Ovarian low and high grade serous carcinomas: hidden divergent features in the tumor microenvironment

Abstract

Only recently low-grade serous carcinoma (LGSOC) of the ovary has been recognized as a disease entity distinct from the more common high-grade serous carcinoma (HGSOC), with significant differences in pathogenesis and clinical and pathologic features. The present study aimed at evaluating whether the different natural histories and patterns of response to therapy demonstrated for LGSOC and HGSOC, along with a diverse genomic landscape, may also reside in the supporting tumor stroma, specifically in the state of differentiation and activation of tumor associated macrophages (TAMs). TAMs play complex roles in tumorigenesis since they are believed to possess both tumor rejecting (M1 macrophages) and tumor promoting (M2 macrophages) activities. Here we showed that, when compared to HGSOC (n = 55), LGSOC patients (n = 25) exhibited lower density of tumor-infiltrating CD68+ macrophage, along with an attenuated M2-skewed (CD163+) phenotype. Accordingly, assessment of intratumoral vascularization and of matrix metalloproteinase 9 expression (a key protein involved in tumor invasion and metastasis) revealed lower expression in LGSOC compared to HGSOC patients, in line with emerging evidence supporting a role for TAMs in all aspects of tumor initiation, growth, and development. In conclusion, results from the present study demonstrate that microenvironmental factors contribute greatly to determine clinical and pathological features that differentiate low and high grade serous ovarian carcinomas. This understanding may increase possibilities and opportunities to improve disease control and design new therapeutic strategies.

Keywords: M1; M2; TAM; ovary; tumor-associated macrophages.

Figure 1. Histological features of LGSOCs and HGSOCs

Low-grade serous carcinoma of the ovary is characterized by relative uniformity of the cells and up to 12 mitoses per 10 high-power fields. High-grade serous carcinoma of the ovary is characterized by pleomorphism, marked nuclear atypia and > 12 mitoses per 10 high-power fields.

Figure 2. Densities of tumor associated macrophages in LGSOC and HGSOC tissue specimens

(A) Representative pictures for immunohistochemical staining of CD68+ macrophages in clinical samples of LGSOC and HGSOC. Magnification 20× and 40×. Scatter plot shows all data points and mean ± SEM for the entire set of patients (n = 25 and n = 55, for LGSOCs and HGSOCs respectively, ***p < 0.0001). Bar graphs depict data (mean ± SEM) following stratification per stage (Early stage, n = 10 and n = 8; Advanced Stage n = 15 and n = 47, for LGSOCs and HGSOCs respectively, ***p < 0.0001). (B) Representative pictures for immunohistochemical staining of CD163+ macrophages in clinical samples of LGSOC and HGSOC. Magnification 20× and 40×. Scatter plot shows all data points and mean ± SEM for the entire set of patients (see above for sample sizes, ***p < 0.0001). Bar graphs depict data (mean ± SEM) following stratification per stage (see above for sample sizes, ***p < 0.0001). (C) Bar charts showing the CD163/CD68 ratio (mean ± SEM) in the entire population (see above for sample sizes, *p = 0.02) and after stratification per stage (see above for sample sizes, *p = 0.049).

Figure 3. Tumor-associated angiogenesis in LGSOC and HGSOC tissue specimens

(A) Representative pictures for immunohistochemical staining of CD31 in clinical samples of LGSOC and HGSOC. Magnification 20×. (B) Scatter plot shows MVD (Microvessel Density, vessels/HPF) values and mean ± SEM for the entire set of patients (see legend to Figure 2 for sample sizes, ***p < 0.0001). Bar graphs depict data (mean ± SEM) following stratification per stage (see legend to Figure 2 for sample sizes, *p = 0.04, ***p < 0.0001). (C) The Spearman rank correlation showed a significant positive correlation between MVD and CD163+ macrophages density (cells/mm²) (n = 80, p < 0.0001).

Figure 4. Matrix metalloproteinases 9 (MMP-9) expression in LGSOC and HGSOC tissue specimens

(A) Representative pictures for immunohistochemical staining of MMP-9 in clinical samples of LGSOC and HGSOC. Magnification 20×. (B) Scatter plot shows MMP-9 IRS (Immunoreactive receptor score) values and mean ± SEM for the entire set of patients (see legend to Figure 2 for sample sizes, **p = 0.006). Bar graphs depict data (mean ± SEM) following stratification per stage (see legend to Figure 2 for sample sizes, *p = 0.04). (C) The Spearman rank correlation showed a significant positive correlation between MMP-9 IRS and CD163+ macrophages density (cells/mm²) (n = 80, *p = 0.04).

Figure 5. E-cadherin expression in LGSOC and HGSOC tissue specimens

(A) Representative images for immunohistochemical staining of E-cadherin in clinical samples of LGSOC and HGSOC. Magnification 20×. (B) Scatter plot shows E-cadherin IRS (Immunoreactive receptor score) values and mean ± SEM for the entire set of patients (see legend to Figure 2 for sample sizes). Bar graphs depict data (mean ± SEM) following stratification per stage (see legend to Figure 2 for sample sizes). (C) There was no correlation between expression of E-cadherin and CD163+ macrophages density (cells/mm2) (n = 80), as showed by the Spearman rank analysis.