**5. Immune-Related Adverse Events**

Immune-related adverse events (irAEs) represent a major cause of treatment discontinuation and a confounding factor in response assessment at imaging. However, the side effects experienced during immunotherapy seem to be better tolerated than the side effects due to cytotoxic therapy; a lower risk of serious (grade III–IV) adverse events has been reported [58].

IrAEs are related to the recruitment of an immune infiltrate in the organs involved. From this perspective, IrAEs can be seen as an "undesirable" sign of immune activation or rather that immunotherapy is acting effectively, as suggested by some studies reporting an association between irAEs and treatment efficacy [59,60].

The first sign of immune activation is the inversion of the liver-to-spleen ratio, which may be accompanied by splenomegaly. Speen activation can be associated with a mild diffuse uptake in bone marrow. In an early phase, reactive lymph nodes may be seen in the basin of the tumor. A sarcoid-like reaction, consisting of an increased uptake in mediastinal hilar lymph nodes and pulmonary granulomatosis, can also occur. These phenomena are transient, self-limiting and require monitoring until their resolution.

IrAEs can affect any organ, particularly the endocrine system's glands (hypophysitis, thyroiditis and adrenalitis). A symmetrical enlargement of the adrenal gland with diffusely increased FDG uptake can be due to adrenalitis (Figure 5). With particular reference to nivolumab-induced thyrotoxicity, in a recent metanalysis by Barroso-Sousa and colleagues [61], incidence rates of 6.5% for hypothyroidism and of 2.5% for thyrotoxicosis are reported. Nivolumab-induced thyroid dysfunction is due to painless thyroiditis, characterized in most cases by an early-onset, transient, thyrotoxic phase, commonly followed by hypothyroidism. A research study investigated the underlying mechanism of the thyroid [62], suggesting that, since normal thyroid tissue expresses PDL1 and PDL2 mRNA and proteins, PD1 pathway blockade impairs immunotolerance and can induce autoimmune thyroiditis in the absence of TRAb positivity. The appearance of a diffusely increased FDG uptake in the thyroid gland in the course of immunotherapy (Figure 6) should be considered suggestive of incipient thyroiditis, even before clinical manifestation and obtaining laboratory findings [63].

**Figure 5.** Adrenalitis: patient with advanced melanoma. Panel (**A**) shows CT (I) and PET/CT (II) during therapy with tyrosine kinase inhibitors; adrenal glands appear normal. Note the appearance of intense hypermetabolism in the adrenal glands (box (**B**), II) accompanied by symmetrical enlargement in CT images (box (**B**), I) during combined nivolumab plus ipilimumab treatment.

**Figure 6.** Nivolumab-induced thyroiditis: (**A**) no uptake in the thyroid gland at baseline. (**B**) appearance of diffuse mild uptake in the thyroid gland in a patient receiving nivolumab for Hodgkin lymphoma who developed thyroiditis.

A threatening irAE that can limit treatment cycles is checkpoint-inhibitor-related pneumonitis (CPI). An accurate diagnosis of CIP can be difficult since, during the treatment of oncological patients, other factors (infections, radiation therapy and other drugs) are often mixed. In recent years, several studies [64,65] have demonstrated the potential of CT radiomics to differentiate CIP from other conditions, such as radiation-induced pneumonitis, leading to the development of the Rad-score [66], a robust model combining 11 imaging histological features with bilateral involvement and sharp borders.

The spleen must be critically checked: spleen enlargement and the inversion of the physiological spleen-to-liver ratio can occur in the course of immunotherapy, similarly to what is observed during conventional rituximab-based regimens, and should not be mistaken for lymphoma involvement.

IrAEs can be asymptomatic; in this context, FDG-PET/CT, being sensitive to foci of active inflammation, including those due to immune activation, has the unique ability to detect irAEs before clinical manifestation.

## **6. Future Perspectives**

The selection of patients suitable to receive immunotherapy relies on several biomarkers (PD-L1 immunohistochemistry, immunohistochemistry for mismatch repair proteins, PCR-based assays for microsatellite instability and sequencing for tumor mutational burden on biopsy); however, they are not perfect and cannot accurately predict patient outcome. Furthermore, the current approach to assessing PD1 status cannot detect heterogeneity over time and across lesions due to the dedifferentiation of tumor clones, which may occur during therapy. To address these limitations, novel PET tracers, designed as antibodies or fragments of antibodies for specific immunotargets (immune-PET), were developed in the past few years. Immuno-PET offers the possibility to non-invasively image in vivo the whole body biodistribution of immune checkpoints and hold the potential to guide treatment decision making.

The first human PD-L1 PET study was conducted with 89Zr -atezolizumab in 22 patients with metastatic non-small-cell lung cancer (NSCLC), bladder cancer and triple-negative breast cancer [67]. The patients were imaged before starting immunotherapy. The study demonstrated a high tracer uptake in normal lymphoid tissue and sites of inflammation. Tumor lesions showed a generally high tumor uptake, with great intra-patient and interpatient heterogeneity. [89Zr]Zr-atezolizumab uptake in tumor lesions correlated with the response to therapy, PFS and OS and outperformed immunohistochemistry on a fresh biopsy in the prediction of clinical response.

CD8 cells play an essential role in the cytotoxic response to tumors boosted by immunotherapy. The results of a clinical study evaluating the CD8 PET tracer 89ZED88082A, a zirconium-89-labeled one-armed antibody in solid tumors, have recently been published [68]. In this study, PET imaging was performed before and during immunotherapy. The pre-treatment biodistribution of the tracer showed specific CD8 targeting, with a high tracer uptake in normal lymphoid tissue. The tracer uptake in tumor lesions was variable within and between patients. The tumor uptake was higher in patients with mismatch-repair-deficient tumors and was correlated with CD8 cell density in tumors stained immunohistochemically. A higher SUVmax in tumor lesions at baseline showed a trend with improved OS. The results of serial 89ZED88082A imaging during immunotherapy showed a great spatial and temporal heterogeneity of the behavior of lesions in responders, providing an insight into the complex dynamic tumor microenvironment. A phase II trial investigating the efficacy of atezolizumab consolidation therapy in high-risk DLBCL is ongoing (HOVON 151) (NCT03850028) (https://clinicaltrials.gov last update posted 18 June 2019, accessed on 25 April 2023). In this study, patients are evaluated with sequential 89Zr-atezolizumab imaging before and after R-CHOP induction therapy, during atezolizumab consolidation therapy and at the time of suspected relapse. The HOVON 151 trial reflects the application of immune-PET in the clinical setting and may widen therapeutic options in lymphomas.

In the future, the promising field of immuno-PET may hopefully improve patient selection for immunotherapy and response assessment, and it may guide the development of new agents.

#### **7. Conclusions**

The criteria for response assessment in lymphoma have deeply evolved in the last decade, assigning an outstanding role to FDG-PET/CT. This path starts from the first lymphomaspecific CT-based criteria and leads towards the PET-based Lugano Classification, which, nowadays, represents the gold standard. The LYRIC criteria, the recent refinement of the

Lugano Classification, were conceived to capture the new patterns of response observed during treatment with novel immunotherapy agents that have entered the clinic. In this context, PET/CT has the unique ability to assess response and uncover immune-related adverse events. PET/CT quantitative parameters, such as MTV and TLG, assessing changes in tumor burden may be useful tools in interpreting the response to immunotherapy.

**Author Contributions:** Conception and design, F.T. and A.C.; literature search, F.T., E.G. and S.C.; writing-original draft preparation, F.T., E.G. and G.G.; writing—review and editing, G.G., A.C.; visualization, S.C.; supervision, G.G. and A.C.; project administration, A.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.
