Abstract CN64
Background
Immune checkpoint inhibitors have brought significant clinical benefits to cancer patients. However, the stimulation of the immune system caused by these drugs might lead to adverse events.
Methods
To estimate the type, frequency, and severity of adverse events in cancer patients undergoing immune checkpoint inhibitors, we conducted a study in two private clinics in Brasilia, DF, Brazil. Data were collected from medical records of individuals with cancer, who were treated between 2013 and 2020 with any immune checkpoint inhibitor (alone or in combination with chemotherapy) and who were 18 years of age or older. Patients whose medical records did not contain information related to the assessment of adverse events were excluded from the study. From each medical record, we extracted types of adverse event e maximum degree of severity for each event, as well as the time between the start of therapy and first appearance of an adverse event. For each type of event, we calculated proportion of occurrence and its 95% confidence interval.
Results
The final sample consisted of 64 participants. Among them, 70.3% presented at least one type of immune-related adverse event. The most frequent events were fatigue (21.9%), itching (20.3%), skin rash (20.3%), diarrhea (20.3%) and hypothyroidism (18.8%). For fatigue, most patients developed grade 1 severity (12.5%), followed by grade 2 (6.3%) and grade 3 (3.1%). For skin rash and pruritus events, most participants were classified as grade 1 (11% for both events), with two individuals (3.1%) developing grade 3 for skin rash. For diarrhea, there was a predominance of grade 1 (12.5%), followed by grade 2 (3.1%) and grade 3 (4.7%). For hypothyroidism, most participants developed grade 1 (12.5%), followed by grade 2 (6.3%). The median time to onset of immune-related adverse events was 13 weeks, ranging from 8 to 54 weeks.
Conclusions
In conclusion, the most frequent immune-related adverse events were fatigue, itching, skin rash, diarrhea and hypothyroidism, most of which were classified as grades 1 and 2. It is possible that these are underestimated of the real proportions in this population, since 12 medical records were excluded for lack of information. More care recording adverse events would make the estimates more precise.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
C.I. Vasques, J. Barbosa Lacerda.
Funding
National Council for Scientific and Technological Development (CNPq), Brazil and Foundation for Research Support of the Federal District (FAP-DF), Brazil.
Disclosure
All authors have declared no conflicts of interest.