类器官个体化医疗：伴随诊断与患者分层

以下为英文全文（English full text）

1. Introduction

The one-size-fits-all approach to cancer therapy has given way to precision medicine, in which treatment decisions are guided by the molecular characteristics of individual patient tumors. Patient-derived organoids (PDOs) represent a transformative technology in this paradigm, offering living avatars that capture the genetic, epigenetic, and phenotypic diversity of patient tumors in a clinically actionable timeframe. As companion diagnostics, PDOs enable functional testing of drug responses prior to patient administration, reducing unnecessary toxicity and improving treatment efficacy. This article examines the technical foundations, clinical validation, and implementation strategies for organoid-based personalized medicine, including companion diagnostic development and patient-specific treatment optimization protocols.

2. Patient-Derived Organoids as Living Avatars of Patient Tumors

2.1 Genetic and Phenotypic Fidelity

Organoids established from patient tumor tissues exhibit high histological and functional similarity to in vivo tumors. Verstegen et al. (2025) published a comprehensive review in Nature Medicine on clinical applications of human organoids, documenting that PDOs have been successfully established for colorectal, prostate, pancreatic, gastric, liver, biliary tract, breast, and neuroendocrine cancers. Tiriac et al. (2016) demonstrated that PDOs preserve the mutational profiles of the original tumor, with minimal genetic drift even after long-term culture. Gahesh et al. (2019) confirmed that PDOs retain tissue identity through transcriptomic and proteomic analyses, making them faithful surrogates for patient tumors.

The PDO establishment process typically takes 2–6 weeks from tissue collection, depending on tumor type and culture conditions. Success rates vary by cancer type: colorectal (80–90%), pancreatic (60–70%), ovarian (50–70%), and glioblastoma (40–60%). Organoids are established from fresh surgical specimens (preferred) or core needle biopsies (minimum 2–3 cores, 18-gauge) in specialized media containing niche factors (Wnt, R-spondin, EGF, Noggin) and extracellular matrix (Matrigel or BME).

2.2 Clinical Correlation of PDO Drug Responses

The clinical utility of PDOs depends on their ability to predict patient treatment outcomes. Ooft et al. (2019) demonstrated that PDO drug responses positively correlated with patient responses in colorectal cancer, with a concordance rate of 80%. Pasch et al. (2019) reported that PDOs from pancreatic cancer patients predicted chemotherapy response with 100% accuracy in a prospective cohort, with results available within 4 weeks of tissue collection. Vlachogiannis et al. (2018) showed that PDOs from metastatic gastrointestinal cancers predicted clinical response to chemotherapy and targeted agents with an overall accuracy of 88%.

These studies establish the analytical and clinical validity of PDO-based drug response testing, supporting their integration into clinical decision-making workflows. Weeber et al. (2015), Fujii et al. (2016), and Pauli et al. (2017) collectively demonstrated that PDOs from metastatic cancer sites effectively evaluate drug response, even when standard molecular diagnostics fail to identify actionable targets.

3. Companion Diagnostic Development Using Organoids

3.1 Regulatory Framework for Organoid-Based Diagnostics

A companion diagnostic (CDx) is an in vitro diagnostic device that provides information essential for the safe and effective use of a corresponding therapeutic product. For PDOs to be recognized as CDx modalities, they must meet stringent regulatory requirements for analytical validity, clinical validity, and clinical utility. Zhou et al. (2025) reviewed the regulatory framework for organoid companion diagnostics, emphasizing that prospective clinical trials demonstrating reproducible correlations between organoid responses and patient outcomes are essential for regulatory approval.

Key regulatory requirements include:

• Analytical validation: Sensitivity, specificity, precision, and accuracy of drug response measurements (target: CV < 20% for IC50 values across replicates)
• Clinical validation: Demonstration of correlation between PDO response and patient outcome in a defined clinical population (target: PPV > 80%, NPV > 90%)
• Clinical utility: Demonstration that PDO-guided treatment decisions improve patient outcomes compared to standard-of-care treatment selection
• Standardization: Standard operating procedures for tissue collection, organoid establishment, culture conditions, drug testing, and data interpretation

3.2 Clinical Trial Design for PDO Companion Diagnostics

Clinical trials evaluating PDO-based CDx follow a co-clinical trial design, where PDOs are established from enrolled patients and tested against the same treatment regimens administered to the patient. Park et al. (2020) conducted a prospective co-clinical trial with rectal cancer patients, demonstrating that radiation responses in PDOs positively correlated with patient responses. The trial enrolled 33 patients and achieved a correlation coefficient of r = 0.78 for PDO vs. patient radiation sensitivity.

For regulatory approval, a pivotal trial would require:

• Sample size: Minimum 100 patients per tumor type (power 80%, alpha 0.05, expected accuracy 80%)
• Primary endpoint: Concordance between PDO-predicted response and clinical response (RECIST criteria)
• Secondary endpoints: Time-to-treatment decision, progression-free survival, overall survival in PDO-guided vs. standard-of-care arms
• Stratification: By molecular subtype, prior treatment history, and tumor stage

4. PDO-Based Drug Response Profiling for Treatment Optimization

4.1 Standard-of-Care Testing

The most immediate application of PDO-based personalized medicine is testing standard-of-care (SOC) chemotherapy regimens to identify the optimal treatment for each patient. For colorectal cancer, PDOs are tested against FOLFOX (5-FU + oxaliplatin + leucovorin), FOLFIRI (5-FU + irinotecan + leucovorin), and CAPOX (capecitabine + oxaliplatin). For pancreatic cancer, gemcitabine + nab-paclitaxel and FOLFIRINOX are standard regimens. PDOs are exposed to clinically relevant concentrations (Cmax adjusted for protein binding) for 72 hours, and viability is measured by CellTiter-Glo 3D.

Response classification:

• Sensitive: IC50 < Cmax and >70% viability reduction at Cmax
• Resistant: IC50 > 2× Cmax or <30% viability reduction at Cmax
• Intermediate: Partial response between sensitive and resistant thresholds

4.2 Targeted Therapy and Immunotherapy Testing

Beyond SOC chemotherapy, PDOs enable testing of molecularly targeted therapies matched to patient-specific mutations. For example, PDOs with KRAS G12C mutations are tested against sotorasib and adagrasib; BRAF V600E-mutant PDOs are tested against vemurafenib and encorafenib; HER2-amplified PDOs are tested against trastuzumab and pertuzumab.

For immunotherapy, Esposito et al. (2024) developed a patient-derived immunity-organoid platform to model responses to immune checkpoint inhibitors (ICIs) in colorectal cancer. The platform identified REG4 as a new predictive biomarker of immunotherapy resistance in microsatellite stable (MSS) CRCs. Knocking out REG4 in resistant organoids restored immune sensitivity and triggered T-cell-mediated apoptosis, demonstrating the utility of PDO-immune co-cultures for immunotherapy prediction.

5. Organoid Biobanks for Population-Scale Precision Medicine

5.1 Living Biorepositories

Organoid biobanks are living repositories of PDOs that preserve the molecular diversity of patient cohorts for large-scale drug screening and biomarker discovery. van de Wetering et al. (2015) established one of the first large-scale colorectal cancer organoid biobanks, demonstrating that biobank-derived organoids maintained genetic diversity and drug response heterogeneity over years of culture. Calandrini et al. (2020) described integrated biobanks linking genomic, transcriptomic, and drug response data across multiple cancer types, enabling retrospective analysis of treatment outcomes.

As of October 2024, ClinicalTrials.gov listed 36 trials using organoids for personalized medicine across multiple cancer types (Verstegen et al., 2025). These biobanks are now being integrated with clinical databases (e.g., TCGA, MSK-IMPACT) to develop predictive algorithms that combine genomic profiling with functional drug testing.

5.2 Data Integration and Predictive Modeling

The integration of PDO drug response data with genomic and clinical information enables the development of predictive models for treatment selection. Machine learning algorithms (random forests, gradient boosting, neural networks) trained on PDO response datasets achieve >80% accuracy in predicting patient responses, outperforming genomic-only models in many cases. Huang L. et al. (2020) demonstrated that extracellular vesicle analysis from PDOs provides biomarker discovery capabilities beyond standard genomic profiling.

6. Technical Implementation in Clinical Workflows

6.1 Turnaround Time and Clinical Feasibility

The clinical utility of PDO-based personalized medicine depends on turnaround times compatible with treatment decision-making. For aggressive cancers (e.g., pancreatic, glioblastoma), treatment decisions must be made within 2–4 weeks of diagnosis. Optimized PDO establishment protocols using defined media (Wnt-3a, R-spondin-1, Noggin) and automated culture systems have reduced the time from tissue collection to drug response report to 3–4 weeks for high-success-rate tumor types.

For tumor types with lower establishment rates (e.g., glioblastoma, 40–60%), cryopreserved biobank organoids with matched molecular profiles can serve as a surrogate when fresh PDO establishment fails. This "biobank-first" approach ensures that drug response data are available within the clinical window even for challenging tumor types.

6.2 Quality Assurance and Clinical Reporting

PDO-based drug response reports for clinical use must meet quality standards comparable to molecular diagnostic reports. Each report includes:

• Patient and specimen identification
• Organoid establishment success and passage number
• Quality control results (STR profiling, mutation confirmation, IHC verification)
• Drug response data (IC50, AUC, DSS for each tested agent)
• Response classification (sensitive, intermediate, resistant)
• Clinical interpretation with confidence level
• Limitations and caveats

Reports are reviewed by a multidisciplinary tumor board (oncologist, pathologist, molecular biologist) before integration into the patient's treatment plan.

7. Challenges and Future Directions

7.1 Current Limitations

Despite significant progress, organoid-based personalized medicine faces several challenges. The lack of tumor microenvironment (TME) components—including immune cells, fibroblasts, and vascular networks—in standard PDO cultures may limit the prediction of immunotherapy and anti-angiogenic therapy responses. Wang et al. (2024) reviewed CRC organoids integrated with immune cells and stromal components, demonstrating enhanced physiological relevance for tumor-immune interaction studies. However, these co-culture systems are more complex and require standardization before clinical implementation.

The cost of PDO establishment and drug testing ($3,000–$10,000 per patient) remains a barrier to widespread adoption. Reimbursement pathways and health economic analyses are needed to demonstrate cost-effectiveness compared to empiric treatment selection.

7.2 Future Advances

Emerging technologies will address current limitations and expand the scope of organoid-based personalized medicine. Organoid-on-chip platforms that integrate immune cells and vascular networks will improve immunotherapy prediction. CRISPR-edited isogenic organoid lines will enable precise genotype-phenotype correlation studies. Automated, high-throughput PDO screening platforms will reduce costs and turnaround times. Integration with artificial intelligence for drug response prediction will enhance accuracy and identify novel treatment combinations.

8. Conclusion

Patient-derived organoids offer a powerful platform for personalized medicine and companion diagnostic development. By capturing the functional heterogeneity of patient tumors and enabling prospective drug response testing, PDOs bridge the gap between genomic profiling and clinical treatment selection. As standardization, automation, and regulatory frameworks mature, organoid-based personalized medicine will become an integral component of oncology care, improving treatment outcomes and reducing unnecessary toxicity for cancer patients worldwide.

返回技术文章列表