The FLT3 inhibitor PKC412 in combination with cytostatic drugs in vitro in acute myeloid leukemia
Abstract An internal tandem duplication of FLT3 (FLT3/ ITD) occurs in approximately 25% of newly diagnosed AML. PKC412 inhibits the growth of leukemic cell lines with FLT3 mutations such as the MV4-11. This study evaluated the in vitro effects of the combination of PKC412 and ara-C or daunorubicin, studying the effect of co-incubation, pre-incubation and sequential incubation of the drugs in patient samples and cell lines. Thirty-three patients with AML were included. Two cell lines were studied; MV4-11 that expresses the FLT3/ITD and HL-60 that does not. In the patient cells PKC412 exerted its effect at concentrations between 0.1 and 2.0 µM. For MV4-11 cells concentrations down to 1 nM were effective. In patient samples, the results of co-incubation of PKC412 with ara-C were synergistic in 5%, additive in 67%, sub additive in 17% and antagonistic in 11% of the cases. In patient cells, incubations with ara-C and PKC412 resulted in synergistic effects in 17% of the FLT3/ITD positive samples compared to 0% synergistic in the FLT3/ITD negative samples (p < 0.01). Antagonistic effects were more common in the FLT3/ITD negative sam- ples. The timing of the drugs had little impact on the effect. In cell lines, antagonistic effects were seen frequently in HL-60 (90%) and less so in MV4-11 (60%) regardless of sequence or timing of the drugs. The combination of daunorubicin and PKC412 resulted in more synergistic and less antagonistic effects compared to combinations with ara-C, in both patient material and cell lines. The combination of Lonafarnib, a farnesyl-transferase inhibitor (FTI) and PKC412 had additive and synergistic effects in both FLT3/ ITD positive and negative cell lines. In conclusion, the combination of PKC412 together with chemotherapeutic drugs is more effective in FLT3/ITD positive AML cells. Antagonistic effects can be seen, especially in patient sam- ples without FLT3/ITD. Also, the combination of PKC412 and the farnesylinhibitor lonafarnib should be further explored. Keywords : Acute myeloid leukemia · FLT3 · PKC412 · Cytarabine · Daunorubicin · Lonafarnib Introduction Fms-like tyrosine kinase 3 (FLT3) is a member of the class III tyrosine kinase receptor family. FLT3 is predominantly expressed on hematopoietic progenitor cells but is also found in other tissues such as placenta, gonads and brain [22, 30]. Interaction with its ligand (FL) results in receptor dimerization, autophosphorylation and subsequent phos- phorylation of cytoplasmic substrates that are involved in signalling pathways regulating the proliferation of pluripo- tent stem cells, early progenitor cells and immature lym- phocytes [21]. A majority of acute myeloid leukemias (AML) and lym- phoblastic leukemias (ALL) express FLT3 [4]. FLT3 is mutated and activated in approx. 30% of all patients with AML [9]. The mutations involve either an internal tandem duplication (FLT3/ITD) in approx. 25% of AML patients or a point mutation in the activating loop in approx. 7% of patients. Both types of mutations result in a ligand-indepen- dent receptor dimerization, phosphorylation and constitu- tive activation of downstream signalling pathways [16]. FLT3/ITD is not expressed in chronic myeloid leukemia, chronic lymphocytic leukemia, non-Hodgkin’s lymphoma or multiple myeloma [38]. The presence of FLT3/ITD in adult AML patients seems to have little or no impact on the ability to achieve complete remission. The most significant impact of FLT3/ITD found in several studies is its association with increased relapse risk, decreased disease-free sur- vival (DFS) and overall survival (OS) [8, 13, 15]. However, two large studies failed to show a decreased OS in AML patients with FLT3/ITD [31, 33]. PKC412 (N-benzoyl-sturosporine) was developed origi- nally as a protein kinase C and vascular growth factor receptor inhibitor and has been used in phase I trials in solid tumors [25, 27]. Subsequently, this compound was found to specifically and potently inhibit the growth of leukemic cell lines expressing FLT3/ITD or an activating loop mutation of FLT3. Moreover, PKC412 prolonged survivial in trans- genic mice with a FLT3/ITD-induced-myeloproliferative- like syndrome [36]. PKC412 has also been used in a phase II trial in advanced AML patients with the FLT3/ITD [32]. In the latter study, 14 of the 20 patients experienced at least a transient 50% reduction in the number of peripheral blasts. This response is comparable to that observed with imatinib in blast crisis of chronic myeloid leukemia [5]. To clarify the role of PKC412 in AML new treatment strate- gies have to be developed, including combinations with conventional cytostatic drugs used in AML. Furukawa et al. recently published results of co-incubation of PKC412 with eight different cytotoxic agents in human leukemia cell lines, showing synergistic results with all agents in the cell lines with an activating FLT3/ITD mutation. In the cell lines expressing wild type FLT3, antagonistic results of co-incubation were frequent [7]. Lonafarnib, SCH-66336, is a farnesyl-transferase inhibi- tor (FTI) that has been shown to inhibit farnesylation and activation of Ras. It has been suggested that Lonafarnib also acts through inhibition of the farnesylation of RhoB, centromere-binding proteins (CENP)-E and -F as well as other not yet identified proteins [26]. Lonafarnib has shown effect as a single drug in a Phase II study in patients with MDS and CMML. Of 42 evaluable patients, 12 responded, 2 of whom had complete response and 10 had a hematolog- ical improvement. In addition, 16 of 37 patients who had bone marrow (BM) blasts of 45% at baseline showed a reduction of 50% in BM blasts [6].We undertook an in vitro study where the combination of PKC412 and cytostatic drugs in samples from AML patients was evaluated. The AML cell lines MV4-11 and HL-60 were examined in a similar way. The study was designed to evaluate the role of different timing of the com- bined drugs. Also, cell lines were also exposed to the combination of FTI and PKC412.
Material and methods
Patient samples and cell lines
Peripheral blood or bone marrow was collected from 33 consecutive patients with AML. The characteristics of the patients are shown in Table 1. Peripheral blood or bone marrow were collected in heparinized tubes before the start of the treatment. The leukemic blast cells were separated by centrifugation (400g, 30 min) on Lymphoprep. The cells were then washed in RPMI 1640. MV4-11 is a FLT3/ITD positive cell line expressing only the mutated version of the gene, making it a model cell line for FLT3/ITD-related research. HL-60 is a cell line derived from a patient with acute promyelocytic leukemia. It expresses wild type FLT3 [28]. The study was approved by the Medical Research Ethics Committee and the institutional review board of the Karolinska University Hospital.
Incubations and culturing
Cells from fresh samples (n = 24) and freshly frozen cells (n = 9) were used for analysis. The cells (2.0 105 cells/ ml) were incubated in a medium consisting of RPMI 1640 supplemented with Glutamax, 10% fetal bovine serum (FBS) and the cystostatic drug. In the PKC412 sensitivity study, we used drug concentrations between 0.001 and 2.0 µM. In the synergism and timing study, we used the following concentrations: PKC412 0.2 µM continuously, ara-C 0.5 µM continuously and daunorubicin 0.15 µM for 1 h. The 1 h incubation of daunorubicin was chosen to mimic the in vivo situation [34]. In the first part of the synergism study, a 4 days co-incubation of PKC412 and
ara-C or daunorubicin was performed. In the second part, two different in vitro timing models were used. In the pre- incubation model, the incubation time for the first drug was 4 days and the second drug was added after 1 day. In the sequential exposure model, the cells were incubated for 2 days with the first drug, then centrifuged and resus- pended and incubated with the second drug for another 2 days. The different incubation models are illustrated in Fig. 1. Cell lines were incubated as described above at a cell concentration of 0.5 10(5)/ml. The following drug concentrations were used for HL-60: Ara-C at 0.05 or 0.1 µM and PKC 412 at 0.2 or 0.5 µM continuously and in the sequential exposure analysis. For daunorubicin, cells were exposed to 0.1 µM of the drug for 1 h. FTI at 0.25–0.5 µM was used continuously in co-incubation with PKC at 0.2–0.5 µM for four days. For MV4-11, Ara-C at 0.01 µM and PKC412 at 0.005 µM were used in the co- and pre-incubation studies and Ara-C at 0.05 µM and PKC412 at 0.01 µM in the sequential incubations. Dauno- rubicin at 0.03 µM was incubated for 1 h. Two different concentrations were used for the FTI-PKC412 4 day co-incubations: 1 nM PKC412 + 250 nM FTI and 5 nm PKC412 + 0.5 µM FTI. The concentrations were chosen on the basis of the IC50 curves.
ATP bioluminescence assay
Extraction of ATP in leukemic cells was performed by mix- ing equal volumes (2 ml) of cell-suspension and 2.5% tri- chloracetic acid (TCA). The extracts were assayed immediately or stored in a freezer ( 20°C) until analysis. For the assessment of cell viability we used a bioluminis- cence assay measuring ATP content [3, 12, 24, 29]. The mea- surements were performed automatically in an Athos Lucy 1 luminometer (Hettich Labinstrument AB, Sweden). An ATP Kit SL 144-041 (Bio Thema, Dalarö, Sweden) was used for the reaction. The amount of ATP (given as nmol ATP/sam- ple) and the percentage ATP in a sample when compared to the drug-free control was calculated. The result at each con- centration represents a mean of two parallel experiments.
Analysis of FLT3 mutations
Polymerase chain reaction (PCR) was performed to identify patient samples with FLT3/ITD and point mutation D835 in the second tyrosine kinase domain (ATKD) as previously described [1]. FLT3/ITD status in the cell lines have been explored previously [28] and were verified with PCR.
The additive model
In order to study the cytotoxic effects on cells incubated with the drug-combination, the additive model was used [11, 35]. This model predicts that the effect of a combination will be equal to the product of the effects of its constituents. For example, if a drug combination was composed of drugs producing a cell viability of 40 and 60%, respectively, the combi- nation would be expected to result in 24% viable cells (0.4 0.6). An observed combination effect that is larger than predicted by the additive model indicates synergism, whereas a smaller effect represents a subadditive effect. A ratio between the observed viability and the viability predicted by
the additive model was calculated for all combinations. If the ratio exceeded 1.2, the interaction was classified as subaddi- tive and if it was below 0.8, the interaction was classified as synergistic. Ratios between 0.8 and 1.2 were considered to indicate additive interactions, and this interval was set to take into account the intra-assay variability [17]. A cell viability value for the combination exceeding the cell viability of the most effective drug alone was classified as antagonism.
Statistics
The t test for independent samples was used to evaluate the differences in effect between FLT3/ITD positive and nega- tive samples in each tested concentration of PKC412 and when comparing the individual drugs and the combination of ara-C/daunorubicin and PKC412 in FLT3/ITD positive and negative samples. We used the Chi-square algorithm to find differences in the distribution of synergistic, additive, subadditive, and antagonistic effects in the FLT3/ITD posi- tive and negative group, respectively. The Pearson correlation test was used to evaluate correlation of sensitivity of individual patient samples to different drug combinations.
Results
FLT3/ITD analysis
The PCR analysis showed that 10 of the 33 patients (30%) were FLT3/ITD positive. Four of these patients were included in the initial experiments to find the optimal concentration PKC412 and 10 were included in the timing were investigated in the synergy study with co-incubation of PKC412 and ara-C. In 11 samples an additive interaction was seen, in four a sub additive and in two an antagonistic
effect. In the seven FLT3/ITD positive samples, five showed an additive, one a synergistic effect and one a sub additive effect. In this patient subgroup, no sample with an antagonistic effect was found. The proportions of each type of response are shown in Fig. 5. PKC412 and daunorubicin were co-incubated in 26 patients, 18 FLT3/ITD negative and 8 FLT3/ITD positive. The combination with daunorubicin (data not shown).
The development of FLT3 inhibitors is an important step in finding new ways to improve the treatment results in AML. Hopefully, ongoing and future clinical trials will clarify how the FLT3 inhibitors should be used in combina- tion with conventional chemotherapy and/or other new non-chemotherapeutic drugs.