Symposium on Hepatic Surgery
Preoperative and Postoperative Care
H. Harlan Stone, M.D.t.'
The liver is the single organ for which medical science has yet to develop some method of artificial substitution. Despite even the most modern and intensive supportive measures, an anhepatic state can be tolerated for only a few hours. This is because the liver is the metabolic center of the body. It is here that nutrients are assimilated, modified, and then stored until needed; waste products and foreign substances are detoxified, catabolized, and/or excreted; and finally, components of certain critical systems, such as the blood coagulation mechanism, are maintained almost in their entirety. Indeed, these complex functions themselves are so poorly understood that any attempt at the extracorporeal assumption of their many responsibilities has to date been absolutely impossible without the incorporation of either homologous or heterologous liver. Nevertheless, there are certain basic functions which can be. measured with respect to level of present capacity and which, in turn, can be temporarily supported by the administration of specific nutrients and/or exogenously produced active compounds.
PREOPERATIVE EVALUATION There are a large number of so-called "liver function tests." Most either quantitate the amount of some enzyme circulating in the peripheral blood, thereby taking this tangential measure of the degree of liver injury to be a valid estimate of the functional hepatic capacity lost, or record some change in a physical or chemical property that has more in keeping with ancient alchemy than with present-day biochemistry and physiology. There are, however, four laboratory tests that have proven to be of considerable value and do reflect with moderate accuracy the overall state of liver function at any given point in time. 3 • 13 These are the blood glucose, the blood ammonia, the serum bilirubin, and serum albumin. Elevations in ammonia and bilirubin indicate significant impairments in catabolism and subsequent excretion of two specific waste products, while hypoglycemia and hypoalbuminemia demonstrate an inability to *Professor of Surgery, Emory University School of Medicine, Atlanta, Georgia
Surgical Clinics of North America-Vol. 57, No.2, April 1977
H. HARLAN STONE
maintain, assimilate, and/or manufacture certain critical substances from the diet taken. Tolerance of major surgery and actual clinical course, irrespective of the operation performed, appears to correlate quite well with these four tests of liver function. 3 Liver histology as noted through biopsy, on the other hand, is seldom reliable. 12 Granted, active hepatitis (whether alcoholic or viral) is uniformly associated with a poor prognosis and has therefore become an almost absolute contraindication to surgery. Exclusive of this single microscopic finding, predictability of outcome and guidance in supportive therapy must always be based upon the four stated l"aboratory tests and the patient's clinical course to date, i.e., presence and intractability of ascites, neurologic disorders of ammonia intoxication, and state of nutrition.
Clinical Assessment Based upon these few laboratory and clinical features, Child has proposed a classification for patients who potentially have sustained some compromise in functional liver reserve (Table 1).3 Although originally designed to select out those cirrhotic patients who would not be good candidates for surgical decompression of the portal venous system, the method can be applied with equal usefulness to all subjects considered for any major operation as well as for those specific individuals for whom liver resection is planned (Table 2).3.10 In Class A, patients have no detectable stigmata of significant liver disease, have normal hepatic reserve, require no special preoperative measures, and can tolerate all forms of major surgery. Regardless of prior disease states, these individuals are normal for all intents and purposes. In Class B, there are definite limitations to the functional reserve of the liver. Only emergency, life-saving operations, should be performed without prior preparation by dietary supplements, ascites control, and occasionally even infusions of salt-poor albumin and fresh frozen plasma. Portosystemic shunting should be considered only if there is splanchnic decompression without diversion of portal flow away from the liver (the Warren distal splenorenal shunt).4 Hepatic resections of greater than 10 to 15 per cent of liver substance are always contraindicated. Table 1. Clinical and Laboratory Classification of Hepatic Function':' CLASS
Functional impainnent Serum bilirubin (mg per cent) Serum albumin (gm per cent) Ascites Neurologic disorders Nutrition Operative mortality
Minimal <2.0 >3.5 None None Excellent <1 per cent
Moderate 2.0 to 3.0 3.0 to 3.5 Easily controlled Minimal Good 10 per cent
Severe >3.0 <3.0 Poorly controlled Moderate to severe Poor, wasted >50 percent
*Adapted from Tables 1-4 and 1-5 in Child, C. G., III: The Liver and Portal Hypertension. Philadelphia, W. B. Saunders Co., 1967.
PREOPERATIVE AND POSTOPERATIVE CARE
Table 2. Basic Liver Functions and Operability Class A
No limitations. Normal response to all operations. Normal ability of liver to regenerate.
Some limitation to liver function. Altered response to all operations, but tolerated well if prepared preoperatively. Limited ability to regenerate new hepatic parenchyma so that all sizeable liver resections are contraindicated.
Severe limitations to liver function. Poor response to all operations regardless of preparatory efforts. Liver resection, no matter what the size, is contraindicated.
In Class C, there is minimal to almost no liver reserve. Such patients are continually in and out ofliver failure, although an occasional patient may appear to move up from a "Class C" to a "Class B" with intensive medical management. Nevertheless, functional liver capacity is so severely restricted that essentially all operations are contraindicated, that is, except for emergency procedures demanded by acute life-threatening disease or injury.
Parenchymal Bulk As a general rule, Class A patients can be compared to individuals who have been deprived of30per cent or less of their hepatic parenchyma. Class B patients emulate liver resections of 50 to 70 per cent, while Class C patients are similar to subjects who have just undergone a 90 to 95 per cent resection. There is one major difference, however. If liver function has been limited by disease preoperatively, regenerative processes have already been maximally stimulated, and thus little improvement in liver function can be expected at any time during the postoperative phase. By contrast, patients who have had hepatic resection and whose livers are otherwise normal almost uniformly regain their preoperative liver reserve within a few weeks following even a relatively massive ablation. The crux is whether parenchymal regenerative mechanisms are essentially spent or are still basically intact. PREOPERATIVE PREPARATION Normal patients and those with known liver disease, yet with a Class A status, require no special diet or medications prior to operation. 3 Class C patients, as a general rule, are never operative candidates, although they should receive supportive measures in an effort to improve their functional status and thereby to make them eventually operable. Thus, it is the Class B patient who warrants energetic and detailed preparations for surgery.
H. HARLAN STONE
Nutrition Patients with compromised liver function should be placed on a diet as high in protein as ammonia tolerance will permit. 4 This nitrogen load can additionally be increased by the oral administration of some nonabsorbable antimicrobial agent (neomycin) in order to suppress the bacterial population of the colon and thereby to limit the intestional absorption of free ammonia and/or ammonium compounds released by these microbes. Should the serum albumin not be elevated by such measures, then infusions of plasma or salt-poor albumin itself may become necessary. The goal is a serum albumin of at least 3.5 gm per cent. 3 • 13
Fluids and Electrolytes Efforts should always be made to totally eliminate, or at least to control effectively, ascites. 3 • 4. 13 A sodium-poor diet and diuretics specifically preventing obligatory sodium reabsorption (spironolactone with or without hydrochlorothiazide) are usually successful if the regimen is adhered to diligently. However, persisting ascites in the face of an otherwise Class B liver function may indicate the need for a more aggressive approach, such as the insertion of a LeVeen peritoneal-jugular shunt. 6 Repeated paracenteses as independent procedures, even though the fluid is filtered and then reinfused intravenously, should be avoided. Because of the humerally influenced positive sodium balance at the expense of excessive potassium excretion, body depletion of potassium and its consequent alkalosis are common. Renal retention of protons and thus a return of the acid-base balance toward normal can be accomplished only if sufficient potassium ions are available for exchange at the renal tubular level. Accordingly, large quantities of potassium must be given to both replete potassium losses as well as to correct the attendant alkalosis. L 3. 4. 8
Coagulopathies Anyone of a multitude of bleeding disorders can occur in the patient with a severely compromised liver function. However, even after a 90 per cent hepatic resection, a rise in the prothrombin time to levels representing as little as 5 per cent of normal activity is only transient. 10. 11 The value of parenterally administered vitamin K to correct this problem has been consistently overemphasized. Even though the prothrombin time may fall into a more acceptable range, other liver produced moities of the coagulation mechanism are not significantly altered. Probably, the most reliable approach has been to give fresh frozen plasma, both during operation as well as in the immediate postoperative period. By this means, almost the entire complement of noncorpuscular clotting elements can be replenished.
Miscellaneous Other measures directed at improvement in overall liver function may indeed be of great benefit, yet there is no universal agreement at present as to their specific indications. The most controversial of all such medications is that group of substances collectively called corticosteroids.
PREOPERATIVE AND POSTOPERATIVE CARE
It is still moot whether these agents are ever indicated and, if so, exactly
which one of the many available compounds is best, how large should the dose be, and how rapidly should it be given. . Again, it cannot be repeated too frequently that active hepatitis must be identified whenever possible. If any doubt exists, provided that a bleeding tendency is not present, needle biopsy of the liver should always be performed.
INTRAOPERATIVE LIVER INJURY Discounting the physical trauma inflicted by retractors or purposeful cutting into liver substance itself, hepatic injury associated with operation is almost always caused by local tissue hypoxia or the anesthetic agent used. 1 • 5. 7. 8 Hypotension secondary to blood loss or to the administration of some vasodilator and derangements in pulmonary ventilation and/or perfusion account for a majority of the cases of parenchymal hypoxia. 1. 5. 8 Although greater stress is usually placed upon the effects of ischemia on other organs (kidney, brain, etc.), the liver is extremely susceptible to injury caused by oxygen lack. It is, however, the tremendous reserve capacity of the normal liver which masks this relatively common cellular injury. Should the patient have prior significant liver disease, the damage caused by hypoxia is much more overt and may of itself cause irreverisble changes, even death. Thus shock, whether due to hemorrhage or pharmacologic agents, must always be minimized. Anesthetic agents appear to cause damage by either direct-and relatively immediate hepatotoxic effects, i.e., chloroform, or by a drug sensitization which in turn is followed by an idiosyncratic reaction, i.e., halothane.!' 7. 8 Since these problems are covered in greater detail elsewhere, mention is made only for completeness. The liver responds and heals after injury as does any other solid viscus. If the wound is physical, then initially there is nonspecific local inflammation which in tum is followed by classical fibroblastic repair. Diffuse injury, on the other hand, as caused by disease, infection, hypoxia, or drugs generally mimics the nodular hyperplasia noted in early phases of cirrhosis. Both inflammatory and regenerative processes are seen throughout liver substance. After major resections, however, inflammation is primarily confined to the cut surface, while parenchymal regeneration through hyperplasia is relatively generalized. 2
POSTOPERATIVE CARE After operation-regardless as to whether on the liver, portal venous system, or elsewhere in the body-care of the patient with compromised liver function (Class B or C) is based upon those same measures applied during the phase of preoperative preparation. Such include: close attention to nitrogen balance, evidences of ammonia intoxication, and level of serum albumin; sodium and water restriction, spironolactone diuretics,
H. HARLAN STONE
and extra aliquots of potassium, always keeping in mind the potential for development of a secondary hypovolemia; the occasional addition of more energetic measures for ascites control; and the generous administration of fresh frozen plasma to treat or prevent the occurrence of any bleeding disorder. Otherwise, care becomes rout~ne for the individual operative procedure performed.
Care Following Hepatectomy In contrast to other liver problems encountered during the postoperative period, hepatic resections may demand relatively massive replacement therapy; yet the need for such is only temporary, since daily requirements for each supplement gradually diminish as liver adaptation and regeneration take place. 2 • 9. 11 The quantity of specific items to be replenished varies considerably. Generally, replacement is based upon per cent of normally functioning liver still present, irrespective as to whether the loss was due to disease, trauma, or surgical excision. However, only if hepatic parenchyma is free of other disease can regeneration be expected to occur rapidly and relatively completely. The exact amount to be given is then further modified according to the weight of the individual patient and his response to what had been given on the day or days before.1 o. 11 IMMEDIATE PROBLEMS. During operation, resection progressively removes increasing numbers of portal outflow tracts. 10. 11 With resections of less than 30 per cent, significant obstruction and thus major splanchnic pooling is rare. However, hepatectomies excising greater amounts of liver lead to a transient, portal hypertension and sizeable sequestration of functional blood volume into the portal venous system. In reality, it is an acute peripheral hypovolemia as a consequence of portal hypervolemia or, in other words, hemorrhagic shock due to so-called "bleeding" into the splanchnic bed. Thus, despite a greater than normal measurable blood volume, the patient progresses relatively rapidly into an otherwise unexplained hypovolemic shock. 10. 11 The quantity of blood lost to the general circulation is roughly proportionate to the massiveness of liver resection (Fig. 1). Based upon the known or calculated preoperative blood volume, the additional blood transfusions required can be anticipated and accordingly given as needed. Nevertheless, individual patients do vary considerably in their responses, as do estimates made by surgeons as to the per cent of liver removed. 10. 11 It is not an acute portal hypertension that warrants treatment, for such is a transient phenomenon and lasts for only 10 to 20 days, at the most. Shunting, in fact, is contraindicated, for it would severely interfere with the delivery of stimuli for liver regeneration carried in the portal blood. 9 Instead, what must be countered is the associated peripheral hypovolemia. To avoid this shock-like state, the central venous pressure should be used as a guide to fluid repletion.lO It is simple and relatively accurate. Approximately half of the additional volume is required during the first 4
PREOPERATIVE AND POSTOPERATIVE CARE
REPLETION OF BLOOD LOST TO SPLANCHNIC SEQUESTRATION
o w a:: 5
Figure 1. Additional blood transfusion required according to per cent of liver resected and patient body weight.
PERCENT LIVER RESECTED
to 8 hours immediately after ligation of the portal venous inflow to the liver. By the fourth postoperative day, splanchnic sequestration no longer increases and then gradually resolves during the week or weeks that follow .10, 11 A significant and potentially dangerous hypoglycemia can suddenly develop during surgery, especially if the resection is greater than 70 per cent of functional liver substance. 10, 11 Generally speaking, the rapidity of onset and the magnitude of fall in blood glucose are proportionate to the preoperative liver glycogen stores and the per cent of hepatic resection. Lesser hepatectomies may fail to produce an obvious hypoglycemia, particularly if routine postoperative fluids contain glucose in at least a 5 per cent concentration. Prevention of central nervous system injury to a low blood glucose (less than 50 mg per cent) is an important consideration with respect to those patients subjected to the more massive resections. 10, 11 It is not so much the rate at which glucose solutions are given or the concentration of glucose in those solutions as it is the grams of glucose provided each hour. In addition, such supplemental therapy adds considerable solute to the urine and can of itself produce troublesome osmotic problems. Thus, a careful check on the adequacy of carbohydrate administration is definitely a must. Blood glucose determinations at hourly intervals are impractical. 10 However, the urine can be tested for glucose spill even more frequently, if necessary. As long as the reading is "a trace" or "1 plus," sufficient exogenous glucose has been given. If the urine sugar test is negative, then an increase either in infusion rate or in concentration of glucose contained must be effected.
H. HARLAN STONE
Continuation of intravenous glucose supplementation may be required for as long as 3 weeks, although oral alimentation will usually provide sufficient free carbohydrate by the tenth postoperative day (Fig. 2).
An intraoperative bleeding tendency rarely occurs unless the quantity of transfusions has been excessive or fresh frozen plasma has been omitted from the treatment regimen. The same is true for the subsequent postoperative course. If, however, bleeding is later noted despite such preventive measures, then reoperation should be considered in an attempt either to ligate the responsible vessel or to remove the ischemic segment of liver initiating the release of fibrinolysins. 1o , 11 The value of antibiotics in preventing wound and intra-abdominal sepsis in these patients has yet to be proven, even though the majority of experienced liver surgeons do recommend beginning one of the cephalosporins prior to the induction of anesthesia. Later Postoperative Problems As mentioned above, glucose infusions are generally required until the second or third postoperative week following major liver resection.10, 11 The oral intake should at first be rich in freely-hydrolyzable carbohydrate, to which the protein content is gradually increased until a high protein diet has been reached. By the end of the third or fourth week, no special nutrients beyond a high protein diet should be required. It must be remembered that the liver is the only endogenous source for human albumin, and thus any major hepatic resection will create a significant hypoalbuminemia. 10, 11 Since the half-life of albumin varies from 8 to 24 hours, depending on the present metabolic state and the influence of protein-sparing carbohydrate, the onset of hypoalbuminemia can be predicted with certainty to appear on the third or fourth postoperative day, or even sooner if the specimen excised represents a mass greater than 70 per cent of original functioning liver. Untreated, hypoalbuminemia leads to ascites, pulmonary congestion, peripheral edema, generalized anasarca, and finally death. Once the full-blown syndrome has developed,
I-~ zw u
DAYS FOLLOWING LIVER RESECTION
Figure 2. Concentration of glucose and duration of infusion according to massiveness of liver resected.
PREOPERATIVE AND POSTOPERATIVE CARE
Preoperative Serum Albumin Concentration 2gm% 3gm% 4gm% 5gm%
80 20 40 60 100 PERCENT LIVER RESECTED Figure 3. First day albumin requirements as based on initial level of serum albumin and per cent of liver substance resected.
even massive infusions of albumin consistently fail to mobilize the extra fluid deposited throughout the body. Death is then almost assured by the time that clinical anasarca has been reached. Thus, it becomes imperative to prevent the development of significant hypoalbuminemia (serum albumin less than 3 gm per cent) and its associated problemsY Salt-poor albumin must be given intravenously as a regular daily aliquot, the amount being determined by the per cent of normal liver remaining, the prior day level of serum albumin, and the patient's body weight (Fig. 3). Formulas are also available to aid in making these calculations. Once therapy has begun, additional supplements may be required to raise the serum albumin concentration, not just to maintain it (Fig. 4). This state of affairs then persists for one to several weeks postoperatively.lo.11 How long appears to be a function of both how much liver was removed as well as how much albumin per given
DESIRED INCREASE IN SERUM ALBUMIN (gm%) Figure 4. gm per cent.
Amount of albumin required to increase a given level of serum albumin up to 5
H. HARLAN STONE
ANTICIPATED DATE OF REGAINING TOTAL CAPACITY FOR AUTOGENOUS SERUM ALBUMIN MAINTENANCE
@ 80 a: z ~ 60
Accordmg loAbsolule First A:Jstoperafive Day Albumin Requirement
(gm/kg body weight)
o Figure 5.
DAYS FOLLOWING LIVER RESECTION Predictability of the need to continue albumin supplements.
mass of body weight was needed during the first few postoperative days (Fig. 5). Like so many other complications, the consequences of hypoalbuminemia are predictable. Once fully established, however, these same problems may become irreversible and in themselves assuredly lethal despite all efforts to the contrary.1O, 11 Regeneration of a normally functioning liver can uniformly be expected, regardless of how massive the resection (i.e., 90 to 95 per cent), provided that the retained liver parenchyma is viable, has an unobstructed duct and blood supply, and was not diseased itself.1° Thus, supportive care is indeed worthwhile until the residual piece ofliver can alone take over two of its most important tasks-maintenance of both blood glucose and serum albumin. Jaundice is usually low-grade and lasts for only 5 to 10 days. 11 Ifmore profound or if persisting into the third week, obstruction of the duct system should be suspected. Other parameters of liver function have seemed to be of little importance and of absolutely no use whatsoever. Nevertheless, monthly liver scans will document an actual increase in liver size which begins as passive congestion, progresses to hypertrophy, and then finally reflects a true cellular hyperplasia. 2 , 10
REFERENCES J. A.: Anesthesia and intraoperative care. In Starzl, T. E., and Putnam, C. W. (eds.): Experience in Hepatic Transplantation. Philadelphia, W. B. Saunders Co., 1969, pp.83-92. 2. Bucher, N. L. R.: Experimental aspects of hepatic regeneration. N. Eng. J. Med., 277:686, 738,1967. 1. Aldrete,
PREOPERATIVE AND POSTOPERATIVE CARE
3. Child, C. G., III: The Liver and Portal Hypertension. Philadelphia, W. B. Saunders Co., 1967, pp. 48-77. 4. Galambos, J. T., Warren, W. D., Rudman, D., et al.: Selective and total shunts in the treatment of bleeding varices. N. Eng. J. Med., 295:1089, 1976. 5. Gelman, S. 1.: Disturbances in hepatic blood flow during anesthesia and surgery. Arch. Surg., 111 :881,1976. 6. LeVeen, H. H., Wapnick, S., Grosberg, S., and Kinney, M. J.: Further experience with peritoneo-venous shunt for ascites. Ann. Surg., 184:574, 1976. 7. Little, D. C., and Wetstone, H. J.: Anesthesia and the liver. Anesthesiology, 25 :815,1964. 8. Loughridge, L.: Anesthesia and the liver. In Scurr, C., and Feldman, S. (eds.): Scientific Foundations of Anesthesia. Chicago, Year Book Medical Publishers, Inc., 1974, pp. 325-329. 9. Starzl, T. E., Francavilla, A., Halgrimson, C. G., et al.: The origin, hormonal nature, and action of hepatotrophic substances in portal venous blood. Surg. Gynec. Obstet., 137:179,1973. 10. Stone, H. H.: Major hepatic resections in children. J. Pediat. Surg., 10:127,1975. 11. Stone, H. H., Long, W. D., Smith, R. B., III, and Haynes, C. D.: Physiologic considerations in major hepatic resections. Am. J. Surg., 117:78, 1969. 12. Sundennan, F. W., and Sundennan, F. W., Jr.: Laboratory Diagnosis of Liver Disease. St. Louis, W. H. Green, 1968, p. 327. 13. Wirthlin, L. S., Urk, H. V., Malt, R. B., and Malt, R. A.: Predictors of surgical mortality in patients with cirrhosis and nonvariceal gastroduodenal bleeding. Surg. Gynec. Obstet., 139:65, 1974.
Emory University School of Medicine 69 Butler Street Atlanta, Georgia 30303