Summary and future perspectives 243 7 thus enhancing susceptibility to PD-1 blockade therapy. Historical outcome- and survival data of chemotherapy-only treated patients are needed to differentiate the distinct effects of chemotherapy and PD-1 blockade therapy. It is noteworthy that conducting a future phase three clinical trial excluding PD-1 blockade therapy in PD-1T-low + PD-L1low patients may face challenges. A subset of patients may not participate in the trial to avoid the risk of being randomized into the chemotherapy treated arm. The reluctance comes from the high demand for ICB treatment, and de-escalation studies encounter a lack of popularity. This sentiment refers to the principal that ‘favoring more extensive treatment is a prevailing preference over opting for less intervention’. However, patients allocated to the chemotherapy-only arm should retain the option to receive PD-1 blockade monotherapy as a second-line intervention upon disease progression. Furthermore, PD-1 blockade overtreatment leads to unnecessary side effects in patients and health care systems deal with a substantial increase in costs. Consequently, it is crucial to improve personalized medicine to guide treatment decisions, thereby reducing costs. The resources saved could, for instance, be allocated to studies evaluating alternative treatment modalities tailored for patients who show primary or secondary resistance to PD-1 blockade therapy. Chapter 5 and 6; Serum and pleural effusion as bio-sources for diagnostic biomarker tests In chapter 5 and 6 of this thesis we investigated alternative bio-sources for biomarker assessment. In current diagnostics, a tissue biopsy is a requisite in the majority of cases and, as such, continues to serve as a golden standard. However, it is an invasive, cumbersome procedure and small tumors may require multiple attempts to secure an adequate tumor tissue sample. In contrast, liquid biopsies offer a minimally invasive approach, enabling the retrieval of information from all tumor sites, thereby circumventing the loss of critical information attributed to tumor heterogeneity. Mass spectrometry (MS) is a powerful tool for analyzing the proteomic landscape within blood samples. In combination with machine learning algorithms, this approach can be used to develop proteomic signatures with predictive capacity. In chapter 5 of this thesis, we performed a MS-based proteomic analysis using pretreatment sera derived from 289 advanced-stage NSCLC patients who received second-line nivolumab treatment, with the aim of constructing a predictive protein signature. By using machine learning, which incorporated spectral data alongside clinical information, we stratified patients into three distinct groups with good (“sensitive”), intermediate, and poor (“resistant”) outcomes following treatment. Our results demonstrated a strong association between the signature and PFS as well
RkJQdWJsaXNoZXIy MTk4NDMw