MTX-induced white matter changes are associated with polymorphisms of methionine-metabolism

Objective: CNS demyelination presenting with CNS white matter changes (WMC) in MRI is a typical side-effect of high-dose methotrexate (MTX) chemotherapy. We speculated that the underlying mechanism may be an inhibition of methionine synthesis due to a MTX-induced lack of 5,10-methylenetetrahydrofolate. We studied whether development of WMC in high-dose MTX-treated 42 patients with primary CNS lymphoma (PCNSL) is associated with functional polymorphisms which might modify the effect of MTX on methionine synthesis: methionine synthase (MTR) c.2756A>G (D919G), 5,10-methylenetetrahydrofolate reductase (MTHFR) c.677C>T (A222V) and transcobalamin 2 (Tc2) c.776C>G (P259R).

Methods: WMC was defined as newly evolving confluent symmetrical hyperintensity of the supratentorial white matter on FLAIR and T2-weighted images as shown in figure 1 (the arrow indicates the defect resulting from the implantation of a reservoir-catheter for intraventricular drug application). Genetic analysis was performed by PCR amplification of genomic DNA followed by restriction enzyme digestion. In combination, a „risk haplotype“ was defined as presence of at least one of the polymorphisms. For statistic analyses, the Chi2 and the exact Fisher test were applied in addition to logistic regression analysis to determine effects of gender and age.

Results: 17 (42%) patients developed WMC. Frequency of WMC was higher among men (not significant) and elder patients (significant; table 1). Each of the three polymorphisms was over-represented among patients with WMC, but after correction for multiple testing, none of the polymorphisms alone referred a significant association with WMC. However, the combined „risk haplotype“ (prevalence of at least one of these polymorphisms) was significantly associated with WMC: It was found in 7 of the 25 (28%) patients without and in 14 of the 17 (82%) patients with WMC (Chi-squar=12.0 and p=0.001; OR=12.0; CI95%=2.2–75.9). Only 3 of the 21 (14%) patients without the risk haplotype“ developed WMC, as opposed to 14 of the 21 (67%) with the risk haplotype. The relative risk of WMC conferred by the risk haplotype was 4.7 (CI95%=1.6–18.4). Logistic regression analysis revealed that this effect did not depend on gender or age (not shown).

Conclusion: These data suggest that the three polymorphisms involved in methionine metabolism synergistically promote neurotoxicity of MTX, lowering methionine synthesis and thereby disposing to WMC in MTX-treated patients.

	Patients without WMC* (n=25)	Patients with WMC* (n=17)	Chi2 (p: Pearson); df=2	OR (CI95%)
Male gender	8 (0.32)	10 (0.59)	2.97 (0.085)	3.04 (0.71–13.51)
Mean age±sd (median); in years	54±12 (55)	62±6 (63)	Test, df=2; p=0.038
MTHFR c.677C>T: TT	3 (0.12)	6 (0.35)	3.3 (0.071)	4.0 (0.69–25.7)
MTR c.2756A>G: AG/GG	4 (0.16)	5 (0.29)	1.1 (0.300)	2.19 (0.40–12.5)
Tc2 c.776C>G:GG	3 (0.12)	7 (0.41)	4.7 (0.029)	5.1 (0.90–25.7)
Combinde risk haplotype	7 (0.28)	14 (0.82)	12.0 (0.001)	12.0 (2.2–75.9)