Renal injury occurs in up to 1.2% of trauma cases in the

Transcription

1 Renal Arterial Injuries: A Single Center Analysis of Management Strategies and Outcomes Sean P. Elliott, Ephrem O. Olweny and Jack W. McAninch* From the Department of Urologic Surgery, University of Minnesota, Minneapolis, Minnesota, and Department of Urology, University of California-San Francisco and San Francisco General Hospital (EOO, JWM), San Francisco, California Purpose: Management of main and segmental renal artery injury following external trauma is controversial. After main renal artery injury the controversy surrounds nephrectomy vs revascularization, whereas after segmental renal artery injury the debate involves operative vs nonoperative management. We reviewed our experience with renal artery injury management at a single trauma center with the goal of identifying optimal clinical management strategies. Materials and Methods: Data on a total of 82 renal artery injuries in 81 patients collected between 1978 and 2006 were retrospectively reviewed. Patient demographics, length of stay, transfusion requirements and injury characteristics (artery subtype, grade, mechanism, and associated parenchymal, venous and nonrenal injuries) were recorded. Management strategies and outcomes for each renal artery injury subtype were compared. Results: Median patient age was 28 years (range 4 to 74) and 90% of the patients were male. Main renal artery injury occurred in 36 of 81 patients (43%) and segmental renal artery injury occurred in 45 (57%). Injury characteristics were similar for each renal artery injury subtype. For main renal artery injury the respective outcomes of nephrectomy vs vascular repair were a mean transfusion of 10,275 vs 6,125 ml (p 0.39), length of stay 18 days for each, mortality rate 26% vs 13%, renal failure rate 8% vs 25% and renal insufficiency/impaired selective function by renal scintigraphy 4% vs 13% (each p not significant). For segmental renal artery injury operative vs nonoperative management was associated with a mean transfusion of 4,994 vs 820 ml (p 0.01), length of stay 29 vs 11 days (p 0.23) and mortality rate 8% vs 6% (p 1.0). Renal failure and impaired selective renal function on scintigraphy were similar between the groups. Conclusions: Nephrectomy for main renal artery injury has outcomes similar to those of vascular repair and it does not worsen posttreatment renal function in the short term. Nonoperative management for segmental renal artery injury results in excellent outcomes. Key Words: kidney, renal artery, wounds and injuries, nephrectomy Renal injury occurs in up to 1.2% of trauma cases in the United States. Of these cases injuries involving the renal vasculature are infrequent, occurring in only 2.5% to 4% of renal injuries. 1 3 Renovascular injuries are associated with multiple nonrenal organ injuries and with a high mortality rate of 19% to 44%. 4 Currently there is no consensus in the published literature on appropriate management strategies for these injuries. Nonoperative management of main renal vein laceration or thrombosis as well as main renal artery thrombosis is appropriate in select cases, although observation of main renal vein avulsion, main renal artery avulsion or main renal artery laceration is usually not indicated due to exsanguination from the renal unit. 5 While there is a role for observation in the stable patient with MRAI, especially after blunt trauma, our experience has been that these patients represent a small minority of all MRAIs, and so observation is not the focus of this report. In the setting of MRAI in an unstable patient with a normally functioning contralateral kidney older as well as contemporary series advocate immediate nephrectomy, given the generally poor outcomes of arterial reconstruction. 6,7 Submitted for publication March 19, * Correspondence: Department of Urology, San Francisco General Hospital, 1001 Potrero Ave., 3A20, San Francisco, California However, specific indications for nephrectomy vs arterial reconstruction are not clearly established. For SRAI nonoperative management is generally favored in otherwise hemodynamically stable patients with good overall outcomes. 5,8 However, a few series describe high complication rates for nonoperative management of SRAI, including urinoma Additionally, some investigators have recommended renal exploration for SRAI with repair or ligation in patients who require laparotomy for other reasons. 1 We retrospectively reviewed our institutional experience with managing renal arterial injuries with the goal of identifying optimal management strategies in specific clinical situations. PATIENTS AND METHODS The Genitourinary Trauma Database at San Francisco General Hospital was searched for patients with renovascular trauma between 1978 and Data collected included patient demographics, injury grade and mechanism, type of vessel injured (main vs segmental artery with or without associated venous injury), shock, hematuria, associated renal laceration and associated injuries. These variables as well as management strategies and outcomes were analyzed for patients with MRAI and SRAI. Outcomes assessed in /07/ / Vol. 178, , December 2007 THE JOURNAL OF UROLOGY Printed in U.S.A. Copyright 2007 by AMERICAN UROLOGICAL ASSOCIATION DOI: /j.juro

2 2452 RENAL ARTERIAL INJURIES cluded survival, feasibility of renal sparing therapy, renal function and complications. They were compared in patients with nephrectomy vs renal sparing treatment for MRAI and in operative vs nonoperative treatment for SRAI. Categorical data were compared using chi-square analysis with the Yates correction. When individual frequencies were small (less than 5), Fisher s exact test was used. Statistical significance was considered at p Logistic regression was performed to analyze for independent predictors of nephrectomy and survival. RESULTS Overall Presentation The 81 patients sustained a total of 82 renal artery injuries. Median age was 28 years (range 4 to 74) with 12 patients (15%) 18 years or younger. Of the patients 90% were male. Etiology was penetrating trauma in 41 of 81 patients (51%) and blunt in 40 (49%). Of the patients 46 (57%) presented with shock (systolic blood pressure less than 90 mm Hg). Gross or microscopic hematuria was present in 60 of 67 evaluable patients (90%) and associated injuries were present in 72 (89%). Associated venous injury was present in 27 patients (33%), while parenchymal laceration was present in 47 (58%). A total of 47 patients (58%) had grade 4 injury, while 34 (42%) had grade 5 injury. Two patients had bilateral injuries, including contralateral grade 3 renal laceration and contralateral SRAI in 1 each (table 1). TABLE 1. Patient and injury characteristics MAI SRAI Totals No. pts (%) 36 (43) 45 (57) 81 Median age (range) 29 (15 74) 27 (4 68) 28 (4 74) No. males (%) 30 (41) 43 (59) 73 (90) No. females (%) 6 (14) 2 (5) 8 (10) No. trauma (%): Penetrating 20 (49) 21 (51) 41 (51) Blunt 16 (40) 24 (60) 40 (49) No. parenchymal 21 (45) 26 (55) 47 (58) laceration (%) No. associated injuries (%) 34 (47) 38 (53) 72 (89) No. shock (%) 26 (57) 20 (43) 46 (57) No. hematuria/total 30/33 (91) 30/34 (88) 60/67 (90) No. (%): Gross 18/30 (60) 15/30 (50) 33/60 (55) Microscopic 12/30 (40) 15/30 (50) 27/60 (45) Absent 3/33 (9) 4/34 (12) 7/67 (10) Unknown 3/36 (8) 11/45 (24) 14/81 (17) No. RAIs (%): Isolated arterial injury 16 (29) 39 (71) 55 (67) Combined arterial 20 (74) 7 (26) 27 (33) venous injury No. injury grade/total No. (%): 4 4/48 (8) 44/48 (92) 48/82 (59) 5 32/34 (94) 2/34 (6) 34/82 (41) Presentation of Penetrating vs Blunt Injury Simultaneous artery and vein injury was more common with penetrating injury. This approached statistical significance (44% vs 23%, p 0.07). Renal surgery was performed for 39 of 41 penetrating injuries vs 16 of 40 blunt injuries (95% vs 40%, p ). The nephrectomy rate for penetrating injury was not statistically different from that for blunt trauma (41% vs 23%, p 0.11). Specifically 10 of 16 blunt MRAIs were characterized by complete renal pedicle avulsion, accounting for our high exploration and nephrectomy rate in this group. Presentation of MRAI vs SRAI Age and sex were no different between patients with MRAI and SRAI. There was also no association between the mechanism of injury (blunt vs penetrating) and type of vascular injury (MRAI vs SRAI) (p 0.49). Shock was more common in patients with MRAI than SRAI (72% vs 44%, p 0.02). There was no difference in hematuria rates or rates of associated injury. Concomitant venous injury occurred in 20 of 36 patients (56%) with MRAI but in only 7 of 45 (16%) with SRAI (p , table 1). Treatment of MRAI vs SRAI Four deaths occurred before definitive intervention, including 2 of 36 in patients with MRAI and 2 of 46 in patients with SRAI. Renal surgery was performed in 31 of 34 initial survivors (91%) with MRAI and in 24 of 44 (55%) with SRAI (p 0.001), whereas close observation was the treatment of choice in 18 of 44 (41%) with SRAI but in only 2 of 34 (6%) with MRAI (p 0.001). The 2 patients who were observed for MRAI had sustained blunt trauma with renal artery thrombosis and a normal contralateral kidney. One patient with MRAI underwent embolization. Due to the small numbers of observations and embolizations in the MRAI group they are not the focus of this report. The nephrectomy rate for MRAI was significantly higher than that for SRAI (23 of 31 or 72% vs 3 of 24 or 13%, p ). All patients with concomitant vein injury were surgically treated regardless of whether the arterial injury was MRAI or SRAI. Of the patients with MRA plus vein injuries 16 of 20 (75%) underwent nephrectomy. This represented the majority of nephrectomies done for MRAI (16 of 23 or 70%, see figure). Vascular ligation with or without partial nephrectomy was performed in 54% of SRAI cases but it was not appropriate for MRAI. Main or segmental arterial embolization to control renal hemorrhage was done in 5 patients, including 1 with MRAI from penetrating trauma and 4 with SRAI, which was penetrating and blunt in 2 each. The survival rate following embolization was 80% (4 of 5 patients). A single male patient died of other injuries soon after embolization for blunt SRAI. The 2 patients who underwent embolization following penetrating SRAI subsequently underwent exploration for other reasons and the kidneys were examined with no further intervention. Outcomes in MRAI (Nephrectomy vs Vascular Repair) The mean transfusion requirement for patients undergoing nephrectomy did not differ significantly from that in patients undergoing vascular repair (10,275 vs 6,125 ml, p 0.64). Likewise mean hospital stay, mortality, renal failure incidence, hypertension, renal insufficiency and nonrenal complications following nephrectomy did not differ statistically from those in patients with vessel repair (table 2). Outcomes in SRAI (Operative vs Nonoperative Management) After SRAI the mean transfusion requirement for patients treated operatively vs those who were observed was 4,994 vs 820 ml (p 0.01). Mean hospital stay in the operative vs

3 RENAL ARTERIAL INJURIES 2453 Interventions following renal artery injury. Asterisk indicates that 1 of 2 patients died of associated injuries following successful embolization to control renal hemorrhage. observation groups was 29 vs 11 days (p 0.23, table 3). All other outcome variables, including death, renal failure, hypertension, renal insufficiency and other complications, did not differ statistically between cases managed operatively and nonoperatively. Multivariate Analysis Logistic regression analysis did not identify any independent predictors of mortality or nephrectomy. Specifically the artery injured (main vs segmental), injury grade, injury mechanism, shock, associated injuries or intervention type (nephrectomy vs renal sparing) did not statistically correlate with survival. The overall mortality rate was 19% (15 of 81 patients). Similarly the artery injured, injury grade, injury mechanism, shock or associated injuries did not statistically correlate with nephrectomy. DISCUSSION Renovascular injury following trauma is relatively uncommon and it occurs most often in multiply injured patients. In our series 89% of patients with renal artery trauma had associated injuries, of which most required operative intervention. Optimal management for RAI remains controversial, particularly due to the low incidence and the lack of prospective studies in the literature. For major MRAIs several groups favor aggressive, early surgical intervention with expedient nephrectomy as part of the damage control paradigm in hemodynamically unstable patients. 1,5,6,12 15 In rare situations in which arterial and/or venous repair is feasible during initial laparotomy for trauma the reported renal salvage rates are consistently low. 2,7,16,17 A recent multicenter analysis of management and outcomes following major renovascular injury at 6 trauma centers showed that attempted arterial repair for grade 5 renovascular injury was associated with significantly poorer outcomes compared to immediate nephrectomy. 6 On this basis the investigators recommended immediate nephrectomy for grade 5 RAI regardless of the mechanism of injury. In our series nephrectomy was more commonly performed than vascular repair but this difference was not statistically significant (p 0.21). In 2 patients nephrectomy was done following failed vascular repair, which was recognized at the initial operation. Independent predictors of nephrectomy were not identified on multivariate logistic analysis. Outcomes with regard to mortality, TABLE 2. Outcomes of nephrectomy vs vessel repair for MRAI Nephrectomy Vessel Repair p Value No. pts 23 8 Mean SD transfusion requirement (ml) 10,275 12,871 6,125 5, Mean SD LOS (days) No. death during hospitalization (%) 6 (26) 1 (12.5) 0.4 No. renal failure (%) 2* (8) 2 (25) 0.29 No. hypertension 1* No. renal insufficiency (Cr mg/dl 1.5 or greater) or impaired renal function 1 (4) 1 (12.5) % or less by renal scan (%) No. other renal related complications during hospitalization (%) 1 (4) No. nonrenal nonfatal complications during hospitalization (%) 4 (17) No. uneventful hospitalization, lost to followup (%) 6 (26) 2 (25) 1 No. no sequelae, Cr 1.5 mg/dl or less at followup 3 (13) 2 (25) 0.59 * The patient who underwent bilateral renovascular surgery had hypertension associated with renal failure. Eg infection.

4 2454 RENAL ARTERIAL INJURIES TABLE 3. Outcomes of operative vs nonoperative SRAD management Operative Nonoperative p Value No. pts Mean SD transfusion requirement (ml) 4,994 6, Mean SD LOS (days) No. death during hospitalization (%) 2 (8) 1 (6) 1 No. renal failure (%) 2* (8) No. hypertension (%) 1* 1 (6) 1 No. renal insufficiency (Cr 1.5 mg/dl or less) or impaired renal function 40% or less by renal scan (%) 2 (8) 3 (16) 0.65 No. persistent urinoma requiring intervention (%) 0 1 (6) 0.46 No. renal related complications during hospitalization (%) 1 (4) 0 1 No. nonrenal nonfatal complications during hospitalization (%) 4 (17) No. uneventful hospitalization, lost to followup (%) 6 (25) 9 (50) 0.11 No. no sequelae, Cr 1.5 mg/dl or less at followup (%) 7 (29) 3 (16) 0.46 * The patient who underwent bilateral renovascular surgery had hypertension associated with renal failure. Excluding urinoma. transfusion requirements, hospital stay, renal failure, renal dysfunction and complications did not differ for nephrectomy vs vessel repair. However, these findings are confounded by the fact that the majority of patients with MRAI who were operated on were victims of penetrating trauma (21 of 31, 68%), had multiple associated injuries and had mortality attributable to associated injuries and not to the choice of nephrectomy vs vessel repair. Additionally, the incidence of many complications was so low that this study was underpowered to detect a difference in rates between the groups. Excluding patients lost to followup, the renal salvage rate in patients undergoing MRA repair in our series (defined by followup serum Cr less than 1.5 mg/dl or post-repair split function greater than 25%) was only 33% (2 of 6). This is consistent with previously published series. 2 Reconstruction of blunt MRAI resulted in immediate intraoperative thrombosis and nephrectomy in 2 of 2 patients. Immediate renal thrombosis was not seen in the reconstruction of MRAI after penetrating trauma but delayed poor graft function occurred in 2 of these 4 patients. Since immediate graft loss developed only after blunt and not penetrating MRAI, it is our belief that blunt trauma may result in an intimal flap or other diffuse, unrecognized arterial injury that causes arterial thrombosis. Because data on renal failure and complications did not differ between reconstruction and nephrectomy, these findings do not support a benefit to vascular repair, given a normally functioning contralateral kidney, particularly in blunt MRAI cases. With regard to SRAI, 1 small series with long-term followup supports nonoperative management in hemodynamically stable patients with excellent outcomes. 8 However, some groups have advocated operative intervention for patients otherwise undergoing laparotomy, 1 while others cited a poor long-term outcome with nonoperative management. 16 Concerns include delayed bleeding and urinoma formation in cases managed nonoperatively. In our series transfusion requirements were higher and there was a trend toward longer hospitalization for patients with SRAI managed operatively but these observations are biased by the mechanism of injury, number of associated injuries and overall patient acuity because patients who underwent renal exploration were undergoing laparotomy for associated injuries. Outcomes in patients with SRAI managed nonoperatively were excellent and only 1 of 18 (6%) required subsequent intervention, that is stenting for persistent urinoma (table 3). These findings support a nonoperative role for SRAI in appropriately selected and staged patients. Our data represent almost 30 years of experience. In that time much has changed in trauma care. In particular the management of renal parenchymal injuries has shifted toward observation of many penetrating wounds and almost all blunt injuries. Similarly the advent of improved imaging techniques and interventional radiology has allowed nonoperative management of almost all blunt abdominal trauma and of an increasing proportion of penetrating abdominal trauma. About half of our patients were victims of blunt trauma. With improved radiographic staging and management in the modern era some of these patients would not likely undergo exploratory laparotomy for their other injuries. As such, renal exploration might be avoided in some cases. However, 10 of 16 patients with blunt MRAI had total renal pedicle avulsion, which is an absolute indication for nephrectomy. 4 Therefore, the management paradigm for MRAI continues to be significantly weighted toward intervention, in contrast to the renal parenchymal trauma paradigm. Our data support observation for all SRAIs if properly staged regardless of whether laparotomy is done for other purposes. Patients with MRAI had equivalent perioperative outcomes after nephrectomy or reconstruction, arguing against renal reconstruction, given the poor functional results. If the abdomen is open, we continue to advocate nephrectomy over observation in the stable patient with MRAI due to the reported 43% incidence of de novo hypertension caused by the thrombosed renal unit. 18 In patients who are increasingly treated nonoperatively for nonrenal injuries some with MRAI and a contained hematoma may be observed with serial hematocrit and computerized tomography. In select cases there may be a role for interventional radiology management for MRAI, that is a covered stent or embolization. In hemodynamically unstable patients or those with renal pedicle avulsion nephrectomy should be performed immediately. CONCLUSIONS In our series nephrectomy for MRAI had outcomes similar to those of vascular repair with regard to transfusion requirements, LOS and mortality. Posttreatment renal function was not measurably worsened by nephrectomy, while renal

5 RENAL ARTERIAL INJURIES 2455 salvage following vessel repair was low, thus, supporting a role for nephrectomy in the management of MRAI. For SRAI transfusion requirements were higher and there was a trend toward longer hospitalization, higher mortality and higher complication rates in patients treated operatively. However, these trends were reflective of the mechanism and overall severity of injury, and not of the specific therapeutic modality used. The excellent outcomes in patients observed following SRAI support a role for conservative management in this subgroup. Abbreviations and Acronyms Cr creatinine LOS length of stay MRAI main renal artery injury SRAI segmental renal artery injury REFERENCES 1. Carroll PR, McAninch JW, Klosterman P and Greenblatt M: Renovascular trauma: risk assessment, surgical management, and outcome. J Trauma 1990; 30: Cass AS, Bubrick M, Luxenberg M, Gleich P and Smith C: Renal pedicle injury in patients with multiple injuries. J Trauma 1985; 25: Wessells H, Suh D, Porter JR, Rivara F, MacKenzie EJ, Jurkovich GJ et al: Renal injury and operative management in the United States: results of a population-based study. J Trauma 2003; 54: Santucci RA, Wessells H, Bartsch G, Descotes J, Heyns CF, McAninch JW et al: Evaluation and management of renal injuries: consensus statement of the renal trauma subcommittee. BJU Int 2004; 93: Santucci RA and Fisher MB: The literature increasingly supports expectant (conservative) management of renal trauma a systematic review. J Trauma 2005; 59: Knudson MM, Harrison PB, Hoyt DB, Shatz DV, Zietlow SP, Bergstein JM et al: Outcome after major renovascular injuries: a Western trauma association multicenter report. J Trauma 2000; 49: Turner WW Jr, Snyder WH 3rd and Fry WJ: Mortality and renal salvage after renovascular trauma. A review of 94 patients treated in a 20 year period. Am J Surg 1983; 146: Bertini JE Jr, Flechner SM, Miller P, Ben-Menachem Y and Fischer RP: The natural history of traumatic branch renal artery injury. J Urol 1986; 135: Husmann DA, Gilling PJ, Perry MO, Morris JS and Boone TB: Major renal lacerations with a devitalized fragment following blunt abdominal trauma: a comparison between nonoperative (expectant) versus surgical management. J Urol 1993; 150: Husmann DA and Morris JS: Attempted nonoperative management of blunt renal lacerations extending through the corticomedullary junction: the short-term and long-term sequelae. J Urol 1990; 143: Moudouni SM, Patard JJ, Manunta A, Guiraud P, Guille F and Lobel B: A conservative approach to major blunt renal lacerations with urinary extravasation and devitalized renal segments. BJU Int 2001; 87: DiGiacomo JC, Rotondo MF, Kauder DR and Schwab CW: The role of nephrectomy in the acutely injured. Arch Surg 2001; 136: Ivatury RR, Zubowski R and Stahl WM: Penetrating renovascular trauma. J Trauma 1989; 29: Narrod JA, Moore EE, Posner M and Peterson NE: Nephrectomy following trauma impact on patient outcome. J Trauma 1985; 25: Nash PA, Bruce JE and McAninch JW: Nephrectomy for traumatic renal injuries. J Urol 1995; 153: Cass AS, Luxenberg M, Gleich P and Smith C: Long-term results of conservative and surgical management of blunt renal lacerations. Br J Urol 1987; 59: Clark DE, Georgitis JW and Ray FS: Renal arterial injuries caused by blunt trauma. Surgery 1981; 90: Haas CA, Dinchman KH, Nasrallah PF and Spirnak JP: Traumatic renal artery occlusion: a 15-year review. J Trauma 1998; 45: 557.