Tao Pan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Tao Pan has a total of 89 co-authors. Here are the top co-authors by number of shared publications:
- Yiding Chen
Department of Breast Surgery and Oncology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China
5 shared publications Chen Lin
Zhejiang Univ, Affiliated Hosp 2, Sch Med, Dept Breast Surg & Oncol, Hangzhou, Zhejiang, Peoples R China
3 shared publications- Yunxiang Zhou
Department of Breast Surgery and Oncology, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, China
3 shared publications - Cheng Guo
Cancer Institute (Key Laboratory of Cancer Prevention and Intervention, China National Ministry of Education), The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, 310009, China
2 shared publications Chi Pan
Zhejiang Univ, Affiliated Hosp 2, Sch Med, Dept Breast Surg & Oncol, Hangzhou, Zhejiang, Peoples R China
2 shared publicationsDongyan Xu
Zhejiang Univ, Affiliated Hosp 2, Sch Med, Dept Breast Surg & Oncol, Hangzhou, Zhejiang, Peoples R China
2 shared publications
Latest Publications
- Open Access
- 29 December 2025
- Published by MDPI AG in Microorganisms
- Vol. 14 (1) , 75
The Impact of Microbiome on Breast Cancer and Regulatory Strategies
Breast cancer, the most prevalent malignant tumor in women, is closely linked to the human microbiota. The microbiome participates throughout breast cancer pathogenesis, including its occurrence, progression, response to anti-tumor therapies, and treatment-related complications. This review examines the central hypothesis that microbiome-driven inflammatory and immune mechanisms shape breast cancer progression through two key pathways: systemic immune-inflammatory regulation and local tumor microenvironment remodeling. Furthermore, microorganisms and their metabolites modulate systemic treatments by interfering with drug metabolism and altering systemic or local immune-inflammatory environments. Targeting the microbiota represents a promising strategy for enhancing anticancer efficacy and reducing treatment-related complications. This review aims to advance the understanding of the etiology and disease progression of breast cancer from the perspective of microbial-regulated inflammation and immunity, offering new insights for its prevention and treatment.
- 29 December 2025
- Published by MDPI AG in Microorganisms
- Vol. 14 (1) , 75
Open AccessThe Impact of Microbiome on Breast Cancer and Regulatory Strategies
Breast cancer, the most prevalent malignant tumor in women, is closely linked to the human microbiota. The microbiome participates throughout breast cancer pathogenesis, including its occurrence, progression, response to anti-tumor therapies, and treatment-related complications. This review examines the central hypothesis that microbiome-driven inflammatory and immune mechanisms shape breast cancer progression through two key pathways: systemic immune-inflammatory regulation and local tumor microenvironment remodeling. Furthermore, microorganisms and their metabolites modulate systemic treatments by interfering with drug metabolism and altering systemic or local immune-inflammatory environments. Targeting the microbiota represents a promising strategy for enhancing anticancer efficacy and reducing treatment-related complications. This review aims to advance the understanding of the etiology and disease progression of breast cancer from the perspective of microbial-regulated inflammation and immunity, offering new insights for its prevention and treatment.
- Open Access
- 29 January 2024
- Published by Elsevier in Clinical Breast Cancer
- Vol. 24 (3) , 191-198
Can We Do Breast-Conserving Surgery Without Intraoperative Frozen Section of Margin?
- 29 January 2024
- Published by Elsevier in Clinical Breast Cancer
- Vol. 24 (3) , 191-198
Open AccessCan We Do Breast-Conserving Surgery Without Intraoperative Frozen Section of Margin?
- Open Access
- 23 January 2024
- Published by Elsevier in Critical Reviews in Oncology/Hematology
- Vol. 195, 104269
Why does HER2-positive breast cancer metastasize to the brain and what can we do about it?
- 23 January 2024
- Published by Elsevier in Critical Reviews in Oncology/Hematology
- Vol. 195, 104269
Open AccessWhy does HER2-positive breast cancer metastasize to the brain and what can we do about it?
- Open Access
- 1 November 2023
- Published by Taylor & Francis in OncoTargets and Therapy
- Vol. ume 16, 939-960
Predictive Significance of Systemic Immune-Inflammation Index in Patients with Breast Cancer: A Retrospective Cohort Study
Background: Peripheral blood inflammation indices, including the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and systemic immune-inflammation index (SII), have become research hotspots in the diagnosis, treatment, and prognosis prediction of breast cancer, whereas existing research findings remain controversial. Methods: Data pertaining to 1808 breast cancer patients were collected retrospectively to analyze the predictive value of NLR/PLR/SII for breast cancer clinicopathological characteristics, chemotherapy response, and relapse. 1489, 258, and 53 eligible breast cancer patients entered into the three analyses, respectively. Logistic regression analyses were used to assess the correlation between these indices and poor response to chemotherapy. A predictive scoring model was established to predict chemotherapeutic responses based upon the odds ratio values of significant variables identified in logistic regression analyses. Results: Higher pretherapeutic NLR/PLR/SII values were significantly correlated with higher tumor stage, triple-negative breast cancer, premenopausal status, and younger age. Logistic regression analyses indicated that pretherapeutic high SII (as a continuous variable or with a cut-off value of 586.40) and HER2-negative status were independent predictors of poor response to neoadjuvant chemotherapy. A first-in-class SII-based predictive scoring model well distinguished patients who might not benefit from neoadjuvant chemotherapy, with an area under the curve of 0.751. In HR-positive cancers, SII was more strongly associated with clinicopathological features and chemotherapy response. In addition, a receiver operating characteristic curve analysis indicated that the specificity of follow-up SII in identifying cancer relapse was greater than 98.0% at a cut-off value of 900. Conclusion: As a predictor of breast cancer, especially in the HR-positive subtype, SII may eclipse NLR/PLR. SII-high patients are more likely to have a worse chemotherapy response and a higher risk of recurrence.
- 1 November 2023
- Published by Taylor & Francis in OncoTargets and Therapy
- Vol. ume 16, 939-960
Open AccessPredictive Significance of Systemic Immune-Inflammation Index in Patients with Breast Cancer: A Retrospective Cohort Study
Background: Peripheral blood inflammation indices, including the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and systemic immune-inflammation index (SII), have become research hotspots in the diagnosis, treatment, and prognosis prediction of breast cancer, whereas existing research findings remain controversial. Methods: Data pertaining to 1808 breast cancer patients were collected retrospectively to analyze the predictive value of NLR/PLR/SII for breast cancer clinicopathological characteristics, chemotherapy response, and relapse. 1489, 258, and 53 eligible breast cancer patients entered into the three analyses, respectively. Logistic regression analyses were used to assess the correlation between these indices and poor response to chemotherapy. A predictive scoring model was established to predict chemotherapeutic responses based upon the odds ratio values of significant variables identified in logistic regression analyses. Results: Higher pretherapeutic NLR/PLR/SII values were significantly correlated with higher tumor stage, triple-negative breast cancer, premenopausal status, and younger age. Logistic regression analyses indicated that pretherapeutic high SII (as a continuous variable or with a cut-off value of 586.40) and HER2-negative status were independent predictors of poor response to neoadjuvant chemotherapy. A first-in-class SII-based predictive scoring model well distinguished patients who might not benefit from neoadjuvant chemotherapy, with an area under the curve of 0.751. In HR-positive cancers, SII was more strongly associated with clinicopathological features and chemotherapy response. In addition, a receiver operating characteristic curve analysis indicated that the specificity of follow-up SII in identifying cancer relapse was greater than 98.0% at a cut-off value of 900. Conclusion: As a predictor of breast cancer, especially in the HR-positive subtype, SII may eclipse NLR/PLR. SII-high patients are more likely to have a worse chemotherapy response and a higher risk of recurrence.
- Open Access
- 16 June 2023
- Published by Wolters Kluwer Health in Medicine
- Vol. 102 (24) , e34018
A prognostic signature associated with cell senescence predicts survival outcomes and strongly associates with immunotherapy and chemotherapy response in breast cancer
The objective of this study is to assess the predictive potency of cell senescence-related genes (CSRGs) in breast cancer (BC) and establish a risk signature. Trascriptome data of CSRGs were obtained from the TCGA and GEO databases. Consensus clustering was used to generate CSRGs-based molecular clusters for BC patients. A CSRGs-derived risk signature was built using multiple Cox regression analyses of differentially expressed genes (DEGs) between clusters. The prognosis, immune infiltration, chemotherapy and immunotherapy response between different risk groups were analyzed and compared. Two molecular clusters of BC patients were generated on the basis of 79 differentially expressed CSRGs, which showed distinct prognosis and immune infiltration. A total of 1403 DEGs between the CSRGs-derived clusters were found, and 10 of them were independent prognostic genes that used to construct a risk signature. The results demonstrated that patients with older age and advanced stage presented with a higher risk scores. In addition, the risk signature was found to be associated with outcomes, immune infiltration, chemotherapy and immunotherapy response. Patients in the low-risk group showed a favorable prognosis and higher immunotherapy response than those in the high-risk group. Finally, we developed a highly stable nomogram that incorporates risk signature, chemotherapy, radiotherapy, and stage variables, enabling accurate prediction of the overall survival (OS) of individual patients. To conclude, the signature derived from CSRGs holds great promise as a biomarker for prognostic assessment of BC and may serve as a valuable tool in guiding immunotherapy.
- 16 June 2023
- Published by Wolters Kluwer Health in Medicine
- Vol. 102 (24) , e34018
Open AccessA prognostic signature associated with cell senescence predicts survival outcomes and strongly associates with immunotherapy and chemotherapy response in breast cancer
The objective of this study is to assess the predictive potency of cell senescence-related genes (CSRGs) in breast cancer (BC) and establish a risk signature. Trascriptome data of CSRGs were obtained from the TCGA and GEO databases. Consensus clustering was used to generate CSRGs-based molecular clusters for BC patients. A CSRGs-derived risk signature was built using multiple Cox regression analyses of differentially expressed genes (DEGs) between clusters. The prognosis, immune infiltration, chemotherapy and immunotherapy response between different risk groups were analyzed and compared. Two molecular clusters of BC patients were generated on the basis of 79 differentially expressed CSRGs, which showed distinct prognosis and immune infiltration. A total of 1403 DEGs between the CSRGs-derived clusters were found, and 10 of them were independent prognostic genes that used to construct a risk signature. The results demonstrated that patients with older age and advanced stage presented with a higher risk scores. In addition, the risk signature was found to be associated with outcomes, immune infiltration, chemotherapy and immunotherapy response. Patients in the low-risk group showed a favorable prognosis and higher immunotherapy response than those in the high-risk group. Finally, we developed a highly stable nomogram that incorporates risk signature, chemotherapy, radiotherapy, and stage variables, enabling accurate prediction of the overall survival (OS) of individual patients. To conclude, the signature derived from CSRGs holds great promise as a biomarker for prognostic assessment of BC and may serve as a valuable tool in guiding immunotherapy.
- Open Access
- 30 March 2023
- Published by Frontiers Media SA in Frontiers in Surgery
- Vol. 10, 1133335
Pathologic complete response of hepatoid adenocarcinoma of the stomach after chemo-immunotherapy: A rare case report and literature review
Background: Hepatoid adenocarcinoma of the stomach (HAS) is a highly malignant subtype of gastric carcinoma with specific clinicopathological features and extremely poor prognosis. We present an exceedingly rare case of complete response after chemo-immunotherapy.Case Description: A 48-year-old woman with highly elevated serum alpha-fetoprotein (AFP) level was found to have HAS verified by pathological examination based on gastroscopy. Computed tomography scan was done and TNM staging of the tumor was T4aN3aMx. Programmed cell death ligand-1 (PD-L1) immunohistochemistry was performed, revealing a negative PD-L1 expression. Chemo-immunotherapy including oxaliplatin plus S-1 and PD-1 inhibitor terelizumab was given to this patient for 2 months until the serum AFP level decreased from 748.5 to 12.9 ng/mL and the tumor shrank. D2 radical gastrectomy was then performed and histopathology of the resected specimen revealed that the cancerous cells had disappeared. Pathologic complete response (pCR) was achieved and no evidence of recurrence has been found after 1 year of follow-up.Conclusions: We, for the first time, reported an HAS patient with negative PD-L1 expression who achieved pCR from the combined chemotherapy and immunotherapy. Although no consensus has been reached regarding the therapy, it might provide a potential effective management strategy for HAS patient.
- 30 March 2023
- Published by Frontiers Media SA in Frontiers in Surgery
- Vol. 10, 1133335
Open AccessPathologic complete response of hepatoid adenocarcinoma of the stomach after chemo-immunotherapy: A rare case report and literature review
Background: Hepatoid adenocarcinoma of the stomach (HAS) is a highly malignant subtype of gastric carcinoma with specific clinicopathological features and extremely poor prognosis. We present an exceedingly rare case of complete response after chemo-immunotherapy.Case Description: A 48-year-old woman with highly elevated serum alpha-fetoprotein (AFP) level was found to have HAS verified by pathological examination based on gastroscopy. Computed tomography scan was done and TNM staging of the tumor was T4aN3aMx. Programmed cell death ligand-1 (PD-L1) immunohistochemistry was performed, revealing a negative PD-L1 expression. Chemo-immunotherapy including oxaliplatin plus S-1 and PD-1 inhibitor terelizumab was given to this patient for 2 months until the serum AFP level decreased from 748.5 to 12.9 ng/mL and the tumor shrank. D2 radical gastrectomy was then performed and histopathology of the resected specimen revealed that the cancerous cells had disappeared. Pathologic complete response (pCR) was achieved and no evidence of recurrence has been found after 1 year of follow-up.Conclusions: We, for the first time, reported an HAS patient with negative PD-L1 expression who achieved pCR from the combined chemotherapy and immunotherapy. Although no consensus has been reached regarding the therapy, it might provide a potential effective management strategy for HAS patient.
- Open Access
- 2 November 2022
- Published by Springer Nature in Scientific Reports
- Vol. 12 (1) , 1-8
Specimen mammography for intraoperative margin assessment in breast conserving surgery: a meta-analysis
In breast conserving surgery (BCS), specimen mammography is one of the most widely used intraoperative methods of assessing margin status. We performed a meta-analysis to evaluate the diagnostic accuracy of specimen mammography. Literature databases including PubMed, Cochrane Library, Web of Science, and EMBASE were searched prior to Jun 2022. A total of 1967 patients were included from 20 studies. A pooled analysis, heterogeneity testing, threshold effect testing, publication bias analysis, and subgroup analyses were performed from extracted data. The pooled weighted values were a sensitivity of 0.55 (95% confidence interval [CI], 0.47–0.63), a specificity of 0.85 (95% CI, 0.78–0.90), a diagnostic odds ratio of 7 (95% CI, 4–12), and a pooled positive likelihood ratio of 3.7 (95% CI 2.6–5.5). The area under the receiver operator characteristic curve was 0.75 (95% CI 0.71–0.78). In the subgroup analysis, the pooled specificity in the positive margin defined as tumor at margin subgroup was lower than the other positive margin definition subgroup (0.82 [95% CI: 0.71, 0.92] vs. 0.87 [95% CI: 0.80, 0.94], p = 0.01). Our findings indicated that specimen mammography was an accurate intraoperative imaging technique for margin assessment in BCS.
- 2 November 2022
- Published by Springer Nature in Scientific Reports
- Vol. 12 (1) , 1-8
Open AccessSpecimen mammography for intraoperative margin assessment in breast conserving surgery: a meta-analysis
In breast conserving surgery (BCS), specimen mammography is one of the most widely used intraoperative methods of assessing margin status. We performed a meta-analysis to evaluate the diagnostic accuracy of specimen mammography. Literature databases including PubMed, Cochrane Library, Web of Science, and EMBASE were searched prior to Jun 2022. A total of 1967 patients were included from 20 studies. A pooled analysis, heterogeneity testing, threshold effect testing, publication bias analysis, and subgroup analyses were performed from extracted data. The pooled weighted values were a sensitivity of 0.55 (95% confidence interval [CI], 0.47–0.63), a specificity of 0.85 (95% CI, 0.78–0.90), a diagnostic odds ratio of 7 (95% CI, 4–12), and a pooled positive likelihood ratio of 3.7 (95% CI 2.6–5.5). The area under the receiver operator characteristic curve was 0.75 (95% CI 0.71–0.78). In the subgroup analysis, the pooled specificity in the positive margin defined as tumor at margin subgroup was lower than the other positive margin definition subgroup (0.82 [95% CI: 0.71, 0.92] vs. 0.87 [95% CI: 0.80, 0.94], p = 0.01). Our findings indicated that specimen mammography was an accurate intraoperative imaging technique for margin assessment in BCS.
- Open Access
- 14 October 2022
- Published by MDPI AG in Metabolites
- Vol. 12 (10) , 973
Simultaneous Determination of Methylated Nucleosides by HILIC–MS/MS Revealed Their Alterations in Urine from Breast Cancer Patients
RNA methylation plays a vital role in the pathogenesis of a variety of diseases including cancer, and aberrant levels of modified nucleosides in RNA were revealed to be related to cancer. Urine is a favored source for biomarker discovery due to the non-invasion to patients. Herein, we developed a sensitive hydrophilic interaction liquid chromatography tandem mass spectrometry (HILIC–MS/MS) method combined with stable isotope dilution for accurate quantification of methylated nucleosides in human urine. With this method, we successfully quantified ten methylated nucleosides in urine samples collected from healthy controls and breast cancer patients. We found N6-methyladenosine (m6A), 2′-O-methyladenosine (Am), N1-methyladenosine (m1A), N6,2′-O-dimethyladenosine (m6Am), N1-methylguanosine (m1G), 2′-O-methylguanosine (Gm), 5-methylcytidine (m5C) and 2′-O-methylcytidine (Cm) were all decreased in early-stage breast cancer patients, and a nomogram prediction model was constructed. Locally advanced breast cancer patients exhibited elevated levels of urinary 2′-O-methylated nucleosides in comparison to early-stage breast cancer patients. Together, we developed a robust method for the simultaneous determination of methylated nucleosides in human urine, and the results revealed an association between the contents of urinary methylated nucleosides and the occurrence of breast cancer, which may stimulate future studies about the regulatory roles of these methylated nucleosides in the initiation and progression of breast cancer.
- 14 October 2022
- Published by MDPI AG in Metabolites
- Vol. 12 (10) , 973
Open AccessSimultaneous Determination of Methylated Nucleosides by HILIC–MS/MS Revealed Their Alterations in Urine from Breast Cancer Patients
RNA methylation plays a vital role in the pathogenesis of a variety of diseases including cancer, and aberrant levels of modified nucleosides in RNA were revealed to be related to cancer. Urine is a favored source for biomarker discovery due to the non-invasion to patients. Herein, we developed a sensitive hydrophilic interaction liquid chromatography tandem mass spectrometry (HILIC–MS/MS) method combined with stable isotope dilution for accurate quantification of methylated nucleosides in human urine. With this method, we successfully quantified ten methylated nucleosides in urine samples collected from healthy controls and breast cancer patients. We found N6-methyladenosine (m6A), 2′-O-methyladenosine (Am), N1-methyladenosine (m1A), N6,2′-O-dimethyladenosine (m6Am), N1-methylguanosine (m1G), 2′-O-methylguanosine (Gm), 5-methylcytidine (m5C) and 2′-O-methylcytidine (Cm) were all decreased in early-stage breast cancer patients, and a nomogram prediction model was constructed. Locally advanced breast cancer patients exhibited elevated levels of urinary 2′-O-methylated nucleosides in comparison to early-stage breast cancer patients. Together, we developed a robust method for the simultaneous determination of methylated nucleosides in human urine, and the results revealed an association between the contents of urinary methylated nucleosides and the occurrence of breast cancer, which may stimulate future studies about the regulatory roles of these methylated nucleosides in the initiation and progression of breast cancer.
- Open Access
- 23 December 2021
- Published by Springer Nature in Research Square
Specimen Mammography for Intraoperative Margin Assessment in Breast Conserving Surgery: A Meta-Analysis
Specimen mammography is one of the widely used intraoperative methods assessing margin status in breast conserving surgery. We performed a meta-analysis to evaluate the diagnostic accuracy of specimen mammography. Literature databases including Pubmed, Cochrane Library, Web of Science and EMBASE were searched prior to May 2020. 18 studies with a total of 1142 patients were included. Data was extracted to perform pooled analysis, heterogeneity testing, threshold effect testing, sensitivity analysis, publication bias analysis and subgroup analyses. The pooled weighted values were a sensitivity of 0.55 (95% CI, 0.45–0.64), a specificity of 0.85 (95% CI, 0.77–0.90), a DOR of 7 (95% CI, 4–11) and a pooled positive likelihood ratio of 3.6 (95% CI 2.4-5.3). The area under the receiver operator characteristic curve was 0.75 (95% CI 0.71-0.78). In the subgroup analysis, the pooled specificity in the positive margin defined as tumor at margin subgroup was lower than the other positive margin definition subgroup (0.79 [95% CI: 0.66, 0.91] vs. 0.88 [95% CI: 0.81, 0.95], p = 0.01). Our findings indicated specimen mammography to be an accurate and intraoperative imaging technique for margin assessment in breast conserving surgery.
- 23 December 2021
- Published by Springer Nature in Research Square
Open AccessSpecimen Mammography for Intraoperative Margin Assessment in Breast Conserving Surgery: A Meta-Analysis
Specimen mammography is one of the widely used intraoperative methods assessing margin status in breast conserving surgery. We performed a meta-analysis to evaluate the diagnostic accuracy of specimen mammography. Literature databases including Pubmed, Cochrane Library, Web of Science and EMBASE were searched prior to May 2020. 18 studies with a total of 1142 patients were included. Data was extracted to perform pooled analysis, heterogeneity testing, threshold effect testing, sensitivity analysis, publication bias analysis and subgroup analyses. The pooled weighted values were a sensitivity of 0.55 (95% CI, 0.45–0.64), a specificity of 0.85 (95% CI, 0.77–0.90), a DOR of 7 (95% CI, 4–11) and a pooled positive likelihood ratio of 3.6 (95% CI 2.4-5.3). The area under the receiver operator characteristic curve was 0.75 (95% CI 0.71-0.78). In the subgroup analysis, the pooled specificity in the positive margin defined as tumor at margin subgroup was lower than the other positive margin definition subgroup (0.79 [95% CI: 0.66, 0.91] vs. 0.88 [95% CI: 0.81, 0.95], p = 0.01). Our findings indicated specimen mammography to be an accurate and intraoperative imaging technique for margin assessment in breast conserving surgery.
- Open Access
- 15 May 2021
- Published by Wiley in Journal of Oncology
- Vol. 2021, 1-13
Prognostic Models for Nonmetastatic Triple-Negative Breast Cancer Based on the Pretreatment Serum Tumor Markers with Machine Learning
Purpose. Triple-negative breast cancer (TNBC) is a heterogeneous and aggressive disease with poorer prognosis than other subtypes. We aimed to investigate the prognostic efficacy of multiple tumor markers and constructed a prognostic model for stage I-III TNBC patients. Patients and Methods. We included stage I-III TNBC patients whose serum tumor markers levels were measured prior to the treatment. The optimal cut-off value of each tumor marker was determined by X-tile. Then, we adopted two survival models (lasso Cox model and random survival forest model) to build the prognostic model and AUC values of the time-dependent receiver operating characteristic (ROC) were calculated. The Kaplan-Meier method was used to plot the survival curves and the log-rank test was used to test whether there was a significant difference between the predicted high-risk and low-risk groups. We used univariable and multivariable Cox analysis to identify independent prognostic factors and did subgroup analysis further for the lasso Cox model. Results. We included 258 stage I-III TNBC patients. CEA, CA125, and CA211 showed independent prognostic value for DFS when using the optimal cut-off values; their HRs and 95% CI were as follows: 1.787 (1.056–3.226), 2.684 (1.200–3.931), and 2.513 (1.567–4.877). AUC values of lasso Cox model and random survival forest model were 0.740 and 0.663 for DFS at 60 months, respectively. Both the lasso Cox model and random survival forest model demonstrated excellent prognostic value. According to tumor marker risk scores (TMRS) computed by the lasso Cox model, the high TMRS group had worse DFS (HR = 3.138, 95% CI: 1.711–5.033, ) and OS (3.983, 1.637–7.214, ) than low TMRS group. Furthermore, subgroup analysis of N0-N1 patients in the lasso Cox model indicated that TMRS still had a significant prognostic effect on DFS (2.278, 1.189–4.346) and OS (2.982, 1.110–7.519). Conclusions. Our study indicated that pretreatment levels of serum CEA, CA125, and CA211 had independent prognostic significance for TNBC patients. Both lasso Cox model and random survival forest model that we constructed based on tumor markers could strongly predict the survival risk. Higher TMRS was associated with worse DFS and OS both in stage I-III and N0-N1 TNBC patients.
- 15 May 2021
- Published by Wiley in Journal of Oncology
- Vol. 2021, 1-13
Open AccessPrognostic Models for Nonmetastatic Triple-Negative Breast Cancer Based on the Pretreatment Serum Tumor Markers with Machine Learning
Purpose. Triple-negative breast cancer (TNBC) is a heterogeneous and aggressive disease with poorer prognosis than other subtypes. We aimed to investigate the prognostic efficacy of multiple tumor markers and constructed a prognostic model for stage I-III TNBC patients. Patients and Methods. We included stage I-III TNBC patients whose serum tumor markers levels were measured prior to the treatment. The optimal cut-off value of each tumor marker was determined by X-tile. Then, we adopted two survival models (lasso Cox model and random survival forest model) to build the prognostic model and AUC values of the time-dependent receiver operating characteristic (ROC) were calculated. The Kaplan-Meier method was used to plot the survival curves and the log-rank test was used to test whether there was a significant difference between the predicted high-risk and low-risk groups. We used univariable and multivariable Cox analysis to identify independent prognostic factors and did subgroup analysis further for the lasso Cox model. Results. We included 258 stage I-III TNBC patients. CEA, CA125, and CA211 showed independent prognostic value for DFS when using the optimal cut-off values; their HRs and 95% CI were as follows: 1.787 (1.056–3.226), 2.684 (1.200–3.931), and 2.513 (1.567–4.877). AUC values of lasso Cox model and random survival forest model were 0.740 and 0.663 for DFS at 60 months, respectively. Both the lasso Cox model and random survival forest model demonstrated excellent prognostic value. According to tumor marker risk scores (TMRS) computed by the lasso Cox model, the high TMRS group had worse DFS (HR = 3.138, 95% CI: 1.711–5.033, ) and OS (3.983, 1.637–7.214, ) than low TMRS group. Furthermore, subgroup analysis of N0-N1 patients in the lasso Cox model indicated that TMRS still had a significant prognostic effect on DFS (2.278, 1.189–4.346) and OS (2.982, 1.110–7.519). Conclusions. Our study indicated that pretreatment levels of serum CEA, CA125, and CA211 had independent prognostic significance for TNBC patients. Both lasso Cox model and random survival forest model that we constructed based on tumor markers could strongly predict the survival risk. Higher TMRS was associated with worse DFS and OS both in stage I-III and N0-N1 TNBC patients.