Contrary to the long-held theory of vessel normalization, TKIs in Renal Cell Carcinoma (RCC) work by ablating blood vessels. This induces massive hypoxia, which surprisingly correlates with a better initial therapeutic response, mirroring clinical observations of necrosis in responding tumors.
The traditional M1 (good) vs. M2 (bad) classification for macrophages is overly simplistic. Single-cell data reveals macrophages exist in fluid, shifting states rather than fixed categories, with cells co-expressing markers from both classical types. This complexity is crucial for understanding the tumor microenvironment.
The therapeutic effect of TKI-induced hypoxia is a double-edged sword. While it causes initial tumor necrosis and response, preclinical models suggest this same hypoxic environment can promote Epithelial-to-Mesenchymal Transition (EMT), a process that may lead to increased metastatic potential over the long term.
Current immunocompetent mouse models for RCC have a major limitation: they are naturally T-cell depleted, with T-cells comprising only 1-10% of immune cells versus 40-70% in human tumors. This makes them excellent for studying myeloid biology but suboptimal for understanding T-cell mediated responses to immunotherapy.
Combination therapies in RCC are not immunologically equivalent. IO-IO (e.g., Ipi/Nivo) primarily promotes a CD8 T-cell response, associated with durable cures. In contrast, IO-TKI combinations stimulate B-cell and T-follicular helper cell responses, representing a fundamentally different mechanism of action within the tumor microenvironment.
A specific macrophage population expressing SPP1 serves as a novel biomarker for true, tissue-level hypoxia. This allows researchers to distinguish drug-induced hypoxia from the 'pseudo-hypoxic' state caused by VHL deficiency, which has been a major challenge in studying RCC biology and treatment response.
The clinical success of re-challenging patients with a different TKI after progression is not just about hitting VEGF harder. TKIs are 'dirty' drugs that also target other kinases like MET and AXL, which are present on immune cells like tumor-associated macrophages (TAMs), providing a distinct immunomodulatory effect.
There is an inherent selection bias in RCC preclinical models. The tumors that successfully grow in mice, whether genetically engineered or patient-derived, tend to be the more aggressive, de-differentiated phenotypes. It is very difficult to model the common, low-grade, indolent clear cell tumors often seen in the clinic.