RENAL ARTERY PSEUDOANEURYSM FOLLOWING LAPAROSCOPIC PARTIAL NEPHRECTOMY

0022-5347/05/1746-2256/0 THE JOURNAL OF UROLOGY® Copyright © 2005 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 174, 2256 –2259, December 2005 Printed in U.S.A.

DOI: 10.1097/01.ju.0000181827.49239.8e

Trauma/Reconstruction/Diversion RENAL ARTERY PSEUDOANEURYSM FOLLOWING LAPAROSCOPIC PARTIAL NEPHRECTOMY DINESH SINGH

AND

INDERBIR S. GILL*, †

From the Section of Laparoscopic and Robotic Surgery, Glickman Urological Institute, Cleveland Clinic Foundation, Cleveland, Ohio

ABSTRACT

Purpose: We describe the presentation, evaluation and management of hemorrhage due to renal artery pseudoaneurysm following laparoscopic partial nephrectomy. Materials and Methods: Of the 345 laparoscopic partial nephrectomies performed by us during a 5-year period 6 patients (1.7%) had postoperative hemorrhage from a renal artery pseudoaneurysm. Patient charts were reviewed to identify pertinent preoperative, intraoperative and postoperative data. Results: Median tumor size was 3.5 cm (range 2.2 to 5), intraoperative blood loss was 175 cc (range 50 to 500), warm ischemia time was 32 minutes (range 30 to 45) and operative time was 3.8 hours (range 2.5 to 5). The mean percent of kidney excised was 31% and pelvicaliceal system entry was suture repaired in all 6 patients. No patient required blood transfusion perioperatively. Average hospital stay was 3.4 days (range 2.5 to 6). Delayed postoperative hemorrhage occurred at a median of 12 days (range 8 to 15). Angiography revealed a renal artery pseudoaneurysm most commonly at a third or fourth order branch (4 and 2 patients, respectively). Percutaneous embolization was successful in each patient. Conclusions: Renal artery pseudoaneurysm is an uncommon complication following laparoscopic partial nephrectomy. These patients often present in delayed fashion. Selective angiographic embolization is the initial treatment of choice. KEY WORDS: kidney; laparoscopy; nephrectomy; postoperative complications; aneurysm, false

Partial nephrectomy has become the standard of care as a nephron sparing option in the properly selected patient with a T1 tumor. Long-term 10-year followup has demonstrated oncological outcomes similar to those of radical nephrectomy in comparably staged patients.1 As experience with laparoscopy has grown, this minimally invasive technique has been applied to partial nephrectomy. As laparoscopic partial nephrectomy (LPN) becomes more widely accepted and used, it is important for clinicians to be familiar with its potential complications.2, 3 To our knowledge, renal artery pseudoaneurysm has not previously been reported following laparoscopic partial nephrectomy. We describe the presentation, evaluation and treatment of delayed hemorrhage from renal artery pseudoaneurysm (RAP) following LPN. METHODS AND MATERIALS

Between September 1999 and July 2004, 345 laparoscopic partial nephrectomies were performed at our institution by a single surgeon using our previously described technique.4 Data were retrospectively reviewed to identify 6 patients who presented with hemorrhage and required angiography. Hospital charts were reviewed to identify characteristics, such as patient age, sex, tumor side, size, location, depth, collecting

system repair, method of renal parenchyma repair, use of any adjunct hemostatic agents, warm ischemia time, operative time, and postoperative and hospital discharge hematocrit. Additionally, radiographic and angiographic findings were reviewed to document the size of the pseudoaneurysm, location in the kidney and level of intrarenal vascular involvement. RESULTS

Six of 345 patients (1.7%) undergoing LPN had postoperative bleeding requiring angiographic evaluation. In all 6 patients angiography demonstrated a renal artery pseudoaneurysm. There were 3 females and 3 males with a median age of 59 years (range 34 to 80), a median body mass index of 27 (range 23 to 32) and a median American Society of Anesthesiologists score of 3 (range 1 to 3). All patients underwent transperitoneal laparoscopic right partial nephrectomy. Four of these 6 patients underwent surgery in 2002 or previously. Median tumor size on computerized tomography (CT) was 3.5 cm (range 2.2 to 5). Three of these 6 patients had a central tumor and 3 had a peripheral tumor. The median depth of intraparenchymal extension was 1.8 cm. Median warm ischemia time was 32 minutes (range 30 to 45), blood loss was 175 ml (range 50 to 500) and operative time was 225 minutes (range 150 to 300). An average of 31% of the kidney (range 15% to 50%) was excised during partial nephrectomy (see table). All 6 patients underwent suture repair of pelvicaliceal system entry with 2-zero polyglactin on a CT-1 needle, followed by hemostatic approximation of the parenchymal de-

Submitted for publication September 15, 2004. * Correspondence and requests for reprints: Section of Laparoscopic and Robotic Surgery, Glickman Urological Institute, A100, Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, Ohio 44195 (telephone: 216-445-1534; FAX: 216-445-7031; e-mail: gilli@ ccf.org). † Financial interest and/or other relationship with Baxter and Pfizer. 2256

3rd 4th

Mean 6

fect with 0-polyglactin on a CTX needle over an oxidized cellulose bolster. Argon beam coagulation was not done in any patient. Floseal ™ was used as a biological hemostatic adjunct in the 2 latter patients as a routine part of the procedure. At the conclusion of the operation hemostasis was confirmed to be excellent in each patient. No patient required blood transfusion perioperatively. Median hospital stay was 3.4 days (range 2.5 to 6). Serial postoperative hemoglobin and hematocrit were stable in all patients prior to discharge home. Pathological evaluation revealed renal cell carcinoma in 5 patients and angiomyolipoma in 1 and all had negative surgical margins. All patients showed 1 or more of certain delayed symptoms at a median of 12 days (range 8 to 15), including flank pain, gross hematuria, dizziness/syncope, fever and/or bloody Jackson-Pratt (JP) drainage. Each patient received blood transfusion of a mean of 6 U (range 3 to 9) at the emergency room elsewhere and/or at our hospital (table 1). Each patient underwent percutaneous selective angiography, which identified an intrarenal arterial pseudoaneurysm as the cause of hemorrhage. The intrarenal artery forming the pseudoaneurysm was most commonly a third (4 patients) or fourth (2)

Not available Not accessible from angiogram 3rd

Overall Pt 272

4.6 350 39 30 10 Dizziness on standing, flank pain 5 Resection apex

Pt 270

4.0 500 45 30 12 Gross hematuria, fever 6 Resection apex 3.0 50 30 15 14 Flank pain

Pt 113 Pt 99

Clinical presentation and angiographic findings

2257

5.0 2.2 200 150 31 33 Not available 50 15 8 Abdominal pain, weakness, Bloody drainage in ⫹ around breath shortness JP drain, fever 9 3 Not accessible from angioResection apex gram 3rd 4th embolization outcome was successful in all. 2.6 120 31 Not available 11 Gross hematuria Pre-angiography blood transfusion (U) 7 Angiographic site Resection apex RAP branch order 3rd Pelvicaliceal suture repair was done in all 6 patients and CT tumor size (cm) Estimated blood loss (ml) Warm ischemia time (mins) % Kidney excised Hemorrhage presentation (postop day) Presentation symptoms

Pt 78 Pt 72

Median 3.5 Median 175 Median 32 Mean 31 Median 12

RENAL ARTERY PSEUDOANEURYSM AFTER LAPAROSCOPIC PARTIAL NEPHRECTOMY

FIG. 1. A, selective renal arteriogram demonstrates pseudoaneurysm (arrow) of third order renal artery branch. B, angiogram after selective embolization shows RAP ablation with coils (arrow).

2258

RENAL ARTERY PSEUDOANEURYSM AFTER LAPAROSCOPIC PARTIAL NEPHRECTOMY

FIG. 2. A, aortogram reveals fourth order RAP (arrow) at apex of resection tumor bed. B, angiogram after selective embolization shows RAP ablation using 1 oxidized cellulose pledget.

order branch (figs. 1 and 2). Selective angiographic embolization was successful in each patient. At a median followup of 25.5 months (range 9 to 33) no patient had any subsequent hemorrhagic episodes. DISCUSSION

Sparse information is available in the literature addressing the complications associated with LPN. However, this is an important issue because the worldwide experience with laparoscopy continues to grow and consequently more urologists gain the skills and experience to perform LPN. We

present the complication of delayed hemorrhage from renal artery pseudoaneurysm and the associated angiographic findings. Although it is an uncommon event, occurring in 6 of 345 patients (1.7%), it is potentially life threatening if not managed appropriately. Renal artery pseudoaneurysm has been previously reported in the open partial nephrectomy literature. In a series of 698 patients undergoing open partial nephrectomy 3 (0.43%) had RAP.5 These 3 patients also presented in delayed fashion at 3 weeks, at 3 months and on postoperative day 5, respectively. Two of these cases were successfully diagnosed and embolized by selective angiography, while in 1 RAP resolved spontaneously. Arterial pseudoaneurysm arises from a transected or punctured artery that leaks into a contained hematoma cavity. We believe that this complication likely happens from 1 of 2 mechanisms. An artery transected partially or end-on during resection could subsequently bleed into a contained space. This is supported by the fact that in 4 patients angiography identified that the RAP was located in the partial nephrectomy bed, particularly near the wedge resection apex. During deep parenchymal resection for the infiltrating, endophytic component of the tumor, larger diameter intrarenal arterial branches are likely to be observed. In 4 of 6 patients the bleeding vessel was a third order branch of the renal artery, which is a deep, substantial artery branch. This transected arterial branch may be partially ligated or in spasm, leading to its nonrecognition in the operating room. As the patient increases activity, the occluding clot may possibly become dislodged, accounting for its delayed presentation. Another possible cause, although less likely, is that a suture needle used to approximate the partial nephrectomy parenchymal defect may initially be inserted suboptimally into the renal parenchyma and then removed and redirected. In so doing, the initial needle pass could potentially have punctured an intrarenal arterial branch. During the ensuing few weeks leakage from the needle puncture hole in the intrarenal artery could enlarge, slowly creating a pulsatile pseudoaneurysm. There were several confounding factors in 2 of our patients that may have contributed to the risk of postoperative hemorrhage. One patient was re-started on clopidogrel bisulfate on postoperative day 4 because of a high risk of cerebrovascular accident. Another patient had hepatitis C, cirrhosis, thrombocytopenia and hypersplenism. Platelet count was 79,000/␮l 2 days prior to surgery and 103,000/␮l on the day of surgery after multiple preoperative platelet transfusions. In this man preoperative azotemia varied from a blood urea nitrogen of 26 to 53 mg/dl (normal 10 to 25) and a serum creatinine of 1.8 to 1.9 mg/dl (normal 0.7 to 1.4). Certain technical caveats are emphasized. A meticulous running suture is placed tightly to oversew all transected blood vessels in the partial nephrectomy bed. An individual figure-of-8 suture is used when necessary to specifically suture ligate a larger transected intrarenal artery. During hemostatic renorrhaphy careful pre-planning of the angle and direction of needle passage through the renal parenchyma is necessary to minimize false punctures and the need for needle redirection. Furthermore, when the renorrhaphy closing suture is tied down, tight closure of the parenchymal defect must be achieved around the bolster. Finally, it is important for the surgeon to laparoscopically re-inspect the surgical site after desufflating the abdomen for 5 to 10 minutes. This can unmask any potential bleeding sites hidden by the pneumoperitoneum effect. We do not administer ketorolac for pain management in patients who undergo LPN due to concerns about platelet aggregation inhibition. The perirenal drain is removed when drainage has been decreased to less than 50 cc daily for 2 to 3 days consecutively, which is usually by day 4 or 5. The higher incidence of RAP in our laparoscopic series

RENAL ARTERY PSEUDOANEURYSM AFTER LAPAROSCOPIC PARTIAL NEPHRECTOMY

(1.7%) compared to the 0.43% incidence in the open series is likely due to multiple factors, including the larger needle used for parenchymal suturing, the somewhat lesser tightness of parenchymal approximation in the laparoscopic technique in the occasional patient and possible under reporting in retrospective studies in the open literature. It is likely that the incidence of clinically insignificant RAP following open or laparoscopic partial nephrectomy is actually even higher, given that only patients presenting with a hemorrhagic complication are likely to undergo angiographic evaluation postoperatively. The management of postoperative RAP requires a high index of suspicion. Patients may present with signs of delayed hemorrhage 1 to 2 weeks after LPN (median 12 days in our study) with symptoms such as gross hematuria, bloody JP drainage, dizziness, syncope, flank pain or fever. Antiplatelet or anti-coagulation medicines must be discontinued. In addition to routine serum tests, gastrointestinal bleeding should be ruled out by guaiac stool testing. Contrast enhanced CT may be done to confirm and assess the size of perinephric bleeding. If the patient remains hemodynamically stable on serial hematocrit assessments during a 48hour observation period, hospital discharge with advice for restricted activity for 2 weeks and close followup is adequate. However, an ongoing need for transfusions or hemodynamic instability is an indication for angiography. Identification of a specific bleeding site, such as a renal artery pseudoaneurysm or an arteriovenous malformation, requires selective embolization. This intervention has minimal morbidity, spares nephrons, avoids reoperative surgical intervention, which risks the possibility of requiring nephrectomy, and usually provides definitive therapy.

2259

CONCLUSIONS

Renal artery pseudoaneurysm is an uncommon but potentially life threatening complication. A high index of suspicion and an understanding of its typical clinical presentation will enable rapid diagnosis. Selective angiographic embolization is the initial treatment of choice. ADDENDUM

Since this manuscript was submitted we have not seen an occurrence of a RAP in 140 additional LPNs. We believe the reason is close attention to the technical caveats described in our article, along with the adjunctive use of the biological hemostatic agent Floseal™. As such, our overall incidence of clinically significant RAP following LPN is 1.2% (6 of 485 cases). REFERENCES

1. Herr, H. W.: Partial nephrectomy for unilateral renal carcinoma and a normal contralateral kidney: 10-year followup. J Urol, 161: 33, 1999 2. Ramani, A. P., Desai, M. M., Steinberg, A. P., Ng, C. S., Abreu, S. C., Kaouk, J. H. et al: Complications of laparoscopic partial nephrectomy in 200 cases. J Urol, 173: 42, 2005 3. Kim, F. J., Rha, K. H., Hernandez, F., Jarrett, T. W., Pinto, P. A. and Kavoussi, L. R.: Laparoscopic radical versus partial nephrectomy: assessment of complications. J Urol, 170: 408, 2003 4. Gill, I. S., Desai, M. M., Kaouk, J. H., Meraney, A. M., Murphy, D. P., Sung, G. T. et al: Laparoscopic partial nephrectomy for renal tumor: duplicating open surgical techniques. J Urol, 167: 469, 2002 5. Albani, J. M. and Novick, A. C.: Renal artery pseudoaneurysm after partial nephrectomy: three case reports and a literature review. Urology, 62: 227, 2003