This retrospective study was conducted with the approval of our institutional review board, and all patients provided written informed consent before treatment. The inclusion criteria were as follows: patients with localized NSCLC, N0M0 disease, who were clinically inoperable or refused surgery, and were treated with SBRT, and for whom the hemoglobin (Hb) levels were available within 2 weeks before SBRT. Between September 2009 and September 2019, 77 patients meeting the inclusion criteria were identified and included in the study. Cases where the pathological diagnosis could not be confirmed were treated as NSCLC if the joint conference of respiratory surgeons, pulmonologists, radiologists, and radiation oncologists came to that consensus.
Planning and irradiation
Before radiation treatment planning, patients were evaluated by four-dimensional computed tomography (4DCT) using Somatom (SIEMENS, Germany) for the amount of movement of the tumor caused by respiration. For 4DCT, real-time positioning management (RPM) system (Varian Medical Systems, USA) was used. Patients with respiratory motion of the tumor ≥ 1.0 cm were planned for the implantation of a fiducial marker which were implanted by bronchoscopy near the tumor. For every such patient, three markers were implanted.
All patients underwent the CT scan under light exhalation breath-hold, and 4DCT were also performed using RPM. The slice thickness was 1.0 or 2.0 mm. Patients were immobilized using Vac-Loc Cushion (CIVCO Medical Solution, USA) with both their arms up. The clinical target volume (CTV) was defined as equal to the gross tumor volume (GTV). In patients with implanted fiducial markers, the internal target volume (ITV) was equal to the CTV. In contrast, summation of the GTVs defined at every respiratory phase of the 4DCT gave the ITV in patients without the fiducial marker. The planning target volume (PTV) was generated by adding 5 mm around the ITV. In principle, the prescribed dose for peripherally situated tumors was 50 Gy in 5 fractions until September 2016 and 48 Gy in 4 fractions after October 2016. The tumors with a central location near organs at risk were treated with 60 Gy in 8 fractions. A central tumor was defined as a tumor whose distance from the proximal bronchial tree was ≤ 2 cm. The dose was prescribed to the isocenter. Leaf margins were modified to cover the PTV by 80% of the prescribed dose. The linear accelerator used was MHCL-15DP (Mitsubishi Electronics, Japan) until September 2015, and TrueBeam (Varina Medical Systems, USA) after October 2015. Treatment planning used of 6–8 beams, including 4–6 non-coplanar beams.
The treatment for patients with implanted fiducial markers was performed under motion tracking using a real-time tumor tracking system (Mitsubishi Electronics, Japan) until September 2015 and SyncTraX (Shimadzu Corporation, Japan) after October 2015. In brief, the system consists of two sets of X-ray tubes under the floor and image intensifiers on the ceiling. The fiducial marker implanted near the tumor is easily visible on the radiograph and is tracked in real time. The position of the marker is recognized as a surrogate of the tumor position. The treatment beam turns on only when the marker is located within designated area. The detailed method has been described in literature earlier .
The treatment for patients without fiducial marker was performed under light free breathing.
The medical charts were reviewed and data pertaining to age, sex, performance status (PS), body mass index (BMI), operability, smoking history (current or past vs. never smoker), the presence of diabetes mellitus (DM), the presence of pathological or cytological confirmation, tumor diameter, irradiation method (respiratory gated or not), and pretreatment Hb levels were obtained.
The World health organization defines BMI < 18.5 as underweight  Base on the BMI (kg/m2) (calculated as follow: body weight (kg) / [height (m)]2), the patients were classified into two groups (BMI < 18.5 and ≥ 18.5).
The survival periods were calculated from the completion of the SBRT.
The associations between Hb levels and other categorical variables were tested using Mann-Whitney U test and the correlation with continuous variables was tested by Spearman's rank correlation coefficient. Local control (LC) and overall survival (OS) rates were calculated using the Kaplan–Meier method, and group comparisons were made using the log-rank test. Univariate and multivariate Cox proportional hazard regression models were used to estimate the LC and OS rates. Variables for which the p-values were < 0.10 in the univariate analysis were included in the multivariate analysis. Receiver-operating characteristic (ROC) analysis was performed to determine the optimal cut-off values for the pretreatment Hb level. A p-value < 0.05 was considered to indicate a statistically significant difference.