https://orcid.org/0000-0003-4444-7958
Figures
Abstract
Background
Immune checkpoint inhibitors (ICIs) are increasingly being used to treat malignancies. Some patients experience immune-related adverse events (irAEs), which may affect any organ/tissue. IrAEs are occasionally fatal and usually have nonspecific symptoms. We developed a three-step application (https://irae-search.com/) to provide healthcare professionals with information on the diagnosis, treatment options, and published reports for 38 categories of irAEs encountered in clinical practice.
Methods
IrAEs reported in ≥5 cases were identified from articles published between October 2018 and August 2020 by searching Japanese (SELIMIC, JAPIC-Q Service, and JMED Plus) and international (MEDLINE, EMBASE, Derwent Drug File) databases. The cases’ symptoms were entered into the application to identify irAEs, which were verified using the reported diagnosis, to evaluate the application’s sensitivity and specificity.
Results
Overall, 1209 cases (1067 reports) were analyzed. The three most common categories of irAEs were pituitary or adrenal disorders (14% of cases), skin disorders (13%), and diabetes mellitus (10%). The top three primary diseases were lung cancer (364 cases), melanoma (286 cases), and renal cell carcinoma (218 cases). The average sensitivity was 90.8% (range 44.4%–100.0%) initially, and improved to 94.8% (range 83.3%–100.0%) after incorporating the symptoms reported in published cases into the application’s logic for two irAE categories. The average specificity was 79.3% (range 59.1% [thyroid disorders]–98.2% [arthritis]).
Citation: Osawa T, Abe T, Kikuchi H, Matsumoto R, Murai S, Nakao T, et al. (2022) Validation of an online application to identify potential immune-related adverse events associated with immune checkpoint inhibitors based on the patient’s symptoms. PLoS ONE 17(3): e0265230. https://doi.org/10.1371/journal.pone.0265230
Editor: Hyunseok Kang, University of California, San Francisco, UNITED STATES
Received: October 13, 2021; Accepted: February 24, 2022; Published: March 15, 2022
Copyright: © 2022 Osawa et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and Supporting information files.
Funding: Ono Pharmaceutical Co., Ltd. (https://www.ono-pharma.com/) funded the study and medical writing support. Employees of Ono Pharmaceutical were involved in study design, data collection and analysis, decision to publish the study, and preparation of the manuscript.
Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: T. Osawa, T. Abe, R. Matsumoto, H. Kikuchi, and S. Murai declared no conflicts of interest. T. Nakao and S. Tanaka, are employees of and hold stock or stock options in Ono Pharma. A. Watanabe is an employee of Ono Pharma. N. Shinohara declared institutional research funds from Ono Pharma, Eisai Pharma, and Pfizer, and speakers’ fees from Takeda Pharmaceutical, Ono Pharma, BMS, MSD, Novartis, and Pfizer. Ono Pharmaceutical Co., Ltd. funded development of the online application and is the joint copyright holder with the corresponding author (T. Osawa). The company markets and is developing immune checkpoint inhibitors related to this work. This does not alter the authors’ adherence to all the PLoS ONE policies on sharing data and materials.
Introduction
Immune checkpoint inhibitors (ICIs) targeting the programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) and cytotoxic T-lymphocyte-associated protein 4 (CTLA4) pathways are among the newest entrants into the arsenal for cancer treatment [1]. Although the indications for ICIs vary, they are now available for treating many types of solid cancers and they offer remarkable benefits for patients with advanced diseases [1]. As a consequence, their use is significantly increasing in various oncological settings.
Although ICIs display proven anticancer efficacy, the enhanced immune response that occurs during treatment with these drugs may elicit immune-related adverse events (irAEs) in various body systems and organs [2–16]. Some irAEs are associated with improved overall survival or progression-free survival, and are hence regarded as a marker for the anticancer efficacy of some ICIs [17, 18]. In such cases, continuing treatment for as long as possible is desired from a therapeutic perspective. However, irAEs sometimes become serious, forcing treatment discontinuation, and may even be fatal if not detected and treated in a timely manner. Therefore, irAEs need careful and specialized management. Based on these concepts, early diagnosis and appropriate management of irAEs are vital in order to avoid serious outcomes and to help continuation of ICI therapy.
IrAEs are sometimes identified by members of the healthcare team, and sometimes worked up based on the patient’s symptoms, vital signs, and associated laboratory tests. Recommendations have been developed to aid the diagnosis and appropriate treatment of irAEs [19–22]. Clinical materials have also been introduced to help raise awareness of these events and to improve the diagnostic accuracy in patients treated with ICIs (see for example: [23–32]). However, the materials were often developed for use by specialists in each medical field, rather than by oncology providers in wider clinical practice. Furthermore, the materials are usually organized according to the type of irAE. This means that the healthcare professionals should use the materials with knowledge of the presence and type of irAE in the affected patient. Moreover, the common complaints (e.g., malaise, fatigue, headache), other symptoms, and laboratory test results are often nonspecific and may be misdiagnosed as something other than an irAE. In these circumstances, the patient may receive symptomatic or stop-gap therapies rather than a systemic treatment targeting the underlying cause. Therefore, there was an unmet need for developing an easy-to-use application to help healthcare professionals quickly identify potential irAEs based on the patient’s subjective symptoms, laboratory tests, or vital signs. In this context, we developed a three-step online application (https://irae-search.com/) that was designed to easily identify over 500 potential irAEs that may be encountered in clinical practice.
In this study, we performed a literature review of published cases of irAEs and used the application to evaluate the irAEs based on the symptoms reported in each case. We report the results of the literature search, and the sensitivity and specificity of the application for detecting 38 categories of irAEs. We also report the improvement in the application’s sensitivity that was achieved by revising the links in the application between symptoms and irAEs with low initial sensitivity.
Methods
Online application
A newly developed web application was used in this study. The application was developed by a software provider (3H Clinical Trial Corporation) in collaboration with the authors. The logic linking the symptoms and irAEs was developed and refined as follows. First, over 500 irAEs were extracted from specialized books of medical diagnosis [33, 34], Japanese Society of Medical Oncology Cancer Immunotherapy Guidelines [22], as well as the Package Inserts, Patient Guides, and Proper Use Guides for ICIs, including those for nivolumab and nivolumab–ipilimumab [35, 36]. Next, the irAEs were classified into 38 categories (Table 1 and S1 Table). Finally, the symptoms were listed for each irAE, and graded into three classes by specificity (highly specific symptoms for a particular irAE; frequently observed symptoms for a particular AE; and other symptoms that occur in many types of irAEs).
The application involves three steps: (1) Select symptoms from choices of common symptoms, (2) Select symptoms by body parts (ambiguous searches are possible), and (3) Select related symptoms and abnormal laboratory tests, which are displayed based on the selections made in steps 1 and 2 (Fig 1A). When a user selects a highly specific or a symptom common to a particular irAE, the application is designed to display the strongly suspected irAE in the final results screen. If the user only selects ‘other’ symptoms, the suspected irAE is only displayed if all of the ‘other’ selected symptoms are commonly linked to the particular irAE. A network map illustrating the links between the symptoms and irAEs is presented in S1 File.
(A) Schematic overview of the three steps of the application for identifying potential immune-related adverse events (irAEs) based on symptoms, laboratory tests, and vital signs. (B) Literature flow chart. (C) Frequencies of irAEs. NA (not applicable) indicates events that could not be classified into the 38 categories of irAEs included in the application. See Table 1 and S1 Table for the list of irAEs and associated diagnoses included in the application.
The application frontside uses a combination of HyperText Markup Language, cascading style sheets, and JavaScript running on a Java SE Development Kit 16.0.1 server and a Java-based Spring Boot framework. Data are stored in a PostgreSQL pgAdmin4 database running on Linux.
Literature search for application validation
In order to validate the application, we assessed its sensitivity and specificity by performing a comprehensive literature search of Japanese (SELIMIC, JAPIC-Q Service, and JMED Plus) and international (MEDLINE, EMBASE, and Derwent Drug File) databases. The search strategies were developed specifically for each database to identify case-reports describing nivolumab (Nivolumab, Opdivo, ONO-4538, BMS-936558, MDX-1106) and adverse events (adverse reactions, adverse events, toxicity, poisoning, safety, drug interactions, carcinogenicity, teratogenicity) published between October 2018 and August 2020. The published abstracts were retrieved. Articles were discarded if they did not describe the symptoms or a diagnosis, reported efficacy only, involved concomitant use with antineoplastic drugs other than an ICI, or described multiple adverse drug reactions. The literature search was conducted by researchers at Ono Pharmaceutical Co., Ltd., and data were extracted from each case-report in a retrospective manner.
Ethics
Ethical approval and consent were not required for the literature searches reported in this article, or for developing the application because it does not record any patient information in a database. Search logs are reviewed for the purpose of system improvements.
Data analysis
Patient characteristics, symptoms, and adverse drug reactions (including the type of irAE) were extracted from each abstract, tabulated in a database, and analyzed descriptively to determine the patient/disease characteristics by researchers at Ono Pharmaceutical Co., Ltd. in a non-blinded manner. The retrieved information was then used to determine the sensitivity and specificity of the application. The sensitivity was calculated as the rate at which a particular irAE was determined by this application to be an irAE that is finally diagnosed by the physician (i.e., the proportion of true positives that are correctly identified by the application) and the specificity was calculated as the rate of nonspecific irAEs (i.e., proportion of true negatives that are correctly identified) (S2 Table). After performing these analyses, it was discovered that some irAEs had a low sensitivity. Therefore, the links between symptoms and irAEs were reviewed and refined based on the information published in each case and clinical experience. As a consequence of the literature review, several new symptoms were added to the application. The sensitivity and specificity were then recalculated using the same cases. Microsoft Excel (Microsoft Corp., Redmond, WA, USA) was used for data aggregation and all analyses. The network map was visualized using the Compound Spring Embedder (CoSE) layout in Cytoscape 3.8.2 (https://cytoscape.org/) [37].
Results
Literature review
The literature search initially retrieved 3828 published reports, of which 1067 were included in the analyses (Fig 1B). These reports included 1209 cases treated with nivolumab or another ICI. The data extracted for these cases are presented in S3 Table. The cases included 835 men and 337 women (not recorded for 37 cases) (Table 2), with a mean age of 65.0 years (range 9–93 years; available in 1063 cases). The three most frequent primary diseases were lung cancer (364 cases), melanoma (286 cases), and renal cell carcinoma (218 cases). Nivolumab (988 cases), nivolumab plus ipilimumab (190 cases), pembrolizumab (63 cases), and ipilimumab (44 cases) were the four most frequently used treatments.
Verification of the application
Fig 1C shows the relative frequencies of irAEs reported in ≥5 cases; this figure includes >95% of cases in which irAEs were identified. IrAEs that occurred in more than ≥5% of cases were pituitary or adrenal disorders (14% of cases); SJS, TEN, pemphigoid, erythema multiforme, and other skin disorders (13%); diabetes mellitus (10%); lower gastrointestinal disorders (9%); interstitial pneumonia and other lung disorders (6%); myositis/myopathy (6%); and nephritis/renal disorders (5%).
The sensitivity and specificity of the application are shown in Table 3, which includes 21 categories of irAEs reported in ≥5 cases. The sensitivity ranged from 44.4% for upper gastrointestinal disorders to 100.0% for pituitary or adrenal disorders, lower gastrointestinal disorders, cholangitis, Vogt–Koyanagi–Harada disease, oral disorders, and cytokine release syndrome. The average sensitivity was 83.3%. The sensitivity was <80% for two categories, namely encephalitis/meningitis etc (51.4%) and upper gastrointestinal disorders (44.4%). The specificity ranged from 59.1% (thyroid disorders) to 98.2% (arthritis), with an average of 79.3%. The specificity was <80% for eight categories: pituitary or adrenal disorders; lower gastrointestinal disorders; myositis/myopathy; nephritis/renal disorders; hepatitis/hepatic disorders; cardiac disorders including myocarditis; serious haematologic disorders; and thyroid disorders.
Refinement of the application for irAEs with low sensitivity
Because the application’s sensitivity was relatively low (<80%) for upper gastrointestinal disorders and encephalitis/meningitis etc, we revised the links between symptoms and irAEs to improve the application’s sensitivity for predicting these categories of irAEs.
S4 Table summarizes the diagnoses and symptoms in reported cases with upper gastrointestinal disorders. Based on the literature search, we identified several common symptoms, including weight decreased (6 of 19 cases), dysphagia (5 of 19 cases), and anorexia (5 of 19 cases), which were originally included in the application but were not linked to upper gastrointestinal disorders. Adding links between these symptoms and this irAE to the application increased the sensitivity from 44.4% to 94.7% (S1 Fig). The specificity remained high (before: 88.3%; after: 83.9%) (Table 3).
Similarly, for encephalitis/meningitis etc, we identified several relatively common symptoms that were not originally linked to this irAE category or were not included as possible symptoms in the application (S5 Table). Dyslalia (2 of 35 cases) was replaced with language disorders (6 of 35 cases), and new links were added for memory impairment (2 of 35 cases) and feeling of weakness/muscular weakness (6 of 35 cases) (S2 Fig). These changes increased the sensitivity for encephalitis/meningitis etc from 51.4% to 85.7%. The specificity remained high (before: 91.0%; after: 87.6%) (Table 3).
The average sensitivity and specificity after implementing these changes were 94.8% and 79.3%, respectively.
Discussion
An online application (irAE Search) was developed with the aim of providing healthcare professionals with an easy-to-use search application to help the user identify potential irAEs based on a combination of symptoms, laboratory tests, and vital signs in the affected patient. It is intended for use by healthcare professionals, including those who are relatively inexperienced in using ICIs or their associated risks. Upon entering the information in three simple steps, the application provides the user with drug use guides and case reports relevant to the most probable irAEs. It has also provides links to many external references and clinically relevant resources. The healthcare professionals can access the application via a web-browser on a computer or via a smartphone/tablet.
In this study, we first verified the irAE categories by performing a systematic literature review, which retrieved 1209 cases. Among irAEs observed in ≥5 cases, the application had a sensitivity reaching 100.0% for several categories of irAEs (pituitary or adrenal disorders, lower gastrointestinal disorders, cholangitis, Vogt–Koyanagi–Harada disease, and oral disorders) Based on these findings, the application is expected to predict all true-positive cases for these categories. However, the application showed relatively low sensitivities for upper gastrointestinal disorders (44.4%) and encephalitis/meningitis etc (51.4%). Revising the links between symptoms and irAEs in the application using the information from the literature review increased the sensitivities for both categories (upper gastrointestinal tract disorders: from 44.4% to 94.7%; encephalitis/meningitis etc: from 54.1% to 85.7%). Finally, the sensitivity improved to >80% for all irAE categories, with an average of 94.8%.
The specificity of the application ranged from 59.1% (thyroid disorder) to 98.2% (arthritis), with an average of 79.3%. Because the specificity represents the accuracy of detecting a true negative case, there is a risk of incorrectly predicting irAEs with low specificity, such as endocrine disorders (59.8% for pituitary or adrenal disorders, 59.1% for thyroid disorders). Furthermore, some symptoms may overlap multiple categories of irAEs. For example, fatigue and anorexia were observed in cases with endocrine disorders. However, they may also occur in cases with other irAEs and are features of the general malaise of cancer patients [38]. For this reason, the specificity was relatively low for some irAEs. Generally, purely diagnostic tests/assays are expected to show high specificity to avoid incorrect diagnoses. However, our application is not intended to be used for purely diagnostic purposes. Rather, it is designed to raise suspicion of potential irAEs and prompt the physician to perform additional tests to help confirm the irAE and to consider appropriate interventions. The probable diagnoses are ordered (descending order) based on the weight applied to the selected symptoms in the application, which also highlights potentially fatal irAEs. Therefore, the low specificity for some irAEs is not necessarily a limitation of the application. In fact, the low specificity may be unavoidable to ensure high sensitivity and avoid overlooking irAEs in clinical practice.
The indications for ICIs are steadily increasing, particularly dual ICI therapy (e.g., nivolumab plus ipilimumab) [39] and their use is expected to expand beyond highly specialized centers. Healthcare professionals outside these centers may be less aware of the risks or signs of irAEs, and they are likely to benefit most from a simple application, such as this one. Thus, this application is valuable for helping healthcare professionals investigate suspected irAEs in patients treated with ICIs.
Limitations
There are some limitations of the application and the present study to discuss. The application is designed to identify potential AEs of ICIs and cannot be applied to AEs caused by other therapeutic reagents. For example, AEs specifically related to tyrosine kinase inhibitors cannot be appropriately identified in patients with renal cell cancer receiving tyrosine kinase inhibitor–ICI combination therapy, a regimen that will be used more frequently in the near future. Because the application is not embedded within electronic medical record systems, healthcare professionals need to be aware of the application to ensure they can access and use it. It is currently only available in Japanese, although an English version is under consideration. For this study, we searched the medical literature using terms to identify nivolumab-treated cases. However, these terms retrieved some cases treated with other ICIs (with or without nivolumab), but these cases were included in the analyses. Cases with concurrent irAEs and cases who received concomitant cytotoxic anticancer therapy were excluded from the study. Because we did not evaluate the sensitivity or specificity for irAEs that occurred in <5 cases, caution should be exercised in such cases. Additionally, because case reports often describe rare diseases/disorders, it is possible that our search overestimated the prevalence of rare disorders or symptoms. Another possible limitation is that the data were extracted in a retrospective manner from published case reports by researchers who were aware of the final diagnosis; these factors may influence the sensitivity or specificity. This study did not investigate the feasibility or utility of the application by healthcare professionals in clinical practice. Future studies, including surveys of the users, may provide insight into its use and opportunities for further development.
Conclusions
In conclusion, irAE Search is an easy-to-use only application that was designed to help healthcare professionals identify potential irAEs in patients treated with ICIs in a timely manner and to facilitate prompt management/treatment. The application showed high sensitivity and moderate-to-high specificity for detecting irAEs in clinical practice. For two categories of irAEs, the sensitivity was improved by incorporating information about symptoms from the published cases. Ongoing development of the application is expected to yield further improvements in the sensitivity and specificity for identifying potential irAEs.
Supporting information
S1 Table. Immune-related adverse events and associated diagnoses included in the online application.
https://doi.org/10.1371/journal.pone.0265230.s001
(PDF)
S2 Table. Assessment of sensitivity and specificity.
https://doi.org/10.1371/journal.pone.0265230.s002
(PDF)
S3 Table. Examples of cases of immune-related adverse events identified by the literature search.
https://doi.org/10.1371/journal.pone.0265230.s003
(PDF)
S4 Table. Diagnoses and symptoms observed in 19 cases with upper gastrointestinal disorders.
https://doi.org/10.1371/journal.pone.0265230.s004
(PDF)
S5 Table. Diagnoses and symptoms observed in 35 cases with encephalitis/meningitis etc.
https://doi.org/10.1371/journal.pone.0265230.s005
(PDF)
S1 Fig. Uplift in sensitivity for predicting upper gastrointestinal disorders after incorporating links to common symptoms described in 19 cases retrieved by the literature search.
https://doi.org/10.1371/journal.pone.0265230.s006
(PDF)
S2 Fig. Uplift in sensitivity for predicting encephalitis/meningitis etc after incorporating links to common symptoms described in 35 cases retrieved by the literature search.
https://doi.org/10.1371/journal.pone.0265230.s007
(PDF)
S1 File. Network map illustrating the links between the symptoms and irAEs.
The symptoms and irAEs are linked using red or grey lines. Red lines indicate high specificity of a symptom for a particular irAE, grey lines indicate low specificity of a symptom to each linked irAE. Symptoms are colored blue, where the intensity of color represents the number of irAEs the symptom is connected to. Light blue symptoms are linked to one or a few irAEs, and are classified as highly specific symptoms. By contrast, dark blue (e.g., pyrexia/chills, nausea/vomiting, and abdominal pain) indicates broad symptoms, which are linked to many irAEs.
https://doi.org/10.1371/journal.pone.0265230.s008
(PDF)
S2 File. Complete list of cases retrieved by the literature search.
https://doi.org/10.1371/journal.pone.0265230.s009
(XLSX)
Acknowledgments
The authors thank Dr Nicholas D. Smith and Dr Junko Isoyama (EMC K.K.) for medical writing support.
References
- 1. Vaddepally RK, Kharel P, Pandey R, Garje R, Chandra AB. Review of indications of FDA-approved immune checkpoint inhibitors per NCCN guidelines with the level of evidence. Cancers (Basel). 2020;12:738. pmid:32245016.
- 2. Alba-Linero C, Alba E. Ocular side effects of checkpoint inhibitors. Surv Ophthalmol. 2021;66:951–959. pmid:33440195.
- 3. Apalla Z, Papageorgiou C, Lallas A, Delli F, Fotiadou C, Kemanetzi C, et al. Cutaneous adverse events of immune checkpoint inhibitors: a literature review. Dermatol Pract Concept. 2021;11:e2021155. pmid:33614223.
- 4. Bukamur H, Katz H, Alsharedi M, Alkrekshi A, Shweihat YR, Munn NJ. Immune checkpoint inhibitor-related pulmonary toxicity: focus on nivolumab. South Med J. 2020;113:600–605. pmid:33140115.
- 5. Cappelli LC, Bingham CO 3rd. Expert perspective: immune checkpoint inhibitors and rheumatologic complications. Arthritis Rheumatol. 2020;73:553–565. pmid:33186490.
- 6. El Sabbagh R, Azar NS, Eid AA, Azar ST. Thyroid dysfunctions due to immune checkpoint inhibitors: a review. Int J Gen Med. 2020;13:1003–1009. pmid:33177863.
- 7. Gopee NH, Gourley AM, Oliphant TJ, Hampton PJ. Toxic epidermal necrolysis occurring with immune checkpoint inhibitors. Dermatol Online J. 2020;26:8fc428f426. pmid:32941713.
- 8. Gürdoğan M, Yalta K. Myocarditis associated with immune checkpoint inhibitors: Practical considerations in diagnosis and management. Anatol J Cardiol. 2020;24:68–75. pmid:32749254.
- 9. Herrmann SM, Perazella MA. Immune checkpoint inhibitors and immune-related adverse renal events. Kidney Int Rep. 2020;5:1139–1148. pmid:32775813.
- 10. Ho AK, Ho AM, Cooksley T, Nguyen G, Erb J, Mizubuti GB. Immune-related adverse events associated with immune checkpoint inhibitor therapy. Anesth Analg. 2021;132:374–383. pmid:33009134.
- 11. Lo Preiato V, Salvagni S, Ricci C, Ardizzoni A, Pagotto U, Pelusi C. Diabetes mellitus induced by immune checkpoint inhibitors: type 1 diabetes variant or new clinical entity? Review of the literature. Rev Endocr Metab Disord. 2021;22:337–349. pmid:33409866.
- 12. Malnick SDH, Abdullah A, Neuman MG. Checkpoint inhibitors and hepatotoxicity. Biomedicines. 2021;9:101. pmid:33494227.
- 13. Pirozzi F, Poto R, Aran L, Cuomo A, Galdiero MR, Spadaro G, et al. Cardiovascular toxicity of immune checkpoint inhibitors: clinical risk factors. Curr Oncol Rep. 2021;23:13. pmid:33415405.
- 14. Sakamoto K, Fukihara J, Morise M, Hashimoto N. Clinical burden of immune checkpoint inhibitor-induced pneumonitis. Respir Investig. 2020;58:305–319. pmid:32713811.
- 15. Tinawi M, Bastani B. Nephrotoxicity of immune checkpoint inhibitors: acute kidney injury and beyond. Cureus. 2020;12:e12204. pmid:33489613.
- 16. Vilariño N, Bruna J, Kalofonou F, Anastopoulou GG, Argyriou AA. Immune-driven pathogenesis of neurotoxicity after exposure of cancer patients to immune checkpoint inhibitors. Int J Mol Sci. 2020;21:5774. pmid:32796758.
- 17. Hussaini S, Chehade R, Boldt RG, Raphael J, Blanchette P, Maleki Vareki S, et al. Association between immune-related side effects and efficacy and benefit of immune checkpoint inhibitors—A systematic review and meta-analysis. Cancer Treat Rev. 2021;92:102134. pmid:33302134.
- 18. Wang D, Chen C, Gu Y, Lu W, Zhan P, Liu H, et al. Immune-related adverse events predict the efficacy of immune checkpoint inhibitors in lung cancer patients: a meta-analysis. Front Oncol. 2021;11:631949. pmid:33732650.
- 19. Brahmer JR, Lacchetti C, Schneider BJ, Atkins MB, Brassil KJ, Caterino JM, et al. Management of immune-related adverse events in patients treated with immune checkpoint inhibitor therapy: American Society of Clinical Oncology clinical practice guideline. J Clin Oncol. 2018;36:1714–1768. pmid:29442540.
- 20. Puzanov I, Diab A, Abdallah K, Bingham CO 3rd, Brogdon C, Dadu R, et al. Managing toxicities associated with immune checkpoint inhibitors: consensus recommendations from the Society for Immunotherapy of Cancer (SITC) Toxicity Management Working Group. J Immunother Cancer. 2017;5:95. pmid:29162153.
- 21. Nagai H, Muto M. Optimal management of immune-related adverse events resulting from treatment with immune checkpoint inhibitors: a review and update. Int J Clin Oncol. 2018;23:410–420. pmid:29516216.
- 22.
Japanese Society of Medical Oncology. Cancer Immunotherapy Guidelines, 2nd ed. Tokyo: Japanese Society of Medical Oncology; 2019. In Japanese.
- 23. Johnson DB, Reynolds KL, Sullivan RJ, Balko JM, Patrinely JR, Cappelli LC, et al. Immune checkpoint inhibitor toxicities: systems-based approaches to improve patient care and research. Lancet Oncol. 2020;21:e398–e404. pmid:32758477.
- 24. da Silva GFM, Landim JIVD, Dos Santos Brasil LT, Plens ICM, Silva ALC, Scheinberg MA, et al. Knowledge gap about immune checkpoint inhibitors among rheumatologists and medical students: a survey. Rheumatol Int. 2020;41:939–942. pmid:32767082.
- 25. Fazer C. Checkpoint inhibitor immunotherapy for head and neck cancer: incorporating care step pathways for effective side-effect management. J Adv Pract Oncol. 2019;10(Suppl 1):37–46. pmid:33014516.
- 26. Hoffner B, Rubin KM. Meeting the challenge of immune-related adverse events with optimized telephone triage and dedicated oncology acute care. J Adv Pract Oncol. 2019;10(Suppl 1):9–20. pmid:33014514.
- 27. Khan OF, Monzon J. Diagnosis, monitoring, and management of adverse events from immune checkpoint inhibitor therapy. Curr Oncol. 2020;27(Suppl 2):S43–S50. pmid:32368173.
- 28. Londoño MC, Reig M. Multidisciplinary clinical approach to cancer patients with immune-related adverse events induced by checkpoint inhibitors. Cancers (Basel). 2020;12:3446. pmid:33228219.
- 29. Naidoo J, Zhang J, Lipson EJ, Forde PM, Suresh K, Moseley KF, et al. A multidisciplinary toxicity team for cancer immunotherapy-related adverse events. J Natl Compr Canc Netw. 2019;17:712–720. pmid:31200355.
- 30. Renna CE, Dow EN, Bergsbaken JJ, Leal TA. Expansion of pharmacist clinical services to optimize the management of immune checkpoint inhibitor toxicities. J Oncol Pharm Pract. 2019;25:954–960. pmid:30975067.
- 31. Rubin KM. Understanding immune checkpoint inhibitors for effective patient care. Clin J Oncol Nurs. 2015;19:709–717. pmid:26583635.
- 32. Wood LS, Moldawer NP, Lewis C. Immune checkpoint inhibitor therapy: key principles when educating patients. Clin J Oncol Nurs. 2019;23:271–280. pmid:31099804.
- 33.
Fuji T, Nara N, editors. Medical Diagnoses, 3rd ed. Tokyo: Igaku-Shoin; 2016. In Japanese.
- 34.
Nagai R, editor. Today’s Diagnosis, 7th ed. Tokyo: Igaku-Shoin; 2015. In Japanese.
- 35.
Ono Pharmaceutical Co. Ltd. [Proper use guide for nivolumab monotherapy]. https://www.opdivo.jp/basic-info/documents [Cited 2021 June 18]. In Japanese.
- 36.
Ono Pharmaceutical Co. Ltd. [Proper use guide for nivolumab plus ipilimumab combination therapy]. https://www.opdivo.jp/basic-info/documents [Cited 2021 June 18]. In Japanese.
- 37. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498–2504. pmid:14597658.
- 38. Thong MSY, van Noorden CJF, Steindorf K, Arndt V. Cancer-related fatigue: causes and current treatment options. Curr Treat Options Oncol. 2020;21:17. pmid:32025928.
- 39. Robert C, Lanoy E, Besse B. One or two immune checkpoint inhibitors? Cancer Cell. 2019;36:579–581. pmid:31951559.