Introduction

The steady increase in the incidence of breast cancer (BC), as well as associated mortality and disability of the population, determines the relevance of the search for effective treatment and prevention of this pathology [1]. In the last decade in oncology, along with chemotherapeutic agents, targeted drugs have been increasingly used for treatment personalization; however, chemotherapy (CT) still occupies a central place in cancer treatment [2, 3]. Despite the fact that many molecular-genetic and immunological assays to search for markers and predictors of response to drug therapy have long been used in oncology, there is individual variability in response to treatment efficacy and safety, depending on the polymorphism of genes in the xenobiotic biotransformation system [4].

Currently, considering the clinical and economic advantages of a personalized approach in medicine, the necessity and feasibility of prescribing pharmacogenetic testing to patients is being widely discussed. However, the justification for conducting these tests must be supported by the results of clinical studies [5]. Among the main enzymes involved in drug metabolism are glutathione S-transferases (GST) GSTM1 and GSTT1, which catalyze detoxification reactions of exogenous and endogenous compounds [6]. The most common polymorphisms are extended gene deletions that inactivate the corresponding enzymes, which, in turn, alters xenobiotic metabolism [7]. Since age, disease stage, and surrogate biological markers can significantly influence treatment effectiveness and prognosis, we decided to test this hypothesis, considering the polymorphic variants of the GSTT1 and GSTM1 genes.

The aim of the study was to evaluate differences in the effectiveness of BC chemotherapy depending on age, stage, and biological tumor subtype, considering the deletion status of the GSTT1 and GSTM1 genotypes in patients.

Materials and methods

In our work, we analyzed data from 132 female patients aged 23 to 79 years (mean age 51.16 ± 12.08 years) who received chemotherapy treatment for BC between 2013 and 2021. CT was administered according to clinical guidelines and included cytotoxic agents such as: cyclophosphamide, doxorubicin, 5-fluorouracil, paclitaxel, docetaxel, and carboplatin.

Genotyping of deletion variants of the GSTT1 (GSTT1-0) and GSTM1 (GSTM1-0) genes was performed using multiplex polymerase chain reaction, followed by analysis of the melting curves of the reaction products (the experimental methodology for GSTM1 and GSTT1 genotyping was described previously [8]). The criteria for evaluating the effectiveness of antitumor treatment were overall survival (OS) and relapse-free survival (RFS) [9]. The Kaplan–Meier method was used to calculate these parameters; the hazard ratio (HR) for relapse development based on a specific characteristic (age, disease stage, biological tumor subtype) was calculated with determination of the 95% confidence interval (CI) using Cox regression. In subgroups with a small number of observations, to eliminate estimation bias and increase the reliability of the results, the bootstrap procedure, based on multiple sample generation using the Monte Carlo method, was used when calculating CIs.

Results

We investigated two polymorphic variants of the GSTM1 and GSTT1 genes in women with BC: "null" — homozygous deletion variant (0/0; hereinafter referred to as GSTT1-0 and GSTM1-0) and "normal" — homozygous or heterozygous variants for the "wild" allele (+/+, 0/+; hereinafter referred to as GSTT1-1 and GSTM1-1). In 110 (83.3%) patients, the wild-type GSTT1 gene was identified, while the null genotype was found in 22 (16.7%) patients. The occurrence of the normal and null GSTM1 genotypes was almost equal: 68 patients (51.5%) and 64 patients (48.5%), respectively. Patient characteristics are presented in Table 1.

Table 1

Distribution of patients by genotypes depending on age, stage, and biological subtype of the tumor

FeatureGSTM1-1, n (%)GSTM1-0, n (%)GSTT1-1, n (%)GSTT1-0, n (%)Total, n (%)Age     Up to 44 years23 (57.5)17 (42.5)36 (27.3)4 (3.0)40 (30.3)After 45 years45 (48.9)47 (51.1)74 (56.1)18 (13.6)92 (69.7)Biological tumor subtypes     Luminal A8 (6.1)10 (7.6)15 (11.4)3 (2.3)18 (13.6)Luminal B HER2-negative23 (17.4)18 (13.6)33 (25.0)8 (6.1)41 (31.1)Luminal B HER2-positive13 (9.8)8 (6.1)18 (13.6)3 (2.3)21 (15.9)HER2-positive6 (4.5)5 (3.8)11 (8.3)—11 (8.3)Triple-negative16 (21.1)16 (21.1)26 (19.7)6 (4.5)32 (24.2)Unknown2 (1.5)7 (5.3)7 (5.3)2 (1.5)9 (6.8)Stage     I1 (0.8)1 (0.8)2 (1.5)—2 (1.5)II29 (22.0)29 (22.0)48 (36.4)10 (7.6)58 (43.9)III38 (28.8)34 (25.6)60 (45.5)12 (9.1)72 (54.6)

To evaluate the effectiveness of CT with different polymorphic variants of GST genes, we analyzed the overall and relapse-free survival of patients both in the general patient group and depending on age, stage, and biological tumor subtype. During the 8-year follow-up period, the median OS was not reached; the median RFS was 62.7 months (5.2 years).

In patients with GSTM1-0, RFS was higher than in carriers of the functional GSTM1-1 genotype: 58.3 (±6.5)% vs. 39.4 (±6.6)%, median 77.1 (±8.4) and 34.2 (±13.0), respectively (HR=0.596, 95% CI 0.369–0.964, p=0.033). A similar trend was observed for OS: with the deletion GSTM1 genotype — 82.0 (±5.2)%, with the wild type — 65.2 (±6.3)%, median not reached; however, this association was not statistically significant (HR=0.6, 95% CI 0.314–1.147, p=0.118). The results of multivariate analysis of OS and RFS depending on patient age, disease stage, surrogate biological subtype, and GSTM1 gene polymorphism are presented in Table 2.

Table 2

Overall and relapse-free survival by age groups, disease stages, biological subtypes depending on GSTM1 gene polymorphism

Feature Overall Survival Relapse-Free Survival   5-year survival (±SE)HR (95% CI)p5-year survival (±SE)HR (95% CI)pAge       Up to 44 yearsN53.4 (±11.4)0.56 (0.209–1.502)0.24334.2 (±11.0)0.426 (0.173–1.05)0.056 D73.3 (±11.4)  60.6 (±12.6)  After 45 yearsN74.1 (±7.1)0.648 (0.273–1.538)0.32145.5 (±8.0)0.686 (0.386–1.219)0.197 D85.5 (±5.5)  60.2 (±7.5)  Biological tumor subtypes       Luminal AN62.5 (±21.3)0.236 (0.021–2.622)0.20150.0 (±20.4)0.382 (0.095–1.543)0.161 D90.0 (±9.5)  80.0 (±12.6)  Luminal B HER2-negativeN66.3 (±10.4)0.374 (0.101–1.383)0.12638.3 (±10.6)0.41 (0.138–0.961)*0.034 D84.7 (±10.3)  60.6 (±11.6)  Luminal B HER2-positiveN75.0 (±15.8)0.021 (0.007–0.039)*0.00132.4 (±15.0)0.472 (0.121–1.835)0.267 D—  51.7 (±18.7)  HER2-positiveN66.7 (±19.2)0.835 (0.109–6.425)0.86344.4 (±22.2)0.829 (0.138–4.981)0.837 D75.0 (±21.7)  50.0 (±25.0)  Triple-negativeN51.6 (±13.3)0.839 (0.304–2.319)0.73542.2 (±12.7)0.60 (0.240–1.501)0.270 D63.5 (±13.1)  47.6 (±14.0)  UnknownN—35.2 (24.04–111.5)*0.001—45.1 (25.7–242.5)*0.001 D62.5 (±21.3)  44.4 (±22.2)  Stage       IIN85.9 (±7.6)0.536 (0.127–2.26)0.38869.0 (±9.8)0.879 (0.338–2.284)0.791 D92.4 (±5.1)  78.1 (±7.9)  IIIN53.2 (±8.6)0.719 (0.344–1.50)0.38024.6 (±7.4)0.521 (0.287–0.887)*0.023 D75.7 (±8.0)  43.6 (±9.1)  

Notes: Statistically significant differences and relative risks are highlighted in bold. * Confidence intervals were calculated using the bootstrap procedure; N represents the wild-type GSTM1 gene; D represents the deleted GSTM1 gene.

As seen in Table 2, the presence of the "null" GSTM1 genotype statistically significantly reduced the risk of relapse in patients with stage III disease (HR=0.521, 95% CI 0.29–0.89, p=0.023); other differences did not have a statistically significant effect on overall and relapse-free survival. When comparing OS and RFS depending on the surrogate biological tumor subtype, it was found that in the group of patients with an unknown biological type, the presence of the GSTM1-0 genotype increased the risk of death and relapse. Among patients with luminal B HER2-positive BC and GSTM1-0, no lethal cases were detected; simultaneously, in women with the luminal B HER2-negative BC subtype, the risk of relapse was reduced in the group with the GSTM1 null genotype. Other differences in overall and relapse-free survival between patients with different biological BC subtypes, as well as depending on patient age based on the presence of the GSTM1 deletion, were statistically insignificant.

When studying the influence of the GSTT1 deletion genotype on OS in BC patients, similar to GSTM1-0, no statistically significant differences were identified; the median for patients with the GSTT1 null genotype was not reached, while for patients with the wild type it was 92.5 months. Statistically significant differences were found in the analysis of RFS: with GSTT1-0, RFS was 70.4 (±10.2)%, and with the functional GSTT1-1 genotype — 45.8 (±5.1)% (HR=0.418, 95% CI 0.191–0.915, p=0.024).

In multivariate analysis, comparing OS and RFS, similar patterns were found as with the GSTM1 gene deletion: the presence of a homozygous null allele genotype of GSTT1 reduces the risk of relapse in patients with stage III disease (RFS with the "normal" genotype 28.0 (±6.1)%, with the "null" genotype 61.4 (±15.3)%; 95% CI 0.098–0.99, p=0.049), without affecting OS (with the "normal" variant it was 59.4 (±6.7)% vs. 76.2 (±14.8)% with the "null" variant; HR=0.402, 95% CI 0.096–1.69, p=0.199) (Fig. 1).

Fig. 1. Overall and relapse-free survival for stage III depending on the deletion polymorphism of the GSTM1 gene

Notes: N — normal genotype; D — homozygous deletion.

In other cases, the presence of deletion polymorphism did not have a statistically significant effect on overall and relapse-free survival. Thus, at stage I, no cases with the GSTT1 null genotype were identified; at stage II: OS with the "normal" GSTT1 genotype was 91.9 (±4.5)%, with the "null" genotype 80.0 (±12.6)% (HR=0.714, 95% CI 0.272–6.7, p=0.713); RFS with the "normal" genotype — 72.6 (±7.1)% vs. 80.0 (±12.6)% in the presence of deletion (HR=0.433, 95% CI 0.099–1.892, p=0.252).

In luminal A and luminal B HER2-negative subtypes, no relapses or lethal outcomes occurred in patients with the GSTT1 deletion genotype. Statistically significant differences were obtained when assessing OS in patients with luminal A (OS with wild type 75.5 (±12.3)%, with null — no lethal cases; HR=0.034, 95% CI 0.02–0.045, p=0.001) and luminal B HER2-negative subtypes (OS with wild type 69.9 (±8.5)% vs. absence of lethal cases with null genotype; HR=0.035, 95% CI 0.025–0.044, p=0.001). Unlike the GSTM1 gene deletion polymorphism, when analyzing patients with an unknown subtype, an opposite trend is observed: with the "normal" genotype, OS is 68.6 (±18.6)%, with the "null" — there were no deaths among patients (HR=0.039, 95% CI 0.015–0.044, p=0.002). In all other cases, no statistically significant differences in OS were found. It should be noted that among patients with the HER2-positive subtype, there were no patients with the GSTT1 null genotype. In the analysis of RFS across all biological subtypes, statistically significant differences were found only for patients with an unknown tumor subtype: with the "normal" genotype 68.6 (±18.6)%, with the "null" — 100% (HR=0.036, 95% CI 0.013–0.041, p=0.001).

No significant differences in OS and RFS were obtained between the groups of patients under 44 years and over 45 years. In the group under 44 years, OS with the "normal" GSTT1 genotype was 60.6±8.9%, with the "null" — 75.0±21.7% (HR=0.338, 95% CI 0.045–2.559, p=0.271); RFS was 42.1±9.0% and 75.0±21.7%, respectively (HR=0.262, 95% CI 0.035–1.958, p=0.160). OS in patients over 45 years, both in the presence of deletion and with the wild type, was 78% (HR=0.696, 95% CI 0.205–2.363, p=0.559); RFS with the functional genotype was 49.1±6.2%, with the null genotype 69.3±15.4% (HR=0.487, 95% CI 0.207–1.149, p=0.093).

Thus, we identified a statistically significant association between the presence of deletion genotypes of GSTM1 and GSTT1 and higher RFS in patients with stage III disease; this cohort of patients benefits from CT compared to patients with the normal functional genotype. Analyzing the biological tumor subtypes, we found that in patients with luminal B HER2-positive BC and the GSTM1-0 genotype, no lethal cases occurred; simultaneously, in women with luminal B HER2-negative BC, the risk of relapse was reduced in the group with the GSTM1 deletion. In patients with luminal A and luminal B HER2-negative BC, OS is higher in the presence of the GSTT1 deletion polymorphism. However, in the unknown subtype, a multidirectional trend was observed: the GSTM1 deletion genotype increased the risk of relapse and death, while the GSTT1 deletion genotype, conversely, reduced the risk of relapse, which is apparently related to the impossibility of selecting personalized treatment.

Discussion

Glutathione S-transferases play an important role in the metabolism of xenobiotics, including cytostatic drugs used for BC treatment, such as doxorubicin/epirubicin, paclitaxel/docetaxel, etc. [10]. Deletion polymorphisms of GST genes can lead to changes in enzymatic activity and contribute to variability in response to chemical compounds. It is known that people with the GST null genotype have a reduced ability to metabolize xenobiotics; as a result, they are more susceptible to developing various multifactorial diseases, including cancer, due to circulating toxic metabolites that cause cell damage. However, this fact may enhance the effectiveness of drug therapy [11]. Therefore, studying gene polymorphisms and their possible impact on the effectiveness of CT for malignant neoplasms is of great interest. Nevertheless, to date, only limited and contradictory information exists on the association of GSTT1 and GSTM1 genotypes with BC outcomes.

Several authors note a reduced risk of disease relapse and lethal outcomes in the presence of deletion polymorphism of the GSTT1 and/or GSTM1 genes [12–14]. Thus, in the work of Mishra A et al., in patients who received neoadjuvant CT for BC, the response rate to CT was higher in those with the GSTT1 and GSTM1 null genotypes; however, the results were not statistically significant, possibly due to the small sample size (44 individuals) [15]. After analyzing 21 articles, Pacholak LM et al. concluded in their systematic review that the deletion variant of the GSTM1 gene increases the effectiveness of CT [16].

On the other hand, some studies did not find associations between GSTT1 and GSTM1 gene polymorphisms and patient survival [17, 18]. A fairly large study in a Chinese population found no link between polymorphic variants of the GSTT1 and GSTM1 genes and patient survival [19].

We found that patients with stage III disease benefit from CT: in the presence of GSTT1 and GSTM1 gene deletions, RFS in these patients is higher.

Our study did not reveal statistically significant differences in the effect of GSTT1 and GSTM1 gene polymorphisms on overall and relapse-free survival of patients depending on the age of disease onset. In our opinion, this is logical and may be due to the fact that the choice of personalized treatment did not depend on age but was based on surrogate biological marker data and disease stage. Although it should be noted that elderly women generally have a better prognosis due to the more frequent occurrence of favorable hormone-dependent BC in this age cohort [20].

The surrogate biological tumor subtype has a significant impact on the choice of drug therapy strategy [21]. In our work, it was found that for the luminal HER2-negative tumor subtype, in the presence of the GSTT1 and GSTM1 deletion genotypes, survival is higher, which is partially comparable to the data of Campos CZ et al. and Almeida M et al. In their work, Campos CZ et al. also noted a reduced risk of relapse in patients with luminal B HER2-negative BC and the GSTM1 null genotype, while Almeida M et al. indicated an association of the heterozygous GSTM1 (1/0) variant and HER2-positive tumor with a poor prognosis and aggressive disease nature [22, 23].

Conclusion

The results of our study showed a significant effect of the GSTT1 and GSTM1 gene deletion polymorphism on relapse-free survival in patients with stage III disease and hormone-dependent breast cancer. Choosing therapy based on pharmacogenetic test data will improve treatment effectiveness, consider the economic feasibility of prescribing certain types of drugs, and reduce the incidence of adverse reactions.