DiGiacomo JC, Rotondo MF, Kauder DR, Schwab CW. The Role of Nephrectomy in the Acutely Injured. Arch Surg. 2001;136(9):1045-1049. doi:10.1001/archsurg.136.9.1045
The high mortality in patients who undergo nephrectomy after trauma is not secondary to the nephrectomy itself but is the consequence of a more severe constellation of injuries associated with renal injuries that require operative intervention.
A retrospective review of all patients identified using International Classification of Diseases, Ninth Revision codes as having sustained renal injuries over a 62-month period.
Seventy-eight patients with renal injuries who underwent exploratory laparotomy were identified.
All medical records were reviewed for patient management, definitive care, and outcome. Based on outcome, patients were assigned to either the survivor or nonsurvivor group. For patients who underwent nephrectomy, intraoperative core temperature changes, estimated blood loss, and operative time were also reviewed.
Seventy-eight patients with renal injuries who underwent exploratory laparotomy were identified. Twenty-nine patients underwent laparotomy with conservative management of the renal injury, of whom 5 (17.2%) died. Twelve patients had renal injuries repaired and all survived. Thirty-seven patients underwent nephrectomy, of whom 16 (43.2%) died. Compared with nephrectomy survivors, nephrectomy nonsurvivors had a significantly lower initial systolic blood pressure, higher Injury Severity Score, higher incidence of extra-abdominal injuries, shorter operative duration, and higher estimated operative blood loss. The nephrectomy survivors' core temperature increased a mean of 0.5°C in the operating room, while the nephrectomy nonsurvivors' core temperature cooled a mean of 0.8°C.
Patients who undergo trauma nephrectomy tend to be severely injured and hemodynamically unstable and warrant nephrectomy as part of the damage control paradigm. That a high percentage of patients die after nephrectomy for trauma demonstrates the severity of the overall constellation of injury and is not a consequence of the nephrectomy itself.
THE REPORTED incidence of renal injury after trauma has traditionally been reported as ranging from 0.8% to 16%, and is more commonly associated with blunt rather than penetrating mechanisms.1- 5 The proportion of injuries from penetrating vs blunt injury is variable and probably is a major contributing factor to the 13.3% to 53.1% incidence range of nephrectomy reported by various authors.6- 14 The implementation of nonoperative management of solid organ injury has become widespread and has decreased the overall need for laparotomy after injury. There remains, however, a population of injured patients who require laparotomy to address issues of blood loss and severe injury. By omitting the least injured patients who previously were surgically explored from those undergoing trauma laparotomy, the new population of operative trauma patients will be more significantly injured and in greater need of urgent operative intervention, further polarizing the spectrum of therapy to nonoperative management of solid organ injury and damage control surgery. The patients in this study were seen at a single urban trauma center that sees a large percentage of penetrating injuries.This retrospective analysis sought to reassess the population of patients who require laparotomy and have renal injuries, with specific attention to those patients who require nephrectomy.
The medical records of all patients with renal injuries over a 62-month period identified retrospectively using International Classification of Diseases, Ninth Revision codes were reviewed. Seventy-eight patients with renal injuries who underwent exploratory laparotomy were identified from a population of 8068 consecutive patients evaluated at an urban level I trauma center.
The medical records were reviewed for initial patient management, definitive care, and outcome. Based on outcome, patients were assigned to either the survivor or nonsurvivor group. The variables of age, systolic blood pressure, pulse rate, and Injury Severity Score (ISS) were subjected to nonparametric analysis (Mann-Whitney test). For patients who underwent nephrectomy, intraoperative core temperature changes, estimated blood loss, and operative time were also subjected to the same nonparametric analysis. Differences were considered statistically significant if P<.05. Hypotension was defined as an initial blood pressure in the trauma receiving area of less than 90 mm Hg. Renal injuries were graded based on the operative findings in accord with the American Association for the Surgery of Trauma organ injury scales.13
Whether radiographic urinary tract studies were performed was noted for all patients. For those patients in whom intravenous urograms (IVUs) were performed, the technique consisted of an intravenous bolus of 1.5 mL/kg of iodinated contrast material and portable abdominal radiographs. Additional intra-abdominal and retroperitoneal injuries identified at the time of laparotomy were also noted. The indications for nephrectomy were categorized as (1) parenchymal injury precluding repair (ie, shattered or ruptured kidney), (2) renal hilar injury, (3) renal vascular injury, or (4) hemodynamic instability.
The hospital course of each patient was reviewed for the development of renal failure and outcome. For those patients who survived for longer than 24 hours, the daily serum urea nitrogen and serum creatinine levels were reviewed until the patient was discharged from the hospital. Renal failure was defined as an increase of the serum creatinine level to higher than 2.5 mg/dL.
Seventy-eight patients with renal injuries who underwent exploratory laparotomy were identified. All were surgically explored through a standard midline incision. Early control of the renal vessels was performed only when the patient was warm and hemodynamically stable in the operating department and had a retroperitoneal hematoma that did not engulf the renal vessels. The indication for laparotomy was penetrating abdominal trauma in 59 cases (gunshot wounds, 43 patients; stab wounds, 14; and shotgun wounds, 2). For the blunt trauma patients, the indications were gross blood by diagnostic peritoneal lavage in 9 patients, computed tomographic (CT) scan findings in 6 patients, and clinical findings of hypotension and abdominal distention in the trauma receiving area in 4 patients.
Preoperative radiographic staging consisted of CT scan in 7 patients (6 blunt injuries and 1 penetrating wound), and 1-view IVU in 2 (both penetrating wounds). None of the 7 patients were hypotensive. While all 7 CT scans accurately assessed the renal units, 1 IVU study falsely demonstrated nonvisualization of a kidney found to have a grade 2 laceration at exploration.
Twenty-nine patients (14 patients with penetrating wounds and 15 patients with blunt injuries) underwent laparotomy with conservative management of the renal injury, of whom 5 (17.2%) died (Table 1). The ISS between the survivors (19.5) and nonsurvivors (35.2) was the only factor that reached statistical significance. Four of the 5 nonsurvivors died within 24 hours of admission, 2 from head injuries and 2 from unresuscitatable shock. The fifth died of multisystem organ failure on hospital day 11.
Twelve patients (11 patients with penetrating wounds and 1 patient with blunt injuries) had renal injuries repaired and all survived (Table 2). The mean ISS was 17.6 and the mean renal injury grade was 2.6 (range, grades 2-3). None of these patients presented with hypotension and all remained warm and hemodynamically stable in the operating room. The method of repair was debridement and simple repair in 11 cases and heminephrectomy in 1 case.
Thirty-seven patients (34 patients with penetrating wounds and 3 with blunt injuries) underwent nephrectomy, of whom 16 (43.2%) died. The mean ISS was 25.4 and the mean renal injury grade was 4.5 (range, grades 3-5). Eight died in the operating room and 8 died in the intensive care unit within 24 hours of admission. The 8 patients who died in the intensive care unit were typically hypothermic, hypotensive, and coagulopathic when they died. There were statistically significant differences between the nephrectomy survivors and nonsurvivors for initial systolic blood pressure (survivors, 102.7 mm Hg; nonsurvivors, 73.5 mm Hg), ISS (survivors, 20.2; nonsurvivors, 32.1), incidence of extra-abdominal injuries (survivors, 43%; nonsurvivors, 75%), intraoperative change of core temperature (survivors, +0.5°C; nonsurvivors, −0.8°C), and estimated operative blood loss (Table 3). The types and frequency of intra-abdominal injuries were similar between the survivors (mean, 2.3; range, 0-5) and nonsurvivors (mean, 2.6; range, 1-8). The ranges for the initial core temperature in the operating room were similar for survivors (33.7°C-36.2°C) and nonsurvivors (31.4°C-36.8°C). The ranges for the last core temperature in the operating room demonstrated minimal overlap between survivors (core temperature, 33.5°C-37.3°C) and nonsurvivors (core temperature, 31.0°C-34.5°C). The distribution of extra-abdominal injuries are listed in Table 4.
The indication for nephrectomy was hilar injury in 25 patients, of whom 15 survived; parenchymal injury precluding repair in 21 patients, of whom 13 survived; and renal vascular injuries in 20 patients, of whom 9 survived. Hemodynamic instability was the most mortal of indications; only 5 of 19 patients with this indication survived. Seven patients had a single indication for nephrectomy, which was hilar injury in 1 instance, parenchymal injury in 5, and hemodynamic instability in 1. The 14 patients with 2 indications for nephrectomy were evenly distributed among the indications. All 9 patients with 3 indications for nephrectomy included hilar injury and renal vascular injury. Five of the 6 patients with all 4 indications for nephrectomy died. One patient underwent nephrectomy as part of the repair of a vena cava injury.
All 21 survivors who underwent nephrectomy had sustained penetrating injury. The mean operative time was 288 minutes (range, 132-515 minutes), the mean estimated blood loss was 4529 mL (range, 400-10 000 mL), and the mean number of red blood cell units transfused was 8.9 (range, 0-40 U). Renal rupture was present in 13 cases, hilar injury in 15 cases, renal vessels injury in 9 cases, and intraoperative hemodynamic instability in 6 cases. Only 1 of these 6 would have been otherwise potentially repairable.
Among the 16 nonsurvivors who underwent nephrectomy, 12 had been shot, 1 had been stabbed, and 3 were injured in motor vehicle crashes. The mean operative time was 128 minutes (range, 50-244 minutes), the mean estimated blood loss was 18 731 mL (range, 3850-80 000 mL), and the mean number of red blood cell units transfused was 23.5 (range, 4-44 U). Renal rupture was present in 8 cases, hilar injury in 10 cases, renal vessels injury in 10 cases, and intraoperative hemodynamic instability in 13 cases. Only 4 had renal injuries that would have been otherwise potentially repairable.
The urology service was consulted intraoperatively in 14 cases. Only 2 patients had transient hypotension in the trauma receiving area, and 2 additional patients had a single episode of transient hypotension in the operating room. None of these patients died. Three patients had grade 2 injuries, 4 had grade 3 injuries, 4 had grade 4 injuries, and 3 had grade 5 injuries. Two patients only received renal explorations and 5 had renal repair (4 simple repairs; 1 heminephrectomy). Seven patients underwent nephrectomy, 6 of whom had been shot and 1 stabbed. Renal rupture was present in 6 cases, hilar injury in 5 cases, and renal vessels injury in 3 cases. The mean ISS for the 7 patients who underwent nephrectomy was 19.4, mean operative time was 317 minutes (range, 183-473 minutes), the mean estimated blood loss was 4693 mL (range, 400-10 000 mL), and the mean number of red blood cell units transfused was 6.6 (range 0-14 U).
The remaining 30 nephrectomies were performed by the trauma service. Twenty-two patients had sustained gunshot wounds, 1 was shot with a shotgun, 4 were stabbed, and 3 were involved in motor vehicle crashes. Sixteen patients died. The mean ISS was 26.7 (survivors, 20.5; nonsurvivors, 32.1). A total of 19 patients were hemodynamically unstable in the operating room. As stated earlier, 1 patient underwent nephrectomy as part of the repair of a vena cava injury.
No survivors developed renal failure or renal insufficiency, regardless of renal injury or intervention. The survivors who underwent nephrectomy had a mean increase in serum creatinine level of 0.2 mg/dL (18 µmol/L).
Thirty-seven of the 78 patients who underwent exploration had a nephrectomy performed, for an operative nephrectomy rate of 47.4%. This is higher than the rates reported by Sagalowsky et al,6 Narrod et al,7 Corriere et al,8 McAninch et al,10 or Nash et al,11 but is consistent with Atala et al9 and Velmahos et al.12 The patient populations are not comparable, however. While the mean ISS of the patients who underwent nephrectomy reported by McAninch et al10 was 39.5, the number of patients who demonstrated hemodynamic instability in the operating room was not given. Neither Sagalowsky et al,6 Narrod et al,7 Corriere et al,8 Atala et al,9 nor Nash et al11 included any ISS scores. In 1998, Velmahos et al12 reported the mean ISS of patients undergoing nephrectomy at 20, significantly lower than the mean ISS of the patients undergoing nephrectomy in this study of 25.4.
The frequency of hypotension or intraoperative hemodynamic instability was also not given in many of these reports. Sagalowsky et al6 and Atala et al9 do not relate this information. Corriere et al8 encountered these conditions present in 2 (7.7%) of 26 patients, Nash et al11 in 6 (23.1%) of 26 patients, and Velmahos et al12 in 5 (29.4%) of 17 patients. In the current study, hemodynamic instability was noted in 19 patients (51.4%).
Mortality rates were only reported by 2 authors. Narrod et al7 had 6 (42.9%) deaths among 14 patients undergoing nephrectomy. Nash et al11 reported the lowest nephrectomy rate (13.3%), but had a mortality rate of 73.1% for those patients. The mortality rate for patients undergoing nephrectomy in the current study was 43.2%. As stated previously, there were no differences between the initial core temperature in the operating room between survivors (core temperature, 33.7°C-36.2°C) and nonsurvivors (core temperature, 31.4°C-36.8°C) and the last core temperature determination in the operating room demonstrated only minimal overlap. Most striking was the determination of core temperature change during laparotomy. Only 1 patient whose core temperature increased during surgery died. This patient warmed from 31.4°C to 32.4°C, but died in the operating room hypothermic and coagulopathic. The lowest core temperature in any survivor was 33.5°C. No patient whose core temperature cooled more than 0.5°C in the operating room survived. The findings here are similar to those of Jurkovich et al,14 who reported no survivors in patients with severe truncal trauma whose lowest core temperature was below 32°C. Krishna et al15 in 1998 reported similar findings for core temperatures below 33°C. That this temperature range represents such a critical threshold for survival is likely caused in large part by the inhibition temperatures below 34°C have on the coagulation pathways and platelet function.16
As a group, the patients who underwent nephrectomy were not significantly different from the other 2 groups, with the exception of the initial blood pressure, frequency of hypotension, and renal injury grade. There were 5 nonsurvivors among the 29 patients who underwent laparotomy but did not have any specific renal intervention. It is not surprising, therefore, that the mean renal injury grade was 2.0 overall, and only ranged from 1 to 3. Of the 3 patients who died of blunt trauma in this subgroup, 2 died of head injuries. The 2 patients who died of penetrating mechanism both had associated liver and extensive retroperitoneal injuries. The absence of nonsurvivors among the 12 patients managed with renal repair should be fully expected, as the decision to salvage an injured kidney must, by definition, be associated with hemodynamic and physiologic normalcy.
Of the 57 patients who were stable in the operating room, 27 required no specific renal intervention, 12 had the renal injury repaired, and 18 underwent nephrectomy. Among these, 14 were evaluated by the urology staff. None of these 14 patients died, as should be expected. Two patients received no intervention, 5 underwent salvage procedures, and 7 underwent nephrectomy despite intraoperative stability, which made intraoperative urology consultation possible. Eleven of the remaining 43 hemodynamically stable patients managed by the trauma surgeon underwent nephrectomy, of whom 3 died. In comparison, 21 patients overall were hemodynamically unstable in the operating room, of whom 19 underwent nephrectomy. The remaining 2 died in the operating room with grade 2 renal injuries. Six of the 19 survived.
One criticism that may be levied against this report is a lack of preoperative radiographic renal evaluation. We believe this criticism is unjustified for 2 reasons. First, the lack of specificity and sensitivity for IVU in the acutely injured has been long recognized. In 1969 Scott et al1 reported the findings of 121 IVUs obtained in patients who had sustained penetrating injury. More than one third of the studies interpreted as normal missed major renal injuries. Scott et al concluded, "This high incidence of falsely negative excretory urograms limits the usefulness of this study as a reliable diagnostic test to exclude renal injury."1(p248) Cass17 reported the findings of 1147 IVUs from a predominantly blunt injured population. Of 1022 patients diagnosed as having renal contusions, 7.4% had IVUs that help aid the incorrect diagnosis of major renal injury. Of the 80 patients with renal lacerations, 20% had normal IVUs and 16% had IVUs that help aid the incorrect diagnosis of unilateral nonfunction. Cass and Luxenberg18 subsequently studied IVU unilateral renal nonfunction and found 15% were associated with renal contusions and 23% were associated with renal lacerations. The remaining 62% were associated with renal rupture or pedicle injury. In 1985, Carroll and McAninch19 pointed out that only 40% to 66% of IVU x-ray films were adequate studies and only 50% to 87% correlated accurately with arteriography or findings at exploration.
Second, the argument that an IVU is needed to confirm a functioning contralateral kidney is fallacious and implies the surgeon would not attempt a difficult salvage on a repairable kidney if a functioning contralateral kidney is present rather than absent. The decision for nephrectomy is "clinical," to be made at the operating table based on the overall clinical situation, not at an x-ray film viewbox. Regardless of whether nephrectomy of an injured kidney will render a patient functionally anephric, attempting renal salvage if repair of that kidney will compromise the patient's ability to survive is not appropriate, as demonstrated by Nash et al.11 Long-term follow-up studies of solitary renal function up to 50 years after unilateral nephrectomy clearly demonstrate that adaptive hyperfiltration in the remaining kidney does not have a deleterious effect and compensates for the loss of functional renal mass.20- 23 The incidence of congenital renal agenesis is estimated at 1 in 1000 to 5000 live births.24,25 The incidence of intrinsic renal disease, especially in the age groups at greatest risk for injury, is also very low. The most common etiologies are diabetic nephropathy, nephrosclerosis, and glomerulonephritis.27,28 The most common inherited renal disease is autosomal dominant polycystic disease, with an estimated prevalence of 1 in 1000 live births.26 Incidence rates of end-stage renal disease are 1 in 2100 American blacks, 1 in 4700 for American Hispanics, 1 in 300 to 900 for Native Americans, 1 in 2300 for Native Canadians, and 1 in 7700 for whites.27,28 It is unlikely that any of these kidneys, whether absent or functionally compromised, will feel "normal" to palpation. The rare patient who is rendered functionally anephric as the result of unilateral nephrectomy can be treated with dialysis and transplantation. Direct inspection of the penetrating injury tract is essential to accurately stage the urinary tract and avoid missed injuries. For the bluntly injured patient, evaluating the kidneys and bladder should accurately stage the urinary tract since blunt ureteral injury is exceedingly rare.
Injured patients rarely require laparotomy solely for the treatment of renal injury. With the broad success of nonoperative management of solid organ injury, those patients who undergo laparotomy tend to be more seriously injured and in need of more aggressive interventions. Patients who require nephrectomy after trauma tend to be the most severely injured and hemodynamically unstable, and warrant nephrectomy as part of the damage control paradigm in order to optimize the opportunity for survival.
Corresponding author: J. Christopher DiGiacomo, MD, Long Island Comprehensive Trauma Center, Department of Surgery, Nassau County Medical Center, 2201 Hempstead Turnpike, East Meadow, NY 11554.