Determination of Capecitabine and Its Metabolites in Plasma of Egyptian Colorectal Cancer Patients

Colorectal cancer (CRC) is a major global health concern, with its incidence continuing to rise worldwide. Clinical manifestations vary widely, ranging from changes in bowel habits to nausea and vomiting. Chemotherapy remains a mainstay of CRC treatment, particularly with agents such as capecitabine (CCB) and 5-fluorouracil (FU). Capecitabine, an orally administered prodrug of FU, is often preferred because it does not require hospitalization for administration. It undergoes a sequential three-step enzymatic conversion: first to 5′-deoxy-5-fluorocytidine (DFCR) by carboxylesterase (CES), then to 5′-deoxy-5-fluorouridine (DFUR) by cytidine deaminase (CDD), and finally to the active form, FU, via thymidine phosphorylase (TP). FU is subsequently inactivated to 5,6-dihydro-5-fluorouracil (DHFU).

Despite standardized dosing, patients frequently display variability in therapeutic response and adverse effects during CCB therapy, likely due to single nucleotide polymorphisms (SNPs) in the enzymes involved in its activation and deactivation pathways. The most common toxicities include hand–foot syndrome and diarrhea.

This study aimed to develop and validate a novel HPLC-UV method for the simultaneous quantification of CCB and its metabolites in human plasma, followed by a correlation study between metabolite concentrations and treatment-related toxicities. A quality-by-design approach using DryLab® 2000 software enabled optimization of the chromatographic conditions, achieving baseline separation of six analytes within a 10-minute gradient run. The method demonstrated excellent linearity over the range of 1–500 µg/mL, with high accuracy, precision, and robustness. Limits of detection (LOD) were 3.0 ng/mL for DHFU and CCB, and 0.3 ng/mL for DFUR, DFCR, and FU; limits of quantification (LOQ) were 10.0 ng/mL for DHFU and CCB, and 1.0 ng/mL for DFUR, DFCR, and FU.

Clinical application revealed significant correlations between DFCR levels and mucositis, and between DFUR levels and hand–foot syndrome, suggesting that this validated analytical method can serve as a predictive tool for CCB-related toxicities, offering a practical alternative to genotyping approaches.

Column and Method Optimization Strategy

To achieve optimal resolution and minimal retention time for CCB and its metabolites, three stationary phases were evaluated sequentially: Hypersil Gold C18, Hypersil Gold C8, and Synergi Fusion-RP.

Given the high polarity of certain metabolites, ion-pair chromatography was employed using tetraalkylammonium salts in the aqueous mobile phase to improve retention and selectivity.

Gradient elution started with 5% methanol (B) in phosphate buffer at pH 2.6 (A), with an initial runtime of 20 min. Optimization simulations using DryLab® varied temperature (25–45 °C) and gradient time (tG). The optimal conditions were obtained at 45 °C and the shortest tG, yielding efficient resolution and minimal runtime.

Final optimized HPLC conditions:

• Mobile Phase A: Acidified water (pH 2.6, phosphate buffer)
• Mobile Phase B: Methanol
• Gradient: 5–100% B (stepwise) over 20 min
• Flow Rate: 1.0 mL/min
• Injection Volume: 25 µL
• Column Temperature: 25 °C
• Detection Wavelengths: 205, 266, and 310 nm.

 Method Validation

The method was validated according to ICH Q2(R1) guidelines for linearity, precision, accuracy, selectivity, and sensitivity.

Calibration ranges:

• DHFU and CCB: 10–20,000 ng/mL
• DFUR and DFCR: 1–20,000 ng/mL
• FU: 1–100,000 ng/mL

Five calibration curves per analyte were constructed on three separate days. Precision and accuracy were assessed using spiked plasma samples and expressed as %CV and %Bias, respectively.

 %Bias = [(measured value − true value)/true value] × 100.

Limits:

• LOD: Signal-to-noise ratio = 3
• LOQ: Signal-to-noise ratio = 10

Selectivity was verified by comparing blank plasma with spiked plasma chromatograms. During routine validation, FU, DFCR, and DFUR were prioritized at 1 × 10³, 1 × 10⁴, and 2 × 10⁴ ng/mL as they represent the activation pathway of CCB.

Plasma Extraction and Patient Application

Sampling and Ethics

The study included 20 colorectal cancer patients. Blood samples (2–3 mL) were collected in EDTA tubes in triplicate at:

• Day 1 – Cycle 1
• Day 14 – Cycle 1
• Day 1 – Cycle 2 (≈ 2 h after oral capecitabine)

The study protocol was approved by the Ethics Committee of the German University in Cairo and the IRB of the National Cancer Institute (NCI, Egypt). Written informed consent was obtained from all participants.

Patients were followed throughout therapy to assess compliance and monitor for adverse effects (notably hand–foot syndrome and mucositis).

Extraction Procedure

The plasma extraction followed a protein precipitation method (adapted from Vainchtein et al.):

1. Add 20 µL of internal standard (CLU, 100 µg/mL) to 160 µL plasma.
2. Vortex for 10 s.
3. Add 20 µL of 99% (v/v) TFA, vortex 1 min at 1400 rpm.
4. Centrifuge 20 min at 14,000 rpm, 4 °C.
5. Transfer supernatant to glass autosampler vials for analysis.

This method ensures rapid processing, high recovery rates, and excellent reproducibility for plasma quantification.