Ethics approval and study design
This retrospective observational study was approved by ethics committee, Aichi Medical University School of Medicine in Japan (application number 2018-H211). This study was also examined and approved by Aichi Cancer Center Hospital, Toyota Memorial Hospital, and Aoyama Hospital, Japan. The study was conducted in accordance with the tenets of the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from all patients.
Patients
Between January 2008 and April 2018, the medical records of patients treated with radiation to their abdominal/pelvic LN with a definitive intent were retrospectively reviewed from 4 institutes. In this study, patients aged 20 to 85 years who developed localised abdominal/pelvic oligometastases in 1–5 LNs after initial treatment were included. The exclusion criteria were as follows: patients with non-epithelial tumours, those with uncontrolled primary lesions, patients treated with palliative intent, those with a short follow-up (<3 months), and those who received re-irradiation. Finally, a total of 113 patients were included.
The initial treatment of the primary disease was either surgery alone (n = 101) or definitive radiotherapy (n = 12). Before undergoing salvage radiotherapy, all the patients underwent whole-body computed tomography (CT) and/or positron emission tomography-CT (PET-CT) to confirm that there were no oligometastases other than those in the abdominal/pelvic LNs. Owing to the presence of oligometastases after initial treatment, 91 patients underwent chemotherapy and 33 underwent concurrent chemotherapy and salvage radiotherapy.
Radiotherapy
A linear accelerator or CyberKnife (CK) was used for salvage radiotherapy. Patients were immobilised in the supine position, and a CT scan with a 1–5-mm slice thickness was obtained for treatment planning. For high-precision radiotherapy, such as intensity-modulated radiotherapy (IMRT) and stereotactic body radiotherapy (SBRT), an external vacuum-type body mould and/or a thermoplastic body mask were used for precise fixation. To avoid large displacement of the gastrointestinal tract during daily treatment, planning CT was performed with the patient in a fasting state, especially when the target was near the stomach. The clinical target volume (CTV) was defined as the gross tumour volume with a 0–0.5-cm margin considering microscopic disease and the anatomical structure. The planning target volume (PTV) was defined as the CTV with a 0.3–1.0-cm margin, considering the internal motion and the setting error in each institution. Moreover, the PTV margin was suitably adjusted to protect organs at risk (OARs). While using the linear accelerators, the abdominal/pelvic LN region was divided into 4 regions: para-aortic, iliac, presacral, and obturator. The nodal area that includes the gross tumour was defined as a prophylactic nodal area. Regarding iliac and obturator regions, the side of the tumour was delineated. In cases of 2 oligometastases or more, prophylactic nodal areas were made smaller so that the irradiation area would not be too large. Prophylactic nodal areas were expanded by 0.5 cm considering setup errors. The main OARs were evaluated by using the equivalent dose in 2 Gy fractions (EQD2) at α/β = 3, including the stomach, small bowel, large bowel, kidneys, and spinal cord. For 3-dimensional conformal radiotherapy (3DCRT), dose constraints were defined as follows: the maximum point dose was 52 Gy to the stomach and small bowel, 62 Gy to the large bowel, and 50 Gy to the spinal cord. For high-precision radiotherapy the dose constraints were defined as follows: the stomach and small bowel received a point dose of 54 Gy (V54) for <3 cm3, the large bowel V64 for <3 cm3, and the kidneys a V50 of <33%. The maximum point dose to the spinal cord was <52 Gy.
Linear accelerator
Sixty-one patients were treated with a linear accelerator. Both the target volume and normal organ structures were contoured using treatment planning systems (either XiO, Electa, CMS, St Louis, MO, USA or Eclipse, Varian Medical System, Palo Alto, CA, USA). All the patients were treated with 10 MV photons, using the 3DCRT or IMRT technique. Patients treated with 3DCRT (n = 47) were prescribed 39.6–70 Gy in 15–37 fractions to the PTV isocenter; the median dose per fraction was 2 Gy (range, 1.8–3 Gy) and the median EQD2 was 50 Gy. In contrast, patients treated with IMRT (n = 14) received 45–70 Gy in 15–35 fractions, where the PTV was covered with a 95% dose; the median dose per fraction was 2 Gy (range, 1.8–3.5 Gy) and the median EQD2 was 61.4 Gy. More detailed data such as treatment modality and dose fractionation are shown in Supplementary Table 1. The prophylactic nodal areas received almost 70–80% of the PTV dose in both 3DCRT and IMRT. Either sequential boost or simultaneous integrated boost technics were used in patients who received prophylactic nodal irradiation with IMRT.
CyberKnife
Fifty-two patients were treated with CyberKnife. All SBRT patients were treated with CyberKnife. Ten patients were treated with less than 4 Gy per fraction, and they were classified as IMRT. CK G4 and M6 (Accuray Inc., Sunnyvale CA, USA) and B were used for SBRT. In addition to using the IRIS variable collimator, the InCise multileaf collimator was used and contributed to reduction in the treatment time for CK-M6. Generally, the tumour was followed-up by using the fiducial-less tracking capability of CK under free breathing. Using a 6-MV photon beam, all treatment plans were generated using the Multiplan treatment planning software (Accuray Inc., Sunnyvale, CA, USA). The dose was prescribed for covering the PTV, ranging from 21 to 60 Gy, and it was fractionated 2 to 30 times with a 60% to 90% isodose line; the median dose per fraction was 6 Gy (range, 2–13.5 Gy) and the median EQD2 was 63.5 Gy.
Data collection and statistical analyses
As the fractionation schedules and dose evaluation method were not standardised, the dose-volume histograms of all the cases were confirmed. Then, the prescription dose of all the cases was re-evaluated by using the EQD2 (D50%) with α/β = 10. The median EQD2 for all the patients was 59.7 Gy (range, 40.3–101.4 Gy); using the median value as the cut-off, the high-dose group was defined as patients with EQD2 ≥ 60 Gy and the low-dose group was defined as those with EQD2 < 60 Gy.
According to the Response Evaluation Criteria in Solid Tumors, the initial tumour response was evaluated based on CT scans [10]. Complete response was defined as a reduction of the lymph node to < 10 mm in the short axis. Partial response was defined as at least a 30% decrease from the baseline. Progressive disease was at least a 20% increase, taking as reference the smallest state. Stable disease was defined as having neither sufficient shrinkage to qualify for partial response nor sufficient increase to qualify for progressive disease. The first follow-up visit was scheduled within 1–4.5 months (median, 2.2 months) after the end of treatment. The disease-free interval (DFI) was defined as the time from the last treatment to salvage radiotherapy. The primary endpoint was the OS. The secondary endpoints included LC, progression-free survival (PFS), and adverse events. OS was measured from the start date of salvage radiotherapy to the date of the last follow-up or death from any cause. Local control (LC) was defined as progression in the target LN as evaluated on CT or PET-CT images. Progression-free survival (PFS) was calculated from the start date of salvage radiotherapy until the date of disease progression or death from any cause.
The OS, LC, and PFS rates were estimated using the Kaplan-Meier method [11]. Log-rank tests were used to compare the estimates of subgroups on univariate analysis. The Cox proportional hazards model was used for multivariate analysis. P-values < 0.05 were statistically significant. Factors that showed a difference with p < 0.1 on univariate analysis were entered into the multivariate analysis. Adverse events were graded according to the National Cancer Institute Common Toxicity Criteria for Adverse Events (version 5.0), and grade ≥ 3 events were counted. As supplementary information, events after re-irradiation were collected as secondary adverse events, separately from the primary adverse events. Re-irradiation here does not refer to the second radiotherapy in the same area as the initial treatment, which was excluded in the study, but to the one performed for oligometastasis of the abdominal/pelvic LNs. All statistical analyses were performed with EZR version 1.33 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), based on the R and R commander [12].