Johns Hopkins Kimmel Cancer Center researchers have developed a novel blood test called GEMINI (Genome-wide Mutational Incidence for Non-Invasive detection of cancer) that combines DNA sequencing and machine learning to detect cancer at an early stage. The test looks for changes in the DNA throughout the genome, which are then analyzed by a machine learning model to distinguish between individuals with cancer and those without.
In a proof-of-concept study published in Nature Genetics, the researchers found that the GEMINI test, when combined with computerized tomography imaging, detected over 90% of lung cancers, including stage I and II diseases. Furthermore, the study indicated that the test could potentially be used to detect other types of cancers, such as liver cancer, melanoma, and lymphoma.
To develop GEMINI, the researchers examined the whole-genome sequences of cancers from 2,511 individuals with various types of cancer. They identified distinct mutation frequencies across the genome in different tumor types, such as lung cancer, which had an average of 52,209 somatic mutations per genome. The researchers also found that genomic regions with a high frequency of mutations in tumor tissue were similar to those found in blood-derived cfDNA (cell-free DNA) from patients with lung cancer, melanoma, and B cell non-Hodgkin lymphoma.
In laboratory tests using samples from the Longitudinal Urban Cohort Ageing Study (LUCAS), the GEMINI test accurately detected cancer by generating a score that reflected the probability of having cancer. The combination of GEMINI and another test called DELFI (DNA evaluation of fragments for early interception) improved the detection of early-stage lung cancer.
Researchers also observed that the GEMINI test could detect abnormalities in cfDNA mutation profiles years before standard diagnoses, highlighting its potential for early cancer detection and monitoring patients during therapy. The test was able to distinguish between different subtypes of lung cancers and detect early liver cancers.
While the results of the study demonstrate the potential of GEMINI as a cost-efficient and scalable method for detecting cancer, the researchers acknowledge the need for larger clinical trials to validate its effectiveness before it can be used in a clinical setting.
In conclusion, the novel GEMINI blood test, which combines DNA sequencing and machine learning, has shown promise in detecting lung and other cancers at an early stage. With further validation through extensive clinical trials, this test could potentially revolutionize cancer detection and improve patient outcomes.