Between April 23, 2021, and November 30, 2023, a total of 122 patients were enrolled in the phase 1 study, with 17 patients in the dose-escalation phase and 105 patients in the dose-expansion phase. Among these patients, 91 had RET fusion-positive NSCLC, 23 had RET-mutant MTC, 7 had RET fusion-positive PTC, and 1 had RET-mutant gastric cancer. For prior treatment, 36.1% (44/122) of patients were treatment-naïve and 63.9% (78/122) of patients were previously treated, including 43.4% (53/122) with chemotherapy, 23.8% (29/122) with MKIs, 16.4% (20/122) with programmed death-1 (PD-1)/programmed death ligand-1 (PD-L1) inhibitors, and 27.9% (33/122) with other systemic therapies. RET fusions were identified by next-generation sequencing (NGS) in 46.7% (57/122) of patients, reverse transcription-polymerase chain reaction (RT-PCR) in 24.6% (30/122) of patients, and other methods, including fluorescence in situ hybridization (FISH) and immunohistochemistry (IHC) in 28.7% (35/122) of patients. The most prevalent fusion partners were KIF5B (54.1%, 66/122) and CCDC6 (15.6%, 19/122). The most common RET mutation was RETM918T (12.3%, 15/122). Detailed patient demographics and baseline characteristics are summarized in Table 1 and Supplementary Table 3.

At the data cut-off date of November 30, 2023, the median follow-up time was 8.28 months (95% confidence interval [CI] 6.57–8.51). A total of 37 patients (30.3%, 37/122) had discontinued SY-5007 treatment due to disease progression or death (n = 29), withdrawal of consent (n = 6), or adverse events (AEs, n = 2). Meanwhile, 69.7% (85/122) of patients remained on treatment (Supplementary Fig. 2).

At the data cut-off date, all patients had received as least one dose of SY-5007 and were evaluable for safety analysis. No dose-limiting toxicity (DLT) was observed, and the maximum tolerated dose (MTD) was not reached. A total of 96.7% (118/122) of patients experienced at least one treatment-related adverse event (TRAE). TRAEs with an incidence of ≥ 50% included increased aspartate aminotransferase (AST, 69.7%), increased alanine aminotransferase (ALT, 56.6%), diarrhea (54.1%), and neutropenia (54.1%), all of which were reversible with appropriate treatment discontinuation. Grade ≥ 3 TRAE occurred in 57.4% (70/122) of patients. The most common grade ≥ 3 TRAEs (incidence of ≥ 5%) were hypertension (22.1%), diarrhea (16.4%), hypertriglyceridemia (6.6%), and neutropenia (6.6%). The TRAEs in each dose group are summarized in Table 2.

TRAE-induced dose interruption and reduction were observed in 56 (45.9%) and 29 (23.8%) patients, respectively. The most common TRAEs (incidence of ≥ 5%) leading to dose interruption were hypertension (12.3%) and diarrhea (10.7%). Hypertension (5.7%) was the primary cause of dose reduction. Dose discontinuation occurred in 2 (1.6%) patients, due to one case of drug eruption and one instance of peripheral arterial thrombosis. Treatment-related serious adverse events (TRSAEs) occurred in 13 (10.7%) patients, with no treatment-related deaths reported.

Forty-three evaluable patients (17 in dose-escalation phase, 26 in dose-expansion phase) were included for PK analysis. After a single-dose administration, the mean time to maximum concentration (Tmax) ranged from 0.5 h to 4.0 h across different dose cohorts, and mean elimination half-life (T1/2) ranged from 6.8 h to 44.4 h. SY-5007 exposure increased in a dose-proportional manner from 20 mg to 160 mg, with maximum concentration (Cmax) ranging from 214.0 to 3468.0 ng/mL and area under the concentration-time curve (AUC) from the time of dosing extrapolated to infinity, based on the last observed concentration (AUCINF_obs) ranging from 2407.0 to 21315.5 hr*ng/mL. Notably, SY-5007 exposure (Cmax and AUC) did not exhibit a significant increase between 160 mg and 200 mg, suggesting saturation at these doses. After multiple-dose administration on Cycle 1 Day 28, SY-5007 exposure (Cmax and AUC) increased significantly with repeated dosing in the 20 mg to 160 mg range. Modest drug accumulation was observed at steady state. (Table 3 and Supplementary Fig. 3).

At the data cut-off date, efficacy assessments were conducted for 116 out of the 122 enrolled patients. Three patients had not yet undergone efficacy evaluation following the initial administration, and three withdrew the consent. Among the efficacy-evaluable patients, 114 had measurable lesions. One patient with gastric cancer, who had a non-measurable lesion, was classified as having progressive disease (PD) due to increased ascites after 2 cycles of SY-5007 treatment. Another patient with NSCLC and a non-measurable lesion was categorized as having stable disease (SD) after 1 cycle of SY-5007 treatment.

Overall, target tumor shrinkage was observed in 95.7% of patients (111/116). The overall objective response rate (ORR) and disease control rate (DCR) were 57.8% (67/116, 95% CI 48.2-66.9%) and 95.7% (111/116, 95% CI 90.2–98.6%), respectively (Fig. 1a). SY-5007 induced rapid and durable responses (Figs. 1b, 2a), with a median time to first response (TTR) of 2.82 months (95% CI 2.75–4.66) and a median duration of response (DoR) of 19.9 months (95% CI 12.8-not evaluated [NE]). The median PFS was 21.1 months (95% CI 13.8-NE), with 29 (23.8%) patients experiencing progression events or death (Supplementary Table 3). The estimated 6-month and 12-month PFS rates were 86.1% (95% CI 77.6–91.6%) and 68.9% (95% CI 55.8–78.8%), respectively (Fig. 2b).

Tumor response in patients with RET-altered solid tumors receiving SY-5007. a Waterfall plots of the maximum tumor size change in 114 efficacy-evaluable patients with measurable lesions in RET-altered solid tumors. b The change in tumor burden over time in patients for whom postbaseline tumor data were available. BID bis in die, DCR disease control rate, MTC medullary thyroid cancer, NSCLC non-small cell lung cancer, ORR objective response rate, PTC papillary thyroid carcinoma, QD quaque die

Efficacy of SY-5007 in patients with RET-altered solid tumors. a The time to response, the duration of treatment, and patient status by the data cut-off date for all enrolled patients with RET-altered solid tumors, according to the dose of SY-5007. b, c Kaplan-Meier curve of the progression-free survival in all patients (b) and NSCLC patients at 160 mg BID (c). Data shown were determined by investigator assessments. d Landscape of gene alterations in ctDNA from patients with progressive disease. The figure illustrates concurrent gene alterations identified in plasma samples from patients at baseline and after disease progression. Each column represents an individual patient for whom data are available. Within each column, colored rectangles indicate specific gene alterations, categorized by alteration type. Newly acquired genetic alterations in patients with progressive disease are highlighted in red font. CI confidence interval, CN copy number, NE not evaluated, NSCLC non-small cell lung cancer, PFS progression-free survival, SNV single nucleotide variation

In the dose-escalation phase, the ORR and DCR were 52.9% (9/17, 95% CI 27.8–77.0%) and 94.1% (16/17, 95% CI 71.3–99.9%), respectively. In the dose-expansion phase, the ORR and DCR were 58.6% (58/99, 95% CI 48.2–68.4%) and 96.0% (95/99, 95% CI 90.0–98.9%), respectively. Specifically, for all patients receiving SY-5007 at 160 mg BID or 200 mg BID, the ORRs were 58.9% (56/95, 95% CI 48.4–68.9%) and 60.0% (6/10, 95% CI 26.2–87.8%), respectively, and DCRs were 94.7% (90/95, 95% CI 88.1–98.3%) and 100.0% (10/10, 95% CI 69.2–100.0%), respectively (Table 4).

Among the 116 efficacy evaluable patients, 40 were treatment-naïve. For these patients, the ORR and DCR were 70.0% (28/40, 95% CI 53.5–83.4%) and 95.0% (38/40, 95% CI 83.1–99.4%), respectively. Of these, 37 received SY-5007 at 160 mg BID, with an ORR of 70.3% (26/37, 95% CI 53.0–84.1%) and a DCR of 94.6% (35/37, 95% CI 81.8–99.3%). Two patients received SY-5007 at 200 mg BID and both achieved a partial response (PR). Among 76 previously treated patients, the ORR was 51.3% (39/76, 95% CI 39.6–63.0%) and the DCR was 96.1% (73/76, 95% CI 88.9–99.2%). For these patients, 58 received SY-5007 at 160 mg BID, resulting in an ORR of 51.7% (30/58, 38.2–65.0%) and a DCR of 94.8% (55/58, 95% CI 85.6-98.9%). Eight patients received SY-5007 at 200 mg BID, with an ORR of 50.0% (4/8, 95% CI 15.7–84.3%) and DCR of 100.0% (8/8, 95% CI 63.1–100.0%, Table 4). Based on a thorough evaluation of safety, PK data, and the encouraging anti-tumor efficacy demonstrated in a substantial sample size, 160 mg BID of SY-5007 was selected as the RP2D.

Further analysis of SY-5007 at RP2D included 100 patients: 70 with RET fusion-positive NSCLC, 23 with RET-mutant MTC, and 7 with RET fusion-positive PTC. The most prevalent fusion partners in NSCLC were KIF5B (72.9%, 51/70) and CCDC6 (17.1%, 12/70), while NCOA4 (42.9%, 3/7) and CCDC6 (28.6%, 2/7) were dominant in PTC. Among RET-mutant MTC patients, the most common mutation was RETM918T mutation (65.2%, 15/23). Regarding treatment history, 45.7% (32/70) of NSCLC patients, 14.3% (1/7) of PTC patients, and 34.8% (8/23) of MTC patients were treatment-naïve (Supplementary Table 4).

Among 64 evaluable NSCLC patients, SY-5007 achieved an ORR of 64.1% (41/64, 95% CI 51.1–75.7%) and a DCR of 96.9% (62/64, 95% CI 89.2-99.6%, Supplementary Table 5). The median TTR was 2.75 months (95% CI 1.01–2.89), and the median DoR was 13.0 months (95% CI 9.26-NE). The median PFS was 15.4 months (95% CI 10.1-NE), and 26.6% of patients experienced progression or death. Estimated 6- and 12-month PFS rates were 89.6% (95% CI 78.2-95.2%) and 67.5% (95% CI 48.3–80.8%), respectively (Fig. 2c). Notably, SY-5007 demonstrated remarkable anti-tumor efficacy in both treatment-naïve and previously treated patients, with an ORR of 71.4% (20/28, 95% CI 51.3–86.8%) and 58.3% (21/36, 95% CI 40.8–74.5%), and a DCR of 100.0% (28/28, 95% CI 87.7–100.0%) and 94.4% (34/36, 95% CI 81.3-99.3%), respectively.

All 23 RET-mutant MTC patients and 7 RET fusion-positive PTC patients receiving SY-5007 at RP2D were evaluable for efficacy assessment, demonstrating an ORR of 52.2% (12/23, 95% CI 30.6–73.2%) and 42.9% (3/7, 95% CI 9.9–81.6%), and DCR of 91.3% (21/23, 95% CI 72.0–98.9%) and 100.0% (7/7, 95% CI 59.0–100.0%), respectively. Both the median DoR and PFS were not reached during the median follow-up of 6.44 months (95% CI 4.60-8.37) for MTC patients and 6.67 months (95% CI 2.79–8.37) for PTC patients and the estimated 12-month PFS rates were 85.2% (95% CI 60.5–95.0%) for MTC and 100.0% (95% CI 100.0–100.0%) for PTC. Additionally, SY-5007 exhibited anti-tumor efficacy in both treatment-naïve and previously treated patients, with an ORR of 75.0% (6/8, 95% CI 34.9–96.8%) and 40% (6/15, 95% CI 16.3–67.7%) for MTC patients, and 0.0% (0/1, 95% CI 0.0-97.5%) and 50.0% (3/6, 95% CI 11.8–88.2%) for PTC patients, respectively (Supplementary Table 5).

Moreover, SY-5007 exhibited notable intracranial anti-tumor efficacy in 3 evaluable patients with baseline measurable central nervous system (CNS) metastases. The intracranial ORR was 66.7% (2/3, 95% CI 9.4–99.2%) and DCR was 100.0% (3/3, 95% CI 29.2–100.0%) as assessed by the Response Assessment in Neuro-Oncology Brain Metastases (RANO-BM) criteria. Notably, in one case report involving a 63-year-old male NSCLC patient with KIF5B-RET fusion, a 10.8 mm intracranial lesion at baseline reduced by 40.7% after 4 weeks and by 56.5% after 12 weeks of treatment with SY-5007 at 160 mg BID (Supplementary Fig. 4), highlighting the promising intracranial efficacy of SY-5007. Ongoing assessments are underway to gain a more comprehensive understanding of its intracranial efficacy.

A retrospective analysis of circulating tumor DNA (ctDNA) profiling was performed using longitudinal plasma samples obtained at baseline, Cycle 1 Day 28, and post-progressive disease. At baseline, blood samples from all patients were subjected for NGS. Among them, 67 (54.9%) patients had detectable RET alterations in ctDNA, including KIF5B-RET (51/122, 41.8%), CCDC6-RET (5/122, 4.1%), NCOA4-RET (3/122, 2.5%), RETM918T mutation (7/122, 5.7%) and other RET variation (1/122, 0.8%). 53 patients did not show detectable RET alterations in ctDNA, and 2 samples were ineligible for ctDNA profiling.

SY-5007 demonstrated significant anti-tumor efficacy in both patients with detectable and undetectable RET variations at baseline, achieving an ORR of 57.4% (35/61, 95% CI 44.1–70.0%) and 60.4% (32/53, 95% CI 46.0-73.5%), respectively, along with a median PFS of 13.8 months (95% CI 10.0–15.5) and NE (p = 0.0011, Supplementary Fig. 5a and Supplementary Table 6). Specifically, within the group of patients with detectable RET alterations, the ORR was 58.7% (27/46, 95% CI 43.2–73.0%) for those with KIF5B-RET, 71.4% (5/7, 95% CI 29.0–96.3%) for those with other types of RET fusions, and 33.3% (3/9, 95% CI 7.5–70.1%) for those with RET mutations. Twenty-three patients with detectable RET alterations at baseline had ctDNA results available at Cycle 1 Day 28, including 20 with KIF5B-RET fusions, 2 with NCOA4-RET fusions, and 1 with a CDCC6-RET fusion. Notably, all 23 patients achieved a PR by radiologic tumor assessment. Longitudinal analysis revealed a rapid decline in the mean variant allele frequency (VAF) of RET alterations in all 23 patients at Cycle 1 Day 28. Remarkably, 21 patients achieved complete RET clearance (Supplementary Fig. 5b, c). Meanwhile, the ORR was 38.1% (16/42, 95% CI 23.6–54.4%) in patients with non-clearance or unknown RET alterations at Cycle 1 Day 28. The median TTR was 0.91 months (95% CI 0.88–0.95), which is shorter compared with the overall population (2.82 months [95% CI 2.75–4.66], Table 4).

TP53 mutation has been identified as the most prevalent concomitant mutation in RET-rearranged NSCLC and is associated with a poor prognosis.27 Our analysis of SY-5007’s efficacy in patients with detectable RET variations at baseline, stratified by TP53 mutation status, revealed that among those with a concomitant TP53 mutation, SY-5007 demonstrated an ORR of 50.0% (11/22, 95% CI 28.2–71.8%) and a median PFS of 10.1 months (95% CI 6.21-NE). In contrast, patients without a concomitant TP53 mutation showed an ORR of 61.5% (24/39, 95% CI 44.6–76.6%) and a median PFS of 15.4 months (95% CI 9.23-NE, Supplementary Table 7), indicating numerically inferior efficacy in patients with concurrent TP53 mutation.

To elucidate the potential mechanisms of resistance to SY-5007, we conducted comprehensive ctDNA profiling. The molecular landscape of concurrent gene alterations, both at baseline and after PD, was assessed in 21 patients, with a median treatment duration of 6.7 months (range 1.3-16.1). Among the patients, 57.1% (12/21) exhibited at least one novel gene alteration after PD, including mutations in TP53 (9.5%, 2/21), FANCM (9.5%, 2/21), PTCH2 (9.5%, 2/21), KRAS (4.8%, 1/21), BRAF (4.8%, 1/21), MLL (4.8%, 1/21), MLL3 (4.8%, 1/21) and EPHA3 (4.8%, 1/21), as well as copy number gain in MYC (4.8%, 1/21). Notably, no treatment-induced novel on-target RET alterations were identified (Fig. 2d), possibly attributed to the relatively short treatment duration, highlighting the potential significance of off-target induced resistance to SY-5007 during this period.