Abstract
Purpose: Alvocidib is a cyclin-dependent kinase 9 inhibitor leading to downregulation of the antiapoptotic BCL-2 family member, MCL-1. Alvocidib has shown clinical activity in a timed sequential regimen with cytarabine and mitoxantrone in relapsed/refractory and newly diagnosed acute myeloid leukemia (AML) but has not been studied in combination with traditional 7+3 induction therapy.
Patients and Methods: A multiinstitutional phase I dose-escalation study of alvocidib on days 1–3 followed by 7+3 (cytarabine 100 mg/m2/day i.v. infusion days 5–12 and daunorubicin 60 mg/m2 i.v. days 5–7) was performed in newly diagnosed AML ≤65 years. Core-binding factor AML was excluded.
Results: There was no MTD on this study; the recommended phase II dose of alvocidib was 30 mg/m2 i.v. over 30 minutes followed by 60 mg/m2 i.v. infusion over 4 hours. There was one dose-limiting toxicity of cytokine release syndrome. The most common grade ≥3 nonhematologic toxicities were diarrhea (44%) and tumor lysis syndrome (34%). Overall, 69% (22/32) of patients achieved complete remission (CR). In an exploratory cohort, eight of nine (89%) patients in complete remission had no measurable residual disease, as determined by a centralized flow cytometric assay. Clinical activity was seen in patients with secondary AML, AML with myelodysplastic syndrome–related changes, and a genomic signature of secondary AML (50%, 50%, and 92% CR rates, respectively).
Conclusions: Alvocidib can be safely administered prior to 7+3 induction with encouraging clinical activity. These findings warrant further investigation of alvocidib combinations in newly diagnosed AML. This study was registered at clinicaltrials.gov identifier NCT03298984.
Translational Relevance
Standard front-line therapy for younger fit patients with newly diagnosed acute myeloid leukemia (AML) includes induction chemotherapy with continuous infusion cytarabine and an anthracycline (i.e., 7+3) with or without targeted agents such as midostaurin or gemtuzumab ozogamicin. However, overall outcomes are poor with 5-year survival rates <50%. Alvocidib is a cyclin-dependent kinase-9 inhibitor that leads to transcriptional suppression of MCL-1, an antiapoptotic BCL-2 family member that is upregulated in AML. Prior studies showed that alvocidib followed by cytarabine and mitoxantrone has antileukemic activity in both newly diagnosed and relapsed/refractory AML. This phase I study revealed that alvocidib 30 mg/m2 i.v. over 30 minutes followed by 60 mg/m2 i.v. over 4 hours on days 1–3 can be administered prior to 7+3 induction in newly diagnosed patients with AML with overall complete remission rates of 69% in patients with nonfavorable cytogenetics. Preliminary clinical activity was seen in secondary patients with AML who have poor outcomes with conventional chemotherapy.
Introduction
Overall outcomes remain poor in patients with acute myeloid leukemia (AML), which affects approximately 20,000 patients and more than 11,000 patients will die each year from this disease in the United States (1). In younger patients (i.e., those <60–65 years), induction therapy with 7+3 remains the mainstay of treatment despite suboptimal long-term outcomes, particularly in patients with adverse-risk disease whereby median overall survival (OS) is <1 year (2). In those with an FLT3 mutation, midostaurin, an oral FLT3 inhibitor, improves OS but not complete remission (CR) rates when added to 7+3 induction (3). However, FLT3 mutations are only present in approximately 25%–30% of newly diagnosed AML. The addition of gemtuzumab ozogamicin, an antibody–drug conjugate targeting CD33, to 7+3 induction was shown to improve event-free survival (EFS) but not OS in newly diagnosed patients with AML ages 50–70 years (4). However, those with adverse-risk disease do not benefit from the addition of gemtuzumab ozogamicin (5). Novel agents are needed to improve clinical outcomes in the front-line treatment of AML, particularly those with adverse-risk disease.
Alvocidib is a potent, nonselective cyclin-dependent kinase (CDK) 9 inhibitor with activity against CDK4, 5, 7, 8, and 11. CDK9 forms a complex with cyclin T1 (PTEF-b) which is recruited by bromodomain-containing protein-4 and mediator to superenhancer DNA complexes to regulate the activity of RNA polymerase II. In turn, RNA polymerase II catalyzes the transcription of genes regulating cell survival and proliferation, such as STAT3, c-MYC, and MCL-1. MCL-1 is upregulated in AML and contributes to leukemia cell survival and resistance to apoptosis (6, 7) MCL-1 is also necessary for survival of leukemic stem cells, the population of cells responsible for minimal residual disease (MRD), and facilitates AML progression (8). Blockade of MCL-1 has antileukemia activity in vivo (6, 8). Inhibition of CDK9 prior to S phase–specific cytotoxic chemotherapy agents may lead to synergistic apoptosis of leukemia cells (9).
Alvocidib has been investigated in a timed sequential therapy (TST) approach in combination with cytarabine and mitoxantrone (ACM) in 405 patients with both newly diagnosed (n = 256) and relapsed/refractory (n = 149) AML with encouraging findings (10–17). To date, there have been two dosing strategies done with alvocidib: a 60-minute intravenous bolus versus a “hybrid” dosing of a 30-minute loading dose followed by a 4-hour infusion. Hybrid dosing was developed to mitigate protein binding of alvocidib and maintain more sustained pharmacokinetic activity of alvocidib (18). A randomized phase II trial of bolus versus hybrid ACM in newly diagnosed poor-risk revealed similar clinical activity between both regimens though a lower dose of hybrid alvocidib was studied compared with previous MTD seen and overall CR was nonsignificantly higher in hybrid arm (74% vs. 62%, respectively; ref. 15). Given the ease of administration, bolus alvocidib was chosen for further development in newly diagnosed AML. Subsequently, a randomized phase II clinical trial of bolus ACM led to higher CR rates compared with one or two cycles of 7+3 induction therapy (70% vs. 57%; P = 0.07) in newly diagnosed patients with AML with nonfavorable-risk cytogenetics (17, 19). However, alvocidib has not been studied in the context of conventional induction therapy with 7+3. Therefore, we designed a phase I dose-escalation study of alvocidib followed by 7+3 in newly diagnosed patients with AML with nonfavorable-risk cytogenetics to assess the safety, MTD, and clinical activity of this regimen. We chose to utilize the hybrid dosing of alvocidib based on our prior experience.
Patients and Methods
Study population
This was a phase I, open-label, multicenter, dose-escalation study of alvocidib followed by cytarabine/daunorubicin (7+3) in patients with AML (NCT03298984). Patients ages 18–65 years with newly diagnosed, previously untreated AML were eligible. Hydroxyurea was permitted prior to enrollment. Full eligibility criteria are listed in the Supplementary Appendix. Patients were excluded if they had previous treatment for AML, acute promyelocytic leukemia, or core-binding factor AML. All patients provided written informed consent. This study was conducted as per the Declaration of Helsinki after approval by ethics committee of each participating center.
Treatment plan
Alvocidib was dose escalated starting at dose level 1: 20 mg/m2 30-minute i.v. bolus followed by 30 mg/m2 i.v. infusion over 4 hours on days 1–3 (Fig. 1). Daunorubicin 60 mg/m2 i.v. bolus over 15 minutes was initiated on days 5–7, and cytarabine 100 mg/m2/day i.v. continuous 24-hour infusion was administered on days 5–11. A bone marrow (BM) aspirate/biopsy was performed on day 14 (±3 days), and patients with residual leukemia (>5% BM blasts and >10% cellularity) were recommended to receive a second induction cycle with alvocidib days 1–3 (same dose level as induction) followed by daunorubicin 45 mg/m2/day i.v. over 15 minutes on days 5–6 and cytarabine 100 mg/m2/day i.v.continuous infusion days 5–9.
CONSORT diagram. *Three subjects were excluded—one subject did not meet inclusion criteria. Two subjects chose treatment with standard of care.
Patients who achieved CR received two to four cycles of consolidation therapy with high-dose cytarabine (HiDAC) 1.5–3 g/m2 i.v. every 12 hours days 1, 3, and 5 upon full hematologic recovery. Allogeneic stem cell transplantation (alloSCT) was permitted after induction.
Toxicity and response assessments
Dose-limiting toxicities (DLTs) were defined on the basis of the NCI CTCAE version 4.03 and outlined in Supplementary Appendix.
Response assessment was performed by a BM aspirate/biopsy at the time of full hematologic recovery or by day 50 and 60 of one versus two induction cycles, respectively, and assessed by standardized European Leukemia Net (ELN) Guidelines (20).
MRD Analysis
MRD was assessed in BM samples at the time of response by a uniform central assay (Hematologics, Inc.) in an exploratory cohort, as has been described previously (21). A total of 200,000 events were analyzed for each sample. The sensitivity of this assay was 0.02%. Details on the methodology of this assay are included in the Supplementary Appendix.
Mitochondrial priming
Leukemia dependence on BH3 member proteins was assessed, as described previously (22). For evaluation of MCL-1 dependence, we utilized the MCL-1–binding protein, MS1, with modifications allowing for improved cell penetrance, termed T-MS1 (23). T-MS1 has higher potency and affinity for MCL-1 than NOXA (24). Details on methodology of this assay are described in Supplementary Appendix.
Statistical analysis
The primary objective of this study was to establish an MTD of alvocidib prior to 7+3 induction. Alvocidib dose was escalated using a 3+3 design (Fig. 1). Successive cohorts of patients (3–6 per cohort) were treated with escalated doses until the MTD was established.
Secondary objectives included assessment of overall response rates [ORR; CR/Cri (complete remission with incomplete recovery) + partial remission (PR)], OS, relapse-free survival (RFS) and EFS. Kaplan–Meier time-to-event analyses was performed on OS, EFS, RFS, and duration of CR. Database lock was on April 30, 2020. All statistical analyses were performed using SAS software version 9.4 or higher.
Results
Patient characteristics
Between December 2017 and September 2019, 32 patients were enrolled to this study. Patient characteristics are shown in Table 1. The median age was 58 years, six (19%) had secondary AML, while 12 (38%) had AML with myelodysplasia-related changes [MRC; defined by MDS-related cytogenetics (25) or history of MDS/chronic myelomonocytic leukemia (CMML)]. By ELN classification, nine (28%) were favorable, seven (22%) were intermediate, and 16 (50%) were adverse-risk. Similarly, 12 (38%) patients had unfavorable-risk cytogenetics by Southwest Oncology Group (SWOG) classification (26). The most common mutations seen in this cohort were NPM1 (31%), ASXL1 (19%), and RUNX1 (16%) mutations.
Patient characteristics.
Safety
The most common overall treatment-emergent adverse events included diarrhea (n = 29, 91%, grade ≥3 = 14, 44%), nausea (n = 20, 63%, grade ≥3 = 0, 0%), vomiting (n = 13, 41%, grade ≥3 = 0, 0%), fatigue (n = 11, 34%, grade ≥3 = 0, 0%), and tumor lysis syndrome (TLS; n = 11, 34%, grade ≥3 = 11, 34%; Supplementary Table S1). The most common ≥grade 3 treatment-emergent adverse events were diarrhea (44%) and TLS (34%; Table 2). These toxicities resolved with supportive interventions as outlined in Patients and Methods and were not considered dose limiting. Diarrhea did not lead to any modifications or delays with alvocidib dosing. One patient experienced a DLT (TLS, acute kidney injury, cytokine release syndrome: CRS) at the highest alvocidib dose level studied (30 mg/m2 i.v. bolus followed by 60 mg/m2 i.v. over 4 hours). Dexamethasone was initiated for CRS with rapid resolution of symptoms. The MTD was not reached and the highest alvocidib dose level was determined to be the recommended phase II dose (RP2D).
Treatment-emergent grade ≥3 nonhematologic toxicities.
An expansion cohort enrolled 23 patients at the RP2D (Table 1). Overall, 30-day and 60-day mortality was 3%. One patient died on day 26 due to septic shock during reinduction. In those who achieved CR, the median time to partial and full neutrophil recovery (i.e., ≥0.5 × 109/L and ≥1 × 109/L, respectively) and partial and full platelet recovery (i.e., ≥50 × 109/L and ≥100 × 109/L, respectively) was 34 and 36 days and 30 and 35 days.
Clinical activity
Among all enrolled patients, ORR and CR rates were 75% and 69%, respectively (Table 3). All patients who achieved a CR achieved full hematologic recovery. Among response-evaluable patients, the ORR and CR rates were 77% and 71%, respectively (one patient died on day 26 of reinduction without a response assessment). Twenty-nine (91%) patients had no evidence of residual leukemia on day 14 assessment (CR: 22/29 = 76% after one cycle of induction) whereas three (9%) received reinduction for residual leukemia on day 14. None of the patients who received reinduction therapy achieved CR. Overall CR rates were 89% (8/9), 71% (5/7), and 56% (9/16) for favorable-, intermediate-, and adverse-risk patients by ELN classification, respectively. Fifty percent of patients with secondary AML (3/6) and AML with MRC (6/12) achieved CR. At the RP2D, 15/23 (65%) achieved CR.
Clinical activity—response assessments.
Of the 22 patients who achieved a CR, 19 (86%) received consolidation therapy with intermediate or HiDAC consolidation (median No. of cycles: 2; range, 1–4). Eleven (34%) proceeded to alloSCT (ELN favorable-risk: n = 3, intermediate-risk: n = 3, adverse-risk: n = 5), all of whom achieved CR with induction therapy. Of the 11 patients who achieved CR and did not receive an alloSCT, five (45%), two (18%), and four (36%) were ELN favorable-risk, intermediate-risk, and adverse-risk, respectively.
MRD exploratory cohort
Twelve (38%) patients on the expansion cohort were included in a centralized MRD flow cytometry assessment. Nine (75%) achieved CR and eight of nine (89%) were determined to be MRD-negative (intermediate-risk: seven of seven; adverse-risk: one of two; Supplementary Table S3).
Genomic signatures predictive of response
A heatmap of genomic signatures obtained at diagnosis by institutional standard next-generation sequencing (NGS) panel outlined by overall mutational landscape is demonstrated in Fig. 2 and by proportion of CR versus no CR in Supplementary Fig. S1. As expected, the majority of patients with NPM1 mutations achieved CR (8/10 = 80% CR). Interestingly, overall CR rate was 83% (5/6) and 80% (4/5) among patients with ASXL1 and RUNX1 mutations, respectively. Only 1/3 (33%) patients with a TP53 mutation achieved CR. Notably, among patients with a previously classified genomic signature specific for secondary AML (i.e., ASXL1, BCOR, EZH2, SF3B1, SRSF2, STAG2, U2AF1, or ZRSR2; ref. 27), 11/12 (92%) achieved CR. Next, we analyzed overall response among patients subdivided into the proposed genomic classification by Papaemmanuil and colleagues (ref. 28; Supplementary Fig. S2). The most common genomically defined subgroups in this cohort were AML with NPM1 mutation (n = 10) and AML with mutated chromatin, RNA-splicing genes, or both (n = 10). CR rates were 80% (8/10) and 90% (9/10) in patients classified as AML with NPM1 mutation and AML with mutated chromatin, RNA-splicing genes, or both, respectively.
Heatmap of genomic signatures and response. A heatmap representing patients with baseline AML mutation achieving CR (green) or no CR (red). Positive and negative AML mutations are shown in white and gray, respectively. Unknown/not tested mutations are shown in black. One patient who achieved a CR without detectable MRD was not assessed for the full panel of 15 genes and was excluded from this analysis.
Clinical outcomes
Figure 3 displays a Swimmer plot of all patients enrolled (n = 32). Of the 22 patients who achieved CR, seven (32%) relapsed to date (median duration of CR: 8.6 months; range, 1.4–13.6 months). Two (18%) patients who achieved CR without MRD relapsed (including one who underwent an alloSCT), while nine (82%) who achieved CR without MRD remain in CR. Eleven (34%) patients died. Causes of death were leukemia-related complications (n = 9), septic shock during reinduction therapy (n = 1) and disseminated mucormycosis after consolidation therapy while in CR (n = 1).
Clinical outcomes. Swimmer plot of best treatment response and survival for all 32 patients (A). The swim lanes represent subjects in the study and indicate their progression and survival in the trial along with response to therapy. The horizontal axis depicts survival in months since the first dose of alvocidib. Colors of the swim lanes depict best response to treatment [orange, CR (MRD-negative), blue, CR/CRi; green, PR; gray, no response]. Solid black circles () on the swim lanes represent earliest CR/CRi or PR, black triangles (
) represent allogenic stem cell transplantation, black circles (
) represent relapse/progression, and red stars represent death. Follow-up data (up to a protocol-specified maximum of 2 years) is still being accrued. OS (B), EFS (C) from day 1 of treatment up to death, relapse, or no response to treatment and RFS (D) defined as time from CR/CRi up to disease relapse or death, for all patients using Kaplan–Meier estimates. Median OS and RFS could not be calculated.
Figure 3 depicts the OS, EFS, and RFS of the 32 patients enrolled on this study. Mean and median duration of follow-up was 11.4 and 9.2 months, respectively. The median OS was not reached because of relatively short duration of follow-up. Landmark 1-year OS was 62.4% [95% confidence interval (CI), 41.9–77.4]. Median EFS was 10.0 months (95% CI, 2.0–NA) while median RFS was not reached. Landmark 1-year and 2-year EFS was 40.9 % (95% CI, 21.9–59.1) and 34.1% (95% CI, 15.4–53.8), respectively. Landmark 1-year RFS was 59.5% (95% CI, 31.7–79.1).
Mitochondrial priming correlates
MCL-1 dependence was assessed by mitochondrial profiling from pretreatment diagnostic BM samples in 27/31 (87%) response-evaluable patients. Median MCL-1 score was 25.2% (range, 7.0%–46.8%). There was no significant correlation of MCL-1 score among CR versus no CR (Supplementary Fig. S3). Twelve (44%) and two (7%) patients had MCL-1 priming scores >30% and >40%, respectively. CR rate was 83% (10/12) and 67% (10/15) among those with MCL-1 dependence > and <30%, respectively. Eleven of 12 patients with genomically defined secondary AML (27) had MCL-1 analysis performed (median MCL-1: 31.5%; range, 7%–38.9%). Of those analyzed, seven (64%) had MCL-1 priming >30%, while four (36%) were <30%.
Discussion
Our findings demonstrate that alvocidib combined with standard 7+3 induction chemotherapy is feasible and effective for younger patients (≤65 years) with newly diagnosed AML. As seen in our previous studies with alvocidib in AML, the most frequent treatment-related AEs were fatigue, nausea, diarrhea, and TLS (11–15, 17, 19). The most significant of these, TLS and diarrhea, were manageable with appropriate prophylaxis and supportive interventions and were not dose limiting. TLS occurred rapidly after the first dose, was predominantly laboratory based, and generally resolved without clinical sequelae. Similarly, we observed a secretory diarrhea associated with alvocidib, as seen previously (10), that occurs rapidly after the first dose and responds to antidiarrheal medications. There was one DLT of CRS leading to acute kidney injury requiring temporary dialysis seen at the highest dose level. Although CRS was not separated from adverse events documented as TLS in previous studies with alvocidib in AML, ≥grade 4 TLS was often accompanied by symptoms of CRS. In fact, infusional alvocidib has been shown to increase proinflammatory cytokines, such as IL6, potentially inciting an inflammatory milieu that can lead to CRS (29). This is the first report to specify alvocidib-associated CRS in AML; however, based on the toxicities seen in prior studies and noted TLS overlap, this is unlikely a new phenomenon and rigorous monitoring for both TLS and CRS is required with alvocidib.
There was not an appreciable delay in hematologic recovery seen in patients on this study despite the addition of alvocidib 3 days prior to 7+3 induction. The median time to partial neutrophil and platelet recovery was 34 and 30 days, respectively, which is similar to CPX-351 (29) and TST (17) and may be comparable with conventional 7+3 treatment regimens (30). Moreover, treatment-related mortality was low with only one (3%) patient expiring within 60 days of treatment. The RP2D from this study—30 mg/m2 i.v. bolus followed by 60 mg/m2 i.v. over 4 hours—is consistent with the MTD of a previous phase I study of alvocidib followed by cytarabine and mitoxantrone (14).
The primary endpoint of this study was to establish an MTD and RP2D of alvocidib followed by 7+3 in newly diagnosed AML. Small numbers of patients enrolled on this study precluded rigorous assessment of clinical activity of this regimen in comparison with historical controls. Achieving CR is associated with longer RFS and OS when compared with nonresponders and those achieving CR without full recovery (31). All of the patients who achieved CR on this study obtained full neutrophil and platelet recovery. The CR of alvocidib followed by 7+3 appears to be similar to the composite CR of TST ACM treatment in newly diagnosed AML (69% vs. 68%) though a direct comparison is not possible due to disparate patient populations studies (15, 17). Although we excluded patients with favorable-risk cytogenetics, it is worth noting that 28% of enrolled patients were favorable-risk by ELN criteria due to either NPM1 mutation (n = 8) or CEBPA biallelic mutation (n = 1). Notably, however, only 9% of patients on this study had residual AML on day 14 BM assessment thus negating the need for reinduction therapy for the vast majority of treated patients. In contrast, 25% and 44% of patients treated with ACM and 7+3, respectively, were found to have residual AML on day 14 in a randomized phase II study (17). Reinduction therapy increases the risk of comorbidities such as organ toxicity, infectious complications, and anthracycline-induced cardiotoxicity, delays hematologic recovery and prolongs hospitalization. Increasing the efficacy of induction therapy without the need for subsequent cytotoxic chemotherapy cycles would be highly advantageous.
Patients with secondary AML and/or AML with MRC have dismal outcomes with conventional chemotherapy. In a randomized phase III trial of cytarabine plus amonafide versus 7+3 in newly diagnosed secondary AML, induction therapy with 7+3 yielded CR rates of 45% and median OS of 7 months (32). Furthermore, a randomized phase III study of CPX-351, liposomal cytarabine and daunorubicin, versus 7+3 in newly diagnosed AML with MRC in patients 60–75 years revealed CR rates of 37% versus 26%, respectively, and approximately 33% of patients in both arms required reinduction therapy (33). In comparison, 50% of the patients with secondary AML or AML with MRC achieved CR on this study though 50% and 42% were <60 years, respectively, which could account for some of these differences.
Lindsley and colleagues previously reported that mutations in one of eight genes (ASXL1, BCOR, EZH2, SF3B1, SRSF2, STAG2, U2AF1, or ZRSR2) is highly specific for secondary AML (27). Although ASXL1 mutations are defined as adverse-risk by ELN criteria, the other seven mutations specific for secondary AML are not specifically listed as adverse-risk by standardized criteria. Recent data suggest that, among older patients with intermediate-risk AML, mutations associated with secondary AML had significantly worse outcomes. A two-class risk assessment from this analysis defined patients with adverse-risk and those with intermediate-risk with secondary AML mutations as “high-risk” disease (34). We noted an encouraging CR rate of 92% (11/12) in patients with a genomic profile consistent with secondary AML. These findings are consistent with previous studies showing particular clinical activity of alvocidib-containing induction regimens in newly diagnosed secondary AML suggesting that the addition of alvocidib to cytotoxic chemotherapy backbones may overcome at least some of the adverse-risk biology of secondary AML (12, 13, 15, 17).
MRD determined by multicolor flow cytometry, quantitative PCR, or NGS is an important prognostic factor impacting the likelihood for relapse and survival after induction and prior to alloSCT (35–38). Multicolor flow cytometric evidence of MRD in CR1 leads to significantly worse outcomes compared with MRD-negative CR after induction therapy (39–41). In a large prospective MRD analysis from the National Cancer Research Institute AML17 trial, only 40% of younger patients treated with diverse “7+3” induction therapy backbones achieved CR without MRD after one cycle. Furthermore, 5-year OS was 63% versus 44% among patients with CR and MRD-negative versus MRD-positive after one cycle of induction therapy (41). These studies reinforce the significant clinical impact of achieving an MRD-negative CR after induction therapy and suggest that new approaches are needed to target MRD. To evaluate alvocidib's role in eliminating MRD, an exploratory cohort of 12 patients were treated at the RP2D to prospectively assess MRD status after one cycle of induction by centralized flow cytometry. Of these 12 patients, nine (75%) achieved CR, and eight (67%) achieved an MRD-negative CR with a detection threshold of <0.02%. New approaches are imperative to target leukemic cells to convert MRD-positive patients to an MRD-negative particularly before alloSCT. Further study is warranted to specifically address whether the addition of alvocidib to 7+3 can decrease MRD-positivity, lead to deeper responses and subsequently improve clinical outcomes.
On the basis of alvocidib's mechanism of action as a CDK9 inhibitor, we hypothesized that patients with AML whose leukemia cells are dependent on MCL-1 for survival may have a predilection for response to alvocidib. Mitochondrial profiling assesses the relative dependence of antiapoptotic BCL-2 peptides in mediating cell survival within a tumor (42). NOXA is a BH3 sensitizer that selectively binds to and antagonizes MCL-1 leading to apoptosis in cells dependent on MCL-1 for survival. A NOXA mimetic peptide (T-MS1) “primes” cells for apoptosis, and high priming scores reflect cells that are considered to be MCL-1 dependent. We previously found that MCL-1 dependence was associated with response to treatment with TST ACM induction in newly diagnosed AML (43). Thus, we performed an exploratory prospective analysis of MCL-1 dependence on response to alvocidib followed by 7+3. We defined the criteria of MCL-1 dependence as a priming threshold of >30%. Although there are no uniform criteria for defining MCL-1 dependence, the 30% threshold is consistent with our revised eligibility criteria for a randomized phase II trial of cytarabine plus mitoxantrone with or without alvocidib in patients with relapsed/refractory MCL-1–dependent AML (44). We did not appreciate any significant differences in response to alvocidib in patients with or without MCL-1 dependence though this exploratory analysis was limited by small numbers of patients in each cohort. Furthermore, we did not assess for dependence on other prosurvival mechanisms such as BCL-2 or BCL-XL dependence which should be further explored in future studies. Nonetheless, 83% (10/12) of patients with MCL-1 dependence achieved CR. Given alvocidib's multi-CDK inhibitory activity and subsequent inhibition of RNA polymerase II, it is likely that alvocidib exerts antileukemia activity more broadly than direct MCL-1 inhibition (9, 45) and may provide a therapeutic advantage over selective CDK9 and MCL-1 inhibitors even in patients considered to be MCL-1 dependent. A randomized phase II clinical trial is warranted comparing standard induction chemotherapy with or without the addition of alvocidib with prospective evaluation and stratification based on MCL-1 dependence.
In conclusion, alvocidib administration prior to 7+3 induction is tolerable, feasible, and showed encouraging clinical activity in newly diagnosed AML with nonfavorable risk cytogenetics. TLS and diarrhea are the most common severe toxicities that can be managed adequately with supportive care measures. Most notably, alvocidib followed by 7+3 led to a 92% CR rate in a genomically defined signature of secondary AML that has traditionally poor clinical outcomes. In an exploratory cohort, high rates of MRD-negative CR rates were obtained with alvocidib followed by 7+3. These data warrant a randomized clinical trial of alvocidib with or without 7+3 in newly diagnosed AML.
Authors' Disclosures
J.F. Zeidner reports grants from Tolero Pharmaceuticals (research funding) during the conduct of the study; personal fees from AbbVie (independent review committee), Agios (advisory board), Bristol Myers Squibb/Celgene (advisory board, consultancy), Daiichi-Sankyo (advisory board), Genentech (advisory board), Pfizer (advisory board), Takeda (advisory board, consultancy), Tolero Pharmaceuticals (advisory board), AsystBio Laboratories (consultancy); grants from AROG (research funding), Bristol Myers Squibb/Celgene (research funding), Forty Seven (research funding), Merck (research funding), Takeda (research funding), and Tolero Pharmaceuticals (research funding) outside the submitted work. D.J. Lee reports other from Tolero (institutional research funding) during the conduct of the study; personal fees from Celgene (consulting); other from AbbVie (institutional research funding), Bayer (institutional research funding), Forty Seven (institutional research funding), Genentech (institutional research funding), and Novartis (institutional research funding) outside the submitted work. M. Frattini reports personal fees from Celgene, Bristol-Myers Squibb, Lin BioSciences, and Cellectis outside the submitted work. K. Kolibaba reports personal fees and other from TG Therapeutics (consultancy, research funding); other from AbbVie (research funding), Acerta (research funding), Celgene (research funding), Cell Therapeutics (research funding), Genentech (research funding), Gilead (research funding), Janssen (research funding), Novartis (research funding), Pharmacyclics (research funding), and Seattle Genetics (research funding) outside the submitted work. S.P. Anthony reports personal fees from Exact Sciences (evidence review member) outside the submitted work; and work for Sumitomo Dainippon Pharma Oncology who is the sponsor of the trial. D. Bearss reports other from Sumitomo Dainippon Pharma Oncology fka Tolero Pharmaceuticals (employee and officer at this company) during the conduct of the study; in addition, D. Bearss has a patent for USPTO 10,682,356 pending to Tolero Pharmaceuticals (combination therapies for treatment of cancer), a patent for USPTO 10,624,880 pending to Tolero Pharmaceuticals (predicting response to alvocidib by mitochondrial profiling), a patent for USPTO 10,568,887 pending to Tolero Pharmaceuticals (combination therapies for treatment of cancer), a patent for USPTO 10,562,925 pending to Tolero Pharmaceuticals (alvocidib prodrugs having increased bioavailability), a patent for USPTO 10,422,788 pending to Tolero Pharmaceuticals (profiling peptides and methods for sensitivity profiling), a patent for USPTO 10,357,488 pending to Tolero Pharmaceuticals (predicting response to alvocidib by mitochondrial profiling), a patent for USPTO 10,267,787 pending to Tolero Pharmaceuticals (profiling peptides and methods for sensitivity profiling), a patent for USPTO 10,259,835 pending to Tolero Pharmaceuticals (alvocidib prodrugs having increased bioavailability), a patent for USPTO 10,132,797 pending to Tolero Pharmaceuticals (profiling peptides and methods for sensitivity profiling), a patent for USPTO 9,901,574 pending to Tolero Pharmaceuticals (predicting response to alvocidib by mitochondrial profiling), and a patent for USPTO 9,758,539 pending to Tolero Pharmaceuticals (alvocidib prodrugs having increased bioavailability); and Tolero Pharmaceuticals was funded by several individual and venture capital investors before its acquisition in 2017 by Sumitomo Dainippon Pharma, and aspects of the trial detailed in this published work were in process before and after this acquisition. Tolero Pharmaceuticals, as a single entity and then as a wholly owned subsidiary of Sumitomo Dainippon Pharma, funded the work described in this manuscript. B.D. Smith reports other from Jazz Pharma (consulting fees), Novartis (consulting fees), Pfizer (consulting fees), Celgene (DSMB), and Agios (consulting fees) outside the submitted work. No disclosures were reported by the other authors.
Authors' Contributions
J.F. Zeidner: Conceptualization, formal analysis, investigation, writing-original draft, writing-review and editing. D.J. Lee: Conceptualization, formal analysis, investigation, writing-original draft, writing-review and editing. M. Frattini: Conceptualization, investigation, writing-review and editing. G.D. Fine: Formal analysis, writing-review and editing. J. Costas: Data curation, writing-review and editing. K. Kolibaba: Data curation, writing-review and editing. S.P. Anthony: Conceptualization, writing-original draft, writing-review and editing. D. Bearss: Conceptualization, writing-original draft, writing-review and editing. B.D. Smith: Conceptualization, investigation, writing-original draft, writing-review and editing.
Acknowledgments
This study was supported by Tolero Pharmaceuticals, acquired by Sumitomo Dainippon Pharma.
The authors would like to thank Judith Karp for her contribution to the study design and development plan of alvocidib. We would like to thank the research staff and all co-investigators at University of North Carolina, Johns Hopkins, and Columbia University. We would also like to thank all patients and their families for trusting us in their care and allowing us to conduct this study.
Sonali Lokhande, MD, a medical writer from Criterion Edge supported by funding from Sumitomo Dainippon Pharma Oncology/Tolero Pharmaceuticals, provided editorial assistance to the authors during preparation of this manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Footnotes
Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).
Prior presentation: These data were presented in part at the annual 2019 European Hematology Association (EHA) meeting in Amsterdam, Netherlands as a poster presentation and at the 2020 Virtual EHA meeting as a poster presentation.
Clin Cancer Res 2021;27:60–9
- Received July 6, 2020.
- Revision received August 26, 2020.
- Accepted September 23, 2020.
- Published first September 30, 2020.
- ©2020 American Association for Cancer Research.