Gender Differences in Cytidine Deaminase Expression/Activity Contributes to Decreased Cytidine Analog Half-life and Likely Contributes to Worse Outcomes with 5-azacytidine or Decitabine Therapy – Author Reply

We thank Ciccoline et al for their attention and comments regarding “Increased CDA expression/activity in males contributes to decreased cytidine analog half-life and likely contributes to worse outcomes with 5-azacytidine or decitabine therapy”(1). Ciccoline et al have not found a sex difference in CDA activity using HPLC-based measurements of CDA enzyme activity in RBC lysates(2). In addition to use of gemcitabine as a substrate, which has a >10-fold higher K_m for CDA than cytidine(3), this methodology is limited by its analysis of RBC lysates, which may not be representative of CDA activity in pharmacologically more relevant tissues, e.g., the liver. Also pharmacologically relevant are the cancer tissues that are the therapeutic targets of therapy. In contrast to use of a single methodology(2), we evaluated sex-differences in CDA using HPLC-based enyzme activity analysis of heparinized plasma, QRT-PCR quantification of CDA expression in murine liver, mass-spectrometric measurement of decitabine levels (a CDA-drug substrate) in murine plasma, and Western blot evaluation of decitabine pharmacodynamic effect. We had consistent results across these model systems and methods, and furthermore, observed sex-differences in CDA expression in various target cancer tissues analyzed by microarray. Here we provide additional verification: CDA was expressed at significantly higher levels in non-cancerous human liver from human males compared to females (Figure 1A). That expression levels of CDA can be more important than CDA-genotype in determining CDA enzyme activity (1) is also illustrated by substantially greater CDA enzyme activity in U937 cells, that express higher levels of the enzymatically less active ‘CC’ CDA variant, compared to cell-lines that contain more active ‘AA’ and ‘AC’ CDA variants expressed at lower levels (Figure 1B).

Figure 1.

A) CDA expression in human liver. Non-cancerous tissue isolated at therapeutic segmentectomy. CDA mRNA measured by QRT-PCR as previously described (primers available on request)(1). RNA amounts normalized with GAPDH as the reference gene. Two independent measurements were performed on each sample. p-value Wilcoxon Test. B) CDA enzyme activity, measured as previously described (1), in cell lines containing the CC, AC and AA variants of CDA. C) CDA protein expression in the same cell lines. CDA protein expression by immunofluoroscence (#365292, Santa Cruz Biotechnology, Santa Cruz, CA).

We agree that our evaluation of clinical impact was limited by its retrospective nature, and sex-based dosage adjustment would be premature. Even so, the clear difference in MDS treatment outcomes despite similar disease biology (cytogenetics and myeloblasts) in the two sexes was highly provocative, and consistent with epigenetic therapeutic effects that by being S-phase dependent, must interact with brief drug half-lives that are further abbreviated by higher CDA-enzyme activity in males. Furthermore, clinical trial results and mathematical modeling indicate that the negative impact of lower drug exposure time is exacerbated in low S-phase fraction malignant disease (e.g., MDS)(1). Given that marginal differences in exposure time can determine treatment success or failure with 5-azacytidine or decitabine(4–5), pharmacodynamic or other dosage guidance is clearly a worthy goal.