From the Society for Clinical Vascular Surgery
Robotic inferior vena caval surgery Victor J. Davila, MD,a Cristine S. Velazco, MD, MS,a William M. Stone, MD,a Richard J. Fowl, MD,a Haidar M. Abdul-Muhsin, MBBS,a Erik P. Castle, MD,b and Samuel R. Money, MD, MBA,a Phoenix, Ariz
ABSTRACT Objective: Inferior vena caval (IVC) surgery is uncommon and has traditionally been performed through open surgical approaches. Renal cell carcinoma with IVC extension generally requires vena cavotomy and reconstruction. Open removal of malpositioned IVC filters (IVCF) is occasionally required after endovascular retrieval attempts have failed. As our experience with robotic surgery has advanced, we have applied this technology to surgery of the IVC. We reviewed our institution’s experience with robotic surgical procedures involving the IVC to determine its safety and efficacy. Methods: All patients undergoing robotic surgery that included cavotomy and repair from 2011 to 2014 were retrospectively reviewed. Data were obtained detailing preoperative demographics, operative details, and postoperative morbidity and mortality. Results: Ten patients (6 men) underwent robotic vena caval procedures at our institution. Seven patients underwent robotic nephrectomy with removal of IVC tumor thrombus and retroperitoneal lymph node dissection. Three patients underwent robotic explantation of an IVCF after multiple endovascular attempts at removal had failed. The patients with renal cell carcinoma were a mean age of was 65.4 years (range, 55-74 years). Six patients had right-sided malignancy. All patients had T3b lesions at time of diagnosis. Mean tumor length extension into the IVC was 5 cm (range, 1-8 cm). All patients underwent robotic radical nephrectomy, with caval tumor thrombus removal and retroperitoneal lymph node dissection. The average operative time for patients undergoing surgery for renal cell carcinoma was 273 minutes (range, 137-382 minutes). Average intraoperative blood loss was 428 mL (range, 150-1200 mL). The patients with IVCF removal were a mean age of 33 years (range, 24-41 years). Average time from IVCF placement until robotic removal was 35.5 months (range, 4.3-57.3 months). Before robotic IVCF removal, a minimum of two endovascular retrievals were attempted. Average operative time for patients undergoing IVCF removal was 163 minutes (range, 131-202 minutes). Intraoperative blood loss averaged 250 mL (range, 150-350 mL). All procedures were completed robotically. The mean length of stay for all patients was 3.5 days (range, 1-8 days). All patients resumed ambulation on postoperative day 1. Nine patients resumed a regular diet on postoperative day 2. One patient with a renal tumor sustained a colon injury during initial adhesiolysis, before robotic radical nephrectomy, which was recognized at the initial operation and repaired robotically. Robotic radical nephrectomy and caval tumor removal were then completed. No blood transfusions were required intraoperatively, but three patients required blood transfusions postoperatively. Conclusions: Although robotic IVC surgery is uncommon, our initial limited experience demonstrates it is safe and efficacious. (J Vasc Surg: Venous and Lym Dis 2016;-:1-6.)
Surgery involving the inferior vena cava (IVC) is rare. There will be w60,000 newly diagnosed renal cell carcinomas (RCCs) in the United States in 2015.1 Concomitant tumor thrombus extension into the IVC is found in 4% to 10% of RRC cases.2 Level 0 tumors are limited to the renal vein, level 1 tumors extend into the vena cava #2 cm above the renal vein, level 2 tumors extend >2 cm
From the Division of Vascular Surgery, Department of Surgery,a and the Department of Urology,b Mayo Clinic. Author conflict of interest: none. Presented at the Forty-fourth Annual Symposium of the Society for Clinical Vascular Surgery, Las Vegas, Nev, March 12-16, 2016. Correspondence: Victor J. Davila, MD, 5777 E Mayo Blvd, Phoenix, AZ 85024 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the Journal policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 2213-333X Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvsv.2016.08.003
above the renal vein, level 3 tumors extend to the hepatic veins but below the diaphragm, and level 4 tumors extend above the diaphragm. An open surgical approach is traditionally used in performing nephrectomy, retroperitoneal lymph node dissection (RPLND), and cavotomy with tumor thrombus removal. Vascular surgeons are frequently called on to perform the vena caval reconstruction after tumor excision. Laparoscopic surgery for RCC has been successfully performed and is widely used in cases without tumor thrombus extension into the IVC. More recently, robotic surgery is allowing surgeons the ability to achieve vascular control in a safer and more effective way than laparoscopy. This has allowed for complete control of the IVC and allowed caval reconstruction to be performed out without the necessity of an open surgical component. Vascular surgeons are also called on to manage deep vein thrombosis, including placement and removal of IVC filters (IVCFs). Although most IVCFs that are placed 1
2
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders ---
today are retrievable by design, less than half are removed.3 Retrieval can be performed endovascularly; however, complications can occur after IVCF placement that require surgical explantation, including filter migration, IVC and adjacent organ perforation, and filter fracture.4 Often, open surgery is required for safe removal given the vascular control that must be obtained during the procedure. RCC with tumor thrombus extension into the IVC and IVCF explantation can both be technically challenging; however, both (and surgery on the IVC in general) have been performed safely with minimally invasive techniques. In this report we describe our experience with robotic IVC surgery.
METHODS After Institutional Review Board approval, databases were queried for patients who underwent robotic surgery involving the IVC from 2011 to 2014. This retrospective review included medical and surgical history, presentation, and demographic information. Operative details included operative time, blood loss, and need for transfusion. Postoperative details included time until regular diet, time to ambulation, and hospital length of stay. Postoperative morbidity and mortality were also noted. All procedures were performed with the da Vinci Si robotic platform (Intuitive Surgical, Inc, Sunnyvale, Calif). In cases of RCC, the primary approach for nephrectomy, RPLND, and tumor thrombus removal is robotic. An open approach is considered if preoperative imaging precludes safe vascular control. The main reason a patient would not be considered for a robotic approach is in cases of level 3 and level 4 tumors, because obtaining robotic-assisted vascular control above the hepatic veins can be very challenging. In patients requiring IVCF explantation, a minimum of two endovascular attempts at removal were required before consideration of robotic removal. Surgical technique for robotic nephrectomy and IVC tumor removal. The patient is placed in the modified right-side-up left-side-down position. Access is gained to the peritoneal cavity using the Veress needle technique in the midline. We place three 8-mm robotic ports in an approximately linear fashion in the midclavicular line while maintaining a distance between ports of 9 or 10 cm, or one hands’ breadth. A liver retractor port is placed in the subxiphoid region through a 5-mm port with two 12-mm assistant ports. The first port is placed just above the umbilicus, and the second is placed midway between the xiphisternum and the umbilicus. The robot is docked, and the location of the fourth robotic port is carefully planned in the lowermost location in the ipsilateral lower abdominal quadrant to avoid clashing with the rest of the robotic arms.
2016
First, electrocautery is used to mobilize the right colon, and the duodenum identified and Kocherized athermally. The vena cava is isolated with blunt dissection and pinpoint electrocautery. The ipsilateral renal artery is divided with an endovascular stapler. The lumbar veins are then identified and ligated. We usually tie the lumbars with 4-0 silk suture and then transect with the robotic scissors. Control of the infrarenal IVC is then obtained. The contralateral renal vein is dissected, and a vessel loop placed around it. The IVC is completely mobilized. The suprarenal IVC is then dissected. A vessel loop is doubly looped to maintain adequate control of the suprarenal vena cava. Next, control is obtained by pulling up on the vessel loops and using Satinsky bulldogs on the renal vein. Rumel tourniquets are tightened on both the suprarenal and infrarenal vena cava by cinching down the vessel loops. The robotic Pott scissors are used to make a cavotomy on the anterior aspect at the level of the renal veins that is extended superiorly until normal IVC is identified. The tumor thrombus is then removed en bloc with the ipsilateral kidney, and the specimen is extracted from the abdominal cavity. The cavotomy is closed with 3-0 or 4-0 Prolene (Ethicon, Somerville, NJ) suture on a small half needle in a running fashion. Immediately before complete closure, the clamps are released to flush air out of the vena cava. The suture line is then secured, and after hemostasis is endured, the clamps and then the vessel loops are removed. Hemostasis in the remaining surrounding areas is obtained with pinpoint electrocautery. A fibrin sealant (Evicel; Ethicon) or a hemostatic matrix (Floseal; Baxter, Deerfield, Ill), or both, can be applied as deemed necessary for further hemostasis. Retrieval of the IVCF is completed in a similar manner with control of the renal veins and vena cava (Fig). Control or ligation of the renal arteries is not performed. After the cavotomy is made, the filter can be seen and removed. In growth of endothelium around the filter may require dissection before removing the filter. If necessary, the tips of the tines of the filter may be broken off and removed individually to facilitate removal from the IVC. The remaining filter can then be removed in its entirety. If the filter is penetrating the wall, the venotomy can be made at this site, and the size can be adjusted to accommodate removal. The filter is then brought out of the abdomen through the 12-mm port. The steps are summarized in Table I.
RESULTS Ten patients (six men) underwent robotic operations involving the vena cava at our institution. All procedures included cavotomy with reconstruction. Reconstruction consisted of primary closure of the cavotomy with
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume
-,
Number
3
-
Fig. Left inset, Planned cavotomy before inferior vena cava filter (IVCF) removal with tines protruding through caval wall. Main illustration, Robotic removal of IVCF. Right inset, Venotomy closure.
running Prolene suture. All procedures were completed robotically, with no conversion to laparoscopy or open laparotomy. At time of surgery, patients were an average age of 55.6 years (range, 24.1-73.6 years; Table II). Average body mass index was 30.8 kg/m2 (range, 21.7-39.4 kg/m2). Preoperative
characteristics consisted of diabetes in three and hypertension in seven. Six patients had previous abdominal surgery, which included hysterectomy in two, and laparoscopic appendectomy, umbilical hernia repair, laparoscopic cholecystectomy, ventral hernia repair without mesh, laparoscopic left radical nephrectomy, and an
Table I. Technical tips and considerations for robotic inferior vena caval (IVC) surgery 1. Ensure proper robotic port placement because this plays a key role in successful completion of this operation. 2. An experienced laparoscopic surgeon should assist at the bedside. This is of extreme importance to temporize bleeding while the team prepares for open conversion, if needed. 3. The surgical team should be prepared for open conversion with all instruments available and ready for immediate use. 4. Suprarenal, infrarenal IVC, and contralateral renal vein should be controlled as early as possible. 5. The right adrenal vein may need to be divided. 6. Proper and secure control of all lumbar veins should be obtained before cavotomy because they often result in bothersome bleeding if missed. This is preferably performed using silk ties in the exact same manner as in open surgery. 7. Excision of part of the IVC wall may be needed if there is tumor thrombus invasion or endothelial growth over the IVC filter (IVCF). 8. We prefer circumferential control of the IVC with vessel loops. Although balloon occlusion can be obtained in the distal IVC, troublesome back bleeding can occur proximally and from unidentified lumbar veins. This is why we believe good dissection and early identification of lumbar veins before cavotomy is important.
4
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders ---
2016
Table II. Demographics and intraoperative details Variablea
Total cavotomy (N ¼ 10)
Thrombus (n ¼ 7)
IVCF retrieval (n ¼ 10)
Male sex
6
5
1
Age, years
55.6 (24.1-73.6)
65.4 (54-73)
Operative time, minutes
240 (163-273)
273 (137-382)
Blood loss, mL
375 (150-1200)
428 (150-1200)
32 (24-41) 163 (131-202) 250 (150-350)
IVCF, Inferior vena cava filter. a Categoric variables are shown as the number and continuous variables as the average (range).
open right partial nephrectomy in one patient each. Preoperative imaging consisted of magnetic resonance imaging in six patients and computed tomography scan in five patients. All patients with IVCF had computed tomography scan and previous attempts at venography as the sole imaging modalities. Operative time averaged 240 minutes (range, 163-273 minutes), and blood loss averaged 375 mL (range, 1501200 mL; Table II). Complications included one colotomy in a RCC patient who had extensive adhesiolysis at the beginning of the case. The colotomy was recognized promptly and repaired robotically. A retroperitoneal hematoma developed in one patient on postoperative day 2, which was managed nonoperatively. All patients resumed ambulation on postoperative day 1. Average time to a regular diet was 1.9 days (range, 1-8 days). No patients required intraoperative transfusion, but three patients required blood transfusions postoperatively: two patients for preoperative anemia and one patient who developed a postoperative retroperitoneal hematoma. The average transfusion requirement per patient was 0.9 units (range, 1-5 units). The average length of stay was 3.5 days (range, 1-8 days). RCC experience. The patients with RCC were an average age of 65.4 years (range, 54-73 years). There were five male patients. Six of the tumors were located on the right side. All tumors were T3 (stage 3). Tumor thrombus extension into the IVC averaged 5.3 cm (range, 1-8 cm). Preoperative imaging showed the average primary tumor measurement in greatest dimension was 7.7 cm (range, 1.9-11 cm). Operative time averaged 273 minutes (range, 137-382 minutes). Blood loss averaged 428 mL (range, 150-1200 mL; Table II). Average length of stay was 3.4 days (range, 2-7 days). IVCF explantation experience. Patients undergoing IVCF explantation were an average age of 32 years (range, 24-41 years). There were two female patients and one male patient. All patients presented with abdominal pain. All patients had retrievable IVCFs, and each patient underwent a minimum of two endovascular attempts of IVCF removal before being considered for robotic explantation. IVCFs were placed an average of 1065 days before explantation (range, 129-1719 days). Operative time averaged 163 minutes (range, 131-202 minutes). Average
blood loss was 250 mL (range, 150-350 mL; Table II). Average length of stay was 3.6 days (range, 1-8 days). Primary findings at the time of explanations leading to failure of endovascular retrieval included robust endothelial ingrowth and perforation/erosion into adjacent organs (liver, duodenum/small bowel, mesentery).
DISCUSSION Minimally invasive surgery has replaced open surgery as the preferred approach in many surgical specialties, including general surgery, gynecology, and thoracic surgery. The laparoscopic approach allows for equivalent long-term outcomes compared with open procedures, with decreased rates of incisional pain, faster return to work, earlier hospital discharge, and earlier resumption of diet.5 Robotic surgery offers the benefits of the minimally invasive approach while allowing the surgeon improved visualization, more freedom to manipulate instruments and tissues, and ergonomic support during long operations. Robotic surgery is rarely used in vascular surgery, primarily because performing vascular anastomoses via laparoscopic methods are technically challenging. Therefore, vascular surgery has not embraced minimally invasive techniques for treating vascular disease apart from endovascular surgery. Endovascular technology has shown excellent results in long-term outcomes, wound infection rates, pain scores, and hospital length of stay, which are many of the same metrics by which laparoscopic and robotic surgery are measured.6 There are cases where the endovascular approach fails, and traditional surgery is required. The da Vinci robotic platform is being used to allow the surgeon to perform complex procedures with similar or improved outcomes compared with laparoscopy and open procedures. Robotic procedures are currently allowing surgeons the ability to precisely perform technically demanding procedures through a minimally invasive approach. Our institution has experience with excision of RCC and associated tumor thrombus by a laparoscopic approach.7 Vena caval surgery using the robotic platform has also been described in RCC.8-10 Our experience shows that these techniques can be applied to surgery on the vena cava with minimal morbidity and mortality. All procedures in our series were completed robotically.
Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume
-,
Number
Davila et al
5
-
Patients recovered well postoperatively, resuming ambulation and diet on postoperative days 1 and 2, respectively. Complications were inherent to the surgery and not specific to the robotic approach. One patient sustained a colotomy during extensive lysis of adhesions from a previous laparotomy, which was identified and treated at the initial operation without subsequent sequelae. A retroperitoneal hematoma developed in another patient who resumed a heparin drip on postoperative day 1 for a known hypercoagulable state. This was managed nonoperatively with blood transfusion, and the patient was discharged on postoperative day 8. There are multiple case reports and a few small series of IVC surgery currently in the literature. Abaza et al8 were the first to describe robotic IVC vascular control with tumor thrombus removal in 2010. They showed that tumor thrombi from 1 cm to 5 cm can be safely extracted from the IVC during nephrectomy, without the need for hand assistance or conversion to an open approach. Gill et al9 described a series of robotic nephrectomy and IVC tumor thrombus removals in nine patients with level 3 tumor thrombi and in seven patients with level 2 tumor thrombi. All procedures were completed robotically in their series as well.9 Finally, Bratslavsky and Cheng10 reported a tumor thrombus removal which extended 11 cm into the IVC. These reports all show that total robotic control of the IVC is possible with a surgical robotic platform even when extensive IVC tumor extension is present. All of these studies reported minimal complications and no perioperative deaths. These reports are limited to surgery for RCC with tumor thrombus extension into the IVC. They have shown that cavotomy and primary closure can be performed robotically. Our series shows that the da Vinci robotic platform can be used to safely operate on the IVC for other indications in addition to RCC. IVCF explantation is a prime example of this. In cases where vascular control can be safely obtained and maintained, this provides a fantastic alternative for IVCF removal when endovascular retrieval fails. Our study did not explicitly compare the outcomes of robotic and open surgical approaches to IVC thrombectomy and reconstruction. However, comparing our results with those of open surgical series demonstrates favorable outcomes. A landmark series by Blute et al2 included 540 patients over a study period of 30 years and reported a variable complication rate of 8.6% to 30%, depending on the thrombus level, with a median length of stay of 7 to 9 days. Recently, Armstrong et al11 published the outcomes of open radical nephrectomy and IVC thrombectomy with reconstruction. They reported an average length of stay of 6 days and an overall early complication rate of 29%. Of note, all of the patients required an intensive care unit admission in the immediate postoperative period.11
IVCFs are known to migrate and cause abdominal pain, and filter erosion into adjacent structures has been reported in up to 20% of IVCF placements.4 Most IVCFs can be removed using an endovascular approach. Etkin et al12 recently described “fall back” techniques for endovascular retrieval of IVCFs. In their series, 268 filters were removed through an endovascular approach, 79% by standard snaring, and 21% with other techniques. These techniques included upsizing the working sheath to 18F, using two sheaths, using a lasso approach, and using endobronchial forceps. The authors comment that the use of endobronchial forceps may injure the IVC wall and, therefore, are used only as a last resort.12 In our experience, failed endovascular retrieval of IVCF involved the use of endobronchial forceps. At least one patient seemed to have significant tissue ingrowth or filter migration into the caval wall, a finding also reported in series of open IVCF explantations.13 The endovascular procedure was aborted (this was the second attempt), and the patient was offered robotic removal. The findings during the robotic case showed that the IVCF had perforated the IVC wall and migrated 1 to 2 mm into the parenchyma of the liver. This was managed relatively easily during the robotic dissection by controlling the perforation with cautery; however, had this been removed by an endovascular approach, any bleeding encountered would be extremely difficult to control and perhaps would have required conversion to laparotomy. Although a great majority of IVCFs can be removed endovascularly, rarely this is not possible, and surgical explanation may be necessary. We are now able to offer patients a minimally invasive alternative to open IVCF removal. Open IVCF explantation has been described as a method for removing IVCFs that are not amenable to endovascular retrieval. Most IVCFs that are currently deployed are designed to be retrievable; however, the indications for use put a time limit on safe endovascular removal, some as short as 2 weeks.12 This is secondary to the endothelial ingrowth that occurs around the anchoring mechanism of the IVCF. Over time, this ingrowth becomes more robust and can lead to perforation of the IVC during endovascular removal. The anchoring mechanism can also contribute to IVC perforation and IVCF migration into adjacent structures. Jia et al4 reported that the IVC penetration rate of IVCF may be as high as 19%. They also found that the duodenum was the most frequently involved organ if penetration was present. These perforations are largely asymptomatic (47%) or may cause abdominal pain (8%). Filter retrieval may also be required in these cases, and an endovascular approach is not always feasible. There are case reports of endovascular retrieval in cases of known IVC perforation and IVCF migration; however; these seem to be the exception rather than the rule.
6
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders ---
These patients will typically require surgical IVCF explantation. Primary tumors of the IVC are rare entities, the most common being leiomyosarcoma.14 Retroperitoneal and visceral sarcomas represent 35% of all retroperitoneal malignancies.15 These retroperitoneal malignancies may involve the IVC through direct extension. Although uncommon, these patients represent another potential application of robotic surgery. As long as sound oncologic principles are strictly adhered to, excision of these malignancies are seemingly possible with the robotic surgical platform, even if the IVC is involved.
CONCLUSIONS IVC surgery is rare. Vascular surgeons can be called on to treat pathology of the IVC, including RCC with tumor thrombus extension and IVCFs not amenable to endovascular retrieval. Our initial experience with robotic ICV surgery supports its continued use and consideration as a minimally invasive surgical technique. We thank Frank Corl for his assistance in creating the medical illustrations for this report.
AUTHOR CONTRIBUTIONS Conception and design: VD, EC, SM Analysis and interpretation: VD, CV, HA-M, EC, SM Data collection: VD, WS, RF, EC Writing the article: VD, CV, HA-M, EC, SM Critical revision of the article: VD, CV, WS, RF, HA-M, EC, SM Final approval of the article: VD, CV, WS, RF, HA-M, EC, SM Statistical analysis: VD, CV, EC Obtained funding: Not applicable Overall responsibility: VD
REFERENCES 1. National Cancer Institute. Surveillance, Epidemiology, and End Results (SEER). Cancer of the kidney and renal pelvis SEER Stat Fact Sheets. Available at: http://seer.cancer.gov/ statfacts/html/kidrp.html. Accessed November 27, 2015. 2. Blute ML, Leibovich BC, Lohse CM, Cheville JC, Zincke H. The Mayo Clinic experience with surgical management, complications and outcome for patients with renal cell carcinoma and venous tumour thrombus. BJU Int 2004;94: 33-41.
2016
3. Angel LF, Tapson V, Galgon RE, Restrepo MI, Kaufman J. Systematic review of the use of retrievable inferior vena cava filters. J Vasc Interv Radiol 2011;22:1522-30.e3. 4. Jia Z, Wu A, Tam M, Spain J, McKinney JM, Wang W. Caval penetration by inferior vena cava filters: a systematic literature review of clinical significance and management. Circulation 2015;132:944-52. 5. Woo Y, Hyung WJ, Pak KH, Inaba K, Obama K, Choi SH, et al. Robotic gastrectomy as an oncologically sound alternative to laparoscopic resections for the treatment of early-stage gastric cancers. Arch Surg 2011;146:1086-92. 6. Moore WS, Kashyap VS, Vescera CL, Quiñones-Baldrich WJ. Abdominal aortic aneurysm: a 6-year comparison of endovascular versus transabdominal repair. Ann Surg 1999;230: 298-306; discussion: 306-8. 7. Martin GL, Castle EP, Martin AD, Desai PJ, Lallas CD, Ferrigni RG, et al. Outcomes of laparoscopic radical nephrectomy in the setting of vena caval and renal vein thrombus: seven-year experience. J Endourol 2008;22:1681-5. 8. Abaza R. Initial series of robotic radical nephrectomy with vena caval tumor thrombectomy. Eur Urol 2011;59:652-6. 9. Gill IS, Metcalfe C, Abreu A, Duddalwar V, Chopra S, Cunningham M, et al. Robotic level III inferior vena cava tumor thrombectomy: initial series. J Urol 2015;194:929-38. 10. Bratslavsky G, Cheng JS. Robotic Assisted radical nephrectomy with retrohepatic vena caval tumor thrombectomy (level III) combined with extended retroperitoneal lymph node dissection. Urology 2015;86:1235-40. 11. Armstrong PA, Back MR, Shames ML, Bailey CJ, Kim T, Lawindy SM, et al. Outcomes after inferior vena cava thrombectomy and reconstruction for advanced renal cell carcinoma with tumor thrombus. J Vasc Surg Venous Lymphat Disord 2014;2:368-76. 12. Etkin Y, Glaser JD, Nation DA, Foley PJ, Wang GJ, Woo EY, et al. Retrievable inferior vena cava filters can always be removed using “fall-back” techniques. J Vasc Surg Venous Lymphat Disord 2015;3:364-9. 13. Rana MA, Gloviczki P, Kalra M, Bjarnason H, Huang Y, Fleming MD. Open surgical removal of retained and dislodged inferior vena cava filters. J Vasc Surg Venous Lymphat Disord 2015;3:201-6. 14. Bower TC, Cherry KJ. Primary and secondary vena cava tumors. In: Stanley JC, Veith FJ, Wakefield TW, editors. Current therapy in vascular and endovascular surgery. 5th edition. Philadelphia: Saunders; 2014. p. 948-52. 15. Contreras CM, Heslin MJ. Soft tissue sarcoma. In: Townsend CM, Beauchamp RD, Evers BM, Mattox KL, editors. Sabiston textbook of surgery. 20th edition. Philadelphia: Elsevier; 2016. p. 754-72.
Submitted May 26, 2016; accepted Aug 12, 2016.
Robotic inferior vena caval surgery Victor J. Davila, MD,a Cristine S. Velazco, MD, MS,a William M. Stone, MD,a Richard J. Fowl, MD,a Haidar M. Abdul-Muhsin, MBBS,a Erik P. Castle, MD,b and Samuel R. Money, MD, MBA,a Phoenix, Ariz
ABSTRACT Objective: Inferior vena caval (IVC) surgery is uncommon and has traditionally been performed through open surgical approaches. Renal cell carcinoma with IVC extension generally requires vena cavotomy and reconstruction. Open removal of malpositioned IVC filters (IVCF) is occasionally required after endovascular retrieval attempts have failed. As our experience with robotic surgery has advanced, we have applied this technology to surgery of the IVC. We reviewed our institution’s experience with robotic surgical procedures involving the IVC to determine its safety and efficacy. Methods: All patients undergoing robotic surgery that included cavotomy and repair from 2011 to 2014 were retrospectively reviewed. Data were obtained detailing preoperative demographics, operative details, and postoperative morbidity and mortality. Results: Ten patients (6 men) underwent robotic vena caval procedures at our institution. Seven patients underwent robotic nephrectomy with removal of IVC tumor thrombus and retroperitoneal lymph node dissection. Three patients underwent robotic explantation of an IVCF after multiple endovascular attempts at removal had failed. The patients with renal cell carcinoma were a mean age of was 65.4 years (range, 55-74 years). Six patients had right-sided malignancy. All patients had T3b lesions at time of diagnosis. Mean tumor length extension into the IVC was 5 cm (range, 1-8 cm). All patients underwent robotic radical nephrectomy, with caval tumor thrombus removal and retroperitoneal lymph node dissection. The average operative time for patients undergoing surgery for renal cell carcinoma was 273 minutes (range, 137-382 minutes). Average intraoperative blood loss was 428 mL (range, 150-1200 mL). The patients with IVCF removal were a mean age of 33 years (range, 24-41 years). Average time from IVCF placement until robotic removal was 35.5 months (range, 4.3-57.3 months). Before robotic IVCF removal, a minimum of two endovascular retrievals were attempted. Average operative time for patients undergoing IVCF removal was 163 minutes (range, 131-202 minutes). Intraoperative blood loss averaged 250 mL (range, 150-350 mL). All procedures were completed robotically. The mean length of stay for all patients was 3.5 days (range, 1-8 days). All patients resumed ambulation on postoperative day 1. Nine patients resumed a regular diet on postoperative day 2. One patient with a renal tumor sustained a colon injury during initial adhesiolysis, before robotic radical nephrectomy, which was recognized at the initial operation and repaired robotically. Robotic radical nephrectomy and caval tumor removal were then completed. No blood transfusions were required intraoperatively, but three patients required blood transfusions postoperatively. Conclusions: Although robotic IVC surgery is uncommon, our initial limited experience demonstrates it is safe and efficacious. (J Vasc Surg: Venous and Lym Dis 2016;-:1-6.)
Surgery involving the inferior vena cava (IVC) is rare. There will be w60,000 newly diagnosed renal cell carcinomas (RCCs) in the United States in 2015.1 Concomitant tumor thrombus extension into the IVC is found in 4% to 10% of RRC cases.2 Level 0 tumors are limited to the renal vein, level 1 tumors extend into the vena cava #2 cm above the renal vein, level 2 tumors extend >2 cm
From the Division of Vascular Surgery, Department of Surgery,a and the Department of Urology,b Mayo Clinic. Author conflict of interest: none. Presented at the Forty-fourth Annual Symposium of the Society for Clinical Vascular Surgery, Las Vegas, Nev, March 12-16, 2016. Correspondence: Victor J. Davila, MD, 5777 E Mayo Blvd, Phoenix, AZ 85024 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the Journal policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 2213-333X Copyright Ó 2016 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvsv.2016.08.003
above the renal vein, level 3 tumors extend to the hepatic veins but below the diaphragm, and level 4 tumors extend above the diaphragm. An open surgical approach is traditionally used in performing nephrectomy, retroperitoneal lymph node dissection (RPLND), and cavotomy with tumor thrombus removal. Vascular surgeons are frequently called on to perform the vena caval reconstruction after tumor excision. Laparoscopic surgery for RCC has been successfully performed and is widely used in cases without tumor thrombus extension into the IVC. More recently, robotic surgery is allowing surgeons the ability to achieve vascular control in a safer and more effective way than laparoscopy. This has allowed for complete control of the IVC and allowed caval reconstruction to be performed out without the necessity of an open surgical component. Vascular surgeons are also called on to manage deep vein thrombosis, including placement and removal of IVC filters (IVCFs). Although most IVCFs that are placed 1
2
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders ---
today are retrievable by design, less than half are removed.3 Retrieval can be performed endovascularly; however, complications can occur after IVCF placement that require surgical explantation, including filter migration, IVC and adjacent organ perforation, and filter fracture.4 Often, open surgery is required for safe removal given the vascular control that must be obtained during the procedure. RCC with tumor thrombus extension into the IVC and IVCF explantation can both be technically challenging; however, both (and surgery on the IVC in general) have been performed safely with minimally invasive techniques. In this report we describe our experience with robotic IVC surgery.
METHODS After Institutional Review Board approval, databases were queried for patients who underwent robotic surgery involving the IVC from 2011 to 2014. This retrospective review included medical and surgical history, presentation, and demographic information. Operative details included operative time, blood loss, and need for transfusion. Postoperative details included time until regular diet, time to ambulation, and hospital length of stay. Postoperative morbidity and mortality were also noted. All procedures were performed with the da Vinci Si robotic platform (Intuitive Surgical, Inc, Sunnyvale, Calif). In cases of RCC, the primary approach for nephrectomy, RPLND, and tumor thrombus removal is robotic. An open approach is considered if preoperative imaging precludes safe vascular control. The main reason a patient would not be considered for a robotic approach is in cases of level 3 and level 4 tumors, because obtaining robotic-assisted vascular control above the hepatic veins can be very challenging. In patients requiring IVCF explantation, a minimum of two endovascular attempts at removal were required before consideration of robotic removal. Surgical technique for robotic nephrectomy and IVC tumor removal. The patient is placed in the modified right-side-up left-side-down position. Access is gained to the peritoneal cavity using the Veress needle technique in the midline. We place three 8-mm robotic ports in an approximately linear fashion in the midclavicular line while maintaining a distance between ports of 9 or 10 cm, or one hands’ breadth. A liver retractor port is placed in the subxiphoid region through a 5-mm port with two 12-mm assistant ports. The first port is placed just above the umbilicus, and the second is placed midway between the xiphisternum and the umbilicus. The robot is docked, and the location of the fourth robotic port is carefully planned in the lowermost location in the ipsilateral lower abdominal quadrant to avoid clashing with the rest of the robotic arms.
2016
First, electrocautery is used to mobilize the right colon, and the duodenum identified and Kocherized athermally. The vena cava is isolated with blunt dissection and pinpoint electrocautery. The ipsilateral renal artery is divided with an endovascular stapler. The lumbar veins are then identified and ligated. We usually tie the lumbars with 4-0 silk suture and then transect with the robotic scissors. Control of the infrarenal IVC is then obtained. The contralateral renal vein is dissected, and a vessel loop placed around it. The IVC is completely mobilized. The suprarenal IVC is then dissected. A vessel loop is doubly looped to maintain adequate control of the suprarenal vena cava. Next, control is obtained by pulling up on the vessel loops and using Satinsky bulldogs on the renal vein. Rumel tourniquets are tightened on both the suprarenal and infrarenal vena cava by cinching down the vessel loops. The robotic Pott scissors are used to make a cavotomy on the anterior aspect at the level of the renal veins that is extended superiorly until normal IVC is identified. The tumor thrombus is then removed en bloc with the ipsilateral kidney, and the specimen is extracted from the abdominal cavity. The cavotomy is closed with 3-0 or 4-0 Prolene (Ethicon, Somerville, NJ) suture on a small half needle in a running fashion. Immediately before complete closure, the clamps are released to flush air out of the vena cava. The suture line is then secured, and after hemostasis is endured, the clamps and then the vessel loops are removed. Hemostasis in the remaining surrounding areas is obtained with pinpoint electrocautery. A fibrin sealant (Evicel; Ethicon) or a hemostatic matrix (Floseal; Baxter, Deerfield, Ill), or both, can be applied as deemed necessary for further hemostasis. Retrieval of the IVCF is completed in a similar manner with control of the renal veins and vena cava (Fig). Control or ligation of the renal arteries is not performed. After the cavotomy is made, the filter can be seen and removed. In growth of endothelium around the filter may require dissection before removing the filter. If necessary, the tips of the tines of the filter may be broken off and removed individually to facilitate removal from the IVC. The remaining filter can then be removed in its entirety. If the filter is penetrating the wall, the venotomy can be made at this site, and the size can be adjusted to accommodate removal. The filter is then brought out of the abdomen through the 12-mm port. The steps are summarized in Table I.
RESULTS Ten patients (six men) underwent robotic operations involving the vena cava at our institution. All procedures included cavotomy with reconstruction. Reconstruction consisted of primary closure of the cavotomy with
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume
-,
Number
3
-
Fig. Left inset, Planned cavotomy before inferior vena cava filter (IVCF) removal with tines protruding through caval wall. Main illustration, Robotic removal of IVCF. Right inset, Venotomy closure.
running Prolene suture. All procedures were completed robotically, with no conversion to laparoscopy or open laparotomy. At time of surgery, patients were an average age of 55.6 years (range, 24.1-73.6 years; Table II). Average body mass index was 30.8 kg/m2 (range, 21.7-39.4 kg/m2). Preoperative
characteristics consisted of diabetes in three and hypertension in seven. Six patients had previous abdominal surgery, which included hysterectomy in two, and laparoscopic appendectomy, umbilical hernia repair, laparoscopic cholecystectomy, ventral hernia repair without mesh, laparoscopic left radical nephrectomy, and an
Table I. Technical tips and considerations for robotic inferior vena caval (IVC) surgery 1. Ensure proper robotic port placement because this plays a key role in successful completion of this operation. 2. An experienced laparoscopic surgeon should assist at the bedside. This is of extreme importance to temporize bleeding while the team prepares for open conversion, if needed. 3. The surgical team should be prepared for open conversion with all instruments available and ready for immediate use. 4. Suprarenal, infrarenal IVC, and contralateral renal vein should be controlled as early as possible. 5. The right adrenal vein may need to be divided. 6. Proper and secure control of all lumbar veins should be obtained before cavotomy because they often result in bothersome bleeding if missed. This is preferably performed using silk ties in the exact same manner as in open surgery. 7. Excision of part of the IVC wall may be needed if there is tumor thrombus invasion or endothelial growth over the IVC filter (IVCF). 8. We prefer circumferential control of the IVC with vessel loops. Although balloon occlusion can be obtained in the distal IVC, troublesome back bleeding can occur proximally and from unidentified lumbar veins. This is why we believe good dissection and early identification of lumbar veins before cavotomy is important.
4
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders ---
2016
Table II. Demographics and intraoperative details Variablea
Total cavotomy (N ¼ 10)
Thrombus (n ¼ 7)
IVCF retrieval (n ¼ 10)
Male sex
6
5
1
Age, years
55.6 (24.1-73.6)
65.4 (54-73)
Operative time, minutes
240 (163-273)
273 (137-382)
Blood loss, mL
375 (150-1200)
428 (150-1200)
32 (24-41) 163 (131-202) 250 (150-350)
IVCF, Inferior vena cava filter. a Categoric variables are shown as the number and continuous variables as the average (range).
open right partial nephrectomy in one patient each. Preoperative imaging consisted of magnetic resonance imaging in six patients and computed tomography scan in five patients. All patients with IVCF had computed tomography scan and previous attempts at venography as the sole imaging modalities. Operative time averaged 240 minutes (range, 163-273 minutes), and blood loss averaged 375 mL (range, 1501200 mL; Table II). Complications included one colotomy in a RCC patient who had extensive adhesiolysis at the beginning of the case. The colotomy was recognized promptly and repaired robotically. A retroperitoneal hematoma developed in one patient on postoperative day 2, which was managed nonoperatively. All patients resumed ambulation on postoperative day 1. Average time to a regular diet was 1.9 days (range, 1-8 days). No patients required intraoperative transfusion, but three patients required blood transfusions postoperatively: two patients for preoperative anemia and one patient who developed a postoperative retroperitoneal hematoma. The average transfusion requirement per patient was 0.9 units (range, 1-5 units). The average length of stay was 3.5 days (range, 1-8 days). RCC experience. The patients with RCC were an average age of 65.4 years (range, 54-73 years). There were five male patients. Six of the tumors were located on the right side. All tumors were T3 (stage 3). Tumor thrombus extension into the IVC averaged 5.3 cm (range, 1-8 cm). Preoperative imaging showed the average primary tumor measurement in greatest dimension was 7.7 cm (range, 1.9-11 cm). Operative time averaged 273 minutes (range, 137-382 minutes). Blood loss averaged 428 mL (range, 150-1200 mL; Table II). Average length of stay was 3.4 days (range, 2-7 days). IVCF explantation experience. Patients undergoing IVCF explantation were an average age of 32 years (range, 24-41 years). There were two female patients and one male patient. All patients presented with abdominal pain. All patients had retrievable IVCFs, and each patient underwent a minimum of two endovascular attempts of IVCF removal before being considered for robotic explantation. IVCFs were placed an average of 1065 days before explantation (range, 129-1719 days). Operative time averaged 163 minutes (range, 131-202 minutes). Average
blood loss was 250 mL (range, 150-350 mL; Table II). Average length of stay was 3.6 days (range, 1-8 days). Primary findings at the time of explanations leading to failure of endovascular retrieval included robust endothelial ingrowth and perforation/erosion into adjacent organs (liver, duodenum/small bowel, mesentery).
DISCUSSION Minimally invasive surgery has replaced open surgery as the preferred approach in many surgical specialties, including general surgery, gynecology, and thoracic surgery. The laparoscopic approach allows for equivalent long-term outcomes compared with open procedures, with decreased rates of incisional pain, faster return to work, earlier hospital discharge, and earlier resumption of diet.5 Robotic surgery offers the benefits of the minimally invasive approach while allowing the surgeon improved visualization, more freedom to manipulate instruments and tissues, and ergonomic support during long operations. Robotic surgery is rarely used in vascular surgery, primarily because performing vascular anastomoses via laparoscopic methods are technically challenging. Therefore, vascular surgery has not embraced minimally invasive techniques for treating vascular disease apart from endovascular surgery. Endovascular technology has shown excellent results in long-term outcomes, wound infection rates, pain scores, and hospital length of stay, which are many of the same metrics by which laparoscopic and robotic surgery are measured.6 There are cases where the endovascular approach fails, and traditional surgery is required. The da Vinci robotic platform is being used to allow the surgeon to perform complex procedures with similar or improved outcomes compared with laparoscopy and open procedures. Robotic procedures are currently allowing surgeons the ability to precisely perform technically demanding procedures through a minimally invasive approach. Our institution has experience with excision of RCC and associated tumor thrombus by a laparoscopic approach.7 Vena caval surgery using the robotic platform has also been described in RCC.8-10 Our experience shows that these techniques can be applied to surgery on the vena cava with minimal morbidity and mortality. All procedures in our series were completed robotically.
Journal of Vascular Surgery: Venous and Lymphatic Disorders Volume
-,
Number
Davila et al
5
-
Patients recovered well postoperatively, resuming ambulation and diet on postoperative days 1 and 2, respectively. Complications were inherent to the surgery and not specific to the robotic approach. One patient sustained a colotomy during extensive lysis of adhesions from a previous laparotomy, which was identified and treated at the initial operation without subsequent sequelae. A retroperitoneal hematoma developed in another patient who resumed a heparin drip on postoperative day 1 for a known hypercoagulable state. This was managed nonoperatively with blood transfusion, and the patient was discharged on postoperative day 8. There are multiple case reports and a few small series of IVC surgery currently in the literature. Abaza et al8 were the first to describe robotic IVC vascular control with tumor thrombus removal in 2010. They showed that tumor thrombi from 1 cm to 5 cm can be safely extracted from the IVC during nephrectomy, without the need for hand assistance or conversion to an open approach. Gill et al9 described a series of robotic nephrectomy and IVC tumor thrombus removals in nine patients with level 3 tumor thrombi and in seven patients with level 2 tumor thrombi. All procedures were completed robotically in their series as well.9 Finally, Bratslavsky and Cheng10 reported a tumor thrombus removal which extended 11 cm into the IVC. These reports all show that total robotic control of the IVC is possible with a surgical robotic platform even when extensive IVC tumor extension is present. All of these studies reported minimal complications and no perioperative deaths. These reports are limited to surgery for RCC with tumor thrombus extension into the IVC. They have shown that cavotomy and primary closure can be performed robotically. Our series shows that the da Vinci robotic platform can be used to safely operate on the IVC for other indications in addition to RCC. IVCF explantation is a prime example of this. In cases where vascular control can be safely obtained and maintained, this provides a fantastic alternative for IVCF removal when endovascular retrieval fails. Our study did not explicitly compare the outcomes of robotic and open surgical approaches to IVC thrombectomy and reconstruction. However, comparing our results with those of open surgical series demonstrates favorable outcomes. A landmark series by Blute et al2 included 540 patients over a study period of 30 years and reported a variable complication rate of 8.6% to 30%, depending on the thrombus level, with a median length of stay of 7 to 9 days. Recently, Armstrong et al11 published the outcomes of open radical nephrectomy and IVC thrombectomy with reconstruction. They reported an average length of stay of 6 days and an overall early complication rate of 29%. Of note, all of the patients required an intensive care unit admission in the immediate postoperative period.11
IVCFs are known to migrate and cause abdominal pain, and filter erosion into adjacent structures has been reported in up to 20% of IVCF placements.4 Most IVCFs can be removed using an endovascular approach. Etkin et al12 recently described “fall back” techniques for endovascular retrieval of IVCFs. In their series, 268 filters were removed through an endovascular approach, 79% by standard snaring, and 21% with other techniques. These techniques included upsizing the working sheath to 18F, using two sheaths, using a lasso approach, and using endobronchial forceps. The authors comment that the use of endobronchial forceps may injure the IVC wall and, therefore, are used only as a last resort.12 In our experience, failed endovascular retrieval of IVCF involved the use of endobronchial forceps. At least one patient seemed to have significant tissue ingrowth or filter migration into the caval wall, a finding also reported in series of open IVCF explantations.13 The endovascular procedure was aborted (this was the second attempt), and the patient was offered robotic removal. The findings during the robotic case showed that the IVCF had perforated the IVC wall and migrated 1 to 2 mm into the parenchyma of the liver. This was managed relatively easily during the robotic dissection by controlling the perforation with cautery; however, had this been removed by an endovascular approach, any bleeding encountered would be extremely difficult to control and perhaps would have required conversion to laparotomy. Although a great majority of IVCFs can be removed endovascularly, rarely this is not possible, and surgical explanation may be necessary. We are now able to offer patients a minimally invasive alternative to open IVCF removal. Open IVCF explantation has been described as a method for removing IVCFs that are not amenable to endovascular retrieval. Most IVCFs that are currently deployed are designed to be retrievable; however, the indications for use put a time limit on safe endovascular removal, some as short as 2 weeks.12 This is secondary to the endothelial ingrowth that occurs around the anchoring mechanism of the IVCF. Over time, this ingrowth becomes more robust and can lead to perforation of the IVC during endovascular removal. The anchoring mechanism can also contribute to IVC perforation and IVCF migration into adjacent structures. Jia et al4 reported that the IVC penetration rate of IVCF may be as high as 19%. They also found that the duodenum was the most frequently involved organ if penetration was present. These perforations are largely asymptomatic (47%) or may cause abdominal pain (8%). Filter retrieval may also be required in these cases, and an endovascular approach is not always feasible. There are case reports of endovascular retrieval in cases of known IVC perforation and IVCF migration; however; these seem to be the exception rather than the rule.
6
Davila et al
Journal of Vascular Surgery: Venous and Lymphatic Disorders ---
These patients will typically require surgical IVCF explantation. Primary tumors of the IVC are rare entities, the most common being leiomyosarcoma.14 Retroperitoneal and visceral sarcomas represent 35% of all retroperitoneal malignancies.15 These retroperitoneal malignancies may involve the IVC through direct extension. Although uncommon, these patients represent another potential application of robotic surgery. As long as sound oncologic principles are strictly adhered to, excision of these malignancies are seemingly possible with the robotic surgical platform, even if the IVC is involved.
CONCLUSIONS IVC surgery is rare. Vascular surgeons can be called on to treat pathology of the IVC, including RCC with tumor thrombus extension and IVCFs not amenable to endovascular retrieval. Our initial experience with robotic ICV surgery supports its continued use and consideration as a minimally invasive surgical technique. We thank Frank Corl for his assistance in creating the medical illustrations for this report.
AUTHOR CONTRIBUTIONS Conception and design: VD, EC, SM Analysis and interpretation: VD, CV, HA-M, EC, SM Data collection: VD, WS, RF, EC Writing the article: VD, CV, HA-M, EC, SM Critical revision of the article: VD, CV, WS, RF, HA-M, EC, SM Final approval of the article: VD, CV, WS, RF, HA-M, EC, SM Statistical analysis: VD, CV, EC Obtained funding: Not applicable Overall responsibility: VD
REFERENCES 1. National Cancer Institute. Surveillance, Epidemiology, and End Results (SEER). Cancer of the kidney and renal pelvis SEER Stat Fact Sheets. Available at: http://seer.cancer.gov/ statfacts/html/kidrp.html. Accessed November 27, 2015. 2. Blute ML, Leibovich BC, Lohse CM, Cheville JC, Zincke H. The Mayo Clinic experience with surgical management, complications and outcome for patients with renal cell carcinoma and venous tumour thrombus. BJU Int 2004;94: 33-41.
2016
3. Angel LF, Tapson V, Galgon RE, Restrepo MI, Kaufman J. Systematic review of the use of retrievable inferior vena cava filters. J Vasc Interv Radiol 2011;22:1522-30.e3. 4. Jia Z, Wu A, Tam M, Spain J, McKinney JM, Wang W. Caval penetration by inferior vena cava filters: a systematic literature review of clinical significance and management. Circulation 2015;132:944-52. 5. Woo Y, Hyung WJ, Pak KH, Inaba K, Obama K, Choi SH, et al. Robotic gastrectomy as an oncologically sound alternative to laparoscopic resections for the treatment of early-stage gastric cancers. Arch Surg 2011;146:1086-92. 6. Moore WS, Kashyap VS, Vescera CL, Quiñones-Baldrich WJ. Abdominal aortic aneurysm: a 6-year comparison of endovascular versus transabdominal repair. Ann Surg 1999;230: 298-306; discussion: 306-8. 7. Martin GL, Castle EP, Martin AD, Desai PJ, Lallas CD, Ferrigni RG, et al. Outcomes of laparoscopic radical nephrectomy in the setting of vena caval and renal vein thrombus: seven-year experience. J Endourol 2008;22:1681-5. 8. Abaza R. Initial series of robotic radical nephrectomy with vena caval tumor thrombectomy. Eur Urol 2011;59:652-6. 9. Gill IS, Metcalfe C, Abreu A, Duddalwar V, Chopra S, Cunningham M, et al. Robotic level III inferior vena cava tumor thrombectomy: initial series. J Urol 2015;194:929-38. 10. Bratslavsky G, Cheng JS. Robotic Assisted radical nephrectomy with retrohepatic vena caval tumor thrombectomy (level III) combined with extended retroperitoneal lymph node dissection. Urology 2015;86:1235-40. 11. Armstrong PA, Back MR, Shames ML, Bailey CJ, Kim T, Lawindy SM, et al. Outcomes after inferior vena cava thrombectomy and reconstruction for advanced renal cell carcinoma with tumor thrombus. J Vasc Surg Venous Lymphat Disord 2014;2:368-76. 12. Etkin Y, Glaser JD, Nation DA, Foley PJ, Wang GJ, Woo EY, et al. Retrievable inferior vena cava filters can always be removed using “fall-back” techniques. J Vasc Surg Venous Lymphat Disord 2015;3:364-9. 13. Rana MA, Gloviczki P, Kalra M, Bjarnason H, Huang Y, Fleming MD. Open surgical removal of retained and dislodged inferior vena cava filters. J Vasc Surg Venous Lymphat Disord 2015;3:201-6. 14. Bower TC, Cherry KJ. Primary and secondary vena cava tumors. In: Stanley JC, Veith FJ, Wakefield TW, editors. Current therapy in vascular and endovascular surgery. 5th edition. Philadelphia: Saunders; 2014. p. 948-52. 15. Contreras CM, Heslin MJ. Soft tissue sarcoma. In: Townsend CM, Beauchamp RD, Evers BM, Mattox KL, editors. Sabiston textbook of surgery. 20th edition. Philadelphia: Elsevier; 2016. p. 754-72.
Submitted May 26, 2016; accepted Aug 12, 2016.
