In Response: The primary aim of our report was to describe and validate a measurable molecular variable that could explain the metabolic conversion of irinotecan (CPT11) to SN38, the pharmacologically active metabolite which exerts chemotherapeutic and toxic effects. In particular, expression of carboxylesterase isoform 2 (CES2) mRNA in peripheral blood mononuclear cells (PBMC) was investigated. Determining the relationship of CES2 mRNA to other pharmacokinetic variables (AUCSN38, AUCCPT11, and AUCSN38-glucuronide) and pharmacodynamic effects (toxicity and response to chemotherapy) was a secondary aim of the study. The results referring to the secondary aim, suggested, but did not conclusively prove a correlation. As stated in the Discussion, the possible correlation must be confirmed by further investigation in a larger number of patients.

In response to the issues raised in the letter by Lim et al., the following comments are offered:

  1. Only very small amounts of CES2 are produced in PBMCs. This was why we decided to adopt RT-PCR as an amplifying test to measure the gene expression. To our knowledge, there are no published data that compare CES2 levels in colon, liver, or PBMCs. Nonetheless, the possibility of having a noninvasive method to predict the systemic metabolism of CPT11 made it worthwhile to test PBMCs. There are other published examples that demonstrate the predictive value of PBMCs as an indicator of systemic metabolism (1). Therefore, we explored the possibility that CES2 expression in PBMCs, although not exclusively responsible for CPT11 activation, was representative of systemic activation.

  2. Even if CPT11 is a poor substrate, CES2 is the enzyme primarily responsible for CPT11 prodrug activation in humans, and this is the rate-limiting metabolic step for SN38 bioavailability. In fact, it has been reported that different amounts of CES2 result in different extents of CPT11 activation (2). As noted above, we tried to evaluate CES2 activity in PBMCs. It was detectable, but the low sensitivity of the test precluded accurate estimation of differences between individuals. This was why we changed to mRNA analysis by RT-PCR.

  3. We were fully aware of the important role of other enzymes of CPT11 metabolism, including UGT1A1, UGT1A7, UGT1A9, and CYP3A4/5. This was presented in the Introduction, and discussed elsewhere in our article. Moreover, in the concluding remarks, we specifically recommended that CES2 mRNA expression should be included in a comprehensive set of markers predictive of irinotecan pharmacokinetics and pharmacodynamics. However, we must emphasize that the research in our article was focused on the activation step that is mainly mediated by CES2.

  4. The activation ratio was calculated as [(AUCSN38 + AUCSN38G) / (AUCCPT11)], where SN38 and SN38G represent free SN38 and SN38 glucuronide, respectively. We believe that considering only the level of free SN38 could bias the estimation of the activation process, since the free SN38 in plasma may strongly depend on the efficiency of the glucuronidation reaction, which has been reported to be influenced by many factors, including the genetic makeup of the patients. Because of the differences in physicochemical properties, CPT11 and its metabolites are likely to be distributed differently in the body. However, in agreement with other studies (3), we found that SN38 and SN38G show very similar patterns of formation, as well as a parallel slow decay in plasma. This supported our rationale for using the sum of free SN38 and SN38G instead of free SN38 alone, to model the pharmacokinetics of CPT11 activation.

  5. The contribution of 5-fluorouracil to toxicity probably deserved more emphasis. Nonetheless, neutropenia and diarrhea, although also attributable to the use of 5-fluorouracil, are mainly associated with CPT11 therapy, and constitute an important dose-limiting toxicity for this drug. Other studies, in which patients were treated with association regimens that included CPT11 and 5-fluorouracil, showed a good correlation between molecular variables (such as the UGT1A1*28 polymorphism that is exclusively associated with CPT11 metabolism), and neutropenia or diarrhea (4). Our findings showed an association trend between CES2 mRNA expression in PBMCs, and toxicity caused by the treatment. We also pointed out that this trend should be confirmed by further studies.

1
Hase I, Imaoka S, Oda Y, et al. Area under the plasma concentration-time curve of inorganic fluoride following sevoflurane anesthesia correlates with CYP2E1 mRNA level in mononuclear cells.
Anesthesiology
 
2000
;
92
:
1661
–6.
2
Xu G, Zhang W, Ma MK, McLeod HL. Human carboxylesterase 2 is commonly expressed in tumor tissue and is correlated with activation of irinotecan.
Clin Cancer Res
 
2002
;
8
:
2605
–11.
3
Gupta E, Lestingi TM, Mick R, et al. Metabolic fate of irinotecan in humans: correlation of glucuronidation with diarrhea.
Cancer Res
 
1994
;
54
:
3723
–5.
4
Iyer L, Das S, Janisch L, et al. UGT1A1*28 polymorphism as a determinant of irinotecan disposition and toxicity.
Pharmacogenomics J
 
2002
;
2
:
43
–7.
American Association for Cancer Research
2006