Swiss researchers have achieved a breakthrough in personalized cancer treatment by successfully demonstrating the clinical feasibility of using comprehensive multiomics analysis to predict optimal therapies for melanoma patients. The landmark study, conducted by the Tumor Profiler research consortium, represents the first time nine molecular biological technologies have been used in parallel to guide real-world cancer treatment decisions.
The eight-year project, led by researchers at University Hospital Zurich, University of Zurich, ETH Zurich, and University Hospital Basel in collaboration with Roche, analyzed tumor samples from 116 melanoma patients using an unprecedented array of molecular technologies. The comprehensive analysis generated 43,000 data points per sample, equivalent to 0.5 terabytes of data, enabling precise treatment decisions within four weeks.
Multiomics Platform Demonstrates Clinical Utility
The study employed nine distinct technologies to create a comprehensive molecular portrait of each tumor: single-cell genomics and transcriptomics, targeted spatial proteomics, cytometry by time of flight, mass spectrometry proteotyping, drug phenotyping, iterative indirect immunofluorescence imaging, targeted next-generation DNA sequencing, and digital pathology. Of these technologies, only NGS and digital pathology are currently part of standard diagnostic protocols.
"These values and information show us that the recommendations from tumor profiling are available within a reasonable period of time and with tangible and directly implementable benefits for the treating physicians," said Nicola Miglino, research assistant at the Department of Medical Oncology and Hematology at University Hospital Zurich and lead author of the study.
The technical turnaround time for the multiomics testing was just two weeks, with the complete process from biopsy to treatment recommendation taking four weeks. This timeline included one week for biopsy preparation and another week for multidisciplinary tumor board discussion.
Superior Treatment Outcomes Observed
In 75% of cases, treating specialists found the tumor profiling recommendations helpful for therapy selection and reported that the data provided substantial information for treatment decisions. More significantly, patients whose treatments were guided by the profiler data demonstrated better response rates and progression-free survival compared to a non-randomized, combined exact- and propensity-matched comparator group who did not participate in the program.
Andreas Wicki, professor of oncology at the University of Zurich and senior physician at University Hospital Zurich, noted that while "the benefit to the overall population is very limited, but for a couple of patients there was a huge benefit." These were individuals who had exhausted standard-of-care therapies and went on to receive specific, biomarker-based treatments, with some patients still alive after four years.
Cost Reduction and Technology Optimization
The economic feasibility of the approach has improved dramatically since the project's inception. When the study began in 2018, the cost for analyzing the first patient was approximately 140,000 Swiss francs, but this has since dropped to roughly one-tenth that level. Researchers also demonstrated through simulation exercises that most patients could benefit from using only two, three, or four technologies rather than the full nine-technology panel.
The treatment recommendations were derived from 54 individual markers known to be associated with tumor response to specific therapies, selected from the vast pool of multiomics data. The majority of the 43,000 biomarkers were not utilized because, as Wicki explained, "we don't understand what the data points mean, and we don't know whether they matter."
Implications for Future Clinical Trials
The study addresses a critical challenge in oncology drug development, where 5,000 to 6,000 compounds are currently in the pipeline with 90% expected to fail. This still leaves 500 to 600 new drugs needing to find their clinical role in the next decade. "There is no way we can test all those drugs in sequence or in parallel in many different molecular groups of patients," Wicki noted.
The diagnostic-based treatment prediction approach could enable clinical trials to move away from seeking largely homogeneous patient populations, potentially making more patients eligible for participation. The practical hope is to avoid "running therapies through a process of trial and error," with even modest improvements in response rates from 60% to 70% or 80% to 90% having considerable impact on patients and healthcare costs.
Expanding to Multiple Cancer Types
The success of the melanoma study has prompted expansion to a second phase involving all Swiss university hospitals and five of the most prevalent solid cancers: melanoma, breast, colorectal, lung, and ovarian cancers. This broader implementation will test the approach across cancers with varying amounts of available tumor tissue.
The research team is now focusing on standardization of clinical testing and bringing these insights into randomized controlled trials for formal assessment. Machine learning models will be essential for leveraging all available data in predicting therapy outcomes, representing a future focus of the Profiler consortium.
"This study is a major step toward data-based medicine," Wicki concluded. "It paves the way for new clinical trials that don't test individual drugs, but actually predict the most effective therapy."
