Carbonic Anhydrase IX (CAIX) Expressing Hypoxic Micro-environment Hampers CD8+ Immune Cell Infiltrate in Breast Carcinoma

Abstract

Background: Hypoxia and necrosis are common features of invasive cancer. The dynamic upregulation of carbonic anhydrase IX (CAIX), triggered by hypoxia-inducible factor 1 (HIF-1) is 1 of the mechanisms supporting cellular adaptation to hypoxia in solid tumors, including breast carcinoma. CAIX activity results in extracellular acidosis and in a profound reorganization of the tumor micro-environment, influencing biological behavior and prognosis. The main focus of our study was to evaluate the mass and distribution of the immune infiltrate, more specifically of CD8+ effector T-cells, in relation with tumoral CAIX expression.

Materials and methods: Formalin-fixed and paraffin-embedded breast carcinoma sections were analyzed following double immunohistochemical staining for CAIX and CD8. Scanned digital slides were evaluated for both labelings, and CD8-related signal was determined within and outside CAIX-positive tumor areas using the HistoQuant (3DHistech) image analysis software. Statistical analysis was performed using GraphPad Prism software.

Results: Of the 34 breast carcinomas, 18 tested partially positive for CAIX. The remaining 16 cases were used as the CAIX-negative control group. Necrotic foci were generally associated with CAIX overexpression, and tumors exhibiting signs of necrosis had a significantly higher rate of relative CAIX expression compared with samples without necrosis (11.47±5.505 vs. without necrosis 3.765±3.5 P-value=0.0216). On the other hand, no statistically significant difference was found when comparing relative CD8+ lymphocyte counts in cases with necrosis as opposed to those where necrosis was absent (134.7±55.7 vs. 97.70±57.25; P value=0.1579). No difference in gross CD8+ T-lymphocyte infiltrate could be measured between CAIX positive and negative samples (98.48±37.32 vs. 95.99±50 P value=0.5928). However, in CAIX-expressing tumors a statistical correlation between the CD8+ T-lymphocyte infiltrate and the extent of CAIX-positive areas was observed. Within the same tumor, CD8+ T-lymphocyte counts showed a significant difference betweeen CAIX+ and CAIX- areas (13.06±9.4 vs. 135.6±62.2 P value <0.0001).

Conclusion: Our measurements demonstrate for the first time that tumor areas with CAIX expression potentially hamper CD8+ T-lymphocyte infiltration in breast carcinoma. The hypoxia-driven adaptive micro-environment likely interferes with the specific response to biological and immune therapies requiring intact effector T-cell response.