Abstract
Immune checkpoint inhibitors have markedly improved outcomes in patients with multiple advanced malignancies. However, their widespread use has markedly increased the incidence of immune-related adverse events (irAEs). irAEs can affect a wide range of organ systems and are characterized by heterogeneous onset, broad toxicity spectra, and complex management requirements, thus ultimately impairing treatment continuation and patient quality of life. This review systematically summarizes the epidemiological features, clinical progression, and current management of irAEs. Existing guidelines largely focus on acute toxicities but have not provided structured strategies for chronic, delayed-onset, or multisystem irAEs. Moreover, clinical practice is hampered by incomplete multidisciplinary collaboration, insufficient training of oncologists, and fragmented treatment pathways, all of which limit the efficacy of irAE management. We propose incorporating irAE management into core oncology training and call for the establishment of comprehensive interdisciplinary frameworks to ensure the standardized long-term use of immunotherapy.
keywords
- Immune checkpoint inhibitors
- immune-related adverse events
- multidisciplinary management
- oncologists
- medical education
Introduction
Immune checkpoint inhibitors (ICIs) restore antitumor immune responses by increasing inhibitory signals on T cells, thereby substantially improving the prognosis of various cancers1–5. However, immune activation can also result in off-target effects leading to immune-related adverse events (irAEs), which are among the major limitations to the safety and continuation of immunotherapy6. irAEs result from immune-mediated attacks on healthy tissues, can affect nearly all organ systems, and notably have nonspecific presentation and unpredictable timing.
ICIs enhance antitumor immunity by blocking the PD-1/PD-L1 and CTLA-4 pathways, but they also disrupt self-tolerance and trigger irAEs. Three major immune mechanisms contribute to irAE pathogenesis: (1) aberrant activation of self-reactive T cells, such as CD8+ T cells targeting bile duct epithelium in hepatitis, or T cells recognizing cardiac myosin or gangliosides in myocarditis and neurologic irAEs7,8; (2) autoantibody-mediated damage, including anti-TPO and anti-ACTH antibodies in thyroiditis and hypophysitis, and anti-platelet or anti-alveolar antibodies in hematologic and pulmonary toxicities9,10; and (3) innate immune dysregulation and cytokine storms, wherein IL-6, IL-17, and TNF-α promote tissue infiltration and injury11–13. These immunotoxic patterns are further shaped by organ-specific immune niches—such as PD-L1 expression in the liver or blood-nerve barrier permeability—and by host-intrinsic factors including HLA haplotypes and gut microbiota diversity.
ICIs have received approval for more than 100 indications by the U.S. Food and Drug Administration (FDA), and the number continues to increase14. With expanding indications and prolonged treatment durations, the clinical manifestations of irAEs continue to evolve. These include concurrent or sequential involvement of multiple organ systems, overlapping symptoms, chronic irAEs that persist after treatment cessation, and delayed-onset irAEs occurring months after treatment discontinuation. Current management strategies largely use an organ-specific, acute toxicity-based approach. However, because they lack guidance regarding disease fluctuation, coordination of treatment rhythms, and long-term toxicity dynamics, they are insufficient for current clinical scenarios.
Oncologists, despite being the primary prescribers of ICIs, often lack formal training in irAE recognition and management, and rely on subspecialty consultation. However, these specialists’ often limited understanding of the timing of immunotherapy and risk profiles can further exacerbate decision-making disjunctions.
Whereas prior high-impact reviews have provided comprehensive insights into both the clinical and mechanistic aspects of irAEs6,15,16, this article offers a complementary perspective by summarizing the overall landscape of irAEs and emphasizing the under-addressed need for structured oncologist education. Rather than focusing on pathogenesis or treatment nuances, we describe practical gaps in real-world irAE management, and we highlight the importance of preparing oncologists for immune toxicity recognition, interdisciplinary coordination, and informed therapeutic decision-making. By framing irAE education as a critical component of immunotherapy implementation, this review is aimed at bridging the gap between knowledge and practice in oncology care. A structured literature search was performed to support the review (search strategy and selection flowchart in Supplementary Methods and Figure S1).
Epidemiology and clinical presentation of irAEs
Incidence of irAEs
The incidence of irAEs varies with ICI type, dosage, and regimen (Table 1). CTLA-4 inhibitors are associated with higher irAE rates than PD-(L)1 inhibitors37. CTLA-4 inhibitors (e.g., ipilimumab) have an overall irAE incidence as high as 60%, and 10%–30% of these irAEs are grade ≥ 3 events, thereby indicating dose dependence38,39. In high-dose adjuvant settings, the incidence of grade ≥ 3 irAEs may exceed 50%38. In contrast, PD-1 inhibitors (e.g., nivolumab and pembrolizumab) cause irAEs in 5%–20% of patients, and grade ≥ 3 events occur in ~10% of patients; these events often include endocrine disorders, pneumonitis, hepatitis, or cutaneous toxicity40,41. PD-L1 inhibitors (e.g., atezolizumab) generally have low toxicity, particularly regarding pulmonary events42,43. Combining CTLA-4 inhibitors and PD-1 inhibitors markedly increases both the incidence and severity of irAEs, such that grade ≥ 3 events occur in more than 50% of patients, usually in early treatment phases1,44,45.
Frequencies of treatment-related irAEs in selected cohorts
Broad organ involvement and heterogeneous toxicity spectrum
irAEs can affect almost any organ system: they commonly involve the skin, gastrointestinal tract, liver, lungs, thyroid, and pituitary glands, but can also affect the heart, nervous system, kidneys, skeletal muscles, and hematopoietic tissues (Figure 1)46. Toxicity profiles vary by ICI type: PD-(L)1 inhibitors tend to cause thyroid dysfunction, pneumonitis, and rash, whereas CTLA-4 inhibitors more commonly cause colitis, hypophysitis, and severe skin reactions46. Combination therapy further amplifies both the toxicity spectrum and severity46.
Systemic distribution of irAEs. This schematic provides a comprehensive overview of irAEs affecting various organ systems. Manifestations are grouped by system and illustrated anatomically: ocular (uveitis, keratitis, and conjunctivitis), neurologic (immune-mediated encephalitis, peripheral neuropathy, myasthenia gravis, and Guillain–Barré syndrome), endocrine (thyroiditis/hypothyroidism/hyperthyroidism, adrenal insufficiency, type 1 diabetes mellitus, and hypophysitis), cardiovascular (myocarditis, pericarditis, arrhythmias, and heart failure), pulmonary (pneumonitis), gastrointestinal (colitis, hepatitis, and pancreatitis), musculoskeletal (arthritis, myositis, and polymyalgia rheumatica), other irAEs (sicca syndrome, sarcoidosis, and multisystem inflammatory syndrome), hematologic (hemolytic anemia, immune thrombocytopenia, aplastic anemia, and hemophagocytic lymphohistiocytosis), renal (glomerulonephritis and acute interstitial nephritis), and dermatologic (rash, pruritus, vitiligo, bullous dermatitis, and Stevens–Johnson syndrome/toxic epidermal necrolysis). HLH, hemophagocytic lymphohistiocytosis; ICI, immune checkpoint inhibitor; irAE, immune-related adverse event; SJS/TEN, Stevens-Johnson syndrome/toxic epidermal necrolysis.
Organ-specific irAE patterns can also be influenced by the tumor type and microenvironment, and the presence of shared immunogenic antigens across tumors and healthy tissues15. For example, melanoma is associated with elevated rates of skin and gastrointestinal irAEs, whereas patients with non-small cell lung cancer develop pulmonary toxicities47.
Risk factors
irAE risk is influenced by interrelated patient-related, treatment-related, and immune-related factors15. The patient factors include performance status, age, lifestyle (diet and exercise), microbiome composition (gut, skin, and lungs), and genetic background. Treatment-related factors, such as ICI type, dose, and regimen, influence the magnitude of immune activation. The immune-related factors include tumor type and microenvironment, immune cell subpopulations, TCR/BCR diversity, and inflammatory mediators, such as IL-648. Whereas some of these factors, such as lifestyle and treatment choices, are modifiable49,50, genetic predisposition is not, thus highlighting the need for multidimensional risk assessment tools.
Onset timing
irAEs can arise at any time during or after treatment, even months after discontinuation15,51. CTLA-4-based regimens are associated with earlier irAE onset than PD-(L)1 monotherapy52. Some irAEs (e.g., rash, diarrhea, and transaminitis) may occur within days or weeks, whereas others (e.g., endocrine dysfunction and neurological irAEs) emerge later. The timing of irAEs also varies by organ; whereas dermatologic and gastrointestinal toxicities typically appear early, hepatic, pulmonary, and neurologic irAEs may arise later or recur post-remission52,53.
Mortality
Fatal irAEs often occur early and progress rapidly, particularly with combination regimens (Table 2). The overall fatality rate is 0.3%–2%69,70, and the drug-specific rates are 0.4% for PD-1 inhibitors and 1.2% for PD-1/PD-L1 plus CTLA-4 combinations70. The types of fatal irAEs also differ. CTLA-4-related deaths are due primarily to colitis (70%), whereas PD-(L)1-related deaths are associated with pneumonitis, hepatitis, and neurotoxicity. Combination regimens are frequently associated with fatal colitis (37%) and myocarditis (25%), the latter of which shows the highest mortality rate (39.7%)70.
Frequencies of treatment-related deaths in selected cohorts
Clinical evolution of irAEs
Although the widespread use of ICIs has substantially prolonged patient survival, improved survivability has introduced new challenges, including the emergence of multisystem, delayed, and chronic irAEs. These findings underscore the extended and complex clinical course of irAEs in terms of timing and presentation.
Multisystem irAEs
Although irAEs typically involve a single organ system, multisystem involvement is not uncommon and may occur concurrently or during steroid tapering. Multisystem irAEs are defined as those affecting multiple distinct organ systems or different tissues within the same system71–73. Among patients receiving PD-(L)1 monotherapy, the incidence is approximately 5%–9%, and these irAEs most frequently involve dermatologic and endocrine toxicities71,74. In combination therapies, the incidence increases to 16%–40%, and gastrointestinal toxicities are relatively common75,76. These cases often include overlapping symptoms that complicate diagnosis and require coordinated multidisciplinary decision-making.
Chronic irAEs
Chronic irAEs are defined as irAEs persisting for more than 3 months after ICI discontinuation73. The most frequently reported chronic irAEs involve the endocrine system, arthritis, xerostomia, neurotoxicity, and ocular toxicity. In contrast, gastrointestinal, hepatic, pulmonary, and renal toxicities are less likely to become chronic and are generally reversible77.
Johnson et al. have proposed a hypothetical framework for the pathophysiological mechanisms of chronic irAEs and have suggested that they are driven primarily by smoldering inflammation and burnout78. Smoldering inflammation refers to persistent immune activation after ICI therapy, wherein continued immune cell infiltration into normal tissues leads to fluctuating or recurrent symptoms, as observed in immune-mediated colitis and arthritis. These conditions often remain responsive to immunosuppressive therapies such as corticosteroids or TNF/IL-6 inhibitors78,79. In contrast, burnout refers to functional impairment of organs or glands caused by early inflammatory injury, in which the inflammatory process has subsided, but irreversible dysfunction persists. Examples include hypothyroidism, hypophysitis, and type 1 diabetes. These toxicities are typically unresponsive to corticosteroid therapy and require lifelong hormone replacement80,81. For certain irAEs, reversibility can often be assessed only retrospectively after treatment.
In a retrospective multicenter cohort study in 387 patients with melanoma receiving adjuvant PD-1 inhibitor therapy, 43% developed chronic irAEs, 96% of which were mild. At a median follow-up of 35 months, 65% of chronic irAEs were ongoing77,82. In another study in 437 patients, 35.7% of irAEs lasted for at least 24 months83. Among patients with non-endocrine chronic irAEs, 60% permanently discontinued ICI therapy, and 76% required corticosteroid treatment (91% of patients with severe irAEs). Additional immunosuppressive agents or other interventions were also necessary for some patients. Nearly one-quarter of patients required hospitalization, and 6% experienced adverse effects due to immunosuppressive therapy.
Although persistent irAEs might indicate favorable antitumor responses84, they can impose substantial functional and quality-of-life burdens on patients85. Studies have linked chronic irAEs to diminished quality of life. In a small cohort study on ICI-induced inflammatory arthritis, the observation of pronounced emotional and physical impairments highlighted the clinical importance of understanding chronic irAEs86. Moreover, the long-term use of immunosuppressants might have additional adverse effects.
Historically, the clinical focus has been primarily on acute irAEs. Key guidelines from the American Society of Clinical Oncology (ASCO)87 and European Society for Medical Oncology (ESMO)88 emphasize acute toxicity management but lack specific recommendations for chronic irAEs. However, growing evidence suggests that chronic irAEs are prevalent and require sustained interventions. The Society for Immunotherapy of Cancer (SITC) has proposed a treatment-oriented classification of chronic irAEs as either active or inactive, according to the need for, and efficacy of, anti-inflammatory interventions73. Importantly, these categories should be viewed as 2 ends of a clinical activity spectrum, with potential overlap between them. The lack of histopathological confirmation in most chronic irAEs complicates the assessment of residual inflammatory activity, thereby increasing the uncertainty in therapeutic decision-making73,78. This uncertainty directly affects treatment decisions regarding continued immunosuppression vs. transition to supportive care, increasing decision-making complexity and communication burden. As indications for ICIs increase in neoadjuvant89,90 and adjuvant settings91, clinicians must better understand chronic irAEs to accurately inform patients regarding the potential long-term risks and benefits.
Delayed irAEs
Delayed irAEs are defined as those occurring >3 months after ICI discontinuation73. These events are often underrecognized and underdiagnosed, possibly because of persistent immune activation and dysregulated feedback despite drug clearance. In the CheckMate 238 trial, delayed irAEs occurred in approximately 4%–6% of patients, and grade ≥ 3 events occurred in 1%–2% of patients. The most common manifestations were colitis, pneumonitis, and rash92,93. In a pooled analysis of 1,567 patients with melanoma treated with pembrolizumab, 14 (0.9%) developed new irAEs more than 2 years after treatment initiation94. The actual incidence might have been underestimated because of the reliance on voluntary reporting. The temporal heterogeneity of these events complicates their attribution and necessitates ongoing clinical vigilance throughout the treatment continuum.
Other long-term risks and special toxicities
ICIs have been reported to accelerate atherosclerotic plaque progression and increase the risk of cardiovascular events, such as myocardial infarction and ischemic stroke, by more than threefold95,96. ICIs might also impair fertility, although the current evidence is limited, and further monitoring and management are needed97–99.
irAEs in special populations: older and autoimmune patients receiving ICI therapy
Although ICIs have transformed cancer therapy, evidence of their safety and efficacy in special populations remains limited. Older patients and those with pre-existing autoimmune diseases are underrepresented in clinical trials, yet real-world data suggest that they might experience distinct patterns of irAEs100. The following section summarizes current insights into irAEs within these vulnerable groups.
In patients with pre-existing autoimmune diseases, irAEs are notably more frequent and exhibit distinct clinical features from those observed in the general population101,102. Across real-world studies and meta-analyses, the overall incidence of irAEs in patients with pre-existing autoimmune diseases ranges from 60% to 62%; 25%–36% experience flares of their underlying autoimmune condition, and 23%–35% develop de novo irAEs. Psoriasis, inflammatory bowel disease, and rheumatoid arthritis are particularly prone to flares. Although most irAEs are grade 1–2 and manageable, approximately one-third of patients require hospitalization, and severe events (grade ≥ 3) are not negligible. Despite the elevated risk, treatment-related mortality remains low (0.07%–0.80%), and only a minority of patients discontinue therapy because of toxicity103–105.
In older patients receiving ICIs, irAEs exhibit distinct characteristics and clinical implications. Although several large real-world studies have shown that the overall incidence of irAEs in older adults is comparable to that in younger patients, certain toxicities—particularly cutaneous, renal, and gastrointestinal toxicities—occur more frequently in older populations106,107. Moreover, in older patients with non-small cell lung cancer (NSCLC), specific severe irAEs, such as pneumonitis, arrhythmias, hepatitis, and colitis, have been associated with markedly elevated risks of mortality and treatment discontinuation, thereby indicating diminished tolerability in this subgroup108,109. Importantly, the occurrence of irAEs in older adults, in contrast to younger populations, does not consistently correlate with improved clinical outcomes; consequently, irAEs might lack predictive value for therapeutic benefit in this age group109. Therefore, when administering ICIs to patients ≥ 75 years of age, careful monitoring and individualized toxicity management are essential. Functional status and comorbidities should be thoroughly evaluated to balance efficacy and safety, particularly in vulnerable populations underrepresented in clinical trials100,106,107,109.
Toxicity profiles of emerging ICIs
Beyond PD-1/PD-L1 and CTLA-4, the principal next-generation ICIs under development target LAG-3, TIM-3, TIGIT, VISTA, BTLA, and the SIRPα–CD47 axis110–116. Most of these agents remain in clinical trials; to date, only relatlimab—co-formulated with nivolumab—has obtained the FDA approval for melanoma.117,118, whereas others, such as tiragolumab, have been discontinued because of inadequate efficacy in small-cell lung cancer119. Progress is constrained by several factors: (i) modest single-agent activity, because clinically meaningful responses generally emerge only when the drugs are combined with PD-1/PD-L1 blockade; (ii) pathway redundancy and compensatory up-regulation of alternative inhibitory axes, which blunt efficacy after a single checkpoint is blocked; (iii) a scarcity of target-specific biomarkers, diluting the proportion of truly responsive patients enrolled in clinical trials; and (iv) a 1.5- to 2-fold increase in grade ≥ 3 irAEs—including rash, hepatitis, pneumonitis, and endocrine dysfunction—when these agents are used with PD-(L)1 inhibitors or chemotherapy120–123. Moreover, SIRPα–CD47 blockade can induce anemia and thrombocytopenia, because CD47 is ubiquitously expressed on hematopoietic cells124.
Diagnosis and current management of irAEs
During ICI treatment, the onset of any new symptoms should prompt consideration of irAEs. However, diagnosing irAEs is challenging, because of their organ-specific presentation and frequent overlap with symptoms of tumor progression, comorbidities, or other treatment-related adverse effects. Early manifestations are often nonspecific, and include fatigue, appetite loss, or mild dermatological changes, which might represent the initial signs of an irAE or reflect unrelated conditions. For example, checkpoint inhibitor-related pneumonitis is often misdiagnosed as an infection, radiation-induced injury, or disease progression. Neurological irAEs, such as encephalitis or polyneuropathy, can be mistaken for central nervous system metastases or underlying neurological disorders. Currently, no standardized diagnostic tests are available, and irAE diagnoses remain largely exclusionary. Differential diagnoses include tumor progression, infections, and other drug-related toxicities87,88,125.
First-line treatment generally involves ICI discontinuation (Figure 2). For grade 2 irAEs and most grade ≥ 3 irAEs, systemic immunosuppressive therapy is required. High-dose corticosteroids are the mainstay treatment, and initial dosing is based on irAE severity and typically comprises prednisolone at 0.5–2.0 mg/kg. The treatment follows a stepwise tapering strategy. In the event of an insufficient early response, dose escalation or the addition of second-line immunosuppressants might be required. After symptoms resolve, steroids are tapered over weeks to months, depending on irAE type and severity87,88,125.
Multidisciplinary management flowchart for irAEs. This flowchart depicts a decision-making model for irAE management based on severity and organ involvement. The process begins with initial recognition and CTCAE-based grading by the ICI-prescribing oncologist. If irAE severity reaches grade ≥ 2 or multiple organs are suspected to be involved, multidisciplinary consultation is initiated. Common subspecialties include rheumatology, pulmonology, gastroenterology, endocrinology, neurology, dermatology, ophthalmology, and cardiology. Management is stratified by CTCAE grade: grade 1 involves continued ICI with monitoring; grade 2 requires ICI withholding and corticosteroids (e.g., prednisone 0.5–1.0 mg/kg/day); and grade ≥ 3 necessitates hospitalization and pulse-dose steroids. Treatment may escalate to second-line immunosuppressants (e.g., mycophenolate, infliximab, or tocilizumab), particularly for steroid-refractory cases. Endocrinopathies are managed with hormone replacement. Final steps involve assessing the treatment response and making informed decisions regarding ICI resumption or further escalation. CTCAE, common terminology criteria for adverse events; ICI, immune checkpoint inhibitor; irAE, immune-related adverse event; TNF, tumor necrosis factor.
Compared with classical autoimmune diseases, irAEs often respond more rapidly to steroids, and most patients show notable improvement within several days126. However, the dynamic nature of irAEs complicates their management. Some patients experience symptom recurrence during steroid tapering or after steroid cessation. Long-term high-dose steroid use carries risks, such as increased infection rates and osteoporosis, that require careful balancing between inflammation control and minimizing steroid-associated toxicity127.
Second-line treatment for irAEs is warranted in patients who do not respond adequately to corticosteroids or who experience steroid-related toxicity. This approach involves immunomodulatory agents tailored to the affected organ system and severity of toxicity. Anti-TNF-α agents such as infliximab are commonly used in steroid-refractory colitis, whereas vedolizumab offers a gut-selective alternative with a potentially lower risk of systemic immunosuppression126,128,129. Mycophenolate mofetil is frequently used in hepatic, renal, or pulmonary irAEs, and IL-6 receptor blockade with tocilizumab has demonstrated efficacy in inflammatory arthritis and myositis130. For B cell-driven complications, including autoimmune cytopenias and neurologic syndromes, rituximab and intravenous immunoglobulin are appropriate options. In cases of severe or multisystem toxicities, agents such as abatacept or cyclophosphamide may be considered. The choice of second-line therapy should be guided by a multidisciplinary team, to balance irAE control with preservation of anti-tumor immunity87,88,125.
Predictive biomarkers for irAE
A range of biomarkers have been identified to predict irAEs in patients receiving ICIs. Hematologic markers such as a low neutrophil-to-lymphocyte ratio (NLR < 3), elevated derived NLR (dNLR > 3), diminished ALC, and clonal expansion of CD8+ T cells (>55) are positively associated with irAE risk. Diminished regulatory T cells and altered PLR, ANC, and MLR also have predictive value131–134. Cytokines, including IL-6, IL-8, IL-17, CXCL9/10/11, and TNF-α, particularly IL-17 in gastrointestinal irAEs, markedly rise after ICI. Lower interferon-γ levels and elevated CRP (>50 mg/L) further correlate with pneumonitis and systemic inflammation135–140.
Autoantibodies such as ANA, anti-TPO, and soluble CTLA-4 (>200 pg/mL) are associated with endocrine and systemic irAEs112–117. Serum proteins such as elevated TSH, troponin, ferritin, and low albumin (<3.6 g/dL) also have predictive significance135,141–144. Genetic markers, particularly HLA-DRB111:01, DQB103:01, and DR4 subtypes, are associated with organ-specific irAEs145–149. MicroRNA variants (e.g., miR-146a rs2910164) and gene expression changes (e.g., CD177 and CEACAM1) further support genetic predisposition150,151.
Finally, gut microbiota patterns—such as enrichment in Faecalibacterium and Firmicutes, depletion of Bacteroidetes—and stool calprotectin levels (>150 mg/g) correlate with gastrointestinal irAEs152–155. Together, these markers offer a promising approach for early irAE risk stratification and individualized ICI management.
Limitations of current management models
Current toxicity grading systems (e.g., Common Terminology Criteria for Adverse Events, CTCAE) are designed primarily for conventional chemotherapy-related toxicities but might be inadequate for assessing irAEs156,157. For instance, the CTCAE grading of rash is based on body surface area, whereas irAE-associated skin reactions often require the evaluation of histological features (e.g., pemphigoid-like lesions) and systemic symptoms158.
The lack of standardized diagnostic and classification frameworks contributes to variability in irAE assessment across clinicians and institutions. Given the diversity in clinical presentation, affected organs, and timing, irAE management typically requires multidisciplinary coordination. However, perspectives on treatment goals differ across specialties. Subspecialists often focus on preserving organ function, whereas oncologists aim to maintain the treatment rhythm within a tolerable toxicity range, while balancing efficacy and safety159–162.
Numerous studies have shown that irAEs are associated with improved survival outcomes, including prolonged progression-free survival (PFS), overall survival (OS), and objective response rate (ORR)71,72,160,163–165. These findings might indicate stronger immune activation. Consequently, oncologists tend to favor active toxicity management over premature therapy discontinuation, even in the face of complex, overlapping, multisystem irAEs. In contrast, subspecialists might be unaware of this correlation and often use conservative strategies, thus overlooking the potential link between irAEs and treatment response. A lack of familiarity with immunotherapy rechallenge criteria and risk mitigation strategies might also lead to recommendations that are disconnected from the broader therapeutic context.
Differences in priorities and risk tolerance among specialties can result in disagreements regarding key decisions, such as treatment interruption, intervention intensity, and therapy scheduling. Clinical decisions are often compromised in the absence of standardized prioritization mechanisms or coordination frameworks. However, such compromises might fail to optimize toxicity control and treatment continuation.
A more fundamental issue lies in the structural disjunction across specialties. Oncologists, the primary agents of immunotherapy, often lack leadership and independent decision-making capacity in irAE management. Faced with organ-specific toxicities, they must rely on specialist consultations, because of their limited immunologic and subspecialty knowledge. In contrast, specialists’ frequent lack of understanding of the broader immunotherapy context can further fragment decision-making. Therefore, treatment strategies are often not based on longitudinal assessments by oncologists, and key decisions, such as when to resume ICI treatment, are frequently not made by those most familiar with the overall therapeutic landscape. A structural dilemma involving no central oversight by specialists and no decisive authority by oncologists therefore undermines treatment coherence and efficiency.
In addition, the current clinical workflow is often fragmented. When multisystem toxicities occur, patients are typically referred to different specialists for separate organ-based management. As a result, patients often experience prolonged diagnostic and treatment timelines, increased communication costs, and neglect of inter-organ toxicity interactions.
One potential solution to this dilemma is oncologist-led coordination models that leverage digital platforms for real-time communication. In our institutional experience, tumor-specific, oncologist-led coordination—supported by an asynchronous encrypted messaging platform—has shown promise in streamlining irAE triage and treatment. Although the full implementation details have been reported elsewhere, this model exemplifies how frontline oncologists can facilitate timely, multidisciplinary input and serve as central integrators of irAE care. This communication-based approach offers a scalable, low-cost solution, particularly for centers lacking embedded multidisciplinary clinics, and can help facilitate earlier recognition and timely management of irAEs, thereby reducing unnecessary ICI interruptions and interruptions of antitumor treatment.
Educational gaps and practical solutions
Although international oncology education systems have become increasingly standardized, and organizations such as the ESMO and ASCO have developed global curricula and promoted multidisciplinary training166–170, current medical education lacks systematic content specific to irAE management. Oncologists are frequently undertrained in this domain. Existing training is often institution-led and consists mainly of lectures or case discussions focusing on individual toxicities without a structured theoretical or procedural framework. Consequently, because most oncologists rely on experience and are poorly equipped to assess and manage complex or dynamic multisystem irAEs, their ability to take initiative in multidisciplinary decision-making is limited.
To address this capability gap, we propose a structured, competency-based re-education framework tailored for oncologists (Figure 3). This framework comprises 5 progressive modules: (1) foundational knowledge, including the pathophysiology of irAEs and toxicity grading based on CTCAE; (2) recognition and differential diagnosis of common irAEs through clinical manifestations, onset timing, and radiographic features; (3) evidence-based treatment strategies, including rational use, tapering, and escalation principles for corticosteroids and second-line immunosuppressants; (4) multidisciplinary collaboration skills to enhance communication and decision-making within multidisciplinary team (MDT); and (5) long-term irAE management and patient education.
Structured education framework for oncologists in irAE management. This framework outlines a multidimensional educational strategy for oncologists managing irAEs. Core modules cover foundational knowledge (e.g., irAE pathogenesis and CTCAE grading), diagnostic reasoning (e.g., symptom triage and differential diagnosis), pharmacologic strategies (e.g., corticosteroid tapering and second-line agents), interdisciplinary collaboration, and long-term patient education. Instructional methods include structured lectures, case-based learning, simulation and role-play, and joint MDT training. Outcome-based assessments include post-session quizzes, case-solving, simulated patient communication, Mini-CEX, and steroid escalation planning. This system is aimed at standardizing oncologist education and enhancing readiness for irAE management across specialties. CTCAE, common terminology criteria for adverse events; ICIs, Immune checkpoint inhibitors; irAE, immune-related adverse event; MDT, multidisciplinary team; Mini-CEX, Mini-clinical evaluation exercise.
The training program emphasizes the integration of theoretical foundations with practical clinical application and uses diverse instructional methods. Case-based teaching focuses on real-world decision-making scenarios, thereby enabling participants to recognize typical presentations and develop appropriate management strategies. Simulation-based training reinforces clinical reasoning and intervention skills in complex situations. For instance, in the simulation of immune-related pneumonitis, trainees are guided to distinguish this condition from infectious pneumonitis, radiation-induced lung injury, and disease progression by integrating symptomatology, imaging findings, and temporal patterns. In the elevated liver enzyme module, trainees assess prior treatment history and lesion distribution to differentiate among immune-related hepatitis, locoregional therapy-induced hepatic injury, and tumor progression. Furthermore, interdisciplinary modules promote coordinated management among oncology, pulmonology, endocrinology, neurology, and gastroenterology specialties, by leveraging regular MDT case discussions.
The evaluation system adopts a multidimensional approach encompassing knowledge acquisition (post-training quizzes and case-based questions), clinical reasoning (Mini-clinical evaluation exercise, Mini-CEX and group case discussions), communication skills [empathy scoring and standardized patient objective structured clinical examination (OSCE)], and practical competencies (e.g., CTCAE grading exercises and immunosuppressive treatment planning). This integrative educational framework is aimed at not only enhancing individual oncologists’ capabilities in managing complex irAEs but also advancing standardization and quality improvement in irAE management across healthcare systems.
Conclusions
IrAEs have become a critical bottleneck in the safe and effective use of ICIs. Although the clinical spectrum of irAEs continues to evolve—ranging from acute to chronic, multisystemic, and delayed manifestations—current management remains fragmented, largely reactive, and insufficiently integrated into oncology training. This review underscores the urgent need for oncologist-led, multidisciplinary frameworks and competency-based re-education to address this complexity. Empowering oncologists with structured knowledge and practical tools will be essential to ensure timely recognition, coordinated intervention, and long-term patient safety in the era of immunotherapy.
Conflict of interest statement
No potential conflicts of interest are disclosed.
Author contributions
Conceived and designed the analysis: Hanping Wang, Haitao Zhao.
Collected the data: Binhe Tian, Yuanmei Yang, Shuman Kuang, Mingjian Piao, Chengjie Li.
Contributed data or analysis tools: Binhe Tian, Yuanmei Yang, Shuman Kuang.
Performed the analysis: Binhe Tian, Yuanmei Yang, Shuman Kuang.
Wrote the paper: Binhe Tian, Yuanmei Yang, Shuman Kuang.
- Received August 4, 2025.
- Accepted August 14, 2025.
- Copyright: © 2025, The Authors
This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License.
References
- 1.↵
- 2.
- 3.
- 4.
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.
- 31.
- 32.
- 33.
- 34.
- 35.
- 36.
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.
- 55.
- 56.
- 57.
- 58.
- 59.
- 60.
- 61.
- 62.
- 63.
- 64.
- 65.
- 66.
- 67.
- 68.
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.
- 112.↵
- 113.
- 114.
- 115.
- 116.↵
- 117.↵
- 118.↵
- 119.↵
- 120.↵
- 121.
- 122.
- 123.↵
- 124.↵
- 125.↵
- 126.↵
- 127.↵
- 128.↵
- 129.↵
- 130.↵
- 131.↵
- 132.
- 133.
- 134.↵
- 135.↵
- 136.
- 137.
- 138.
- 139.
- 140.↵
- 141.↵
- 142.
- 143.
- 144.↵
- 145.↵
- 146.
- 147.
- 148.
- 149.↵
- 150.↵
- 151.↵
- 152.↵
- 153.
- 154.
- 155.↵
- 156.↵
- 157.↵
- 158.↵
- 159.↵
- 160.↵
- 161.
- 162.↵
- 163.↵
- 164.
- 165.↵
- 166.↵
- 167.
- 168.
- 169.
- 170.↵










