In an interim analysis of a single-institution study reported in JAMA Network Open, Pavlos Msaouel, MD, PhD, and colleagues found that the use of a mobile device application for patient reporting of adverse reactions to immune checkpoint inhibition was feasible. It permitted identification of patient reports most likely to require clinical action and required no additional care team staffing.

As stated by the investigators: “Immune checkpoint inhibitors can produce distinct toxic effects that require prompt recognition and timely management. [The aim of the study was to] develop a technology-enabled, dynamically adaptive protocol that can provide the accurate information needed to inform specific remedies for immune toxic effects in patients treated with immune checkpoint inhibitors.”

Pavlos Msaouel, MD, PhD

Study Details

The study enrolled 50 patients with genitourinary cancers treated with immune checkpoint inhibitors at The University of Texas MD Anderson Cancer Center between September 2019 and September 2020. A smartphone application was downloaded onto patients’ personal mobile devices 1 week before initiation of immune checkpoint inhibitor treatment; it prompted patients to report symptoms at least three times per week. Each time a particular symptom report triggered a prespecified symptom alert threshold, the application informed patients to call their clinical care team, as well as sent automated e-mails to the clinical care team. Patients with symptoms likely attributable to immune checkpoint inhibitor toxicity were evaluated and underwent clinical management.

The primary outcome measure of the interim analysis was feasibility, measured as patient and care team adherence and the lack of need to increase care team staffing. Patients were considered adherent on a particular day if they had completed a survey on either that day or within the previous 2 days. Care team adherence was assessed as the time to respond to an alert, with the percent of responses within 3 days and 7 days being evaluated.  

Key Findings

Among the 50 patients, 47 had at least one follow-up visit and were included in the analysis. Among these patients, the median age was 65 years (range = 37–86 years); 39 (83%) were men; and 39 (83%) had metastatic cancer, most commonly urothelial cell carcinoma (n = 22) and renal cell carcinoma (n = 22).

The median follow-up was 63 days (interquartile range [IQR] = 35.5–122 days). A total of 1,502 surveys were completed by the 47 patients over 4,552 patient-days, with patients submitting a median of 24 surveys. Median patient adherence was 74% (IQR = 60%–86%).

Symptom reports from patients led to 409 automatic alerts for 593 individual symptoms. The median time to care team response to alerts was 19 hours (IQR = 3–80 hours), with 73% of alerts reviewed within 3 days and 89% within 7 days.

The most frequently reported symptoms were fatigue (40%), pain (21%), and diarrhea (21%). Arthralgia and myalgia were reported in 17% of all surveys but accounted for the highest number of symptom alerts (n = 142).

Of the 593 symptom alerts, 51 (9%) were linked to an adverse event and required acute clinical intervention. The weighted overall positive predictive value for an alert resulting in acute intervention was 6.9%. Symptoms with the highest positive predictive values were nausea/vomiting (26%), dizziness (21%), and shortness of breath (14%). The symptoms most likely to result in alerts not requiring intervention were arthralgia and myalgia, fatigue, and cough.

The investigators concluded: “The findings of this cohort study suggest an acceptable and fiscally sound method can be developed to create a dynamic learning system to detect and manage immune-related toxic effects.”

Christopher Logothetis, MD, and Dr. Msaouel, of the Department of Genitourinary Medical Oncology, The University of Texas MD Anderson Cancer Center, are the corresponding authors of the JAMA Network Open article.

Disclosure: The study was supported by the National Cancer Institute and the Ronin Project. For full disclosures of the study authors, visit jamanetwork.com.