ALBERT

Description: Unconjugated monoclonal antibodies that target hematopoietic differentiation antigens have been developed to treat hematologic malignancies. Although some of these have activity against chronic lymphocytic leukemia and hairy cell leukemia, in general, monoclonal antibodies have limited efficacy as single agents in the treatment of leukemia. To increase their potency, the binding domains of monoclonal antibodies can be attached to protein toxins. Such compounds, termed immunotoxins, are delivered to the interior of leukemia cells based on antibody specificity for cell surface target antigens. Recombinant immunotoxins have been shown to be highly cytotoxic to leukemic blasts in vitro, in xenograft model systems, and in early-phase clinical trials in humans. These agents will likely play an increasing role in the treatment of leukemia.

Description: Abstract 839 Although most children with ALL are cured, treatment is associated with multiple toxicities and outcome after relapse is poor. New therapies are needed to overcome drug resistance and reduce non-specific toxicities of chemotherapy. CD22 is a B-lineage differentiation antigen expressed on most B-lineage ALL blasts. The anti-CD22 immunotoxin RFB4(dsFv)-PE38 CAT-3888 (BL22) was recently shown to have clinical activity with an acceptable safety profile in children with ALL (Blood 2007;110:262a). We undertook a Phase I trial of a modified agent with higher CD22 binding affinity (CAT-8015 or HA22). Methods: Patients 6 months to 24 years of age with relapsed or refractory CD22 + B-lineage ALL or non-Hodgkin lymphoma were eligible for enrollment into this Phase I trial. CAT-8015 was administered at doses of 5, 10, 20, or 30 mcg/kg every-other-day for 6 doses every 21 days for up to 6 cycles. One patient was enrolled at each of the first 3 dose levels (5, 10, 20 mcg/kg) with standard 3+3 dose escalation commencing at 30 mcg/kg. All patients received acetaminophen, ranitidine and diphenhydramine to mitigate infusion-related symptoms, and prophylaxis for central-nervous-system leukemia with intrathecal hydrocortisone, cytarabine and methotrexate. Patients at high risk for tumor lysis syndrome received standard prophylaxis. Results: Seven patients with ALL (6 precursor-B, 1 mature B-cell) 5 to 17 years of age (median, 10) were treated on the clinical trial. All patients had been heavily pre-treated and had baseline cytopenias due to active malignancy and thus were not evaluable for hematologic toxicities. The most common adverse events observed to date have been hyperbilirubinemia, transaminase elevations, hypoalbuminemia, elevated creatinine, febrile neutropenia, abdominal pain, pyrexia, hypertension, microscopic proteinuria, hemoglobinuria, hypoxia and pleural effusion. Two of 4 patients treated at 30 mcg/kg experienced Grade 3 or greater toxicity consistent with capillary leak: 1 with Grade 3 pleural effusion and hypoxia and 1 with Grade 4 vascular leak syndrome. All toxicities attributed to CAT-8015 were reversible. Clinical activity was demonstrated in 4 of 7 subjects. One patient treated at 10 mcg/kg had a complete remission by morphology and flow cytometry. Three patients met the protocol definition for hematologic activity (blood count improvement). One of these patients developed high-titer neutralizing antibodies. Two patients met the protocol definition for stable disease. The patient treated at the lowest dose level had progressive disease. Conclusions: CAT-8015 appears to be active against chemotherapy-refractory ALL. Strategies to predict and/or prevent vascular leak syndrome are currently being developed. Disclosures: No relevant conflicts of interest to declare.

Description: To determine whether the multidrug resistance gene MDR1could act as a selectable marker in human subjects, we studied engraftment of peripheral blood progenitor cells (PBPCs) transduced with either MDR1 or the bacterial NeoR gene in six breast cancer patients. This study differed from previous MDR1 gene therapy studies in that patients received only PBPCs incubated in retroviral supernatants (no nonmanipulated PBPCs were infused), transduction of PBPCs was supported with autologous bone marrow stroma without additional cytokines, and a control gene (NeoR) was used for comparison with MDR1. Transduced PBPCs were infused after high-dose alkylating agent therapy and before chemotherapy with MDR-substrate drugs. We found that hematopoietic reconstitution can occur using only PBPCs incubated ex vivo, that theMDR1 gene product may play a role in engraftment, and that chemotherapy may selectively expand MDR1 gene-transduced hematopoietic cells relative to NeoR transduced cells in some patients.

Description: A recombinant immunotoxin, BL22 has shown clinical efficacy against hairy cell leukemia and other B-cell non-Hodgkin’s lymphomas. BL22 contains an antibody-derived domain that recognizes CD22, and PE38, a truncated Pseudomonas exotoxin A domain, for inhibition of protein synthesis. Previous studies have shown that Pseudomonas exotoxin A-based immunotoxins exert their cytotoxicity by apoptosis and additional, yet uncharacterized mechanisms. We investigated the cytotoxicity of BL22 in two mantle cell lymphoma lines, NCEB-1 and Granta-519, and in the diffuse, large-cell lymphoma line SU-DHL-4. Presence of the CD22 target was confirmed by flow cytometry. Treatment of NCEB-1 cells to 1 μg/ml BL22 for 24-32h reduced the numbers of viable cells in comparison to cells exposed to a non-binding immunotoxin. Induction of apoptosis, determined by activation of caspase-3 and exposure of phosphatidylserine, was not detected before 48h of exposure. Incorporation of bromo-deoxyuridine into DNA was, however, almost completely inhibited after 24h. Correspondingly, the numbers of G1 phase cells were increased upon treatment with BL22 as determined by flow cytometry. G1-arrest resulted from decreased enzyme activities of cyclin-dependent kinases (cdk)-2 and 4 as assessed in vitro with respective substrates, histone H1 and the retinoblastoma susceptibility (RB) protein. Immunoblotting revealed markedly reduced amounts of cyclins A, E, D1 and D3 whereas cdk2 and cdk4 protein expression levels remained unchanged. Expression of the RB protein was decreased, and the protein was hypophosphorylated in immunotoxin-treated cells. Pulse-labeling with [35S]-labeled amino acids, followed by immunoprecitation of cyclins E and D1 confirmed that BL22 inhibited synthesis of these cyclins. Thus, immunotoxin-induced cell cycle arrest results from insufficient synthesis of regulatory cyclins required for progression through G1 into S-phase. Similar results were obtained in Granta-519 cells. SU-DHL-4 diffuse, large-cell lymphoma cells were highly sensitive to BL22, and G1 arrest was observed as early as after 8–24h in these cells. Expression of cyclins A, D3 and E was decreased while cyclin D1 was not expressed in SU-DHL-4 cells. Interestingly, cell cycle arrest did not prevent from subsequent induction of apoptosis in lymphoma cells continuously exposed to the immunotoxin. In conclusion, we have characterized cell cycle arrest as a previously unknown mechanism that contributes to the cytotoxicity of Pseudomonas exotoxin-based immunotoxins.

Description: The immunoglobulin superfamily receptor translocation associated 2 (IRTA2) gene encodes a cell surface receptor homologous to the family of Fc receptors. Because of the restricted expression of the mRNA in B cell-linage cells, IRTA2 is a new potential target for the immunotherapy of B cell malignancies. To study the function and expression of the IRTA2 gene product, we produced monoclonal antibodies (MAbs) specific to IRTA2. We successfully made 36 MAbs from a mouse immunized with IRTA2-encoding plasmid DNA. All the MAbs reacted with IRTA2-extracellular domain-Fc fusion protein in ELISA and 293T cells expressing IRTA2 in a fluorescence-activated cell sorter (FACS) analysis. Binding of each MAb to a series of deletion mutants of IRTA2-Fc fusion protein and mutual competition of the MAbs to the binding to the whole extracellular domain of IRTA2-Fc fusion protein revealed that the panel of MAbs recognizes more than 10 epitopes on IRTA2. There are at least 1, 1, 2, 4, and 2 epitopes on extracellular domain 1 (ED1), ED3, ED1-3, ED4-9, and whole extracellular domain, respectively. Twenty five out of the 36 MAbs are specific to IRTA2 but 11 others showed cross reactivity to other members of the IRTA family. Among 36 MAbs produced, 3 MAbs (F25, F56 and F119) were used for further analysis based on their specific reactivity with recombinant IRTA2 expressed on 293T cells and lack of cross reactivity with other IRTA family members. In a FACS analysis, MAbs F56 and F119 detected IRTA2 expression on 6/7 B-cell non-Hodgkin’s lymphoma and 1/6 Burkitt’s lymphoma cell lines. Reverse transcriptase-polymerase chain reaction experiments and Western blotting using MAb F25 confirmed the expression profile. We also analyzed blood samples of 11 patients with hairy cell leukemia (HCL), 11 patients with chronic lymphocytic leukemia (CLL), and 4 patients with follicular lymphomas (FL) by FACS. We detected IRTA2 expression on 100% HCL (11/11), 55% CLL (6/11) and 0% of FL cells (0/4). Our results provide the first evidence that the IRTA2 is expressed on the surface of HCL and CLL cells as well as some human lymphoma cell lines. We expect IRTA2 will be established as a new target for immunotherapy.

Description: B cells of chronic lymphocytic leukemia (CLL) are long-lived in vivo, possibly because of defects in apoptosis. We investigated BL22, an immunotoxin composed of the Fv portion of an anti-CD22 antibody fused to a 38-kDa Pseudomonas exotoxin-A fragment. B cells from 22 patients with CLL were immunomagnetically enriched (96% purity) and were cultured with BL22 or an immunotoxin that does not recognize hematopoietic cells. The antileukemic activity of BL22 was correlated with CD22 expression, as determined by flow cytometry. BL22 induced caspase-9 and caspase-3 activation, poly(adenosine diphosphate [ADP]-ribose)polymerase (PARP) cleavage, DNA fragmentation, and membrane flipping. Cell death was associated with the loss of mitochondrial membrane potential and the down-regulation of Mcl-1 and X-chromosomal inhibitor of apoptosis protein (XIAP). Furthermore, BL22 induced a proapoptotic 18-kDa Bax protein and conformational changes of Bax. Z-VAD.fmk abrogated apoptosis, confirming that cell death was executed by caspases. Conversely, interleukin-4, a survival factor, inhibited spontaneous death in culture but failed to prevent immunotoxin-induced apoptosis. BL22 cytotoxicity was markedly enhanced when combined with anticancer drugs including vincristine. We also investigated HA22, a newly engineered immunotoxin, in which BL22 residues are mutated to improve target binding. HA22 was more active than BL22. In conclusion, these immunotoxins induce caspase-mediated apoptosis involving mitochondrial damage. Combination with chemotherapy is expected to improve the efficacy of immunotoxin treatment. (Blood. 2004;103:2718-2726)

Description: Abstract 2443 Background: Complete remissions have been observed in some but not all children in an ongoing Phase 1 trial of the anti-CD22 recombinant immunotoxin moxetumomab pasudotox (m. pasudotox, HA22, CAT-8015) for relapsed and refractory ALL and NHL. Objectives: To develop ALL/NHL cell lines resistant to m. pasudotox in attempt to uncover possible mechanisms of immunotoxin resistance and to use this information to try to improve clinical outcomes. Methods: Using the ALL cell line HAL-01 and the Burkitt NHL cell line CA46, we generated two cell lines (HAL-01-R and CA46-R, respectively) that were resistant to killing by m. pasudotox by repeated exposure to sub-lethal doses of m. pasudotox. We studied the basis of their resistance. Results: Cytotoxicity was markedly reduced in the resistant cell lines (Table) and m. pasudotox was unable to ADP-ribosylate and inactivate elongation factor-2 (EF2). In HAL-01-R this was due to a low level of DPH4 mRNA and protein, which prevented diphthamide biosynthesis and rendered EF2 refractory to m. pasudotox. Analysis of the promoter region of the DPH4 gene showed that the CpG island was heavily methylated. Incubation of sensitive cells with the methylation inhibitor 5-azacytidine prevented the emergence of resistant cells. A second resistance mechanism in the diphthamide pathway was identified in the CA46-R line. There was a small decrease in CD22 site density and internalization rate in CA46-R cells, which were insufficient to account for m. pasudotox resistance. In contrast, protein synthesis measured by 3H-leucine incorporation was not reduced in CA46-R cells suggesting a defect in diphthamide modification of EF2. Consistent with this hypothesis was the finding that m. pasudotox catalyzed the incorporation of ADP-ribose into EF2 in extracts of sensitive CA46 cells, but not in CA46-R cells, although EF2 protein levels were similar in the two cell lines. We examined the expression of the six genes required for diphthamide synthesis (DPH1–5 and WDR85). Only the levels of WDR85 mRNA and protein were substantially decreased in CA46-R cells. We used shRNA to knock down WDR85 mRNA and protein levels and tested the cells for sensitivity to m. pasudotox. The IC50 increased 10-fold from 0.6 ng/ml to 6.0 ng/ml. Conversely, we restored WDR85 protein expression into CA46-R cells by transduction of resistant cells with a WDR85 expressing vector, which partially restored sensitivity to m. pasudotox. Altogether, these findings show that a reduction in WDR85 protein is sufficient to cause m. pasudotox resistance. To study why WDR85 expression is reduced in CA46-R cells, a high resolution SNP array analysis of the WDR85 gene was performed. A homozygous deletion at Chr9:140,468,104–140,610,581 (Build: GRCh37/19) was detected, which accounts for the inability of the cells to make WDR85 protein. Conclusion: We isolated two immunotoxin-resistant ALL/NHL cell lines and found that resistance was associated with reduced expression of diphthamide synthesis genes required to correctly post-translationally modify EF2. As a consequence, m. pasudotox was unable to ADP-ribosylate and inactivate EF2. Reduced expression of DPH4 in HAL-01-R was related to hypermethylation and could be prevented by 5-azacytidine. Reduced expression of WDR85 in CA46-R was due to a deletion of the gene. These represent novel mechanisms of immunotoxin resistance that will be of interest to explore in future studies. This study was supported by the Intramural Research Program of the NIH, National Cancer Institute (NCI), Center for Cancer Research and by a Cooperative Research and Development Agreement between MedImmune, LLC and the NCI. Disclosures: No relevant conflicts of interest to declare.

Description: Abstract 3442 Poster Board III-330 BL22 is a 63 kDa anti-CD22 recombinant immunotoxin containing truncated Pseudomonas exotoxin and variable domains from an anti-CD22 antibody. Patients with multiply relapsed/refractory hairy cell leukemia received BL22 and achieved 19 (61%) complete remissions (CRs) and 6 (19%) partial responses (PRs) in phase I testing, and 17 (47%) CRs plus 9 (25%) PRs in phase II testing (n=36), for overall response rates (ORR) of 72-81%. The average dose/cycle was the same for phase I and II (29 vs 33 ug/Kg x3). The dose for phase II was 40 ug/Kg x3 initially and 30 ug/Kg x3 for retreatment, but retreatment was held if patients had hematologic remission (HR, neutrophils ≥ 1500/mm3, Hgb ≥ 11 g/dL, and platelets ≥ 100,000/mm3) after cycle 1. Disease-free survival (DFS, CR duration) for phase II has not yet been reached at a median of 32 (range 4-62) months, with 12 (71%) of 17 CRs still ongoing. Considering all 36 CRs from phase I and II testing, median DFS was 33 (3-112) months with 15 (42%) of 36 CRs ongoing. Patients in CR usually underwent bone marrow biopsy every 6 months for 2 years and yearly thereafter, and after relapsing usually remained in HR. In fact, the median HR duration of these patients has not yet been reached at 42 (range 4-112) months, with 24 (67%) of the 36 patients remaining in HR or CR. Outcomes were better for those with pre-BL22 spleens measuring ≤ 200 mm in height than those with either prior splenectomy or spleens 〉 200 mm, in terms of CR (68% vs 34%, p=0.007), ORR 95% vs 48%, p=0.000003), and DFS (median 69+ vs 27 mo, p=0.002). In contrast, CR rates or DFS was not related to whether patients had 1 (n=8) years of response to their last course of purine analog (p=0.5-0.75). Of 69 patients who received BL22, 8 (12%) had a completely reversible hemolytic uremic syndrome (HUS) and all maintained normal renal function after a median 80 (9-112) months of follow-up. We conclude that BL22 is highly active producing durable remissions in chemoresistant HCL, particularly in patients with limited disease burden. Testing is underway with a high-affinity version of BL22, called HA22 (CAT-8015). Disclosures Kreitman: NIH: Patents & Royalties. Off Label Use: BL22 is a recombinant immunotoxin which targets CD22+ cells. FitzGerald:NIH: Patents & Royalties. Pastan:NIH: Patents & Royalties.

Description: Abstract 2488 Hairy cell leukemia (HCL) is a B-cell malignancy which is thought to originate in most patients after the B-cell contacts antigen. To determine whether certain HLA types are preferentially expressed in HCL, HLA class I and class II allele frequencies at low resolution were collected from 247 HCL patients including 233 Caucasian, 5 Black, 5 Hispanic, 2 Asian and 2 Hawaiian/Pacific Islanders. Out of 494 total alleles from the 247 patients, the most frequent were HLA-A*02 (145, 29%), HLA-B*07 (58, 12%), HLA-C*07 (141, 29%), and HLA-DRB1*11 (76, 15%). In comparison with normal donors, only HLA-DRB1*11 was preferentially expressed in HCL, with a population frequency among the 233 Caucasians of 70 (30%), compared to 17% of a database of USA Caucasians (n=61655, p

Description: Abstract 2611 CD22 is expressed on the surface of B cell hematologic malignancies such as acute lymphoblastic leukemia (ALL). CD22 is a Siglec family lectin present on B cells, starting at the pre-B cell stage of development, but is not expressed on plasma cells. CD22 consists of 7 extracellular Ig domains and is found in 2 isoforms, one of which is missing the second and third N-terminal Ig domains. We generated CAR modified T cells containing anti-CD22 extracellular binding motifs fused to intracellular signaling domains for T cells activation (CD3 zeta) or costimulation (CD28 or 4-1BB). The binding motifs were derived from scFvs that targeted a membrane distal epitope of CD22, Ig domain 3, (BL22 and a higher affinity HA22 motif) or that bound a more membrane proximal, Ig domains 5–7, of CD22 (m971). The CAR constructs we generated were second-generation (CD28 and CD3 zeta; or, 4-1BB and CD3 zeta) or third generation (CD28, 4-1BB and CD3 zeta signaling domains). A CH2CH3 spacer domain from IgG1 was added in some constructs to examine the impact of extending the scFv-derived binding domain away from the transduced T cell membrane. In vitro cellular cytotoxicity and cytokine release experiments with 4 B cell-ALL cell lines (REH, SEM, NALM6, KOPN8) as well as the CD22 (+)ve Daudi and Raji cell lines were performed. Our results demonstrate that addition of the CH2CH3 domain did not improve tumor lysis and that standard affinity BL22 and higher affinity HA22-derived scFv epitopes were equivalent. With regard to signaling domains, second generation constructs were better than third generation constructs both in vitro and in vivo. In comparison between second generation constructs, CD28 containing domains outperformed 4-1BB with regard to lytic activity and cytokine release. Most surprising was the activity of the m971-derived scFv binding epitope. m971-CAR had significantly higher killing activity, a far more robust cytokine release profile, and superior in vivo activity. NSG mice were injected i.v. with 0.5× 106 NALM6-GL cells (pre-B cell ALL line engineered to express luciferase). Three days later, when disease was evident, mice were treated with 1×107 CAR+ T cells, and then followed by bioluminescent imaging to measure disease burden. The m971 CAR was significantly more potent at tumor clearance than our previously developed most active construct expressing the HA22-derived scFv domain (Figure 1). Disease progressed rapidly when non-transduced T cells were used (mock). We are currently examining the activity of different signaling domains on m971 CAR efficacy in vivo and directly comparing the anti-CD22 m971 CAR to the CD19 CAR currently being evaluated in clinical trials. These studies will guide future anti-CD22 CAR-based anti-leukemia immunotherapy trials. Disclosures: No relevant conflicts of interest to declare.