OncoGraph:一种利用蛋白标志物监测癌症复发和治疗反应的患者特异性血液检测

Background: ctDNA-based MRD assays are increasingly used for post-treatment monitoring but remain costly and technically demanding, limiting global accessibility, particularly in low- and middle-income countries (LMICs). Protein tumor markers (PTMs) are inexpensive and clinically effective in select cancers (e.g., PSA for prostate, AFP for liver), but most malignancies lack disease-specific markers, and existing markers have restricted applicability with substantial inter-patient variability, limiting individualized monitoring. We developed OncoGraph, a patient-specific approach that identifies abnormally elevated PTMs and tracks their longitudinal dynamics to monitor recurrence and treatment response. This approach offers an affordable, scalable solution suited for LMIC healthcare systems. Methods: Baseline blood samples from newly diagnosed cancer patients were analyzed by OncoSeek, a validated multi-cancer early detection assay quantifying a predefined panel of common PTMs (e.g., AFP, CEA). For individuals testing positive, PTMs exceeding internally validated thresholds were designated as patient-specific PTMs. These markers were longitudinally tracked after surgery or during therapy to monitor recurrence, response, or progression. Using OncoGraph, we evaluated postoperative recurrence in 24 liver cancer patients and treatment monitoring in 23 advanced lung cancer and 116 lymphoma patients receiving chemotherapy or targeted therapy. Results: In liver cancer (n=24), 13 patients had elevated baseline PTMs (all AFP). After surgery, OncoGraph-positive patients had a higher recurrence rate (4/5 vs. 1/8) and shorter median recurrence-free survival compared with OncoGraph-negative patients (385 days vs. not reached; HR 12.8; P < 0.01).In advanced lung cancer (n=23), 17 patients carried PTMs elevated at baseline (CEA and/or CYFRA21-1). After a median of 44.5 days after treatment initiation (range, 31-63), patients with significant PTM decrease had longer median progression-free survival (438 vs. 191.5 days; P < 0.05).In lymphoma (n=116), only eight patients had baseline-abnormal PTMs (all CA125). After two treatment cycles, all eight showed marked OncoGraph decrease, fully concordant with interim PET/CT responses (partial or complete response) assessed after 2-4 cycles. Conclusion: OncoGraph provides a patient-specific framework for PTM-based monitoring, addressing limitations of traditional PTMs through patient-specific marker selection. By selecting patient-specific informative PTMs and tracking their longitudinal dynamics, OncoGraph delivers early, clinically actionable signals of recurrence risk and treatment response across multiple cancer types. Its low cost, minimal technical requirements, and broad applicability make it a practical and scalable monitoring solution, particularly for LMIC healthcare settings.