Spatially-resolved CD8⁺ and CD34⁺ computational immunohistochemistry indicators improve post-nephrectomy risk...

Clear cell renal cell carcinoma (ccRCC) is a heterogeneous tumor that may progress after nephrectomy as local or distant disease. The ccRCC tumor microenvironment (TME) hinges on two complementary pillars of immune response and angiogenesis. We aimed to assess if spatial CD8 and CD34 profiles at the tumor-stroma interface predict progression-free survival (PFS).

We retrospectively analyzed 214 ccRCC patients treated at Vilnius University Hospital Santaros Klinikos (2009 - 2019). Immunohistochemistry for CD8 and CD34 was performed on surgical tumor excision samples, and digital image analysis was performed to quantify cell densities and vessel areas relative to their spatial patterns within the tumor-stroma interface. The impact on PFS of CD8⁺ immunogradient and CD34⁺ area fraction was tested.

Two prognostic models were developed: a Baseline model using variables available after surgery and a Disease-Course model additionally incorporating follow-up information. In the Baseline model, higher CD8_m_CM independently predicted shorter PFS (HR 3.24, p = 0.001), together with tumor size >4.95 cm and coagulative tumor necrosis, while female sex predicted longer PFS. In the Disease-Course model, local recurrence was the strongest adverse predictor (HR 17.69, p < 0.001), while both spatial biomarkers remained independently associated with PFS: higher CD8_m_CM predicted shorter PFS (HR 5.14, p = 0.001), whereas higher VAF_m_CM predicted longer PFS (HR 0.32, p = 0.014).

Spatial patterns of immune and vascular components at the tumor-stroma interface independently predict PFS in patients with ccRCC following nephrectomy and may improve current risk stratification schemes in both the immediate postoperative setting and during follow-up.