Sage Journals: Discover world-class research

Abstract

Objective:

To investigate the comprehensive landscape of hepatotoxic adverse events (AEs) associated with immune checkpoint inhibitors (ICIs), with a special focus on evaluating the potential risk of lethal hepatotoxic AEs.

Introduction:

The widespread adoption of ICIs has markedly improved the prognosis for patients with advanced cancer. However, this therapeutic advance is accompanied by the risk of immune-related adverse events (irAEs), especially hepatotoxic AEs, which particularly affect patients with pre-existing liver diseases or hepatobiliary cancers.

Methods:

Reports in the FAERS database from Q1 2014 to Q3 2024 were collected. The characteristics of ICI-related hepatotoxic AEs were summarized. Disproportionality analysis was conducted to calculate reported odds ratios for assessing signals of hepatotoxic AEs. Additionally, logistic regression was employed to evaluate patient-related factors contributing to an increased risk of these AEs.

Results:

Hepatotoxic AEs increased yearly and occurred primarily in patients with hepatobiliary tumors. CTLA-4 inhibitors exhibited the highest incidence of AEs. In contrast, PD-1 inhibitors had the shortest median time to AE onset. Abnormal hepatic function was a common AE, whereas Stauffer’s syndrome was identified as a rare AE. The risk of hepatotoxic AEs was notably elevated by combination immunotherapy and the concurrent use of specific drugs. Despite variations in the safety profiles of different ICI regimens, these differences did not significantly influence the risk of fatal hepatotoxicity. Furthermore, older men who experienced other AEs were found to be at higher risk for developing fatal hepatotoxicity.

Conclusion:

The safety profiles of different ICIs vary widely, necessitating individualized drug selection based on patient-specific factors.

Keywords

adverse events hepatotoxicity immune checkpoint inhibitor pharmacovigilance analysis FAERS database

Introduction

Traditional chemotherapy primarily combats cancer by inhibiting cell proliferation and inducing apoptosis.¹ As research into tumors has deepened, the discovery of the tumor microenvironment, particularly its immune components, has unveiled mechanisms of immune regulation and identified targets for immunotherapy.^2,3 Besides therapies such as cytokines and adoptive cell therapy, immune checkpoint inhibitors (ICIs) have emerged as a cornerstone in the treatment of various advanced malignancies.⁴

Since the U.S. Food and Drug Administration (FDA) approved ipilimumab for the treatment of metastatic melanoma in 2011, ICIs have been increasingly used for the treatment of a variety of advanced malignancies.^5,6 ICIs are a class of monoclonal antibodies that are primarily directed against cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed death receptor 1 (PD-1), or its ligand (PD-L1). By reversing the dysfunctional state of T cells and modulating the tumor microenvironment, ICIs enhance the immune response against tumors.^7–9 ICIs have improved the survival of many patients with advanced cancer, and as a result, were awarded the 2018 Nobel Prize in Physiology or Medicine.¹⁰ This success marks a promising new era in oncology, especially for patients who have not responded to conventional therapies.

Immune checkpoints are critical for self-tolerance and preventing autoimmune reactions. However, by blocking these checkpoints, ICIs can lead to immune-related adverse events (irAEs) when the immune system attacks healthy tissues.¹¹ irAEs cause tumor cells to release proteins, including self-antigens, which activate T-cells and may result in damage to normal tissues.¹² Studies have shown that the incidence of irAEs varies widely among patients and affects organs such as the skin, endocrine glands, gastrointestinal tract, liver, and lungs.¹³ While most irAEs are manageable, some can be serious or life-threatening. Safety analyses have shown that ICI-related adverse events occur in 54%–76% of patients, with life-threatening cases accounting for 14.1%–28.6%.¹⁴ Therefore, vigilant monitoring and timely management of adverse events are essential to ensure that ICI therapy is safe and effective, optimizing patient prognosis and maintaining quality of life.

The increased use of ICIs has led to a rise in the incidence of irAEs, which has become a major clinical concern. As of 2018, the International Immune Cancer Registry reported 13,051 cases of irAEs in 12,648 patients.¹⁵ ICI-related hepatotoxicity affects 0%–30% of patients, with severe cases ranging from 0% to 20%. This is particularly significant for patients with hepatocellular carcinoma (HCC) or pre-existing liver disease, as hepatotoxic AEs can worsen liver disease, compromise treatment outcomes, and affect patient survival and quality of life.^16,17

The U.S. Food and Drug Administration Adverse Event Reporting System (FAERS) is an important global database used to monitor the safety of FDA-approved drugs after they are marketed.¹⁸ Our study analyzed hepatotoxic AEs for various ICI regimens using FAERS data to identify potential signals of hepatotoxicity and associated risk factors, including those leading to fatal outcomes. This analysis provides valuable insights for ICI pharmacovigilance studies.

Materials and methods

Data acquisition and de-duplication

All data in ASCII format from the first quarter of 2004 to the third quarter of 2024 were downloaded from the FAERS database. Data cleaning followed the FDA’s official guidelines. AE reports were sorted by CASEID, FDA_DT, and PRIMARYID fields in the DEMO table. For reports with the same CASEID, the report with the latest FDA_DT value was retained, and for reports with the identical CASEID and FDA_DT, the report with the highest PRIMARYID value was selected. After the data de-duplication, the reports were removed based on the CASEID of the deletion report lists. It is important to note that all patient information within the FAERS database is anonymized, ensuring individuals cannot be identified through specific personal details. Consequently, obtaining informed consent and ethical approval for this study was not required.

AE signal detection and analysis

AE reports for FDA-approved ICIs were extracted for the first quarter of 2014 through the third quarter of 2024, focusing on “major suspect” drugs. These included six PD-1 inhibitors (nivolumab, pembrolizumab, cemiplimab, dostarlimab, toripalimab, tislelizumab), three PD-L1 inhibitors (atezolizumab, durvalumab, avelumab), and two CTLA-4 inhibitors (ipilimumab, tremelimumab), identified by the DRUGNAME and PROD_AI3 fields (Table S1).

Hepatotoxicity AEs were determined using the “Hepatobiliary Diseases” High-Level Group Terminology (HLGT) and the Medical Dictionary for Regulatory Activities (MedDRA) Preferred Terminology (PT) in the System Organ Classification (SOC; Table S2). The control group consisted of all other non-ICI drugs in the FAERS database. Disproportionality analyses were performed using Reporting Odds Ratio (ROR), Reporting Proportion Ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPN), and Multi-Item Gamma Poisson Shrinker (MGPS) methods to detect potential AE signals, with larger values indicating a stronger correlation between the drug and the AE.^19,20 The calculation formula is shown in Table 1. The results were evaluated in conjunction with the reports from the VigiAccess database, literature review, and case reports.

Table 1.

Two-by-two contingency table for combinations of ICIs and AE and the framework for calculating disproportionality.

Item	Target AE	Other AE	Total
Target drug	a	b	a + b
Other drug	c	d	c + d
Total	a + c	b + d	N = a + b + c + d
Methods		Calculation formulas	Thresholds
ROR		$R O R = \frac{(a / c)}{(b / d)} = \frac{a d}{b c}$ $S E (l n R O R) = \sqrt{(\frac{1}{a} + \frac{1}{b} + \frac{1}{c} + \frac{1}{d})}$ $95 % C I = e^{l n (R O R) \pm 1.96 \sqrt{(\frac{1}{a} + \frac{1}{b} + \frac{1}{c} + \frac{1}{d})}}$	a ⩾ 3 and 95% CI (lower limit) > 1 suggests that 1 positive signal was generated.
PRR		$P R R = \frac{a / (a + b)}{c / (c + d)}$ $S E (l n P R R) = \sqrt{\frac{1}{a} - \frac{1}{a + b} + \frac{1}{c} - \frac{1}{c + d}}$ $95 % C I = e^{l n (P R R) \pm 1.96 \sqrt{\frac{1}{a} - \frac{1}{a + b} + \frac{1}{c} - \frac{1}{c + d}}}$ $x^{2} = \frac{{(a d - b c)}^{2} (a + b + c + d)}{(a + b) (a + c) (c + d) (b + d)}$	a ⩾ 3 and 95% CI (lower limit) > 1 suggests that 1 positive signal was generated.
BCPNN		$I C = l o g_{2} \frac{p (x, y)}{p (x) p (y)} = l o g_{2} \frac{a (a + b + c + d)}{(a + b) (a + c)}$ $E (I C) = l o g_{2} \frac{(a + γ 11) (a + b + c + d + α) (a + b + c + d + β)}{(a + b + c + d + γ) (a + b + α 1) (a + c + β 1)}$ $V (I C) = \frac{1}{{(l n 2)}^{2}} {\begin{cases} [\frac{(a + b + c + d) - a + γ - γ 11}{(a + γ 11) (1 + a + b + c + d + γ)}] \\ + [\frac{(a + b + c + d) - (a + b) + α - α 1}{(a + b + α 1) (1 + a + b + c + d + α)}] \\ + [\frac{(a + b + c + d) - (a + b) + β - β 1}{(a + b + β 1) (1 + a + b + c + d + β)}] \end{cases}}$ $γ = γ 11 \frac{(a + b + c + d + α) (a + b + c + d + β)}{(a + b + α 1) (a + c + β 1)}$ $I C - 2 S D = E (I C) - 2 \sqrt{V (I C)}$ where α1 = β1 = 1, α = β = 2, γ11 = 1	95% CI (lower limit) > 0 suggests generation of 1 positive signal
MGPS		$E B G M = \frac{a (a + b + c + d)}{(a + c) (a + b)}$ $95 % C I = e^{l n (E B G M) \pm 1.96 \sqrt{(\frac{1}{a} + \frac{1}{b} + \frac{1}{c} + \frac{1}{d})}}$	95% CI (lower limit) > 2 suggests generation of 1 positive signal

a: number of cases of specific adverse events caused by the target drug; b: number of cases of adverse events other than specific adverse events caused by the target drug; c: number of cases of specific adverse events caused by medicines other than the target drug; d: number of cases of adverse events other than specific adverse events caused by medicines other than the target drug; N: total number of adverse events included in the analyzed data.

AR: Adverse event; ROR: Reported odds ratio; PRR: Proportional reporting ratio; BCPNN: bayesian confidence propagation neural network; MGPS: multi-item gamma poisson shrinker; CI: confidence interval.

Statistical analysis of hepatotoxic AE occurrence times, after excluding outliers, included plotting cumulative distribution curves and comparing median occurrence times across different ICIs. Outliers and missing data for gender, age, and body weight were excluded before annotating common concomitant medications (chemotherapeutic agents, targeted therapeutic agents, proton pump inhibitors (PPIs), aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs), metformin and other oral antidiabetics, glucagon-like peptide-1 (GLP-1) receptor agonists, insulin, antibiotics, immune-enhancers, immunosuppressants and glucocorticoids, statins, angiotensin converting enzyme inhibitor(ACEIs)/angiotensin receptor blockers (ARBs; Tables S3–S5)). Univariate and multivariate logistic regression analyses were conducted to explore factors influencing hepatotoxic AEs and fatal hepatotoxic AEs induced by ICIs.

Statistical analysis

Statistical analyses and figure illustrations were performed using R version 4.4.0. Descriptive analysis summarized the demographic and clinical characteristics of AE cases, with categorical variables as frequencies (percentages) and continuous variables as mean (SD) or median (IQR). The time to hepatotoxic AE onset was calculated from the medication start date to the AE onset date.

The Kruskal-Wallis test compared the median time to hepatotoxic AE onset across ICIs, with p < 0.05 indicating significance. Significant results prompted Dunn’s test for pairwise comparisons, adjusting the p value threshold to <0.05/number of two-by-two comparisons. Univariate and multivariate logistic regressions identified risk factors for hepatotoxic AEs and fatal outcomes, with p < 0.05 as the significance level.

Results

Demographic and clinical characteristics of AE reports caused by ICIs

This study processed AE reports related to ICIs by deduplicating and screening, reducing the DEMO dataset from 21,964,449 to 18,274,926 records. The analysis focused on 14,275,339 records since 2014, identifying 162,122 cases with 431,389 AE reports primarily attributed to ICIs from January 2014 to September 2024 (Figure 1).

Figure 1.

Flowchart of data processing for reports of hepatotoxic AEs caused by ICIs in the FAERS database.

As shown in Table 2, hepatotoxic AEs were noted in 12,891 (7.95%) of these cases, with 14,657 hepatotoxicity-related AEs reported. Demographically, males comprised 55.36% of hepatotoxic AE cases compared to 53.23% overall. Patients over 65 years old represented the largest group, especially in hepatotoxic AE cases at 44.53%. The mean weight for all cases was 71.25 ± 21.22 kg, slightly higher than hepatotoxic AE cases at 67.74 ± 19.59 kg. Physicians reported most cases, particularly 58.95% of hepatotoxic AEs, while consumer reports were less common. PD-1 inhibitors were the most used drugs but slightly less so in hepatotoxic AE cases. PD-L1 inhibitors showed a slightly higher rate in hepatotoxic AE cases. CTLA-4 inhibitors were more prevalent in hepatotoxic AE cases. Japan had the highest proportion of hepatotoxic AE reports at 40.88%, whereas the U.S. had fewer hepatotoxic AE cases despite having the most reports overall. About 22.09% of hepatotoxic AE cases resulted in death, slightly lower than the overall death rate of 25.16%. PD-1 inhibitors had the highest proportion of lethal hepatotoxicity within overall hepatotoxicity (22.57%), with notable differences between drugs (e.g. 26.07% for nivolumab vs 2.72% for tislelizumab). The lethal risk for PD-L1 inhibitors was similar to that for PD-1 inhibitors (22.11%), but the variation between atezolizumab (22.99%) and durvalumab (19.84%) were less pronounced. The lethal risk was slightly lower for CTLA-4 inhibitors (19.23%), but significantly higher for tremelimumab (30.00%).

Table 2.

Characteristics of the patients with immune checkpoint inhibitor-related adverse events in the FAERS database.

Indicator	Total cases (N = 162,122)	Cases with Hepatotoxic AEs (N = 12,891)	Cases without hepatotoxic AEs (N = 149,231)
Sex (n (%))
Female	54,942 (33.89)	4318 (33.5)	50,624 (33.92)
Male	86,290 (53.23)	7137 (55.36)	79,153 (53.04)
Unknown	20,890 (12.89)	1436 (11.14)	19,454 (13.04)
Age group (n (%))
<18 years	373 (0.23)	31 (0.24)	342 (0.23)
18–64 years	47,453 (29.27)	4547 (35.27)	42,906 (28.75)
⩾65 years	63,920 (39.43)	5741 (44.53)	58,179 (38.99)
Unknown	50,376 (31.07)	2572 (19.95)	47,804 (32.03)
Age (year)
N (missing)	50,376	2572	47,804
Mean (SD)	64.93 (12.85)	64.47 (12.89)	64.98 (12.84)
Median (Q1,Q3)	67.00 (58.00,74.00)	66.00 (57.00,73.61)	67.00 (58.00, 74.00)
Weight (kg)
N (missing)	117,652	8508	109,144
Mean (SD)	71.25 (21.22)	67.74 (19.59)	71.63 (21.36)
Median (Q1, Q3)	68.03 (57.00, 81.65)	65.00 (54.62, 67.74)	69.00 (57.50, 82.00)
Reporting year (n (%))
2014	1799 (1.11)	71 (0.55)	1728 (1.16)
2015	4497 (2.77)	228 (1.77)	4269 (2.86)
2016	9288 (5.73)	406 (3.15)	8882 (5.95)
2017	13,336 (8.23)	690 (5.35)	12,646 (8.47)
2018	16,007 (9.87)	1087 (8.43)	14,920 (10)
2019	18,380 (11.34)	1265 (9.81)	17,115 (11.47)
2020	17,004 (10.49)	1383 (10.73)	15,621 (10.47)
2021	18,981 (11.71)	1750 (13.58)	17,231 (11.55)
2022	21,469 (13.24)	1836 (14.24)	19,633 (13.16)
2023	21,325 (13.15)	2063 (16)	19,262 (12.91)
2024	20,036 (12.36)	2112 (16.38)	17,924 (12.01)
Type of reporter (n (%))
Physician	69,248 (42.71)	7599 (58.95)	61,649 (41.31)
Pharmacist	32,991 (20.35)	2499 (19.39)	30,492 (20.43)
Other health-professional	18,334 (11.31)	969 (7.52)	17,365 (11.64)
Lawyer	56 (0.03)	6 (0.05)	50 (0.03)
Consumer	40,033 (24.69)	1757 (13.63)	38,276 (25.65)
Unknown	1460 (0.9)	61 (0.47)	1399 (0.94)
Reporting countries (n (%))
United States	60,781 (37.49)	2036 (15.79)	58,745 (39.37)
Japan	35,435 (21.86)	5270 (40.88)	30,165 (20.21)
France	11,699 (7.22)	1049 (8.14)	10,650 (7.14)
China	7352 (4.53)	726 (5.63)	6626 (4.44)
Germany	5999 (3.7)	576 (4.47)	5423 (3.63)
Canada	3599 (2.22)	182 (1.41)	3417 (2.29)
Italy	3336 (2.06)	277 (2.15)	3059 (2.05)
United Kingdom	3082 (1.9)	267 (2.07)	2815 (1.89)
Australia	2872 (1.77)	186 (1.44)	2686 (1.8)
India	2610 (1.61)	114 (0.88)	2496 (1.67)
Other	25,357 (15.64)	2208 (17.13)	23,149 (15.51)
Outcome
Death (n (%))
Yes	40,784 (25.16)	2848 (22.09)	37,936 (25.42)
No	121,338 (74.84)	10,043 (77.91)	111,295 (74.58)
Life-threatening (n (%))
Yes	9922 (6.12)	1210 (9.39)	8712 (5.84)
No	152,200 (93.88)	11,681 (90.61)	140,519 (94.16)
Caused/prolonged hospitalization (n (%))
Yes	63,743 (39.32)	6932 (53.77)	56,811 (38.07)
No	98,379 (60.68)	5959 (46.23)	92,420 (61.93)
Disabling/incapacitating (n (%))
Yes	3315 (2.04)	321 (2.49)	2994 (2.01)
No	158,807 (97.96)	12,570 (97.51)	146,237 (97.99)
Congenital anomaly (n (%))
Yes	49 (0.03)	9 (0.07)	40 (0.03)
No	162,073 (99.97)	12,882 (99.93)	149,191 (99.97)
Therapeutic regimen
PD-1 (1890/8373, 22.57)	113,325 (69.9)	8373 (64.95)	104,952 (70.33)
nivolumab (1123/4308,26.07)	63,432 (39.13)	4308 (33.42)	59,124 (39.62)
pembrolizumab (743/3738,19.88)	45,768 (28.23)	3738 (29)	42,030 (28.16)
cemiplimab (15/125,12.00)	1752 (1.08)	125 (0.97)	1627 (1.09)
dostarlimab (3/24,12.50)	582 (0.36)	24 (0.19)	558 (0.37)
toripalimab (2/31,6.45)	441 (0.27)	31 (0.24)	410 (0.27)
tislelizumab (4/147, 2.72)	1350 (0.83)	147 (1.14)	1203 (0.81)
PD-L1 (685/3098,22.11)	34,063 (21.01)	3098 (24.03)	30,965 (20.75)
atezolizumab (520/2262,22.99)	21,129 (13.03)	2262 (17.55)	18,867 (12.64)
durvalumab (145/731,19.84)	10,882 (6.71)	731 (5.67)	10,151 (6.8)
avelumab (20/105,19.05)	2052 (1.27)	105 (0.81)	1947 (1.3)
CTLA-4 (273/1420,19.23)	14,734 (9.09)	1420 (11.02)	13,314 (8.92)
ipilimumab (264/1390,18.99)	14,563 (8.98)	1390 (10.78)	13,173 (8.83)
tremelimumab (9/30,30.00)	171 (0.11)	30 (0.23)	141 (0.09)

Hepatotoxic AEs and total AEs caused by ICIs from 2014 to 2024

As illustrated in Figure 2(a), the proportion of cases with hepatotoxic AEs has risen over time, increasing from 3.9% to 10.5%. In the first three quarters of 2024 alone, 2112 hepatotoxic AE cases were reported, accounting for 10.5%, marking a significant rise compared to the previous decade.

Figure 2.

Distribution of hepatotoxic AEs and non-hepatotoxic AEs caused by ICIs. (a) Distribution of hepatotoxic AEs and non-hepatotoxic AEs caused by ICIs between 2014 and 2024. (b) Distribution of hepatotoxic AEs and non-hepatotoxic AEs caused by various types of ICIs.

Figure 2(b) shows variability in hepatotoxic AE proportions by ICI type. Among PD-1 inhibitor cases (total 104,952), 7.4% (8373 cases) involved hepatotoxic AEs. For PD-L1 inhibitors (total 30,965 cases), this figure was 9.1% (3098 cases). CTLA-4 inhibitors had 9.6% (1420 out of 13,314 cases) associated with hepatotoxic AEs. The reporting rate of hepatotoxic AEs across 11 ICIs ranged from 4.1% to 17.5%, with tremelimumab showing the highest rate at 17.5%. Most drugs maintained a lower incidence, typically under 10%. According to the VigiAccess database, the percentage of reported hepatobiliary disorders varied among different ICIs, with dostarlimab having the highest percentage at 13% and toripalimab having the second highest percentage at 8%. The percentage of reports for other drugs varied between 3% and 6%.

The occurrence of hepatotoxic AEs caused by ICIs

The overall ROR for ICIs and hepatotoxic AEs was 4.38 (95% CI: 4.31–4.46), indicating a significant correlation. Tremelimumab had the highest ROR at 23.34 (95% CI: 16.13–33.77), followed by toripalimab at 15.43 (95% CI: 10.78–22.08). Most other ICIs showed RORs between 3.66 and 7.44, all with positive signals for hepatotoxic AEs. Similar trends were observed using PRR, BCPNN, and MGPS methods (Table S6).

Among tumor types, hepatobiliary tumors had the highest incidence of hepatotoxic AEs at 20.7%, followed by skin tumors at 11.1%, and renal/urinary tract tumors at 10.2% (Figure 3(a)). Hepatotoxic AEs were mainly classified as hepatobiliary disorders (84.0%), with immune system disorders and injuries/surgical complications following (Figure 3(b)). Top hepatotoxic AEs in frequency included hepatic function abnormalities, hepatitis, liver disorders, and immune-mediated hepatic disorders, most commonly associated with nivolumab, pembrolizumab, atezolizumab, durvalumab, and ipilimumab (Figure 3(c)).

Figure 3.

Analysis of hepatotoxic AEs in the FAERS database. (a) Number and percentage of hepatotoxic AEs and total AEs for tumors of different organ systems. (b) Number and percentage of hepatotoxic AEs at different SOC levels. (c) Distribution of positive hepatotoxic AEs across different ICIs by frequency. (d) Distribution of positive hepatotoxic AEs across different ICIs by ROR.

Specific AEs like Stauffer’s syndrome (strongly linked to nivolumab) and portal vein embolism (linked to atezolizumab) showed strong ROR signals. Hepatic rupture was notably associated with pembrolizumab and atezolizumab. Certain rare events such as biloma did not show positive ROR results across ICIs. Nivolumab also showed strong associations with acute hepatitis B, viral hepatitis, and liver transplant rejection, while pembrolizumab was linked to hepatic calcification, congestive liver disease, and hepatic cysts. Tislelizumab was associated with acute liver failure, and atezolizumab with various conditions including esophageal variceal bleeding and chronic hepatitis B (Figure 3(d)).

Figure S1 highlights that atezolizumab had the largest variety of adverse events reported, with autoimmune hepatitis, hepatitis, immune-mediated hepatic disorder, and immune-mediated hepatitis among the most frequently intersecting AEs across multiple ICIs.

Time to onset of hepatotoxic AEs due to ICIs

After excluding outliers based on the interquartile range, statistical analysis revealed a median time of 92 days for the occurrence of hepatotoxic AEs caused by ICIs. As shown in Figure 4(a), most hepatotoxic AEs (9121 cases) occurred within the first 90 days post-administration, with occurrences gradually decreasing over time: 4103 cases between 90 and 180 days, 2290 cases between 180 and 270 days, and 1607 cases between 270 and 360 days. Notably, hepatotoxic AEs continued to occur even more than a year after treatment initiation.

Figure 4.

Analysis of time to occurrence of hepatotoxic AEs in the FAERS database. (a) Distribution of the time to occurrence of hepatotoxic AEs. Median time to occurrence of hepatotoxic AEs: 92 days. (b) Median time to occurrence of hepatotoxic AEs for PD-1, PD-L1, and CTLA-4 inhibitors. (c) Median time to occurrence of hepatotoxic AEs for 11 ICIs. (d) Cumulative probability curves for the occurrence of hepatotoxic AEs for PD-1, PD-L1, and CTLA-4 inhibitors. (e) Cumulative probability curves for the occurrence of hepatotoxic AEs for the 11 ICIs.

Figure 4(b) indicates that PD-1 inhibitors had the shortest median time to hepatotoxic AE onset at 91 days, compared to 94 days for PD-L1 inhibitors and 96 days for CTLA-4 inhibitors. Among 11 ICIs (Figure 4(c)), toripalimab showed the quickest onset at 0 days, while dostarlimab had the longest at 100 days, with other drugs ranging from 73 to 99 days. Most ICIs had a median time above 90 days.

Log-rank and Kruskal-Wallis tests confirmed significant differences in the median time to hepatotoxic AE occurrence among ICI types (p < 0.05; Figure 4(d) and (e)). Pairwise comparisons revealed that PD-1 inhibitors had a significantly shorter median time to AE onset compared to CTLA-4 and PD-L1 inhibitors (p < 0.0167). Additionally, toripalimab had a significantly shorter median time to AE onset compared to the other 10 ICIs (p < 0.00091).

Analysis of factors influencing hepatotoxicity due to ICIs

To ensure data integrity, records with missing values and outliers (age > 120 years, weight > 400 kg) were excluded. Univariate logistic regression analysis identified several factors as statistically significant for the occurrence of hepatotoxic AEs: hepatobiliary cancer, body weight, ICI therapeutic regimen, targeted therapy, chemotherapy, antibiotics, aspirin, PPIs, statins, other oral hypoglycemic agents, immunopotentiators, and immunosuppressants (p < 0.05). Multivariate logistic regression further determined that hepatobiliary cancer, body weight, ICI therapeutic regimen, targeted therapy, antibiotics, PPIs, other oral hypoglycemic agents, immunosuppressants, and immunopotentiators were independent predictors (p < 0.05). Specifically, body weight, antibiotics, and immunopotentiators acted as protective factors against hepatotoxic AEs, while hepatobiliary cancer, combination ICI therapy, targeted therapy drugs, PPIs, other oral hypoglycemic agents, and immunosuppressants were risk factors (Figure 5(a) and (b)).

Figure 5.

Analysis of factors influencing the occurrence of hepatotoxic AEs in the FAERS database. (a–b) Univariate and multivariate logistic regression analysis of factors influencing the occurrence of hepatotoxic AEs. (c–d) Univariate and multivariate logistic regression analysis of factors influencing the occurrence of fatal hepatotoxic AEs.

For fatal hepatotoxicity, univariate analysis found that hepatobiliary cancer, other AEs, age, gender, chemotherapy, other oral hypoglycemic drugs, and immunosuppressants were significant (p < 0.05). Multivariate analysis confirmed that hepatobiliary cancer, other AEs, gender, and immunosuppressants remained significant (p < 0.05). Hepatobiliary cancer, accompanying other AEs, male gender, and immunosuppressants were independent risk factors for fatal outcomes (Figure 5(c) and (d)). Figure S2 demonstrated the effect of baseline information and medication use on the occurrence of AE in hepatic function abnormal and immune-mediated hepatitis at the PT level.

Discussion

The clinical application of ICIs has highlighted the necessity for identifying and managing hepatotoxic AEs.^21–25 This study is a retrospective, real-world pharmacovigilance analysis of large-scale cases with ICIs-associated hepatotoxic AEs based on the FAERS database. It is the first to comprehensively investigate the impact of combination medications on both overall and fatal hepatotoxic AEs. The extensive data available in FAERS provide robust statistical support and encompass diverse populations, facilitating the detection of new or rare AE signals and drug safety profiles.

The precise mechanism behind ICI-induced hepatotoxicity remains unclear but likely involves disrupting liver-specific immune tolerance. Molecules like PD-L1 expressed on hepatic nonparenchymal cells and CTLA-4 on regulatory T cells (Tregs) help suppress effector T cell activity. ICIs block these negative regulators, potentially triggering autoimmune responses that can cause liver injury.¹¹ While PD-1 and PD-L1 inhibitors primarily suppress T cell function in peripheral tissues, such as the liver, CTLA-4’s effects are concentrated on the initial phase of the immune response within lymph nodes. This distinction leads to broader activation of the immune system by CTLA-4 inhibition. Blocking CTLA-4 not only activates effector T cells (Teffs) but also diminishes the immunosuppressive activity of Tregs, resulting in systemic immune activation and autoimmune reactions. Consequently, CTLA-4 blockade is associated with extensive immune activation and a higher incidence of side effects. In contrast, PD-1/PD-L1 inhibition targets the activation of T cells specifically within the tumor microenvironment, leading to a more confined range of side effects.²⁶

Hepatotoxic AEs appear to be more prevalent among males and older populations, although this finding is not consistent across all studies.^3,27 The influence of these demographic factors on the risk of hepatotoxicity warrants further investigation in future studies. Additionally, patients with hepatobiliary tumors are at a higher risk for developing hepatotoxic AEs, likely due to compromised liver function resulting from liver metastases or tumor infiltration.²⁸ These AEs can lead to severe outcomes, with approximately 22.09% being fatal. Notably, hepatitis induced by ICIs plays a significant role in these fatal AEs, particularly within groups receiving combination therapies.²⁹ Therefore, it is essential to focus on high-risk populations to facilitate early identification and appropriate intervention.

All ICIs exhibited signals for hepatotoxic AEs, with tremelimumab showing the strongest signal and nivolumab the weakest. The proportion of hepatotoxic AE reports has increased in the FAERS database over time. Similarly, the incidence of immune-related hepatitis in the VigiBase database has also shown a significant rise over time, suggesting that ICI-induced hepatotoxic AEs are becoming more pronounced.³⁰ This chronological increase is closely associated with the expansion of ICI indications and the growing popularity of combination therapies. CTLA-4 inhibitors were linked to the highest rate of hepatotoxic AEs, whereas PD-1 inhibitors had the lowest rates. The results from a clinical study evaluating ipilimumab in the treatment of unresectable or metastatic melanoma indicated that hepatotoxicity induced by CTLA-4 inhibitors may be dose-related.³¹ The notably low fatal hepatotoxicity rate for tislelizumab (2.72%) could reflect differences in patient demographics (e.g, a higher proportion of Asian patients) or drug structural optimizations (e.g, Fc segment modification aimed at reducing intrahepatic T-cell overactivation).^32,33 The elevated risk for tremelimumab can be attributed to factors such as small sample size (n = 30), mechanism-specificity (direct activation of Teffs instead of Tregs), and biased selection of indications (for HCC patients).^34,35 Furthermore, hepatobiliary disorders accounted for 84.0% of all hepatotoxic AE reports, indicating a primary focus on liver-related issues but also suggesting potential impacts on other organ systems. Therefore, future multi-omics studies (e.g, hepatic-intestinal axis microbial metabolite assays) are needed to reveal mechanisms of cross-organ toxicity.

Stauffer syndrome is a rare paraneoplastic phenomenon associated with renal cell carcinoma (RCC) that manifests as abnormal liver function. A notable case report described the progression of liver injury in an RCC patient undergoing combination immunotherapy with ipilimumab and nivolumab, which was suspected to be related to Stauffer syndrome. This case highlights the importance of closely monitoring liver function in RCC patients, particularly those receiving immunotherapy.³⁶ Portal vein thrombosis linked to atezolizumab treatment has not been previously reported, suggesting this could represent a rare AE that warrants attention in future pharmacovigilance efforts. Furthermore, instances of hepatic tumor rupture bleeding following pembrolizumab therapy or its combination with bevacizumab have been documented, underscoring the necessity for continuous monitoring of patients with liver tumors to prevent such complications after ICI treatments.^37,38 The use of ICIs in organ transplant recipients is limited due to studies indicating an increased risk of organ rejection, estimated between 36% and 54%. An illustrative case involved a patient who experienced severe graft rejection after receiving nivolumab for recurrent HCC following liver transplantation.³⁹ Therefore, when considering ICI therapy for cancer patients with a history of organ transplantation, it is essential to remain vigilant for signs of immune rejection and carefully balance the potential risks against therapeutic benefits.

Understanding the timing and cumulative probability of hepatotoxic AEs following ICI therapy is crucial for grasping the temporal distribution of these AEs. Most hepatotoxic AEs occur within the first 3 months after initiating ICI treatment, consistent with previous findings. However, the risk persists beyond this period, necessitating a dynamic monitoring approach for these AEs.⁴⁰ PD-1 inhibitors, including toripalimab, tislelizumab, and pembrolizumab, tend to induce hepatotoxic AEs earlier than PD-L1 and CTLA-4 inhibitors, demanding increased vigilance during the initial treatment phases. For example, the median time to onset of hepatotoxicity caused by toripalimab was reported as 1.1 months (range: 0.1–7.9 months), which is notably shorter than what is indicated in drug inserts. This discrepancy could be attributed to the limited data available in the FAERS database, given toripalimab’s relatively recent introduction to the market. This situation underscores the differences between real-world clinical practice and controlled clinical trial settings.

This study compared the risk of all-grade hepatotoxicity between monotherapy and combination regimens involving ICIs. It found that combination immunotherapy demonstrated a higher risk of all-grade hepatotoxicity, consistent with the epidemiological profile of irAEs.⁴¹ Notably, there was no significant difference in the risk of lethal hepatotoxicity across all ICI treatment regimens. However, in patients with advanced melanoma, the combination of nivolumab with ipilimumab resulted in a relatively high incidence (17%) of grade 3/4 hepatotoxic AEs.⁴² Additionally, adding targeted therapy to immunotherapy increased the likelihood of hepatotoxic AEs. Despite this, neither targeted therapy nor chemotherapy significantly affected the occurrence of lethal hepatotoxic AEs. Meta-analysis revealed that liver toxicity occurred in 5.77% of ICI monotherapy cases, 13.62% when ICIs were combined with chemotherapy, and 16.07% when ICIs were used alongside targeted therapy drugs.⁴⁰ Age and gender had minimal influence on the risk of all-grade hepatotoxic AEs, but older males faced a higher risk of fatal hepatotoxic AEs. Patient body weight may influence the dosage of the ICI regimen, potentially affecting the occurrence of hepatotoxicity.^31,42,43 Fixed-dose ICIs might provide a safer option for patients with higher body weights.⁴⁴ These findings underscore the importance of considering both the type of regimen and patient-specific factors when assessing the risk of hepatotoxic AEs associated with ICIs.

Antibiotics can influence the risk of hepatotoxic AE by altering gut flora, and studies have shown that they might increase the risk of hepatobiliary irAEs in lung cancer patients.^45–47 PPIs may increase the risk of hepatotoxic AEs, such as cholestasis.⁴⁸ Both antibiotics and PPIs could potentially diminish the efficacy of ICIs.⁴⁹ In a pharmacovigilance analysis, aspirin use in cancer patients treated with ICIs did not significant affect the risk of hepatotoxic AEs, which is similar to our results.⁵⁰ Statins do not significantly impact the occurrence of hepatotoxic AE when used alongside immunotherapy but may improve ICI outcomes by modulating inflammation and the immune microenvironment.^51–53 This suggests that combing ICIs with statins may offer a manageable safety profile for tumor patients with concomitant liver disease. Hepatotoxic AEs might occur in diabetic patients who are using hypoglycemic agents along with other medications.⁵⁴ Our study found that the combined use of oral hypoglycemic agents increased the risk of hepatotoxic AEs. Immune enhancers appear to reduce the incidence of hepatotoxic AEs, whereas immunosuppressants could increase the risk or lead to fatal hepatotoxic AEs, possibly influenced by the patient’s underlying disease conditions. Tumor patients with autoimmune diseases or those who have undergone organ transplants may face higher irAE risks during ICI treatment. Although irAEs can indicate effective immunotherapy responses and be managed through the use of immunosuppressive agents.^55–57

The study recognizes several limitations inherent in the FAERS database, primarily due to its reliance on spontaneous reporting. This approach can introduce various biases, including underreporting, delayed reports, duplicate entries, and inaccurate information. Even when employing large datasets and conventional statistical methods, challenges such as handling missing values and outliers may still result in biases. The absence of baseline patient information and detailed specifics about drug treatments can further complicate causal inference. Given these limitations, the results from the FAERS database are best suited for monitoring adverse reaction signals and generating hypotheses rather than providing precise estimates of actual AE frequencies. These findings require external validation to confirm their reliability. Additionally, variations among reporters can affect the consistency and comparability of the data reported, which poses another layer of complexity in interpreting the results accurately.

In future studies, it is recommended to include patients undergoing immunotherapy in combination with localized treatments such as radiotherapy and interventional therapy. Engaging in real-world studies will assist in uncovering previously unknown AE signals, thereby offering a more comprehensive understanding of AE profiles associated with ICIs. This broader perspective can significantly contribute to enhancing clinical AE management strategies and guiding the selection of therapeutic regimens.

Conclusions

Although hepatotoxic AEs from ICIs are relatively infrequent, a significant proportion of these cases can be fatal, underscoring the importance of their identification and management. Patients with hepatobiliary tumors should be especially vigilant. The incidence of these AEs has increased, likely due to the expanded variety and use of ICIs in combination therapies. Abnormal hepatic function is the most common type of hepatotoxic AE, though rare conditions like Stauffer syndrome also warrant attention. Most AEs occur within the first 3 months of treatment, with PD-1 inhibitors notably showing the earliest onset of hepatotoxicity. While early-onset hepatotoxic AEs are critical, delayed occurrences should also be closely monitored. Combination therapies may elevate the risk of hepatotoxic AEs, particularly concerning the heightened risk of lethal AEs in elderly males and those already experiencing other AEs. By translating pharmacovigilance data into actionable clinical knowledge, this study bridges the gap from signal detection to risk prevention and control. It underscores the necessity for continuous monitoring and tailored intervention strategies to mitigate the risks associated with ICI therapies.

Supplemental Material

sj-docx-1-iji-10.1177_03946320251343943 – Supplemental material for Evaluation of immune checkpoint inhibitor-associated hepatotoxic adverse events: A pharmacovigilance analysis based on the FAERS database

Supplemental material, sj-docx-1-iji-10.1177_03946320251343943 for Evaluation of immune checkpoint inhibitor-associated hepatotoxic adverse events: A pharmacovigilance analysis based on the FAERS database by Zhihong Chen, Junwei Zhang, Lei Zhang, Yaoge Liu, Ting Zhang, Xinting Sang, Yiyao Xu and Xin Lu in International Journal of Immunopathology and Pharmacology

Footnotes

Acknowledgements

The authors acknowledged the data support from the FAERS database.

Author contributions

X.L. and Y.X. conceived and designed the research. Material preparation and data collection were performed by Z.C., J.Z., and L.Z. Data analysis was performed by Z.C., Y.L., T.Z., and X.S. The manuscript was drafted by Z.C. and revised by all authors. All authors read and approved the final version of the manuscript.

Data availability

Data will be made available on request.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors acknowledged the grants from the National High Level Hospital Clinical Research Funding under Grant [number 2022-PUMCH-C-049, 2022-PUMCH-A-237].

ORCID iD

Xin Lu

Supplemental material

Supplemental material for this article is available online.

References

Karati

Kumar

(2024) Molecular insight into the apoptotic mechanism of cancer cells: An explicative review. Current Molecular Pharmacology 17: e18761429273223.

Liu

Xie

Yang

, et al. (2023) HPVTIMER: A shiny web application for tumor immune estimation in human papillomavirus-associated cancers. iMeta 2(3): e130.

Peng

Lin

Jiang

, et al. (2024) CTLs heterogeneity and plasticity: Implications for cancer immunotherapy. Molecular Cancer 23(1): 58.

Manohar

(2024) At the crossroads of TNF α signaling and cancer. Current Molecular Pharmacology 17: e060923220758.

Cameron

Whiteside

Perry

(2011) Ipilimumab: First global approval. Drugs 71(8): 1093–1104.

Kong

Zhang

Chen

, et al. (2024) Immune checkpoint inhibitors: Breakthroughs in cancer treatment. Cancer Biology & Medicine 21(6): 451–472.

Wojtukiewicz

Rek

Karpowicz

, et al. (2021) Inhibitors of immune checkpoints—PD-1, PD-L1, CTLA-4—new opportunities for cancer patients and a new challenge for internists and general practitioners. Cancer Metastasis Reviews 40(3): 949–982.

Gao

Bai

Lin

, et al. (2023) Gamma delta T-cell-based immune checkpoint therapy: Attractive candidate for antitumor treatment. Molecular Cancer 22(1): 31.

Glaviano

Lau

HSH

Carter

, et al. (2025) Harnessing the tumor microenvironment: Targeted cancer therapies through modulation of epithelial-mesenchymal transition. Journal of Hematology & Oncology 18(1): 6.

10.

Ledford

Else

Warren

(2018) Cancer immunologists scoop medicine Nobel prize. Nature 562(7725): 20–21.

11.

Shojaie

Ali

Iorga

, et al. (2021) Mechanisms of immune checkpoint inhibitor-mediated liver injury. Acta Pharmaceutica Sinica B 11(12): 3727–3739.

12.

Mangani

Yang

Anderson

(2023) Learning from the nexus of autoimmunity and cancer. Immunity 56(2): 256–271.

13.

Chen

, et al. (2018) Comparative safety of immune checkpoint inhibitors in cancer: Systematic review and network meta-analysis. BMJ 363; k4226.

14.

Zhong

Chen

, et al. (2021) Immune-related adverse events: Promising predictors for efficacy of immune checkpoint inhibitors. Cancer Immunology Immunotherapy 70(9): 2559–2576.

15.

Zhang

Zhou

, et al. (2018) Immune-related adverse events from combination immunotherapy in cancer patients: A comprehensive meta-analysis of randomized controlled trials. International Immunopharmacology 63: 292–298.

16.

Remash

Prince

McKenzie

, et al. (2021) Immune checkpoint inhibitor-related hepatotoxicity: A review. World Journal of Gastroenterology 27(32): 5376–5391.

17.

Parlati

Sakka

Retbi

, et al. (2023) Burden of grade 3 or 4 liver injury associated with immune checkpoint inhibitors. JHEP Reports 5(12): 100880.

18.

Dhodapkar

Shi

Ramachandran

, et al. (2022) Characterization and corroboration of safety signals identified from the US Food and Drug Administration Adverse Event Reporting System, 2008-19: Cross sectional study. BMJ 379: e071752.

19.

Capece

Fischietti

Verzella

, et al. (2013) The inflammatory microenvironment in hepatocellular carcinoma: A pivotal role for tumor-associated macrophages. BioMed Research International 2013: 187204.

20.

Sakaeda

Tamon

Kadoyama

, et al. (2013) Data mining of the public version of the FDA Adverse Event Reporting System. International Journal of Medical Sciences 10(7): 796–803.

21.

Liu

, et al. (2023) Hepatotoxicity in immune checkpoint inhibitors: A pharmacovigilance study from 2014–2021. PLoS One 18(3): e0281983.

22.

Yan

Zhao

, et al. (2023) Hepatic failure associated with immune checkpoint inhibitors: An analysis of the Food and Drug Administration Adverse Event Reporting System database. Cancer Medicine 12(8): 9167–9174.

23.

Yang

Wang

Zhou

, et al. (2023) Hepatotoxicity associated with immune checkpoint inhibitors in clinical practice: A study leveraging data from the US Food and Drug Administration’s Adverse Event Reporting System. Clinical Therapeutics 45(2): 151–159.

24.

Liu

Zhang

, et al. (2024) Hepatitis-related adverse events associated with immune checkpoint inhibitors in cancer patients: An observational, retrospective, pharmacovigilance study using the FAERS database. Frontiers in Pharmacology 15: 1383212.

25.

Zhou

Zhang

, et al. (2024) Hepatitis associated with immune checkpoint inhibitors-based combinations of other therapies: A real-world pharmacovigilance analysis based on the FDA adverse event reporting system (FAERS) database. Cancer Immunology Immunotherapy 74(1): 25.

26.

Boutros

Tarhini

Routier

, et al. (2016) Safety profiles of anti-CTLA-4 and anti-PD-1 antibodies alone and in combination. Nature Reviews Clinical Oncology 13(8): 473–486.

27.

Jiang

, et al. (2024) Risk factors of immune-mediated hepatotoxicity induced by immune checkpoint inhibitors in cancer patients: A systematic review and meta-analysis. Current Oncology 31(11): 7129–7143.

28.

Pinter

Scheiner

Pinato

(2023) Immune checkpoint inhibitors in hepatocellular carcinoma: Emerging challenges in clinical practice. The Lancet Gastroenterology & Hepatology 8(8): 760–770.

29.

Wang

Salem

Cohen

, et al. (2018) Fatal toxic effects associated with immune checkpoint inhibitors: A systematic review and meta-analysis. JAMA Oncology 4(12): 1721.

30.

Vozy

De Martin

Johnson

, et al. (2019) Increased reporting of fatal hepatitis associated with immune checkpoint inhibitors. European Journal of Cancer 123: 112–115.

31.

Ascierto

Del Vecchio

Robert

, et al. (2017) Ipilimumab 10 mg/kg versus ipilimumab 3 mg/kg in patients with unresectable or metastatic melanoma: A randomised, double-blind, multicentre, phase 3 trial. The Lancet Oncology 18(5): 611–622.

32.

Desai

Deva

Lee

, et al. (2020) Phase IA/IB study of single-agent tislelizumab, an investigational anti-PD-1 antibody, in solid tumors. Journal for Immunotherapy of Cancer 8(1): e000453.

33.

Zhang

Geng

Hao

, et al. (2022) Tislelizumab: A modified anti-tumor programmed death receptor 1 antibody. Cancer Control 29: 10732748221111296.

34.

Sangro

Gomez-Martin

De La Mata

, et al. (2013) A clinical trial of CTLA-4 blockade with tremelimumab in patients with hepatocellular carcinoma and chronic hepatitis C. Journal of Hepatology 59(1): 81–88.

35.

Keam

(2023) Tremelimumab: First approval. Drugs 83(1): 93–102.

36.

Denaro

Garrone

Occelli

, et al. (2021) Unusual fatal outcome following administration of a combination of anti-PD1 and anti-CTLA4 in metastatic renal cell carcinoma: Liver toxicity case report and a literature review. European Journal of Case Reports in Internal Medicine 8: 002639.

37.

Igari

Yamamoto

Kikuchi

, et al. (2024) A case of ruptured hepatic metastases during pembrolizumab administration for cutaneous malignant melanoma. Anais Brasileiros de Dermatologia 99(3): 450–451.

38.

Park

Yoo

Kim

, et al. (2024) Fatal intratumoral hemorrhage in a patient with hepatocellular carcinoma following successful treatment with atezolizumab/bevacizumab: A case report. World Journal of Clinical Cases 12(22): 5177–5183.

39.

Gassmann

Weiler

Mertens

, et al. (2018) Liver allograft failure after nivolumab treatment—A case report with systematic literature research. Transplantation Direct 4(8): e376.

40.

Peeraphatdit

Wang

Odenwald

, et al. (2020) Hepatotoxicity from immune checkpoint inhibitors: A systematic review and management recommendation. Hepatology 72(1): 315–329.

41.

Zheng

Huang

Lin

, et al. (2023) Hepatotoxicity of immune checkpoint inhibitors: What is currently known. Hepatology Communications 7(3): e0063.

42.

Sznol

Ferrucci

Hogg

, et al. (2017) Pooled analysis safety profile of nivolumab and ipilimumab combination therapy in patients with advanced melanoma. Journal of Clinical Oncology 35(34): 3815–3822.

43.

Zhao

Suryawanshi

Hruska

, et al. (2017) Assessment of nivolumab benefit–risk profile of a 240-mg flat dose relative to a 3-mg/kg dosing regimen in patients with advanced tumors. Annals of Oncology 28(8): 2002–2008.

44.

Le Louedec

Leenhardt

Marin

, et al. (2020) Cancer immunotherapy dosing: A Pharmacokinetic/pharmacodynamic perspective. Vaccines 8(4): 632.

45.

Jing

Chen

, et al. (2022) Association of antibiotic treatment with immune-related adverse events in patients with cancer receiving immunotherapy. Journal for Immunotherapy of Cancer 10(1): e003779.

46.

Pinato

Howlett

Ottaviani

, et al. (2019) Association of prior antibiotic treatment with survival and response to immune checkpoint inhibitor therapy in patients with cancer. JAMA Oncology 5(12): 1774.

47.

Pinato

Mishra-Kalyani

, et al. (2023) Association between antibiotics and adverse oncological outcomes in patients receiving targeted or immune-based therapy for hepatocellular carcinoma. JHEP Reports 5(6): 100747.

48.

Zeng

Dai

Zhou

, et al. (2021) Hepatotoxicity-related adverse effects of proton pump inhibitors: A cross-sectional study of signal mining and analysis of the FDA Adverse Event Report System database. Frontiers in Medicine 8: 648164.

49.

Cortellini

Di Maio

Nigro

, et al. (2021) Differential influence of antibiotic therapy and other medications on oncological outcomes of patients with non-small cell lung cancer treated with first-line pembrolizumab versus cytotoxic chemotherapy. Journal for Immunotherapy of Cancer 9(4): e002421.

50.

Yang

Liu

, et al. (2023) The association between aspirin use and immune-related adverse events in specific cancer patients receiving ICIs therapy: Analysis of the FAERS database. Frontiers in Pharmacology 14: 1259628.

51.

Verma

Huang

Hutchinson

, et al. (2022) Statin use and severe acute liver injury among patients with elevated alanine aminotransferase. Clinical Epidemiology 14: 1535–1545.

52.

Zeng

Deng

Luo

, et al. (2024) Advances in statin adverse reactions and the potential mechanisms: A systematic review. Journal of Advanced Research. Epub ahead of print 14 December 2024. DOI: 10.1016/j.jare.2024.12.020.

53.

Wang

Liu

, et al. (2024) Statin therapy: A potential adjuvant to immunotherapies in hepatocellular carcinoma. Frontiers in Pharmacology 15: 1324140.

54.

Hua

Lam

, et al. (2024) The use of potentially interacting supplement–drug pairs in adults with type 2 diabetes: A large population-based cohort study in the UK Biobank. Diabetes Research and Clinical Practice 211: 111658.

55.

Haanen

Ernstoff

Wang

, et al. (2020) Autoimmune diseases and immune-checkpoint inhibitors for cancer therapy: Review of the literature and personalized risk-based prevention strategy. Annals of Oncology 31(6): 724–744.

56.

Han

Fitzgerald

Lee

, et al. (2022) Association between toxic effects and survival in patients with cancer and autoimmune disease treated with checkpoint inhibitor immunotherapy. JAMA Oncology 8(9): 1352.

57.

Ibis

Aliazis

Cao

, et al. (2023) Immune-related adverse effects of checkpoint immunotherapy and implications for the treatment of patients with cancer and autoimmune diseases. Frontiers in Immunology 14: 1197364.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.79 MB