Genetic variations in CYP2A6 predict efficacy of neoadjuvant letrozole concurrent with chemotherapy in locally advanced hormone receptor-positive breast cancer

Abstract

Purpose

We conducted a retrospective exploratory analysis of prospective cohort to investigate the impact of CYP2A6 metabolizer status on predicting the efficacy and prognosis of neoadjuvant chemotherapy combined with letrozole.

Method

Three allelic variants that related to the abolished or reduced activity of CYP2A6 enzyme and 7 other single nucleotide polymorphisms (SNPs) in the CYP2A6 locus were examined from whole blood samples before neoadjuvant treatment, and patients were classified into 3 metabolizer status subgroups (‘normal’, ‘intermediate’ and ‘slow’ metabolizers) according to CYP2A6 genotypes.

Result

Our study included 56 patients received neoadjuvant chemotherapy combined with letrozole. The pCR rate in the CYP2A6 intermediate/slow metabolizers (29.2%) was higher than that in the normal metabolizers (12.5%; p = 0.046). Survival analysis suggested CYP2A6 intermediate/slow metabolizers may have a slightly lower risk of DFS events compared to normal metabolizers. These findings suggested that CYP2A6 metabolizer status might play a crucial role in predicting the efficacy and prognosis of adding letrozole to neoadjuvant chemotherapy.

Conclusion

The CPY2A6 gene polymorphism may be a novel potential predictive and prognostic biomarker for postmenopausal hormone receptor-positive breast cancer patients receiving neoadjuvant chemotherapy combined with letrozole. Our findings have clinical implications for guiding precision endocrine therapy in hormone receptor-positive breast cancer patients.

Key Message

• Two significant outcomes are: CYP2A6 metabolizer status serves as an independent predictive factor for pCR among patients undergoing neoadjuvant chemotherapy concurrently with letrozole.

• CYP2A6 intermediate/slow metabolizers may have a slightly lower risk of DFS events compared to normal metabolizers.

Graphical Abstract

Introduction

Breast cancer is currently the most common malignant tumor, of which 60%–75% are hormone receptor-positive subtype. Letrozole, a third-generation aromatase inhibitor, is a standard of care for postmenopausal women with hormone receptor-positive early breast cancer (eBC), markedly reducing their risk of recurrence and death when compared to tamoxifen [1,2].

Letrozole pharmacologically blocks the aromatase enzyme in the liver, muscle, fat, and bone, thereby inhibiting the conversion of androgen to estrogen and decreasing circulating estrogen to undetectable levels for most postmenopausal women. Further studies have confirmed that letrozole inhibits aromatase activity, modulates the tumor microenvironment, and lowers estrogen content there as well in hormone receptor-positive breast cancer [3–5]. The P024 trial was a study of neoadjuvant endocrine therapy, in which 324 postmenopausal hormone receptor-positive patients with locally advanced breast cancer were given 4 months of neoadjuvant letrozole or tamoxifen, respectively. The letrozole group showed better overall response rate (ORR; 55% vs.. 36%, p < 0.001) and a higher rate of breast conservation surgery (45% vs. 35%, p = 0.022) than the tamoxifen group [6]. However, the mechanism of individual difference in efficacy is still unclear. Therefore, we need to find a way to accurately separate those who respond to letrozole from those who do not at initial diagnosis.

Drug metabolizing enzymes are involved in the biotransformation of drugs in the body, thus allowing them to exert important pharmacological effects. The pharmacological activity of a drug is altered when it is transformed by metabolizing enzymes. Most drugs lose their activity during this process, which is called inactivation. Letrozole is metabolized primarily by the CYP enzyme system and catalyzed to its inactive product CGP 446453 by CYP2A6. The CYP2A6 enzyme, expressed mainly in the liver, is encoded by the CYP2A6 gene located on human chromosome 19q13.2. CYP2A6 gene polymorphisms such as single nucleotide polymorphism (SNP) or CYP2A6 gene deletion mutations may result in altered expression or activity of CYP2A6 metabolizing enzymes [7–9]. Moreover, the blood concentrations of letrozole have been demonstrated to be influenced by CYP2A6 gene polymorphism, with patients in the CYP2A6 slow metabolism group having considerably higher letrozole blood levels than those in the CYP2A6 normal metabolism group [10]. However, the relationship between CYP2A6 gene polymorphisms and letrozole efficacy had not yet been reported and merits further exploration.

Neoadjuvant therapy is an important part of the systemic treatment for breast cancer, and the treatment response of the neoadjuvant therapy provides oncologists information on the prediction of efficacy and prognosis for certain regimens. Previous research has indicated that chemotherapy resistance is promoted in a variety of cancer cells by high estrogen levels or increased expression of estrogen receptor (ER), while cancer cells react better to cytotoxic agents in the context of low estrogen levels or decreased expression of ERs [11,12]. Thus, patients with locally advanced hormone receptor-positive breast cancer who are indicated for neoadjuvant therapy may theoretically benefit from neoadjuvant chemotherapy combined with letrozole, since letrozole has been shown to be capable of reducing estrogen levels [3,5,13]. However, the clinical benefits of adding letrozole to chemotherapy remain controversial, and further research is needed to identify patients who may benefit from this combination. We propose the hypothesis that CYP2A6 genotype status might play an important role in predicting the efficacy and prognosis of neoadjuvant chemotherapy combined with letrozole for breast cancer, which may be a new potential biomarker for neoadjuvant letrozole endocrine therapy in breast cancer.

Method

Patients and study design

We conducted a retrospective study of breast cancer patients from prospective neoadjuvant clinical trial SHPD002 (NCT02221999), who received neoadjuvant therapy at Department of Breast Surgery, Renji Hospital, between July 2014 and October 2018. The main inclusion criteria were female postmenopausal patients, histologically confirmed invasive breast cancer with positive ER or progesterone receptor (PR) and enough blood samples for CYP2A6 genotype detection. The main exclusion criteria were inflammatory breast cancer or metastatic breast cancer.

In the neoadjuvant setting, patients received chemotherapy concurrently with letrozole (2.5 mg, orally, one tablet per day). Trastuzumab was given to HER2-positive patients once a week, with a loading dose of 4 mg/kg and a maintenance dose of 2 mg/kg. For all patients, weekly paclitaxel was intravenously given at a dose of 80 mg/m² on days 1, 8, 15, and 22 of each 28-day cycle and cisplatin at a dose of 25 mg/m² on days 1, 8, and 15 of each 28-day cycle, for a total of 4 cycles. All patients underwent surgery sequentially after neoadjuvant therapy. Adjuvant endocrine therapy was administered to all patients.

The retrospective study was approved by the Independent Ethics Committee of Renji Hospital, School of Medicine, Shanghai Jiao Tong University (approval number, KY2021-095-B).

Collection of specimens and clinicopathological data

About 5 ml of whole blood was prospectively collected from patients before neoadjuvant treatment and stored in EDTA anticoagulant tube at 4 °C. After separation, the specimen was sent to Renji Hospital Biobank and stored at −80 °C for future use.

All clinicopathological data were prospectively collected before neoadjuvant treatment. Body mass index (BMI) was calculated as weight in kilogram divided by the square of height in meter, and the value of 25 was used to split patients into two groups. All tissues were histologically diagnosed by the Department of Pathology, Renji Hospital. ER, PR, Ki-67 and HER2 were examine on formalin-fixed paraffin-embedded tumor samples from biopsy. Hormone receptor positivity was defined as ≥1% tumor cell nuclei stained for ER or PR by immunohistochemistry (IHC), and 10% was used for dichotomization. HER2 positivity was defined as IHC 3+ or amplification by fluorescence in situ hybridization according to the American Society of Clinical Oncology/College of American Pathologists recommendations 2018 [14]. In terms of Ki67 index, we used the value of 20% to subgroup separation.

Deoxyribonucleic acid extraction and polymerase chain reaction assays

DNA extraction was performed using the TIANamp Genomic DNA Kit (Tiangen Biotech Co., Ltd., Beijing, China). The pending genotypes were selected from 1000 Genomes Project and literature findings. A total of 3 classical genotypes [CYP2A6*4, *1 A (51 A) (rs1137115), *9 (rs28399433)] and 7 other SNPs (rs56113850, rs61663607, rs8192720, rs7250713, rs7256108, rs8102683, rs8192728) of CYP2A6 were selected in public database, with minor allele frequency >0.1 in Han population. Genotyping of CYP2A6 was accomplished with several different methods. The CYP2A6*4 allele is characterized as the whole deletion of this gene, and the genotype was detected by 2-step quantitative PCR assays [15]. The first step of PCR was used to amplify the target fragment, and the second step of PCR was used to distinguish CYP2A6*4 from other genotypes. We then assessed additional variants that do not alter the CYP2A6 gene structure, including *1 A (51 A) (rs1137115), *9 (rs28399433), rs56113850, rs61663607, rs8192720, rs7250713, rs7256108, rs8102683 and rs8192728. Genotyping was performed with Nanodispenser Spectro CHIP chip spotting and Mass ARRAY Compact System (Agena Bioscience) by Shanghai Benegene Biotechnology Co., LTD. Primers for PCR and single base extension were designed by using Assay Design Suite V2.0 online (Agena). Detailed primer sequences were provided in Supplementary Table 1. Genotypes were determined by Typer software version 4.0 (Agena) using default settings after auto clustering. Unidentified specimens were used to ensure that all measurements of all clinical outcomes were performed blinded.

Table 1.

CYP2A6 metabolizer status classified by CYP2A6 genotype.

CYP2A6 genotype	CYP2A6 metabolizer status	N (%)
*1/*1	Normal	8 (14.28%)
*1/*9	Intermediate	15 (26.78%)
*1/*1 A(51A)	Intermediate	15 (26.78%)
*1 A(51A)/*1A(51A)	Slow	7 (12.50%)
*1 A(51A)/*9	Slow	1 (1.79%)
*1/*4	Slow	1 (1.79%)
*1 A(51A)/*4	Slow	5 (8.93%)
*9/*4	Slow	1 (1.79%)
*4/*4	Slow	3 (5.36%)

CYP2A6 genotyping

The CYP2A6 classical genotypes, including CYP2A6*4, *1 A (51 A) (rs1137115) and *9 (rs28399433), were related to the abolished or reduced activity of CYP2A6 enzyme [16–18], in which CYP2A6*9 and *1 A (51 A) were considered a ‘decrease of function (D)’ allele whereas *4 was considered a ‘loss of function (L)’ allele. On these premises, the enrolled patients were categorized into three kinds of metabolizers: (1) ‘normal metabolizers’ were defined as having neither a D nor an L allele (i.e. *1/*1); (2) ‘intermediate metabolizers’ were defined as only one D allele [i.e. *1 A (51 A) or *9]; (3) ‘slow metabolizers’ had either an L allele or two D alleles [e.g. *1/*4 or *1 A (51 A)/*1A (51 A)] (Table 1). Besides, each SNP was explored in different comparison models. For example, “C” was the minor allele of the SNP locus for CYP2A6 rs56113850. Both TC and CC were combined and compared against TT as reference for the dominant model. CC was compared against TC plus TT for the recessive model. As for the additive model, both TC vs. TT and CC vs. TT were analyzed with TT used as reference. In the exploratory analysis, in addition to the intermediate/slow metabolizer subgroup, we divided the normal metabolizers into normal and quick metabolizer subgroups based on rs56113850 status: quick metabolizer was defined as patients containing rs56113850 CC genotype, whereas normal metabolizer was defined as those containing rs56113850 TT/TC genotype [19,20].

Outcomes

The outcomes of this study were pCR, defined as the absence of invasive cancer in the breast samples obtained at the time of surgery (ypT0/is), and DFS, denoted as the time from surgery to the first occurrence of local or regional recurrence, contralateral breast cancer, distant metastasis, second primary malignancy or death from any cause. Adverse events (AEs) were assessed during study period and graded according to Common Terminology Criteria for Adverse Events (CTCAE) version 4.01. Multivariate adjustments were made for the following variables: age, tumor size, estrogen receptor (ER) status, progesterone receptor (PR) status, HER2 status, and body mass index (BMI), to account for potential confounding effects in the analysis.

Statistical analysis

The chi-squared test was conducted to compare categorical variables. Multivariate logistical regression was used to derive odds ratios (ORs) with 95% confidence interval (CI) and calculate the association of each genotype with efficacy and toxicities. To provide a quantitative tool to predict the individual probability of accomplishing pCR, we built the nomogram based on the multivariable logistic analysis among all patients included in this study. The predictive performance of the nomogram was assessed within the same cohort using calibration curve and receiver operating characteristic (ROC) curve. The estimated median follow-up was calculated using the reverse Kaplan-Meier method. Disease-free survival rates were compared by Kaplan-Meier curves, examined by log-rank test. Cox proportional hazard regressions were performed to derive hazard ratios (HRs) with 95% CIs. Well-recognized clinicopathological characteristics were incorporated in the multivariate logistic and Cox regression analysis, including age, tumor size, ER, PR, HER2, BMI, and neoadjuvant regimen. To assess the robustness of our study, Propensity Score Matching (PSM) method was used in the sensitivity analysis [21]. All statistical analyses were performed using R language version 3.6.1 (www.r-project.org). All tests were two-sided, and p < 0.05 was considered statistically significant.

Result:

Study population

A total of 56 patients receiving neoadjuvant therapy were enrolled in this study. Of the 56 patients, 22 (39.29%) had HER2-positive breast cancer and received trastuzumab targeted therapy. Among all the patients, 8 (14.29%) were classified into normal metabolizers, 30 (53.57%) were categorized into intermediate metabolizers, and 18 (32.14%) were grouped into slow metabolizers. No association was found between CYP2A6 metabolizer status and clinicopathological features (Table 2).

Table 2.

Association between CYP2A6 metabolizer status and clinicopathological parameters of breast cancer patients.

	CYP2A6 metabolizer status N (%)		p-value
	Normal	Intermediate/slow
Age (years)
<60	3 (37.50)	21 (43.75)	0.741
≥60	5 (62.50)	27 (56.25)
T stage
1 ∼ 2	1 (12.50)	12 (25.00)	0.438
3 ∼ 4	7 (87.50)	36 (75.00)
N stage
0 ∼ 1	7 (87.50)	40 (83.33)	0.766
2 ∼ 3	1 (12.50)	8 (16.67)
ER status
<10%	3 (37.50)	7 (14.58)	0.117
≥10%	5 (62.50)	41 (85.42)
PR status
<10%	2 (25.00)	5 (10.42)	0.248
≥10%	6 (75.00)	43 (89.58)
HER2 status
Negative	5 (62.50)	29 (60.42)	0.911
Positive	3 (37.50)	19 (39.58)
Ki67 status
<20%	2 (25.00)	7 (14.58)	0.458
≥20%	6 (75.00)	41 (85.42)

Abbreviations: ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2.

CYP2A6 metabolizer and pCR

Among 56 patients in this study, 15 patients (26.79%) achieved pCR. The multivariate analysis revealed that CYP2A6 intermediate/slow metabolizers were more likely to achieve pCR compared with normal metabolizers (29.17% vs. 12.50%; adjusted OR = 33.168, 95% CI 1.064–1033.970; p = 0.046; Table 3), albeit no difference in the univariate logistic analysis (OR = 2.882, 95% CI 0.324–25.646; p = 0.342). Additionally, lower ER expression level (adjusted OR = 0.963, 95% CI 0.934–0.994; p = 0.019) and HER2 status (adjusted OR = 9.695, 95% CI 1.561–60.213; p = 0.015) were also independent predictive factors for pCR (Table 3). Correlations between other CYP2A6 genotypes and pCR were shown in Supplementary-Table 2.

Table 3.

Multivariate regression analysis for predicting factors of pCR rate.

Variables	Comparison	adjusted OR (95%CI)	p-value
CYP2A6 genotype predicted metabolizer status	Intermediate/slow vs. normal	33.168 (1.064–1033.970)	0.046 *
Age	47–71	1.038 (0.914–1.180)	0.564
Tumor size	1.5%–16	0.890 (0.556–1.427)	0.630
ER status	0%–95%	0.963 (0.934–0.994)	0.019 *
PR status	0%–90%	0.988 (0.952–1.025)	0.512
HER2 status	Positive vs. negative	9.695 (1.561–60.213)	0.015 *
BMI	17.263–35.796	0.903 (0.692–1.179)	0.453

Note: OR and 95%CI were analyzed by multivariate logistic regression and adjusted by age, ER status, PR status, HER2 status, tumor size, BMI and neoadjuvant regimen. Age, ER status, PR status, tumor size, and BMI were included as continuous variables, with the ranges of these variables presented in the ‘Comparison’ column, and the other variables were included as categorical ones.

^*p < 0.05.

Abbreviations: pCR, pathological complete response; OR, odds ratio; CI, confidence interval; ER, estrogen receptor; PR, progesterone receptor; HER2, human epidermal growth factor receptor 2; BMI, body mass index.

Based on the interaction between CYP2A6 metabolizer status and neoadjuvant regimen, a nomogram was established, incorporating 7 variables including CYP2A6 metabolizer status, and clinicopathological features to predict pCR (Figure 1a). The nomogram was constructed and evaluated within the same patient cohort. The calibration curves showed a strong agreement between the predicted probabilities of the model and the observed pCR results (Figure 1b). Additionally, we evaluated the predictive value of CYP2A6 metabolizer status for pCR by comparing the ROC curves of different models. Model 1 combined clinicopathological features and neoadjuvant regimen with the area under curve (AUC) of 0.878. Model 2 incorporated clinicopathological features, neoadjuvant regimen and CYP2A6 metabolizer status with the AUC of 0.907, which is numerically superior to Model 1 without the inclusion of CYP2A6 metabolizer status (Figure 1c).

Figure 1.

Model establishment and assessment for pCR prediction in the chemotherapy concurrent with letrozole subgroup. (a) Nomogram for estimation of pathological complete response; (b) calibration curves of the nomogram; (c) receiver operating characteristic curves of different predictive models for pathological complete response. The blue line exhibited model 1 (AUC 0.907, incorporating age, tumor size, ER status, PR status, HER2 status, BMI and CYP2A6 metabolizer status). The red line showed model 4 (AUC 0.878, including all the factors in model 1 except CYP2A6 metabolizer status). Abbreviation: AUC, area under the curve; BMI, body mass index; ER, estrogen receptor; HER2, human epidermal growth factor receptor 2; PR, progesterone receptor.

CYP2A6 metabolizer and survival

The median follow-up time of all patients was 43 (interquartile range 40–48) months. A total of 6 events occurred in the intermediate/slow group, while 2 events occurred in the normal group. In terms of CYP2A6 metabolizer status, the risk of DFS events in the intermediate/slow group was numerically lower than in the normal group (intermediate/slow vs. normal: 87.50% vs. 75.00%; Log-rank p = 0.049; adjusted HR = 0.390, 95% CI 0.038–4.024, p = 0.429; Figure 2).

Figure 2.

Kaplan-Meier estimates of disease-free survival (DFS) according to CYP2A6 metabolizer status in all patients. Abbreviations: CI, confidence interval; DFS, disease-free survival; HR, hazard ratio.

CYP2A6 and safety

In terms of hematological or biochemical AE, the data was available for 55 patients (98.21%) (Table 4). The most common AEs (>10% of patients) reported for normal and intermediate/slow metabolizers were total bilirubin increased (normal, 42.86%; intermediate/slow, 27.08%), hyperlipidemia (40.00%; 46.15%), aspartate aminotransferase (AST) increased (28.57%; 39.58%), hypercholesterolemia (0%; 17.95%) and alanine aminotransferase (ALT) increased (14.28%; 20.83%). The most common grade 3 or 4 AE (>5% of patients) reported for normal versus intermediate/slow metabolizers was hyperlipidemia (0%; 5.13%). A numerically higher incidence was observed in the intermediate/slow group for alanine aminotransferase increased, aspartate aminotransferase increased, hypercholesterolemia, and hyperlipidemia without statistically significant difference (all p > 0.05).

Table 4.

Treatment-related hematological and biochemical adverse events.

	Patients participated in the evaluation (n = 55)
	Normal (n = 7)		Intermediate/slow (n = 48)
Term	Grade 3/4	All Grades	Grade 3/4	All Grades
Alanine aminotransferase increased	0 (0%)	1 (14.28%)	0 (0%)	10 (20.83%)
Aspartate aminotransferase increased	0 (0%)	2 (28.57%)	0 (0%)	19 (39.58%)
Total bilirubin increased	0 (0%)	3 (42.86%)	0 (0%)	13 (27.08%)
	Patients participated in the evaluation (n = 44)
	Normal (n = 5)		Intermediate/slow (n = 39)
	Grade 3/4	All Grades	Grade 3/4	All Grades
Hypercholesterolemia	0 (0%)	0 (0%)	0 (0%)	7 (17.95%)
Hyperlipidemia	0 (0%)	2 (40.00%)	2 (5.13%)	18 (46.15%)

Note: The documented adverse events were available in 55 patients for alanine aminotransferase increased, aspartate aminotransferase increased and total bilirubin increased, and in 44 patients for hypercholesterolemia and hyperlipidemia.

Discussion

This study is the first to explore the significance of CYP2A6 gene polymorphism in predicting the efficacy and prognosis of neoadjuvant chemotherapy combined with letrozole among postmenopausal women with hormone receptor-positive breast cancer. We found that CYP2A6 intermediate/slow metabolizer status was associated with superior efficacy of neoadjuvant chemotherapy plus letrozole. Moreover, a nomogram including CYP2A6 metabolizer status was first reported to predict the efficacy of neoadjuvant therapy.

Previous research suggests that high estrogen levels or ER expression promote chemotherapy resistance, while low estrogen levels enhance drug efficacy [11,12]. Thus, neoadjuvant chemotherapy combined with letrozole may benefit hormone receptor-positive breast cancer patients by reducing estrogen levels. However, the clinical benefits of adding letrozole to neoadjuvant chemotherapy remain controversial. For example, a randomized clinical trial with 101 postmenopausal patients showed a significantly higher pathological complete response (pCR) rate of 25.2% in the chemotherapy plus letrozole group compared to 10.2% in chemotherapy alone (p = 0.049) [22]. Conversely, the CBCSG-036 trial of 249 patients found no significant difference in pCR rates between the two treatment groups (7.2% in the neoadjuvant chemotherapy concurrent with letrozole group vs. 4.0% in the neoadjuvant chemotherapy group, p = 0.278) [23]. Therefore, more research is warranted to identify the potential candidates who can benefit from the addition of letrozole to neoadjuvant chemotherapy, thereby achieving precision treatment and improving survival outcomes.

In patients treated with neoadjuvant chemotherapy plus letrozole, CYP2A6 intermediate/slow metabolizers achieved a higher pCR rate than normal metabolizers, suggesting a correlation between CYP2A6 metabolizer status and the efficacy of letrozole. Till date, no studies had directly compared the relationship between CYP2A6 metabolizer status and the efficacy of neoadjuvant regimen. Intriguingly, several publications focused on the role of CYP2A6 polymorphisms in letrozole drug metabolism. Hertz et al.’s study found that non-exonic variants identified through GWAS, including rs56113850, accounted for approximately 15% of the overall variability in both transformed and unadjusted letrozole concentrations [24]. Desta et al. recruited breast cancer patients who were assigned to two years of oral treatment with letrozole (2.5 mg/day) after surgery and tested for drug plasma concentrations along with germline CYP2A6 genotypes (n = 261). They found that the letrozole blood concentrations were higher in patients with CYP2A6 intermediate/slow metabolizer status than those with normal metabolizer status [normal metabolizer vs. intermediate metabolizer vs. slow metabolizer = 81.2 (0–217.5) vs. 112.4 (52.7–1256.9) vs. 152.1 (50.0–268.6) ng/ml, p < 0.0001], which indicated that CYP2A6 metabolizer status may be related to letrozole blood concentrations [10]. Moreover, Dombernowsky et al. conducted a phase II clinical trial, which enrolled 551 patients with locally advanced, locoregionally recurrent or metastatic breast cancer. Patients were randomized to letrozole 2.5 mg/day or letrozole 0.5 mg/day. The objective response rate, as the primary endpoint, was significantly better in the letrozole 2.5 mg/day group than that in the letrozole 0.5 mg/day group (24% vs. 13%, p = 0.04) [25]. Notably, neither of them examined the relationship between CYP2A6 metabolizer status and efficacy. Nevertheless, these results at least partially corroborate our findings by indicating a relationship between CYP2A6 metabolizer status/letrozole dosage and efficacy, in which different CYP2A6 metabolizer status with the same dose of letrozole or different doses of letrozole might lead to different effect due to its different blood concentrations.

Basic research has also indirectly elucidated the relationship between estrogen inhibition and breast cancer efficacy. Bonelli et al. found that in hormone-sensitive breast cancer cell line MCF-7, the proliferation of testosterone-driven breast cancer cells was correspondingly inhibited with increasing concentrations of letrozole [26]. In addition, Huang et al. revealed that estrogen could interfere with the activity of paclitaxel and other antitumor drugs by increasing intracellular p26Bcl-2 protein expression and inhibiting paclitaxel-induced apoptosis in MCF-7 cells [27]. The findings imply that estrogen may decrease the pharmacological activity of paclitaxel, and therefore further suppression of estrogen levels may improve chemosensitivity. Sui et al. pointed out that the addition of fulvestrant, a selective ER down-regulator, could completely reverse the resistance of ER+ BCap37 cells to paclitaxel, and additionally sensitizes ER+ MCF-7 and T47D cell lines to the treatment of paclitaxel [28]. The aforementioned study indirectly supports our findings that, in the subgroup of letrozole combined with chemotherapy, the benefit favored the CYP2A6 intermediate/slow metabolizers rather than the normal metabolizers. According to the above, we hypothesize that the higher efficacy of letrozole combination in CYP2A6 intermediate/slow metabolizers may be due to low activity of CYP2A6 enzyme, attenuating letrozole metabolism, elevating letrozole blood levels, further inhibiting estrogens in circulating blood as well as in local tumor microenvironment, and thereby sensitizing breast cancer to chemotherapy. However, further translational research is needed to prove this concept in the future.

Nowadays, we had not seen any studies reporting the association between CYP2A6 metabolizer status and survival outcomes in breast cancer patients. Our survival analysis implied that CYP2A6 intermediate/slow metabolizers may have a trend of lower risk of DFS events compared to normal metabolizers, although the result did not reach statistical significance in the multivariate analysis. Studies of CYP2A6 polymorphisms in other tumors indirectly support our findings. An oral fluoropyrimidine, S-1, is a commonly used agent as adjuvant chemotherapy for gastric cancer. S-1 differs from letrozole since letrozole is metabolized to an inactive substance by CYP2A6 enzyme, while the main substance in S-1, tegafur, is metabolized to an active 5-fluorouracil by CYP2A6 enzyme. Previous studies have described the association between CYP2A6 polymorphisms and S-1 pharmacokinetics, and many of the CYP2A6 variants leading to diminished enzyme activity are associated with reduced metabolism of tegafur [29]. Jeong et al. included postoperative patients who were diagnosed as stage II-III gastric cancer and treated with the same dose of S-1 in the adjuvant setting [30]. The researcher analyzed the wild-type allele (W) (CYP2A6*1) and four variant alleles (V) (CYP2A6*4, *7, *9, *10) that abolish or reduce this enzyme activity, and tested the impact of CYP2A6 polymorphisms on the outcomes of adjuvant S-1 treatment. The results showed that the 3-year relapse-free survival of patients with variant alleles was lower than that of patients with wild-type alleles (V/V vs. W/V vs. W/W = 72.5% vs. 83.1% vs. 95.9%, p = 0.032). In parallel, our exploratory study revealed a similar trend that stronger CYP2A6 metabolizer enzyme activity was associated with worse survival outcomes in patients undergoing neoadjuvant chemotherapy with letrozole. Taken together, our study and previous research indicate that CYP2A6 metabolizer status can affect the blood concentrations of letrozole, accordingly resulting in the efficacy and survival benefit.

There are some limitations of this study. First, our data analysis was based on a relatively small sample size. Although we observed significant associations in univariate analysis, the lack of statistical significance in multivariate analysis may be due to limited statistical power. Further studies with larger cohorts are needed to validate these findings. Secondly, the nomogram model was developed and evaluated within the same small dataset (n = 56), which may increase the risk of overfitting and could partly contribute to the high AUC observed. External validation in an independent patient cohort is required before the nomogram model can be considered for clinical application. However, as an exploratory study, all of these findings point to at least a partial picture of intrinsic relationship between CYP2A6 genotype and neoadjuvant letrozole treatment. In addition, our study examined a limited number of CYP2A6 genotypes that are common in Asians, and some genotypes with low variation rates in Asians were excluded, such as *10 with a variation rate of 0.4-4.3% and *12 with a variation rate of 0-0.8% [9]. Even so, these genotypes might have little impact on the results of this study due to their low variation rates.

This finding may have important implications for precision therapy once it is validated. For example, tailoring the letrozole dose based on CYP2A6 activity may help achieve comparable systemic exposure, and for patients with normal metabolizer status, alternative aromatase inhibitors not primarily metabolized by CYP2A6, such as anastrozole or exemestane, might be more suitable. So further validation in larger, independent cohorts is needed. In the treatment of advanced breast cancer, letrozole remains one of the key agents [31–33], and metabolic genotype may similarly influence efficacy by altering plasma concentration in this setting. Therefore, CYP2A6-related biomarkers may hold potential value for guiding personalized endocrine therapy and warrant further investigation in these patients. In summary, the results of this retrospective study were exploratory, and prospective, large-sample clinical studies with increased detection of polymorphic loci are needed for further in-depth and comprehensive research, and ultimately inform precision treatment strategies.

Conclusion

To the best of our knowledge, this is the first study to reveal that CYP2A6 metabolizer status is a novel biomarker for predicting pCR in hormone receptor-positive postmenopausal breast cancer patients receiving neoadjuvant chemotherapy combined with letrozole, with pCR rate higher for CYP2A6 intermediate/slow metabolizers than that for normal metabolizers. There was a similar trend of survival benefit for this subgroup of patients. Our results may help to further illuminate the logic for the precise and individualized application of neoadjuvant letrozole and guide treatment strategies.

Funding Statement

This study was funded by National Natural Science Foundation of China (No. 82173115 and 82103695), Shanghai Rising-Star Program (No. 22QC1400200), Shanghai Municipal Health Commission Health Industry Clinical Research Special Project (No. 202340085), Science and Technology Commission of Shanghai Municipality (No. 20DZ2201600), Clinical Research Innovation Nurturing Fund of Renji Hospital and United Imaging (No. 2021RJLY-002), and Nurturing Fund of Renji Hospital (No. PYIII20-09 and RJPY-LX-002).

Ethics approval and consent to participate

Our study had adhered to the principles stated in the ‘Declaration of Helsinki’, and was performed under the provisions of the Ethics Committee of Renji Hospital, School of Medicine, Shanghai Jiao Tong University (approval number, KY2021-095-B). The informed consent was waived under the provisions of the Ethics Committee of Renji Hospital.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.