Skin manifestations including rash/pruritus and mucositis are the most frequent irAEs associated with immune checkpoint inhibitors. Approximately 47–68% of patients treated with anti-CTLA-4 antibodies [23] and 30–40% of patients treated with anti-PD-1/anti-PD-L1 antibodies [27] experience skin toxicities of any grade.

The typical rash is reticular, maculopapular, and slightly erythematosus, especially developed on the trunk and extremities, and may be pruritic (Fig. 33.2). The majority of immuno-related rashes can be treated with topical creams based on corticosteroids. Itching, if persistent, can be treated with oral antihistamines. Grade 2 rash, especially if pre-existing, requires treatment with oral steroids. Grade 3–4 requires discontinuation of treatment and therapy with intravenous (IV) corticosteroids. A consultation with a dermatologist is recommended, as is a skin biopsy if possible.

Stevens-Johnson syndrome is a fairly rare occurrence that can be fatal (Fig. 33.3). In these patients, IV treatment with corticosteroids is always recommended. If there is no improvement in symptoms, the use of immunosuppressive agents in combination with corticosteroids should be considered [28]. Vitiligo is a common manifestation and generally occurs late following treatment with immune checkpoint inhibitors. It is characterized by a perivascular lymphocytic infiltrate in the deep dermis layer in proximity to melanocytes and is considered as a marker of response to treatment [29]. Oral mucositis is more common with anti-PD1 antibodies than with ipilimumab and usually requires treatment with topical corticosteroids and lidocaine, to be used after excluding the presence of candidiasis (Fig. 33.4).

Diarrhea is one of the most common toxicities in patients receiving immunotherapy and must be diagnosed and treated in an adequate and timely manner to prevent it from developing into a serious adverse event. Generally, diarrhea appears after about 6 weeks of treatment [13] and is more often related to treatment with anti-CTLA-4 than with anti-PD-1 agents. Diarrhea has been reported in about 30% of patients with melanoma treated with ipilimumab, being grade 3–4 in less than 10% [30]. In a phase II study, the rate of severe diarrhea was higher in patients treated with ipilimumab 10 mg/kg than with 3 mg/kg (10% vs 1%) [31]. Similarly, in a phase III trial comparing ipilimumab 3 mg/kg and 10 mg/kg in patients with metastatic melanoma, GI toxicity was higher with the 10 mg/kg dose, with diarrhea reported in 79% of patients treated with 10 mg/kg versus 64% of patients in the 3 mg/kg arm. Grade 3–4 diarrhea was almost twice as common with ipilimumab 10 mg/kg compared with 3 mg/kg (34% vs 18%) [32]. Grade 3–4 diarrhea only occurs in 1–2% of patients treated with anti-PD-1 agents. GI toxicity is the main irAE with the combination of nivolumab plus ipilimumab, with an incidence of about 40%, including 9% of patients experiencing severe diarrhea and colitis [9, 10].

Enteritis, in some cases without diarrhea, can cause obstruction of the small intestine [33]. On biopsy, histology shows an edematous mucosa with rich infiltrate of neutrophils and/or lymphocytes [29, 30]. To date, there are no effective treatments to prevent the development of diarrhea and/or colitis. The use of budesonide to treat enteritis has been assessed but did not show any significant benefit [34].

About 10% of patients treated with anti-CTLA-4 experience clinically significant endocrinopathy [35], while the incidence of endocrine disorders in patients treated with nivolumab is 14%, with 2% of events grade 3–4 in severity [4]. Less than 5% of patients experienced grade 3–4 endocrine toxicity with combination treatment [32]. The main types of endocrine toxicity derive from inflammation of the thyroid, pituitary, or adrenal glands and are often difficult to identify because they typically present with non-specific symptoms, such as fatigue, nausea, headache, and visual changes.

Liver toxicity typically occurs in less than 10% of patients [29]. Among patients receiving ipilimumab 3 mg/kg, severe, life-threatening, or fatal hepatotoxicity occurred in 2%, while the incidence of all grade hepatitis was 2.3% with nivolumab monotherapy. However, occurrence was higher (13%) in patients treated with ipilimumab plus nivolumab in combination. Very often, patients are asymptomatic and the only indicator of toxicity is increased transaminases. A worsening of total bilirubin is less often found and usually occurs late. Generally, liver toxicity begins after about 8–12 weeks of treatment.

Although clinical trials have excluded patients with a history of hepatitis B and C, these patients have been treated in clinical practice. Although data are limited, the presence of previous viral hepatitis does not seem to increase the risk of hepatotoxicity [38, 39]. Before starting treatment with immune checkpoint inhibitors, it is mandatory to assess liver function and markers for hepatitis B or C virus. In all patients with HBsAg positivity, early antiviral treatment may be needed in high viral load (HBV DNA).

In patients with grade 1 toxicity (i.e., asymptomatic increases in transaminases and hyperbilirubinemia), treatment can be continued, and liver function tests should be monitored until resolution. Grade 2 events require the interruption of immunotherapy until resolution. Oral prednisone 1 mg/kg/day should be administered. In patients with grade 3–4 events, treatment should be permanently discontinued; high-dose IV glucocorticoids (methylprednisolone 2 mg/kg/day) are recommended. Mycophenolate 500 mg every 12 h should be considered if liver enzymes are still elevated after 48 h of treatment. If no improvement occurs in the following 5–7 days, tacrolimus at the dosage of 0.10–0.15 mg/kg/day is recommended. Infliximab is not recommended because of its hepatotoxicity.

Grade 3 or higher pneumonitis has been reported in 5–7% of patients with NSCLC treated with nivolumab or pembrolizumab [40]. However, the incidence of symptomatic pneumonitis is only 1% with ipilimumab [41]. About 7% of patients treated with nivolumab plus ipilimumab had pneumonitis, with only 1% of grade 3–4 severity [32]. A recent systematic review and meta-analysis suggest that the overall incidence of pneumonitis during PD-1 inhibitor monotherapy was 2.7% (95% CI: 1.9–3.6%) for all grades and 0.8% (95% CI: 0.4–1.2%) for grade 3 or higher pneumonitis. The incidence was higher in patients with NSCLC compared with melanoma for pneumonitis of all grades (4.1% vs 1.6%; p = 0.002) and grade 3 or higher (1.8% vs 0.2%; p < 0.001). The incidence of pneumonitis was more frequent during combination therapy than monotherapy for all grades (6.6% vs 1.6%; p < 0.001) and grade 3 or higher (1.5% vs 0.2%; p = 0.001) pneumonitis [42, 43]. Several factors, such as pre-existing lung damage due to tumor burden, smoking, chronic obstructive pulmonary disease, pulmonary fibrosis, and variable expression of PD-L1 on normal lung tissues, may play a role in this although the exact cause of this difference is still unknown.

Pulmonary toxicity should be considered whenever patients present with symptoms of a respiratory infection, new-onset cough, or wheezing. In symptomatic patients, a chest CT scan and a pulmonary consultation are recommended, as is initiation of oral or IV corticosteroids. In patients with moderate or severe symptoms, a diagnostic bronchoscopy that can assess whether there is widespread lymphocytic infiltration is advised. If the toxicity is grade 1 with only asymptomatic radiological signs, treatment suspension of 2–4 weeks until resolution of radiographic findings should be considered. If grade 2, suspension of treatment is mandatory and treatment with oral corticosteroids should be initiated. In severe or recurrent cases, IV treatment with methylprednisolone 2 mg/kg or equivalent may be appropriate, and immune treatment should be permanently discontinued.

Although rare, there have been documented cases of pulmonary sarcoidosis in patients treated with immune checkpoint inhibitors [44].

Renal failure has been reported in patients treated with ipilimumab, anti-PD-1s, and combination therapy. Immune checkpoint inhibitors can cause acute kidney injury that presents similar to other drug-induced tubulointerstitial nephritis. Cortazar et al. analyzed data from published phase II and III trials enrolling 3695 patients and reported an overall incidence of acute kidney failure of 2.2%, with 0.6% grade 3–4 events. The median duration until the appearance of kidney injury was 13 weeks [45]. Acute kidney failure occurred more frequently in patients treated with nivolumab plus ipilimumab (4.9%) than in patients treated with anti-PD-1 monotherapy. The most common signal of acute kidney failure was an increase in serum creatinine, which occurred in all affected patients. Pyuria (68%) and hematuria (16%) were also frequently noted in patients with renal impairment. In patients with suspected immune-related renal failure, a renal consultation should be sought early and a biopsy performed if possible. In patients with grade 2–3 renal toxicity, treatment should be suspended and steroids are recommended. Treatment can be resumed if symptoms improve to grade 1 severity. High doses of corticosteroids and treatment discontinuation are recommended for grade 4 toxicity.

Based on data in prospective studies, the overall incidence of any grade of neurotoxicity is 3.8% with anti-CTLA-4 antibodies, 6.1% with anti-PD-1 antibodies, and 12.0% with combination therapy [46]. However, most of these neurological adverse events are grades 1–2 and consist of non-specific symptoms such as headache (55%), dysgeusia (13%), or dizziness (10%). The incidence of grade 3–4 events was 1% across all treatments. On the basis of the nervous system area involved, neurotoxicity can be classified as encephalopathy, myelopathy, pure meningitis, meningoradiculitis, Guillain-Barre-like syndrome, peripheral neuropathy, or myasthenic syndrome. The spectrum of neurological symptoms appears to be highly heterogeneous, including headache, fever, tiredness or weakness, confusion, memory problems, sleepiness, hallucinations, seizures, and stiff neck. In the literature, several cases of serious toxicity have been reported, such as reversible posterior encephalitis syndrome [47], Guillain-Barre syndrome [48], myasthenia gravis, transverse myelitis [49], and demyelinating polyneuropathy [50].

All serious neurological irAEs should be treated with high-dose corticosteroids, and a neurologist should be consulted for differential diagnosis and additional therapy.

Arthralgia and arthritis are the most commonly reported rheumatic and musculoskeletal irAEs associated with immune checkpoint treatment. The incidence of arthralgia secondary to nivolumab in phase III trials ranges from 5 to 16% [51], and similar rates have been reported with ipilimumab monotherapy [52]. With combination therapy, incidence of arthralgia was about 10% [32]. Thus, although arthritis is a very common manifestation of autoimmune disease, it has not been reported as frequently as an adverse event in clinical trials of checkpoint inhibitors as might have been expected. One reason for this could be that there are several mutually exclusive ways to code musculoskeletal adverse events in the current system. For example, arthralgia, arthritis, joint effusion, and musculoskeletal pain are all potential options for coding the same event. Furthermore, its incidence could be underrepresented because most clinical trials report only grade 3–4 toxicity and arthritis is often considered to be less severe than other adverse events [53]. No observational studies that monitored patients for inflammatory arthritis with confirmation by a rheumatologist have been reported. However, in a case series of nine patients treated with immune checkpoint inhibitors and who developed inflammatory arthritis, clinical presentation was variable and involved large and small joints, with or without systemic involvement (colitis, urethritis) and autoantibody detection. Patients were treated with corticosteroids and some also required methotrexate and/or anti-TNF treatment [54]. The same series also reported on four patients who developed sicca syndrome while receiving checkpoint inhibition therapy. Symptoms of dry mouth were more severe than dry eyes in all four patients. Three had positive anti-nuclear antibodies. One patient was positive for La/SS-B antibodies and had parotitis treated with 6 weeks of prednisone that resulted in complete resolution of symptoms. All patients were negative for Ro antibodies [54].