Estrogen receptor alpha and androgen receptor are commonly expressed in well-differentiated liposarcoma

Abstract

Background: Liposarcoma (LS) is the second-most common type of soft-tissue sarcoma. Despite advances in knowledge and treatment of this disease, there remains a need for more effective LS therapy. Steroid hormone receptors regulate metabolism in adipocytes. Estrogen receptor alpha (ER), progesterone receptor (PR), and androgen receptor (AR) have been implicated in the pathophysiology of other cancer types. We sought to comprehensively determine temporal expression patterns of these receptors in LS.

Methods: We analyzed 561 histologically subtyped LS specimens from 354 patients for expression of ER, PR, and AR by immunohistochemistry (IHC) using diagnostic-grade reagents and protocols. The fractions of positively stained tumor cells were scored within each specimen. IHC scores were compared across LS subtypes using the Kruskal-Wallis test, and subtypes were compared using Dunn's post-hoc test. Ages of patients with receptor-positive vs. -negative LS were compared by t-test. Genders and races were compared for hormone receptor positivity using Fisher's exact test and Chi-square analysis, respectively. Recurrence-free survival was compared between receptor-positive and negative patients by log-rank test. p< 0.05 was considered significant.

Results: ER and AR were frequently expressed in LS, while few tumors expressed PR. Most of the ER + and AR + samples were of the well-differentiated LS subtype. A smaller fraction of de-differentiated LS expressed ER or AR, but expression was common within well-differentiated regions of tumors histologically classified as de-differentiated LS. In LS specimens from patients who underwent multiple surgeries over time, receptor expression frequently changed over time, which may be attributable in part to intratumor heterogeneity, varying degrees of de-differentiation, and biopsy bias. ER and AR were frequently co-expressed. Receptor status was not significantly associated with gender or race, but AR and PR expression were associated with earlier age at diagnosis. Receptor expression was not associated with altered recurrence-free survival.

Conclusions: ER and AR are commonly expressed in LS, particularly in well-differentiated tumors. These data warrant further functional study to determine receptor function in LS, and the potential efficacy of anti-hormone therapies for the treatment of patients with LS.

Figure 1

Steroid hormone receptor expression in LS. Sections of LS were stained using antibodies against A) ER, B) AR, or C) PR. Shown are two representative microscopic fields that were scored as receptor-positive or -negative. Scale bar in (A) is 50 μm.

Figure 2

Steroid hormone receptor expression is most common in WDLS. A) LS specimens were scored for% positively-stained nuclei for ER, PR, and AR, then classified by histological subtype and binned according to score as indicated. B) ER and AR scores were compared between specimen subtypes. Colored bars indicate mean + SD. *p< 0.0001 by Dunn’s post-hoc test. C) Venn diagrams illustrating the number of WDLS and DDLS specimens with co-expression of hormone receptors using a threshold of 1% or 10% positively-stained nuclei.

Figure 3

Changes in tumor hormone receptor status over time. Fifty-three patients with WDLS and/or DDLS for whom specimens were obtained from ≥2 surgeries ≥6 months apart were evaluated. In cases where multiple specimens were obtained from the same tumor at the same time point, the% positively-stained nuclei were averaged across specimens. The% positively-stained nuclei and histological classification of each tumor are noted by color intensity and hash marks as indicated in legend.

Figure 4

Intratumor heterogeneity in degree of differentiation contributes to heterogeneity in hormone receptor expression. A) Frequencies of hormone receptor expression among WDLS specimens obtained from tumors classified as DDLS. B) Tumors for which 2 specimens were obtained from different regions at the same time point were evaluated for concordance in hormone receptor expression using a threshold of 10% for positivity. Scatterplots show% positive nuclei for ER and AR in tumors for which one DDLS specimen and one WDLS specimen were available. C) Representative liposarcoma specimens showing regions of receptor-positivity and -negativity. D) Hormone receptor status for patients with WDLS/DDLS as determined using any LS specimen obtained from any surgery.