Skip to main content
NCBI home page
The Breast : Official Journal of the European Society of Mastology logoLink to The Breast : Official Journal of the European Society of Mastology
. 2026 Mar 13;87:104749. doi: 10.1016/j.breast.2026.104749

Immunohistochemical biomarkers to predict adjuvant chemotherapy response in patients with early breast cancer in the MATADOR trial (BOOG 2005-02)

Mark Opdam a, Nigel Ultee a, Jill JJ Geenen b,c, Annelot GJ van Rossum a, Ingrid AM Mandjes d, Sara Balduzzi e, Ingrid Hofland f, Jelle Wesseling a,g,h, Renee X Menezes i,j, Hendrika M Oosterkamp k, Sabine C Linn a,l,m,⁎; the MATADOR trialists' group for the Dutch Breast Cancer Research Group (BOOG)
PMCID: PMC13185652  PMID: 41864056

Abstract

Background

Intensified anthracycline-based regimens are as effective as standard anthracycline-based chemotherapy with taxanes in unselected high-risk patients with early breast cancer and come with their own toxicity profiles. Many biomarkers linked to taxane resistance have been identified, but none are used clinically. A predictive biomarker for taxane resistance or sensitivity would help personalise treatment.

Methods

In the randomised controlled trial MATADOR (ISRCTN61893718), 664 patients with pT1-3, pN0-3 breast cancer were treated with either docetaxel, doxorubicin, and cyclophosphamide (TAC) or dose-dense schedule of doxorubicin and cyclophosphamide (ddAC). IHC protocols for 13 previously proposed biomarkers (ABCB1, AR, BCL2, CyclinD1, EZH2, GSTP1, pH2AX, Ki67, MAPT, P53, Thioredoxin, TUBB, and TUBB3) were tested in all 577 clinical high-risk patients. The prognostic and predictive value was assessed in the total group, and in the hormone receptor-positive HER2-negative and in the triple-negative (TN) subgroup.

Results

Patients with TN EZH2high tumours have improved RFS after TAC treatment (n = 83 adjusted hazard ratio [adjHR] 0.35 [0.14–0.83]), while with EZH2low RFS improved after ddAC (n = 16 adjHR 6.31 [0.79–50.5]; Pinteraction0.01). Similar findings are observed for TUBB when stromal tumour-infiltrating lymphocytes (sTILs) are included in the model. TUBBhigh have improved outcome after TAC treatment (n = 48 adjHR 0.26 [0.08–0.80]), while for TUBBlow improved with ddAC (n = 47 adjHR 1.94 [0.58–6.50]; Pinteraction0.03).

Conclusions

Patients with TN EZH2low tumours benefit more from ddAC, while EZH2high have a better outcome after TAC. High tubulin expression may predict docetaxel benefit if adjusted for sTILs. Further research with more patients is needed to find the best use of these biomarkers.

Clinical trial identification

MATADOR, ISRCTN61893718, BOOG 2005-02, CKTO 2004-04.

Keywords: Early breast cancers, Prognostic and predictive biomarkers, Taxanes, EZH2, Tubulins

Highlights

  • •

    High EZH2 is a predictive marker for taxane benefit over ddAC in patients with early triple-negative breast cancer.

  • •

    High β-tubulin isotypes can also predict taxane benefit in the MATADOR trial when corrected for sTIL score.

  • •

    ABCB1, AR, BCL2, CycD1, GSTP1, pH2AX, Ki67, MAPT, P53 and also ThioR IHC did not show an interaction with treatment.

1. Introduction

Millions of women worldwide are diagnosed with breast cancer every year and hundreds of thousands die from the disease [1]. Taxanes, such as docetaxel and paclitaxel, are an important class of chemotherapy drugs that have been widely used in the treatment of breast cancer for many years [2]. They play a crucial role in the treatment of both early and metastatic breast cancer [3,4]. Taxanes added to an anthracycline-based chemotherapy regimen reduces the risk of recurrence on average by 14% [5]. Although this benefit is substantial and clinically meaningful, it also suggests that not all patients derive benefit from the addition of taxanes and might achieve better outcomes with an alternative chemotherapy regimen. Moreover, taxanes are associated with peripheral neuropathy in up to 47% of patients [6,7]. Therefore, it is essential to identify women who may not benefit from taxane treatment. Conversely, a biomarker that can enrich for patients likely to derive substantial benefit from a taxane is equally important. Especially considering the quest for anthracycline-free regimens to avoid potential serious late anthracycline-related toxicities such as heart failure and myeloproliferative syndromes [8].

Currently, there is no biomarker used in the clinic that identifies patients who are unlikely to benefit from the addition of a taxane to a chemotherapy regimen. Likewise, no biomarker is clinically available that enriches for patients likely to derive substantial benefit from a taxane. Several resistance mechanisms have been described over the years, and numerous biomarkers that play a role in taxane resistance have been reported in the literature [9]. However, no biomarker has yet reached clinical implementation. Before a biomarker can be clinically implemented, validation in an independent series is of utmost importance, but this is often lacking [10].

The MATADOR trial is a multicentre randomised controlled trial in which patients with early breast cancer are randomised between two chemotherapy regimens, either a chemotherapy regimen containing docetaxel, doxorubicin, and cyclophosphamide (TAC) or dose-dense schedule of doxorubicin and cyclophosphamide (ddAC) [11]. Importantly, both regimens are equally effective in the general population, making this trial an ideal setting to test biomarkers that can identify patient groups with taxane resistance or sensitivity. To achieve this properly, we will first test the biomarkers for their prognostic value independent of the allocated treatment. Second, we will test the biomarkers for survival prediction in patients treated with either dose-dense doxorubicin and cyclophosphamide or a taxane-based regimen. Finally, and most importantly, we will investigate if there is an interaction between the biomarker and the treatment arms.

In our literature search, we identified biomarkers that could explain resistance or sensitivity to one of the two chemotherapy regimens used in the MATADOR trial. We selected biomarkers for which reliable immunohistochemistry antibodies were available, as this technique is widely used in diagnostic laboratories and increases the likelihood of successful future implementation. We selected the following 13 biomarkers for which, based on literature, an interaction with treatment could be expected.

  • -

    Drug efflux via the efflux transporter P-glycoprotein (ABCB1 also known as MDR1) [12].

  • -

    Androgen receptor (AR) [[13], [14], [15]].

  • -

    B-cell lymphoma apoptosis regulator (BCL2) [12,[16], [17], [18], [19]].

  • -

    Cyclin D1 (CycD1) [20,21].

  • -

    Enhancer of zeste homolog 2 (EZH2) [22,23].

  • -

    Glutathione S-transferase P1 (GSTP1) [24].

  • -

    H2A histone family member X (pH2AX) [25,26].

  • -

    Marker of Proliferation Ki-67 (Ki67) [[27], [28], [29]].

  • -

    Microtubule-associated protein Tau (MAPT) [30].

  • -

    Tumour protein P53 (P53) [18,31,32].

  • -

    Thioredoxin-1 (ThioR) [12].

  • -

    β-tubulin isotype class I (TUBB) [12,33,34].

  • -

    β-tubulin isotype class III (TUBB3) [30,[34], [35], [36]].

In this manuscript, we report the prognostic and predictive value of the above-mentioned biomarkers, with the ultimate goal of identifying a biomarker that interacts with the treatments given in the MATADOR trial. Since stromal tumour-infiltrating lymphocytes (sTILs) have recently been discovered as a predictive biomarker for taxane benefit, we will also test whether the predictive utility of the biomarkers is independent of sTILs [37]. The results of this study will provide evidence of which biomarker is most promising for future clinical implementation.

2. Material and methods

The MATADOR trial (Microarray Analysis in breast cancer to tailor Adjuvant Drugs or Regimens, ISRCTN61893718) is a prospective, randomised, multicentre, non-blinded phase III trial conducted in the Netherlands during 2004-2012. Twenty-nine centres participated in this study. The primary objective was to predict benefit in recurrence-free survival (RFS) of either dose-dense or docetaxel-containing, anthracycline-based adjuvant therapy by discovering a gene expression profile that can predict recurrence [37]. The first results and the results of the SNP and toxicity data were published elsewhere [11,38]. Patients with pT1-3, N0-3 breast cancer were randomised (1:1) between six adjuvant cycles of doxorubicin (60mg/m2) and cyclophosphamide (600mg/m2 every 2 weeks), ddAC, and six cycles of docetaxel (75mg/m2), doxorubicin (50mg/m2) and cyclophosphamide (500mg/m2 every 3 weeks), TAC. The study protocol and amendments were approved by the ethical committee of the Netherlands Cancer Institute and the institutional review boards of the participating centres. The study was performed in accordance with Good Clinical Practice guidelines and with the Declaration of Helsinki. All patients provided written informed consent. Patients were followed for disease relapse for 10 years at regular intervals. Evaluation for relapse included physical examination and mammography. Adverse events were graded according to CTCAE version 3.0.

2.1. Immunohistochemistry biomarkers

The immunohistochemistry biomarkers tested were selected based on a literature search on ‘standard IHC markers’, ‘taxane resistance’, and ‘dose-dense sensitivity/resistance’. Supplemental Table 1 gives an overview of the biomarkers used in this study, the antibody used, and the scoring method. To calculate an interobserver kappa score, all immunohistochemistry stains were scored by two researchers in an online environment [39]. Cut-off values were either used from literature or the median value in the MATADOR trial was used. Supplemental Table 2 gives an overview of the cut-off values used.

2.2. RNA expression

The RNA analysis of the tumour samples of the MATADOR series has been previously and the results have been deposited in a public repository [37,40].

2.3. Statistics

The database was closed on 9 March 2023. Immunohistochemistry biomarker analyses were performed in the clinically high-risk patients based on modified Adjuvant! Online classification [41]. For each immunohistochemistry biomarker associations with recurrence-free survival (RFS) and overall survival (OS) were modelled using Cox proportional hazards regression. A multivariable adjusted hazard ratio (adjHR) for each biomarker was determined with adjustment for N-status, grade, and treatment, and stratification for immunohistochemical subtype and T-stage. All models with a significant biomarkers were tested for proportional hazard violation. Due to the validation nature of our study and because all biomarkers have a biological background, we did not correct the P values for multiple testing. Patients with missing values for a biomarker were excluded from the analysis of that specific biomarker.

3. Results

In this study, we included 577 clinically high-risk patients, with a median follow-up of 10 years and a total of 142 events. For 553 (96%) patients, at least one of the 13 immunohistochemistry (IHC) biomarkers (ABCB1, AR, BCL2, CycD1, EZH2, GSTP1, pH2AX, Ki67, MAPT, P53, ThioR, TUBB, TUBB3) was successfully assessed on tumour tissue (Fig. 1). The patient characteristics are shown in Table 1. Of the total patients, 474 (82%) had scores for all 13 IHC biomarkers, and 79 (14%) patients had scores for one or more, but not all, IHC biomarkers. For the remaining 24 (4%) patients, no IHC score was available. A more detailed per-patient analysis showed that most patients had IHC scores for all biomarkers, and that missing scores tended to occur in the same subset of patients (Supplemental Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

MATADOR patient flow chart. The clinical risk of recurrence was assessed using the modified Adjuvant! Online classification [41]. Success rate was calculated by dividing the number of patients with a score for that biomarker by the total number of patients in the analysis. TAC: docetaxel doxorubicin cyclophosphamide regimen, ddAC: dose-dense doxorubicin cyclophosphamide regimen, IHC: immunohistochemistry, HR+HER2-: Hormone receptor-positive HER2-negative, TN: triple-negative.

Table 1.

Characteristics of patients with or without available immunohistochemistry stains. For all variables, number of patients is shown with percentage of the total, except for stromal tumour-infiltrating lymphocytes (sTILs). For sTILs, mean and median scores are reported together with the range between first quartile (Q1) and third quartile (Q3). IHC: immunohistochemistry, ddAC: dose-dense doxorubicin cyclophosphamide regimen, TAC: docetaxel doxorubicin cyclophosphamide regimen, HR+: Hormone receptor-positive, RFS: recurrence-free survival.
MATADOR included patients n = 577
With all IHC stainings With 1-12 IHC stainings No IHC stainings
Clinical pathological features 474 (82.1) 79 (13.7) 24 (4.2)
Treatment ddAC 242 (51.1) 40 (50.6) 11 (45.8)
TAC 232 (48.9) 39 (49.4) 13 (54.2)
Age groups <50 years 216 (45.6) 30 (38.0) 15 (62.5)
≥50 years 258 (54.4) 49 (62.0) 9 (37.5)
Surgery breast conserving 256 (54.0) 32 (40.5) 13 (54.2)
mastectomy 218 (46.0) 47 (59.5) 11 (45.8)
T stage T1 209 (44.1) 29 (36.7) 11 (45.8)
T2 240 (50.6) 39 (49.4) 12 (50.0)
T3 23 (4.9) 10 (12.7) 1 (4.2)
T4 1 (0.2) 1 (1.3) 0 (0.0)
missing 1 (0.2) 0 (0.0) 0 (0.0)
N stage N0 74 (15.6) 11 (13.9) 5 (20.8)
N+ 400 (84.4) 68 (86.1) 19 (79.2)
Grade good 19 (4.0) 7 (8.9) 2 (8.3)
intermediate 200 (42.2) 43 (54.4) 12 (50.0)
poor 233 (49.2) 25 (31.6) 9 (37.5)
missing 22 (4.6) 4 (5.1) 1 (4.2)
Histology ductal 396 (83.5) 46 (58.2) 20 (83.3)
lobular 52 (11.0) 30 (38.0) 3 (12.5)
other 19 (4.0) 3 (3.8) 1 (4.2)
missing 7 (1.5) 0 (0.0) 0 (0.0)
Subtype HR+HER2- 365 (77.0) 69 (87.3) 21 (87.5)
Triple-negative 92 (19.4) 8 (10.1) 3 (12.5)
HER2-positive 17 (3.6) 2 (2.5) 0 (0.0)
sTILs median [Q1 – Q3]
mean		 10
14.4		 [5−20] 5
11.6		 [5−10] 7.5
9.1		 [5−10]
RFS event no 353 (74.5) 63 (79.7) 19 (79.2)
yes 121 (25.5) 16 (20.3) 5 (20.8)
Open in a new tab

For each staining, we show a representative negative and positive case in Supplemental Fig. 2. Most IHC scores had an interobserver kappa score >0.7. Only Thioredoxin was more difficult to evaluate, with a corresponding kappa score of 0.6 (Supplemental Table 3). The correlation between protein IHC and RNA expression of the corresponding gene is classified from none (Pearson correlation <0.2 for ABCB1, TUBB, and pH2AX) to strong (>0.6 for TUBB3, GSTP1, Ki67, and MAPT). Correlations between IHC scores and/or stromal tumour-infiltrating lymphocytes are mainly driven by hormone receptor status (Supplemental Fig. 3, Supplemental Fig. 4).

3.1. Prognostic value of the IHC biomarkers

The prognostic value of each biomarker, tested in a multivariable model, can depend on the subgroups in which the biomarker is tested and whether it is applied as a continuous or binary variable. All significant associations with recurrence-free survival (RFS) are marked with a red diamond in Fig. 2 and will be discussed below. For all the other biomarkers no significant association was seen (Supplemental Fig. 5).

Fig. 2.

Fig. 2

Open in a new tab

Prognostic value of IHC biomarkers with significant results for recurrence-free survival. Each biomarker is tested in all patients (left), patients with hormone receptor-positive HER2-negative tumours (HR+HER2-: middle) and with triple-negative tumours (TN: right). Continuous biomarker (row 1-3), and binary score for the biomarker with the cut-off point indicated (row 4-6). Furthermore the dose-dense doxorubicin cyclophosphamide (ddAC) treated patients are analyzed separately (in row 2 and 5), and for docetaxel doxorubicin cyclophosphamide (TAC) treated patients (in row 3 and 6). For each multivariable model the number of patients and events is given. For CyclinD1 the cut-off for HR+HER2-is adjusted because of a strong association with immunohistochemical subtype (indicated by #). The adjusted hazard ratio (adjHR) for the biomarker is depicted with the diamond and if significant marked in red. The 95% confidence interval (95% CI) for the biomarker is shown in the last column. The adjusted hazard model included N-status, grade, tumour size, immunohistochemical subtype, and treatment if not specified.

A binary ABCB1 score is a significant prognostic factor for a worse survival in the MATADOR patients randomised to the ddAC treatment arm (adjHR: 1.74[1.03 – 2.92] P = 0.04). This association is also found in the subgroup of patients with a HR+HER2-tumour (including both treatments; adjHR: 1.58 [1.05 – 2.37] P = 0.03, and only ddAC treatment; adjHR: 2.13 [1.20 – 3.80] P = 0.01). For the patients with TN tumours, ABCB1 is associated with improved RFS, which contrasts with the effect observed in patients with HR+HER2-tumours. Continuous ABCB1 score for patients with TN tumours treated with TAC was predictive for improved RFS (adjHR [95% confidence interval]: 0.33 [0.13 – 0.83] P = 0.02). CYCD1 with a binary score is also associated with an improved RFS for the TAC-treated patients with TN tumours (adjHR: 0.30 [0.09 – 0.99] P = 0.05). EZH2 is positively associated with RFS in TAC-treated patients in the TN subgroup (continuous adjHR: 0.82 [0.68 – 0.99] P = 0.03, binary cut-off adjHR: 0.16 [0.04 – 0.64] P = 0.01). P53 is only a significant biomarker in patients with HR+HER2-tumours after ddAC treatment with a continuous score (adjHR: 0.89 [0.79 – 1.00] P = 0.05). The last biomarker with significant prognostic value is TUBB3. For TUBB3, the continuous score in all patients is associated with a worse RFS in patients treated with ddAC (adjHR: 1.07 [1.01 – 1.14] P = 0.03). The same effect for TUBB3 is seen with a binary cut-off in ddAC-treated patients (adjHR: 1.71 [1.02 – 2.84] P = 0.04). In patients with TN tumours, TUBB3 is significantly associated with outcome, but only after ddAC treatment (continuous score adjHR: 1.17 [1.03 – 1.33] P = 0.01; binary score adjHR: 3.11 [1.08 – 8.96] P = 0.04).

All other biomarkers are not significantly associated with RFS in these subgroups. In a multivariable model including sTILs and with OS as endpoint, the results are similar (Supplemental Fig. 6). In addition, continuous BCL2 score in all patients also results in a significant association with improved overall survival (adjHR: 0.93 [0.86 – 0.99] P = 0.02, and ddAC-treated adjHR: 0.91 [0.84 – 0.99]).

3.2. Predictive value of EZH2, ABCB1 and TUBB

To test if a biomarker can be used to predict the optimised treatment, we group the patients into low and high based on the predefined cut-off for each biomarker and calculate the interaction with treatment. The multivariable prediction model is tested for all patients, for patients with HR+HER2-tumours and for patients with TN tumours (Fig. 3).

Fig. 3.

Fig. 3

Open in a new tab

Predictive value of IHC biomarkers for recurrence-free survival. Biomarker split into low and high with predefined cut-off. Number of events and cases is calculated in each cohort. Adjusted hazard ratio (adjHR) depicted together with the 95% confidence interval (CI) of the multivariable model including N-status, grade, tumour size, and immunohistochemical subtype. Pinteraction (Pint) between biomarker and treatment is shown in the last column with significant interactions shown with a red diamond. For BCL2, CyclinD1, GSTP1 and MAPT the cut-offs are not the same for all subgroups because of a strong association with immunohistochemical subtype (indicated by #). (A) All patients. (B) Patients with hormone receptor-positive, HER2-negative tumours (HR+HER2-). (C) Patients with triple-negative tumours (TN).

EZH2 can predict the preferred treatment for patients with TN tumours (Fig. 3C). Patients with EZH2high have an improved RFS after TAC treatment (adjHR: 0.35 [0.14 – 0.83] P = 0.02), while the point estimate for patients with EZH2low suggests they derive more benefit from ddAC (adjHR: 6.31 [0.79 – 50.5] P = 0.08, Pinteraction = 0.01). The effect remains significant if sTILs are included in the model (Supplemental Fig. 7C) and also with OS as endpoint (Supplemental Fig. 8C). EZH2 is not a predictive biomarker in patients with HR+HER2- tumours (Fig. 3B, Supplemental Fig. 7B, Supplemental Fig. 8B).

In the TN subgroup, TUBB significantly interacts with treatment if sTILs are included in the multivariable analysis (Supplemental Fig. 7C). In the RFS model with sTILs included, patients with TUBBhigh have more benefit from TAC (adjHR: 0.26 [0.08 – 0.80] P = 0.02), while in patients with TUBBlow the point estimate suggests that they might have more benefit from ddAC (adjHR: 1.94 [0.58 – 6.50] P = 0.28; Pinteraction = 0.03). In the RFS model without sTILs, the interaction is not significant (Pinteraction = 0.07), nor with OS as endpoint (Pinteraction = 0.22). TUBB3high is also associated with an improved RFS after TAC treatment (adjHR 0.29 [0.10 – 0.87] P = 0.03; Fig. 3C). In the TUBBlow subgroup, no treatment effect is seen and the interaction is not significant (adjHR 1.01 [0.36 – 2.79] P = 0.99; Pinteraction = 0.13).

All other biomarkers do not show a significant interaction with treatment, nor in the models with sTILs included, nor with OS as endpoint (Supplemental Fig. 7 and Supplemental Fig. 8).

Because accumulating data suggest that sTILs are an important biomarker for taxane response [37,42,43], we further examined the relationship between sTILs and the most promising IHC biomarkers (EZH2 and tubulins) on outcome. Based on the observation that high expression of both tubulin markers was associated with a preference for TAC in patients with triple-negative tumours, we concluded that both isoforms may contribute to taxane efficacy (Fig. 3C). For that reason we combined both tubulin markers and analysed them together. In Fig. 4 we grouped patients based on the biomarker and sTILs and calculated the number of events for both treatment arms. For patients with TN tumours and both sTILs and EZH2 high the strongest benefit of taxane treatment is seen (Fig. 4A). Two events in 25 cases treated with TAC, compared to 10 events in 25 cases after ddAC treatment (P = 0.02). A signal in the same direction is observed when both tubulin markers are used as biomarker instead of EZH2. In patients with TN tumours with high sTILs, high TUBB and high TUBB3, no RFS events were observed after treatment with TAC (n = 6), while almost all patients treated with ddAC had an event (6 out of 7; P = 0.005; Fig. 4A). In the other subgroups, no clear difference between treatment arms in RFS events was seen (HR+HER2-tumours Fig. 4B).

Fig. 4.

Fig. 4

Open in a new tab

Recurrence-free survival events in patients grouped based on stromaltumour-infiltrating lymphocytes (sTILs) and IHC biomarker score. Total number of patients for all four groups is shown on the y-axis, with ddAC in blue and TAC in red, patients with events shown in dark. EZH2 as IHC biomarker in left panel. TUBB and TUBB3 as combined biomarker in right panel. (A) Patients with triple-negative tumours, sTILs scored ≥20% as positive. (B) Patients with hormone receptor-positive HER2-negative tumours, sTILs scored ≥10% as positive.

4. Discussion

In the MATADOR randomised trial, we tested thirteen biologically relevant biomarkers to identify patients who might benefit more from either taxane-containing chemotherapy or from dose-dense chemotherapy regimens without a taxane. We find two biomarkers that appear to be relevant for treatment choice in patients with TN tumours. High EZH2 is predictive of improved outcome after taxane-containing therapy, while patients with low EZH2 treated with ddAC have an improved outcome. Likewise, patients with TN tumours and high tubulin expression fare better after TAC than after ddAC.

EZH2 is a marker for higher-grade and TN subtype [22], as observed in the MATADOR series. The previously found relation with poor survival is not seen in patients with TN tumours allocated to the TAC arm [44,45]. The high correlation with Ki67 suggests that EZH2 could be a proliferation marker. The correlation between EZH2 and grade strengthens this claim and has been observed previously [22]. Nonetheless, the EZH2 score has a significant interaction with treatment (corrected for grade), while the most used proliferation marker in breast cancer, Ki67 does not (although the point estimates are comparable). EZH2 expression is also correlated with sTILs; however, the significant interaction between EZH2 and treatment remains when stratified for sTILs and also with OS as endpoint. This demonstrates the added predictive value of EZH2 on top of sTILs. Of interest might be a previously found relation between the ability of EZH2 to modulate the tumour microenvironment. A working mechanism that has been described is that EZH2 suppresses CD8+T-cell infiltration and that this can be reversed by taxane treatment [46,47]. Taxane treatment re-activates the immune cells by releasing extracellular vesicles and induces immunogenic cell death [42,48]. This hypothesis is in line with our finding that the largest difference in events between TAC and ddAC treatment is seen in patients with both high EZH2 and high sTILs (Fig. 4A).

Tubulins are a well-known preclinical marker for taxane response, but have never been clinically validated in the context of sTILs and hormone receptor status. Microtubules are formed by heterodimers of α/β isoforms of tubulins and play a critical role in cell division [49]. Taxanes can stabilise the microtubules and thereby interfere with the cell division of tumour cells. Mutations and isotype composition of the tubulins can result in resistance to microtubule-stabilising agents [50]. Previous studies are in line with our results that tubulins are important biomarkers for taxane benefit. In the TAX-303 phase III trial, doxorubicin was compared to docetaxel in patients with locally advanced or metastatic breast cancer [35]. Patients with high TUBB3 had a significantly better response to docetaxel compared to doxorubicin, without taking into account hormone receptor-status or sTILs. The improved RFS in patients with high TUBB3 that we observed (Fig. 3C) suggests that it could also be used as a biomarker in the early setting. The only other study carried out in the (neo)adjuvant setting randomised patients to either ixabepilone or paclitaxel and evaluated the interaction of TUBB3 expression with taxane treatment [36]. This study did not find an interaction between TUBB3 and treatment and this may be explained by the fact that both treatment arms contained spindle poisons. Others found an association between high TUBB3 and disease progression after taxane treatment, without comparison to another treatment [34,51,52]. In studies without a comparator arm, that contains no treatment or a treatment with a different mechanism of action, it is impossible to distinguish between prognostic and predictive value of the biomarker [53]. In the MATADOR trial patients with TN tumours and high TUBB also show a worse outcome after ddAC treatment. The lack of association between the two tubulin biomarkers (Supplemental Fig. 3) and the association with different clinical features (Supplemental Fig. 4), suggests that a different group of patients is identified. The strong correlation of TUBB3 with ENSG00000258947 could indicate high specificity of the antibody, while the TUBB antibody may be less specific because of the lack of correlation of TUBB protein expression with its corresponding mRNA expression. Alternatively, post-translational modifications or other reasons may explain the lack of correlation between mRNA and protein [54]. Nevertheless, both biomarkers show a similar signal with treatment response in patients with TNBC and this observation could indicate that it is not one particular tubulin isoform that is important for the taxanes to work. For that reason we have combined both tubulin IHC biomarkers and when both biomarkers are high, the interaction with treatment increases, but consequently decreases the number of patients in this subgroup (data not shown). The observation that tumours with high expression of tubulins have more benefit from taxanes is in line with the original mechanism of the drug [55]. That only a significant interaction is found for TUBB in patients with TN tumours after correcting for sTILs status suggests that sTILs status is a more important predictor for taxane benefit [37,42].

Overexpression of ABCB1 is associated with multidrug resistance [56,57] and also in the MATADOR series we observe decreased recurrence-free and overall survival, but only in patients with HR+HER2-tumours. In contrast, continuous ABCB1 score in patients with TN tumours are linked to improved survival. This highlights the importance to stratify for the patient subgroup in which the prognostic biomarker is tested [36].

For all thirteen biomarkers there is substantial agreement in scores between observers, indicating that the scoring can reliably be performed. Most biomarkers have at least a moderate correlation with the RNA expression of the corresponding gene, demonstrating that the antibody probably is on target. On the other hand, the lack of correlation between RNA expression and IHC score for some biomarkers can be explained by the fact that the RNA is derived from all the tissue in and around the tumour. In addition, post-transcriptional processes can influence the protein expression [58].

One of the limitations of this study was the small number of events in the subgroups. However, the discrepant results observed in some biomarkers across subgroups indicated that subgroup analyses are necessary. Notably, the smallest subgroup (TN) showed the most significant signals, suggesting that important biomarkers can be more easily identified in a more homogeneous small cohort than in a heterogeneous large cohort. Currently, patients receive chemotherapy regimens tailored to their immunohistochemical subtypes, whereas this was not yet implemented at the start of the MATADOR trial. Meanwhile, regimens have evolved, however anthracyclines and taxanes still remain the most commonly used chemotherapeutic agents [59]. In this study we test 13 IHC biomarkers individually, with or without sTILs on tumour material collected before the start of treatment. We did not test these biomarkers in combination with each other. We believe that a single IHC biomarker has the most potential for clinical implementation. However, we also recognise that resistance mechanisms are complex, involve many proteins, and that treatment pressure can lead to the and outgrowth of underrepresented clones. Furthermore, a binary cut-off is necessary for treatment prediction, but there is no consensus for most biomarkers. For biomarkers with conflicting cut-offs found in literature, we have used the median score found in the MATADOR series. We have to acknowledge that these cut-offs are not optimised for predicting treatment interaction. However, if we would have used a data-driven approach to determine the optimal cut-offs, then the current study could not have been a validation study. Therefore, we chose to use predefined cut-offs and validate previously suggested biomarkers in this randomised controlled trial.

5. Conclusion

In the MATADOR trial, EZH2 IHC has shown potential as a predictive biomarker for patients with TN breast cancer. Specifically, patients with low EZH2 benefit more from ddAC, while patients with high EZH2 have improved survival after TAC, independent of N-status, grade, tumour size, and sTILs. TUBB and TUBB3 also appear to be promising biomarkers; high tubulin expression may predict docetaxel benefit if adjusted for sTILs. All of the other ten biomarkers tested failed to show a significant interaction with treatment. Further research with a larger number of patients is needed to determine the optimal use of these biomarkers.

CRediT authorship contribution statement

Mark Opdam: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Nigel Ultee: Writing – review & editing, Visualization, Software, Formal analysis, Data curation. Jill J.J. Geenen: Writing – review & editing, Writing – original draft, Methodology, Investigation. Annelot G.J. van Rossum: Writing – review & editing, Resources, Project administration, Methodology, Investigation, Conceptualization. Ingrid A.M. Mandjes: Writing – review & editing, Resources, Project administration, Investigation, Data curation, Conceptualization. Sara Balduzzi: Writing – review & editing, Validation, Resources, Data curation. Ingrid Hofland: Writing – review & editing, Resources. Jelle Wesseling: Writing – review & editing, Resources. Renee X. Menezes: Writing – review & editing, Writing – original draft, Software, Resources, Investigation, Formal analysis, Data curation. Hendrika M. Oosterkamp: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Sabine C. Linn: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization.

Disclosure

S.C.L. and H.M.O. received an unrestricted institutional research grant from Sanofi and Amgen for the MATADOR study (ISRCTN61893718). H.M.O. also received institutional research support from Roche. M.K. received institutional research funding from Roche, BMS, and AstraZeneca. Additionally, S.C.L. received institutional research support from Agendia, Amgen, AstraZeneca, Eurocept, Genentech, Immunomedics (now Gilead), Roche, Sanofi, and TESARO (now GSK). S.C.L. serves on the advisory board for AstraZeneca, with compensation paid to the institute, and for Cergentis on a pro bono basis. S.C.L. also receives funding for educational activities from Daiichi Sankyo, paid to the institute. H.M.O. is an advisory board member for Roche, Pfizer, and Novartis. M.K. serves on the advisory boards for AstraZeneca, BMS, Roche, MSD, and Daiichi, with compensation paid to the institute.

Ethical approval

The study protocol and amendments were approved by the ethical committee of the Netherlands Cancer Institute [40]. The study adhered to Good Clinical Practice guidelines and the Declaration of Helsinki. All patients gave written informed consent to participate in the trial and to use their surgically removed tumour tissue for translational research.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the authors used Microsoft Copilot, based on the GPT-4 architecture for English editing. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Funding

This study was supported by the Dutch Cancer Society (CKTO 2004-04) and received unrestricted research grants from Sanofi and Amgen. The funders did not influence the study design, data collection, analysis, interpretation, manuscript writing, or the decision to submit the paper for publication.

Acknowledgements

First, we extend our heartfelt thanks to all the patients and their families. We are also grateful to the study teams at the participating centres, all members of the MATADOR trialists' group (Supplemental Table 4), the Data Centre of the Netherlands Cancer Institute for data collection, the Core Facility – Molecular Pathology and Biobanking of the Netherlands Cancer Institute for assistance with immunohistochemistry stains, the Genomics Core Facility of the Netherlands Cancer Institute for their work on RNA sequencing, Willemijn Vader, and the Dutch Breast Cancer Research Group (BOOG) for coordinating the study.

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.breast.2026.104749.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Supplementary Tables 1–3 and Figs. 1-8. Anonymised patient data that support the finding of this study is shared in a Mendeley repository [40] and RNA sequencing data on GEO (GSE167977). Also on Mendeley a bioinformatics file with the code for generation of the figures is shared. Additional information is available from the lead contact upon request.

Multimedia component 1
mmc1.docx (15.5MB, docx)

References

  • 1.Organization WH. Breast cancer. Updated 13-04-2024. Accessed 2-July-2024, https://www.who.int/news-room/fact-sheets/detail/breast-cancer.
  • 2.King K.M., Lupichuk S., Baig L., et al. Optimal use of taxanes in metastatic breast cancer. Curr Oncol. May 2009;16(3):8–20. doi: 10.3747/co.v16i3.377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Bedard P.L., Di Leo A., Piccart-Gebhart M.J. Taxanes: optimizing adjuvant chemotherapy for early-stage breast cancer. Nat Rev Clin Oncol. Jan 2010;7(1):22–36. doi: 10.1038/nrclinonc.2009.186. [DOI] [PubMed] [Google Scholar]
  • 4.Palmieri C., Krell J., James C.R., et al. Rechallenging with anthracyclines and taxanes in metastatic breast cancer. Nat Rev Clin Oncol. Oct 2010;7(10):561–574. doi: 10.1038/nrclinonc.2010.122. [DOI] [PubMed] [Google Scholar]
  • 5.Braybrooke J., Bradley R., Gray R., et al. Anthracycline-containing and taxane-containing chemotherapy for early-stage operable breast cancer: a patient-level meta-analysis of 100 000 women from 86 randomised trials. Lancet. 2023;401 doi: 10.1016/S0140-6736(23)00285-4. 1277-1292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Martin M., Pienkowski T., Mackey J., et al. Adjuvant docetaxel for node-positive breast cancer. N Engl J Med. 2005;352(22):2302–2313. doi: 10.1056/NEJMoa043681. [DOI] [PubMed] [Google Scholar]
  • 7.von Minckwitz G., Kümmel S., Vogel P., et al. Intensified neoadjuvant chemotherapy in early-responding breast cancer: phase III randomized GeparTrio study. JNCI: J Natl Cancer Inst. 2008;100(8):552–562. doi: 10.1093/jnci/djn089. [DOI] [PubMed] [Google Scholar]
  • 8.Anthracycline-containing and taxane-containing chemotherapy for early-stage operable breast cancer: a patient-level meta-analysis of 100 000 women from 86 randomised trials. Lancet. Apr 15 2023;401(10384):1277–1292. doi: 10.1016/s0140-6736(23)00285-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Murray S., Briasoulis E., Linardou H., Bafaloukos D., Papadimitriou C. Taxane resistance in breast cancer: mechanisms, predictive biomarkers and circumvention strategies. Cancer Treat Rev. Nov 2012;38(7):890–903. doi: 10.1016/j.ctrv.2012.02.011. [DOI] [PubMed] [Google Scholar]
  • 10.Diamandis E.P. Cancer biomarkers: can we turn recent failures into success? JNCI: J Natl Cancer Inst. 2010;102(19):1462–1467. doi: 10.1093/jnci/djq306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.van Rossum A.G.J., Kok M., van Werkhoven E., et al. Adjuvant dose-dense doxorubicin-cyclophosphamide versus docetaxel-doxorubicin-cyclophosphamide for high-risk breast cancer: first results of the randomised MATADOR trial (BOOG 2004-04) Eur J Cancer. Oct 2018;102:40–48. doi: 10.1016/j.ejca.2018.07.013. [DOI] [PubMed] [Google Scholar]
  • 12.Noguchi S. Predictive factors for response to docetaxel in human breast cancers. Cancer Sci. 2006;97(9):813–820. doi: 10.1111/j.1349-7006.2006.00265.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Loibl S., Müller B.M., von Minckwitz G., et al. Androgen receptor expression in primary breast cancer and its predictive and prognostic value in patients treated with neoadjuvant chemotherapy. Breast Cancer Res Treat. Nov 2011;130(2):477–487. doi: 10.1007/s10549-011-1715-8. [DOI] [PubMed] [Google Scholar]
  • 14.Castellano I., Allia E., Accortanzo V., et al. Androgen receptor expression is a significant prognostic factor in estrogen receptor positive breast cancers. Breast Cancer Res Treat. Dec 2010;124(3):607–617. doi: 10.1007/s10549-010-0761-y. [DOI] [PubMed] [Google Scholar]
  • 15.Witzel I., Loibl S., Wirtz R., et al. Androgen receptor expression and response to chemotherapy in breast cancer patients treated in the neoadjuvant TECHNO and PREPARE trial. Br J Cancer. Dec 2019;121(12):1009–1015. doi: 10.1038/s41416-019-0630-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Lee K.H., Im S.A., Oh D.Y., et al. Prognostic significance of bcl-2 expression in stage III breast cancer patients who had received doxorubicin and cyclophosphamide followed by paclitaxel as adjuvant chemotherapy. BMC Cancer. Apr 12 2007;7:63. doi: 10.1186/1471-2407-7-63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Gasparini G., Barbareschi M., Doglioni C., et al. Expression of bcl-2 protein predicts efficacy of adjuvant treatments in operable node-positive breast cancer. Clin Cancer Res. Feb 1995;1(2):189–198. [PubMed] [Google Scholar]
  • 18.Mottolese M., Benevolo M., Del Monte G., et al. Role of P53 and BCL-2 in high-risk breast cancer patients treated with adjuvant anthracycline-based chemotherapy. J Cancer Res Clin Oncol. 2000/11/01 2000;126(12):722–729. doi: 10.1007/PL00008478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Callagy G.M., Webber M.J., Pharoah P.D., Caldas C. Meta-analysis confirms BCL2 is an independent prognostic marker in breast cancer. BMC Cancer. May 29 2008;8:153. doi: 10.1186/1471-2407-8-153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Abdelrahman A.E., Fathy A., Elsebai E.A., Nawar N., Etman W.M. Prognostic impact of Apaf-1, Cyclin D1, and AQP-5 in serous ovarian carcinoma treated with the first-line chemotherapy. Ann Diagn Pathol. Aug 2018;35:27–37. doi: 10.1016/j.anndiagpath.2018.02.005. [DOI] [PubMed] [Google Scholar]
  • 21.Li Z., Cui J., Yu Q., Wu X., Pan A., Li L. Evaluation of CCND1 amplification and CyclinD1 expression: diffuse and strong staining of CyclinD1 could have same predictive roles as CCND1 amplification in ER positive breast cancers. Am J Transl Res. 2016;8(1):142–153. [PMC free article] [PubMed] [Google Scholar]
  • 22.Guo S., Li X., Rohr J., et al. EZH2 overexpression in different immunophenotypes of breast carcinoma and association with clinicopathologic features. Diagn Pathol. Apr 26 2016;11:41. doi: 10.1186/s13000-016-0491-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Puppe J., Opdam M., Schouten P.C., et al. EZH2 is overexpressed in BRCA1-like breast tumors and predictive for sensitivity to high-dose platinum-based chemotherapy. Clin Cancer Res. Jul 15 2019;25(14):4351–4362. doi: 10.1158/1078-0432.Ccr-18-4024. [DOI] [PubMed] [Google Scholar]
  • 24.Arai T., Miyoshi Y., Kim S.J., et al. Association of GSTP1 expression with resistance to docetaxel and paclitaxel in human breast cancers. Eur J Surg Oncol. Jul 2008;34(7):734–738. doi: 10.1016/j.ejso.2007.07.008. [DOI] [PubMed] [Google Scholar]
  • 25.Nagelkerke A., van Kuijk S.J., Sweep F.C., et al. Constitutive expression of γ-H2AX has prognostic relevance in triple negative breast cancer. Radiother Oncol. Oct 2011;101(1):39–45. doi: 10.1016/j.radonc.2011.07.009. [DOI] [PubMed] [Google Scholar]
  • 26.Wilson A.J., Holson E., Wagner F., et al. The DNA damage mark pH2AX differentiates the cytotoxic effects of small molecule HDAC inhibitors in ovarian cancer cells. Cancer Biol Ther. Sep 15 2011;12(6):484–493. doi: 10.4161/cbt.12.6.15956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Kim T., Han W., Kim M.K., et al. Predictive significance of p53, Ki-67, and Bcl-2 expression for pathologic complete response after neoadjuvant chemotherapy for triple-negative breast cancer. J Breast Cancer. 2015;18(1):16–21. doi: 10.4048/jbc.2015.18.1.16. 3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Tham Y.L., Gomez L.F., Mohsin S., et al. Clinical response to neoadjuvant docetaxel predicts improved outcome in patients with large locally advanced breast cancers. Breast Cancer Res Treat. Dec 2005;94(3):279–284. doi: 10.1007/s10549-005-9020-z. [DOI] [PubMed] [Google Scholar]
  • 29.Vincent-Salomon A., Rousseau A., Jouve M., et al. Proliferation markers predictive of the pathological response and disease outcome of patients with breast carcinomas treated by anthracycline-based preoperative chemotherapy. Eur J Cancer. Jul 2004;40(10):1502–1508. doi: 10.1016/j.ejca.2004.03.014. [DOI] [PubMed] [Google Scholar]
  • 30.Pentheroudakis G., Batistatou A., Kalogeras K.T., et al. Prognostic utility of β-tubulin isotype III and correlations with other molecular and clinicopathological variables in patients with early breast cancer: a translational Hellenic cooperative oncology group (HeCOG) study. Breast Cancer Res Treat. 2011/05/01 2011;127(1):179–193. doi: 10.1007/s10549-011-1427-0. [DOI] [PubMed] [Google Scholar]
  • 31.Hamilton A., Larsimont D., Paridaens R., et al. A study of the value of p53, HER2, and Bcl-2 in the prediction of response to doxorubicin and paclitaxel as single agents in metastatic breast cancer: a companion study to EORTC 10923. Clin Breast Cancer. Oct 2000;1(3):233–240. doi: 10.3816/CBC.2000.n.020. discussion 241-2. [DOI] [PubMed] [Google Scholar]
  • 32.Wahl A.F., Donaldson K.L., Fairchild C., et al. Loss of normal p53 function confers sensitization to taxol by increasing G2/M arrest and apoptosis. Nat Med. Jan 1996;2(1):72–79. doi: 10.1038/nm0196-72. [DOI] [PubMed] [Google Scholar]
  • 33.Hasegawa S., Miyoshi Y., Egawa C., et al. Mutational analysis of the class I beta-tubulin gene in human breast cancer. Int J Cancer. Sep 1 2002;101(1):46–51. doi: 10.1002/ijc.10575. [DOI] [PubMed] [Google Scholar]
  • 34.Tommasi S., Mangia A., Lacalamita R., et al. Cytoskeleton and paclitaxel sensitivity in breast cancer: the role of beta-tubulins. Int J Cancer. May 15 2007;120(10):2078–2085. doi: 10.1002/ijc.22557. [DOI] [PubMed] [Google Scholar]
  • 35.Galmarini C.M., Treilleux I., Cardoso F., et al. Class III β-Tubulin isotype predicts response in advanced breast cancer patients randomly treated either with single-agent doxorubicin or docetaxel. Clin Cancer Res. 2008;14(14):4511–4516. doi: 10.1158/1078-0432.Ccr-07-4741. [DOI] [PubMed] [Google Scholar]
  • 36.Saura C., Tseng L.M., Chan S., et al. Neoadjuvant doxorubicin/cyclophosphamide followed by ixabepilone or paclitaxel in early stage breast cancer and evaluation of βIII-tubulin expression as a predictive marker. Oncologist. 2013;18(7):787–794. doi: 10.1634/theoncologist.2013-0075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Opdam M., van Rossum A.G.J., Hoogstraat M., et al. Predictive gene expression profile for adjuvant taxane benefit in breast cancer in the MATADOR trial. iScience. 2024/06/29/2024 doi: 10.1016/j.isci.2024.110425. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.van Rossum A.G.J., Kok M., McCool D., et al. Independent replication of polymorphisms predicting toxicity in breast cancer patients randomized between dose-dense and docetaxel-containing adjuvant chemotherapy. Oncotarget. Dec 26 2017;8(69):113531–113542. doi: 10.18632/oncotarget.22697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Slide score. http://www.slidescore.com
  • 40.Opdam M., van Rossum A.G.J., Menezes R.X., Kok M., Oosterkamp H.M, Linn S.C. MATADOR: a prospective, randomised, phase III biomarker trial in breast cancer patients. Mendeley Data. 2024;2 doi: 10.17632/n24jkp487r.2. [DOI] [Google Scholar]
  • 41.Cardoso F., van’t Veer L.J., Bogaerts J., et al. 70-Gene signature as an aid to treatment decisions in early-stage breast cancer. N Engl J Med. 2016;375(8):717–729. doi: 10.1056/NEJMoa1602253. [DOI] [PubMed] [Google Scholar]
  • 42.Vennin C., Cattaneo C.M., Bosch L., et al. Taxanes trigger cancer cell killing in vivo by inducing non-canonical T cell cytotoxicity. Cancer Cell. Jun 12 2023;41(6):1170–1185.e12. doi: 10.1016/j.ccell.2023.05.009. [DOI] [PubMed] [Google Scholar]
  • 43.Carson W.E., 3rd, Shapiro C.L., Crespin T.R., Thornton L.M., Andersen B.L. Cellular immunity in breast cancer patients completing taxane treatment. Clin Cancer Res. May 15 2004;10(10):3401–3409. doi: 10.1158/1078-0432.Ccr-1016-03. [DOI] [PubMed] [Google Scholar]
  • 44.Alford S.H., Toy K., Merajver S.D., Kleer C.G. Increased risk for distant metastasis in patients with familial early-stage breast cancer and high EZH2 expression. Breast Cancer Res Treat. Apr 2012;132(2):429–437. doi: 10.1007/s10549-011-1591-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Reijm E.A., Timmermans A.M., Look M.P., et al. High protein expression of EZH2 is related to unfavorable outcome to tamoxifen in metastatic breast cancer. Ann Oncol. 2014/11/01/2014;25(11):2185–2190. doi: 10.1093/annonc/mdu391. [DOI] [PubMed] [Google Scholar]
  • 46.Duan R., Du W., Guo W. EZH2: a novel target for cancer treatment. J Hematol Oncol. 2020/07/28 2020;13(1):104. doi: 10.1186/s13045-020-00937-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Peng D., Kryczek I., Nagarsheth N., et al. Epigenetic silencing of TH1-type chemokines shapes tumour immunity and immunotherapy. Nature. 2015/11/01 2015;527(7577):249–253. doi: 10.1038/nature15520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Lau T.S., Chan L.K.Y., Man G.C.W., et al. Paclitaxel induces immunogenic cell death in ovarian cancer via TLR4/IKK2/SNARE-Dependent exocytosis. Cancer Immunol Res. Aug 2020;8(8):1099–1111. doi: 10.1158/2326-6066.Cir-19-0616. [DOI] [PubMed] [Google Scholar]
  • 49.Kanakkanthara A., Miller J.H. βIII-tubulin overexpression in cancer: causes, consequences, and potential therapies. Biochim Biophys Acta Rev Cancer. 2021;1876(2) doi: 10.1016/j.bbcan.2021.188607. 2021/12/01/ [DOI] [PubMed] [Google Scholar]
  • 50.Kavallaris M. Microtubules and resistance to tubulin-binding agents. Nat Rev Cancer. 2010/03/01 2010;10(3):194–204. doi: 10.1038/nrc2803. [DOI] [PubMed] [Google Scholar]
  • 51.Paradiso A., Mangia A., Chiriatti A., et al. Biomarkers predictive for clinical efficacy of taxol-based chemotherapy in advanced breast cancer. Ann Oncol. May 2005;16(Suppl 4):iv14–iv19. doi: 10.1093/annonc/mdi902. [DOI] [PubMed] [Google Scholar]
  • 52.Wang Y., Sparano J.A., Fineberg S., et al. High expression of class III β-Tubulin predicts good response to neoadjuvant taxane and Doxorubicin/cyclophosphamide-based chemotherapy in estrogen receptor–negative breast cancer. Clin Breast Cancer. 2013/04/01/2013;13(2):103–108. doi: 10.1016/j.clbc.2012.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Simon R. Clinical trial designs for evaluating the medical utility of prognostic and predictive biomarkers in oncology. Méd Sur. Jan 1 2010;7(1):33–47. doi: 10.2217/pme.09.49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Liu Y., Beyer A., Aebersold R. On the dependency of cellular protein levels on mRNA abundance. Cell. Apr 21 2016;165(3):535–550. doi: 10.1016/j.cell.2016.03.014. [DOI] [PubMed] [Google Scholar]
  • 55.Yvon A.-M.C., Wadsworth P., Jordan M.A. Taxol suppresses dynamics of individual microtubules in living human tumor cells. Mol Biol Cell. 1999;10(4):947–959. doi: 10.1091/mbc.10.4.947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Trock B.J., Leonessa F., Clarke R. Multidrug resistance in breast cancer: a meta-analysis of MDR1/gp170 expression and its possible functional significance. JNCI: J Natl Cancer Inst. 1997;89(13):917–931. doi: 10.1093/jnci/89.13.917. [DOI] [PubMed] [Google Scholar]
  • 57.Mechetner E., Kyshtoobayeva A., Zonis S., et al. Levels of multidrug resistance (MDR1) P-glycoprotein expression by human breast cancer correlate with in vitro resistance to taxol and doxorubicin. Clin Cancer Res. Feb 1998;4(2):389–398. [PubMed] [Google Scholar]
  • 58.Ravo M., Mutarelli M., Ferraro L., et al. Quantitative expression profiling of highly degraded RNA from formalin-fixed, paraffin-embedded breast tumor biopsies by oligonucleotide microarrays. Lab Invest. Apr 2008;88(4):430–440. doi: 10.1038/labinvest.2008.11. [DOI] [PubMed] [Google Scholar]
  • 59.Cardoso F., Kyriakides S., Ohno S., et al. Early breast cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2019/08/01/2019;30(8):1194–1220. doi: 10.1093/annonc/mdz173. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.docx (15.5MB, docx)

Articles from The Breast : Official Journal of the European Society of Mastology are provided here courtesy of Elsevier

RESOURCES