Fibroblast Activation Protein-Targeted Radioligand Therapy with ¹⁷⁷Lu-EB-FAPI for Metastatic Radioiodine-Refractory Thyroid Cancer: First-in-Human, Dose-Escalation Study

Abstract

Purpose:

Fibroblast activation protein (FAP) is a promising target for tumor treatment. In this study, we aimed to investigate the safety and efficacy of the albumin binder-conjugated FAP-targeted radiopharmaceutical, ¹⁷⁷Lu-EB-FAPI (¹⁷⁷Lu-LNC1004), in patients with metastatic radioiodine-refractory thyroid cancer (mRAIR-TC).

Patients and Methods:

This open-label, non-randomized, first-in-human, dose-escalation, investigator-initiated trial had a 3+3 design and involved a 6-week ¹⁷⁷Lu-LNC1004 treatment cycle in patients with mRAIR-TC at 2.22 GBq initially, with subsequent cohorts receiving an incremental 50% dose increase until dose-limiting toxicity (DLT) was observed.

Results:

¹⁷⁷Lu-LNC1004 administration was well tolerated, with no life-threatening adverse events observed. No patients experienced DLT in Group A (2.22 GBq/cycle). One patient experienced grade 4 thrombocytopenia in Group B (3.33 GBq/cycle); hence, another three patients were enrolled, none of whom experienced DLT. Two patients experienced grade 3 and 4 hematotoxicity in Group C (4.99 GBq/cycle). The mean whole-body effective dose was 0.17 ± 0.04 mSv/MBq. Intense ¹⁷⁷Lu-LNC1004 uptake and prolonged tumor retention resulted in high mean absorbed tumor doses (8.50 ± 12.36 Gy/GBq). The mean effective half-lives for the whole-body and tumor lesions were 90.20 ± 7.68 and 92.46 ± 9.66 hours, respectively. According to RECIST, partial response, stable disease, and progressive disease were observed in 3 (25%), 7 (58%), and 2 (17%) patients, respectively. The objective response and disease control rates were 25% and 83%, respectively.

Conclusions:

FAP-targeted radioligand therapy with ¹⁷⁷Lu-LNC1004 at 3.33 GBq/cycle was well tolerated in patients with advanced mRAIR-TC, with high radiation dose delivery to the tumor lesions, encouraging therapeutic efficacy, and acceptable side effects.

Translational Relevance.

Fibroblast activation protein (FAP) is a promising target for diagnosing and treating numerous malignant tumors. ¹⁷⁷Lu-EB-FAPI (LNC1004) is a FAP-targeted radiopharmaceutical conjugated to Evans blue (EB, albumin binder). In this study, FAP-targeted radioligand therapy with ¹⁷⁷Lu-LNC1004 was well tolerated in patients with advanced metastatic radioiodine-refractory thyroid cancer, with high radiation doses delivered to the tumor lesions. This treatment showed encouraging therapeutic efficacy with few side effects. Therefore, prospective, randomized, controlled, multicenter clinical trials are warranted.

Introduction

Cancer-associated fibroblasts (CAF) are central players in immune regulation, abundant in the tumor microenvironment, that have been found to contribute to cancer progression and metastasis (1, 2). Fibroblast activation protein (FAP), a type II transmembrane cell surface serine protease belonging to the dipeptidyl peptidase family, is highly expressed in several epithelial carcinoma CAFs (3, 4). Since the first evaluation of radioiodine (¹³¹I)-labeled FAP-specific mAb F19 in patients with metastatic colorectal cancer, FAP has been identified as a potential target for cancer imaging and treatment because of its limited expression in normal tissues (5). Recently, various radiolabeled quinoline-based FAP inhibitors (FAPI) have been developed that facilitate FAP-positive tumor imaging (6–8) and are effective for use in FAP-targeted radioligand therapy (RLT; refs. 9–12). The direct structural modification of FAPIs to enhance tumor uptake and retention while reducing or maintaining non-target tissue accumulation is an optimal method for therapeutic radiopharmaceutical development. However, the relatively rapid washout from the tumor limits their therapeutic potential, especially when labeled with radionuclides with long half-lives, such as lutetium 177 (¹⁷⁷Lu; refs. 13–15). Moreover, achieving significant enhancement of pharmacokinetic properties through subtle structural modifications may be challenging.

Serum albumin has emerged as a versatile carrier for therapeutic agents because it is the most abundant in plasma (16). Because the association of bioactive drugs with albumin is reversible, the albumin–drug complex functions as a drug reservoir that enhances drug distribution and bioavailability (17). Evans blue (EB) is a promising albumin-binding moiety that exhibits a relatively high affinity for binding site 1 on serum albumin, with prolonged blood circulation and therapeutic efficacy following the conjugation of targeting molecules with EB derivatives (18, 19). Recently, several EB-modified FAPI (EB-FAPI) radiopharmaceuticals exhibited remarkable tumor growth suppression with negligible side effects in preclinical studies (20, 21).

Approximately 5%–15% of differentiated thyroid cancer (DTC) and 50% of metastatic DTC develop radioactive iodine–refractory DTC, which loses avidity to ¹³¹I and is associated with disease aggressiveness (22). Anaplastic thyroid cancer (ATC) is a highly aggressive malignancy with a mean survival time of <6 months following diagnosis. Medullary thyroid cancer (MTC), arising from the parafollicular C cells of neuroendocrine origin, accounts for approximately 2% of all thyroid cancers (23, 24). Many of these patients with thyroid cancer become radioiodine-refractory (RAIR) and cannot benefit from ¹³¹I therapy. Tyrosine kinase inhibitors (TKI) are the standard treatment for metastatic RAIR thyroid cancer (mRAIR-TC); however, managing patients that progress even after TKI treatment present a considerable clinical challenge. Previous studies have reported that FAP expression correlates with thyroid cancer metastasis and progression. Our previous study further revealed intense FAPI accumulation in mRAIR-TC tumors by evaluating the diagnostic ability of ⁶⁸Ga-FAPI PET/CT (25). Thus, investigating FAP-targeted RLT in the management of mRAIR-TC is of great clinical significance to develop an effective alternative to the current standard-of-care practice.

Therefore, this first-in-class study explored the safety, tolerability, dosimetry, and efficacy of a novel FAP-targeted therapeutic radiopharmaceutical, ¹⁷⁷Lu-EB-FAPI (¹⁷⁷Lu-LNC1004), in patients with mRAIR-TC with disease progression after TKI treatment.

Patients and Methods

This single-center, open-label, non-randomized, first-in-human, dose-escalation investigator-initiated trial was conducted at The First Affiliated Hospital of Xiamen University, China. This study was approved by the Institutional Review Board and performed in accordance with the Declaration of Helsinki. The study is registered at ClinicalTrials.gov (registration number: NCT05410821). The multidisciplinary tumor board affiliated with the hospital approved the use of ¹⁷⁷Lu-LNC1004. All patients provided written informed consent.

Patient selection

The inclusion criteria were adult patients (>18 years) with (i) histologically confirmed metastasized thyroid cancer (including DTC, MTC, and ATC), with a pathological subtype of DTC diagnosed as mRAIR-DTC based on previous ¹³¹I therapy results; (ii) unresectable tumors; (iii) post-TKI therapy disease progression; and (iv) tumor lesions showing increased radiotracer uptake on ⁶⁸Ga-FAPI-46 PET/CT (defined as a maximum standardized uptake value ≥10 in more than 50% of tumor lesions). The exclusion criteria were as follows: (i) serum creatinine level >150 μmol/L; (ii) hemoglobin level <8.0 g/dL; (iii) white-cell count <2.0×10⁹/L; (iv) platelet count <50×10⁹/L; (v) total bilirubin level >3 times the upper limit of the normal range and serum albumin level <2.0 g/dL; (vi) cardiac insufficiency, including carcinoid heart valve disease, severe allergy, or hypersensitivity to radiographic contrast material; (vii) claustrophobia; and (viii) pregnancy or breastfeeding.

Study design

This study had a classic 3+3 dose-escalation design and was conducted over 6 weeks. On the basis of previous clinical data of ¹⁷⁷Lu-labeled EB derivatives, including EB-prostate-specific membrane antigen (PSMA) and EB-TATE, the initial dose was set at 2.22 GBq (60 mCi), whereas subsequent cohorts received an incremental 50% dose increase until dose-limiting toxicity (DLT) was observed. Initially, three patients were included in the first-dose group (2.22 GBq/cycle). If none of the three patients in a cohort experienced DLT, another three patients were enrolled at the next higher dose level. If one of the three patients experienced DLT at a certain dose level, three more patients were enrolled at the same dose level. The highest dose with no more than one of the six patients experiencing DLT was the MTD. Adverse events (AE) were graded using the Common Terminology Criteria for Adverse Events Version 5 (CTCAE 5.0). DLT was defined as any ¹⁷⁷Lu-LNC1004-related AE ≥ grade 3 (G3). The study flowchart is shown in Supplementary Fig. S1.

The primary endpoint assessed the safety and MTD of ¹⁷⁷Lu-LNC1004 RLT in patients with mRAIR-TC. The secondary endpoints were dosimetry and determination of the preliminary treatment efficacy of ¹⁷⁷Lu-LNC1004.

¹⁷⁷Lu-LNC1004 synthesis and administration

¹⁷⁷Lu-labeling of EB-FAPI was performed using previously published methods (26). First, the LNC1004 [1, 4, 7, 10-tetra-azacyclododecane-1, 4, 7, 10-tetraacetic acid (DOTA)-EB-FAPI] precursor was incubated with the required radioactivity of ¹⁷⁷LuCl₃ (ITM Isotope Technologies) in 0.5 mol/L sodium acetate buffer (adjusted to pH 5.5 with hydrochloric acid). After the mixture was incubated at 95°C for 30 minutes, it was diluted with 0.9% saline solution and sterile filtered. Furthermore, the samples were tested for quality, endotoxins, and sterility. Radiochemical purity, determined by thin-layer and high-performance liquid chromatography, was consistently higher than 98%.

No fasting, special diet, or other specific preparations were required on the day of ¹⁷⁷Lu-LNC1004 administration. Before treatment, 4 mg of ondansetron was administered to prevent nausea and vomiting. Patients received intravenous hydration (1 L of 0.9% saline) 30 minutes before ¹⁷⁷Lu-LNC1004 administration. The radiopharmaceutical ¹⁷⁷Lu-LNC1004 was diluted with 100 mL of 0.9% saline, which was co-administered slowly via an intravenous infusion for 20–30 minutes. Symptoms and vital parameters were monitored before, during, and after treatment. The treatment regimen was planned for up to 2 cycles, with an inter-cycle time interval of 6 weeks.

Scintigraphy and SPECT/CT imaging

The kinetics of ¹⁷⁷Lu-LNC1004 were determined using planar whole-body scintigraphy in the anterior and posterior projections at 1, 4, 24, 48, 72, 96, and 168 hours after administration of the radiopharmaceutical using a double-head γ-camera (Symbia T16; Siemens) equipped with a low-energy high-resolution parallel-hole collimator and a 15% energy window set symmetrically over the 113 keV and 208 keV photopeaks. The table speed was 10 cm/min, and the matrix was 256 × 1,024, producing 2.4 × 2.4 mm pixels. Furthermore, whole-body SPECT/CT was performed at 72 hours after administration. CT was performed for attenuation correction without a contrast agent (tube voltage, 130 kVp; tube current-time product, 17 mAs; beam pitch, 1.5; section width, 5 mm), and SPECT scans were acquired at 128 angles over 360° and 25 seconds per stop. Images were iteratively reconstructed and corrected for attenuation and scatter (Flash 3D Siemens, 4 subsets and 8 iterations; Gaussian intersection smoothing filter; attenuation coefficient, 0.15/cm). The image matrix was 128 × 128, resulting in a cuboid voxel length of 4.8 mm.

Dosimetry protocol

Dosimetry was estimated in the first treatment cycle for each patient according to a procedure reported previously (26–28). First, seven serial planar whole-body scintigraphy and one whole-body SPECT/CT scans were performed per patient. As the patients were not allowed to empty their bladders before the first scan, the total-body counts acquired immediately after the injection were defined as 100% of the administered activity. Regions of interest were selected and drawn manually over the source regions on the acquired scintigraphy images by the same physicist in collaboration with a nuclear medicine physician, who selected suitable lesions that had the highest uptake for dosimetry. Images were then analyzed using the Hermes system (Hermes Medical Solutions). SPECT/CT scans were reconstructed and quantified using the Hermes SUV SPECT software. The mean absorbed doses to organs and tumors were estimated using the built-in Organ Level Internal Dose Assessment/Exponential Modeling (OLINDA/EXM) version 2.1 software. The International Commission on Radiological Protection 89 adult model and the spheres model were used for normal organs and tumor lesions, respectively (both included in OLINDA/EXM 2.1). The model was adapted to individual normal-organ and tumor volumes obtained from the latest CT scan of the patient. The following parameters were assessed using the dosimetry protocol: uptake as a fraction of the administered activity (%AA), effective half-life (h), residence time (h), mean absorbed organ (mGy/MBq), and tumor dose (Gy/GBq; Supplementary Fig. S2).

Clinical, radiologic, and laboratory follow-up

The complete blood counts, liver and kidney function results, and tumor marker levels (including thyroglobulin and thyroglobulin antibody) were systematically determined before and during post-treatment follow-up in all patients every 2 weeks. Patient records were reviewed for any incidence of hematological, gastrointestinal, renal, hepatic, or other AEs, and the grade was assigned according to the CTCAE 5.0. All circumstances that resulted in the cessation or delay of treatment were documented. Changes in circulating tumor marker levels were also recorded. According to the RECIST version 1.1, tumor response was monitored using functional imaging (⁶⁸Ga-FAPI-46 PET/CT; slice thickness, 5 mm) at 6 weeks after each cycle, and the objective response rate (ORR) and disease control rate (DCR) were assessed.

Statistics analyses

All statistical analyses were performed using SPSS (SPSS Statistics for Windows, version 21.0; IBM Corp.) and GraphPad Prism software (GraphPad Software). Normally distributed quantitative data were expressed as the mean ± SD, whereas non-normally distributed data were expressed as the median with interquartile range.

Data availability

The data generated in this study are available within the article and its Supplementary Data Files. Further data generated in this study are not publicly available due to patient privacy but are available upon reasonable request from the corresponding author.

Results

Safety and tolerability of ¹⁷⁷Lu-LNC1004

A total of 12 patients were enrolled between May 2022 and January 2023. Patient characteristics are summarized in Table 1. Briefly, three patients received 2.28 ± 0.10 GBq, six patients received 3.50 ± 0.09 GBq, and three patients received 4.80 ± 0.28 GBq. No life-threatening AEs, clinically detectable pharmacological effects, or immediate AE-related significant changes in vital signs or laboratory examination results were observed. In Patient 4 (Group B), the superficial cervical lymph node and clavicle metastatic lesions developed swelling 1 hour after injection, which persisted for several days and subsided on day 7, followed by bleeding and necrosis of the clavicle lesions. In Patients 5 and 6 (Group B), pain symptoms in osseous metastases increased the day after treatment and NSAIDs were administered to Patient 5, whose pain symptoms persisted for 6 days, but gradually eased thereafter. Transient fatigue was also reported in Patients 2 (Group A) and 4 (Group B). Notably, Patient 8 (Group B) showed an improvement in self-reported physical capacity and quality of life after treatment.

Table 1.

Patient characteristics.

Patient number	Age (y)	Sex	Pathologic subtype (genomic status)	Metastases	Time from initial diagnosis (y)	Relevant previous surgery	Previous course of 131I therapy	Previous cumulative dose of 131I (GBq)	TKI treatment	Other relevant treatment	ECOG performance status	Baseline tumor markers	TRAE
1	52	M	PTC (BRAFV600E−)	LN, oss, pul, sub	4	Total thyroidectomy	3	20.35	Sorafenib	125I seed implantation	1	Tg 36.1 ng/mL	Ache in lesions
2	53	F	PTC	LN, pul	9	Total thyroidectomy, cervical LN dissection, tracheotomy	2	16.65	Sorafenib	None	1	Tg 18.3 ng/mL	None
3	58	M	PTC	LN, pul	9	Total thyroidectomy	4	25.90	Anlotinib	None	1	Tg 27.5 ng/mL	None
4	35	M	PTC (BRAFV600E+)	LN, oss	10	Total thyroidectomy, cervical LN dissection, tumor excision and internal fixation of right clavicle	4	24.05	Apatinib, anlotinib	None	2	Tg 18,507 ng/mL	Swelling, bleeding, and necrosis in superficial lesions
5	42	M	PTC	LN, oss, pul, hep, pan	13	Total thyroidectomy, tumor excision and internal fixation of the vertebra	6	59.20	Lenvatinib	EBRT to osseous metastases	3	Tg 960 ng/mL	“Flare” phenomenon
6	55	F	MTC (BRAFV600E−, RET−)	Oss	5	Total thyroidectomy	0	0.00	Anlotinib	None	3	CEA 167.23 ng/mL	“Flare” phenomenon
												CT 5,648 pg/mL
7	48	F	FTC	Oss, pul, hep	4	Total thyroidectomy, tumor excision and internal fixation of the vertebra	5	44.40	Sorafenib	None	2	Tg 3,371 ng/mL	None
8	72	F	PTC for the primary tumor, ATC for mediastinal LN metastases	LN, pul	6	Total thyroidectomy, endobronchial stent placement	0	0.00	Anlotinib	None	3	TgAb 54,219 IU/mL	None
9	32	M	PTC	Pul, LN	14	Total thyroidectomy, cervical LN dissection	5	37.00	Sorafenib	None	1	TgAb 22,088 IU/mL	None
10	56	M	PTC	Pul, LN	11	Total thyroidectomy, cervical LN dissection	2	7.40	Sorafenib	None	1	Tg 1,285 ng/mL	None
11	64	M	PTC	Pul, LN, sub	5	Total thyroidectomy	2	12.95	Sorafenib	None	1	Tg 101 ng/mL	None
12	33	M	PTC	Rel, LN, pul	5	Total thyroidectomy	4	29.60	Anlotinib	None	1	Tg 646 ng/mL	None

Abbreviations: ATC, anaplastic thyroid cancer; CEA, carcinoembryonic antigen; CT, calcitonin; EBRT, external-beam radiation therapy; ECOG, Eastern Cooperative Oncology Group; FTC, follicular thyroid cancer; hep, hepatic; LN, lymph node; MTC; medullary thyroid cancer; oss, osseous; pan, pancreatic; PTC, papillary thyroid cancer; pul, pulmonary; Rel, relapse; sub, subcutaneous; Tg, thyroglobulin; TgAb, thyroglobulin antibody; TRAE, treatment‐related adverse event.

No patients experienced hepatotoxicity or nephrotoxicity following ¹⁷⁷Lu-LNC1004 RLT administration, whereas hematotoxicity was observed in all 3 groups. In Group A, a grade 1 (G1) short-term and self-limiting thrombocytopenia occurred in Patient 2 after the first RLT cycle, which reappeared alongside G1 leukopenia and G2 neutropenia after the second RLT cycle. Thrombocytopenia was observed in two of the three patients in Group B, one of which was recorded as G4 (Patient 6) and led to the involvement of another three patients (Table 2). Despite experiencing G1 thrombocytopenia and leukopenia after the first RLT cycle, Patient 5 recovered spontaneously. However, the thrombocytopenia worsened after the second RLT cycle, which was concomitant with G1 anemia, and required drug intervention. Similarly, in Patient 7, the G1 thrombocytopenia deteriorated to G2 and induced hematotoxicity after 2 RLT cycles, with concomitant G2 leukopenia and G1 anemia. In Patient 8, G1 thrombocytopenia occurred after the first RLT cycle. Interestingly, hematotoxicity was reversible after treatment and was not observed after the second RLT cycle. Two patients (Patient 12 and 10) in Group C experienced G4 and G3 thrombocytopenia after the first and second RLT cycles, which terminated further dose escalation (Table 2). Patient 12 further experienced G3 leukopenia and G4 neutropenia and required drug treatment and dose reduction at the second RLT cycle. However, the hematotoxicity observed was reversible after the second RLT cycle. Therefore, the recommended dose for future clinical trials was found to be 3.33 GBq/cycle.

Table 2.

Hematotoxicity, hepatotoxicity, and nephrotoxicity before and 6 weeks after ¹⁷⁷Lu-LNC1004 RLT, according to CTCAE 5.0.

		Before therapy		After the first RLT		After the second RLT
Group	Patient number	CTCAE term	Grade	CTCAE term	Grade	CTCAE term	Grade
A	1	Anemia	1	Anemia	1
(n = 3)	2			Thrombocytopenia	1	Thrombocytopenia	1
						Leukopenia	1
						Neutropenia	2
	4			Anemia	1	Leukopenia	1
						Neutropenia	1
	5			Thrombocytopenia	1	Thrombocytopenia	2
				Leukopenia	1	Leukopenia	1
						Anemia	1
B	6	Anemia	2	Anemia	2	Anemia	2
(n = 6)		Hypoalbuminemia	1	Hypoalbuminemia	1	Hypoalbuminemia	1
				Leukopenia	2	Leukopenia	1
						Thrombocytopenia	4
	7	Thrombocytopenia	1	Thrombocytopenia	1	Thrombocytopenia	2
		Neutropenia	1			Neutropenia	1
						Leukopenia	2
						Anemia	1
						Hypoalbuminemia	1
	8	Anemia	1	Hypoalbuminemia	2
		Hypoalbuminemia	2	Thrombocytopenia	1
	10			Thrombocytopenia	1	Thrombocytopenia	3
C	12			Leukopenia	3
(n = 3)				Neutropenia	4
				Thrombocytopenia	4

Note: Group A: 2.22 GBq; Group B: 3.33 GBq; Group C: 4.99 GBq. No adverse event was observed in Patients 3, 9, and 11; hence, these patients were not included in this table.

Abbreviations: CTCAE: Common Terminology Criteria for Adverse Events; RLT: radioligand therapy.

Post-therapeutic ¹⁷⁷Lu-LNC1004 whole-body scans

¹⁷⁷Lu-LNC1004 showed relatively high uptake in the blood pool 1 hour after administration, as indicated by a strong signal in the heart region and major vessels. Among the normal organs, ¹⁷⁷Lu-LNC1004 showed increased uptake in the liver, spleen, and kidneys, and was excreted to the bladder via the kidneys. The tumor uptake of ¹⁷⁷Lu-LNC1004 was clearly visible at 1 hour after administration. Except for the increased activity in blood circulation, the in vivo biodistribution pattern of the ¹⁷⁷Lu-LNC1004 was identical to that of the pre-treatment ⁶⁸Ga-FAPI PET/CT. In addition, post-treatment whole-body and SPECT/CT scan analysis revealed significant uptake and retention of ¹⁷⁷Lu-LNC1004 in tumor lesions on delayed imaging 7 days after injection in all patients. Representative examples are shown in Fig. 1.

Figure 1.

A representative example of a patient who underwent ¹⁷⁷Lu-LNC1004 therapy. A, A 36-year-old man with progressive metastatic radioiodine-refractory differentiated thyroid cancer began treatment with multi-kinase inhibitors but still experienced disease progression. ⁶⁸Ga-FAPI-46 PET/CT was performed for patient screening before ¹⁷⁷Lu-LNC1004 therapy and revealed intense ⁶⁸Ga-FAPI-46 uptake in metastatic lesions. B, After administrating 3.33 GBq of ¹⁷⁷Lu-LNC1004, intense radiotracer uptake was observed in the metastatic lesions (arrows) on post-therapeutic whole-body scintigraphy (WBS; anterior views) from 1 to 168 hours after injection, which was consistent with ⁶⁸Ga-FAPI-46 uptake presented in the PET/CT.

Dosimetry evaluation of ¹⁷⁷Lu-LNC1004

Dosimetric parameters, specifically percentage uptake (% uptake), effective half-life, residence time, and mean absorbed dose for the total body, blood pool, liver, kidneys, and metastatic sites were determined after ¹⁷⁷Lu-LNC1004 administration. The kinetics of ¹⁷⁷Lu-LNC1004 observed after administration are shown in Fig. 2A. First, the percentage of uptake of the total body over 1 hour was determined to be 100%, and the curve declined at an approximately constant rate. Moreover, the blood pool and liver curves demonstrated an initial rapid washout between the first scan and 24 hours after administration, followed by a slower decline. The initial uptake and washout rates of ¹⁷⁷Lu-LNC1004 in the kidneys decreased slowly. The effective half-lives of ¹⁷⁷Lu-LNC1004 in the total body, blood pool, liver, and kidneys were 90.20 ± 7.68, 74.35 ± 6.28, 82.73 ± 6.64, and 101.00 ± 6.40 hours, respectively (Fig. 2B). The kinetics and effective half-life of ¹⁷⁷Lu-LNC1004 in the salivary glands, pancreas, spleen, and urinary bladder are shown in Supplementary Fig. S3A and S3B.

Figure 2.

Kinetics and dosimetry results of normal organs and metastases and comparative results from metastases (bone, lymph node, and other) in patients treated with ¹⁷⁷Lu-LNC1004: A, Uptakes in the percentage of administered activity (%); B, Effective half-life (h); C, Residence time (h); D, Kinetics of various types of metastases; E, Effective half-life of various types of metastases (h); F, Residence time of various types of metastases (h); and G, Absorbed dose of various types of metastases. les, lesions; LN, lymph node; met, metastases. Other involved recurrences, subcutaneous and visceral metastases.

The residence time represents the total number of disintegrations that occur during an integration time per unit administered activity. The residence time of ¹⁷⁷Lu-LNC1004 in the total body, blood pool, liver, and kidneys was 123.65 ± 13.85, 6.37 ± 2.36, 5.52 ± 1.89, and 2.43 ± 0.98 hours, respectively (Fig. 2C). As a result of calculations from these kinetic parameters, the mean absorbed doses of total body, heart wall, red marrow, liver, and kidneys were 0.17 ± 0.02 mSv/MBq, 0.82 ± 0.38 mSv/MBq, 0.11 ± 0.03 mSv/MBq, 0.29 ± 0.13 mSv/MBq, and 1.32 ± 0.69 mSv/MBq, respectively. The whole-body effective dose of ¹⁷⁷Lu-LNC1004 was 0.17 ± 0.04 mSv/MBq. The residence time and mean absorbed doses of the other organs are shown in Supplementary Fig. S3C and Table 3.

Table 3.

Estimated absorbed dose after intravenous administration of ¹⁷⁷Lu-LNC1004.

Target organ	177Lu-LNC1004 (mSv/MBq)
Adrenals	0.1673 ± 0.0219
Brain	0.1363 ± 0.0166
Breasts	0.1538 ± 0.0185
Esophagus	0.1452 ± 0.0167
Eyes	0.1376 ± 0.0191
Gallbladder wall	0.1498 ± 0.0177
Left colon	0.1474 ± 0.0180
Small intestine	0.1453 ± 0.0179
Stomach wall	0.1483 ± 0.0172
Right colon	0.1448 ± 0.0170
Rectum	0.2011 ± 0.2069
Heart wall	0.8238 ± 0.3847
Kidneys	1.3213 ± 0.6855
Liver	0.2923 ± 0.1323
Lungs	0.2157 ± 0.2570
Pancreas	0.4649 ± 0.5792
Prostate	0.1355 ± 0.0124
Ovaries	0.1590 ± 0.0154
Salivary glands	0.4294 ± 0.8479
Red marrow	0.1136 ± 0.0258
Osteogenic cells	0.2503 ± 0.3021
Spleen	0.9918 ± 0.3781
Testes	0.1299 ± 0.0121
Thymus	0.1464 ± 0.0180
Urinary bladder wall	0.4875 ± 0.1815
Uterus	0.1665 ± 0.0230
Total body	0.1654 ± 0.0223
Effective dose	0.1676 ± 0.0363

For tumor lesions, initial ¹⁷⁷Lu-LNC1004 uptake in all metastases was 0.33% (0.22%–0.70%), and the uptake in bone metastases (0.30%; 0.22%–0.71%) was slightly lower than that of lymph node (0.37%; 0.26%–0.70%) and other (0.37%; 0.16%–1.38%) metastases. The metastases washout rate declined slowly, which was similar to that of the kidneys (Fig. 2D). The effective half-life of ¹⁷⁷Lu-LNC1004 in all metastases was 92.46 ± 9.66 hours, and the effective half-life in bone metastases (98.82 ± 11.89 hours) was slightly longer than that of lymph node (91.77 ± 9.27 hours) and other (87.77 ± 5.12 hours) metastases (Fig. 2E). Although the residence time of ¹⁷⁷Lu-LNC1004 in metastases (0.42 hours; 0.29–0.99 hours) was shorter than that observed of the above organs (Fig. 2F), it exhibited the highest mean absorbed dose (8.50 ± 12.36 Gy/GBq; range, 1.32–58.72 Gy/GBq; Fig. 2G). Specifically, the residence time in bone, lymph node, and other metastases was 0.62 (0.38–2.73 hours), 0.39 (0.25–0.93 hours), and 0.36 hours (0.18–1.17 hours), and the mean absorbed dose was 5.12 ± 2.50 Gy/GBq (range, 2.31–7.62 Gy/GBq), 5.95 ± 6.80 Gy/GBq (range, 1.32–25.03 Gy/GBq), and 15.27 ± 20.52 Gy/GBq (range, 2.10–58.72 Gy/GBq), respectively.

The radiation emission curves are shown in Supplementary Fig. S4. When the patients were at a distance of 1 m, the average exposure of the 3 groups recorded at 4 hours after treatment was 5.95 ± 0.32 μSv/h (Group A), 10.53 ± 2.22 μSv/h (Group B), and 14.77 ± 2.84 μSv/h (Group C). When the patients were at a distance of 3 m, the average exposure of the 3 groups recorded at 4 hours was 1.38 ± 0.16 μSv/h (Group A), 2.02 ± 0.15 μSv/h (Group B), and 2.72 ± 0.28 μSv/h (Group C). The radiation emission curves declined over time, such that the average exposure of the three groups at 1 and 3 m at 48 hours after administration was 4.22 ± 0.42 μSv/h (Group A), 7.00 ± 1.34 μSv/h (Group B), and 9.91 ± 1.42 μSv/h (Group C), and 1.09 ± 0.18 μSv/h (Group A), 1.46 ± 0.19 μSv/h (Group B), and 1.95 ± 0.15 μSv/h (Group C), respectively.

Efficacy after ¹⁷⁷Lu-LNC1004 RLT

Evaluation according to RECIST 1.1 in all patients at 6 weeks after the second RLT cycle revealed partial response in three patients (Patients 4, 5, and 6) and stable disease in the other seven patients, with an ORR and DCR of 25% (3/12) and 83% (10/12), respectively. In Patient 4, the target lesions in the right clavicle, rib, and lymph nodes significantly decreased after two RLT cycles. Notably, obstructive pneumonia caused by the enlarged left hilar lymph node metastases resolved (Fig. 3A). In Patient 5, the target lesion of the liver metastases and a non-target lesion of the pleural fluid resolved, and the sum of the longest diameters (SLD) of the target lesions in the bone and lung metastases declined (Fig. 3B). However, two patients (Patients 1 and 8) exhibited disease progression after 2 RLT cycles, whereby the SLD increased in Patient 8 and new subcutaneous metastases were detected in Patient 1. Details of patient outcomes are summarized in Fig. 4 and Supplementary Table S1. In addition, ⁶⁸Ga-FAPI-46 PET/CT-derived SUV_max and changes in SUV_max before and after ¹⁷⁷Lu-LNC1004 RLT in the target lesions are summarized in Supplementary Fig. S5 and Supplementary Table S2. Interestingly, a significant decrease in ⁶⁸Ga-FAPI-46 uptake in tumor lesions was observed in patients with partial response, whereas an increased uptake was observed in patients with progressive disease.

Figure 3.

Two representative patients with a favorable therapeutic response after 2 cycles of ¹⁷⁷Lu-LNC1004. A, A 36-year-old man with disease progression after TKI treatment received ¹⁷⁷Lu-LNC1004 RLT at 3.3 GBq/cycle. Baseline ⁶⁸Ga-FAPI-46 PET/CT revealed intense ⁶⁸Ga-FAPI-46 uptake in most metastatic lesions, including lymph node, vertebral, and clavicular metastases (left, arrows). Obstructive pneumonia was also observed in the left lung due to the enlarged lymph nodes. After 2 treatment cycles, restaging ⁶⁸Ga-FAPI-46 PET/CT revealed a significant reduction in tumor size and radiotracer uptake in most of the metastatic lesions (right, arrows). Furthermore, the non-target lesion of obstructive pneumonia resolved. B, A 42-year-old man with disease progression after TKI treatment underwent ¹⁷⁷Lu-LNC1004 therapy. Baseline ⁶⁸Ga-FAPI-46 demonstrated intense uptake in multiple bone metastases, left hilar lymph node, and liver metastases (left, arrows). Follow-up ⁶⁸Ga-FAPI-46 PET/CT revealed a significant reduction in tumor size and radiotracer uptake in these metastatic lesions (right, arrows) after 2 cycles of ¹⁷⁷Lu-LNC1004 therapy.

Figure 4.

Best percentage of changes from baseline in the sum of the largest diameter of target lesions (n = 12). ATC, anaplastic thyroid cancer; DTC, differentiated thyroid cancer; MTC, medullary thyroid cancer; PD, progressive disease; PR, partial response; SD, stable disease.

Discussion

This novel prospective study provided the first evidence of the feasibility of ¹⁷⁷Lu-LNC1004 RLT in patients with mRAIR-TC with disease progression after TKI treatment and the first-in-human data demonstrated a favorable safety profile with few serious AEs. The 3+3 dose-escalation study design revealed 3.33 GBq as the optimal dose for future clinical trials.

To date, ¹³¹I has been the most common adjuvant therapy for metastatic DTC; however, patients with advanced DTC have de novo resistance or become resistant to ¹³¹I. Moreover, medullary and anaplastic thyroid carcinomas failed to respond to ¹³¹I therapy because of a lack of iodine metabolism-related gene expression (29). Although TKI treatment for mRAIR-TC has undergone rapid evolution over the last decade, drug resistance remains a major obstacle in improving prognosis. Currently, no other therapy is available for patients with terminal cancer who exhaust or refuse conventional treatment options. Increased FAP expression has shown a positive correlation with dedifferentiation and aggressive outcomes of thyroid cancer (30). In our previous study, approximately 96% of patients with metastatic thyroid cancer showed intense ⁶⁸Ga-FAPI uptake in tumor lesions (25). Moreover, 92% (12/13) of patients with mRAIR-TC screened with ⁶⁸Ga-FAPI-46 PET/CT for this trial met the criteria specified in the protocol for study enrollment (SUV_max ≥10 in over 50% of lesions). These data suggest that FAP-targeted radiotherapy with radiolabeled FAPIs may provide a potential therapeutic approach for mRAIR-TC.

Although previous studies have used FAP-targeted radionuclide therapy with therapeutic-nuclide radiolabeled FAPI-46 in various advanced cancers, the in vivo half-life of radiotherapeutic agents may not be long enough for sufficient tumor uptake and retention (13, 14, 31). Therefore, to optimize the efficacy of FAPI RLT, the biological half-life of radiotherapeutic agents must be adjusted to the therapeutic radionuclides, such as ¹⁷⁷Lu. In our study, the effective half-life of LNC1004 (whole-body) was 90.20 ± 7.68 hours, which was longer than that of ¹⁷⁷Lu-FAP-2286 reported by Baum and colleagues (35.00 ± 9.00 hours; ref. 27). Consequently, whole-body radiation exposure from ¹⁷⁷Lu-LNC1004 reached 0.17 ± 0.04 mSv/MBq, which was 6.54-fold greater than ¹⁷⁷Lu-FAPI-46. Nevertheless, ¹⁷⁷Lu-LNC1004 distribution in the liver and kidneys slightly increased, which was 2.13- and 1.49-fold higher than that of ¹⁷⁷Lu-FAPI-46 (31). Radiation exposure to the kidneys from ¹⁷⁷Lu-LNC1004 was similar to ¹⁷⁷Lu-FAP-2286; however, the absorbed dose to the liver from ¹⁷⁷Lu-LNC1004 was 4.14-fold that of ¹⁷⁷Lu-FAP-2286. Ballal and colleagues (27) assessed ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂, which enhances the tumor-absorbed dose via FAPI dimerization, and found that the whole-body and liver absorbed doses from ¹⁷⁷Lu-LNC1004 were similar to those of ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂, whereas the absorbed dose in the kidneys was lower for ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂ (0.30 ± 0.29 mSv/MBq) than for ¹⁷⁷Lu-LNC1004. Notably, the ¹⁷⁷Lu-LNC1004 absorbed dose in the kidneys when used without renal protection was comparable with that of ¹⁷⁷Lu-PSMA-617 and ¹⁷⁷Lu-DOTATATE when co-administered with amino acids (lysine/arginine) for renal protection (32, 33). Despite the slight increase in radiation exposure of ¹⁷⁷Lu-LNC1004 to the liver and kidneys compared with ¹⁷⁷Lu-FAPI-46, the dose limits for these organs were not exceeded (34). Interestingly, none of the patients experienced hepatotoxicity or nephrotoxicity. These findings indicated that the increased radiation exposure associated with prolonged blood circulation was acceptable.

Moreover, the average radiation exposure of the three groups at 1 and 3 m for 4 hours after treatment in this study was less than that of ¹³¹I therapy and below the Nuclear Regulatory Commission release criteria (35). The radiation emission of patients who received 2.22 GBq and 3.33 GBq of ¹⁷⁷Lu-LNC1004 at 1 m 4 hours after administration was less than 9 μSv/h, below which patients were discharged when administered ¹⁷⁷Lu-PSMA-617 and ¹⁷⁷Lu-DOTATATE (36, 37). Despite the exposure of the 4.99 GBq group at 1 m for 4 hours after treatment being higher than this threshold, 48 hours were sufficient for the radiation emission to decline below 9 μSv/h. Although the time periods may vary, 3 days is generally considered sufficient to resume public activities after ¹⁷⁷Lu-LNC1004 exposure.

When bound to albumin, ¹⁷⁷Lu-LNC1004 showed prolonged blood circulation, which consequently increased energy deposition in the red marrow compared with ¹⁷⁷Lu-FAP-2286 and ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂. Indeed, subacute G3–4 hematologic toxicity was observed in 25% of patients, especially those receiving the higher dose (4.99 GBq/cycle). Therefore, the occurrence rate of hematological toxicity of ¹⁷⁷Lu-LNC1004 was higher than ¹⁷⁷Lu-FAPI-46, ¹⁷⁷Lu-FAP-2286, and ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂. This suggests that 3.33 GBq may be the MTD, whereby increasing the dose beyond this level may bring more risks than benefits. Although Patient 6 in the 3.33 GBq group experienced G4 thrombocytopenia, it may have been related to the close association of this AE with widespread bone metastases, rather than with the dose itself. Nevertheless, the overall rate of G3–4 hematological toxicity was lower than that of ⁹⁰Y-FAPI-46. On the basis of the generally accepted maximum absorbed dose of 2 Gy in bone marrow (34), up to 18 GBq of ¹⁷⁷Lu-LNC1004 (accumulated dose) may be tolerated by one patient. Therefore, in accordance to the results of our study, up to four cycles of ¹⁷⁷Lu-LNC1004 at 3.33Gq/cycle is feasible for RLT. However, given the poor correlation between the mean absorbed dose in the bone marrow and the development of hematologic toxicity, a thorough therapeutic regimen must be planned. In addition, we will conduct long-term safety monitoring according to a pre-specified list of late radiation adverse reactions of special interest for up to 5 years or until death for all the recruited patients. After the completion of our initial dose escalation phase study, we will fix the therapy dose in a following expansion study with a larger cohort/longer follow-up/more diverse patient selection.

Compared with FAP-2286 and DOTAGA.(SA.FAPi)₂, the longer tumor retention of LNC1004 allows for coordination with longer half-life therapeutic radionuclides, including ¹⁷⁷Lu and ²²⁵Ac. Alternatively, ⁹⁰Y has the advantage of high-branching-ratio β⁻ emission, which allows high-dose deposition within defined tumor lesions. Although the β-particle energy of ⁹⁰Y was higher than that of ¹⁷⁷Lu in a previous study, the prolonged blood retention of ¹⁷⁷Lu-LNC1004 led to an increased tumor-absorbed dose, which was 7.08-fold that of ⁹⁰Y-FAPI-46 (1.20 Gy/GBq; ref. 14). The absorbed dose in bone metastases was similar to that of ¹⁷⁷Lu-FAPI-2286, and the mean tumor absorbed dose was similar to that of ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂. Although treatment efficacy was not the main objective of the present study, and despite patients only receiving two treatment cycles with limited follow-up time, we observed radiographic disease control in 83% of the patients, along with signs of tumor response, which was superior to the efficacy of ¹⁷⁷Lu-FAPI-46 (31). Furthermore, a post-RLT flare phenomenon has been previously reported (27); thus, the symptoms observed in Patients 5 and 6 may potentially be related to existing metastases localized in the area of pain exacerbation. In addition, 25% of patients achieved partial response and 58% exhibited stable disease after two treatment cycles according to RECIST 1.1 criteria. In addition, our findings indicated that most PR lesions experience a decrease in ⁶⁸Ga-FAPI uptake after ¹⁷⁷Lu-LNC1004 therapy, which suggests that the changes in ⁶⁸Ga-FAPI uptake before and after therapy could be used for predicting the therapeutic efficacy of ¹⁷⁷Lu-LNC1004. However, further investigation in future clinical trials is required. Overall, the preliminary tumor response to ¹⁷⁷Lu-LNC1004 in this study was superior to that of previous studies using ¹⁷⁷Lu-FAP-2286 in patients with diverse adenocarcinomas and ⁹⁰Y-FAPI-46 in patients with advanced sarcoma (38). Despite two patients exhibiting intense ⁶⁸Ga-FAPI-46 uptake in the metastases at baseline, disease progression still occurred after two treatment cycles. The contradiction between intense tumor uptake and unfavorable therapeutic efficacy may be attributed to mutations in DNA damage repair and checkpoint genes (39). Possible measures to enhance the efficacy of RLT include shortening the interval between treatments or evaluating combinations with other treatments, such as TKI or immune checkpoint inhibitor therapy.

To date, the study regarding the prevalence of FAP expression in different thyroid cancer subtypes is rare. Zhu and colleagues (40) demonstrated that FAP expression correlated with worse clinicopathological features in papillary thyroid cancer (PTC). Our previous study exhibited intense FAP expression in PTC metastatic lesions (25). In the present study, FAP-targeted PET/CT imaging demonstrated intense ⁶⁸Ga-FAPI uptake in DTC, MTC, and ATC lesions (Supplementary Fig. S6); however, the differences in FAP expression between thyroid cancer subtypes remain unknown and studies reporting the association between tumor genomic status and FAP expression are limited. Although one study reported that BRAF mutation was significantly positively correlated with FAP expression in PTCs (40), a similar correlation in MTC, ATC, or other gene mutations (including RAS and RET/PTC) has yet to be found. Further investigating the pathological and genomic landscape of thyroid cancer may help improve the efficacy of RLT.

This study has some limitations. First, the patient population that received RLT with ¹⁷⁷Lu-LNC1004 as the last line of treatment on a compassionate-use basis was small and homogeneous. Second, our study lacked long-term assessment of the side effects; however, such observations will be implemented in subsequent studies. Finally, ¹⁷⁷Lu-FAPI-46 RLT was not used as a control in this study; hence, the data in our study can only be compared with results from the literature.

In conclusion, this study provides the first evidence of the safety, feasibility, and dosimetry of ¹⁷⁷Lu-LNC1004 RLT with escalating doses in patients with mRAIR-TC. Overall, ¹⁷⁷Lu-LNC1004 was well tolerated by all patients with high radiation doses delivered to mRAIR-TC lesions. The preliminary data on tumor responses are encouraging and support the potential therapeutic role of ¹⁷⁷Lu-LNC1004 RLT in treating mRAIR-TC. Therefore, prospective randomized controlled multicenter clinical trials are warranted in the future.

Supplementary Material

Supplementary Data 1

Figure S1-S6, Table S1-S3

This article is featured in Selected Articles from This Issue, p. 4699

Authors' Disclosures

No disclosures were reported.

Authors' Contributions

H. Fu: Data curation, software, formal analysis, investigation, methodology, writing–original draft, project administration, writing–review and editing. J. Huang: Resources, supervision, project administration. T. Zhao: Software, formal analysis, methodology. H. Wang: Investigation, visualization, project administration. Y. Chen: Visualization, project administration. W. Xu: Validation, project administration. Y. Pang: Visualization, project administration. W. Guo: Visualization, project administration. L. Sun: Supervision. H. Wu: Supervision, writing–review and editing. P. Xu: Visualization, project administration. B. Su: Project administration. J. Zhang: Conceptualization, supervision, methodology, writing–review and editing. X. Chen: Conceptualization, supervision, funding acquisition, methodology, writing–review and editing. H. Chen: Conceptualization, resources, data curation, supervision, funding acquisition, validation, methodology, project administration, writing–review and editing.