Page 5 - manuscript_ijb05590

1. Introduction

Contemporary oncology research has undergone transformative advancement

through increasingly sophisticated tumor modeling platforms. While conventional 2D

monolayer cultures facilitate high-throughput compound screening, they fundamentally

lack the multidimensional complexity of native tissue ecosystems—failing to

recapitulate critical physiological features such as oxygen/nutrient gradients, metabolic

crosstalk, and spatially organized cell-cell communication 1–3 (Figure 1). Three-

dimensional (3D) models address critical limitations by preserving key architectural

features such as extracellular matrix (ECM)-cell crosstalk and the niche architecture of
4
human malignancies . Nevertheless, standard 3D systems remain inadequate for
capturing dynamic tumor microenvironment (TME) remodeling or essential

mechanical stimuli (e.g., interstitial fluid pressure, vascular shear stress) that drive

5
tumor progression (Figure 1). Genetically engineered mouse models (GEMMs) and
patient-derived xenografts (PDXs) partially overcome these shortcomings. GEMMs

recapitulate tumor initiation and progression within native microenvironments,

faithfully modeling molecular and systemic interactions across tumorigenesis,

6
metastasis, and therapeutic response while preserving human cancer characteristics .
PDX models, established by engrafting patient-derived tumor fragments into

immunodeficient mice, retain TME architecture and replicate the pathological,

histological, and genomic profiles of original tumors while maintaining drug response

fidelity 7,8 . Nevertheless, both models face significant scalability, time, and cost

9
constraints, hindering clinical translation . Furthermore, these static systems
particularly fail to recapitulate hemodynamic parameters (blood/interstitial flow)

10
essential for studying metastasis, immune recruitment, and drug pharmacokinetics
(Figure 1), highlighting the unmet need for dynamically perfusable platforms.

Conventional tumor models, while valuable for foundational insights, are

increasingly superseded by microfluidic-integrated tumor-on-a-chip platforms that

overcome their limitations through interdisciplinary innovation (Figure 1):

(1) Organoid-on-a-chip: Organoids are generated by isolating normal or cancer stem

4