Safety and efficacy of intralesional rituximab injection versus involved site radiation therapy in primary ocular adnexal MALT lymphoma: study protocol for a multicentre randomised controlled trial Authors: Yaoming Liu, Yang Gao, Ling Jin, Qian Li, Xiaohui Wang, Yujie Wu, Jichao Zhou, Wei Liu, Lili Wang, Yuying Dong, Yajing Zheng, Lu Chen, Jinmiao Li, Shicai Su, Junjie Tang, Yinghao Wang, Weifeng Huang, Chao Cheng, Cheng Li, Ningyu An, Jia Tan, Hong Chen, Huiyu Guo, Yi Wang, Qing Zhou, Fei Han, Honglei Liu, Yu Zhu, Rong Lu Published: 2024 ABSTRACT AbstractIntroductionInvolved site radiation therapy (ISRT) is a widely used treatment for primary ocular adnexal mucosa-associated lymphoid tissue (MALT) lymphoma with control rates over 85%; however, its complications could reach 30%–50%. Intralesional rituximab injection has shown promising results with remission rates between 65% and 100% and minimal complications. This study aimed to conduct a multicentre randomised controlled clinical trial to compare the safety and efficacy of intralesional rituximab injection versus ISRT on primary ocular adnexal MALT lymphoma.Methods and analysisThis is a multicentre randomised controlled trial and will be conducted at eight ophthalmic departments and institutes in China. 108 patients with primary ocular adnexal MALT lymphoma will be enrolled and randomised to undergo intralesional rituximab injection or ISRT. Comprehensive examinations will be performed before and after the treatment. The primary outcome is the cumulative occurrence rate of complications of grade ≥2 within 5 years after treatment commencement. The main secondary outcome is overall response rate, and progression-free survival, time to next treatment and overall survival will also be compared between the groups.Ethics and disseminationEthical approval has been obtained from the Ethics Committee of the Zhongshan Ophthalmic Center (ID: 2023KYPJ125-3; V.20240818) and the ‘5010 Plan’ evaluation committee at Sun Yat-sen University, Guangzhou, China, and all subcentres. All the participants will be required to provide written informed consent. The results will be disseminated through scientific meetings and published in peer-reviewed journals.Trial registration numberNCT06190301. STRENGTHS AND LIMITATIONS OF THIS STUDY In this randomised controlled trial, we aim to assess a treatment option that has fewer complications and comparable therapeutic effects for ocular adnexal mucosa-associated lymphoid tissue (MALT) lymphoma. In this randomised controlled trial, we aim to assess a treatment option that has fewer complications and comparable therapeutic effects for ocular adnexal mucosa-associated lymphoid tissue (MALT) lymphoma. This study is a multicentre study that will speed up enrolment and the conclusions will be of wide applicability. This study is a multicentre study that will speed up enrolment and the conclusions will be of wide applicability. The study is limited by only including Asian ethnicity and patients with stages T1-3 of ocular adnexal MALT lymphoma. The study is limited by only including Asian ethnicity and patients with stages T1-3 of ocular adnexal MALT lymphoma. Introduction Mucosa-associated lymphoid tissue lymphoma (MALT lymphoma) is, according to the definition of the WHO, a unique type of lymphoma, accounting for 7%–8% of newly diagnosed lymphomas.1 2 It is more common in women, with a median diagnosis age of 65.3 Ocular adnexa are the most common primary site for non-gastric MALT lymphoma.4 In the Asian region, MALT lymphomas account for up to 80%–90% of ocular adnexal lymphomas,5 6 much higher than the proportion of MALT lymphomas in other parts of the body. Ocular adnexal MALT lymphomas can invade the conjunctiva, orbit, lacrimal gland, lacrimal sac and eye muscles. Clinically, they often manifest as painless tumours, with varying presentations depending on the lesion location. They lack specificity, and diagnosis mainly relies on histopathological examination.1 MALT lymphoma is a type of indolent tumour. Most progress slowly, with a 5-year overall survival rate over 90%, and a 10-year survival rate of 75%–80%. It can relapse after treatment, affecting the same organ (in 60% of cases) or other extranidal sites.1 However, a few cases may spread outside the orbit or transform into high-grade invasive lymphomas, affecting the patient’s life. Due to the painless clinical course of MALT lymphoma and the high remission rate of traditional treatments, modern treatment strategies focus on ‘non-toxic’ methods to minimise side effects while maintaining efficacy. However, for the treatment of ocular adnexal MALT lymphoma, there are currently no clear guidelines or consensus. Clinicians mainly adopt individualised treatments based on different locations and extents, including surgical excision, radiotherapy, chemotherapy, monoclonal antibody immunotherapy and antibiotic treatment.5711 The primary purpose of surgical excision is to obtain tumour samples for pathological examination and to make a clear diagnosis, rather than aiming for complete removal. This is because complications and organ dysfunction from eye tumour removal surgery can be more destructive than the lymphoma itself.5 Research has shown that the extent of MALT lymphoma surgical removal does not affect survival rates, and extensive excisions that damage surrounding tissues should be avoided.12 Hence, surgical excision alone is insufficient to thoroughly treat MALT lymphoma. Involved site radiation therapy (ISRT) in the treatment of ocular adnexal MALT lymphoma is the most extensively researched and used method, with local control rates exceeding 85%.5 13 However, due to the uniqueness of the eye, complications from radiotherapy are as high as 30%–50%, including cataracts, corneal ulcers, dry eye disease, conjunctivitis, radiation retinopathy and radiation optic neuropathy. These can lead to significant vision impairment and dysfunction in patients.1416 The most common complication of ISRT is radiation-induced cataract. Cataract usually developed at 2–5 years after irradiation, and the risk of cataract was found to increase by 15% per year.17 18 Some reports also indicate that up to 40% of patients with ocular adnexal MALT lymphoma who underwent radiation therapy experienced recurrence on the opposite side or at a distant site.1 Thus, the focus of research on the treatment of ocular adnexal MALT lymphoma is to maintain a high remission rate while minimising the toxic side effects of treatment. Rituximab is one of the most used monoclonal antibodies in patients with lymphoma, significantly improving the survival rate of patients with lymphoma.6 It is a chimeric human/mouse anti-CD20 antibody, acting on B lymphocytes. It binds to the CD20 antigen on both normal cells and tumour cells. The CD20 antigen is involved in the proliferation, activation, differentiation and signal transduction of B cells. According to literature reports, local injection of rituximab for the treatment of ocular adnexal MALT lymphoma has shown good therapeutic effects, with remission rates ranging from 65% to 100%,1923 and almost no serious complications were observed. However, the number of relevant studies is limited, and the number of cases in individual studies currently does not exceed 20, all of which are single centre and most of which are retrospective studies, and almost none have established control groups. Only one single-centre prospective trial exists in this field,21 and these studies are focused exclusively on conjunctival lymphomas.1923 Therefore, the local application of rituximab has shown promising prospects, but higher-level evidence is still needed. This project proposes to establish a prospective, multicentre, randomised, controlled clinical trial, including nine domestic centres, to compare the safety and efficacy of intralesional rituximab injection versus ISRT for the treatment of primary ocular adnexal MALT lymphoma. The aim is to provide high-level clinical evidence for the treatment of ocular adnexal MALT lymphoma and to offer patients treatment options that have fewer complications and comparable therapeutic effects. Methods and analysis Objectives The trial aims to compare the safety and efficacy of intralesional rituximab injection versus ISRT in the treatment of primary ocular adnexal MALT lymphoma, to offer patients treatment options that have fewer complications and comparable therapeutic effects. Design and setting This study adopts a multicentre, central randomisation, open-label, positive parallel control design. The study will be conducted at Zhongshan Ophthalmic Center (ZOC), Sun Yat-sen University as the principal centre, in joint partnership with six other hospitals and eye institutes, including Xi'an People’s Hospital (Xi'an Fourth Hospital), The First Affiliated Hospital of Zhengzhou University, Peking University Third Hospital, Hainan Eye Hospital, The First Affiliated Hospital of Jinan University and People’s Hospital of Ningxia Hui Autonomous Region as subcentres. The trial settings are presented in figure 1. The actual start date for the study is 11 January 2024, and the end date is anticipated to be 31 December 2029. Study flow chart. Eligibility Inclusion criteria Age between 18–75 years old. Age between 18–75 years old. Ocular adnexal MALT lymphoma with a comprehensive diagnosis based on pathology, clinical manifestations and biological characteristics according to the WHO classification of tumours of haematopoietic and lymphoid tissues (2017).24 25 Ocular adnexal MALT lymphoma with a comprehensive diagnosis based on pathology, clinical manifestations and biological characteristics according to the WHO classification of tumours of haematopoietic and lymphoid tissues (2017).24 25 Based on the tumour, node, metastasis (TNM) staging of ocular adnexal lymphoma (table 1), patients with stages T1-3 of ocular adnexal MALT lymphoma without involvement of intraconal compartment are included. Based on the tumour, node, metastasis (TNM) staging of ocular adnexal lymphoma (table 1), patients with stages T1-3 of ocular adnexal MALT lymphoma without involvement of intraconal compartment are included. Willing to participate the trial and sign the informed consent form. Willing to participate the trial and sign the informed consent form. Table 1Definitions of TNM for ocular adnexal lymphoma, AJCC Cancer Staging Manual, eighth editionPrimary tumour (T)TXLymphoma extent not specifiedT0No evidence of lymphomaT1Lymphoma involving the conjunctiva alone without eyelid or orbitalT2Lymphoma with orbital involvement with or without conjunctival involvementT3Lymphoma with preseptal eyelid involvement with or without orbital involvement and with or without conjunctival involvementT4Orbital adnexal lymphoma and extraorbital lymphoma extending beyond the orbit to adjacent structures, such as bone, maxillofacial sinuses and brain.Regional lymph nodes (N) NXInvolvement of lymph nodes not assessed N0No evidence of lymph node involvement N1Involvement of lymph node region or regions draining the ocular adnexal structures and superior to the mediastinum (periauricular, parotid, submandibular and cervical nodes) N1aInvolvement of a single lymph node region superior to the mediastinum N1bMetastasis in a single ipsilateral regional lymph node based on lymph node biopsy N2Metastasis in a single ipsilateral regional lymph node, >3 cm in greatest dimension, or in bilateral or contralateral lymph nodes N3Diffuse or disseminated involvement of peripheral and central lymph node regionsDistant metastasis (M) M0No evidence of involvement of other extranodal sites M1aNoncontiguous involvement of tissues or organs external to the ocular adnexa (eg, parotid glands, submandibular gland, lung, liver, spleen, kidney, breast) M1bLymphomatous involvement of the bone marrow M1cBoth M1a and M1b involvementAJCCAmerican Joint Committee on Cancer Definitions of TNM for ocular adnexal lymphoma, AJCC Cancer Staging Manual, eighth edition AJCCAmerican Joint Committee on Cancer Exclusion criteria Based on the TNM staging of ocular adnexal lymphoma (table 1), patients staged as T2/T3 with involvement of intraconal compartment and patients staged as T4. Based on the TNM staging of ocular adnexal lymphoma (table 1), patients staged as T2/T3 with involvement of intraconal compartment and patients staged as T4. Patients who have previously received local or systemic radiation, chemotherapy or drug treatment specifically for ocular adnexal MALT lymphoma. Patients who have previously received local or systemic radiation, chemotherapy or drug treatment specifically for ocular adnexal MALT lymphoma. Presence of cataract and is anticipated to require surgical treatment within a certain period after enrolment; existing cataract affects visual field testing and fundoscopic examination; vision affected by cataract is <20/40. Presence of cataract and is anticipated to require surgical treatment within a certain period after enrolment; existing cataract affects visual field testing and fundoscopic examination; vision affected by cataract is <20/40. In addition to the need for treatment of ocular adnexal MALT lymphoma, there is a requirement for other ocular procedures (eg, full-thickness corneal transplant or retinal surgery) or an anticipated need for another emergent ocular surgery. In addition to the need for treatment of ocular adnexal MALT lymphoma, there is a requirement for other ocular procedures (eg, full-thickness corneal transplant or retinal surgery) or an anticipated need for another emergent ocular surgery. Complicated with other ocular diseases: including corneal abnormalities or existing corneal infections, iridocorneal endothelial syndrome, anterior segment dysgenesis, true microphthalmos, uveitis, glaucoma, ocular trauma and retinal disorders such as central retinal vein occlusion, central retinal artery occlusion, retinal detachment and so on. Complicated with other ocular diseases: including corneal abnormalities or existing corneal infections, iridocorneal endothelial syndrome, anterior segment dysgenesis, true microphthalmos, uveitis, glaucoma, ocular trauma and retinal disorders such as central retinal vein occlusion, central retinal artery occlusion, retinal detachment and so on. The statue of hepatitis B virus (HBV) or human immunodeficiency virus (HIV) infection. The statue of hepatitis B virus (HBV) or human immunodeficiency virus (HIV) infection. Need for long-term use of local or systemic steroids. Need for long-term use of local or systemic steroids. Patients already enrolled in other drug clinical trials. Patients already enrolled in other drug clinical trials. Pregnant or breastfeeding women. Pregnant or breastfeeding women. Serious systemic diseases: advanced cardiac disease, kidney disease, respiratory disease, or other malignant tumours. Serious systemic diseases: advanced cardiac disease, kidney disease, respiratory disease, or other malignant tumours. Inability to understand the research content. Inability to understand the research content. If both eyes of a patient meet the inclusion criteria, the right eye will be selected for participation in the study, while the left eye will receive the same treatment, but its data will not be included in this research. Each patient can have only one eye participating in the study. If both eyes of a patient meet the inclusion criteria and the patient is randomised to the ISRT group, the eye with the larger lesion or with more symptomatic disease is treated first, and in the case of well tolerability, the other eye is irradiated. Recruitment This trial is a multicentre study. All subcentres are tertiary hospitals in China and could ensure adequate participants. First screening will be conducted in the outpatient clinics at all subcentres. After confirming the eligibility, the trial will then be proceeded. Allocation of interventions A central randomisation system will be used to generate a random allocation sequence with a fixed block size. The subjects in each block unit will be evenly assigned (1:1) to the control group (ISRT) and the experimental group (intralesional rituximab injection). The random sequence will be generated by an independent person who will not be involved in the implementation of the trial. All subjects will be randomly assigned in order of enrolment according to the random list. Masking This trial involves localised drug treatment or radiation intervention, making it impossible to implement a blinded approach. Therefore, an open-label design is adopted. Technicians and other persons who are responsible for screening and outcome assessment are unaware of the group assignment throughout the entire study period. The trial statisticians will be blinded to the randomisation results until the completion of statistical analysis. Intervention methods All surgeries will be performed by professors or experienced senior attending doctors at each subcentre. Intervention group (intralesional rituximab injection)1923 After a pathological tissue biopsy confirms positive margins, local treatment with rituximab is administered. The lesion is injected with rituximab once a week for a total of four times, followed by an injection once a month for a total of five times. If the tumour still has residue after 6 months, continue with monthly injections for up to a year. After a pathological tissue biopsy confirms positive margins, local treatment with rituximab is administered. The lesion is injected with rituximab once a week for a total of four times, followed by an injection once a month for a total of five times. If the tumour still has residue after 6 months, continue with monthly injections for up to a year. Each dose includes 1–3 mL of undiluted rituximab (10 mg/mL). Each dose includes 1–3 mL of undiluted rituximab (10 mg/mL). Before administering the drug to the lesion, the eyelid edge is disinfected with a 5% iodopovidone gauze, and local anaesthesia is applied using 2 drops of 0.4% Novesina. Small volumes of the solution are then repeatedly injected into visible or image-indicated lesions using a 30-gauge needle. Care is taken not to inject the solution into the blood vessels. The entire procedure lasts about 10 min. Before administering the drug to the lesion, the eyelid edge is disinfected with a 5% iodopovidone gauze, and local anaesthesia is applied using 2 drops of 0.4% Novesina. Small volumes of the solution are then repeatedly injected into visible or image-indicated lesions using a 30-gauge needle. Care is taken not to inject the solution into the blood vessels. The entire procedure lasts about 10 min. Tobramycin eye drops are applied, and used three times a day, continuously for 4 days, followed by bandaging for 1 day. Tobramycin eye drops are applied, and used three times a day, continuously for 4 days, followed by bandaging for 1 day. See details in online supplemental file 1. Control group (ISRT) After the pathological tissue biopsy confirms positive margins, ISRT is used. The total dosage is 24–30 Gy, divided into 12–15 fractions. The treatment is given continuously for 4–5 days, one time a day, followed by a 2-day break; then it continues for another 4–5 days, one time a day, with a 2-day break; and then for another 4–5 days, one time a day, until the treatment course is completed.5 13 26 See details in online supplemental file 1. Outcome measurement Primary outcome The cumulative incidence of ≥grade 2 complications within 5 years after treatment commencement: Grade 0: none. Grade 1: asymptomatic cataracts; mild keratitis; mild local pain; mild fever; mild allergic/inflammatory reactions of the skin/conjunctiva. Grade 2: symptomatic cataracts; moderate keratitis or corneal ulcers; minor retinal lesions or glaucoma; moderate allergic/inflammatory reactions of the skin/conjunctiva with or without keratitis, requiring corticosteroids and/or antibiotics; dry eye requiring artificial tears; iritis with photophobia; moderate pain; moderate fever. Grade 3: cataract requiring surgery; severe keratitis or corneal ulcers; severe retinal lesions or retinal detachment; severe optic nerve pathology; severe glaucoma; severe allergic/inflammatory reactions of the skin/conjunctiva; severe pain; high fever, requiring hospitalisation or extended hospital stay. Grade 4: panophthalmitis; blindness. Grade 5: death directly resulting from treatment. If a patient experiences treatment-related complications not mentioned above, they will be graded according to the WHO’s general criteria for grading treatment toxicity (Common Terminology Criteria for Adverse Events v5.0): Grade 1: mild; asymptomatic or minor; clinical or diagnostic observations only; no treatment required. Grade 2: moderate; requires minor, local, or non-invasive treatment; limitations on instrumental activities of daily living (ADL) appropriate to age. Grade 3: severe or of significant medical importance but not immediately life-threatening; leading to hospitalisation or extending hospital stay; causing disability; limitations on self-care ADL. Grade 4: life threatening; requires emergency treatment. Grade 5: death associated with AE. ADL: Instrumental ADL refer to activities such as cooking, buying clothes, using the telephone and managing finances. Self-Care ADL refer to activities like bathing, dressing and undressing, eating, washing, taking medication and not being bedridden. Secondary outcomes Primary efficacy evaluation indicator: Overall response rate. Overall response rate. Secondary efficacy evaluation indicators: Progression-free survival. Progression-free survival. Time to next treatment. Time to next treatment. Overall survival. Overall survival. Other prespecified and exploratory outcomes Types and frequencies of adverse events. Types and frequencies of adverse events. Tumour size change. Tumour size change. Visual acuity. Visual acuity. Intraocular pressure (IOP). Intraocular pressure (IOP). Visual field. Visual field. Visual evoked potential (VEP). Visual evoked potential (VEP). Treatment cost. Treatment cost. Examinations Baseline data and follow-up examination items are as follows (figure 2): Demographic data: including name, ID number, sex and education level. The schedule for subject follow-up. Medical history, surgical history and medications in use. A full-body examination or assessment includes vital signs, height and weight measurement, complete blood count, blood chemistry analysis, liver and kidney function tests, urinalysis and an ECG. Optometry and visual acuity: using an ETDRS LogMAR visual acuity chart (Precision Vision, Villa Park, Illinois, USA), in accordance with the refractive error study in children research protocol; at a test distance of 4 m, the naked eye and BCVA will be evaluated and recorded in the sitting position. IOP measurement: each visit requires an IOP evaluation in both eyes of the patient and the average of the three measurements will be taken using non-contact Tonometry. Slit-lamp examination: the ophthalmological examination includes slit-lamp examination (BQ-900, Haag Streit, Switzerland) and fundus examination. Anterior segment photography: a camera attached to the slit lamp microscope captures high-resolution images of the front part of the eye. Mydriatic fundus photography: a fundus camera will be used (model can be defined according to the subcentre) to take a photo of the fundus after dilation of the pupils. Endothelial cell counting is assessed using endokeratoscope (SP-2000P, Topcon, Japan). Orbital ultrasound is performed with an OTI Scan 3000 unit (OTI, Ophthalmic Technology, Toronto, Ontario, Canada) using a 10 MHz frequency probe in a supine position. Each subcentre can choose the instrument model. Orbital MRI was performed on a 1.5 T (General Electric, Waukesha, Wisconsin, USA) or a 3 T superconducting system (Siemens Medical Solutions, Erlangen, Germany). Axial, coronal and sagittal spin-echo T1W images, as well as axial T2W images were obtained. In addition, contrast-enhanced axial, coronal and sagittal T1WI scans were performed with fat suppression. Each subcentre can choose the instrument model. Visual field examination will be performed using SITA standard 24–2 programme by Humphrey Field Analyzer Mark 2/3 (Carl Zeiss Meditec, Dublin, California, USA). It is required to report false positive rate <15%, false negative rate <15% and fixation loss <20%. The reliability of visual field inspection, and MD and pattern SD values will be recorded. Full-field ERG was performed following the International Society for Clinical Electrophysiology of Vision protocol.27 Each subcentre can choose the instrument model. The VEP examination is conducted according to the International Federation of Clinical Neurophysiology and American Society of Electroencephalography guidelines.28 29 Each subcentre can choose the instrument model. Adverse events are collected at each visit. Sample size calculation The sample size is calculated based on the primary outcome, namely the cumulative incidence of ≥grade 2 complications within 5 years after treatment commencement. The cumulative incidence of complications in the control group receiving ISRT over 5 years is assumed to be 30% based on the following: The American Academy of Ophthalmology Ophthalmic Technology Assessment Committee Oculoplastic and Orbit Panel conducted a systematic assessment of the efficacy and safety of existing treatments for ocular adnexal lymphoma.5 They included 27 studies with 2009 patients, of which approximately 30% reported various types of ≥grade 2 complications (median follow-up: 2.7–11.4 years). This study anticipates that the cumulative incidence of ≥grade 2 complications in the experimental group over 5 years will be reduced to 8%.192330 With a two-sided significant level of 0.05 and a power (1–β) of 80%, the target sample size for each group is estimated to be 46 using the software PASS V.16.0 (NCSS, LLC, USA). Considering an approximate 15% dropout rate, the final sample size for each group is determined to be 54, for a total of 108 cases. The main efficacy endpoint is the overall response rate within 5 years after treatment commencement. Non-inferiority testing will be employed. Studies reported an average overall response rate of approximately 95% for ocular adnexal MALT lymphoma treated with ISRT.1 13 This study, based on the lower incidence of complications with rituximab relative to radiotherapy, establishes the clinical tolerance for the efficacy difference between these two treatment modalities, known as the non-inferiority margin (∆), set at −15%. Based on a two-sided alpha of 0.05 and an 80% power of the test, the required sample size per group is 46. The sample size, calculated based on safety indicators, is deemed sufficient to meet the non-inferiority hypothesis for efficacy. Monitoring Data monitoring The data monitoring will be performed by professional data administrators in clinical research centre of ZOC and will be independent of the researchers, who have no competing interests. Interim analysis Interim analysis will not be conducted. Harms The risks include common risks in radiation therapies and rituximab use. This study does not involve additional treatment methods and protocols. The radiation therapy and localised injection of rituximab in the study are standard clinical treatments, which patients might undergo even if they do not participate in the study. The treatment methods for ocular adnexal MALT lymphoma include surgery, radiation therapy, chemotherapy, anti-infective treatment and monoclonal antibody immunotherapy. Sole surgery often cannot achieve complete tumour removal, and local tumour control is not satisfactory. Chemotherapy is commonly used in advanced tumours and requires intravenous drug administration with significant systemic side effects. Anti-infective therapy is only suitable for patients positive for specific pathogens. Therefore, the treatment protocol in this trial will not add any additional risk or burden. Auditing The trial will not involve new drugs or new devices and will be conducted under the guidance of the ethics committee of all subcentres. Data collection and management All the original data will be stored in the Electronic Data Capture (EDC) system. The principal centre is eligible to view the data of all subcentres, but revision will not be allowed. The data will be directly entered into the database by researchers and used as raw data. All raw data information must be kept by the researcher in the subject’s file. Any changes from the raw data will be documented in the EDC system. Statistical plan Population for statistical analysis Modified intent-to-treat population: refers to participants who receive randomisation and interventions have at least one follow-up visit. People will be included in the analysis according to their original group assignment regardless of the actual intervention received. Modified intent-to-treat population: refers to participants who receive randomisation and interventions have at least one follow-up visit. People will be included in the analysis according to their original group assignment regardless of the actual intervention received. Per-protocol population: refers to participants who complete the trial according to the trial protocol completely. Per-protocol population: refers to participants who complete the trial according to the trial protocol completely. Safety assessment population (SAP): all participants who have at least one safety assessment record constitute the safety analysis population of this study. Safety assessment population (SAP): all participants who have at least one safety assessment record constitute the safety analysis population of this study. Statistical software SAS V.9.4 software or other appropriate software will be used for statistical analysis. Basic principles All statistical inferences use two-sided tests. A statistically significant test level is defined as 0.05, and the CI estimate for parameters uses a 95% CI. Missing data Efficacy analysis uses multiple imputation for supplementation or sensitivity analysis. Dropout analysis Comparison of overall dropout rates and dropout rates due to adverse events will use Pearson χ2 test or Fisher’s exact test. Descriptive statistics Mean and SD or median with IQR will be reported for continuous data. Categorical data will be described as frequency and percentages. No baseline comparisons between trial arms will be conducted. Safety analysis The difference of cumulative incidence of ≥grade 2 complications within 5 years after treatment commencement between trial arms will be tested by Pearson χ2 test. Logistic regression will be used to evaluate the adjusted intervention effect by including baseline demographic and clinical characteristics: age, gender and TNM stage. Also, the centre effect will be assessed. Efficacy indicator (overall response rate) The primary efficacy indicator is the overall response rate, with an assumption of non-inferiority. When the lower limit of the 95% interval of the difference in overall response rates between treatment groups (experimental group − control group) is greater than non-inferiority margin of −15%, it can be judged that the efficacy of the experimental group is not inferior to the control group. If the lower limit of the 95% interval is greater than 0, it can be determined that the efficacy of the experimental group is superior to the control group. Overall safety analysis Use Pearson χ2 test or Fisher’s exact probability method to compare the occurrence of all adverse events (including but not limited to the aforementioned treatment-related ophthalmic adverse events) between the two groups, and list all adverse events that occurred in this trial and reported abnormal laboratory tests, clinical examinations, and so on. Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research. Study completion and termination Study completion When the study is completed, the investigator will hold a clinical study report meeting, revise and sign the final clinical study report. The investigator will report the completion of the study to the ethics committee. Discontinuation and early termination of the study Subjects will be withdrawn from the study if: they do not comply with protocol or for safety reasons; female patients who intend to become pregnant or who have been confirmed to be pregnant; they participate in other clinical trials; serious adverse events occur during the study. If a subject decides to withdraw from the study, the subject will be advised to complete the last follow-up and required tests prior to withdrawal. Patient and public involvement Neither the patients nor the public are involved in our research’s design, conduct, reporting or dissemination plans. Ethics and dissemination Ethics approval Ethical approval has been obtained from the Ethics Committee of the ZOC (ID: 2023KYPJ125-3; V.20240818) and the ‘5010 Plan’ evaluation committee at Sun Yat-sen University, Guangzhou, China and all subcentres. The registration identifiers of other centres are listed as follows: Xi'an People’s Hospital (ID: KJLL-H-K-2024016), The First Affiliated Hospital of Zhengzhou University (ID: 2023-KY-1265-003), Peking University Third Hospital (ID: M2023426), Hainan Eye Hospital (ID: 2023-022-03), The First Affiliated Hospital of Jinan University (ID: KY-2023-201) and People’s Hospital of Ningxia Hui Autonomous Region (ID: 2023-KFKT-015). Informed consent Researchers must obtain signed informed consent to confirm that the subjects fully understand the content of the study and participate voluntarily before the trial starts. Confidentiality The content of this clinical study is confidential information. Any information about this study, including the study design, methods, results, and so on, is within the scope of confidentiality which cannot be discussed with persons outside the study. Dissemination policy The results will be disseminated through scientific meetings and published in peer-reviewed journals. Statement on post-hoc protocol amendments This study has undergone several post-hoc protocol amendments, all of which have been documented in ClinicalTrials.gov (trial registration number: NCT06190301) and were approved by the relevant ethics committees. These amendments were made to enhance the study design and ensure the rigour of our findings. supplementary material