TARGETED therapy in lung cancer to achieve a good prognosis is of the utmost importance because of the difficulty associated with detection in the early stages of the condition.
Clinicians are faced with difficulty in the early detection of lung and bronchus cancer because it often presents with no symptoms and progression of the disease is cannot be monitored early on. Additionally, individualised treatment is hard to initiate because of the nature of immunotherapy and varying success of treatment between individuals. The authors of the study, from Case Western Reserve University in Cleveland, Ohio, USA, commented on the importance of the method developed for the study, “…the nodule may appear larger after therapy because of another reason… but the therapy is actually working. Now, we have a way of knowing that.”
In collaboration with six academic institutions, researchers worked on an artificial intelligence (AI) model targeted to ascertain the lung cancer patients in whom immunotherapy would be the most beneficial, which also confers economic advantages by reducing unnecessary costs.
Development of this AI tool was built upon previous research which identified different responses to treatment from cancer tumour types. Data from 50 CT scans were used to establish mathematical methods which then identified the size and texture in the tumour after a small number of immunotherapy cycles. Specific changes in tumours identified by patterns were associated with positive response to treatment and higher survival rates. The tumours which exhibited the most obvious texture changes were those that had the greatest response to immunotherapy. To further this research the authors intend to use the AI method with CT scans from different sites around the body, and test individuals who have undergone treatment with different immunotherapy agents.