Phase I Trial of Bortezomib (PS-341; NSC 681239) and “Nonhybrid” (Bolus) Infusion Schedule of Alvocidib (Flavopiridol; NSC 649890) in Patients with Recurrent or Refractory Indolent B-cell Neoplasms Free

Despite numerous therapeutic options (e.g., cytotoxic chemotherapy, monoclonal antibodies, radioimmunotherapy, and combinations thereof) for patients with indolent B-cell non-Hodgkin lymphoma (NHL), cures largely remain elusive. Allogeneic stem cell transplantation (SCT) using reduced-intensity conditioning (RIC) regimens, which carry a lower treatment-related mortality (TRM) risk than myeloablative regimens, can theoretically produce cures via a graft-versus-lymphoma effect, but definitive evidence of efficacy of this strategy is lacking (1). Similarly, while proteasome inhibitors and immunomodulatory drugs have dramatically altered the therapeutic landscape in multiple myeloma, the disease remains incurable aside from allogeneic SCT. Although RIC regimens have lowered the high TRM-associated with myeloablative conditioning, convincing evidence is lacking that allogeneic SCT improves survival compared with autologous SCT, and the former is only recommended in the context of clinical trials (2). Novel therapeutic approaches are, therefore, clearly needed for patients with these diseases.

Bortezomib, the first-in-class proteasome inhibitor, received regulatory approval in the United States in 2003 after large phase II and III trials demonstrated promising single-agent activity in patients with relapsed or refractory multiple myeloma (3, 4). Its mechanism of action (MOA) is partly mediated through NF-κB inhibition that results in apoptosis, decreased angiogenic cytokine expression, and interference with the tumor microenvironment. Additional MOAs include c-Jun N-terminal kinase activation and growth factor expression modulation (5). Subsequent combination phase III studies established the drug for frontline multiple myeloma therapy (6). On the basis of a 33% response rate in a large multicenter phase II trial (7), bortezomib was also approved for patients with relapsed or refractory mantle cell lymphoma (MCL). More recently, regimens combining bortezomib with chemoimmunotherapy have yielded high response rates (83%–88%) in patients with relapsed/refractory indolent B-cell NHL and MCL (8, 9).

As perturbations of the cell cycle are almost universal in human malignancies, the cyclin-dependent kinases (CDK) have become attractive targets for cancer therapy (10, 11). Alvocidib (flavopiridol), the first CDK inhibitor to enter the clinic, globally represses transcription by non-ATP–competitive inhibition of positive transcription elongation factor b (P-TEFb, CDK9/cyclin T1; ref. 12), inducing downregulation of the short-lived antiapoptotic protein myeloid cell leukemia-1 (Mcl-1), which may represent the principal MOA of the drug in malignant hematopoietic cells (13). Indeed, in multiple myeloma cells, for which Mcl-1 is an essential survival protein (14), alvocidib induces apoptosis through Mcl-1 transcriptional repression and downregulation (15).

Alvocidib has been administered via a variety of schedules (10, 11). While the single-agent activity in patients with MCL or multiple myeloma using 1-hour intravenous bolus administration was modest to nonexistent (16, 17), marked clinical efficacy was noted in studies in patients with genetically high-risk chronic lymphocytic leukemia (CLL) using a pharmacologically derived “hybrid” administration schedule (18, 19). However, hybrid schedule results were disappointing in a phase I trial in patients with relapsed/refractory acute leukemias (20). Similarly, hybrid dosing did not improve results compared with a 1-hour bolus infusion in a phase I trial of alvocidib, fludarabine, and rituximab that reported an overall response rate of 82% in patients with MCL, B-NHL, or CLL (21). Finally, a recent randomized phase II comparison of bolus and hybrid dosing of alvocidib (followed sequentially by cytarabine and mitoxantrone) in patients with newly diagnosed poor-risk acute myeloid leukemia (AML) showed comparably encouraging results for both schedules, prompting selection of a bolus administration for successor studies (22).

Malignant cells are highly susceptible to a strategy in which survival signaling and cell-cycle regulatory pathways are simultaneously disrupted (23). Furthermore, Mcl-1 accumulation is an undesirable molecular consequence of proteasome inhibitor exposure, attenuating proapoptotic effects, and arguing that targeting Mcl-1 may increase the effectiveness of these agents (24). In support of these concepts, bortezomib and alvocidib interact synergistically to induce apoptosis in various leukemic cells (25, 26). In this context, alvocidib also inhibits IκB kinase, and by extension, NF-κB (27), and in this way may cooperate with bortezomib in triggering malignant hematopoietic cell death (26).

Given these observations, a multicenter phase I trial of this novel drug combination was initiated in patients with recurrent or refractory B-cell neoplasms. Patients received bortezomib by intravenous push followed by a 1-hour infusion of alvocidib on days 1, 4, 8, and 11. On the basis of encouraging results in the CLL trial (18, 19), a hybrid schedule of alvocidib as a 30-minute bolus infusion followed by a 4-hour infusion on days 1 and 8 was explored and the MTD determined (28). During this time, the original nonhybrid schedule of this trial was put on hold without reaching the MTD and accrual was resumed after completion of the hybrid schedule (28). The results of the original nonhybrid schedule are described in this article.

Bortezomib (PS-341; NSC 681239; Millennium Pharmaceuticals, Inc.) and alvocidib (flavopiridol; NSC 649890; Sanofi-Aventis Pharmaceuticals, Inc.) were obtained through the Cancer Therapy Evaluation Program (CTEP) Pharmaceutical Management Branch, NCI (Bethesda, MD).

Patients must have had a diagnosis of relapsed or refractory follicle center lymphoma (follicular or diffuse), MCL, marginal zone B-cell lymphoma (splenic, nodal, or extranodal), lymphoplasmacytoid lymphoma/immunocytoma, plasma cell myeloma, plasmacytoma, plasma cell leukemia, or Waldenstrom macroglobulinemia with no potential curative therapeutic options. Patients with a history of a central nervous system involvement were ineligible. Prior autologous SCT, but not prior allogeneic SCT, was allowed. Prior bortezomib was also allowed.

Additional eligibility criteria included being at least 18 years of age, an Eastern Cooperative Oncology Group performance status (ECOG) score of 1 or less, no grade 2 or greater neuropathy, preserved kidney and liver function, hemoglobin levels of 8 g/dL or higher, an absolute neutrophil count of 1.5 × 10⁹/L or greater, and platelet counts of 100 × 10⁹/L or greater.

This study was designed as a phase I, nonrandomized, dose-escalation study to determine the MTD for the bortezomib and alvocidib combination, where alvocidib is administered as a 1-hour infusion. The dose of alvocidib ranged from 15 mg/m² at dose level 1 to 90 mg/m² at dose level 9. The dose of bortezomib for dose level 1 was 1.0 mg/m² and for dose levels 2–9 was 1.3 mg/m². On days 1, 4, 8, and 11 of a 3-week cycle, bortezomib was administered via intravenous push over 3 to 5 seconds, after which alvocidib was administered as an intravenous infusion over 1 hour. Treatment was repeated every 3 weeks.

Disease status was assessed after the first 6 weeks of treatment and every 6 to 8 weeks thereafter. Patients experiencing a response or stable disease were allowed to continue treatment indefinitely at the investigator's discretion. Patients received full supportive care including herpes zoster prophylaxis. No specific precautions for tumor lysis syndrome (TLS) or cytokine release syndrome (CRS) were used before August 2007, but upon resumption of accrual to the nonhybrid regimen in May 2010, specific measures were applied. Pretreatment with allopurinol and overnight in-hospital observation for CRS and TLS monitoring were strongly encouraged for all patients. Clinical discretion was used to determine whether more stringent tumor lysis precautions were indicated on the basis of tumor burden and risk. For any signs of CRS, patients were treated with dexamethasone and prophylaxis was encouraged for those with circulating tumor cells documented by routine blood count.

The standard 3+3 dose-escalation design was used, with a dose level expansion of up to 6 evaluable patients if a dose-limiting toxicity (DLT) was noted. The MTD was defined as the highest dose level at which fewer than 2 of 6 patients experienced a DLT.

A DLT was initially defined as any of the following occurring during the first cycle of treatment and determined to be possibly, probably, or definitely related to study treatment: (i) grade 4 anemia; (ii) grade 4 absolute neutrophil count or platelet toxicity of greater than one week duration; (iii) grade 3 or greater febrile neutropenia; and (iv) any grade 3 or greater nonhematologic adverse events except grade 3 or greater nausea and vomiting without prophylactic and/or symptomatic treatment, grade 3 or greater diarrhea without maximum opioid and/or octreotide treatment, and grade 3 or greater infections, opportunistic infection with grade 2 or greater lymphopenia (specifically herpes zoster or herpes simplex infection). A protocol amendment (October 2008) modified the DLT definition to include instances in which both drugs are omitted due to dose modifications on at least two days of the scheduled drug administration during the first cycle.

NCI Common Terminology Criteria for Adverse Events (CTCAE) v3.0 was used for reporting adverse events until April 1, 2011, when mandatory conversion to CTCAE v4.0 supervened.

Response criteria used were based on the nature of the disease as follows: (i) patients with lymphomas were evaluated using the NCI-sponsored Working Group Lymphoma Response Criteria (29); (ii) patients with plasma cell myeloma or plasmacytoma were evaluated according to European Group for Blood and Bone Marrow Transplant (EBMT) criteria (30); (iii) patients with plasma cell leukemia were evaluated according to the criteria of Vela-Ojeda and colleagues (31); and (iv) patients with Waldenstrom macroglobulinemia were evaluated according to the Second International Workshop on Waldenstrom macroglobulinemia criteria (32).

Samples from the first 29 enrolled patients were obtained for alvocidib pharmacokinetic analysis. The samples were obtained before and after treatment on cycle 1 day 1 and cycle 3 day 8 according to the following schedule: preinfusion, immediately postinfusion, and 1, 2, 4, 8, 12, and 24 hours postinfusion. Alvocidib concentrations were analyzed as previously described (28).

Paired bone marrow aspirates were obtained from multiple myeloma patients only: (i) before initial study treatment, and (ii) on day 2 of cycle 1, approximately 24 hours after initial treatment. CD138⁺ multiple myeloma cells were enriched from bone marrow as previously described (28).

Protein extraction and Western blot analyses with primary antibodies to pJNK, Mcl-1, XIAP, and GAPDH were done as previously described (28). A PARP-specific antibody (BD Pharmingen) was also included. An Odyssey Imager (LI-COR Biosciences) was used to quantify binding of IRDye 680LT-conjugated secondary antibodies (LI-COR Biosciences).

RelA belongs to a family of transcription factors of NF-κB complex. Cytospin preparations of enriched CD138⁺ plasma cells were prepared from study patients before and during treatment. The cells were fixed, stained, and visualized by fluorescence microscopy as previously described (28). Analysis of NF-κB within the nucleus was performed using the Definiens Developer v1.5 software suite (Definiens). The nuclear region was determined with automatic threshold segmentation on the DAPI and Alexa 488 stains; further refined for plasma cells by the size and shape of the nuclei. Mean pixel intensity of the Alexa 488 signal was determined from the nuclei of the plasma cells.

Basic demographics, adverse events, DLTs, dose levels, and clinical responses were summarized by basic descriptive statistics such as frequency, proportion, mean, median, and range. Pharmacokinetic and pharmacodynamic analyses including all biomarkers and p65/ReIA were limited to descriptive statistics.

Studies were performed after Institutional Review Board approval and in accordance with an assurance filed with and approved by the Department of Health and Human Services. Informed consent was obtained from each patient. The ClinicalTrials.gov trial registration ID is NCT00082784.

A total of 44 patients, 12 female and 32 male, were enrolled on study between March 2004 and July 2012 (Table 1). The study participants included 35 Caucasians (including 4 Hispanics) and 9 African Americans (including 1 Hispanic). The median age of patients was 64.5 years (range: 40–79). Twenty-four patients were diagnosed with multiple myeloma, 2 were diagnosed with Waldenstrom macroglobulinemia, and 18 were diagnosed with NHL (5 of whom had MCL, an NHL subset). The median number of prior treatments was 3, with a range of 1 to 10. Ten patients had received prior autologous SCT and 10 patients had prior bortezomib therapy. Patients received a median of 5 cycles of study treatment, with a range of 1 to 16 cycles.

Myelosuppression was a common hematologic toxicity and typically manifested itself as grade 3 leukopenia (30%), lymphopenia (25%), neutropenia (34%), or thrombocytopenia (27%; Table 2). Grade 4 neutropenia (23%) and thrombocytopenia (14%) were also commonly observed. The most common nonhematologic toxicities included grade 3 diarrhea (20%) and fatigue (16%). Grade 3 pain and grade 3 sensory neuropathy were slightly less commonly experienced. One patient experienced grade 4 sensory neuropathy. A patient with follicular lymphoma developed grade 3 oral mucositis/esophagitis (which lasted more than 7 days) on day 2 of cycle 1, associated with both paraneoplastic pemphigus and a grade 4 CD4 lymphocyte count without HIV infection. The prior treatment regimen of the patient was rituximab and bendamustine. The most common grade 2 hematologic and nonhematologic toxicities, occurring in at least 20% of patients, are listed in Table 2. Before initiation of prophylactic antiviral treatment with acyclovir or comparable agents, 2 patients experienced grade 3 herpes zoster.

In the initial enrollment phase to the nonhybrid schema, 1 DLT was seen at dose level 5, necessitating expansion to 6 DLT-evaluable patients (Table 3). No other DLTs were noted in that expansion cohort. Patient enrollment continued through 3 DLT-evaluable patients at dose level 7 before the trial was interrupted and switched to the evaluation of the hybrid schema. Upon return to the nonhybrid schema, the 3+3 dose-escalation design of the study allowed progression through dose level 9. DLT accumulation at subsequent higher dose levels required a deescalation to lower dose levels until the MTD was reached at dose level 5. Grade 3 DLTs, experienced by 1 patient each, included back pain, fatigue, peripheral neuropathy, febrile neutropenia, TLS, diarrhea, and esophagitis/oral mucositis. TLS was manifested by short-lived laboratory perturbations (elevated phosphate and uric acid) resolving within 12 hours without sequelae or recurrence. One patient (dose level 9) experienced neutropenia that prompted dose omission of 2 scheduled doses of both agents (due to grade 4 ANC on day 4 and grade 3 ANC on day 8). Although grade 4 neutropenia lasting less than 7 days did not meet DLT criteria, omission of 2 scheduled doses of both investigational agents for this adverse event did meet DLT criteria. The MTD for a schedule of drug administration on days 1, 4, 8, and 11 of a 21-day cycle was determined to be 1.3 mg/m² bortezomib (intravenous push) and 40 mg/m² alvocidib (1-hour infusion).

Although this study was not powered to assess response, 3 complete responses (CR) and 10 partial responses (PR) were observed among the 39 patients evaluable for response (overall response rate = 33%; Table 4). The overall response rate was similar for both NHL patients (33%) and multiple myeloma patients (32%), with a 50% response rate in a limited sample of patients with Waldenstrom macroglobulinemia. CRs were experienced by 2 (one of whom had MCL) of the 15 NHL patients (13%) and 1 of the 24 multiple myeloma patients (4%). The multiple myeloma patient with the CR had previously been treated with bortezomib. PRs were achieved in 26% of the patients—3 with NHL (20%), 6 with multiple myeloma (27%), and 1 with Waldenstrom macroglobulinemia (50%). A total of 20 patients had stable disease (SD) as best response and 6 patients had progressive disease (PD). Among the patients not evaluable for response, 3 patients came off treatment for adverse events before the disease assessment interval, 1 patient came off treatment in cycle 1 due to a second cancer (renal) and did not have a response assessment, and 1 patient came off treatment in cycle 1 for convenience (moved far from the treatment center) and did not have a response assessment.

Samples were collected from the first 29 patients enrolled to the nonhybrid schedule. Alvocidib pharmacokinetics was evaluated using a noncompartmental (model-independent) approach. A sample immediately at the end of the 1-hour alvocidib infusion was only collected in 11 of the 29 patients. Consequently, peak alvocidib concentration (C_MAX) could be reported for these 11 patients. Of these 11 patients, only 9 had sufficient data to calculate a terminal elimination rate for the determination of half-life, clearance, and AUC_INF. No lag in C_MAX was observed. Maximum alvocidib concentration was observed immediately at the end of the infusion (Supplementary Fig. SS1). Cycle 1 day 1 alvocidib pharmacokinetic parameters per dose level are shown in Supplementary Table S1.

To determine the feasibility of measuring pharmacodynamic markers as a predictor of response, pre- and posttreatment samples (24 hours after the first doses of drugs) were obtained from the multiple myeloma patients for quantitative analysis of p65/RelA nuclear localization by immunofluorescence staining (Fig. 1) and for quantitative analysis of pJNK, Mcl-1, PARP, and XIAP by Western blotting (Fig. 2). Changes in expression of these proteins were previously observed in cells coexposed to alvocidib and bortezomib in vitro (25).

Cytospin preparations of enriched CD138⁺ plasma cells were prepared from the pre- and posttreatment bone samples from 8 patients with multiple myeloma including 2 patients from the PR group (DS027 and DS046), 5 patients from the SD group (DS011, DS047, DS050, DS055, and SG015), and 1 patient from the PD group (DS023). NF-κB activation was measured by monitoring changes in nuclear p65/RelA localization, as a function of mean pixel intensity, in pre- and posttreatment samples as shown in Fig. 1A. Minimal changes in nuclear RelA localization were detected in 3 patient samples (DS027, DS050, and DS023). Downregulation of nuclear p65/RelA after treatment was observed in 3 patient posttreatment samples (DS011 with 38% lower expression, DS047 with 31% lower expression, and DS055 with 29% lower expression). Upregulation of nuclear RelA was observed in 2 patients (DS046 with 44% higher expression and SG015 with 149% higher expression).

Expression levels of Mcl-1, pJNK, PARP, and XIAP in the pre- and posttreatment CD138⁺ cells from 4 patients (DS047, DS050, DS055, and SG015) in the SD group were determined by Western blot analysis (Fig. 2). Although some samples exhibited high magnitude changes in the posttreatment sample in comparison with the pretreatment sample (e.g., an 81% decrease of PARP for patient DS055; a 500% increase in XIAP for patient DS047), there was no correlation of the changes with clinical activity as consistent increases or decreases in expression were not observed for any of the proteins across all of the patients.

In the original study of the alvocidib/bortezomib regimen initiated in March 2004, 29 patients were first enrolled to the nonhybrid schedule at dose levels 1 to 7 without reaching the MTD. In 2007, the protocol underwent a CTEP-recommended amendment to proceed to a hybrid schedule based on promising data in CLL (19). Results from the hybrid schedule trial have been published (28). A total of 16 patients were treated, consisting of 11 male and 5 female patients. Nine patients had NHL (6 patients had MCL subtype), 6 had multiple myeloma, and 1 had an extramedullary plasmacytoma. Treatment was on a 21-day cycle: bortezomib by intravenous push on days 1, 4, 8, and 11 and alvocidib on days 1 and 8 by 30-minute bolus infusion followed by a 4-hour continuous infusion. The MTD was established as 1.3 mg/m² for bortezomib and 30 mg/m² for alvocidib (both the 30-minute bolus and 4-hour infusions). There were 2 CRs (12%) and 5 PRs (31%) observed (overall response rate = 44%). After determining the MTD for the hybrid schedule (28), data from the prehybrid schedule were reevaluated and determined to demonstrate a similar response and toxicity profile. With the agreement of CTEP, the nonhybrid schema trial was reopened to patient enrollment in May of 2010.

The results of this study indicate that this schedule of the combination of bortezomib and alvocidib can be administered safely to patients with recurrent or refractory indolent B-cell neoplasms (including NHL and multiple myeloma). The MTD was determined to be 1.3 mg/m² of bortezomib (intravenous push) followed by 40 mg/m² of alvocidib (1-hour infusion) on days 1, 4, 8, and 11 on a 21-day cycle (Table 3). In a previous hybrid study of these agents in the same patient population, the MTD was found to be 1.3 mg/m² of bortezomib (intravenous push) on days 1, 4, 8, and 11 followed by a hybrid schedule of 30 mg/m² of alvocidib (30-minute infusion) and then 30 mg/m² of alvocidib (4-hour infusion) on days 1 and 8 of a 21-day cycle (28).

The most common hematologic toxicities for both the current and previous study involved myelosuppression (including grade 3 leukopenia and lymphopenia) and grades 3 and 4 neutropenia and thrombocytopenia (Table 2; ref. 28). Likewise, the nonhematologic toxicity profile for both studies was very similar with grade 3 diarrhea, fatigue, and sensory neuropathy most frequently reported. One of the DLTs in the current regimen was grade 3 laboratory TLS in 1 patient with Waldenstrom macroglobulinemia at dose level 7 (Table 2). The event was managed with intravenous fluids and laboratory abnormalities normalized within 12 hours. This patient received a total of 6 cycles of the treatment and achieved a PR. No TLS was observed with the hybrid schedule (28). DLTs for both studies included grade 3 fatigue and febrile neutropenia, while the previous study also reported a single case of grade 3 aspartate aminotransferase elevation. Two patients on the present study developed herpes zoster (Table 2), but no additional episodes were reported following prophylactic antiviral therapy. Similarly, no episodes of herpes zoster were reported in the hybrid study, which included prophylactic antiviral therapy. (28). Patients received a median of 5 cycles of treatment on the current study (Table 1) compared with a median of 4 cycles of treatment on the previous study (28). Thus, both study regimens displayed a similar safety profile.

Although the primary objective of this phase I study was not to determine the efficacy of the treatment regimen, 3 CRs (NHL = 2, multiple myeloma = 1) and 10 PRs (NHL = 3, multiple myeloma = 6, Waldenstrom macroglobulinemia = 1) were observed for an overall CR+PR response rate of 13 patients out of 39 evaluable patients (33%; Table 4). This compares to an overall CR+PR response rate of 7 patients of 16 evaluable patients (total = 44%, NHL = 33%, multiple myeloma = 57%) in the previous study of the hybrid regimen (28). As the multiple myeloma patient in the present study who achieved a CR had previously progressed on bortezomib, this raises the possibility that either alvocidib exhibited single-agent activity, or, as previously observed in in vitro studies (25), it may have cooperated with bortezomib to lower the cell death threshold. Previously reported response rates for bortezomib as a single agent in relapsed/refractory disease are approximately 33% for MCL (7), 35% multiple myeloma (3), and 13.3% for indolent NHL including follicular lymphoma, marginal zone lymphoma, and small lymphocytic lymphoma (33). Because this study was not powered to characterize efficacy, firm conclusions about the activity of the combination of bortezomib and alvocidib in comparison with bortezomib as a single agent cannot be drawn. Nevertheless, the response rates in this phase I trial, which involved multiply-treated relapsed/refractory patients, are encouraging and warrant further study.

This nonhybrid schedule produced comparable pharmacokinetic parameter values as the hybrid schedule (28). Furthermore, the pharmacokinetics of the nonhybrid schedule were consistent with published results using 1-hour infusion schedule (34). Alvocidib pharmacokinetics in the absence of bortezomib were not evaluated in this clinical trial. Consequently, direct comparison cannot be made. However, pharmacokinetic parameters are in very close agreement with a previous study (34, 35) for dose-normalized C_MAX, dose-normalized AUC, half-life, and clearance. Together, these results suggest that the administration of bortezomib immediately before the alvocidib infusion does not affect its disposition.

On the basis of previous findings demonstrating that bortezomib and alvocidib interact synergistically to induce apoptosis in human leukemia cells (25, 26), it was plausible to postulate that coadministration of these agents in vivo to patients with B-cell malignancies might induce analogous perturbations in apoptotic-regulatory pathways in tumor cells as those previously observed in vitro. To assess the feasibility of acquiring the samples, and to begin evaluating pharmacodynamic markers of biologic effects of treatment, CD138⁺ plasma cells were obtained from the bone marrow of the patients with multiple myeloma, both before and 24 hours after treatment with bortezomib and alvocidib. Because bortezomib can inactivate NF-κB, which may activate the JNK pathway (26) as well as downregulate XIAP (36), translocation of the p65/RelA subunit of NF-κB to the nucleus and the accumulation of pJNK and XIAP were selected as candidate markers. Similarly, because alvocidib can attenuate Mcl-1 cytoprotective effects by preventing proteasome inhibitor-mediated Mcl-1 accumulation (25), total Mcl-1 protein levels were monitored. Finally, PARP cleavage was used as an indicator of apoptosis.

Of 24 multiple myeloma patients in the study, immunofluorescence staining for p65/RelA was conducted on 8 paired plasma cell samples and Western blot analysis for pJNK, Mcl-1, PARP, and XIAP was conducted on 4 paired plasma cell samples. Among the samples assayed for nuclear p65/RelA expression, there was no clear pattern of expression associated with disease status. Although a decrease in expression was observed in 3 samples (all PRs), an increase in expression was observed in 2 samples (1 PR and 1 SD; Fig. 1). Similarly, among the 4 paired samples (all SD) analyzed by Western blot analysis, and despite the fact that single-agent alvocidib has been shown to diminish Mcl-1 expression (37), only one sample showed a decrease in expression. While this result was unanticipated, it is possible that in this instance, diminished Mcl-1 proteasomal degradation may have predominated. Although a major mode of action of the bortezomib/alvocidib regimen is presumed to be apoptosis induction, only sample SG015 (from a patient with SD) displayed enhanced posttreatment PARP cleavage. Whether the inconclusive results seen here reflect the small sample size, differing drug doses, or the possibility that these molecules may not be robust markers of therapeutic activity cannot be determined in the context of this phase I trial. While bone marrow sampling was mandatory for patients with myeloma, it proved logistically cumbersome, thereby limiting the number of samples obtained. These questions could possibly be answered in a more highly powered phase II trial with larger patient numbers and more uniform drug dosing.

In conclusion, this phase I study has determined the MTD for the combination of bortezomib and alvocidib administered as a 1-hour infusion in patients with recurrent or refractory indolent B-cell cancers. Both the observed hematologic and nonhematologic toxicities were similar to those observed with bortezomib as a single agent combined with a hybrid dose schedule of alvocidib (28). The dosing regimen employed here resulted in 3 CRs and 10 PRs in a heavily pretreated patient population. The frequency of these CRs and PRs was encouraging and similar to that observed in a previous trial of these agents that employed a more cumbersome hybrid dosing schedule (28). Because the simpler regimen used in the current trial demonstrated comparable activity to the hybrid dosing scheme, and because the nature and frequency of the toxicities associated with the regimens did not differ substantively, the results reported here argue for selection of the nonhybrid regimen for successor combination trials of alvocidib. As the toxicities of the current regimen were not inconsequential, plans are under development for a new phase I trial of alvocidib and a next-generation orally active proteasome inhibitor, MLN9708, administered in conjunction with the nonhybrid schedule of alvocidib in patients with multiple myeloma who have progressed following any proteasome inhibitor treatment.

G.D. Roodman reports receiving a commercial research grant from Eli Lilly and is a consultant/advisory board member for Amgen. No potential conflicts of interest were disclosed by the other authors.

Conception and design: B. Holkova, A.D. Colevas, W.D. Figg, J.D. Roberts, S. Grant

Development of methodology: B. Holkova, M. Kmieciak, V. Ramakrishnan, C.J. Peer, L.A. Doyle, D. Coppola, J.D. Roberts, S. Grant

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): B. Holkova, M. Kmieciak, E.B. Perkins, R.C. Baz, G.D. Roodman, R.K. Stuart, C.J. Peer, L. Kang, M.B. Tombes, E. Shrader, C. Weir-Wiggins, W.D. Figg, D. Coppola, J.D. Roberts, D. Sullivan

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): B. Holkova, M. Kmieciak, R.C. Baz, W. Wan, C.J. Peer, J. Dawson, E. Shrader, C. Weir-Wiggins, H. Sankala, K.T. Hogan, A.D. Colevas, W.D. Figg, D. Coppola, S. Grant

Writing, review, and/or revision of the manuscript: B. Holkova, M. Kmieciak, P. Bose, R.C. Baz, G.D. Roodman, R.K. Stuart, V. Ramakrishnan, C.J. Peer, M.B. Tombes, E. Shrader, H. Sankala, K.T. Hogan, A.D. Colevas, L.A. Doyle, W.D. Figg, D. Coppola, D. Sullivan, S. Grant

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): B. Holkova, J. Dawson, C. Honeycutt, M.B. Tombes, E. Shrader, M. Wellons

Study supervision: B. Holkova, R.C. Baz, C. Honeycutt, C. Weir-Wiggins, A.D. Colevas, L.A. Doyle, J.D. Roberts, D. Sullivan

Other (helped in acquistion of image analysis data): L. Kang

Other (study implementation, subject, site, and data management): M.B. Tombes

The authors thank Sarah Kolla (Virginia Commonwealth University) for assistance with the pharmacodynamic studies, Erin Gardner (NCI) for her assistance with the pharmacokinetic studies, and Joe Johnson (H. Lee Moffitt Cancer Center & Research Institute) for his valuable assistance in the Analytical Microscopy Core Facility supported by NCI P30 CA76292. Services in support of the research project were generated by the Virginia Commonwealth University Massey Cancer Center Biostatistics Shared Resource, supported, in part, with funding from NCI Cancer Center Support Grant P30 CA016059.

The study was supported by grants from the NIH [NCI R01 CA93738, NCI R01 CA100866, NCI R21 CA110953, NCI R01 CA167708, NCI P50 CA130805 (Lymphoma SPORE), NCI P30 CA016059 (Cancer Center Support Grant to Massey Cancer Center), NCI P30 CA76292 (Cancer Center Support Grant to H. Lee Moffitt Cancer Center & Research Institute), NCI N01 Contract HHSN261201100100C for the Southeast Phase 2 Consortium, NCRR M01 RR000065 (GCRC), and National Center for Advancing Translational Sciences KL2TR000057 (CTSA)] and an award from the Multiple Myeloma Research Foundation.

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.