The EGFR pathway drives most NSCLC cases. Non-small cell lung cancer accounts for roughly 85% of all lung cancer diagnoses, making it the leading cause of cancer-related death worldwide. Mutations in the epidermal growth factor receptor (EGFR) gene - particularly exon 19 deletions and the L858R point mutation - fuel abnormal cell growth and have transformed treatment, as targeted drugs called tyrosine kinase inhibitors (TKIs) can selectively block this signal.
Resistance inevitably develops. EGFR-TKIs such as gefitinib, erlotinib, and osimertinib initially produce impressive responses, but virtually all patients eventually develop acquired resistance. This resistance arises through multiple overlapping mechanisms: secondary EGFR mutations like T790M and C797S restore drug-insensitive signaling, bypass pathways such as MET and AXL take over, and tumor cells can transform into a more invasive state through epithelial-mesenchymal transition (EMT).
Single-target strategies have a fundamental limitation. Because resistance is multifaceted rather than monolithic, targeting one mechanism at a time often only triggers the emergence of the next. Next-generation TKIs and combination regimens have advanced the field but still face narrow activity spectra, the risk of selecting for new resistant clones, and compounded side effects. This creates a clear need for agents that can simultaneously address multiple resistance pathways.
Plant-derived molecules offer built-in polypharmacology. Unlike synthetic drugs engineered for a single target, natural compounds evolved to interact with multiple biological systems. This inherent ability to modulate several pathways at once makes them attractive candidates for overcoming the complex network of EGFR TKI resistance. This review systematically evaluated more than 33 representative compounds spanning four major chemical classes: alkaloids, terpenoids, flavonoids, and polyphenols.
The review focused on mechanisms relevant to resistance. A structured literature search covering publications from 2010 to 2025 was conducted across PubMed, Google Scholar, and Web of Science. Studies were selected based on whether they examined defined natural compounds acting on EGFR signaling or TKI resistance in NSCLC specifically, and were critically appraised for experimental rigor and translational relevance.
The central aim is rational combination therapy. Rather than cataloging natural compounds as standalone alternatives to TKIs, this review frames them as sensitizing agents - backbone partners in next-generation combinations designed to prevent or reverse resistance. The strategic logic is to use their multi-target capability to suppress the core EGFR axis alongside the bypass pathways and microenvironmental factors that sustain tumor survival when TKIs are used alone.
EGFR is a receptor that relays growth signals into the cell. The protein sits on the cell surface with a domain that binds growth factors outside and a kinase domain inside. When activated, it triggers two major downstream cascades: the RAS/RAF/MEK/ERK pathway and the PI3K/Akt/mTOR pathway, both of which promote cell survival and division. In NSCLC, mutations make EGFR constitutively active even without growth factor binding, locking cells in a growth state.
Normal cells tightly regulate EGFR activity. Under healthy conditions, activated EGFR is rapidly pulled off the cell surface by endocytosis, tagged for destruction by ubiquitin ligases, and degraded in lysosomes - reducing its half-life to under an hour. Protein phosphatases also directly deactivate the receptor, and negative feedback loops prevent sustained signaling. This multi-layered control prevents runaway growth.
Cancer cells disrupt these control mechanisms. In NSCLC, mutant EGFR is poorly internalized and inadequately ubiquitinated, so it persists on the cell membrane and keeps signaling. The tumor microenvironment can suppress the phosphatases that would normally restrain it. Chronic TKI exposure further dysregulates feedback loops, allowing cells to adapt and survive through alternative signaling nodes, explaining why resistance eventually develops even to highly effective drugs like osimertinib.
Berberine and piperlongumine show preclinical synergy with TKIs. Berberine, a compound found in several medicinal plants, inhibits EGFR phosphorylation and downstream Akt/mTOR signaling while also countering MET-amplification-driven resistance, showing synergy with EGFR-TKIs in cell and animal models. Piperlongumine works through a dual mechanism - directly inhibiting both wild-type and mutant EGFR while also promoting degradation of the anti-apoptotic protein Mcl-1, and demonstrating activity in osimertinib-resistant tumor models.
Other alkaloids use distinct resistance-targeting strategies. Matrine modulates the tumor microenvironment by suppressing IL-6 and the JAK1/STAT3 pathway, synergizing with the TKI afatinib. Oxymatrine directly suppresses phosphorylation of multiple EGFR variants. Sanguinarine employs an unusual redox approach, generating reactive oxygen species that specifically degrade the T790M resistance mutant. Griffithazanone A targets PIM1 kinase upstream of the resistance network, enhancing apoptosis and synergizing with both gefitinib and osimertinib.
Clinical translation faces serious pharmacokinetic hurdles. Despite mechanistic promise, berberine suffers from poor oral bioavailability and rapid elimination. Piperlongumine's human pharmacokinetics are largely uncharacterized. Sanguinarine's ROS-generating mechanism raises safety concerns about off-target oxidative damage to normal tissues that have not been rigorously evaluated. These gaps represent fundamental barriers that must be addressed before clinical development becomes feasible.
Celastrol induces EGFR degradation but carries toxicity concerns. This triterpenoid suppresses EGFR by promoting its degradation, blocking downstream PI3K/Akt and RAS/MAPK signaling, and shows synergistic effects with gefitinib and osimertinib in preclinical models. Advanced nano-drug systems that co-deliver celastrol and gefitinib have been developed to overcome its well-documented narrow therapeutic index and associated risks of hepatotoxicity and cardiotoxicity, though these formulations remain at an early experimental stage.
Ginsenoside Rg3 has both preclinical and preliminary clinical data. A retrospective clinical study reported that combining EGFR-TKIs with Rg3 significantly prolonged progression-free survival and improved objective response rates. Mechanistically, Rg3 enhances TKI sensitivity by inhibiting protective autophagy, suppresses EGFR-mediated proliferation, modulates the immune microenvironment by reducing PD-L1 glycosylation, and targets the HIF-1alpha/VEGF/EGFR axis in combination nano-formulations. Its major limitation is exceptionally poor oral bioavailability, requiring pharmaceutical delivery solutions before the clinical signal can be confirmed prospectively.
Beta-elemene, ursolic acid, and betulinic acid round out the terpenoid class. Beta-elemene uses multiple mechanisms including EZH2 downregulation, AMPK/MAPK activation, and lncRNA H19 modulation to overcome TKI resistance, though its established clinical form is an intravenous emulsion reflecting negligible oral bioavailability. Ursolic acid directly binds EGFR and simultaneously reduces PD-L1 expression. Betulinic acid directly blocks the ATP-binding site of EGFR, synergizing with TKIs - but like other triterpenoids, both suffer from extreme aqueous insolubility that necessitates nano-formulation strategies still in early development.
Quercetin exemplifies the classic natural product paradox. This widely studied flavonoid directly inhibits EGFR phosphorylation, blocks downstream ERK/MEK and Akt pathways, induces autophagy-dependent cell death, and simultaneously targets bypass receptors including c-Met, Her-2, AXL, and IGF1R. In resistant cell lines it shows synergistic cell killing with erlotinib. However, its oral bioavailability is negligible, and the concentrations effective in culture are pharmacologically unattainable in humans - making advanced nano-formulations not just improvements but essential prerequisites.
Luteolin and apigenin demonstrate strong multi-pathway mechanisms. Luteolin directly binds and inhibits EGFR while also blocking downstream ERK/MEK and Akt/mTOR pathways and modulating autophagy, showing synergy with erlotinib in resistant models. Apigenin inhibits EGFR phosphorylation and disrupts metabolic adaptations in T790M-mutant cells. Both compounds share the class-defining limitation of extremely poor oral bioavailability due to rapid flavonoid metabolism, rendering their impressive in vitro profiles largely irrelevant without delivery solutions.
Dihydromyricetin and hydroxygenkwanin target EGFR protein stability. Rather than simply blocking EGFR activity, these compounds promote EGFR ubiquitination and degradation - an approach that should in principle circumvent many resistance mutations by reducing the receptor itself. Hydroxygenkwanin concurrently suppresses downstream STAT3, Akt, and ERK pathways. While the mechanism is pharmacologically attractive, both compounds lack comprehensive human pharmacokinetic data, and their promotion of ubiquitin-mediated degradation raises concerns about unintended disruption of essential cellular proteins.
Curcumin has the broadest evidence base but the worst pharmacokinetics. Extensively studied curcumin inhibits EGFR phosphorylation and multiple bypass receptors including c-Met, Her-2, and AXL, showing synergistic apoptosis induction with erlotinib. However, its oral bioavailability is essentially zero due to rapid metabolism and poor stability. The translational focus has therefore shifted toward engineered analogs and hybrids (such as WZ35 and various CP compounds) that retain the multi-target pharmacophore while being drug-like, and toward antibody-conjugated nano-delivery systems.
Resveratrol and EGCG target EGFR through distinct mechanisms. Resveratrol inhibits EGFR phosphorylation and downstream pathways while also inhibiting drug efflux transporters that contribute to resistance - a particularly relevant mechanism for combination therapy. EGCG (from green tea) directly blocks the EGFR ATP-binding site, leading to receptor internalization and degradation, and suppresses downstream MAPK, PI3K/Akt, and STAT3 pathways. Both are constrained by poor bioavailability; resveratrol analogs like TMS and DMU-212, and EGCG conjugates, have been developed as potential successors.
Gallic acid, ellagic acid, shikonin, and silibinin complete the landscape. Gallic acid promotes EGFR degradation via the proteasome while modulating immune pathways including PD-L1. Ellagic acid shows notable selectivity - anti-proliferative activity against TKI-resistant cells while sparing sensitive ones. Shikonin generates reactive oxygen species that degrade EGFR and synergizes with TKIs, though its potent ROS induction creates toxicity concerns. Silibinin reverses EMT by modulating microRNAs and inhibiting EGFR downstream pathways, with an improved water-soluble formulation (silibinin-meglumine) already developed.
Poor pharmacokinetics are the principal barrier, not lack of efficacy. The central challenge is not that natural compounds fail to work in the laboratory, but that the effective concentrations demonstrated in cell culture are pharmacologically unattainable in human plasma using the native compounds. Most compounds - particularly flavonoids, polyphenols, and many terpenoids - suffer from extremely low aqueous solubility, rapid phase II metabolism, and swift systemic clearance, rendering much reported in vitro synergy clinically irrelevant without deliberate pharmaceutical intervention.
Safety assumptions are inadequately tested for oncology use. Claims of favorable safety profiles are typically based on limited acute toxicity assays or historical dietary use, not rigorous oncology-focused studies. Critical gaps include chronic toxicity, organ-specific liabilities (hepatotoxicity with high-dose EGCG or celastrol, for example), mechanism-based risks from ROS-promoting agents, disruption of essential protein degradation pathways, and drug-drug interactions with co-administered TKIs through CYP450 enzyme modulation.
Two engineering strategies can bridge the translational chasm. First, nanotechnology-based platforms - particularly ligand-targeted nanoparticles and exosomes - can enhance solubility, protect compounds from metabolism, enable tumor-selective delivery, and allow controlled co-delivery with TKIs. Second, medicinal chemistry-driven optimization can create synthetic analogs or prodrugs that retain the core multi-target pharmacophore while significantly improving chemical stability, potency, and pharmacokinetic properties. Both approaches are exemplified by ongoing work with celastrol, Rg3, curcumin, and resveratrol derivatives.
Future research must shift from description to problem-solving. Rather than generating more mechanism-finding studies, the field needs rigorous translational evidence using pharmacologically relevant models - patient-derived organoids and xenografts treated with the intended clinical formulations, not just high-dose cell culture experiments. Early-phase clinical trials should be designed primarily to establish pharmacokinetic profiles and maximum tolerated doses of novel formulations in combination with standard TKIs, not just to look for tumor responses.
AI and computational tools can accelerate lead identification. Artificial intelligence can speed up the identification of promising natural product-inspired leads and predict their pharmacokinetic and toxicological properties (ADMET) before expensive experimental testing. Network pharmacology and molecular docking can guide rational analog design, as already demonstrated for EGCG and quercetin derivatives. This computational layer can help filter the enormous chemical diversity of natural products toward the most developable candidates.
Biomarker-driven patient selection will be essential. Not all patients will benefit equally from any given natural compound combination. Developing mechanistic biomarkers - such as specific EGFR mutation profiles, expression levels of bypass proteins like MET or AXL, or immune microenvironment characteristics - is vital for enriching clinical trials with patients most likely to respond. Ginsenoside Rg3, which has preliminary clinical data, represents a compelling candidate for a prospective biomarker-driven phase I/II trial assessing combination safety and efficacy.
Natural compounds offer a unique strategic angle against resistance. Their inherent polypharmacology allows simultaneous modulation of the core EGFR axis and multiple resistance-associated bypass pathways - a theoretical advantage over sequential single-target therapies that cannot be easily replicated with conventional synthetic drugs. This review catalogues compelling preclinical evidence across more than 33 compounds representing four major chemical classes, establishing a strong rationale for continued development.
The future lies in engineered successors, not crude extracts. The direct clinical application of natural compounds as dietary supplements or herbal preparations is unlikely to achieve the consistent pharmacological exposures needed for meaningful anti-cancer activity. Instead, the path forward requires synthetic analogs and nano-formulations that preserve the multi-target logic of the parent molecules while conferring drug-like properties including reliable oral bioavailability, stability, and a manageable safety profile.
A convergence of disciplines is required. Realizing the therapeutic potential of natural compounds against EGFR TKI-resistant NSCLC demands integration of natural product biology, medicinal chemistry, nanotechnology, computational science, and rigorous clinical oncology. Success will be measured not by the number of mechanisms identified in the laboratory, but by the development of validated, next-generation combination regimens that deliver meaningful clinical benefit to patients whose tumors have stopped responding to current targeted therapies.