Hyperglycemic Hyperosmolar Nonketotic Dehydration in Relation To Blood Ketone Body Ratio in Partially Hepatectomized Rabbits Toshio Nakatani, MD, Kazuhiro Yasuda, MD, and Kazue Ozawa, MD, Kyoto, JaPan
Due to the development of techniques for diagnosing small hepatomas, the Use of single segmentectomy of the liver is increasing. On the other hand, there are still many patients who require extended hepatic lobectomy for advanced hepatoma. In recent years, due to intensive preoperative, intraoperative, and postoperative management, the mortality rate after partial hepatectomy for hepatic malignant diseases has markedly decreased to only 2 percent in our surgical dePartment, even though most of the cases in Japan are accompanied by liver cirrhosis. Although partial hepatectomy can now be performed safely, there are still many problems to be solved in regard to postoperativemanagement of patients who have undergone massive hepatic resection. Among them is the Occurrence of postoperative hypoglycemia during the phase immediately after massive hepatic resection; which has been recog: nized as one of the most serious problems, In the early era of extensive hepatic resection, intensive administration of glucose solution had been recommended, often with a hypertonic solution to avoid hypoglycemia after massive hepatic resection [1:4], However, it has been reported that glucose intolerance developed in the early critical phase after massive hepatic resection in patients and animals [5,6]. It has also been reported that in 70 percent hepatectomized rabbits, the remnant liver predominantly utilizes fatty acids as an energy substrate rather than glucose in the phase immediately after the operation when the hepatic energy charge level (adenosine triphosphate plus 0.5 adenosine diphosphate)/(adenosine triphosphate plus adenosine diphosphate plus adenosine monophosphate) and the blood ketone body ratio (acetoacetate to ~-hydroxybutyrate in arterial blood) are decreased [7-9]. From the Second Department of Surgery, Faculty of Medicine, Kyoto University, Kyoto, Japan. Requests for reprints should be addressed to Toshio Nakatani, MD, Second Department of Surgery, Faculty of Medicine, Kyot0 University, 54, Shogoin-Kawara-cho, Sakyo-ku, Kyoto, 606, Japan.
Volume 155, April 1988
Therefore, the question arose of whether it is reasonable to administer a glucose solution intensively in the early period after massive hepatic resection. Moreover, several years ago, we encountered an instance of hyperglycemicnonketotic dehydration after right hepatic lobectomy and the patient died. Findings suggest th e presence of intracellular metabolic blocks which impede not only glucose synthesis but also glucose utilization in the critical state of hepatic function. The liver is the main organ that controls the blood glucose level. However, when hepatic function has severely deteriorated, it seems that the liver is no longer able to control the blood glucose level. Recent studies from our laboratory have indicated that the arterial blood ketone body ratio, which reflects the redox state of the liver mitochondria, is a reliable indicator of hepatic energy charge and should be used as a parameter to estimate the dynamic changes occurring in the energy status of the remnant liver [10]. In the present study, the ability of the liver to control the blood glucose level was studied in relation to the blood ketone body ratio in partially hepatectomized rabbits during postoperative continuous administration of a hyperosmolar glucose solution. Material and Methods Healthy, young male rabbits weighingbetween 2.1 and 2.4 kg were maintained on a diet of Clea CR-2 (Nippon Haigoshiro, Osaka, Japan) and water ad libitum preoperatively for about 2 weeks, and then they were fasted the night before operation. The rabbits were anesthetized with an intravenous injection of 15 mg/kg body weight of sodium 5-allyl-5-(1-methylbutyl)-2-thiobarbiturate and fixed supine on a surgical board. The rabbits thet were given a glucosesolution postoperatively were maintained in the supine position throughout the experiment. For 70 percent hepatectomy, the right anterior, left anterior, and right posterior lobeswere resected. In sham-
559
Nakatani et al
TABLE I
1000
Obstructive Jaundice 900
Changes in Plasma Osmotic Pressure 20 Hours After 70 Percent Hepatectomy With or Without Biliary Obstruction
Group
Rabbits
Osmotic Pressure (mOsm/liter)*
6
296 -t- 3
7 9
330 4- 8 378 4- 15 p <0.005
800 ,=a
Untreated normal Partial hepatectomy (70 % )Without biliary obstructiont With biliary obstruction w StatiStical significance
700
N --~ 600 ;>
70% Hepatectomy
.._1
Bile Duct Obstruction% at 12 hours /
r 0 r.1
O O
500
' Statistical values indicate the mean 4- standard error of the mean. - p <0.001 versus untreated rabbits. Blood glucose level was maintained at about 300 mg/dl for 20 hours. wBlood glucose level reached 790 mg/dl at 20 hours. 11p <0.01 versus partially hepatectomized rabbits without biliary obstruction.
400
300
r~
70% Hepatectomu
~
~5
200 ~ 100
0
~.-..~
~7 ' ~ 7
Sham operation 0
4
8
12
16
20
Hours after 70% HeDatectomy Figure 1. Changes in blood glucose levels during postoperative hypertonic glucose administration in 70 percent hepatectomized rabbits with or without biliary obstruction and in sham-operated rabbits. Values at each point indicate the number of animals studied.
tyrate were measured enzymatically [11,12]. The blood sugar level was determined by the o-toluidine method [13], the plasma bilirubin level" was measured by the method of Malloy and Evelyn [14], and the plasma osmotic pressure was measured using cryoscopy. Results were expressed as the mean 4- standard error of the mean. The significance of differences in mean values among the three groups was determined by analysis of variance. The significance between two of three groups was determined by modified Student's t test only when a significant difference among the groups was determined by analysis of variance [15]. Results
operated rabbits, laparotomy and mobilization of the liver were performed. In the first experiment, a 25 percent glucose solution was administered continuously at a rate of 2.7 ml/kg/hour by way of an infusion pump through a catheter inserted into the marginal ear vein beginning a t the end of the operation. External biliary drainage was established in another group of 70 percent hepatectomized rabbits by inserting a catheter into the common bile duct. During postoperative administration of a hyperosmolar glucose solution, the external biliary drainage catheter was obstructed at 12 hours to induce obstructive jaundice. In the second experiment, some of the rabbits that underwent 70 percent hepatectomy, with or without biliary obstruction, were given a single intravenous injection of 5 units of insulin approximately 15 hours after the operation during postoperative hypertonic glucose infusion. For assays of blood sugar, blood ketone bodies, and plasma bilirubin levels, arterial blood samples were obtained from a catheter inserted into the femoral artery. Preparation of the blood samples for measurement of acetoacetate and fl-hydroxybutyrate was performed as described previously [8]. Acetoacetate and f~-hydroxybu-
560
Figure 1 shows the changes in blood glucose levels during postoperative continuous hypertonic glucose administration in the 70 percent hepatectomized rabbits with or without biliary obstruction and in the s h a m - o p e r a t e d rabbits. In the s h a m - o p e r a t e d and 70 percent hepatectomized rabbits without biliary obstruction, the blood gIucose level was maintained a t a b o u t 200 and 300 mg/dl, respectively, for 20 hours. However, in the 70 percent hepatectomized rabbits with biliary obstruction beginning at 12 hours, the blood glucose level increased sharply after obstruction and reached 790 mg/dl at 20 hours. In this group, the plasma total bilirubin level increased significantly to 1.07 mg/dl during the 8 hour period of biliary obstruction c o m p a r e d with a level of 0.14 mg/dl in the u n t r e a t e d rabbits (p <0.0001) and a level of 0.28 mg/dl in the 70 p e r c e n t hepatectom i z e d r a b b i t s w i t h o u t b i l i a r y o b s t r u c t i o n (p <0.001). Severe glucosuria was observed throughout the experiment, and osmotic diuresis of about 4 ml/kg/hour for 8 to 12 hours was followed by anuria or oliguria in the rabbits with biliary obstruction. Plasma osmotic pressure increased to 378 m O s m / liter in the h e p a t e c t o m y with jaundice group 20 hours after the operation, as shown in T a b l e I. Urine
The American Journal of Surgery
Hyperglycemic Hyperosmolar Nonketotic Dehydration
1.0
9
i
i
i
i
J
i
~
i
)
Insulin,i.v.(5u.)
T
/
sulin
800 0.8
"%%
.o (13
>
,~,, 70% ~-, Hepatectomy ~',T I ~a.j2
0.6
"0 o nn
r0 0.4 r
.-.. 600 500
~
Hepatectomy Bile Duct
no
0.2
", ",, ~9
I
Obstruction / at 12 hoursj I
I
I
I
I
0
4
8
12
16
20
Hours after Operation Figure 2. Changes in blood ketone body ratios In 70 percent hepatectomlzed rabbits with or without blllary obstruction and In sham-operated rabbits during postoperative hypertonlc glucose administration. Values at each point Indicate the number Of animals Studied.
ketone bodies were not detected in any Ofthe groups throughout the experiment. In the preliminary experiment, it was shown that the arterial oxygen pressure, arterial carbon dioxide pressure, hydrogen ion concentration, and hematocrit value in the partially hepatectomized rabbits were not significantly different compared with those of the nonhepatectomized rabbits. The time courses of changes in the blood ketone body ratio are shown in Figure 2. Twelve hours after the Operation, the blood ketone body ratios decreased similarly to about 0.5 in both 70 percent hepatectomized groups with or without biliary obstruction. After biliary obstruction, the ratio decreased to 0.i9 at 20 hours in the jaundiced rabbits, whereas the ratio remained at 0.34 in the group without biliary 6bstruction. The effects of insulin administration 5ii the blood glucose levels of these rabbits are compared in Figure 3. In the shamoperated rabbits, the blood glucose level of 180 rag/ dl decreased significantly to 113 mg/dl 2 hours after insulin administration. In the 70 percent hepatectomized rabbits, it decreased from 308 to 248 mg/dl. On the other hand, in rabbits with biliary obstruc-
Volume 155, April 1988
"" 70%HeDatectomy
11
With obstructive Jaundice
70% Hepatectomy
Insulin ~!--)
.......
9"~ 300 0 0
ea 200
a~:-O
lO0
I
d,,2..1
400
(.9
70%
(-)
~-
>
o0
"0 0
0
7O0
8
o
(p
Sham i ~ 0 ooeratlOn I
I
I
0
30
60
I
(-) 0
(+)
I
] 20
rain)
Minutes after Insulin Administration Figure 3. Effects of insulin administration on blood glucose levels In 70 percent hePatectomlzed rabbits with or without biliary obstruction during postoperative hyperlonlc glucose administration. Approximately 15 hours after operation, a single Intravenous Injection of 5 units of Insulin was administered. The closed triangles, circles, and boxes Indicate the blood glucose levels without Insulin administration. The open triangles, circles, and boxes Indicate the levels after insulin administration. Each point represents the mean value of five or more animals.
lion, already increased blood glucose levels continued to increase without responding to insulin.
Comments Studies in our laboratory have shown that the blood ketone body ratio, which closely correlates with the hepatic mitochondrial redox state, reflects hepatic energy status in hemorrhagically shocked, jaundiced, and partially hepatectomized animals [16-18]. When clinically applied, the ratio can be regarded as a marker of hepatic energy status or the redox potential of the liver mitochondria and, therefore, might reflect the metabolic integrity of the entire body [19,20]. It has been reported that the blood ketone body ratio is normally maintained above 0.7. In a previous study of partially hepatectomized patients, when the ratio decreased to below 0.4, multiPle organ failure developed [19]. Only some of these patients recovered by cross-hemodialysis with pig or baboon liver using an interposed cuprophan membrane [21]. When the ratio further
561
Nakatani et al
decreased to below 0.25, all of the patients died. Therefore, it seems that a blood ketone body ratio of 0.25 is the point of no return of hepatic function. In the first experiment in our study, blood glucose levels in two groups of rabbits, in which the same amount of liver was resected with or without jaundice, showed marked differences at 20 hours. After the induction of obstructive jaUndice, the blood ketone body ratio significantly decreased to 0.19 and a marked glucose intolerance developed. This phenomenon clearly indicates that the blood glucose level during posthepatectomy glucose infusion does not correlate with the size of the remnant liver, but instead, negatively correlates with the changes in the blood ketone body ratio, which reflects the hepatic mitochondrial redox state. Reduced availability 0f hepatic oxygenation or blood flow could decrease the blood ketone body ratio. However, in the preliminary experiment, bepatectomized rabbits showed comparable arterial oxygen pressure to nonhepatectomized rabbits. In partially hepatectomized rats, it has been reported that the arterial Oxygen pressure, arterial carbondioxide pressure, hydrogen ion concentration, and hematocrit values were not significantly different compared with those of nonhepatectomized rats [22]. It has als0 been reported that hepatic blood flow per gram of liver tissue increased after hepatectomy in partially hepatectomized pigs [23]. In the present study, the biliary congestion shown to be due to common duct obstruction was considered not so severe to disturb liver perfusion. However, biliary congestion did affect liver mitochondrial metabolism, as the hepatic mitochondrial redox state had already been reduced when the common bile duct was obstructed. When the blood ketone body ratio markedly decreases, the hepatic mitoch0ndrial redox state is considered to be severely reduced. In such a condition, hepatic glucose metabolism would be impeded. Recent studies from our laboratory have shown that in 70 percent hepatectomized rabbits with a blood ketone body ratio of around 0.4, the hepatic concentrations of glucose-6-phosphate, fructose-6-phosphate, pyruvate, and lactate increased to approxim a t e l y 200 p e r c e n t of the values in the sham-operated control group before a glucose load [24]. It has been suggested that the reduced mitochondrial redox state, which is reflected in a decreased blood ketone body ratio, inhibits pyruvate dehydrogenase complex which regulates the entrance of pyruvate into the Krebs' cycle and also inhibits a few enzymes that regulate the Krebs' cycle activity, resulting in the inhibition of further oxidation of glucose [25,26]. Therefore, we believe that although the blood ketone body ratio is maintained at around 0.4, the blood glucose level is barely under hepatic control. However, once the ratio decreases further toward 0.25 or even lower, the ability of the liver to metabolize administered glucose
562
would be severely impeded. In such a crisis of hepatic energy status, gluconeogenesis is considered to be inhibited. Therefore, if exogenous glucose is not administered, severe hypoglycemiawill develop. On th e other hand, once exogenous hypertonic glucose solution is administered, the blood glucose level increases markedly, resulting in hyperosmolar dehydration. In such a severely reduced redox state, it is very difficult to control the blood glucose level. Insulin resistance was observed in the presence of deteriorated hepatic function in the second experiment. When the blood ketone body ratio is severely decreased, the liver does not seem to be able to correct the hyperglycemia, even with a large amount of insulin. In such a hepatic energy crisis, not only is glucose Utilization inhibited as already ediscussed herein, but glycogen synthesis also becomes inhibited: Therefore, marked insulin resistance develops. When the blood ketone body ratio is severely decreased to lower than 0.4 in patients, glucose solution should be adminiStered carefully with frequent measurement of the blood glucose level. It has been reported that the remnant liver predominantly utilizes endogenous fatty acids as an energy substrate when the blood ketone body ratio moderately decreases but remains above 0.4 in the early period after massive hepatectomy because glucose is not utilized efficiently in such a condition [7,8]. As insulin inhibits the mobilization of fatty acids from the adipose tissue, it should not be administered when the blood ketone body rati o is below 0.4 after partial hepatectomy. When hyperglyCemia develops after major hepatectomy, the blood glucose level should be corrected by decreasing the glucose soluti0n, not by administration of insulin. Bivins et al [27] recommends administering a a minimal dose of insulin with half saline to correct the hyperglycemia. If these steps are not taken, severe hyperglycemia will result in hyperglycemic hyperosmotar nonketotic dehydration. Summary
The ability of the liver to control blood glucose levels was studied in relation to the blood ketone body ratio, which reflects the hepatic mitochondrial redox state. In partially hepatectomized rabbits, postoperative continuous administration of a hyperosmolar glucose solution revealed the following findings: (1) The occurrence Of glucose intolerance during postoperative hypertonic glucose administration was more dependent on the severity of the reduced hepatic mitochondrial redox state, as reflected by the blood ketone body ratio, than on the size of the liver remnant. (2) Insulin resistance developed when the redox state of the liver mitochondria was severely reduced. (3) It was very difficult to maintain the blood glucose level within normal range when the redox state was severely reduced. In patients, insufficient glucose administration can
The American Journal of Surgery
Hyperglycemic Hyperosmolar Nonketotic Dehydration
easily result in hypoglycemia, and excess administration can result in hyperglycemia, which might, in turn, result in hyperglycemic hyperosmolar nonketotic dehydration. References 1. McDermott WV, Ottinger LW. Elective hepatic resection. Am J Surg 1966; 112: 376-81. 2. McDermott WV, Weber AL. Major hepatic resection: diagnostic techniques and metabolic problems. Surgery 1963; 54: 56-66. 3. Ochsner JL, Meyers BE, Ochsner A. Hepatic Iobectomy. Am J Surg 1971; 121: 273-82. 4. Pinkerton JA, Sawyers JL, Foster JH. A study of the postoperative course after hepatic Iobectomy. Ann Surg 1971; 173: 800-11. 5. Ida T, Ozawa K, Honjo I. Glucose intolerance after massive liver resection in man and other mammals. Am J Surg 1975; 129: 523-7. 6. Ozawa K, Ida T, Yamada T, Honjo I. Significance of glucose tolerance as prognostic sign in hepatectomized patients. AmJ Surg 1976; 131: 541-6. 7. Nakatani T, Ozawa K, Asano M, Ukikusa M, Kamiyama Y, Tobe T. Changes in predominant energy substrata after hepatectomy. Life Sci 198i; 28: 257-64. 8. Nakatani T, Ozawa K, Asano M, Ukikusa M, Kamiyama Y, Tobe T. Differences in predominant energy substrata in relation to the resected hepatic mass in the phase immediately after hepatectomy. J Lab Clin Mad 1981; 97: 887-98. 9. Nakatani T, Yasuda K, Ozawa K, Kawashima S, Tobe T. Effects of (+)-octanoylcarnitine on deoxyribonucleic acid synthesis in regenerating rabbit liver. Clin Sci 1982; 62: 295-7. 10. Ukikusa M, Ozawa K, Shimahara Y, Asano M, Nakatani T, Tobe T. Changes in blood ketone body ratio. Their significance after major hepatic resection. Arch Surg 1981; 116: 781-5. 11. Mellanby J, Williamson DH. Acatoacetate. In: Bergmeyer HU, ed. Methods of enzymatic analysis. New York: Academic Press, 1974: 1840-3. 12. Williamson DH, Mellanby J. D-(-)-3-hydroxybutyrate. In: Bergmeyer HU, ed. Methods of enzymatic analysis. New York: Academic Press, 1974: 1836-9.
Volume 155, April 1988
13. Hultman E. Rapid specific method for determination of aldosaccharide in body fluids. Nature 1963; 183: 108-9. 14. Malloy HT, Evelyn KA. The determination of bilirubin with the photoelectric colorimeter. J Biol Chem 1937; 119: 481-90. 15. Wallenstein S, Zucker CL, Fleiss J. Some statistical methods useful in circulation research. Circ Shock 1981; 8: 503-17. 16. Yamamoto M, Tanaka J, Ozawa K, Tobe T. Significance of acetoacetate/fl-hydroxybutyrate ratio in arterial blood as an indicator of the severity of hemorrhagic shock. J Surg Res 1980; 29: 124-31. 17. Tanaka J, Ozawa K, Tobe T. Significance of blood ketone body ratio as an indicator of hepatic cellular energy status in jaundiced rabbits. Gastroenterology 1979; 76: 691-6. 18. Ozawa K, Fujimoto T, Nakatani T, Asano M, Aoyama H, Tobe T. Changes in hepatic energy charge, blood ketone body ratio and indocyanine green clearance in relation to DNA synthesis after hepatectomy. Life Sci 1982; 31: 647-53. 19. Ozawa K, Aoyama H, Yasuda K, et al. Metabolic abnormalities associated with postoperative organ failure. A redox theory. Arch Surg 1983; 118: 1245-51. 20. Ketone body ratio: an index of multiple organ failure? Lancet 1984; 1: 25-6. 21. Ozawa K, Kamiyama Y, Kimura K, et al. Clinical experience of postoperative hepatic failure treatment with pig or baboon liver cross-hemodialysis with an interposed membrane. Artif Organs 1982; 6: 433-46. 22. Tanaka A, Noguchi M, Taki Y, et al. The influence of hemodilution in normal and hepatectomized rats in relation to hepatic energy metabolism. Am J Mad Sci 1987; 293: 354-60. 23. Kahn D, Van Hoorn-Hickman R, Terblanche J. Liver blood flow after partial hepateotomy in the pig. J Surg Res 1984; 37: 290-4. 24. Irie R, Kohno Y, Aoyama H, et al. Impaired glucose tolerance related to changes in the energy metabolism of the remnant li,er after major hepatic resection. J Lab Clin Med 1983; 101: 692-8. 25. Williamson DH, Lund P, Krebs HA. The redox state of free nicotinamide-adenine dinucleotide in the cytoplasm and mitochondria of rat liver. Biochem J 1967; 103: 514-27. 26. Lehninger AL. The tricarboxylic acid cycle and the phosphogluconate pathway. In: Lehninger AL, ed. Biochemistry. New York: Worth, 1977: 444-76. 27. Bivins BA, Hyde GL, Sachatello CR, Griffen WO. Physiopathology and management of hyperosmolar hyperglycemic nonketotic dehydration. Surg Gynecol Obstet 1982; 154: 534.
563