| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Laboratory of Experimental Surgery and Surgical Research, Cardiothoracic Surgery Clinic, Laboratory of Histology and Embryology, and ||First Clinic of Surgery, School of Medicine, Democritus University of Thrace, Alexandroupolis, Greece; and Laboratory of Biochemistry, University General Hospital of Alexandroupolis, Alexandroupolis, Greece. #These authors contributed equally to this work.
Address correspondence and reprint requests to Petros Ypsilantis, DVM, PhD, Laboratory of Experimental Surgery and Surgical Research, University General Hospital of Alexandroupolis, Dragana 68100 Alexandroupolis, Greece. Address e-mail to pipsil{at}med.duth.gr.
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Eighteen healthy male rabbits were endotracheally intubated and sedated with propofol 2% (Group P), sevoflurane (Group S) or sevoflurane while receiving Intralipid 10% (Group SI). Sedation lasted 48 h or until death (Group P) or the maximum surviving period of Group P (Groups S and SI). The initial propofol infusion rate (20 mg · kg–1 · h–1) or sevoflurane concentration (1.5%) was adjusted, if needed, to maintain a standard level of sedation. Blood biochemical analysis was performed in serial blood samples and histologic examination in the heart, lungs, liver, gallbladder, kidneys, urinary bladder, and quadriceps femoris muscle at autopsy.
RESULTS: The mortality rate was 100% (surviving period, 26–38 h) for Group P, whereas 0% for Groups S and SI. The initial propofol infusion rate had to be increased up to 65.7 ± 4.6 mg · kg–1 · h–1 and sevoflurane concentration up to 4%. Serum liver function indices, lipids and creatine kinase were significantly increased (P < 0.05) in Groups P and SI and lactate was increased only in Group P, whereas amylase was increased in all groups. In Group P, histologic examination revealed myocarditis, pulmonary edema with interstitial pneumonia, hepatitis, steatosis, and focal liver necrosis, cholangitis, gallbladder necrosis, acute tubular necrosis of the kidneys, focal loss of the urinary bladder epithelium, and rhabdomyolysis of skeletal muscles; in Group S, low-grade bronchitis and incipient inflammation of the liver and the kidneys; and in Group SI, low-grade bronchitis, liver steatosis and hepatitis, and incipient inflammation of the gallbladder, kidneys, and urinary bladder.
CONCLUSIONS: Continuous infusion of 2% propofol at large doses for the sedation of rabbits undergoing prolonged mechanical ventilation induced fatal multiorgan dysfunction syndrome similar to the propofol infusion syndrome seen in humans. Our novel findings including lung, liver, gallbladder, and urinary bladder injury were also noted. The role of propofol's lipid vehicle in the manifestation of the syndrome was minor. Sevoflurane proved to be a safe alternative medication for prolonged sedation.
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Propofol is commercially formulated in an oil-in-water emulsion. Long-term propofol infusions have been associated with serum lipid increases, which may lead to lipid deposition (8). Lipemia itself can impair mitochondrial oxygen uptake and precipitate problems of oxygen utilization which are considered key pathogenic mechanisms of PRIS (3). Different propofol formulations (1%, 2%, and 6%) have been tested for long-term administration schedules to reduce the volume and the amount of lipids (9,10). In one study, no problems were encountered when children 4–17 mo of age were sedated for 6–17 h in the ICU with a new formulation of propofol 6% in Lipofundin MCT/LCT (11). Nevertheless, the role of propofol's lipid vehicle in the occurrence of PRIS has not been elucidated.
In this project we sought to investigate the possible side effects of propofol 2% sedation at large doses in prolonged mechanical ventilation in rabbits, and to determine whether these are related to propofol's lipid vehicle, on the basis of blood biochemical profile and histopathologic findings of vital and other organs. We used our formerly described model of propofol tolerance, in which large propofol doses were infused for up to 36 h to maintain a standard level of sedation (12).
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Eighteen healthy male New Zealand White rabbits, 6 mo of age, weighing 4.0–4.5 kg, were used. They were housed individually in wire cages under controlled environmental conditions (20°C–22°C room temperature, 50%60% relative humidity, 12 h light/12 h dark cycle). The facilities were in compliance with Directive 86/609/EEC. The animals were provided with 125 g of commercial pelleted diet (certified rabbit chow no. 51, EL.VI.Z., Xanthi, Greece) per day and tap water ad libitum.
Experiment 1: Effect of Propofol 2% Sedation on Rabbits Under Prolonged Mechanical Ventilation
A group of animals (Group P, n = 6) was endotracheally intubated, mechanically ventilated, and sedated by continuous infusion of propofol 2% for 48 h or until death. Propofol was administered at an initial rate of 20 mg · kg–1 · h–1 which was adjusted, when needed, to maintain a light level of sedation (see below) according to our formerly described model of propofol tolerance (12). Blood samples were collected every 12 h and analyzed for a number of biochemical variables, indicative of vital organs function. At the end of the experiment, autopsy was performed and the heart, lungs, liver, gallbladder, kidneys, urinary bladder, and part of the quadriceps femoris muscle were excised to be subjected to histologic examination. The propofol concentration was measured in bile samples collected at autopsy.
Experiment 2: Effect of Sevoflurane Sedation on Rabbits Under Prolonged Mechanical Ventilation
To distinguish the effects of propofol sedation from those of prolonged mechanical ventilation, a group of animals (Group S, n = 6) was endotracheally intubated, mechanically ventilated, and sedated by an alternative sedative drug, the inhaled anesthetic sevoflurane for the maximum survival period of Group P. The initial sevoflurane concentration was 1%, which was adjusted, if needed, to maintain a light level of sedation. Blood biochemical and histologic examinations were performed as in Group P.
Experiment 3: Effect of Intralipid 10% on Sevoflurane-Sedated Rabbits Under Prolonged Mechanical Ventilation
To study the effect of propofol's 2% lipid vehicle (Intralipid 10%) on sedated rabbits under prolonged mechanical ventilation, a group of animals (Group SI, n = 6) was endotracheally intubated, mechanically ventilated, and sedated by administration of the inhaled anesthetic sevoflurane while receiving Intralipid 10% for the maximum surviving period of Group P. The initial sevoflurane concentration was 1%, which was adjusted, if needed, to maintain a light level of sedation. Intralipid 10% was administered by continuous IV infusion at a rate which was adjusted hourly according to the average hourly propofol 2% infusion rate in Group P. Blood biochemical and histologic examinations were performed as in Group P.
Preparation of the Animals
After overnight fasting, the animals were premedicated with a mixture of xylazine (5 mg/kg, IM) and atropine (0.05 mg/kg, IM), and 20 min later, were anesthetized with ketamine hydrochloride (50 mg/kg, IM) to be endotracheally intubated. Artificial ventilation was then initiated using a pressure ventilator (ADS 1000 Veterinary Anesthesia Delivery System/Engler, FL). The initial settings (flow rate, 12 L/min; respiratory rate, 30 breaths/min; peak inspiratory pressure, 12 cm H2O; and inspiration triggering, –2 cm H2O) provided a tidal volume of around 16 mL/kg and were adjusted, if needed, to maintain arterial Paco2 between 30 and 40 mm Hg. The initial 40% Fio2 (fraction of inspired O2) was adjusted if needed to achieve Pao2 over 90 mm Hg or Spo2 >95%.
A 6-lead electrocardiogram, invasive arterial blood pressure, heart rate and Spo2 (Dash 3000 patient monitor/Marquette, Milwaukee, WI) as well as respiratory rate and body temperature (SDI Vet Ox 4700 Plus monitor/SDI, Kansas City, MO) were continuously measured, and arterial blood gases together with the acid-base status and the levels of electrolytes and glucose were assessed every 2 h (GEM Premier 3000/IL, Lexington, MA). NaHCO3 solution was administered IV if metabolic acidosis occurred (30 mL NaHCO3 8% was administered within 1 h when pH < 7.25 and HCO3– <15 mEq/L). Electrolytes and glucose levels were corrected if necessary. Electrocardiogram electrodes were attached on the lateral aspects of shoulders and thighs. A central ear artery and two marginal ear veins were catheterized (23-gauge polyethylene catheter) to provide intraarterial and IV access, respectively. Spo2 and body temperature were measured using rectal electrodes. The animals were placed at lateral recumbency on a heated-surface operation table and were covered with an isothermic blanket to keep them normothermic; their position was changed (left lateral to right lateral) every 6 h. The urinary bladder was catheterized using a 10 CH Foley catheter to facilitate urine collection. During the experiment, the animals received either Ringer's lactated saline solution IV for electrolyte and fluid support or dextrose 5% saline when glucose levels decreased to <80 mg/dL (normal range 75–140 mg/dL). The basic rule that guided fluid administration management was "urine volume + blood sample volume = total drug volume + saline volume + hydroxyethyl starch volume." Hydroxyethyl starch (Voluven/Freefex) replaced blood sample volume, and saline replaced all other losses. Each animal also received 300 IU tinzaparin sodium (Innohep/Leo), SC every 12 h for thrombosis prophylaxis.
Propofol Administration
Propofol was administered in the form of 2% oil-in-water injectable emulsion (Diprivan 2% inj., Zeneca Spa/Italy). At the first signs of awakening from general anesthesia (restoration of corneal reflex), the animals began to receive propofol by continuous IV infusion at an initial infusion rate (IR) of 20 mg · kg–1 · h–1 using an infusion pump (Module DPS Orchestra I.S., Fresenious Vial S.A., Brezins, France). The IR was adjusted, whenever this was necessary, by steps of 5 mg · kg–1 · h–1, to maintain the desired level of sedation. The level of sedation was assessed based on reflex response every 30 min; earlier assessments were made only if clinical signs indicating awakening of the animal (25% increase of heart rate and/or arterial blood pressure from previous measurement, any increase of respiratory rate from ventilator setting) were present. The criteria for the desired level of sedation are listed in Table 1. To avoid propofol overdose, if the IR had to be reduced to maintain an adequate level of sedation, attempts were made to decrease the IR (in steps of 5 mg · kg–1 · h–1) every 15 min until the appearance of initial signs of awakening (increased heart rate, increased arterial blood pressure, restoration of palpebral reflex, restoration of pina and pedal reflexes at light painful stimuli); in these cases the IR was again increased back to the previous safe point.
|
Sevoflurane Administration
Sevoflurane (Sevorane/Abbott, Maidenhead, Berkshire, UK) was administered using a conventional vaporizer at an initial rate of 1.5%, which was adjusted if needed to maintain light level of sedation (Table 1).
Intralipid 10%
In Diprivan 2%, propofol is formulated in Intralipid 10% (Fresenius Kabi, Runcorn, Cheshire, UK), which consists of soybean oil (100 mg/mL), glycerol (22.5 mg/mL), egg lecithin (12 mg/mL), with a pH around 8 and an osmolarity around 300 mOsm/kg water.
Measurement of Blood Biochemical Variables
Arterial blood samples were collected in heparinized syringes and analyzed for pH, Paco2, Pao2, HCO3–, oxygen saturation, potassium, sodium, calcium, glucose, base excess, hematocrit and lactate levels using a blood gas/electrolyte/metabolite/coagulation/co-oxymetry analyzer (GEM Premier 3000/IL, Milan, Italy). Venous blood samples were collected in test tubes and centrifuged at 3000 rpm for 15 min to obtain serum. The following variables were measured (ILAB 600 analyser, Instrumentation Laboratory SpA, Milano, Italy): urea, creatinine, aspartate aminotrasferase (AST), alanine aminotransferase (ALT), lactate dehydrogenase (LDH), total protein content, albumins, globulins, total bilirubin, direct bilirubin, indirect bilirubin, 
-glutamine transferase (-GT), cholesterol, triglycerides, amylase, C-reactive protein, creatine kinase (CK), and CK-MB. Cardiac troponin I was also measured using the AIA 21 system analyzer (TOSOH Bioscience SRL/Tessenderlo, Belgium) according to the manufacturer's instructions.
Histologic Examination
After the animals had died, the heart, lungs, liver, gallbladder, kidneys, urinary bladder, and part of the quadriceps femoris muscle were excised and fixed in 10% neutral buffered formaldehyde, then embedded in paraffin wax, sectioned serially at 4 µm, and stained routinely with hematoxylin-eosin. Slides were subjected to histologic examination under an Olympus BX40 microscope by two independent researchers in a blinded fashion.
Measurement of Propofol Concentration in Bile
Propofol was measured in bile samples using the method of high-performance liquid chromatography, as previously described (12).
Statistical Analysis
Before the study, based on pilot experiments, a power calculation was performed with 90% power and 
error set at 0.05 (two-sided). We estimated that for the detection of 100% difference in the most important biochemical variables (mean difference ± sd, 23 ± 14 IU/L for AST, 40 ± 9 IU/L for ALT, 240 ± 73 IU/L for LDH, 61 ± 20 mg/dL for cholesterol, 77 ± 10 mg/dL for triglycerides and 1500 ± 250 IU/L for CK) a maximum number of six rabbits per group would be required. Data from blood biochemical analysis were presented as means ± sd and subjected to repeated measures analysis of variance. Differences between means within time points in the same group were tested using the Bonferroni test and between groups for the same time point using the Student's t-test. A probability of P < 0.05 was considered statistically significant.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Propofol IR Adjustments
Mean propofol IR adjustments in Group P are shown in Figure 1. Adjustments followed a five-phase pattern of infusion as previously described (12): the initial IR remained steady for 1.4 ± 0.3 h (Phase 1), then was progressively increased for 10.2 ± 3.8 h (Phase 2) up to a plateau for 3.3 ± 1.4 h (Phase 3), then was gradually decreased for 12.3 ± 1.9 h (Phase 4), and finally remained steady for 6.0 ± 2.4 h until the death of the animals (Phase 5). Initial IR (20 mg · kg–1 · h–1) was increased up to 65.7 ± 4.6 mg · kg–1 · h–1 (228.3% ± 22.9% of the initial IR).
|
Blood Biochemistry
There was a significant increase (P < 0.05) of serum AST, ALT, LDH, total protein content, globulins, total bilirubin, direct bilirubin, indirect bilirubin, cholesterol, triglycerides, and CK levels from the 12th hour, of -GT and lactate from the 24th hour and of amylase levels at the 36th hour after the onset of propofol infusion. Transient hypoglycemia was also noted (Table 3).
|
Autoptic—Histopathologic Findings
Heart.
At autopsy, the heart was in a contracted state. Histologic examination revealed infiltration of acute inflammatory cells between the myocardial fibers and small foci of myofibrial degeneration (Fig. 2A). Small foci of incipient rhabdomyolysis were also noted.
|
Lungs.
Major macroscopic findings of the lungs at autopsy included lung enlargement and congestion and pink frothy edema fluid effusing from lung sections and filling the tracheal cannula. The lungs had milky tincture. Histologic examination of the lungs revealed interstitial pneumonia and pulmonary edema (Fig. 2B). Alveolar spaces were completely lined by hyperplastic alveolar type II epithelial cells and were filled with alveolar macrophages. Sparse interstitial lymphoplasmacytic inflammation with some neutrophils and eosinophils were also observed. There was focal pulmonary edema and hemorrhage.
Liver.
The liver, at autopsy, had milky tincture. Histologic examination revealed low-grade active hepatitis with lymphocytic inflammation scattered through lobule (Fig. 2C). There was low-grade cholestasis (bile droplets in hepatocytes and bile plugs in canaliculi). The bile ducts were surrounded by neutrophils, whereas there were no neutrophils in lumens. Focal hepatocyte necrosis and low-grade fatty change (steatosis) were also noted.
Gallbladder.
At autopsy, the gallbladder was dark green. Histologic examination revealed necrosis of the three layers of the gallbladder's wall (mucosa, muscularis, and serosa layers) (Fig. 2D).
Kidneys.
Acute tubular necrosis with granular and eosinophilic hyaline casts in tubules and interstitial edema were observed in the kidneys (Fig. 2E). Characteristic findings were the loss of cell nuclei and the preservation of cell outline of tubular epithelium.
Urinary bladder.
Histologic examination of the urinary bladder revealed focal loss of epithelium and submucosal edema (Fig. 2F). There was no necrosis or hemorrhage.
Skeletal muscle.
Histologic examination of the quadriceps femoris muscle revealed rhabdomyolysis. Acute necrotic reaction in skeletal muscle with swelling, loss of striations and vacuoles were noted. Many nuclei were degenerated and a mild acute inflammation was noticed (Fig. 2G).
Bile propofol concentration.
Mean propofol concentration in bile samples was 10.33 ± 2.25 mg/L.
Experiment 2: Effect of Sevoflurane Sedation on Rabbits Under Prolonged Mechanical Ventilation
Mortality Rate, Clinical Course
All the animals survived their intended survival period (38 h). During the experiment, no deviation of vital signs was noted apart from an increase in heart rate (>200 bpm) from the 12.3 ± 4.3 h after the onset of sevoflurane administration. Cardiorespiratory data are presented in Table 2.
Sevoflurane Concentration Adjustments
The initial sevoflurane concentration (1.5%) had to be increased up to 4.0% during the experiment (Table 2) to maintain the desired level of sedation.
Blood Biochemistry
There was a significant increase (P < 0.05) of serum CK levels from the 12th hour and of amylase from the 24th hour after the onset of sevoflurane administration. Total protein content and albumins decreased from the 12th hour. Transient fluctuations of glucose, Ca, and lactate levels were also noted (Table 4). CK levels were significantly lower (P < 0.05) than in Group P.
|
Autoptic—Histopathologic Findings
Lungs.
At autopsy, no lesion was found. Histologic examination revealed low-grade bronchitis, characterized by the presence of inflammatory cells (mainly eosinophils) around the bronchi.
Liver.
No lesion was found at autopsy, whereas at histology, there were a few inflammatory cells (lymphocytes) at the portal tracts.
Kidneys.
No lesions were found in three of six animals. In the rest of the animals, the kidneys had reddish color at autopsy, and microscopic examination showed few scattered lymphocytes in the renal parenchyma.
Heart, gallbladder, urinary bladder and skeletal muscle.
No lesions were found.
Experiment 3: Effect of Intralipid 10% on Sevoflurane-Sedated Rabbits Under Prolonged Mechanical Ventilation
Mortality Rate, Clinical Course
All the animals survived their intended survival period (38 h). During the experiment, no deviation of vital signs was noted apart from an increase in heart rate from around the fifth hour from the onset of sedation.
Sevoflurane Concentration Adjustments
The initial sevoflurane concentration (1.5%) had to be increased up to 4.0% during the experiment (Table 2) to maintain the desired level of sedation.
Blood Biochemistry
Total protein content, globulins, total bilirubin, direct bilirubin, indirect bilirubin, cholesterol, triglycerides, and CK were gradually increased from the 12th hour, whereas albumins decreased from the 12th hour of the experiment. Serum AST, ALT, LDH, and -GT levels increased from the 24th hour. Transient glucose, K and Ca fluctuations were also noted (Table 5). Triglyceride and CK levels were significantly lower (P < 0.05) than in Group P.
|
Autoptic—Histopathologic Findings
Lungs.
At autopsy, no lesion was found. Histologic examination revealed low-grade bronchitis, characterized by the presence of inflammatory cells (mainly eosinophils) around the bronchi.
Liver.
At autopsy, the liver had a slight milky tincture. At histology, there were low-grade fatty changes in the hepatocytes (steatosis) and low-grade active hepatitis characterized by inflammatory cells at the portal tracts and hepatocytes.
Gallbladder.
There were a few inflammatory cells at the submucosal layer.
Kidneys.
At autopsy, the kidneys had a reddish color. Microscopic examination showed few scattered lymphocytes in the renal parenchyma.
Urinary bladder.
There were a few inflammatory cells at the submucosal layer and edema.
Heart, skeletal muscle.
No lesions were found.
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In our attempt to determine whether these detrimental effects were due to prolonged mechanical ventilation, irrespective of the sedative drug used, we kept mechanically ventilated rabbits under sevoflurane sedation for the maximum survival period of the propofol-sedated rabbits. Inhaled anesthetics have only recently been proposed as alternative medication for sedation in the ICU (13,14). Isoflurane has been tested for prolonged sedation in critically ill patients, using the anesthetic conserving device (AnaConDa/Hudson RCI, Upplands Vasby, Sweden) with promising results (13); however, at present no relevant studies on the application of sevoflurane are available. This is the first report of safe sevoflurane sedation under prolonged mechanical ventilation at subanesthetic concentrations using a traditional vaporizer. The sevoflurane concentration had to be increased from 1.5% up to 4% to maintain the desired level of sedation. There are no reports of sevoflurane tolerance after prolonged administration. The mechanism of action, and hence that of tolerance development, of inhaled anesthetics is under investigation; there is evidence that the -aminobutyric acid system and the central benzodiazepine receptors are involved in the mechanism of isoflurane action (15). All animals survived the 38-h study period, with tachycardia being the only clinical symptom after the 12th hour. Blood biochemical analysis showed an increase of serum amylase, whereas histologic examination revealed a minimally affected liver histology and mild bronchitis, possibly because of mechanical ventilation.
Having established a safe sedation regimen for keeping rabbits under prolonged mechanical ventilation, we infused Intralipid in sevoflurane-sedated rabbits to investigate the possible side effects of propofol's lipid vehicle. Apart from tachycardia, noted after the first few hours of sedation, no serious deviation in vital signs was observed, and all animals survived the 38-h study period. There was an increase of serum lipids, direct and indirect bilirubin, and hepatic function indices. Histologic examination revealed low-grade steatosis and inflammation of the liver and low-grade bronchitis and incipient inflammation of the kidneys and the urinary bladder. Our results, when compared with those of the propofol-treated animals, show that propofol's lipid solvent is responsible for only a small part of the catastrophic manifestations of the complex PRIS.
The pathogenic mechanism of PRIS has not been fully elucidated. Prolonged infusion of propofol at large doses has been shown to impair the oxidation of free fatty acids (16), which are the primary energy source for muscle. The subsequent imbalance between energy demand and supply leads to peripheral and cardiac muscle necrosis (17–20). In our study, a profound increase of CK blood levels from the 12th hour after the onset of propofol infusion indicated extended myocytolysis. This was supported by the false positive blood reaction in urinalysis by an automated analyzer, in the absence of erythrocytes at microscopic examination (data not presented), most likely secondary to the passage of myoglobin through the glomerulus (21). Histologic examination of skeletal muscles eventually demonstrated the presence of extended rhabdomyolysis.
Cardiac failure, characterized by progressive bradycardia leading to asystole, has been ascribed as the major cause of death in patients with PRIS (1). In our study, monitoring of cardiac function in propofol-sedated animals did not show any conduction defect or bradycardia, but, on the contrary, there was tachycardia during the last hours of the experiment. Cardiac muscle injury markers cardiac troponin I and CK-MB, were not increased and, at autopsy, the heart was found in a contracted state, distinguishing myofibrial degeneration from myocardial infarction (22). Finally, histologic examination revealed myocarditis with only small foci of myofibrial degeneration and incipient rhabdomyolysis. Thus, the relatively small extent of cardiac injury could clarify the absence of terminal bradycardia which is characteristic of PRIS in patients.
In our study, the fatal outcome of propofol-sedated animals was attributed to acute pulmonary edema based on the clinical course, the blood laboratory results (arterial blood gas analysis) and finally the autoptic and histologic findings. Although one cannot exclude the cardiac origin of the pulmonary edema noted, the concurrent presence of interstitial pneumonia suggested the primary involvement of the lungs in the pathogenesis of the syndrome. In previous reports, respiratory diseases preceded PRIS, constituting the main reason for admission to the ICU, and were not consequences of propofol administration (1).
Rhabdomyolysis has been shown to induce renal failure, a main feature of PRIS. Myoglobin toxicity is related to mechanical "plugging" of the renal tubules, direct toxic effects, and alterations in renal blood flow (21). In our study, acute tubular necrosis with granular and eosinophilic hyaline casts (most probably myoglobin) in tubules was found at histologic examination of the kidneys. Surprisingly, blood biochemical parameters related to renal function were not altered, indicating late injury of the kidneys.
Microscopic examination of the urinary bladder of propofol-sedated rabbits revealed focal loss of epithelium and submucosal edema. This is the first report of urinary bladder damage after propofol administration and could have been due to toxic accumulation of propofol or its metabolites in the urine, although no such measurement was performed.
There are contradictory reports on the cause of hypertriglyceridemia noted after prolonged infusion of propofol in ICU patients; some authors associate it with the lipid solvent of propofol (10,23,24) and impaired fat metabolism in critically ill patients (25), and others with propofol per se (26). In our study, serum triglycerides were substantially increased in the propofol group, whereas only moderately increased in the intralipid group, despite similarly elevated cholesterol levels in both groups. This suggests the primary involvement of propofol in the unique increase of serum triglycerides on the grounds of intralipid-induced hyperlipemia. Impaired ß-oxidation of free fatty acids induced by propofol (16) could be related to the marked increase of serum triglycerides, although it has not been determined whether propofol directly affects the mobilization of free fatty acids from triglycerides. Liver dysfunction may have adversely affected lipid kinetics, as well. Liver damage was more serious in the propofol group in which focal hepatocyte necrosis and cholangitis were noted in addition to low-grade hepatitis and steatosis. There are only a few reports of hepatocellular injury related to propofol administration in the ICU setting (20,27,28). Observations have been based on blood laboratory results, whereas in only one case, diagnosis was bioptic, revealing 10% zone III liver necrosis with fatty change (28).
Critically ill patients in the ICU are at risk of developing bile stasis or acalculous cholecystitis (29–34). However, the contribution of potential risk factors, which include fasting, total parenteral nutrition, sedation, mechanical ventilation, infection and shock, has not been elucidated. This is the first report that implicates propofol in the induction of profound gallbladder damage. Apart from the significant increase in direct and indirect bilirubin levels, histologic examination revealed extended gallbladder necrosis, probably due to the toxic necrotizing effect of propofol on this organ (propofol bile level around 10 mg/L); this must have led to cholestasis, and accumulation of propofol in the bile, because propofol does not alter the sphincter of Oddi pressure profile (35). The toxic necrotizing effect in various organs, such as the liver, the gallbladder, and the urinary bladder could be related to the pathogenic mechanism responsible for muscle necrosis due to an unsatisfied energy demand. Alternatively, accumulated propofol or its metabolites propofol glucuronide, 1-quinol glucuronide, and 4-quinol glucuronide (36) could also be implicated in the toxic effects noted.
Prolonged infusion of propofol has been associated with pancreatitis (37), whereas in children, even short-term sedation with propofol increases pancreatic enzyme levels without, though, any clinical symptom of pancreatitis (38). In ICU propofol-sedated adult patients, standard doses of the drug do not induce pancreatic dysfunction (39), but hypertriglyceridemia-related pancreatitis is often encountered (40). In our study, serum amylase was increased at the end of the study period in propofol-sedated rabbits, indicative of incipient pancreatitis. Although the lipid solvent must have contributed to this finding, one cannot exclude the potential role of sevoflurane as well, because amylase levels were also increased in both the Intralipid and the sevoflurane groups. However, no pancreatic tissue was examined to verify the diagnosis.
Progressive severe lactic acidemia, a major feature of PRIS, was noted in the propofol-treated animals. Metabolic acidosis could not be corrected despite the administration of sodium bicarbonate (up to 120 mL NaHCO3 8% during the experiment in aliquots of 30 mL in hourly infusions). The increase of plasma lactate may have been due to either excessive production in states of inadequate cellular oxygen delivery, defects of cellular metabolic pathways or decreased lactate metabolism.
It has been reported that a propofol infusion period of longer than 48 h is required for PRIS to develop, but in one report, features of PRIS were noted only 3 h after propofol infusion at large doses for neurosurgery anesthesia (41). In our study, blood biochemical analysis indicated liver dysfunction and rhabdomyolysis as early as 12 h after the onset of propofol infusion. The early indications of PRIS coincided with the remarkable increase in the propofol infusion rate.
Our study has limitations. A more objective and sensitive method for evaluating the depth of sedation, such as the Bispectral Index or the electroencephalogram, would offer more precise adjustments of the sedative infusion rate or concentration. Because of the higher metabolic rate of rabbits, we used an initial propofol infusion rate of 20 mg · kg–1 · h–1, increased up to 220%, which is disproportionally higher than that reported in PRIS cases in humans (>4 mg · kg–1 · h–1) (3). This is a limitation of the study in terms of dosing resemblance with humans, but, conversely, it may explain the early onset of PRIS development in this species. Subsequently, the early mortality of rabbits may not have allowed incipient pathologic situations to fully develop. As a result, not all the PRIS characteristics were manifested in the extent and severity seen in humans. Additionally, there should be a species-dependent difference in adverse effects of propofol responsible for the timing of different organs' toxicity.
In conclusion, the continuous infusion of 2% propofol at large doses for the sedation of rabbits undergoing prolonged mechanical ventilation induced fatal multiorgan dysfunction syndrome resembling PRIS. Our novel findings, including lung, liver, gallbladder, and urinary bladder injury, may guide physicians to further investigate the potential side effects of propofol in the sedation of ICU patients. The role of propofol's lipid vehicle in the development of PRIS proved to be minor, and was confined mainly to hyperlipemia and liver dysfunction. Finally, sevoflurane was a viable alternative sedation choice for prolonged mechanical ventilation in this rabbit model.
|
Footnotes |
|---|
Supported by Laboratory of Experimental Surgery and Surgical Research, School of Medicine, Democritus University of Thrace, Alexandroupolis, Greece.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
