Dose escalation based on 18F-FDG PET/CT response in definitive chemoradiotherapy of locally advanced esophageal squamous cell carcinoma: a phase III, open-label, randomized, controlled trial (ESO-Shanghai 12)

Definitive chemoradiotherapy has established the standard non-surgical treatment for locally advanced esophageal cancer. The standard dose of 50–50.4 Gy has been established decades ago and been confirmed in modern trials. The theorical advantage of better local control and technical advances for less toxicity have encouraged clinicians for dose escalation investigation. 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) have the potential to tailor therapy for esophageal patients not showing response to CRT and pioneers the PET-based dose escalation. The ESO-Shanghai 12 trial is a prospective multicenter randomized phase 3 study in which patients are randomized to either 61.2 Gy or 50.4 Gy of radiation dose by PET response. Both groups undergo concurrent chemoradiotherapy with paclitaxel/cisplatin regimen for 2 cycles followed by consolidation chemotherapy for 2 cycles. Patients with histologically confirmed ESCC [T1N1-3M0, T2-4NxM0, TxNxM1 (Supraclavicular lymph node metastasis only), (AJCC Cancer Staging Manual, 8th Edition)] and without any prior treatment of chemotherapy, radiotherapy or surgery against esophageal cancer will be eligible. The primary endpoints included overall survival in PET/CT non-responders (SUVmax > 4.0) and overall survival in total population. Patients will be stratified by standardized uptake volume, gross tumor volume and tumor location. The enrollment could be ended, when the number of PET/CT non-responder reached 132 and the total population reached 646 for randomization. This trial has been approved by the Fudan University Shanghai Cancer Center Institutional Review Board. Trial results will be disseminated via peer reviewed scientific journals and conference presentations. Trial registration The trial was initiated in 2018 and is currently recruiting patients. Trial registration number NCT03790553.

Background According to the Global Burden of Disease Study 2019, esophageal cancer (EC) accounted for approximately 535,000 new cancer cases and 498,000 deaths in 2019 worldwide [1]. The 5-year standardized net survival rate ranges between 10 and 30% globally [2]. There are two distinct histopathologic subtypes of EC: esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma. These subtypes vary in terms of incidence, risk factors, location, and age of diagnosis [3], and ESCC is the dominant subtype in Asian populations [4]. Definitive chemoradiotherapy (CRT) has been established as the standard nonsurgical treatment for locally advanced EC. Measures to achieve better local control and overall survival, including radiation dose escalation, chemotherapy regime optimization, and the combination of target agents, have been explored for decades. However, few of these measures have demonstrated positive results in randomized phase III trials. The ESO-Shanghai 12 trial (NCT03790553) was designed to evaluate dose escalation in ESCC chemoradiation based on the response on 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography/computed tomography (PET/CT), with the aim of establishing an approach to precision radiation therapy for EC.

Standard dose for CRT
The phase III RTOG-8501 trial demonstrated that 50 Gy radiation with concurrent chemotherapy is superior to radiotherapy of 64 Gy alone, thus establishing definitive concurrent CRT as the standard nonsurgical treatment scheme for locally advanced EC [5]. The phase III randomized trial RTOG-9405 compared a high dose of 64.8 Gy and a low dose of 50.4 Gy in definitive concurrent CRT. The results showed that a high dose did not improve survival or local control, with the absolute value of median overall survival and the 2-year survival rate being lower in the high-dose group than in the lowdose group [6]. During the past decade, radiotherapy has entered the era of three-dimensional conformal radiation therapy (3DCRT); however, the standard radiation dose of 50 Gy does not present preferable survival characteristics regardless of whether it is combined with concurrent systemic therapy [7][8][9]. With the wide application of intensity-modulated radiotherapy (IMRT), radiation oncologists are interested in re-investigating dose escalation in EC. The highly anticipated European CON-CORDE [10] and ARTDECO [11] studies, along with a Chinese phase III trial [12], have further demonstrated that increasing the radiation dose for non-selected populations does not offer benefits in definitive CRT for EC in the modern era. However, high doses (≥ 60 Gy) have been widely used in Asia and have demonstrated positive results in Asian populations and ESCC subtypes in retrospective studies [13,14], and meta-analyses have further contributed to the debate [15,16]. Moreover, our previous drug comparison phase III trial [17] using involved-field irradiation-based IMRT up to a total dose of 61.2 Gy demonstrated favorable survival outcomes (Table 1).

Failure patterns of CRT in EC
Researchers have tried to explore failure patterns of EC chemoradiotherapy. Button et al. [18] reported the first failure site in 145 patients receiving chemoradiotherapy for EC. The irradiation field included the gross tumor volume (GTV), plus 3 cm of the proximal and distal normal esophagus and a 1.5-cm margin in the transversal position. Eighty-five patients were identified as treatment failure, including 55 patients with in-field failure, 13 patients with distant metastases, and 14 patients with both. Only 3 patients had relapse observed at the edge of the irradiation field. Welsh et al. [19] retrospectively analyzed 239 patients who received CRT for EC, among whom 50% had local failure and 48% had distant metastases. Of all the cases of local failure, 90% occurred in the GTV, 27% in the clinical target volume (CTV), and 12% in the primary tumor volume (PTV). In the ESO-Shanghai 1 trial, the radiotherapy techniques were involved-field irradiation (no lymph node preventive irradiation) and IMRT, and the radiotherapy dose was 61.2 Gy/34 Fx. Of the 436 trial participants, 258 (59.2%) experienced treatment failure; 37 patients (8.5%) showed failure of irradiation of the lymph nodes in the field, of which 7 cases (1.6%) had simple lymph node failure [20]. Therefore, considering recurrence within the irradiation field accounts for half of cases of CRT failure in EC, dose escalation may be an effective approach to achieve a good prognosis.

Ethics and dissemination:
This trial has been approved by the Fudan University Shanghai Cancer Center Institutional Review Board. Trial results will be disseminated via peer reviewed scientific journals and conference presentations. Trial registration The trial was initiated in 2018 and is currently recruiting patients. Trial registration number NCT03790553.

Feasibility and effectiveness of high-dose radiotherapy in EC
The esophagus may be the most important dose-limiting organ. Under the guidance of PET/CT, Yu et al. [21] irradiated more than 50% of the tumor area with the maximum standardized uptake volume (SUV) before radiotherapy. They found that it was safe for patients to receive 70 Gy/25 Fx at the same time and that the short-term toxicity was tolerable. In a phase 1/2 trial, Chen et al. [22] reported that chemoradiotherapy with a simultaneous integrated boost of radiotherapy (at doses of 50.4 Gy to subclinical areas at risk and 63.0 Gy to the gross tumor and involved nodes, all given in 28 fractions) for patients with locally advanced EC was well tolerated, with encouraging local control. If the target volume dose can be increased without affecting the tolerance of normal tissue, local control and the overall therapeutic effect may be improved in EC radiotherapy [23]. In recent years, the progress of radiation technology has ensured the safety and effectiveness of increasing the local radiotherapy dose. Also, because involved-field irradiation entails less toxicity, its use may increase the feasibility of increasing the dose.

Tumor response assessment
In EC, evaluation of the tumor response to neoadjuvant therapy can not only be used to predict prognosis but also to detect non-responders of CRT and allow adjustment of treatment strategy. According to the MUNICON trial, patients who show a clinical response to induction treatment (chemotherapy-or chemoradiotherapy) have a better prognosis [24]. At present, several methods are available to evaluate the treatment response; among them, ultrasonic gastroscopy and 18 F-FDG PET/CT have high accuracy [25]. Consequently, improving induction chemotherapy to increase the proportion of treatment responders, as well as intensifying the radiotherapy, could be strategies to improve local disease control or survival. Several phase II trials [26][27][28][29] have shown that PET/CT has the potential to tailor therapy for patients not showing an early response to chemotherapy and have pioneered PET-directed EC neoadjuvant therapy ( Table 2).

Design
The ESO-Shanghai 12 trial is a prospective, multicenter, randomized phase 3 study, in which patients are randomized to receive radiation at a dose of 61.2 Gy or 50.4 Gy based on the 18 F-FDG PET/CT response (stratified by SUV max > 4 and SUV max ≤ 4). Both arms will undergo concurrent chemoradiotherapy with a paclitaxel/cisplatin (TP) regimen for two cycles followed by consolidation chemotherapy for two cycles ( Fig. 1).

Overall survival in PET/CT non-responders (time frame: 2 years)
The time between the start of the study treatment (day 1) and death from any cause or the last followup (for patients who are alive at the end of the study) in patients who have an SUV max > 4 on PET/CT at 28 radiotherapy fractions. 2. Overall survival in the intention-to-treat (ITT) population (time frame: 2 years) The time between the start of the study treatment (day 1) and death from any cause or the last followup (for patients who are alive at the end of the study) in the ITT population.

Secondary endpoints
1. Local control in PET/CT non-responders, the ITT population, and PET/CT responders (time frame: 2 years) The time between the start of the study treatment (day 1) and local recurrence (including primary tumor recurrence and regional lymph node failure).

years)
A quality-of-life score will be obtained based on the answers to the questionnaire. 5. Questionnaire: EORTC-QLQ-OES18 (time frame: 2 years) A quality-of-life score will be obtained according to the answers to the questionnaire. 6. Exploration of predictive and prognostic biomarkers.

Patient selection Inclusion criteria
To be eligible for inclusion in this study, patients must fulfill all the following criteria: 1. Participating in the study voluntarily and able to sign the informed consent form. 2. Aged between 18 and 75 years, and of either sex.

Exclusion criteria
The following patients will be ineligible for this study.

Study treatment
The treatment plan is shown in Fig. 1. Patients will receive radiotherapy combined with concurrent chemotherapy. Radiotherapy will begin on day 1, concurrent with the beginning of cycle 1 of chemotherapy. Radiotherapy will be delivered with photons (≥ 6 MV) to a total dose of 50.4 Gy in 28 fractions or 61.2 Gy in 34 fractions. 18 F-FDG PET/CT will be performed for all patients at baseline and at 28 radiotherapy fractions (± 3 days), and then patients will be randomized after assessment. Patients will be stratified according to the SUV max (≤ 4 or > 4), GTV (≤ 40 cm 2 or > 40 cm 2 ) and tumor location [cervical/upper thoracic location with heart dose ≤ 10 Gy or (middle/lower thoracic location or heart dose > 10 Gy)] (Fig. 2).

Pre-PET/CT patient preparation
• Patients will not undergo a barium meal examination at least 1 week before the PET/CT scan. • Prior to injection, the patient must fast for at least 6 h.
• Blood glucose measurement will be performed before the injection of 18 F-FDG, and blood glucose levels should be less than 140 mg/dL. • The height and weight of patients will be measured using calibrated and medically approved devices.

Injection of 18 F-FDG
• Patients will receive an 18 F-FDG dose of 3.7 MBq/ Kg of body weight in accordance with the manufacturer's recommendations. Patients will be kept in a quiet and dimly lit room before and after the injection. • A saline flush of 1 mL will follow the 18 F-FDG injection. • The exact time of dose calibration will be recorded using a global time recording device, which will be consistently used throughout the study for time recording. The exact time of injection will be noted and recorded to permit the correction of the administered dose for radioactive decay. Furthermore, any of the dose remaining in the tube or syringe, or any 18 F-FDG that is spilled during the injection will be recorded. The 18 F-FDG injection will be performed using an intravenous catheter and three-way stopcock.
3. PET/CT imaging All PET exams will include three trans-axial, wholebody series, attenuated and non-attenuated, corrected PET and CT images.
• The scan will be conducted 60 ± 10 min after the injection of 18 F-FDG. • The patient will be instructed to empty their bladder immediately before the image acquisition. • A spiral CT scan using the low-dose technique (120 kV, 140 mA, 5-mm slice thickness) will be conducted first, followed by a PET emission scan from the distal femur to the top of the skull. • A PET emission scan covering the same transverse field of view will be obtained immediately after the CT scan. The PET data will be reconstructed using a Gaussian filter iterative (iterations 4; subsets 8; image size 168) for the reconstruction of emission images. The CT data will be used for attenuation correction of the PET images, and fused images will be displayed on a workstation.

Reporting of PET findings and SUV calculations
SUVs are commonly used in clinical practice in addition to visual assessments. The SUV is a measurement of the uptake in a tumor normalized on the basis of a distribution volume. It is calculated as follows: In this calculation, Act voi is the activity measured in the volume of interest (see "Definitions for volumes of interest (VOI) and regions of interest (ROI)"), Act administered is the administered activity corrected for the physical decay of 18 F-FDG at the start of acquisition, and BW is body weight. The patient's height, weight, and sex will be reported to allow for other SUV normalizations (lean body mass, body surface area).
• PET/CT images will be analyzed independently by two senior nuclear medicine physicians using a multimodality computer platform (Syngo, Siemens, Knoxville, TN, USA). For any inconsistent or equivocal interpretations, another experienced radiologist will be invited to the discussion to reach a consensus.

Radiotherapy
IMRT will be required. The radiation plan for all patients will involve 61.2 Gy/ 34 Fx, at 1.8 Gy/ day for 5 days per week. The standardized-dose group will receive a radiation dose of 50.4 Gy/28 Fx, while the high-dose group will receive 61.2 Gy of radiation. The patients will be immobilized in the supine position and receive a contrast-enhanced planning CT scan 0.5 cm under the cricothyroid membrane (for thoracic tumors) or basis cranii (for cervical tumors) below the kidney. IMRT will be delivered to all patients using 6 MV photons. The radiation target volume will be delineated by the field involved.

GTV
The GTV will be defined as any visible primary tumor and metastatic lymph nodes. The primary tumor will be delineated using esophagography, esophagoscopy, contrast-enhanced thoracic CT, and 18 F-FDG PET/CT). Metastatic lymph nodes will be identified using biopsy, increased uptake of FDG on PET/CT, or based on the following radiographic criteria: nodes of ≥ 1 cm in the shortest axis in the intrathoracic and intra-abdominal region or nodes of ≥ 0.5 cm in the shortest axis along the recurrent nerve.

CTV
The CTV will be defined as the GTV plus 3 cm of the proximal and distal normal esophagus without lateral margins.

PTV
The PTV will be generated by applying a 1-cm margin to the CTV.
Tissue inhomogeneity correction will be adopted for planning purposes. The criteria for dose distribution will be as follows: 95% of the PTV to receive ≥ 99% of the prescribed dose; 99% of the PTV to receive ≥ 95% of the prescribed dose; < 2 cm 3 of the PTV to receive ≥ 120% of the prescribed dose; and < 1 cm 3 of the PTV to receive ≥ 110% of the prescribed dose. The highest and lowest dose points inside the PTV will be recorded.

Normal organ contouring and dose restrictions
Normal organs, including the spinal cord, heart, and right and left lungs, will be contoured on each slice of the planning CT with no planning margin. The spinal cord dose constraint must not be exceeded for any reason. The heart contours will extend from the beginning of the right atrium and right ventricle (the pulmonary artery trunk, ascending main aorta, and superior vena cava will be excluded) down to the apex of the heart. The lung volume will be defined as the total lung minus the PTV.
The priority order of consideration for normal organ dose restrictions will be as follows: 1. Spinal cord: the highest dose point must be of less than 45 Gy. 2. Lung: the volume of the lung (PTV excluded) receiving 20 Gy must be equal to or less than 30% of the total lung volume, and the mean lung dose must be equal to or less than 15 Gy at the same time. 3. Heart: the mean dose must be less than 40 Gy.

Dose modifications
It is strongly recommended that the normal organ dose constraints are not exceeded. If any dose constraint needs to be exceeded to achieve adequate coverage of the PTV, the physician will decide whether the dose should be modified, or whether the patient should be excluded from the trial. The acceptable violations of dose modification will be as follows: 92-95% of the PTV to receive ≥ 99% of the prescribed dose and normal organ dose restrictions (except the spinal cord) exceeding 5-10%.

Radiotherapy interruption
If the any of the following toxicities are observed, radiotherapy will be delayed until the toxicity improves to grade 2 or lower.
If the following toxicity is observed, radiotherapy will be delayed until complete recovery.
• Mediastinal or thoracic infection with fever over 38.5 °C.
The suspension of radiotherapy will be permitted for 2 weeks at most; if the patient does not recover within 2 weeks, radiotherapy will be terminated.

Radiotherapy quality assurance
The IMRT planning CT of the first three patients from each participating institution will be sent to Fudan University Shanghai Cancer Center for central quality assurance to ensure that the center complies with the specific study requirements for delineation, planning, and dose distribution.

Chemotherapy
Patients in both groups will receive two cycles of concurrent chemotherapy with radiotherapy followed by two cycles of consolidation chemotherapy after CRT. Each cycle of chemotherapy will last for 28 days (4 weeks). The drugs to be used include: paclitaxel 135 mg/m 2 /day, intravenously guttae (IVGTT) over 3 h, day 1; and cisplatin 25 mg/m 2 /day, IVGTT, days 1-3.

Chemotherapy interruption and dose modification
If the following toxicities are observed on day 1, chemotherapy will be delayed until toxicity improves to grade 1 or lower.
A delay to chemotherapy of 2 weeks at most will be allowed; if the patient does not show sufficient improvement within this time, chemotherapy will be terminated.
Chemotherapy dose modifications will be based on the greatest toxicity during the last cycle. Any patients for whom chemotherapy dose modifications are required will receive the modified dose during subsequent cycles. If modifications are needed, the doses of paclitaxel and cisplatin will be decreased by 25% from the planned dose for the first cycle and by 50% for the second cycle. Dose modifications will be permitted twice at most; for patients who still require dose modification after this point, chemotherapy will be terminated.
The criteria for dose modification are as follows:

Dose modification of cisplatin
• Febrile neutropenia (ANC < 0.5 × 10 9 /L and fever over 38.3° C or over 38.0° C for 1 h). • Peripheral neuropathy of grade 2 or higher. • Serum creatinine > 3 times the upper limit of normal.
Adverse events will be evaluated according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE V.4.0). All adverse events that occur during the course of the trial (from the start of treatment until 28 days after the end of treatment), regardless of their relatedness to the study medication, will be recorded. Adverse events that occur more than 28 days after the end of treatment will only be recorded if they are relevant.

Quality of life
Quality of life (QOL) is measured as secondary endpoints in this trial and will be assessed using hard-copy versions of the EORTC core questionnaire, the EORTC QLQ-C30 (version 3.0) [30], the disease-specific module for EC, and the QLQ-OES18 [31]. The following assessment points have been chosen to describe QOL changes across time: before the start of treatment (baseline), at the end of radiotherapy (28 or 34 fractions), before two consolidative chemotherapy cycles, and at each follow-up. The EORTC-QLQ-C30 and EORTC-QLQ-OES18 questionnaires will be completed by patients and checked by physicians at clinic visits to minimize the number of missing items and assessments. Time windows of ± 3 weeks will be applied for follow-up assessment.

Adverse events
Adverse events will be accessed according to National Cancer Institute Common terminology for adverse events (NCI-CTCAE) (version 4.0) [32] and will be monitored throughout the study.

Translational research
The clinical trial will involve the collection of tissue samples and blood samples for future translational research and the development and/or validation of biomarkers. Trial participants will be asked for additional optional consent to participate in this aspect of the study. The standard tissue sample will consist of pretreatment biopsy of tumor tissue. The blood samples will include 2 × 5 ml ethylenediamine tetraacetic acid (EDTA) blood samples, one of which will be collected at the time of pretreatment, postconcurrent and preconsolidation chemotherapy, and the other after recurrence. All samples will be stored at the FUSCC BioBank for future translational research.

Statistical analysis
For patients with residual tumors of esophagus, it is estimated that the 2-year survival rate of low-dose radiotherapy group and high-dose radiotherapy group is 11% and 27%, according to literature [33,34]. The expected enrollment time is 7 years, and the last enrolled patients were followed up for 2 years. According to the ratio of 1:1, bilateral α = 0.05, power = 0.80.
When 5% of the patients are expected to fall off, 66 cases in each group and 132 cases in total were included in the PET/CT non-responder group. For non-selected population, the 2-year survival rate is estimated to be 40% of the standard-dose radiotherapy group and 49% of the high-dose radiotherapy group, based on literature data [6,35]. According to the same enrollment and follow-up time conditions, 323 cases are included in each group, and the total number of samples is determined to be 646. The double end-point fixed sequence method was used to test the PET/CT non-responder group first, and then the overall intention-to-treat population. When the number of PET/ CT non-responder reached 132 and the total population reached 646, the enrollment could be ended. Fixed sequence is applied in statistics of primary endpoint.

Discussion
Radiation dose has always been a concern in radiation oncology, in different tumor types and has been debated for several decades [36]. The theorical advantage of better local control and technical advances for less toxicities have encouraged clinicians for dose   escalation investigation. Randomized phase III trials have added evidence that high dose does not bring advantage to unselected patients receiving dCRT. At the same time, the evaluation of tumor response by a variety of imaging methods can not only predict the prognosis, but also detect the non-responders and adjust neoadjuvant treatment strategy. In dCRT, the ongoing British SCOPE2 study [37] compare the effects of standard drugs and alternative combinations used in chemotherapy for patients who do not respond to standard drug chemotherapy in the early stage of treatment, as well as compares the effects of standard dose and high-dose radiotherapy. Our current study is dedicated to SCC histology only and compares radiation dose without switching chemotherapy regime. Our response assessment is post radiation rather than early PET response, aiming to investigate the PET-direct radiotherapy in a pure way.
Last not least, we abandon elective nodal irradiation and apply involved field irradiation, which is less toxic. The detailed comparison is in Table 3. Esophageal cancer is on the way to the era of immunotherapy, despite the footstone of chemoradiation in locally advanced disease [38][39][40]. Precise radiotherapy based on biology and imaging will provide new basis for individualized therapies in the future [41].