WHAT IS ALREADY KNOWN ON THIS TOPIC

Lymphocyte activation gene 3 (LAG-3) is a novel target for immune checkpoint blocking. However, regarding the correlation between LAG-3 expression within tumors and patient prognosis, some studies have reached conflicting conclusions. Currently, there is a lack of tools to dynamically monitor LAG-3 expression levels to further explore the significance of LAG-3 in tumor therapies.

WHAT THIS STUDY ADDS

In this study, we construct a novel peptide-based positron emission tomography imaging agent, 68Ga-NOTA-XH05, which can assess LAG-3 expression repeatedly and non-invasively. Additionally, we demonstrate that non-invasive imaging of LAG-3 on activated immune effector cells holds the potential as a suitable biomarker for early response in cancer immunotherapy.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

68Ga-NOTA-XH05 is expected to be a successful and non-invasive method to monitor LAG-3 during oncoimmunotherapy. With more confirmation, 68Ga-NOTA-XH05 can offer crucial insights into the role of LAG-3 expression during immunotherapy and may assist in clinical decision-making.

Introduction

Blockade of inhibitory immune checkpoints represents a pivotal approach in the field of immunotherapy. Immune checkpoint inhibitors (ICIs) that specifically target cytotoxic T-lymphocyte-associated protein-4 (CTLA-4), programmed cell-death protein-1 (PD-1), and its ligand PD-L1 have been successfully employed in clinical settings.1 Despite demonstrating efficacy in certain patients, the durable response rates achieved through CTLA-4 and PD-1/PD-L1 inhibition remain limited, thereby driving research efforts toward novel ICIs.2 Lymphocyte activation gene-3 (LAG-3) has emerged as a promising therapeutic target, with relatlimab, an LAG-3 antibody, having gained approval for the treatment of unresectable or metastatic melanoma when used in combination with nivolumab.3

LAG-3, a member of the immunoglobulin superfamily, is expressed on activated CD4+ and CD8+ T cells, regulatory T cells, B cells, natural killer cells, dendritic cells, and macrophages.4 5 It exhibits higher affinity toward major histocompatibility complex II (MHC II) compared with CD46 and has been reported to potentially interact with galectin-3, LSECtin, α-synuclein, and FGL1 as additional ligands.4 By interacting with MHC II molecules, LAG-3 downregulates T cell proliferation and cytokine secretion while playing a crucial role in facilitating tumor immune evasion.7

LAG-3 is commonly expressed on exhausted CD4+ and CD8+ T cells8 9, and its presence in tumors has been associated with unfavorable clinical outcomes in various human tumor types.10 Paradoxically, several clinical reports have demonstrated that patients exhibit a favorable prognosis when LAG-3 expression is detected on tumor-infiltrating lymphocytes,11 suggesting that LAG-3 can serve as a marker for T cell activation, reflecting the infiltration of tumors and initial suppression of tumor growth by T cells. Given the conflicting observations regarding the prognostic value of LAG-3 expression in patients with cancer, further investigation into the significance of LAG-3 expression in tumors is warranted.

In clinical studies, the assessment of immune checkpoint (IC) expression is commonly performed using immunohistochemistry (IHC) on tumor biopsies.12 13 However, IC expression measured on tumor biopsies only represents a limited fraction of the overall tumor and may not accurately reflect spatial heterogeneity.14 Moreover, due to dynamic changes in ICs during disease progression, repeated invasive biopsies are impractical and unsafe for assessing temporal heterogeneity.15 Furthermore, the lack of standardization across different studies regarding antibodies used, positive thresholds applied, and scoring systems employed undermines the reliability of IHC as a method.16 These limitations underscore the pressing need for new approaches to evaluate IC expression in tumors.

Positron emission tomography (PET) enables non-invasive and quantitative whole-body imaging of specific molecules using various imaging agents. ICs-targeted PET imaging can serve as an adjunct diagnostic method for comprehensive and dynamic assessment of specific IC expression in tumors and the entire body, facilitating clinical decision-making regarding the use of ICIs.17 For instance, PET tracers for PD-L1 imaging have demonstrated superior correlation with treatment outcomes compared with IHC in clinical trials.18 Currently, several monoclonal antibody-based PET tracers targeting LAG-3 have been developed and used in preclinical or clinical studies to evaluate LAG-3 expression in mice or patients.19 20

Compared with large molecules like antibodies, small molecule drugs such as peptides exhibit favorable pharmacokinetics and tissue distribution patterns. They possess the ability to rapidly penetrate target tissues while being efficiently cleared from non-target tissues and blood circulation. These characteristics facilitate same-day patient examinations and reduce radiation exposure. Additionally, polypeptides are generally less toxic and immunogenic, offering great flexibility for chemical modification and radiolabeling.21 Zhai et al reported a cyclic peptide Cyclo (CVPMTYRAC) with a high affinity for LAG-3, demonstrating a dissociation constant of 0.66±0.35 μM.22 In this study, we developed a novel tracer named 68Ga-NOTA-XH05 by labeling Cyclo (CVPMTYRAC) with gallium-68. Our findings indicate that this tracer can effectively reflect the expression level of LAG-3 in B16-F10 murine melanoma. Furthermore, we demonstrated that 68Ga-NOTA-XH05 can evaluate the response to Toll-like receptor 9 (TLR9) agonist therapy as well as its abscopal effects in melanoma treatment outcomes. In the future, 68Ga-NOTA-XH05 may offer valuable support in identifying suitable candidates for anti-LAG-3 therapy and assessing the effectiveness of immunotherapy.

Materials and methods

Additional materials and methods sections, including in vitro cell binding and blocking essay, immunofluorescence analysis, and flow cytometry analysis, are provided in online supplemental materials 1.

Synthesis and evaluation of NOTA-XH05

Cyclic peptide cyclo (CVPMTYRAC) (denoted as XH05) was synthesized using standard Fmoc chemistry.23 The purified XH05 was coupled with the chelating agent p-SCN-Bn-NOTA to obtain NOTA-XH05. Briefly, cyclo (CVPMTYRAC) and p-SCN-Bn-NOTA were mixed and dissolved in DMF at a molar ratio of 1:1, then nine times triethylamine was added, and the reaction was performed for 4 hours at room temperature. The product was diluted with equal volume of water, purified by high-performance liquid chromatography (HPLC) and lyophilized to obtain NOTA-Bn-Cyclo (CVPMTYRAC), refered to as NOTA-XH05. The Biacore surface plasmon resonance (SPR) interaction analysis was performed to define the binding kinetics between NOTA-XH05 and recombinant mouse CD223/LAG3 protein (YMD30401, AntibodySystem) or recombinant mouse CD223/LAG3 (EHD30401, AntibodySystem) immobilized as a target antigen. The binding affinity was indicated by the KD representing the equilibrium dissociation degree (M), which was calculated from the ratio of the dissociation rate kd (s−1) to the association rate constant ka (M−1 s−1).

Radiolabeling and identification of 68Ga-NOTA-XH05

300 µL sodium acetate (0.25 mol/L) and 1 µL NOTA-XH05 (5 nmol/µL) were added to 1 mL 68GaCl3 (111–148 MBq), and the mixture was heated to 95℃ and reacted for 10 min. The product was purified by Sep-Pak C18 Light Cartridge to obtain 68Ga-NOTA-XH05. The radiochemical purity of 68Ga-NOTA-XH05 was measured through HPLC. Stability for 68Ga-NOTA-XH05 was conducted after incubation of the tracer with phosphate buffered saline (PBS) or serum for 4 hours or 2 hours.

Establishment of syngeneic melanoma model

All animal studies were conducted under the guidelines approved by the Institutional Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology. Female C57BL/6 and BALB/c-nu mice (6–8 weeks) were obtained from Beijing HFK Bioscience. The C57BL/6 mice were numbered and divided into experimental group and control group by random number method. B16-F10 murine melanoma cells were purchased from a cell bank (Shanghai, China) and cultured in RPMI 1640. B16-F10 cells (1×106) were prepared in a 1:1 (v:v) ratio in Matrigel (Corning) and injected subcutaneously into the right shoulder or bilateral shoulders of the mice to establish unilateral and bilateral tumor models.

Treatment study

TLR9 agonist, CpG-ODN 1826, was purchased from Sangon Biotech, and reconstituted in PBS. When the major axis of tumors reached 5–6 mm, C57BL/6 mice in the experimental group and BALB/c-nu mice received intratumoral injections of CpG-ODN 1826 (50 µg) on the right side every other day for a total of three times. In the control group, the same volume of PBS was injected intratumorally at the same time. The body weight of the mice and the tumor volume were measured every 2 days from the first treatment. The tumor volume was calculated as 0.5×major diameter×minor diameterˆ2. The mice were euthanized when the tumor volume was ≥1500 mm3 (for bilateral subcutaneous tumor model mice, the tumor volume of either side was ≥1500 mm3), when they lost more than 20% of their weight before tumor implantation, or when they had difficulties in movement, eating, or drinking.

68Ga-NOTA-XH05 PET imaging and biodistribution study

Mice were injected intravenously with 68Ga-NOTA-XH05 (2.96–4.44 MBq) 1 hour before PET imaging. Images were obtained on a small animal PET/CT (NovelMedical, Beijing, China) after 15 min scan (5 min CT scan+10 min PET scan). After decay and scatter correction, the images were reconstructed by three-dimensional ordered subsets expectation maximization (OSEM3D) algorithm. NMSoft-AIWS V.1.7 (NovelMedical, Beijing, China) software was used to draw the region of interest (ROI) of the tumor to obtain the uptake value. An ROI was also drawn around the descending aorta to obtain the blood uptake value to calculate the tumor to blood ratio (TBR). At the end of the PET/CT scan, the mice were euthanized, and the organs of interest were harvested and weighed. The radioactive counts were measured with a γ counter, and the tracer accumulation of each tissue or organ was calculated after decay correction and noted by the percentage of injected dose per gram of tissue (%ID/g).

Statistic analysis

Statistical analyses were performed using GraghPad Prism V.8 software. Quantitative data were expressed as the mean±SD with all error bars denoting the SD. Unpaired Student’s t-test was used for comparisons between treated and untreated samples. Pearson correlation teat was performed to correlate LAG-3 expression determined by flow cytometry to tumor uptake of the tracer. One-way ANOVA was used to compare the uptake of tracer by resting T cells, activated T cells, and blocked activated T cells, as well as to compare TBR of treatment group, control group, and T cell depletion group or responders, non-responders and control mice. Two-way ANOVA was adopted in comparisons of tumor volume and TBR calculated in bilateral tumor model. P values <0.05 were considered statistically significant.

ResultsIn vitro characterization of 68Ga-NOTA-XH05

NOTA-XH05 was characterized using mass spectrometry, with a calculated m/z value of 746.1 for [M+2 hour]2+ ion of C63H94N16O18S4; the experimental value was found to be 746.19 (online supplemental figure S1). The binding affinity between NOTA-XH05 and recombinant mouse LAG-3 protein was determined through SPR assay (figure 1B), and the binding rate constant (ka) was found to be 230.7 M−1s−1, the dissociation rate constant (kd) was 2.39×10−3 s−1, and the calculated binding affinity constant (KD) was 10.36 µM. Although the affinity of NOTA-XH05 decreased compared with that of the original peptide, it still exhibited strong binding ability toward mLAG-3 protein. Furthermore, the binding affinity constant between NOTA-XH05 and recombinant human LAG-3 protein was determined to be 12.41 µM (online supplemental figure S2).

Figure 1

In vitro characterization of 68Ga-NOTA-XH05. (A) Synthetic scheme of 68Ga-NOTA-XH05. (B) The binding affinity of NOTA-XH05 to recombinant mouse LAG-3 protein was tested by SPR. (C) HPLC chromatogram of 68Ga-NOTA-XH05 before (blue) and after incubation in PBS (red) and serum (green) for 4 hours and 2 hours, respectively. (D) Cell uptake and blocking essay showed the binding specificity of 68Ga-NOTA-XH05 to LAG-3 (n=3). **p<0.01, ***p<0.001. PBS, phosphate buffered saline.

By conducting HPLC analysis, the radiochemical purity of 68Ga-NOTA-XH05 after purification was determined to be greater than 99%. The retention time of the tracer was approximately 13.45 min (figure 1C). Following incubation in PBS and serum for 4 hours and 2 hours, respectively, the proportion of intact tracer exceeded 99%, indicating that 68Ga-NOTA-XH05 was stable and suitable for subsequent experiments (figure 1C). The hydrophilic nature of the tracer was confirmed by a log p value of −1.59±0.02. In vitro evaluation using cell uptake and blocking assays demonstrated stronger binding affinity of 68Ga-NOTA-XH05 toward activated T cells compared with resting T cells (0.41%±0.05% vs 0.24%±0.05%, p=0.0034). However, pretreatment with LAG-3 blockers significantly reduced the binding affinity of 68Ga-NOTA-XH05 toward activated T cells (0.09%±0.01%, p=0·0001), thus confirming the specificity of this tracer (figure 1D).

68Ga-NOTA-XH05 uptake correlates with LAG-3 expression

The in vivo imaging ability of 68Ga-NOTA-XH05 to visualize LAG-3 was evaluated in B16-F10 murine melanoma-bearing mice. Once the major axis of tumors reached 7–8 mm, mice were intravenously injected with 68Ga-NOTA-XH05, and small animal PET imaging was conducted at 0.5, 1, 1.5, and 2 hours postinjection (figure 2A). At 0.5 hours, clear visualization of B16-F10 tumors was achieved with a peak uptake of 0.90%ID/g±0.06%ID/g; subsequently, tumor uptake gradually decreased (figure 2B). The T/B, tumor to muscle ratio and tumor to liver ratio reached the highest at 1 hour, which were 1.54±0.31, 8.22±1.60 and 1.18±0.23, respectively (figure 2C). Therefore, 1-hour time point was selected for subsequent experiments’ imaging purposes. Moreover, 68Ga-NOTA-XH05 predominantly accumulated in the kidney, bladder, gallbladder, and partial intestinal tract (figure 2A). Furthermore, following PET imaging at 1 hour, organs of interest were harvested for biodistribution evaluation after sacrificing the mice (figure 2D). Consistent with PET findings, renal uptake exhibited the highest value (5.87%ID/g±1.83%ID/g), indicating renal excretion of the tracer. The small intestine and colon exhibited low levels of uptake (0.49%ID/g±0.08%ID/g and 0.39%ID/g±0.03%ID/g, respectively), along with the significant gallbladder uptake observed in PET imaging, suggesting that 68Ga-NOTA-XH05 may also undergo hepatobiliary excretion, thus being excreted with intestinal content.

Figure 2

68Ga-NOTA-XH05 clearly visualizes B16-F10 tumors and its uptake correlates with LAG-3 expression. (A) Representative PET/CT images of B16-F10 bearing mice injected with 68Ga-NOTA-XH05 after 0.5, 1, 1.5, 2 hours. Tumors are circled and marked with a ‘T’, kidneys are marked with a ‘K’, and gallbladders are indicated with a white arrow and marked with a ‘GB’. (B) Uptake of 68Ga-NOTA-XH05 in tumor (blue), blood (red), muscle (green) and liver (purple) at different time points after injection (n=4). (C) Tumor to blood ratio (T/B, red), tumor to muscle ratio (T/M, green) and tumor to liver ratio (T/L, purple) at different time points after injection (n=4). (D) Biodistribution analysis of 68Ga-NOTA-XH05 accumulation in selected organs. Mice were sacrificed 1 h post-injection (n=4). (E) Correlation plot of tumor uptake of 68Ga-NOTA-XH05 vs LAG-3 expression (MFI) on tumors from mice undergoing PET imaging, followed by subsequent ex vivo analysis by flow cytometry (n=10). PET, positron emission tomography.

The tumor uptake demonstrated a positive correlation with the flow cytometry-quantified expression of LAG-3 (R2=0.52, p=0.0184) (figure 2E), indicating that PET imaging using 68Ga-NOTA-XH05 can provide important information for analyzing the expression of tumor LAG-3 and its dynamic changes. Notably, there was a positive association between tumor uptake and the expression of LAG-3 on CD3+ T cells (R2=0.45, p=0.0351) (online supplemental figure S4A), as well as CD4+ T cells (R2=0.54, p=0.0155) (online supplemental figure S4B), further revealing the value of 68Ga-NOTA-XH05 in evaluating the immune status within tumors.

68Ga-NOTA-XH05 PET imaging reflects response to immunotherapy

The method of immunotherapy employed in this study involved intratumoral injection of CpG oligonucleotide (CpG-ODN, CpG). CpG is a synthetic oligodeoxynucleotide containing unmethylated cytosine-guanine sequence motifs that can activate toll-like receptor 9 (TLR9) present on antigen-presenting cells, thereby simulating the immunostimulatory activity of bacterial DNA and triggering intracellular signaling and immune activation.24 Intratumoral injection of CpG can induce the infiltration of antitumor T cells by activating dendritic cells, leading to an abscopal effect. This unique property of CpG has prompted its investigation as an immune adjuvant in clinical research, demonstrating promising outcomes when combined with ICIs therapy and enhancing the efficacy of tumor vaccines.25–27

Here, mice bearing B16-F10 tumor (~5–6 mm) were administered CpG or PBS intratumorally on days 0, 2, 4 and imaged on day 8 (figure 3A). Specific 68Ga-NOTA-XH05 PET imaging uptake was calculated by dividing tumor uptake by descending aorta uptake to derive a TBR as a measure of specifically signal. In C57BL/6 mice, CpG treated tumors had a TBR of 2.23±0.39, which was significantly higher than that of PBS treated mice (TBR=1.35±0.30, p<0.0001) (figure 3B,C). In immunodeficient BALB/c-nu mice, TBR was significantly lower than that in CPG-treated C57BL/6 mice (TBR=1.08±0.14, p<0.0001). Following imaging, C57BL/6 mice were sacrificed and ex vivo radioactivity was assessed by γ counter, which confirmed PET analysis (online supplemental figure S5). At the same time, the uptake of spleen was significantly decreased in the CpG group (0.28%ID/g±0.04%ID/g vs 0.39%ID/g±0.08%ID/g, p=0.0145). In addition, although the difference was not significant, the tumor local lymph node uptake was lower in the CpG group. The larger mass of local lymph nodes and spleen in the CpG group may be the reason for the lower percentage of injected dose per gram of tissue (online supplemental figure S6A,S7A).

Figure 3

Tumor uptake of 68Ga-NOTA-XH05 increases after CpG therapy. (A) Timeline of drug injection and PET imaging. (B) Representative PET/CT images of CpG-treated and control C57BL/6 mice and CpG-treated BALB/c-nu mice on day 8. Tumors are circled and marked with a ‘T’. (C) Individual TBR of CpG-treated and control C57BL/6 mice and CpG-treated BALB/c-nu mice (n=4–9). ****p<0.0001. PET, positron emission tomography.

Tumor volume was monitored from the day of the first treatment (day 0), and some tumors in the CpG group responded well to the treatment while others did not respond (figure 4A,B). The mice whose tumor volume reached 500 mm3 on or before day 8 were considered to be non-responders. After grouping according to the response to CpG, it was found that the TBR of the responders (TBR=1.99±0.10) was higher than that of the control group (TBR=1.35±0.30, p=0.0041) but lower than that of the non-responders (TBR=2.54±0.40, p=0.0341), indicating that 68Ga-NOTA-XH05 could effectively distinguish the responders from the non-responders (figure 4C,D).

Figure 4

68Ga-NOTA-XH05 PET imaging reflects response to immunotherapy. (A, B) Individual (A) and average (B) tumor volume of treated responders (green), treated non-responders (red), and control mice (black). (C) Representative PET/CT images of treated responders, treated non-responders, and control mice on day 8. Tumors are circled and marked with a ‘T’. (D) Individual TBR of treated responders, treated non-responders, and control mice. *p<0.05, **p<0.01, ***p<0.001. PET, positron emission tomography; TBR, tumor to blood ratio.

Verification of LAG-3 by immunofluorescence and flow cytometry

Immunofluorescence and flow cytometry were used to validate the PET imaging findings. At the end of observation, mice were euthanized and B16-F10 tumors were harvested, and HE staining and immunofluorescence staining were performed. HE staining showed that there was a large amount of necrosis in the tumor of the responders while the necrotic area was less in the non-responders and the control group (figure 5A). Immunofluorescence showed that there was a large number of T cells infiltrating in both the responders and the non-responders, which was significantly higher than that in the control group, and the expression of LAG-3 in the non-responders was higher than that in the responders, which was consistent with the PET analysis (figure 5B). The difference of CD3 and LAG-3 expression among the responders, the non-responders and the control group suggested that the ineffective CpG therapy might be due to the overexpression of LAG-3 in the tumors, which made T cells exhausted so that even if a large number of T cells infiltrated, they could not exert their antitumor effect. This result reflected the influence of LAG-3 expression level on the efficacy of tumor immunotherapy and suggested the importance of distinguishing between different LAG-3 expression states.

Figure 5

HE staining and immunofluorescence staining of tumors of B16-F10 bearing mice after CpG therapy. (A) HE staining of tumor tissue sections of B16-F10. There was a large amount of necrosis in the tumor of the responders while tumors of the non-responders and the control mice mainly contained tumor cells. Scale bar=1 mm. (B) Immunofluorescence staining showed colocalization (yellow) of LAG-3 (red) and CD3 (green) in tumor tissue sections of B16-F10. T cells infiltration in both the responders and the non-responders was significantly higher than that in the control group, and the expression of LAG-3 in the non-responders was higher than that in the responders. DAPI (blue) was used to visualize cell nucleus. Scale bar=100 μm.

To further investigate the changes in the expression of LAG-3 on different T cell subsets in the tumor after CpG treatment, flow cytometry analysis was performed on day 8. The proportions of different T cell subsets in T cells and the expression of LAG-3 were measured. The CD8+ to CD4+ T cell ratio in the CpG group was 1.95±0.43, which was significantly higher than that in the control group (1.12±0.18, p=0.0014), indicating that after CpG treatment, killer T cells in the tumor increased (figure 6A). The proportion of LAG-3+ T cells in the tumors was significantly higher in the CpG group than in the control group (23.66%±7.27% vs 11.16%±7.06%, p=0.0090) (figure 6B). Analysis of T cell subsets showed that the proportion of LAG-3+ CD8+ T cells in the CpG group was significantly increased (14.97%±4.18% vs 4.27%±2.22%, p=0.0002) while there was no significant difference in the proportion of LAG-3+ CD4+ T cells (p=0.1513) (figure 6C). The expression of LAG-3 in CD4+ (MFI=2747.56±290.75 vs 2207.00±279.00, p=0.0055) and CD8+ (MFI=1216.00±106.61 vs 903.80±138.83, p=0.0005) T cells was significantly increased in the CpG group (figure 6D). Moreover, during the dissection of the mice, it was found that the local lymph nodes and spleen of the CpG group were significantly enlarged, suggesting that intratumoral injection of CpG had an effect on the local and systemic immunity of the mice (online supplemental figure S6A,S7A). To explore the changes in local and systemic immunity in mice, flow cytometry analysis of local lymph nodes and spleen was further performed, and the changes in T cells and T cell subsets in the local lymph nodes and spleen were similar to those in the tumors, confirming the speculation above (online supplemental figure S6,S7).

Figure 6

Flow cytometry analysis of LAG-3 expression on T cells and T cell subsets within B16-F10 tumors treated by intratumoral CpG injection on day 8. (A) The ratio of CD8+ T cells to CD4+ T cells in tumors of CpG-treated (n=9) or control mice (n=5). (B, C) Expression of LAG-3 (cumulative %) on T cells (B) and T cell subsets (C) in tumors of CpG-treated (n=9) or control mice (n=5). (D) Expression of LAG-3 (MFI) on T cell subsets in tumors of CpG-treated (n=9) or control mice (n=5). **p<0.01, ***p<0.001.

68Ga-NOTA-XH05 PET imaging reflects abscopal effects of CpG

Flow cytometry showed that the systemic immune status of mice was also changed after CpG treatment, reflecting the abscopal effect of CpG intratumoral injection. To verify whether this change could affect untreated tumors, and explore whether 68Ga-NOTA-XH05 could reflect this effect, we chose a model that recapitulates an in situ tumor adjuvant vaccination strategy.25 B16-F10 tumors were implanted in both axillae of the mice, and only the right tumors (~5–6 mm) were intratumorally injected with CpG or PBS on days 0, 2, 4 (figure 7A). Tumor volume was monitored from the day of the first treatment (day 0) (figure 7B). In the CpG group, the growth rate of the left tumor without CpG injection was significantly lower than that of the control group, as well as the growth rate of the right tumor treated with CpG, confirming the abscopal effect of CpG injection (figure 7C).

Figure 7

68Ga-NOTA-XH05 PET imaging monitors abscopal effects of CpG. (A) Timeline of drug injection and PET imaging. Only the right tumors were intratumorally injected with CpG (n=8) or PBS (control, n=6), in the case of B16-F10 tumors implantation in both axillae of the mice. (B, C) Individual (B) a average (C) tumor volume of right tumors of CpG-treated mice (blue), left tumors of CpG-treated mice (green), right tumors of control mice (red), left tumors of control mice (purple). (D) Representative PET/CT images of CpG-treated and control mice on day 2 and day 8. Tumors are circled and marked with a ‘T’. (E) Individual TBR of CpG-treated mice (blue), left tumors of CpG-treated mice (green), right tumors of control mice (red), left tumors of control mice (purple) on day 2 and day 8. *p<0.05, **p<0.01, ****p<0.0001. PET, positron emission tomography; TBR, tumor to blood ratio; PBS, phosphate buffered saline.

68Ga-NOTA-XH05 PET imaging was performed on days 2 and 8 (figure 7D,E). On day 2, there was no significant difference in the TBR of the left tumors between the CpG group and the control group (p>0.9999). However, on day 8, the TBR of the left tumors in the CpG group (TBR=2.64±0.42) exhibited a significant increase compared with that on day 2 (TBR=1.43±0.15, p<0.0001), and it was also significantly higher than that of the left tumor in the control group (TBR=1.90±0.51, p=0.0030). These findings indicated that 68Ga-NOTA-XH05 could effectively reflect the abscopal effect of CpG intratumoral injection. Furthermore, on day 8, there was a significant elevation in TBR of right tumors within the CpG group compared with those in the control group (2.22±0.34 vs 1.59±0.33, p=0.0169), consistent with results obtained from unilateral tumor model.

At the end of observation, mice were euthanized and tumors were harvested for HE staining and immunofluorescence staining. HE staining showed that there was a large amount of necrosis within the tumors of both sides in the CpG group while there was less necrosis within the tumors in the control group (online supplemental figure S8A). Using ki67 as a tumor cell proliferation marker, immunofluorescence staining showed that the expression of ki67 in the tumors in the CpG group was significantly lower than that in the control group, indicating that the proliferation of tumor cells in the bilateral tumors in the CpG group was slowed (online supplemental figure S8B). The abscopal effect of CpG intratumoral injection therapy was confirmed by similar responses of tumors on two sides. Immunofluorescence staining for CD3 and LAG-3 showed that the expression of CD3 and LAG-3 in bilateral tumors was increased in the CpG group, which was consistent with the 68Ga-NOTA-XH05 PET imaging results, indicating that the injection of CpG into one tumor could affect the immune status in the untreated tumor, and increase the infiltration and activation of T cells (online supplemental figure S9A). In addition, CD4+ (online supplemental figure S9B) and CD8+ (online supplemental figure S9C) T cells in T cell subsets were significantly increased in bilateral tumors in the CpG group.

Discussion

The introduction of ICIs targeting CTLA-4 and PD-1/PD-L1 has revolutionized the field of immuno-oncology. Despite their sustained clinical efficacy, there is significant heterogeneity in patient response, with only a minority experiencing benefits from these therapies.28 Consequently, there is an increasing focus on identifying novel ICIs such as LAG-3, which plays a crucial role in cancer-specific T cell activation and may contribute to tumor escape within the tumor microenvironment. Ongoing clinical trials are currently investigating the potential of targeting LAG-3.11 Assessing IC expression within tumors holds promise for selecting patients who are more likely to respond favorably to ICIs and predicting treatment outcomes. Nuclear imaging has emerged as a valuable non-invasive and repeatable modality for detecting IC expression in various studies.29

In this study, we successfully developed a novel PET imaging agent, 68Ga-NOTA-XH05, by using gallium-68 for the labeling of Cyclo (CVPMTYRAC), which exhibits high affinity toward LAG-322. The labeling process of 68Ga-NOTA-XH05 was straightforward and resulted in high radiochemical purity following purification. This agent enables repeated, quantitative, and non-invasive imaging assessment of LAG-3 expression with sensitivity and specificity. Furthermore, our findings demonstrate the potential utility of non-invasive imaging of LAG-3 on activated T cells as a promising biomarker for early response evaluation in cancer immunotherapy.

The B16-F10 melanoma subcutaneous tumor models were subjected to in vivo small animal PET imaging study, which revealed clear visualization of the tumor at 0.5-hour postinjection of 68Ga-NOTA-XH05. The optimal contrast was observed at 1 hour, while a decrease in contrast was noted at 2 hours. Flow cytometry analysis demonstrated a positive correlation between tumor uptake of 68Ga-NOTA-XH05 and the mean fluorescence intensity of LAG-3 in the tumor, as well as LAG-3 expression on CD3+ T cells. This indicates the potential of this radiotracer for dynamic monitoring of LAG-3 expression within tumors.

LAG-3 expression is generally considered to be associated with poor prognosis because of its suppressive effect on the immune response, but contradictory results have been observed in some studies, including studies of breast cancer,30–32 esophageal adenocarcinoma,33 gastric cancer34 and non-small cell lung cancer.35 Moreover, a meta-analysis found that high LAG-3 expression level was associated with a better overall survival in several tumors, particularly in early-stage disease.36 In these cases, LAG-3 can be considered a marker of T cell activation, which is also reflected in some studies of immunotherapy. A study of advanced gastric cancer patients treated with nivolumab showed that the LAG-3 expression level on T cells correlated with the efficacy of nivolumab therapy.37 Single-photon emission CT (SPECT) studies with a 99mTc-labled single-domain antibody targeting LAG-3 also showed increased tracer uptake in MC38 tumor after PD-1 antibody treatment.38 In this study, we obtained similar results in B16-F10 tumor-bearing mice treated with CpG. On the day 8 after the first CpG intratumoral injection, 68Ga-NOTA-XH05 PET imaging showed that the TBR of CpG group was significantly higher than that of the control group, demonstrating the potential of LAG-3 as a predictive biomarker of effective response for oncoimmunotherapy. In addition to its expression on the cell surface, LAG-3 can also produce soluble LAG-3 after proteolytic cleavage. Some studies have shown that plasma sLAG-3 level increases in a dose-dependent way during ICIs treatment.39 However, there was no significant difference in blood uptake of 68Ga-NOTA-XH05 between the CpG group and the control group in this study. It may be related to the limited detection ability of the tracer. Whether 68Ga-NOTA-XH05 can reflect the level of sLAG-3 in vivo or its combination with plasma sLAG-3 can judge the therapeutic effect of immunotherapy is worthy of further study.

We further investigated whether there exists a disparity in LAG-3 expression between the responders and non-responders using 68Ga-NOTA-XH05, considering that only a subset of mice responded to CpG stimulation. Notably, our results revealed a significantly higher TBR in non-responders compared with responders. Consistently, fluorescence staining exhibited an abundance of T cells with excessive LAG-3 expression infiltrating the non-responder group, suggesting that LAG-3 might impede the tumor-killing function of T cells. These findings underscore the dual role of LAG-3 in tumor development. On one hand, LAG-3 is expressed on activated T cells and its expression level may partially reflect the antitumor efficacy of activated T cells during early-stage immunotherapy. On the other hand, as an immunosuppressive molecule, overexpression of LAG-3 may hinder T cell functionality leading to exhaustion. Utilization of 68Ga-NOTA-XH05 could provide valuable insights into discerning LAG-3 expression patterns associated with T cell activation and exhaustion following immunotherapy and aid in determining whether combining LAG-3 inhibitors would be beneficial for treatment.

Additionally, we established a bilateral B16-F10 subcutaneous tumor model to investigate the potential of 68Ga-NOTA-XH05 in assessing the abscopal effect of intratumoral CpG injection. PET imaging using 68Ga-NOTA-XH05 revealed no significant difference in TBR between the non-injected side of the CpG group and control group during early stage. However, at later stage, there was a notable increase in uptake of 68Ga-NOTA-XH05 on the non-injected side, with TBR surpassing that of the control group. These findings are consistent with suppressive effect of CpG on both therapeutic and abscopal sites, highlighting 68Ga-NOTA-XH05’s ability to assess abscopal effects. This pattern resembles a previous PET imaging study utilizing OX40 as an indicator for T cell activation; however, unlike our study where immune response persisted due to absence of tumor regression, uptake of 64Cu-DOTA-AbOX40 decreased later due to tumor regression.40

There are still unresolved issues in this study that require further investigation. The affinity of 68Ga-NOTA-XH05 to hLAG-3 is comparable to that of mLAG-3, and polypeptide small molecule probes typically exhibit good safety profiles,21 which holds promise for human applications. However, additional studies on its efficacy and safety in mice with humanized immune systems as well as clinical trials are warranted. Furthermore, while the uptake of 68Ga-NOTA-XH05 increased in melanoma receiving immunotherapy, higher uptake might be associated with poorer efficacy. Conversely, a preclinical study on LAG-3 PET imaging in lung cancer demonstrated that the combination treatment group using PD-1 antibody and STING agonist exhibited higher tracer uptake compared with monotherapy and control groups, which correlated with improved therapeutic effects.41 The differences in tumor growth indicated by elevated LAG-3 expression could be attributed to different tumor types and treatment regimens leading to distinct stages of tumor development. Further research is needed to determine the applicability of 68Ga-NOTA-XH05 across various tumors and treatment protocols as well as establish a threshold for tracer uptake indicative of T cell activation and exhaustion for more accurate evaluation purposes. Although these studies present challenges, given that LAG-3 is an effective target for immunotherapy, elucidating the clinical significance of increased LAG-3 expression following immunotherapy can help ascertain whether combining anti-LAG-3 therapy is necessary for enhancing efficacy.

In conclusion, we have developed a novel peptide-based PET tracer, 68Ga-NOTA-XH05, which specifically targets LAG-3. This innovative tracer enables non-invasive detection of LAG-3 expression in melanoma and facilitates the evaluation of immunotherapy efficacy. As LAG-3 received more and more attention as an immunotherapeutic target and drugs targeting it are expected to undergo rapid growth in the future, improved strategies such as non-invasive PET are valuable tools for investigating the expression patterns of targets in tumor as well as the therapeutic effects of newly developed targeted therapies, including antibodies and small inhibitors. In the future, imaging of immunotherapy biomarkers may be used for patient stratification and therapeutic monitoring.

Data availability statement

Data are available on reasonable request. All data relevant to the study are included in the article.

Ethics statementsPatient consent for publication

Not applicable.

Ethics approval

Animal experiments were approved by the Institutional Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology ((2022)3259).