Moxetumomab pasudotox in heavily pre-treated patients with relapsed/refractory hairy cell leukemia (HCL): long-term follow-up from the pivotal trial

This multicenter, open-label, single-arm study enrolled patients from 32 centers across 14 countries. Adults (aged ≥ 18 years) with histologically confirmed HCL or variant HCL and an indication for therapy were eligible for this study. An indication for therapy was defined as one or more of the following: neutrophils < 1.0 × 10³/µL, platelets < 100 × 10³/µL, hemoglobin < 10 g/dL, or symptomatic splenomegaly. Patients must have been treated with a minimum of two systemic therapies, including two courses of a PNA, or one course of a PNA followed by rituximab or a BRAF inhibitor. Patients must not have had prior treatment with moxetumomab pasudotox and were required to have an Eastern Cooperative Oncology Group (ECOG) performance status ≤ 2. Patients with impaired renal or hepatic function, active second malignancies requiring treatment, and uncontrolled intercurrent illnesses such as an ongoing or active infection were excluded. Detailed inclusion and exclusion criteria can be found in the study protocol.

Moxetumomab pasudotox was administered at a dose of 40 µg/kg by intravenous (i.v.) infusion over 30 min on Days 1, 3, and 5 of a 28-day cycle for up to six cycles, or until documentation of MRD-negative CR, disease progression, initiation of an alternate therapy, or unacceptable toxicity.

If during treatment and prior to completion of 6 cycles of therapy, blood counts were consistent with CR for at least 4 weeks, an interim disease assessment could have been performed at the discretion of the investigator. If a CR without MRD was documented, treatment was discontinued.

Patients were given prophylactic treatments for renal insufficiency (fluids and aspirin), arterial thrombosis (low-dose aspirin), and hypersensitivity reactions (hydroxyzine, acetaminophen, and ranitidine). Details can be found in the study protocol. Patients were encouraged to drink at least 3 L of fluid per day on Days 0–8 of the treatment cycle.

Adequate pneumocystis carinii pneumonia prophylaxis was provided for patients receiving corticosteroids ≥ 10 mg of prednisone daily or equivalent, CD4 lymphocyte count < 0.2 × 10³/µL, or at the discretion of the investigator. Viral prophylaxis was provided for patients receiving chronic corticosteroids or with lymphopenia.

The primary endpoint was the durable CR rate, defined as CR with hematologic remission (HR) lasting > 180 days; it was calculated as the ratio of the number of patients achieving durable CR over the total number of patients treated. CR was determined by blinded independent central review (BICR) and local investigators based on pathology (no evidence of leukemic cells in both the peripheral blood and bone marrow by routine hematoxylin and eosin [H&E] staining), radiography with CT or MRI (no hepatomegaly, splenomegaly, or lymphadenopathy) and HR (neutrophils ≥ 1.5 × 10³/µL, platelets ≥ 100 × 10³/µL, and hemoglobin ≥ 11.0 g/dL, without transfusions or growth factors for ≥ 4 weeks). Resolution of splenomegaly required a maximum spleen diameter of < 17 cm or > 25% shorter than at baseline. Relapse was defined as the loss of any criteria needed for best response, including asymptomatic reappearance of hairy cells in the bone marrow by H&E staining. Additional details can be found in the study protocol.

The secondary endpoints included CR rate, objective response (OR) rate, time to and duration of CR and OR, MRD, progression-free survival (PFS) rates, safety/tolerability, immunogenicity, and pharmacokinetics.

MRD was locally assessed during treatment using quantitative flow cytometric analysis of peripheral blood samples or bone marrow aspirate, according to the institution’s protocols, and by immunohistochemistry (IHC) on the bone marrow biopsy tissue. Local study sites received instructions on sample collection and submission for analysis. Samples for peripheral blood were collected at screening, prior to Cycles 3 and 5 only, at the end of treatment, at 181 days after end of treatment, 12 and 18 months after end of treatment, and yearly thereafter. Bone marrow aspirate was collected at screening, at the end of treatment for all patients, and 181 days after end of treatment for patients in PR or CR. Local MRD status was then confirmed by BICR evaluation of H&E staining and IHC on the bone marrow biopsy tissue. IHC involved staining for the presence of phenotypical B-cell (CD20) and HCL (CD79a, annexin A1, DBA.44, and PAX-5) antigens. MRD negativity was defined as the absence of HCL phenotype cells in the bone marrow biopsy by IHC and in the bone marrow aspirate/blood by flow cytometry.

Safety was evaluated from enrollment until 30 days after the last moxetumomab pasudotox dose was administered. The evaluation included adverse events (AEs), serious AEs (SAEs), vital signs, electrocardiograms, physical examinations, and clinically meaningful changes in laboratory tests. The National Cancer Institute Common Terminology Criteria v4.03 were used for grading, and the Medical Dictionary for Regulatory Activities v20.0 was used for coding AEs.

The intent-to-treat (ITT) population, including all patients who received moxetumomab pasudotox, was used to evaluate efficacy, whereas the safety population, including all patients who received ≥ 1 dose of moxetumomab pasudotox, was used to evaluate safety. Both populations included the same 80 patients.

The sample size was based on the historical CR rate of rituximab against relapsed HCL requiring treatment of ≤ 13% [11] and the target durable CR for moxetumomab pasudotox of ≥ 28%. Using the exact binomial test, it was estimated that a sample size of 77 patients would provide 90% power to detect a difference between 13 and 28% durable CR rates, with a two-sided significance level of 0.05. The 95% confidence interval (CI) for durable CR was estimated using the exact probability method (Clopper–Pearson exact interval); if the lower bound of the 95% CI was above 13%, the durable CR rate was deemed significantly higher than the historical control. The Kaplan–Meier method was used to estimate the duration of CR, OR, and PFS. In these analyses, patients alive with no documented relapse prior to data cutoff, dropout, or initiation of an alternate therapy, were censored on the last date of disease or hematologic assessment.

Eighty patients were enrolled and treated with moxetumomab pasudotox (Table 1); the first patient received the first dose of moxetumomab pasudotox on May 2, 2013, and the last patient received the last dose on May 30, 2016 (data cutoff on April 29, 2019). Patients had received a median of three lines of prior systemic therapy. Thirty-nine (49%) patients had PNA-refractory disease, and 30 (38%) patients were unfit for PNA re-treatment (Table 1). Among these 30 patients, 20 patients had grade 4 neutropenia at baseline and 19 patients had a significant ongoing infection in their medical history (i.e., cellulitis, diverticulitis, hepatitis B virus, herpes zoster, sinusitis, and wound infection) and/or experienced a grade ≥ 3 infection (i.e., pneumonia and sepsis) or febrile neutropenia during screening. In all cases, the infections were adequately controlled by the start of treatment.

Fifty patients (63%) completed six cycles of treatment, and 37 patients (46%) completed the study phase, defined as being followed up to Day 181 after the last treatment (irrespective of whether or not they completed treatment). Of the 30 patients (38%) that did not complete six cycles of treatment, 12 had a CR with MRD negativity; 12 discontinued due to AEs; two had disease progression; one died as a result of septic shock that was unrelated to treatment; and three discontinued for other reasons (Additional file 1: Figure S1).

Overall, 43 patients (54%) discontinued the study: Three (4%) withdrew consent, two (3%) were lost to follow up, four (5%) died, and 34 (43%) discontinued for other reasons, mainly due to disease progression or because they started an alternative anti-cancer therapy (Additional file 1: Figure S1). All deaths were considered unrelated to study treatment.

Disease responses, as assessed by BICR and the investigators for the ITT population, are presented in Table 2. At a median follow-up of 24.6 months (range: 1.2–71.7 months), the durable CR rate (CR with HR > 180 days) was 36% (29 patients; 95% CI, 26–48%). This surpassed the target rate of 28% and was significantly higher than the historical control (lower bound of 95% CI > 13%). Moreover, the CR rate with HR ≥ 360 days was 33% (26 patients; 95% CI 22–44%). Thirty-three patients had CR, 8 of whom of had a spleen size > 13 cm (range: 13.2–16.1 cm); all 8 of these patients remained in HR at 180 days, and 7 of them remained in HR at 360 days.

In the majority of cases (51 out of 80 patients, 64%), response assessments were concordant between BICR and local settings. Among the 33 BICR-categorized CRs, 29 were also investigator-categorized CRs, including 28 durable CRs; among the 42 investigator-categorized CRs, 13 were not reported as CR by BICR. The discrepancies were due to differences in assessment of bone marrow involvement, technical challenges (e.g., unavailable/unevaluable H&E slides), and incomplete disease assessments by local investigators. Some level of variability is to be expected with low levels of disease (e.g., 1–2% involvement).

Based on BICR, 60 patients (75%; 95% CI, 64–84%) had an OR, of which 33 (41%; 95% CI, 30–53%) had a CR; both the median CR duration and median HR duration from the onset of CR were 62.8 months (Fig. 1a, b). Landmark rates of CR at 6, 12, 18, 24, 36, 48, and 60 months were 94%, 77%, 74%, 70%, 61%, 61%, and 61%, respectively. The median OR duration was 66.7 months, and the median HR duration from the onset of HR was 45.8 months (Fig. 1c). Finally, median PFS was 41.5 months if a progression event was considered as the loss of any CR criteria; median PFS was longer (71.7 months) if a progression event had to be accompanied by loss of HR (Fig. 1d).

Notably, the median estimates for the duration of response (e.g., CR and HR) were only reached at the last event, when a single patient was at risk. For this reason, the landmark analyses provide a more robust measure of treatment efficacy than the median estimates. Nevertheless, these analyses should be interpreted with caution owing to the small number of patients at the later time points.

Overall, 27 (82%) of the 33 patients with a CR also demonstrated MRD negativity (BICR), representing 34% of all patients. Both the median duration of CR (62.8 vs 12.0 months; Fig. 2) and median duration of HR from CR (62.9 [95% CI: NE] vs 12.0 months [95% CI 5.88–NR months]) were longer in MRD-negative patients than in MRD-positive patients.

For the primary analysis, safety data were collected up to 30 days after the last dose of moxetumomab pasudotox was administered. SAEs were followed until resolution or death. All patients had completed at least 6 months of post-treatment follow-up by the date of the primary analysis cutoff. Therefore, limited safety data were collected after this date (no new information on SAEs), and safety results in this long-term follow-up analysis remain largely unchanged [29].

Overall, 79 subjects (99%) experienced at least one AE, of which the most common were peripheral edema (39%), nausea (35%), and fatigue (34%) (Additional file 2: Table S1). The most commonly occurring grade 3–4 events were decreased lymphocyte count in 16 patients (20%), anemia in 8 patients (10%), and asymptomatic hypophosphatemia in 8 patients (10%). AEs not reported as related to study treatment led to study discontinuation in 4 patients (5%): glioblastoma, grade 5 pneumonia, grade 1 memory impairment, and grade 4 sepsis syndrome. Secondary malignancies were not a safety issue in this study. There was one report of glioblastoma and multiple myeloma each, and six cases of non-melanoma skin cancers (1 basal cell carcinoma, 2 lipomas, 2 seborrheic keratoses, and 1 squamous cell carcinoma), but these were not reported as related to moxetumomab pasudotox treatment.

The most frequent treatment-related AEs were nausea (22 patients, 28%), peripheral edema (21 patients, 26%), headache (17 patients, 21%), and pyrexia (16 patients, 20%) (Table 3). Treatment-related grade 3–4 AEs were reported in 24 patients (30%) and treatment-related SAEs in 14 (18%). Overall, treatment-related AEs led to study drug discontinuation in 8 patients (10%): hemolytic uremic syndrome (HUS; 4 patients, 5%), capillary leak syndrome (CLS; 2 patients, 3%), and increased blood creatinine (2 patients, 3%). Notably, HUS and CLS events were generally reversible.

Key laboratory findings are reported in  Additional file 3: Table S2. All serum creatinine laboratory values stayed within normal limits up to 12 months post-treatment, and there was no decline in renal function over time ( Additional file 4: Figure S2). With a median follow-up of 24.6 months, four deaths were reported: two due to HCL progression and two due to an AE (1 each of pneumonia and septic shock). The fatal pneumonia was reported in a 79-year-old female 30 days after the last administration of study medication. Given the long delay between this event and the last administration of study medication, the event was considered not to be related to moxetumomab pasudotox by the investigator. The septic shock was reported in a 70-year-old male, known to have grade 3 neutropenia at baseline. One hundred and twenty-nine days after starting study medication, this patient was hospitalized for septic shock due to Staphylococcus and died 4 days later. In light of the baseline neutropenia and the 4 months of therapy with infectious complications, this event was considered not to be related to moxetumomab pasudotox by the investigator. Therefore, overall no deaths were considered to be related to moxetumomab pasudotox administration.