Applications and research progress of PET/CT in dynamic monitoring of immunotherapy in melanoma

Abstract

Background

Melanoma is a highly aggressive malignancy, and conventional therapies have limited efficacy in metastatic cases. The advent of immune checkpoint inhibitors (ICIs) has significantly improved patient prognosis. However, challenges such as heterogeneous responses, resistance, and immune-related adverse events (irAEs) necessitate reliable tools for early efficacy prediction and dynamic assessment.

Main body

As a molecular imaging modality that integrates structural and functional information, Positron Emission Tomography/Computed Tomography (PET/CT)—particularly through ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) metabolic imaging—can identify atypical response patterns during immunotherapy, including pseudoprogression, hyperprogression, and immune-detached responses (iDRs). It also enables accurate evaluation of therapeutic efficacy using criteria such as PERCIST and PERCIMT. Moreover, PET/CT shows unique value in the early detection of irAEs, with metabolic alterations (e.g., thyroiditis or colitis) preceding clinical symptoms. In recent years, novel probes targeting immune components such as PD-1/PD-L1 and CD8⁺ T cells (Immuno-PET) have further enhanced the capacity for real-time monitoring of the tumor immune microenvironment.

Conclusions

PET/CT represents a valuable imaging modality for comprehensive assessment of immunotherapy in melanoma. With continued refinement of response criteria and the development of immune-targeted tracers, PET-based imaging is expected to further facilitate personalized and dynamic immunotherapy strategies.

Introduction

Malignant melanoma is a highly aggressive tumor originating from melanocytes in the skin or mucous membranes, characterized by early metastasis and strong invasiveness. It accounts for approximately 90% of all skin cancer-related deaths [1]. In recent years, the global incidence of melanoma has continued to rise, with an estimated 332,000 new cases and 58,000 deaths reported in 2022 [2]. For patients with early-stage melanoma, surgical resection combined with radiotherapy can lead to a favorable prognosis. However, treatment of metastatic melanoma remains a major clinical challenge. Conventional chemotherapy agents exhibit limited specificity, transient efficacy, and significant toxicity, resulting in a 5-year survival rate that has long remained around 10% [3].

This therapeutic dilemma has been markedly improved by advances in targeted and immunotherapeutic agents. In particular, with in-depth research on tumor immune mechanisms and the clinical application of immune checkpoint inhibitors (ICIs) have fundamentally transformed the treatment landscape of melanoma [4, 5]. Currently, ICIs approved for melanoma mainly target two critical immune checkpoints: PD-1 and CTLA-4. These agents exert their therapeutic effects through targeted monoclonal antibodies (including nivolumab, pembrolizumab, and ipilimumab) that specifically inhibit tumor immune evasion mechanisms, consequently restoring T cell-mediated antitumor immunity. Multiple phase III clinical trials have demonstrated that ICIs, either as monotherapy or in combination regimens, can significantly improve survival outcomes in patients with advanced melanoma. Median progression-free survival (PFS) and overall survival (OS) are increased by approximately two to three times compared to conventional treatments, establishing ICIs as the standard first-line therapy for unresectable or metastatic melanoma [6–11].

However, not all patients benefit from immunotherapy [12, 13]. In clinical practice, about 50% of melanoma patients show either no response to ICIs or develop resistance after initial treatment success [14, 15], significantly limiting treatment options. Additionally, immune-related adverse events (irAEs)—such as immune pneumonitis, hepatitis, and colitis—require careful attention [16–18]. Meanwhile, the high cost of immunotherapy may restrict accessibility for some patients. Therefore, identifying responsive patients at an early stage, predicting therapeutic efficacy, and dynamically monitoring immune responses have become focal points of current research.

As a molecular imaging modality that integrates both morphological and functional information, positron emission tomography/computed tomography (PET/CT) has attracted increasing attention in the management of melanoma. By using radiotracers such as ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) to assess glucose metabolism in tumor cells, PET/CT not only enables precise anatomical localization of primary and metastatic lesions, but also detects early metabolic changes following immunotherapy at the imaging level. This facilitates evaluation of therapeutic response, supports clinical decision-making, and may assist in identifying potential responders. Thus, PET/CT has emerged as a valuable tool in the context of melanoma immunotherapy. This review systematically summarizes the current applications and recent advances of PET/CT in the assessment of immunotherapeutic efficacy in melanoma (Fig. 1).

Fig. 1.

PET/CT in Melanoma Immunotherapy: A Flowchart of This Review. The framework is categorized into four modules: Module I: Immunotherapy Response Evaluation: Focusing on the utility of ¹⁸F-FDG PET/CT in assessing treatment response, with a comparison of specialized metabolic criteria including PERCIST, PERCIMT, and PECRIT. Module II: Immune-Related Adverse Events (irAEs) Monitoring: Highlighting the role of ¹⁸F-FDG PET/CT in the early detection and characterization of multi-organ toxicities, such as immune-related thyroiditis and colitis. Module III: Immunotherapy Efficacy Prediction: Summarizing the prognostic value of various PET-derived parameters and recent research progress in identifying biomarkers for long-term clinical outcomes. Module IV: Development and Application of Novel Immuno-PET Probes: Exploring the transition from non-specific glucose metabolism imaging to target-specific molecular profiling using next-generation immune tracers

Application of ¹⁸F-FDG

PET enables comprehensive evaluation of immunotherapy response patterns

Immunotherapy response can be evaluated by monitoring tumor metabolic changes on ¹⁸F-FDG PET scans. In patients responding to treatment, PET typically shows significant decreases in FDG uptake, reductions in standardized uptake value (SUV), or even complete resolution of metabolic activity to background levels, indicating a complete metabolic response. In contrast, patients with treatment failure or disease progression are generally characterized by persistent or increased FDG uptake, or the emergence of new hypermetabolic lesions. While metabolic resolution may not represent a definitive cure, it has been validated across some studies as an imaging phenotype predictive of favorable clinical outcomes [19, 20].

Due to the unique mechanism of immunotherapy—activating the immune system to combat tumors—its exhibits distinct response patterns that differ from conventional treatments. The following three atypical response patterns are particularly noteworthy:

Pseudoprogression

Characterized by initial lesion enlargement or new lesion emergence early during treatment, followed by stabilization or regression on subsequent imaging [21]. This phenomenon is more frequently observed with anti-CTLA-4 therapy and less commonly with anti-PD-1/PD-L1 treatment, with an overall incidence below 10% across cancer types [22, 23].

Hyperprogression

Refers to a rapid increase in tumor burden over a short period. Although no universally accepted definition exists, it is often characterized by a marked increase in tumor growth rate (TGR) or tumor volume. For instance, Champiat et al. defined it as a more than twofold increase in TGR during treatment [24], while Kato et al. incorporated criteria such as short time to treatment failure (TTF < 2 months) and accelerated tumor growth [25]. Other evaluation methods include the TGKR index [23, 26, 27] and criteria proposed by Matos et al. based on tumor volume and new lesions in multiple organs [28, 29]. Reported incidence rates of hyperprogression range from 4% to 29%, likely due to variability in study populations [23, 30], and it is generally associated with poor prognosis [31].

Immune-detached response (iDR)

Refers to a mixed response in which some lesions shrink while others, often in different organs, enlarge [21]. The incidence of iDR is approximately 10% [32]. As some patients with iDR may still benefit from immunotherapy, certain studies propose that iDR could serve as a surrogate marker for treatment efficacy and favorable prognosis [32, 33].

These atypical response patterns underscore that radiographic progression does not necessarily indicate treatment failure. Unlike CT that focuses on anatomical changes, PET-CT provides comprehensive metabolic assessment, potentially avoiding unnecessary treatment changes. For example, a case report described a stage IV melanoma patient receiving nivolumab/ipilimumab combination therapy. Although CT scans showed primary tumor growth and new lesions (meeting progression criteria), PET-CT confirmed pseudoprogression by revealing complete metabolic resolution in all lesions. The patient subsequently continued with nivolumab monotherapy and achieved sustained disease control [34]. While current international guidelines (e.g., ESMO and NCCN) primarily endorse FDG PET/CT for initial staging and clarifying indeterminate findings [35, 36], the unique clinical challenges of immunotherapy—such as atypical response patterns—suggest a broader utility. Therefore, we propose that in patients with multiple lesions, PET/CT serves as a valuable adjunct for follow-up monitoring to support more optimized and personalized treatment decisions.

Evolution of response evaluation criteria based on new immunotherapy response patterns

The appearance of atypical response patterns to immunotherapy has revealed limitations in conventional tumor evaluation criteria, driving continuous updates to assessment systems. Currently, two main categories of response criteria exist for immunotherapy evaluation: morphological criteria and metabolic criteria.

In terms of morphological criteria, since the release of RECIST 1.1 in 2009, a series of modified versions have been proposed, including irRC, irRECIST, iRECIST, imRECIST, and itRECIST [37]. The RECIST criteria assess tumor burden by measuring the longest diameter of target lesions, offering simplicity and good reproducibility. However, RECIST has notable limitations: for tumor types such as lymphomas or sarcomas, lesion size may not change significantly even with effective treatment, leading to potential inaccuracies in evaluation.

Regarding metabolic criteria, substantial advancements have been achieved since the initial EORTC guidelines were established in 1999. Subsequent refined protocols including PERCIST, PERCIMT, imPERCIST, and iPERCIST have progressively evolved (Table 1). The original EORTC framework relied principally on body weight-normalized standardized uptake values (SUVbw), a metric potentially influenced by fluctuations in patient body composition and temporal variations in scan acquisition protocols. Significant improvements were introduced in PERCIST 1.0 [38]: it adopts standardized uptake value normalized to lean body mass (SUL) instead of SUVbw to minimize interference from body fat, and introduces the SULpeak concept—defined as the average SUL within a 1.2 cm-diameter spherical volume centered on the most metabolically active point—which demonstrates better reproducibility and stability compared to traditional SUVmax. To further accommodate the unique response patterns observed with immunotherapy, such as pseudoprogression and immune-related inflammatory changes, the PERCIMT criteria revised the definition of progressive metabolic disease (PMD). Rather than automatically interpreting new lesions as disease progression, PERCIMT considers factors such as the number, size, and total metabolic burden of lesions, thus reducing the likelihood of misclassifying pseudoprogression [39]. Building upon this framework, imPERCIST and iPERCIST provided further optimization for the evaluation of new lesions in the context of immunotherapy [40, 41]. Moreover, the PECRIT criteria introduced an innovative approach by integrating both morphological and metabolic parameters, offering a novel and potentially more comprehensive method for assessing clinical benefit from immune-based therapies [19].

Table 1.

Efficacy evaluation criteria based on metabolic response

Metabolism-Based Criteria	Complete Metabolic Response (CMR)	Partial Metabolic Response (PMR)	Stable Metabolic Disease (SMD)		Progressive Metabolic Disease (PMD)	New Lesions
EORTC (1999) [42]	Metabolic activity of all lesions reduced to background levels	After one chemotherapy cycle, tumor SUV decreases by at least 15–25%; after multiple cycles, > 25%	SUV increase < 25% or decrease < 15%	Tumor FDG uptake increases > 25% or maximum tumor FDG increases > 20%, or new metastatic lesions appear		Directly classified as PMD
PERCIST (2009) [38]	Metabolic activity of all lesions reduced to background levels	SULpeak of hottest lesion decreases > 30% and absolute decrease > 0.8	Neither PMD nor PMR/CMR	SULpeak of hottest lesion increases ≥ 30% and absolute increase > 0.8		Directly classified as PMD
PERCIMT (2018) [39]	Metabolic activity of all lesions reduced to background levels	Partial disappearance of lesions with no new lesions	Neither PMD nor PMR/CMR	≥ 4 new lesions (diameter < 1 cm) or ≥ 3 new lesions (diameter > 1 cm) or ≥ 2 new lesions (diameter > 1.5 cm)		Classified as PMD based on quantity and diameter (see left column)
imPERCIST(2019) [40]	Same as PERCIST	Total SULpeak of selected lesions (up to 5 most metabolically active lesions) decreases ≥ 30%	Neither PMD nor PMR/CMR	SULpeak of selected lesions increases ≥ 30%		Not directly classified as PMD; new lesions may be included in total SULpeak calculation. PMD requires total SULpeak increase ≥ 30%.
iPERCIST(2019) [43]	Same as PERCIST	Same as PERCIST	Neither PMD nor PMR/CMR	SULpeak of hottest lesion increases ≥ 30%. Initial detection of new lesions is UPMD; confirmed as CPMD if lesions persist or worsen on follow-up.		Requires repeat PET after 4–8 weeks (if lesions persist or metabolic activity increases, classified as CPMD; if resolved or decreased, classified as PMR or SMD).
Morphological and Metabolic Combined Criteria	Clinical benefit			No clinical benefit
PECRIT(2017) [19]	RECIST 1.1: Complete Response (CR) (all target lesions disappear; target lymph nodes < 1 cm; no new lesions)	RECIST 1.1: Partial Response (PR) (sum of target lesion diameters decreases > 30%)	RECIST 1.1: Stable Disease (SD) but SUV peak change > 15.5%	RECIST 1.1: Stable Disease (SD) but SUV peak change < 15.5%		RECIST 1.1: Progressive Disease (PD) (sum of target lesion diameters increases > 20% and ≥ 5 mm, or new lesions appear).

Note: UPMD means unconfirmed progressive metabolic disease. CPMD means confirmed progressive metabolic disease

The prognostic value of current response criteria is still under debate. Some studies show that PERCIMT performs well in metastatic melanoma, with its disease control rate significantly linked to overall survival (P = 0.003), while other criteria (e.g., imPERCIST5) did not show the same significance (P = 0.12) [43, 44]. However, other research suggests that both imPERCIST5 and PERCIST5 effectively differentiate responders from non-responders in survival outcomes [45]. These inconsistencies may stem from differences in study populations, follow-up periods, and other variables.

Overall, all current response criteria revisions follow a core principle—avoiding early discontinuation of effective therapies due to unusual response patterns [37]. For instance, PERCIMT allows up to four new lesions, enabling continued treatment for patients who may still benefit [46]. However, it is important to note that these metabolic criteria are currently utilized primarily as exploratory tools in academic research settings rather than as primary clinical standards. To transition toward ‘best practice,’ large-scale prospective validation remains essential to confirm their clinical utility and harmonize evaluation protocols. In clinical practice, the choice of criteria should align with study goals and be combined with imaging and lab data for accurate interpretation.

PET monitoring of immune-related adverse events

While ICIs have markedly improved melanoma outcomes, they are linked to irAEs. Clinical data shows that although the three main ICIs—ipilimumab, pembrolizumab, and nivolumab—have generally manageable safety profiles, the incidence of adverse events remains relatively high, with most being grade 1–2 events (Table 2). For example, the KEYNOTE-002 study reported that the most common irAEs associated with pembrolizumab were fatigue (25.8%), pruritus (22.1%), rash (15.2%), and gastrointestinal symptoms [8]. Data from Chinese populations indicate that thyroid dysfunction (26.2%) and liver function abnormalities (23.3%) are among the most frequently observed adverse events with pembrolizumab treatment [9].

Table 2.

Incidence of Adverse Events Associated with Immune Checkpoint Inhibitors in Several Clinical Studies

Immune Checkpoint Inhibitor	Clinical Trial	Overall Treatment-Related Adverse Events (%)	Grade 3–4 Immune-Related Adverse Events (%)
Ipilimumab	NCT00094653 [6]	58.9%(301/511)	11.4%(58/511)
Pembrolizumab	keynote-001 [7]	86%(562/655)	17%(114/655)
	keynote-002 [8]	70.6%(252/357)	12.3%(44/357)
	keynote-151 [9]	85.4%(99/103)	12.6%(13/103)
Nivolumab	CheckMate 037 [10]	67.5%(181/268)	8.2%(22/268)
Ipilimumab plus Nivolumab	CheckMate 067 [11]	95.5%(299/313)	55%(172/313)

As a functional imaging modality, ¹⁸F-FDG PET/CT can monitor irAEs by detecting characteristic metabolic changes across multiple organ systems. In clinical practice, this modality can identify endocrine manifestations such as thyroiditis (diffuse thyroid uptake) and hypophysitis (pituitary uptake), gastrointestinal involvement including colitis (segmental or diffuse colonic uptake) and hepatitis (diffuse hepatic uptake), as well as respiratory complications like pneumonitis (focal or diffuse pulmonary uptake). The technique also detects musculoskeletal and lymph node abnormalities, evidenced by symmetrical joint/muscle uptake in arthritis and symmetrical lymph node uptake in sarcoid-like reactions, providing comprehensive evaluation of irAEs [47].

Studies have shown organ-specific variations in the incidence of irAEs detected by PET/CT: thyroiditis (15%–22%) and colitis (10%–22%) occur most frequently, while severe events such as myocarditis are relatively rare [48–50]. Although PET/CT demonstrates limited sensitivity for dermatitis and nephritis, it offers crucial advantages in early detection and quantitative evaluation. Notably, irAEs exhibit organ-specific timing: thyroiditis typically emerges early (median onset: 8 weeks), whereas colitis and pneumonitis show wider variability (2–16 weeks) [47]. Advanced analytical approaches, including CNN-based automated segmentation and SUVX% quantification, enable PET/CT to identify irAEs significantly before clinical manifestation—providing lead times of 81 days for colitis, 48 days for pneumonitis, and 27 days for thyroiditis [47]. However, most available evidence comes from non-Chinese studies, with limited large-scale PET/CT data on irAEs in Chinese melanoma populations. This distinction is critical because Chinese patients predominantly present with acral and mucosal subtypes, which exhibit distinct biological behaviors and higher risks of visceral metastasis compared to the cutaneous subtypes prevalent in Western cohorts [51, 52]. Furthermore, Chinese populations have shown a higher incidence of specific irAEs, such as immune-related hepatotoxicity during immunotherapy [9]. Consequently, there is an augmented clinical demand for PET/CT in China to provide high-sensitivity systemic monitoring and early toxicity detection. Moreover, PET/CT has several limitations: its sensitivity is insufficient for detecting irAEs such as dermatitis and myocarditis; medications like metformin can induce physiological intestinal uptake, leading to false positives; and sarcoid-like reactions (SLRs) can exhibit imaging features that overlap with metastases, complicating differential diagnosis [47, 48, 53, 54]. Furthermore, although some studies suggest a potential positive correlation between irAEs and treatment response, this association remains to be validated by further research [50, 55]. Additionally, the non-specificity of ¹⁸F-FDG makes it susceptible to confounding by concomitant inflammatory processes. Pre-existing conditions such as chronic obstructive pulmonary disease or concurrent pneumonia can manifest as intense tracer uptake, which may be indistinguishable from immune-related pneumonitis or tumor progression.

Based on current evidence, the following strategies are recommended in clinical practice: perform PET/CT monitoring every 12–16 weeks, with particular attention to high-risk organs such as the thyroid, lungs, and gastrointestinal tract; and emphasize a comprehensive assessment that integrates clinical symptoms, laboratory tests, and, when necessary, histopathological confirmation.

PET in predicting the efficacy of immunotherapy

Multiple studies have demonstrated that PET-derived parameters, such as metabolic tumor volume (MTV) and total lesion glycolysis (TLG), are significantly associated with patient survival outcomes [56–60]. The temporal value of these assessments is significant, as different imaging windows provide distinct clinical insights. In a study involving 85 melanoma patients treated with ICIs, the MTV measured three months after treatment initiation (MTVpost) was identified as the most powerful prognostic indicator: patients with high MTVpost had a median overall survival (OS) of only 16 months, whereas those with low MTVpost had a median OS exceeding 60 months (P = 0.0003) [56]. This late-on-treatment assessment reflects the durable therapeutic effect after the initial inflammatory phase. Other studies have shown that early metabolic response (MR) and metabolic flare (MF), observed as early as one week after treatment initiation, can also predict long-term outcomes. Patients with positive MR or MF had a median progression-free survival (PFS) of over 38 months, significantly longer than the 2.8 months observed in the metabolically stable group [60]. These findings suggest that PET-based metabolic features may serve as sensitive early predictors of immunotherapy efficacy.

However, variability exists across studies, primarily due to several key factors. Firstly, treatment regimen differences play a crucial role, as varying mechanisms of action between immunotherapies (e.g., anti-PD-1 monotherapy compared with anti-PD-1/CTLA-4 combination therapy) may lead to distinct metabolic response patterns [57, 61]. Secondly, patient population heterogeneity plays a crucial role, with PET parameters demonstrating differential predictive value across melanoma subtypes - while SUVmax shows significant association with overall survival in mucosal melanoma, parameters like total MTV (TMTV) and bone marrow-to-liver ratio (BLR) prove more predictive in cutaneous melanoma [62]. Methodological variations further contribute to this heterogeneity, including differences in PET timing (baseline, 1 week, or 3 months post-treatment), assessment metrics (localized parameters such as SUV/SUL versus whole-body metrics like wMTV/TLG), and follow-up durations (ranging from 6 months to several years) [60]. The choice of timing is a critical trade-off: while ultra-early imaging (1–2 weeks) allows for early intervention, it carries a higher risk of being confounded by transient inflammatory flares. In contrast, later scans (3 months) are more reliable for confirming survival benefits but may delay necessary treatment switches. Additionally, the use of different evaluation criteria (RECIST, EORTC, or PERCIMT) complicates cross-study comparisons, with some evidence suggesting PERCIMT’s superior performance in early response prediction (P = 0.045) [63], while other studies highlight the independent prognostic value of baseline MTV [58]. These multifaceted factors collectively account for the observed discrepancies in study outcomes. The absence of standardized imaging protocols continues to hinder cross-study comparability. This underscores an urgent need for the international harmonization of PET scanning timeframes and evaluation metrics, alongside large-scale prospective studies to validate existing retrospective findings. Such efforts are essential for establishing “best practices” and integrating PET-derived biomarkers into routine clinical monitoring.

Regarding model optimization, multiparametric approaches have shown clear advantages. Compared to single-lesion indicators such as SUVmax or SUVmean, whole-body tumor burden parameters like TMTV offer a more comprehensive reflection of the tumor microenvironment and thus provide superior prognostic value [58, 62]. Furthermore, the metabolic activity of immune-related organs such as the spleen and bone marrow—quantified via the spleen-to-liver ratio (SLR) and BLR—has been linked to immunotherapy response and survival outcomes. For instance, a high baseline BLR has been associated with poor prognosis, and transcriptomic analyses in these patients have revealed enrichment of regulatory T-cell markers [64]. However, the prognostic value of dynamic changes in spleen metabolism remains controversial. One study reported that increased splenic FDG uptake two weeks post-treatment correlated with favorable response [65], while another found that an SLR increase of more than 25% at three months was associated with poor outcomes [66]. These conflicting findings highlight the need for further research to clarify the clinical relevance of metabolic changes in immune organs. To improve predictive accuracy, recent studies have increasingly incorporated multi-omics approaches. For example, combining clinical indicators such as LDH and CRP with transcriptomic data has been shown to enhance predictive performance [58, 66]. In the future, the integration of artificial intelligence for whole-body metabolic profiling may facilitate personalized treatment strategies.

Notably, the predictive value of PET has also been extended to the field of cell-based therapies. A study on tumor-infiltrating lymphocyte (TIL) therapy demonstrated that early CMR is a strong predictor of long-term survival. Among patients who achieved partial response (PR) according to RECIST, those with concurrent CMR had significantly better outcomes: median OS was not reached in PR/CMR patients, compared to only 28.7 months in PR/non-CMR patients (P = 0.0016) [67]. This finding underscores the unique advantage of PET in evaluating the efficacy of emerging cellular immunotherapies.

Development and application of novel immuno-PET probes

Conventional imaging modalities, such as ¹⁸F-FDG PET/CT, have limitations in evaluating responses to immunotherapy, as they fail to accurately capture the actual and dynamic changes within the tumor immune microenvironment (TME). With the deepening understanding of immune regulatory mechanisms and advances in radiolabeling technologies, molecular probes targeting specific immune components—referred to as immuno-PET probes—have emerged as a research focus, offering new strategies for precision monitoring in cancer immunotherapy [68].

The core value of immuno-PET lies in its ability to provide noninvasive, dynamic, and comprehensive monitoring of key immunotherapy processes. Currently, immuno-PET is mainly applied in the following areas: it enables pretreatment efficacy prediction through imaging of CD8⁺ T cells or PD-L1 expression to identify patients most likely to benefit from immunotherapy [69–71]; facilitates early treatment assessment (typically within 1–2 weeks) by distinguishing pseudoprogression from true progression, thus preventing unnecessary treatment interruption [72, 73]; and evaluates the immunomodulatory effects of combination therapies (e.g., radiotherapy with ICIs) by tracking T cell infiltration and verifying abscopal effects. Furthermore, specialized probes like ⁶⁸Ga-FAPI can identify tumor-associated fibroblast infiltration or suppressive T cell populations [74], offering insights into immunosuppressive tumor microenvironments and potential targets to overcome therapeutic resistance.

The table below summarizes several key immune targets and representative immuno-PET probes (Table 3). It is important to note that the development of current immuno-PET agents is based on various tumor models, among which melanoma is one of the most widely used for validation. Given the heterogeneity in immune checkpoint expression profiles across different tumor types, it is essential to verify target expression in specific models—using techniques such as flow cytometry, Western blotting, or immunohistochemistry—prior to probe selection and application.

Table 3.

Key targets and example probes

Mechanism	Target	Representative Probes	Clinical Value
Immune Checkpoint	PD-1/PD-L1	89Zr-atezolizumab [71], 18F-BMS-986,192 [75]	Assess PD-1/PD-L1 expression heterogeneity; optimize ICI selection
	PD-L1 plus CTLA-4 (dual)	131I-KN046 [76]	Theranostic probe for dual-target imaging and therapy
T Cells & Function	CD8⁺ T cells	89Zr-Df-IAB22M2C [77], 68Ga -NODAGA-SNA006 [78]	Differentiate immune-inflamed, excluded, and desert tumors; predict ICI efficacy
	Granzyme B	68Ga-granzyme B [72, 79, 80]	Evaluate T-cell cytotoxic activity; early identification of pseudoprogression
	T-cell activation	18F-Ara-G [81] (nucleoside tracer)	Monitor T-cell activation; predict treatment response
Immune Microenvironment	Fibroblast Activation Protein (FAP)	68Ga-FAPI [82]	Assess fibrotic/immunosuppressive TME; predict ICI resistance
	Hypoxia-related targets	18F-FMISO [83]	Identify immunosuppressive hypoxic regions; guide combination therapies

Despite its promising prospects, the clinical application of immuno-PET probes remains largely in the experimental or early clinical phases (Phase I/II) and faces centralized challenges. Technically, achieving high specificity is hindered by inaccessible intracellular targets (e.g., FOXP3) and overlapping surface markers across immune subsets [68]. From a translational perspective, divergent pharmacokinetics—ranging from full antibodies to small molecules—complicate the standardization of imaging timeframes. Furthermore, the definitive correlation between these experimental imaging findings and long-term clinical endpoints requires more robust validation.

Future developments are shifting toward multi-target synergistic imaging (e.g., dual-labeled CD8/PD-L1 probes) and theranostic applications. For instance, a dual-target PD-L1/CTLA-4 probe labeled with ¹³¹I has shown potential in both accurate imaging and enhancing immunotherapy efficacy by promoting T-cell activation [76]. Additionally, integrating AI-driven radiomics with multi-omic profiles (pathological and transcriptomic) could unlock deeper biomarkers within the dynamic tumor microenvironment.

Ultimately, the maturity of this field depends on transitioning these agents into clinical ‘best practices’ through large-scale Phase III trials and standardized protocols. This will support the emergence of an ‘imaging-guided dynamic immunotherapy’ model, enabling real-time adaptation of therapeutic strategies for truly personalized melanoma care.

Conclusion and future perspectives

The widespread application of ICIs in melanoma treatment has elevated the importance of PET/CT as a dual-modality imaging tool for comprehensive evaluation of both therapeutic efficacy and irAEs. Notably, ¹⁸F-FDG PET/CT is currently the most validated modality for capturing atypical response patterns, such as pseudoprogression, by monitoring systemic metabolic shifts that precede anatomical changes. Beyond conventional assessment, parameters like MTV and TLG provide a more robust reflection of total tumor burden and patient prognosis than single-lesion metrics.

Despite its promise, the standardization of PET in the context of immunotherapy remains a challenge. While ¹⁸F-FDG PET remains the cornerstone for metabolic monitoring, it often struggles to differentiate between inflammation and true progression. This limitation has paved the ‘path forward’ for novel immuno-PET probes. By enabling the target-specific visualization of immune components (e.g., PD-1/PD-L1, CD8⁺ T cells), these experimental tracers offer higher biological specificity, though they require further large-scale clinical validation.

Furthermore, the potential cost-effectiveness of PET/CT monitoring is worth noting. Although it increases initial imaging expenses, identifying non-responders early may help avoid the costs associated with continuing ineffective therapies. Additionally, early detection of irAEs could reduce the financial burden of managing severe complications, potentially leading to a more favorable cost-benefit profile in long-term patient management.

Looking ahead, efforts should focus on the standardization of PET protocols, integration with biological biomarkers, and the acceleration of clinical translation of immuno-PET. These advances will pave the way for truly individualized immunotherapy guided by molecular imaging.

Author contributions

Jianying Gu: Conceptualization, Project administration, Funding acquisition; Dengfeng Cheng: Conceptualization, Project administration, Resources; Chuanyuan Wei: Conceptualization, Project administration, Funding acquisition; Lilu Xie: Writing - Original Draft; Dai Shi: Writing - Original Draft; Jiayi Huang: Writing - Original Draft; Wujian Mao: Writing - Review & Editing; Shaoluan Zheng: Writing - Review & Editing, Supervision.

Funding

This study was funded by National Natural Science Foundation of China (82203528, 81972559, 82272891), China Postdoctoral Science Foundation (2022M710769, 2022TQ0072).

Data availability

All data and materials referenced in this review are obtained from publicly available sources, as cited in the manuscript.

Declarations

Ethics approval and consent to participate

This article is a literature review and does not involve human or animal subjects; therefore, ethics approval and participant consent are not required.

Consent for publication

Not applicable.

Competing interests

All authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.