different time points, achieved peripheral blood F-cells > 90%. This is of tremendous benefit, especially in patients with SCD where the sickling process is interfered with by the fetal hemoglobin molecules. Moreover, following engraftment, patients with SCD experienced a reduction in vaso-occlusive crises (VOC); on study time without VOC were 3.8 months, 7.8 months, and 16.6 months, and there was no evidence of ongoing hemolysis at the time of this report, suggesting stabilization of the red cells by the fetal hemoglobin levels.

Another important finding of this study was that allele editing of BCL11A was durable and was seen in bone marrow CD34-positive cells (range: 41.8% to 88.1%), supporting the proof of concept that allelic editing of BCL11A for reactivation of fetal hemoglobin in patients with TDT and SCD is feasible.

To summarize, current data for the CLIMB THAL-111 and CLIMB SCD-121 studies show stable engraftment of CRISPR/Cas9 modified CD34-positive cells for at least 6-12 months. Targeting of BCL11A resulted in high fetal hemoglobin levels, which in TDT patients was enough to achieve transfusion independence and in SCD patients enough to practically eliminate incidence of VOCs.

Mark A. Schroeder, MD: GVHD Studies

Chronic GVHD occurs in up to 70% of patients after allogeneic stem cell transplant (alloSCT) of which approximately 50% will respond to standard treatment with corticosteroids. Ibrutinib is approved by the FDA for second-line treatment of steroid‑refractory chronic GVHD, based on a phase II study in patients with severe oral and skin chronic GVHD. However, there are currently no standard second‑line therapies due to lack of large, randomized phase III trials. Thus, effective treatment options for patients with chronic GVHD after corticosteroid failure are lacking.

REACH3
At the ASH 2020 annual meeting, Zeiser and colleagues reported results from the randomized phase III REACH3 study of the JAK inhibitor ruxolitinib vs best available therapy (BAT) in 329 patients 12 years of age and older with severe steroid-refractory or steroid-dependent chronic GVHD and myeloid engraftment. Patients were randomized to either ruxolitinib 10 mg twice daily or BAT plus steroids in both arms, with or without a calcineurin inhibitor. In the primary efficacy period, patients were treated for 6 months, and in the extension period, patients could be treated for up to 3 years. The primary endpoint was ORR at 6 months. Secondary endpoints included 6-month failure-free survival and modified Lee Symptom Scale.

The ORR at 6 months was 49.7% in the ruxolitinib arm vs 25.6% with BAT (odds ratio [OR]: 2.99; P < .001), of which most were PRs (43.0% with ruxolitinib and 22.6% with BAT). Median duration of response was not reached in the ruxolitinib and 6.24 months with BAT. Best ORR with ruxolitinib was 76.4% and 60.4% with BAT (OR: 2.15). I would like to highlight that in chronic GVHD, CRs are largely uncommon, and that was the case in the ruxolitinib arm (CR rate: 12%). Compared with BAT, treatment with ruxolitinib improved symptoms by the modified Lee Symptom Scale: 24.2% of patients had a 7-point reduction from baseline vs 11.0% of patients at 6 months (OR: 2.62; P = .0011). OS results were not mature in this analysis of REACH3 and, therefore, were not reported.

The safety profile with ruxolitinib was similar to what has been previously reported in acute GVHD. In REACH3, hematologic AEs were a concern. All-grade anemia was reported in 29.1% of patients in the ruxolitinib arm (grade ≥ 3: 12.7%) vs 12.7% with BAT (grade ≥ 3: 7.6%). All-grade thrombocytopenia was reported in 21.2% of patients in the ruxolitinib arm (grade ≥ 3: 15.2%) vs 14.6% with BAT (grade ≥ 3: 10.1%). As a result, more than 50% of patients in the ruxolitinib arm underwent dose modifications or discontinued treatment. There was also a signal for increased fungal infections with ruxolitinib vs BAT (11.5% vs 5.7%, respectively), with the majority being of grade 3 vs grade 1/2. Because of this, patients receiving ruxolitinib should closely be monitored for fungal infections, and prophylaxis against fungal infection should be considered in patients to be started on this therapy.

The takeaways from the REACH3 study, for me, were that this is the first successful randomized controlled trial for the management of patients with chronic GVHD who are failing steroids. In addition, compared with BAT alone, the combination of BAT and ruxolitinib was associated with significantly improved responses (ORR: 49.7% vs 25.6%; P < .0001), failure-free survival (not reached vs 5.7 months; P < .0001), and symptoms by modified Lee Symptom Scale score (24.2% vs 11.0%; P = .0011). The safety profile of ruxolitinib was as expected. Based on these results, I think that ruxolitinib ultimately will be approved by the FDA for the treatment of steroid‑refractory chronic GVHD.

ROCKstar
At ASH 2020, ROCK2 inhibition was explored as a novel approach to treating chronic GVHD. Cutler and colleagues presented top-line results from the phase II ROCKstar study of belumosudil, a ROCK2 inhibitor, in 132 patients who were 12 years of age or older and had a diagnosis of chronic GVHD after previous alloSCT and who had received 2-5 previous lines of systemic therapy. Patients in this study had a median age of 56 years and had received a median of 3 previous lines of therapy, and 67% had severe chronic GVHD. Approximately 50% of patients had at least 4 involved organs, and most were refractory to their last line of therapy.

Patients received belumosudil 200 mg either once daily or twice daily. Similar to the REACH3 study, the primary endpoint was ORR at 6-months (per the 2014 National Institutes of Health chronic GVHD consensus criteria). Secondary endpoints included safety, response duration, Lee Symptom Scale score, and survival.

At the 6-month time point, ORRs for belumosudil with once-daily vs twice-daily dosing were comparable: 73% vs 77%, respectively. Similar to ruxolitinib, CRs were uncommon in organs, with only 7 patients achieving a CR in all affected organs. In contrast to JAK inhibition, ROCK2 inhibition yielded responses in organs with fibrotic disease. And, similar to JAK inhibition, the Lee Symptom Scale was improved with belumosudil in 42% of patients in the daily arm and 36% of patients in the twice-daily arm. Median time to response was 4 weeks, and the median duration of response was 50 weeks. Responses to belumosudil were observed in 78% (n = 74) of patients who had previous ibrutinib failure, and in 68% (n = 38) of patients that had received previous ruxolitinib.

The safety profile for belumosudil was acceptable and in line with what is expected in patients being treated for chronic GVHD. AEs in ≥ 20% of patients included fatigue (38%), diarrhea (33%), nausea (31%), cough (28%), and upper respiratory tract infections (27%). Grade ≥ 3 AEs in ≥ 5% of patients, included pneumonia (n = 10), hypertension (n = 8), and hyperglycemia (n = 6); 2 patients experienced reactivation of Epstein-Barr virus and cytomegalovirus; and 67% (n = 89) of patients experienced drug-related AEs, with 5% (n =7) experiencing serious AEs.

There was a total of 8 deaths on study (4 at each dosing schedule). Deaths in once-daily dosing were attributed (1 each) to aspiration pneumonia, hemoptysis, multiple organ dysfunction syndrome/septic shock, or acute myeloid leukemia relapse. Deaths in the twice-daily dosing arm were attributed to cardiac arrest (n = 2), infection (n = 1), or respiratory failure (n = 1).

My takeaways from the ROCKstar study was that in patients with chronic GVHD—severe in 67% of cases—and 2-5 previous lines of systemic therapy, belumosudil yields clinically meaningful results and had a safety profile consistent with previous reports. Overall, responses were seen in more than 70% of patients, across all key subgroups and in all affected organ systems (including with fibrotic disease). Survival outcomes were very encouraging with a 12-month failure-free survival rate of 58% and a median duration of response of 50 weeks. Also, important to note that responses were seen in patients who had failed previous ibrutinib or ruxolitinib treatment. Moreover, unlike previous reports with JAK inhibitors in GVHD, AEs were notable due to lack of cytopenias.

Based on these results, I think belumosudil will likely be approved for chronic GVHD that has progressed after ≥ 2 previous lines of therapy.

GRAVITAS‑119 study
Acute GVHD is a common complication after alloSCT, and until recently, there were no FDA-approved drugs for steroid-refractory disease. Ruxolitinib was recently approved as a treatment for steroid‑refractory acute GVHD based on data from a phase II study and the ensuing confirmatory randomized phase III study. An outstanding question in the field is whether JAK inhibition could potentially be used to prevent GVHD from happening. Despite available rationale for using JAK inhibitors early, the main concern is off-target effects on the regeneration of stem cells after an alloSCT. There were 2 studies exploring this very question at ASH 2020, one by Pidala and colleagues, reporting results from a cohort of 12 patients receiving JAK2 inhibitor pacritinib plus sirolimus and low-dose tacrolimus, and the GRAVITAS‑119 study by Choe and colleagues investigating itacitinib, a JAK1-specific inhibitor, as prophylaxis for acute GVHD in 65 patients. I would like to focus on the larger of the two: GRAVITAS‑119 study.

The single-arm, open-label GRAVITAS-119 study enrolled patients 18 years of age or older with hematologic malignancies who were candidates for reduced-intensity conditioning and/or alloSCT. Itacitinib was given at 200 mg once daily plus tacrolimus/methotrexate or cyclosporine A/mycophenolate mofetil with or without antithymocyte globulin (ATG). At Day 90 after transplantation, itacitinib dose was lowered to 100 mg once daily for the next 90 days and then tapered. Patients were followed for 6 months for GVHD and OS. The primary endpoint was the percentage of patients with hematologic recovery at Day 28, defined as absolute neutrophil count (ANC) ≥ 500/mm³ for 3 consecutive measurements; platelets ≥ 20,000/mm³ and no platelet transfusion in the previous 3 days. Secondary endpoints included GVHD-free and relapse-free survival, transplant-related mortality, OS, acute and chronic GVHD, and safety.

Approximately 98.5% of patients achieved both ANC and platelet recovery by Day 28, and all patients had achieved hematologic recovery by Day 31. Median time to ANC recovery was 17 days—which is what we would expect with a reduced intensity conditioning alloSCT—and the median time to platelet recovery was 14 days.

With the caveat of small numbers, in this reduced-intensity conditioned population of patients, there was a low incidence of grade III/IV acute GVHD, reported at approximately 5% of patients in both the ATG and no-ATG groups. At 12 months, rates of moderate to severe chronic GVHD were lower with ATG, at 10.4% vs 31.3% without ATG. Of importance, the combined rate of GVHD and relapse-free survival was 23.1% with no ATG vs 60.9% with ATG. The 12-month OS estimate were comparable between groups (74.3% with no ATG vs 82.6% with ATG). These findings need be interpreted with caution given small numbers, but the rate of relapse or progression at 12 months did not appear to be affected.

In terms of safety, the most common grade ≥ 3 hematologic AEs included thrombocytopenia (49.3%), anemia (29.2%), and decreased blood counts (29.2%), and 8 (12.3%) patients developed grade ≥ 3 febrile neutropenia. Grade ≥ 3 nonhematologic events of diarrhea, hypertension, and hypertriglyceridemia were observed in 15.4%, 13.8%, and 12.3% of patients, respectively. AEs led to dose modifications or discontinuation in 23.1% of patients.

As I alluded to earlier, one key concern with the use of JAK inhibitors after alloSCT is an increased risk of infections. Reassuringly, in this small study, there did not appear to be an increased in the risk of viral, bacterial, or fungal infections with itacitinib. Nevertheless, all-grade cytomegalovirus infections were seen in 6 patients without ATG and in 5 patients with ATG.

In this study, 2 patients had secondary graft failure and had to undergo a second transplant: one patient with myelodysplastic syndrome had previously received a transplant with an 8/8-matched unrelated donor, and one patient with acute lymphoblastic lymphoma had previously received a transplant with an 8/8-matched related donor.

The take-home message from the GRAVITAS-119 study is that JAK1 inhibition with itacitinib plus calcineurin-based regimens can be safely given as prophylaxis for GVHD after a reduced-intensity conditioning alloSCT. Preliminary data from this study, although in a small number of patients, suggest that there might be a reduced incidence of severe acute GVHD (~ 5%) by Day 180 after transplant. For me, the occurrence of secondary graft failure in 2 patients is concerning, despite being expected for matched unrelated donor transplants with reduced intensity conditioning, and this will need to be further evaluated with additional follow-up.

More ASH 2020 Conference Coverage on the CCO Web site!
Remember to download the Capsule Summary slidesets for the key studies and review a CME-certified Expert Analysis text module with additional studies on our Web site! For more information on the different nonmalignant disorders discussed in this commentary be sure to visit our on-demand Webcasts of recent live Webinars on ITP, β-thalassemia, and GVHD!