The Long-Term Sequelae and Clinical Workload Resulting from Toxicities of Immune Checkpoint Inhibitor (ICI) Therapy in Patients with Cancer Authors: Veera Nurmela, Anni Juntunen, Satu Tiainen, Janne Martikainen, Sanna Pasonen-Seppänen, Aino Rönkä Published: 2026 ABSTRACT AbstractBackgroundUp to half of cancer patients treated with immune checkpoint inhibitors (ICIs) develop immune-related adverse events (irAEs). However, the long-term consequences of irAEs remain poorly characterized.ObjectiveThis study investigated the likelihood of chronic irAEs, challenges related to their corticosteroid treatment and clinical workload associated with their management.MethodsWe retrospectively analyzed 230 patients who received ICIs at Kuopio University Hospital, Finland, between the years 2015 and 2024.ResultsIrAEs were chronic in 60.9% of patients who developed toxicities. Thyroiditis was the most common chronic AE (40%). Of the patients with irAE+, 42.9% were treated with systemic corticosteroids and steroid-related AEs were observed in 58% of them. Corticosteroid usage for ≥ 3 months was associated with an increased risk of steroid-related AEs (< 3 months 38.2% vs ≥ 3 months 78.8%, p < 0.001). An average of 5.3 contacts with a medical oncologist was needed for the management of an irAE, with the highest workload associated with hepatitis (10.0 contacts/irAE) and colitis (6.1 contacts/irAE).ConclusionsThe characterization of the long-term sequelae of irAEs and the associated clinical workload is crucial for proper patient education and clinical resource planning.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s40801-026-00561-x. Key Points 61% of patients with immune-related adverse events (irAE) had a chronic toxicity (i.e. toxicity persisting for over 12 weeks).43% of the irAEs required treatment with corticosteroids, with 49% of them having prolonged corticosteroid therapy of over 12 weeks. This was associated with a higher risk of steroid-related adverse events.Management of irAEs caused a significant clinical workload, requiring an average of 5.3 contacts with a medical oncologist per irAE. Introduction Immune checkpoint inhibitors (ICIs) are an established component of cancer therapy. ICIs can induce long-lasting antitumor responses. However, over 50% of ICI-treated patients suffer from therapy-related toxicities, that is, immune-related adverse events (irAEs) [1]. The treatment of irAEs can be complex and requires commitment from both the patient and cancer care unit. In addition, the recurrence rate of irAEs is relatively high and their clinical management is often complicated [2]. To plan effective irAE management, a better understanding of the consequences of different types of irAEs and their associated clinical workloads is essential. Up to 40% of irAEs become chronic [3, 4]. The concept of chronicity in terms of irAEs lacks a universal definition; however, chronic irAEs are most commonly defined as toxicities that continue for 12 weeks after irAE onset or ICI discontinuation [5–7]. Acute and chronic toxicities may vary according to their clinical manifestations. In a previous study, acute AEs tended to arise more frequently in visceral tissues, such as the gastrointestinal tract or lungs [3]. Endocrine, dermatological and nervous system irAEs appear to be particularly prone to chronicity, often owing to irreversible organ damage or their smoldering nature [3, 8]. The administration of systemic corticosteroids remains the cornerstone of irAE management [9]. Prompt initiation of high-dose corticosteroid therapy is critical in patients with severe or highly symptomatic irAEs. Moreover, irAEs with a relapsing–remitting course may require multiple courses of corticosteroid therapy. Although effective tapering of steroids is recommended after clinical resolution of irAEs, standardized tapering protocols are lacking. Consequently, corticosteroid tapering schedules are often prolonged in clinical practice. These factors may significantly increase the risk of corticosteroid-associated complications, such as osteoporosis and hyperglycemia [10] in patients with irAEs. In addition to immunosuppressive therapy, the clinical management of irAEs consists of outpatient follow-up, hospitalization and involvement of other medical specialties. The divergent therapeutic strategies for acute and chronic irAEs may result in a differential burden on healthcare resources. However, the long-term sequelae and healthcare implications associated with different types of irAEs have not been systematically characterized in the literature. In this study, we aimed to evaluate in detail the burden of chronicity, challenges associated with corticosteroid therapy for irAEs and the total clinical workload regarding their medical management. With an annual increase in the number of patients treated with ICIs, understanding the consequences of irAEs is important for assessing the effectiveness of ICI therapy. Materials and Methods Patients We retrospectively collected data from all patients (n = 235) whose cancer was treated with ICI therapy in the metastatic setting at Kuopio University Hospital, Cancer Center, Finland, between January 1, 2015 and December 31, 2022, and in the adjuvant setting between January 1, 2015 and September 24, 2024. Patients who had received ICIs in clinical trials (n = 4) were excluded, leaving a total of 231 patients in the cohort. ICIs coupled with chemotherapy or tyrosine kinase inhibitors (TKIs) were included. The last follow-up date was March 4, 2025. The list of therapies administered to the patients is shown in the Electronic Supplementary Material (ESM), Table S1. We manually searched the basic demographic, treatment, imaging and survival data from the electronic patient registry of Kuopio University Hospital with permission from the local Research Ethics Committee (IOPredict, permission number 1482/2019) and the organizational permission of Kuopio University Hospital (5654210, 916/2022). This study was conducted in accordance with the principles of the Declaration of Helsinki. Because of the retrospective design of this study, informed consent was not required. Immune-Related Adverse Events (irAEs) An irAE was defined as a typical toxicity induced by ICI therapy, occurring after ICI therapy initiation following the exclusion of differential diagnoses (such as infections). Biopsy confirmation of the target irAE tissue was included when available. Toxicities were graded using the Common Terminology Criteria for Adverse Events (CTCAE), version 5.0. IrAEs that resolved within 12 weeks of the start of the event were defined as acute. Based on a commonly used criteria in previous studies [5, 7], IrAEs that did not resolve within 12 weeks after the start of the event were considered chronic. IrAEs were considered resolved when the patient was asymptomatic and without irAE-targeted medication, and possible markers, such as laboratory measures, were within the reference range. The planned tapering of corticosteroids in otherwise asymptomatic patients was not included in the duration of an irAE. Patients whose follow-up period was < 12 weeks were excluded from chronicity analyses. The following four types of contact were considered when counting the number of irAE-related contacts with a medical oncologist: outpatient visits, telephone calls, silent (paper) consultations and hospitalizations. We included all contacts after the occurrence of an irAE in which the adverse effects were addressed. Contacts not addressing irAEs were excluded, even if the patient was experiencing an AE at that time. We did not include contact with primary care professionals, oncology nurses, or non-oncology specialists in our analysis. In the case of hospitalization, a single inpatient period was counted as one contact, regardless of its duration. If there were multiple reasons for hospitalization, we included only inpatient episodes in instances where the irAEs alone would have been considered severe enough to warrant hospitalization. Survival Analyses Overall survival (OS) was calculated from the date of ICI treatment initiation to the date of death from any cause or the date of the last follow-up visit, and death was counted as an event. Progression-free survival (PFS) was calculated from the date of ICI initiation to cancer progression, death from any cause, or end of follow-up, whichever occurred first, and cancer progression or death was counted as an event. The median follow-up time was calculated using the inverse Kaplan-Meier method. The median follow-up time was 37.6 months for metastatic patients, 47.1 months for adjuvant patients and 37.6 months for all patients. Statistical Analyses All data were analyzed using SPSS statistics software (Armonk, NY, USA), version 27. Differences between groups were analyzed using the chi-squared or Fisher’s exact tests. The numbers of consultations and contacts were compared using the independent t-test and Levene’s test for equality. The median hospitalization and corticosteroid time were compared using the median test with Yates’s continuity correction. Survival curves were plotted using the Kaplan–Meier method. Statistical significance was set at p < 0.05. Results Patient Characteristics In total, 230 patients were included in this study. The key baseline characteristics of the patients are summarized in Table 1; 68.3% were male and the median age at the start of ICI therapy was 66.7 years (95% confidence interval [CI] 64.8–68.1). The most prevalent cancer types in the cohort were non-small cell lung cancer (NSCLC, 43.5%), melanoma (26.1%) and renal cell carcinoma (RCC, 13.0%). ICI monotherapy was administered to 76.5% of the patients and dual ICI therapy to 4.3% of the patients; 19.1% received ICI in combination with other treatments. Table 1Main basic characteristics of the whole study cohort (n = 230), patients with irAEs (n = 128) and patients with irAEs divided into subgroups stratified by AE chronicity (acute n = 50, chronic n = 78)Patient cohort, n (%)p-valueAcute vs chronicAll230 (100)irAE+128a (56)Acute irAE50 (39)Chronic irAE 78 (61)Male gender, n (%)157 (68.3)88 (68.7)35 (70.0)53 (67.9)0.807Age [years], median66.766.564.968.10.070PS (WHO), n (%) 0135 (58.7)82 (64.1)33 (66.0)49 (62.8)0.715 ≥ 195 (41.3)46 (35.9)17 (34.0)29 (37.2)Cancer type, n (%) NSCLC100 (43.5)49 (38.3)20 (40.0)29 (37.2)0.749 Melanoma60 (26.1)39 (30.5)17 (34.0)22 (28.2)0.487 RCC30 (13.0)18 (14.1)3 (6.0)15 (19.2)0.036 Head and neck13 (5.7)8 (6.3)2 (4.0)6 (7.7)0.400 Urothelial11 (4.8)3 (2.3)1 (2.0)2 (2.6)1.000 MSI high colon9 (3.9)6 (4.7)4 (8.0)2 (2.6)0.208 Lymphoma5 (2.2)3 (2.3)2 (4.0)1 (1.3)0.560 TNBC2 (0.9)2 (1.6)1 (2.0)1 (1.3)1.000ICI treatment, n (%) Single ICI176 (76.5)100 (78.1)41 (82.0)59 (75.6)0.738 Combi ICI10 (4.3)7 (5.5)2 (4.0)5 (6.4) ICI + otherb44 (19.1)21 (16.4)7 (14.0)14 (18.0)Statistically significant results are shown in boldAE adverse event, ICI immune checkpoint inhibitor, irAEs immune-related adverse events, MSI microsatellite instability, NSCLC non-small cell lung carcinoma, PS performance status, RCC renal cell carcinoma, TNBC triple negative breast cancer, WHO World Health OrganizationaPatients with at least 3 months’ follow-up (FU) from the start of irAE (11 patients excluded due to short FU)bOther = chemotherapy or tyrosine kinase inhibitor (TKI) Main basic characteristics of the whole study cohort (n = 230), patients with irAEs (n = 128) and patients with irAEs divided into subgroups stratified by AE chronicity (acute n = 50, chronic n = 78) Statistically significant results are shown in bold AE adverse event, ICI immune checkpoint inhibitor, irAEs immune-related adverse events, MSI microsatellite instability, NSCLC non-small cell lung carcinoma, PS performance status, RCC renal cell carcinoma, TNBC triple negative breast cancer, WHO World Health Organization aPatients with at least 3 months’ follow-up (FU) from the start of irAE (11 patients excluded due to short FU) bOther = chemotherapy or tyrosine kinase inhibitor (TKI) Two hundred and two patients were treated in the metastatic setting and 28 in the adjuvant setting. NSCLC (47.5%), melanoma (18.3%) and RCC (14.9%) were the most common cancer types in metastatic patients, and melanoma was most common in adjuvant patients (82.1%). Metastatic patients were in worse physical condition compared with adjuvant patients (World Health Organization performance status [WHO PS] of ≥ 2: 45.5% vs 10.7%, p < 0.001), and they had received systemic corticosteroids prior to ICI initiation more often than adjuvant patients (37.1% vs 14.3%, p = 0.022). When comparing the frequency of chronic diseases, chronic obstructive pulmonary disease (COPD) was more common among metastatic patients than among adjuvant patients (19.8% vs 3.6%, p = 0.035). A more detailed description of the patient characteristics, divided into adjuvant and metastatic patients, is provided in the ESM, Table S2. Frequency of irAEs and Their Chronicity At least one irAE was observed in 60.4% (139/230) of the patients. For irAE chronicity analyses, 11 patients were excluded owing to a follow-up time shorter than 3 months, leaving 128 patients in the irAE+ cohort. The baseline characteristics of the 128 patients did not significantly differ from those of the entire cohort (Table 1). The frequency of irAEs was found to be higher in adjuvant patients than in metastatic patients (82.1% vs 57.4%, p = 0.012; ESM, Table S2). Of the patients with irAEs, 60.9% (78/128) had at least one chronic AE (Table 1). Chronic toxicities were more common than acute irAEs in patients with RCC (19.2% vs 6.0%, p = 0.036; Table 1). Other demographic background factors were distributed equally between the patients with acute and chronic toxicities (Table 1). Chronic toxicities were also equally common between adjuvant and metastatic patients (60.9% vs 61.0%, p = 0.994). Persistence of Chronic irAEs Beyond 12 Weeks Next, we evaluated the persistence of chronic toxicities at the 6-month time point and at the end of follow-up. Out of the 128 irAE+ patients, 112 patients completed 6 months of follow-up, and 50.9% (57/112) of them still presented with at least one chronic AE. This figure was 50.8% (65/128) at the end of follow-up. Among the 116 irAE+ patients with metastatic cancers and sufficient follow-up, 50.0% (45/90) presented with an AE at 6-month follow-up and 44.0% (51/116) at the end of follow-up. Among 23 irAE+ patients receiving adjuvant ICI therapy, 54.5% (12/22) presented with an AE at 6-month follow-up and 60.9% (14/23) at the end of follow-up. Types and Severity of Chronic irAEs Altogether, 196 irAE events were recorded, with 159 detected in the metastatic patient group and 37 in the adjuvant patient group. Colitis was observed more frequently (18.9% vs 5.4%, p = 0.046) and myositis less frequently (1.9% vs 10.8%, p = 0.025) in metastatic patients, compared with adjuvant patients, respectively (ESM, Table S3). Otherwise, irAEs were distributed evenly between these groups. For chronicity analysis, only irAEs with follow-up of ≥ 3 months were included (n = 178). Chronic irAEs were more common (52.8%, 94/178) than acute ones (47.2%, 84/178; Table 2). There was no difference in severity of acute and chronic toxicities (CTCAE grade I–II: 81.0% vs 81.9%, respectively, p = 0.869). The most frequent chronic AEs were thyroiditis (39.4%) and hepatitis (10.6%; Table 2). Endocrine AEs constituted 47.9% of chronic AEs. The most frequent acute AEs were dermatitis (39.3%) and colitis (27.4%; Table 2). The irAEs that most likely became chronic were nervous system AEs (100%), pancreatitis (100%) and type 1 diabetes mellitus (100%; Table 2). The irAEs least likely to become chronic were arthritis (0%), pneumonitis (0%) and colitis (14.8%; Table 2). Table 2Sites and frequencies of immune-related adverse events (irAEs) divided into subgroups based on chronicitySite of irAEn (%)All178 (100)Acute84 (47)Chronic94 (53)p-value (acute vs chronic)Thyroiditis42 (100)5 (11.9)37 (88.1)< 0.001Dermatitis41 (100)33 (80.5)8 (19.5)< 0.001Colitis27 (100)23 (85.2)4 (14.8)< 0.001Hepatitis16 (100)6 (37.5)10 (62.5)0.416Nervous system8 (100)0 (0.0)8 (100)0.007Myositis7 (100)2 (28.6)5 (71.4)0.449Nephritis6 (100)2 (33.3)4 (66.7)0.685Pancreatitis5 (100)0 (0.0)5 (100)0.061Arthritis5 (100)5 (100)0 (0.0)0.022Pneumonitis3 (100)3 (100)0 (0.0)0.103Myocarditis3 (100)1 (33.3)2 (66.7)1.000Diabetes Type 13 (100)0 (0.0)3 (100)0.248Hematologic2 (100)1 (50.0)1 (50.0)1.000Other endocrinea5 (100)0 (0.0)5 (100)0.061Otherb5 (100)3 (60.0)2 (40.0)0.668Statistically significant results are shown in boldaHypophysitis, hypocortisolismbXerostomia, cholangitis, iritis Sites and frequencies of immune-related adverse events (irAEs) divided into subgroups based on chronicity Statistically significant results are shown in bold aHypophysitis, hypocortisolism bXerostomia, cholangitis, iritis Survival of Metastatic Patients by irAE Chronicity The survival of patients with metastatic cancer was assessed in three subgroups based on their irAE status: (1) no irAEs, (2) acute irAEs and (3) chronic irAEs. When all patients were included in the analysis, the median OS and PFS times were 15.4 months, 21.7 months and 29.7 months (p = 0.009), and 4.2 months, 5.6 months and 10.3 months (p = 0.03), respectively. However, when survival was analyzed at 3- and 6-month landmarks, the difference no longer reached significance, suggesting immortal time bias (ESM, Table S4). Survival according to the four most common irAE sites (dermatitis, colitis, thyroiditis and endocrine disorders) is shown in the ESM, Table S4. No specific irAE site was associated with a better prognosis than other sites. Corticosteroid Therapy for irAEs Of the 196 reported AEs, 45.9% (90/196) were treated symptomatically, 42.9% (84/196) required systemic corticosteroid therapy and 11.2% (22/196) of irAEs did not require any treatment. Of the most common irAEs, only 6.5% (3/46) of thyroiditis cases and 29.5% (13/44) of dermatitis cases were treated with systemic corticosteroids, whereas corticosteroids were often required for the treatment of colitis (59.4%; 19/32), hepatitis (76.5%; 13/17) and nervous system toxicities (100%; 11/11). Corticosteroid therapy was used in 45.2% (38/84) of acute AEs and 36.2% (34/94) of chronic AEs. Among the most common chronic AEs, corticosteroids were required in 0.0% (0/37) of thyroiditis, 90.0% (9/10) of hepatitis, 100.0% (8/8) of nervous system toxicities and 25.0% (2/8) of dermatitis cases. No significant differences were observed in the corticosteroid treatment patterns between the adjuvant and metastatic groups (ESM, Table S5). We then analyzed the duration of irAE-specific corticosteroid therapy. For this purpose, we calculated the number of days of irAE-associated corticosteroid therapy after the occurrence of an irAE. In 49.3% (33/67) of the patients, corticosteroid therapy lasted for ≥ 3 months, and therapy duration tended to be longer for patients with chronic irAEs than acute irAEs (corticosteroid ≥ 3 months: 75.0% (24/32) vs 46.2% (12/26), p = 0.024). There was no significant difference in the corticosteroid usage for ≥ 3 months between metastatic (44.8%, 26/58) and adjuvant (77.8%, 7/9) patient groups (p = 0.083). Next, we assessed the median duration of corticosteroid therapy in ICI-treated irAE+ patients in comparison with irAE− patients regardless of the steroid indication. For this, we calculated the number of days taking corticosteroids after the initiation of ICI therapy to the end of follow-up. Overall, the ICI-treated patients spent a median of 41.5 days on corticosteroids, corresponding to 8.6% of their follow-up time. Among irAE+ patients, the median duration of corticosteroids was 56 days (10.6% of follow-up), compared with 29 days (8.4% of follow-up) in patients without irAEs (p = 0.015 for duration). The respective figures were 47 days (5.0% of follow-up) for acute irAEs and 68.5 days (8.1% of follow-up) for chronic irAEs (p = 0.587). There was no significant difference in the number of days on corticosteroid between all metastatic (45 days, 10.9%) and adjuvant (17 days, 3.8%) patients (p = 0.313). In the metastatic patient group, patients with irAEs spent 67 days (12.4% of follow-up) and patients without irAEs 32 days (8.8% of follow-up) on corticosteroids (p = 0.044). Among adjuvant patients, the respective figures were 28 days (6.6%) and 0 days (0.0%, p = 0.326). Finally, we analyzed the incidence of steroid-induced AEs (Table 3). Overall, 58.2% (39/67) of the irAE patients treated with corticosteroids developed AEs from steroids. These were equally common between metastatic (55.2%, 32/58) and adjuvant (77.8%, 7/9) patient groups (p = 0.285). Longer corticosteroid usage (≥ 3 months) for irAEs was associated with an increased risk of steroid-related AEs (< 3 months: 38.2% vs ≥ 3 months: 78.8%, p < 0.001). The most common steroid-related AEs were infections (n = 26), swelling (n = 9), new osteoporotic diagnosis and/or osteoporotic fracture (n = 7), new diagnosis or worsening balance of type 2 diabetes (n = 8), adrenal insufficiency (n = 6) and other AEs (n = 12). Table 3Duration and complications of corticosteroid treatmentAllan = 67 (%)< 3 months corticosteroid n = 34 (%)≥ 3 months corticosteroidn = 33 (%)p-valueNo. of complications of corticosteroids 028 (41.8)21 (61.8)7 (21.2)< 0.001 ≥ 1 39 (58.2)13 (38.2)26 (78.8)Type of complication Infection26 (38.8)10 (29.4)16 (48.5)0.087 Swelling9 (13.4)2 (5.9)7 (21.2)0.083 T2DM8 (11.9)2 (5.9)6 (18.2)0.150 Adrenal insufficiency6 (9.0)2 (5.9)4 (12.1)0.427 Osteoporosis-diagnosis and/or fracture7 (10.4)0 (0.0)7 (21.2)0.005 Otherb12 (17.9)4 (11.8)8 (24.2)0.183Statistically significant results are shown in boldT2DM type 2 diabetes mellitusaAmong patients receiving corticosteroid treatment to an immune-related adverse eventbElectrolyte imbalance, insomnia, muscle weakness, sweating Duration and complications of corticosteroid treatment Statistically significant results are shown in bold T2DM type 2 diabetes mellitus aAmong patients receiving corticosteroid treatment to an immune-related adverse event bElectrolyte imbalance, insomnia, muscle weakness, sweating Contacts and Consultations In total, 1038 irAE-related treatment contacts with a medical oncologist were recorded, corresponding to an average of 5.3 contacts per irAE (n = 196). Most of the contacts were phone calls (623/1038, 60%), followed by outpatient visits (287/1038, 27.6%), hospitalizations (72/1038, 6.9%) and silent consultations (56/1038, 5.4%), which corresponded to averages of 3.2, 1.5, 0.4 and 0.3 contacts per irAE, respectively. There were, on average, 6.1 contacts per chronic AE and 4.6 contacts per acute AE (p = 0.024), which mostly consisted of phone calls (3.9 vs 2.7 per irAE, p = 0.046). On average, there were 6.1 contacts per AE in the adjuvant patients and 5.5 contacts per AE in the metastatic patients (p = 0.194). The distribution of contacts was similar between irAEs in adjuvant and metastatic patients in all categories (Table 4). Table 4Absolute and average number of contacts needed for the management of irAEsContacts (N, average)AllOutpatient visitsPhone callsSilent consultationsHospitalizationsAll irAEs, n = 1961038, 5.3287,1.5623, 3.256, 0.372, 0.4Chronic irAEs, n = 94575, 6.1157, 1.7365, 3.925, 0.328, 0.3Acute irAEs, n = 84387, 4.6113, 1.3227, 2.723, 0.324, 0.3irAE location Colitis, n = 32196, 6.140, 1.3126, 3.914, 0.416, 0.5 Thyroiditis, n = 46169, 3.766, 1.483, 1.811, 0.29, 0.2 Dermatitis, n = 44161, 3.778, 1.875, 1.74, 0.14, 0.1 Hepatitis, n = 17170, 10.021, 1.2130, 7.66, 0.413, 0.8 Other, n = 57342, 6.082, 1.4209, 3.721, 0.430, 0.5Adjuvant, n = 37227, 6.144, 1.2164, 4.47, 0.212, 0.3Metastatic, n = 159881, 5.5243, 1.5459, 2.949, 0.360, 0.4irAEs immune-related adverse events Absolute and average number of contacts needed for the management of irAEs irAEs immune-related adverse events Regarding individual AEs, the management of hepatitis was associated with an average of 10.0 contacts per AE, followed by colitis with an average of 6.1 contacts per AE. The leading type of contact for both hepatitis and colitis was phone calls, with 7.6 and 3.9 average contacts, respectively. Management of the most common irAEs, thyroiditis and dermatitis, required an average of 3.7 contacts each. Hospital care for irAEs was required in 15.7% (36/230) of ICI-treated patients and 25.9% (36/139) of those developing an irAE. Hospitalization was equally common in acute and chronic irAEs (21.4% vs 16.0%, p = 0.348). The median hospitalization was 2.5 days for acute irAEs and 5 days for chronic irAEs (p = 0.611). There was no difference in the hospitalization rates between the metastatic (15.3%, 31/202) and adjuvant (17.9%, 5/28) groups. Of the 196 toxicities, inpatient care was needed for 22.4% (44/196) of the cases. The most common irAEs leading to hospitalization were colitis (30.4%, 7/23) and pneumonitis (100.0%, 3/3) (ESM, Table S6). To evaluate the burden of irAE-related hospitalization in comparison with the total hospitalization burden of ICI-treated patients, we calculated the number of days spent in the hospital, irrespective of the cause of admission from ICI initiation until the end of follow-up in patients without irAEs, outpatients with irAEs and those hospitalized because of an irAE. In metastatic patients, median hospitalization times were 5 days (n = 86), 4 days (n = 85) and 17 days (n = 31), respectively (p < 0.001). In pairwise comparisons, the difference was significant when comparing the no irAE and hospitalized patient groups (p < 0.001), and the outpatient and hospitalized patient groups (p < 0.001). The hospitalization duration for adjuvant patients was 1 day (n = 5), 0 days (n = 18) and 33 days (n = 5) (p = 0.029). Consultations regarding the treatment of irAEs from medical oncology to other specialties were conducted via in-person visits, phone calls, hospital admissions or silent consultations. The total number of consultations for irAEs was 158, corresponding to an average of 0.81 consultations per irAE (ESM, Table S6). Among these, phone consultations were the most frequent, constituting 83.5% (132/158) of all consultations. The most common individual AE leading to consultation was thyroiditis with 1.4 consultations per AE, followed by hepatitis with 0.7 consultations per AE (ESM, Table S7). Discussion This study aimed to characterize the long-term effects of ICI therapy on immune-related adverse events and the clinical workload associated with their management. Our results indicate that irAEs are frequently associated with long-term sequelae. First, the risk of irAE chronicity was high at 61% in this cohort. Second, the irAE treatment periods with systemic corticosteroids lasted for over 3 months in half of the patients and were associated with a 58% risk of steroid complications. Finally, when assessing the clinical workload, we found that approximately five contacts with a medical oncologist were required for the treatment of one irAE. Because robust risk factors for irAEs are lacking, the characterization of long-term complications and the clinical workload generated by irAEs is crucial for proper patient education and clinical resource planning. Of the irAE+ patients in our study, 61% developed chronic AEs. The presence of an irAE at 3-month follow-up predicted poor resolution, as the proportion of irAE+ patients was 50% at 6-month follow-up and 47% and at the end of follow-up (median follow-up, 37.5 months). Comparison with other reports is complicated by the differential study cohorts and definitions of chronicity; however, similar or slightly lower probabilities of chronicity with poor resolution likelihood during further follow-up have been reported. For example, a study including melanoma patients treated with adjuvant ICIs showed that 43.2% of patients developed chronic irAEs, and in 85% of the patients, the chronic AE persisted to last follow-up (median follow-up 529 days) [3]. Similarly, in another real-world study including 437 patients, 35.2% of the irAEs persisted for ≥ 6 months with a low likelihood of irAE resolution at 1-year and 2-year time points [8]. It is important to note that chronic irAEs comprise a wide spectrum of toxicities. In our study, 40% of the chronic toxicities were thyroiditis, which can often be rendered asymptomatic through replacement therapy. However, chronicity was also commonly associated with hepatitis, dermatitis or neurological toxicities, that is, inflammatory AEs with a smoldering nature involving symptomatic flares and the long-term need for optimization of immunosuppressive treatment. Different types of AEs may also have differing functional impacts on patients. For example, while the symptom burden and quality of life (QoL) effects of ICI-induced thyroiditis appear similar to or even lower than that of non-exacerbated autoimmune diseases [11], chronic nervous system irAEs are associated with severe and permanent neurological disability [12]. The absence of clear categorization and standardization across endocrine and inflammatory toxicities, however, complicates the assessment of chronicity, further impacting their effective and time-sensitive management. Future studies are necessary to determine patient-specific experiences with different types of chronic AEs. When comparing adjuvant and metastatic patients, we observed an equal irAE distribution, with a similar tendency for chronicity. Interestingly, however, we found that the incidence of irAEs was more common in the adjuvant group (82%) than in the metastatic group (57%). irAE chronicity is particularly noteworthy in adjuvant therapy patients, given the curative intent of their cancer treatment and their longer, often normal, life expectancies. Therefore, in adjuvant treatment, the benefit of a statistical reduction in the recurrence risk by ICI therapy could be outweighed by toxicity concerns. The burden of chronic AEs within metastatic patients is more difficult to assess due to the many confounding factors during the course of the disease, such as symptomatic progression or toxicities of follow-line therapies. Interestingly, the development of an irAE has previously been associated with improved survival outcomes of metastatic patients, suggesting that irAEs could be considered as a trade-off for survival benefit [13, 14]. In a study by Tong et al., the survival benefit was especially associated with chronic compared with transient irAEs [15]. Our initial results appeared similar to this study; however, the 3-month and 6-month landmark analyses revealed a substantial immortal time bias [8]; hence, we could not confirm this finding. Long-lasting corticosteroid periods for the treatment of irAEs in our cohort were common; almost half of the patients with acute AEs and 75% of those with chronic AEs received corticosteroids for ≥ 3 months. Importantly, corticosteroid treatment lasting ≥ 3 months was observed to be a risk factor for corticosteroid-related adverse effects. Along with swellings, infections were the most common AE, but due to confounding factors, such as the immunosuppressive effect of the cancer itself or possible chemotherapy treatments increasing the risk of infections, the figure is likely an overestimation. Alarmingly, 15% of irAE+ patients in our study treated with corticosteroids developed osteoporosis, some of which might have been prevented with early use of calcium and vitamin D and consideration of bone-targeted therapy. Guidelines do not offer specific tapering strategies for the glucocorticoid treatment of irAEs, which may result in unnecessarily long steroid periods. Protocols such as the previously proposed corticosteroid tapering strategy “fast first, then slowly” may be effective, reducing hospital stays and the median corticosteroid time [16]. Rapid tapering of corticosteroids, however, is often challenging, especially in patients with metastatic cancers, as corticosteroids are commonly used not only for the treatment of irAEs but also for symptom palliation and as antiemetics in patients with cancer. Indeed, when we analyzed the overall corticosteroid use in ICI-treated patients, we found a median duration of 32 days even without irAEs. Besides the risks of corticosteroid-associated complications, we and others have previously shown that there is emerging evidence suggesting that corticosteroid use may be associated with poorer outcomes among patients with irAEs [17, 18]. Given this complexity, close surveillance and prophylaxis of AEs from the start of corticosteroid administration are of importance. To our knowledge, this is the first study to describe the clinical workload associated with irAE treatment. We observed that treatment of a single adverse effect required an average of five contacts with hospital physicians. Hepatitis and colitis, AEs that are often severe, require high-dose corticosteroids and are associated with relapses during follow-up, were particularly burdensome. Colitis was the most common irAE, leading to specialty consultation (gastroenterology), followed by thyroiditis (endocrinology). While most irAE-related contacts were phone calls and outpatient visits, we also observed a high frequency of hospitalizations owing to irAEs. Importantly, hospitalization owing to irAEs significantly increased the overall inpatient time of ICI-treated patients, especially in adjuvant settings. This contrasts with a previous analysis in which disease progression, rather than an irAE, was the main reason for admission and accounted for more days of hospitalization [19]. Notably, these findings reflect the workload from a single-center perspective and likely underestimate the true number of patient contacts, for example, interactions with oncology nursing staff were not included in the analysis. The management of irAEs may vary significantly from one cancer care unit to another, depending on the resources, experience and size of the unit. For effective management of toxicities, providing irAE-treating oncologists with structured training may reduce corticosteroid periods, hospitalizations and the need for specialty consultations. Moreover, at least in larger units, a multidisciplinary consultation team may be an option for appropriate irAE management [20]. One limitation of this study was its retrospective, single-center design, which increased the risk of bias. Additionally, the modest size of the cohort, especially the adjuvant group, restricted the statistical power of the findings and limited the possibility of performing detailed stratified analyses, for example, according to irAE severity. In addition, the inclusion of combinatorial therapies (ICIs + chemotherapy or TKIs) in the cohort may have caused potential misclassification of AEs, however, the AEs from chemotherapy are usually distinct from irAEs, and the number of combinatorial therapies with TKIs was very low (n = 2). However, as the study included all ICI-treated patients from 2015 to 2022, it provided an accurate real-world description of ICI-treated patients. Other strengths of this study include the detailed characterization of the clinical treatment burden of irAEs, including contact, consultation and corticosteroid usage. Conclusions We showed that chronic irAEs are common but comprise mostly endocrine toxicities with likely a reasonable effect on QoL. Future prospective studies with longer follow-up and larger cohorts are required to assess patient QoL and the risk–benefit profile of chronic irAEs more specifically. Furthermore, novel approaches, for example based on peripheral blood markers [21], to predict the risk of developing irAEs and chronic irAEs are needed. Our results also demonstrate that effective steroid tapering is necessary to avoid complications if corticosteroid treatment for an irAE is required. Importantly, we provide a description of the clinical workload associated with irAEs, uncovering the most burdensome types of irAEs and the spectrum of contacts with a medical oncologist required for their management. These results may serve as a starting point for planning effective irAE management and resource distribution in cancer centers providing ICI treatment. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 195 KB) Abbreviations Adverse event Acetylsalicylic acid Arteriosclerosis obliterans Coronary artery disease Confidence interval Chronic obstructive pulmonary disease Common Terminology Criteria for Adverse Events Diabetes mellitus type 2 Follow up Hazard ratio Immune-checkpoint inhibitor Immune-related adverse event Intravenous Median overall survival Median progression-free survival Microsatellite instability Non-small cell lung cancer Overall survival Peripheral arterial disease Progression-free survival Per os Performance status Quality of life Renal cell carcinoma Tyrosine kinase inhibitor Triple negative breast cancer World Health Organization Funding This work was supported by State Research Funding for university-level health Research, Kuopio University Hospital, Wellbeing Service County of North Savo (project number 5654210), Cancer Foundation of Finland, Cancer Society of North Savo, The Finnish Medical Foundation (7074), Cancer Society of North Savo (20240026) and the Paavo Koistinen Foundation (202300011). Declarations Conflicts of Interest AR reports personal fees for speaking at symposia or financial support for attending conferences from Amgen, Bristol-Myers Squibb, Merck, Novartis, Pfizer and Roche. ST reports personal fees for speaking at symposia, consulting or financial support for attending conferences from Astra Zeneca, BMS, Merck, MSD, Novartis, Pfizer and Roche. None of the financial support was related to the present study. All other authors declare that they have no competing interests. Availability of Data and Material The datasets used and/or analyzed during this study are available from the corresponding author upon reasonable request. Ethics Approval This study was conducted in accordance with the Declaration of Helsinki (1964). The study received ethical approval from the Medical Research Ethics Committee of Wellbeing Services County of North Savo (permission number 1482/2019) and organization-level permission from Kuopio University Hospital (research permit 5654210, 916/2022). Due to the retrospective nature of the study, the Medical Research Ethics Committee of Wellbeing Services County of North Savo waived the informed consent requirement (permission number 1482/2019). Consent to Participate Not applicable. Consent for Publication Because of the retrospective design of this study, informed consent was not required. Code Availability Not applicable. Author Contributions All authors contributed to the study conceptualization and methodology. Investigation, data curation and formal analysis were performed by Veera Nurmela, Anni Juntunen and Aino Rönkä. The original draft was written by Veera Nurmela, Anni Juntunen and Aino Rönkä. All authors critically reviewed, edited and approved the final manuscript. Project was administrated and supervised by Aino Rönkä. Footnotes Veera Nurmela and Anni Juntunen contributed equally. References Associated Data Supplementary Materials Supplementary file1 (PDF 195 KB)