Patient characteristics

This analysis included 47 patients who were included in the safety analysis set (SAF) of the Japanese P1/2 study and 153 patients who received CPX-351 and included in the Intention-to-treat analysis (ITT) of the global P3 study. In the Japanese study, there were 4.3% (n = 2) of t-AML compared to 19.6% (n = 30) in the global study, and AML from MDS with prior HMA treatment was 10.6% (n = 5) in Japan compared to 32.7% (n = 50) in the global study; AML from MDS without prior HMA treatment was 74.5% (n = 35) in Japan compared to 13.7% (n = 21) in the global study. The number of cases with platelets less than 50,000/μL at baseline was 47% (n = 22) in the Japanese study, and 62% (n = 95) in the global study (Table 1). The CR/CRi rate in the first cycle of CPX-351was 51.1% in the Japanese study and 37.9% in the global study.

Table 1 Patient characteristics in the Japanese P1/2 study and global P3 studyAdverse events

The most frequently observed adverse event in both the Japanese and the global studies was febrile neutropenia (85.1% in the Japanese study and 70% in the global study), and pneumonia (25.5% in the Japanese study and 30.1% in the global study) (Table 2). The adverse events identified in the Japanese study were hypokalemia (21.3%), and those identified in the global study were respiratory failure (7.2%) and hypoxia (19.6%) (Fig. 1). Skin disorder was occurred in 38 (80.9%; 4 was G3) in the Japanese study and 124 (81.0%; 14 was G3) in the global study (Supplemental table S2). Of these cases, systemic administration of corticosteroids was required in 4 cases in the Japanese study, and 7 cases in the global study. Most of the other skin disorders could be managed with topical steroids alone. Two cases were also reported as cytarabine syndrome in Japan.

Table 2 Most frequent AEs in the Japanese P1/2 study and global P3 studyFig. 1

Most frequent AEs in the Japanese P1/2 study and Global P3 study. ≥ G3 grade 3 to 5; < G3 grade 1 to 2

Time to platelets and neutrophils recovery from induction treatment

The changes in platelet and neutrophil counts during the induction and consolidation therapy in the Japanese study were shown in Fig. 2. For all CPX-351-treated patients, the median time to platelets recovery to 50,000/μL and 100,000/μL were 36 days (interquartile range (IQR) 29–64) and 36 days (IQR 29–73), and the median time to neutrophils recovery to 500/μL and 1000/μL were 36 days (IQR 29–43) and 36 days (IQR 31–51) in the Japanese study respectively. In the global study, the median time to platelets recovery to 50,000/µL and 100,000/µL were 37 days (IQR 34–54) and 44 days (IQR 36-NA) and the median time to neutrophils recovery to 500/µL and 1000/µL were 36 days (IQR 29–48) and 41 days (IQR 34–53) (Table S3). On the other hand, among patients who achieved CR or CRi in the first cycle of CPX-351, the median time to platelet recovery to 50,000/µL and 100,000/µL was 36 days (IQR 29–47) and 36 days (IQR 29–55) in the Japanese study and 37 days (IQR 34–44) and 42.5 days (IQR 35–49) in the global study. Similarly, for CR/CRi patients, the median time to recovery of neutrophils to 500/µL and 1,000/µL was 36 (IQR 29–47) and 36.5 (IQR: 31.0–47.0) days and 35 (IQR 29–41) and 38 (IQR 34–43) days, respectively (Table S3).

Fig. 2

Neutrophils and platelets counts during treatment in Japanese P1/2 study. A Change of neutrophils during induction therapy. B Change of platelets during induction therapy. C Change of Neutrophils during consolidation therapy. D Change of platelets during consolidation therapy. I1 induction cycle 1, I2 induction cycle 2, C1 consolidation cycle 1, C2 consolidation cycle 2, EoT end of treatment, CR complete remission, CRi complete remission with incomplete blood count recovery

The changes in platelet and neutrophil counts during induction therapy in the Japanese cohort showed similar trends in neutrophil counts in both CR/CRi and non-CRi cases, but platelet counts in non-CRi cases were delayed compared to CR cases. To examine factors delaying neutrophil and platelet recovery, a univariate analysis was performed on the population that achieved CR/CRi after one cycle of induction therapy in the global cohort, but no factors delaying neutrophil and platelet recovery were identified (Supplemental Table S1).

Time to FN and pneumonia occurrence

The median time to FN occurrence was 8 days in the Japanese study and 11 days in the global study (Fig. 3A), and the median time to pneumonia occurrence was 17 days and 20 days, respectively (Fig. 3B). The median time to recovery from FN was 25 days in the Japanese study and 5 days in the global study (Fig. 3C), and the median time to recovery from pneumonia was 53 days and 20 days, respectively (Fig. 3D), with a tendency toward longer periods in Japanese study. Regarding the use of G-CSF, 25.5% (n = 12) in Japan and 33.3% (n = 51) in the global study received G-CSF. Most of the G-CSFs used were short-acting, while 6 (3.9%) global patients received pegfilgrastim. In the Japanese study, the main reason for G-CSF use was for leukopenia/neutropenia in 9 cases, for treatment of pneumonia in 2 cases, and for bacteremia in 1 case. Three of the patients were administered G-CSF in order to proceed to the next course.

Fig. 3

Occurrence and recovery from febrile neutropenia and pneumonia in Japanese P1/2 study and global P3 study. A Occurrence of febrile neutropenia in the Japanese P1/2 study and global P3 study. B Occurrence of pneumonia in the Japanese P1/2 study and global P3 study. C Recovery from febrile neutropenia in the Japanese P1/2 study and global P3 study. D Recovery from pneumonia in the Japanese P1/2 study and global P3 study

Infection prophylaxis

The number of patients who received infection prophylaxis during the initial induction therapy period is shown (Table 3). In the Japanese study, 43 patients (91.5%) received infection prophylaxis, of which 40 patients (85.1%) received antibiotic prophylaxis, with levofloxacin (n = 20, 42.6%) the most commonly used. In addition, 13 patients (27.7%) received combination therapy with sulfonamide and trimethoprim. For fungal prophylaxis, 43 patients (91.5%) received antifungals, with fluconazole (n = 24, 55.8%) the most frequently used. Only 2 patients (4.3%) received acyclovir.

Table 3 Infection prophylaxis in the Japanese P1/2 study and global P3 study

On the other hand, in the global study, 129 (84.3%) received infection prophylaxis, of which 100 patients (65.4%) received antibiotic prophylaxis, most frequently with cefepime (n = 29, 19.0%), followed by levofloxacin (n = 24, 16.3%) and ciprofloxacin (n = 24, 16.3%). Only 2 patients (1.3%) received combination therapy with sulfonamide and trimethoprim, while 89 patients (58.2%) received antifungal prophylaxis and 99 patients (64.7%) received antiviral prophylaxis (Table 3).

Gastrointestinal adverse events and emetic prophylaxis

In the Japanese study, nausea occurred in 16 (34.0%; one was G3), while in the global study, nausea occurred in 75 (49.0%; one was G3). Vomiting occurred in 4 (8.5%; all were G1 or 2) patients in the Japanese study and 39 (25.5%; one was G3) in the global study (Fig. 1, Table 2).

In the Japanese study, all cases received anti-emetic prophylaxis, with 45 (95.7%) receiving a 5-HT3 receptor antagonist (5-HT3-RA) and 2 (4.3%) receiving dexamethasone only or combination of neurokinin 1 receptor antagonist (NK1-RA) and dexamethasone. Of the cases in which 5-HT3-RA were used, 24 (51.1%) were used 5-HT3-RA alone, 12 (25.5%) with NK1-RA, 4 (8.5%) with dexamethasone, and 4 (8.5%) were given in combination with triplet (5-HT3-RA, NK1-RA and dexamethasone). Aprepitant was administered on consecutive days 1–5 in 10 cases (71.4%), on days 1–3 in 3 cases (21.4%), and on day 1 only in 1 case (7.1%). In the global study, 127 (83.0%) received anti-emetic prophylaxis. Of the cases in which 5-HT3-RA were used, 40 (26.1%) were received 5-HT3-RA alone, 65 (42.5%) were received 5-HT3-RA and dexamethasone. No cases received NK1-RA (Table 4).

Table 4 Anti-emetic prophylaxis in the Japanese P1/2 study and global P3 studyChange of LVEF during CPX-351 treatment in Japanese P1/2 study

Changes in LVEF during the treatment period are shown in Fig. 4; There were no cases in which CPX-351 treatment intervention resulted in a sudden decrease in LVEF; however, one (4.8%) patient had an LVEF < 50% at least once during the treatment period. This patient had an LVEF of 45% on day 39 of the first cycle of remission induction therapy, but there was no evidence of cardiac disease, and the LVEF recovered to 54% at the day 45 measurement (Fig. 4).

Fig. 4

Change in LVEF during CPX-351 treatment in Japanese P1/2 study. A Change in LVEF during CPX-351 treatment in all Japanese P1/2 study. B Change in LVEF in a patient with decreased LVEF during CPX-351 treatment in Japanese P1/2 study. I1 induction cycle 1, I2 induction cycle 2, EoT end of treatment