Circulating T-cell receptor repertoire and clinicopathological correlations in breast cancer patients: immune repertoire analysis from the VGH-TAYLOR study Authors: Ling-Ming Tseng, Chi-Cheng Huang, Ji-Lin Chen, Yi-Fang Tsai, Ta-Chung Chao, Wen-Chi Wu, Pei-Ju Lien, Yen-Shu Lin, Chin-Jung Feng, Yen-Jen Chen, Jiun-I. Lai, Jen-Hwey Chiu, Chih-Yi Hsu, Chun-Yu Liu Published: 2025 ABSTRACT Sequencing of the T-cell receptor (TCR) repertoire can reflect immune status and monitor treatment responses. Here, we aimed to characterize the TCRβ repertoire and its correlation with clinical and prognostic significance in breast cancer. A total of 856 peripheral blood samples were collected from female breast cancer patients for TCR sequencing. At baseline, TCR clonality was elevated in patients with stage IV disease compared with earlier-stage breast cancer, and higher clonality within the stage IV cohort was associated with poorer overall survival. Compared with the luminal A subtype, luminal B2 breast cancer exhibited lower baseline TCR Shannon diversity. Following both adjuvant and neoadjuvant chemotherapy, we observed increased TCR convergence and clonality accompanied by reduced Shannon diversity. Remarkably, these repertoire changes were most evident in patients treated with taxane or anthracycline plus taxane regimens in the adjuvant setting, and with anthracycline plus taxane or platinum-based regimens in the neoadjuvant setting. Notably, trastuzumab was associated with increased clonality and reduced diversity in HER2-enriched tumors, but not in Luminal B2. Lower TCR richness in post-treatment blood samples was associated with patients achieving pathologic complete response in the neoadjuvant setting. This study provides comprehensive circulating TCR repertoire metric profiles across the heterogeneous population of breast cancer patients.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13058-025-02172-w. Introduction Breast cancer is the most common malignancy in women and is a heterogeneous disease that exhibits diverse clinicopathological features, responses to therapy, and clinical outcomes [1]. Apart from traditional systemic therapies such as chemotherapy, endocrine therapy, and anti-human epidermal growth factor receptor 2 (HER2)-targeted therapy for heterogeneous breast cancer, other molecular targeted therapies are selected for the treatment of specific breast cancer subgroups. These include poly ADP ribose polymerase (PARP) inhibitors for germline BRCA-mutated, HER2-negative breast cancer and phosphoinositide 3-kinase (PI3K) inhibitors for PIK3CA-mutated breast cancer. Furthermore, the development of immunotherapy for breast cancer shows promising results in achieving durable responses and improving overall survival [2]. Improving the local immune response and recovering immunosurveillance mediate the long-term effects of anticancer therapies [3, 4]. Immune cell signatures and tumor-infiltrating lymphocytes (TILs) have been proposed as predictive and prognostic biomarkers of treatment response [5]. The immune repertoire refers to the collection of T-cell receptors (TCR) and B-cell receptors in the circulatory system. TCR is a highly diverse heterodimer, usually composed of α and β chains, that recognizes peptide-MHC epitopes to initiate adaptive immune responses. Several studies have indicated that there is heterogeneity in the TCR repertoire in various cancer types, including lung cancer, hepatocellular carcinoma, and breast cancer [6–8]. Analysis of TCR repertoire using high-throughput next-generation sequencing to estimate TCR diversity, clonality, and convergence can reflect immune responses [9]. These TCR repertoire metrics have potential applications in cancer biology and the development of translational implications [10]. Notably, the correlation between the TCR repertoire and genetic alterations, such as EGFR and homologous recombination repair (HRR) genes, has been investigated [11, 12]. The TCR repertoire of tumor-infiltrating or circulating T cells could serve as biomarkers for diagnosis and prognosis [13–16]. Nevertheless, substantial intratumor heterogeneity of TCR clonality has been observed in different regions of the same tumor. The majority of T-cell clones may be restricted to specific tumor regions [17–19]. Liquid biopsy can be used to obtain baseline information on patients less invasively and monitor dynamic alterations in cancer progression and response to treatment [20]. In this study, we explored the T-cell repertoire in patients with breast cancer, highlighting its potential association with clinicopathological features. We used TCR sequencing to analyze TCR diversity, clonality, and convergence, providing insights into immune responses in breast cancer. By characterizing the TCR profile of breast cancer subtypes, as well as the relationship of the TCR repertoire with clinicopathological features, genetic alterations, and response to chemotherapy, we aimed to better understand immune responses in breast cancer and explore potential implications for the development of personalized therapeutic strategies. Methods Participants and sample collection The VGH-TAYLOR study encompasses a broad range of clinical scenarios, covering various breast cancer subtypes and treatment settings. Written informed consent was obtained from all the patients, and the study was approved by the ethics committee of the Institutional Review Board of Taipei Veterans General Hospital, Taipei, Taiwan. This study was conducted in compliance with the principles of the Declaration of Helsinki. A total of 856 peripheral blood samples were collected from female breast cancer patients aged > 20 years, Eastern Cooperative Oncology Group Performance score ≤ 3, and life expectancy ≥ 3 months in the VGH-TAYLOR study (Table S1; NCT04626440) [21]. Patients were enrolled between November 22, 2018, and December 28, 2020, with a median follow-up of 57 months. The schedule of blood collection timepoints for each group is summarized (Table S2). Eligible subjects were classified into three groups: Group 1 (adjuvant setting), Group 2 (neoadjuvant setting), and Group 3 (stage IV breast cancer). Peripheral blood was collected at baseline (visit 1) in all patients. For Group 1, a second blood sample (visit 2) was obtained after surgery and upon completion of the prescribed adjuvant chemotherapy regimen (Table S3–S4). For Group 2, the second sample (visit 2) was collected following completion of the entire neoadjuvant chemotherapy course and after definitive surgery (Tables S5–S6). Group 3 patients provided only a baseline blood sample prior to the initiation of systemic therapy. Clinicopathological parameters were recorded, including tumor stage, age at diagnosis, tumor size, histological grade, recurrence, and expression of estrogen receptor (ER), progesterone receptor (PR), HER2, and Ki-67. Tumor size was defined as the diameter of the primary breast tumor, based on clinicopathological reports (AJCC pathological T stage for stage I–III) at diagnosis. For metastatic patients, tumor size was determined from imaging of the primary tumor at the time of diagnosis, when available. Overall survival was defined as the time from the date of initial diagnosis to death from any cause or last follow-up. Recurrence-free survival was defined as the time from diagnosis to the first recurrence of invasive breast cancer (local, regional, or distant) or death due to breast cancer; patients who died without recurrence from non-breast cancer causes were censored at the date of death or last follow-up. Event-free survival was defined as the time from diagnosis to the first occurrence of any event, including invasive disease recurrence, progression, a second primary cancer, or death from any cause, whichever occurred first [22, 23]. Chemotherapy regimens Breast cancer patients in this study received neoadjuvant or adjuvant chemotherapy according to the National Comprehensive Cancer Network (NCCN) guidelines in effect between 2018 and 2022 [24, 25]. In the adjuvant setting (Group 1), patients were most commonly treated with taxane-based regimens (docetaxel or paclitaxel, 4–6 cycles), anthracycline-based regimens (doxorubicin or epirubicin, 4–6 cycles), or sequential anthracycline followed by taxane regimens, typically administered over a total of 8 cycles. For patients receiving neoadjuvant chemotherapy (Group 2), the predominant protocol consisted of taxane-based treatment followed by an anthracycline-containing regimen. In HER2-positive breast cancer, a heart-sparing regimen of six cycles of docetaxel plus carboplatin in combination with trastuzumab (TCH or TCarbo) was frequently used. Platinum-containing regimens (e.g., carboplatin) were also administered in selected patients with triple-negative breast cancer (TNBC), consistent with NCCN recommendations. All chemotherapy protocols were delivered at the discretion of the treating physician and tailored to tumor biology, clinical stage, and individual patient factors. Assessment of breast cancer subtypes ER and PR status were determined by immunohistochemistry (IHC) on formalin-fixed, paraffin-embedded tumor tissue, with positivity defined as nuclear staining in > 1% of tumor cells. HER2 expression was scored from 0 to 3 + by IHC, and HER2 gene amplification was assessed by fluorescence in situ hybridization (FISH). Tumors with an IHC score of 3 + , or a score of 2 + with confirmed HER2 amplification by FISH, were classified as HER2-positive. All assessments were performed in accordance with ASCO/CAP guidelines [26]. According to the 2013 St. Gallen consensus [27], breast cancer is classified into five subtypes based mainly on the expression of ER, PR, HER2, and Ki-67. These subtypes were luminal A (ER-positive, PR ≥ 20%, HER2-negative, Ki-67 < 14%, and/or nuclear grade 1 or 2), luminal B1 (ER-positive, PR < 20%, HER2-negative, Ki-67 ≥ 14%), luminal B2 (ER-positive and HER2-positive), HER2-enriched (HER2-positive, ER-negative, PR-negative), and TNBC (ER-negative, PR-negative, HER2-negative). TCR sequencing TCRβ sequencing was performed at a depth of approximately 1.0–1.5 million reads per sample using the Oncomine™ TCR Beta-LR Assay, a targeted amplicon-based sequencing platform optimized for TCR β-chain profiling from total RNA [28] (Thermo Fisher Scientific, MA, USA). Briefly, RNA was extracted from the buffy coat using MagMAX Kits and quantified using the Qubit RNA HS Assay Kit. Next, 25 ng of DNase-treated RNA was reverse-transcribed into cDNA using the SuperScript IV VILO Master Mix. Then, 5 μL of cDNA was amplified using Ion AmpliSeq HiFi Mix. The resulting amplicons and Oncomine TCR Beta-LR Panel were digested with FuPa Reagent and ligated to an adaptor. Libraries were purified with Agencourt AMPure XP Reagent, washed with freshly prepared 70% ethanol, and eluted according to the manufacturer’s instructions. The samples were sequenced using the Ion GeneStudio S5 System and the Ion 530 Chip. Data were analyzed and reported using Ion Reporter Software (version 5.6), which provided clonality, richness, convergence, Shannon diversity, and the usage frequencies of clones, variable (TRBV) genes, and joining (TRBJ) genes. Clonality was defined as 1–evenness, a metric reflecting the degree of clonal expansion, while richness was defined as the number of unique clonotypes and reported descriptively. Convergence was calculated as the aggregate frequency of clonotypes that are identical in amino acid sequence but differ in nucleotide sequence, commonly arising from chronic antigen stimulation such as from tumor antigens [29, 30]. Because RNA input and sequencing depth were standardized across all samples, no additional computational downsampling was performed, and repertoire metrics were derived directly from the normalized data provided by the Ion Reporter pipeline. The assay amplified from the FR1 region of the TCRβ variable gene to the constant region, generating amplicons that spanned the CDR1–CDR3 regions. V(D)J gene assignment and CDR3 sequence annotation were performed by the Ion Reporter Software using the IMGT reference database. From this output, we extracted the CDR3 nucleotide length distribution as well as TRBV and TRBJ gene usage frequencies. TCR metrics (convergence, clonality, richness, and diversity) were evaluated, with their definitions summarized (Table S7) and schematic illustrations provided (Fig. S1A–B). Targeted sequencing DNA was extracted from formalin-fixed, paraffin-embedded tumor samples for next-generation sequencing. All tumor tissue samples used for targeted DNA sequencing were collected at the time of initial diagnosis, prior to the initiation of any systemic therapy [21, 31]. The Oncomine Comprehensive Assay v3 (Thermo Fisher Scientific, Waltham, MA) was used to identify 161 cancer-related genes, including single-nucleotide variants, copy number variations, gene fusions, and indels. Variant calling and annotation were performed as previously described [32], Mutations in PIK3CA and HRR genes in breast cancer patients were determined. Based on previous reports [33, 34], HRR candidate genes included ARID1A, ATM, ATR, ATRX, BAP1, BARD1, BLM, BRCA1/2, BRIP1, CHEK1/2, FANCA/C/D2/E/F/G/L, MRE11, MRE11A, NBN, PALB2, PTEN, RAD50, RAD51, RAD51B, RAD51C, RAD51D, and WRN. Statistical analysis All statistical analyses were performed using the SPSS for Windows software (version 22.0, SPSS, Chicago, IL, USA). Data are expressed as mean ± standard deviation (SD). Pearson’s correlation was used to analyze the correlation between TCR repertoire and clinicopathological parameters. Normality analysis was conducted prior to further statistical analysis. Based on the data's normality, the independent t-test or Mann–Whitney U test was used for comparing two independent samples, while the paired t-test or Wilcoxon signed-rank test was applied for paired samples. For multiple comparisons, one-way ANOVA or the Kruskal–Wallis test was used. If multiple comparisons were significant, pairwise subgroup comparisons were conducted using post hoc analysis, with the Bonferroni correction applied to adjust alpha values by dividing the original alpha value by the number of tests. Survival curves for breast cancer patients were generated using the Kaplan–Meier method and compared with the log-rank test. Optimal cutoff values for TCR metrics were determined using maximally selected rank statistics implemented in the survminer R package (version 0.5.0) [35]. Cox proportional hazards regression analysis was applied to evaluate the impact of clinicopathological factors on overall survival. Statistical significance was defined as a p-value less than 0.05, unless otherwise specified. Results TCRβ clonality was associated with late stage and worse survival in patients with treatment-naïve breast cancers The 856 enrolled subjects included into three groups: Group 1 (scheduled for surgery as first-line treatment, followed by adjuvant therapy), Group 2 (scheduled for neoadjuvant therapy as first-line treatment, followed by surgery), and Group 3 (diagnosed with de novo or treatment-naïve recurrent stage IV breast cancer) (Fig. 1A). To profile the baseline (visit 1) TCRβ repertoire in breast cancer, blood samples from treatment-naïve breast cancer patients were analyzed (Table S1). These patients, with a median age of 56 years (interquartile range: 47–65 years), included 167 patients (19.6%) with luminal A, 395 (46.4%) with luminal B1, 88 (10.3%) with luminal B2, 85 (10.0%) with HER2-enriched, and 117 (13.7%) with TNBC subtypes. TCR sequencing data revealed an inverse correlation between TCR clonality and both richness and Shannon diversity (Figs. 1B and S2A). TCR clonality showed a positive correlation with convergence and age, but no significant association with tumor size (Figs. 1C, D, and S2B). Notably, high TCR clonality was associated with stage IV breast cancer (Fig. 1E). Additionally, Shannon diversity was positively correlated with richness (Fig. S2C). Patients with advanced age and stage IV breast cancer exhibited lower TCR Shannon diversity (Figs. S2D and E). The correlations of the different metrics were summarized in a heatmap (Fig. S2F). Given the prognostic relevance of ER-low tumors [36], we stratified the ER-positive group into ER-low (1–10%) and ER-high (> 10%) subgroups. We then compared TCRβ repertoire metrics among ER-high, ER-low, and ER-negative (< 1%) cases and found no significant differences in TCR metrics (Fig. S2G–J). We further examined the relationship between baseline TCR metrics and patient outcomes, finding that high baseline TCR clonality was associated with poorer overall survival, while high TCR convergence showed a trend toward improved overall survival in stage IV breast cancer, but not in non-metastatic disease (Figs. 1F–I and S2K–N). Cross-tabulation analyses showed that the distribution of breast cancer subtypes did not differ significantly across high and low groups for any TCR repertoire metric (all Pearson’s χ2 P > 0.10, data not shown). To assess the independent prognostic value of baseline TCR clonality in stage IV breast cancer, we conducted univariate and multivariate Cox regression analyses and identified baseline clonality as an independent predictor of overall survival (Table S8). These findings indicate that higher baseline TCR clonality is associated with advanced stage and reduced TCR richness, and serves as an independent predictor of poorer survival in stage IV breast cancer.Fig. 1Correlation between the baseline TCRβ clonality and clinicopathological parameters. A Study scheme of the VGH-TAYLOR study. A total of 856 breast cancer patients were recruited, including those receiving adjuvant or neoadjuvant therapies, as well as patients with stage IV disease (both de novo and recurrent). B–E Correlations of TCRβ clonality with richness, convergence, patient age, and AJCC stage in patients with treatment-naïve breast cancers. F–I Associations of baseline TCRβ convergence, clonality, richness, and Shannon diversity with overall survival in stage IV breast cancer. For multiple comparisons in (E), six post-hoc tests were performed, with p-values < 0.0083 considered statistically significant after adjustment Correlation between the baseline TCRβ clonality and clinicopathological parameters. A Study scheme of the VGH-TAYLOR study. A total of 856 breast cancer patients were recruited, including those receiving adjuvant or neoadjuvant therapies, as well as patients with stage IV disease (both de novo and recurrent). B–E Correlations of TCRβ clonality with richness, convergence, patient age, and AJCC stage in patients with treatment-naïve breast cancers. F–I Associations of baseline TCRβ convergence, clonality, richness, and Shannon diversity with overall survival in stage IV breast cancer. For multiple comparisons in (E), six post-hoc tests were performed, with p-values < 0.0083 considered statistically significant after adjustment Profiling of the TCRβ repertoire in treatment-naïve breast cancers We analyzed the relationship between the baseline TCRβ repertoires of circulating T cells and breast cancer subtypes by comparing the convergence, clonality, richness, and diversity of TCR in different breast cancer subtypes. There were no statistically significant differences in overall survival between breast cancer subtypes (Fig. S3A), based on time from diagnosis to death from any cause or last follow-up. The results showed no statistical difference in the convergence, clonality, and richness of TCR among the breast cancer subtypes (Fig. 2A–C). Notably, patients with luminal B2 breast cancer had a lower TCR diversity than those with luminal A subtype (Fig. 2D). CDR3 is a variable region of TCR that results from random recombination of variable, diversity, and joining gene segments [37]. We compared CDR3 length, usage of variables, and joining genes among breast cancer subtypes. The results showed that the breast cancer subtypes shared similar CDR3 lengths (Fig. 2E). TRBV1, TRBV6-3, TRBV7-1, TRBV12-1, TRBV12-2, TRBV21-1, TRBV26, and TRBJ2-2P were not detected in TCRβ of patients with breast cancer. TRBV28 and TRBJ2-1 were the most frequently used variables and joining genes, respectively, in all subtypes (Fig. 2F and G). However, we observed a trend of decreased diversity in variable genes in the luminal B2 subtype compared to luminal A (Fig. S3B–C). Additionally, the usage frequencies of the TRBV5-1, TRBV7-8, TRBV11-2, TRBV15, TRBJ2-1, and TRBJ2-4 genes differed among the subtypes. (Fig. S3D–I). Fig. 2The profile of the baseline TCRβ repertoires of circulating T cells in breast cancer subtypes. A–D The profiles of baseline TCRβ convergent frequency, clonality, richness, and Shannon diversity of circulating T cells across breast cancer subtypes. E–G The usage frequencies of CDR3 lengths, TRBV genes, and TRBJ genes. For multiple comparisons in (D), ten post-hoc statistical tests were conducted, with p-values less than 0.005 considered statistically significant after adjustment for multiple comparisons The profile of the baseline TCRβ repertoires of circulating T cells in breast cancer subtypes. A–D The profiles of baseline TCRβ convergent frequency, clonality, richness, and Shannon diversity of circulating T cells across breast cancer subtypes. E–G The usage frequencies of CDR3 lengths, TRBV genes, and TRBJ genes. For multiple comparisons in (D), ten post-hoc statistical tests were conducted, with p-values less than 0.005 considered statistically significant after adjustment for multiple comparisons Correlation between the TCRβ repertoire and HRR gene mutations HRR is a DNA repair process that protects the cells from DNA damage. Deficiency in the HRR pathway, termed generation, thereby influencing the immune microenvironment deficiency, is associated with breast cancer [38]. In the VGH-TAYLOR cohort, 22.4% of the patients harbored one or more HRR gene mutations, with PALB2 and PTEN being the most frequently mutated genes (Fig. 3A). We assessed the relationship between HRR mutations and the baseline TCR repertoire and found that the number of HRR gene mutations had no impact on TCR metrics, including convergence, clonality, richness, and diversity (Fig. 3B–E). However, when examining the correlation of each TCR metric with individual HRR candidate genes (alteration rate > 5%), we observed that patients with ARID1A mutations exhibited lower TCR richness and had a trend decrease in Shannon diversity compared to those with the wild-type (Fig. 3F–I and S4A–F).Fig. 3Correlation between the baseline TCRβ repertoire and HRR gene mutations. A HRR gene mutation profile in VGH-TAYLOR cohort. B–E The profile of baseline TCRβ convergent frequency, clonality, richness, and Shannon diversity in patients with wild-type, one mutation, and more than one mutation of HRR gene. F–I The profile of TCRβ convergent frequency, clonality, richness, and Shannon diversity in patients with wild-type and ARID1A mutation Correlation between the baseline TCRβ repertoire and HRR gene mutations. A HRR gene mutation profile in VGH-TAYLOR cohort. B–E The profile of baseline TCRβ convergent frequency, clonality, richness, and Shannon diversity in patients with wild-type, one mutation, and more than one mutation of HRR gene. F–I The profile of TCRβ convergent frequency, clonality, richness, and Shannon diversity in patients with wild-type and ARID1A mutation Correlation between the TCRβ repertoire and PIK3CA mutations Mutations in PIK3CA, a gene encoding the catalytic subunit of PI3Kα, are common in breast cancer [39]. Recently, a study has successfully retrieved a patient-derived TCR with reactivity to target cells expressing mutant but not wild-type PIK3CA from peripheral blood mononuclear cells in a breast cancer patient [40]. We aimed to investigate the association between PIK3CA mutations and the baseline TCR repertoire. In the VGH-TAYLOR cohort, PIK3CA mutations were detected in approximately 40% of the patients (Fig. S5A). The PIK3CA hotspot mutations (H1047R, E545K, E542K, and H1047L) or other mutations showed no statistically significant differences in the TCR repertoire compared to the wild-type (Fig. S5B–E). Correlation of the circulating TCRβ repertoire with adjuvant chemotherapy responses (Group 1) The critical role of the immune system in conventional chemotherapy for breast cancer has become increasingly apparent. We analyzed the dynamic changes during both adjuvant and neoadjuvant chemotherapy. In the adjuvant setting, blood samples were collected before surgery and after adjuvant therapy for TCR sequencing (Fig. 4A). The results indicated increased TCR convergence and clonality, alongside decreased diversity of circulating TCRs following adjuvant chemotherapy (Fig. 4B–I). To further analyze the correlation between treatment regimens and changes in the TCR repertoire, treatments were categorized into chemotherapy (anthracycline, taxane, or both), platinum-based therapy, immunotherapy, and trastuzumab. Multivariate linear regression was performed to assess potential confounding, with ΔClonality as the dependent variable (Table S9). Patients receiving taxane or anthracycline plus taxane showed increased TCR convergence and clonality, together with reduced Shannon diversity (Fig. S6A–D). No significant repertoire changes were detected in paired samples for platinum-based therapy or immunotherapy, likely due to limited sample size (Fig. S6E–H). In the trastuzumab cohort, increased TCR convergence and clonality were observed only in the HER2 subtype, but not in luminal B2 (Fig. S6I–L). Dynamic changes in TCR metrics in unpaired and paired adjuvant samples were summarized (Table 1). These findings indicate that adjuvant chemotherapy, particularly taxane- or anthracycline–taxane–based regimens, drives peripheral T-cell clonal expansion and reduced diversity, with trastuzumab-associated repertoire changes restricted to the HER2-enriched subtype but not observed in luminal B2.Fig. 4The change in the TCRβ repertoire during adjuvant chemotherapy (Group 1). A The scheme of blood sample collection in adjuvant setting. B–E The profile of TCRβ convergent frequency, clonality, richness, and Shannon diversity in patient samples before and after adjuvant chemotherapy. F–I The corresponding profiles in paired adjuvant samples before and after adjuvant chemotherapyTable 1Comparison of systemic therapies administered in the adjuvant setting (Group 1)TotalTotalVisit 1Visit 2PairedVisit 1Visit 2PairedTCR metricsVisit 1Visit 2Mean ± SDMean ± SDp-valueNMean ± SDMean ± SDp-valueChemotherapyConvergenceAnthracycline546230.024 ± 0.0320.026 ± 0.0230.394210.021 ± 0.0220.027 ± 0.0230.191Taxane546420.024 ± 0.0320.034 ± 0.035*0.008390.021 ± 0.0210.034 ± 0.034*0.001A + T546830.024 ± 0.0320.029 ± 0.027*0.001780.021 ± 0.0200.028 ± 0.027*0.002ClonalityAnthracycline557230.283 ± 0.1110.344 ± 0.1360.040210.266 ± 0.0970.354 ± 0.135*0.010Taxane557420.283 ± 0.1110.335 ± 0.107*0.002410.284 ± 0.1320.333 ± 0.108* < 0.001A + T557830.283 ± 0.1110.340 ± 0.109* < 0.001780.270 ± 0.1090.341 ± 0.111* < 0.001RichnessAnthracycline5572327,434 ± 18,42323,345 ± 18,5320.2502127,430 ± 21,56822,985 ± 18,9570.633Taxane5574227,434 ± 18,42323,745 ± 14,4510.3274130,146 ± 24,96023,687 ± 14,6260.341A + T5578327,434 ± 18,42325,325 ± 15,8400.4897826,987 ± 18,33426,058 ± 15,9970.949Shannon diversityAnthracycline5572310.176 ± 2.2158.867 ± 2.6990.016219.890 ± 2.7618.65 ± 2.7220.113Taxane5574210.176 ± 2.2159.461 ± 1.967*0.0084110.126 ± 2.7329.48 ± 1.988*0.012A + T5578310.176 ± 2.2159.487 ± 2.056*0.0017810.295 ± 2.0929.498 ± 2.107*0.001Platinum-basedConvergencePlatinum54630.024 ± 0.0320.048 ± 0.037*0.04530.039 ± 0.0350.048 ± 0.0370.250ClonalityPlatinum55730.283 ± 0.1110.350 ± 0.1320.31230.252 ± 0.0570.35 ± 0.1320.168RichnessPlatinum557327,434 ± 18,42336,658 ± 13,9820.261346,327 ± 11,00836,658 ± 13,9820.552Shannon diversityPlatinum557310.176 ± 2.2159.849 ± 2.3180.647311.563 ± 0.6239.849 ± 2.3180.246ImmunotherapyConvergenceICI54640.024 ± 0.0320.017 ± 0.0110.83040.015 ± 0.0120.017 ± 0.0110.749ClonalityICI55740.283 ± 0.1110.345 ± 0.0660.15340.213 ± 0.090.345 ± 0.0660.185RichnessICI557427,434 ± 18,42311,470 ± 11,909*0.044422,193 ± 20,83411,470 ± 11,9090.315Shannon diversityICI557410.176 ± 2.2158.472 ± 1.9110.09349.739 ± 2.5038.472 ± 1.9110.473TrastuzumabLum B2 subtypeConvergenceTrastuzumab40220.022 ± 0.0290.026 ± 0.0250.195210.024 ± 0.0360.023 ± 0.0210.157ClonalityTrastuzumab40220.276 ± 0.1490.331 ± 0.1390.092210.294 ± 0.1690.331 ± 0.1430.111RichnessTrastuzumab402226,642 ± 17,46028,807 ± 18,2570.5942123,456 ± 17,96029,056 ± 18,6690.235Shannon diversityTrastuzumab402210.073 ± 2.9889.763 ± 2.7450.614219.319 ± 3.3529.760 ± 2.8130.489HER2 subtypeConvergenceTrastuzumab36180.027 ± 0.0350.042 ± 0.038*0.027170.021 ± 0.0220.037 ± 0.036*0.003ClonalityTrastuzumab36180.272 ± 0.1110.368 ± 0.104*0.003170.295 ± 0.1310.373 ± 0.112*0.025RichnessTrastuzumab361827,819 ± 19,41820,625 ± 89230.4721724,508 ± 18,05020,932 ± 84100.468Shannon diversityTrastuzumab361810.158 ± 2.3928.978 ± 1.629*0.026179.299 ± 2.6518.944 ± 1.7230.432A + T, anthracycline plus taxane; ICI, Immune checkpoint inhibitor therapy. *, Statistical significance. P-values were considered statistically significant at P < 0.05, except for the 'Total Visit 1 vs Visit 2' comparisons in the chemotherapy setting, where P < 0.0083 was considered significant after adjustment for multiple comparisons The change in the TCRβ repertoire during adjuvant chemotherapy (Group 1). A The scheme of blood sample collection in adjuvant setting. B–E The profile of TCRβ convergent frequency, clonality, richness, and Shannon diversity in patient samples before and after adjuvant chemotherapy. F–I The corresponding profiles in paired adjuvant samples before and after adjuvant chemotherapy Table 1Comparison of systemic therapies administered in the adjuvant setting (Group 1)TotalTotalVisit 1Visit 2PairedVisit 1Visit 2PairedTCR metricsVisit 1Visit 2Mean ± SDMean ± SDp-valueNMean ± SDMean ± SDp-valueChemotherapyConvergenceAnthracycline546230.024 ± 0.0320.026 ± 0.0230.394210.021 ± 0.0220.027 ± 0.0230.191Taxane546420.024 ± 0.0320.034 ± 0.035*0.008390.021 ± 0.0210.034 ± 0.034*0.001A + T546830.024 ± 0.0320.029 ± 0.027*0.001780.021 ± 0.0200.028 ± 0.027*0.002ClonalityAnthracycline557230.283 ± 0.1110.344 ± 0.1360.040210.266 ± 0.0970.354 ± 0.135*0.010Taxane557420.283 ± 0.1110.335 ± 0.107*0.002410.284 ± 0.1320.333 ± 0.108* < 0.001A + T557830.283 ± 0.1110.340 ± 0.109* < 0.001780.270 ± 0.1090.341 ± 0.111* < 0.001RichnessAnthracycline5572327,434 ± 18,42323,345 ± 18,5320.2502127,430 ± 21,56822,985 ± 18,9570.633Taxane5574227,434 ± 18,42323,745 ± 14,4510.3274130,146 ± 24,96023,687 ± 14,6260.341A + T5578327,434 ± 18,42325,325 ± 15,8400.4897826,987 ± 18,33426,058 ± 15,9970.949Shannon diversityAnthracycline5572310.176 ± 2.2158.867 ± 2.6990.016219.890 ± 2.7618.65 ± 2.7220.113Taxane5574210.176 ± 2.2159.461 ± 1.967*0.0084110.126 ± 2.7329.48 ± 1.988*0.012A + T5578310.176 ± 2.2159.487 ± 2.056*0.0017810.295 ± 2.0929.498 ± 2.107*0.001Platinum-basedConvergencePlatinum54630.024 ± 0.0320.048 ± 0.037*0.04530.039 ± 0.0350.048 ± 0.0370.250ClonalityPlatinum55730.283 ± 0.1110.350 ± 0.1320.31230.252 ± 0.0570.35 ± 0.1320.168RichnessPlatinum557327,434 ± 18,42336,658 ± 13,9820.261346,327 ± 11,00836,658 ± 13,9820.552Shannon diversityPlatinum557310.176 ± 2.2159.849 ± 2.3180.647311.563 ± 0.6239.849 ± 2.3180.246ImmunotherapyConvergenceICI54640.024 ± 0.0320.017 ± 0.0110.83040.015 ± 0.0120.017 ± 0.0110.749ClonalityICI55740.283 ± 0.1110.345 ± 0.0660.15340.213 ± 0.090.345 ± 0.0660.185RichnessICI557427,434 ± 18,42311,470 ± 11,909*0.044422,193 ± 20,83411,470 ± 11,9090.315Shannon diversityICI557410.176 ± 2.2158.472 ± 1.9110.09349.739 ± 2.5038.472 ± 1.9110.473TrastuzumabLum B2 subtypeConvergenceTrastuzumab40220.022 ± 0.0290.026 ± 0.0250.195210.024 ± 0.0360.023 ± 0.0210.157ClonalityTrastuzumab40220.276 ± 0.1490.331 ± 0.1390.092210.294 ± 0.1690.331 ± 0.1430.111RichnessTrastuzumab402226,642 ± 17,46028,807 ± 18,2570.5942123,456 ± 17,96029,056 ± 18,6690.235Shannon diversityTrastuzumab402210.073 ± 2.9889.763 ± 2.7450.614219.319 ± 3.3529.760 ± 2.8130.489HER2 subtypeConvergenceTrastuzumab36180.027 ± 0.0350.042 ± 0.038*0.027170.021 ± 0.0220.037 ± 0.036*0.003ClonalityTrastuzumab36180.272 ± 0.1110.368 ± 0.104*0.003170.295 ± 0.1310.373 ± 0.112*0.025RichnessTrastuzumab361827,819 ± 19,41820,625 ± 89230.4721724,508 ± 18,05020,932 ± 84100.468Shannon diversityTrastuzumab361810.158 ± 2.3928.978 ± 1.629*0.026179.299 ± 2.6518.944 ± 1.7230.432A + T, anthracycline plus taxane; ICI, Immune checkpoint inhibitor therapy. *, Statistical significance. P-values were considered statistically significant at P < 0.05, except for the 'Total Visit 1 vs Visit 2' comparisons in the chemotherapy setting, where P < 0.0083 was considered significant after adjustment for multiple comparisons Comparison of systemic therapies administered in the adjuvant setting (Group 1) A + T, anthracycline plus taxane; ICI, Immune checkpoint inhibitor therapy. *, Statistical significance. P-values were considered statistically significant at P < 0.05, except for the 'Total Visit 1 vs Visit 2' comparisons in the chemotherapy setting, where P < 0.0083 was considered significant after adjustment for multiple comparisons Correlation of the circulating TCRβ repertoire with neoadjuvant treatment responses and clinical outcomes (Group 2) Neoadjuvant chemotherapy was administered to treat locally advanced cancer prior to surgery. In the neoadjuvant setting, blood samples were collected before the initiation of therapy and after the completion of surgery (Fig. 5A). Patients receiving neoadjuvant chemotherapy showed increased convergence and clonality and decreased Shannon diversity of the circulating TCR repertoire (Fig. 5B–I). Moreover, patients receiving anthracycline plus taxane or platinum-based regimens showed increased TCR convergence and clonality, along with decreased Shannon diversity in the neoadjuvant setting (Fig. S7A–F). Notably, in the trastuzumab neoadjuvant cohort, increased TCR clonality and reduced Shannon diversity were observed only in the HER2 subtype, but not in luminal B2 (Fig. S7G–J). Dynamic changes in TCR metrics in unpaired and paired neoadjuvant samples were summarized (Table 2). Pathologic complete response (pCR) following neoadjuvant chemotherapy is associated with better outcomes in breast cancer [41]. Accordingly, we further analyzed the association between pCR and circulating TCR repertoires during neoadjuvant chemotherapy. Patients who achieved pCR had improved recurrence-free survival and event-free survival (Fig. S7K). The results showed no differences in TCR convergence, clonality, or Shannon diversity between non-pCR and pCR patients in both pre- and post-treatment blood samples. Patients who achieved pCR exhibited lower post-treatment TCR richness compared with non-pCR patients (Fig. 5J–M). However, the magnitude of richness changes from baseline to post-treatment did not differ significantly between the two groups (Fig. S7L), suggesting that the observed difference reflects the absolute post-treatment repertoire status. These results demonstrate that neoadjuvant chemotherapy, particularly anthracycline plus taxane or platinum-based regimens, induces clonal expansion and reduced diversity of circulating T cells, with trastuzumab-associated effects observed only in the HER2-enriched subtype but not in luminal B2, and that lower post-treatment TCR richness. A reduction in TCR richness in post-treatment patients may be associated with neoadjuvant chemotherapy-induced pCR.Fig. 5The change in the TCRβ repertoire during neoadjuvant chemotherapy (Group 2). A The scheme of blood sample collection in neoadjuvant setting. B–E The profile of TCRβ convergent frequency, clonality, richness, and Shannon diversity in patient samples before and after neoadjuvant chemotherapy. F–I The corresponding profiles in paired neoadjuvant samples before and after neoadjuvant chemotherapy. J–M The change in TCRβ convergent frequency, clonality, richness, and Shannon diversity before and after neoadjuvant chemotherapy in pCR and non-pCR patientsTable 2Comparison of systemic therapies administered in the neoadjuvant setting (Group 2)TotalTotalVisit 1Visit 2PairedVisit 1Visit 2PairedTCR metricsVisit 1Visit 2Mean ± SDMean ± SDp-valueNMean ± SDMean ± SDp-valueChemotherapyConvergenceAnthracycline10620.026 ± 0.0310.03 ± 0.0170.42420.017 ± 0.0100.03 ± 0.017NATaxane106210.026 ± 0.0310.056 ± 0.052*0.005190.042 ± 0.0490.057 ± 0.0550.055A + T106460.026 ± 0.0310.044 ± 0.0570.053330.026 ± 0.0310.048 ± 0.063*0.001ClonalityAnthracycline10820.280 ± 0.1270.288 ± 0.0920.78520.249 ± 0.1120.288 ± 0.092NATaxane108210.280 ± 0.1270.344 ± 0.1070.013200.298 ± 0.1330.348 ± 0.1080.088A + T108460.280 ± 0.1270.340 ± 0.130*0.008340.278 ± 0.1320.338 ± 0.14*0.007RichnessAnthracycline108227,223 ± 21,66637,001 ± 34,6770.603224,088 ± 10,66437,001 ± 34,677NATaxane1082127,223 ± 21,66623,950 ± 13,7030.7832027,379 ± 16,21924,487 ± 13,8310.430A + T1084627,223 ± 21,66623,784 ± 15,7630.5703429,944 ± 21,84425,358 ± 16,6050.397Shannon diversityAnthracycline108210.269 ± 2.43710.587 ± 2.5120.868210.916 ± 2.11610.587 ± 2.512NATaxane1082110.269 ± 2.4379.427 ± 1.8040.0862010.187 ± 2.2919.392 ± 1.8440.120A + T1084610.269 ± 2.4379.399 ± 2.3120.0443410.405 ± 2.6309.581 ± 2.446*0.038Platinum-basedConvergencePlatinum109400.026 ± 0.0300.057 ± 0.064* < 0.001320.036 ± 0.0440.060 ± 0.069*0.004ClonalityPlatinum111400.279 ± 0.1260.322 ± 0.114*0.032340.255 ± 0.1160.327 ± 0.119* < 0.001RichnessPlatinum1114027,051 ± 21,48625,119 ± 14,7790.9723433,674 ± 20,26024,745 ± 15,3040.050Shannon diversityPlatinum1114010.281 ± 2.4159.807 ± 1.9880.2683411.039 ± 2.1629.711 ± 2.090* < 0.001TrastuzumabLum B2 subtypeConvergenceTrastuzumab20150.043 ± 0.0540.050 ± 0.0450.330120.047 ± 0.0570.050 ± 0.0510.470ClonalityTrastuzumab20150.330 ± 0.1260.374 ± 0.0950.268120.368 ± 0.1210.372 ± 0.0960.892RichnessTrastuzumab201522,124 ± 17,37525,157 ± 15,7520.6101224,883 ± 18,88428,463 ± 15,9820.553Shannon diversityTrastuzumab20159.054 ± 2.1759.025 ± 1.8000.967129.053 ± 2.2919.191 ± 1.8790.823HER2 subtypeConvergenceTrastuzumab21140.023 ± 0.0170.037 ± 0.0450.73490.023 ± 0.0240.045 ± 0.0530.164ClonalityTrastuzumab22140.244 ± 0.1320.335 ± 0.126*0.046100.235 ± 0.0960.331 ± 0.134*0.007RichnessTrastuzumab221429,012 ± 16,45116,440 ± 12,964*0.0171029,422 ± 15,43119,914 ± 13,6340.232Shannon diversityTrastuzumab221410.867 ± 2.6968.919 ± 2.216*0.0311010.817 ± 2.5969.443 ± 2.2290.061A + T, anthracycline plus taxane; NA, not available. *, Statistical significance. P-values were considered statistically significant at P < 0.05, except for the 'Total Visit 1 vs Visit 2' comparisons in the chemotherapy setting, where P < 0.0083 was considered significant after adjustment for multiple comparisons The change in the TCRβ repertoire during neoadjuvant chemotherapy (Group 2). A The scheme of blood sample collection in neoadjuvant setting. B–E The profile of TCRβ convergent frequency, clonality, richness, and Shannon diversity in patient samples before and after neoadjuvant chemotherapy. F–I The corresponding profiles in paired neoadjuvant samples before and after neoadjuvant chemotherapy. J–M The change in TCRβ convergent frequency, clonality, richness, and Shannon diversity before and after neoadjuvant chemotherapy in pCR and non-pCR patients Table 2Comparison of systemic therapies administered in the neoadjuvant setting (Group 2)TotalTotalVisit 1Visit 2PairedVisit 1Visit 2PairedTCR metricsVisit 1Visit 2Mean ± SDMean ± SDp-valueNMean ± SDMean ± SDp-valueChemotherapyConvergenceAnthracycline10620.026 ± 0.0310.03 ± 0.0170.42420.017 ± 0.0100.03 ± 0.017NATaxane106210.026 ± 0.0310.056 ± 0.052*0.005190.042 ± 0.0490.057 ± 0.0550.055A + T106460.026 ± 0.0310.044 ± 0.0570.053330.026 ± 0.0310.048 ± 0.063*0.001ClonalityAnthracycline10820.280 ± 0.1270.288 ± 0.0920.78520.249 ± 0.1120.288 ± 0.092NATaxane108210.280 ± 0.1270.344 ± 0.1070.013200.298 ± 0.1330.348 ± 0.1080.088A + T108460.280 ± 0.1270.340 ± 0.130*0.008340.278 ± 0.1320.338 ± 0.14*0.007RichnessAnthracycline108227,223 ± 21,66637,001 ± 34,6770.603224,088 ± 10,66437,001 ± 34,677NATaxane1082127,223 ± 21,66623,950 ± 13,7030.7832027,379 ± 16,21924,487 ± 13,8310.430A + T1084627,223 ± 21,66623,784 ± 15,7630.5703429,944 ± 21,84425,358 ± 16,6050.397Shannon diversityAnthracycline108210.269 ± 2.43710.587 ± 2.5120.868210.916 ± 2.11610.587 ± 2.512NATaxane1082110.269 ± 2.4379.427 ± 1.8040.0862010.187 ± 2.2919.392 ± 1.8440.120A + T1084610.269 ± 2.4379.399 ± 2.3120.0443410.405 ± 2.6309.581 ± 2.446*0.038Platinum-basedConvergencePlatinum109400.026 ± 0.0300.057 ± 0.064* < 0.001320.036 ± 0.0440.060 ± 0.069*0.004ClonalityPlatinum111400.279 ± 0.1260.322 ± 0.114*0.032340.255 ± 0.1160.327 ± 0.119* < 0.001RichnessPlatinum1114027,051 ± 21,48625,119 ± 14,7790.9723433,674 ± 20,26024,745 ± 15,3040.050Shannon diversityPlatinum1114010.281 ± 2.4159.807 ± 1.9880.2683411.039 ± 2.1629.711 ± 2.090* < 0.001TrastuzumabLum B2 subtypeConvergenceTrastuzumab20150.043 ± 0.0540.050 ± 0.0450.330120.047 ± 0.0570.050 ± 0.0510.470ClonalityTrastuzumab20150.330 ± 0.1260.374 ± 0.0950.268120.368 ± 0.1210.372 ± 0.0960.892RichnessTrastuzumab201522,124 ± 17,37525,157 ± 15,7520.6101224,883 ± 18,88428,463 ± 15,9820.553Shannon diversityTrastuzumab20159.054 ± 2.1759.025 ± 1.8000.967129.053 ± 2.2919.191 ± 1.8790.823HER2 subtypeConvergenceTrastuzumab21140.023 ± 0.0170.037 ± 0.0450.73490.023 ± 0.0240.045 ± 0.0530.164ClonalityTrastuzumab22140.244 ± 0.1320.335 ± 0.126*0.046100.235 ± 0.0960.331 ± 0.134*0.007RichnessTrastuzumab221429,012 ± 16,45116,440 ± 12,964*0.0171029,422 ± 15,43119,914 ± 13,6340.232Shannon diversityTrastuzumab221410.867 ± 2.6968.919 ± 2.216*0.0311010.817 ± 2.5969.443 ± 2.2290.061A + T, anthracycline plus taxane; NA, not available. *, Statistical significance. P-values were considered statistically significant at P < 0.05, except for the 'Total Visit 1 vs Visit 2' comparisons in the chemotherapy setting, where P < 0.0083 was considered significant after adjustment for multiple comparisons Comparison of systemic therapies administered in the neoadjuvant setting (Group 2) A + T, anthracycline plus taxane; NA, not available. *, Statistical significance. P-values were considered statistically significant at P < 0.05, except for the 'Total Visit 1 vs Visit 2' comparisons in the chemotherapy setting, where P < 0.0083 was considered significant after adjustment for multiple comparisons Discussion In this study, we comprehensively profiled the circulating TCRβ repertoire in a large cohort of breast cancer patients using baseline peripheral blood samples from the VGH-TAYLOR study. Higher TCR clonality was associated with advanced-stage disease and poorer survival outcomes. Both adjuvant and neoadjuvant chemotherapy induced dynamic changes in the circulating TCR repertoire, characterized by increased convergence and clonality and decreased diversity. Notably, these changes were most pronounced under taxane or anthracycline-plus-taxane regimens in the adjuvant setting, and under anthracycline-plus-taxane or platinum-based regimens in the neoadjuvant setting. Trastuzumab-induced repertoire alterations were confined to the HER2-enriched subtype but were not observed in luminal B2 tumors. Moreover, reduced TCR richness following neoadjuvant chemotherapy was correlated with pCR. Together, these findings underscore the potential of circulating TCR repertoire metrics as non-invasive biomarkers for assessing disease burden and treatment responsiveness in breast cancer. Numerous companies offer TCR sequencing platforms that differ in starting materials (DNA or RNA), library preparation methods (multiplex PCR, 5' rapid amplification of cDNA ends [5' RACE], targeted enrichment), and sequencing levels (bulk or single-cell) [42]. The choice of assay substantially influences results and their comparability with other datasets. Bulk RNA sequencing captures only a subset of the universal TCR repertoire because of technical and biological constraints, and it often requires higher RNA input than some other methods, although it remains efficient and robust. Single-cell sequencing provides higher resolution and enables detection of rare clonotypes, but it is limited by higher costs, greater complexity, and the need for specialized infrastructure [43]. In this study, we profiled only the TCRβ chain using a bulk RNA-based assay, which enabled robust assessment of clonality, diversity, and convergence in peripheral blood. However, this approach does not capture α–β chain pairing, which is essential for determining antigen specificity [44]. We utilized the Oncomine™ TCR Beta-LR Assay, a multiplex PCR-based platform optimized for bulk RNA samples, offering broad TRBV and TRBJ coverage and reliable detection of clonal expansions and convergence. A key limitation is the potential underrepresentation of rare or low-frequency clonotypes; while the assay captures a substantial fraction of the repertoire, subtle convergence patterns may be missed. Although deeper sequencing could improve rare clone detection, practical constraints such as cost and diminishing returns often limit its application. Single-cell or paired α/β TCR sequencing offer higher resolution and antigen-level insights, but come with increased technical complexity and resource demands [45, 46]. This trade-off underscores the importance of choosing a platform aligned with the study’s objectives and scale. The immune system plays a role in prognosis and chemotherapy responses in breast cancer. However, there are limited comprehensive studies on TCRβ repertoire profiling in breast cancer. In the VGH-TAYLOR breast cancer cohort, TRBV28 and TRBJ2-1 were commonly used in TCRβ V(D)J gene rearrangements. Our results revealed that treatment-naïve patients with luminal B2 breast cancer had a lower diversity of TCRβ repertoires compared to luminal A subtype. Similarly, recent studies have indicated that the diversity of Vβ and Jβ genes in TCR repertoires is affected by HER2 expression in tumors [16]. The updated guidelines from the 2023 St. Gallen Consensus [47] and the International Ki67 Working Group [48] introduce more refined methodologies for biomarker assessment and treatment recommendations. These updated definitions, particularly for distinguishing luminal A-like from luminal B-like tumors, may lead to the reclassification of certain cases and could influence the observed associations between TCR repertoire metrics and clinicopathological features or treatment responses. Future studies may benefit from incorporating these evolving standards to enable more precise subtype stratification. Our analysis focused on circulating blood TCR repertoires; however, corresponding TILs data were not available for this cohort. Incorporating TILs data in future studies could provide complementary insights into localized immune dynamics. Homologous recombination deficiency (HRD) induces genomic instability and neoantigen generation, thereby influencing the immune microenvironment and treatment response [49]. Liao et al. reported that patients with HRD have lower TCR diversity and richness of TCR compared to patients with homologous recombination proficiency in anthracycline-, cyclophosphamide-, and taxane-sensitive TNBC [50]. Although the definition of HRR-related genes and the role of single mutations in driving HRD remain debated, we hypothesized that HRD-associated neoantigens could promote clonal expansion of reactive T cells and potentially enrich circulating TCR diversity. While multiple HRR mutations were not associated with significant changes in TCR metrics, patients with ARID1A mutations exhibited reduced circulating TCR richness. This observation should be interpreted with caution, as it is based on a single mutation and a single TCR metric (richness), and its biological implications for tumor–immune interactions remain unclear. ARID1A, a key component of the SWI/SNF chromatin remodeling complex, is frequently mutated, leading to genomic instability [51]. Its effects on the immune microenvironment are context-dependent: ARID1A deficiency can promote immune evasion and resistance to immune checkpoint blockade in some tumors, induce myeloid-derived suppressor cell–mediated immune suppression in others, or enhance immunogenicity and ICB sensitivity, as observed in gastrointestinal cancers [52–55]. Neoantigens derived from hotspot mutations have emerged as promising targets for cancer immunotherapy [56]. Chandran et al. demonstrated that the PIK3CA mutation can generate a public neoantigen that is immunogenic and targetable by TCR gene therapy, providing an innovative approach for cancer treatment [40]. Recent studies have identified the PIK3CA mutation-driven immune signature as a prognostic model for predicting responses to immunotherapy [57]. However, the relationship between PIK3CA mutations and immunogenicity remains unclear, and there is a lack of clinical trial evidence demonstrating the benefits of immunotherapy for breast cancer patients with PIK3CA mutations. In this study, we evaluated the association between PIK3CA mutations and the TCR repertoire and found no statistically significant differences among PIK3CA mutations in treatment-naïve breast cancer patients. Quantifying the TCR repertoire could measure tumor antigen-specific T cells and serve as a potential biomarker for chemoimmunotherapy response [29, 58]. Previous studies have shown significantly higher TCR diversity in peripheral blood compared to tumoral and peritumoral tissues in hepatocellular carcinoma [59]. Furthermore, changes in circulating TCR diversity following therapy may reflect pharmacodynamic effects related to a patient’s overall immune responsiveness, beyond localized anti-tumor responses in the tumor microenvironment [60, 61]. At the tumor level, higher TCR clonality after neoadjuvant chemotherapy has been observed in pCR patients compared with non-pCR patients, and tumor TCR clonality is further associated with longer disease-free survival [62]. Consistently, Park et al. also demonstrated that TCR diversity in TILs decreased following neoadjuvant chemotherapy in patients who achieved pCR [63]. In peripheral blood, TCR diversity decreases during neoadjuvant chemotherapy and is associated with a pCR, with regimens comprising combinations of anthracyclines, cyclophosphamide, taxanes, and occasionally carboplatin, trastuzumab, pertuzumab, or fluorouracil [16]. Chun et al. assessed changes in T-cell subset quantities and TCR clonal repertoire diversity in patients with metastatic TNBC undergoing treatment with pembrolizumab in combination with either paclitaxel or capecitabine. They reported that the combination of pembrolizumab with either paclitaxel or capecitabine resulted in similar patterns of circulating TCR metrics, with minimal changes in TCR repertoire richness and clonality during treatment [64]. In comparison, our study examined adjuvant and neoadjuvant chemotherapy regimens (primarily anthracycline and taxane-based) in early-stage breast cancer. We found that these regimens increased TCR convergence and clonality while simultaneously decreasing the diversity of circulating TCRs in breast cancer patients. Notably, a decrease in TCR richness post-treatment was associated with achieving a pCR following neoadjuvant chemotherapy. Conclusions Our comprehensive profiling of circulating TCRβ repertoires in breast cancer reveals that repertoire dynamics reflect both disease burden and treatment effects. Increased convergence and clonality, along with reduced diversity, were associated with advanced disease, poor survival, and specific chemotherapy regimens. These findings highlight the utility of the circulating TCR repertoire in monitoring treatment dynamics and its potential as a prognosis marker in breast cancer. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1