Our Science – Kreitman Website
Robert J. Kreitman, M.D.
The Clinical Immunotherapy Section of the Laboratory of Molecular Biology is focused on targeted therapy using both recombinant immunotoxins for hematologic malignancies, particularly hairy cell leukemia (HCL), and rituximab plus chemotherapy for HCL. The recombinant immunotoxins contain truncated forms of Pseudomonas exotoxin (PE) fused to anti-CD25 or anti-CD22 Fv fragments. The anti-CD22 recombinant immunotoxin HA22, a high-affinity version of BL22, is being developed for the treatment of HCL, chronic lymphocytic leukemia (CLL), non-Hodgkin's lymphoma (NHL), and, in collaboration with Dr. Wayne in NCI's Pediatric Oncology Branch, pediatric acute lymphoblastic leukemia (ALL). The anti-CD25 recombinant immunotoxin LMB-2 is currently undergoing phase II testing in conjunction with chemotherapy in patients with adult T-cell leukemia (ATL), where complete remissions (CRs) have been observed. Recently, we have taken advantage of our large HCL population to perform clinical trials with combinations of chemotherapy and rituximab to answer questions relevant to the optimal therapy of newly diagnosed and multiply relapsed HCL, and also to better understand the behavior of HCL in immunotoxin-treated patients. In the laboratory, we are using clinical samples from patients to answer questions about treatment efficacy and toxicity, to better understand the biology and origin of HCL, and to model in vivo how to combine recombinant immunotoxins with chemotherapy to effectively treat a more aggressive disease like ATL.
Development of anti-CD22 recombinant immunotoxins for CD22+ B-cell malignancies. To effectively target CD22+ drug-resistant B-cell malignancies without chemotherapy, BL22 was originally created using the Fv fragment of the anti-CD22 MAb RFB4 fused to a truncated form of PE called PE38. BL22 in phase I-II testing achieved CR rates of 47-61% in chemoresistant HCL, with overall response rates (ORR) of 72-81%. A completely reversible hemolytic syndrome (HUS) was observed in 8 (12%) of 69 HCL patients. BL22 was less effective in chronic lymphocytic leukemia (CLL), probably due to low CD22 expression. We used hot spot mutagenesis to increase the affinity of BL22. The resulting protein, called HA22, CAT-8015 or moxetumomab pasudotox, contains THW instead of SSY in the antigen binding site of VH. This resulted in 14-fold increased binding affinity for CD22, due to lower off-rate, and up to >100-fold increased cytotoxicity. HA22 had antitumor activity in murine xenograft studies, and an acceptable safety profile in cynomolgus monkeys. Phase I testing was begun to determine safety and efficacy in HCL. Sixteen patients were enrolled at 5-40 ug/Kg every other day for 3 doses (QOD x3) and 26 patients received 50 ug/Kg QOD x3 in an expanded MTD cohort without any dose-limiting toxicity. The ORR in 42 evaluable patients was 88%, with responses observed at all dose levels, including 23 (55%) CRs. All but 1 CR lasted >1 year, with median disease-free survival not yet reached at 26 months. A multicenter phase I trial in CLL and NHL is underway, with 16 patients enrolled. Activity has been seen at doses lower than those used in HCL and accrual continues. In pediatric ALL, a much more aggressive disease, 4 (24%) of 17 achieved CR with a more dose-intense administration of HA22. We are discussing with MedImmune and CTEP plans for a multicenter phase III trial of HA22 in HCL. Additional goals include completing the separate phase I trials of HA22 in CLL/NHL and ALL, the latter in collaboration with Dr. Wayne in NCI's Pediatric Branch, and based on the results consider additional trials including pivotal phase III testing as a single agent or combined with chemotherapy.
Development of anti-CD25 recombinant immunotoxin LMB-2 for CD25+ leukemias. We have continued to enroll HCL patients ineligible to receive HA22 on a phase II trial of LMB-2, and have achieved major responses including an ongoing CR. To investigate immunotoxin combinations with chemotherapy, we decided to target ATL where rapid disease progression and immunogenicity limit efficacy of LMB-2 as a single agent. We developed a mouse model showing that gemcitabine reduced tumor interstitial soluble CD25 (sCD25) and showed synergistic antitumor activity combined with LMB-2. In an ongoing phase II trial, ATL patients receive fludarabine and cyclophosphamide prior to LMB-2, with ORR 50% including 25% CRs in 8 evaluable patients. Our goals are to complete this trial with limited LMB-2 drug remaining, and if activity is promising, propose to have additional LMB-2 made for additional studies or develop a new agent for CD25 targeting.
Development of MAb-chemotherapy combinations for early and relapsed/refractory HCL. For the past >20 years, cladribine alone (or less commonly pentostatin alone) has been the standard 1st and 2nd line treatment of HCL, but appears non-curative in most patients. To determine if rituximab should be added to cladribine, newly diagnosed or once-relapsed HCL patients are randomized to cladribine with either immediate or >6-month delayed rituximab, and minimal residual disease (MRD) at 6 months and other time points measured. So far 60 patients have been enrolled to this 150-patient trial, with the goal to compare both MRD rates at 6 months and long-term blood MRD-free survival. To study pentostatin-rituximab and bendamustine-rituximab combinations in HCL for the first time prospectively, a randomized trial is underway in multiply relapsed HCL with 2-way crossover between the 2 arms. Our clinical goal is to determine the best regimen from the standpoint of toxicity and efficacy, and determine the capacity of either regimen for eradicating HCL MRD.
Laboratory research connected with HA22 and other therapies for HCL. A major goal is to improve the efficacy of CD22 targeting by further engineering of the HA22 structure. In collaboration with Ira Pastan's lab, the toxin was further truncated to little more than the ADP-ribosylating domain III, producing a liposomal protease resistant mutant of HA22 called HA22-LR which we found to have cytotoxicity >10-fold improved for primary CLL cells and equal toward HCL compared to HA22. Identification and mutation of 8 immunogenic epitopes in domain III led to HA22-LR-8M, which is also more cytotoxic than HA22 toward CLL. This molecule could improve the CR rate of HA22 toward HCL by preventing immunogenicity, and also improve activity toward CLL/NHL. A 2nd goal is to sequence immunoglobulin rearrangements unique to each HCL patient, to study HCL biology and to design patient-specific PCR assays for MRD. We have previously found this test (RQ-PCR) able to detect 1 HCL in 106 normal cells. We now have 6 HCL patients treated with BL22 or HA22 who are RQ-PCR-negative up to 8 years after achieving CR, and additional patients negative after receiving rituximab combined with chemotherapy. We plan to test additional patients to determine correlations between MRD and clinical outcome. We are also using RNA/DNA from these samples to better characterize B- and T- cell repertoire in patients treated with immunotoxins compared to chemotherapy and/or rituximab, and to study fundamental questions related to HCL biology, including those related to the V600E mutation recently discovered in HCL.
This page was last updated on 10/29/2013.