Medical charts of all patients were reviewed through the electronic hospital information system by the study coordinator (YS). Primary lesions accounted for 83% and 77% of cases in the balloon and nonballoon groups, respectively (Table 1). Both groups were comparable in terms of patients’ age, gender, location of tumor, type of tumor, history of recurrence, and type of resection (Table 2). The outcomes we compared in this study included intraoperative blood loss, amount of blood transfusion, operative time, and postoperative wound drainage (Table 3). There was no difference between the two groups in achieving surgical margins for primary malignant lesions. No patient was lost to followup for at least 6 months after surgery with a median followup of 38 months (range, 7-106 months) in the balloon group and 43 months (range, 9-127 months) in the nonballoon group.
In the intervention group in which aortic balloon occlusion was applied, the diagnoses consisted of 13 primary benign tumors, 12 primary malignant tumors, and five metastatic carcinomas. Nineteen of 30 patients had lower lumbar spine involvement to the level of the L5 vertebra, seven patients were affected to the level of L4, whereas another four were affected to the level of L3. Among the control group, in which an aortic balloon was not applied, the diagnoses included 14 primary benign tumors, six primary malignant tumors, and six metastatic carcinomas. Eighteen of 26 patients had L5 involvement, five patients were affected to the level of L4, whereas another three were affected to the level of L3. Of the six patients with primary malignant tumors in the control group, three patients had Stage 2 hypertension, one patient had a tortuous aorta, one patient was younger than 12 years old, and one patient had a history of aortic occlusion during previous surgery. Of the seven patients with giant cell tumors in the control group, five patients had Stage 2 hypertension, one patient had low renal artery bifurcation, and one patient had unstable aortic plaque.
Of the 30 patients in the intervention group, 11 were considered to have a high risk for perioperative bleeding as a result of a history of recurrent malignancy. Of the 26 patients in the control group, five were recurrent cases. Among these five patients, three patients had Stage 2 hypertension and two patients had balloon occlusion in previous surgeries. In general, transfusion was considered when hemoglobin was or was estimated to be < 8 g/dL.
The intraoperative blood loss was estimated by the sum of the volume of mechanical suction and absorption of dressings and sponges. The need for blood transfusion was determined by assessing the hemodynamic status of the patient; anesthesiologists monitored intraoperative blood pressure, pulse rate, and hemoglobin values.
Unlike the control group in which operating time is recorded from incision to wound closure, the start time for the balloon group was recorded from the moment of balloon insertion, which is done before incision. The postoperative drainage volume was recorded by nurses at the end of the first 12 hours after surgery. The perioperative complications were collected from the patients’ records and the late balloon-related complications including aortic aneurysms and renal dysfunctions were evaluated by abdominal CT scan and a blood creatinine test, respectively, at the time of followup until 6 months after surgery.
An independent t-test was used to compare the mean age between the two groups. Pearson’s chi square test and Fisher’s exact test were performed to compare the demographic data including gender, location of the tumor, type of tumor, etc. A Mann-Whitney U test was applied to compare clinical outcomes, including intraoperative blood loss, operating time, volume of transfusion, and postoperative wound drainage. The level of significance was p < 0.05. Statistical analyses were conducted using IBM SPSS Statistics 22.0 (IBM, Armonk, NY, USA) and GraphPad Prism 7.02 (GraphPad, La Jolla, CA, USA).
The implementation of aortic balloon occlusion was associated with lower median intraoperative blood loss in surgery of sacral tumors involving the lower lumbar spine (2000 mL, range, 400-6000 mL versus 2650 mL, range, 550-6800 mL, respectively; median difference, 605 mL; 95% confidence interval [CI], 100-1500 mL; p = 0.035) (Table 3). Median volumes of blood product transfusion and packed red blood cells were lower with the application of balloon occlusion, from 2300 mL (range, 400-6000 mL) to 1600 mL (range, 600-5400 mL; median difference, 700 mL; 95% CI, 20-1360; p = 0.029) and from 1200 mL (range, 200-3000 mL) to 800 mL (range, 200-3000 mL; median difference, 400 mL; 95% CI 20-760; p = 0.018), respectively.
No difference was found in the total operative time between the two groups (median time, 215 minutes, range, 110-430 minutes for the balloon group versus 225 minutes, range, 115-340 minutes for the control group; median difference, 10 minutes; 95% CI, -40 to 30 minutes; p = 0.902).
Balloon-related vascular complications included local hematoma at the puncture site in five patients, femoral artery spasm in three patients, lower limb ischemia in one patient, and femoral artery pseudoaneurysm in one patient. Hematomas were watched carefully but no further treatment was performed. Femoral artery spasms were all reversible after removal of the catheter sheath. The patient with limb ischemia was treated with embolectomy, resulting in preservation of the limb. One patient who developed a femoral artery pseudoaneurysm at the puncture site 2 weeks after retrieval of the catheter sheath was successfully treated by transarterial thrombin injection and a covered stent, which can repair the damaged vessel with a fabric coating. Two patients were diagnosed with acute kidney injury related to the inappropriate position of the balloon. Both patients recovered after 1 week of hemodialysis and strict fluid and electrolyte monitoring. In addition, delayed postoperative bleeding, defined as excessive wound drainage that required surgical exploration or immediate embolization, occurred in two patients in the balloon group and in one patient in the nonballoon group. All of them were successfully treated by emergent embolization. No complications related to large amounts of blood loss such as ischemic heart disease and cerebral vascular accident were determined in either group. No perioperative mortality was determined in this cohort. No other long-term adverse events such as abdominal aneurysm or chronic renal failure were reported regarding the aortic balloon occlusion.
Massive intraoperative bleeding during sacrectomy is a major challenge for orthopaedic surgeons . Several approaches used to minimize bleeding include absorbable hemostatic agents [8, 30], hemostatic devices [6, 22], hypotensive epidural anesthesia , and preoperative arterial embolization [12, 20, 31]. Although embolization is effective in limiting blood flow in major arteries, complications such as gluteal or lumbosacral necrosis may still occur as a result of extensive embolization and irreversible blood supply [5, 24]. The use of an aortic balloon seems promising because it can directly decrease blood flow in the distal aorta  while the impact on arterial blood supply to the adjacent normal structures after balloon deflation is minor. However, there have been no studies that have evaluated whether the balloon could still function when sacral tumors extend cranially into the lower lumbar spine. In our study, we found that the aortic balloon occlusion was associated with reduced intraoperative blood loss and transfusion use in lumbosacral tumor resections. Notably, the total operative time (including balloon insertion) was not prolonged compared with the control group. We also found that vascular events were the most common balloon-related complications with acute kidney injury being a newly occurring complication that has not been previously reported.
Our study had some limitations as a result of its retrospective nature. First, the assignment of patients to the balloon and control groups was not random. Patients with potentially higher risks of bleeding were clinically indicated for balloon use. However, some of those patients did not receive balloon occlusion as a result of prioritized contraindications for safety reasons. In effect, the exclusion of those patients allowed for a more normalized distribution of patient characteristics, which makes the two groups of patients more comparable. Second, the variation in anesthetic techniques among anesthesiologists could affect the estimation of bleeding volume and calculation of transfusion volume, which may lead to inevitable observer differences. Third, the two groups are not directly comparable with respect to diagnosis as a result of combined histologic types. In addition, combining primary tumors with metastatic lesions is another limitation of the study, which could be solved by expanding the number of patients and evaluating primary and metastatic tumors separately in future studies. We also had some soft tissue lesions, which further complicates the analysis, but in all instances, there was bony involvement. We included no patients who had soft tissue neoplasms without bony involvement. Lastly, a followup of 6 months may not be sufficient to evaluate long-term balloon-related complications.
In patients who had aortic balloon occlusion, there was an associated reduction in mean intraoperative blood loss, mean volume of blood product transfusion, and mean volume of packed red blood cell transfusion compared with those who did not receive the balloon. Aortic balloon occlusion has been shown to reduce blood loss in a few studies on sacropelvic tumors [10, 15, 17, 26, 29, 32] (Table 4). One study on a large series of 215 patients reported that the mean intraoperative blood loss volume (2236 mL) was lower than that of the patients without occlusions (3935 mL) . Another study of a total of 137 patients demonstrated that use of balloon occlusion reduced intraoperative blood loss and transfusion (678 mL and 351 mL versus 1619 mL and 1106 mL) . In our study, we expanded the indication for an aortic balloon to include tumors extending into the lumbar region of the spine and showed its efficacy in controlling bleeding. According to our observations, iliolumbar arteries arising from the internal iliac arteries, the median sacral artery, collaterals from the external iliac arteries as well as segmental lumbar arteries arising more cephalad from the aorta contribute to the blood supply for sacral-lumbar tumors of the spine. As such, more elaborate and precise positioning of the balloon just caudal to the renal artery bifurcation is a necessity. In this way, the blood flow to multiple pairs of segmental lumbar arteries could be maximally reduced. This preoperative arterial embolization has also been shown to successfully reduce intraoperative blood loss . Importantly, superselective embolization, which is crucial to the optimization of bleeding control, needs to be performed by an experienced interventional radiologist with a high level of expertise in lumbosacral tumor procedures [3, 18, 21]. Furthermore, the cost-effectiveness analysis of the aortic balloon occlusion versus preoperative arterial embolization has not, as of yet, been elucidated.
As shown in our results, the use of the balloon does not prolong the total operative time of lumbosacral tumor surgery. Because the insertion of the aortic balloon takes approximately 30 minutes, the time required for tumor resection is essentially shortened, which is consistent with previous studies. Previous case-control studies on sacral tumor operations have shown that use of an aortic balloon reduces operative time [15, 26]. Both studies determined a historical group of patients who did not use balloon occlusion as the control. Furthermore, because the balloon is inserted before the operative incision is performed, it might be reasonable to add the insertion time to the operative time, which can more accurately reflect the total operative time in which the patient is under general anesthesia and surgical intervention. According to our understanding, aortic balloon occlusion as well as preoperative embolization  could reduce operative time mainly through improving visualization of the operative field and accessibility of lesions. Interestingly, a study revealed that hypotensive epidural anesthesia could reduce blood loss with the added benefit of reducing operative time in pelvic and sacral bone tumor resections . Although the potential risk for adverse events with the use of hypotensive epidural anesthesia remains to be evaluated, it may serve as an alternative approach for patients contraindicated for an aortic balloon.
In our study, complications were common in patients who had the aortic balloon. Minor vascular events including local hematoma and femoral artery spasm occurred in eight (27%) of the 30 patients, which was higher than that (6%) reported in a large series of 911 patients after intraaortic balloon pump treatment . Major vascular events requiring surgical intervention occurred in two (7%) of 30 patients (one femoral artery pseudoaneurysm and one ischemia of a limb), which was comparable to an incidence rate of 9% reported in a prospective study on percutaneous coronary angioplasty . These complications might be related to the use of a thickened 12-Fr sheath leaving a large hole in the artery after retrieval. Some suture-mediated closure devices are available to minimize vascular complications . Aside from the vascular complications at the puncture site, we also identified two patients with postoperative acute kidney injury, which has not been reported as a balloon-related complication in previous studies (Table 4). We suspect that the balloon slid cranially and obstructed at least one side of the renal artery during the position change after insertion. The resulting renal ischemia, in addition to volume insufficiency and adverse drug effects, may have impaired renal function and caused the acute kidney injury. As such, we suggest that an angiogram be performed to identify balloon location after any position changes during the operation. Embolization has a good safety profile in patients with bone tumors in general [12, 18]. However, acute kidney injury, which is induced by contrast rather than renal ischemia, has been mentioned in a previous study of patients with metastatic bone tumors undergoing embolization . Although the overall incidence is low, this severe complication must be considered, especially in patients with preexisting renal impairment, when opting to use embolization. Besides, we also notice that severe complications including gluteal or lumbosacral necrosis could possibly occur if superselective embolization was not achieved [5, 24]. We propose that combining limited preoperative superselective embolization with intraoperative aortic balloon occlusion may potentially optimize their respective efficacies and minimize their complications.
In conclusion, this study demonstrated that placement of the aortic balloon at a level just caudal to the renal artery bifurcation was associated with lower intraoperative blood loss and need for transfusion in lumbosacral tumor resections. However, procedure-specific complications were common and there was no benefit to total operative time. We suggest that the surgical procedures still need to be further refined to minimize complications. We also recommend that prospective studies be undertaken to confirm the efficacy of aortic balloon occlusion in surgery for lumbosacral tumors.
We thank Ms Yanchun She for assistance with data collection.
1. Bilsky MH, Boland PJ, Panageas KS, Woodruff JM, Brennan MF, Healey JH. Intralesional resection of primary and metastatic sarcoma involving the spine: outcome analysis of 59 patients. Neurosurgery. 2001;49:1277–1286.
2. Boriani S, Chevalley F, Weinstein JN, Biagini R, Campanacci L, Delure F, Piccill P. Chordoma of the spine above the sacrum–treatment and outcome in 21 cases. Spine. 1996;21:1569–1577.
3. Chen K, Zhou M, Yang H, Qian Z, Wang G, Wu G, Zhu X, Sun Z. Pre-operative embolization facilitating a posterior approach for the surgical resection of giant sacral neurogenic tumors. Oncol Lett. 2013;6:251–255.
4. Cheng EY, Ozerdemoglu RA, Transfeldt EE, Thompson RC. Lumbosacral chordoma–prognostic factors and treatment. Spine. 1999;24:1639–1645.
5. di Summa PG, Schaffer C, Zaugg P, Bauquis O, Raffoul W. Lumbar artery perforator (LAP) flap: a salvage tool for extended lumbo-sacral necrosis after bilateral internal iliac arteries embolization. Case Rep Plast Surg Hand Surg. 2016;3:20–24.
6. Filippakis GM, Leandros M, Albanopoulos K, Genetzakis M, Lagoudianakis E, Pararas N, Konstandoulakis MM. The use of spray electrocautery to control presacral bleeding: a report of four cases. Am Surg. 2007;73:410–413.
7. Freeman AK, Thorne CJ, Gaston CL, Shellard R, Neal T, Parry MC, Grimer RJ, Jeys L. Hypotensive epidural anesthesia reduces blood loss in pelvic and sacral bone tumor resections. Clin Orthop Relat Res. 2017;475:634–640.
8. Germanos S, Bolanis I, Saedon M, Baratsis S. Control of presacral venous bleeding during rectal surgery. Am J Surg. 2010;200:e33–35.
9. Giles TD, Berk BC, Black HR, Cohn JN, Kostis JB, Izzo JL Jr, Weber MA. Expanding the definition and classification of hypertension. J Clin Hypertens (Greenwich). 2005;7:505–512.
10. Guo W, Ji T, Tang XD, Yang Y. Outcome of conservative surgery for giant cell tumor of the sacrum. Spine. 2009;34:1025–1031.
11. Holm NR, Sindberg B, Schou M, Maeng M, Kaltoft A, Bottcher M, Krusell LR, Hjort J, Thuesen L, Terkelsen CJ, Christiansen EH, Botker HE, Kristensen SD, Lassen JF; CLOSE-UP study group. Randomised comparison of manual compression and FemoSeal vascular closure device for closure after femoral artery access coronary angiography: the CLOSure dEvices Used in everyday Practice (CLOSE-UP) study. EuroIntervention. 2014;10:183–190.
12. Hosalkar HS, Jones KJ, King JJ, Lackman RD. Serial arterial embolization for large sacral giant-cell tumors: mid- to long-term results. Spine (Phila Pa 1976). 2007;32:1107–1115.
13. Kresowik TF, Khoury MD, Miller BV, Winniford MD, Shamma AR, Sharp WJ, Blecha MB, Corson JD. A prospective study of the incidence and natural history of femoral vascular complications after percutaneous transluminal coronary angioplasty. J Vasc Surg. 1991;13:328–333; discussion 333–335.
14. Li D, Guo W, Tang X, Ji T, Zhang Y. Surgical classification of different types of en bloc resection for primary malignant sacral tumors. Eur Spine J. 2011;20:2275–2281.
15. Luo Y, Duan H, Liu WL, Min L, Shi R, Zhang WL, Zhou Y, Tu CQ. Clinical evaluation for lower abdominal aorta balloon occluding in the pelvic and sacral tumor resection. J Surg Oncol. 2013;108:148–151.
16. Meharwal ZS, Trehan N. Vascular complications of intra-aortic balloon insertion in patients undergoing coronary reavscularization: analysis of 911 cases. Eur J Cardiothorac Surg. 2002;21:741–747.
17. Mi C, Lu HL, Liu H. Surgical excision of sacral tumors assisted by occluding the abdominal aorta with a balloon dilation catheter: a report of 3 cases. Spine. 2005;30:E614–E616.
18. Nair S, Gobin YP, Leng LZ, Marcus JD, Bilsky M, Laufer I, Patsalides A. Preoperative embolization of hypervascular thoracic, lumbar, and sacral spinal column tumors: technique and outcomes from a single center. Interv Neuroradiol. 2013;19:377–385.
19. Pazionis TJ, Papanastassiou ID, Maybody M, Healey JH. Embolization of hypervascular bone metastases reduces intraoperative blood loss: a case-control study. Clin Orthop Relat Res. 2014;472:3179–3187.
20. Prabhu VC, Bilsky MH, Jambhekar K, Panageas KS, Boland PJ, Lis E, Heier L, Nelson PK. Results of preoperative embolization for metastatic spinal neoplasms. J Neurosurg. 2003;98:156–164.
21. Rossi C, Ricci S, Boriani S, Biagini R, Ruggieri P, De Cristofaro R, Roversi RA, Khalkhali I. Percutaneous transcatheter arterial embolization of bone and soft tissue tumors. Skeletal Radiol. 1990;19:555–560.
22. Saurabh S, Strobos EH, Patankar S, Zinkin L, Kassir A, Snyder M. The argon beam coagulator: a more effective and expeditious way to address presacral bleeding. Tech Coloproctol. 2014;18:73–76.
23. Stephens GC, Schwartz HS. Lumbosacral chordoma resection–image integration and surgical planning. J Surg Oncol. 1993;54:226–232.
24. Suzuki T, Shindo M, Kataoka Y, Kobayashi I, Nishimaki H, Yamamoto S, Uchino M, Takahira N, Yokoyama K, Soma K. Clinical characteristics of pelvic fracture patients with gluteal necrosis resulting from transcatheter arterial embolization. Arch Orthop Trauma Surg. 2005;125:448–452.
25. Tang X, Guo W, Yang R, Tang S, Ji T. Risk factors for blood loss during sacral tumor resection. Clin Orthop Relat Res. 2009;467:1599–1604.
26. Tang XD, Guo W, Yang RL, Tang S, Dong S. Use of aortic balloon occlusion to decrease blood loss during sacral tumor resection. J Bone Joint Surg Am. 2010;92:1747–1753.
27. Verlaan JJ, Kuperus JS, Slooff WB, Hennipman A, Oner FC. Complications, secondary interventions and long term morbidity after en bloc sacrectomy. Eur Spine J. 2015;24:2209–2219.
28. Wuisman P, Lieshout O, Sugihara S, van Dijk M. Total sacrectomy and reconstruction–oncologic and functional outcome. Clin Orthop Relat Res. 2000;381:192–203.
29. Yang L, Chong-Qi T, Hai-Bo S, Zhang L, Tian-Fu Y, Hong D, Fu-Xing P. Appling the abdominal aortic-balloon occluding combine with blood pressure sensor of dorsal artery of foot to control bleeding during the pelvic and sacrum tumors surgery. J Surg Oncol. 2008;97:626–628.
30. Zhang CH, Song XM, He YL, Han F, Wang L, Xu JB, Chen CQ, Cai SR, Zhan WH. Use of absorbable hemostatic gauze with medical adhesive is effective for achieving hemostasis in presacral hemorrhage. Am J Surg. 2012;203:e5–8.
31. Zhang K, Chen K, Zhou M, Chen H, Lu J, Yang H. Extremely large giant-cell tumor of sacrum with successful resection via posterior approach. Spine J. 2015;15:1684–1685.
32. Zhang L, Gong Q, Xiao H, Tu C, Liu J. Control of blood loss during sacral surgery by aortic balloon occlusion. Anesth Analg. 2007;105:700–703.
© 2018 Lippincott Williams & Wilkins LWW
33. Zhang YD, Guo W, Yang RL, Tang XD, Yan TQ, Ji T. Malignant pelvic tumors involving the sacrum: surgical approaches and procedures based on a new classification. Orthop Surg. 2016;8:150–161.