Nilotinib in patients with systemic mastocytosis: analysis of the phase 2, open-label, single-arm nilotinib registration study

Enrollment criteria for the phase 2, multicenter 2101 trial have been previously described (H. M. Kantarjian et al. 2007). Briefly, adult patients with hematologic malignancies were recruited into 6 parallel treatment arms. Patients who met the standard disease criteria for SM (at least 1 major and 1 minor or 3 minor criteria for SM) (Valent et al. 2003) and needed treatment were recruited into the SM arm of the study and were assessed for efficacy according to a Simon two-stage minimax design (Simon 1989). The major criterion for SM is the presence of multifocal clusters of mast cells in the bone marrow. The minor criteria are the presence of spindle-shaped mast cells in the marrow, elevated serum tryptase levels, abnormal expression of CD2 or CD25, and the presence of the KIT D816V mutation (Valent et al. 2001). Key inclusion criteria included World Health Organization (WHO) performance status ≤2. Exclusion criteria included disease infiltration into the central nervous system and impaired cardiac function (including a left ventricular ejection fraction <45 %, congenital long QT syndrome, or QTc >450 ms).

Recruitment occurred between June 9, 2005, and March 31, 2006. Data on the significance of improvements in clinical findings (C-findings) as a measure of treatment response and clear definitions of the various types of SM were published after initiation of the study (Gotlib et al. 2013; Horny et al. 2008). To incorporate these data into the study, C-findings were collected, and patients were categorized as having ASM, ISM, or other SM subtype retrospectively based on WHO criteria for these SM subtypes (Horny et al. 2008). Per the WHO criteria, patients meeting SM criteria who had one or more C-findings were classified as having ASM as long as the criteria for MCL were not met.

Informed consent was obtained from all patients according to institutional guidelines. The study was conducted in accordance with the Declaration of Helsinki; the protocol and all amendments were reviewed and approved by the ethics board or institutional review board at each participating trial center.

Nilotinib 400 mg twice daily was selected as the study dose based on safety and tolerability and pharmacokinetic and preliminary efficacy data from the phase 1 portion of this study (H. Kantarjian et al. 2006). One treatment cycle consisted of 28 days of continuous dosing with nilotinib 400 mg twice daily. Patients remained on nilotinib treatment until disease progression or unacceptable toxicity that precluded additional therapy. Patients responding to therapy were allowed to enter an extension study after 24 months of treatment and continue receiving drug.

Dose modifications, including dose reductions and dose interruptions, were allowed for patients who developed intolerance to nilotinib. Patients with treatment interruptions longer than 21 days were discontinued from the study, except those with hematologic toxicity, in whom interruptions were permitted for up to 42 days. Patients discontinuing due to adverse events (AEs) were monitored until resolution or stabilization of the AE. All patients were monitored for AEs for 28 days after the final dose of nilotinib.

The primary efficacy endpoint was the overall response rate (ORR), defined as the proportion of patients with a minor response or better lasting a minimum of 4 weeks (Valent et al. 2003). Response was evaluated using improvements in laboratory findings (for all patients) and ASM response criteria (for the ASM subgroup). For patients with ASM, response evaluations were based on findings in bone marrow, peripheral blood, and extramedullary sites of disease and improvements in C-findings (Valent et al. 2007). C-findings were defined as hematologic abnormalities [absolute neutrophil count (ANC) <1.0 × 10⁹/L, hemoglobin level <10 g/dL, or platelet count <100 × 10⁹/L], hepatomegaly with ascites and impaired liver function, portal hypertension, palpable splenomegaly with hypersplenism, malabsorption with hypoalbuminemia and/or weight loss, and skeletal lesions (such as large-sized osteolyses and/or severe osteoporosis with pathological fracture). The definitions of complete or incomplete remission and pure clinical response included complete resolution of 1 C-finding and no progression in other C-findings. Complete remission also required disappearance of organomegaly, decrease of serum tryptase level to <20 ng/mL, and loss of all mast cell infiltrates in affected organs. Incomplete remission required >50 % decrease in serum tryptase level and/or a decrease in mast cell infiltrates and/or visible reduction of organomegaly. Pure clinical response involved no response in mast cell infiltrates, serum tryptase level, or organomegaly. Good partial response was defined as >50 % regression in ≥1 C-findings and no progression in other C-findings. Minor response was defined as ≤50 % regression of ≥1 C-findings and no progression in other C-findings. Stable disease was defined by C-finding parameters falling within a constant range. Progressive disease was defined as ≥1 C-finding with progression.

Secondary efficacy endpoints included overall survival, which was estimated using the Kaplan–Meier method. The analysis of overall survival included all deaths occurring during treatment or after discontinuation of study drug. Assessment of KIT mutation status was performed by Sanger sequencing after polymerase chain reaction (PCR) amplification, and FIP1L1–PDGFRA translocation status was determined by nested reverse-transcriptase PCR at baseline and/or at various time points throughout the study. The ORR stratified by KIT D816V mutation status at baseline or at any time was assessed.

Laboratory parameters were assessed to help characterize the quality of response in responders in the ASM subgroup and in all patients with other SM subtypes. For each laboratory parameter, values were recorded at baseline and the time of best response. The “best value” was the lowest level of tryptase and mast cell burden and the highest level for other parameters. The median baseline and best value were calculated for the responder population. The difference between these median values was then used to determine the percentage improvement in each parameter.

Toxicity was assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) version 3.0 (National Cancer Institute 2006). AEs, serious AEs (SAEs), and hematologic and nonhematologic laboratory abnormalities were assessed.

Baseline demographic and disease characteristics of the 61 patients enrolled in the SM arm of the 2101 study are listed in Table 1. Patients were retrospectively assigned to SM subgroups based on WHO 2008 criteria (Tefferi and Vardiman 2008). The patient population was heterogeneous. Most patients had ASM [n = 37 (60.7 %)] or ISM [n = 19 (31.1 %)]; of the 5 remaining patients, 3 had MCL, 1 had SM-acute myeloid leukemia (AML), and 1 had SM with an associated clonal hematologic non–mast cell lineage disease (AHNMD). At study entry, 49 patients (80.3 %) met 1 major and 1 minor disease criteria for SM; 11 patients (18.0 %) met 3 minor criteria for SM; and 1 patient (1.6 %) had a major protocol violation, defined as entering the study but not meeting the predefined SM disease group classification. In the ASM group (n = 37), C-findings were as follows: 15 patients had skeletal involvement only; 11 patients had cytopenias only; 3 patients had hepatosplenomegaly only; 3 patients had skeletal lesions accompanied by hepatosplenomegaly; and 5 patients had malabsorption, skeletal lesions, and hepatosplenomegaly. The median time since first diagnosis of SM was 25.9 months (range 1.2–287.3 months). The median patient age was 52 years (range 29–79 years). The median time since diagnosis was similar in patients with ASM and those with ISM (25.9 vs 27.4 months, respectively). At baseline, 36 patients (59.0 %) had a KIT D816V mutation, and 5 patients (8.2 %) had no KIT mutation; in 20 patients (32.8 %), KIT mutation analysis was missing. All patients had at least 1 sample evaluated for KIT mutation status at any point during the study. Of the 25 patients with no or missing mutation data at baseline, 12 (48.0 %) subsequently tested positive for KIT D816V. Using data obtained at any point during the study, 48 patients (78.7 %) had a KIT D816V mutation, and 13 patients (21.3 %) did not; no patients had a D816Y mutation. All of the 53 patients tested for the FIP1L1–PDGFRα fusion at any time during the study had a negative result.

Baseline characteristics of these retrospectively determined subgroups reflected differences in overall disease status. For example, the proportion of patients with WHO performance status of ≥2 was higher in the ASM group (18.9 %) than in the ISM group (5.3 %). Fewer patients in the ASM group had no prior treatment for SM (40.5 vs 52.6 % in the ISM group). Baseline tryptase levels were generally higher in the ASM group (27.3 % of patients had levels >200 ng/mL at baseline vs 12.5 % of patients in the ISM group). Anemia (hemoglobin level <10 g/dL), neutropenia (ANC <2 × 10⁹/L), and thrombocytopenia (platelet count <100 × 10⁹/L) were all more common in the ASM group than the ISM group (18.9 vs 0 %, 8.1 vs 0 %, and 21.6 vs 0 %, respectively). The proportion of patients with bone marrow mast cell infiltration >30 % was also greater in the ASM group (20.6 %) than in the ISM group (5.9 %). Constitutional and mediator-related symptoms were more common in patients with ASM than in those with ISM (59.5 vs 42.1 and 54.1 vs 47.4 %, respectively), whereas urticarial and cutaneous symptoms were less common in the ASM group than the ISM group (29.7 vs 42.1 and 56.8 vs 68.4 %, respectively).

At the time of data cutoff (January 19, 2009), 8 patients had entered the extension study, and 53 patients had discontinued treatment (Table 2). The most common reason for discontinuation was AEs [n = 18 (29.5 %)]. Diarrhea, fatigue, and pain in extremity (3.3 % each) were the AEs most frequently leading to discontinuation. Eleven patients (18.0 %) experienced grade 3/4 AEs leading to discontinuation. Each grade 3/4 AE occurred in only 1 patient. Eleven patients (18 %) had disease progression that led to discontinuation.

The median total daily dose of nilotinib received was close to the planned total daily dose of 800 mg [787.6 mg (range 318.9–953.5 mg)]. Median treatment duration was 232 days (range 3–1274 days). One-third of patients (32.8 %) were treated with nilotinib for a minimum of 12 months, and 18.0 % received nilotinib for ≥24 months. Nearly half of all patients had a nilotinib dose reduction [n = 29 (47.5 %)] and/or interruption [n = 30 (49.2 %)]. The most common AEs requiring dose adjustment or interruption were nausea (16.4 %), diarrhea (14.8 %), vomiting (13.1 %), and fatigue (6.6 %). Dose interruptions lasted a median of 13.5 days (range 2–47 days). Interruptions constituted a median of 5.1 % of total dosing days (range 0.2–42 %).

In the ASM subgroup, a minor response or better was achieved by 8 patients (21.6 %) receiving nilotinib 400 mg twice daily, all of whom had a KIT D816V mutation at any time during the study (Table 3). Incomplete remission and pure clinical response were each achieved by 1 patient, while good partial response and minor response were each achieved by 3 patients (8.1 %). Stable disease was observed in 16 patients (43.2 %), and 3 patients (8.1 %) had disease progression as their best response. Ten patients were not evaluable for response, in part because some patients did not have baseline C-finding data, which were collected retrospectively after response criteria for ASM became available (Gotlib et al. 2013).

In patients with ASM, responses were evaluated based on baseline KIT D816V mutation status (Table 3). Of 22 patients with mutated KIT D816V at baseline, 6 (27.3 %) achieved an overall response to treatment, including 1 pure clinical response (4.5 %), 2 good partial responses (9.1 %), and 3 minor responses (13.6 %). Nine patients with a baseline KIT D816V mutation (40.9 %) had stable disease, and 1 patient (4.5 %) had disease progression. Of 15 patients with no baseline mutation or missing KIT D816V status, 2 (13.3 %) had an overall response to treatment [including 1 incomplete remission (6.7 %) and 1 good partial response (6.7 %)], 46.7 % had stable disease, and 13.3 % had disease progression as their best response. However, 7 of 15 patients with no or missing KIT mutation status at baseline tested positive for a D816V mutation in subsequent analyses.

The baseline characteristics of ASM responders were largely similar to those of the total ASM population, but some differences were observed (Table 1). For instance, all responders had the major criterion for SM (vs 78.4 % of all patients). Further, the proportion of patients with constitutional symptoms was lower in responders than in the total ASM population (37.5 vs 59.5 %), and the proportion of patients with bone marrow mast cell infiltration <10 % was higher in responders (62.5 vs 47.1 %). The median time since SM diagnosis was longer for responders than for all patients (34.5 vs 25.9 months), and presence of the KIT D816V mutation at any time was more common in responders (100 %) compared with the total ASM population (78.4 %).

In responders in the ASM subgroup, the parameter with the highest degree of change was ANC; median best increase from baseline was 94.7 % (Fig. 1). Furthermore, when best values were compared with baseline values, the proportion of mast cells found in bone marrow decreased 70.0 %. Moderate change was seen in platelet counts (43.5 % increase) and tryptase levels (29.8 % decrease). There was little change in hemoglobin and albumin levels during the study (increases of 11.5 and 15.1 %, respectively).

Because the ASM criteria for response cannot be used in patients with other SM subtypes, patients in other subgroups were evaluated based on changes in laboratory parameters. In the ISM subgroup (n = 17 evaluable at baseline), the proportion of mast cells found in the bone marrow decreased 80 %, and ANC decreased by 64.6 % (Fig. 2). Platelet counts increased by 19.5 %. As in the ASM subgroup, hemoglobin and albumin levels showed minimal change during the study (increases of 4.3 and 4.5 %, respectively). The median tryptase level in the ISM group increased by 22.6 %; however, the interquartile range of tryptase levels at baseline and best value during the study showed a trend for decreasing tryptase levels.

Because patients in the “other” subgroup had different SM designations, patient laboratory values were evaluated individually rather than as a group (Fig. 3). All of these patients, except 1 patient with MCL (denoted as MCL-3 in the figure), had a KIT D816V mutation at any time. In general, most patients in the “other” subgroup experienced improvements in albumin and neutrophil levels and platelet counts. The patient with SM-AML experienced a decrease in hemoglobin, while all other patients experienced an increase in hemoglobin during the study. Because of missing values, trends in mast cell infiltration and tryptase level in this population could not be determined.

During the study, 11 deaths occurred: 9 in the ASM subgroup and 2 in the “other” subgroup (both patients with MCL). Four deaths occurred during treatment or within 28 days of study drug discontinuation, while 7 deaths occurred during long-term follow-up. Ten deaths were due to disease progression, and 1 cause of death was listed as dyspnea and weakness. Of patients with KIT D816V mutations at baseline (n = 36) or at any time (n = 48), 5 deaths were reported. The rate of overall survival at 24 months was 81.2 % (95 % CI 70.6–91.8 %), with the median overall survival not yet reached after a median of 34.7 months of follow-up (Fig. 4).

All patients reported at least 1 AE, which was not unexpected given the severe nature of the underlying disease. The most frequent study drug-related AEs included nausea (39.3 %), headache (32.8 %), fatigue (29.5 %), diarrhea (27.9 %), vomiting (26.2 %), and pruritus (24.6 %; Table 4). Gastrointestinal and skin toxicities were consistent with the overall safety profile for nilotinib (Kantarjian et al. 2007; Tasigna package insert. 2015). Newly occurring or worsening grade 3/4 hematologic laboratory abnormalities included thrombocytopenia (10.3 %), anemia (10.0 %), and neutropenia (6.9 %; Table 5). The most frequent newly occurring or worsening grade 3/4 biochemical laboratory abnormalities included decreased serum phosphate (17.2 %), increased serum lipase (17.0 %), increased serum bilirubin (13.3 %), and increased serum alanine aminotransferase (10.2 %). Study drug-related SAEs of any grade were experienced by 14 patients (23.0 %). Diarrhea (4.9 %), vomiting (3.3 %), and dehydration (3.3 %) were the only study drug-related SAEs that occurred in more than 1 patient. Patients were carefully monitored for cardiac changes occurring on study. A QTcF increase >60 ms from baseline occurred in two patients but did not result in treatment discontinuation in either patient. No patients had a QTcF >480 ms.