¹⁷⁷Lu-LNC1004 Radioligand Therapy in Patients with End-stage Metastatic Cancers: A Single-Center, Single-Arm, Phase II Study

Cancer-associated fibroblasts (CAF) are the most prevalent stromal cells in the tumor microenvironment (TME) and constitute a substantial portion of the cellular composition of tumor tissues (1). Fibroblast activation protein (FAP), a member of the dipeptidyl peptidase protein family, is highly expressed on CAFs across >90% of epithelial tumors (2, 3). FAP-expressing CAFs promote tumor development and metastasis by influencing extracellular matrix remodeling, intracellular signaling, angiogenesis, epithelial-to-mesenchymal transition, and immunosuppression (4). Given its minimal expression in normal tissues, FAP has been recognized as a pan-tumor target (5).

Although significant progress has been made in improving the outcomes of patients with cancer through various therapeutic strategies, several challenges persist, particularly because of cancer therapy resistance, which often culminates in relapse, metastasis, or ultimately cancer mortality. Therefore, effective novel therapies for metastatic cancers are urgently required. TME components such as CAFs and the extracellular matrix are major contributors to therapy resistance (6). FAP-targeted radioligand therapy (RLT) targets both FAP-positive CAFs and neighboring cancer and stromal cells within the TME through crossfire effects (7). Recently, various FAP-targeted radioligands, including FAP inhibitors (FAPI) and FAP-binding peptides, have been developed to improve tumor imaging and FAP-targeted RLT (8–11). However, rapid clearance from tumors limits their therapeutic potential, particularly when labeled with radionuclides with long half-lives, such as lutetium-177 (¹⁷⁷Lu) and yttrium-90 (⁹⁰Y; refs. 12, 13).

Serum albumin has emerged as a versatile carrier of therapeutic agents because of its abundance in plasma and reversible binding to bioactive drugs, which allows the albumin–drug complex to function as a reservoir, enhancing drug distribution and bioavailability (14). Evans blue (EB), a promising albumin-binding moiety with relatively high affinity for binding site 1 on serum albumin, prolongs blood circulation and improves therapeutic efficacy when conjugated with targeting molecules (15). In our previous phase I study in patients with metastatic radioiodine-refractory thyroid cancer, a ¹⁷⁷Lu-EB–FAPI (¹⁷⁷Lu-LNC1004), a FAP-targeted radiopharmaceutical conjugated to EB, demonstrated encouraging therapeutic efficacy with acceptable side effects at a dose of 3.33 GBq/cycle (16). These preliminary results prompted investigation of the potential role of ¹⁷⁷Lu-LNC1004 RLT in treating other types of solid tumors. Therefore, this study explored the initial clinical experience with ¹⁷⁷Lu-LNC1004 RLT at a fixed dose of 3.33 GBq/cycle in patients with end-stage metastatic cancers and investigated the efficacy and safety of ¹⁷⁷Lu-LNC1004 RLT in patients with various solid tumors who had progressed despite standard therapies.

This prospective, single-center, single-arm, open-label, nonrandomized, investigator-initiated phase II trial was conducted at The First Affiliated Hospital of Xiamen University, China. The study protocol was approved by the Institutional Review Board and performed in accordance with the Declaration of Helsinki. The study was registered at ClinicalTrials.gov (registration number: NCT05963386). A multidisciplinary tumor board affiliated with the hospital approved the use of ¹⁷⁷Lu-LNC1004. Written informed consent was obtained from all patients.

All patients underwent a ⁶⁸Ga-FAPI-46 PET/CT screening assessment to confirm high FAP expression, which was conducted following established practice guidelines (17). Briefly, the dose of intravenously injected ⁶⁸Ga-FAPI-46 was calculated based on patient weight [1.8–2.2 MBq (0.05–0.06 mCi)/kg]. One hour after administration, patients underwent noncontrast agent-enhanced PET/CT using a hybrid PET/CT scanner (Discovery MI; GE Healthcare; RRID: SCR_025461). All acquired data were transferred to the Advantage Workstation (version AW 4.7; GE Healthcare; RRID: SCR_025461) and reconstructed using the Bayesian penalized likelihood reconstruction algorithm (Q.Clear; GE Healthcare; RRID: SCR_025461). Further details are provided in the Supplementary Materials. The inclusion criteria were (i) progressive metastatic cancers in patients ages >18 years; (ii) unresectable tumors; (iii) exhaustion of all approved therapies, as determined by the multidisciplinary tumor board; and (iv) tumor lesions with increased radiotracer uptake on ⁶⁸Ga-FAPI-46 PET/CT [defined as a maximum standardized uptake value (SUV_max) ≥10 in >50% of metastatic lesions]. The exclusion criteria were (i) serum creatinine levels >150 μmol/L; (ii) hemoglobin levels <8.0 g/dL; (iii) white blood cell count <2.0 × 10⁹/L; (iv) platelet count <50 × 10⁹/L; (v) total bilirubin levels >3 times the upper normal limit and serum albumin levels <2.0 g/dL; (vi) cardiac insufficiency, including carcinoid heart valve disease, severe allergy, or hypersensitivity to radiographic contrast materials; (vii) claustrophobia; and (viii) pregnancy or breastfeeding.

¹⁷⁷Lu labeling of LNC1004 was performed as previously reported (16). Briefly, the precursor LNC1004 [1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-EB–FAPI] was incubated with the required radioactivity of ¹⁷⁷LuCl₃ (Isotope Technologies Munich) in 0.5 mol/L sodium acetate buffer (adjusted to pH 5.5 with hydrochloric acid) at 95°C for 30 minutes, followed by dilution with 0.9% saline solution and sterile filtration. The samples were tested for quality, endotoxins, and sterility. Radiochemical purity, determined using thin-layer and high-performance liquid chromatography, consistently exceeded 98%.

No specific preparations, fasting, or dietary restriction were required on the day of ¹⁷⁷Lu-LNC1004 administration. Patients received 4 mg ondansetron before treatment to prevent nausea and vomiting. ¹⁷⁷Lu-LNC1004 was diluted in 100 mL of 0.9% saline and administered slowly via intravenous infusion over 20 to 30 minutes. Symptoms and vital parameters were monitored before, during, and after treatment. The treatment regimen was scheduled to consist of four cycles, with a 6-week interval between each cycle. Treatment was discontinued if patients experienced ≥ G3 adverse events (AE) persisting >4 weeks without remission or rapid disease progression that reduced tolerance for further RLT.

¹⁷⁷Lu-LNC1004 kinetics were determined following previously established methods (16). Planar whole-body scintigraphy in the anterior and posterior projections were acquired 1, 4, 24, 48, 72, 96, and 168 hours after ¹⁷⁷Lu-LNC1004 infusion 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 and 208 keV photopeaks. Whole-body single-photon emission computed tomography (SPECT)/CT was also performed 72 hours after administration. CT was performed without a contrast agent for attenuation correction (tube voltage, 130 kVp; tube current-time product, 17 mAs; beam pitch, 1.5; section width, 5 mm). SPECT scans were acquired at 128 angles over 360° and 25 seconds per stop. The 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 consisted of 128 × 128 pixels, with a cuboid voxel length of 4.8 mm. The tumor-absorbed dose was estimated during the first treatment cycle for each patient according to previous reports (10, 16, 18).

All patients underwent routine complete blood count and liver and kidney function assessment before and during posttreatment follow-up every 2 weeks. In the event of significant toxicity, blood tests were repeated weekly until resolution. Patient records were reviewed for any hematologic, gastrointestinal, renal, hepatic, or other AEs, graded according to Common Terminology Criteria for Adverse Events version 5.0. All circumstances that resulted in treatment cessation or delay were documented. The investigators assessed potential relationships between AEs and either disease progression or RLT.

According to the RECIST version 1.1, tumor response was monitored using functional imaging modalities, including ⁶⁸Ga-FAPI-46 PET/CT and contrast-enhanced CT/MRI. Assessments with ⁶⁸Ga-FAPI-46 PET/CT were performed 6 weeks after each cycle, whereas evaluations using contrast-enhanced CT/MRI were performed at a 12-week interval. The primary endpoint was the post-RLT RECIST radiologic response, as assessed by the study investigators without central review, and the post-RLT RECIST objective response was defined by response category between baseline and restaging after RLT. Post-RLT disease control was defined as either complete response, partial response (PR), or stable disease (SD) according to RECIST 1.1. The disease control rate (DCR) was calculated as the proportion of patients achieving post-RLT disease control. The best RECIST objective response to RLT was recorded as the most favorable response category observed between baseline and any interim timepoint during RLT, including post-RLT restaging. The secondary endpoints were progression-free survival (PFS) and overall survival (OS). PFS was recorded from RLT initiation until either RECIST progression, death, or last follow-up. OS was recorded from RLT initiation until death or last follow-up as of August 2024. Dosimetric evaluation of serial quantitative SPECT/CT images was performed as a secondary objective to estimate the absorbed dose of ¹⁷⁷Lu-LNC1004 in tumor lesions.

All statistical analyses were performed using SPSS (version 29.0; IBM Corp.; RRID: SCR_002865) and GraphPad Prism (version 10.2; GraphPad Software; RRID: SCR_002798). Normally distributed quantitative data are expressed as the mean ± SD, whereas nonnormally distributed data are expressed as the median with an IQR. For categorical variables, the χ² test was used to compare numbers and percentages. Multivariate logistic regression was used to identify associations between response and clinicopathologic parameters. Pre- and post-RLT SUV_max and percentage changes in SUV_max among the PR, SD, and progressive disease (PD) groups were analyzed using the paired sample t test and one-way ANOVA. Kaplan–Meier plots were generated for PFS and OS. A log-rank test was used to test for the difference in survival rates between the groups. Cox proportional hazards modeling was used to estimate the risk of disease persistence/progression and death from any cause after adjusting for age and all other factors that had been proven to be associated with PFS and OS in multivariate analyses. All tests were two tailed, and P values < 0.05 were considered statistically significant.

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

A flowchart diagram and patient characteristics are shown in Fig. 1, Table 1, and Supplementary Table S1. Among the 50 patients initially screened between June 2022 and April 2024, 9 patients did not meet the eligibility criteria for FAP imaging and 13 patients did not undergo ¹⁷⁷Lu-LNC1004 treatment because of rapid disease deterioration, a change in management, or other reasons. Overall, 28 (56%) patients were treated, receiving a total of 63 RLT cycles. All 28 patients completed one ¹⁷⁷Lu-LNC1004 RLT cycle. At the time of analysis, one patient was still undergoing active RLT. Among the 28 patients, 19 (68%), 10 (36%), and 6 (21%) received two, three, and four ¹⁷⁷Lu-LNC1004 RLT cycles, respectively. The reasons for discontinuation included tumor progression (4/28, 14%), rapid disease deterioration or death (10/28, 36%), thrombocytopenia (4/28, 14%), and other therapeutic regimen changes (3/28, 11%).

Among the 28 patients, 15 (54%) were female and 13 (45%) had metastatic radioiodine-refractory thyroid cancer. The other tumor entities were breast cancer (3/28, 10%), neuroendocrine carcinoma (2/28, 7%), sarcoma (2/28, 7%), non–small cell lung cancer (2/28, 7%), gastric cancer (1/28, 4%), colon cancer (1/28, 4%), nasopharyngeal cancer (1/28, 4%), renal cancer (1/28, 4%), esophageal cancer (1/28, 4%), and neuroendocrine prostate cancer (1/28, 4%). All patients had advanced-stage disease with involvement of multiple organs; metastases were most commonly in the lymph nodes, lungs, and bone. Most patients underwent a median of two previous systemic treatment lines before ¹⁷⁷Lu-LNC1004 RLT and experienced disease progression during their last regimen. Among the 28 patients, 24 (86%) had an Eastern Cooperative Oncology Group (ECOG) performance status of ≥2. The median cumulative activity of ¹⁷⁷Lu-LNC1004 was 7.0 GBq, with a median administered radioactivity of 3.1 GBq/cycle and a median RLT duration of 3 months.

¹⁷⁷Lu-LNC1004 exhibited relatively high uptake in the blood within 1 hour of administration, as indicated by an intense signal in the heart and major blood vessels. Tumor uptake of ¹⁷⁷Lu-LNC1004 was visible 4 hours after administration. SPECT/PET imaging revealed the in vivo biodistribution pattern of ¹⁷⁷Lu-LNC1004 at 4 hours after administration identical to that of ⁶⁸Ga-FAPI-46 at 1 hour after administration. Serial posttherapeutic total-body and SPECT/CT scan analyses revealed significant tumor retention of ¹⁷⁷Lu-LNC1004 at 1, 2, 3, 4, and 7 days after injection in all patients. This remained intense 2 weeks after administration, indicating prolonged tumor retention. Representative examples are shown in Fig. 2A and B. Dosimetry analysis was conducted in 25 of 28 (89%) patients, with results summarized in Supplementary Table S2. The mean absorbed dose for tumor lesions was 4.69 ± 3.83 Gy/GBq, with bone, lymph node, lung, and other metastases at 4.57 ± 1.98 Gy/GBq, 4.88 ± 4.39 Gy/GBq, 6.47 to 6.75 Gy/GBq, and 3.77 ± 1.74 Gy/GBq, respectively (Fig. 2C).

¹⁷⁷Lu-LNC1004 RLT was generally well tolerated, with no immediate AEs recorded during injection and no treatment-related deaths (Table 2; Supplementary Table S3). Hematotoxicity was the most common AE reported during treatment, which mainly included anemia, thrombocytopenia, and leukopenia. Among all recorded AEs (n = 121), 88 (73%) were identified through laboratory examination, whereas 33 (27%) were based on clinical symptoms. Among the 28 patients, 19 (68%) experienced grade 3/4 AEs. Grade 3/4 hematotoxicity was observed in 12 (43%) patients and included anemia (6/28, 21%), thrombocytopenia (8/28, 29%), leukopenia (2/28, 7%), and neutropenia (2/28, 7%). Grade 3/4 general disorders were recorded in 13 (46%) patients and included dysphagia (3/28, 11%), tumor pain (11/28, 39%), and dyspnea (1/28, 4%). No case of grade 3/4 hepatotoxicity or nephrotoxicity was observed. The investigators attributed grade 3/4 hematotoxicity to RLT in 6 (21%) patients. One male patient experienced grade 3 thrombocytopenia after one RLT cycle, which led to RLT discontinuation and subsequent recovery of thrombocyte count after 4 weeks of treatment. Another male patient with extensive bone metastases experienced grade 3 thrombocytopenia after one RLT cycle and recovered after 3 weeks of treatment, followed by a change in therapeutic regimens. A female patient with widespread osseous lesions developed grade 4 thrombocytopenia after two RLT cycles, which led to RLT discontinuation, with subsequent recovery of thrombocyte count after 4 weeks of treatment. Grade 3 thrombocytopenia was also observed in a male patient after two RLT cycles. However, this patient died of complications arising from coronavirus disease 2019–associated acute respiratory distress syndrome. Additionally, leukopenia and neutropenia were recorded in one male patient (grade 3 after three RLT cycles) and one female patient (grade 3 after two RLT cycles). Hematotoxicity in the male patient was reversible after 2 weeks of treatment and was not observed in subsequent RLT. Despite subsequent posttreatment recovery of leukocyte and neutrophil counts, disease progression led to RLT discontinuation in the female patient. Finally, ¹⁷⁷Lu-LNC1004–related grade 3 bone pain flare was noted in two (7%) patients with osseous metastases, which increased 1 and 5 days after treatment, persisted for 2 days, and subsequently improved with NSAIDs and oxycodone.

Tumor responses according to the RECIST are summarized in Supplementary Table S4. Among the 28 patients, 8 (29%) were not evaluated using RECIST, as one patient had not yet reached the scheduled point for restaging at the time of analysis and seven patients died before radiologic response evaluation. Among the 20 remaining patients, 4 (20%) achieved PR and 9 (45%) demonstrated SD (Fig. 3), with an objective response rate of 20% and a DCR of 65%. Representative cases are outlined in Fig. 4. Among the 28 patients, 5 (18%) did not achieve a best RECIST response under RLT, as they died before the first imaging timepoint. Among the 23 remaining patients, 6 (26%) achieved PR and 13 (57%) demonstrated SD (Supplementary Table S4). The ⁶⁸Ga-FAPI-46 PET/CT–derived SUV_max and percentage change in SUV_max (before and after ¹⁷⁷Lu-LNC1004 RLT) in the target lesions are presented in Supplementary Table S5. Notably, a significant reduction in ⁶⁸Ga-FAPI-46 tumor uptake was observed in patients with PR, SD, and PD. However, no significant correlation was found between the percentage change in SUV_max and the RECIST responses (P = 0.352).

Univariate analyses of the association between the clinicopathologic and RECIST responses are presented in Supplementary Table S6. No significant differences were observed between the DCR and non-DCR groups in terms of sex, age, ECOG performance status, time from initial diagnosis, previous lines of therapy, organ involvement, or bone metastases. Nevertheless, patients who received >2 RLT cycles, with hemoglobin level >117 g/L, leukocyte count ≤10 × 10⁹/L, and thrombocyte count ≤239 × 10⁹/L at baseline tended to demonstrate a more favorable DCR than other patients, although no significant difference was observed. Notably, evaluable patients with an SUV_max ≤15 (65.0%) in targeted lesions had a better DCR than that of patients with an SUV_max >15 (P = 0.022). Multivariate analyses revealed that the mean SUV_max of RECIST-targeted lesions was independently associated with a better DCR (P = 0.021).

The median follow-up period was 18.1 months (0.9–24.6 months). Among the 28 patients, 16 (57%) were followed for >6 months after initiating RLT. The median PFS was 4.0 (95% confidence interval, 0.8–7.2) months (Fig. 5A). During the follow-up period, 19 of 28 (68%) patients died, and the median OS was 6.3 (95% confidence interval, 0.8–11.7) months (Fig. 5B). Upon response category stratification, the median PFS for patients with disease control (PR and SD) was significantly longer at 16.3 months (log-rank P < 0.001) than that of PD (1.9 months) and non-evaluable patients (1.7 months; Fig. 5C). Similarly, the median OS for patients achieving disease control (PR and SD) was significantly longer at 18.4 months (log-rank P < 0.001) than that of PD (6.1 months) and non-evaluable patients (1.7 months; Fig. 5D).

No significant differences in PFS or OS were observed in terms of sex, time from initial diagnosis, previous lines of local therapy, or baseline thrombocyte count (Supplementary Table S7). Moreover, no significant difference in PFS or OS was observed between patients with and without bone metastasis. However, PFS and OS were significantly associated with age (P = 0.041 and 0.005, respectively) and baseline ECOG performance status (P = 0.026 and 0.009, respectively). Patients who received ≤2 lines of systemic therapy before ¹⁷⁷Lu-LNC1004 had significantly better PFS and OS than those of patients who received >2 lines of prior treatments (P = 0.014 and 0.006, respectively). Additionally, compared with other patients, those with baseline leukocyte counts ≤10 × 10⁹/L had more favorable PFS and OS (P < 0.001 and 0.004, respectively) and those with hemoglobin levels >117 g/L had more favorable PFS (P = 0.028) but not OS (P = 0.113). Patients with ≤3 organs of tumor involvement demonstrated superior OS (P = 0.015), although no significant difference in PFS was observed (P = 0.195). Patients with a mean SUV_max ≤15 of RECIST-targeted lesions also exhibited better PFS (P = 0.038), although no significant difference in OS was observed. Notably, patients who received >2 cycles of ¹⁷⁷Lu-LNC1004 RLT had better prognoses than those of patients who received ≤2 cycles (median PFS, 16.3 vs. 2.0 months, P < 0.001; median OS, 18.4 vs. 3.1 months, P = 0.010, respectively).

Multivariate analysis revealed ≤2 lines of previous systemic therapy, ≤3 organ involvement, baseline hemoglobin >117 g/L, and baseline leukocyte count ≤10 × 10⁹/L as independent prognostic factors of superior PFS (P = 0.005, 0.003, 0.004, and <0.001, respectively) and OS (P = 0.002, 0.002, 0.019, and 0.005, respectively; Supplementary Fig. S1). Meanwhile, an SUV_max of targeted lesions ≤15 and >2 ¹⁷⁷Lu-LNC1004 RLT cycles were only revealed as independent prognostic factors for superior PFS (P = 0.004 and 0.033, respectively). Sex, age, and ECOG performance status were not identified as independent prognostic factors for PFS or OS (all P > 0.05).

This prospective phase II study expands beyond our previous investigation of metastatic radioiodine-refractory thyroid cancer, incorporating an extended cohort of various solid tumors with a longer follow-up period to evaluate objective tumor response, safety, and dosimetry after multiple ¹⁷⁷Lu-LNC1004 RLT cycles. We demonstrated the promising antitumor efficacy and favorable safety profile of RLT with ¹⁷⁷Lu-LNC1004 in patients with various end-stage solid tumors who had progressed after standard treatment. Post-RLT RECIST evaluation showed a PR and SD in 20% and 45% of patients, respectively, with an objective response rate of 20% and a DCR of 65%. Additionally, >50% of patients were followed up for >6 months after RLT, with a median PFS of 4.0 months and OS of 6.3 months.

The primary goal of therapy in patients with progressive and extensive metastatic disease is to achieve tumor control and prevent further deterioration or death, particularly after exhausting other available therapeutic options. Previous studies have demonstrated the feasibility of FAP-targeted RLT for various advanced tumors using radiopharmaceuticals based on small FAPIs, including ¹⁷⁷Lu- or ⁹⁰Y-labeled FAPI-04/46 (19, 20). However, the therapeutic efficacy of FAPI-04/46 is limited by its relatively short tumor retention and rapid washout, even when used with radionuclides with long half-lives. Subsequent research has focused on modifying the chemical structure of current FAPI molecules to prolong tumor retention (21, 22). Ballal and colleagues (23) assessed the dimeric FAPI molecule, ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂ (¹⁷⁷Lu labeled dotaga anhydride homodimeric squaramide.FAPi), in advanced radioiodine refractory–differentiated thyroid cancer, which enhances the tumor-absorbed dose via dimerization. Although the dosimetry analysis revealed tumor-absorbed doses of up to 10.8 Gy/GBq, efficacy was only assessed using biochemical response and clinical response (visual analog score and global pain assessment), lacking RECIST response to objectively measure changes in tumor size (23). Baum and colleagues (10) evaluated the feasibility of ¹⁷⁷Lu-radiolabeled FAP–targeted cyclic peptide FAP-2286 in diverse adenocarcinomas. The overall dose delivered by ¹⁷⁷Lu-FAP-2286 to tumor lesions (0.4–10.6 Gy/GBq) was comparable to that of other FDA-approved radiopharmaceuticals [including ¹⁷⁷Lu-DOTATATE (tetraazacyclododecanetetraacetic acid–DPhe1-Tyr3-octreotate) and ¹⁷⁷Lu-PSMA-617 (PSMA, prostate-specific membrane antigen); refs. 24, 25], with 2 of 11 patients demonstrating SD after the first treatment cycle. Recently, eight patients with advanced metastatic sarcoma underwent multiple cycles of ¹⁷⁷Lu-FAP-2286 RLT, and 80% (4/5) of patients showed PR according to RECIST (26). The prolonged tumor retention of ¹⁷⁷Lu-LNC1004 allows it to work effectively with therapeutic radionuclides with longer half-lives, which is critical for achieving high therapeutic efficacy and a robust safety profile (16). Upon biodistribution and dosimetric analysis, we found that ¹⁷⁷Lu-LNC1004 exhibited superior tumor retention than that of ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂ and ¹⁷⁷Lu-FAP-2286, with comparable tumor-absorbed doses (10, 16, 23). Moreover, the DCR of ¹⁷⁷Lu-LNC1004 RLT in the present study (65%) was equivalent to that of ¹⁷⁷Lu-FAP-2286 in patients with sarcoma and diverse adenocarcinomas (18%–80%; refs. 10, 26), as well as ⁹⁰Y-FAPI-46 in patients with sarcoma, solitary fibrous tumor, and pancreatic cancer (50%–82%; refs. 11, 12, 27).

In radiotheranostics, quantitative PET/CT with an imaging radiotracer can identify the abundance of a specific molecular target in each cancer lesion, hence determining suitable patients for companion therapeutic radiopharmaceuticals. In neuroendocrine tumors, higher ⁶⁸Ga-DOTATATE uptake in tumors (measured using the SUV_max) has been associated with better therapeutic response to peptide receptor radionuclide treatment and better prognoses (28, 29). In the current study, the SUV_max of targeted lesions derived from baseline ⁶⁸Ga-FAPI-46 PET/CT significantly influenced treatment response. However, unlike peptide receptor radionuclide treatment studies, a lower SUV_max of targeted lesions was identified as an independent factor for predicting a better DCR (29). This aligns with previous response prediction findings using FAPI-PET in patients with esophageal squamous cell carcinoma receiving chemoradiotherapy (30) and may be attributable to the radiation resistance caused by the increased FAP expression in the TME (31). Furthermore, tumors with high FAP expression are likely to demonstrate more aggressive, rapid disease development. Therefore, the therapeutic response to ¹⁷⁷Lu-LNC1004 may not necessarily be related to the FAPI SUV_max in tumor lesions (32). In our previous study, we observed a significant reduction in ⁶⁸Ga-FAPI-46 uptake in tumor lesions after two ¹⁷⁷Lu-LNC1004 RLT cycles in patients with PR and increased uptake in patients with PD (16). This phenomenon was also observed in patients with gastric adenocarcinoma receiving chemotherapy (33). However, the current study demonstrated a reduction in ⁶⁸Ga-FAPI-46 uptake in tumor lesions in all patients, regardless of tumor response, consistent with a recent ¹⁷⁷Lu-FAP-2286 peptide-targeted radionuclide therapy (PTRT) study involving patients with sarcoma (26). Furthermore, no significant difference in the percentage change in the SUV_max between patients with PR and PD was observed (58.3% vs. 39.8%), although patients with PR demonstrated a higher percentage decrease than that of patients with PD, similar to that reported by Miao and colleagues (34). However, increased FAPI uptake in inflammatory and fibrotic tissues early after radiotherapy may represent a potential interpretation drawback (35, 36). Thus, further studies evaluating the clinical application of FAPI-PET for assessing residual cancer are warranted.

A previous study revealed that a fixed dose of 3.33 GBq/cycle of ¹⁷⁷Lu-LNC1004 is well tolerated with acceptable side effects (16). Hence, we employed repeated ¹⁷⁷Lu-LNC1004 RLT at 3.33 GBq/cycle to achieve the maximum treatment efficacy while ensuring patient safety. As patients had exhausted all available on-label or evidence-based treatment options, and the most prevalent ECOG score was ≥3, ¹⁷⁷Lu-LNC1004 was offered under compassionate use to achieve antitumor effects with manageable toxicity. Nevertheless, advanced tumor stage, extensive pretreatment, and RLT-related side effects likely contributed to a high rate of treatment discontinuation, which may have compromised therapeutic efficacy results. Preventive interventions to improve quality of life and mitigate treatment-related AEs are essential for patients with end-stage metastatic cancer, many of whom already suffer from disease-related symptoms. Acute toxicities or immediate (e.g., allergic) reactions to RLT were not observed. However, grade 3/4 hematotoxicities related to ¹⁷⁷Lu-LNC1004 were noted in 21% of patients during follow-up. Prolonged blood circulation of ¹⁷⁷Lu-LNC1004 when bound to serum albumin increased radiation deposition in the red bone marrow when compared with ¹⁷⁷Lu-FAP-2286 and ¹⁷⁷Lu-DOTAGA.(SA.FAPi)₂ (16). However, >50% of patients had preexisting hematotoxicity because of limited hematologic reserves in this tumor stage, induced by prior chemotherapy or bone marrow infiltration. Therefore, severe hematotoxicity may be attributable to end-stage cancer or tumor progression rather than to ¹⁷⁷Lu-LNC1004 RLT. Previous ¹⁷⁷Lu-PSMA-617 or ¹⁷⁷Lu-DOTATATE studies also identified hematotoxicity, especially thrombocytopenia, as a relevant side effect, with the frequency of grade 3/4 thrombocytopenia varying from 7.9% to 13% and 2% to 16.6%, respectively (37–40). Additionally, RLT-related grade 3/4 thrombocytopenia was observed in 19% of the patients who received multiple treatment cycles (11). In the present study, the hematotoxicity profile of ¹⁷⁷Lu-LNC1004 was comparable to that of previous cohorts, with only 7% of patients experiencing AEs that led to RLT discontinuation, a rate comparable to that for ¹⁷⁷Lu-PSMA-617 but lower than that for ¹⁷⁷Lu-DOTATATE and ⁹⁰Y-FAPI-46 (11, 37, 39). Furthermore, previous studies found that liver and kidney radiation doses did not exceed the critical range, and no patient experienced hepatotoxicity or nephrotoxicity (16). Therefore, repeated ¹⁷⁷Lu-LNC1004 RLT cycles seem feasible and safe.

Studies examining the survival of patients with end-stage malignant tumors treated with ¹⁷⁷Lu- or ⁹⁰Y-labeled FAP–targeted ligands remain limited. In patients with relapsed or refractory tumors treated with ¹⁷⁷Lu-FAPI-46, the median PFS and OS were 3.0 and 4.0 months, respectively (20). Additionally, the OS of patients who received ¹⁷⁷Lu-FAP-2286 RLT ranged from 4.4 to 7.8 months (10, 26). Although Yadav and colleagues (41) reported a clinical DCR reaching 95% in patients with end-stage breast cancer treated with the ¹⁷⁷Lu-DOTAGA-FAPI dimer, the median PFS and OS were 8.5 and 12 months, respectively. Unlike malignant tumors of epithelial origin, intense FAP expression is mostly observed in sarcoma cells, which may explain their theoretically favorable response to FAP-targeted RLT. In sarcoma and other cancer entities, the DCR of patients treated with ⁹⁰Y-FAPI-46 ranged from 38% to 82%, the PFS varied from 3.4 to 7.6 months, and OS was 10.0 months (11, 27). In the present study, patients had end-stage tumors that continued to progress despite multiple lines of treatment and presented with poor status. Patients with an ECOG status >2 had a median PFS and OS of only 3.0 months, mainly because of their high tumor burden and poor tolerance to RLT, whereas those with an ECOG status ≤2 were associated with superior PFS and OS. Moreover, tumors involving >3 organs were associated with inferior OS, whereas those involving ≤3 organs were independent factors associated with superior PFS and OS. Furthermore, fewer lines of systemic therapy prior to RLT were identified as an independent factor associated with better PFS and OS. This may stem from tumor heterogeneity, in which differentiated tumor lesions respond to therapy, whereas more aggressive tumor lesions can selectively survive because of therapeutic resistance (42). These findings indicate that early ¹⁷⁷Lu-LNC1004 RLT should be considered rather than waiting for further patient deterioration and warrant further verification in a clinical trial.

In this study, patients with higher tumor uptake of ⁶⁸Ga-FAPI-46 at baseline demonstrated poorer PFS than that of patients with lower ⁶⁸Ga-FAPI-46 uptake, aligning with previous reports linking high FAP expression in cancerous lesions with relatively poor survival (27, 43). Additionally, patients who completed the full RLT regimen (4 cycles) had better PFS and OS than those who completed 1 to 2 cycles, and >2 cycles of ¹⁷⁷Lu-LNC1004 RLT was an independent factor for better PFS. Achieving a high accumulated radiation dose in cancerous lesions is known to be crucial in RLT. Therefore, discontinuing RLT prematurely may compromise the therapeutic efficacy of ¹⁷⁷Lu-LNC1004. In patients unable to tolerate all RLT cycles, incorporating additional therapeutic modalities may prove valuable.

Notably, we observed a link between anemia and poor PFS, with baseline hemoglobin levels >117 g/L identified as an independent predictor of better PFS and OS. Anemia is a major cause of tumor hypoxia that influences the radiosensitivity of tumor cells, and pretreatment anemia has been well established as a prognostic factor in patients with head and neck cancer and non–small cell lung cancer undergoing radiotherapy (44–46). Similarly, patients with higher leukocyte counts at baseline displayed inferior survival, consistent with findings by Schernberg and colleagues (47). Leukocytosis (defined as a leukocyte count >10 × 10⁹/L) is associated with a poor prognosis in various cancers (48, 49). Increased biological inflammation before chemoradiotherapy can lead to immunosuppression, thereby creating conditions that favor tumor development and ultimately reduce survival (50). These findings provide new perspectives for investigating the combination of RLT with immune checkpoint blockade or radiosensitizing immunotherapy as a definitive treatment modality with curative intent.

This study had some limitations. First, only a small patient population received ¹⁷⁷Lu-LNC1004 RLT on a compassionate-use basis and lacked a control group. Second, metabolic responses measured using PET Response Criteria in Solid Tumors were unavailable because of the lack of post-RLT ¹⁸F-FDG ([¹⁸F]2-fluoro-2-deoxy-D-glucose) PET/CT imaging. Finally, the study lacked long-term assessment of RLT-related side effects, though long-term safety monitoring is scheduled based on a predefined list of late radiation AEs for up to 5 years or until death for all patients.

In conclusion, FAP-targeted RLT using ¹⁷⁷Lu-LNC1004 at a fixed dose of 3.33 GBq/cycle was well tolerated with a low toxicity profile. RECIST disease control was achieved in nearly half of patients with various end-stage metastatic tumors following ¹⁷⁷Lu-LNC1004 RLT and positively associated with improved PFS and OS. These findings highlight the potential of FAP-targeted RLT for treating end-stage metastatic cancer and underscore the need for further improvements and prospective assessment in a larger, more diverse patient cohort.

X. Chen reports being a co-founder of and holding shares in Yantai Lannacheng Biotechnology Co., Ltd. No disclosures were reported by the other authors.

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

This work was funded by the National Natural Science Foundation of China (82102094 and 82422039), Fujian Research and Training Grants for Young and Middle-aged Leaders in Healthcare, Key Scientific Research Program for Young Scholars in Fujian (2021ZQNZD016), Fujian Natural Science Foundation for Distinguished Young Scholars (2022D005), Fujian Natural Science Foundation for Youth Innovation (2022J05314), Academician Workstation Program of the First Affiliated Hospital of Xiamen University (XDFY-AW-2406-004), National University of Singapore (NUHSRO/2020/133/Startup/08, NUHSRO/2023/008/NUSMed/TCE/LOA, NUHSRO/2021/034/TRP/09/Nanomedicine, NUHSRO/2021/044/Kickstart/09/LOA, and 23-0173-A0001), National Medical Research Council (MOH-001388-00, CG21APR1005, MOH-001500-00, and MOH-001609-00), Singapore Ministry of Education (MOE-000387-00 and MOET32023-0005), and National Research Foundation (NRF-000352-00).