Objective To investigate the safety and feasibility of en-bloc resection of a primary sacral chordona based on a 3-dimensional printing model.Methods 31 patients with primary sacral chordoma underwent en-bloc resection via a onestage posterior approach or combined anterior and posterior approaches in our oncology department from January 2013 to December 2014.They comprised 21 males and 10 females of mean age (49.2±12.5) years (range,26-67 years).Preoperative 3-D printing models were created by 3D printing technology,it included tumor tissue,the around vascular and nerves involved in sacral chordoma.The sacral chordomas were en-bloc resection with decompression and internal fixation.Results With the mean (29.0±6.8)months follow-up (range from 19 to 41),all patients underwent en bloc excision via 26 cases with posterior approach,5 cases combined posterior and anterior approaches in one stage.The mean operative time and estimated blood loss were (275.0±58.1) min and (3 250.0±1 304.4) ml,respectively.The visual analogue scale (VAS) score was (5.6±1.9) in average (range from 3 to 9) at preoperation,and (2.0±1.5) at post-operation,which was significantly lower than that of preoperation,and the pain was relief obviously.There were 13 cases in grade C,11 cases in grade D,7 cases in grade E of American Spinal Injury Association (ASIA) grade neurological function before surgery,compared with the pre-operation,there were 5 cases in grade C,6 cases in grade D,20 cases in grade E of post-operation,which was significantly improved.MSTS (Musculoskeletal Tumor Society) 93 score was 6-29 points (20.0％-96.7％) at the follow-up 3 months after surgery,with the average of (19.8 ± 5.8) points,which excellent in 8 cases,good in 14 cases,general in 5 cases,poor in 4 cases.Two cases of dysporia for the reasons of resecting on one side of the S1,2 nerve roots involved by the sacral chordoma,after sacrificing the nerve root of complete tumor resection,the urine left dysfunctional,while the pain of other 29 patients were thoroughly relief after surgery.The ones were relieved with the disturbance of sensation of the perineum before the operation.2 cases were recovery of leakage of cerebrospinal by the drainage of lumbar cistern with normal temperature.One hypostatic pneumonia patient was cured by anti-inflammatory.One with the urinary infection got better by the effective bladder irrigation,which had diabetics mellitus with the bladder stoma before.1 case of skin necrosis due to vascular thrombosis before operation,recevied flap translocation half month after surgery,got recovery 3 months later.Only one underwent tumor resection for the recurrence at 15 months follow-up.Conclusion It is feasible and safe to perform en bloc resection of primary sacral chordoma.This is the most effective means of managing method of the marginal resection of the tumor.Preoperative 3-D printing modeling enables better anatomical understanding of the relationship between the tumor,and can avoid vascular and nerves tissue injury,which can also assist in planning the surgical procedure,and be worth recommendation.

Objective:To propose and verify a surgical classification system for the axial primary malignant and aggressive benign tumor.Methods:The CZH surgical classification system was originally developed for the axial primary malignant and aggressive benign tumor. The CZH surgical classification system includes seven types, according to the anatomic features and the extension of tumor violation. A total of 136 patients (79 males and 57 females) with axial primary malignant and aggressive benign tumor from multiple tertiary centers who received surgery from July 2006 to July 2019 were included. The average age was 44.40±17.55 years (8-83 years) old. There were 99 malignant tumors and 37 aggressive benign tumors included. The number of patients with each classification was presented as followed, Type I 13, Type II 15, Type IIIa 3, Type IIIb 20, Type IVa 43, Type IVb 12, Type Va 21, Type Vb 3, Type VI 2, Type VIIa 3 and Type VIIb 1. Surgical procedures were selected according to different types in classification. The inter- and intra-observer consistencies were evaluated by the Kendall's W test. The VAS, Frankel score, overall survival and recurrence free survival were recorded during the follow-up. Results:The inter- and intra-observer consistent coefficient was 0.973 and 0.996, respectively ( P<0.05). The single posterior approach was adopted for the Type II tumors. Other patients underwent surgery by the combined antero-posterior approach. The majority in anterior approach (113 cases) was the modified submandibular approach. The reconstruction modes included anterior "T" shape titanium mesh (112 cases) or the 3D printed prothesis (7 cases) combined with the posterior occipto-cervical fusion (92 cases) or the pedicle screw system (44 cases). The average surgical duration and the volume of intraoperative bleeding was 348.40±136.14 min (60-760 min) and 1 225.69±859.40 ml (80-4 000 ml), respectively. The operation duration and volume of intraoperative bleeding among each type were with statistical difference. The patients with Type IV, V tumors had longer operation duration than those with Type II tumors. Those with Type V and VII tumors had longer operation duration than those with Type I tumors. The patients with Type V tumors had more intraoperative bleeding than those with Type I-IV tumors. The average preoperative VAS score was 4.15±2.25 and then was reduced significantly to 0.62±0.71 and 0.38±0.59 at one and three months after operation, respectively. The Frankel score was also significantly ameliorated at one and three months postoperatively. There were 22 postoperative complications (16.2%). The complications included cerebral spinal fluid leak (12.5%), dysphagia and/or dysphonia (7.4%), dyspnea (5.1%), wound infection (3.7%), wound hemorrhage (2.2%) and pharyngeal dehiscence (1.5%). The incidence of postoperative complication was 25.9% in Type IV-VII tumors, while 11.8% in Type I-III tumors. Conclusion:CZH surgical classification system was verified with high observer consistency. This classification system could assist surgeons to select proper surgical approaches, resection modes and reconstruction modes, and thus ensure the safety of surgery and reduce the recurrence. The tumors in Type IV, V and VII may be with more challenging for surgeons. The incidence of postoperative complication in Type IV-VII tumors may be higher than that in Type I-III tumors.