European Journal of Cardiovascular Medicine

With over a million cases per year, breast cancer is the leading cause of mortality for women globally. Breast carcinoma has the largest incidence of malignant tumors among women globally, which accounts for 15% of all malignant tumor-related mortality, ranks sixth in terms of malignant tumor death rate, and tends to increase quickly [1]. The tissue microenvironment plays a crucial role in regulating cell survival, proliferation, migration, polarization, and differentiation [2,3]. The histological traits, metastatic patterns, molecular characteristics, prognosis, and treatment responses of breast cancer vary widely.

In all likely cases of breast cancer, established clinicopathological prognostic factors such as histological tumor grade, lymph node metastases, human epidermal growth factor receptor 2/neu protooncogene (HER2-neu), progesterone receptor (PR), and estrogen receptor (ER) status are routinely considered. The stroma has a crucial role in controlling and altering tumor invasion and metastasis. Novel indicators for determining the prognosis of invasive breast cancer are stromal markers [4]. In many epithelial cancers and phyllodes tumors, stromal CD10 expression has been linked to biological aggressiveness [5]. CD10 is a zinc dependent metalloproteinase frequently expressed in bone marrow lymphoid stem cells, pro-B lymphoblasts, mature neutrophils, various lymphoma subtypes, renal cell carcinoma and endometrial stromal cell carcinoma. No systematic studies have been conducted so far in Kerala evaluating CD 10 expression in breast carcinoma. Hence the present study is undertaken to quantify the prevalence of CD10 in carcinoma breast and evaluate their associations with ER, PR, HER2, composite profiles including triple-negative phenotypes, and invasion-related features.

This cross-sectional study was carried out in a tertiary care centre in South India. The ethical approval for the study was obtained from the Institutional Ethical Committee. A cross-sectional analysis of 151 consecutive primary breast carcinoma cases was performed with clinicopathological characterization (age, Modified Bloom–Richardson grade, DCIS grade, lymphovascular invasion, perineural invasion) and immunohistochemistry for ER, PR, HER2 and stromal CD10. All slides were evaluated by senior histopathologists. Tumor node metastasis classification and staging of cases was done as per American Joint Committee on Cancer Guidelines [6]. The most suitable tissue block of breast cancer cases was selected for immunohistochemical evaluation. Stromal expression of CD 10 was studied on entire section of selected breast cancer blocks for all cases. More than 10% of stromal cells with cytoplasmic & membranous staining were taken as positive Descriptive statistics summarized distributions; associations between CD10 and categorical covariates were evaluated using chi-square tests with corresponding p-values and degrees of freedom. Concordance between CD10 and receptor-defined categories was assessed with Cohen’s kappa. Pre-specified subgroup analyses included triple-negative and composite ER/PR/HER2 phenotypes

The present study analyzed 151 cases. The average age of patients in this study was 57 years, with a range of 30 to 80 years. Among the 151 cases included in the study sample, 103 (68.2%) cases demonstrated CD10 expression in tumor stromal cells on immunohistochemistry. Analysis of the Modified Bloom–Richardson (MBR) grading among the 151 breast carcinoma cases revealed that the majority of tumors, 98 cases (64.9%), were classified as grade 2, 31 cases grade 1 (20.5%) and 22 cases (14.2%) grade 3. In the analysis of the ductal carcinoma in situ (DCIS) component, the majority of tumors, 126 cases (83.4%), showed absence of a DCIS component. Low-grade DCIS was identified in 9 cases (6%), while high-grade and intermediate-grade DCIS were each observed in 8 cases (5.3%). Lymphovascular invasion (LVI) analysis revealed that 19 cases (12.6%) exhibited evidence of LVI, while 132 cases (87.4%) showed no lymphovascular involvement, indicating that a relatively small proportion of tumors demonstrated invasion into lymphatic or vascular channels.

Only 2 cases (1.3%) demonstrated the presence of PNI, while the vast majority, 149 cases (98.7%), showed absence of perineural involvement, indicating that PNI is an uncommon feature in the current dataset. ER positive expression was noted in 97 cases (64.2%), while 54 cases (35.8%) were ER negative. This indicates that a majority of tumors in the present cohort express estrogen receptors. In the evaluation of HER2/neu expression among 151 breast carcinoma cases, 31 cases (20.5%) were HER2/neu positive, while 120 cases (79.5%) were negative for HER2 overexpression. This finding indicates that the majority of tumors in the studied population lacked HER2 amplification, representing the predominance of non-HER2-driven subtypes.PR expression was noted in 83 cases (55%) while 68 cases (45%) are PR negative.

Within the ER-positive, PR-positive, HER2-negative subgroup, a near-equal split in CD10 expression was observed, with 77 cases (51%) demonstrating CD10 positivity and 74 cases (49%) lacking CD10 expression among a total of 151 cases. This balanced distribution indicates that CD10 is expressed in roughly half of the luminal (HER2-negative) tumors, suggesting no clear predominance of stromal CD10 expression within this molecular phenotype in the presented cohort.

Within the ER-positive, PR-positive, HER2-positive subgroup, only 12 of 151 cases (7.9%) were observed, while the remaining 139 cases (92.1%) did not express this profile, indicating that this phenotype constituted a small minority of the cohort. The low prevalence aligns with population-level distributions where HR+/HER2+ disease is less common than HR+/HER2− tumors, typically comprising under 10–12% of breast cancers depending on registry and period, reinforcing that triple-positive status is comparatively infrequent in unselected series.

In the ER-negative, PR-negative, HER2-positive subgroup, 19 of 151 cases (12.6%) exhibited this receptor profile, while 132 cases (87.4%) did not, indicating that the HER2-enriched phenotype without hormone receptor expression comprised a modest minority of the cohort.

In this cohort, triple-negative breast cancer (ER−/PR−/HER2−) constituted 35 of 151 cases (23.2%), while 116 cases (76.8%) were non–triple-negative, indicating that nearly one-quarter of tumors fell into the triple-negative phenotype.

On evaluating CD10 expression across various MBR grades, the findings demonstrated an increasing tendency of CD10 positivity with higher grades. Specifically, grade 1 lesions exhibited 54.8% CD10 positivity, whereas grade 2 and grade 3 lesions showed 71.4% and 72.7% positivity, respectively. This association is not statistically significant (p value 0.199); indicating that CD10 expression did not show a significant relationship with MBR grade in the present data.

The relationship between ductal carcinoma in situ (DCIS) grade and CD10 expression displayed varying rates of positivity among subgroups. Low-grade DCIS exhibited 55.6% positivity, high-grade DCIS showed 75% positivity, intermediate DCIS had 75% positivity, while cases with absent DCIS pattern demonstrated a positivity of 68.3%. The chi-square value was 1.01 with 3 degrees of freedom and a p-value of 0.800, indicating an absence of statistically significant association. Overall, CD10 expression appeared independent of DCIS grade, showing similar positivity rates across all subgroups.

Among the cases evaluated for lymphovascular invasion (LVI), 84.2% of LVI-positive tumors showed CD10 positivity compared to 65.9% of LVI-negative tumors. Although CD10 expression seemed higher in tumors with lymphovascular invasion, the chi-square test yielded a value of 2.57 with 1 degree of freedom and a p-value of 0.109. This result indicates that the association between LVI and CD10 expression was not statistically significant.

Assessment of CD10 expression in relation to perineural invasion (PNI) showed that all PNI-positive cases (100%) exhibited CD10 positivity, whereas 67.8% of PNI-negative cases were CD10 positive. Although this marked difference was noted, statistical evaluation using the chi-square test produced a value of 0.95 with 1 degree of freedom and a p-value of 0.331, which is not statistically significant. This indicates that while CD10 expression was universally observed in PNI-positive cases, the small number of such cases likely limited the strength of the statistical association.

Figure 3: Association between MBR Grade and CD 10 positivity

In cases displaying an ER-positive, PR-positive, and HER2-negative immunoprofile, CD10 positivity was found in 62.3% of cases, whereas 74.3% of CD10 positivity was observed among ER/PR-positive but HER2-negative tumors. The chi-square test returned a value of 2.5 with 1 degree of freedom and a p-value of 0.114, indicating a non-significant relationship. These findings suggest that CD10 expression does not show a meaningful correlation with ER and PR co-expression in the absence of HER2 amplification

ER, PR, HER2-positive immunoprofile showed a strong positive relationship with CD10 expression (p value = 0.010). In this group, all cases (100%) were CD10 positive. In contrast CD 10 positivity was seen in only 65.6% of cases which were not positive for all the three above immunomarkers. This demonstrates that CD10 expression is significantly associated with hormonal receptor-positive and HER2-positive breast tumors, suggesting a possible correlation between CD10 activity and aggressive tumor biology in this phenotype.The kappa value of 0.077 indicates only slight agreement and is close to chance, implying minimal concordance between the ER-positive/PR-positive/HER2-positive profile and CD10 expression.

For the ER-negative, PR-negative, and HER2-positive subtype, CD10 positivity was observed in 68.4% of cases, while 68.2% of CD10 positivity appeared among cases which were not HER2 Neu enriched. The chi-square result was 0 with a p-value of 0.983, confirming no statistically significant relationship between this receptor pattern and CD10 expression. Thus, CD10 expression appears consistent regardless of ER and PR negativity in HER2-enriched tumors. The kappa value of 0.001 indicates virtually no agreement beyond chance between the ER‑negative/PR‑negative/HER2‑positive profile and CD10 expression, reflecting negligible concordance.

In triple-negative breast cancer (TNBC) cases, 71.4% showed CD10 positivity compared to 67.2% positivity in non-triple-negative cases. The chi-square test yielded a value of 0.22 with a p-value of 0.641, showing no statistical significance. Although there was a mild numerical increase in CD10 expression among triple-negative tumors, the difference was not statistically meaningful. This indicates that CD10 expression is not a distinct feature associated with the triple-negative subtype in this dataset. The kappa value of 0.025 denotes only slight agreement, essentially indicating concordance no better than chance between triple‑negative status and CD10 expression.

The present study comprehensively evaluated CD10 expression in primary breast carcinomas from a Kerala cohort, linking descriptive clinicopathological profiles with immunohistochemical marker prevalence and their associations with standard prognostic receptors including ER, PR, and HER2. The dataset comprised 151 cases with a broad adult age spectrum, a predominance of invasive carcinoma of no special type, and a balanced distribution across key clinicopathologic strata to enable robust analyses of CD10 in relation to tumor biology.

Acrossanalysis, CD10 positivity was common at 68.2% (103/151) in the present study. A few studies found low to moderate variation in CD10 expression: some reported CD10 expression as high as 80% [7, 8]. Statistical testing with chi-square indicated that most associations between CD10 and clinicopathological covariates were not statistically significant, although specific profiles such as ER+/PR+/HER2-positive showed a significant association with CD10 expression.

Grading by the Modified Bloom–Richardson system demonstrated an intermediate-grade predominance with Grade 2 comprising 98 cases representing 64.9%, Grade 1 comprising 31 cases representing 20.5%, and Grade 3 comprising 22 cases representing 14.6%, supporting the interpretation that most tumors in this series demonstrate moderate histologic aggressiveness and present a balanced substrate for correlation analyses. This finding was consistent with results of studies by Mohammadizadeh et.al and Iwaya et.al [9, 10].

In situ and invasion-related features further characterized baseline pathology. An associated DCIS component was absent in 126 tumors representing 83.4%, with low grade present in 9 cases representing 6.0%, intermediate in 8 cases representing 5.3%, and high grade in 8 cases representing 5.3%, indicating that the great majority presented with invasive disease without substantial in situ remnants. Lymphovascular invasion was identified in 19 cases representing 12.6%, while 132 cases representing 87.4% lacked LVI, consistent with reported frequencies in unselected invasive cohorts and underscoring the clinical salience of LVI-positive subsets. Perineural invasion was rare, present in 2 cases representing 1.3%, and absent in 149 cases representing 98.7%, reinforcing that PNI is an uncommon histologic feature in breast carcinoma relative to other solid tumors

Association analysis between CD10 expression and clinicopathologic variables demonstrated broadly null results with some trends. Invasion-related features in the present analysis, such as LVI, showed trends toward higher CD10 in LVI-positive tumors without statistical significance (84.2% vs 65.9%, p = 0.109), whereas Fernandes et al. reported significant associations with LVI (p = 0.008) and necrosis (p = 0.008) alongside grade and nodal status, supporting the concept that CD10-enriched stroma accompanies invasive and necrotic tumor ecologies [11].The discrepancy likely reflects power constraints, as the present LVI-positive subset was modest, and potential whole-section versus TMA sampling differences where focal stromal hotspots may be preferentially captured. Similarly, perineural invasion was too rare in the present series to evaluate robustly, and literature rarely focuses on PNI in breast carcinoma relative to GI and head–neck malignancies, limiting comparative interpretations [11, 12].

Receptor-based associations provided the central inferential insights of the study. Composite receptor profiles yielded differential findings with one notable significant association. The notable significant association in the present analysis was between the ER+/PR+/HER2+ -positive profile and CD10 expression (100% vs 65.5%, chi-square 6.08, p=0.014), indicating that combined ER/PR activation with HER2 signaling may coexist with CD10-rich stromal states in this cohort. Prior literature frequently emphasizes CD10 enrichment in HER2-positive and triple-negative subtypes as markers of aggressive biology rather than in triple-positive luminal/HER2 hybrid phenotypes [11, 12, 13]

The ER+/PR+/HER2− profile showed CD10 positivity at 62.3% versus 74.3% in the corresponding negative category, with chi-square 2.5 and 1 degree of freedom, p = 0.114, indicating no significant association, and a kappa of − 0.121 implying agreement below chance and no meaningful concordance.In the ER−/PR−/HER2+ subgroup, CD10 positivity was 68.4% versus 68.2% in the comparator, with chi-square 0 and 1 degree of freedom, p=0.983, demonstrating no association and kappa of 0.001 reflecting virtually no agreement beyond chance, indicating that CD10 is not informative within HER2-enriched hormone receptor–negative tumors in this series.

The triple-negative analysis showed CD10 positivity at 71.4% compared with 67.2% in non-triple-negative tumors; chi-square was 0.22 with 1 degree of freedom and p = 0.641, indicating no significant association and suggesting that CD10 is not a distinctive hallmark of triple-negative phenotype.

Overall, literature synthesis suggests stromal CD10 is frequently expressed and often correlates with adverse phenotypes – ER/PR negativity, HER2 positivity, higher grade, LVI, nodal metastasis, and TNBC – though not uniformly across all cohorts, with variability driven by scoring criteria, specimen selection, and population biology. The present study’s predominantly null associations, with the singular significant enrichment in the triple-positive group, imply that in this Kerala population CD10 may be a common stromal feature across receptor classes rather than a discriminator of specific molecular subtypes.

This study delineates a landscape in which CD10 stromal expression is frequent but largely non-discriminatory across most receptor-defined categories, with a single significant enrichment in triple-positive disease.

1. Fan Y, He S: The characteristics of tumour microenvironment in triple-negative breast cancer. Cancer Manag Res. 2022, 14:1-17. 10.2147/CMAR.S316700

2. Jana SH, Jha BM, Patel C, Jana D, Agarwal A. CD10-a new prognostic stromal marker in breast carcinoma, its utility, limitations and role in breast cancer pathogenesis. Indian J Pathol Microbiol. 2014; 57:530–6. doi: 10.4103/0377-4929.142639.

3. Sharma AK, Hemrajani D, Harsh A, Pancharia A, Yadav M. An observational study of stromal cd10 expression in Breast carcinoma and its association with clinicopathological parameters at SMS Medical College and Hospital. Asian Pacific Journal of Cancer Biology. 2024 Mar 13;9(1):31-5.

4. Puri V, Jain M, Thomas S. Stromal Expression of CD10 in Invasive Breast Carcinoma and Its Correlation with ER, PR, HER2‐neu, and Ki67. International journal of breast cancer. 2011;2011(1):437957.

5. Tian F, Zhao J, Shi QF, Zhang DC, Wang YY, Sun JP, Wang Y, Yang SA, Zhang C. Association between CD10 expression and phyllodes tumor: A retrospective case control study. Medicine. 2023 Nov 3;102(44):e35677.

6.Amin MB, Greene FL, Edge SB, Compton CC, Gershenwald JE, Brookland RK, Meyer L, Gress DM, Byrd DR, Winchester DP. The eighth edition AJCC cancer staging manual: continuing to build a bridge from a population‐based to a more “personalized” approach to cancer staging. CA: a cancer journal for clinicians. 2017 Mar;67(2):93-9

7. Gaffoor N, Krishnamurthy J: Stromal expression of CD10 in breast carcinoma and its association with known prognostic factors - a tissue microarray-based study. J Lab Physicians. 2023, 15:354-60. 10.1055/s- 0043-1761925

8. Taghizadeh-Kermani A, Jafarian AH, Ashabyamin R, Seilanian-Toosi M, Pourali L, Asadi M, Mashhadi L: The stromal overexpression of CD10 in invasive breast cancer and its association with clinicopathologic factors. Iranian journal of cancer prevention. 2014, 7:17.

9.Mohammadizadeh F, Salavati M, Moghaddam NA: CD10 expression in the stromal component of invasive breast carcinoma: a potential prognostic determinant. J Res Med Sci. 2012, 17:194-9.

10. Iwaya K, Ogawa H, Izumi M, Kuroda M, Mukai K: Stromal expression of CD10 in invasive breast carcinoma: a new predictor of clinical outcome. Virchows Arch. 2002, 440:589-93.

11. Fernandes J, Gupta S, Sharma S, Kumar A, Bansal S, Kaur R, et al. Stromal expression of CD10 in breast carcinoma and its association with known prognostic factors—A tissue microarray-based study. J Lab Physicians. 2023;15(1):61-69. doi:10.1055/s-0043-1761925.

12.Shakil S, Memon A, Ahmed M, Samad A, Soomro I. Stromal CD10 expression in invasive breast carcinoma: correlation with ER, PR, HER2 and tumor grade in mastectomy cases. J Shalamar Med Coll. 2021;13(4):152-156

13. Paul A, Radhakrishnan A, Vinodchandra J, Anandi D, Das L, George B, et al. Stromal expression of CD10 in invasive breast carcinoma and its association with ER, PR and HER2/neu: A cross-sectional IHC study of 50 cases. J Acad Med Sci. 2022;6(6):86-90.