In recent years, there has been a drastic change in the treatment landscape of several advanced cancers. Cancer immunotherapies have shown to be a viable treatment option for many patients who would have otherwise had a poor prognosis with traditional treatments, such as chemotherapy.
So far the most successful subtype of cancer immunotherapies have been immune checkpoint inhibitors, which have become first- or second-line treatments in many standard of care guidelines across several indications such as advanced lung cancer or melanoma. Currently approved immune checkpoint inhibitor drugs block proteins CTLA-4 (cytotoxic T-lymphocyte-associated protein 4), PD-1 (programmed death 1) or PD-L1 (programmed death-ligand 1). By this immune checkpoint blockade, they remove the immune system’s brakes to recognise and kill cancer cells.
Improved outcomes, but with new kinds of challenges
While undoubtedly bringing benefit for many patients in terms of improved outcomes, these novel therapies don’t come without challenges. One of the key challenges is caused by the very mechanism that helps the body’s T-cells identify and attack cancer cells. The mechanism behind their efficiency is also the mechanism causing many, and sometimes severe, adverse events. Overactivation of the immune system may cause many unwanted effects as the immune system may also target some non-harmful, normal cells. Hence, the adverse events from immune checkpoint inhibitors resemble autoimmune diseases. The adverse events of immune checkpoint inhibitors can arise at any time during treatment, sometimes even when a patient’s therapy has ended, and they may affect any organ or tissue of the human body. Management of these novel toxicities is crucial for providing patients with safe therapies and achieving the best possible outcomes.
Digital health interventions in detecting immune-related adverse events
Most immune-mediated adverse events are reversible, when detected and treated early (1). It is crucial that healthcare professionals learn to recognise symptoms of potential immune-related adverse events early on. In many cases, the visibility of patients’ symptoms for their care team is limited to certain time points; either during a scheduled appointment, or in an unscheduled visit, such as a visit to the ER with severe symptoms. To get timely visibility to patients’ symptoms, patients need to be able to report their symptoms continuously, and healthcare professionals need to be able to review, evaluate and triage patients based on their status. It is highly important that patients proactively communicate any potential new symptoms that may be related to toxicities during and after their treatment periods.
This very challenge has been in the works in our product development teams. Since 2015, we have been diligently working with a number of leading European institutes and cancer care centers in the development of algorithms for detecting the development of immune-mediated toxicities. The developed algorithms alert care teams when symptoms start to develop unfavourably or are severe, prioritise patient reports, and enable healthcare professionals to do early interventions for patients. The algorithms also educate both patients and healthcare professionals to take necessary proactive actions to help intervene as early as possible.
The development has already led to exciting results. For example, an oncologist is capable of detecting a patient having pseudoprogression. The algorithms are also capable of identifying, for example, a developing severe gastrointestinal toxicity early on while patients are still at home, urging the patient to seek attention at the outpatient clinic or emergency room.
From Reactive to Predictive Toxicity Management
Some of our latest developments in the fight against immune-related toxicities include the utilisation of predictive models based on machine learning. Machine learning is applied in the detection of onset and continuation of toxicities patients may receive from immune checkpoint inhibitor therapies. Based on the patient’s therapy and the historical trend of symptoms, Kaiku Health algorithms are capable of predicting the occurrence of up to 15 symptoms that may indicate the development of immune-related adverse events, such as skin and pulmonary toxicities. This is an important development in our continuous effort to turn digital health interventions from reactive to predictive.
There’s still limited knowledge on adverse events when treating patients with cancer immunotherapies and on understanding which patients will benefit from these novel therapies. We truly believe that utilising cutting edge technology to better understand and manage toxicities is crucial in improving safety of these novel therapies further and to help in reaching ever better patient outcomes. We will only get there by working closely with leading cancer centers, research institutes and life sciences companies.
Henri Virtanen, Chief Product Officer
1) Haanen JBAG et. al, on behalf of the ESMO Guidelines Committee, Management of toxicities from immunotherapy: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up, Annals of Oncology, Volume 28, Issue suppl_4, July 2017, Pages iv119–iv142, https://doi.org/10.1093/annonc/mdx225