Search

Rituximab has greatly improved the prognosis of B cell malignancies (BCM). However, several resistance mechanisms have been reported, including the inhibition of apoptosis, complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC) as well as modulation of CD20 antigen and increased complement-related cytotoxicity inhibition factor (CD55). The calicheamicin-conjugated anti-CD22 monoclonal antibody (mAb), CMC544, is one of the new promising agents, and effective on several types of BCM in vitro, especially when used in combination with rituximab. However, details regarding the cytotoxic effects of CMC544 and mechanisms triggering cell death, particularly when it is combined with rituximab, have not been elucidated. We studied cell cycle features of CMC544 on BCM used alone or in combination with rituximab, and analyzed quantitative alteration of target molecules such as CD20, CD22 and CD55.(Materials and Methods) The cell lines used in this study were CD22-positive Daudi and Raji cells, and CD22-negative Jurkat and NB4 cells. Cells obtained from 8 patients with BCM were also used. CMC544, unconjugated anti-CD22 mAb (G5/44), and free NAC-calicheamicin DMH were provided by Wyeth Pharmaceutical Co., Ltd. The effects of CMC544 were analyzed by incubating the cells at the concentrations of 0–100μg/ml for 0–96h, then by measuring cell count, cell viability, 3H-thymidine incorporation and cell cycle distribution on flow cytometry (FCM) as well as by video microscopic observation. The amounts of CD20, CD22 and CD55 antigens were analyzed by FCM, laser scanning microscope. The process of CMC544-induced cell death was assessed by Apocyto®. Three possible mechanisms of the combined effect of CMC544 and rituximab were investigated separately by direct effect of the mAb, CDC and ADCC. (Results) Analysis of the cell cycle distribution indicated that CD22-positive cells were temporally arrested at the G2/M phase after 6–12h incubation with CMC544, and the hypodiploid proportion increased after 24–72h. CD22-positive cells enlarged after 12–24h incubation with CMC544. Apoptotic morphological changes such as bleb formation and cell shrinkage were observed in 35–57% of the cells. These changes were not observed after incubation with G5/44. Apocyto® stain showed a continuous pattern from early to late apoptosis. Similar effects were observed in clinical samples that expressed CD22. The amounts of CD22 and CD55 were significantly reduced 3h after incubation with CMC544, while the CD20 expression was maintained 24h after incubation with CMC544. Incubation of the cells for 12–24h with CMC544 and then for 12hr with rituximab induced 1.2–2.1 times more anti-proliferation, apoptosis, CDC and ADCC than simultaneous incubation.(Conclusion) The effect of CMC544 was clearly observed as G2/M arrest following increase of the hypodiploid proportion in the cell cycle distribution, and this was a useful method to investigate the effect of CMC544. Combination of CMC544 and rituximab enhanced the cytotoxic effect, and sequential administration was more effective. The reduction of CD55 and continued expression of CD20 after the incubation with CMC544 supported the above observation. Further investigations are required to find the most effective treatment regimen to overcome drug resistance of B cell malignancies.

Background: The conserved Aurora family kinases, a family of mitotic serine/threonine kinases, have three members (Aurora-A, -B and -C) in mammalian cells. The Aurora kinases are involved in the regulation of cell cycle progression, and alterations in their expression have been shown to associate with cell malignant transformation. Aurora A localizes to the centrosomes during anaphase, and it is required for mitotic entry. Aurora B regulates the formation of a stable bipolar spindle-kinetochore attachment in mitosis. The function of Aurora-C in mammalian cells has not been studied extensively. In this study, we investigated that human leukemia cells expressed all 3 Aurora kinases at both protein and mRNA level, and the mechanisms of cell cycle regulation by knock down of Aurora C in leukemia cells. Methods: In this study, we used the 7 human leukemia cell lines, K562, NB4, HL60, U937, CEM, MOLT4, SUP-B15 cells. The expression levels of mRNA and proteins of Aurora kinases were evaluated by RT-PCR and western blot. The analysis of proliferation and cell cycle were performed by MTT assay and FCM, respectively. Results: The mRNA of Aurora-A and Aurora-B are highly expressed in human leukemia cell lines (K562, NB4, HL60, U937, CEM, MOLT4, SUP-B15 cells), while the mRNA of Aurora C is not only expressed highly in all cells. In contrast, an increase in the protein level of the 3 kinases was found in all cell lines. These observations suggested posttranscriptional mechanisms, which modulate the expression of Aurora C. In cell cycle analysis by flow cytometory, the knock down of Aurora C by siRNA induced G0/G1 arrest and apoptosis in leukemia cells, and increased the protein levels of p27Kip1 and decreased Skp2 by western blot. In MTT assay, it was revealed that the growth inhibition of leukemia cells transfected with siRNA Aurora C compared with leukemia cells untransfected with siRNA Aurora C. Moreover, We showed that Aurora C was associated with Survivin and directly bound to Survivin by immunoprecipitation and western blot. Conclusion: We found that human leukemia cells expressed all 3 members of the Aurora kinase family. These results suggest that the Aurora kinases may play a relevant role in leukemia cells. Among these Aurora kinases, Aurora C interacted with Survivin and prevented apoptosis of leukemia cells, and induced cell cycle progression. Our results showed that Aurora-C may serve as a key regulator in cell division and survival. These results suggest that the Aurora C kinase may play an important role in leukemia cells, and may represent a target for leukemia therapy.

The clinical result of has been significantly improved by arsenic trioxide (ATO) in acute promyelocytic leukemia (APL). ATO induces apoptosis of various cells and also bone marrow damage in relapsed or refractory patients who had frequently received anti-leukemia agents. Therefore, neutropenia persists over a longer period of time after ATO treatment than after retinoic acid treatment. The neutropenia might be associated with a high incidence of infection. On the other hand, salvarsan, an arsenic-containing compound, had been used for the treatment of syphilis and trypanosomiasis. However, ATO has not been analyzed with regard to its potential as an anti-microbial agent and anti-leukemic agent. We analyzed the in vitro anti-microbial effects of ATO, and studied the incidence and severity of infection among APL patients who were treated with ATO. (Methods) ATO discs were specially prepared for the disc diffusion method. Bacteria isolated from clinical samples were suspended in McFarland 0.5, and uniformly applied onto appropriate culture media with an ATO disc. After incubation in humidified air with 5%CO2 at 37°C for 24 hours, the bacterial growth-inhibitory zone was measured. In other experiments, bacteria and APL NB4 cells were co-cultured in 5%FCS-RPMI medium containing ATO (10−6–10−7M). The number of bacteria was counted after 24-hour incubation. (Patients) Thirty-eight patients with relapsed/refractory APL were treated with ATO (0.15mg/kg/day): 30 patients were treated with ATO alone, and 8 were treated with a combination of ATO and chemotherapy. Thirty-four cases achieved complete remission. The body temperature, duration of febrile neutropenia, neutrophil count, immunoglobulin level, focus of infection, and pharmacokinetics of ATO were studied. The incidence and kinds of viral infection were also analyzed. (Results) ATO significantly inhibited in vitro growth of 17 different bacteria isolated from clinical samples. The growth of E. coli, S. aureus (MSSA), H. influenzae, and St. pneumoniae among others was inhibited by the concentration of ATO that is clinically used. Fever with temperatures above 38°C was observed in 16 patients and the fever persisted for 4.4±2.2 (mean±SD) days. The duration of neutropenia of (Conclusion) ATO inhibit bacterial growth in vitro, and did not increase clinical bacterial infection while the duration of neutropenia in ATO-treated patients was relatively long. On the other hand, ATO increased the incidence of herpes zoster infection, especially among patients who concomitantly received ATO and chemotherapy.

[Background] Homeobox (Hox) genes are grouped together in 4 clusters, A to D. Recent studies has shown that the Hox proteins are important in the control of differentiation and proliferation in hematopoietic cells. It is reported that HoxA10 has been expressed in all types acute myeloid leukemias (except for APL) but not in acute lymphoid leukemias, implicating an important role as a regulator of myeloid progenitors. Moreover, overexpression of HoxA10 induses the proliferation of early progenitor cells and leads to the development of myeloid leukemias. However, we found that Bcr-Abl inhibitors increased the expression of HoxA10 gene in CML cells. In this study, we investigated the mechanism controlling HoxA10 in CML cells treated with Bcr-Abl inhibitors. [Methods] The cells used in this study were Ph1 positive CML cells, K562 and Meg01, and U937 cells for a Ph1 negative cell. We used AMN107 and BMS354825 for Bcr-Abl inhibitors, LY294002 for a PI3K inhibitor, PP2 for a Src kinase inhibitor, and SB203580 for a p38 MAP kinase inhibitor. For analysis of HoxA10 mRNA and protein, RT-PCR and western blot were performed in K562, Meg-01 and U937 cells, which untreated or treated with AMN107, BMS354825, LY294002, PP2, or SB203580 respectively. We then attempted to localize the intracellular locations of HoxA10 in K562 and Meg01, which untreated or treated with AMN107, BMS354825, or LY294002 by using conforcal fluorescence microscopy. Finally, for analysis of proliferation in K562, Meg-01 and U937 transfected with siRNA HoxA10, MTT assays were performed in untransfected or transfected K562, Meg-01 and U937 treated with or without AMN107, BMS354825, orLY294002. [Results] Both K562 and Meg01 cells expressed HoxA10 mRNA and protein at lower level compared to U937 cells. Interestingly, treatment with AMN107, BMS354825, or LY294002 increased the expression of HoxA10 mRNA and protein in both K562 and Meg01 cells. In contrast, treatment with PP2 or SB203580 did not affect the expression of HoxA10 mRNA and protein in both K562 and Meg01 cells. The fluorescence of HoxA10 was more strongly observed in the area corresponding to the cell’s cytoplasm than nucleus, and the treatment with AMN107, BMS354825, or LY294002 increased the fluorescence in nucleus of K562 and Meg01 cells in a time-dependent manner. In K562 and Meg01 cells transfected with the siRNA HoxA10, treatment with AMN107 or BMS354825 slightly inhibited the proliferation compared to K562 and Meg01 transfected with control siRNA. [Conclusions] In this study, we showed that Bcr-Abl inhibitors induced the expression of HoxA10, and HoxA10 was regulated by PI3K/P[Kappa[B pathway in CML cells. This finding indicates a new insight in the regulation of cell proliferation via the PI3K/PKB signal pathway in CML cells.

[Background] FoxM1, a member of the Fox transcription factor family, plays an important cell cycle regulator of both the transition from G1 to S phase and progression to mitosis. FoxM1 expression was also found to be up-regulated in some solid tumors (basal cell carcinomas, hepatocellular carcinoma, and primary breast cancer). These results suggested that FoxM1 plays a role in the oncogenesis of malignancies. However, it is unknown whether FoxM1 expression contributes to the development or progression of leukemia cells. Therefore, we investigated whether and how FoxM1 regulated the cell cycle of leukemia cells. [Methods] The cells used in this study were human leukemia cell lines, K562, HL60, U937 cells. For analysis of FoxM1 mRNA, RT-PCR was performed in all cell lines. For analysis of proliferation and mitotic regulatory proteins (p27, p21, Skp2, Cdc25B, Cyclin D1, Survivin, and Aurora kinase B) in leukemia cells, MTT assays and western blot were performed in all cell lines untransfected or transfected with siRNA FoxM1, respectively. For cell cycle analysis, flow cytometory analysis was performed in leukemia cells untransfected or transfected with siRNAFoxM1 by PI staining. [Results] In all leukemia cell lines, the expression of FoxM1B mRNA were significantly higher than normal MNCs. In K562, HL60, and U937 cells transfected with the siRNA FoxM1, suppression of FoxM1 caused a mean 71% (range 62 to 80%) reduction in S phase cells and a mean 4.4-fold (range 3.2 to 5.6-fold) increase in G2/M phase cells compared to untransfected cells. MTT assay demonstrated that the proliferation of the siRNA FoxM1 transfected cells was inhibited compared to the untransfected cells at 2, 3, 4, or 5 days after siRNA FoxM1 transfection. FoxM1 has been reported to regulate transcription of essential mitotic regulatory genes. We showed that FoxM1 knockdown by siRNA in leukemia cells reduced protein and mRNA expression of Aurora kinase B, Survivin, Cyclin D1, Skp2 and Cdc25B, while increased protein expression of p21and p27 in RT-PCR and western blot analysis. [Conclusions] In this study, we report in the first time that FoxM1 is overexpressed in myeloid leukemia cells. These results demonstrated that expression of FoxM1 is an essential transcription factor for growth of leukemia cells, and regulate expression of the mitotic regulators, Cdc25B, Cyclin D1, Survivin, Aurora kinase B, and p21. Moreover, we showed that FoxM1 regulated the expression of Skp2 protein, which is known to promote degradation of the cell cycle regulator p27. Our study found that inhibition of FoxM1 expression in leukemia cells suppressed their growth in vitro. Therefore, FoxM1 might be a new potential target of therapy for leukemias. We will have further study whether the level of FoxM1 expression in leukemia cells is correlated with patient survival or sensitivity for chemotherapy.

[Background] The proto-oncogene Akt/PKB is activated in many human cancers including leukemias, and regulates the cell survival and apoptosis. Akt/PKB is activated by growth factors or the phosphatidylinositol-3,4,5-triphosphate via PI3K, phosphorylates many substrates for cell survival or proliferation. The tumor suppressor PTEN prevents Akt/PKB phosphorylation and activation. On the other hand, the PH domain leucine-rich repeat protein phosphatase (PHLPP) dephosphorylates Akt/PKB, induces apoptosis, and suppresses tumor growth. It has been reported that the expression levels of PHLPP are significantly reduced in several colon cancer and glioblastoma cell lines, and Akt/PKB phosphorylation is elevated. Therefore, in these cell lines, the reduction of PHLPP causes to promote tumor growth. In this study, we have investigated the expression and function of PHLPP in leukemia cells. [Methods] The cells used in this study were human leukemia cell lines, K562, HL60, U937, CEM, and MOLT4 cells. Primary acute lymphoblastic leukemia (6ALL (6 L2)), myelodysplastic syndromes (8 MDS (8 RAEB)), and acute myeloblastic (12 AML (3 M1, 4 M2, 2 M3, 2 M4, 1 M5)) cells were obtained from the peripheral blood. Human normal mononuclear cells (MNCs) were isolated from peripheral blood (PB) of healthy volunteers after obtaining informed consents. For analysis of PHLPP mRNA, RQ-PCR was performed in all cell lines and clinical samples before and after chemotherapy. For analysis of proliferation in all cell lines, MTT assays and western blot were performed in all cell lines untransfected or transfected with siRNA pHLPP. [Results] In all cell lines and clinical samples obtained from patients with leukemia, the expression of PHLPP mRNA were significantly higher than normal MNCs. In K562, HL60, U937, CEM and MOLT4 cells transfected with the siRNA PHLPP, it is shown that the proliferation of these cells was stimulated and the phosphorylation state of Akt was markedly increased compared to the untransfected cells. The mean PHLPP gene expression was for AML: 3.57, MDS: 2.18, and ALL 3.14 relative to normal MNCs. Mean PHLPP expression was significantly higher in AML, MDS and ALL compared to normal MNCs (P < 0.01). Moreover, when CR status after chemotherapy, mean PHLPP expression levels were significantly reduced compared to unCR status after chemotherapy in MNCs from PB of patients with leukemia. [Conclusions] In this study, we showed in the first time that the PHLPP expression levels were high compared to normal MNCs in leukemia cells. Quantitive analysis of PHLPP gene expression in PB appeared to be a useful tool in discriminating abnormal levels in AML, MDS and ALL patients. Moreover, it is serves as an ideal target for MRD monitoring in PB. However, it is unclear why PHLPP, which is reported that promotes apoptosis and suppresses tumor growth, is at high expression level in leukemia cells.

Purpose: The protein p27 is an important regulator of cell cycle. An increase in p27 causes proliferation of cells, while a decrease in p27 induces quiescence of cells. p27 is regulated by transcriptional, translational and proteolytic mechanisms. Among them, an importnant mechanism in the regulation of p27 is proteolysis. Kinase interacting with stathmin, KIS, is a serine / threonine kinase, and regulates cell cycle progression through the phosphorylation of p27 on serine 10. The S10 phosphorylation on p27 plays an important role in p27 degradation. KIS that phosphorylates p27 on S10 and its role in the regulation of cell cycle progression have not been defined in leukemia cells. In this study, we investigated the role of KIS in leukemia cells. Materials and Methods: We examined the biological significance of KIS expression in K562, NB4, U937, CEM, MOLT4, and SUP-B15 leukemia cells and relationship with the G1 regulaters, such as p27. Moreover, we generated the lentivirus vector inserted with the dsDNA of KIS small interfering RNA (siRNA), and effects of down-regulation of KIS by siRNA transfection were investigated in leukemia cells. Results: RT-PCR and western blot analysis showed high KIS expression in all leukemia cells. The p27 phosphorylation on S10 was significantly lower in the leukemia cells transduced with KIS siRNA and depletion of KIS enhanced growth arrest. The down-regulation of KIS induced G1 arrest in cell cycle, but not apoptosis. In cell cycle analysis, control leukemia cells showed 42.3 ± 1.8 %, but leukemia cells transfected with KIS siRNA showed 67,1 ± 2.1 % in G1 fraction. Moreover, these cells significantly showed small population in S ansd G2 fractions compared compared with controls. Conclusion: These findings suggest KIS expression promotes cell cycle progression in leukemia cells. Depletion of KIS using siRNA in leukemia cells induced cell cycle arrest in G1 phase compared with control cells. In this study, we showed that KIS might be the target for the molecular therapy.

Cyclooxygenase-2 (COX-2) is reported to be an important cellular target for therapy in malignancies. The growth inhibitory effects of COX-2 inhibitors on malignancies have been demonstrated to be through not only COX-2 dependent, but also independent mechanisms. In this study, we showed that etodolac, COX-2 inhibitor, induced apoptosis via COX-2 independent pathway, and investigated the molecular details of etodolac-induced apoptosis in Burkitt’s lymphoma cells. In Daudi and Raji Burkitt’s lymphoma cell lines, which expressed no COX-2 enzyme, etodolac more strongly induced apoptosis compared to meloxicam. Moreover, Etodolac did not induce apoptosis to normal B lymphocytes. For the pathway of etodolac-induced apoptosis, reduction of anti-apoptotic bcl-2 mRNA and Bcl-2 protein, activation of Caspase-9 and -3, down-regulation of caspase inhibitors, c-IAP-1 and Survivin were involved. Moreover, EBER-1 and -2 expression in EBV positive Daudi and Raji cells were reduced to result in down-regulation of Bcl-2 by treatment with etodolac. It has been reported that etodolac has stereoisomers, R- and S- etodolac. We found that racemate of etodolac more strongly induced apoptosis in Daudi and Raji cells compared to R- or S- etodolac. In conclusion, our findings indicated etodolac inhibited EBERs expression and induced apoptosis via a Bcl-2-regulated pathway. Moreover, racemate of etodolac more effectively induced apoptosis than R- and/or S- etodolac. Therefore, these activities of etodolac potentially extend to the treatment of patients with Burkitt’s lymphoma resistant to chemotherapy.

Background: The therapy with ATO induced high complete remission and maintained for long survival for patients with relapsed or refractory APL. While the clinical effect of ATO against APL was confirmed, its pharmacokinetics has yet to be clarified. In most reports on pharmacokinetics of ATO, the arsenic concentrations were measured as total arsenic. We investigated the pharmacokinetics of arsenic species in Japanese patients with relapsed or refractory APL treated with ATO. Patients and Methods: In the prospective study, from 12 patients with APL treated with ATO, the blood and urine for the pharmacokinetic data were collected and subsequently stored frozen until analysis. ATO (0.15 mg/kg) was intravenously administered once daily over 2 hours to until bone marrow remission to a maximum of 60 days. The plasma and urine were collected on day 1 and after 1, 2 and 4 weeks. Inorganic arsenic (AsIII and AsV) and the major metabolites monomethylarsonic acid (MAAV) and dimethylarsinic acid (DMAAV) in plasma and urine were quantified by HPLC/ICP-MS. Results: Ten of 12 patients (83%) achieved complete remission (CR). Six of ten (60%) who achieved CR were negative in the post-treatment RT-PCR test. For two patients the blood and urine were collected also during consolidation. The plasma concentrations of inorganic arsenic on day 1 reached the Cmax (mean 22.6±11.4 ng/mL) immediately after completion of administration followed by a biphasic elimination while the appearance of methylated metabolites in the blood was delayed. During the repeated administration, the plasma concentrations of inorganic arsenic reached the steady-state. The Cmax of inorganic arsenic on week 4 was similar (mean 23.2±10.2 ng/mL) but the elimination was delayed. As a result, the AUC increased about 2-fold (from mean 211.8±55.1 to 474.8±192.6 ng/mL), and the clearance declined (from 0.7±0.2 to 0.4±0.1 L/kg/h) but no marked change was observed in volume of distribution. In contrast, the MAAV and DMAAV concentrations increased in relation to increased administration frequency fold (from mean 3.1±1.6, 5.4±2.9 to 10.9±4.7, 21.4±12.3 ng/mL). The plasma concentrations of arsenobetaine, an organic arsenic compound derived from seafood, remained almost constant (about 2 mg/mL) during the study period. The urinary excretion rates of AsIII and AsV remained almost constant after week 1, suggesting that the steady-state was attained. In contrast, a tendency to increase with administration frequency was observed in the excretion rates of MAAV and DMAAV after week 4 (from mean 17.4±11.2, 19.4±8.5 to 19.6±10.0, 21.1±9.5 %). The total arsenic excretion rate remained at ~60% of dose after week 1. Conclusion: ATO is metabolized when administered intravenously to APL patients and methylated metabolites were promptly eliminated from the blood and excreted into urine after completion of administration, indicating no measurable accumulation of ATO in the blood.

Cyclooxygenase-2 (COX-2) is reported to regulate apoptosis and to be an important cellular target for therapy. In this study, we demonstrated that etodolac, a COX-2 inhibitor, inhibited proliferation and induced apoptosis in myeloma cell lines (RPMI 8226 and MC/CAR cells), expressing the COX-2 enzyme. In both cell lines, etodolac more strongly induced apoptosis compared with thalidomide or meloxicam. Etodolac induced down-regulation of bcl-2 protein and mRNA, activation of caspase-9, -7 and -3, down-regulation of caspase inhibitors, cIAP-1 and survivin, and loss of mitochondrial membrane potential in a dose-dependent manner. In addition, our data demonstrated that when myeloma cells were coincubated with 50 mM etodolac on bone marrow stromal cells (BMSC), myeloma cell adhesion to BMSC was significantly inhibited compared with thalidomide or meloxicam coincubation, and the adhesion molecules VLA-4, LFA-1 (CD11a), CXCX4, and CD44 were suppressed on myeloma cells treated with etodolac. Moreover, we found that 100 mM R-etodolac, S-etodolac, and the combination of R- and S-etodolac, which are the stereoisomers of etodolac, slightly inhibited the proliferation of myeloma cells, while 50 to 100 mM etodolac significantly inhibited the proliferation of myeloma cells. In conclusion, our findings indicate that etodolac induced apoptosis via a bcl-2 dependent pathway, suppressed the expression of adhesion molecules, and inhibited myeloma cell adhesion to BMSC compared with thalidomide or meloxicam. Thus, the activities of etodolac potentially extend to the treatment of patients with myeloma resistant to standard chemotherapy, including thalidomide.

Several studies including ours have suggested that lack of CD56 expression in multiple myeloma (MM) defines a unique patient subset with poorer prognosis. However, the mechanism underlying this aggressive behavior of CD56− MM has not been well elucidated. In this study, we sorted out both CD56− and CD56+ fractions from MM cell lines or patients with MM, and investigated their different responsiveness to interleukin-6 (IL-6) or insulin like growth factor-I (IGF-I), and tried to clarify the course of action in cell cycle distribution. After stained with PE-CD56, CD56− and CD56+ fractions in KMS-21-BM and U-266 cell lines were isolated by the cell sorter, and cultured separately either in the presence or absence of IL-6 (2 ng/ml and 10 ng/ml, respectively). Although CD56− cells in both KMS-21-BM and U-266 cell lines responded significantly to IL-6 (P=0.001 and 0.009, respectively), CD56+ cells did not. Ki-67+ cells in CD56− KMS-21-BM cells, that were significantly fewer than that in CD56+ ones (P=0.0003), increased significantly upon 24-hour incubation with IL-6 (P

Recently, arsenic trioxide has increasingly been used for relapsed acute promyelocytic leukemia. However, it is known to have several adverse effects, including acute cardiac toxicities.To determine cardiac toxicities resulting from arsenic trioxide therapy in patients with relapsed or refractory acute promyelocytic leukemia.Phase II clinical prospective cohort study.A university hospital in Hamamatsu, Japan.8 patients with relapsed acute promyelocytic leukemia.Arsenic trioxide, 0.15 mg/kg of body weight, administered daily by 2-hour infusion for a maximum of 60 days.Continuous monitoring with ambulatory electrocardiography.Five patients (63%) achieved complete remission. During induction therapy with arsenic trioxide, prolonged QT intervals were observed in all patients. Ventricular premature contractions were noticed during 8 of 12 courses of therapy. Four patients developed nonsustained ventricular tachycardia and required treatment with antiarrhythmic agents.Cardiac toxicity occurs during arsenic trioxide therapy in patients with acute promyelocytic leukemia. Such patients should be monitored for prolonged QT intervals and ventricular arrhythmia.