Why metastatic NSCLC is lethal: Despite modern treatments including targeted therapy and immunotherapy, metastatic NSCLC (mNSCLC) remains the leading cause of cancer death worldwide. The challenge lies in the extreme molecular and cellular heterogeneity of these tumors - what works for one patient may fail entirely for another.
Scope of this review: This 62-page comprehensive review synthesizes evidence from across the cancer biology literature on three key topics: the histopathological and genetic landscape of mNSCLC, the clinical relevance of metastatic patterns, and the roles of tumor-associated cells in shaping the immunosuppressive tumor microenvironment.
Why the TME matters: Cancer cells do not exist in isolation. The tumor microenvironment (TME) - comprising immune cells, fibroblasts, blood vessels, and extracellular matrix - actively shapes tumor behavior, drug resistance, and metastatic potential. Understanding the TME is essential for developing effective immunotherapies.
Clinical relevance: The review emphasizes that advances in molecular profiling (including TCGA analyses), single-cell sequencing, and clinical trial data have now enabled a detailed understanding of how mNSCLC exploits its microenvironment for immune evasion - knowledge that directly guides therapy selection.
Adenocarcinoma (ADC): The most common subtype (40-60%), adenocarcinoma arises from type II alveolar cells. It is characterized by glandular formation, lepidic growth patterns, and mucin production. ADC is most commonly found peripherally in the lung and shows distinct patterns including acinar, papillary, micropapillary, and solid growth. TTF-1 and CK7 are key immunohistochemical markers.
Squamous cell carcinoma (SCC): The second most common subtype (25-30%), SCC arises from central airways and shows keratinization and intercellular bridges. Unlike ADC, SCC has fewer actionable mutations in receptor tyrosine kinases, and is marked by p40 and p63 IHC positivity. It is strongly associated with smoking history.
Large cell carcinoma (LCC): A diagnosis of exclusion, LCC lacks the defining features of ADC or SCC, representing about 10-15% of NSCLC. Poorly differentiated, it shows aggressive behavior and limited therapeutic options.
Classification importance: Histological subtype directly influences treatment decisions. ADC patients should undergo full molecular profiling for EGFR, ALK, ROS1, KRAS, and other alterations. SCC patients rarely carry EGFR mutations but may have FGFR1 amplification or DDR2 mutations. Accurate subtyping is the foundation of precision therapy.
EGFR mutations: The most common actionable alteration in ADC (30-40%), EGFR mutations occur primarily as exon 19 deletions (60%) or L858R point mutations (35%). Third-generation osimertinib is the preferred treatment, now extending to adjuvant therapy after resection. EGFR-mutant tumors have high rates of brain and adrenal metastasis.
KRAS mutations: Present in 20-30% of ADC, KRAS G12C is the most common single point mutation in NSCLC. Long considered undruggable, selective KRAS G12C inhibitors (sotorasib, adagrasib) are now approved. KRAS mutations are associated with smoking and often co-occur with STK11 mutations, which impair immunotherapy response.
ALK, ROS1, and RET fusions: Gene fusions involving ALK (3-7%), ROS1 (1-2%), and RET (1-2%) define distinct mNSCLC subgroups highly responsive to their respective tyrosine kinase inhibitors. These fusions are more common in never-smokers and ADC histology. Brain metastases are particularly prevalent in ALK-positive disease.
TP53 and tumor suppressors: TP53 is mutated in 52% of ADC and 79% of SCC - making it the most frequently altered gene in NSCLC. Unlike EGFR or ALK, TP53 loss is not currently directly targetable but influences prognosis and may modulate response to therapy. Co-mutations in KEAP1, NF1, and STK11 further shape therapeutic sensitivity.
The metastatic cascade: Metastasis requires tumor cells to invade local tissue, enter the bloodstream (intravasation), survive in circulation, exit blood vessels at distant sites (extravasation), and colonize new organs. Each step involves complex interactions between tumor cells and immune cells, extracellular matrix, and stromal components.
Epithelial-mesenchymal transition (EMT): EMT is a process where epithelial tumor cells lose their cell-cell adhesion and gain migratory mesenchymal properties - a critical early step in metastasis. TGF-beta, IL-6, and hypoxia all drive EMT in NSCLC. EMT also confers resistance to targeted therapy and chemotherapy.
Organ-specific metastasis (organotropism): Different mNSCLC molecular subtypes show preferences for specific organs. EGFR-mutant tumors metastasize to brain (up to 70%), bones, and adrenal glands. ALK-positive tumors show high brain and liver tropism. TP53 and SMARCA4 mutations associate with bone and CNS metastasis respectively, as shown by MSK-MET cohort analyses of 25,000 patients.
Tumor evolutionary timing: TRACERx study data revealed that certain mutations (MDM2 amplification, TP53) are clonal and present before metastatic divergence, while others (HIST1H3B amplification) are subclonal - present only in a minority of cells with metastatic advantage. This evolutionary perspective guides identification of which genetic events drive early vs. late metastatic spread.
Key immune cell players: The mNSCLC TME contains tumor-associated macrophages (TAMs), regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs), cancer-associated fibroblasts (CAFs), and cytotoxic T lymphocytes (CTLs). The balance between immune-activating and immune-suppressing cell types determines whether the tumor can evade immune elimination.
TAM polarization: Macrophages in the TME are predominantly polarized toward the M2 phenotype (alternatively activated), which promotes tumor growth, angiogenesis, and immune suppression rather than anti-tumor immunity. M2 TAMs secrete IL-10, TGF-beta, and VEGF - all of which suppress CTL activity.
Cancer stem cells (CSCs): A small population of tumor cells with stem-like properties - called cancer stem cells - are particularly resistant to chemotherapy and radiation, highly migratory, and can generate new tumor masses. CSCs express CD44, CD133, and ALDH markers and are enriched in metastatic sites.
Non-coding RNAs in immune evasion: Long non-coding RNAs (lncRNAs) and miRNAs regulate immune checkpoint expression, macrophage polarization, and T cell function within the TME. For example, certain lncRNAs upregulate PD-L1 expression on tumor cells, reducing T cell-mediated killing.
First-line pembrolizumab: For PD-L1-high mNSCLC (TPS 50% or above), pembrolizumab monotherapy is standard first-line treatment based on KEYNOTE-024, achieving 5-year OS rates of 31.9% vs 16.3% for chemotherapy - the most dramatic immunotherapy survival benefit in NSCLC.
Combination chemoimmunotherapy: For PD-L1-low or negative patients, combining checkpoint inhibitors with chemotherapy improved outcomes across subtypes. KEYNOTE-189 (pembrolizumab plus chemotherapy in non-squamous NSCLC) and KEYNOTE-407 (squamous NSCLC) demonstrated consistent OS benefits.
Dual checkpoint blockade: CheckMate-227 tested nivolumab plus ipilimumab (anti-PD-1 plus anti-CTLA-4). Benefit was observed in TMB-high patients, demonstrating that dual blockade can overcome resistance seen with single-agent PD-1 inhibition in certain molecular contexts.
Biomarker limitations: PD-L1 TPS and TMB are currently the best biomarkers for immunotherapy selection but are imperfect. Patients with low PD-L1 can respond and some high-PD-L1 patients do not. Next-generation biomarkers integrating TME composition, gene expression signatures, and radiomics are urgently needed.
Targeting the TME directly: Rather than blocking immune checkpoints alone, future strategies aim to reprogram the TME - converting immunosuppressive M2 macrophages to immune-activating M1 states, depleting MDSCs, and eliminating immunosuppressive Tregs. These approaches could benefit patients who fail checkpoint therapy.
Cancer stem cell targeting: Eliminating the CSC population that drives metastasis and resistance is a high-priority goal. Agents targeting CSC surface markers (anti-CD44, anti-CD133 antibodies) or CSC signaling pathways (Wnt, Notch, Hedgehog inhibitors) are in early clinical development.
Non-coding RNA therapeutics: lncRNA and miRNA modulators are emerging as potential therapeutics to disrupt immune evasion. Antisense oligonucleotides and RNA interference approaches can target specific non-coding RNAs that maintain immunosuppression in the mNSCLC TME.
Multi-target combination strategies: The future of mNSCLC treatment lies in rational combinations of targeted therapy (EGFR/ALK inhibitors), immunotherapy (checkpoint blockade), and TME-modifying agents - personalized based on each patient's molecular profile, TME composition, and metastatic pattern. Biomarker-driven clinical trial designs will be essential to test these complex combinations efficiently.