As reported by Global Cancer Burden statistics in 2020, colorectal cancer (CRC) ranked third amongst prevalent cancers (10.0 %) and second among causes of cancer-related deaths globally (9.4 %) [1]. Currently, CRC patients can be effectively treated through surgical resection, chemotherapy, radiotherapy, and targeted therapy, but clinical outcomes of most CRC patients remain poor [2], [3], [4]. Immune checkpoint blockade (ICB) therapy is a giant leap forward in cancer treatment, but less than 10 % of high microsatellite instability/deficient mismatch repair type CRC patients show a significant response to immune checkpoint inhibitors (ICI), and the efficacy of most microsatellite stable/proficient mismatch repair patients is poor [3], [5]. Colonoscopy is still the gold standard for diagnosing colon cancer, but the symptoms of early-stage CRC are often not obvious. Therefore, exploring non-invasive detectable biomarkers can help improve early diagnosis, prognostic risk prediction, and evaluation of the effectiveness of immune therapy.

Natural killer (NK) cells, as representative immune effector cells, play a potential role in cancer immunotherapy [6]. The study of Walle et al. [7] demonstrated that NK cells can exert anti-tumor effects by coordinating with CXCL8. Choi et al. [8] utilized the combination of polymers with NK cells for immunotherapy purposes. Currently, NK cell-based immunotherapy against CRC is an advanced immune treatment strategy, mainly including monoclonal antibody therapy that enhances NK cell cytotoxicity, cytokine therapy that enhances NK cell-mediated CRC killing, and adoptive therapy of NK cells [9]. In addition to their therapeutic effects, NK cells have prognostic value for CRC patients. Shembrey et al. [10] found that NK cell-related genes can predict recurrence in patients with colon cancer. In the study by Ishwar et al. [11], DNA methylation patterns of NK cells were identified as diagnostic markers for CRC. Therefore, NK cells have important significance in predicting the prognostic survival status of CRC patients.

However, there has been no study on the subtypes of NK cell-associated genes concerning the prognosis, survival status, and immune landscape of CRC. In this study, unsupervised clustering was utilized to divide CRC into different subtypes based on NK cell-associated genes among differentially expressed genes (DEGs) of CRC, and evaluated differences in prognosis, immune infiltration, and gene mutations amongst varying subtypes. We also elucidated the differences in gene enrichment pathways among different subtypes. Furthermore, we identified potential drugs that could improve the prognosis of CRC patients through small-molecule drug screening. This study provided an important reference for personalized treatment plans for CRC patients by dividing NK cell-associated CRC subtypes (The content of the process is shown in the Schematic diagram 1).