Advertisement

Tumor escape in the microenvironment of penile carcinoma

Login to Access Video or Poster Abstract: MP44-18
Sources of Funding: none.

Introduction

In the complex interplay between cancer and the hosts immune system, the tumor is attacked by the natural anti-tumor response of cytotoxic T-cells (CTL). However, the tumor has several escape mechanisms: 1) regulatory T-cells (Tregs) inhibit CTLs, 2) aberrant HLA expression by the tumor cells misleads CTLs and 3) Programmed Death Ligand 1 (PD-L1) on tumor cells or on tumor infiltrating macrophages (TIM) deactivate CTLs. Clinically, HPV-positive penile cancers have a favorable outcome, presumably through micro-environmental factors. This study aims to gain insight in immunological factors, their biological interactions and their prognostic value for lymph node metastases and disease specific survival (DSS).

Methods

We revised histological sections of 213 penile cancer patients surgically treated between 2000 and 2009. HPV-status, different levels of HLA-expression and PD-L1-expression on tumor, stroma and TIM had previously been identified by our study group. Sections were stained for macrophage-marker CD163, CTL-marker CD8, and Treg-marker FoxP3. Macrophages were scored binary as high or low numbers present in the tumor cell fields. For T-cell balance, the CTL/Treg ratio was used for stroma and tumor. These parameters were included in multivariable regression models testing the prognostic value for lymph node involvement (LN+ or LN-) and DSS. To account for influences between the included parameters, interaction terms were also included.

Results

Scoring T-cell ratios (CTL/Treg), PD-L1 and CD163 was possible in 175, 200 and 208 samples respectively. HPV typing was known in all 213 tumors, HLA-expression was determined in 168 tumors. Multivariable analyses showed three independent prognostic factors for both lymph node status and DSS: 1) PD-L1+ TIM (unfavorable), PD-L1 margin expression (favorable), and a high intra-tumoral CTL/Treg ratio in the presence of PD-L1+ TIM (favorable). This indicates that PD-L1+ TIM and PD-L1 margin expression were independently predictive in absence of HPV, the interaction between CTL/Treg ratio and PD-L1+ TIM was predictive irrespectively of HPV-status.

Conclusions

A PD-L1 expression pattern predominantly at the tumor-stroma margin predicts good prognosis, while the negative predictive value of PD-L1+ TIM appear to be compensated by a high CTL/Treg ratio. All independent prognostic factors are PD-L1 related parameters. These results strengthen the rationale for anti-PD-1/PD-L1 immunotherapy in penile carcinoma.

Funding

none.

Authors
Sarah Ottenhof
Rosa Djajadiningrat
Helene Thygesen
Jeroen de Jong
Simon Horenblas
Ekaterina Jordanova
back to top