Assays for the Analysis of P-Glycoprotein in Acute Myeloid Leukemia and CD34 Subsets of AML Blasts P Sonneveld and E Wiemer

Home , Efflux (microbiology)

Assays for the analysis of P-glycoprotein in acute myeloid leukemia and CD34 subsets of AML blasts P Sonneveld and E Wiemer

Department of Hematology, University Hospital Rotterdam, The Netherlands

Expression of the multidrug resistance (MDR) phenotype is an more frequently expressed by relapsing AML, indicating its independent prognostic variable in acute myeloid leukemia. role in refractory disease. Approximately 43–57% of the patients have P-glycoprotein (P- gp) expression. A major drawback with the interpretation of P- For this reason, many investigators have studied P-gp gp data in AML is the lack of coherence with different analytical expression in AML patients with the purpose of identifying its assays. We have focused our efforts of P-gp detection on flow exact role as a prognostic factor in AML. The outcome of these cytometry using a dual technique of P-gp staining with anti- studies is not yet settled, for a large part because of the many bodies for the extracellular epitope (MRK16) and a functional inconsistencies between these studies. Major discrepancies analysis of P-gp using the rhodamine efflux assay and the that still need to be uniformly defined include (1) the varia- effect of P-gp inhibitors such as SDZ PSC 833. This technique was combined with the staining of lineage-specific antigens bility of the P-gp assays that are used; (2) the criteria for posi- such as CD34, CD56 and c-kit. In this way, various subsets of tivity; (3) the identification of subpopulations of cells that have AML cells can be identified such as MRK 16؉/؊, CD34؉/؊ blasts. different P-gp expression; and (4) functional tests as opposed These cells can be sorted for further analysis, such as the mol- to detection assays. ecular expression of P-gp and other pleiotropic drug resist- We have performed prospective studies of P-gp expression ance genes. in AML patients that demonstrated the high variability Keywords: MDR; resistance; acute myeloid leukemia; cyclosporin; SDZ PSC 833 between patients and we have attempted to identify thresholds for positivity in relation to other prognostic parameters.31,32 However, in order to accomplish a reliable and reproducible assay for P-gp in disseminated hematologic malignancies such Introduction as AML and multiple myeloma, these measurements typically should be performed in well defined subpopulations of tumor Acute myeloid leukemia (AML) is usually treated with sequen- cells which can be selected by lineage-specific or tumor-spe- tial cycles of chemotherapy, followed by intensification and/or cific antigens. We have performed several studies in which consolidation therapy. In younger patients the probability of we compared different methods for the analysis of P-gp in long-term survival, associated with intensive chemotherapy multiple myeloma and AML.19,26,31–35 Part of these results including high-dose cytarabine and/or allogeneic or autolog- were previously published. A summary of these results will be ous stem cell transplantation is approximately 50% and may presented here. approach 80% in selected groups such as those with the t(8;21) and t(15;17) karyotypes.1–9 Approximately 25% of younger patients and 40–50% of elderly patients with AML RNA-based assays are primary refractory to conventional chemotherapy. The majority of these patients have a very low probability of sur- In our early studies of mdr1, we used the RNAse protection vival. In addition, many initially responsive patients relapse assay to determine the levels of mdr1RNA in bone marrow within 1–2 years.1,4,10–12 and blood samples of patients with acute and chronic leukem- The cause of primary or acquired refractoriness to chemo- ias and multiple myeloma. The technical details of this assay therapy in AML is only partly understood. and its modifications have been described in several publi- Among many different mechanisms of drug resistance, plei- cations.19,26,31–35 The RNAse protection assay offers the otropic or multidrug resistance (MDR) has drawn specific advantage of specificity for mdr1 and the possibility to detect attention because of its possible significance for simultaneous different splicing products. It also gives information on the resistance to various cytostatic agents. MDR is encoded by the expression of not only mdr1, but also mdr3, which is fre- mdr1 gene which is situated at chromosome 7q21. P-glyco- quently expressed in non-Hodgkin’s lymphomas (NHL) and protein (P-gp) is a plasma membrane which acts as an ATP- chronic lymphocytic leukemia (CLL).36 There are however dependent drug efflux pump in cells exposed to cytotoxic several serious disadvantages of the RNAse protection assay. agents such as daunorubicin (DNR), vinca alkaloids The problem which really limits the application of this assay (vincristine, vinblastine), and taxanes. High expression of is the large number of cells needed for RNA isolation and to mdr1 in cell lines can be transfered to sensitive cell lines by perform the assay with appropriate controls. Generally gene transfection or induced by exposure of sensitive cells to 10 × 106 nucleated cells are required for a complete analysis. increasing concentrations of one of these drugs.13,14 Because of the large amount of RNA that can be obtained, MDR expression is frequently observed in untreated AML the sensitivity of the assay is generally good. Even low cells, and may therefore be an important prognostic pheno- expression levels of mdr1RNA such as present in the RPMI type in poorly responding patients.15–30 In addition, MDR is 8226 DOX1 cell line can be detected and possible false-positive and false-negative results can be partly eliminated by making titrations. All this is however based on the assumption Correspondence: P Sonneveld, University Hospital Rotterdam, Room that a homogeneous cell population is used, which is not con- L 407, Dr Molewaterplein 40, 3015 GD Rotterdam, The Netherlands taminated with normal white blood cells or other cells which Received 24 October 1996; accepted 2 February 1997 may have different expression levels of mdr1, such as for Analysis of P-glycoprotein in AML P Sonneveld and E Wiemer 1161 example CD56+ natural killer (NK) cells or T cells. Such a pure Some laboratories have tried to solve this problem by introd- and uniform sample of tumor cells can not be obtained in ucing a ‘staining intensity scale’, in order to separate samples most patients, and this makes the RNAse protection assay an with modest or high intensity.40 The evaluation is however unattractive assay for large scale analysis of mdr1. Another subject to personal interpretation and is always restricted to disadvantage with the use of RNAse protection or Northern several hundreds of cells. Also, the fixation technique is of blot assays is the fact that it will require at least several days great influence on the staining intensity, with formaldehyde or even weeks to obtain a result. The assays are laborious and resulting in dull staining and acetone generally producing frequently not accessable for clinicians. more bright staining. For this reason, slides from different lab- Many of these objections can also be used against the more oratories cannot easily be compared except if identical fix- recently developed RT-PCR assay, which also analyses RNA ation and staining procedures are used. Still, immunocytoch- expression. However, this assay has been developed and emistry is an important method of P-gp analysis, in particular adapted to be used with relatively low numbers of cells. In for diagnostic procedures and for staining of tissue slides and our hands approximately 105 AML cells are needed for a bone marrow biopsies, provided that the procedure will be reliable assay. This requirement offers the possibility to per- standardized according to internationally accepted guidelines. form the assay on preselected cells, for example cells which This is certainly a prerequisite for large clinical trials where have been sorted on the basis of specific membrane antigens. immunocytochemistry is the only method which meets wide In those tumors, in which lineage- and tumor-specific antigens acceptance and applicability. Only then can the value of this are expressed, cells can be sorted and purified. In our group assay be fully appreciated. Immunocytochemistry can be this has been done with CD38+ multiple myeloma cells and combined with counterstaining cytochemistry such as May– CD56+ NK cells. Such an approach could also be taken to Gru¨nwald–Giemsa or acid phosphatase, in order to further select and to purify to some extent AML cells by using the define the morphology of positive cells. CD34 antigen, or in a similar approach to select for B lineage differentiation markers in acute lymphoblastic leukemia, such as CD19 or CD20. Flow cytometry-based assays of P-glycoprotein These assays are all based on expression levels of mdr1RNA. The results are not predictive for the actual Flow cytometry has provided the clinical hematologist with a expression of the relevant protein, P-glycoprotein and the drug whole range of new diagnostic tests, including a better pheno- efflux function. Although it has been claimed that RT-PCR is typic characterization of the various subclasses of acute leuke- a reliable assay for the functional expression of mdr1, no sub- mias. For the detection of MDR in acute leukemias, flow cyto- stantial study has yet been published to prove this point.37 metry is probably the most valuable tool which is presently Therefore, the main advantage of mdr1RNA based assays available. The reason for this is the multiplicity of analyses remains their specificity for mdr1. Especially when such an that can be performed simultaneously on single cells or on assay is combined with (functional) methods of P-glycoprotein samples containing large numbers and/or subpopulations of analysis, the combined results can be used to provide definite cells. Flow cytometric analysis of P-glycoprotein can be per- proof that the resistance is caused by mdr1 rather than by formed with different approaches in order to obtain different other ABC transporters. outcome parameters which each represent either P-gp expression or function.18,31,34,43 Qualitative expression of P-gp using monoclonal antibodies Immunocytochemistry for P-glycoprotein detection that recognize extracellular epitopes of P-gp can be performed in unfixed cells. Such antibodies are MRK16, 4E3, or MM417. AML like many other hematological tumors is characterized FITC-labeled secondary antibodies are used to produce a flu- by an easy access to tumor cells during all phases of active orescence signal. Instead Texas Red can also be used, thus disease. This is one reason why the expression of P-gp in AML allowing triple fluorescence while reducing the background has been studied extensively. Moreover, AML is treated with fluorescence. The results are usually expressed as a proportion several mdr1-related drugs, including anthracyclines, mitox- of cells that are positive, ie show fluorescence above the back- antrone and etoposide. ground level. The main disadvantage of this analysis is the Immunocytochemistry of AML blasts using monoclonal variable definition of positivity between different investigators. antibodies against P-glycoprotein such as MRK16, C219, JSB1, Any study using this assay should perform calibration stan- C494 and 4E3 has been performed in many studies.38–42 Gen- dards such as specified beads to define the flow cytometer. In erally, such a procedure can not be applied on unseparated addition control cell lines, preferably stable, well established bone marrow or blood cells, and a purification step is neces- sensitive and MDR cell lines, should always be analyzed with sary. After Ficoll separation, cytospins can be prepared which the clinical sample in order to correct for interassay variation. can be stained. We have performed this procedure in over It should be noted that standard cell lines should have a level 250 AML and MM samples with a satisfying staining intensity of P-gp expression which is comparable to clinical AML even if thawed, viable cells were used after cryopreservation. samples, in order to avoid any amplification difficulties during The main advantage of this procedure is, that the morphology the analysis. Such cell lines are for example the human mye- remains intact and that P-gp staining can be evaluated in indi- loma cell line RPMI 8226dox1 and the KB8. vidual cells. It has been found recently, however, that not all Quantitative analysis of P-gp is meant to determine the level antibodies give uniform results with immunocytochemistry. of P-gp expression. This result is obtained by measuring the The C219 monoclonal antibody, for example, which binds to mean or median channel shift of a previously gated cell popu- a cytoplasmic epitope of P-gp, produces a high background lation, when comparing the relative fluorescence of P-gp-spe- staining in AML blasts which subsequently can be interpreted cific Moab and the matched control antibody. Usually the as (false) positive samples. Kolmogorov–Smirnov statistics are used to calculate the stain- A major disadvantage of immunocytochemistry in general, ing intensity. However, this approach may insufficiently but for P-gp in particular is its lack of quantification criteria. address the wide variability of P-gp expression. Consequently, Analysis of P-glycoprotein in AML P Sonneveld and E Wiemer 1162 cells with either a high or a low level of expression may have used a fixed interval of 90 min between the initial incubation a small contribution to the overall sample assessment. of rhodamine in the presence of SDZ PSC 833, followed by Samples which have a small proportion of highly P-gp- a washing procedure, until the moment of intracellular rhoda- positive cells may incorrectly be regarded as negative. For a mine 123 assessment. Again, the analysis of clinical samples correct and reliable analysis of P-gp expression different should be performed simultaneously with standard cell lines. antibodies may be employed. The results are usually expressed as the efflux ratio of rhoda- Flow cytometry further has a definitive advantage in its mine or another dye/drug as a function of SDZ PSC 833 or capacity to analyze large numbers of cells and to recognize another modulator at a fixed concentration (Figure 1). Parti- different subpopulations of cells with heterogeneous P-gp cular attention should be given to the presence of protein expression. (serum) in the conditioned medium, since many modulators The most specific analysis of P-gp with flow cytometry is are partly inactivated when bound to serum proteins. the measurement of the efflux function of the protein. Usually Interestingly, the functional assay of P-gp using the rhoda- these techniques are based on the determination of either flu- mine 123 efflux ratio is not always in agreement with the orescent dye or drug accumulation or the efflux measurement MRK16-based P-gp detection on the same cells, using flow of these in intact cells. As fluorescent dyes, rhodamine 123, cytometry. Leith et al43 have observed both MRK16- DiOC2, calcein-AM or Hoechst 33342 can be used. The flu- positive/efflux negative and MRK16-negative/efflux-positive orescent anthracyclines daunorubicin, doxorubicin and idaru- cases, which may reflect the presence of alternative efflux bicin may also serve this purpose. The choice for a functional mechanisms which have not yet been identified. probe of P-gp depends on several factors. Rhodamine 123 is a mitochondrial dye which is transported across the mitochondrial lipid membrane as well as across the plasma mem- Double fluorescence of P-glycoprotein with lineage-specific brane, while calcein is a cytoplasmic marker, which is trans- antigens ported across the plasma membrane only. Hoechst 33342 is a DNA binding dye, and its fluorescence may be quenched For a better understanding of the role of P-gp for AML prog- as it is intercalated. DiOC2 is a membrane potential indicator nosis, it is necessary to investigate P-gp expression and func- and seems a reliable dye for P-gp transport function in cell tion in specific subsets of cells, for example cells that express membranes. These dyes further differ with respect to the flu- immature antigens (CD34, c-kit, HLA-DR), or cells with other orescent properties, the equilibration characteristics, the rate specific phenotypes such as favorable/unfavorable karyotypes, of membrane transport and the optimal pH at which they are in vitro spontaneous proliferation or other poor prognostic transported. The dyes may be used in intact, unfixed cells to measure the maximum or equilibrium intracellular concentration (accumulation assay) or to measure the efflux through P-gp after prior exposure of these cells to the dye, followed by placement in a dye-free medium. At present it is not known which dye is preferable, although most investigators have been using rhodamine 123. Of particular interest is the fact that viable cells have to be used. This means that before analysis, fresh or cryopreserved AML cells should be put through Ficoll–Hypaque and that dead cells should be excluded from the flow analysis. There is however a considerable day-to-day variation of this assay, in particular if absolute fluorescence units are used to calculate the outcome of dye accumulation or efflux assays, for which reason the inter-sample variation may impose serious difficulties in the analysis of large numbers of clinical samples. This problem can at least be partly evaded by using P-gp-blocking agents or reversal agents such as verapamil, cyclosporin or its non-immunosuppressive derivative, SDZ PSC 833. Of these, SDZ PSC 833 is a specific inhibitor of P-gp which completely blocks the efflux of rhodamine 123 at concentrations of 1–2 ␮M. The absolute ratio of the intracellular accumulation of rhodamine 123 with or without SDZ PSC 833 represents an objective and reproducible indicator of the P-gp-mediated efflux of the dye. Still, prob- lems may then be encountered in samples with only slightly increased P-gp expression, where no difference may be observed in samples if SDZ PSC 833 is added. Therefore the use of efflux ratio instead of accumulation ratio is preferable in particular for cells with a low level of resistance. The proper interpretation of this assay depends on the predefined times that are allowed for efflux of rhodamine 123 in vitro in the presence or absence of the reversal agent, once the cells are Figure 1 Comparison of different analytical techniques for mdr1 resuspended in rhodamine 123-free medium. Unfortunately, as performed in Ficoll–Hypaque purified AML blast cells from patients with untreated AML. (a) Daunorubicin intracellular accumulation the rate of efflux is not equivalent in all AML samples, and ratio with/without SDZ PSC 833: relation with MRK 16 staining. (b) for this reason one should perform the analysis ‘on-line’ or Rhodamine 123 intracellular fluorescence ratio with/without SDZ PSC choose a time when the maximum efflux is reached. We have 833: relation with daunorubicin accumulation ratio. Analysis of P-glycoprotein in AML P Sonneveld and E Wiemer 1163

Figure 2 Double fluorescence analysis of MRK16 and anti-CD34 antibodies in bone marrow blast cells isolated from untreated patient with de novo AML as compared with these cells stained with matched control IgGs for both antibodies. phenotypes. At present there is only limited information on of these antigens, usually the differentiation-specific antigen whether these adverse characteristics are independently should be directly labeled, for example with FITC or phyco- expressed or if there exists a possible association. Those stud- erythrin. P-gp should be detected with a monoclonal antibody ies that were published mostly used separate procedures for that recognizes an extracellular epitope such as MRK16 with the analysis of differentiation antigens or other genes and P- a secondary antibody that carries another fluorescence label, gp. Generally, arbitrary cut-off levels were used to define posi- for example Texas Red. Details of the technical assay have tive samples in such studies and therefore they do not provide been published before.31,32 conclusive evidence about co-expression of P-gp with other It is important to adjust for background fluorescence of the adverse prognostic antigens. There are however several CD34-antibody’s label when analyzing MRK16. For the same possible approaches to investigate this issue. reason it was not possible in our hands to combine rhodamine First, double expression of P-gp with other extracellular epi- or daunorubicin accumulation with MRK16. Several investi- topes can be performed such as P-gp with CD34 or c-kit in gators have attempted to combine P-gp detection together AML, or P-gp with CD38 or CD19 in multiple myeloma with cytoplasmic or nuclear antigens or they have used P-gp (Figure 2). In order to perform a technically reliable assay, one antibodies that recognize intracellular epitopes. In both cases, Analysis of P-glycoprotein in AML P Sonneveld and E Wiemer 1164 the plasma membrane has to be permeabilized, which may significance of mdr1, it should be investigated in AML cells lead partly to destruction of P-gp epitopes. We have combined only, thus excluding analytical techniques based on bulk MRK16 with cytoplasmic light chain detection in multiple samples. Moreover, all significant prognostic variables should myeloma samples using Permeafix (Ortho, Raritan, NJ, USA), be studied in the same samples including conventional kary- and we have found that although the assay can be performed otyping, analysis of differentiation antigens (CD34), morpho- to recognize P-gp expressing clonal plasma cells, there may logic classification (M0) and other variables. In addition it be a loss of MRK16 related fluorescence (Sonneveld, unpub- seems preferable to investigate other mechanisms of MDR, as lished results). Furthermore, it is important to define the opti- well as cytarabine dose and metabolism. mal incubation conditions for the P-gp antibodies and to select the preferred order of incubation steps. In our experience both fresh or viably cryopreserved cells can be used to perform these studies.34 With these studies we have observed that References CD34-positive subsets of AML blasts have a higher expression of P-gp but not MRP (multidrug resistance-associated protein). 1 Yates W, Glidewell O, Wiernik PH, Cooper MR, Steinberg D, Dosik H, Levy R, Hoagland C, Henry P, Gottlieb A, Cornell C, In contrast, c-kit-positive, CD34-negative cells did not express Berenberg J, Hutchison JL, Raich P, Nissen N, Ellison RR, Felick P-gp. Similar studies in multiple myeloma and in natural killer R, James GW, Falkson G, Silver RT, Hauani F, Green NM, Hender- cells revealed that CD38 or CD56 analysis can also be com- son E, Lone L, Holland JF. Cytosine arabinoside with daunorubicin bined with MRK16.35 These examples show that P-gp or adriamycin for therapy of acute myelocytic leukemia: a CALGB expression in subsets may vary. study. Blood 1982; 60: 454–462. Secondly, an important advantage of flow cytometry for the 2 Cassileth PA, Lynch E, Hines JD, Oken MM, Mazza JJ, Bennett JM, McGlare PB, Edelstein M, Harrington DP, O’Connell MJ. Varying detection of P-gp seems to be the possibility to identify mul- intensity of postremission therapy in acute myeloid leukemia. tiple subsets of hematopoietic cells, using lineage-specific or Blood 1992; 79: 1924–1930. differentiation antigens. Those cells can be separated using 3 Mayer JR, Davis RB, Schiffer CA, Berg DT, Powell BL, Schulman positive or negative sorting procedures. These sorting tech- P, Omura GA, Moore JO, McIntyre OR, Frei E. Intensive postre- niques will result in a homogeneous population of cells, mission chemotherapy in adults with acute myeloid leukemia. allowing a further analysis, such as molecular tests for mdr1 New Engl J Med 1994; 331: 896–903. 4 Rees JKH, Gray RG, Swirsky D, Hayhoe FGJ. Principal results of gene expression or in situ hybridization. the Medical Research Council’s 8th Acute Myeloid Leukemia trial. Lancet 1986; 2: 1236–1241. 5 Buchner T, Urbanitz D, Hiddeman W, Ruhl H, Ludwig WD, Clinical relevance of P-gp in AML Fischer J, Aul HC, Vaupel HA, Kuse R, Zeile G. Intensified induction and consolidation with or without maintenance chemo- Many authors have observed a relation of P-gp expression therapy for acute myeloid leukemia (AML): two multicenter studies of the German AML Cooperative Group. J Clin Oncol 1985; 3: with a worse outcome of remission induction treatment, ie the 1583–1589. complete remission rate (CR) or even survival. The results of 6 Vogler WR, Winton EF, Gordon DS, Raney MR, Go B, Mayer L. these prospective or retrospective studies are summarized in A randomized comparison of postremission therapy in acute myel- Table 1. There are now clear indications that P-gp expression ogenous leukemia: a South-Eastern Cancer Study Group trial. is associated with an adverse probability of CR and overall Blood 1984; 63: 1039–1045. survival, although it remains to be clarified if P-gp is an inde- 7 Zittoun RA, Mandelli F, Willemze R, Witte TD, Labar B, Resegotti L, Leoni F, Damasio E, Visani G, Papa G, Caronia F, Hayat M, pendent prognostic factor or whether it coincides with other Strickmans P, Rotoli B, Leoni P, Peetermans M, Dardenne M, 16,18,37,44–47 variables. The details of these analyses will be dis- Vegna M, Petti M, Solbu G, Suciu S. Autologous or allogeneic cussed elsewhere. It is evident that the analyses performed so bone marrow transplantation compared with intensive chemo- far have used different techniques for mdr1 analysis and differ- therapy in acute myelogenous leukemia. New Eng J Med 1995; ent clinical endpoints. For a valid evaluation of the prognostic 332: 217–223. 8Lo¨wenberg B, Van Putten WLJ, Verdonck LF, Dekker AW, De Gast CG, Willemze R, Zwaan FE, De Planque H, Abels J, Sonneveld P, Van der Lelie J, Goudsmit R, Sizoo W, Hagenbeek A. Autolog- Table 1 Multivariate analysis of mdr1 in untreated acute ous bone marrow transplantation in acute myeloid leukemia in myeloid leukemia first remission: results of a Dutch prospective study. J Clin Oncol 1990; 8: 287–294. Ref. No. mdr1- Adverse 9 Bishop JF, Matthews JP, Young GA. A randomized study of high- patients positive factors for dose cytarabine in induction in acute myeloid leukemia. Blood (%) survival 1996; 87: 1710–1717. 10 Dombret H, Chastang C, Fenaux P, Reiffers J, Bordessoule D, Bou-

16 abdallah R, Mandelli F, Ferrant A, Auzanneau G, Tilly H. A con- Campos 150 47 mdr1, CD34 Karyotype trolled study of recombinant human granulocyte colony-stimulat- age NE 43,44 ing factor in elderly patients after treatment for acute myelogenous Leith 171 55 mdr1, Karyotype leukemia. New Engl J Med 1995; 332: 1678–1683. secondary NE 11 Lo¨wenberg B, Zittoun R, Kerkhofs H, Jehn U, Abels J, Debusscher AML CD34 NE 45 L, Cauchie C, Peetermans M, Solbu G, Suciu S. On the value of Hunault 124 40 mdr1, age, Karyotype intensive remission-induction chemotherapy in elderly patients of karyotype NS 65+ years with acute myeloid leukemia: a randomized phase III Age NS 46 study of the EORTC Leukemia Group. J Clin Oncol 1989; 7: Del Poeta 158 43 mdr1, CD7, Karyotype 1268–1274. age NE 12 Hamblin TJ. Disappointments in treating acute leukemia in the CD34 NE 47 elderly (commentary). New Engl J Med 1995; 332: 1712–1713. Sonneveld 130 57 mdr1, age CD34 NS 13 Badley G, Jurunka PF, Ling V. Mechanisms of multidrug resist- karyotype ance. Biochem Biophys Acta 1988; 948: 87–128. 14 Gottesmann MM, Pastan I. The multidrug transporter, a double- NE, not evaluated; NS, not significant. edged sword (review). J Biol Chem 1988; 263: 12163–12166. Analysis of P-glycoprotein in AML P Sonneveld and E Wiemer 1165 15 Arceci RJ. Clinical significance of P-glycoprotein in multidrug drug resistant cells with high proliferative capacity determine resistance malignancies. Blood 1993; 81: 2215–2222. response to therapy in acute myeloid leukemia. Leukemia 1995; 16 Campos L, Guyotat D, Archimbaud E, Calmard-Oriol P, Tsuruo 9: 1025–1031. T, Trancy J, Treille D, Fiere D. Clinical significance of multidrug 32 Boekhorst PAW te, Wittebol S, Hagemeijer A, Lo¨wenberg B, Son- resistance P-glycoprotein expression on acute nonlymphoblastic neveld P. Predominance of the multidrug resistance (MDR1) leukemia at diagnosis. Blood 1992; 73: 473–476. phenotype in acute myeloid leukemia cells is associated with an 17 Gekeler V, Frese G, Noller R, Handgretinger R, Wilisch A, Schmidt immature (CD34) phenotype. Blood 1993; 10: 3157–3162. H, Muller CP, Dopfer R, Klingebiel T, Diddens H. MDR1-P-glyco- 33 Nooter K, Herweijer H. Multidrug resistance (mdr) genes in human protein, topoisomerase, and glutatione-S-transferase ␲ gene cancer. Br J Cancer 1991; 63: 663–669. expression in primary and relapsed acute adult and childhood leu- 34 Broxtermann HJ, Sonneveld P, Feller N, Ossenkoppele GJ, Wa¨hrer kaemias. Br J Cancer 1992; 66: 507–551. DCR, Eekman CA, Schoester M, Lankelma J, Pinedo HM, Lo¨wen- 18 Guerci A, Merlin JL, Missoum N, Feldmann L, Marchal S, Witz F, berg B, Schuurhuis GJ. Quality control of multidrug resistance Rose C, Guerci O. Predictive value for treatment outcome in acute assays in adult acute leukemia: correlation between assays for P- myeloid leukemia of cellular daunorubicin accumulation and P- glycoprotein expression and activity. Blood 1996; 87: 4809–4816. glycoprotein expression simultaneously determined by flow cyto- 35 Sonneveld P, Schoester M, De Leeuw K. Clinical modulation of metry. Blood 1995; 82: 2147–2153. multidrug resistance in multiple myeloma. J Clin Oncol 1994; 12: 19 Herweijer H, Sonneveld P, Baas F, Nooter K. Expression of MDR1 1584–1591. and MDR3 multidrug-resistance genes in human acute and 36 Sonneveld P, Nooter K, Burghouts JThM, Herweijer H, Adriaansen chronic leukemias and association with stimulation of drug HJ, Van Dongen JJM. High expression of the mdr3 multidrug- accumulation by cyclosporin. J Natl Cancer Inst 1990; 28: resistance gene in advanced stage chronic lymphocytic leukemia. 1133–1140. Blood 1992; 79: 1496–1500. 20 Ino T, Miyazaki H, Isogai M, Nomura T, Tsuzuki M, Tsuruo T, 37 Futscher BW, Blake LL, Gerlach JH, Grogan ThM, Dalton SW. Ezaki K, Hirano M. Expression of P-glcyoprotein in de novo acute Quantitative polymerase chain reaction analysis of mdr1 mRNA myelogenous leukemia at initial diagnosis: results of molecular in multiple myeloma cell lines and clinical specimens. Anal and functional assays, and correlation with treatment outcome. Biochem 1993; 213: 414–421. Leukemia 1994; 8: 1492–1497. 38 Brophy NA, Marie JP, Rojas VA, Warnke RA, McFall PJ, Smith SD, 21 Lamy T, Goasguen JE, Mordelet E, Grulois I, Dauriac C, Brenou Sikic BI. Mdr1 gene expression in childhood acute lymphoblastic B, Chaperon J, Fauchet R, Le Prise PY. P-glycoprotein (P-170) and leukemias and lymphomas: a critical evaluation by four tech- CD34 expression in adult acute myeloid leukemia (AML). Leuke- niques. Leukemia 1994; 8: 327–335. mia 1994; 8: 1879–1883. 39 Van Kapel J, Nooter K, Oostrum RG, Sonneveld P. Western immu- 22 List A, Spier CM, Cline A, Doll DC, Garewal H, Morgan R, Sand- noblotting enhanced chemiluminescence (ECL) detection of P-gly- berg AA. Expression of the multidrug resistance gene product (P- coprotein in vinca-alkaloid resistant cell-lines and human leu- glycoprotein) in myelodysplasia is associated with a stem cell kemic cells. In: Kaspers GJL, Pieters R, Twentyman PR, Weisenthal phenotype. Br J Haematol 1991; 78: 28–34. LM, Veerman AJP (eds). Drug Resistance in Leukemia and Lym- 23 Marie JP, Zittoun R, Sikic BI. Multidrug resistance (MDR1) gene phoma. Harwood Academic: Chur, Switzerland, 1993, pp 77–87. expression in adult acute leukemias: correlation with treatment 40 Grogan T, Dalton W, Rybski J, Spier C, Meltzer P, Richter L, Glea- son M, Pindur J, Cline A, Scheper R. Optimization of immunocyto- outcome and in vitro drug sensitivity. Blood 1991; 78: 586–592. chemical P-glycoprotein assessment in multidrug-resistant plasma 24 Miwa H, Kita K, Nishii K, Morita N, Takakura N, Ohishi K, Mah- cell myeloma using three antibodies. Lab Invest 1990; 63: 815– mud N, Kageyama S, Fukumoto M, Shirakawa S. Expression of + 824. MDR1 gene in acute leukemia cells: association with CD7 acute 41 Kandel RA, Campbell S, Noble-Topham S, Bell R, Andrulis IL. Cor- myeloblastic leukemia/acute lymphoblastic leukemia. Blood relation of P-glycoprotein detection by immunohistochemistry 1993; 82: 3445–3449. with mdr-1 mRNA levels in osteosarcomas. Diagn Molec Pathol 25 Musto P, Mellilo L, Lombardi G, Matera R, Di Giorgio G, Carot- 1995; 4: 59–65. enuto M. High risk of early resistant relapse for leukaemic patients 42 Haddad G, Thorner PS, Bradley G, Dalton WS, Ling V, Chan HSL. with presence of multidrug resistance associated P-glycoprotein A sensitive multilayer immunoalkaline phosphatase method for positive cells in complete remission. Br J Haematol 1991; 78: detection of P-glycoprotein in leukemia and tumor cells in the 50–53. bone marrow. Lab Invest 1994; 71: 595–603. 26 Nooter K, Sonneveld P, Oostrum R, Herweyer H, Hagenbeek A, 43 Leith CP, Chen IM, Kopecky KJ, Appelbaum FR, Head DR, God- Valerio D. MDR1 gene in blast cells from patients with acute mye- win JE, Weick JK, Willmann CL. Correlation of multidrug resist- locytic leukemia is associated with decreased anthracycline ance (MDR1) protein expression with functional dye/drug efflux accumulation that can be restored by cyclosporin-A. Int J Cancer in acute myeloid leukemia by multiparameter flow cytometry: 1990; 45: 263–268. identification of discordant MDR−/efflux+ and MDR+/efflux− 27 Paietta E, Andersen J, Racevskis J. Multidrug resistance gene cases. Blood 1995; 86: 2329–2336. (MDR1) transcript or CD34 antigen expression levels do not pre- 44 Leith CP, Kopecky KJ, Chen IM, Godwin JE, Appelbaum FR, Head dict for complete remissions in de novo adult acute myeloid leuke- DR, Willman CL. MDR1 expression is highly predictive for mia (AML): an Eastern Cooperative Oncology Group study. Blood achievement of complete remission (CR) in acute myeloid leuke- 1994; 84: 3771 (Abstr.). mia (AML) in the elderly: a Southwest Oncology Group Study. 28 Pirker R, Wallner J, Geissler K, Linkesch W, Haas OA, Bettelheim Blood 1995; 86: 269a (Abstr.). P, Hopfner M, Scherrer R, Valent P, Havelec L. MDR1 gene 45 Hunault H, Zhou D, Ramon S, Delmer A, Cadiou M, Perrot JY, expression and treatment outcome in acute myeloid leukemia. J Marie JP, Zittoun R. MDR expression in acute myeloid leukemia Natl Cancer Inst 1991; 83: 708–712. (AML): major prognosis significance in multivariate analysis. 29 Sato H, Gottesmann MM, Goldstein LJ, Pastan I, Block AM, Sand- Blood 1995; 86: 269a (Abstr.). berg AA, Preisler HD. Expression of the multidrug resistance gene 46 Del Poeta G, Stasi R, Aronica G, Venditti A, Cox MC, Bruno A, in myeloid leukemias. Leukemia Res 1990; 14: 11–22. Buccisano F, Masi M, Tribalto M, Amadori S, Papa G. Clinical 30 Solary E, Bidan JM, Calvo F, Chauffert B, Caillot D, Mugneret F, relevance of P-glycoprotein expression in de novo acute myeloid Gauville C, Tsuruo T, Carli PM, Greg H. P-glycoprotein expression leukemia. Blood 1996; 87: 1997–2004. and in vitro reversion of doxorubicin resistance by verapamil in 47 Sonneveld P, Schoester M, Smit E, Lo¨wenberg B. Expression of the clinical specimens from acute leukemia and myeloma. Leukemia multidrug resistance phenotype MDR1 but not MRP is the single 1991; 5: 592–597. most important prognostic factor in elderly patients with acute 31 Boekhorst PAW te, Lo¨wenberg B, Van Kapel J, Sonneveld P. Multi- myeloid leukemia. Br J Haematol 1996; 93: 141 (Abstr.). Analysis of P-glycoprotein in AML P Sonneveld and E Wiemer 1166