Effect of gemtuzumab ozogamicin on acute myeloid leukaemia blast ...

short report

Effect of gemtuzumab ozogamicin on acute myeloid leukaemia blast cells in vitro, as a single agent and combined with other cytotoxic agents K. L. Morris, J. A. Adams and J. A. Liu Yin University Department of Clinical Haematology, Manchester Royal Infirmary, Manchester, UK

Received 1 June 2006; accepted for publication 17 August 2006 Correspondence: Prof. JA Liu Yin, Clinical

Summary The effect of gemtuzumab ozogamicin (GO) alone, or combined with lowdose cytarabine or etoposide, on the proliferation of acute myeloid leukaemia blast cells in vitro was investigated. GO alone induced a dose-dependent inhibition of proliferation although an increase in apoptosis was only seen in a minority of patients. A correlation was found between PgP function and GO sensitivity but not between CD33 or PgP expression and GO. Combinations of GO with varying concentrations of cytarabine or etoposide were additive in inhibiting proliferation, reducing cell viability and increasing apoptosis.

Haematology, Cobbett House, Manchester Royal infirmary, Oxford Road, Manchester, M13 9WL, UK. E-mail: [email protected]

Keywords: gemtuzumab ozogamicin (Mylotarg), acute myeloid leukaemia, apoptosis, co-incubation, multi drug resistance.

Current treatment strategies induce remission in the majority of acute myeloid leukaemia (AML) patients; however up to 70% subsequently relapse, highlighting the need to develop novel therapies. Gemtuzumab ozogamicin (GO) consists of a humanised anti-CD33 antibody (Hp67.6) conjugated to the cytotoxic antibiotic calicheamicin (CAL). CD33 is expressed on the cell surface in 80–90% of AML patients but not on pluripotent haematopoietic stem cells or non-haematological tissues (Linenberger et al, 2001). On binding to the CD33 antigen, GO is internalised, the CAL portion is released and binds to the minor groove of DNA causing breaks in the structure and, ultimately, cell death (Damle, 2004). GO has been used in the treatment of relapsed CD33 positive AML (Sievers et al, 2001) and has been investigated for use in CD33 positive acute lymphoblastic leukaemia (ALL) (Golay et al, 2005). Combination chemotherapy regimens for induction using GO and high-dose chemotherapy in patients younger than 60 years are currently being evaluated in the UK Medical Research Council Trial AML 15. However, some AML patients, particularly the elderly, may respond poorly to high-dose chemotherapy due to dose-limiting toxicity (Walter et al, 2003). Novel low-dose chemotherapy regimens in older patients are therefore warranted.

Materials and methods Samples, GO and controls Blast cells from 27 AML patients, 26 at presentation and one at relapse, were taken with informed consent. GO (humanised

anti-CD33 antibody Hp67.6 conjugated to CAL), HcT-MO1 (control antibody against the Mucin 1 (MUC-1) antigen conjugated to CAL), the humanised anti-CD33 antibody Hp67.6 (without CAL) and unconjugated CAL were kindly provided by Wyeth, Philadelphia, PA, USA.

Cell culture Blasts were isolated by density gradient centrifugation and cultured in McCoy’s 5A medium supplemented with 15% fetal calf serum (FCS), granulocyte-macrophage colonystimulating factor (GM-CSF, 100 ng/ml), stem cell factor (SCF) and interleukin 3 (IL-3, 10 ng/ml) at 37C in 5% CO2, 5% O2 and 90% N2. In initial experiments, blasts from 11 patients were cultured with GO, HcT-MO1 or Hp67.6 (without CAL) at 1–1000 ng/ml. Unconjugated CAL was added at 0Æ01–10 ng/ml, i.e. approximately the amount of CAL that would be present in the equivalent dose of 1–1000 ng/ml GO, (CAL comprises approximately 1% of the total molecular weight of GO). In further experiments, AML blasts from 27 patients were cultured with either cytarabine or etoposide (0Æ1–100 ng/ml) (David Bull Laboratories, Warwick, UK) alone or in combination with GO (1, 10 and 100 ng/ml). Cell proliferation was assessed by tritiated thymidine (3H-T) uptake at 96 or 120 h incubation. The inhibitory effects of GO or cytarabine or etoposide alone at the relevant doses were summated to obtain an expected effect value for combinations, any significant differences between expected effect values and actual effect values were evaluated by Wilcoxon signed rank statistics.

ª 2006 The Authors Journal Compilation ª 2006 Blackwell Publishing Ltd, British Journal of Haematology, 135, 509–512 doi:10.1111/j.1365-2141.2006.06326.x

Short Report

Apoptosis, CD33 quantification and PgP status Apoptosis was measured at 96 or 120 h incubation using the Annexin V-phycoerythrin (PE) apoptosis detection kit (BD Biosciences, Oxford, UK). The absolute number of CD33 molecules per cell was quantified by flow cytometry (Quantibrite; BD Biosciences). Blast cells were incubated with the antibody MRK-16 to determine the external expression levels of P-glycoprotein (PgP). Multidrug resistance (MDR) status was measured using the quantitative multidrug resistant assay kit MultiDrugQuant TM (Alexis Corporation, Nottingham, UK), This functional assay is designed to quantify the amount of drug resistance relating to MDR1 (PgP) as a percentage of the total drug resistance. The vinblastine-resistant lymphoid cell line, CEM-VLB100 was used as a positive control.

Results and discussion Effect of GO alone on proliferation of AML blasts Gemtuzumab ozogamicin induced a dose-dependent inhibition of proliferation in vitro in all samples studied. The mean inhibition for GO and each control (n ¼ 11 ± SEM) is shown in Fig 1. GO at 1000 ng/ml resulted in a mean % inhibition of proliferation of 79Æ5% ± 7Æ2 (range 42–97Æ1) compared with GO at 1 ng/ml 22Æ7% ± 7Æ7 (range 0–60Æ5). Interestingly the HcT-M01 control conjugate also inhibited proliferation but to a significantly lesser extent than GO (except at 1000 ng/ml, student’s paired t-test (P ¼ 0Æ1), suggesting the presence of the MUC-1 antigen on the surface of AML blasts. MUC-1 is one of a heterogeneous group of glycosylated proteins (Leong et al, 2004) found by Brossart et al (2001) to be expressed in 67% of AML patients. CAL

alone at the equivalent concentration found in GO, also resulted in a dose-dependent inhibition of proliferation, particularly apparent at the equivalent dose to GO at 1000 ng/ml. At this dose the mean % inhibition of proliferation with CAL alone was not significantly different than GO or the HcT-M01 control (student’s paired t-test P ¼ 0Æ1 and 0Æ2 respectively), however at lower doses, inhibition by CAL was significantly lower than that of GO. This suggests that CAL alone (without conjugation to CD33) at high-doses is able to inhibit proliferation of AML blasts in vitro. Similar findings were reported by Zwaan et al (2003). No significant effect on proliferation was observed in response to the Hp67.6 antibody alone.

Apoptosis Van der Velden et al (2004) observed a dose-dependent increase in apoptosis levels after culturing AML blasts with GO for 72 and 96 h. We observed a similar dose-dependent increase in apoptosis levels after 96 or 120 h incubation in 10 of 27 patients studied (data not shown). Absence of apoptosis in response to GO may be due to cell cycle status of cultured AML blasts, cells in G0 phase appear less able to take up GO (Jedema et al, 2004).

CD33 quantification The absolute number of CD33 molecules per cell ranged from 925 to 11 140, with a median expression of 4873. Using 5000 as a cut-off level, no significant difference between low CD33 expression (5000) and proliferation or apoptosis was observed at any dose of GO studied. Jedema et al (2004) reported no correlation between CD33 expression and response rates in vivo in a cohort of 12 AML patients receiving GO. However, an in vitro study carried out by Walter et al (2005) did show a quantitative relationship between CD33 expression and GO-induced cytotoxicity in myeloid cell lines. A CD33-independent mechanism of uptake by endocytosis at higher concentrations of GO was reported by Jedema et al (2004), which may explain why GO and high concentrations of CAL alone are still effective even in patients with low levels of CD33.

PgP status

Fig 1. Inhibition of proliferation in response to gemtuzumab ozogamicin (GO) and controls. Acute myeloid leukaemia blasts were cultured for 96 or 120 h in the presence of GO at various concentrations (Open squares), HcT-M01 (closed squares), Calicheamicin (CAL) alone (open triangles) or Hp67.6 antibody alone (closed triangles). Cell proliferation was then assessed using the 3H-T uptake assay. Results are represented as % Inhibition of proliferation (mean ± SEM n ¼ 11) CAL alone was added at equivalent concentration to the CAL portion in GO i.e. 1% of the GO concentration.

510

Thirteen out of 27 patients (48%) expressed a PgP ratio of MRK-16: control antibody of >1Æ3 and were deemed MRK-16 positive (Pallis & Das Gupta, 2005) of which five were functionally positive for PgP. Two patients were classed as functionally positive for PgP without staining positive for MRK-16 antibody. No significant difference between MRK-16 status and inhibition of blast proliferation by GO was observed. However, there was a significant difference between sensitivity to GO at 10 ng/ml in PgP functionally negative

ª 2006 The Authors Journal Compilation ª 2006 Blackwell Publishing Ltd, British Journal of Haematology, 135, 509–512

Short Report

Inhibition of proliferation (% Compared to cell only control)

(A)

45 40 35 30 25 20 15 10 5 0

P = 0·06 (n/s)

1

1

10

10

Concentration of cytarabine (ng/ml) GO alone at 1ng/ml

Inhibition of proliferation (% Compared to cell only control)

(B)

Cytarabine alone

Observed Cytarabine + GO

40 35

P = 0·1 (n/s)

30 25 20 15 10 5 0 1

1

10

10

Concentration of Etoposide (ng/ml) GO alone at 1ng/ml

Etoposide alone

Observed Etoposide + GO

Fig 2(A, B). % Inhibition of proliferation with gemtuzumab ozogamicin (GO) at 1 ng/ml in combination with low-dose cytarabine or etoposide (1 or 10 ng/ml). Cells were incubated with GO at 1 ng/ml in the presence/absence of cytarabine or etoposide (1 and 10 ng/ml) for 96 or 120 h. Cell proliferation was then assessed using the 3H-T uptake assay. Mean GO alone at 1 ng/ml is represented by lower portion (light grey) and mean cytarabine or etoposide alone (1 or 10 ng/ml) are represented by the upper portion (black), the two are summated to demonstrate the expected effect value when the two were combined. The mean observed effect value for each combination is shown by the dark grey bars. n ¼ 27.

patients vs. PgP functionally positive patients (Student’s unpaired t-test, P ¼ 0Æ04), suggesting that PgP function is an important factor in sensitivity to GO.

Additive effect on proliferation when GO is combined with cytarabine or etoposide Figures 2A and B show the effect of GO alone (at 1 ng/ml) or in combination with cytarabine or etoposide at 1 and 10 ng/ml respectively, on proliferation of blasts from 27 patients. Cytarabine alone (1 or 10 ng/ml) induced a dose-dependent inhibition of proliferation in 25/27 patients studied whereas 26/27 patients were sensitive to etoposide (1 or 10 ng/ml). Combination of GO (1 ng/ml) with cytarabine or etoposide (1–10 ng/ml) resulted in a greater inhibition of proliferation than each drug alone at the relevant dose. The % inhibition of proliferation of each patient with each drug alone was summated to give a mean expected % inhibition for combinations (shown by the grey and black portions). No significant differences between expected and observed values (measured by the Wilcoxon signed ranks statistics) were seen with

combinations of GO at 1 ng/ml with cytarabine or etoposide 1 or 10 ng/ml, indicating that the combination at these doses was additive. Gemtuzumab ozogamicin at 1 ng/ml in combination with cytarabine at 1 or 10 ng/ml resulted in a 33Æ2 and 39Æ6% inhibition of proliferation respectively. These combinations were comparable to the effect of GO alone at 10 ng/ml (40Æ1%, n ¼ 27) and were not significantly different when compared with the student’s paired t-test (P ¼ 0Æ9 and 0Æ2 respectively). Similar effects were observed when GO at 1 ng/ml was combined with etoposide at 1 or 10 ng/ml respectively (26Æ8 and 33Æ4% inhibition). The combination of etoposide at 10 ng/ ml with GO at 1 ng/ml was not significantly different when compared with GO alone at 10 ng/ml (student’s paired t-test, P ¼ 0Æ2) These results suggest that by combining GO with lowdose cytarabine (1 or 10 ng/ml) or etoposide (10 ng/ml), the dose of GO may be reduced 10-fold in order to produce the same effect. Therefore, it is likely that low-dose GO in combination with cytarabine or etoposide also at low-doses may offer an effective chemotherapy regimen for AML patients in vivo, especially useful for elderly patients not fit to receive intensive chemotherapy. The data reported here thus provide a rationale for the combination of Mylotarg (3 mg/m2) with low-dose cytarabine, for the treatment of elderly patients as proposed in the forthcoming UK AML 16 Trial.

Acknowledgements The authors would like to thank Wyeth Pharmaceuticals (St Davids PA, USA) for providing gemtuzumab ozogamicin and controls. J.A.A. is the Robert Whiteson Memorial Trust Fellow.

References Brossart, P., Schneider, A., Dill, P., Schammann, T., Grunebach, F., Wirths, S., Kanz, L., Buhring, H.J. & Brugger, W. (2001) The epithelial tumor antigen MUC1 is expressed in haematological malignancies and is recognised by MUC1-specific cytotoxic T-lymphocytes. Cancer Research, 61, 6846–6850. Damle, N.K. (2004) Tumour-targeted chemotherapy with immunoconjugates of calicheamicin. Expert Opinion on Biological Therapy, 4, 1445–1452. Golay, J., Di Gaetano, N., Amico, D., Cittera, E., Barbui, A.M., Giavazzi, R., Biondi, A., Rambaldi, A. & Introna, M. (2005) Gemtuzumab Ozogamicin (Mylotarg) has therapeutic activity against CD33+ acute lymphoblastic leukaemias in vitro and in vivo. British Journal of Haematology, 128, 310–317. Jedema, I., Barge, R.M., Van Der Velden, V.H., Nijmeijer, B.A., Van Dongen, J.J., Willemze, R. & Falkenburg, J.H. (2004) Internalization and cell cycle-dependent killing of leukemic cells by Gemtuzumab Ozogamicin: rationale for efficacy in CD33-negative malignancies with endocytic capacity. Leukemia, 18, 316–325. Leong, C.F., Raudhawati, O., Cheong, S.K., Sivagengei, K. & Noor Hamidah, H. (2004) Epithelial membrane antigen (EMA) of


511

Short Report MUC1 expression in monocytes and monoblasts. Pathology, 35, 422–427. Linenberger, M.L., Hong, T., Flowers, D., Sievers, E.L., Gooley, T.A., Bennett, M.S., Leopold, L.H., Applebaum, F.R. & Bernstein, I.D. (2001) Multidrug-resistance phenotype and clinical responses to Gemtuzumab Ozogamicin. Blood, 98, 988–994. Pallis, M. & Das Gupta, E. (2005) Flow cytometric measurement of phenotypic P-glycoprotein. Methods in Molecular Medicine, 111, 167–181. Sievers, E.L., Larson, R.A., Stadtmauer, E.A., Estey, E., Lowenberg, B., Dombret, H., Karanus, C., Theoblad, M., Bennet, J., Sherman, M.L., Berger, M.S., Eten, C.B., Loken, M.R., Van Dongen, J.M., Bernstein, I.D. & Appelbaum, F.R. (2001) Efficacy and safety of gemtuzumab ozogamicin in patients with CD33-positive acute myeloid leukaemia in first relapse. Journal of Clinical Oncology, 19, 3244–3254. Van der Velden, V.H.L., Boeckx, N., Jedema, I., Marvelde, J.G., Hoogeveen, P.G., Boogaerts, M. & Van Dongen, J.J.M. (2004) High

512

CD33-antigen loads in peripheral blood limit the efficacy of gemtuzumab ozogamicin (Mylotarg) treatment in acute myeloid leukaemia. Leukemia, 18, 983–988. Walter, R.B., Raden, B.W., Hong, T., Flowers, D.A., Bernstein, I.D. & Linenberger, M.L. (2003) Multidrug resistance protein attenuates gemtuzumab ozogamicin-induced cytotoxicity in acute myeloid leukaemia cells. Blood, 102, 1466–1472. Walter, R.B., Raden, B.W., Kamikura, D.M., Cooper, J.A. & Bernstein, I.D. (2005) Influence of CD33 expression levels and ITIM-dependent internalization on gemtuzumab ozogamicin-induced cytotoxicity. Blood, 105, 1295–1302. Zwaan, C.M., Reinhardt, D., Jurgens, H., Huismans, D.R., Hahlen, k., Smith, O.P., Biondi, A., van Wering, E.R., Feingold, J. & Kaspers, G.J.L. (2003) Gemtuzumab ozogamicin in paediatric CD33-positive acute lymphoblastic leukaemia: first clinical experiences and relation with cellular sensitivity to single agent calicheamicin. Leukemia, 17, 468–477.
