| Sumario: | Simple Summary: Metronomic chemotherapy—a regimen based on continuous low doses of anticancer drugs—represents a valuable treatment option for patients with advanced gastrointestinal cancers who have already received multiple lines of therapy. However, identifying in advance which patients are most likely to benefit from this approach remains an unsolved challenge. In this study, we analyzed blood samples from patients enrolled in the COMET clinical trial to identify protein markers—specifically cytokines, small signaling molecules measurable in plasma—associated with treatment outcomes at baseline, before therapy begins. Using a systematic computational strategy, we identified three panels of three cytokines each, all showing a statistically significant ability to stratify patients by prognostic risk. As this was a single-arm trial without a control group, these findings cannot establish whether the identified cytokines are truly predictive of the benefits from metronomic chemotherapy specifically, or whether they reflect the underlying differences in prognosis. These results are therefore hypothesis-generating and require prospective external validation before any clinical translation. They represent a step toward precision medicine in metronomic chemotherapy and provide a rationale for future biomarker-stratified trials. Background/Objectives: Metronomic chemotherapy offers a well-tolerated option for heavily pretreated metastatic gastrointestinal cancer patients, but reliable prognostic biomarkers for patient selection are lacking. This study aimed to identify exploratory circulating cytokine signatures associated with outcomes in patients treated with metronomic chemotherapy. Methods: We analyzed plasma samples from 34 patients enrolled in the COMET trial (EudraCT 2007-000065-38), a phase II study of metronomic UFT, cyclophosphamide, and celecoxib. An 88-cytokine Luminex® (Luminex Corporation, Austin, TX, USA) panel was measured at baseline and at four treatment timepoints. Partial least squares discriminant analysis identified candidate biomarkers, followed by systematic combinatorial analysis using Manciu's method to construct 3-cytokine composite risk scores. Results: Twenty-one patients (61.8%) experienced progressive disease and 13 (38.2%) achieved stable disease. Six biomarkers showed significant discriminative power: IL-16, MCP-4, THBS-2, Eotaxin-1, PDGF-AB/BB, and TRAIL. Three 3-cytokine panels achieved statistically significant risk stratification (all p < 0.05), with hazard ratios for overall survival ranging from 2.59 to 6.24. For the representative IL-16 + MCP-4 + THBS-2 panel, high-risk patients showed a median PFS of 2.0 vs. 4.0 months (HR 3.24, p = 0.0046) and a median OS of 5.8 vs. 11.1 months (HR 4.19, p = 0.0010). Conclusions: This exploratory pharmacodynamic biomarker analysis identifies three 3-cytokine panels associated with prognostic risk stratification in metronomic chemotherapy for metastatic gastrointestinal cancer. As this single-arm trial cannot distinguish prognostic from predictive value, findings are hypothesis-generating. Prospective external validation is required before clinical translation, and exploration in combination with immune checkpoint inhibitors is warranted.
|
|---|