Primary AML may be induced by environmental toxins in the same way as chemotherapy-induced genotoxic damage leads to therapy-related AML (t-AML). The genotypes of MDR1 and CYP3A4 genes may be important in predisposing to both primary and t-AML. Hoffmeyer et al (PNAS, 2000) found that a silent single nucleotide polymorphism, C3435T was associated with higher plasma concentrations of digoxin after oral dosing. The T allele increases effective exposure to toxins, and may predispose to renal tumors. We previously reported (ASH 2002) an association of t-AML with C3435T in a small sample (17 pts). The cytochrome P450 (CYP) enzymes including CYP3A4 metabolize drugs and toxins, and can activate procarcinogens. Some studies found that an A to G substitution in the CYP3A4 promoter (CYP3A4*1B) “protects” against t-AML, while the wild type allele was associated with t-AML (including our study, ASH 2002).

Aims: To determine if either the MDR1 or CYP3A4 polymorphism affects predisposition to develop primary AML, and to determine ethnic differences in the frequency of the polymorphisms.

Methods: The MDR1 C3435T polymorphism was analyzed using PCR and Sau3A1 digestion. CYP3A*1B was studied using PCR and hybridization to a wild type or mutant 19-mer oligonucleotide probe. We studied primary and t-AML patients (68 Jewish and 43 Arab for C3435T, and 81 Jewish and 37 Arab for CYP3A1*1B), and control DNA samples of the same ethnic origin (58 Caucasian Jewish and 44 Arab and 88 Ethiopian Jews).

Results: The TT genotype at position 3435 of MDR1 was found more frequently in AML patients (both primary and t-AML) than in controls (p value: 0.0302). There was an additive effect of the T allele, with fewest AML patients having the CC genotype, more with CT and the most with the TT genotype (P value: 0.0015). The relative risk of primary AML is 0.5975 (C.I- (0.3135, 1.1389) for patients with the CC genotype versus those with CT/TT. The relative risk of t-leuk is 0.1180 (C.I- (0.0073 1.8952) for patients with the CC genotype versus those with CT/TT. Of 23 t-AML patients analyzed, none had the CC genotype. Because the allele may affect response to therapy, we analyzed survival. Using Kaplan-Meier analysis, we found no statistically significant difference in the survival of AML patients of different MDR1 C3435T genotypes. Among controls, there was a statistically significant difference in the frequency of the different genotypes among Arabs, Caucasian Jews and Ethiopian Jews. Arab individuals had a lower frequency of the CC genotype and more frequent TT, Ethiopian Jews had a high frequency of the CC genotype (p value: 0.0072). Among AML patients of different ethnicity, there was no difference in genotype. The CYP3A*1B polymorphism was found in approximately 15% of the Israeli population (Caucasian Jews and Arabs alike). In comparison, 52% of Ethiopian Jews carried CYP3A*1B. 9/118 primary AML patients (7%) carried CYP3A*1B and only 1 (2.4%) of 41 t-AML patients was found to carry CYP3A4*1B (p value < 0.05).

Conclusions: The MDR1 C3435T polymorphism is associated with a higher risk of developing primary as well as t-AML. Arabs may be at higher risk for AML due to a high frequency of the T allele. In addition, the CYP3A4*1B polymorphism might be protective against primary AML, and we confirm previous reports that it may protect against t-AML in adults. Lastly, Ethiopian Jews have a high frequency of the C allele of the MDR1 C3435T polymorphism, and high frequency of CYP3A4*1B, which may underlie the relatively low incidence of AML in this ethnic group in Israel.

Author notes

Corresponding author

Sign in via your Institution