ALBERT

Description: Overexpression of Bcl-2 is a potential mechanism for chemoresistance in acute leukemia and has been associated with unfavorable clinical outcome. We hypothesized that down-regulation of Bcl-2 would restore chemosensitivity in leukemic cells. To test this hypothesis, we performed a phase 1 study of G3139 (Genasense, Genta, Berkeley Heights, NJ), an 18-mer phosphorothioate Bcl-2 antisense, with fludarabine (FL), cytarabine (ARA-C), and granulocyte colony-stimulating factor (G-CSF) (FLAG) salvage chemotherapy in patients with refractory or relapsed acute leukemia. Twenty patients with refractory or relapsed acute myeloid leukemia (AML) or acute lymphoblastic leukemia (ALL) were enrolled. G3139 was delivered by continuous infusion on days 1 to 10. FLAG chemotherapy was administered on days 5 to 10. Common side effects of this combination included fever, nausea, emesis, electrolyte imbalance, and fluid retention that were not dose limiting. Plasma pharmacokinetics of G3139 demonstrated steady-state concentration (Css) within 24 hours. Of the 20 patients, 9 (45%) had disease response, 6 (5 AML, 1 ALL) with complete remission (CR) and 3 (2 AML and 1 ALL) with no evidence of disease but failure to recover normal neutrophil and/or platelet counts or to remain in remission for at least 30 days (incomplete remission). Bcl-2 mRNA levels were down-regulated in 9 of the 12 (75%) evaluable patients. This study demonstrates that G3139 can be administered safely with FLAG chemotherapy and down-regulate its target, Bcl-2. The encouraging clinical and laboratory results justify the current plans for a phase 3 study in previously untreated high-risk AML (ie, age at least 60 years).

Description: Background: Although complete remission rates for AML are near 70% with combination induction and consolidation chemotherapy, most patients will relapse and die from the disease or treatment complications. New agents with unique mechanisms are needed. One such class of therapeutics are fusion proteins consisting of protein synthesis inactivating peptide toxins fused to tumor cell selective ligands. DT388-IL3 is one such fusion protein. Rationale: In preclinical studies, DT388-IL3 was cytotoxic to the IL3 receptor expressing leukemia cell lines but not toxic to IL3 receptor negative cell lines. This agent was less toxic to normal progenitors and not toxic to early hematopoietic stem cells. The majority of AML progenitors overexpress IL3 receptors. Animal model work in mice bearing human leukemia cells has demonstrated anti-leukemia efficacy which is dose dependent with this agent. Toxicities in monkeys include vascular leak syndrome and pancytopenia observed only at the highest doses. The MTD in monkeys was estimated at 60mcg/kg/day. We report preliminary data on the use of DT388-IL3 fusion protein in humans from an ongoing phase I trial. Pharmacokinetics; clinical and immune response to this novel fusion protein are also being followed. Patients and Methods: Patients with refractory AML were eligible. The first dose level was qd M-W-F X six doses of DT388-IL3 at 4mcg/kg/day with dose escalation planned for subsequent patients. Patients with progression of disease or unacceptable study drug toxicity were to be removed from the study. Toxicity was graded according to NCI CTCAE version 3.0. Three patients have been treated with DT388-IL3. Serum samples were collected and will be assayed for anti-DT388-IL3 antibodies prior to and after treatment. Blood samples were obtained to measure circulating levels of active DT388-IL3 and its half life. Patient blasts were also collected prior to treatment for later analysis of expression of IL3 receptors. Result: Two patients tolerated the treatment schedule(of six doses) without any significant toxicities. Mild fever, headaches, nausea were noted. Both of these patients had progression of disease-one during treatment and one on day 15 bone marrow biospy. The above mentioned patients died secondary to disease complications at 2 weeks and 18 weeks after their last dose of the study agent respectively. DT388-IL3 levels on these two patients post infusion were below the the reliable detectable limits of the assay. The third patient became febrile and hypotensive after the first dose. The hypotension persisted and she did not receive any further doses. This patient is alive 5 weeks later with supportive care alone. DT388-IL3 levels following this patient’s dose are as follows: 2min post infusion 34.3ng/ml, 30min post infusion 1.9ng/ml, 60min post infusion 0.075ng/ml, 120min post infusion 0.003ng/ml, 240min post infusion undetectable. Conclusion: Preclinical/animal studies suggest that DT388-IL3 has anti-leukemia efficacy. Preliminary data from our ongoing phase I trial reveals minimal study agent related toxicity and no life threatening complications at this first dose level. Dose escalation is planned as per protocol.

Description: Bcl-2 acts as an important regulator of the mitochondrial pathway of apoptosis and promotes resistance of MM cells to chemotherapy. The Bcl-2 antisense oligonucleotide G3139 specifically targets Bcl-2 and may enhance the anti-tumor efficacy of Dex and Thal. In this trial G3139 was administered at 5 mg to the first 3 Pts and then 7 mg/kg/d by IVCI for 7d of 21d cycle. On day 4, Pts started Dex 40 mg daily for 4 d and Thal 100-400 mg as tolerated. After 3 cycles, responding Pts continued G3139 on a 5-week cycle with Dex 20 mg x 4d and Thal at the tolerated daily dose for up to 1 yr with an optional second yr for responding Pts. Thirty-three Pts treated to date had the following characteristics: median age 60 yrs (range: 28- 76), 22 males; 16 Pts had complex karyotypes; 14 Pts had B2M 〉 2.5 g/dl; LDH 〉1.5 normal in 7 Pts; Cr 〉1.5 mg/dl in 6 Pts; platelets

Description: Therapeutic options for chronic lymphocytic leukemia (CLL) at relapse are limited because of myelosuppressive toxicity. Denileukin diftitox (ONTAK®, Ligand Pharmaceuticals) is a genetically engineered fusion protein comprising the enzymatically active domain of diphtheria toxin and the full length sequence of interleukin-2 (IL-2) targeting malignancies expressing the medium and high affinity IL-2 receptors. We designed a phase II study to evaluate the efficacy of ONTAK® in patients with fludarabine-refractory CLL, which is a follow-up to the previously published study (Frankel, et al, Clin. Cancer Res.2003; 9:3555). Denileukin diftitox was administered at a dose of 18μg/kg IV daily for 5 days every 3 weeks, for a maximum of 8 cycles. Thirteen patients have been treated so far, with 10 patients being evaluable for response (completed ≥ 3 cycles). Median age was 59 years (range 44–84), and 62% (8/13) were Rai stage III-IV, with a median of 3 prior therapies (range 1–6). The overall response was 40%, with 1 CR (10%, duration of response 5+ months) and 3 PR (30%, duration of response 3+, 3+ and 4+ months). Two responding patients (both PR) are still on study, while two (1 CR, 1 PR) were removed from study because of toxicities after 7 and 5 cycles, respectively. Four patients (40%) had progressive disease after cycles 3, 4, 4, and 7, respectively. One patient has completed four cycles and restaging studies are pending. Of the 3 patients not evaluable for response, two are still on study (having not completed 3 cycles), while one refused further treatment after 4 doses of cycle one. The grade 3/4 toxicities encountered were: neutropenia 4/13, thrombocytopenia 4/13, vascular leak syndrome 3/13, left ventricular cardiac dysfunction 1/13, hypotension 2/13, tachyarrhythmia 3/13, elevated PT 1/13, fatigue 1/13, rash 1/13, SIADH 1/13, constipation 1/13, vomiting 2/13, petechiae 1/13, transient elevation of GGT 1/13, transient elevation of AST/ALT 7/13, hyperglycemia 4/13, electrolyte imbalance 8/13, infection and/or febrile neutropenia 4/13, insomnia 1/13, visual disturbance 1/13, dyspnea 2/13, hypoxia 2/13. We conclude that denileukin diftitox has activity in CLL, with toxicities that can be managed with adequate premedication and close monitoring.

Description: To develop a cytotoxic drug which targets the IL-3 receptor (IL-3R) on human AML cells we previously developed a fusion protein containing a truncated form of diphtheria toxin which lacks the native binding site (DT388) fused to human IL-3. This molecule kills leukemic progenitors from many AML patients while showing little toxicity to normal hematopoietic progenitors. However, some AML samples showed little or no cell kill after exposure to this molecule. To attempt to improve the cytotoxicity of DT388IL-3 two variants of the toxin were constructed which contain alterations in the IL-3 residues that are designed to enhance binding affinity to the IL-3R. The two variants, DT388IL3[K116W] and DT388IL3[D125-133], have substitution of a hydrophobic tryptophan group at the 116 position and an eight amino acid deletion from the C-terminus of the IL-3 molecule, respectively, while the catalytic and translocation domains of DT388 remain unchanged. These variant DT388IL3 molecules and the unmodified ‘native’ fusion toxin were compared for their ability to kill AML colony forming cells (AML-CFC) from the peripheral blood of 13 newly-diagnosed AML patients and myeloid CFC from 3 normal bone marrows (NBM). AML and NBM cells were cultured for 24h with or without fusion toxin at concentrations varying between 1 and 250 ng/ml prior to plating in CFC assays. Little or no AML-CFC kill was observed for 3/13 samples. The mean % AML-CFC kill for the remaining 10 AML samples ranged from 21 – 61% for the lowest and highest concentration of native DT388IL3 tested and was significantly higher (P90% kill achieved for 6 of these at concentrations as low as 1 ng/ml. The concentration of DT388IL3[Kll6W] required to achieve ≥50% kill of AML-CFC was on average ≥5-fold lower than the concentration of native toxin required to achieve the same effect. Thus, the variant DT388IL3 molecules tested show enhanced cytotoxic activity against AML progenitors with little change in the toxicity profile against normal hematopoietic precursors. In particular, the DT388IL3[K116W] variant warrants further testing against more primitive normal and leukemic progenitors as a potentially promising new therapeutic agent for AML. Such studies are underway.

Description: Patients with relapsed AML over the age of 60 have a poor prognosis. Gemtuzumab ozogamicin (GO) has been approved for older pts in first relapse, although many pts who attain complete remission (CR) do not fully recover normal platelet count (so-called CRp). In vitro studies have shown that oblimersen down-regulates Bcl-2 in AML cells and enhances apoptotic cell death induced by GO. We conducted a Phase 2 study to evaluate the safety and efficacy of GO combined with oblimersen for older pts with AML. Eligibility requirements included: age ≥ 60 yrs; AML in 1st relapse; ≥ 3 mos 1st CR duration; ≥ 25% CD33-positive AML cells. Pts received oblimersen at a dose of 7 mg/kg/d for 7 days by CIV beginning on days 1 and 15; GO was given at a dose of 9 mg/m2 IV over 2 hrs on days 4 and 18. A total of 48 pts were enrolled (ITT population), all of whom received at least 1 dose of oblimersen; 9 pts failed to receive the required 2 doses of GO (per-protocol population, n=39). The median age was 67 (range, 59 to 88 yrs). Duration of 1st CR: 〈 6 mos: 7 pts; (15%); 6 to 12 mos: 29 pts (60%); 〉 12 mos: 12 pts (25%). No. of prior regimens: 1 (17 pts, 35%); 2 or 3 (26 pts, 54%); ≥ 4 (5 pts, 10%). Among treated pts, 79% completed 21 days of protocol therapy. Overall, 12 pts achieved a major response, either CR (n=5) or CRp (n=7), for an ITT response rate of 25% and a per-protocol response rate of 31%. The median time to remission was 52 days. Ten of the 12 responders survived 〉 6 mos, whereas only 6 non-responders survived ≥ 6 mos. Serious adverse events for the oblimersen/GO combination were qualitatively similar to those reported for GO alone and included among other reactions: Grade 3-4 febrile neutropenia (42%) or thrombocytopenia 33%; nausea; fever; rigors, and dyspnea. Treatment-emergent adverse reactions led to discontinuation of protocol therapy in 10 pts (21%). The most common serious adverse event was febrile neutropenia (25%). One pt (2.1%) died during treatment (sepsis) and 16 pts (33%) died within 30 days of last study medication (infection, bleeding, respiratory failure, progressive AML, and other disease-related complications). No episodes of VOD were observed. Oblimersen can be safely combined with GO; however, pts enrolled in this study appear to have had more unfavorable characteristics at entry compared with prior studies using GO alone in pts with relapsed AML. Therefore, assessment of an incremental benefit from the addition of oblimersen will require a randomized trial.

Description: Over-expression of Bcl-2 has been linked to acquired dex resistance in MM cells. Conversely, reduction of Bcl-2 using oblimersen has been shown to enhance dex-induced apoptosis in MM cell lines and in fresh MM cells in culture (Liu et al., Blood101:4105, 2003). Preliminary non-randomized clinical studies have suggested that oblimersen may potentiate the activity of VAD and dex/thalidomide in patients (pts) with resistant MM. We have conducted a randomized, multinational, Phase 3 trial of pts with advanced MM using dex given with or without oblimersen. Eligibility requirements included: relapsed or refractory MM; measurable M-protein in serum or urine; ≤ 6 prior regimens; ECOG PS ≤ 3; serum Cr ≤ 1.5 mg/dL; ANC 〉 1,000/ml; plts ≥ 50,000/ml. Exclusions: previous allogeneic transplant; other significant medical disease. Pts were stratified according to number of prior treatments (1 or 2 vs. ≥ 3 regimens), prior autologous stem cell transplant (yes/no), and primary resistance to vs. relapse from primary treatment. All pts received dex 40 mg p.o. x 4 days during weeks 1, 2 and 3. Pts randomized to oblimersen received 7 mg/kg/d x 7 days by IV infusion, beginning 3 days before dex treatment during the 1st and 3rd weeks, and during all subsequent dex cycles. Beginning on week 5, additional 4-day dex cycles were repeated in stable or responding pts every 3 weeks, for a maximum of 12 months. All pts received prophylactic H-2 antagonists, TMP/SMZ, and a bisphosphonate. Response assessments were conducted prior to each treatment cycle. The primary endpoint of the study was to compare time-to-disease progression between the 2 treatment groups. Secondary endpoint comparisons included overall survival, event-free survival at 6 months, overall response rates, response duration, clinical benefit, and safety. A total of 224 pts were randomized into the study. All pts have completed 〉 12 mos follow-up, and final results of this trial will be presented.

Description: Oblimersen sodium (Ob, Genasense™) is an 18 mer phosphorothioate antisense directed against bcl-2, an antiapoptotic protein that mediates chemoresistance in malignant cells. In order to assess whether detectable intracellular concentrations (ICs) of Ob are achievable in vivo, we have developed a sensitive and specific ELISA-based assay. This assay involves Ob hybridization to a 5′-end overhang of a 3′-biotinylated capture probe ligated to a digoxigenin (Dig)-labeled probe, and detection by an anti-Dig-alkaline phosphatase system. In K562 cell extract, the assay was linear within 50–2000 pM range with a limit of quantification (LOQ) of 50 pM (equivalent to 5.0 fmol/100 μL). The withtin-run coefficients of variation (CVs) in 4 spiked concentrations were between 3–7% in 6 replicates with accuracy values between 93–109%. The between-run CVs were between 6–12% and accuracy values between 97–102%. The specificity of the assay was demonstrated by low cross reactivity with mismatched oligonucleotides and putative 3′-end metabolites shortened by 1, 2 or 3 nucleotides. Validation of IC measurement was performed in vitro in K562 cells treated with fluorescent labeled Ob conjugated to oligofectamine. At Ob concentrations between 0.1 – 10 μM, Bcl-2 mRNA downregulation measured by real-time RT-PCR occurred efficiently. Nonlinear regression analysis of a dose-response curve showed that 50% Bcl-2 downregulation (IC50) occurred at approximately 0.29 μM, corresponding to an Ob IC concentration of 37 pmole/mg protein. Cellular uptake was confirmed by microscopy and flow cytometry. To validate these results in vivo, we measured Ob IC in bone marrow samples from untreated AML pts aged 〉 60 yrs enrolled on the phase I study OSU 0164. These pts were induced with Ob 7 mg/kg/d CIVI on days 1–10, cytarabine 100 mg/m2/d CIVI on days 4–10 and daunorubicin administered iv at two dose levels (45 mg/m2/d IV and 60 mg/m2/d) on days 4–6. Among 21 pts assessable for clinical response and Bcl-2 levels, at pretreatment, Bcl-2 copy numbers (normalized to ABL) were higher among 12 pts who achieved a CR (median 85,325; range 19,120–149,100) than among 9 non-responsive (NR) pts (32,100 bcl-2 /abl copies; range 1,488–163,500) (P=.04; Mann-Whitney test). Following 72 hr Ob infusion, a decrease (−38%) in median Bcl-2/ABL mRNA copies in CR patients and an increase (+115%) in Bcl-2/ABL copies in NR pts (P=.002; Mann-Whitney test) were observed by real time RT-PCR. A trend in higher median IC of Ob was observed in CR pts (17.0 pmole/mg protein; range 1.5–30.0) as compared to NR pts (4.4 pmole/mg protein; range 0.33–28.0) (P=.06; Mann-Whitney test). Six of 7 pts with IC above the median obtained a CR. No differences were observed in the Ob plasma PKs between the CR pts [median steady state concentration (Css) 2.8 μg/mL, area-under-the-curve (AUC) 772 μg*hr/mL and clearance (Cl) 9.6 L/hr) and the NR pts (median Css 3.4 μg/mL, AUC 752 μg*hr/mL, Cl 6.4 L/hr). Although the number of samples analyzed was small, our data suggest that, despite interpatient variability of both Bcl-2 mRNA expression and Ob uptake, this antisense can be successfully delivered to pts and result in clinically relevant target downregulation. A Cancer and Leukemia Group B phase III AML study to characterize prospectively the interplay of IC levels of the Ob and Bcl-2 downregulation is in progress.

Description: DT388–GM-CSF, a targeted fusion toxin constructed by conjugation of human granulocyte-macrophage colony-stimulating factor (GM-CSF) with the catalytic and translocation domains of diphtheria toxin, is presently in phase I trials for patients with resistant acute myeloid leukemia. HL-60/VCR, a multidrug-resistant human myeloid leukemia cell line, and wild-type HL-60 cells were used to study the impact of DT388–GM-CSF on metabolism of ceramide, a modulator of apoptosis. After 48 hours with DT388–GM-CSF (10 nM), ceramide levels in HL-60/VCR cells rose 6-fold and viability fell to 10%, whereas GM-CSF alone was without influence. Similar results were obtained in HL-60 cells. Examination of the time course revealed that protein synthesis decreased by about 50% and cellular ceramide levels increased by about 80% between 4 and 6 hours after addition of DT388–GM-CSF. By 6 hours this was accompanied by activation of caspase-9, followed by activation of caspase-3, cleavage of caspase substrates, and chromatin fragmentation. Hygromycin B and emetine failed to elevate ceramide levels or induce apoptosis at concentrations that inhibited protein synthesis by 50%. Exposure to C6-ceramide inhibited protein synthesis (EC50∼5 μM) and decreased viability (EC50 ∼6 μM). Sphingomyelinase treatment depleted sphingomyelin by about 10%, while increasing ceramide levels and inhibiting protein synthesis. Diphtheria toxin increased ceramide and decreased sphingomyelin in U-937 cells, a cell line extremely sensitive to diphtheria toxin; exposure to DT388–GM-CSF showed sensitivity at less than 1.0 pM. Diphtheria toxin and conjugate trigger ceramide formation that contributes to apoptosis in human leukemia cells through caspase activation and inhibition of protein synthesis.

Description: The α subunit of the human granulocyte-macrophage colony-stimulating factor (GM-CSF) receptor has several isoforms that result from alternative splicing events. Two forms, α-1 and α-2, have intracytoplasmic sequences that are identical within a membrane-proximal domain but differ completely distally. Variant and mutated GM-CSF receptor α subunits, along with the β subunit (βc protein) were expressed in M1 murine leukemia cells. and the ability of the receptors to signal for differentiation events and to activate Jak/Stat signaling pathways was examined. All cell lines expressing both α and βc proteins exhibited high-affinity binding of radiolabeled human GM-CSF. Receptor α subunits with intact membrane-proximal intracellular domains could induce expression of the macrophage antigen F4/80 and down-regulate the expression of CD11b. Addition of recombinant human GM-CSF to cells expressing α-1 subunits induced the expression of CD86 and tyrosine phosphorylation of Jak-2 and its putative substrates SHPTP-2, Stat-5, and the GM-CSF receptor βc subunit. Cells containing α subunits that lacked a distal domain (term-3) or had the alternatively spliced α-2 distal domain showed markedly decreased ability to support tyrosine phosphorylation of Jak-2 and its substrates or to up-regulate CD86. Ligand binding induced stable association of the α-1 subunit and βc protein. In contrast, the α-2 subunit did not stably associate with the βc subunit. These data identify potential molecular mechanisms for differential signaling of the α-1 and α-2 proteins. The association of unique signaling events with the 2 active GM-CSF α subunit isoforms offers a model for variable response phenotypes to the same ligand.