Harnessing proteomics for precision medicine in lung cancer

Lung cancer continues to impose a substantial burden on society. While progress has been made through prevention, new treatments, and, in some countries, screening, its impact remains profound. The emergence of targeted therapies for oncogene-mutated tumors and immune checkpoint inhibitors have contributed to significant improvements in patient survival in late-stage cases. However, not all patients have a suitable targetable mutation, not all respond to the prescribed treatment, and among those that do, many relapse. As a result, less than one-third of lung cancer patients in Sweden survive longer than 5 years after the diagnosis. There is a pressing need for continued development of precision medicine strategies to deliver the right treatment to the right patient.

Genomics has been a driving force in cancer research and has become one of the key diagnostic tools in oncology, alongside immunohistochemistry. However, it has become evident that evaluation of individual marker expression and genome- level analyses may be insufficient to fully decipher cancer complexity and make lung cancer, if not curable, at least a manageable chronic disease. Proteomics has more recently emerged as a valuable tool for characterizing the phenotype and providing insights into cancer mechanisms and vulnerabilities. Thus, this thesis aimed to elucidate lung cancer biology at the proteome level and to demonstrate the clinical utility of proteomics methods.

In Study I, we analyzed the proteomic landscape of non-small-cell lung cancer (NSCLC) using in-depth mass spectrometry (MS)-based proteomics on 141 resected tumors. We complemented the analysis with genomic, transcriptomic, epigenomic, and clinical data and described six proteomic subtypes of NSCLC. The subtypes reflected the expected histological groupings but exhibited distinct immune infiltration patterns and oncogene-driver mechanisms. We proposed mechanistic insights and therapeutic vulnerabilities underlying the observed phenotypes and developed classification methods to enable the study of new samples and cohorts in the context of these findings.

In Study II, we investigated the post-translational modification farnesylation using lung cancer cell line models. Farnesylation and its inhibition using farnesyltransferase inhibitors (FTIs) play a role in cancer by determining the subcellular localization and activity of several cancer-related proteins. We used MS-based proteomics to identify farnesylated proteins, conduct a proteome- wide assessment of protein localization and relocalization in response to FTI treatment, and analyze global FTI effects on protein abundance. Our study revealed clinically relevant targets beyond Ras family proteins for which FTIs were initially developed and provided a comprehensive data resource for further exploration by the research community.

In Study III, we demonstrated the performance of MS-based proteomics applied to the most common type of preserved tissue in the clinic - formalin-fixed, paraffin-embedded (FFPE). We analyzed 15 lung adenocarcinoma samples using label-free proteomics on three different modern MS instruments. We achieved deep proteome coverage, quantifying over 7,000 proteins in less than 45 min of analysis time on the two newest instruments. We also demonstrated the potential clinical utility of the generated data by exploring the identification and quantification of drug targets and potentially predictive biomarkers.

In Study IV, we designed a protein-based predictive biomarker panel for immune checkpoint inhibitor (ICI) treatment selection in NSCLC and developed a complementary MS-based targeted proteomics assay. The panel covered tumor microenvironment features related to immune cell infiltration, interferon gamma signaling, antigen presentation, proteasome composition, and other immune- related processes and sample composition properties. We demonstrated the clinical relevance of the panel using a publicly available transcriptomics dataset from an ICI clinical trial. Our findings included distinct survival trends in the biomarker-stratified groups and complementarity to the PD-L1 expression status that is commonly used as a biomarker for ICI treatment selection. We then validated the targeted proteomics assay's analytical performance on an independent cohort of biopsy samples.

Together, these studies have contributed to a deeper understanding of lung cancer biology at the proteome level and support ongoing efforts to integrate proteomics into clinical practice. By generating clinically relevant data and demonstrating the feasibility of proteomics in various settings, this work offers a foundation for future precision medicine applications.

List of scientific papers

I. Janne Lehtio, Taner Arslan, loannis Siavelis, Yanbo Pan, Fabio Socciarelli, Olena Berkovska, Husen M. Umer, Georgios Mermelekas, Mohammad Pirmoradian, Mats Jönsson, Hans Brunnström, Odd Terje Brustugun, Krishna Pinganksha Purohit, Richard Cunningham, Hassan Foroughi Asl, Sofi Isaksson, Elsa Arbajian, Mattias Aine, Anna Karlsson, Marija Kotevska, Carsten Gram Hansen, Vilde Drageset Haakensen, Åslaug Helland, David Tamborero, Henrik J. Johansson, Rui M. Branca, Maria Planck, Johan Staaf, and Lukas M. Orre. Proteogenomics of non-small cell lung cancer reveals molecular subtypes associated with specific therapeutic targets and immune-evasion mechanisms. Nature Cancer 2, 1224-1242 (2021). https://doi.org/10.1038/s43018-021-00259-9

II. Yanbo Pan*, Olena Berkovska*, Soumitra Marathe, Georgios Mermelekas, Greta Gudoityte, Amare D. Wolide, Taner Arslan, Brinton Seashore-Ludlow, Janne Lehtio, and Lukas M. Orre. Functional-proteomics-based investigation of the cellular response to farnesyltransferase inhibition in lung cancer. iScience 28, 111864 (2025). https://doi.org/10.1016/j.isci.2025.111864

III. Olena Berkovska*, Igor Schliemann*, Mahnaz Nikpour, Georgios Mermelekas, Janne Lehtio, and Lukas M. Orre. Label-free mass spectrometry-based proteomics delivers rapid, in-depth proteome-wide profiling of FFPE tissue. [Manuscript]

IV. Olena Berkovska*, Georgios Mermelekas*, Soumitra Marathe, Marija Karadzovska-Kotevska, Mats Jönsson, Maria Planck, Janne Lehtiö, and Lukas M. Orre. Development of a targeted proteomics assay for immunotherapy response prediction in lung cancer. [Manuscript]

*These authors contributed equally