Multidrug Resistance in Acute Myeloid Leukemia:
Potential New Therapeutics
I n acute myeloid leukemias (AMLs),
multidrug resistance (MDR) is fre-
Many investigators working to im-
prove cancer therapy have tried to use
various immunologic approaches to
Although there have been signiﬁcant
advances in the treatment of AML, only
20%230% of these patients reach long-
quently, but not always, caused by the killing cancer cells because such ap- term disease-free survival (14). The
MDR1 gene product, a 170- to 180- proaches might circumvent chemother- addition of Pgp expression predicts an
kDa glycoprotein known as P-glyco- apeutic drug resistance. One of the even poorer prognosis in the elderly,
protein (Pgp) or human MDR1 protein assumptions of an immunologic ap- those patients who have relapsed, and
(1). It is characterized by broad re- proach is that cells that are resistant to patients with secondary AML (11,15).
sistance to several structurally, chemi- chemotherapeutics should not be cross- When considering radioimmunother-
cally, and pharmacologically distinct resistant to immunotherapy because the apy for the treatment of any type of
chemotherapeutic compounds. Many mechanisms of action are so widely cancer, one must keep in mind three
theories explain how Pgp affects different. Unfortunately, several groups important factors: the target tumor,
MDR, including the ‘‘drug pump’’ have shown that MDR cells seem to be isotope, and antibody (16). Because
model, in which Pgp hydrolyzes aden- resistant to several forms of immuno- leukemias are easily accessible and
osine triphosphate to actively pump logic therapies, including complement- have well-deﬁned differentiation anti-
drugs out of the cell (2,3). As many mediated cytotoxicity (5,6), immunotoxin, gens, they are ideal for the use of
MDR cells have been found to have gelonin attached to monoclonal anti- radiolabeled antibodies (17). AML is
bodies (mAbs) (8), interleukin 2 (9), characterized by the expression of
and drug immunoconjugates (10–12). CD13, CD15, CD33, and CD117
See page 1546
Another immunologic treatment is (18). CD33 is a 67-kDa cell-surface
the use of mAbs to deliver radioisotopes glycoprotein found on myeloid leuke-
increased intracellular pH and de- directly to the tumor cells. At the present mia cells and on committed myelo-
creased plasma membrane potentials, time, the use of radiolabeled antibodies monocytic and erythroid progenitor
another explanation of the MDR in the treatment of AML with Pgp cells. It is not found on lymphoid,
mechanism of Pgp is the altered- expression has not been well studied, if nonhematopoietic, or mature myeloid
partition model. This model takes into at all. In this issue of The Journal of cells, making CD33 a prime target for
account how the alterations in in- Nuclear Medicine, Kersemans et al. (13) antibody binding (19,20).
tracellular pH and membrane potential use a strategy to overcome MDR using Different isotopes have been at-
that accompany the overexpression of mAbs that have been labeled with 111In tached to antibodies based on their
Pgp indirectly affect the partitioning modiﬁed with a nuclear localizing half-life and the type of particle they
of the drugs, but Pgp does not directly sequence (NLS). They showed that the emit. For single-cell kill, with the least
pump the drugs out of the cell (4–6). addition of the peptide sequence cytotoxicity to normal cells, the ideal
Other drug-resistance proteins, which CGYGPKKKRKVGG, which contains isotope emission should have a path
are associated with hematologic can- the NLS PKKKRKV of SV-40 large length that is the same as that of the
cers that use similar mechanisms, T-antigens, allowed the antibody to be target cells (21). 90Y and 131I emit
include the MDR-associated protein, internalized and to translocate to the b-particles, which have a relatively long
the lung resistance protein, and the nucleus. Once in the nucleus, Auger elec- range and a low linear-energy transfer.
breast cancer resistance protein (7). trons from the 111In were cytotoxic to the Unfortunately, these properties make
parental HL60 cell line and drug-resistant b-particles an excellent choice for solid
(mitoxantrone-insensitive) HL60-MX-1 tumors but not for hematologic cancers,
cell line and to the primary AML speci- in which antibodies cannot bind to all
Received May 12, 2008; revision accepted the tumors cells at sufﬁcient levels (21).
May 16, 2008.
mens that were expressing a diversity of
For correspondence or reprints contact: Jeffrey MDR phenotypes. These results suggest This long particle range results in the
H. Weisburg, Stern College for Women, Yeshiva
University, New York, NY 10016.
that 111In-anti-CD33 mAbs that are mod- irradiation of target tumor cells as well
E-mail: firstname.lastname@example.org iﬁed with an NLS may be a new treatment as surrounding cells, causing normal
COPYRIGHT ª 2008 by the Society of Nuclear
for patients who have drug-resistant cells to be killed. Alpha-particles, on
DOI: 10.2967/jnumed.107.050153 myeloid leukemias. the other hand, have a shorter range
AML MDR AND POTENTIAL NEW THERAPEUTICS • Weisburg 1405
(about the size range of 3–5 times that have levels of antigen that are compa- multidrug-resistant cells, is promising.
of the target cells) than b-particles and rable to those expressed in the cell The results demonstrated in the article
a much higher linear energy transfer. lines. Kersemans et al. (13) reported by Kersemans et al. (13) show that
Only one a-particle is needed to cross that the HL-60 and HL-60-MX-1 cells there may be a treatment to overcome
the nucleus to induce cell death by contained 3–4 · 104 CD33 per cell. MDR in AML and perhaps other types
causing double-strand breaks in the With this number of sites per cell, they of hematologic cancers.
DNA, making a-particles more efﬁcient determined that at a speciﬁc activity of
than b-particles at killing a single cell. 8 MBq/mg, 10% of the radioactivity is
Jeffrey H. Weisburg
The shorter range of a-particles is localized to the cell surface, 70% is in Yeshiva University
beneﬁcial in 2 ways. First, nonspeciﬁc the cytoplasm, and 20% is in the New York, New York
cytotoxicity to nearby healthy cells nucleus. It has been previously re-
should be reduced, and second, because ported that myeloid and monocytic
the path length of a-particles is shorter, leukemias contain only 1 · 104 CD33
they would be an effective single-cell per cell (27). A possible drawback
killer (22). Auger electrons have a high- with CD33 as the target is that these 1. Chang G. Multidrug resistance ABC transporters.
FEBS Lett. 2003;555:102–105.
linear-energy transfer and an even levels can be decreased among differ- 2. Croop JM, Guild B, Gros P, Houseman DE.
smaller path length than do a-particles, ent cells lines and in patient samples, Genetics of multidrug resistance: relationship of
making Auger electrons also suitable preventing enough Auger electrons a coned gene to the complete multidrug resistant
phenotype. Cancer Res. 1987;47:5982–5988.
for the treatment of hematologic can- from entering the nucleus to kill the 3. Hamada H, Tsuruo T. Characterization of the
cers. High antigen densities are essential cell. To overcome these lower CD33 ATPase activity of the Mr 170,000 to 180,000
because a large number of decays is vital levels of different targets, a higher membrane glycoprotein (P-glycoprotein) associ-
ated with multidrug resistance in K562/ADM
for strong cytotoxicity (23). The dose, speciﬁc activity may be needed to cells. Cancer Res. 1988;48:4926–4932.
meaning what kills the cells, is de- increase the percentage inside the 4. Wadkins RM, Roepe PD. Biophysical aspects of
pendent on the energy of the particle and nucleus. However, this could reduce P-glycoprotein-mediated multidrug resistance. Int
Rev Cytol. 1997;171:121–165.
the half-life. Unlike a- and b-radiation, immunoreactivity. Another aspect to 5. Weisburg JH, Curcio M, Caron PC, et al. The
Auger emitters need to be attached to the consider is the pharmacokinetics of multidrug resistant phenotype confers immunolog-
individual tumor cells and enter the the antibody with the attached poly- ical resistance. J Exp Med. 1996;183:2699–2704.
6. Weisburg JH, Roepe PD, Dzekunov S, Scheinberg
nucleus. Optimum results with Auger peptide chain; are they different from DA. Intracellular pH and multidrug resistance
electron emitters were seen when they those of the antibody alone? Pharma- regulate complement-mediated cytotoxicity of nu-
were bound to the DNA, causing double- cokinetics must be determined for these cleated cells. J Biol Chem. 1999;274:10877–10888.
7. Hirose M, Hosoi E, Hamano S, Jalili A. Multidrug
stand breaks (24). novel constructs. resistance in hematological malignancy. J Med
In the study by Kersemans et al. The attachment of the NLS to the Invest. 2003;50:126–135.
(13), the authors were able to specif- antibody may invigorate interest in 8. McGrath MS, Rosenblum MG, Philips MR,
Scheinberg DA. Immunotoxin resistance in multi-
ically target the nucleus of the HL-60 radioimmunotherapy and seems to drug resistant cells. Cancer Res. 2003;63:72–79.
and HL-60-MX-1 cells by the addition have many applications. The use of 9. Drach J, Gsur A, Hamilton G, et al. Involvement of
of the NLS onto the antibodies, a-emitters on NLS-tagged antibodies P-glycoprotein in the transmembrane transport of
IL-2, IL-4 and interferon-g in normal human T
making the increase in nuclear radio- is noteworthy. The NLS-tagged anti- lymphocytes. Blood. 1996;88:1747–1754.
activity 4- to 6-fold greater than that of bodies could bring the isotope directly 10. Naito K, Takeshita A, Shigeno K, et al. Calichea-
the antibodies without the NLS. With to the nucleus to ensure a ‘‘direct hit.’’ micin-conjugated humanized anti-CD33 monoclo-
nal antibody (gemtuzumab, zogamicin, CMA-676)
this increase in nuclear localized Possible problems associated with shows cytocidal effect on CD33-positive leukemia
radiation, survival of the targeted cells a-particle emitters such as 212Bi and cell lines, but is inactive on P-glycoprotein-express-
was signiﬁcantly decreased. The au- 213Bi are the short half-lives, making it ing sublines. Leukemia. 2000;14:1436–1443.
11. Linenberger ML, Hong T, Flowers D, et al. Multidrug-
thors also saw this increased cytotox- difﬁcult for an antibody to reach the resistance phenotype and clinical response to
icity in the primary AML samples that tumor cells in time to have an effect, gemtuzumab ozogamicin. Blood. 2001;98:988–994.
expressed the different types of MDR and the longer path lengths, affecting 12. Sievers EL, Larson RA, Stadtmauer EA, et al.
Efﬁcacy and safety of gemtuzumab ozogamicin in
transporters (13). This result is excit- some normal bystander cells (21). The patients with CD33-positive acute myeloid leuke-
ing because other methods to over- issue of short half-life could be solved mia in ﬁrst relapse. J Clin Oncol. 2001;19:3244–
come drug resistance in AML have by using 225Ac, which has a half-life 3254.
13. Kersemans V, Cornelissen B, Minden MD,
fallen short. Unsuccessful results have of 10.5 d (28). To reduce the killing of Brandwein J, Reilly RM. Drug-resistant AML
been observed in clinical trials using normal cells, the NLS on the antibody cells and primary AML specimens are killed by
111In-anti-CD33 monoclonal antibodies modiﬁed
speciﬁc MDR1 inhibitors such as would bring the a-particle directly to
with nuclear localizing peptide sequences. J Nucl
PSC-883 or quinine (25,26). the nucleus of the cancerous target, Med. 2008;49:1546–1554.
Given the dependence of cytotoxic- and less antibody would be required. 14. Cassileth PA, Harrington DP, Appelbaum FR, et al.
ity on high levels of antibody binding, The addition of the NLS to a radio- Chemotherapy compared with autologous or alloge-
nic bone marrow transplantation in the manage-
it is necessary to determine whether labeled antibody, allowing it to enter ment of acute myeloid leukemia in ﬁrst remission.
the samples of patients with AML the nucleus and increase the killing of N Engl J Med. 1998;339:1649–1656.
1406 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 49 • No. 9 • September 2008
15. Legrand O, Simonin G, Perrot J-Y, et al. Pgp and human myeloid progenitor cells. Leuk Res. 1984;8: addition to daunorubicin and cytarabine in the
MRP activities using calcein-AM are prognostic 521–534. treatment of elderly patients with previously un-
factors in adult acute myeloid leukemia patients. 21. Mattes MJ. Radionuclide-antibody conjugates for treated acute myeloid leukemia (AML), in relation to
Blood. 1998;91:4480–4488. single-cell cytotoxicity. Cancer. 2002;94(4 suppl): MDR status at diagnosis. Blood. 2005;106:2646–
16. Goldenberg DM. Target therapy of cancer with 1215–1223. 2654.
radiolabeled antibodies. J Nucl Med. 2002;43:693– 22. Burke JM, Jurcic JG, Scheinberg DA. Radio- 26. Solary E, Drenou B, Campos L, et al. Quinine as
713. immunotherapy for acute leukemia. Cancer Con- a multidrug resistance inhibitor: a phase 3 multicentric
17. Burke JM, Jurcic JG. Radioimmunotherapy of trol. 2002;9:106–113. randomized study in adult de novo acute myelogenous
leukemia. Adv Pharmacol. 2004;51:185–208. 23. Michel RB, Andrews PM, Castillo ME, Mattes MJ. leukemia. Blood. 2003;102:1202–1210.
18. Todd WM. Acute myeloid leukemia and related In vitro cytotoxicity of carcinoma cells with 111In- 27. Tanimoto M, Scheinberg DA, Cordon-Cardo C,
conditions. Hematol Oncol Clin North Am. 2002; labeled antibodies to HER-2. Mol Cancer Ther. et al. Restricted expression of an early myeloid
16:301–319. 2005;4:927–937. and monocytic cell surface. Leukemia. 1989;3:339–
19. Andrews RG, Torok-Storb B, Bernstein ID. 24. Buchegger F, Perillo-Adamer F, Dupertuis YM, 348.
Myeloid-associated differentiation antigens on Delaloye AB. Auger radiation target into DNA: 28. Miederer M, McDevitt MR, Sgouros G, et al.
stem cells and their progeny identiﬁed by mono- a therapy perspective. Eur J Nucl Med Mol Pharmacokinetics, dosimetry, and toxicity of the
clonal antibodies. Blood. 1983;62:124–132. Imaging. 2006;33:1352–1363. targetable atomic generator, 225Ac-HuM195, in
20. Grifﬁn JD, Linch D, Sabbath K, et al. A mono- 25. van der Holt B, Lowenberg B, Burnett AK, et al. The nonhuman primates. J Nucl Med. 2004;45:129–
clonal antibody reactive with normal and leukemic value of the MDR1 reversal agent PSC-833 in 137.
In the article ‘‘SPECT/CT,’’ by Buck et al. (J Nucl Med. 2008;49:1305–1319), the legends of Figures 1, 5, and 6 contain
errors. In the legend of Figure 1, ‘‘99mTc-iodobenzamide’’ should be ‘‘123I-ioﬂupane (DaTSCAN [GE Healthcare]).’’ In
the legend of Figure 5, panel B is described as a 99mTc-MIBI scan but is in fact a 99mTc-pertechnetate scan. In the legend
of Figure 6, ‘‘99mTc-MIBG’’ should be ‘‘123I-MIBG.’’ The authors regret the errors.
AML MDR AND POTENTIAL NEW THERAPEUTICS • Weisburg 1407