News

17/11/2021 - Press release

Artificial intelligence for selecting the best lung cancer treatment

Hospital del Mar, in collaboration with Roche, is working on a pioneering artificial intelligence tool to improve the diagnosis and prognosis of patients with small cell lung cancer, one of the cancers with the worst outlook. Based on digitised images from patient biopsies and survival data, the aim is to test whether the tool can reliably predict the benefit of the various treatments. This would allow more effective selection of the most appropriate type of treatment. At a later stage, the aim is to evaluate the specific mutations present in each tumour and to analyse whether digital imaging is capable of predicting these

Hospital del Mar, in collaboration with Roche, has initiated the development of a pioneering artificial intelligence device to improve the diagnosis and prognosis of patients suffering small cell lung cancer. This is one of the lung cancers with the worst prognosis and having such a reliable tool will allow medical teams to determine the best treatment for each type of patient.

There is currently no other similar tool for this type of tumour. The Hospital del Mar team, made up of the Medical Oncology and Anatomical Pathology services, intends to design an algorithm that, based on digitised images of analysed patient biopsies and survival data, can generate an artificial intelligence tool capable of effectively predicting the patient's prognosis and possible treatment response. It will also determine whether the case being studied could have a positive long-term response to anti-disease drugs.

Currently, data from 350 patients at the Hospital del Mar, collected over the last ten years, is already available and will begin to feed the algorithm in order to develop its potential. As Dr Joan Gibert, bioinformatician at the Anatomical Pathology Service and researcher at the Molecular Cancer Therapy Research Group at the Hospital del Mar Medical Research Institute (IMIM-Hospital del Mar), comments, "Artificial Intelligence (AI) algorithms identify patterns in Anatomical Pathology images that are used to make predictions. These models acquire knowledge based on examples and have proven to be very useful for detecting features that are difficult or impossible to see at first glance."

According to Nuria Soler, Director of Healthcare Solutions at Roche Spain, "This project will allow us to advance towards a better early diagnosis of the disease, as until now only 20% of these tumours are diagnosed at an early stage. Our priority is always the patients, and our way of offering them the best care is through innovation and research. Proof of this is this project in which we are working together with clinicians, providing them with the most advanced technology for the benefit of patients."

Deciding on the most appropriate treatment

Although small cell lung cancer treatment is changing with the introduction of immunotherapy, chemotherapy remains a key component in the management of this disease. This is why understanding the patterns of response to this therapy is crucial for improving the management of these patients. Dr. Edurne Arriola, head of the lung cancer section in the Medical Oncology Service at Hospital del Mar and an IMIM researcher, points out the advantages of having this kind of tool available for a type of cancer with few therapeutic alternatives. It affects 15% of lung cancer patients and the average 5-year survival rate is very low, at less than 10%. "This tool would allow us to determine the most appropriate treatment based on the anatomical pathology image, as well as survival data. It would also allow us to predict the benefit of treatment with chemotherapy or its combination with immunotherapy, based on how the algorithm interprets this data. And, finally, it would enable us to predict whether the patient would have a potential long-term response to the treatment", she stresses.

At present, the team developing this tool is also trying to determine the capacity of the liquid biopsy technique to detect prognostic and predictive markers of treatment benefit based on an analysis of the specific mutations of each tumour. The plan is to add this information to the algorithm to enhance its diagnostic ability. The initial results are expected in one year.

Personalised care and cutting-edge cancer research at Hospital del Mar

Personalised care for cancer patients at Hospital del Mar is provided through pioneering and cutting-edge work in multidisciplinary functional units specific to each kind of tumour. The units, comprising professionals specialising in each cancer type, offer the best therapeutic options in a model of shared decision-making with the patient. Nurse managers guide patients through the diagnostic and therapeutic process. This quality care is combined with ground-breaking cancer research at the Hospital del Mar Medical Research Institute (IMIM) and an extensive programme of clinical trials. The research areas focus on furthering immunotherapy and liquid biopsy, searching for biomarkers and new therapeutic targets, and developing new surgery and radiotherapy strategies to improve efficacy and patient quality of life. This research generates almost 200 articles in scientific publications each year, two out of three of which are in high-impact journals. This state-of-the-art care and research are the basis for teaching excellence at the Hospital del Mar Campus.

More news