Phase II Study of Bryostatin 1 and Vincristine for Aggressive Non-Hodgkin Lymphoma Relapsing after an Autologous Stem Cell Transplant

Abstract

Bryostatin 1, isolated from a marine bryozoan, enhances the efficacy of cytotoxic agents through modulation of the protein kinase C pathway and is active in combination with vincristine for diffuse large B-cell lymphoma. Further, the apoptotic frequency of peripheral blood T lymphocytes as determined by flow cytometry may predict which patients will respond to this combination. We tested the efficacy and safety of bryostatin 1 50μg/m² given over 24 hours and vincristine 1.4 mg/m² on days 1 and 15 every 28 days in aggressive B-cell non-Hodgkin lymphoma relapsing after autologous stem cell transplantation. End points included tumor response, toxicity and survival. Responses were correlated with an increase in apoptotic frequency of CD5+ cells by flow cytometry using annexin V staining. Fourteen patients were enrolled with 13 being evaluable for a response. The overall response rate was 31% with 2 patients achieving a complete response. The most common toxicities were grade 3 lymphopenia (7 patients), grade 3 to 4 neutropenia (2 patients), and grade 3 hypophosphatemia (2 patients). Median progression-free and overall survivals for all patients were 5.7 and 21.4 months respectively. One patient demonstrated an increase in T-cell apoptotic frequency, also achieving a complete response. Bryostatin 1 and vincristine has efficacy in select patients with aggressive non-Hodgkin lymphoma. Future investigations of agents targeting the protein kinase C pathway may benefit from early response assessment using flow cytometry to evaluate T-cell apoptosis.

1. Introduction

High-dose chemotherapy followed by an autologous stem cell transplantation is the standard of care for relapsed diffuse large B-cell lymphoma and is utilized in select patients with indolent non-Hodgkin lymphoma (NHL) as well [1,2]. However, patients relapsing after this modality have a poor prognosis with few therapeutic options. Agents with novel targets and a limited toxicity profile are therefore needed for this heavily pretreated population.

Bryostatin 1 is a macrocytic lactone isolated from the marine invertebrate Bugula neritina. Short-term exposure activates protein kinase C, a kinase family involved in cell proliferation and apoptosis. Prolonged exposure decreases protein kinase C activity through selective binding of the regulatory domain. In addition to its direct cytotoxic effects, other potential biologic activities include the direct stimulation of hematopoietic precursors and the activation and differentiation of a variety of normal myeloid and lymphoid cells [3-10]. Growth inhibition and differentiation specifically have been demonstrated in NHL cell lines. Subsequent phase I evaluations demonstrated minimal toxicity associated with bryostatin 1. Myalgias were dose-limiting whether bryostatin 1 was administered intermittently over 1 hour or as a prolonged infusion [11-14]. Efficacy, however, was limited with either dosing schedule in larger studies limited to NHL patients [15,16].

Bryostatin 1 enhances the activity of a variety of cytotoxic agents, especially vincristine in a diffuse large B-cell lymphoma cell line and subsequently in a xenograft model [17-25]. The authors further demonstrated down-regulation of multidrug resistance gene expression and no inhibition of normal bone marrow growth [24,25]. In light of the preclinical data suggesting synergistic efficacy and its limited myelotoxicity, bryostatin 1 may be most useful as a chemotherapy adjunct. We combined bryostatin 1 with vincristine for patients with relapsed B-cell malignancies based on our encouraging phase I experience with bryostatin 1 as a 24-hour infusion followed by bolus vincristine [26].

Given the 8 month median time to maximal response, an assay to predict which patients may respond would be beneficial. We previously observed a clinical response or prolonged stable disease in each of the 5 patients who demonstrated an increase in apoptotic frequency in CD5+ cells by annexin V staining. Based on the regimen's activity and safety profile observed in the phase I trial, the current phase II trial was designed to test this novel combination in NHL patients relapsing after autologous transplantation and to further evaluate the ability of flow cytometry determined apoptotic frequency to predict a clinical response.

2. Results

2.1. Patient Characteristics

Fourteen patients were registered to the study between September 2003 and February 2007. Accrual was terminated early based on the mutual decision of the sponsor (National Cancer Institute / Cancer Treatment Evaluation Program) and lead investigators to halt further study of bryostatin 1 given the more potent bryostatin analogs in development [27]. Baseline patient characteristics are summarized in Table 1. Patients were heavily pretreated with a median of 4 prior regimens including autologous stem cell transplantation. All patients had received prior vincristine.

Table I.

Patient Characteristics

	(n =14)
Male: Female	9:5
Median Age	57 (48-71)
ECOG performance status
0	11
1	2
2	1
Histology
Diffuse large B cell	4
Grade 3 Follicular	3
Transformed	6
Mantle cell	1
Mean durations (months):
Diagnosis to transplant	37
Transplant to relapse	23
Diagnosis to enrollment	81
Prior regimens	4 (2-7)

Abbreviations: ECOG-Eastern Cooperative Oncology Group

2.2. Clinical Responses

Thirteen of 14 patients were evaluable for a response. One patient was recognized as having central nervous system parenchymal involvement after registration and she was removed from protocol. Median duration of treatment was 3.6 (range, 1.0-9.5) months. Two patients achieved a complete response and 2 achieved partial responses for an overall response rate of 30.8% (95% CI, 9.1-61.4%). The median time to response was 4 months (range, 3-6) for these patients. Time to progression was 9, 48+, 4, and 3 months respectively. In addition, six patients had stable disease with a median time to progression of 5.5 months (range, 2-12+), 2 of which subsequently underwent sibling matched allogeneic stem cell transplantation using reduced intensity conditioning. One recipient ultimately died of graft versus host disease and the other remains in remission > 4 years later.

Median follow-up was 11.5 (range, 1.8-55.1) months. Eight patients had documented dates of progression. Figure 1 shows the Kaplan-Meier plot of PFS and OS. Censored observations are indicated by dots on the plot. The proportion that had not progressed or died at 12 months was 0.32 (95% CI, 0.09-0.58) and was the same at 24 months, with a median PFS of 5.7 (95% CI, 3.6- NR) months. The proportions surviving at 12 and 24 months were 0.54 (95% CI, 0.25-0.76) and 0.27 (95% CI, 0.07-0.53), with a median survival of 21.4 (95% CI, 8.5-NR) months.

Figure 1.

Kaplan–Meier plot for survival.

2.3. Adverse Events

As demonstrated in Table 2, toxicities were mild and self-limited. Despite universal prior vincristine exposure, only 1 patient developed grade 3 neuropathy. This patient's gait instability, consistent with a large fiber peripheral neuropathy, was felt to be secondary to vincristine and resolved upon discontinuation of the agent. Only one patient experienced grade 3 myalgias and required a 20% dose reduction after cycle 1. The same patient also experienced grade 4 neutropenia necessitating a 1 week treatment delay after cycle 2. One other episode of neutropenia was observed, grade 3 in severity occurring after cycle 1. The patient tolerated 3 additional cycles without difficulty despite not being dose reduced or delayed. There was one episode of grade 3 pancreatitis, but imaging studies revealed this was due to peri-pancreatic lymphomatous progression. An upper extremity deep venous thrombosis was related to the patient's central venous catheter and felt to be unrelated to the study agents. Two episodes of grade 3 hypophosphatemia occurred, both in the absence of other metabolic abnormalities or evidence of hemolysis.

Table II.

Toxicity profile (n=14 patients) (evaluable cycles=49)

	Grade (number of patients)
Toxicities	3	4
Hematologic
Neutropenia	1	1
Lymphopenia	7
Non-hematologic
Neuropathy	1
Thrombosis	1
Myalgia	1
Pancreatitis	1
Hypophosphatemia	2

2.4. Laboratory Studies

To assess the effect of treatment on apoptosis as a marker of response, we used annexin V immunostaining detected by flow cytometry. The analysis of the relationship between apoptotic fraction and response compared to pretreatment was done for both CD5+ T-cells and CD19+ B-cells on 8 and 7 patients respectively. Apoptotic fraction was dichotomized as increased or decreased. All p-values from the Kruskal-Wallis exact test were found to be non-significant due to small sample sizes (data not shown). Figure 2 shows the annexin V frequency dynamics of CD5+ cells after treatment. The majority of patients demonstrated a decrease in apoptotic frequency at the 6-hour time point with a persistent decrease in apoptotic frequency at 30 hours. One patient had an overall increase in apoptotic frequency at the 6-hour time point and a further increase at 30 hours. This was one of the two patients who achieved a CR.

Figure 2.

Annexin V frequency dynamics in peripheral blood CD5+ cells. For one patient, the apoptotic frequency increased at 6 and 30 hr (gray dashed lines). For the remainder (black lines), the direction was below the initial value.

3. Discussion

Bryostatin 1 is the prototype of a novel family of protein kinase C modulators with a variety of antitumor and biologic modification properties. Despite its limited single agent activity, bryostatin 1 has demonstrated the ability to potentiate the cytotoxicity of several chemotherapy agents. In the diffuse large B-cell lymphoma line WSU-DLCL2, bryostatin 1 combined with vincristine demonstrated an increase in apoptosis and growth inhibition as compared to either agent alone [25]. The authors further demonstrated suppression of the anti-apoptotic protein bcl-2 and induction of the tumor-suppressor protein p53. Bryostatin 1 was subsequently shown to reverse the multidrug resistance phenotype and promote vincristine accumulation restoring antitumor activity in a mouse xenograft model using the same cell line [24]. In our phase I trial, myalgias were dose limiting and no significant myelotoxicity was observed. Further, complete and partial responses as well as stable disease > 6 months duration were noted in 10 of 25 patients despite most patients having received prior vincristine. Included were 2 patients with diffuse large B-cell lymphoma who had relapsed after an autologous stem cell transplant [26].

The patient population in the current trial closely mimics the cell line phenotype used in preclinical studies and is further based on those who obtained durable responses in our phase I trial. Unfortunately, the strict eligibility criteria resulted in slow accrual. Nonetheless, the tolerability of the regimen was again demonstrated here with limited myelosuppression and an overall paucity of grade 3 or 4 events. Objective responses were observed in this heavily pretreated population. Two patients remain in remission, one after an allogeneic stem cell transplant and the other without any further therapy. Compared to our phase I experience, the median time to maximal response was shorter in the current study (8 months vs. 4 months respectively) as was the time to progression in patients achieving an objective response (25 months vs. 16 months respectively). This may relate to patients being more heavily pre-treated and therefore increasing resistant to additional therapy in the current study.

Our previous demonstration of a correlation of clinical responses with an increase in the apoptotic frequency in CD5+ cells was only reproduced in 1 patient. It is possible that analysis at later time points may have yielded additional information not collected as part of this trial. It is difficult to make any conclusions regarding the use of T lymphocyte apoptotic frequency to predict a clinical response to bryostatin 1 and vincristine due to the small numbers included. However, future investigations of agents targeting the protein kinase C pathway should consider the further evaluation of this assay.

Combination therapy of protein kinase C inhibitors with other cytotoxic agents should be investigated given the demonstrable responses and lack of clinically significant hematologic and non-hematologic toxicity observed in our trial. Preclinical data suggest the superiority of bryostatin 1 in combination with CHOP as compared to CHOP alone [21]. In combination with fludarabine for indolent NHL and chronic lymphocytic leukemia, bryostatin 1 did not appear to add significant toxicity to full dose fludarabine [28]. Up-regulation of CD20 in lymphoma cell lines by bryostatin 1 has been demonstrated as well, increasing the cell's susceptibility to rituximab [29,30]. For these reasons, clinical investigations using agents targeting the protein kinase C pathway in combination with chemoimmunotherapeutic regimens incorporating anti-CD20 agents are warranted. This strategy may lead to an increase in efficacy of standard regimens for untreated patients and reverse drug resistance in those who are previously treated.

In summary, this study demonstrates the efficacy of bryostatin 1 and vincristine in select patients with heavily pretreated NHL, including prior autologous stem cell transplantation. Minimal toxicity of bryostatin 1 when used in combination with cytotoxic agents continues to be observed. While assurance of continued clinical development of bryostatin 1 is remote, we are aware that more potent bryostatin analogs (e.g., bryologs) are entering the clinical arena for further testing. Our study further supports the clinical development of this class of potentially interesting and important modulators of protein kinase C.

4. Methods

4.1. Study Design

This multi-institutional phase II study was conducted between 2003 and 2007 at University Hospitals Case Medical Center in Cleveland Ohio, University of Florida Shands Cancer Center in Gainesville Florida, Henry Ford Medical Center in Detroit Michigan, and Oregon Health Sciences University in Portland Oregon. The protocol was sponsored by the National Cancer Institute, which supplied bryostatin 1 for all study patients. The primary objective was to evaluate the overall response rate when bryostatin 1 is given in combination with vincristine in patients with aggressive B-cell NHL who had progressed or relapsed after an autologous bone marrow or peripheral stem cell transplant. Secondary objectives included the evaluation of the toxicity of the regimen, progression free survival and overall survival. Laboratory correlatives included the measure of apoptosis in CD5+ and CD19+ peripheral blood lymphocytes using annexin V staining and flow cytometry to determine if an increase in apoptotic response (with two or more consecutive apoptotic fractions) was predictive of outcome. This protocol was approved by the review board of each participating institution prior to patient entry. All patients signed informed consent.

4.2. Patient selection

All patients with biopsy proven aggressive B-cell NHL relapsing or progressing after autologous stem cell transplantation were eligible. Patients were required to have measurable disease with a lymph node or tumor mass ≥ 2 cm; age ≥ 18 years; a life expectancy untreated of > 8 weeks; Eastern Cooperative Oncology Group performance status of ≤ 2; absolute neutrophil count (ANC) ≥1,250/μL; platelets ≥ 50,000/μL; hemoglobin ≥ 8.5 g/dL; serum creatinine ≤ 2.0 mg/dl; serum bilirubin≤ 2.0 mg/dl; and AST(SGOT) /ALT(SGPT) ≤3 times the upper limit of normal.

Patients did not receive previous anti-cancer therapy for at least 4 weeks and had recovered from all treatment-related toxicity. Patients could not have had lymphomatous central nervous system involvement or HIV infection. Patients who had an uncontrolled intercurrent infection or with ≥ grade 2 peripheral neuropathy were not eligible.

4.3. Drug Administration

Patients were treated with a 50μg/m² 24-hour intravenous infusion of bryostatin 1 through a central venous catheter on day 1 followed immediately by a 1.4 mg/m² (maximum of 2.0 mg) bolus intravenous injection of vincristine. Treatment was repeated every 2 weeks with 2 treatments (4 weeks) constituting one cycle of therapy. For patients with responding or stable disease, or freedom from progression at 6 months, treatment continued but bryostatin 1 and vincristine were given every 3 weeks with 6 weeks being considered a cycle of therapy. Standard anti-emetics and the administration of blood products were allowed. The use of corticosteroids was prohibited.

4.4. Dose Modification

After the first treatment, patients were able to receive full dose therapy if just prior to subsequent cycles; the ANC was ≥ 1,000/μL, platelet count ≥ 50,000/μL, and in the absence of grade 3/4 toxicity (National Cancer Institute Common Toxicity Criteria, version 3.0). Up to a 2-week delay was permissible during the first 6 cycles of therapy and up to a 3-week delay was permissible beyond cycle 7 to allow recovery of blood counts to treatment levels. Patients experiencing hematologic toxicity had a stepwise dose modification first with a 50% dose reduction in vincristine, followed by a 20% dose reduction in bryostatin 1 for the 2^nd reduction, and finally elimination of vincristine if a 3^rd modification was needed. Additional dose modifications included 50% reductions in vincristine dosing for grade 2 hyperbilirubinemia or peripheral neuropathy. Additional dose modifications in bryostatin 1 included a 20% dose reduction for grade 3 myalgias.

4.5. Response Criteria

Response evaluations, consisting of complete history, physical examination, and imaging studies, were performed after every 2 cycles of chemotherapy. To be considered assessable, patients had to receive 2 cycles of chemotherapy unless progressive disease was documented during the first 8 weeks. Responses were categorized using the International Workshop Criteria [31]. Objective response assessment was performed bimonthly up to 12 months after beginning therapy based on delayed responses observed in our phase I trial [26]. Patients who experienced disease progression were taken off study. Patients who achieved a complete remission (CR) received two additional cycles and were then observed without further therapy.

4.6. Statistical Design

This trial was designed to test the null hypothesis that the true overall response rate was at most 10%. The smallest response rate that would indicate that bryostatin 1 and vincristine was worth further study in this population was 30%. The design was generated based on the parameters and assumptions of Simon's minimax two-stage design [32] with α=0.10 and β=0.10 (power=0.90). The study design required a maximum of 25 patients, where interim analysis would be performed after 16 patients had been accrued and followed for at least 8 weeks for response. At this interim analysis, the study was to be terminated early if ≤1 of the first 16 patients responded. At the time of the final analysis based on results from 25 patients, >4 responses were required to indicate that this regimen warranted further evaluation in this patient population.

Progression free survival (PFS) was defined as the time from the date of initial treatment to the date of disease progression or death from any cause. Patients who went off treatment due to other reasons (eg, consolidative therapy) or who died without disease progression were censored at that time. Overall survival (OS) was defined as the date of initial treatment to the date of death from any cause. The distributions of these time-to-event end points were each estimated using the Kaplan-Meier method [33].

4.7. Correlative Studies

Peripheral blood from patients treated at University Hospitals Case Medical Center was collected for apoptosis assays at the following time points: pretreatment (T 0); 6 h (T +6 h); 24 h (end of bryostatin 1infusion, T +24 h); and 6 h after bolus vincristine administration (T +30 h). Isolation of CD5+/CD19- and CD5-/CD19+ lymphocytes and flow cytometry with annexin V immunostaining were performed as per previously published methods [26]. In the text, cells expressing CD5 but not CD19 are referred to as CD5+, while CD5-negative/CD19-positive cells are referred to as CD19+.