📄 other

Interstitial Lung Disease Associated with Immune Checkpoint Inhibitors in Gastrointestinal Cancer Treatment: A Pharmacovigilance Analysis of the FDA Adverse Event Reporting System (FAERS) Database and EudraVigilance Database

This pharmacovigilance study analyzing FAERS and EudraVigilance data reveals that immune checkpoint inhibitor-induced interstitial lung disease in gastrointestinal cancer patients exhibits distinct, agent-specific risk profiles regarding gender, onset timing, and geographic variation, underscoring the need for risk-stratified monitoring to mitigate its high fatality rate.

Original authors: ruiying Chen, Pengkhun Nov, Jiqiang Li

Published 2026-07-25

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: ruiying Chen, Pengkhun Nov, Jiqiang Li

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine your body's immune system as a highly trained security team, constantly patrolling the neighborhood to spot and stop intruders like viruses or cancer cells. Sometimes, this team gets too sleepy and lets the bad guys slip by. Enter Immune Checkpoint Inhibitors (ICIs), which are like a loud, enthusiastic coach shouting, "Wake up! Go get 'em!" to the security team. These drugs have been a game-changer for treating many cancers, including those in the stomach and gut. However, just like a coach who yells too loud might accidentally scare the neighbors, these drugs can sometimes make the immune team attack the wrong targets. One of the most dangerous mix-ups is when the immune system starts attacking the lungs, causing a condition called Interstitial Lung Disease (ILD). It's like the security team accidentally setting off a fire alarm inside the house itself. While we know this happens in lung cancer patients, scientists have been wondering: does this same "friendly fire" happen in patients with stomach and gut cancers, and if so, who is most likely to get hurt?

A team of researchers from Zhujiang Hospital decided to investigate this by digging through a massive digital filing cabinet called the FAERS database, which collects millions of reports of side effects from drugs around the world. They also cross-checked their findings with a European database called EudraVigilance to make sure they weren't missing anything. They looked specifically at patients with gastrointestinal (digestive system) cancers who had been treated with these "wake-up call" drugs and had developed lung trouble.

Here is what they found in their report:

The "Who" and "What"
Out of the thousands of reports they sifted through, they identified 588 specific cases where patients with digestive cancers developed ILD after taking these drugs. The most common culprits were drugs named Nivolumab and Pembrolizumab. Interestingly, while men made up the majority of the reported cases (about 72%), the data suggested that women might actually be at a higher relative risk for certain specific drugs. It's like a storm that hits a city mostly, but the damage per person is actually worse in a specific neighborhood.

The "When" Matters
Timing was a huge clue. The researchers noticed that different drugs caused trouble at different times. One drug, Ipilimumab, was like a sprinter; it caused lung issues very quickly, mostly within the first 30 days of treatment. On the other hand, drugs like Nivolumab and Pembrolizumab were more like marathon runners; they could cause problems at any time, from the first month all the way to over a year later. This suggests that doctors can't just check the lungs once and forget about it; they need to keep watching for different reasons depending on which drug is being used.

The "Where" and "Who Else"
The study also found that age and location played a role. Most of the patients who got sick were older, over 65 years old. This makes sense, as older lungs might be more fragile. The reports came mostly from Japan, the United States, and China. The researchers noticed that the specific types of lung trouble reported varied by country. For instance, Japan reported a lot of a specific type of lung inflammation called "bronchiolitis" linked to Nivolumab, while the US saw different patterns. This hints that genetics or local medical practices might influence how these drugs affect people.

The Bottom Line
The study concludes that while these drugs are powerful tools against digestive cancers, they carry a real risk of lung damage that looks different depending on the patient's age, gender, the specific drug used, and even where they live. The researchers suggest that because the risk isn't the same for everyone, doctors need to be extra careful and monitor patients differently based on these factors. They also pointed out that this study has limits—it relies on voluntary reports, so we don't know the exact percentage of people who get sick, just the patterns of those who were reported. However, this "map" of risks helps doctors know where to look and when to be most alert, potentially saving lives by catching lung trouble early.

Technical Summary: Interstitial Lung Disease Associated with Immune Checkpoint Inhibitors in Gastrointestinal Cancer Treatment

Problem Statement
While immune checkpoint inhibitors (ICIs) have revolutionized the treatment of gastrointestinal (GI) malignancies, they carry the risk of immune-related adverse events (irAEs), most notably interstitial lung disease (ILD). Although ILD is well-characterized in non-small cell lung cancer (NSCLC) and melanoma, its specific clinical profile, risk patterns, and safety signals within the GI cancer population remain insufficiently explored. Existing clinical trial data are often limited by sample sizes and strict inclusion criteria, failing to capture the heterogeneity of real-world populations. Furthermore, the unique pathophysiology of GI tumors, potentially mediated by the "gut-lung axis," suggests that GI patients may face distinct ILD risk profiles compared to other tumor types. There is a critical need to evaluate real-world safety signals to guide individualized monitoring and risk management.

Methodology
This study conducted a retrospective pharmacovigilance analysis using two major spontaneous reporting databases: the FDA Adverse Event Reporting System (FAERS) and the EudraVigilance database.

Data Source: FAERS data covering the period from the first quarter of 2004 to the fourth quarter of 2023 (with the study abstract noting a 2015–2025 scope, though the methods specify extraction up to Q4 2023) were utilized.
Data Processing: Reports were preprocessed using SAS 9.4. Adverse events were mapped to the Medical Dictionary for Regulatory Activities (MedDRA) preferred terms and system organ classes.
Statistical Analysis: The study employed disproportionality analysis algorithms, specifically the Reporting Odds Ratio (ROR), Proportional Reporting Ratio (PRR), and Empirical Bayes Geometric Mean (EBGM), to quantify safety signals. A signal was considered significant if the lower limit of the 95% confidence interval (CI) for the ROR was greater than 1 and the number of cases ( $a$ ) was at least 3.
Stratification: Analysis was stratified by demographics (age, gender), specific ICI agents (e.g., nivolumab, pembrolizumab, ipilimumab), time-to-onset (TTO), geographic region, and clinical outcomes (fatal vs. non-fatal, serious vs. non-serious).

Key Results

Case Characteristics: A total of 588 ICI-ILD cases in GI malignancies were identified in the FAERS database. The median patient age was 70.5 years, with 65.14% of patients aged ≥65. Males constituted the majority of reports (71.6%). The primary tumor types were gastric cancer (32.7%) and esophageal cancer (22.8%).
Drug Involvement: Nivolumab was the most frequently implicated agent (58.3% of cases), followed by pembrolizumab (24.8%).
Temporal Patterns: The median time-to-onset was 58 days. Risk distribution varied by drug class:
- Anti-CTLA-4 (Ipilimumab): Showed a highly concentrated risk window within the first 30 days of treatment.
- Anti-PD-1/PD-L1 (Nivolumab, Pembrolizumab): Demonstrated a more prolonged risk distribution, with nivolumab showing definitive signals even beyond 360 days.
Demographic Disparities: While males had higher absolute case numbers, gender-stratified ROR analysis indicated that females exhibited higher relative risk signals for specific agents (e.g., avelumab). Elderly patients (≥65 years) demonstrated the most pronounced association signals overall.
Geographic Specificity: Reporting patterns varied significantly by region. Japan reported the highest volume of nivolumab-related cases, with distinct signals for bronchiolitis. The United States showed prominent signals for pembrolizumab and ipilimumab. China was the exclusive source for reports on domestically developed PD-1 inhibitors (tislelizumab, toripalimab).
Severity: 99.8% of reported cases were classified as serious adverse events, with fatal outcomes occurring in 33.2% of cases. Nivolumab accounted for the highest number of fatal ILD cases (221).
Validation: Analysis of the EudraVigilance database corroborated the FAERS findings, confirming statistically significant links between ILD and nivolumab, pembrolizumab, ipilimumab, and tislelizumab, with strong signals for immune-mediated lung disease.

Significance and Claims
The authors claim this is the first large-scale retrospective analysis systematically evaluating the disease-specific distribution of ICI-related ILD in gastrointestinal cancer patients based on FAERS data. The study's significance lies in:

Filling a Research Gap: It addresses the paucity of data regarding ICI-ILD specifically in GI malignancies, distinct from the well-studied NSCLC population.
Risk Stratification: The findings highlight distinct risk profiles based on age, gender, specific ICI agents, and tumor type (particularly gastric and esophageal cancers).
Clinical Guidance: The identification of heterogeneous time-to-onset patterns (early vs. prolonged) suggests that monitoring strategies should be temporally stratified based on the mechanism of action of the specific ICI used.
Real-World Evidence: By utilizing spontaneous reporting data, the study captures safety signals in complex, real-world populations that may be missed in controlled clinical trials.

Limitations
The paper explicitly acknowledges several limitations inherent to spontaneous reporting systems:

Lack of Denominator Data: The study cannot calculate true incidence rates or relative risks in the total treated population, only reporting frequencies.
Reporting Bias: Voluntary reporting leads to potential under-reporting and selection bias influenced by market time, geography, and reporting propensity.
Missing Covariates: Critical clinical data such as baseline pulmonary function, smoking history, concomitant medications, and ECOG performance status were unavailable.
Diagnostic Ambiguity: Without confirmatory data (e.g., HRCT, biopsy), it is difficult to distinguish ICI-induced ILD from tumor progression or radiation pneumonitis, potentially leading to indication bias.

Conclusion
The study concludes that ICI-related ILD in GI oncology presents with distinct, agent-specific, and demographic-specific risk profiles. Given the high fatality rate and heterogeneous time-to-onset, the authors advocate for risk-stratified monitoring to optimize immunotherapy safety. They propose future directions including prospective multicenter studies to establish diagnostic standards, biomarker discovery for high-risk populations, and further exploration of the gut-lung axis mechanisms.

More like this