Prone Position

The prone position is often used for operations involving the spine [1] and posterior fossa [2], and for certain urologic and lower gastrointestinal procedures [3]. This position provides excellent surgical access and, with proper positioning, a decrease in extradural vein and cerebrospinal fluid pressures [1]. The main complications associated with the prone position include: ocular and auricular injuries, musculoskeletal injuries, venous air embolism, compression of limb neurovascular bundles, ischemia of the skin at pressure points, and excessive joint flexion and extension [2]. Additional rare complications have also been reported [4-7]. However, the overall complication rate is low [6,8]. We report two cases with rare complications of the prone position. Case Reports Case 1 A 13-yr-old male (height 153 cm, weight 40 kg) with a history of partial complex seizures treated for 3 yr with carbamazepine presented for surgical repair of a cervical cavernous hemangioma. Surgery was indicated due to recurring symptoms (numbness of the left upper and lower limbs) and increasing size of the hemangioma at the level of C1-2 as measured by magnetic resonance imaging. The patient had never received an anesthetic and there was no family history of anesthetic related complications. The review of systems was negative except for exercise-induced asthma which was treated with albuterol. All preanesthetic laboratory results were within normal limits. General anesthesia was induced with thiopental 250 mg, fentanyl 100 micro gram, and vecuronium 5 mg and maintained with isoflurane 0.2%-0.8%, N2 O/O2 (FIO2 approximate 0.35), fentanyl, and vecuronium. The patient was placed in a three-pin Mayfield head fixation, then turned to the prone position on blanket rolls with his arms tucked along his sides; he remained in this position throughout the surgery (7 h). Monitoring included the routine anesthesia monitors, direct arterial blood pressure, esophageal temperature, urinary output, mass spectrometry, precordial Doppler, and somatosensory evoked potentials. High-dose glucocorticoids and vancomycin were administered during surgery. The patient was hemodynamically stable and there were no hypoxemic episodes. The lowest recorded body temperature was 35.6 degrees C. Blood transfusion was not required. The results of arterial blood gas determinations are shown in Table 1 and Figure 1.Table 1: Case 1: Results of Arterial Blood Gases (All Drawn from a Radial Arterial Line) During Surgery and in the Intensive Care UnitFigure 1: Case 1, Association between pHa (PH) and bicarbonate during surgery and intensive care unit stay.Due to the length of the procedure and facial edema, the patient remained tracheally intubated and was transferred to the intensive care unit (ICU). Upon arrival in the ICU, ventilation was controlled, blood pressure was 130/60 mm Hg, heart rate was 120 bpm and the patient's temperature was 36.4 degrees C. In response to the severe metabolic acidosis (see Table 1), the patient was hyperventilated and repeat intravenous boluses of sodium bicarbonate were administered (total dose of 95 mEq). Lactate was 14.4 mmol/L (normal range 0.93-1.65), chloride 102 mEq/L, and the anion gap was 27. Body temperature increased to 38 degrees C over the next few hours, but remained stable thereafter. The following morning the patient was awake and alert. The metabolic acidosis had decreased significantly and lactate decreased to 3.1 mmol/L. His trachea was extubated. Laboratory tests revealed creatinine kinase (CK) of 976 U/L (normal range 94-491). Aspartate aminotransferase was 47 U/L (normal range 20-40), total bilirubin 1.7 mg/dL with direct bilirubin of 0.4 mg/dL. Prothrombin time and activated prothrombin time (aPTT) were normal. All biochemical abnormalities returned to normal without treatment within the next few days. The patient was discharged from the hospital 10 days after surgery. Case 2 A 37-yr-old female (height 169 cm, weight 90 kg) was admitted for surgical correction of idiopathic scoliosis. Past medical history was unremarkable except for recently diagnosed peptic ulcer disease treated with cimetidine. She had previously received a general anesthetic and had had no problems. Eight units of blood for autologous transfusion were predonated over several weeks. Preoperative hemoglobin was 12.8 g/dL and the platelet count was normal. A two-stage surgical procedure was performed. The first included anterior thoracolumbar fusion with discectomies, bone grafts, and femoral rings. During this 10-h procedure, the patient was positioned in a modified right lateral decubitus position. She remained hemodynamically stable and received 400 mL of autologous blood. She was transferred to the ICU for observation and discharged to the orthopedic floor the following day without complications. Of note were postoperative CK levels of 4630 U/L (normal range 38-176) which decreased to 2997 U/L 15 h later. Six days later, the patient underwent the second stage operation which included posterior T3-L4 fusion, iliac crest bone grafts, and insertion of Isole rods and sublaminar wires. Preoperative laboratory tests included: hemoglobin (Hb) 10.5 g/dL, normal platelet count, glucose 129 mg/dL, creatinine 0.5 mg/dL, Na 139 mEq/L, K 3.9 mEq/L, alkaline phosphatase 146 U/L (normal range, 81-213), AST 59 U/L (normal range, 12-31) and total bilirubin 0.3 mg/dL. General anesthesia was induced with thiopental 375 mg, succinylcholine 100 mg, and oxymorphone 0.3 mg, and maintained with N2 O/O (2) (FIO2 approximate 0.35) isoflurane 0.5%-1%, vecuronium, and oxymorphone. Monitoring was the same as in Case 1. The patient was placed in the prone position for 12 h on the Children's Hospital of Philadelphia four-poster frame. Both arms were abducted less than 90 degrees and the elbows were flexed. Both upper extremities were carefully padded. The axilla was specifically noted to be acceptably positioned and the abdomen was free of any pressure. Weight was borne on the anterior-superior iliac bones and the chest. The nipples were directed to the midline. Systolic blood pressure was more than 90 mm Hg and the heart rate was between 100 and 130 bpm during surgery. The lowest temperature was 35.4 degrees C. Urine output was 50-100 mL/h. The patient received 10.7 L lactated Ringers' solution, 1.7 L of 5% albumin, 8 U of autologous blood, 2 U of homologous blood, and 6 U of cell saver blood. The lowest Hb during surgery was 6.4 g/dL. Upon admission to the ICU she was sedated and mechanically ventilated. With a FIO2 0.4, PaO2 was 173 mm Hg, PaCO2 38 mm Hg, pH (a) 7.39, HCO3 23 mEq/L, base excess 1 mEq/L, and Hb was 12.4 g/dL. Urine output was adequate (0.5-1.5 mL centered dot kg-1 centered dot h-1) but its color was red-brown. Results of increased liver function tests, urine myoglobin, and serum CK are shown in Table 2. Due to suspected rhabdomyolysis with myoglobinuria, intravenous sodium bicarbonate was administered to maintain urine pHa > 6.5 [9] and diuresis was induced with mannitol. Urine output was 3200 mL in the first 24 h and 5600 mL on the following day. To correct coagulation abnormalities, 4 U of fresh frozen plasma and vitamin K were administered. The patient's trachea was extubated on the first postoperative day and she was discharged from the hospital 9 days after surgery.Table 2: Case 2: Main Laboratory Results After Admission to the Intensive Care Unit (ICU)Discussion This report presents several uncommon complications including lactic acidosis, myoglobinuria, and abnormally increased CK, which may have been associated with the combined effect of prolonged prone positioning and general anesthesia. Increased lactate levels support a diagnosis of tissue hypoperfusion. Compression of the major arteries of the legs may be associated with the prone position [2]. However, no signs of muscle ischemia could be detected and serum CK was only mildly increased in Case 1. In the absence of hypotension, hypoxemia, or severe anemia, hypoperfusion of intraabdominal organs could be implicated as a cause of lactic acidosis. A commonly overlooked cause of lactic acidosis is visceral ischemia [10], but clinical signs of visceral ischemia may be minimal [11]. Lactic acidosis can resolve rapidly (half-life of 1 h) [12] or have a prolonged course (with a half-life as long as 18 h) [13]. In this case, the lactate level decreased from 14.4 mmol/L to 3.1 mmol/L over 12 h. McCormack et al. [14] demonstrated a substantial increase in portal vein lactate and a decrease in bicarbonate and pHa 30 min after the release of mesenteric occlusion. This can occur as acids, including lactate, are flushed from the ischemic bowel and may explain the worsening of metabolic acidosis measured after arrival of the first case in the ICU compared to that measured late in surgery. A diagnosis of malignant hyperthermia was considered in Case 1. However, of the classic triad of skeletal muscle rigidity, mildly increased body temperature, and metabolic acidosis, only the latter was present. The common denominator in all cases of malignant hyperthermia is hypermetabolism [15]. A clinical consequence is increased CO2 production, which frequently is the earliest and most sensitive indicator [16]. No increase in CO2 production was observed in this patient. Abnormalities in liver function tests (LFTs) may indicate liver hypoperfusion due to hepatic artery and portal vein compression in both patients. The small abdominal girth of the first patient, along with mild abdominal pressure, could potentially compress vessels between the operating Table andthe vertebral column. Likewise, compression of these vessels may have also been responsible for the abnormal LFTs in the second patient. We are unaware of any studies investigating liver blood flow during surgery in the prone position or measuring LFTs after prolonged cases as presented (7 and 12 h). Disseminated intravascular coagulation (DIC) was considered in the second case since the patient received a total of 16 U of blood during surgery [17]. However, a normal aPTT and normal platelet count throughout her postoperative course make this diagnosis unlikely. The aPTT is a more sensitive screening test for DIC than prothrombin time [18] and platelet count is abnormal in more than 90% of patients with DIC [19]. The coagulopathy was probably secondary to massive transfusion, although the role of liver injury due to hypoperfusion cannot be excluded. Both patients had increased CK levels. This increase was mild in the first case, but substantial in the second, and was associated with myoglobinuria. This mechanism of injury has been reported in patients who maintained the same position for many hours [20]. Although there are reports about similar injuries during anesthesia and surgery [21-24], only two are related to the prone position [25,26]. The association between rhabdomyolysis and operative position was studied by Targa et al. [27]. They found that rhabdomyolysis was associated with the lateral position and long-lasting surgery. However, their mean surgical time was less than 3.5 h. The measured serum CK is invariably increased in rhabdomyolysis and the release of this enzyme from skeletal muscle is a fairly specific marker of injury, particularly when increases are extreme [28]. CK levels usually peak within 24 h and decrease rapidly thereafter [29]. A similar trend was observed in Case 2 where the patient had a substantial increase of CK levels after her first surgery during which she was in the modified right lateral position for 10 h. Since the mechanism of injury in Case 2 could be pressure-stretch myopathy, rather than visceral hypoperfusion and ischemia [23], this patient's injury can be related to lengthy operative procedures in the lateral or prone positions. The treatment of rhabdomyolysis and myoglobinurea has been reviewed in detail elsewhere [9,30]. Succinylcholine can cause rhabdomyolysis and myoglobinuria in patients with muscle disease [31]. Muscle disease is an important predisposing factor that should be suspected in patients with scoliosis [32]. Therefore, succinylcholine may have contributed to the increased CK in Case 2. An association between surgical repair of idiopathic scoliosis and postoperative rhabdomyolysis and myoglobinuria has not been reported. Therefore, it is difficult to incriminate the extensive surgery as the sole contributor to this condition. In conclusion, we have reported several uncommon complications associated with prolonged surgery in the prone position. Mechanisms of injury may be related to visceral hypoperfusion and ischemia, pressure related rhabdomyolysis and myoglobinuria, or both. The outcome was favorable in both patients.