Surufatinib plus toripalimab for patients with advanced solid tumors and disease progression after prior immunotherapy: an open-label multi-cohort phase 2 trial

Abstract

Background

Immune checkpoint inhibitor (ICI) plus an anti-angiogenic agent has shown antitumor activity in patients whose disease had progressed on a prior ICI but further evidence across tumor types is needed.

Methods

This phase 2 trial evaluated the efficacy and safety of surufatinib (anti-angiogenic multi-tyrosine-kinase inhibitor, 250 mg orally, once daily) plus toripalimab (programmed-death-1 [PD-1] inhibitor, 240 mg intravenously, once every three weeks) in adult patients with advanced solid tumors and progression on a PD-1/programmed death ligand 1 inhibitor. The primary endpoint was investigator-assessed objective response rate (ORR) per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.

Results

A total of 28 patients were enrolled and received treatment. Two of 27 evaluable patients achieved partial response, for an ORR of 7.4% (95% CI: 0.9–24.3); the disease control rate was 66.7% (95% CI: 46.0–83.5). Median progression-free survival was 3.9 months (95% CI: 1.4–4.2) and overall survival was 13.0 months (95% CI: 8.6–21.6; median follow-up: 20.9 months). Efficacy findings were consistent when assessed per immune-related RECIST. Eleven (39.3%) patients experienced grade ≥ 3 treatment-related adverse events (TRAEs), most frequently increased blood pressure (n = 3, 10.7%) and liver damage (n = 2, 7.1%). TRAEs led to surufatinib discontinuation in two (7.1%) patients. There were no treatment-related deaths.

Conclusions

ICI-based combination rechallenge with surufatinib-toripalimab conferred signal-generating activity in patients with advanced solid tumors, with a manageable safety profile. Studies in larger populations are warranted.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00262-026-04419-7.

Introduction

Immune checkpoint inhibitors (ICIs), predominantly monoclonal antibodies targeting programmed death 1 (PD-1) and its ligand (PD-L1), have significantly improved outcomes of various malignancies [1]. Despite durable responses in some patients [2–5], most eventually develop resistance with limited subsequent treatment options and poor prognosis [6].

Rechallenge with ICIs has emerged as a clinically relevant strategy in selected patients, but the term “ICI rechallenge” encompasses several distinct scenarios, including retreatment with the same ICI after treatment discontinuation, switching to an alternative immune-based regimen, and rechallenge with an ICI-containing combination [6–9]. Available evidence remains limited and heterogeneous, with most data derived from retrospective analyses or exploratory cohorts. In recent years, clinically meaningful responses to ICI-based rechallenge have been reported in selected patients with melanoma, renal cell carcinoma (RCC), and non-small cell lung cancer (NSCLC), including prospective studies such as LEAP-004 and KEYNOTE-146, and retreatment analyses from KEYNOTE-010 and KEYNOTE-024 [2, 3, 10, 11]. However, these encouraging findings were observed mainly from those ICI-sensitive malignancies, whilst prospective evidence for ICI-based rechallenge in gastrointestinal (GI) malignancies remains particularly scarce.

Among various rechallenge approaches, combining anti-angiogenic therapy with an ICI has a biologically convincing rationale and has shown activity in several tumor types. The vascular endothelial growth factor (VEGF) signaling contributes to an immunosuppressive tumor microenvironment, and VEGF/-VEGF receptor blockade may help enhance the activity of anti–PD-1/PD-L1 antibodies through modifying the innate immunity, promoting type 1 T-helper cell signaling, and transforming the immunosuppressive tumor microenvironment [12, 13]. However, this strategy is best regarded as an ICI-based combination rechallenge approach rather than rechallenge with ICI alone.

Toripalimab is an anti–PD-1 antibody approved for a variety of advanced malignancies, and surufatinib is an oral multi-kinase inhibitor with anti-angiogenic and immunomodulatory properties [14, 15]. Previous clinical studies demonstrated a well-tolerated safety profile of surufatinib-toripalimab combination and encouraging anti-tumor activity in patients with multiple types of advanced solid tumors [16–21]. We report here the findings of the prior immunotherapy cohort from a prospective, multicohort phase 2 study (NCT04169672), in which we evaluated the efficacy and safety of surufatinib plus toripalimab in patients with advanced solid tumors who experienced disease progression on prior anti–PD-1/PD-L1 therapy. Given that a substantial proportion of the enrolled population had GI malignancies, this study also provided exploratory prospective evidence in such a clinical setting for which prospective data remain limited.

Methods

Study design and participants

This was an open-label, single-arm, multicohort, phase 2 study evaluating the safety and efficacy of surufatinib-toripalimab combination therapy in patients with advanced solid tumors in China. The present report included results for the prior immunotherapy cohort, in which we intended to assess the clinical activity of ICI-based rechallenge with surufatinib-toripalimab in patients who had received prior anti–PD-1/PD-L1 therapy and subsequently experienced disease progression.

Adult patients (age 18–75 years) with histologically/cytologically confirmed, unresectable, or metastatic advanced solid tumors who had previously received an anti–PD-1/anti–PD-L1 antibody for at least 3 months before experiencing disease progression were enrolled. PD-L1 expression in tumor samples (fresh or archival) was assessed prior to study treatment by immunohistochemistry (VENTANA SP263 rabbit monoclonal primary antibody; BenchMark ULTRA System, Roche, AZ, USA) and measured as the combined positive score (CPS) [22]. Patients who had received prior anti-angiogenic inhibitors were excluded, and previous anticancer treatment-related toxicities should have recovered to grade ≤ 1 prior to enrollment. A full account of the study eligibility criteria is provided in Supplementary Table S1.

The study was performed according to the principles of the Declaration of Helsinki, good clinical practice guidelines, and local laws/regulations of China. The protocol and its amendments, and patient informed consent were approved by local institutional review boards and independent ethics committees at all participating sites (Approved on Aug 19, 2019, approval letter No. 2019YW119). All patients provided written informed consent to participate prior to screening.

Procedures

All patients received surufatinib (250 mg orally, once daily) plus toripalimab (240 mg intravenously, once every 3 weeks) continuously until disease progression, death, unacceptable toxicity, patient withdrawal, or loss to follow-up (whichever occurred first), but for a maximum of 24 months dosing of toripalimab.

Tumor response was assessed by the investigator per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 [23] and immune-related RECIST (irRECIST) [24] every 6 weeks from the first dose of the study drugs to week 48, and every 12 weeks thereafter until disease progression, death, unacceptable toxicity, or study discontinuation (whichever occurred first). Initial findings of complete response (CR) or partial response (PR) were confirmed ≥ 4 weeks later. Patients were followed for survival every 12 weeks from treatment discontinuation until death, loss to follow-up, withdrawal of informed consent, or end of study (whichever occurred first).

All treatment-emergent adverse events (TEAEs), serious adverse events (SAEs), and prespecified TEAEs of special interest were recorded (using Medical Dictionary for Regulatory Activities version 25.1) throughout the study, from signing of informed consent to 90 days after the last administration of the study drugs. TEAE severity was graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0 [25].

Outcomes

The primary efficacy endpoint was objective response rate (ORR) assessed by the investigator according to RECIST 1.1. Secondary endpoints were ORR according to irRECIST; duration of response (DoR), progression-free survival (PFS), and disease control rate (DCR) according to both RECIST 1.1 and irRECIST; and overall survival (OS).

Safety endpoints included incidence and severity of TEAEs; grade ≥ 3 TEAEs; SAEs; treatment-related TEAEs (TRAEs); TEAEs of special interest; and TEAEs leading to permanent treatment discontinuation, and dose interruption or modification. Changes from baseline in targeted vital signs, clinical laboratory test results, and ECOG PS were also summarized.

Statistical analyses

The sample size was based on feasibility. The planned enrolment was 20–30 patients for the cohort of patients who had received prior immunotherapy. ORR and DCR were analyzed separately based primarily on the efficacy-evaluable analysis (EEA) set (all patients who had received at least one dose of the study drug and had at least one valid post-baseline tumor assessment); 95% exact confidence intervals (CIs) were calculated using the Clopper–Pearson method. Analyses of PFS, OS, and safety were conducted using the intention-to-treat (ITT) set (all patients who received at least one dose of the study drugs). DoR was analyzed based on patients with confirmed CR or PR. Median values of all time-to-event variables were estimated using the Kaplan–Meier method. Prespecified subgroup analyses were performed according to PD-L1 CPS. Post hoc analyses were conducted to assess efficacy according to the duration of prior immunotherapy (≤ 6 months vs. > 6 months) in all patients, and efficacy in the subpopulation with gastric/gastro-esophageal junction adenocarcinoma (GC/GEJC).

All statistical analyses were performed using SAS Enterprise Guide 8.3 (64-bit). This trial is registered with ClinicalTrials.gov (NCT04169672).

Results

Patient characteristics

Among the 34 patients who were assessed for eligibility, 28 were enrolled and received at least one dose of the study drugs (ITT set). At the data cut-off date (28 February 2023), all 28 patients had discontinued surufatinib-toripalimab, due primarily to radiological disease progression (n = 21, 75.0%). A patient disposition diagram for this study is provided in Supplementary Fig. S1.

Table 1 summarizes the patient baseline demographic and disease characteristics. The patients had a mean ± standard deviation age of 57.5 ± 9.1 years (82.1% aged < 65 years) and were predominantly male (n = 20, 71.4%). The majority had stage IV disease (n = 27, 96.4%), and the most frequent tumor type was GC/GEJC (n = 9, 32.1%). Per protocol, all 28 patients had received one line of prior ICI therapy, predominantly in the first-line setting (n = 17, 60.7%), and the majority had a duration of prior ICI of > 6 months (n = 17, 60.7%).

Table 1.

Summary of patient baseline and disease characteristics

Characteristic	Prior immunotherapy cohort
Data are n (%) unless otherwise specified	N = 28
Age, years
Mean ± standard deviation	57.5 ± 9.1
Median (range)	58 (37–75)
Category, n (%)
< 65 years	23 (82.1)
≥ 65 years	5 (17.9)
Sex
Male	20 (71.4)
Female	8 (28.6)
Race
Asian	28 (100)
ECOG PS
0	5 (17.9)
1	23 (82.1)
Tumor stage at screening
III	1 (3.6)
IV	27 (96.4)
PD-L1 CPS
< 1	8 (28.6)
≥ 1 to < 5	4 (14.3)
≥ 5 to < 10	4 (14.3)
≥ 10 to < 50	7 (25.0)
≥ 50	4 (14.3)
Missing	1 (3.6)
Primary tumor type
GC/GEJC	9 (32.1)
ESCC	4 (14.3)
NSCLC	4 (14.3)
CRC	3 (10.7)
BTC	2 (7.1)
Other*	6 (21.4)
No. of involved organs
1	3 (10.7)
2	11 (39.3)
≥ 3	14 (50.0)
Prior anti-tumor therapy lines received
1	17 (60.7)
≥ 2	11 (39.3)
Prior radiotherapy received	12 (42.9)
Prior anti-tumor surgery/operation received	16 (57.1)
Prior anti–PD-1/–PD-L1 immunotherapies received
Tislelizumab	7 (25.0)
Sintilimab	5 (17.9)
Camrelizumab	4 (14.3)
Toripalimab	4 (14.3)
Pembrolizumab	2 (7.1)
Nivolumab	1 (3.6)
Other†	5 (17.9)
Treatment line for prior immunotherapy
1	17 (60.7)
2	7 (25.0)
≥ 3‡	4 (14.3)
Duration of prior immunotherapy
≤ 6 months	11 (39.3)
> 6 months	17 (60.7)
Best efficacy of prior immunotherapy
PR	11 (39.3)
SD	8 (28.6)
Unknown	9 (32.1)

BTC, biliary tract cancer; CPS, combined positive score; CRC, colorectal carcinoma; ECOG PS, Eastern Cooperative Oncology Group performance score; ESCC, esophageal squamous cell carcinoma; GC/GEJC, gastric/gastro-esophageal junction adenocarcinoma; NSCLC, non-small cell lung cancer; PD-1, programmed death 1; PD-L1, programmed death ligand 1; PR, partial response; SD, stable disease

^*One patient each with cervical squamous cell carcinoma, nasopharyngeal cancer, squamous cell anal carcinoma, salivary gland-type tumor, melanoma, and small intestinal mucinous adenocarcinoma

^†One patient each received AK105, CS1003, KL-A167, SHR1701, and GR1405

^‡Line of immunotherapy was third line in one (3.6%) patient and fourth line in three (10.7%) patients

Efficacy

The EEA set comprised 27 (96.4%) patients (one patient had no post-treatment assessment). For the primary endpoint (Table 2), two patients had a best overall response per RECIST 1.1 of PR, for an ORR of 7.4% (95% CI: 0.9–24.3), with DoR of 10.2 months and 10.8 months, respectively (Fig. 1). Sixteen patients achieved stable disease (SD), providing a DCR of 66.7% (95% CI: 46.0–83.5). Two patients had received prolonged treatment of longer than 24 months (colorectal cancer, PD-L1 CPS ≥ 10, duration of prior ICI ≤ 6 months) and 21 months (GC/GEJC, PD-L1 CPS < 10, duration of prior ICI > 6 months), and one patient for longer than 12 months (cervical squamous cell carcinoma, PD-L1 CPS ≥ 10, duration of prior ICI > 6 months). Tumor response was consistent when assessed by irRECIST (Table 2).

Table 2.

Summary of investigator-assessed tumor response data per RECIST and irRECIST in patients in the prior immunotherapy cohort (EEA set)

Data are n (%) unless otherwise stated	RECIST 1.1	irRECIST
Best overall response in the EEA set (N = 27)
(ir)CR	0	0
(ir)PR	2 (7.4)	3 (11.1)
(ir)SD	16 (59.3)	15 (55.6)
(ir)PD	9 (33.3)	9 (33.3)
Not evaluable	0	0
ORR	2 (7.4)	3 (11.1)
95% CI*	0.9–24.3	2.4–29.2
DCR	18 (66.7)	18 (66.7)
95% CI*	46.0–83.5	46.0–83.5
DoR, months†
Median	NE	10.2
95% CI†	10.2–NE	4.2–NE

CI, confidence interval; CR, complete response; DCR, disease control rate; DoR, duration of response; EEA, efficacy-evaluable analysis; ir, immune-related; NE, not estimable; ORR, objective response rate; PD, progressive disease; PR, partial response; RECIST 1.1, Response Evaluation Criteria in Solid Tumors version 1.1; SD, stable disease

^*Calculated using the Clopper–Pearson method

^†Calculated using the Kaplan–Meier method

Fig. 1.

Swimmer plot showing tumor response (investigator assessed per RECIST 1.1) over time in patients in the prior immunotherapy cohort (EEA set). Tumor type and prior immunotherapy duration are noted > 6 m, duration of prior immunotherapy treatment of more than 6 months; ≤ 6 m, duration of prior immunotherapy treatment of up to 6 months; BTC, biliary tract cancer; CPS, combined positive score; CRC, colorectal carcinoma; CSCC, cervical squamous cell carcinoma; ESCC, esophageal squamous cell carcinoma; GC/GEJC, gastric adenocarcinoma/gastro-esophageal junction adenocarcinoma; NPC, nasopharyngeal cancer; NSCLC, non-small cell lung cancer; PD, disease progression; PD-L1, programmed death ligand 1; PR, partial response; RECIST 1.1, Response Evaluation Criteria in Solid Tumors version 1.1; SCAC, squamous cell carcinoma of the anal canal; SD, stable disease; SIMA, small intestinal mucinous adenocarcinoma; SGT, salivary gland tumor

By the data cutoff date, after a median follow-up of 20.9 months (95% CI: 16.1–25.0), 24 (85.7%) patients in the ITT set had experienced a PFS event (by RECIST 1.1 or irRECIST) and 17 (60.7%) patients had died. Median (ir)PFS was 3.9 months (95% CI: 1.4–4.2) when assessed by RECIST 1.1 and 4.1 months (95% CI: 1.4–5.4) by irRECIST (Table S9, Fig. 2a). Median OS was 13.0 months (95% CI: 8.6–21.6), and the 12-month survival rate was 59.3% (95% CI: 37.5–75.8) (Fig. 2b).

Fig. 2.

Kaplan–Meier plot of a investigator-assessed PFS per RECIST 1.1 and b OS in patients in the prior immunotherapy cohort (ITT set).

CI, confidence interval; ITT, intention-to-treat; OS, overall survival; PFS, progression-free survival; RECIST 1.1, Response Evaluation Criteria in Solid Tumors version 1.1

Tumor response to surufatinib-toripalimab based on PD-L1 CPS yielded a numerically higher RECIST-1.1-assessed ORR and DCR in the CPS ≥ 10 (n = 11) versus CPS < 10 (n = 15) subgroups (EEA set). ORR was 18.2% (95% CI: 2.3–51.8) for CPS ≥ 10 versus 0% (95% CI: 0.0–21.8) for CPS < 10; DCR was 72.7% (95% CI: 39.0–94.0) versus 60.0% (95% CI: 32.3–83.7), respectively. Results were consistent when assessed by irRECIST (Supplementary Table S2).

Median PFS was similar for the PD-L1 CPS ≥ 10 (n = 11) and CPS < 10 (n = 16) subgroups when assessed per RECIST 1.1, but was numerically longer in the CPS ≥ 10 subgroup when assessed per irRECIST (Supplementary Table S3). Median OS was numerically longer in the PD-L1 CPS ≥ 10 subgroup versus CPS < 10 (14.5 vs. 12.9 months). In addition, 6-, 9-, and 12-month OS rates were higher in the PD-L1 CPS ≥ 10 subgroup (Supplementary Table S4).

The results from a post hoc analysis of efficacy according to duration of prior ICI are provided in the Supplement (Tables S5, S6, S7). In patients with a duration of prior ICI > 6 months (n = 17) or ≤ 6 months (n = 10), DCRs were 76.5% (95% CI: 50.1–93.2) and 50.0% (95% CI: 18.7–81.3), median PFS was 4.2 months (95% CI: 1.4–6.9) and 2.0 months (95% CI: 0.8–4.1), and median OS was 13.0 months (95% CI: 10.4–21.6) and 8.2 months (95% CI: 3.9–NE), for the > 6 months and ≤ 6 months subgroups, respectively.

Among the nine patients with GC/GEJC evaluated for efficacy, a post hoc analysis showed a best overall response of SD in five patients per RECIST 1.1, for a DCR of 55.6% (95% CI: 21.2–86.3), and the median PFS was 2.7 months (95% CI: 1.4–5.5). Tumor response and PFS assessed by irRECIST were available in Supplementary Table S8. Median OS was 13.7 months (95% CI: 3.6–NE), with a 12-month OS rate of 62.5% (95% CI: 22.9–86.1).

Safety

At the data cutoff date, the median duration of exposure to surufatinib-toripalimab was 4.1 months (range, 0.8–24.6). TEAEs were reported for 25 (89.3%) patients; all were considered by the investigator to be related to the study treatment (TRAEs) (Table 3). The most frequent TRAEs (in ≥ 20% of patients) were blood thyroid stimulating hormone increased (n = 13, 46.4%), occult blood positive (n = 11, 39.3%), and proteinuria (n = 11, 39.3%). Grade ≥ 3 TRAEs occurred in 11 (39.3%) patients; those observed in two or more patients were increased blood pressure (n = 3, 10.7%) and liver damage (n = 2, 7.1%).

Table 3.

Overview of treatment-related TEAEs in the ITT set

	Surufatinib-toripalimab
Data are n (%)	N = 28
Any treatment-related TEAE	25 (89.3)
Grade ≥ 3	11 (39.3)
Surufatinib-related TEAE	25 (89.3)
Grade ≥ 3	11 (39.3)
Leading to surufatinib dose interruption or reduction	12 (42.9)
Toripalimab-related TEAE	22 (78.6)
Grade ≥ 3	5 (17.9)
Leading to toripalimab dose interruption or delay	5 (17.9)
Leading to discontinuation of either study drug	2 (7.1)
Leading to discontinuation of surufatinib	2 (7.1)
Leading to discontinuation of toripalimab	0
Leading to death	0
Treatment-related SAE	8 (28.6)
Surufatinib-related SAE	8 (28.6)
Toripalimab-related SAE	5 (17.9)
irAE related to toripalimab	8 (28.6)
Grade ≥ 3	2 (7.1)

Treatment-related TEAEs by preferred term, reported in ≥ 20% of patients*†	Any grade	Grade ≥ 3
Any	25 (89.3)	11 (39.3)
Blood thyroid stimulating hormone increased	13 (46.4)	0
Proteinuria	11 (39.3)	1 (3.6)
Occult blood positive	11 (39.3)	0
Conjugated bilirubin increased	10 (35.7)	0
Urinary occult blood positive	8 (28.6)	0
Diarrhea	7 (25.0)	1 (3.6)
Anemia	7 (25.0)	1 (3.6)
Aspartate aminotransferase increased	7 (25.0)	0
Blood bilirubin increased	7 (25.0)	0
Blood urea increased	7 (25.0)	0
Tri-iodothyronine free decreased	7 (25.0)	0
Hypertriglyceridemia	6 (21.4)	1 (3.6)
Thrombocytopenia	6 (21.4)	0

TEAE, treatment-emergent adverse event; ir, immune-related; ITT, intention-to-treat; SAE, serious adverse event

^*Patients with more than one TEAE within a particular preferred term were counted only once for that preferred term

^†Coded according to the Medical Dictionary for Regulatory Activities version 25.1

TRAEs leading to dose interruption or reduction of either surufatinib or toripalimab occurred in 13 (46.4%) patients. Surufatinib was discontinued due to TRAEs in two (7.1%) patients; no TRAEs led to toripalimab discontinuation. Treatment-related SAEs were reported for eight (28.6%) patients; all but one, liver injury (n = 2, 7.1%), occurred in just one patient each.

Immune-related AEs (irAEs) attributable to toripalimab were recorded for eight (28.6%) patients, and were grade ≥ 3 in two (7.1%) patients. One patient with baseline diabetes developed grade 2 immune-related pancreatitis followed several days later by grade 3 diabetic ketoacidosis after four treatment cycles, which resolved after insulin therapy and supportive treatment. The other patient developed immune-mediated lung disease after 24 treatment cycles, which improved in both symptoms and radiographic findings after toripalimab interruption and corticosteroid treatment. No irAEs resulted in permanent discontinuation of toripalimab or death.

Discussion

Surufatinib-toripalimab combination therapy demonstrated modest antitumor activity with a manageable safety profile in patients with ICI-pretreated advanced solid tumors. Across tumor types, the ORR was 7.4%, and median PFS and OS were 3.9 months and 13.0 months, respectively. To our knowledge, prospective data on ICI-based rechallenge in GI cancers remain extremely limited, and this study provides preliminary exploratory evidence in a cohort enriched for GI malignancies.

In recent clinical trials, clinically meaningful benefit of ICI-based rechallenge with an anti-angiogenic agent plus an anti–PD-1/PD-L1 antibody has been observed. In the phase 2 LEAP-004 study, lenvatinib plus pembrolizumab rechallenge conferred an ORR of 21.4% (three CRs, 19 PRs) in 104 patients with advanced, unresectable melanoma, with median PFS and OS being 4.2 and 14.0 months, respectively [10]. In the phase 1/2 KEYNOTE-146 study, lenvatinib-pembrolizumab provided an ORR of 62.5% and a DCR of 92.3%, with a median DoR of 12.5 months and a median investigator-assessed PFS of 12.2 months in 103 patients with advanced RCC who had progressed on a PD-1/PD-L1 inhibitor [11]. These promising results, however, are predominantly derived from ICI-sensitive malignancies. In contrast, an international, multicohort phase 2 trial demonstrated that regorafenib plus pembrolizumab conferred ORRs of 5.9% and 11.1%, and SD rates of 48.5% and 63.0% in patients with advanced hepatocellular carcinoma (HCC) previously treated with atezolizumab plus bevacizumab (Cohort 1) or other ICI-based regimens (Cohort 2), respectively. Median PFS were 2.8 months in Cohort 1 and 4.2 months in Cohort 2 [26]. These results were numerically comparable to those from our study, although cross-trial comparisons are inherently limited. Taken together, the available prospective data suggest that, in some GI malignancies, the benefit of ICI-based rechallenge may be reflected at least in part by disease stabilization rather than by objective tumor regression.

Since that not all patients respond to ICI-based rechallenge, it is important to identify those patients most likely to benefit. PD-L1 CPS is the primary metric for tumor PD-L1 expression in many types of solid tumor [22]. The KEYNOTE-061 and KEYNOTE-062 studies demonstrated improved OS with pembrolizumab versus chemotherapy particularly in the subpopulations of patients with advanced GC/GEJC and a PD-L1 CPS ≥ 10 [27, 28]. In the present study, ORR and DCR were numerically higher in patients with PD-L1 CPS ≥ 10 versus < 10. However, these findings should be interpreted cautiously because of the small sample size and lack of statistical validation. In addition, PD-L1 expression in tumor samples was assessed before study treatment but both fresh samples and archival samples were permitted. Given the dynamic nature of PD-L1 expression, temporal heterogeneity in PD-L1 assessment cannot be excluded. Hence, these data cannot support a predictive claim that PD-L1 CPS ≥ 10 is a validated threshold for benefit in this rechallenge setting.

The present findings also suggest numerically greater benefit including DCR, median PFS and OS in patients with a prior ICI duration of > 6 months, consistent with a real-world French national database study showing that initial ICI treatment ≥ 3 months in patients with NSCLC was associated with improved OS after ICI rechallenge, possibly due to consolidation of immune memory during the initial treatment course [29]. However, an additional consideration in interpreting this finding is the eligibility requirement for prior immunotherapy exposure of ≥ 3 months, which may have introduced selection bias of enriching patients with more indolent disease biology or relative sensitivity to prior immune checkpoint blockade. This eligibility criterion was intentional as the study was designed to focus on a population more consistent with secondary/acquired resistance rather than primary resistance. Thus, the findings should not be generalized to all patients with ICI-refractory disease. Post-ICI strategies for primary refractory disease may differ from those for acquired resistance, as suggested by studies such as SWOG S1616 in melanoma [30].

In a small post-hoc subgroup of patients with advanced GC/GEJC, surufatinib-toripalimab showed limited antitumor activity, with stable disease as the best overall response in five of nine evaluable patients (DCR, 55.6% by RECIST 1.1), a median PFS of 2.7 months, and a median OS of 13.7 months. While ICI combined with chemotherapy is current first-line standard of care for advanced GC/GEJC with any PD-L1 expression, post-progression options remain limited to ramucirumab plus paclitaxel or chemotherapy alone for human epidermal growth factor receptor 2 (HER2)-negative disease, or trastuzumab deruxtecan for HER2-positive disease [31–33]. Given the exploratory analysis nature, these findings should not be overinterpreted and do not constitute tumor-specific evidence, but they may provide preliminary exploratory support for further evaluation of an ICI-containing combination strategy in selected patients with GC/GEJC after prior ICI failure.

The safety profile of surufatinib-toripalimab combination therapy in this rechallenge setting was manageable, with no new safety signals compared with an early-phase trial of the same combination [16] or with previous data for the individual agents in patients with advanced solid tumors [34–37]. The more frequently observed adverse events, such as hypertension, proteinuria, and occult blood positivity, were generally consistent with the known toxicity profile of surufatinib. irAEs, which were attributed to toripalimab, occurred in 28.6% of patients, but were predominantly mild (grade 1 or 2) in severity. Grade ≥ 3 irAEs were uncommon, with one case each of diabetic ketoacidosis and immune-mediated lung disease. The diabetic ketoacidosis occurred in a patient with baseline diabetes and was preceded by grade 2 immune-related pancreatitis, and the event resolved after supportive treatment; the immune-mediated lung disease improved after toripalimab interruption and corticosteroid treatment. Nevertheless, more detailed characterization of the timing, management, and reversibility of irAEs will be important in future rechallenge studies.

The study has several limitations, including the open-label, single-arm design, small sample size, inclusion of heterogeneous tumor types, and the lack of independent review of tumor response. In particular, the combination of small sample size and multiple malignancies enrolled makes it difficult to correlate clinical outcome (efficacy and safety) with any particular tumor type. All subgroup analyses, including those according to PD-L1 CPS, prior ICI duration, and GC/GEJC subgroup, should be considered exploratory and hypothesis-generating only. At the same time, data on biomarkers such as microsatellite instability/mismatch repair status and tumor mutational burden, which have been widely accepted as biomarkers for predicting the effectiveness of immunotherapy [38, 39], were unavailable in this cohort. This incomplete biomarker characterization may also limit the interpretation of response heterogeneity and biomarker subgroup findings of our study.

In conclusion, this study provided initial prospective data on the efficacy and safety of ICI-based combination rechallenge in a cohort enriched for GI malignancies whose disease had progressed on a prior ICI. All findings should be regarded as exploratory and signal-generating, and further evaluation in larger, biomarker-informed, tumor-specific prospective studies is warranted.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BTC

Biliary tract cancer

CI

Confidence interval

CPS

Combined positive score

CR

Complete response

CSCC

Cervical squamous cell carcinoma

CRC

Colorectal carcinoma

DCR

Disease control rate

DoR

Duration of response

ECOG PS

Eastern cooperative oncology group performance status

EEA

Efficacy-evaluable analysis

ESCC

Esophageal squamous cell carcinoma

GC/GEJC

Gastric adenocarcinoma/gastroesophageal junction adenocarcinoma

GI

Gastrointestinal

HCC

Hepatocellular carcinoma

HER2

Human epidermal growth factor receptor 2

ICI

Immune checkpoint inhibitor

irAE

Immune-related adverse event

irRECIST

Immune-related response evaluation criteria in solid tumors

ITT

Intention-to-treat

NE

Not estimable

NPC

Nasopharyngeal cancer

NSCLC

Non-small-cell lung cancer

ORR

Objective response rate

OS

Overall survival

PD

Progressive disease

PD-1

Programmed cell death protein 1

PD-L1

Programmed cell death-ligand 1

PFS

Progression-free survival

PR

Partial response

RCC

Renal cell carcinoma

RECIST v1.1

Response evaluation criteria in solid tumors version 1.1

SAE

Serious adverse event

SCAC

Squamous cell carcinoma of the anal canal

SD

Stable disease

SIMA

Small intestinal mucinous adenocarcinoma

SGT

Salivary gland tumor

TEAE

Treatment-emergent adverse event

TRAE

Treatment-related adverse event

VEGF

Vascular endothelial growth factor

Author contributions

Lin Shen collaborated with the sponsor to conceive and design the study. Panpan Zhang, Ming Lu, Lijie Song, Yanqiao Zhang, Zhendong Chen, Chunjiao Wu, Zhiping Li and Xing Zhang enrolled the patients and collected data. Qian Xu participated in the final data analysis. All authors contributed to the data interpretation, reviewed the manuscript and approved the final version of the manuscript for submission.

Funding

This work was supported by Shanghai Committee of Science and Technology, China (the Science and Technology Innovation—Biomedical Science and Technology Support Project [Grant No. 20S11907500]) and HUTCHMED Limited.

Data availability

The authors declare that all data supporting the findings of this study are available within the paper and its supplementary information files.

Declarations

Conflict of interest

Qian Xu, Jinghong Zhou, Haiyan Shi, Panfeng Tan, Songhua Fan and Weiguo Su are employees of HUTCHMED Limited.

Ethical approval

The study was performed according to the principles of the Declaration of Helsinki, good clinical practice guidelines, and local laws/regulations of China. The protocol and its amendments, and patient informed consent were approved by local institutional review boards and independent ethics committees at all participating sites (Approved on Aug 19, 2019, approval letter No. 2019YW119).

Consent to participate

All patients provided written informed consent to participate prior to screening.