Investigating treatment effects in the microenvironment of renal cell cancer with venous tumor thrombus: Translational analysis of the NAXIVA trial.

Abstract

368 Background: 10-15% of patients with renal cell carcinoma (RCC) develop venous tumour thrombus (VTT). The presence of VTT makes curative surgery technically challenging, with high peri-operative mortality. VTT contains renal cancer cells that are genetically similar to the parent tumour, but less is known about the non-cancer cell components of the VTT microenvironment. The NAXIVA trial demonstrated that neoadjuvant axitinib can reduce the extent of VTT, with 7/20 patients showing an improvement in VTT level by Mayo classification. This provided the first level II evidence for the use of axitinib in this setting. Non-responders had lower microvessel density and exhausted T cell phenotypes in baseline biopsies. NAXIVA offers a unique opportunity to study the VTT microenvironment and the biological response to axitinib in humans to inform treatment combinations in future trials. Methods: NAXIVA (NCT03494816) was a single arm, multi-centre, UK based phase 2 study. 20 patients with resectable clear cell RCC were treated with up to 8 weeks of axitinib prior to surgery. Translational samples were taken at baseline and at time of surgery. Additional untreated VTT and primary tumour samples were obtained from the ARTIST study. Samples were analysed by immunofluorescence microscopy and RNA sequencing. Results: Untreated VTT has a complex microenvironment. There are established CD31+ microvessels surrounded by SMA+ pericytes and stromal cells. Extensive immune infiltrates are present, including CD8+ and CD4+ T cells, CD68+ macrophages and MPO+ neutrophils. Consistent with previous studies, we find high levels of viable tumour cells in VTT as marked by CA9 and Ki67. Following 8 weeks of axitinib treatment there was a significant reduction in microvessel density in both the primary tumour and VTT, consistent with the mechanism of action of axitinib. This was particularly seen in a CD34+ subset of vessels, which we hypothesise marks younger vessels that may be more sensitive to axitinib. There were no clear changes in T cell or myeloid cell densities. However, Gene Ontology analysis of RNA-seq data showed upregulation of myeloid inflammation signatures upon treatment, which suggests there may be functional changes in the myeloid cells. We are investigating the significance of these genes in more detail. Conclusions: The VTT microenvironment is highly organised and resembles the parent tumour. The complex nature of the VTT microenvironment provides rationale for future trials, potentially combining immunotherapy with axitinib as used in the metastatic setting. The potential increase in myeloid cell activation is significant, as previous studies have shown myeloid activation may alter the effects of immunotherapy. Understanding the interactions between the cells present, and how they change on treatment is therefore critical for the success of these therapies. Clinical trial information: NCT03494816.