We recruited Iraqi women with postmenopausal status diagnosed with BC according to the National Comprehensive Cancer Network (NCCN) clinical practice guideline.¹⁴ These women were randomly recruited from the oncology department of Imam Al-Hussein Medical City after obtaining consent.

Women aged 50-78 years old received PTX as monotherapy at 80 mg/m², administered as a 1-hour intravenous infusion every week for at least two cycles, with no other concurrent diseases. The patients were included only if their medical records were available.

Exclusion criteria comprised premenopausal women diagnosed with breast cancer; patients with a history of cardiovascular disease, diabetes mellitus, hepatic dysfunction, or hyperthyroidism; and patients taking medications known to affect the pharmacokinetic or pharmacodynamic properties of paclitaxel, such as ketoconazole, erythromycin, rifampin, phenytoin, abaloparatide, abametapir, abatacept, and abciximab.

This cross-sectional observational study was conducted from December 2024 to March 2025 and involved 150 women with BC who received PTX. These women were assessed individually in the second week of PTX therapy using a questionnaire to obtain demographic data, including age and body mass index, and the probability and severity of PTX's side effects. At the same time, neutrophilia count and BC markers were assessed.

Approximately 6 mL of blood was withdrawn from each patient and divided into two parts: 2 mL was placed in an EDTA tube for the absolute neutrophil count and genetic analysis, and 4 mL was placed in a gel tube for serum collection to assess BC markers.

The serum CA15.3 and CEA levels were measured using the chemiluminescent microparticle immunoassay technology¹⁶ with the aid of the ARCHITECT i1000SR immunoassay analyzer and CA15.3 and CEA kits (Abbott, USA).

Differential complete blood count was performed to assess haematological parameters, particularly the ANC, a critical indicator of bone marrow function and chemotherapy-induced myelosuppression in BC patients receiving PTX therapy. This test was performed using an automated Swelab Alfa Haematology Analyser (Boule Diagnostics AB, Sweden).

The severity of PTX-induced neutropenia is commonly graded using a scale derived from the National Cancer Institute Common Toxicity Criteria. This scale grades neutropenia into four scores based on the ANC: score 1, ANC of ≥ 1.5 to <2×10⁹/L; score 2, ANC of ≥1.0 to <1.5×10⁹/L; score 3, ANC of ≥0.5 to <1.0×10⁹/L; and score 4, ANC <0.5×10⁹/L.¹⁷

The PTX-induced oral mucositis was assessed using the World Health Organization scale, which combines both clinical and functional/symptoms-based examination. Oral mucositis was graded on a 5-point scale: 0, no oral mucositis; 1, erythema and soreness; 2, ulcers with ability to eat solids; 3, ulcers requiring a liquid diet; and 4, ulcers with alimentation not possible.¹⁸

The severity of paclitaxel-induced peripheral neuropathy was graded according to the WHO Common Toxicity Criteria for Peripheral Neuropathy. The criteria utilize a 5-grade scale: 0 (none), 1 (mild paresthesia and/or loss of reflexes), 2 (moderate paresthesia and/or mild objective weakness), 3 (severe or intolerable paresthesia and/or marked motor impairment), and 4 (life-threatening or paralysis).¹⁹

Genomic DNA was isolated from whole-blood samples using the AddPrep Genomic DNA Extraction Kit (AddBio, Korea) following the manufacturer's instructions. The allele-specific PCR method was utilized to identify CYP2C8 (rs10509681) gene polymorphisms, employing the Mastercycler Gradient Thermal Cycler (Eppendorf, Germany). This is a widely used SNP genotyping approach that offers superior specificity and is cost-effective in distinguishing alleles based on single-nucleotide variations.²⁰

For each SNP, two allele-specific reverse primers (one for each allele) and one common forward primer were built according to Dr Hassan Abo Almaali. The primer sequences for this gene are forward 5-'GAAGACAGGGTGCTCTGGA-3' and reverse (Allele T) 5'-TCCGTGCTACATGATGACAA-3', and (Allele C) 5'-TCCGTGCTACATGATGACAG-3', with a product size of 821 bp. Each PCR reaction was conducted in a total volume of 25 µL, comprising 3 µL of DNA template, 2 µL of each forward and reverse primer, 5 µL of 5× Master Mix (AddBio, Korea), and 13 µL of distilled water. The PCR parameters for rs10509681 included a duplicate run, with an initial denaturation at 95 °C for 5 minutes, followed by 35 cycles comprising 25 seconds at 95 °C, 25 seconds at 55 °C for annealing, and 1 minute at 72 °C, culminating in a final extension at 72 °C for 5 minutes.

After amplification, 5 µL of each PCR product was combined with 3 µL of loading dye and applied to a 1.5% agarose gel prepared with 1× TBE buffer and 0.5 µg/mL ethidium bromide. Electrophoresis was conducted at 100 V for 35 minutes. DNA bands were visualised under UV light with a transilluminator and photographed using a digital camera. Band sizes were estimated by comparing them to a 100-1000 bp DNA ladder. Two independent evaluators validated all outcomes.

The Statistical Package for the Social Sciences (SPSS 26) was used to perform statistical analysis. The Shapiro-Wilk test was used first to assess the normality of data distribution, guiding subsequent test selection. For normally distributed numerical data, results are reported as mean ± SD. For non-normally distributed data, including BC markers, data are presented as median ± IQR and analyzed with nonparametric tests, specifically the Kruskal-Wallis test with post hoc-Dunn's test. Non-numerical categorical data, including PTX-related adverse effects, are presented as counts and percentages, and analyzed using the Chi-square test. The allele distribution of the CYP2C8 T>C genes was analysed with the Hardy-Weinberg equation and tested using nonparametric tests as well as the Chi-square test. Bivariate correlation-Pearson test was employed to assess the association of CYP2C8 C>A/T rs11572080 genetic variation with the efficacy and safety of PTX. A p-value less than 0.05 was considered statistically significant.

The mean age and BMI of women with BC were found to be 52.02 years ± 10.66 and 29.36 Kg/m² ± 5.15, respectively.

The serum levels of BC markers were measured to investigate the efficacy of PTX, and the medians of these markers (CA 15.3 and CEA) were about 17.95 U/ml and 2.37 ng/ml, respectively.

The neutropenia induced by PTX was observed in varying severities in all BC women, with score 3 being predominant. The mucosa erythema, soreness, and difficulty eating were common characteristics of oral mucositis observed in most women with BC treated with PTX, and scores 1 and 2 were highly prevalent among those patients. Scores 1 and 2 of peripheral neuropathy, characterised by paraesthesia, weakness, and reduced tendon reflexes, were commonly diagnosed in women with BC who were treated with PTX, as presented in Table 1.

The serum levels of BC markers, including CA 15.3 and CEA, were measured to investigate the effect of different CYP2C8 T>C genotypes on the PTX efficacy. The serum levels of CA 15.3 were significantly lower in Iraqi women with BC, who carried either the CC or TC genotypes compared to those with the TT genotype, at P < 0.05.

The serum levels of CEA were also significantly lower in women who carried either the TT or TC genotype compared to those with the CC genotype, at P < 0.05, as shown in Table 3. There was a significant association between the CYP2C8 T>C genetic variation and the serum level of BC markers, as explained in Table 4.

The high frequency of PTX-related adverse effects, including neutropenia, oral mucositis, and peripheral neuropathy, was assessed to identify the possible impacts of different CYP2C8 T>C genotypes on the PTX safety.

The frequency of neutropenia was significantly higher in women carrying either the TC or CC genotype compared to those with the TT genotype, and scores 3 and 4 were commonly observed in these women at P < 0.05. This indicates an increase in the concentration of PTX in the blood due to impaired CYP2C8-mediated drug metabolism. This suggests that PTX affects neutrophil proliferation. Oral mucositis in scores 2 and 3 appeared more frequently in women with either the TC (53%) or CC (72%) genotype compared to those with the TT (28%) genotype at P<0.05.

Scores 3 and 4 of peripheral neuropathy were significantly more common in women with either the TC or CC genotype compared to those with the TT genotype at P < 0.05, as shown in Table 5. There was a significant association between the CYP2C8 T>C genetic variation and the frequency of neutropenia and oral mucositis, as indicated in Table 6.

We acknowledge that the cross-sectional observational design limits our ability to establish causal relationships between genotype and treatment response. We were unable to collect additional variables, which may limit our analysis.