Potentials of Non-Invasive 18F-FDG PET/CT in Immunotherapy Prediction for Non–Small Cell Lung Cancer

The immune checkpoint inhibitors (ICIs), by targeting cytotoxic-T-lymphocyte-associated protein 4, programmed cell death 1 (PD-1), or PD-ligand 1, have dramatically changed the natural history of several cancers, including non–small cell lung cancer (NSCLC). There are unusual response manifestations (such as pseudo-progression, hyper-progression, and immune-related adverse events) observed in patients with ICIs because of the unique mechanisms of these agents. These specific situations challenge response and prognostic assessment to ICIs challenging. This review demonstrates how 18F-FDG PET/CT can help identify these unusual response patterns in a non-invasive and effective way. Then, a series of semi-quantitative parameters derived from 18F-FDG PET/CT are introduced. These indexes have been recognized as the non-invasive biomarkers to predicting the efficacy of ICIs and survival of NSCLC patients according to the latest clinical studies. Moreover, the current situation regarding the functional criteria based on 18F-FDG PET/CT for immunotherapeutic response assessment is presented and analyzed. Although the criteria based on 18F-FDG PET/CT proposed some resolutions to overcome limitations of morphologic criteria in the assessment of tumor response to ICIs, further researches should be performed to validate and improve these assessing systems. Then, the last part in this review displays the present status and a perspective of novel specific PET probes targeting key molecules relevant to immunotherapy in prediction and response assessment.

ICIs reactivate the patient's immune system through T lymphocytes against tumor cells. This mechanism differs from that of cytotoxic or targeted agents, which could result in the development of inflammation at the tumor sites and subsequent anti-tumor responses (Missailidis, 2008). These responses are of unique patterns, pseudo-progression, hyper-progression (HPD), long duration of response, and disease regression, which continue after treatment discontinuation, especially pseudoprogression and HPD, which can influence the treatment evaluation in many clinical cases and predict the entirely reverse prognosis. Moreover, such an increased immune system activity can activate T immune cells in a variety of normal tissues, causing autoimmune side effects, also known as immune-related adverse events (irAEs). Therefore, the novel manifestations, previously seldom seen during conventional therapies and frequently described under immunotherapy, have challenged responsive and prognostic assessment to ICIs. Nevertheless, a substantial number of patients still do not derive benefit from checkpoint inhibitors. Clinical data showed that the responding rate of ICIs is closely correlated with biomarker expression in tumor specimens, for example, PD-L1 status (Borghaei et al., 2015;Herbst et al., 2016;Reck et al., 2016), tumor mutational burden (TMB) , microsatellite instability-high (MSI-H) (Gregg et al., 2019;Zhao et al., 2019), and mismatch repair-deficient (MMRdeficient) (Gregg et al., 2019;Zhao et al., 2019). In particular, PD-L1 expression and TMB have been recognized as the valuable predictive factors for NSCLC patients with ICIs: a recent metaanalysis demonstrated that NSCLC patients with high PD-L1 expression did benefit the most from a single-agent ICI treatment in the first-line setting (Passiglia et al., 2021), and another metaanalysis resulted in a proven benefit in overall survival (OS) in favor of ICIs in the TMB-high NSCLC population . MSI-H and MMR-deficient are proposed as the predictors for anti-PD-1/anti-PD-L1 immunotherapy efficacy in colorectal cancer . Given that MSI-H lung cancers are extremely rare, MSI testing/MMR immunohistochemistry (IHC) is not considered as routine in lung cancer closely associated with high expression of PD-1/PD-L1 and shows durable responses to PD-1 blockade (Gregg et al., 2019). However, the heterogeneity within tumors limits the effectiveness of predictive value of these biomarkers based on IHC and next-generation sequencing (NGS) of tumor specimens. Additionally, since not all advanced NSCLC patients are suitable for tumor biopsy, biopsy specimens are not always readily available for the pathological assay. Thus, noninvasive biomarkers are urgently needed for the prediction of potential responsiveness of ICIs.
2-Deoxy-2-( 18 F)-fluoro-D-glucose ( 18 F-FDG) positron emission tomography/computed tomography (PET/CT) takes advantage of non-invasively evaluating the glucose metabolic level to determine the tumor burden all over the body. Parameters derived from 18 F-FDG PET/CT, relevant to glucose metabolism on the molecular level, are demonstrated to be correlated with the treatment response and survival, which shows a potential capability of recognizing and tracing biomarkers related to responsiveness and prognosis. Therefore, 18 F-FDG PET/CT has been widely used to determine the tumor stage and drug efficacy including the responsiveness of ICIs in NSCLC patients (Na et al., 2017;Liao et al., 2020).
Meanwhile, new positron molecular probes targeting key molecules associated with immunotherapy have been developed and would be used to improve the assessing specificity and effectiveness for NSCLC patients treated with immunotherapy.
Thus, this review summarizes the potential applications of 18 F-FDG PET/CT as a non-invasive method for the prediction and tracking of responsiveness and prognosis of ICIs in NSCLC patients. A perspective of novel PET molecular probes in immunotherapy prediction and response assessment is also discussed.

Identification of Unusual Response Patterns
The mechanism of ICIs leads to unique manifestations during and after immunotherapy. Pseudo-progression and HPD belong to the unusual response patterns, which occur more frequently in ICIs than in other treatments (chemotherapy, target treatment, angiogenic inhibitors, etc.) (Iacovelli et al., 2015;Mellema et al., 2015;Barrón et al., 2018;Ferrara et al., 2018;Zhou et al., 2020).
Immune system-related pseudo-progression was first described in melanoma patients receiving ipilimumab (an anti-CTLA-4 antibody) in 2008. Pseudo-progression is described as a process with an increase in tumor size or metabolic tumor burden or the appearance of new lesions before the occurrence of a subsequent decrease or stability (Costa et al., 2021). Among patients with advanced NSCLC, the rate of pseudo-progression is 3-5% (Gettinger et al., 2015;Ferrara et al., 2018;Fujimoto et al., 2019), which is lower than that among patients with melanoma, for whom the equivalent rate was reported as up to 10% (Hodi et al., 2008). Besides the primary manifestations above (varying tumor size and metabolic tumor burden, emergence of new lesions), other rare forms of pseudo-progression include pleural or pericardial effusion (Kanazu et al., 2018;Song et al., 2019), intestinal perforation , and brain pseudoprogression (Melian et al., 2018). The certain mechanisms of pseudo-progression are unknown. There are some hypotheses trying to explain this pattern, including delayed and weaker activation of the immune system, immune cell infiltration, intracellular and vasogenic edema, and inflammation and intra-tumor hemorrhage (Fujimoto et al., 2019) (Gerwing et al., 2019). The chronic activation of immune system results in delay control of tumor growth, which can be shown as the elevation of FDG uptake in tumor sites with or without increase in tumor size and appearance of new lesions during the early period after administration of ICI agents. In contrast, immune cell infiltration and activation by checkpoint blockade antibodies influence the tumor immune microenvironment and then might promote the glycolysis of cancer cells, leading to the increase in FDG uptake at the early treatment phase (Tomita et al., 2018;Tomita et al., 2020). In terms of the diverse hypotheses for pseudo-progression, the mechanisms of glucose-metabolic elevation are completely different, but the FDG-uptake increase and the subsequent decrease or stability indeed can be observed through multiple PET/CT scans to help distinguish pseudo-progression from other responses. It is reported that pseudo-progression could occur between weeks 4 and 20 from baseline (Evangelista et al., 2019;Costa et al., 2021) and this time interval from baseline to pseudo-progression tends to be shorter than that from baseline to response. Additionally, patients with pseudo-progression would experience a markedly superior survival to those with normal response situations (Kurra et al., 2016;Solinas et al., 2017;Zhou et al., 2020). Thus, imaging reassessment is suggested after 4-8 weeks (never more than 12 weeks) (Costa et al., 2021).
HPD was depicted as a process of expansive growth or change in the rate of tumor progression that is grossly different from baseline, causing a detrimental effect on the patients. It is another atypical response pattern of ICIs and becomes a big challenge in clinical practices because of the higher incidence and worse prognosis. HPD's molecular mechanisms are still unclear. Champiat et al. concluded the following probable explanations: 1) expansion of regulatory T cells; 2) exhaustion of compensatory T cells; 3) modulation of protumorigenic immune cell subsets; 4) oncogenic pathway activation; and 5) activation of aberrant inflammation (Champiat et al., 2018). The expansive growth and proliferation of tumors inevitably induce the raising of tumoral glucose consumption leading to the rapid and sustained increase in FDG uptake. The incidence of HPD in NSCLC patients is about 5-19.2% , and HPD could be apparently associated with a poor prognosis . Multi-center retrospective research with 406 advanced NSCLC patients demonstrated patients experiencing HPD treated with ICIs had significantly lower overall survival (OS) compared with patients with normal progressive disease (median OS, 3.4 months [95% confidence interval (CI), 2.8-7.5 months] vs. 6.2 months (95% CI, 5.3-7.9 months); hazard ratio, 2.18 [95% CI, 1.29-3.69]; P 0.003). Nevertheless, it is difficult to differentiate HPD from other response patterns within 8 weeks of starting treatments and the second early imaging evaluation is proposed after this period (more than 8 weeks after immunotherapy) to properly adjust therapeutic programs (Costa et al., 2021).
In fact, there is no consensus on the definitions for pseudoprogression and HPD, and their molecular mechanisms are still unclear. The range of their incidences is quite wide, which reflects the heterogeneity of cancer subtypes and therapies (type of agents and dose).
Reactivation of the immune system through the use of ICIs has led to the development of a new series of side effects named irAEs. The precise pathophysiology underlying irAEs is not clearly understood. Due to the whole immunity enhancement, irAEs can potentially occur in any system of the whole body and can vary in involved systems, incidence, and severity of grade according to each particular agent and its dose and tumor types. irAEs commonly manifest within the first 3 months of administration for most drugs and can also occur at any time during treatment, even after discontinuation of ICI therapy (Costa et al., 2021). These effects could be mistaken as the disease progression, including HPD resulting in inappropriate strategies. Although most irAEs are mild and can be managed through transient immunosuppression with corticosteroids, high-grade events often require hospitalization and specialized treatment because some of them are life-threatening. In addition, multiple studies reported that patients who experienced irAEs demonstrate marked improvements in progression-free survival (PFS), OS, and overall response rate compared to those lacking toxicity (Das and Johnson, 2019). Therefore, early and accurate identification of irAEs is essential to aid in the diagnosis and management of patients and to reduce associated morbidity. Some effects can be apparent on 18 F-FDG PET/CT imaging. A correct interpretation of 18 F-FDG PET/CT images could be complicated for imaging manifestations of the unusual patterns and adverse reactions. However, the whole-body scan is also essential to discover these situations and know their characteristic localizations, especially when some effects are not always associated with the presence of clinical symptoms and signs. Thus, PET imaging with 18 F-FDG can ensure close clinical monitoring and medical intervention when necessary to avoid serious complications. In some situations, imaging features from 18 F-FDG PET/CT can effectively improve the diagnosis of particular effects. Taking the intestinal adverse effect for example, after carefully considering and ruling out other possibilities such as inflammatory bowel disease or continuing metformin before undergoing PET/CT, diffuse or segmental FDG uptake in bowel with or without thickening wall could be recognized as the toxicity of ICIs ( Figure 1). Additionally, multiple scans (before, during, and after treatment) based on assessment criteria (see 2.1.3) can further contribute to distinguishing among these situations. Table 1 displays the common irAEs, including the features in PET/CT to help to recognize these effects.
The identification of pseudo-progression, HPD, and irAEs challenges the immunotherapy in NSCLC. According to the demonstrations above, during the early period of immunotherapy, the alterations of FDG uptake based on pseudo-progression and HPD may be similar. However, their mechanisms of the FDG uptake are entirely different. Additionally, some irAEs could further confound the diagnosis. Taking the sarcoidosis or sarcoid-like reaction, for example, these FDG avid enlarged lymph nodes may be misdiagnosed as lymph node metastases. Thus, the finds in 18 F-FDG PET/CT imaging need to be compared with those of the subsequent PET/CT scans to confirm the kind of response (in 2.1.3, the assessment criteria based on 18 F-FDG PET/CT displayed how to use PET/CT scans to confirm the response status). The quantitative parameters based on 18 F-FDG PET/CT could help in differentiation (the role of parameters from 18 F-FDG PET/CT in differentiation is shown in 2.1.2). As mentioned above, the recommended imaging assessment time differs among pseudo-progression, HPD, and irAEs. There is no uniform recommendation optimum time yet when those responses or events could be differentiated well. Thus, the complexity of these responses and events leads to the discernment process requiring a comprehensive analysis combined with multiple information such as clinical symptoms, signs, laboratory examinations, and imaging manifestations, which can be well implemented by the model of the multidisciplinary team (MDT) including nuclear medicine physicians.

Parameters in Prognostic Evaluation for Immunotherapy
Generally, there are three imaging assessing approaches based on 18 F-FDG PET/CT: visual interpretation and estimation of relative uptake, assessment of uptake over a defined time using semi-
Frontiers in Genetics | www.frontiersin.org February 2022 | Volume 12 | Article 810011 5 quantitative methods, and assessment of uptake from the time of injection to a defined endpoint using kinetic analysis (Shankar et al., 2006). Visual assessment is of subjectivity, and the results could vary among readers. Additionally, FDG accumulation is influenced by uptake time, blood glucose concentration, and partial-volume effects (Shankar et al., 2006). The visual comparison among different scans is unsuitable and less rigorous for evaluation. Full kinetic quantitative analysis can provide an absolute rate for FDG metabolism, which is independent of imaging time and contributes to the investigation of various components of glucose metabolism, such as transport and phosphorylation. Nevertheless, because of the complexity of such an approach, including patient compliance issues and the requirement for arterial blood sampling or dynamic imaging of a blood-pool structure to obtain a precise input function, the kinetic analysis has been used infrequently for treatment evaluation of malignancy (Weber et al., 1999).
The semi-quantitative method is a relatively objective approach to evaluate the metabolic tumoral status. Semiquantitative parameters-standardized uptake value (SUV)have been used to determine tracer uptake in attenuationcorrected PET images in 18 F-FDG PET/CT, mainly representing the extent of glucose metabolism. However, it was reported that SUV based on 18 F-FDG PET/CT is associated with molecules relevant to some signaling pathways, for instance, the mTOR pathway. Previous studies have shown SUV of FDG is significantly associated with PD-L1 expression (Takada et al., 2017), and the activation of the protein kinase B (AKT)-mTOR pathway increases PD-L1 protein expression in NSCLC (Kaira et al., 2014). Then SUV from 18 F-FDG PET/CT could be relevant to the prognosis of immunotherapy in NSCLC.
The maximum SUV(SUV max ) is widely used in clinical practices because of its high repeatability. However, SUV max , only representing the highest metabolic point in the volume of interest (VOI) determination, ignores the intra-tumoral and inter-tumoral heterogeneity.
Other semi-quantitative parameters such as mean SUV (SUV mean ), peak SUV(SUV peak ), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) can reflect more information on biological behaviors such as heterogeneity of tumors, therapeutic responsiveness, and prognosis. SUV mean was defined as the average SUV related to the tumor burden. SUV peak was the highest average SUV calculated within a spherical VOI (usually 1-cm-diameter region). MTV represents the metabolically active portion of the tumor, obtained through several SUV-based segmentation techniques. TLG is the product of SUV mean x MTV. MTV and TLG, which consider volumetric information and reflect the whole-body tumor burden, have been more frequently proposed as prognostic factors for immunotherapy in NSCLC patients (Seban et al., 2020a;Seban et al., 2020b;Castello et al., 2020;Seban et al., 2020c;Polverari et al., 2020;Monaco et al., 2021). The latest relevant studies on 18 F-FDG PET-derived parameters for predicting the response and prognosis of ICIs in NSCLC patients in the recent 3 years are displayed in Table 2. These research works demonstrated the value of baseline semi-quantitative parameters of 18 F-FDG PET/CT in the responsiveness and prognosis of NSCLC patients treated with ICIs. However, under the semi-quantitative method, FDG accumulation is still affected by many technical, physiological, and pathological factors. Thus, the imaging-based comparative response evaluations during treatment for immune-treatment have been adopted.
It should be noted that SUL peak is consistently employed as the only semi-quantitative parameter in functional criteria (see 2.1.3). SUV can be normalized to body mass, lean body mass, or body surface area. SUV normalized to lean body mass is named SUL. SUV normalized to body mass is the commonest parameter in clinical practices but cannot consider the relatively lower 18 F-FDG accumulation in fatty tissues (Wahl et al., 1992) and are more dependent on body habitus across populations than SUL and SUV corrected for body surface area (Evangelista et al., 2020). Moreover, normalization to body surface area or lean body mass potentially reduces the effect of weight loss (which may occur during therapy) on subsequent SUV determinations. In addition, lean body mass may be the better method because of the availability of sex-specific corrections (Sugawara et al., 1999). Peak standardized uptake values normalized by lean body mass (SUL peak ) were the highest average SUL calculated within a spherical VOI in the site of the most metabolically active tumor manifestation and of high repeatability. Therefore,  1. Sum of the maximum diameter of lesions increased by >20% over the smallest achieved sum of maximum diameter. The appearance of one or more new lesions is always considered progression Or 2. Then, 18 F-FDG PET (SCAN 2) assessment is required after 3-4 weeks: i) when percent change in SULpeak per PERCIST criteria is more than 15.5%, the SMD will be confirmed 2. SD/SMD; then, 18 F-FDG PET (SCAN 2) assessment is required after 3-4 weeks: ii) when the percent change in SULpeak is less than or equal to 15.5%, this pattern will be confirmed as PMD. Complete resolution of FDG uptake within all lesions to a level of less than or equal to that of the mean liver activity and that is indistinguishable from the background (bloodpool uptake) Reduction of at least 30% in the sum of SUL peak of all target lesions detected at baseline and an absolute drop of 0.8 SUL peak units Neither CMR, PMR, nor PMD.
1. Increase in at least 30% in the sum of SUL peak of all target lesions detected at baseline, and an absolute increase in 0.8 SUL peak units with exclusion of infection/treatment effect Target lesion: up to five measurable target lesions, typically the five hottest lesions among all lesions, including new lesions, and no more than two per organ Patients with appearance of new FDG-avid lesions must be confirmed 4-8 weeks later PMR or SMD with appearance of new FDGavid lesions at 4-8 weeks must be reevaluated PMR or SMD with appearance of new FDGavid lesions at 4-8 weeks must be reevaluated  A significant correlation between the grade of uptake of the traces and the response assessed after 3 months of nivolumab was observed Li et al., 2020 Zr-N-sucDfpembrolizumab Preclinical experiment (healthy cynomolgus monkeys) Preferential uptake in the lymphoid tissues, including the lymph nodes, spleen, and tonsils, was shown England et al., 2018(England et al., 2018 Zr-DF nivolumab Preclinical experiment There was highly specific binding of 89 Zr-DF nivolumab to activated T-cell infiltrating tumors in humanized murine models Cole et al., 2017(Cole et al., 2017  PET/CT images clearly showed that 89 Zr-Df-F (ab')2 possessed superior pharmacokinetics and imaging contrast over the radiolabeled full antibody, with much earlier and higher tumor uptake (5.5 times more at 2 h after injection) and much lower liver background (51% reduction at 2 h after injection) Truillet et al., 2018(Truillet et al., 2018  SUL peak has been used as an effective index for evaluating malignancies treatment, including immunotherapy.

Metabolic-Based Response Assessment Criteria
Imaging-based response assessment is essential for the clinical management of immunotherapy. The role of the assessment system is as follows: 1) identifying non-responders; 2) confirming the response patterns; and 3) detecting irAEs.
Morphologic criteria are widely adopted in clinical trials and practices. These anatomic/structural assessing systems determine therapeutic effectiveness on the ground of the change of targetlesion size and the presence of new lesions principally. However, these anatomic response criteria cannot timely discern response without lesion volumetric changes such as cavitation, cystic change, intra-tumoral hemorrhage, and reduction in vascularity, which might occur after the use of immunotherapeutic agents and are considered physiopathologic features of a specific response to them (Costa et al., 2021). For example, as quite extensively used anatomic response criteria, Response Evaluation Criteria in Solid Tumors version 1.0 or version 1.1 (RECIST 1.0 or RECIST 1.1) define tumor responses into a series of four bins of response [complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD)]. Nevertheless, some tumor sizes shrink slowly, but patients live for a long period with SD evaluated by RECIST 1.0/1.1, so this stable-disease response could have highly beneficial outcomes (Benjamin et al., 2007;Llovet et al., 2008;Van den Abbeele, 2008). Thus, with such reductionism through RECIST, the potentially valuable information that may be important is lost (Wahl et al., 2009). An additional consideration for RECIST is that the response evaluation from the baseline and follow-up studies is based on a precise estimate, which requires repeatability and accuracy of the assessment. However, more misclassifications and variance in response are seen when a different reader assesses the baseline and follow-up studies (Erasmus et al., 2003;Wahl et al., 2009).
FDG uptake of malignant tissues depends on the tumor's functional status, such as rate of glycolysis, expression of glucose transport, level of proliferation, and association with particular molecules. Thus, 18 F-FDG is a marker of tumoral activity. In terms of the functional volume, metabolic extent, and wholebody burden of tumor, functional criteria with 18 F-FDG PET/CT to assess immunotherapy responses have been adapted for lung cancer and melanoma patients. There are at least four FDG PET/ CT criteria ( Table 4): 1) Early Prediction of Response to Immune Checkpoint Inhibitor therapy (PECRIT) (Cho et al., 2017): PECRIT is the combining functional and anatomic assessing system based on the RECIST 1.1 and PET Response Criteria in Solid Tumors (PERCIST), which requires two PET/CT scans (first scan: before therapy; second scan: between days 21 and 28 on therapy). According to PECRIT, initial SD, defined by RECIST 1.1, should be assessed by 18 F-FDG PET/CT (SCAN 2) after 3-4 weeks. On SCAN 2, when the percent change in SUL peak per PERCIST criteria is more than 15.5%, the stable metabolic disease (SMD) will be confirmed, or when the percent change in SUL peak is less than or equal to 15.5%, this pattern will be confirmed as progressive metabolic disease (PMD). As shown in Table 3, PECRIT admits the initial increase in FDG uptake during the early immunotherapeutic period and could effectively distinguish pseudo-progression and real progression by the particular threshold value of SUL peak percent change (15.5%). Moreover, this evaluation criteria combined with morphologic and functional criteria to effectively reduce assessment cost. Cu-NOTAipilimumab-F (ab')2 Preclinical experiment (mice) PET imaging with both 64 Cu-NOTAipilimumab and 64 Cu-NOTA-ipilimumab-F (ab')2 was able to localize CTLA-4+ tissues Ehlerding et al., 2017(Ehlerding et al., 2017 64 Cu-DOTA-ipilimumab Preclinical experiment (mouse models of NSCLC) 64 Cu-DOTA-ipilimumab can correctly localize the tumor, but a link was found with the receptor on the cell surface rather than in the intracellular domain Antiinterferon-γ Activated lymphocytes inside tumor lesions Gibson et al., 2018(Gibson et al., 2018 Zr-anti-IFN-γ Preclinical experiment (mouse with mammary tumors) The activation status of cytotoxic T cells is annotated by 89 Zr-anti-IFN-γ PET, providing valuable non-invasive insight into the function of immune cells in situ Protease granzyme B Cytotoxic CD8 + T cells and natural killer cells Larimer et al., 2017(Larimer et al., 2017 68 Ga-NOTA-GZP Preclinical experiment (human melanoma specimens) Granzyme B PET imaging can serve as a quantitatively useful predictive biomarker for efficacious responses to cancer immunotherapy Interleukin-2 Tumor/tissue infiltrating T lymphocytes Markovic et al., 2018(Markovic et al., 2018  However, measurement of all tumor lesions, as suggested by this PECRIT, is not only time-consuming but also sometimes impossible, especially in patients with advanced disease from an aggressive malignancy, which may be associated with a high number of metastatic lesions and/or diffuse organ infiltration. In addition, although data from PECRIT original research demonstrated that sensitivity, specificity, and accuracy of this algorithm to predict response at 4 months after initiation of therapy were 100, 93.3, and 95.0%, respectively, PECRIT is just established through the study based on 20 advanced melanoma patients. Therefore the threshold value of SUL peak could be further validated or statistically calculated based on large samples with multiple cancer types. 2) PET Response Evaluation Criteria for Immunotherapy (PERCIMT) (Anwar et al., 2018): PERCIMT, which demands two (pretherapeutic and posttherapeutic) PET/CT scans, suggests the number of newly emerged FDG-avid lesions on posttherapy PET/CT as cut-off value to identify the treatment failure in patients under immunotherapy which is quite different from the other criteria. Compared to other criteria, PERCIMT simplifies the assessment protocol by focusing on the limited number of new lesions. In the original study of PERCIMT, an absolute number of four new 18 F-FDG-avid lesions, irrespective of their size, gave a reliable indication of nonresponse to treatment, and when the cut-off of newly emerged FDG-avid lesions with functional size < 1.0 cm was four, this assessing system showed the high sensitivity (84%) and specificity (100%). However, this research was also limited to the relatively small number of patients (N 41), single cancer type (metastatic melanoma), and monotherapy (ipilimumab).
3) Immune PERCIST (iPERCIST) (Goldfarb et al., 2019): iPERCIST, which was adapted from PERCIST and the immune RECIST (iRECIST), demands two or three PET/CT scans (pretreatment, 2 months later after immunotherapy and 4-8 weeks after the second scan when necessary). iPERCIST's dual-time-point evaluation is designed to differentiate the unconfirmed progressive metabolic disease (UPMD) from the confirmed progressive metabolic disease (CPMD). Then, the UPMD (PMD at SCAN-2) needs to be reevaluated 4-8 weeks later at SCAN-3. The continuation of treatment after the first evaluation by SCAN-2 was according to the physician's judgment, metabolic response and the deterioration of clinical status, as described in iRECIST. If this UPMD (PMD at SCAN-2) is still considered as UPMD at SCAN-3, the UPMD at SCAN-3 is finally confirmed as CPMD. Patients with complete metabolic response (CMR), partial metabolic response (PMR), stable metabolic disease (SMD), or UPMD followed by PMR or SMD were classified as responders. Patients with UPMD FIGURE 1 | A 58-year-old male patient with metastatic intrahepatic cholangiocarcinoma was treated with the third line of ipilimumab. After four cycles of treatment, compared with 18 F-FDG PET/CT maximum intensity projection (MIP) image (A) before treatment, the posttreatment 18 F-FDG PET/CT (B-D), (I,J) demonstrates the following: 1) multiple soft tissue density nodules in the peritoneum, which were larger in volume and higher in FDG uptake than before ((A,B) hollow arrows); 2) multiple nodules without FDG uptake in both lungs, some of which are larger in volume than before and some of which have no change compared with before (cannot be shown in MIP images-(A,B)); 3) the lymph nodes with increased FDG uptake in mediastinal areas and bilateral hilar, which were smaller in volume and FDG-uptake lower than before ((A,B) hollow triangles); and 4) multiple foci of increased FDG uptake in bones: the FDG-uptake degree in vertebral body of lumbar 4 was significantly lower than before ((A) gray triangle); the right acetabular lesion was the new lesion ((B) gray triangle); the remaining ostial lesions were higher than those before ((A,B) black line arrows). According to PERCIMT criteron: SMD was confirmed for the number of the newly FDG-positive lesions is less than 4. Meanwhile, in the light of imPERCIST5, PMD was also evaluated for the sum of SUL peak of the patient's top five target lesions after treatment was more than 30% higher than that of the top five target lesions before treatment. As a result of clinical follow-up, the patient was confirmed PD and those lesion above were validated as metastases. PD was determined for the appearance of new lesion based on PECRIT. PMD was determined for new FDG-avid lesion at SCAN-2, which can be considered as UPMD in the line with iPERCIST. But the patient didn't receive the next 18 F-FDG PET/CT between days 21 and 28 after treatment or 4-8 weeks later, so it could not be evaluated according to PECRIT and iPERCIST. Follow-up results showed the PFS for ICI was 6 months. Additionally, this posttreatment PET/CT displayed: 1) newly patchy ground glass density foci in both lungs, with increased FDG uptake (C-PET axial image, D-CT axial image); 2) diffuse increased FDG uptake in ascending colon, transverse colon, descending colon, sigmoid colon and rectum (i-PET axial image) which is new compared with the previous one, and the corresponding colonic walls were not significantly thickened ((J)-CT axial image). The patchy ground glass shadows of both lungs gradually disappeared during chest CT follow-up ((E-H)-CT axial images). Combined with clinical information, the patchy ground glass shadows of both lungs were diagnosed as immunerelated pneumonia. Under the administration of MDT, the diffuse increase of glucose metabolism in the colon was considered as immune-related colitis.
Frontiers in Genetics | www.frontiersin.org February 2022 | Volume 12 | Article 810011 associated with clinical deterioration or UPMD followed by CPMD were classified as non-responders. iPERCIST more effectively avoids the interference from pseudo-progression and HPD as the third scan could be administrated. The original clinical trial of iPERCIST showed the reliable ability to identification for responsiveness: responders continued treatment for a mean of 10.7 months (range 3.8-26.3), OS was longer for responders than that for non-responders (19.9 vs. 3.6 months, log-rank P 0.0003), and the 1-year survival rates were 94% for responders and 11% for non-responders (Goldfarb et al., 2019). Nevertheless, this original trial was also developed through small sample research (N 28), single cancer type (NSCLC), and monotherapy (nivolumab, a PD-1 blocker), and the prognostic value of iPERCIST should be confirmed in large prospective multicentric studies. 4) Immunotherapy-modified PERCIST5 (imPERCIST5) (Ito et al., 2019): imPERCIST5 requires several PET/CT scans, and especially when PMR or SMD is with newly emerged lesions, several scans should be administrated to reevaluate. The key difference of imPERCIST5 lies in the interpretation of new lesions on the posttreatment scan. In imPERCIST5, tumor response was assessed by the change in the sum of SUL peak of up to five lesions. New lesions do not necessarily result in a scan to be classified as PMD, for the appearance of new lesions that resolved spontaneously and were probably inflammatory in nature. Thus, imPERCIST5 could efficiently discern real progression from pseudo-progression and irAEs. According to the original clinical analysis, imPERCIST remained prognostic (hazard ratio, 3.853; 95% confidence interval, 1.498-9.911; P 0.005) (Ito et al., 2019). Of note, in the original study of imPERCIST5, whereas imPERCIST5 reduces overdiagnosis of progressive disease, new lesions in patients with PMR or SMD by imPERICST5 were eventually found to have metastases in 55% of the cases. Thus, if the patients have the decrease or no change in metabolism of existing target lesions as well as the appearance of new lesions, the prognosis appears indeterminate. Then biopsy should be considered before any change in treatment. imPERCIST5 was still produced based on a clinical study with only 60 metastatic melanoma patients treated with ipilimumab. A case study with assessment on PERCIMT and imPERCIST5 is provided in Figure 1.
These assessment criteria above discovered the limitations of current morphologic criteria, such as response without lesion volumetric changes and individual measurement differences among readers. Then, they all highlight the value of 18 F-FDG PET/CT (e.g., the high repeatability and accuracy of measurement for semi-quantitative parameters among readers) in responsiveness evaluation and prognosis in malignancies with immunotherapy, and three of the four assessment systems suggested more than one posttreatment PET/CT scan. Furthermore, they proposed their own resolutions to overcome the limitations of anatomic criteria: PECRIT and iPERCIST focused on the overall metabolic level in the whole body; PERCIMT emphasized the number of the newly FDG-avid lesions; and imPERCIST5 stressed the limited target lesions' FDG-uptake values. However, it must be admitted that these criteria are in diversity and just verified in small samples at present. Only one of them was developed from a study based on NSCLC patients. Thus, large-sample multi-center validation and further improvement are necessary in the future. Moreover, given the diversity of the metabolic level among NSCLC subtypes (glucose metabolism level of lung adenocarcinoma is lower than that of other NSCLC pathological subtypes (Lopci et al., 2016)), the dedicated evaluation criteria oriented to NSCLC or NSCLC subtypes should be developed. In addition, the increase in radiation dose and high cost of PET/CT scan also become the reasons restricting the application of assessment criteria based on 18 F-FDG PET/CT. Above all, the lack of specificity of FDG is the main limitation for 18 F-FDG PET/CT. It is reported that an increased consumption of glucose may be due either to lymphocytes' activation, which reflects the adaptive host response to tumor and is associated with a more favorable outcome, or to neutrophils' activation, promoting an inflammatory cascade leading to tumor progression and dismal prognosis (De Larco et al., 2004). Therefore, the posttherapeutic assessment is needed more than once to confirm the right status.

Specific PET Probes for Immunotherapy
New molecular probes arise that directly target the key molecules of immune checkpoint pathways and immune responses. They can not only image the PD-1/PDL-1/CTLA-4 status or lymphocyte infiltrations of tumor tissue and better understand the uptake and distribution of agents/molecules and their mechanisms, but also be useful to monitor patients initially presenting sensitivity to ICIs and then showing an acquired resistance after a variable period of time.
The novel PET probes and their research progression are presented in Table 4. Fundamental research demonstrated the high target specificity and fulfilling affinity and adequate tumor penetration of these probes. The tracers under clinical settings have authenticated the safety of humans. During the imaging experiments, the uptake of tracers has shown sufficient resolution, which can assess potential heterogeneity within each lesion and among lesions from the same patients. Furthermore, in some studies concerning the assessment of response to immunotherapy, significant correlations were displayed between the grade of uptake of tracers and the response. Beyond the scope of PET imaging, promising molecular structures can also be targeted using singlephoton emission tomography ligands (Table 4) and provide another potential and feasible choice for response evaluation and prediction of prognosis in malignancies with immunotherapy. However, there are still limitations such as high hepatic and splenic uptake that affect the detection of abdominal lesions, long-circulating half-life, and radiation exposure for patients. In addition, most of them are still under the preclinical research phase, and only preliminary human studies have been carried out for some probes. Therefore, further research, improvement, and validation of specific probes to ICIs will continue in the future.
for NSCLC patients with ICIs, especially adopting semiquantitative parameters and functional assessing criteria. Novel PET probes targeting key molecules relevant to ICIs and treatment response provide the promising perspective to more specifically screen oncological population that will benefit from immunotherapy and continuously assess ICI response in the future.

AUTHOR CONTRIBUTIONS
XL was responsible for outline design and writing articles. ML assisted in the writing of this review. RW designed the outline. JZ was responsible for the selection of the theme and revising the draft.