For example, a new imaging technique that examines the location of mitochondria in lung cancer can be developed to distinguish highly aggressive tumors from less aggressive tumors. An emerging strategy in early cancer detection is to screen for multiple cancers simultaneously (see Moving Toward Minimally Invasive Cancer Screening, p. 65). However, such tests generate a wealth of information, and a limitation so far has been the accurate and timely analysis of results from these tests. AI-based approaches have immense potential of overcoming this limitation. Artificial intelligence is also contributing to other aspects of cancer science and medicine. Researchers are already leveraging AI-based analyses of large genomic datasets in classifying tumors at a molecular level (687), determining tumor heterogeneity (688), diagnosing primary and metastatic cancers (689), identifying biomarkers for treatment selection (690), and predicting overall survival (691), among other applications across the cancer care continuum (692). Another exciting application of AI in cancer medicine is its potential to create models that can predict how patients will respond to a treatment. In a recent study, researchers used clinical and genomic data from 700 patients with cancer, who were treated with ICIs, to develop a prediction model based on machine learning—a type of AI that is programmed to learn over time from the data provided to make predictions or decisions (693). The model accurately predicted how patients with melanoma, gastric cancer, and bladder cancer would have responded to the ICI treatment. Furthermore, the model’s predictions were more accurate compared to those based on biomarkers currently being used in the clinic to predict a patient’s response to the ICI treatment (693). Cancer science and medicine are on the verge of an AI-driven revolution. However, it is important to be cognizant of the fact that AI-assisted approaches can introduce unintended biases in analyses and can further widen inequities in the burden of cancer experienced by medically underserved populations (694) (see Sidebar 24, p. 66). It is vital that the datasets used to train AI-based models accurately and proportionally represent population groups affected by the type of cancer being studied. Wearable Technologies Wearable technology, also called wearables, is a category of electronic devices that can be worn as accessories, embedded in clothing, implanted in the user's body, or affixed on the skin. Wearables are hands-free microcomputers with the ability to send and receive data via the Internet as well as to perform a variety of functions, such as counting steps or monitoring heart rate. An exciting area of cancer research is the use of wearables that can be implanted on or inside the patient’s body for delivering drugs effectively and automatically (695,696). Patients with cancer, especially elderly patients, often have to take many medications. According to one study of patients with cancer ages 70 years or older, 61 percent of the study’s participants were taking five or more medications a day (697). Many of the anticancer therapeutics are pills taken orally, which can cause distress and inconvenience for patients, resulting in patients not taking their medication on time or at all (698), thus increasing the risk of adverse health outcomes (697). Researchers recently reported the development of a wearable that can be attached to skin for delivering anticancer drugs (699). The study showed that the wearable successfully delivered anticancer drugs in a preclinical mouse model of melanoma and prevented recurrence of cancer. Importantly, there was no noticeable effect of the delivered drugs on other organs in the body (699). Another focus of ongoing research is examining the potential of wearables to diagnose cancer. Recently, researchers reported the development of a flexible patch containing a miniaturized ultrasound scanner. The patch can be attached to a bra, which when worn scans the breast tissue. The ability to attach the patch in six different positions on the bra allows the entire breast to be imaged (700). Although the patch needs to be tested in a large number of individuals, such innovative applications of wearables are opening exciting new frontiers in cancer science and medicine. Smart watches and health and fitness (activity) trackers are among the most commonly used wearables. According to a 2020 report, one in five U.S. adults regularly wears a smart watch or a fitness tracker, and more than half of those who do support sharing of data from these devices with their health care providers and with researchers (701). Researchers are testing the potential use of smart watches and activity trackers in clinical cancer care in several ways (702). Findings from a number of studies have shown that use of wearables among cancer patients and survivors helps encourage physical activity (703); predict whether patients receiving active cancer treatment are at an increased risk of hospitalization (704); identify patients who are fit for certain types of cancer treatments (705); examine association of sleep patterns with overall survival (706); and determine quality of life (707). Future studies with large numbers of patients will further strengthen the utility of wearables in cancer science and medicine as an exciting new frontier that can measure biometric parameters in patients and deliver care remotely (708), thus harnessing technological advances to further expand and improve telemedicine. Tackling Difficult-toTreat Cancers The consistently declining rates of U.S. cancer deaths in recent years underscore the unprecedented progress against cancer Envisioning the Future of Cancer Science and Medicine AACR Cancer Progress Report 2023 147
RkJQdWJsaXNoZXIy NTkzMzk=