Vasculogenic mimicry in head and neck squamous cell carcinoma

Head and neck squamous cell carcinoma (HNSCC) is a common malignancy worldwide. HNSCC typically consists of tumors in the oral cavity, hypopharynx, oropharynx, nasopharynx, and larynx, of which oral squamous cell carcinoma (OSCC) is the most common subsite. The treatment options depend on the tumor site and stage, among others. The multimodality treatment generally includes surgery, radiotherapy with possible adjuvant chemotherapy, immunotherapy, and targeted therapy. Unfortunately, patients with HNSCC have a relatively poor 5-year survival rate of around 50%–60% in Nordic countries. Also, the incidence rates of HNSCC are increasing in the same region, including Finland. Therefore, there is a need to identify new prognostic factors and therapeutic targets to eventually improve the survival rates of HNSCC patients. Tumors require oxygen via the blood supply to grow and metastasize. Angiogenesis is a well-known phenomenon wherein new vessels develop from preexisting ones to support tumor progression. In 1999, a novel vascularization phenomenon called vasculogenic or vascular mimicry (VM) was described in uveal melanoma. In VM, aggressive tumor cells reorganize and form de novo fluidconducting tube-like structures to provide oxygen and nutrients for rapidly proliferating tumor cells. A positive VM status has been linked with poor prognosis in several solid tumors. Furthermore, this status is associated with adaptive resistance to cancer therapeutics. Therefore, the present doctoral dissertation aims to evaluate the prognostic role of VM in HNSCC; validate methodological approaches to study VM using different staining procedures; in vitro and in vivo models; and assess the effect of relevant anti-angiogenic drugs on VM in HNSCC. Previously, the immunohistochemically validated VM-positivity was linked to shorter survival in many solid tumors, but there was no consensus regarding HNSCC. Therefore, in Study I we conducted a systematic review and meta-analysis investigating the prognostic value of VM in HNSCC and esophageal squamous cell carcinoma (ESCC). Approximately one-third of the studied tumor samples were reported as VM-positive (periodic acid Schiff [PAS]-positive and endothelial cell marker-negative structures). We found a statistically significant correlation between shorter overall survival (OS) and VM positivity in HNSCC and ESCC. Furthermore, VM was associated with more metastasis in the locoregional lymph nodes and advanced stage. These results highlight that VM could represent a potential prognosticator and an interesting therapeutic target in HNSCC. Despite the promising clinical utility of VM, there is a lack of in vitro and in vivo models to identify VM. Furthermore, the traditional definition of VM has been enriched with the observation that aggressive tumor cells can transdifferentiate into endothelial cell-like phenotype with vascular properties. Thus, Study II aimed to address these limitations by conducting a comparative analysis using a variety of staining procedures and in vitro analysis. The study included clinical samples from oral tongue squamous cell carcinoma (OTSCC) patients treated surgically at Oulu University Hospital from 1990 to 2010. Interestingly, we found a mosaic pattern in tumor vasculature, wherein VM-forming tumor cells expressed tumor-specific and endothelial cell markers in patient samples and in vitro. Hence, intratumoral mosaicism might be another staining approach when identifying tumor vasculature or VM. Furthermore, we elucidated whether cell line origin or tumor-mimicking matrix impacts VM formation. Specifically, we used 13 primary and metastatic HNSCC cell lines and cultured them on top of murine-based Matrigel or human-derived Myogel, which revealed cell line-dependent variations in VM formation capability and morphology. Finally, we proposed zebrafish larvae as a potential model to study VM in vivo. Based on the recent literature, VM has been linked with resistance mechanisms behind anti-angiogenic treatment (AAT), which has shown limited success in managing OSCC. Hence, Study III aimed to explore the efficacy of selected anti-angiogenic drugs in targeting VM. The in vitro protocol was based on our findings in Study II; the treatment included sorafenib, sunitinib, and axitinib, all of which have been studied as monotherapy in clinical trials to treat HNSCC. Interestingly, while these drugs impaired the tube formation of endothelial cells, they generally showed a limited effect on tumor cell tubulogenesis. Sorafenib inhibited the tube formation of metastatic OSCC cell lines, which we previously reported to co-express vascular endothelial markers. We also demonstrate a consistently preferential anti-cancer effect of certain drugs over others on tumor cell proliferation, apoptosis, and cell migration. These data support the concept that VM might underlie drug resistance in OSCC, encouraging further studies using in vivo and clinical settings.