In recent years, immunotherapy has become one of the pillars of cancer treatment. Extensive research in cancer biology has led to a deeper understanding of cancer cell survival mechanisms and the importance of the tumor microenvironment. Not only are we gradually uncovering the essential principles of cancerogenesis and immune escape, but we are also using this knowledge therapeutically. One of the most groundbreaking findings is definitely the discovery of tumor cells´ exploitation of immune checkpoints and the introduction of immune checkpoint inhibitors (ICIs).

A normally functioning immune system should be able to distinguish the body´s own antigens from foreign or mutated self-proteins and this executive role is borne by T cells (Buchbinder and Desai, 2016; Jelinek and Hajek, 2016). The vital part of the recognition process is the interaction between a T cell receptor (TCR) and a major histocompatibility complex (MHC) on the surface of the antigen presenting cell (APC), which displays the antigen in question. In addition, there are many other co-stimulatory or co-inhibitory pathways (immune checkpoints) that help to determine the fate of the T lymphocyte. Stimulatory pathways lead to T cells´ activation, maturation, expansion and the inhibition of their apoptosis, so they can eliminate defective or autoreactive cells, whereas inhibitory pathways have the opposite effect, causing T cell anergy and exhaustion and inducing tolerance (Buchbinder and Desai, 2016; Jelinek et al., 2018; Hradska et al., 2020).

The two most important negative regulators of T cell immune function are cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) and programmed death 1 (PD-1) immune checkpoints (Buchbinder and Desai, 2016). CTLA-4 helps to prevent autoreactivity in the early stages of immune response, especially in the lymph nodes, and can be found in the cytoplasm of naïve T cells and on the surface of regulatory T cells (Tregs). Upon binding to its cognate ligands B7-1 and B7-2 on the APCs, it leads to T cell anergy (Buchbinder and Desai, 2016; Matsuki and Younes, 2016; Jelinek et al., 2017; Hradska et al., 2020). PD-1 is typical for “exhausted” T cells that have experienced high levels of stimulation, for example, during chronic infection or cancer, leading to their suboptimal function. However, various immune cells such as B cells and natural killer cells also show PD-1 expression. PD-1 ligands PD-L1 and PD-L2 can be found on various hematopoietic (lymphocytes, APCs) and non-hematopoietic cells (endothelial, parenchymatic cells), but are also induced on many types of tumor cells. PD-1/PD-L1/2 pathway is effective later in the immune response, primarily in peripheral tissues and, under normal circumstances, maintains peripheral tolerance during the course of infection and prevents autoimmunity (Buchbinder and Desai, 2016; Wang et al., 2018).

The rationale behind immune checkpoint blockade is to reinvigorate the immune system, which is inhibited by overexpression of immune checkpoint inhibitor molecules on the surface of cancer cells, to restore T cell cytotoxicity and to destroy defective cells (Jelinek et al., 2018; Hradska et al., 2020). Immune checkpoint inhibitors, monoclonal antibodies (mAbs) targeting PD-1, PD-L1/2 and CTLA-4, were introduced in 2011 and changed the landscape of especially solid tumor cancer treatment so fundamentally that this breakthrough led to the award of the Nobel Prize in 2018 jointly to James P. Allison and Tasuku Honjo (Smyth and Teng, 2018). In hemato-oncology, only two molecules have been approved by the FDA (US Food and Drug Administration) - nivolumab (anti-PD-1 mAb) for the treatment of relapsed/refractory (RR) classical Hodgkin lymphoma (cHL) and pembrolizumab (anti-PD-1 mAb) also for the treatment of RR cHL and for RR primary mediastinal large B cell lymphoma (PMBL) as well (Ansell et al., 2015; Armand et al., 2016, 2018a, 2018b; Chen et al., 2017). However, this is just the tip of the iceberg as plenty of ICIs are investigated in many other hematological malignancies such as non-Hodgkin lymphomas (NHLs), multiple myeloma (MM), acute myeloid leukemia (AML), myelodysplastic syndrome (MDS) and chronic lymphocytic leukemia (CLL).

The new anti-cancer strategy came with a new specific type of toxicity which is known as immune related adverse events (irAEs) (Haanen et al., 2017; Fan et al., 2021). The pathophysiological mechanism is not yet fully understood, but irAEs are considered to be T cell mediated and mark the reinvigorated immune system which may cause organ-specific inflammation, tissue damage and autoimmunity (Chhabra and Kennedy, 2021). Activation of the complement may also play a role in the process (Roth et al., 2021). The most frequently affected tissues are skin, gastrointestinal system, endocrine glands, liver and lungs, but irAEs may involve any organ system. They usually occur within weeks to 3 months after the initiation of ICI therapy and are mild to moderate in severity (Haanen et al., 2017). However, lethal cases are also observed, so it is very important for physicians to recognize potential problems early and to take appropriate measures as soon as possible. IrAEs are classified according to the CTCAE (Common Terminology Criteria for Adverse Events) scoring system.

There are many reviews describing ICIs´ immune related toxicity in solid tumors or in cancer treatment in general. Nevertheless, ICIs are gradually building their position in the treatment of hematological malignancies and, in this article, we would like to overview the incidence of irAEs in this specific setting and also offer a brief insight into the management of irAEs. We took into account all available safety data from blood cancer ICI prospective clinical trials (both monotherapy and combination therapy) until April 2021 using Pubmed database and Google Scholar. The results are divided by affected organ systems.

Dermatologic toxicity is the most common type of irAE in hematological malignancies. Maculopapular rash, dermatitis and pruritus represent the typical manifestation, but erythema nodosum, vitiligo, xerosis, psoriasis or even exfoliative dermatitis, skin necrosis and Stevens-Johnson syndrome were also observed (Badros et al., 2017; Khouri et al., 2018; Mateos M.-V. et al., 2019, Mateos et al., 2019 M. V.; Diefenbach et al., 2020; Song et al., 2020).

Rash and pruritus occur in 0–17% patients receiving ipilimumab (anti-CTLA-4 mAb) in monotherapy and are mostly mild (grade 1–2). Interestingly, there was no skin toxicity attributed to ipilimumab administered in the post-alogeneic stem cell transplant setting in the study conducted by Davids et al. (2016). Incidence was higher and irAEs slightly more severe in combination therapy, either with another ICI such as nivolumab (18–36%) or with other immunotherapeutic agents (6–61%) (Khouri et al., 2018; Diefenbach et al., 2020). Double immune checkpoint blockade even led to grade 4 Stevens-Johnson syndrome in one patient treated for cHL (Diefenbach et al., 2020).

Nivolumab (anti-PD-1 mAb) was associated with comparable all-grade skin toxicity in monotherapy (9–47%) and in combination with other targeted therapeutics (13–31%) (Armand et al., 2018a; Zinzani et al., 2019; Diefenbach et al., 2020; Maruyama et al., 2020). However, grade 3–4 cases were more common in the second cohort. Lower skin toxicity was observed in 2 combination studies of nivolumab and chemotherapy (AVD - doxorubicin, vinblastine, dacarbazine) for the treatment naïve cHL, where only 0–6% of patients experienced skin irAEs (Ramchandren et al., 2019; Bröckelmann et al., 2020).

Another FDA-approved PD-1 inhibitor, pembrolizumab, caused dermatologic irAEs in 0–33% of patients (Khodadoust et al., 2016; Armand et al., 2019). In the CITN-10 study, it led to worsening of erythema and pruritus in patients with Sezary syndrome, but not in mycosis fungoides (Khodadoust et al., 2016). The flare reaction was associated with high expression of PD-1 on circulating Sézary cells, mostly remitted within 12 weeks with only symptomatic treatment and topical corticosteroids and did not lead to treatment discontinuation. Pembrolizumab combined with immunomodulatory drugs (IMiDs) and dexamethasone was extensively investigated in multiple myeloma. Skin related toxicity was not very frequent (2–10%), but could be considered more severe, because one patient died of treatment-related grade 5 Stevens-Johnson syndrome (D’Souza et al., 2019; Mateos M. V. et al., 2019, Mateos et al., 2019 M.-V.). Interestingly, one myeloma patient developed vitiligo on ICI treatment, which is almost exclusively seen in melanomas (Badros et al., 2017; Ransohoff and Kwong, 2017; Chhabra and Kennedy, 2021).

Anti-PD-L1 mAbs are not yet so widely investigated as anti-PD-1 mAbs, but the skin toxicity seems comparable with rash and dry skin reported from available clinical trial data (Kazandjian et al., 2021; Ribrag et al., 2021). (Table 1)

Grade 1 (affecting <10% of body surface area = BSA) and grade 2 (10–30% BSA) dermatologic irAEs are typically managed with topical emollients, topical steroids and/or antihistamines, avoiding skin irritants or sun exposure and usually do not lead to ICI treatment discontinuation (Haanen et al., 2017; Chhabra and Kennedy, 2021). More severe cases are associated with affection of >30% BSA and cooperation with a dermatology specialist and punch biopsy is recommended. Grade 3 cases are managed with oral (prednisolone 0.5–1 mg/kg) or intravenous (methylprednisolone 0.5–1 mg/kg) systemic steroids and, after resolving the symptoms to grade 1 or mild grade 2, ICIs can be restarted after thorough risk/benefit assessment and discussion with the patient. Steroids are recommended to be weaned over 1–2 and 2–4 weeks respectively, because of the risk of relapse. Grade 4 is a life-threatening situation associated with symptoms such as erythema, purpura or epidermal detachment, requires rapid administration of intravenous steroids (methylprednisolone 1–2 mg/kg) and permanent discontinuation of ICI treatment.

Patients receiving ICI treatment relatively often experience irAEs affecting endocrine organs - mostly the thyroid gland, pituitary gland, adrenal glands and pancreas (Chang et al., 2019; Chhabra and Kennedy, 2021). Anti-CTLA-4 mAbs are connected with higher rates of hypophysitis, whereas PD-1 inhibitors more frequently induce hypothyroidism, hyperthyroidism or thyroiditis. Type 1 diabetes mellitus and adrenal insufficiency are less common.

Ipilimumab use in hematological malignancies, both in monotherapy or combination therapy, was associated with low incidence of thyroid dysfunction (0–6%) (Davids et al., 2016; Khouri et al., 2018). Only one case of suspected mild hypophysitis was reported in a phase 1 study for RR B-NHL (Ansell et al., 2009). This patient with a history of type 2 diabetes mellitus developed hyperglycemia during upper respiratory tract infection, anorexia, anemia, weight loss and erectile dysfunction. Laboratory findings uncovered low serum testosterone, low-normal gonadotropin levels and a normal MRI pituitary scan.

Both FDA-approved PD-1 inhibitors, nivolumab and pembrolizumab, showed similar thyroid toxicity. Hypothyroidism occurred in 0–29% and 0–17% of cases, respectively, while hyperthyroidism was reported in 0–13% and 0–17% of cases, respectively (Khodadoust et al., 2016; D’Souza et al., 2019; Younes et al., 2019; Zinzani et al., 2019; Maruyama et al., 2020). Adrenal insufficiency was also observed in both of these agents with the maximum incidence of 6% (Badros et al., 2017; Armand et al., 2018a; Barta et al., 2019; Ribrag et al., 2019). Type 1 diabetes mellitus was associated with nivolumab use in two RR cHL clinical trials (1 patient each) (Armand et al., 2018a; Maruyama et al., 2020). The rare endocrine irAE, hypoparathyroidism, affected one patient with cHL while receiving camrelizumab (anti-PD-1 mAb) and decitabine (hypomethylating agent) combination treatment (Liu et al., 2021). There was no endocrine toxicity observed in anti-PD-L1 based hematological clinical trials and in all available studies investigating ICI use in acute leukemia or myelodysplastic syndrome (Ravandi et al., 2019; Cassaday et al., 2020; Kazandjian et al., 2021; Ribrag et al., 2021). (Table 2)

Endocrine disorders during ICI therapy are usually mild to moderate (grade 1 and 2), but in a minority of cases, can be severe or life-threatening (Haanen et al., 2017; Chang et al., 2019; Chhabra and Kennedy, 2021). Thyroid-targeting irAEs are seen in up to 80% of cases caused by anti-thyroid antibodies. Hyperthyroidism is often transient and within 3–6 weeks usually leads to permanent hypothyroidism which does not resolve after ICI discontinuation and requires life-long thyroid hormone replacement. It is recommended to perform thyroid function tests (TSH = thyroid-stimulating hormone, fT4 = free thyroxine, fT3 = free tri-iodothyronine) every cycle for the first 3 months of anti-PD-(L)1 treatment and every second cycle thereafter. Cortisol level should be taken when TSH is falling and/or symptoms suggesting secondary (central) hypothyroidism occur. It can be a sign of hypopituitarism and potentially life-threatening central adrenal insufficiency when rapid corticosteroid replacement prevents development of adrenal crisis. In the rare cases of symptomatic thyreotoxicosis, beta-blockers (propranolol or atenolol) are recommended. Antithyroid drugs are not routinely used unless Graves´ disease is diagnosed. In the case of painful thyroiditis, prednisolone in the dosage of 0.5 mg/kg should be considered. Thyroid disorders usually do not lead to ICI discontinuation.

Blood glucose levels should be monitored regularly in patients treated with ICIs because of the danger of de novo type 1 diabetes mellitus (Haanen et al., 2017; Chang et al., 2019). Patients with pre-existing type 2 diabetes mellitus can develop diabetic ketoacidosis. These complications should be managed according to local practice; high-dose steroids are not recommended. Affected patients can restart ICI treatment after they are stable and insulin therapy is adjusted. ICI discontinuation does not lead to the restitution of the endogenous insulin level.

Primary adrenal insufficiency (PAI) differs from abovementioned central adrenal insufficiency by elevated ACTH (adrenocorticotropic hormone) and low cortisol level (Chang et al., 2019). The difference is also in management, as PAI requires rapid mineralocorticoid replacement in addition to glucocorticoids.

ICI related liver injury shares many features with autoimmune hepatitis. In most cases, it presents with asymptomatic elevation of alanine aminotransferase (ALT), aspartate aminotransferase (AST) or total bilirubin, but severe hepatitis and liver failure associated with fever, malaise, jaundice and change of stool color can also occur (Tian et al., 2018; Chhabra and Kennedy, 2021).

The incidence of abnormal ALT and AST levels is comparable between CTLA-4 and PD-1 inhibitors. The most prominent treatment-related hepatotoxicity was observed in a phase 1/2 clinical trial investigating the efficacy of ipilimumab, nivolumab and brentuximab vedotin (BV) in RR cHL patients - the rate of ALT/AST increase was as high as 48%/39% for the ipilimumab + BV cohort and 47%/32% for the nivolumab + BV cohort (Diefenbach et al., 2020). All of these events were grade 1 to 2 in severity. Singular grade 3 cases were detected in the triple therapy group. Slightly less toxic anti-PD-1 mAbs appeared to be pembrolizumab and camrelizumab with only 0–8% and 0% proportion of patients experiencing liver dysfunction, respectively (Khodadoust et al., 2016; Armand et al., 2019; Mei et al., 2020; Liu et al., 2021). Hepatotoxicity was also reported in anti-PD-L1 hemato-oncological trials, but the cohorts were small, so the results should be regarded with caution (Kazandjian et al., 2021; Ribrag et al., 2021). (Table 3)

When considering hepatic irAE, other causes of liver injury such as medication, autoimmunity, viral infection and alcohol should be ruled out in the first place (Haanen et al., 2017; Tian et al., 2018; Chhabra and Kennedy, 2021). Imaging methods (CT = computer tomography, MRI = magnetic resonance imaging, Doppler ultrasound) or liver biopsy could help in disputable situations. Management is dependent on severity of damage. Grade 1 (1-3x ULN = upper limit of normal) cases are closely monitored at least once weekly and ICI treatment can continue. When ALT/AST levels reach grade 2 (3-5x ULN), ICI should be temporary withheld and oral corticosteroid therapy initiated (prednisolone 1 mg/kg). Grade 3 and 4 liver injuries (5-20x ULN and >20 ULN, respectively) lead to permanent ICI discontinuation and systemic steroid treatment. In grade 3 cases with ALT/AST<400, normal bilirubin, INR (International Normalized Ratio) and albumin oral prednisolone 1 mg/kg is suggested. Higher dosage (2 mg/kg) of prednisolone should be administered when there is ALT/AST>400, elevated bilirubin or INR or decreased albumin. Grade 4 toxicity is managed by (methyl)prednisolone 2 mg/kg. Steroid-refractory patients might benefit from mycophenolate mofetil (500–1000 mg twice a day), tacrolimus or anti-thymocyte globulin. Infliximab is not recommended due to its potential hepatotoxic effect.

Gastrointestinal (GI) irAEs occur frequently, especially during anti-CTLA-4 mAb treatment (Rocha et al., 2019; Chhabra and Kennedy, 2021). Their clinical presentation can be variable with diarrhoea, abdominal pain, hematochezia, weight loss, fever, vomiting, mouth ulcers, anal lesions and even extra-intestinal manifestations reported (Haanen et al., 2017). Upper GI symptoms (nausea, vomiting, dysphagia, epigastric pain) are less common (Haanen et al., 2017; Rocha et al., 2019).

In hemato-oncological clinical trials, ipilimumab showed similar GI toxicity as in solid tumors. Diarrhoea was the most usual manifestation, observed in a maximum of 56 and 61% patients receiving ipilimumab monotherapy or combination therapy, respectively (Ansell et al., 2009; Diefenbach et al., 2020). Colitis was reported in a maximum of 7 and 5%, respectively (Zeidan et al., 2018; Diefenbach et al., 2020). Double immune checkpoint blockade (ipilimumab + nivolumab) was associated also with gastritis and a severe case of autoimmune pancreatitis (Diefenbach et al., 2020; Armand et al., 2021).

Anti-PD-1 and anti-PD-L1 agents seemed to be less toxic to the GI tract (Chhabra and Kennedy, 2021). Diarrhoea was also the most commonly reported GI symptom, but the incidence did not exceed 33% (Ribrag et al., 2021). Other types of manifestation included colitis, typhlitis, enteritis, proctitis, proctosigmoiditis, pancreatitis, cholecystitis and duodenitis (Khodadoust et al., 2016; Armand et al., 2019; Ravandi et al., 2019; Diefenbach et al., 2020; Maruyama et al., 2020). No grade 5 GI ICI-related toxicity has been reported so far. (Table 4)

When considering GI toxicity of ICIs, bacterial or viral infection should be excluded at first (Haanen et al., 2017; Rocha et al., 2019; Chhabra and Kennedy, 2021). GI involvement associated with the main diagnosis is not as frequently seen in hematological malignancies as in solid oncology, when metastases could cause symptoms similar to GI irAEs. Laboratory tests, abdominal X ray or CT scan and endoscopic methods could help in dubious cases. The benefit of biopsy should be carefully assessed due to the risk of perforation. GI irAEs may mimic idiopathic inflammatory bowel disease.

Grade 1 colitis is defined by < 4 liquid stools per day and the treatment is symptomatic with mainly fluid and electrolyte repletion (Haanen et al., 2017; Chhabra and Kennedy, 2021). When grade 1 colitis persists >14 days or grade 2 (4–6 liquid stools) persists >3 days, ICI treatment should be withheld and oral steroids (prednisolone 0.5–1 mg/kg or budesonide 9mg once daily) immediately initiated. If no improvement is seen after 72 h or in grade 3–4 cases (≥7 liquid stools per day), intravenous (methyl)prednisolone 1–2 mg/kg is recommended with possible escalation to infliximab 5 mg/kg after another 72 h. Other possible immunosuppressive treatment options include mycophenolate mofetil (500–1000 mg twice a day) or tacrolimus. Patients experiencing grade 3–4 GI toxicity are not usually re-challenged with ICI therapy because of the high risk of relapse, but it could be discussed on an individual basis.

Pulmonary toxicity of ICIs does not occur very often, but can be rapidly progressive and even fatal when present (Chhabra and Kennedy, 2021). Generally, it is more typical for anti-PD-(L)1 mAbs in comparison to anti-CTLA-4 inhibitors. Pneumonitis is one of the most common reasons for ICI discontinuation and in hematological malignancies, it is the leading treatment-related cause of death. Clinical symptoms include dyspnea, cough and chest pain and the median time to onset is 10–12 weeks after the initiation of ICI treatment.

The incidence of pneumotoxicity for ipilimumab monotherapy was 0–11%, and 5–12% for the combination with nivolumab (Ansell et al., 2009; Davids et al., 2016; Diefenbach et al., 2020; Armand et al., 2021). There was one grade 5 case in each group (Davids et al., 2016; Diefenbach et al., 2020). Nivolumab and pembrolizumab caused pneumonitis/interstitial lung disease/upper respiratory tract infection in 0–24% and 0–13% of patients, respectively (Ansell et al., 2015; Badros et al., 2017; Armand et al., 2019; Maruyama et al., 2020). There were 4 treatment-related deaths reported for the nivolumab cohort, while none for pembrolizumab (Lesokhin et al., 2016; Daver et al., 2019; Ramchandren et al., 2019; Diefenbach et al., 2020). No pneumotoxicity was observed in either of the anti-PD-L1 mAb clinical trials (avelumab, durvalumab + tremelimumab) (Kazandjian et al., 2021; Ribrag et al., 2021). (Table 5).

Pneumonitis is a diagnosis per exclusionem, so other conditions such as pulmonary embolism, pneumonia, viral infection or cancer progression should be excluded in the first place (Chhabra and Kennedy, 2021). The diagnostic method of choice is a high-resolution CT scan where ground glass opacities, organizing pneumonia, interstitial pneumonia or hypersensitivity pneumonitis could support the diagnosis of ICI-related lung toxicity (Haanen et al., 2017; Chhabra and Kennedy, 2021). Bronchoscopy with bronchoalveolar lavage may help to rule out infection. Lung biopsy is recommended only in dubious cases and it should be performed by video-assisted thoracoscopic surgery (Haanen et al., 2017; Delaunay et al., 2019).

Grade 1 pneumonitis is characterized by radiological findings without clinical symptoms (Haanen et al., 2017; Delaunay et al., 2019; Chhabra and Kennedy, 2021). These patients should be closely monitored and the delay of treatment considered. Oral prednisone 1 mg/kg and ICI interruption are recommended when the symptoms arise (grade 2). After recovery, steroids should be tapered over a period of 4–6 weeks due to possible relapse. Patients can be re-challenged with ICI when the dose of oral prednisone is ≤ 10 mg. Grade 3 and 4 cases of pneumonitis are associated with new or worsening hypoxia and life-threatening breathing difficulties. These patients should be hospitalized and high-dose intravenous (methyl)prednisolone (2–4 mg/kg/day) initiated. If there is no improvement after 2 days, infliximab (5 mg/kg), mycophenolate mofetil (1000 mg twice a day) or cyclophosphamide are to be added. Slow steroid tapering over at least 6 weeks is also recommended upon recovery to avoid relapse. In grade ≥3 cases when the infectious etiology can´t be quickly dismissed, broad-spectrum antibiotics are to be initiated alongside steroids. Restarting ICI therapy in patients experiencing severe pneumonitis (grade 3–4) is not recommended.

Immune checkpoint blockade completely revolutionized the landscape of solid tumor oncology treatment. However, its journey in hemato-oncology has been much slower and more difficult. The initial great success in cHL and later PMBL, which led to accelerated FDA approval of nivolumab and pembrolizumab in these indications, was overshadowed by the general hold that was placed on large clinical trials investigating the promising combination of ICIs and IMiDs in the multiple myeloma setting due to suspicious treatment-related toxicity. Moreover, dual immune checkpoint blockade did not fulfil expectations and did not show increased benefit for patients suffering from hematological malignancies, only greater toxicity was reported. Aside from cHL and PMBL, ICI monotherapy did not prove to be effective either. Nevertheless, the future of ICIs in hemato-oncology seems to lie in combinations, either with conventional chemotherapy or other targeted agents such as mAbs, CAR T cells, PI3κ inhibitors, antibody-drug conjugates or bispecific antibodies as many ongoing studies and their promising efficacy results demonstrate. Therefore, hemato-oncologists should be prepared to work with ICIs more often and learn to recognize, diagnose and manage their specific irAEs.

Generally, the incidence and distribution of irAEs is approximately similar between solid tumor oncology and hemato-oncology. However, the exact rates may differ due to smaller amounts of treated patients and the fact that most of the safety data came from phase 1 and 2 clinical trials. The only available phase 3 safety data come from the controversial KEYNOTE-183 and KEYNOTE-185 studies (Mateos M.-V. et al., 2019; Usmani et al., 2019). The first one investigated the benefit of adding pembrolizumab to pomalidomide and dexamethasone in RR MM patients while the second one was for ND MM and studied lenalidomide and dexamethasone with or without pembrolizumab. These combinations proved toxic, when both clinical trials were suspended prematurely because of higher treatment-related mortality rates in experimental arms. Even if no specific adverse event of the ICI and IMiD combination has been unveiled to date, it is worth mentioning that the only two cases of grade 5 ICI-related myocarditis reported from prospective ICI blood cancer trials came from these two (1 from each). Investigation of another combination, daratumumab (anti-CD38 mAb) with PD-1 or PD-L1 inhibitors, was abandoned based on the unsatisfying results from a phase 1b/2 LUC2001 study of atezolizumab with or without daratumumab for advanced or metastatic non-small-cell lung cancer (NCT03023423) (Cohen et al., 2021). The cohort which received both agents did not reach superior outcome and there was also an imbalance of deaths in the experimental arm even if they weren´t considered treatment-related. Although this study was targeting solid tumors, the decision of the investor affected all similar ongoing studies, including hemato-oncological (mainly MM again).

Agents acting on PD-1 axis are used in hemato-oncology more often than CTLA-4 inhibitors, so irAEs such as hyper/hypothyroidism or pneumonitis may occur more frequently. Nevertheless, blood cancer specialists should be aware of all possible types of irAEs. Mostly, it is a diagnosis per exclusionem, so the physician is required to eliminate all other feasible causes. The majority of irAEs are mild to moderate, can be managed symptomatically or by oral steroids and mainly do not lead to permanent ICI discontinuation. On the other hand, fatal and potentially lethal irAEs were also reported, especially myocarditis, pneumonitis, neurologic events, hypophysitis, acute kidney injury and Stevens-Johnson syndrome. These events usually require permanent cessation of ICI treatment, rapid administration of high dose corticosteroids or other immunosuppressants (infliximab, mycophenolate mofetil, tacrloimus etc.) and multidisciplinary approach.