Letters to the Editor
176
Association of folate transporter SLC19A1 polymorphisms with the outcome of multiple myeloma after chemotherapy and tandem autologous transplantation Leukemia (2007) 21, 176–178. doi:10.1038/sj.leu.2404455; published online 26 November 2006
Patients with multiple myeloma (MM) have a median survival of 3 years when treated with standard-dose chemotherapy. An improved survival of approximately 18 months has been shown with high-dose therapy coupled with stem cell rescue (autologous stem cells transplantation, ASCT). However, despite initial good responses, most patients relapse.1 Traditional prognostic factors account for only 15–20% of outcome heterogeneity in MM patients.1 New prognostic factors can also be considered relevant for the therapeutic response, such as single nucleotide polymorphisms (SNPs) in genes involved in the transport of antitumor drugs, and their impact could be established by a pharmacogenetic approach. The reduced folate carrier (RFC1 or SLC19A1) plays an important role in transport of 5-methyltetrafolate into the cells and a lower expression has been associated with the resistance to folate antagonist drugs.2 Pharmacogenetic studies investigating several targets of folate-related drugs used in the treatment of hematological malignancies have been recently reviewed and no clear evidence for the involvement of several known SLC19A1 polymorphisms emerged.3 However, two new SNPs, in exon 4 (254C4T, responsible for a synonymous substitution P232P, rs12659) and in exon 7 (233G4T, rs1051296), have been identified but not yet studied for the impact on gene expression or function. An SLC influence on the outcome of dexamethasone–doxorubicin (adriamycin)–vincristine (DAV) chemotherapy followed by ASCT in patients with MM could be suggested from an increased sensitivity to doxorubicin as found in breast cancer cell lines transfected with SLC19A3.4 We report the results of a pharmacogenetic study on 120 MM patients originating from the Italian population, enrolled between May 1999 and December 2004 in three different Italian centers. All of them were treated with the same chemotherapy regimen (dexamethasone 40 mg/m2, days 1–4; doxorubicin 50 mg/m2, day 1; vincristine 1 mg, day 1) therapy. Cyclophosphamide (4 g/m2) and G-CSF (5 mg/kg/day) were used to mobilize peripheral blood stem cells. Melphalan (200 or 100 mg/m2, for patients over 65 years) was used as a conditioning regimen for each ASCT. A median of 4.5 106 CD34 þ cells/kg was infused. All patients gave informed consent to therapy and transplantation. Disease status before chemotherapy or transplant and response to treatment were defined according to the European Group for Blood and Marrow Transplantation (EBMT). Complete response (CR) was defined as disappearance of M-component from both serum and urine (evaluated by immunofixation), associated with less than 5% plasma cells in bone marrow. Patients were defined as in partial response (PR) when the initial monoclonal protein value was reduced to less than 50% with stability of lytic bone lesions, correction of hypercalcemia, anemia and hypoalbuminemia. Stable disease (SD) was defined as the failure to achieve either CR or PR, while progression disease (PD) required at least a 25% increase in paraprotein or the development of new bone lesions. Proper staging of myeloma for evaluating prognosis and establishing the treatment plan was achieved with the Durie– Salmon system that has been the most widely used myeloma staging system since 1975. At the time of diagnosis, hemoglobin,
creatinine, b2 microglobulin and albumin serum levels were measured. MM patients were consecutively selected whenever practically feasible; they were informed and gave written consent to participate in the study and allow their biological samples to be genetically analyzed. Approval of the study was given by the local ethical committee. DNA was extracted from whole blood of each patient. Genotyping was carried out using the Taqman assay (ABI, Applied Biosystems, Foster City, CA, USA). The MGB Taqman probes and primers were synthesized by ABI (TaqMan SNP Genotyping Assays) that also provided standard PCR profile and reaction conditions. Homozygotes for the most frequent allele of each SNP were set as the reference class. Unconditional logistic regression was applied to calculate odds ratio (OR) and 95% confidence intervals (CIs). All ORs were adjusted by age and gender. Overall survival (OS) was calculated in months from diagnosis up to June 2006, considering a maximum of 60 months for each patient. Cox regression analysis was applied to determine the contribution of age, sex, DAV and ASCT response on survival rate that was represented as Kaplan–Meyer survival curves (95% confidence limit). The SNPAlyze software package based on the expectation-maximization (EM) algorithm was used to estimate Lewontin’s linkage disequilibrium (LD) coefficient (D0 ) and the square of correlation coefficient (r2). A total of 67 male and 53 female subjects were enrolled, with a mean age of 60.579.7 years at diagnosis (range, 34–83 years). The genotype frequencies obtained are in agreement with Hardy–Weinberg equilibrium and strong LD was observed between the two SNPs with values of D0 40.957 and r240.752. Two major haplotypes were found accounting for 93% of all haplotypes: TG (48%) and CT (45%). The distributions of genotypes between male and female subjects were not significantly different. Patient’s response, evaluated 1 month after DAV therapy grouping CR þ PR versus SD þ PD, was not associated to any of the SLC19A1 polymorphisms (Table 1). At the time of analysis, most patients underwent tandem ASCT. One month after second transplantation, the response of 95 valuable individuals was categorized into two classes: improved (CR þ PR ¼ 66%) versus stable or progressed (SD þ PD ¼ 33%). The distributions of the responses after transplantation were not significantly different between 55 male and 40 female subjects. An improvement (27.4% CR þ PR versus 15.2% SD þ PD) was significantly associated with the SLC19A1 T-233T genotype (OR ¼ 0.24, 95% CI, 0.1–0.9, P ¼ 0.04; Table 1) taking into account age (P ¼ 0.09), sex (P ¼ 0.07) and CT response (P ¼ 0.82). Up to now, survival analysis could only be performed on 64 patients (exposed to DAV therapy and tandem ASCT). Our preliminary data showed an OS mean value of 44 months (6–60 months) and suggested a longer OS for patients with SLC19A1 T-233T genotype (Figure 1, for log-rank test P ¼ 0.05) that also showed a higher percentage of living people (93 vs 64%, P ¼ 0.03) with respect to patients with SLC19A1 G-233T and G-233G genotypes. The adjusted hazard ratio (HR) was 0.11 (95% CI, 0.01–1.01), computed with a Cox regression analysis, taking into account age, gender, DAV and ASCT response and
Letters to the Editor
177 Table 1 Genotype distributions of SLC19A1 polymorphisms and ORs for MM patients, related to DAV and ASCT response
ORa (95% CI)
P-value
4 (14.3) 16 (57.1) 8 (28.6)
1 1.58 (0.4–5.9) 2.27 (0.5–9.9)
0.50 0.27
7 (22.6) 16 (51.6) 8 (25.8)
1 1.48 (0.5–4.8) 1.86 (0.5–6.6)
0.50 0.34
N SD+PD (%) b
SLC19A1 254T4C, P232P T/T 25 (28.4) T/C 41 (46.6) C/C 22 (25) SLC19A1 233G4Tb G/G 29 (34.1) G/T 36 (42.4) T/T 20 (23.5)
N CR+PR (%)
b
N SD+PD (%)
0.70
0
20
40
months
60
Observed: SLC19A1_233 = 0
Observed: SLC19A1_233 = 1
Predicted: SLC19A1_233 = 0
Predicted :SLC19A1_233 = 1
Figure 1 Kaplan–Meier plot for overall survival after 60 months from the diagnosis among MM patients by SLC19A1 –233G4T genotype. SLC19A1_233 ¼ 0 includes G-233G þ G-233T genotypes (n ¼ 49) and SLC19A1_233 ¼ 1 (n ¼ 15) includes T-233T genotypes.
c
OR (95% CI)
b
SLC19A1 254T4C, P232P T/T 16 (25.4) 9 (28.1) T/C 30 (47.6) 16 (50) C/C 17 (27) 7 (21.9) SLC19A1 233G4T G/G 18 (29) 14 (42.4) G/T 27 (43.6) 14 (42.4) T/T 17 (27.4) 5 (15.2)
0.80
0.60
ASCT response b
overall survival
0.90
DAV response N CR+PR (%)
1.00
1 0.64 (0.2–2.1) 0.26 (0.1–1.1)
0.61 0.13
1 0.60 (0.2–1.8) 0.24 (0.1–0.9)
0.37 0.04
Abbreviations: ASCT, autologus stem cell transplantation; CI, confidence interval; DAV, doxorubicin (adriamycin)–vincristine; MM, multiple myeloma; OR, odds ratio. a Unconditional logistic regression analysis adjusting for age and gender. b For some polymorphisms, numbers do not sum up to the totals due to genotyping failure (due to low fluorescence signal intensity in the TaqMan assays). c Unconditional logistic regression analysis adjusting for age, gender and DAV response.
traditional factors involved in MM prognosis. DAV and ASCT response resulted in significant association with a shorter OS, that is, non-responders showed a shorter OS; also hemoglobin and b2 microglobulin resulted associated with OS as previously reported from other authors. On the contrary, OS did not seem to be influenced by creatinine and albumin levels. Similar results were obtained for SLC19A1 C-254C_T-233T haplotype associated with a longer OS in line with LD data (Table 2). SLC19A1 T-233T genotype is associated with an improved response after treatment with melphalan and ASCT. Although a specific role of SLC19A1 in melphalan transport was not demonstrated, our result could be interpreted in the sense that this carrier might contribute to the uptake of the conditioning drug, as it was shown for other members of the SLC protein family.5 Moreover, SLC19A1 T-233T seems to be associated with a longer survival in line with the precedent result; that is, a good ASCT response influence the disease progression and outcome. Also the LD analysis and haplotype inference were exploited, that is of particular importance in order to better characterize the role of the candidate variants with unknown functional effect.6 The two SNPs resulted in strong linkage disequilibrium, and, although the small group of patients tested, SLC19A1 haplotype (C232C_T-233T) seems to be associated with a longer OS also suggesting that an altered folate metabolism could decelerate disease progression and a fatal outcome. In fact, rapidly dividing cancer cells have an increased folate requirement for maintaining DNA synthesis.7 Finally, folate involvement in MM has also been observed in a study on a specific polymorphism of 5,10-a´methylenetetrahydrofolate-reductase (MTHFR), responsible for folate metabolism and promoting DNA methylation. In particular, p16 hypermethylation, seemed
Table 2 Overall survival based on Cox regression model after 60 months from the diagnosis Log likelihood ¼ 51.167 HR
LR w2 ¼ 34.26 s.e.
1.01 0.04 Agea Gender 1.13 0.64 DAV response 4.17 2.35 ASCT response 3.18 1.83 0.29 0.17 Hemoglobina 1.92 1.12 Creatininea a 22.78 20.13 b2 microglobulin 0.53 0.46 Albumina 0.12 SLC19A1 233G4Tb 0.11
Prob4w2 ¼ 0.0001
z
P4|z|
95% CI
0.27 0.22 2.54 2.01 2.07 1.12 3.54 0.74 1.95
0.790 0.828 0.011 0.044 0.039 0.261 0.000 0.462 0.051
0.94–1.08 0.37–3.41 1.38–12.57 1.03–9.80 0.09–0.94 0.61–6.02 4.02–128.79 0.09–2.87 0.01–1.01
Abbreviations: ASCT, autologus stem cell transplantation; CI, confidence interval; DAV, doxorubicin (adriamycin)–vincristine; HR, hazard ratio. a At the time of diagnosis. b The HR shows the relative risk of death of patients with T-233T genotype relative to patients with G-233G+G-233T genotypes.
to play a role in pathogenesis of myeloma and, for some authors, in overall survival.8 In any case, further analyses must be performed to confirm our preliminary data. This is the first report demonstrating a predictive value of SLC19A1 genotypes for the outcome of DAV regimen followed by autologous transplantation in MM patients. These findings might have a significant impact on the management of MM patients, as alternative drug combinations become available. Therefore, we conclude that genetic variability may influence the outcome of MM therapy and pharmacogenetics may provide a tool for improving individualization of the therapy. However, the results cannot be generalized, as prognostic factors identified for a specific protocol can be of less importance for other protocols.
Acknowledgements We thank Pier Paolo Fattori, MD, Anna Maria Mianulli, MD, and Manuela Imola, MD, for their contribution to patient recruitment. We acknowledge the support of Mrs Alison Frank in revising the English language.
V Maggini1,7, G Buda2,7, S Galimberti2, E Conidi1, N Giuliani3, F Morabito4, G Genestreti5, P Iacopino6,
Letters to the Editor
178
V Rizzoli3, R Barale1, M Petrini2 and AM Rossi1 Department of Biology, Pisa University, Pisa, Italy; 2 Department of Oncology, Transplants and Advanced Technologies, Section of Hematology, Pisa University, Pisa, Italy; 3 Department of Internal Medicine and Biomedical Science, Hematology and BMT Center, Parma University, Parma, Italy; 4 Haematology Service – S Annunziata Hospital, Cosenza, Italy; 5 Department of Oncology, Division of Onco-Hematology, Infermi Hospital, Rimini, Italy and 6 Bone Marrow Transplant Unit, Hematology Department, AO Reggio Calabria, Italy E-mail:
[email protected] 7 These two authors contributed equally to this study. 1
References 1 Barlogie B, Shaughnessy J, Tricot G, Jacobson J, Zangari M, Anaissie E et al. Treatment of multiple myeloma. Blood 2004; 103: 20–32.
2 Moscow JA. Methotrexate transport and resistance. Leuk Lymphoma 1998; 30: 215–224. 3 Robien K, Boynton A, Ulrich CM. Pharmacogenetics of folaterelated drug targets in cancer treatment. Pharmacogenomics 2005; 6: 673–689. 4 Liu S, Huang H, Lu X, Golinski M, Comesse S, Watt D et al. Downregulation of thiamine transporter THTR2 gene expression in breast cancer and its association with resistance to apoptosis. Mol Cancer Res 2003; 1: 665–673. 5 Huang Y, Sadee W. Membrane transporters and channels in chemoresistance and sensitivity of tumor cells. Cancer Lett 2006; 239: 168–182. 6 Crawford DC, Nickerson DA. Definition and clinical importance of haplotypes. Annu Rev Med 2005; 56: 303–320. 7 Kelemen LE. The role of folate receptor alpha in cancer development, progression and treatment: Cause, consequence or innocent bystander? Int J Cancer 2006; 119: 243–250. 8 Chiusolo P, Farina G, Putzulu R, Reddiconto G, Fiorini A, De Stefano V et al. Analysis of MTHFR polymorphisms and P16 methylation and their correlation with clinicalbiological features of multiple myeloma. Ann Hematol 2006; 85: 474–477.
Treatment of P190 Bcr/Abl lymphoblastic leukemia cells with inhibitors of the serine/threonine kinase CK2
Leukemia (2007) 21, 178–180. doi:10.1038/sj.leu.2404460; published online 2 November 2006
Chronic myelogenous leukemia (CML) and Ph-positive acute lymphoblastic leukemia (ALL) are caused by the fusion of the BCR gene to the ABL gene, which encodes a tyrosine kinase. The P190 and P210 are two main chimeric forms of the Bcr/Abl protein, with the P190 frequently associated with cases of ALL and P210 predominantly occurring in CML.1 Development of inhibitors such as Imatinib, which specifically target the deregulated Bcr/Abl tyrosine kinase, has revolutionized the treatment of CML.2 However, Philadelphia-positive ALL has a much less durable response to Imatinib and treatment options remain limited.3 Therefore, the identification of signal transduction pathways that are perturbed by Bcr/Abl has become an important area of investigation, aiming to find additional molecular targets for treating this type of cancer.4 We have previously identified the catalytic subunit of the serine/threonine kinase CK2 as an interacting partner for the Bcr/ Abl oncoprotein.5 CK2 activity is elevated in many kinds of tumors.6,7 Very specific inhibitors of CK2 have been developed. In previous experiments, we showed that one of these inhibitors, 4,5,6,7-tetrabromobenzimidazole (TBB), inhibited the growth of malignant pre-B lymphoblastic leukemia cells that express Bcr/ Abl P190.5 However, this required relatively high concentrations of 60 mM TBB, to achieve this effect. We therefore compared TBB to a newly developed CK2 inhibitor, 2-dimethylamino-4,5,6,7-tetrabromo-1H-benzimidazole (DMAT). DMAT has the lowest K(i) value of all known CK2 inhibitors, and is cell permeable.8 Based on reported ranges of efficacy,7,8 DMAT was tested at 5, 10 and 20 mM. We used the previously described PLC1 cell line, which was established from a P190 Bcr/Abl transgenic mouse that developed pro/pre-B lymphoblastic leukemia/lymphoma.5 PLC1 cells were standardly maintained on irradiated primary mouse embryonic fibroblast feeder layers (MEFs). Cell viability was assessed on aliquots from individual wells using the trypan blue dye exclusion method. As shown in
Figure 1a, 10 mM DMAT was very effective in inhibiting the proliferation of the lymphoblastic leukemia cells over a period of 2 days, and was superior to treatment with 60 mM TBB. At 22 h, there were no viable cells remaining in the wells treated with 20 mM DMAT (not shown). Treatment of two other leukemia lines derived from different individual BCR/ABL transgenic mice confirmed the effect of this drug on the leukemia cells (Figure 1b). We also treated normal primary mouse fibroblasts with 10 mM DMAT to assess possible toxicity of this drug. As shown in Figure 1c, the drug inhibited proliferation of these cells, but did not cause significant cell death. Also, kidney, bone marrow and liver of mice treated for 25 days with twice daily intraperitoneal injections of 25 mg/kg DMAT were histologically normal (results not shown). We routinely grow lymphoblastic leukemia cells in the presence of stroma to model drug treatment in the context of a defined microenvironment. However, as stroma is known to provide protection to lymphoblastic leukemia cells, we also evaluated the efficacy of DMAT in inhibiting proliferation of the leukemic cells under less protective conditions, in the absence of stroma. Moreover, drug studies performed by others usually do not include stromal support. As shown in Figure 2a, a large drop in viability was observed during the initial 2 days of treatment with 10 mM DMAT when the cells were co-cultured with stroma, but viable cells remained at 30 h. In contrast, when the cells were treated with DMAT in the absence of stromal support, almost no viable cells remained after 30 h of drug treatment (Figure 2b). This shows that as a monotreatment, DMAT is able to effectively eradicate these malignant lymphoblasts. Because of the potential problems of emerging drug resistance, therapy for human cancer is moving away from drug monotherapy, and towards treatment with more than one drug. Imatinib is the treatment of choice for CML, and therefore we evaluated the combined use of DMAT with Imatinib. As shown in Figure 3a, treatment with 10 mM Imatinib or 10 mM DMAT alone caused a substantial drop in cell viability over a period of
