Metabolic and Nutritional Disorders Following the Administration of Immune Checkpoint Inhibitors: A Pharmacovigilance Study

Abstract

Background: Although several metabolic and nutritional disorders (MNDs) have been reported in the recipients of immune checkpoint inhibitors (ICIs), these events have not been fully captured and comprehensively characterized in real-world population.

Objectives: To provide complete metabolic and nutritional toxicity profiles after ICIs (single and combined) initiation through an integrated big database.

Methods: Reporting odds ratios (ROR) and information component (IC) based on statistical shrinkage transformation were utilized to perform disproportionality analysis using the US Food and Drug Administration Adverse Events Reporting System. Both ROR and IC were used to calculate disproportionality when compared with the whole database, but only ROR was used when comparison was made for different ICI strategies. Only when both the lower limits of 95% confidence intervals (CIs) for ROR (ROR₀₂₅) and IC (IC₀₂₅) exceeded specified threshold values (1 and 0, respectively) was regarded as a signal.

Results: A total of 29,294,335 records were involved and 8,662 records were for MNDs in patients exposed to ICIs. Statistically significant association was detected between ICIs use and total MNDs (IC₀₂₅/ROR₀₂₅ = 1.06/2.19). For monotherapy, three ICI monotherapies (anti-PD-1, anti-PDL-1, and anti-CTLA-4) were all disproportionately associated with MNDs. Statistically significant differences in reporting frequencies also emerged when comparing anti-PD-1 with anti-PD-L1/anti-CTLA-4 monotherapy, with RORs of 1.11 (95%CI 1.01-1.21), and 1.35 (95%CI 1.23-1.48), respectively. Notably, combination therapy was associated with a higher reporting frequency of theses toxicities compared to monotherapy with a ROR of 1.56 (95%CI 1.48-1.64). Additionally, disproportionality analysis at High-level Group Term level highlighted eight broad entities of MNDs. Further disproportionality analysis at Preferred Term level indicated a wide range and varied strength of signals. For ICI monotherapy, nivolumab and pembrolizumab showed the broadest spectrum of MNDs. For combination therapy, a variety of signals were detected for nivolumab + ipilimumab therapy even comparable to two PD-1 monotherapies.

Conclusion: Metabolic and nutritional complications could be provoked by ICI monotherapy (especially anti-PD-1) and further reinforced by combination therapy. Clinicians and patients should be informed about these potential risks that might be encountered in real-world practice. Aforehand education and regular monitoring of related biochemical parameters (calcium, sodium, potassium, protein) are recommended to ensure better cancer survivorship.

Keywords: FAERS database; disproportionality analysis; immune checkpoint inhibitors; information component; metabolic and nutritional disorders; pharmacovigilance study; reporting odds ratio.

Figure 1

Safety profile of metabolism and nutrition diseases according to different ICI regimens*. *Each cell of the safety profile contains the values of IC₀₂₅ and ROR₀₂₅(lower end of the 95% confidence interval of IC and ROR, respectively); NIVOL, nivolumab; PEMBR, pembrolizumab; ATEZO, atezolizumab; AVELU, avelumab; DURVA, durvalumab; IPILI, ipilimumab; POLY1, nivolumab+ pembrolizumab+ ipilimumab; POLY2, nivolumab+ ipilimumab; POLY3, pembrolizumab+ ipilimumab. To obtain robust results and reduce the false positive signals, signal values were only calculated for complications with at least 3 records. A signal was defined as both IC₀₂₅ >0 and ROR₀₂₅ >1 and highlighted in blue.