Clinicopathological and molecular correlates of clinical benefit from disitamab vedotin (RC48), a HER-2-targeting antibody-drug conjugate, in metastatic urothelial carcinoma: a multi-center, real-world study

Abstract

Introduction

Disitamab Vedotin (DV), a HER-2-targeting antibody-drug conjugate, has exhibited substantial clinical advantages in individuals with metastatic urothelial carcinoma (mUC). Nonetheless, the extent of therapeutic efficacy in real-world scenarios for mUC patients and the specific patient profiles with the most benefit from DV remain unclear.

Methods

In this multi-center, retrospective, observational study, a total of 30 patients with mUC were included between April 2021 and March 2024 from three distinct hospitals, along with clinicopathological features and targeted sequencing data. Our analysis encompassed the assessment of the objective response rate (ORR), progression-free survival (PFS), overall survival (OS), and the incidence of treatment-related adverse events (TrAEs).

Results

In the general population, the median follow-up time was 12.1 months, and the median OS was not reached with a 12-month OS rate of 86.7%. The median PFS was 12.2 months (95% confidence interval [CI]: 6.6–17.7 months), with an ORR of 42.3% (95% CI: 23.4%-63.1%) and a DCR of 73.1% (95% CI: 55.2%-88.4%). Higher HER-2 expression is correlated with prolonged clinical benefit from DV (Log-rank P = 0.082. HER-2 IHC 2+ & 3 + vs. IHC 1+: Median PFS: 13.1 vs. 4.6 months; Response rate: 60.0% vs. 0.0%). History of durable response to platinum-based chemotherapy can also imply longer PFS and better response rate (Log-rank P = 0.086. Median PFS: 14.2 vs. 3.6 months. Response rate: 50.0% vs. 10.0%). Moreover, higher mutation burden (TMB) is closely related to better outcomes from DV therapy (Log-rank P = 0.011. Median PFS: 12.8 vs. 4.6 months. Response rate: 71.4% vs. 11.1%. Firth’s penalized multivariable Cox regression: P = 0.015. Hazard Ratio: 0.165. 95% CI: 0.026–0.717).

Conclusion

In conclusion, patients with higher HER-2 expression, positive responses to chemotherapy, or elevated TMB are more likely to benefit from DV. These discoveries support the rationale for employing HER-2 antibody-drug conjugates (ADCs) for mUC patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-025-14870-x.

Introduction

Urothelial carcinoma (UC), the malignant tumor arising from the epithelium lining the ureter, urinary bladder, and urethra, is a highly heterogeneous disease. Though for most patients, the standard care of neoadjuvant cisplatin-based chemotherapy followed by radical surgery, might curb the disease within a non-lethal state, a considerable fraction of UC patients would progress into metastatic urothelial carcinoma (mUC), with few therapeutical choices and poor clinical outcome [1].

According to the latest guideline, platinum-based chemotherapy and immune checkpoint inhibitors (ICI) are mainstream choices for patients with mUC [1]. Solid clinical data showed that standard chemotherapy and immunotherapy could merely provide mUC patients with limited clinical benefit, with median progression-free survival (mPFS) of only 7.8 and 2.0 months, respectively [2, 3]. The rise of antibody-drug conjugates (ADC), nevertheless, has brought novel opportunities. In a Phase III trial, treatment with Nectin-4 targeting ADC Enfortumab Vedotin (EV) and pembrolizumab resulted in significantly better outcomes than chemotherapy in patients with untreated locally advanced or metastatic UC [4]. The TROP-2 targeting ADC also exerted durable clinical benefit in previously treated mUC patients, with median PFS and OS of 5.4 months and 10.9 months, respectively [5].

Disitamab Vedotin (DV), or RC48, is a novel ADC of the anti-HER-2 antibody conjugated with monomethyl auristatin E (MMAE) via a cleavable linker [6]. Exerting anti-tumor activity via microtubule-targeting toxicity, DV exhibited durable clinical benefits and a sound safety profile for mUC patients. The recent data reported from two single-arm, phase II trials (RC48-C005 and RC48-C009) showed an objective response rate (ORR) of 50.5% and a median PFS of 5.9 months [7, 8]. Also, the combination of DV with checkpoint blockade, a commonplace regimen in clinical practice, was confirmed with sound safety and clinical benefit in early-stage trials and real-world data [9–11]. Hence, The United States Food and Drug Administration (FDA) has granted Breakthrough Therapy designation for DV as the second-line treatment of HER-2-positive patients with mUC who have previously received platinum-containing chemotherapy. Also, in China, DV has been approved in clinical practice for HER-2-positive mUC patients by the National Medical Products Administration (NMPA), and the Phase III trial comparing standard chemotherapy plus ICI with DV plus ICI is now recruiting (NCT05911295), which indicates its profound clinical potential. However, due to limited data and the heterogeneous population, no potential biomarkers of DV responder are currently available [8, 12, 13]. In real-world scenarios, to what extent DV can benefit previously treated mUC patients with complex medication history, and the very characteristics that exceptional responders to DV might harbour remain pending issues that desire to be confirmed, not only for current clinical practice but also for future trial design regarding the combination of ADCs and standard care [14].

In this retrospective observational study, we portrayed the DV treatment in mUC patients in real-world scenarios with multi-institutional data. By collecting data from the electronic health record and profiling the genomic landscape, we explored the putative clinical and genomic correlates of clinical benefit from DV, which might foster future patient stratification and bring insights into the development of ADC drugs.

Materials and methods

Patient inclusion

From April 2021 to March 2024, we searched for and included the UC patients who were prescribed for DV treatment, with primary sites including the renal pelvis, ureter, and bladder. This study involves three independent institutions: Fudan University Shanghai Cancer Center (FUSCC), Zhongshan Hospital Fudan University (ZSHS), and Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine (XHHS). We identified a total of 37 patients who had received DV treatment, as documented in the Electronic Health Records (EHR). However, five patients were excluded due to the absence of essential follow-up data. Additionally, two patients were also excluded because they were not presented with metastatic disease at the time of their DV treatment. Consequently, 30 patients with metastatic urothelial carcinoma confirmed with radiology findings or biopsy were ultimately included in our analysis (Supplementary Fig. 1A). All the patients were evaluated by the immunohistochemistry assay or the next-generation sequencing via the fresh-frozen or formalin-fixed, paraffin embedded tissue.

Endpoint evaluation and follow-up process

As a retrospective, real-world, observational study, treatment efficacy, progression-free survival (PFS), and overall survival were included in this investigation as primary study endpoints. Treatment efficacy was evaluated based on the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 [15] by board-certified radiologists retrospectively, based on the computed tomography (CT) or magnetic resonance imaging (MRI) scan generated during the evaluation of patients.

All the discharged patients have been followed up via telephone to confirm their vital status. PFS and OS were defined as the duration from the start of DV treatment to the first evidence of disease progression, death, or the latest follow-up of enrolled patients. Previous treatment history, treatment-related adverse events (TrAEs), and demographic features such as age and gender were also collected for each patient using EHR data. Details of the clinicopathological characteristics of patients enrolled have been presented in Table 1.

Table 1.

Baseline characteristics of mUC patients enrolled in this study

Variables	Number of Patients		Variables	Number of Patients
Age			Best confirmed overall response
	≥ 65	21 (70.0%)		CR	3 (11.5%)
	< 65	9 (30.0%)		PR	8 (30.8%)
Gender				SD	8 (30.8%)
	Male	19 (63.3%)		PD	7 (26.9%)
	Female	11 (36.7%)		NA	4
Surgery preceding DV treatment			Events
Radical Cystectomy/Radical nephroureterectomy		20 (66.7%)	Death		6 (20.0%)
	TURBT/Diagnostic cystoscopy	10 (33.3%)	Disease progression		15 (50.0%)
Metastatic site			TrAEs
Lung/Liver		6 (20.0%)	PNS symptoms (Hypoesthesia, paresthesia, and alopecia)		13 (43.4%)
Pelvic/retroperitoneal lymph nodes		19 (63.3%)	Elevated alanine transaminase		1 (3.3%)
Other distant metastasis sites		5 (16.7%)	Not present		16 (53.3%)
HER-2 IHC			Previous treatment
	1+	6 (20.0%)		Chemotherapy	3 (10.0%)
	2+	10 (46.7%)	Checkpoint inhibitors		5 (16.7%)
	3+	4 (13.3%)		Both	19 (63.3%)
	NA	6 (20.0%)	Treatment-naive		3 (10.0%)

CR Complete response, PR Partial response, SD Stable disease, PD Progressive disease, NA Not applicable, PNS Peripheral nervous system symptoms.

Immunohistochemistry evaluation of HER-2

For HER-2 evaluation, the Ventana anti-HER-2/Neu (4B5) rabbit monoclonal antibody and ultra-View Universal DAB Detection Kit (Roche) were used for staining and testing [16], and the protocol has been described previously [17]. Two experienced and independent pathologists reviewed the IHC scores according to the 2023 edition of the American Society of Clinical Oncology (ASCO)/College of American Pathologists (CAPs) guidelines in breast cancer [18].

Next-generation sequencing

Patients from the XHHS and the FUSCC cohort received next-generation sequencing to evaluate possible genomic correlates of clinical benefit from DV treatment. For patients from the XHHS cohort, the samples were sequenced via the 1123-gene ChosenOne™ panel by ChosenMed Technology (Beijing) Co., Ltd. This panel’s detailed sample preparation and quality control process has been described previously [19]. For patients from the FUSCC cohort, the FUSCC Genitourinary Cancer sequencing panel was used for somatic mutation analysis in fresh-frozen or FFPE tissue. This panel was aimed to detect mutations, small insertions and deletions in 618 genes, copy-number alterations in 270 genes, and structural variants in 27 genes. The detailed protocol has been previously described [20].

All the evaluated tissue were surgery specimens from radical surgery or transurethral resection of bladder tumor (TURBT), except for one patient with a biopsy specimen (PT-DV003) and one patient with a cell-free DNA sample (PT-DV015). Seventeen patients finally received corresponding genomic profiling. And only the samples from surgery and biopsy (n = 16) were included for final analysis. A tumor mutation burden above 10.0 mutations per megabase is considered as TMB^high, as a widely accepted practice [21]. A list of the detailed data needed to reproduce the findings in this study has been included in the Supplementary Table1, including the overall and progression-free survival time, the demographic data, and the somatic mutations of each patient.

Statistical analysis

For survival analysis, we used Kaplan-Meier analysis by performing a Log-rank test to explore the potential correlates of clinical benefit from DV treatment. The multivariable Cox regression model was also used to validate TMB as an independently correlated with response to DV, and visceral metastasis and addition of ICI were included in the analysis, for they might be key confounding factors in mUC patients [22]. Given the limited sample size, the Firth’s penalized partial likelihood correction was applied to the Cox regression model using the R package coxphf, as previously described [23, 24]. The ORR was calculated as the combined total of complete response (CR) and partial response (PR) rates. The disease control rate (DCR) was calculated as the percentage of cases showing CR, PR, and stable disease (SD) among the evaluable cases. The comparison of response rate to DV treatment between two groups is done by Fisher’s exact test for the difference in the proportion of patients presented with CR or PR between each group. The Clopper-Pearson exact method determines the 95% confidence interval for ORR and DCR [25]. A P < 0.05 was considered statistically significant in our study. All the statistical analyses were performed using IBM SPSS Statistics V.22.0 and R V.4.3.2 (http://www.rproject.org/).

Results

Characteristics of enrolled patients and overall outcome of DV treatment

Following patient inclusion and data quality control, thirty mUC patients were included in this study (Supplementary Fig. 1 A). They included 63.3% of male (n = 19) and 36.7% of female (n = 11) patients, with a median age of 67.5 years. Before the initiation of DV treatment, 90.0% of all the patients (n = 27) received at least one failed course of systemic therapy due to disease progression, with 10.0% (n = 3) for platinum-based chemotherapy, 16.7% (n = 5) for checkpoint blockade, and 63.3% (n = 19) for both (Table 1). 56.7% of patients (n = 17) received concurrent checkpoint blockade in addition to DV treatment (Fig. 1A).

Fig. 1.

Overall Treatment Outcome of DV in mUC patients. A Swimmer plot demonstrating all the treatment, disease and vital status of patients receiving DV treatment. B Kaplan-Meier plot demonstrating the overall estimate of progression-free survival and overall survival following DV treatment, with a corresponding 95% confidence interval. CR, complete response. PR, partial response. SD, stable disease. PD, progressive disease. NE, not evaluated

According to EHR and follow-up data, we evaluated the treatment toxicity, efficacy, and long-term implications following DV treatment. During the patient inclusion since February 2021, the median follow-up time was 12.1 months. 66.7% of patients (n = 20) quit DV treatment either because of progressive disease or toxicity (Fig. 1A). The most prevalent treatment-related adverse events (TrAEs) of DV treatment are peripheral nervous symptoms, such as paresthesia and alopecia (43.4%, n = 13, Table 1). Among 26 patients with adequate radiographic data, the ORR was 42.3%, with the 95% confidence interval (CI) between 23.4% and 63.1%, and the disease control rate (DCR) was 73.1% (95% CI: 55.2%−88.4%) (Table 1). Three patients exhibited complete responses (CR). Our follow-up data reported the median PFS as 12.2 months (95% CI: 6.6–17.7 months), and the median OS was not reached following DV treatment, with a 12-month OS rate of 86.7% (Fig. 1B).

Higher HER-2 expression levels and previous history of response to platinum-based chemotherapy are correlated with benefits from DV treatment

For regular clinicopathological factors, we found higher HER-2 expression (IHC 2 + and IHC 3+) was primarily associated with prolonged PFS and response rate compared with patients of HER-2 IHC 1+ (Fig. 2A. Log-rank P = 0.082. Median PFS: 13.1 vs. 4.6 months. Response rate: 60.0% vs. 0.0%. Fisher’s exact test, P = 0.019). Moreover, for patients receiving platinum-based chemotherapy with radiologically evaluable target lesions, the initially successful response can be correlated with better outcomes from DV. If the patients showed durable relief (CR or PR) from chemotherapy, they may also again benefit more from DV treatment (Fig. 2B. Log-rank P = 0.086. Median PFS: 14.2 vs. 3.6 months. Response rate: 50.0% vs. 10.0%. Fisher’s exact test, P = 0.176).

Fig. 2.

Clinicopathological correlates of DV treatment(A-B) Kaplan-Meier plot and percentage bar plot illustrating the predictive value of HER-2 immunohistochemistry level and previous outcome in platinum-based chemotherapy for the following result of DV treatment. C A representative patient significantly responding to DV (PT-DV001), with previous response to neoadjuvant chemotherapy. Arrows indicating the compared lesions. Dx, diagnosis. TURBT, transurethral resection of bladder tumor. PET-CT, positron emission tomography

PT-DV001, for example, previously received neoadjuvant chemotherapy and was assessed with partial response on initial evaluation. However, even if the end-stage patient showed progressive disease following checkpoint blockade and the DV treatment was prematurely terminated due to peripheral neuropathy, the patient exhibited significant remission at the metastatic site of the lymph node at the obturator internus after the brief administration of DV. Long-term follow-up showed that the patient managed to sustain a disease-free status for a duration rarely seen in mUC patients (Fig. 2C).

Higher TMB is closely related to better outcomes from DV treatment

For patients with adequate genomic data (n = 16), we found higher TMB (above 10.00 Mut/Mb) was significantly correlated with longer PFS (Fig. 3A, Log-rank P = 0.011. Median PFS: 12.8 vs. 4.6 months) and better response rate (Fig. 3B. Response rate: 71.4% vs. 11.1%. Fisher’s exact test, P = 0.035). Simultaneously, we found patients with durable clinical benefits either exhibiting higher TMB or expressing more neoantigen (Fig. 3C). The only exception of the TMB^low patients presenting good clinical response might be attributed to the fact that the sample for TMB evaluation was cell-free DNA, instead of the resected sample (PT-DV015) [26]. Multivariable Cox regression with Firth’s penalized likelihood correction validated that higher TMB is correlated with lower risk of disease progression, independent of putative confounding factors like the addition of checkpoint blockade or visceral metastasis (Fig. 3D. Multivariable Cox regression, P = 0.015. Hazard Ratio: 0.165. 95% CI: 0.026–0.717).

Fig. 3.

Genomic correlates of DV treatment. A Kaplan-Meier plot illustrating the progression-free survival of patients with different TMB levels. B Percentage bar plot demonstrating the different response rate of patients with different TMB. C Scatter plot showing the relationship between progression-free survival, neoantigen burden, best confirmed overall response to DV, and tumor mutation burden. D Forest plot demonstrating the multivariable Cox regression model validating TMB is independently correlated with clinical benefit from DV. E An exceptional responder to DV treatment with high TMB. Representative image of MRI scan showing changes in tumor volume before and after DV treatment. TURBT, transurethral resection of bladder tumor. Arrows indicating the compared lesions

As a typical case, a patient (PT-DV005), with a TMB of 13.30 (Mut/Mb), was presented with multiple metastases (Fig. 3E & Supplementary Fig. 4), and the patient refused to receive chemotherapy. However, following an initial challenge from DV, the patient was assessed as complete response (Fig. 3E & Supplementary Table1). The exceptional responder previously mentioned (PT-DV001) was also confirmed with a high TMB of 14.53 Mut/Mb (Fig. 2C & Supplementary Table1).

Discussion

Like DV in China, the FDA has approved EV plus pembrolizumab for the treatment of patients with locally advanced or metastatic urothelial carcinoma who are ineligible for cisplatin-based chemotherapy, with its revolutionary value for mUC management [27]. As ADCs are endowed with great clinical potential and are bound to alter the future management of cancer, it would be pivotal to comprehensively evaluate the patients before making treatment decisions. While it is established that HER2 expression predicts response to HER2-targeted therapies and that TMB is associated with response to immunotherapy [28, 29], our study uniquely provides real-world evidence that both biomarkers may also inform clinical benefit from DV in the treatment of mUC. Notably, our analysis identifies TMB as a potential predictor of DV efficacy—an association that has not been extensively explored in the context of ADCs for mUC. This finding raises the possibility that TMB, beyond its role in predicting immunotherapy response, may reflect broader tumor-intrinsic properties relevant to ADC sensitivity. Moreover, by evaluating these biomarkers in a real-world setting outside the constraints of clinical trial selection, our study adds clinically meaningful insights into how HER2 expression and TMB could be used to optimize patient selection and personalize treatment strategies in a population with limited therapeutic options.

Initially, we reported the overall clinical outcome of DV treatment in the general real-world mUC population. The general efficacy and safety profile was generally satisfactory and was similar to previous trials and retrospective analyses of DV in mUC, with an ORR of 42.3% and median PFS of 12.2 months [8, 16, 30, 31]. Moreover, this study’s key finding and purpose is the explorative analysis of potential biomarkers of DV. Similar to findings in biomarker analysis of phase II and III trials in successful ADCs like Enfortumab Vedotin and Trastuzumab Deruxtecan [32, 33], we found higher expression of HER-2 by IHC assay was associated with prolonged progression-free survival. Therefore, the traditionally HER-2 low (IHC 0&1+) patients may not be suitable DV candidates, as was suggested by the two recent trials [8].

In addition, the combination of DV with checkpoint inhibitors may not necessarily improve the overall efficacy of DV treatment, as was indicated in a previous retrospective study by Chen et al. [30]. In our study, 56.7% of patients received DV in combination with pembrolizumab. We acknowledge that, under such circumstances, TMB could plausibly reflect sensitivity to the checkpoint inhibitor rather than to the antibody-drug conjugate itself. Given the retrospective nature of our study and the limited sample size, we are underpowered to statistically disentangle the individual contributions of DV and pembrolizumab to clinical outcomes. Nonetheless, we have revised the Discussion to reflect this ambiguity in interpretation.Importantly, we also observed that the association between TMB and clinical benefit remained directionally consistent—even among patients not receiving concurrent checkpoint blockade—though the subgroup size was small. This raises the possibility that TMB may also reflect tumor-intrinsic features, such as genomic instability or increased neoantigen burden, that enhance sensitivity to antibody-drug conjugates independently of immune checkpoint activity [34]. Interestingly, based on EHR data, we found previous successful outcomes from standard chemotherapy may herald ideal outcomes following DV treatment. Both DV and chemotherapy function as cytotoxic therapies by nature, so consistent genomic determinants and microenvironmental milieu, which are vital to the outcome from chemotherapy, might have led to success in the second challenge of DV [35]. We hypothesize that a prior favorable response to platinum-based chemotherapy may serve as a surrogate marker for underlying tumor biology—such as a more chemosensitive phenotype or a more active tumor immune microenvironment. These factors could similarly enhance responsiveness to DV, which exerts its effect through targeted cytotoxic payload delivery [36]. Moreover, chemotherapy-responsive tumors may exhibit higher proliferative rates, intact apoptotic pathways, or greater permeability to cytotoxic agents—all of which could also potentiate the efficacy of antibody-drug conjugates. Supporting this, prior studies have shown that patients who respond well to systemic chemotherapy often have tumor microenvironment features that remain permissive to subsequent therapeutic interventions, including ADCs.

Most importantly, to our knowledge, this study marks the first study investigating the genomic correlates of clinical benefits from DV through matched sequencing data. We found that TMB was closely related to clinical outcomes from DV treatment. As TMB has served as a robust standard for the immunogenicity of tumor cells and a biomarker for immune checkpoint blockade [21], its predictive power carried over into the scenario of ADC treatment. This conclusion is also similar to a recent multi-center retrospective analysis of the efficacy of Enfortumab Vedotin in real-world mUC patients (the UNITE study) [37]. Moreover, preclinical data by Huang et al. and Wu et al. suggested that DV treatment can effectively activate the cGAS-STING pathway and potentiate the response to immunotherapy in vitro and in vivo, simultaneously forming immunological memory and preventing further tumor recurrence [38, 39]. Hence, it would be reasonable to use TMB as a surrogate for DV recipients to evaluate its immunological fragility, by which DV challenge could elicit a secondary immune response through cytotoxic machinery.

As a retrospective, real-world analysis, our cohort size was constrained by the availability and completeness of both clinical and genomic data, which inevitably impacted the statistical power and generalizability of our findings. The limitation of our study also lies in the missing data during the collection of patients, and probably the consequent restricted potential for extrapolation. The statistical power is also limited, given the small patient population. As a real-world study, the evaluation of disease relief and progression might also be subject to personal bias and missing data, which may account for a slight difference in our data from the result of high-volume clinical trials regarding overall and progression-free survival [31]. Moreover, key factors such as PD-L1 data, transcriptome data and ECOG performance status should also be considered when such data are available [22]. Therefore, large-scale and comparative clinical sampling of DV recipients is still necessary to validate our findings and to understand the potential immunological and genomic dynamics underlying the course of DV treatment. However, this study is strengthened by detailed and matched genomic, pathologic, and clinical data, which effectively showed that TMB may be able to predict better outcomes from DV treatment in the future.

In conclusion, our investigation explored putative biomarkers for DV responders, such as higher HER-2 expression, preceding success in chemotherapy, and higher TMB. Our data, therefore, provided both clinicians and industry with further evidence and insights regarding the patient stratification and pharmacodynamics of novel applications of DV treatment.

Abbreviations

HER-2

Human Epithelial Growth Factor Receptor 2

mUC

Metastatic urothelial carcinoma

ICI

Immunocheckpoint inhibitors

mPFS

Median progression-free survival

ADC

Antibody-drug conjugate

EV

Enfortumab Vedotin

DV

Disitamab Vedotin

MMAE

Monomethyl auristatin E

ORR

Objective response rate

DCR

Disease control rate

FDA

Food and Drug Administration

FUSCC

Fudan University Shanghai Cancer Center

ZSHS

Zhongshan Hospital Fudan University

XHHS

Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine

EHR

Electronic health record

RECIST

Response Evaluation Criteria in Solid Tumors

CT

Computed tomography

MRI

Magnetic resonance imaging

TrAEs

Treatment-related adverse events

CR

Complete response

PR

Partial response

SD

Stable disease

PD

Progressive disease

IHC

Immunohistochemistry

TMB

Tumor mutation burden

Authors’ contributions

Z. Liu, K. Jin and Z. Xu for acquisition of data, analysis and interpretation of data, statistical analysis and drafting of the manuscript; H. Zeng, X. Su, J. Xu, Y. Ding, H. Liu, and Y. Zhu for technical and material support; J. Xu, Z. Wang, and Y. Chang for study concept and design, analysis and interpretation of data, drafting of the manuscript, obtained funding and study supervision. All authors read and approved the final manuscript.

Funding

This study was funded by grants from National Natural Science Foundation of China (82272786, 82272930, 82372793, 82373276), China Postdoctoral Science Foundation (BX20230091, 2024M760547), Shanghai Municipal Natural Science Foundation (22ZR1413400, 23ZR1440300, 23ZR1411700, 24SF1901304), Shanghai Sailing Program (24YF2705400, 24YF2706000), Chenguang Program of Shanghai Education Development Foundation and Shanghai Municipal Education Commission (24CGA07), Shanghai Municipal Health Bureau Project (202340127, 20234019, 20224Y0232), Beijing Xisike Clinical Oncology Research Foundation (Y-Gilead2024-PT-0098), Fudan University Shanghai Cancer Center for Outstanding Youth Scholars Foundation (YJYQ201802), Shanghai Anticancer Association EYAS PROJECT (SACA-CY22B02, ZYJH202309, SACA-CY24C10), Youth Foundation of Zhongshan Hospital affiliated with Fudan University (ZSZP202405), Chongqing Postdoctoral Program for Innovative Talents (CQBX202426) and Key Project of Young Doctor Incubation Program of the Second Affiliated Hospital of Army Medical University (2024YQB024). All these study sponsors have no roles in the study design, collection, analysis, or interpretation of data.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study was approved by the Clinical Research Ethics Committee of Zhongshan Hospital and Fudan University (No. B2024-085). Written informed consent was obtained from each patient.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.