Phase I Study of Bryostatin 1 and Fludarabine in Patients with Chronic Lymphocytic Leukemia and Indolent (Non-Hodgkin's) Lymphoma

Purpose: Preclinical studies suggested that bryostatin 1 might potentiate the therapeutic effects of fludarabine in the treatment of hematologic malignancies. We undertook a phase I study to identify appropriate schedules and doses of bryostatin 1 and fludarabine to be used in phase II studies.

Experimental Design: Patients with chronic lymphocytic leukemia (CLL) or indolent lymphoma received fludarabine daily for 5 days and a single dose of bryostatin 1 via a 24-hour continuous infusion either before or after the fludarabine course. Doses were escalated in successive patients until recommended phase II doses for each sequence were identified on the basis of dose-limiting toxic events.

Results: Bryostatin 1 can be administered safely and tolerably with full dose fludarabine (25 mg/m²/d × 5). The recommended bryostatin 1 phase II dose is 50 μg/m² for both sequences, bryostatin 1 → fludarabine and fludarabine → bryostatin 1. The combination is active against both CLL and indolent lymphomas with responses seen in patients who had been previously treated with fludarabine. Correlative studies do not support the hypothesis that bryostatin 1 potentiates fludarabine activity through down-regulation of protein kinase C in target cells.

Conclusions: Bryostatin 1 can be administered with full dose fludarabine, and the combination is moderately active in patients with persistent disease following prior treatment. In view of the activity of monoclonal antibodies such as the anti-CD20 monoclonal antibody rituximab in the treatment of CLL and indolent lymphomas, the concept of combining bryostatin 1 and fludarabine with rituximab warrants future consideration.

Systemic treatment options for chronic lymphocytic leukemia (CLL) and indolent (non-Hodgkin's) lymphoma have been expanding in recent years and now include oral alkylating agents with or without corticosteroids; purine nucleoside analogues (fludarabine, 2-chlorodeoxyadenosine); combination chemotherapy, generally combining an alkylating agent or purine nucleoside analogue with a corticosteroid and cytotoxic agent(s) of other classes; monoclonal antibodies (rituximab, alemtuzumab); radiolabeled monoclonal antibodies (ibritumomab tiuxetan; tositumomab); and, for selected patients, high-dose therapy with autologous or allogeneic hematopoietic stem cell rescue. Nevertheless, most patients are treated with palliative intent, and new therapies are needed.

Bryostatin 1 is a macrocyclic lactone isolated from the marine invertebrate bryozoan, Bugula neritina, which has shown promise as an anticancer agent (1). As a single agent, bryostatin 1 displayed antitumor activity against both murine and human tumor cell lines in vitro and against transplantable murine tumors in vivo, including activity against B lymphoid tumors (1–5). After extensive phase II evaluation, however, it is evident that bryostatin 1 has minimal single agent activity (6).

In addition to direct cytotoxic effects, bryostatin 1 exerts a variety of other biological actions including hematopoietic progenitor cell stimulation (7), differentiation induction (8), neutrophil activation (9), immune cell activation (10, 11), induction of platelet aggregation (12), and potentiation of hematopoietic growth factor activity (13, 14). These properties have supported the investigation of bryostatin 1 as a biological response modifier.

Two lines of preclinical investigation suggest that the combination of bryostatin 1 and a purine nucleoside analogue might be clinically useful. Bhatia et al. proposed a general mechanism in which the exposure of leukemic cells to a differentiation stimulus following exposure to a DNA-damaging agent led to a marked potentiation of apoptosis (15). In an analogous manner, we demonstrated that exposure of leukemic cells to fludarabine followed by bryostatin 1 led to enhanced apoptosis and highly synergistic antiproliferative effects in monocytic leukemia cells (16). On the other hand, Mohammad et al. showed that the reverse sequence, i.e., bryostatin 1 followed by 2-chlorodeoxyadenosine, led to enhanced leukemic cell growth inhibition in vitro and in vivo (17). In addition, this sequence was associated with enhanced activity in a murine CLL xenograft model (18).

The biochemical basis for such synergistic interactions remains unclear. However, several lines of evidence have focused on bryostatin 1–related modulation of protein kinase C (PKC). PKC is an intracellular signal transduction molecule involved in diverse cellular processes, particularly cell growth, cell differentiation, and survival (19). Bryostatin 1 induces initial PKC activation followed by diminished enzyme activity that presumably stems from proteasomal degradation (20). This may promote differentiation which, in the context of prior DNA damage, potentiates apoptotic responses, or, when it precedes DNA damage, may prevent PKC-mediated cytoprotective responses. Other biochemical effects that may be PKC-mediated have been observed including down-regulation of the multidrug resistance protein mdr1 (21), enhanced expression and/or release of IFN-γ (22) and tumor necrosis factor-α (23), and Bcl-2 phosphorylation (24).

Collectively, these findings provided a basis for clinical studies of bryostatin 1 in combination with the purine nucleoside analogue fludarabine. Single agent bryostatin 1 phase I studies have involved weekly or biweekly administration of infusions ranging from 1 to 72 hours; maximum tolerated doses (MTD) according to these schedules have ranged from 25 to 120 μg/m² (25–28). Dose-limiting toxicity (DLT) has consistently been a transient myalgia of uncertain etiology that may be refractory to narcotic analgesics. This toxicity has uniformly been cumulative in that the onset follows two or more doses. Fludarabine is generally administered at a dose of 25 mg/m²/d for 5 days every 4 weeks. Common toxicities include myelosuppression, which may be cumulative, and infection, particularly pneumonia and opportunistic infections.

In light of these considerations, a phase I study of the combination of bryostatin 1 and fludarabine in patients with CLL (including small lymphocytic lymphoma, SLL) and indolent lymphoma was undertaken. The goals of this trial were to establish the MTD for bryostatin 1 administered in conjunction with a standard 5-day course of fludarabine, to characterize the regimen's toxicities, to identify the recommended doses for phase II trials, to obtain preliminary evidence of the regimen's anticancer activity, and to examine the regimen's pharmacodynamic effects. As preclinical studies supported both sequences of administration, bryostatin 1 → fludarabine and fludarabine → bryostatin 1, and because DLTs might differ according to the sequence of administration, the phase I trial involved parallel studies, one of each sequence.

Drug sources and formulation. Bryostatin 1 (NSC 339555) was provided by the Division of Cancer Treatment and Diagnosis of the National Cancer Institute. The formulation components were a 10 mL flint vial containing 0.1 mg (100 μg) bryostatin as a lyophilized cake or powder and 5 mg of povidone (as a bulking agent) lyophilized from 40% t-butanol. The formulation was reconstituted with 1 mL sterile diluent containing 60% polyethylene glycol 400, 30% dehydrated ethanol, and 10% polysorbate 80. The resulting solution was diluted with 9 mL of 0.9% sodium chloride injection, and then further diluted in 0.9% sodium chloride or 5% dextrose in water to a final concentration in the range of 0.15 to 0.75 μg/mL. Infusion sets were assembled of glass or polyolefin, not polyvinylchloride, components, and tubing was primed with the formulated product in order to minimize adsorption. The bryostatin 1 formulation was administered by syringe pump or standard infusion controller. If administered via peripheral venous access, a concurrent infusion of 2 L of 0.9% sodium chloride was given in order to minimize phlebitis at the site of infusion. Fludarabine was purchased from commercial sources.

Eligibility criteria. Eligibility criteria included: (a) age ≥18 years and able to give informed consent; (b) either CLL stage I with symptoms and/or bulky lymphadenopathy, or stages II to IV according to the Rai et al. classification (29), with or without prior therapy, or indolent (non-Hodgkin's) lymphoma, progressive or relapsed following chemotherapy. Indolent lymphomas were considered to be lymphomas corresponding to one of the following revised European-American Lymphoma classification (30) categories: B cell CLL/prolymphocytic leukemia/lymphoma (CLL/SLL), lymphoplasmacytoid lymphoma (Waldenström's)/immunocytoma, mantle cell lymphoma, follicular types of follicle center cell lymphoma, marginal zone B cell lymphoma, splenic marginal zone lymphoma, hairy cell leukemia, T cell CLL/prolymphocytic leukemia, large granular lymphocyte leukemia, and mycosis fungoides/Sezary's syndrome. (c) Zubrod performance status ≤2; (d) granulocytes ≥1.0 × 10⁹/L, hemoglobin ≥8 gm/dL, and platelets ≥75 × 10⁹/L; (e) calculated or actual creatinine clearance ≥40 mL/minutes; (f) aspartate aminotransferase, alkaline phosphatase ≤2.5× the upper limit of normal and total bilirubin ≤2.0 mg/mL; (g) RBC direct antibody (Coombs)–test negative; (h) no prior bone marrow or peripheral stem cell transplantation (added as a study amendment to exclude patients with compromised myeloid reserve); (i) no systemic chemotherapy within 3 weeks prior to study treatment, and no ongoing requirement for systemic glucocorticoid or immunoglobulin therapy; (j) no intercurrent medical condition that in an investigator's opinion would compromise treatment or assessment of toxicity; (k) no known malignant CNS disease; (l) not pregnant or nursing and willing to use a medically accepted form of birth control. The characteristics of the patients enrolled are shown in Table 1.

Treatment plan. Patients received either a single dose of bryostatin 1 via a 24-hour continuous infusion followed immediately by fludarabine daily for days (bryostatin 1 → fludarabine), or fludarabine daily for 5 days followed by a single dose of bryostatin 1 via a 24-hour continuous infusion (fludarabine → bryostatin 1; Fig. 1). Initially, patients were assigned at random to one of the two sequences. This policy led to delays in accrual, however, as both sequences might be on enrollment hold pending observation for toxicity in a single sequence, and the protocol was amended so that enrollment to the two sequences became independent. During treatment, patients underwent weekly evaluation by a research nurse with monitoring of peripheral blood cell counts and serum chemistries (aspartate aminotransferase, alkaline phosphatase, lactate dehydrogenase, bilirubin, creatinine, and glucose) and monthly evaluation by a physician. Treatment was repeated every 4 weeks. The protocol specified dose modifications for patients experiencing significant treatment-related toxicities. Disease status was assessed every 2 months. Patients experiencing a response or stable disease were allowed to continue treatment indefinitely. During the study, however, it became apparent that many patients were eligible to continue treatment for many months, and that many of these patients had achieved a stable disease state but were experiencing progressive lymphopenia with a risk of opportunistic infection. At this time, the protocol was amended with an advisory suggesting limiting treatment to six to nine courses. Prophylactic hematopoietic growth factors were not permitted during the first course of treatment. Otherwise, patients received full supportive care. Prophylactic antibiotics to prevent opportunistic infections (which have been seen with fludarabine use; ref. 31) were permitted. An option existed to re-treat patients who experienced a response, discontinued treatment, and subsequently experienced disease progression. Dose escalation of an individual patient's dose was not permitted, but there was an option to re-enroll a patient only once at a higher fludarabine, but not bryostatin 1, dose.

Toxicity evaluation. All adverse events were characterized by nature, severity, and relationship to study treatment according to the National Cancer Institute Common Toxicity Criteria v2.0. Adverse events considered to be possibly, probably, or definitely related to treatment were scored as toxicities. For patients with baseline blood cell counts less than normal, hematologic toxicity was scored according to the National Cancer Institute Working Group criteria (32). Myalgia was scored as: none, grade 0; mild pain not interfering with daily activities, grade 1; moderate pain, or pain or analgesics producing some interference with daily activities, grade 2; severe pain, or pain or analgesics severely interfering with daily activities, grade 3; and disabling, grade 4.

Dose levels, definition of DLT, and identification of MTDs. Initial and subsequent dose levels were identical for each sequence, but dose level escalation was conducted independently (Table 2). Initial dose levels incorporated fludarabine at half the standard dose and, based on previous single agent experience (6, 25–28), a moderate dose of bryostatin 1. Fludarabine was escalated to full dose in two successive dose levels, and subsequently, the bryostatin 1 dose was escalated to the MTD. DLT was defined as any grade 4 toxicity except lymphopenia or any grade 3 toxicity except neutropenia, leukopenia, lymphopenia, thrombocytopenia, anemia, hyperglycemia, or alopecia during the first course of treatment. MTD was defined as the highest dose level with DLT in one third or fewer patients. Patients were enrolled in cohorts of three according to a modified “three + three” dose escalation scheme designed to identify the highest bryostatin 1 dose for each sequence that could be administered with full dose fludarabine (33).

Response evaluation. For patients with CLL, response was characterized according to a modification of the National Cancer Institute Working Group criteria (32). For patients with lymphoma, response was categorized as complete or partial response, stable disease, or progression of disease according to criteria based on the sum of the products of bidimensional tumor measurements (partial response as >50% reduction, progression of disease as >25% increase).

Correlative laboratory studies. Blood samples for correlative studies were obtained during the first treatment cycle from patients with CLL receiving the bryostatin → fludarabine schedule in which peripheral blood leukocytes were ≥80% leukemia cells and who were being treated at one of the three participating centers (Virginia Commonwealth University, Richmond, VA). One sample was obtained prior to initiation of the first bryostatin 1 infusion, and another was obtained within 1 hour of the conclusion of the first bryostatin 1 infusion and prior to administration of fludarabine. Peripheral blood mononuclear cells were isolated from blood samples by centrifugation over Ficoll-Hypaque. Total cellular PKC activity was assessed using a commercially available kit as previously described (34). Apoptosis was assessed as the fraction of cells in the sub-G₁, hypodiploid population as determined by flow cytometry following staining with propidium iodide or by terminal nucleotidyl transferase–mediated nick end labeling assay as previously described (35).

Statistical analysis. The significance of differences between experimental variables was determined using Student's t test for unpaired observations.

Patients. Fifty-nine patients were enrolled between November 5, 1998 and July 3, 2003 (Table 1). There were approximately equivalent numbers of patients with CLL/SLL and lymphoma. Patients were predominantly male, in middle or late life, of good performance status, and with extensive prior treatment; slightly fewer than 50% of patients had received prior fludarabine.

DLT and MTDs. Neutropenia was the most common DLT and largely defined the MTD for both sequences, bryostatin 1 → fludarabine and fludarabine → bryostatin 1 (Table 2). For each sequence, bryostatin 1 could be administered safely and tolerably with full dose fludarabine (25 mg/m²/d × 5), and the bryostatin 1 MTD for each sequence was 50 μg/m². DLT was more frequent in patients with CLL/SLL than in patients with lymphoma (Table 3), but the study was not designed to compare the tolerability of the combination by diagnosis, and this difference may not be significant.

Other toxicities and tolerability. Most patients received multiple treatment courses (Table 1), and treatment was well tolerated with toxicities that were transient and/or manageable. There was no remarkable difference in overall toxicity between the two sequences. Myelosuppression, particularly anemia, neutropenia, and lymphopenia, was common (Table 4). Infectious complications were relatively uncommon (Table 5). Nine patients received prophylactic sulfamethoxazole/trimethoprim. Other common grade 3 or greater toxicities included hyperbilirubinemia, fatigue, and hyperglycemia (Table 5). Severe myalgia was less frequent than in single-agent studies in which bryostatin 1 was more frequently administered (6, 25–28).

Disease response. Although most patients had received extensive prior treatment (Table 1), responses were common with either treatment sequence and in patients with either CLL/SLL or lymphoma (Table 6). Response rates were slightly higher with the sequence bryostatin 1 → fludarabine and in patients with lymphoma. However, as the study was not powered to assess response by treatment or diagnosis, the significance of these differences is uncertain. There was no apparent relationship between response rate and dose level (data not shown). One complete and seven partial responses were observed among 25 patients who had received prior fludarabine; however, as intermittent treatment is a common clinical practice in the management of CLL/SLL and indolent lymphoma, it cannot be assumed that most or all of these patients were fludarabine-refractory. Two partial responses were observed among eight patients who had received fludarabine within 2 months of study enrollment and therefore might be presumed to be fludarabine-refractory. Another patient experienced dramatic but transient resolution of bilateral pleural effusions, ascites, and scrotal and bilateral lower leg edema that did not qualify as a partial response.

Correlative laboratory studies. The effects of in vivo exposure to bryostatin 1 on the ex vivo response to fludarabine were examined in CLL cells obtained from five patients randomized to the bryostatin 1 → fludarabine sequence (Fig. 2). Cells from all five patients showed increased apoptosis following ex vivo exposure to 10 μmol/L fludarabine, and cells from two of five patients (patient nos. 1 and 3) displayed an increase in apoptosis following in vivo exposure to bryostatin 1 (P < 0.05 in each case). Cells from only one patient (no. 3) showed an increase in fludarabine-mediated apoptosis that was specific to the post-in vivo bryostatin exposure state. In no case did in vivo exposure to bryostatin 1 diminish the ex vivo sensitivity of cells to fludarabine.

Leukemic cell total PKC activity was assayed in samples obtained prior to and shortly following the infusion of bryostatin 1 in 12 patients with CLL (Fig. 3). In three patients, total leukemic cell PKC activity increased significantly following bryostatin 1 infusion (patient nos. 3, 11, and 12; P < 0.05). In no sample was there a significant reduction in CLL cell total PKC activity. There also was no correlation between the effects of leukemic cell PKC activity and clinical response (data not shown).

The results of this study indicate that full dose fludarabine (25 mg/m²/d × 5) can be administered in conjunction with bryostatin 1 safely and tolerably in patients with CLL and indolent lymphoma who have received extensive prior treatment other than bone marrow and/or stem cell transplantation. The recommended bryostatin 1 dose for phase II study is 50 μg/m² for both sequences, bryostatin 1 → fludarabine and fludarabine → bryostatin 1. The toxicities of the combination are similar to those of single-agent fludarabine, except that hyperbilirubinemia, fatigue, and hyperglycemia may be more common. Although preclinical studies had suggested that administration sequence might be important in the clinical effects of the combination, the recommended phase II bryostatin 1 dose is the same and toxicities were similar for each sequence.

Fludarabine, with or without corticosteroids, has been associated with response (complete remission + partial remission) rates in the range of 50% to 60% in CLL (36–38) and 50% to 100% in indolent lymphoma (39–41). The rationale for the addition of bryostatin 1 to fludarabine was that the combination might be more active than fludarabine as a single agent. The response rates in this study were in the lower end of these ranges, but it would not be appropriate to conclude that the combination is without promise as this phase I study was not designed to assess activity. Factors that might diminish apparent response rates include a study population with an adverse prognosis due to extensive prior treatment, and administration of disparate drug doses. However, if bryostatin 1 does potentiate the actions of fludarabine, one might expect to observe the activity of the combination in patients refractory to fludarabine as a single agent. In fact, some evidence of such activity was observed. Response rates were higher in the bryostatin 1 → fludarabine sequence; but the study was not designed to compare the activity of the two sequences, and no conclusions about relative activity can be reliably drawn.

A similarly promising activity for bryostatin 1 as a potential biological response modifier was seen in a phase I study of the combination of bryostatin 1 and vincristine (42).

The present study was based on the concept that bryostatin potentiates the proapoptotic effects of fludarabine, possibly through down-regulation of PKC activity (16). However, other studies have suggested that bryostatin 1 may antagonize the proapoptotic effects of fludarabine (43). In the present study, leukemic cells exposed to bryostatin 1 in vivo showed neither increased nor reduced sensitivity to fludarabine in subsequent ex vivo assay (Fig. 2). This may reflect differences among the cell lines used for preclinical investigations and fresh human leukemia cells, the inability to achieve adequate plasma bryostatin 1 concentrations in vivo, the inability to mimic in vivo fludarabine plasma concentrations ex vivo, or a combination of these factors.

Preclinical studies suggested that bryostatin 1–related potentiation of the proapoptotic effects of fludarabine might result from the down-regulation of PKC activity. In a prior study, we showed a trend toward down-regulation of total PKC activity in peripheral blood mononuclear cells obtained 72 hours after the initiation of bryostatin 1 infusions in patients with solid tumors or lymphoma receiving bryostatin 1 as a single agent (44). However, no down-regulation of total PKC activity in leukemic cells was observed in the present study (Fig. 3). This finding may reflect that, in order to avoid the confounding effects of fludarabine, PKC activity was assayed shortly after the completion of the 24-hour bryostatin 1 infusion, which contrasts with the later sampling time of 72 hours employed in the previous study. In fact, the pharmacodynamics of bryostatin 1 with regard to PKC are quite complex (7, 45). When assayed within a few hours of bryostatin 1 exposure, total cellular PKC may show up-regulation. Changes in total PKC activity reflect potentially disparate effects on different PKC isoforms. Furthermore, PKC activity in specific cellular compartments may be more relevant to clinical effects than changes in total cellular PKC. Finally, the differences may reflect the disparate in vivo responses of CLL cells and normal peripheral blood mononuclear cells to bryostatin 1.

In the interval following the completion of this trial, the first clinically relevant method for sample analysis of bryostatin 1 pharmacokinetics was described (46). In a limited report of clinical results, continuous infusion of bryostatin (20 μg/m²/d × 14 days) resulted in concentrations of ∼90 pmol/L; measured on days 8 to 15 of the infusion. None of the four patients had bryostatin 1 concentrations above the lower limit of quantitation (∼55 pmol/L) at the end of the first day of infusion.¹⁰

These values stand in contrast to bryostatin 1 concentrations employed in the preclinical studies that formed the basis for this clinical study (16, 17), which ranged from 10 to 100 nmol/L. The results of the pharmacokinetic analysis raises the possibility that the effects of bryostatin 1 observed in preclinical studies may not be apparent in patients due to inadequate plasma concentrations.

Recently, a large number of clinical investigations involving new strategies for the treatment of CLL/SLL and indolent lymphoma have been implemented. An important development has been the observation that monoclonal antibodies such as the anti-CD20 monoclonal antibody rituximab have significant activity in the treatment of B cell disorders, including B cell CLL/SLL and lymphoma. Currently, optimal therapeutic strategies involving established agents remain to be defined, and differences may be emerging for different diseases within this group. For example, in CLL/SLL, the combination of fludarabine and alkylating agents is very active but potentially too toxic, whereas the combination of fludarabine and rituximab is very active and tolerable (47). In indolent lymphoma, the combination of fludarabine and rituximab is also very active and tolerable, and the combination of rituximab with classical nonfludarabine combination chemotherapy such as cyclophosphamide/doxorubicin/vincristine/prednisone is tolerable and extremely active (48). Moreover, it has been shown that when leukemia and lymphoma progress on treatment, malignant cells are likely to retain the surface expression of CD20. Consequently, it is possible that rituximab may be an appropriate component of combination therapy for patients with “rituximab-refractory” disease (49). For these reasons, “second line” treatment regimens commonly involve rituximab or other monoclonal antibody agents.

Although the clinical activity of the combination of bryostatin 1 and fludarabine in CLL and indolent lymphoma has not been definitively assessed, it is unlikely that its activity will approach the activity of combinations of conventional chemotherapeutic agents with rituximab or other monoclonal antibodies. Thus, further clinical investigation of the bryostatin 1/fludarabine regimen may be hard to justify or implement. However, recent preclinical studies suggest that bryostatin 1 may`-regulate the expression of CD20 in malignant cells, raising the possibility that bryostatin 1 might potentiate the proapoptotic effects of rituximab (50). These observations have lead to a clinical trial of bryostatin 1 and rituximab in patients with refractory CLL (see http://www.clinicaltrials.gov/show/NCT00087425). Thus, (a) preclinical and clinical results with bryostatin 1 and fludarabine, (b) clinical results with the fludarabine/rituximab regimen (47), and (c) preclinical results with the bryostatin 1/rituximab combination (50) collectively provide a rationale for the investigation of a combination involving all three agents.