2. Academic Validation
  3. Circulating cytokine profiling and clustering identify biomarker predicting efficacy of ICI in combination with chemotherapy

Circulating cytokine profiling and clustering identify biomarker predicting efficacy of ICI in combination with chemotherapy

  • Cancer Lett. 2025 Oct 28:631:217918. doi: 10.1016/j.canlet.2025.217918.
Xiaotian Xu 1,  Yiling Li 1,  Shiran Sun 2,  Xianlong Lin 3,  Wenfeng Zhang 1,  Yue Wu 1,  Baojun Wei 1,  Danfei Xu 1,  Cuiling Zheng 1,  Hezhi Fang 4,  Wei Cui 5
Affiliations

Affiliations

  • 1 Department of Clinical Laboratory, State Key Laboratory of Molecular Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China.
  • 2 Department of Radiation Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China.
  • 3 Key Laboratory of Laboratory Medicine, Ministry of Education, Zhejiang Provincial Key Laboratory of Medical Genetics, School of Laboratory Medicine and Life Sciences, Wenzhou Medical University, Wenzhou, 325035, Zhejiang, China.
  • 4 Department of Clinical Laboratory, State Key Laboratory of Molecular Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China. Electronic address: [email protected].
  • 5 Department of Clinical Laboratory, State Key Laboratory of Molecular Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China. Electronic address: [email protected].
PMID: 40701131 DOI: 10.1016/j.canlet.2025.217918
Abstract

The combination of chemotherapy can enhance the efficacy of immune checkpoint inhibitors (ICIs), but requires precise patient stratification and biomarker screening. Cytokines influence immunotherapy outcomes, and multiplex cytokine profiling aids in identifying predictive biomarkers for ICIs. We analyzed 1331 plasma samples (1025 untreated pan-cancer patients and 306 healthy controls), including 238 receiving ICIs plus chemotherapy. Cytokine clusters were identified via non-negative matrix factorization. Cluster effected on early response and progression-free survival (PFS) were evaluated, and a Cytokine-based ICI Survival Index (CISI) was developed. The effect of specific cytokines on anti-programmed death 1 (PD1) treatment was verified in vivo. Thus, three inflammatory clusters were identified: Cluster 1 (high IFN-γ/IL-8/IL-1β, proinflammatory), Cluster 2 (high IL-6), and Cluster 3 (high IL-5/IL-17, Th2 activation). Cluster 3 showed superior PFS (HR = 2.44/3.84, p = 0.00011) and response rates (85.42 % vs. 54.33 %/61.90 %, p = 0.00075) versus Clusters 1&2. High IFN-γ/IL-8 predicted poorer outcomes. The CISI model, incorporating cytokine clusters and clinical variables (treatment, IL-10, monocyte-to-lymphocyte ratio, and M stage), outperformed conventional biomarkers programmed death-ligand 1 (PD-L1) and IL-8 in predictive efficiency [Concordance indexes (C-indexes) = 0.75 vs. 0.55 and 0.56]. In vivo studies confirmed the effects on anti-PD1 efficacy by characteristic cytokines in clusters. In conclusion, our cytokine clustering based on multi-cytokine profiles and CISI model predicted prognosis and immunotherapeutic response in tumor patients, providing new insights into personalized Cancer therapy strategies.

Keywords

Cytokine; Immunotherapy; Machine learning; Pan-cancer; Prognosis.

Figures
Products