Table 1.

Mechanisms of resistance, therapeutic strategies, model applications, and model characteristics for each immunotherapy resistance model

Model category	Mechanisms of tumor immune resistance	Intervention strategies	Model applications	Model characteristics
In vitro-induced immunotherapy resistance models	Sustained IFN-γ signaling activation → upregulation of ISGs → T cell exhaustion[10]	SAP chemical inducer → reversal of sustained IFN-γ signaling → overcomes melanoma therapy resistance	Preliminary screening of immunotherapy resistance targets and high-throughput screening of immunotherapy candidate drugs	Advantages: Short cycle time, easy expansion, rapid validation of immune resistance mechanisms Limitations: Lack of TME components, and potential loss of inherent tumor heterogeneity during prolonged passaging
	Acidosis potentiates PD-L1 expression → IFN-γ + low pH → PD-L1 upregulation → immune evasion[14]	NaHCO₃ + Anti-PD-L1 → neutralizes tumor acidity → enhanced T cell Infiltration
	Cold tumor phenotype → low TMB + MDSCs/tregs infiltration → primary immunotherapy resistance[16]	Targeting MS4A4A + Anti-PD-1 → restoring therapeutic sensitivity in CT26/B16-F10 tumors
	Antigen loss and defective antigen presentation[24]	STING agonist-loaded nanoparticles + Anti-PD-1 → NK cell activation
	Tyro3/CDK9 overexpression mediates anti-PD-1 resistance in breast cancer[15]	Tyro3/CDK9 Inhibitors + ICIs → therapeutic sensitization
	Glucocorticoids impair immunotherapy efficacy[22]	Restrict dexamethasone use during ICB therapy in glioma patients with peritumoral edema
	Elevated levels of immunosuppressive molecules → immune exhaustion[28]	Toripalimab → Elevated CD8 + /CD4 + T cell ratio with increased tumor cell apoptosis → reversal of immunotherapy resistance in PDO models
In vivo immunotherapy resistance induction models	MHC Class I/II downregulation → reduced T cell infiltration → immunotherapy resistance in lung cancer[35]	Radiotherapy + NLRP3 Agonist + anti-PD-1 → upregulation of MHC Class I/II → reversal of therapy resistance	Simulation of clinical progressive immunological resistance and evaluation of combination therapies	Advantages: Preservation of intact tumor microenvironment Limitations: Model construction affected by multifactorial variation with inherent human mouse immune system inconsistency
	TAM Kinase (Tyro3/Axl/MerTK) activation → M2-like macrophage polarization → immunotherapy resistance in breast cancer[15]	TAM receptor inhibitors → reversal of M2 phenotypic polarization
	TGFβ / Notch pathway activation → tregs expansion + NK cell reduction[39]	Dual pathway inhibitor + anti-PD-1 → synergistic inhibition of colorectal cancer progression
	STK11 deficiency → MDSCs accumulation + PD-L1 downregulation → resistance to therapy in CT26 models[43]	Targeting the STK11 pathway → restoration of CD8⁺ T cell function
	Serpinf1 overexpression → elevated FFA levels → CD8⁺ T cell dysfunction[34]	Orlistat → inhibition of FFA synthesis → reversal of anti-PD-1 resistance
PDX immunotherapy resistance models	Effector immune cells with functional defects[50]	HDAC inhibitors/antiangiogenic agents + ICI → enhanced response to immunotherapy for ovarian cancer SAP chemical inducer → tumor growth inhibition	Personalized immunotherapy response prediction and elucidation of clinical resistance mechanisms	Advantages: Preservation of patient-specific genomic characteristics Limitations: High risk of graft-versus-host disease, prolonged experimental duration, elevated maintenance costs, and uncertainties in human–mouse HLA matching
Targeted genetic engineering models of immunotherapy resistance	IL-4I1 overexpression → depletion of essential amino acids and production of toxic metabolites → CD8⁺ T cell dysfunction[60]	Early IL-4I1 detection → prediction of anti-PD-1 response	Functional validation of specific genes and discovery of immunotherapy resistance biomarkers	Advantages: Precise construction of customized immune resistance models with study specificity Limitations: Off-target risks and inability to recapitulate polygenic co-occurrence-mediated therapy resistance
	N-MYC overexpression → upregulates PD-L1 → immunotherapy resistance in neuroblastoma[65]	N-MYC inhibitor + Anti-PD-1 therapy → overcoming immunotherapy resistance
	miR-20a-5p upregulation → inhibition of NPAT in CD8⁺ T cells → TNBC immunotherapy resistance[62]	miR-20a-5p as a predictive biomarker for TNBC
	FOXA1 silencing → promoting PD-L1 expression → nasopharyngeal cancer resistance[69]	FOXA1 overexpression + Atezolizumab → sensitization to immunotherapy
Host–microbiota interaction-mediated resistance models	Colonization of immunotherapy-resistant gut microbiota: gut microbiota of transplantation-resistant patients → inhibition of anti-PD-1 efficacy by an unfavorable gut microbiota for immunotherapy[75]	Baicalin enriches Akkermansia spp. and Clostridia_UCG-014 spp. → Increased production of TNFɑ and IFNγ by CD8⁺ T cells and diminished immunosuppression by Tregs[80]	Microbiota–immune crosstalk investigation	Advantages: Recapitulation of the microbiota–immune axis Limitations: Lack of standardization (unresolved dosage protocols and inconsistent transplantation times)
		Pectin → increased butyrate production → enhanced T cell function → increased sensitivity to anti-PD-1 therapy[81]