Study protocol for prospective multi-institutional phase III trial of standard of care therapy with or without stereotactic ablative radiation therapy for recurrent ovarian cancer (SABR-ROC)

This is a phase 3, multicenter, randomized, prospective study to evaluate whether the addition of SABR to the standard of care therapy significantly improves the 3-year overall OS in patients with ROC. Participants are stratified according to the following criteria before randomization: (1) platinum sensitivity, absence of ascites, normal level of CA125, ECOG performance status of 0–1; 0–3 vs. 4; (2) site of lymph node recurrence: no vs. yes; and (3) PARP inhibitor use vs. non-use. This study randomly assigns the participants to Arm 1 and Arm 2 at a ratio of 1:2 (Fig. 1). Participants assigned to Arm 1 will continue to receive the current salvage therapy suitable for individuals by considering factors such as the location and size of recurrence, and the patient’s comorbidities, at the discretion of the physician. Patients are recommended to receive standard care therapy according to the National Comprehensive Cancer Center (NCCN) guidelines for ROC. Participants in Arm 2 will undergo SABR for recurrent lesions that are clearly identified on routine imaging tests. After SABR, additional standards of care therapy will be continued as planned at the discretion of the physician. Additional salvage therapy will continue before or after SABR, as planned, at the discretion of the physician. After SABR is performed on recurrent lesions identified at the time of registration, additional SABR may be performed on newly developed recurrent lesions during follow-up at the physician’s discretion (no limit on the number of times). This trial is anticipated to run for 5 years, with a 2-year enrollment phase and a 3-year follow-up. The first participant was recruited in November 2022. Participant registration is expected to be completed by December 2024, and the final results will be available after 2027. This protocol has been approved by the Korean Gynecologic Oncology Group (KGOG) and KROG and has been designated as KGOG 3064 and KROG 2204. This trial was registered at ClinicalTrials.gov (NCT05444270) on June 29, 2022.

Patients who meet all of the following criteria will be included in this study: (1) pathologically proven epithelial ovarian cancer; (2) patients must have completed treatment for the primary tumor (maximal debulking operation and adjuvant platinum-based chemotherapy according to the stage); (3) number of metastases allowed: ≤10 (adjacent lesions can be counted as a single lesion if possible to be included in a single RT treatment plan; left cervical lymph node, right cervical lymph node, center of the left lung, periphery of the left lung, left pleura, center of the right lung, periphery of the right lung, right pleura, mediastinal lymph node, left lobe of the liver, right lobe of the liver, perihepatic space, spleen, perisplenic space, within two vertebrae above or below based on the spine with the lesion, abdominal cavity, pelvic cavity, paraaortic lymph node, and pelvic lymph nodes); (4) maximum diameter of each metastatic site of gross tumor ≤ 5 cm; (5) age ≥ 19 years old; (6) sufficient bone marrow function on tests performed within 60 days of registration; (7) Zubrod Performance Status Score ≤ 2 at the time of registration; and (8) patients must submit the informed consent form related to the study before participating in the study.

Patients who meet any of the following criteria will be excluded from this study: (1) brain metastasis; (2) diffuse peritoneal carcinomatosis; (3) exudative, bloody, or cytologically proven malignant pleural or pericardial effusion; (4) a history of RT; (5) lesions unsuitable for targeting because of unclear borders; (6) co-existing or underlying invasive cancer (excluding thyroid cancer, cervical CIS, basal cell carcinoma of skin, and early gastric cancer) that has not achieved disease-free status for ≥ 3 years; and (7) serious comorbidities.

The supporting data for calculating the number of participants required for this study were based on the results of a retrospective analysis of the treatment results of patients with ROC at the Yonsei Cancer Center (unpublished data). In this analysis, the 3-year survival rates were 74.4% and 58.5% in the IFRT and non-radiation treatment groups, respectively. Based on this, the type I error rate was set at 0.05; statistical power, 80%; 1-year dropout, 5%; randomization ratio, 1:2; accrual time, 2 years; and follow-up, 3 years. As a result, the number of patients needed was calculated as 90 for Arm 1 and 180 for Arm 2.

Patients eligible for enrollment are considered through multidisciplinary discussions with oncologists. The investigator explains the background, purpose, and procedures of this study to the patients and their guardians and gives them sufficient time to decide whether to participate in the study. If the patient agrees to participate in the study, informed consent is obtained. Patients are informed that they can indicate their intention to withdraw from the study at any time and that their participation in the study will be immediately discontinued upon withdrawal. To minimize the possibility of coercion or undue influence, the investigators explain that the research patient’s decision to participate in the research is voluntary, they can refuse to participate in the research or withdraw participation at any time during the research period, and there will be no disadvantages for the next treatment.

The following screening tests must be performed to for patients to participate in this clinical trial: clinical examination, history taking and physical examination, complete blood count (once, routine test volume), routine chemistry to evaluate hepatic and renal function (once, routine test volume), computed tomography (CT) or positron emission tomography (PET)-CT to confirm the size and location of the recurrent lesion, and tumor marker CA125. However, these need not be conducted if they have been performed within 2 months of registration.

This study requires only a photon beam of 6 MV or more: both intensity modulated RT (IMRT) and volumetric modulated arc therapy with linear accelerators are acceptable, and TomoTherapy (Madison, WI, USA) and CyberKnife (Sunnyvale, CA, USA) can be used. All participants should have a CT-based treatment plan, and CT slices should be 3 mm apart or smaller, if necessary. Multiple CT scans are also permitted if there are two different treatment sites such that images cannot be obtained simultaneously. It is recommended that all treatment sites be targeted with the patient in a single treatment position. However, the patient’s position can be altered when RT is administered to the lungs and extremities. All lesions that may move during breathing should be evaluated using 4D CT, fiducial markers, or fluoroscopy. In addition, for lesions with movement of ≥ 5 mm, breathing control techniques, including the use of an abdominal compression device, active breathing control, breathing suppression, gating, and tracking, or internal target volume techniques, are all recommended. All treatments in this study require an image guidance technique that can identify the target of treatment in three perpendicular directions (superior/inferior, left/right, and anterior/posterior). Cone Beam CT, Mega-Voltage CT, dual fixed-position in-room kV imaging system, in-room diagnostic CT, and TomoTherapy approaches are all accepted. Target volumes are defined as follows: The gross tumor volume (GTV) includes all recurrent sites observed on planning CT and additional PET-CT. The clinical target volume is not defined due to the nature of this study. To account for setup error and movement of internal organs, a margin of 3–5 mm around the GTV is set to generate the planning target volume (PTV). Multileaf collimator shielding is used based on the PTV.

The recommended dose fractions are listed in Table 1. In the case of 6–9 fractions, the prescribed doses of 5 and 10 fractions, along with dose constraints, are applied proportionally. The prescription isodose surface is selected to include 90% of the PTV on the prescription isodose curve. Doses less than 90% of the prescription dose are limited to the outer boundary of the PTV. As an exception, 70% coverage is allowed for some PTVs if small lesions are collectively formed and have irregular shapes. However, maximum efforts should be made to ensure that the GTV is covered by 100% of the prescribed isodose surface. The prescription isodose surface should be determined from 60 to 100% of the maximum dose in the PTV. The maximum dose should be 100–166.67%, based on the prescription isodose surface, and is normalized to 100%. If a considerable volume of OARs is included in the PTV, a treatment plan using 10 fractions is recommended, and in this case, PTV-EVAL can be set separately to evaluate the PTV coverage. PTV-EVAL is defined as the remaining part of the predefined PTVs, excluding OARs (bowel, duodenum, rectum, bladder, etc.) during treatment planning. However, in this case, even if the evaluation of the target is based on PTV-EVAL, the minimum dose of the PTV should be ≥ 70% of the prescribed dose. If it is difficult to achieve the R_50% conditions, excluding the evaluation of the ratio of the prescription isodose volume, it is evaluated as an acceptable treatment plan, even if it is not met. However, the R_50% and D_2cm are evaluated using the existing PTV, and lowering the OAR dose should be prioritized over dose fall-off in this case. When a treatment plan with PTV-EVAL is evaluated, prior consultation should be communicated to the research headquarters (Yonsei Cancer Center, Seoul, South Korea) to prevent decision as an unacceptable deviation.

Adverse events to be recorded include hematological, gastrointestinal, renal/genitourinary, dermatological, and other adverse effects and will be evaluated based on the Common Terminology Criteria for Adverse Events (CTCAE) version 5. An institutional investigator will report all serious adverse events to the institutional review board of each participating center and the primary investigator.

After the completion of RT, observation and clinical examination will be conducted as follows: interview with study staff at 3, 6, 9, 12, 18, 24, 30, and 36 months after registration, history taking, physical examination, and evaluation of all possible side effects. Related tests can be prescribed if necessary, imaging tests (either CT or PET-CT) including the treatment site or suspected recurrence area, tumor marker CA125 test at every visit after registration, and survey (using the EORTC quality of life questionnaire (QLQ)-C30 questionnaire before randomization, and at 12, 24, and 36 months after registration for all participants; for participants in Arm 2, Cancer Therapy Satisfaction Questionnaire (CTSQ) is also used before randomization, and at 12, 24, and 36 months after registration).

The principal investigator is responsible for managing the clinical trial data. Investigators will complete the case report forms using a web-based system (https://​www.​mytrial.​co.​kr/​). Data monitoring will be performed by a clinical research organization (KGOG, Seoul, South Korea) and the clinical research coordinator of each participating center. Documents generated from this trial will be stored at each participating center for 10 years after trial completion with the approval of the institutional review board, after which personal information will be discarded.

For patients who agree to participate, 20 unstained slides are sectioned from formalin-fixed paraffin-embedded blocks of tumor tissue during the primary debulking operation in participating institutions. The slides are sent to research headquarters for next-generation sequencing. Tumor samples already stored at each institution are collected for genome analysis, and there is no plan to acquire additional tissue for this study. Except for the 20 slides, tissue blocks will be stored in the pathology department of each institution, and the research headquarters will keep them for up to five years after genomic analysis.

To develop a patient selection and prediction model in which SABR can improve clinical outcomes, artificial intelligence (AI)-based radiomics analysis is performed to predict the molecular subtype using the patient’s CT and/or PET-CT images. Imaging data are collected from each institution and transferred to the research headquarters for analysis.

The primary endpoint is OS (failure or death from any cause). Secondary endpoints include control of existing recurrence sites, new recurrences, adverse events, chemotherapy-free intervals, health-related quality of life, discovery of treatment-refractory molecular subtypes through next-generation sequencing, development of a prediction model using AI-based radiomics, and analysis of other clinically meaningful descriptive variables described in the protocol.

The primary hypothesis of this study is that the addition of SABR to the standard of care therapy in patients with ROC would improve the OS from 58.5 to 74.42%, corresponding to a hazard ratio of 0.55. Survival is estimated by Kaplan–Maier method, and survival failure is defined as the subject’s death from any cause. The log-rank test is used to compare the distribution of survival estimates between the Arm 1 and Arm 2 participants. The survival period is defined as the period from the date of randomization to the date of death. Imaging tests are performed every 3 months for the 1st year, every 6 months for 3 years, and every 6 months thereafter until disease progression. The Cox proportional hazards regression model is used to analyze factors affecting the OS along with treatment methods.