To the editor:

Recently, Dulucq and coworkers reported that multidrug resistance gene (MDR1) polymorphisms were associated with major molecular responses (MMRs) to standard-dose imatinib in chronic myeloid leukemia (CML).1  Significantly more patients homozygous for allele 1236T achieved a MMR. Patients with this genotype had also higher imatinib concentrations. These findings suggested a better response in patients with the TT genotype.

We studied the same most prevalent ABCB1 gene single nucleotide polymorphisms (SNPs), C1236T, G2677T/A, and C3435T, with respect to molecular response in a cohort of 46 early chronic phase CML patients receiving high-dose imatinib (800 mg) according to the Hemato-Oncologie voor Volwassenen Nederland (HOVON)–51 protocol.2,3  No significant differences in baseline characteristics including age, Sokal risk score, and dose of cytarabine were apparent among the different MDR1 genotypes. Genotyping was performed using Taqman allelic discrimination assays on an ABI Prism 7000 Sequence detection system (Applied Biosystems), with 2 allele-specific minor groove binding probes for each SNP. The distribution of the allelic variants of each of the 3 SNPs were in Hardy-Weinberg equilibrium. However, each combination of 2 SNPs was in strong linkage disequilibrium (P < .001). The overall incidences of a MMR and complete molecular response (CMR) were, respectively, 78% and 41% at 2 years from diagnosis, after a median follow-up of 46 months (range, 32-60 months). Among the patients homozygous for 1236C, a cumulative incidence of MMR of 92% after 1 year was observed, compared with 52% and 50% in patients with genotype CT or TT, respectively (P = .02; Table 1), which effect remained statistically significant in multivariate analysis. Hazard ratios to achieve a CMR for patients harboring the CT and TT alleles were, in multivariate analysis, respectively, 0.25 (0.10-0.63) and 0.27 (0.08-0.97), P = .01, indicating a 4-fold reduction of the probability to obtain a CMR. Also patients homozygous for 3435T and 2677T showed lower probabilities to obtain a MMR and CMR (Table 1). Summarizing, molecular response in CML patients receiving high-dose imatinib strongly depended on SNP-genotype, with the TT-genotype associated with worse response. These findings may be explained by enhanced clearance of imatinib by the 2677TT genotype, which leads to an amino acid substitution and thereby increased transport activity.4  Enhanced clearance significantly associated with the TT-genotype and less dose reduction was reported earlier by Gurney et al.5  In addition, recent clinical findings by Ni et al are also in agreement with these results, who studied cytogenetic resistance in patients treated with 400 mg imatinib.6  These observations are in contrast with the study by Dulucq et al.1  Strikingly, responses in the French study were strongly related to imatinib plasma levels.1  Therefore, pharmacokinetic resistance due to a variety of mechanisms affecting plasma levels rather than tumor cell resistance due to P-glycoprotein activity may be suggested as an alternative explanation for these results.

Table 1

Cumulative incidence of molecular response, by genotype

Allelic variantNo. (%)Major molecular response (MMR)
Complete molecular response (CMR)
At 1 y, % (SE)Univariate
Multivariate
At 1 y, % (SE)Univariate
Multivariate
HR (95% CI)P*HR (95% CI)P*HR (95% CI)P*HR (95% CI)P*
All patients 46 59 (7)     22 (6)     
C1236T, n = 43            
    CC 12 (28) 92 (8) .02 .03 42 (14) .06 .01 
    CT 23 (53) 52 (10) .32 (.14-.71)  .31 (.13-.71)  13 (7) .36 (.16-.82)  .25 (.10-.63)  
    TT 8 (19) 50 (18) .33 (.12-.89)  .40 (.14-1.15)  25 (15) .48 (.17-1.39)  .27 (.08-.97)  
G2677T, n = 41            
    GG 10 (24) 90 (9) .11 .21 40 (15) .13 .05 
    GT 23 (56) 61 (10) .45 (.20-1.05)  .49 (.20-1.16)  22 (9) .46 (.20-1.08)  .33 (.13-.85)  
    TT 8 (20) 38 (17) .36 (.13-.99)  .42 (.15-1.21)  13 (12) .35 (.11-1.13)  .23 (.06-.88)  
C3435T, n = 44            
    CC 10 (23) 90 (9) .04 .06 50 (16) .10 .04 
    CT 24 (55) 58 (10) .37 (.16-.83)  .35 (.15-.82)  17 (8) .39 (.17-.92)  .33 (.13-.81)  
    TT 10 (23) 40 (15) .31 (.12-.83)  .35 (.12-.98)  10 (9) .39 (.13-1.17)  .24 (.07-.83)  
Allelic variantNo. (%)Major molecular response (MMR)
Complete molecular response (CMR)
At 1 y, % (SE)Univariate
Multivariate
At 1 y, % (SE)Univariate
Multivariate
HR (95% CI)P*HR (95% CI)P*HR (95% CI)P*HR (95% CI)P*
All patients 46 59 (7)     22 (6)     
C1236T, n = 43            
    CC 12 (28) 92 (8) .02 .03 42 (14) .06 .01 
    CT 23 (53) 52 (10) .32 (.14-.71)  .31 (.13-.71)  13 (7) .36 (.16-.82)  .25 (.10-.63)  
    TT 8 (19) 50 (18) .33 (.12-.89)  .40 (.14-1.15)  25 (15) .48 (.17-1.39)  .27 (.08-.97)  
G2677T, n = 41            
    GG 10 (24) 90 (9) .11 .21 40 (15) .13 .05 
    GT 23 (56) 61 (10) .45 (.20-1.05)  .49 (.20-1.16)  22 (9) .46 (.20-1.08)  .33 (.13-.85)  
    TT 8 (20) 38 (17) .36 (.13-.99)  .42 (.15-1.21)  13 (12) .35 (.11-1.13)  .23 (.06-.88)  
C3435T, n = 44            
    CC 10 (23) 90 (9) .04 .06 50 (16) .10 .04 
    CT 24 (55) 58 (10) .37 (.16-.83)  .35 (.15-.82)  17 (8) .39 (.17-.92)  .33 (.13-.81)  
    TT 10 (23) 40 (15) .31 (.12-.83)  .35 (.12-.98)  10 (9) .39 (.13-1.17)  .24 (.07-.83)  

SE indicates standard error; HR, hazard ratio; and CI, confidence interval.

*

P values are for the comparison of the molecular response rate between the different allelic variants of each genotype.

Adjusted for Sokal risk group and cytarabine dose.

In conclusion, molecular resistance in CML patients receiving high-dose imatinib appeared strongly associated with the TT-genotype of ABCB1, suggesting a role for P-glycoprotein–mediated drug efflux in residual malignant hematopoietic progenitor cells that may possibly account for persistent molecular residual disease.

Acknowledgments: This work was supported by the Queen Wilhelmina Fund (KWF) Kankerbestrijding for support of data management. Novartis Oncology Netherlands provided support for the standardization and centralization of real-time quantitative polymerase chain reaction.

Contribution: W.D., B.v.d.H., J.J.W.M.J., G.J.O. and J.J.C. were responsible for the initial design of present analysis, actual evaluation, and writing the paper All authors were responsible for the design of the HOVON study, treatment of patients, critical review of the paper, suggestions for additional analysis, and writing the final paper. W.D. and B.v.d.H. contributed equally to the manuscript.

Conflict-of-interest disclosure: P.S., G.J.O., and J.J.C. have received consulting fees from Novartis Oncology. The remaining authors declare no competing financial interests.

Correspondence: J.J. Cornelissen, PhD, MD, Erasmus University Medical Center, Department of Hematology, Groene Hilledijk 301, 3075 EA Rotterdam, The Netherlands; e-mail: j.cornelissen@erasmusmc.nl.

1
Dulucq
 
S
Bouchet
 
S
Turcq
 
B
, et al. 
Multidrug resistance gene (MDR1) polymorphisms are associated with major molecular responses to standard-dose imatinib in chronic myeloid leukemia.
Blood
2008
, vol. 
112
 
5
(pg. 
2024
-
2027
)
2
Deenik
 
W
van der Holt
 
B
Verhoef
 
GEG
, et al. 
Dose-finding study of imatinib in combination with intravenous cytarabine: feasibility in newly diagnosed patients with chronic myeloid leukemia.
Blood
2008
, vol. 
111
 
5
(pg. 
2581
-
2588
)
3
Deenik
 
W
Janssen
 
JJWM
van der Holt
 
B
, et al. 
Efficacy of escalated imatinib combined with cytarabine in newly diagnosed patients with chronic myeloid leukemia.
Haematologica
2010
, vol. 
95
 
6
(pg. 
914
-
921
)
4
Kim
 
RB
Leake
 
BF
Choo
 
EF
, et al. 
Identification of functionally variant MDR1 alleles among European Americans and African Americans.
Clin Pharmacol Ther
2001
, vol. 
70
 
2
(pg. 
189
-
199
)
5
Gurney
 
H
Wong
 
M
Balleine
 
RL
, et al. 
Imatinib disposition and ABCB1 (MDR1, P-glycoprotein) genotype.
Clin Pharmacol Ther
2007
, vol. 
82
 
1
(pg. 
33
-
40
)
6
Ni
 
LN
Li
 
JY
Miao
 
KR
, et al. 
Multidrug resistance gene (MDR1) polymorphisms correlate with imatinib response in chronic myeloid leukemia [published online ahead of print March 4, 2010].
Med Oncol