REMOTE DIGITAL MANAGEMENT OF METASTATIC RENAL CANCER PATIENTS USING SUNITINIB

Introduction: Renal cell carcinomas (RCC) are the seventh most common histological type of cancer in the world. Those with inoperable or metastatic RCC often undergo treatment with targeted agents and/or checkpoint inhibitors. It is common that patients using tyrosine kinase inhibitors present severe adverse (TKI) symptoms that impact their quality of life and clinical outcome. Doctors and nurses are unaware of these symptoms most of the time. New modalities for monitoring patients outside the hospital setting may impact clinical management and improve oncology care. Objective: To manage renal cancer patients using sunitinib with the help of an artificial intelligence remote monitoring bot. Methods: The project was approved by the local IRB . Once registered in the monitoring system, the patient will receive daily messages with guidance on medication use and the management of side effects. The flow also allows the identification of warning signs by the monitoring team, with the possibility of direct and immediate contact with the patient. A health team (nurse and oncologist) will follow and confirm the patient-bot interaction. This interaction takes place through the CleverCare monitoring software, which has modules for collecting, processing information, personalized support via SMS, targeting actions, the dashboard for epidemiological data, and broadcast messages for alerts. The collected data are used to feed the Healthcare Data Science Platform. Results: We identified the main side effects related to the use of sunitinib and created an interaction flow for the bot. The flow set and remember the patient for the time to take the pills, blood collection, and future visits appointments, and reinforce prevention actions such as water intake, show how to measure blood pressure and the application of skincare creams for palmar – plantar erythrodysesthesia prevention. In addition, it can ask specific questions about major TKI treatment complications: nausea and vomiting, fatigue, diarrhea, hand-foot syndrome, mucositis, hypertension, hypothyroidism, bleeding. On this platform, information is analyzed using machine learning algorithms, artificial intelligence, and natural language processing. Three patients validated the flow and referred to improvement in their confidence in the treatment. A prospective randomized trial will evaluate the impact of this intervention on RCC clinical care. Conclusion: Remote digital symptom monitoring may be considered for implementation as a part of high quality cancer care.