Cyclophosphamide, Alvocidib (Flavopiridol), and Rituximab, a Novel Feasible Chemoimmunotherapy Regimen for Patients with High-Risk Chronic Lymphocytic Leukemia

Abstract

Alvocidib has demonstrated efficacy in high-risk chronic lymphocytic leukemia (CLL) patients. In this phase I study, we combined cyclophosphamide, alvocidib and rituximab (CAR) in a schema designed to mitigate tumor lysis syndrome (TLS) seen previously with alvocidib. Nine nucleoside analog-naïve, high-risk patients received escalating doses of CAR therapy. Dose limiting toxicity was not experienced. No instances of TLS were observed. Patient responses included three complete remissions and four partial remissions. CAR was tolerable and active in high-risk CLL patients without TLS toxicity. With continued monitoring of toxicities, a phase Ib/II study of this combination as frontline therapy is warranted.

Introduction

Despite significant changes and progress in the treatment of chronic lymphocytic leukemia (CLL) patients, specific genomic [del(17p13.1), del(11q22.3), un-mutated immunoglobulin heavy chain variable region (IgV_H)] and clinical [age ≥70 years and β2microglobulin (B2M) level ≥ 4] risk factors continue to be associated with poor clinical outcomes [1, 2]. Chemoimmunotherapy has become the standard of care in frontline CLL therapy secondary to improvements in progression-free survival (PFS) and overall survival (OS) [3]. However, standard regimens are not curative and efforts are ongoing to optimize therapy for CLL patients.

In contrast to standard chemoimmunotherapies, alvocidib (flavopiridol), a cyclindependent kinase (CDK) inhibitor, has been shown to be effective in high-risk groups of CLL patients and does not promote the same cellular immune suppression typically seen with chemotherapy agents, such as fludarabine. Combined analysis of two early clinical phase I/II trials of alvocidib [4-6] included 112 heavily pre-treated patients [36% with del(17p13.1) and 33% with del(11q22.3)] demonstrated an impressive overall response rate (ORR) of 46% and a median progression-free survival (PFS) of approximately 10 months. There were no significant differences in ORR or PFS among cytogenetic groups [7] or in patients older versus younger than age 70 [8], suggesting efficacy of alvocidib in these high-risk populations.

However, during the phase I trials, the dose limiting toxicity (DLT) of alvocidib was hyperacute tumor lysis syndrome (TLS), occurring in 48% of patients, with 19% requiring dialysis [9]. To limit TLS, eligibility was modified in the phase II trial [6], by restricting enrollment to patients with white blood cell (WBC) count <200×10⁹/L, implementing aggressive TLS prophylaxis, and reducing cycle length and number of treatments per cycle. With these modifications, more patients completed therapy, and the severity of TLS decreased. The overall rate of TLS on the phase II trial was 44% with 6% requiring dialysis. TLS occurred most frequently in patients with Rai Stage III/IV, female gender, adenopathy ≥10cm, elevated WBC count, increased B2M, decreased albumin, and higher plasma levels of alvocidib–glucuronide (a glucuronidated metabolite of alvocidib) [9]. Concerns related to the onset of acute TLS portends to a potential limitation to its use. Therefore, we combined cyclophosphamide, alvocidib, and rituximab (CAR) with the aims of developing an effective regimen for high-risk CLL patients, while limiting toxicities and demonstrating potential feasibility of administration as an out-patient.

Materials and Methods

Patients

Patients were enrolled on the National Cancer Institute (NCI)-sponsored and The Ohio State University institutional review board-approved study following written informed consent. Enrollment criteria included: age over 17 years, symptomatic CLL or SLL by NCI criteria [1] with poor-risk genetic or clinical risk factors [presence of del(17p13.1), del(11q22.3), unmutated IgV_H (≥98% homology), age over 70 years and/or elevated B2M(≥4)], no prior therapy with purine analogs, Eastern Cooperative Oncology Group (ECOG) [10] performance status less than 3, no active infection, and adequate renal and hepatic function. Patients with cytopenias were not excluded from participation.

Study Design, Treatment Plan, and Dose Escalation Schema

The study was designed as a traditional 3×3 phase I model, where 3-6 patients were enrolled at each dose level. The maximum tolerated dose (MTD) was defined as the dose level below which 2 or more of a cohort of 6 patients experienced dose-limiting toxicity (DLT), as defined in the following section. At the MTD, the study initially planned for an expansion cohort of 12 patients for a total of 18 patients treated for further evaluation of safety for phase II trials. Table 1 provides a detailed dose schedule description for each cohort.

Table 1.

Dose Escalation Schema for Cyclophosphamide, Alvocibib, and Rituximab Regimen

Cohort	Rituximab	Cyclophosphamide	Alvocidib
1 (n=3)	100mg IV C1D1; 375mg/m2 IV C1D2 & C2-6D1	300mg/m2 IV C1-6D1-3	30mg/m2 IV (30min), 30 mg/m2 IV (4hrs) C1D8; C2-6D1&8
2 (n=3)			30mg/m2 IV (30min), 50 mg/m2 IV (4hrs) C1D8, C2-6D1&8
3 (n=3)		375mg/m2 IV C1-6D1-3

IV=intravenous, C=cycle, D=day

Dose Limiting Toxicity

DLT was defined as the occurrence of any of the following events during the first two cycles of therapy when judged to be clinically significant (as further defined below) and possibly, probably, or definitely related to the study treatment. Adverse events that were clearly the result of disease progression, concomitant medical illness, or accidental injury were not considered DLT. Toxicity grading was performed using the CTEP Active Version of the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) [11]. DLT for this study was defined as follows: (1) any grade 3 or 4 non-hematologic toxicity (with the exceptions of: reversible electrolyte or glucose abnormalities, liver function test abnormalities that returned to ≤ grade 2 in < 5 days, and TLS not requiring dialysis); (2) grade 3 or 4 hematologic toxicity that did not resolve to 20% of baseline by day 35 of therapy (unless disease-related); and (3) grade 3 infection or febrile neutropenia.

Supportive Care

All patients received allopurinol (300mg daily) for the duration of the trial. For the first 2 doses of alvocidib, patients received inpatient pretreatment hydration, urine alkalization, and prophylactic phosphate binder therapy, rasburicase (4.5mg) 2 hours prior to dose. Other supportive care measures were as detailed in previous reports [4, 6].

Response Assessment

International Workshop on Chronic Lymphocytic Leukemia Guidelines [1] were used for assessment of response. For patients not demonstrating disease progression, response status was evaluated after Cycles 2, 4, and 6 (if applicable) and 2 months following therapy completion.

Pharmacokinetics

Plasma pharmacokinetics (PK) of alvocidib and alvocidib-glucuronide were assessed in each patient on Day 8 of both Cycle 1 and 2. Plasma samples were collected pre-dose and at 0.5, 4.5 and 24 hours after the start of alvocidib infusion. PK samples were analyzed using methods as described previously [12, 13]. Non-compartmental PK parameters were determined for alvocidib and alvocidib-glucuronide in Phoenix WinNonlin v.6.3. Population PK parameter estimates for both agents were determined by incorporating plasma concentration data versus time, body weight, and sex from this study into a previous dataset and population PK/pharmacodynamic model for alvocidib-induced TLS [14].

Statistics

The primary endpoint of the study was to define the MTD of CAR, using the traditional 3×3 phase I design as described above in the Study Design section. Secondary endpoints included describing response rates, PFS, and OS following administration of CAR. Common hematologic and non-hematologic toxicities are summarized with frequencies by dose level and for all patients. Baseline characteristics are described using frequencies and medians with ranges for categoric and continuous variables, respectively. PFS was measured from the date of first treatment to the date of progression or death from any cause, censoring patients alive at last follow-up. OS was measured from the date of first treatment to the date of death or date of last follow-up. Best response, PFS, and OS are provided for each patient. Observed PK parameters were evaluated and summarized in graphic and table format. No inferential statistical tests of hypotheses were planned due to the small sample size.

Results

Patient Characteristics

A group of 9 patients were treated on Cohorts 1 (n=3), 2 (n=3), and 3 (n=3). Median age was 55 years (range 42-77) and 7 patients were male. Patients had the following high-risk features: del(17p13.1) (n=3), del (11q22.3) (n=4), un-mutated IgV_H (n=6), age >70 years (n=2), and B2M>4 (n=1). Only 3 of the patients had previous treatment, all with alemtuzumab, and 1 of these patients had also received 4 weekly doses of rituximab as treatment for autoimmune hemolytic anemia. All patients completed 6 planned cycles of therapy with exception of one patient in Cohort 2, who stopped therapy secondary to lack of response. Baseline patient characteristics are summarized in Table 2.

Table 2.

Pretreatment Patient Characteristics

Characteristic	Patients (n=9)
Median Age in Years (range)	55 (43-77)
Male	7
Caucasian	8
ECOG Performance Status [10]
0	7
1	2
Number of Prior Treatments
0	6
1	2
2	1
Rai Stage at Treatment
Intermediate Risk (I-II)	7
High Risk (III-IV)	2
Del(17p)	3
Del(11q)	4
Complex Cytogenetics	3
Unmutated IgVH status (n=7)	6
β2microglobulin >4	1
Bulky Adenopathy >5cm	0
Median Leukocyte Count × 109/L (range)	66.9 (3.4-217.0)
Median Hemoglobin, g/dL (range)	11.8 (9.5-14.2)
Median Platelet Count × 109/L	138 (75-250)

Toxicity

The CAR regimen was well tolerated by the patients on this trial and DLT was not observed. The most common toxicities were hematologic, including: anemia (n=8), neutropenia (n=7), leukopenia (n=6), and thrombocytopenia (n=6). The most common non-hematologic toxicities included: fatigue (n=5), electrolyte disturbances (n=5), and gastrointestinal (GI) toxicity [diarrhea (n=7), abdominal discomfort (n=5), nausea/vomiting (n=4), and liver dysfunction (n=4)]. The majority of grade 3 and 4 toxicities were hematologic: anemia (grade 3: n=1), leukopenia (grade 3: n=2, grade 4: n=2), neutropenia (grade 4: n=6), and thrombocytopenia (grade 3: n=2). No patients experienced grade 4 non-hematologic toxicity and only one grade 3 non-hematologic, treatment-related adverse event was reported (electrolyte disturbance). One patient experienced grade 3 diarrhea that resolved within 24 hours during Cycle 1, but had received two doses of kayexylate, a potassium binder that induces diarrhea. Therefore, this was not considered a treatment-related adverse event and was not considered a DLT. There were three episodes of grade 2 cytokine release syndrome (CRS; 2 patients in Cohort 2 and 1 patient in Cohort 3). All toxicities resolved by the end of therapy. Notably, no patients experienced TLS or infection. The most common adverse events are summarized in Supplemental Table 1.

Response

All nine patients were evaluable for response, as summarized in Table 3. All patients demonstrated cytoreduction following initial dose of cyclophosphamide and rituximab as represented by a decrease from the initial median ALC on Cycle 1, Day 1 of 62.2 (range=3.0-215.3) to a median ALC of 2.25 (range=0.71-26.1) on Cycle 1, Day 8 of therapy as depicted in Supplemental Figure 1. Three patients achieved complete response (CR), 4 achieved partial response, 1 had stable disease, and 1 had progressive disease (PD). The patient with PD had previously been treated with alemtuzumab and had three high-risk factors, including an age of 77 years and presence of del(17p13.1) and un-mutated IgV_H cytogenetics. This patient remains alive at 19.1 months following first treatment. In fact, all 9 patients are still alive with a median follow-up of 15.6 months (range: 4.6-22.7 months).

Table 3.

Response by Risk Factor and Dose Level

Dose Level	UPN	High Risk Factors					Response	PFS (mos)	OS (mos)
		Del(17p)	Del(11q)	Un-mut IgVH	Age >70	B2 M≥4
1	1	Y	N	Y	Y	N	CR	22.7	22.7
	2	N	Y	U	N	N	PR	7.4†	22.1
	3	N	N	Y	N	N	SD	5.1†	10.1
2	4	Y	N	U	N	N	CR	6.2	6.2
	5	N	Y	Y	N	N	PR	16.8†	17.6
	6	Y	N	Y	Y	Y	PD	3.4†	19.1
3	7	N	Y	Y	N	N	CR	15.6	15.6
	8	N	N	Y	N	N	PR	10.3	10.3
	9	N	Y	N	N	N	PR	4.6	4.6

^†Progressive Disease

Un-mut=un-mutated, B2M=beta-2-microglobulin, Y=yes, N=no, U=unknown, CR=complete response, PR=partial response, SD=stable disease, PD=progressive disease, UPN=unidentified patient number, PFS=progression-free survival, OS=overall survival

Pharmacokinetics

Plasma concentration was calculated on Day 8 of Cycles 1 and 2 (See Supplemental Figure 2). Observed maximum concentrations and areas under the plasma concentration-time curves, along with population PK parameter estimates are summarized in Supplemental Table 2. Overall, PK data was similar in this study to previous reports by our group [5, 6].

Conclusions

This phase I feasibility study showed that the CAR regimen was well tolerated and active when used as frontline therapy for CLL patients considered to have high-risk genomic and clinical characteristics. Despite high-risk genomic and clinical characteristics, 7 of the 9 patients achieved a response and 1 patient achieved SD. These results are impressive and promising in this patient population typically resistant to standard chemotherapy. Notably, 2 of the 3 patients on this trial with del(17p13.1), and hence dysfunctional p53, achieved a CR, which supports in vitro [15] and in vivo data [6, 7] that alvocidib can induce cell apoptosis in a manner independently of the p53 pathway. In comparison, among the patients with del(17p13.1) karyotype in the landmark CLL8 trial, no patients who received FC (n=29) and only one patient who received FCR (n=22) achieved a CR [3] . With CAR therapy in the current study, all patients (n=4) with del(11q22.3) achieved response (CR=1, PR=3), which supports data suggesting that this subgroup benefits from cyclophosphamide and chemoimmunotherapy [16, 17]. Although the one patient with progressive disease had 3 high-risk factors [age>70, del(17p13.1), and un-mutated IgV_H status], he remains alive at 19.1 months following treatment with this regimen. This individual’s CLL has since been proven to be highly refractory as his CLL has progressed through three subsequent standard regimens.

In contrast to early phase I/II studies with alvocidib, patients in this trial did not experience severe TLS. The novel dosing schedule of this trial, which did not introduce alvocidib until Cycle 1, Day 8, allowed for cytoreduction of CLL burden by cyclophosphamide and rituximab, presumably lowering the risk for TLS. However, this may also be due to the fact that patients treated in the early clinical trials were relapsed/refractory versus treatment-naïve and had a higher Rai Stage, indicating more volume of disease versus the generally lower Rai Stage in this trial (Rai Stage I/II: n=7), leading to better tolerance of the regimen. Other previously reported factors associated with TLS after treatment with alvocidib [9], including female gender, adenopathy ≥10cm, elevated WBC count >200×10⁹/L, and increased B2M, were also not heavily represented in this clinical trial. Lastly, the first 2 doses of alvocidib were given as an inpatient with aggressive supportive care, which likely helped to provide for early recognition and treatment of TLS.

This study closed enrollment early, without allowing for enrollment of 3 additional patients at the highest dose level as intended with a traditional 3×3 phase I design. The trial was disadvantaged by rate of accrual, potentially related to the requirement for hospitalization while receiving the first 2 doses of alvocidib, which caused many insurance companies to deny participation in the trial. However, this regimen was designed to allow transition to outpatient administration of therapy following the first 2 doses of alvocidib minimizing inpatient time requirement.

Although toxicity assessment was limited by small patient numbers, combined therapy with CAR was well tolerated. The severe side effects of grade 3/4 TLS and CRS were not observed with the novel timing of alvocidib (with delayed introduction until Cycle 1, Day 8) allowing for cytoreduction by rituximab and cyclophosphamide. This trial’s importance lies in this novel concept of delayed introduction of the CDK inhibitor in combination therapy with the goal of limiting toxicity. This patient group of high cytogenetic and clinical risk CLL patients have very limited effective treatment options and development of regimens, such as CAR therapy, that appear to have clinical activity should be further investigated to provide options for this tough-to-treat patient population. Secondary to potential safety and efficacy suggested by this trial, other phase I/II trials are currently underway at our institution using this concept in chemoimmunotherapy regimens with alvocidib and other novel CDK inhibitors.

The major barrier alvocidib has faced in clinical development and pharmaceutical support has been toxicity, such as TLS. With the innovative concept demonstrated in this trial, significant interest has emerged in the form of pharmaceutical support for this drug. The right to develop and commercialize alvocidib has recently been purchased by Tolero Pharmaceuticals, Inc. The company has publicly announced its intentions to proceed with phase II investigation of alvocidib in CLL patients in the near future.

In summary, the CAR combination reported in this trial could offer advantages over standard frontline therapies for CLL patients in this high-risk group. Given other potential benefits of the CDK inhibitors such as decreased infectious complications and potentially decreased risk of myelodysplasia, further study of this regimen is warranted in the setting of a phase Ib/II clinical trial with continued monitoring for toxicities.

Abbreviations

B2M

beta 2 microglobulin

CAR

cyclophophamide, alvocidib, rituximab

CDK

cyclin-dependent kinase

CLL

chronic lymphocytic leukemia

CR

complete response

CRS

cytokine release syndrome

DLT

dose limiting toxicity

ECOG

Eastern Cooperative Oncology Group

FC(R)

fludarabine, cyclophosphamide, (rituximab)

IgV_H

immunoglobulin heavy chain variable region

MTD

maximum tolerated dose

NCI

National Cancer Institute

PD

progressive disease

PFS

progression-free survival

PK

pharmacokinetic

ORR

overall response rate

OS

overall survival

TLS

tumor lysis syndrome

WBC

white blood cell