Development of a machine learning-based radiomics signature for estimating breast cancer TME phenotypes and predicting anti-PD-1/PD-L1 immunotherapy response

doi:10.21203/rs.3.rs-3104002/v1

Download PDF

Research Article

Development of a machine learning-based radiomics signature for estimating breast cancer TME phenotypes and predicting anti-PD-1/PD-L1 immunotherapy response

https://doi.org/10.21203/rs.3.rs-3104002/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 29 Jan, 2024

Read the published version in Breast Cancer Research →

You are reading this latest preprint version

Backgrounds:

Since breast cancer patients respond diversely to immunotherapy, exploration of novel biomarkers for precisely predicting clinical response are urgently required to enhance therapeutic efficacy. The purpose of our present research was to construct and independently validate a biomarker of tumor microenvironment (TME) phenotypes via a machine learning-based radiomics way. The interrelationship between the biomarker, TME phenotypes and recipients’ clinical response would also be revealed.

Methods

In this retrospective multi-cohort investigation, five separate cohorts of breast cancer patients were recruited to measure breast cancer TME phenotypes via a radiomics signature, which was constructed and validated by integrating RNA-seq data with DCE-MRI images for predicting immunotherapy response. Initially, we constructed TME phenotypes using RNA-seq of 1089 breast cancer patients in the TCGA database. Then, parallel DCE-MRI images and RNA-seq of 94 breast cancer patients obtained from TCIA were applied to develop a radiomics-based TME phenotypes signature by Random Forest in machine learning. In an internal validation set, the repeatability of radiomics signature was validated. Two additional independent external validation sets were analyzed to reassess this signature. The Immune phenotype cohort (n = 158) divided enrolled subjects into immune-inflamed and immune-desert phenotypes based on CD8 cell infiltration; these data were utilized to examine the relationship between the immune phenotypes and this signature. A final Immunotherapy-treated cohort with 77 cases who received anti-PD-1/PD-L1 treatment was utilized to evaluate the predictive efficiency of this signature in terms of clinical outcomes.

Results

The TME phenotypes of breast cancer was separated into two heterogeneous clusters: Cluster A, a "immune-inflamed" cluster, containing substantial innate and adaptive immune cell infiltration, and Cluster B, a "immune-desert" cluster, with modest TME cell infiltration. We constructed a radiomics signature for the TME phenotypes ([AUC] = 0.855; 95% CI: 0.777–0.932; P < 0.05) and verified it in an internal validation set (0.844; 0.606-1; P < 0.05). In the known immune phenotypes cohort, the signature can identified either immune-inflamed or immune-desert tumor (0.814; 0.717–0.911; P < 0.05). In the Immunotherapy-treated cohort, patients with objective response had higher baseline radiomics scores than those with stable or progressing disease (P < 0.05); moreover, the radiomics signature deserved an AUC of 0.784 (0.643–0.926; P < 0.05) for predicting immunotherapy response.

Conclusions

Our imaging biomarker, a practicable radiomics signature, is beneficial for predicting the TME phenotypes and clinical response in anti-PD-1/PD-L1-treated breast cancer patients. The "immune-desert" phenotype belonging to “cold tumor” should be provoked for transforming into "immune-inflamed" phenotype namely as "hot tumor".

machine learning

radiomics signature

breast cancer

tumor microenvironment

As demonstrated by the most updated regional report, breast cancer incidence among Chinese women overtook lung cancer incidence for the initial time to make it the most prevalent cancer, accounting for around 19.9% of newly diagnosed cancers¹. Minimizing the incidence of recurrence and metastasis five years after surgery, as well as length of life, are essential concerns and obstacles for breast cancer². Present breast cancer treatment have hit a plateau; novel therapeutic strategies are sought³.In 2019, the U.S. Food and Drug Administration initially authorized the anti-programmed cell death ligand 1 (PD-L1) inhibitor Atezolizumab for the first-line management of PD-L1-positive advanced triple-negative breast cancer, marking a new adventure in immunotherapy for breast cancer⁴. Latest studies, however, have demonstrated that immunotherapies, for example immune checkpoint inhibitors (ICIs), show unsatisfactory efficacy across the board for breast cancer patients^5–7.

The failure to identify patients via the landscape of tumor microenvironment (TME) may be the primary cause of these dismal outcomes. Meanwhile, the complete spectrum of TME phenotypes in breast cancer currently unexplained. The majority of prior investigations have emphasized on either one or two subtypes of TME breast cancer cells^8–11, which may have resulted in a skewed knowledge of the breast cancer TME. Since TME cells exhibit robust intercellular communication, it is more logical to regard them as a whole.

The emergence of next-generation sequencing has enabled the investigation of the TME. Given the fact that RNA-seq of tumor samples invariably indicates the presence of TME cell types, scientists proposed a variety of genetic expression profile-based measurement of the TME cell abundance in tumor tissue^12–14. However, conventional TME evaluations often involve the acquisition of tissue biopsies following surgery. It would be beneficial to search out non-invasive techniques for evaluating the TME landscape.

Radiomics is the technique of interpreting medical images and converting them into numerical data^15,16. High-dimensional imaging data offer abundant information regarding tumor phenotypes that are governed not merely by the inherent physiological processes of cancer cells, but also by the TME, such as the makeup and infiltrating patterns of tumor-infiltrating immune cells¹⁷. The purpose of radiomics is to produce image-mediated predictors that act as tools to reveal the association between certain imaging and TME phenotypes and to gain knowledge of breast cancer biology in order to improve therapeutic governance^18,19. Radiomics precedes biopsy testing as it benefits of being physically noninvasive, enabling longitudinal estimation of the tumor and its surrounding milieu, description of spatial variability, as well as disease progression.

Immunotherapy has drastically altered the strategy for treating cancers such as breast cancer^20–22. Unfortunately, lower than 10% of individuals with specific breast cancer react to the treatment, which is regrettably indicative of the broad variation in patient response to immunotherapy^23,24. Therefore, it is necessary to develop strategies for determining which patients are most likely to react to immunotherapy. It is demonstrated that preexisting intratumoral and peritumoral TME cell infiltration associated with anti-programmed cell death protein (PD)-1 and anti-PD-L1 immunotherapy response^25,26. There have been characterized three distinct immunological phenotypes: immune-inflamed, immune-excluded, and immune-desert. The immune-inflamed phenotype is defined by tumor cells expressing PD-L1, an enormous prevalence of infiltrating immune cells in the TME, and an increased frequency of tumor-infiltrating lymphocytes (TILs), and these tumors are immunotherapy-receptive^14,27. In contrast, immune-excluded and immune-desert phenotypes are regarded as inflammation-free.

We anticipated that MRI radiomics may capture microstructural variations noninvasively across breast cancer TME phenotypes in order to find new determinants of immunotherapy success. We seek to (i) build TME phenotypes for breast cancer and (ii) develop a radiomics signature for the TME phenotypes of breast cancer in terms of determining clinical outcomes of patients administered with anti-PD-1/PD-L immunotherapy.

Protocol and data sources

In this multi-cohort investigation, retrospective genomics and radiomics analysis was carried out on five distinct cohorts of breast cancer patients 18 years or older (Fig. 1). The Cancer Genome Atlas (TCGA) and Cancer Imaging Archive (TCIA) databases provided the information for the Gene expression analysis cohort, the Radiomics discovery cohort, and the Radiomics validation cohort. The Radiomics discovery cohort and the Radiomics validation cohort both originate from the same data collection namely TCGA-BRCA, as the former cohort containing 80% cases and the later one with 20% cases. The Immune phenotype cohort and the Immunotherapy-treated cohort were extracted from the database of the Guangdong Provincial People's Hospital. The Gene expression analysis cohort was comprised of transcriptome data from 1089 patients (the transcriptomic data and relevant clinicopathological variables are included in Table S1), for describing the landscape of the breast cancer TME and revealing its heterogeneity. The Radiomics discovery cohort consists of 94 patients whose DCE-MRI and RNA-seq data are available, and all cases permitted prediction of TME phenotypes by radiomics analysis using RNA-seq data and accompanying MR images.

In conjunction with one internal and two separated external validation sets, the radiomics signature was validated. The Radiomics validation cohort (internal validation set) consisted of 24 patients, and this dataset was utilized to confirm the congruence between the radiomics signature (from the Radiomics discovery cohort) and the TME phenotypes in this independent dataset. The Immune phenotype cohort (external validation set 1) consisted of 158 patients grouped into the two most severe tumor immune phenotypes based on their CD8 immunohistochemistry (IHC) staining findings (CD8 20% was deemed positive)^28,29: the immune-inflamed phenotype and the immune-desert phenotype, regardless of the treatment provided. This dataset was employed to reveal the match between the breast cancer radiomics signature and tumor immunophenotype. All 77 breast cancer patients in the Immunotherapy-treated cohort (external validation set 2) underwent anti-PD-1/PD-L1 treatment. This cohort was utilized to identify the relationship between radiomics signature and adherence to the Response Evaluation Criteria for Solid Tumors (RECIST) version 1.1 clinical response. Our research was authorized by the review committee of Guangdong Provincial People's Hospital and conducted in compliance with the Helsinki Declaration's ethical standards. Due to the retroactive character of the investigation, the committee waived the informed consent requirement. All trial data were de-identified and made anonymous.

Calculation of TME cell abundance

Previous study²⁷ constructed a compendium of breast cancer TME genes associated with specific microenvironment cell subpopulations by modifying two gene signatures, CIBERSORT³⁰ and MCP-Counter³¹. This compilation includes 324 genes that represented 22 categories of immune cells and 40 genes that represented endothelial cells and fibroblasts. Using single sample gene enrichment analysis (ssGSEA,"GSVA" in R) with expression data, this result was used in our work to compute each cell subpopulation of each sample by measuring their abundance.

TME phenotypes clustering

To establish the optimal number of stable breast cancer TME subtypes, k-means ("kmeans" in R) clustering, Nbclust testing ("NbClust" in R, index = "all"), and Silhouette analysis ("Silhouette" in R) was used. Prior to clustering, we scaled each sample in order to cluster them according to the compositional pattern of each TME cell type. For heatmap visualization, the data was reordered based on the k-means clustering findings and scaled the original ssGSEA results prior to displaying (“pheatmap in R”).

Prognostic value of TME phenotypes

Kaplan-Meier survival curve was plotted via the "ggsurvplot" in R, the prognostic value of each cell subpopulation in each TME cluster of the total cohort was measured. The best cutoff value was used in each investigation to classify cell abundance as high or low ("cutp" in R).

MR image analysis and feature extraction

Radiomics features were affected by equipment and acquisition conditions, hence we preprocessed the obtained DICOM images: (i) Z-score normalization. Normalize the intensity of all MRI images using the Z-score calculation (Caret package); (ii) Resampling. The Simple ITK software was used to normalize the voxel spacing by standardizing the voxel size to 1.0\(\times\)1.0\(\times\)1.0 mm³.

The segmentation of the imaging was accomplished semi-automatically. Initially, two radiologists (** and **, with ** and ** years of experience in breast cancer MR diagnosis, respectively) utilized ITK-SNAP software (www.itk-snap.org) to manually draw tumor borders on MRI and establish agreement on tumor outlines; disputes were addressed by consensus. To get quantitative data from the TME, we next utilized Python to automatically augment the lesions depending on the findings of the lesion border identification. On each side of the tumor border (i.e., the outer and inner sides of the border), an expanding and contracting 2 mm peripheral ring was constructed, culminating in a 4-mm ring³². Large-size blood arteries, neighboring tissues, and air spaces were all excluded if there was no tumor cell invasion (Fig. 2).

We extracted features for each area of interest, i.e., the intratumoral and peritumoral regions, using Python, pyradiomics. A final collection of 479 quantitative features, including 14 shape features, 90 first-order intensity features, and 375 second-order and higher-order texture features, was extracted. The picture features were described in detail in Supplemental Tables S2 and S3.

By assessing all radiomics features extracted using intra- and inter-class correlation coefficients (ICC), inter- and intra-observer consistency and repeatability were obtained. Radiologists ** and ** retrieved features from thirty randomly chosen patients. Radiologists ** performed the identical processes two weeks later. Selecting features with an ICC greater than 0.75 ensured the reliability of extracted features.

Construction of radiomics signature

In the Radiomics discovery cohort, we created a random forest model to predict the TME phenotypes; five-fold cross-validation picked the best model. This model was applied to the Radiomics validation cohort using the Youden index to determine the best threshold, hence maximizing the overall sensitivity and specificity.

Feature significance specifies which features were relevant and enabled the enhancement of a model via feature selection. We calculated feature significance using the random forest algorithm from the scikit-learn package (Python).

Statistical analysis

The Student-t test, Wilcoxon test, and Kruskal-Wallis test were used to compare continuous variables, and the Shapiro-Wilk test was employed to determine the normality of the distribution prior to comparison. For categorical variables, Fisher's exact test was utilized. The Kaplan-Meier method was used to analyze survival, and the log-rank test was used to compare survival across clusters. ICCs were used to determine inter- and intra-observer consistency in lesion segmentation. A random forest classifier model was used to categorize the TME phenotypes. For the evaluation of model performance, the area under the subject operating characteristic curve (AUC) and other performance assessment metrics (Accuracy, sensitivity, specificity, positive predictive value and negative predictive value) were used to determine if Radiomics score (rad-score) could classify patients into two groups based on the abundance of TME cell infiltration. Spearman's correlation coefficient was utilized to determine the relationship between the 20 most significant radiomics features and the abundance of 24 cell infiltration in TME.

In the Immunotherapy-treated cohort, patients were classified into high or low scoring groups using the median value of rad-score. The duration of follow-up was determined from the beginning of immunotherapy. Clinical response was classified as complete response, partial response, stable disease, or progressive disease based on RECIST version 1.1 and assessed at six months; immunotherapy response for each patient was presented in Supplemental Table S4. Two-sided P < 0.05 was considered statistically significant. All statistical analyses were performed in R (Version 3.5.0).

Role of funding sources

The fundings of the study make contribution to the establishment of breast cancer dataset, but has no roles in study design, data collection, data analysis, data interpretation.

Demographic and clinical characteristics

Total 1442 subjects with mean age as 57 ± 13 years were included in our present research, Among them, 1089 patients (mean age, 58.5 ± 13.2 years) was assigned into the Gene expression analysis cohort, 94 patients (mean age, 54.8 ± 11.2 years) in the Radiomics discovery cohort, 24 patients (mean age, 53 ± 13 years) in the Radiomics validation cohort, 158 patients in the Immune phenotype cohort (mean age, 51.1 ± 11.1 years), and 77 patients in the Immunotherapy-treated cohort (mean age, 50.8 ± 9.1 years). The population characteristics of these five cohorts were exhaustively described in Table 1.

Table 1

Characteristics of patients in the Gene expression analysis, Radiogenomics discovery cohort, Radiogenomics validation cohort, Immune phenotype cohort and Immunotherapy-treated cohort
Variables	Gene expression analysis cohort		Radiogenomics discovery cohort		Radiogenomics validation cohort		Immune phenotype cohort		Immunotherapy-treated cohort
	n = 1089		n = 94		n = 24		n = 158		N = 77
	N	%	N	%	N	%	N	%	N	%
Age (years)
18–60	600	55.1	62	66	18	75	121	76.6	67	87
> 60	489	44.9	32	34	6	25	37	23.4	10	13
Laterality
Left	567	52.1	50	53.2	10	41.7	89	56.3	51	66.2
Right	522	47.9	44	46.8	14	58.3	69	43.7	26	33.8
Status
Alive	938	86.1	91	96.8	24	100	NA	NA	NA	NA
Dead	151	13.9	3	3.2	0	0	NA	NA	NA	NA
OS(years)
≤ 1	168	15.4	6	6.3	1	4.2	NA	NA	NA	NA
> 1 ≤3	481	44.2	44	46.3	6	25	NA	NA	NA	NA
> 3 ≤5	187	17.2	24	25.3	8	33.3	NA	NA	NA	NA
> 5years	253	23.2	21	22.1	9	37.5	NA	NA	NA	NA
Pathologic stage
I	180	16.5	20	21.3	8	33.3	6	3.8	0	0
II	620	57	59	62.7	15	62.5	57	36.1	19	24.7
III	248	22.8	15	16	1	4.2	23	14.5	12	15.6
IV	20	1.8	0	0	0	0	2	1.3	0	0
X	0	0	0	0	0	0	0	0	0	0
NA	21	1.9	0	0	0	0	70	44.3	46	59.7
Estrogen receptor
Positive	802	73.6	78	83	19	79.2	119	75.3	15	19.5
Negative	237	21.8	16	17	5	20.8	39	24.7	58	75.3
Indeterminate	/	/	0	0	0	0	0	0	0	0
NA	50	4.6	0	0	0	0	0	0	4	5.2
Progesterone receptor
Positive	693	63.6	72	76.6	15	62.5	100	63.3	15	19.5
Negative	343	31.5	22	23.4	9	37.5	58	36.7	58	75.3
Indeterminate	/	/	0	0	0	0	0	0	0	0
NA	53	4.9	0	0	0	0	0	0	4	5.2
Human epidermal growth factor receptor 2
Positive	164	15.1	11	11.7	2	8.4	118	74.7	6	7.8
Negative	559	51.3	48	51.1	14	58.3	40	25.3	67	87
Indeterminate	190	17.4	0	0	0	0	0	0	0	0
NA	176	16.2	35	37.2	8	33.3	0	0	4	5.2
NA: Not Available

Landscape of TME phenotypes in breast cancer

The reference TME compendium developed by prior research consisted of 364 genes covering 24 TME cell subpopulations to comprehensively define the TME phenotypes of breast cancer. This finding was used to measure the abundance of 24 TME cellular subpopulations in each sample (Supplementary Table S5). Then, we ran k-means clustering on the TME phenotypes of breast cancer. All 1089 breast cancer TME phenotypes were categorized into two heterogeneous clusters (Fig. 3A). The ideal and stable number of clusters was two (Fig. 3B and Supplementary Figure S1). Cluster A, the "immune-inflamed" cluster ("hot tumor"), was shown as relatively large infiltration of immune cells. Cluster B, the "immune-desert" cluster ("cold tumor"), was distinguished by poor TME cell infiltration. In furthermore, we characterized the distribution of TME cell subpopulations in order to determine their relative proportions within clusters. The relative percentage of non-immune and innately inactivated immune cells grew in the Cluster B, but the relative weight of innate and adaptive immune cells increased in the Cluster A. (Fig. 3C). The distribution of clinical variables was not random within the Cluster A and B. (Supplementary Figure S2)

Validation of breast cancer TME clusters

To ensure that the two consensus clusters were not influenced by the analysis algorithm, we measured the abundance of cell infiltration for both phenotypes using MCP-counter (Fig. 3D, Supplementary Table S6) and estimated tumor purity using ESTIMATE (Fig. 3E) and ABSOLUTE, Leukocytes Unmethylation for Purity(LUMP), IHC (Supplementary Table S7) to validate the stability and robustness of the ssGSEA results. Infiltration of immune cells was much more prevalent in the Cluster A, whereas endothelial cells and fibroblasts were significantly more prevalent in the Cluster B. This result was in line with the findings of the ssGSEA. Additionally, the tumor purity of the Cluster A was much worse than that of the Cluster B. (Fig. 3E). Further, we analyzed the distribution of molecular subtypes of breast cancer in clusters, locating triple-negative breast cancer mostly in the Cluster A and other molecular subtypes in the Cluster B. (Fig. 3F, Supplementary Figure S2). In conclusion, we revealed that breast tumors had two diverse TME phenotypes. The characteristics of the two clusters were demonstrated in Fig. 3G.

Prognostic significance of TME cells in breast cancer

Given the significance of the TME in prognosis of breast cancer, we further studied the prognostic significance of TME clusters. The Cluster A had considerably superior OS (P < 0.05) (Fig. 4A). In addition, the prognostic importance of each cell subpopulation was investigated (Fig. 4B). A better prognosis was indicated by a larger degree of immune cell infiltration, including immunosuppressive cell infiltration, in the Cluster A, the Cluster B, and the whole cohort.

Activation of immune modulators in breast cancer

The activation of immune checkpoint molecules in response to immunological stimulation was a potentially significant intrinsic immune escape mechanism, thus we chose CD274, CTLA4, IDO1, Siglec-9, PDCD1, PDCD1LG2, LAG3, and HAVCR2 to evaluate the immune activity of each cluster. Wilcoxon test indicated that all eight immune checkpoint-related genes were substantially over-expressed in the Cluster A. (Fig. 5A). Axoneme, axoneme assembly, and cilium movement pathways were activated in the Cluster A, according to the GSEA. (Fig. 5B).

Construction of Radiomics Signature

The radiomics signature for predicting TME phenotypes was developed using random forest machine learning. Table S8 displayed the specific details of the top 20 features that contributed the most in the evaluation of feature importance. In the Radiomics discovery cohort, the radiomics signature deserved an AUC of 0.855 (95% confidence interval [CI]: 0.777–0.932, P < 0.05) for identifying TME phenotypes (Fig. 6A). The optimal cut-off value calculated for the Radiomics discovery cohort based on the Youden’s index was 0.528. Using this threshold value, patients' tumors were classified as "hot" or "cold." The confusion matrix was used to further test the performance of the model, and Figs. 6B displayed the sensitivity, specificity, and other evaluation metrics, which all suggested that the signature had excellent predictive performance.

Validation of radiomics signature

The radiomics signature had a comparable capacity to predict TME phenotypes in the Radiomics validation cohort, with an AUC of 0.844 (95% CI, 0.606-1.000, P < 0.05) (Fig. 6A), and when the optimal cut-off point was set at 0.569, the sensitivity and specificity could approach 0.80 and 0.89, respectively (Fig. 6B).

The efficacy of the radiomics signature to predict TME phenotypes in the Immune phenotype cohort revealed an AUC of 0.814 (95% CI, 0.717–0.911, P < 0.05). (Fig. 6A). The assessment metrics further illustrated the radiomics signature's strong prediction ability (Fig. 6B). Patients in the immune-inflamed (CD8+) group exhibited substantially higher rad-scores than those in the immune-desert (CD8-) group (P < 0.01) (Fig. 6C). Furthermore, the radiomics signature performed better in both HER2(+) and HER2(-)/HR(+) groupings (Supplementary Figure S3).

The follow-up period for the Immunotherapy-treated cohort was 6 months following the initiation of immunotherapy. The AUC for the radiomics signature prediction capacity for objective response (complete and partial response) and no response (stable disease and progressing disease) to immunotherapy was 0.784 (95% CI, 0.643–0.926, P < 0.01) (Fig. 6A). However, there was no statistical significance (P > 0.05) in the capacity of the radiomics signature to identify disease control (stable disease, partial response, and complete response) and progressing disease. Patients with objective response had higher rad-scores than those with stable disease and progressive disease (P < 0.05) (Fig. 6C), whereas patients with disease control did not have higher rad-scores than those with progressive disease (P > 0.05).

Relationship between the radiomics signature and immune checkpoint

Radiomics predictors could distinguish between cases with positive (tumor proportion score ≥ 1+) and negative PD-L1 expression in the Immune phenotype cohort (P < 0.01). (Fig. 6C). Specifically, immune infiltration was significantly higher in the PD-L1-positive cases than the PD-L1-negative one. CD274 expression was greatly enhanced in the immune-inflamed phenotype than in the immune-desert one, as seen in Fig. 5A. Thus, the consistency of our results demonstrates the superior predictive power of our radiomics signature for the TME phenotypes. However, the radiomics predictor did not demonstrate the capacity (P > 0.05) to differentiate between other clinicopathological factors (e.g., TNM stage, breast cancer molecular subtype) (Supplementary Figure S4).

Interpretability of radiomics features

The linkage between the top 20 contributing radiomics features and the abundance of 24 immune cell and stromal cell infiltrates was examined in the Radiomics discovery cohort (Fig. 7, Table S9). The findings showed a significant link between most cell types and radiomics features (P < 0.05), and this correlation was most apparent and largely negative between Neutrophils, Nk cells, Plasma cells, and T cells CD4 memory activated and radiomics features. Due to similar importance of each radiomics feature, the correlation level with the abundance of 24 cellular subpopulation was distributed evenly.

Our work makes use of massive TCGA breast cancer data to identify two heterogeneous breast cancer TME subtypes and related clinical significance. The Cluster A is a so-called "hot tumor," whereas the Cluster B is a "cold tumor." We emphasize the outstanding features of clusters that may cause immune escape: enhanced expression of immune checkpoint markers in the Cluster A and deficiencies in innate immune cell recruitment in the Cluster B. Our current analysis is congruent with the findings from Xavier Tekpli and Wen Huang et.^13,14 and is consistent with the immunologic principles outlined in a prior article³³. Our findings have substantial implications for clinical translation; especially, they may aid in the identification of individuals who will benefit from the ICIs therapy. We discover a "hot tumor" cluster in breast cancer and find that elevated expression of immune checkpoint markers in this cluster may contribute to immunological escape. Although most clinical studies have demonstrated that ICIs have less than 10% success in triple negative breast cancer^6,24, these patients often receive numerous rounds of immunotherapy without first assessing the immune checkpoint proteins. Particularly, the effectiveness of PD-L1 inhibitors in the first-line monotherapy may reach 25%⁶, which matches the proportion of "hot tumors" in our current analysis. It is expected that certain breast cancer cells that respond to ICIs would also respond to chemotherapy. As a consequence, the ICIs-sensitive cells are removed after many lines of chemotherapy, which is not reflected in the levels of immune checkpoint proteins following surgery. Given ICIs' more focus effectiveness over chemotherapy, we suggest that ICIs should be used sooner in the "hot tumor" of breast cancer.

Although immunotherapy is increasingly employed in breast cancer, PD-L1 expression is the most widely used biomarker associated with tumor immune checkpoint treatment; yet, for the majority of tumors, PD-L1 measured by IHC assays are unsatisfactory as a biological marker for the anti-PD-1/PD-L1 therapy³⁴. As a result, new biomarkers for predicting and monitoring patient response to immunotherapy are required, which is an essential step toward the age of precision immuno-oncology. Our work seeks to address this requirement by presenting an MRI-based biomarker that, given the ubiquitous availability and routine use of MRI, may be relevant and accessible. We created a radiomics signature of TME cell infiltration from MR images and examined the relationship between the imaging features, transcriptome data, TME phenotypes, and clinical response to immunotherapy in our research. Furthermore, the radiomics signature of the TME phenotypes was confirmed in three different cohorts, showing its relationship with the immune phenotypes while predicting clinical outcomes in patients treated with the anti-PD-1/PD-L1. A research published in Lancet Oncol by Roger Sun et al. ³² developed a radiomics signature of tumor immune infiltration from CT scans that may be used to predict patient immunotherapy performance. In contrast to that study, Roger Sun's study assessed tumor immune infiltration by the abundance of infiltrating CD8 cells. Although current articles examining TILs by IHC typically use CD8 + TILs levels to represent TILs^35–37, immune cells located within mesenchyme and cancer nest calculated by transcriptomic data include lymphocytes and various phagocytes. However, TILs within IHC generally refer to lymphocytes within the mesenchyme, which indicates that the infiltration of immune cells calculated by the transcriptome in a more broader range.

In this work, the radiomics signature comprises mostly of textural features that may objectively, statistically, and multidimensionally shows tumor biology and inherent heterogeneity. An direct interpretation of this feature is that homogeneous and low-density tumor and peripheral rings are linked with the enhanced immune cell infiltration³². To further explain the biological aspects of the radiomics features, we correlated the top 20 extracted significant radiomics features with the 24 TME cell infiltration abundances. We observed a substantial association between the majority of cell populations and the radiomics features, suggesting that our extracted features may represent the phenotypes of breast cancer TME. Notably, a recent study³² examined features extracted from the intratumoral region and its peritumoral region when treated by breast cancer immunotherapy, finding a link between immune cell infiltration (CD8) into the tumor and a CT-based radiomics signature, which was consistent with our findings. Furthermore, other study³⁸ concluded that peritumoral textural features might indicate TME, lending credence to our findings.

A rising number of studies^19,39,40 have explored radiomics as a predictor of immune infiltration or immunological pathways in recent years, although few patients in these research underwent immunotherapy. Another work employed machine learning to predict overall survival and responsiveness to immunotherapy, demonstrating a link between radiomics and genetics or biology in lung cancer and gives theoretical foundation for future research⁴⁰. Our work provides more evidence of the relationship between the MR radiomics, the TME phenotypes, and the outcomes of the anti-PD-1/PD-L treatments.

The immue-inflamed phenotype and the immune-desert phenotype are the only two immune phenotypes that we decided to examine in our research. This decision was taken so that the rad-score could be classified as high or low and the findings could be analyzed appropriately. We focused on the degree of immune and stromal cell infiltration in breast cancer TME since a lack of immune infiltration has been linked to poor immunotherapy response^25,41. To account for the spatial distribution of immune and stromal cells in tumor TME, we displayed two areas independently, tumor and peripheral margin, and utilized radiomics signature to predict the distinct phenotypes of TME based on imaging features for both. As we expected, future prospective direction might incorporate these three phenotypes, as well as possible additional immune phenotypes, as we have gained a better knowledge of breast cancer and its microenvironment immune function¹².

Our research has certain shortcomings. First, there are more than 24 different kinds of stromal cells, making it challenging to cover all phenotypes properly. Three kinds of these cell subpopulations were indicated—adaptive and activated innate immune cells, inactivated innate immune cells, and non-immune cells—was a partial solution to this issue. The second is the cohort's heterogeneity. The objective of choosing diverse cohorts was to provide a general method for characterizing breast cancer that would capture its underlying behavior. Different cohorts were used for training and assessment to prevent overfitting. Consequently, we sought to homogenize the data by establishing quality standards, pre-selecting pictures based on the reconstruction algorithm, and taking image acquisition parameters into account. However, these constraints reduced the number of eligible patients. Due to the retrospective nature of our work, validation of the findings requires further examination in a large prospective study.

Taken together, our research demonstrates that the TME phenotypes of breast cancer may be categorized into two distinguished heterogeneous clusters. Radiomics might be a noninvasive, cost-effective, and dependable tool for determining TME phenotypes and clinical response to immunotherapy in patients. Although the outcomes need to be replicated in a larger prospective test, they show the possibility of developing non-invasive biomarkers in immunotherapy.

TME: tumor microenvironment; PD-L1: anti-programmed cell death ligand 1; ICIs: immune checkpoint inhibitors; TCGA: the Cancer Genome Atlas; TCIA: the Cancer Imaging Archive; IHC: Immunohistochemistry; RECIST: Response Evaluation Criteria for Solid Tumors; AUC: area under the subject operating characteristic curve; LUMP: Leukocytes Unmethylation for Purity

Ethics approval and consent to participate

Our research was authorized by the review committee of Guangdong Provincial People's Hospital and conducted in compliance with the Helsinki Declaration's ethical standards.

Consent for publication

Due to the retroactive character of the investigation, the committee waived the informed consent requirement. All trial data were de-identified and made anonymous.

Availability of data and materials

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

Competing interests

The authors declare that they have no competing interests.

Funding

This study was supported in part by the Key-Area Research and Development Program of Guangdong Province (grant numbers: 2021B0101420006), National Key R&D Program of China (grant numbers: 2021YFF1201003), Regional Innovation and Development Joint Fund of National Natural Science Foundation of China (grant numbers: U22A20345), National Natural Science Foundation of China (grant numbers: 82071892), National Natural Science Foundation of China (grant numbers: 82271941), Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application (grant numbers: 2022B1212010011), High-level Hospital Construction Project (grant numbers: DFJHBF202105).

Authors' contributions

Changhong Liang, Zaiyi Liu and Xinhua Wei conceived and supervised the study; Xiaorui Han designed experiments and collected imaging data; Qingru Hu made statistical analysis; Huifen Ye performed bioinformatic analysis; Xiaorui Han and Zhihong Chen constructed and validated radiomics signature; Xiaorui Han and Yuan Guo wrote the manuscript and made results visualization; Changhong Liang made manuscript revisions.

Acknowledgments

Not applicable.

Zheng, R.; Zhang, S.; Zeng, H.; Wang, S.; Sun, K.; Chen, R.; Li, L.; Wei, W.; He, J. Cancer Incidence and Mortality in China, 2016. J. Natl. Cancer Cent. 2022, 2 (1), 1–9. https://doi.org/10.1016/j.jncc.2022.02.002.
Sittenfeld, S. M. C.; Zabor, E. C.; Hamilton, S. N.; Kuerer, H. M.; El-Tamer, M.; Naoum, G. E.; Truong, P. T.; Nichol, A.; Smith, B. D.; Woodward, W. A.; Moo, T.-A.; Powell, S. N.; Shah, C. S.; Taghian, A. G.; Abu-Gheida, I.; Tendulkar, R. D. A Multi-Institutional Prediction Model to Estimate the Risk of Recurrence and Mortality after Mastectomy for T1-2N1 Breast Cancer. Cancer 2022, 128 (16), 3057–3066. https://doi.org/10.1002/cncr.34352.
Van Baelen, K.; Geukens, T.; Maetens, M.; Tjan-Heijnen, V.; Lord, C. J.; Linn, S.; Bidard, F.-C.; Richard, F.; Yang, W. W.; Steele, R. E.; Pettitt, S. J.; Van Ongeval, C.; De Schepper, M.; Isnaldi, E.; Nevelsteen, I.; Smeets, A.; Punie, K.; Voorwerk, L.; Wildiers, H.; Floris, G.; Vincent-Salomon, A.; Derksen, P. W. B.; Neven, P.; Senkus, E.; Sawyer, E.; Kok, M.; Desmedt, C. Current and Future Diagnostic and Treatment Strategies for Patients with Invasive Lobular Breast Cancer. Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 2022, 33 (8), 769–785. https://doi.org/10.1016/j.annonc.2022.05.006.
Narayan, P.; Wahby, S.; Gao, J. J.; Amiri-Kordestani, L.; Ibrahim, A.; Bloomquist, E.; Tang, S.; Xu, Y.; Liu, J.; Fu, W.; Song, P.; King-Kallimanis, B. L.; Hou, S.; Gong, Y.; Kalavar, S.; Ghosh, S.; Philip, R.; Goldberg, K. B.; Theoret, M. R.; Blumenthal, G. M.; Kluetz, P. G.; Sridhara, R.; Pazdur, R.; Beaver, J. A. FDA Approval Summary: Atezolizumab Plus Paclitaxel Protein-Bound for the Treatment of Patients with Advanced or Metastatic TNBC Whose Tumors Express PD-L1. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 2020, 26 (10), 2284–2289. https://doi.org/10.1158/1078-0432.CCR-19-3545.
Gianni, L.; Huang, C. S.; Egle, D.; Bermejo, B.; Zamagni, C.; Thill, M.; Anton, A.; Zambelli, S.; Bianchini, G.; Russo, S.; Ciruelos, E. M.; Greil, R.; Semiglazov, V.; Colleoni, M.; Kelly, C.; Mariani, G.; Del Mastro, L.; Maffeis, I.; Valagussa, P.; Viale, G. Pathologic Complete Response (PCR) to Neoadjuvant Treatment with or without Atezolizumab in Triple-Negative, Early High-Risk and Locally Advanced Breast Cancer: NeoTRIP Michelangelo Randomized Study. Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 2022, 33 (5), 534–543. https://doi.org/10.1016/j.annonc.2022.02.004.
Emens, L. A. Breast Cancer Immunotherapy: Facts and Hopes. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 2018, 24 (3), 511–520. https://doi.org/10.1158/1078-0432.CCR-16-3001.
Schmid, P.; Salgado, R.; Park, Y. H.; Muñoz-Couselo, E.; Kim, S. B.; Sohn, J.; Im, S.-A.; Foukakis, T.; Kuemmel, S.; Dent, R.; Yin, L.; Wang, A.; Tryfonidis, K.; Karantza, V.; Cortés, J.; Loi, S. Pembrolizumab plus Chemotherapy as Neoadjuvant Treatment of High-Risk, Early-Stage Triple-Negative Breast Cancer: Results from the Phase 1b Open-Label, Multicohort KEYNOTE-173 Study. Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 2020, 31 (5), 569–581. https://doi.org/10.1016/j.annonc.2020.01.072.
Loi, S.; Dushyanthen, S.; Beavis, P. A.; Salgado, R.; Denkert, C.; Savas, P.; Combs, S.; Rimm, D. L.; Giltnane, J. M.; Estrada, M. V.; Sánchez, V.; Sanders, M. E.; Cook, R. S.; Pilkinton, M. A.; Mallal, S. A.; Wang, K.; Miller, V. A.; Stephens, P. J.; Yelensky, R.; Doimi, F. D.; Gómez, H.; Ryzhov, S. V.; Darcy, P. K.; Arteaga, C. L.; Balko, J. M. Correction: RAS/MAPK Activation Is Associated with Reduced Tumor-Infiltrating Lymphocytes in Triple-Negative Breast Cancer: Therapeutic Cooperation Between MEK and PD-1/PD-L1 Immune Checkpoint Inhibitors. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 2019, 25 (4), 1437. https://doi.org/10.1158/1078-0432.CCR-18-4264.
Nalio Ramos, R.; Missolo-Koussou, Y.; Gerber-Ferder, Y.; Bromley, C. P.; Bugatti, M.; Núñez, N. G.; Tosello Boari, J.; Richer, W.; Menger, L.; Denizeau, J.; Sedlik, C.; Caudana, P.; Kotsias, F.; Niborski, L. L.; Viel, S.; Bohec, M.; Lameiras, S.; Baulande, S.; Lesage, L.; Nicolas, A.; Meseure, D.; Vincent-Salomon, A.; Reyal, F.; Dutertre, C.-A.; Ginhoux, F.; Vimeux, L.; Donnadieu, E.; Buttard, B.; Galon, J.; Zelenay, S.; Vermi, W.; Guermonprez, P.; Piaggio, E.; Helft, J. Tissue-Resident FOLR2+ Macrophages Associate with CD8+ T Cell Infiltration in Human Breast Cancer. Cell 2022, 185 (7), 1189-1207.e25. https://doi.org/10.1016/j.cell.2022.02.021.
Li, H.; Yang, P.; Wang, J.; Zhang, J.; Ma, Q.; Jiang, Y.; Wu, Y.; Han, T.; Xiang, D. HLF Regulates Ferroptosis, Development and Chemoresistance of Triple-Negative Breast Cancer by Activating Tumor Cell-Macrophage Crosstalk. J. Hematol. Oncol.J Hematol Oncol 2022, 15 (1), 2. https://doi.org/10.1186/s13045-021-01223-x.
Santoni, M.; Romagnoli, E.; Saladino, T.; Foghini, L.; Guarino, S.; Capponi, M.; Giannini, M.; Cognigni, P. D.; Ferrara, G.; Battelli, N. Triple Negative Breast Cancer: Key Role of Tumor-Associated Macrophages in Regulating the Activity of Anti-PD-1/PD-L1 Agents. Biochim. Biophys. Acta Rev. Cancer 2018, 1869 (1), 78–84. https://doi.org/10.1016/j.bbcan.2017.10.007.
Zeng, D.; Li, M.; Zhou, R.; Zhang, J.; Sun, H.; Shi, M.; Bin, J.; Liao, Y.; Rao, J.; Liao, W. Tumor Microenvironment Characterization in Gastric Cancer Identifies Prognostic and Immunotherapeutically Relevant Gene Signatures. Cancer Immunol. Res. 2019, 7 (5), 737–750. https://doi.org/10.1158/2326-6066.CIR-18-0436.
Tekpli, X.; Lien, T.; Røssevold, A. H.; Nebdal, D.; Borgen, E.; Ohnstad, H. O.; Kyte, J. A.; Vallon-Christersson, J.; Fongaard, M.; Due, E. U.; Svartdal, L. G.; Sveli, M. A. T.; Garred, Ø.; OSBREAC; Frigessi, A.; Sahlberg, K. K.; Sørlie, T.; Russnes, H. G.; Naume, B.; Kristensen, V. N. An Independent Poor-Prognosis Subtype of Breast Cancer Defined by a Distinct Tumor Immune Microenvironment. Nat. Commun. 2019, 10 (1), 5499. https://doi.org/10.1038/s41467-019-13329-5.
Xu, Q.; Chen, S.; Hu, Y.; Huang, W. Landscape of Immune Microenvironment Under Immune Cell Infiltration Pattern in Breast Cancer. Front. Immunol. 2021, 12, 711433. https://doi.org/10.3389/fimmu.2021.711433.
Liu, Z.; Wang, S.; Dong, D.; Wei, J.; Fang, C.; Zhou, X.; Sun, K.; Li, L.; Li, B.; Wang, M.; Tian, J. The Applications of Radiomics in Precision Diagnosis and Treatment of Oncology: Opportunities and Challenges. Theranostics 2019, 9 (5), 1303–1322. https://doi.org/10.7150/thno.30309.
Lambin, P.; Rios-Velazquez, E.; Leijenaar, R.; Carvalho, S.; van Stiphout, R. G. P. M.; Granton, P.; Zegers, C. M. L.; Gillies, R.; Boellard, R.; Dekker, A.; Aerts, H. J. W. L. Radiomics: Extracting More Information from Medical Images Using Advanced Feature Analysis. Eur. J. Cancer Oxf. Engl. 1990 2012, 48 (4), 441–446. https://doi.org/10.1016/j.ejca.2011.11.036.
Lambin, P.; Leijenaar, R. T. H.; Deist, T. M.; Peerlings, J.; de Jong, E. E. C.; van Timmeren, J.; Sanduleanu, S.; Larue, R. T. H. M.; Even, A. J. G.; Jochems, A.; van Wijk, Y.; Woodruff, H.; van Soest, J.; Lustberg, T.; Roelofs, E.; van Elmpt, W.; Dekker, A.; Mottaghy, F. M.; Wildberger, J. E.; Walsh, S. Radiomics: The Bridge between Medical Imaging and Personalized Medicine. Nat. Rev. Clin. Oncol. 2017, 14 (12), 749–762. https://doi.org/10.1038/nrclinonc.2017.141.
Mohammadi, A.; Mirza-Aghazadeh-Attari, M.; Faeghi, F.; Homayoun, H.; Abolghasemi, J.; Vogl, T. J.; Bureau, N. J.; Bakhshandeh, M.; Acharya, R. U.; Abbasian Ardakani, A. Tumor Microenvironment, Radiology, and Artificial Intelligence: Should We Consider Tumor Periphery? J. Ultrasound Med. Off. J. Am. Inst. Ultrasound Med. 2022, 41 (12), 3079–3090. https://doi.org/10.1002/jum.16086.
Su, G.-H.; Xiao, Y.; Jiang, L.; Zheng, R.-C.; Wang, H.; Chen, Y.; Gu, Y.-J.; You, C.; Shao, Z.-M. Radiomics Features for Assessing Tumor-Infiltrating Lymphocytes Correlate with Molecular Traits of Triple-Negative Breast Cancer. J. Transl. Med. 2022, 20 (1), 471. https://doi.org/10.1186/s12967-022-03688-x.
Agostinetto, E.; Losurdo, A.; Nader-Marta, G.; Santoro, A.; Punie, K.; Barroso, R.; Popovic, L.; Solinas, C.; Kok, M.; de Azambuja, E.; Lambertini, M. Progress and Pitfalls in the Use of Immunotherapy for Patients with Triple Negative Breast Cancer. Expert Opin. Investig. Drugs 2022, 31 (6), 567–591. https://doi.org/10.1080/13543784.2022.2049232.
D-mannose facilitates immunotherapy and radiotherapy of triple-negative breast cancer via degradation of PD-L1 - PubMed. https://pubmed.ncbi.nlm.nih.gov/35181605/ (accessed 2022-12-04).
Neoadjuvant Chemotherapy and Immunotherapy in Luminal B-like Breast Cancer: Results of the Phase II GIADA Trial - PubMed. https://pubmed.ncbi.nlm.nih.gov/34667023/ (accessed 2022-12-04).
Emens, L. A.; Cruz, C.; Eder, J. P.; Braiteh, F.; Chung, C.; Tolaney, S. M.; Kuter, I.; Nanda, R.; Cassier, P. A.; Delord, J.-P.; Gordon, M. S.; ElGabry, E.; Chang, C.-W.; Sarkar, I.; Grossman, W.; O’Hear, C.; Fassò, M.; Molinero, L.; Schmid, P. Long-Term Clinical Outcomes and Biomarker Analyses of Atezolizumab Therapy for Patients With Metastatic Triple-Negative Breast Cancer: A Phase 1 Study. JAMA Oncol. 2019, 5 (1), 74–82. https://doi.org/10.1001/jamaoncol.2018.4224.
Cortes, J.; Cescon, D. W.; Rugo, H. S.; Nowecki, Z.; Im, S.-A.; Yusof, M. M.; Gallardo, C.; Lipatov, O.; Barrios, C. H.; Holgado, E.; Iwata, H.; Masuda, N.; Otero, M. T.; Gokmen, E.; Loi, S.; Guo, Z.; Zhao, J.; Aktan, G.; Karantza, V.; Schmid, P.; KEYNOTE-355 Investigators. Pembrolizumab plus Chemotherapy versus Placebo plus Chemotherapy for Previously Untreated Locally Recurrent Inoperable or Metastatic Triple-Negative Breast Cancer (KEYNOTE-355): A Randomised, Placebo-Controlled, Double-Blind, Phase 3 Clinical Trial. Lancet Lond. Engl. 2020, 396 (10265), 1817–1828. https://doi.org/10.1016/S0140-6736(20)32531-9.
Franzén, A. S.; Raftery, M. J.; Pecher, G. Implications for Immunotherapy of Breast Cancer by Understanding the Microenvironment of a Solid Tumor. Cancers 2022, 14 (13), 3178. https://doi.org/10.3390/cancers14133178.
Sofopoulos, M.; Fortis, S. P.; Vaxevanis, C. K.; Sotiriadou, N. N.; Arnogiannaki, N.; Ardavanis, A.; Vlachodimitropoulos, D.; Perez, S. A.; Baxevanis, C. N. The Prognostic Significance of Peritumoral Tertiary Lymphoid Structures in Breast Cancer. Cancer Immunol. Immunother. CII 2019, 68 (11), 1733–1745. https://doi.org/10.1007/s00262-019-02407-8.
Xiao, Y.; Ma, D.; Zhao, S.; Suo, C.; Shi, J.; Xue, M.-Z.; Ruan, M.; Wang, H.; Zhao, J.; Li, Q.; Wang, P.; Shi, L.; Yang, W.-T.; Huang, W.; Hu, X.; Yu, K.-D.; Huang, S.; Bertucci, F.; Jiang, Y.-Z.; Shao, Z.-M.; AME Breast Cancer Collaborative Group. Multi-Omics Profiling Reveals Distinct Microenvironment Characterization and Suggests Immune Escape Mechanisms of Triple-Negative Breast Cancer. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 2019, 25 (16), 5002–5014. https://doi.org/10.1158/1078-0432.CCR-18-3524.
Zhao, S.; Ma, D.; Xiao, Y.; Li, X.-M.; Ma, J.-L.; Zhang, H.; Xu, X.-L.; Lv, H.; Jiang, W.-H.; Yang, W.-T.; Jiang, Y.-Z.; Zhang, Q.-Y.; Shao, Z.-M. Molecular Subtyping of Triple-Negative Breast Cancers by Immunohistochemistry: Molecular Basis and Clinical Relevance. The Oncologist 2020, 25 (10), e1481–e1491. https://doi.org/10.1634/theoncologist.2019-0982.
Loi, S.; Michiels, S.; Adams, S.; Loibl, S.; Budczies, J.; Denkert, C.; Salgado, R. The Journey of Tumor-Infiltrating Lymphocytes as a Biomarker in Breast Cancer: Clinical Utility in an Era of Checkpoint Inhibition. Ann. Oncol. 2021, 32 (10), 1236–1244. https://doi.org/10.1016/j.annonc.2021.07.007.
Han, X.; Cao, W.; Wu, L.; Liang, C. Radiomics Assessment of the Tumor Immune Microenvironment to Predict Outcomes in Breast Cancer. Front. Immunol. 2021, 12, 773581. https://doi.org/10.3389/fimmu.2021.773581.
Huang, Y.; Wei, L.; Hu, Y.; Shao, N.; Lin, Y.; He, S.; Shi, H.; Zhang, X.; Lin, Y. Multi-Parametric MRI-Based Radiomics Models for Predicting Molecular Subtype and Androgen Receptor Expression in Breast Cancer. Front. Oncol. 2021, 11, 706733. https://doi.org/10.3389/fonc.2021.706733.
Sun, R.; Limkin, E. J.; Vakalopoulou, M.; Dercle, L.; Champiat, S.; Han, S. R.; Verlingue, L.; Brandao, D.; Lancia, A.; Ammari, S.; Hollebecque, A.; Scoazec, J.-Y.; Marabelle, A.; Massard, C.; Soria, J.-C.; Robert, C.; Paragios, N.; Deutsch, E.; Ferté, C. A Radiomics Approach to Assess Tumour-Infiltrating CD8 Cells and Response to Anti-PD-1 or Anti-PD-L1 Immunotherapy: An Imaging Biomarker, Retrospective Multicohort Study. Lancet Oncol. 2018, 19 (9), 1180–1191. https://doi.org/10.1016/S1470-2045(18)30413-3.
Faivre-Finn, C.; Vicente, D.; Kurata, T.; Planchard, D.; Paz-Ares, L.; Vansteenkiste, J. F.; Spigel, D. R.; Garassino, M. C.; Reck, M.; Senan, S.; Naidoo, J.; Rimner, A.; Wu, Y.-L.; Gray, J. E.; Özgüroğlu, M.; Lee, K. H.; Cho, B. C.; Kato, T.; de Wit, M.; Newton, M.; Wang, L.; Thiyagarajah, P.; Antonia, S. J. Four-Year Survival With Durvalumab After Chemoradiotherapy in Stage III NSCLC-an Update From the PACIFIC Trial. J. Thorac. Oncol. Off. Publ. Int. Assoc. Study Lung Cancer 2021, 16 (5), 860–867. https://doi.org/10.1016/j.jtho.2020.12.015.
Darvin, P.; Toor, S. M.; Sasidharan Nair, V.; Elkord, E. Immune Checkpoint Inhibitors: Recent Progress and Potential Biomarkers. Exp. Mol. Med. 2018, 50 (12), 1–11. https://doi.org/10.1038/s12276-018-0191-1.
So, Y. K.; Byeon, S.-J.; Ku, B. M.; Ko, Y. H.; Ahn, M.-J.; Son, Y.-I.; Chung, M. K. An Increase of CD8+ T Cell Infiltration Following Recurrence Is a Good Prognosticator in HNSCC. Sci. Rep. 2020, 10 (1), 20059. https://doi.org/10.1038/s41598-020-77036-8.
Song, Z.; Yang, L.; Zhou, Z.; Li, P.; Wang, W.; Cheng, G.; Chen, R.; Chang, L.; Zhang, Y.; Guan, Y.; Xia, X.; Yi, X.; Zhou, R.; Chen, M. Genomic Profiles and Tumor Immune Microenvironment of Primary Lung Carcinoma and Brain Oligo-Metastasis. Cell Death Dis. 2021, 12 (1), 106. https://doi.org/10.1038/s41419-021-03410-7.
CD8+ tumor-infiltrating lymphocytes within the primary tumor of patients with synchronous de novo metastatic colorectal carcinoma do not track with survival - PubMed. https://pubmed.ncbi.nlm.nih.gov/32953115/ (accessed 2022-12-06).
Yang, B.; Zhou, L.; Zhong, J.; Lv, T.; Li, A.; Ma, L.; Zhong, J.; Yin, S.; Huang, L.; Zhou, C.; Li, X.; Ge, Y. Q.; Tao, X.; Zhang, L.; Son, Y.; Lu, G. Combination of Computed Tomography Imaging-Based Radiomics and Clinicopathological Characteristics for Predicting the Clinical Benefits of Immune Checkpoint Inhibitors in Lung Cancer. Respir. Res. 2021, 22 (1), 189. https://doi.org/10.1186/s12931-021-01780-2.
Xu, N.; Zhou, J.; He, X.; Ye, S.; Miao, H.; Liu, H.; Chen, Z.; Zhao, Y.; Pan, Z.; Wang, M. Radiomics Model for Evaluating the Level of Tumor-Infiltrating Lymphocytes in Breast Cancer Based on Dynamic Contrast-Enhanced MRI. Clin. Breast Cancer 2021, 21 (5), 440-449.e1. https://doi.org/10.1016/j.clbc.2020.12.008.
Khorrami, M.; Prasanna, P.; Gupta, A.; Patil, P.; Velu, P. D.; Thawani, R.; Corredor, G.; Alilou, M.; Bera, K.; Fu, P.; Feldman, M.; Velcheti, V.; Madabhushi, A. Changes in CT Radiomic Features Associated with Lymphocyte Distribution Predict Overall Survival and Response to Immunotherapy in Non-Small Cell Lung Cancer. Cancer Immunol. Res. 2020, 8 (1), 108–119. https://doi.org/10.1158/2326-6066.CIR-19-0476.
Van Damme, H.; Dombrecht, B.; Kiss, M.; Roose, H.; Allen, E.; Van Overmeire, E.; Kancheva, D.; Martens, L.; Murgaski, A.; Bardet, P. M. R.; Blancke, G.; Jans, M.; Bolli, E.; Martins, M. S.; Elkrim, Y.; Dooley, J.; Boon, L.; Schwarze, J. K.; Tacke, F.; Movahedi, K.; Vandamme, N.; Neyns, B.; Ocak, S.; Scheyltjens, I.; Vereecke, L.; Nana, F. A.; Merchiers, P.; Laoui, D.; Van Ginderachter, J. A. Therapeutic Depletion of CCR8+ Tumor-Infiltrating Regulatory T Cells Elicits Antitumor Immunity and Synergizes with Anti-PD-1 Therapy. J. Immunother. Cancer 2021, 9 (2), e001749. https://doi.org/10.1136/jitc-2020-001749.

No competing interests reported.

Supplementarymaterialstable.xlsx
FigureS1.tif
FIGURE S1. Determination of the optimal cluster number of breast cancer TME phenotypes. A-D, Silhouette analysis for KMeans clustering on the breast cancer data. E, Nbclust test of the breast cancer data.
FigureS2.tif
FIGURE S2. The distribution of different clinical factors of Cluster A and B. The distributions of age and molecular type are significantly different between the two clusters, whereas the values of the other clinicopathological features are similar. * ****, P<0.0001; *, P<0.05; NS, no significant.
FigureS3.tif
Figure S3. Performance of radiomics signature in the Immune phenotype cohort. A, Diagnostic efficacy of radiomics signature in different molecular subtypes. B, Evaluation metrics of radiomics signature in different molecular subtypes. * ACC, accuracy; SEN, sensitivity; SEP, specificity; PPV, positive predictive value; NPV, negative predictive value.
FigureS4.tif
Figure S4. The efficiency of radiomics signature for differentiating various TNM stages (left) and molecular subtypes (right).

Download PDF

Journal Publication

published 29 Jan, 2024

Read the published version in Breast Cancer Research →

Editorial decision: Major revision
17 Sep, 2023
Reviews received at journal
24 Aug, 2023
Reviewers agreed at journal
10 Aug, 2023
Reviewers invited by journal
06 Aug, 2023
Editor assigned by journal
28 Jun, 2023
Submission checks completed at journal
28 Jun, 2023
First submitted to journal
24 Jun, 2023

You are reading this latest preprint version

Development of a machine learning-based radiomics signature for estimating breast cancer TME phenotypes and predicting anti-PD-1/PD-L1 immunotherapy response

Status:

Journal Publication

Version 1

Abstract

Backgrounds:

Methods

Results

Conclusions

Figures

Introduction

Materials

Protocol and data sources

Calculation of TME cell abundance

TME phenotypes clustering

Prognostic value of TME phenotypes

MR image analysis and feature extraction

Construction of radiomics signature

Statistical analysis

Role of funding sources

Results

Demographic and clinical characteristics

Landscape of TME phenotypes in breast cancer

Validation of breast cancer TME clusters

Prognostic significance of TME cells in breast cancer

Activation of immune modulators in breast cancer

Construction of Radiomics Signature

Validation of radiomics signature

Relationship between the radiomics signature and immune checkpoint

Interpretability of radiomics features

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1