Current Strategies in Surgical Nutrition
0039--6109/91 $0.00
+ .20
Nutritional Support of Patients with Cancer of the Gastrointestinal Tract
John M. Daly, MD, * H.P. Redmond, BSc, FRCSI, t Michael D. Lieberman, MD,+ and Lori Jardines, MD§
Among hospitalized patients, those with cancer have the highest incidence of protein-calorie malnutrition. Significant malnutrition occurs in more than 50% of cancer patients undergoing major operative procedures. Clinically significant malnutrition is a result of diminished nutrient intake, increased nutrient losses, and tumor-induced derangements in host metabolism. In the absence of adequate exogenous nutrients, the body utilizes endogenous substrates to satisfY the ongoing requirements of both host and tumor for energy and protein. In patients with malignant obstruction of the gastrointestinal tract, the tumor itself may induce diminished nutrient intake. Present-day treatment modalities, including gastrointestinal resection, chemotherapy, and radiotherapy, compound these metabolic derangements, which further increase the risk of Significant postoperative morbidity and death. The presence of malnutrition in cancer patients has major prognostic significance. In a review of more than 3000 cancer patients, DeWys and colleagues found better survival rates in those patients without weight loss compared with patients who had lost just 6% of their body weight. 13 Numerous other studies have noted increased postoperative morbidity and mortality rates associated with malnutrition. The development and refinement of enteral and parenteral nutrition support techniques have provided the opportunity for studying the relation between nutritional supplementation and postoperative prognosis. Therefore, because cancer patients are often malnourished and because malnutrition is associated with increased postoperative morbidity, nutritional support is often instituted to improve From the Division of Surgical Oncology, Department of Surgery, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania *Jonathan E. Rhoads Professor of Surgery, and Chief, Division of Surgical Oncology tResearch Fellow :j:American Cancer Society Fellow in Surgical Oncology §Assistant Professor of Surgery
Surgical Clinics of North America-Vol. 71, No.3, June 1991
523
524
JOHN
M.
DALY ET AL.
nutritional status and potentially reduce the risks of postoperative complications. This article addresses the role of nutritional therapy in patients with cancer of the gastrointestinal tract.
NUTRITIONAL STATUS AND CANCER The term cancer cachexia encompasses a varied group of symptoms and signs including inanition, anorexia, weakness, organ dysfunction, and eventually death. Although this cachexia is clearly related to the presence of malignancy, there is no consistent relationship between the development of cachexia and the stage of the disease, tumor histology, primary site, or duration of illness. The cause of cancer cachexia is thought to be multifactorial (Table 1). Anorexia Anorexia associated with cancer cachexia develops for a number of reasons. First, local effects of tumor that diminish nutrient intake are often involved, particularly when the tumor arises from or impinges on the gastrointestinal tract. Tumors arising in the pharynx and the esophagus can cause various degrees of dysphagia secondary to partial or complete obstruction. Gastric neoplasms may cause early satiety or outlet obstruction, and enteric or colonic obstruction can result in intestinal obstruction, blind loop syndrome, and fistulas. Second, alterations in taste perception and appetite contribute to diminished nutrient intake. Patients often have an increased threshold for sweet taste and a concomitant decrease in sour and salty taste thresholds. Deficiencies in trace elements such as zinc have been implicated in these changes in some instances. Third, patients with extensive hepatic involvement by tumor may develop nausea and anorexia secondary to high lactic acid levels produced by anaerobic tumor metabolism. Fourth, neuroendocrine control of appetite is centered in the lateral hypothalamus, and a number of factors may alter afferent and efferent stimuli to and from this area. Tryptophan, a precursor of the neurotransmitter serotonin, has been shown to be elevated in brain tissue of tumor-bearing animals, and alterations in serotonin levels are associated with anorexia. Finally, chemotherapeutic agents can act centrally, in the chemoreceptor trigger zone, or locally on the gastrointestinal tract, inducing nausea and vomiting or stomatitis, mucositis, glossitis, and pharyngitis. Table 1. Factors Related to Cancer Cachexia Anorexia Local tumor effects Altered taste perception Elevated lactic acid levels Neuroendocrine factors Chemotherapeutic agents Abnormal carbohydrate metabolism Abnormal lipid metabolism Abnormal protein metabolism
SUPPORT IN GASTROINTESTINAL CANCER
525
The extent of the malnourished state associated with malignant disease is often greater than expected from the decrease in dietary intake. Furthermore, the degree and type of malnutrition vary in the cancer patient. Several studies have demonstrated that the presence of tumor appears to affect the resting energy expenditure (REE). Although it appears that some cancer patients have altered resting and total energy expenditure, those factors that determine elevation or depression of energy expenditure remain unclear. Dempsey et al studied 173 patients with gastrointestinal malignancies, noting that gastric cancer patients were more likely to be hypermetabolic than were patients with pancreatic and hepatobiliary tumors. 12 Curative resection results in a return to normal levels of REE, but patients who undergo palliative resection fail to achieve this normalcy. Carbohydrates Abnormalities in host carbohydrate metabolism are common in cancer patients. Peripheral utilization of glucose, hepatic gluconeogenesis, and consequently glucose turnover have been demonstrated to be abnormal. 10, 22, 23, 34 Altered glucose tolerance was recognized many years ago and was suggested as early as 1919 to be a possible diagnostic aid in patients with gastrointestinal symptoms. 47 Glucose intolerance has been correlated with insulin resistance 32, 49 and may involve an attenuated insulin secretory response. 20 Recent studies in sarcoma-bearing rats suggest that supplemental insulin administration may preserve host lean body mass and even influence survival. 37, 45 Cancer patients appear to be glucose intolerant and also have been found to synthesize glucose at an elevated rate. Recently, Shaw and Wolfe 50 noted that patients with gastrointestinal tumors had elevated rates of basal hepatic glucose production, They noted a direct relation between the extent of the tumor burden and the increase in gluconeogenesis. Furthermore, patients with the largest tumor burdens failed to suppress their own endogenous glucose production during glucose infusions. In studies on patients with sarcomas, those workers also demonstrated that administration of glucose did not suppress endogenous protein catabolism. 51 The reason for the elevated rates of hepatic gluconeogenesis is unclear. Increased levels of appropriate substrates from tissue breakdown in the periphery may fuel a substrate-driven process. Increased gluconeogenesis occurs in cachectic patients who are challenged with the substrates alanine, glycerol, or lactate. 33,56,57 Plasma concentrations oflactate are often elevated in cancer patients. At least part of the explanation resides with the host hepatocytes themselves, as hepatocytes isolated from glycogen-depleted tumor-bearing rats not only produce more glucose from available substrate than do hepatocytes from similar nontumor-bearing controls but also respond by producing more glucose when challenged with alanine and lactate. 46 Such increases in the hepatic capacity to synthesize glucose have also been demonstrated by Gutman and associates 17 as well as by Hammond and Balinsky.18 These findings are consistent with radioisotope tracer studies demonstrating increases in glucose recycling. 16, 33 The Cori cycle is a metabolically futile cycle in which glycolytic production of lactate serves as substrate for
526
JOHN
M.
DALY ET AL.
hepatocyte glucose production. Because the energy required to synthesize glucose outweighs the energy created during glycolysis, increased use of this pathway by an organism can lead to significant waste of energy. Holroyde et al found levels of Cori cycle activity to be increased in cancer patients compared with starved controls.23, 24 Burt and Brennan also demonstrated increased rates of glucose turnover in 12 patients with localized esophageal carcinoma. 7 Gluconeogenic activity is often inappropriately high in cancer patients compared with non tumor-bearing controls. Eden et al studied eight cancer patients and seven control patients after a 14-day period of enteral nutrition. 15 In both the basal and the re-fed states, cancer patients had increased levels of glucose turnover and recycling. The implications for energy loss in such a situation are substantial. The increased glucose flux observed amounted to 42% of the daily glucose intake in the cancer patient population, or between 250 and 350 kcal per day.
Lipids Alterations in lipid metabolism in cancer patients include changes in body fat composition and increased lipid mobilization. Decreases in total body fat are common in these patients and may be related to insulin deficiency. Decreased levels oflipoprotein lipase correlate with the degree of weight loss, Increased lipolysis allows glycerol and fatty acids to be utilized for gluconeogenesis and ~mergy production. Fatty acids have been implicated as the main substrates in some patients with neoplastic disease, and these patients fail to suppress lipolysis after glucose administration.
Protein Although abnormal carbohydrate cycling has been implicated as the central mechanism for host nutritional depletion, abnormalities in protein metabolism occur and are interrelated. Intuitively, severe wasting of host muscle mass along with depletion of visceral and circulating proteins (especially albumin) occurs secondary to a combination of increased protein breakdown and decreased protein synthesis. Yet studies in both animals and humans suggest that a more complex derangement of protein metabolism is induced in the tumor-bearing host. Both tumor-bearing animals and persons with cancer seem to be unable to respond appropriately to diminished nutrient supplies. 58 However, Kawamura et al found that sarcoma-bearing rats incorporated radiolabeled tyrosine into total body protein at a rate 33% faster than controls. 28 At the same time, incorporation of tyrosine into skeletal muscle (gastrocnemius and rectus abdominis) was decreased from 6% to 10%. The reconciliation of increased total protein flux in the face of decreased skeletal muscle synthesis may come from increases in hepatic fractional protein synthesis. In studies comparing sarcoma-bearing rats with nontumor-bearing pair-fed controls, Norton et al43 demonstrated that despite equal nitrogen balance between the groups, the tumor-bearing animals had increased total protein flux and diminished muscle protein synthesis but increased hepatic protein synthesis. Protein synthesis within the tumor did not fluctuate with nutritional state.
SUPPORT IN GASTROINTESTINAL CANCER
527
In humans, a similar pattern of deranged protein metabolism has been documented. Norton et al42 noted that three of seven cancer patients had elevated levels of labeled glycine incorporated into protein. In the face of decreased nutrient intake, even the normal levels observed in the other four cancer patients become suspect. Heber and associates found increased levels of whole-body protein turnover and increased excretion of 3-methylhistidine (evidence of muscle breakdown) in a group of 12 patients with nonsmall-cell lung cancer.21 Jeevanandam et al,26 in a study comparing cancer patients with non tumor-bearing malnourished patients and healthy controls, found markedly increased levels of protein turnover in the cancer group. Evidence of decreased skeletal muscle protein synthesis and increased hepatic protein synthesis is consistent in both animals and humans. Warren et al 54 used isolated hepatocytes from sarcoma-bearing rats to demonstrate enhanced synthesis of both secretory and fixed cellular proteins compared with pair-fed controls. Furthermore, these increases were proportional'to tumor load. This group of researchers postulated an increase in hepatic protein degradation to explain their finding of equivalent protein levels in tumor-bearing and non tumor-bearing livers. Karlberg et aP7 have examined albumin production in a similar model and found it to be elevated. However, increases in the rate of albumin degradation and subsequent expansion of the intravascular volume result in hypoalbuminemia. Concomitant increases in muscle protein degradation may playa role in muscle wasting (and overall protein flux), as evidenced by increased levels of urinary 3-methylhistidine in studies of both animals and humans with tumor. Emery et aP6 documented an 80% increase in the intramuscular level of free 3-methylhistidine in mice bearing tumors. Thus, in the tumorbearing organisms, levels of protein flux depend on the relative rates of synthesis and degradation in both skeletal muscle and liver. Theologides52 has proposed that a tumor-secreted peptide is responsible for the changes in intermediary metabolism that occur in the tumor-bearing host. However, it appears unlikely that tumors themselves are universally capable of inducing the array of metabolic abnormalities described above. Although in isolated instances tumors elaborate products with anorexigenic effects (such as carcinoids [which elaborate serotonin] and small-cell lung carcinomas [which produce bombesin]), no tumor-derived product with a systemic effect has yet been identified. 31, 38 Recently, studies have evaluated the host-produced factors associated with the endogenous immune response to a tumor. These factors, or cytokines, are produced by immunocytes as a paracrine communication and are integral to a coordinated immune response. Interleukins 1, 2, 4, and 6; interferons alpha, beta, and gamma; and lymphotoxin all have powerful effects on cells of the immune system. One such macrophage-elaborated cytokine, tumor necrosis factor-alpha (TNF) or cachectin, appears to exert its effects beyond the immune system and produces many of the systemic effects seen in cachexia. Tumor necrosis factor was first defined as the primary antitumor agent found in the serum of BCG-primed, lipopolysaccharide-treated mice and has since been found to be cytotoxic or cytostatic to a wide variety of
528
JOHN
M.
DALY ET AL.
tumors in vivo and in vitro. 2 , 9 Interest in TNF as a possible mediator of cancer cachexia grew from observations that trypanosomiasis in rabbits was accompanied by severe tissue wasting and hypertriglyceridemia. Beutler and Cerami attributed these effects to TNF after demonstrating that TNF was capable of inducing hypertriglyceridemia and carcass fat loss in mice by inhibiting transcription of the gene for lipoprotein lipase. 5 As reduced levels of lipoprotein lipase activity have been identified in the plasma of cancer patients, and as increased levels of TNF have been found both in the serum of cancer patients and as a product of monocytes isolated from cancer patients, investigation of TNF as a mediator of cancer cachexia has intensified. 1, 3 Receptors for TNF have been found on metabolically important liver, adipose, and muscle cells. In vitro studies of cultured cells from these tissues have demonstrated mixed results depending on isolation techniques, length of exposure to TNF, and other factors. In both animals and humans, TNF has been administered to normal and noncachetic tumor-bearing controls. In animal studies, rats given sublethal doses of TNF have demonstrated increased liver amino acid uptake, glucose turnover, and levels of both acute-phase reactants and stress hormones. Concomitantly, recipient animals demonstrate diminished appetite with little change in anabolic hormone levels. When antibody to TNF has been administered to animals given endotoxin or live Escherichia coli, a protective effect has been noted, ranging from abrogation of the stress response (smaller doses) to survival after previously defined lethal challenge. 6, 35, 53 In humans, Warren et al55 have measured the acute response to TNF in healthy volunteers. The changes included a rise in C-reactive protein, a decrease in serum zinc levels, and a doubling of the forearm amino acid flux. This last effect was primarily attributable to increases in the gluconeogenic amino acids alanine and glutamine. Michie et al36 examined the relation of endotoxin to TNF in healthy volunteers. They found a brief pulse of TNF in the subjects' plasma after bolus endotoxin administration, followed by the characteristic increases in temperature, heart rate, ACTH, and epinephrine. Significantly, administration of ibuprofen blunted these responses while having no effect on TNF levels. Emerging parallels between the metabolic changes found in cancer cachexia and after the administration of cytokines led Kern and Norton to speculate about the mechanisms for host cachexia. 29 They proposed that cytokines elaborated by activated immunocytes in response to tumor have secondary effects on host organs. In the short term, these effects would promote an acute-phase response, rerouting nutrients from the periphery to the liver. However, over the long term, such teleologic strategies backfire, resulting in anorexia and widespread abnormalities in carbohydrate, protein, and lipid metabolism. CONSEQUENCES OF MALNUTRITION IN THE CANCER PATIENT Significant perioperative risk factors in cancer patients include age, obesity, chronic illness, and malnutrition. The consequences of malnutrition for the tumor-bearing host have been well documented.
SUPPORT IN GASTROINTESTINAL CANCER
529
Table 2. The Nutritionally High-Risk Patient Gross underweight: weight-for-height below 80% of standard Gross overweight: weight-for-height above 120% of standard Recent loss of 12% or more of usual body weight Alcoholism No oral intake for more than 10 days on simple intravenous solutions Protracted nutrient losses Malabsorption syndromes Short-gut syndromes/fistulas Renal dialysis Draining abscesses, wounds Increased metabolic needs Protracted fever and infection Trauma
In human studies, evidence of increased morbidity and mortality rates with depressed immunocompetence exists. Poor wound healing, increased w0und infection rates, and prolonged postoperative ileus and hospital stay have all been linked to poor nutritional status in cancer patients. Preoperative assessment of high-risk patients, in particular evaluating reversible factors such as malnutrition, is essential for optimal outcome (Table 2). Nutritional assessment techniques include a detailed history, noting the degree of weight change as well as dietary and alcohol intake, and a thorough physical examination (Table 3). A global clinical assessment stratifies patients into one of three nutritional categories: normal status, mild to moderate malnutrition, or severe malnutrition. Other measures studied are serum albumin and transferrin levels. Total circulating lymphocyte counts, delayed cutaneous hypersensitivity, grip strength, creatinineheight index, and anthropometry are used to a much lesser extent. A number of authors have combined indices of nutritional status to predict postoperative sequelae. The prognostic nutritional index, described by Buzby et al,8a combines serum albumin and transferrin levels, body fat (skinfold thickness), and delayed cutaneous hypersensitivity skin testing. Pettigrew and Hill45a compared body composition measurements, grip strength, prognostic indices, and clinical judgment. A careful assessment Table 3. Assessment of Nutritional Status History and physical examination Extent of disease Weight loss Diet history Anthropometric analysis Laboratory evaluation Hepatic protein synthesis Liver function tests BUN/creatinine Cholesterol/triglyceride Amylase Electrolytes Glucose Immune function Estimation of energy expenditure
530
JOHN
M.
DALY ET AL.
of medical risk, noting in particular cardiorespiratory disease and preexisting sepsis, as well as nutritional state, was as effective as other currently used indications of risk.
NUTRITIONAL SUPPORT The challenge of nutritional support is to prevent or reverse the cachexia of malignancy. The consequences of malnutrition are severe and include impairment of immune function with increased susceptibility to infection and perhaps tumor growth, increased perioperative morbidity, and ultimately death. When one is attempting to treat malnutrition, the patient with cancer must have a complete nutritional assessment, determination of the optimal route of nutrient administration, and meticulous monitoring of nutritional therapy. Standard methods for assessing nutritional status include a thorough history and physical examination, anthropometric measurements, laboratory evaluation of specific visceral organ function, measurement of immune function, and estimation of energy expenditure. Completion of a thorough history and physical examination is important in terms of documenting the extent of disease, identifYing coexisting illness, and quantifying the rate and degree of weight loss and associated protein-calorie deficits. A loss of more than 10% of the usual body weight correlates with increased morbidity and mortality rates. Anthropometry is a noninvasive method of determining body composition based on measurements of skinfold thickness (total body fat) and midarm muscle circumference (lean body mass). Anthropometrics alone has limited value in diagnosing malnutrition, predicting morbidity, or monitoring the effectiveness of nutritional therapy. Laboratory determination of hepatic protein synthesis (i. e., albumin, thyroxine-binding prealbumin, transferrin), liver function tests, blood urea nitrogen, creatinine, triglycerides, cholesterol, amylase, electrolytes, and glucose levels should be performed on all cancer patients being considered for nutritional support. Immune competence, as determined by delayed-type hypersensitivity, may be measured. Finally, patients with severe nutritional deficits should have periodic measurements of REE and 24-hour nitrogen balance. No single clinical or laboratory test accurately predicts the severity of malnutrition. In fact, clinical judgment is superior or equal to single objective measures in predicting nutritionally associated complications. Multiparameter nutritional indices have been developed to identifY patients at increased risk of nutritionally related morbidity and death. One useful predictive index, the prognostic nutritional index, relates four variables (serum albumin, serum transferrin, triceps skinfold, and delayed-type hypersensitivity) to postoperative morbidity and mortality rates. 8a Once it has been shown that volitional oral intake is inadequate and the patient's need for nutritional maintenance or repletion has been evaluated, the route for nutrient delivery must be determined. The alternatives are enteral or intravenous delivery or both. Enteral and parenteral techniques are equally efficacious in meeting nutritional requirements. Enteral feeding is the preferred method of nutritional support in
531
SUPPORT IN GASTROINTESTINAL CANCER
Table 4. Contraindications to Enteral Nutritional Support Absolute Complete mechanical intestinal obstruction Severe diarrhea High-output external fistula (>500 ml/day) Severe acute pancreatitis Shock Relative Immediate postoperative or poststress period «12 hour) Acute severe enteritis Complete absence of small intestine
terms of preserving gastrointestinal architecture and preventing bacterial translocation from the gut. The specific contraindications for enteral feeding are listed in Table 4. Enteral feedings may be delivered by a variety of techniques. It is preferable to administer nutrients distal to the ligament of Treitz to avoid tHe complication of aspiration pneumonia and gastric ileus. This may be accomplished by placement of a small (S-F) silicone rubber nasoduodenal tube, whose position is always confirmed radiographically, or by surgical construction of a feeding jejunostomy. If it is anticipated that the duration of enteral support will be longer than 4 weeks, a jejunostomy is performed. Gastric feeding is acceptable via nasogastric tube or gastrostomy in an alert patient with an intact gag mechanism who has a patent gastrointestinal tract. The gastrostomy may be placed percutaneously with the aid of endoscopy or by an operative Stamm procedure. Complications related to enteral therapy include abdominal wound infection, bowel obstruction, tube dislodgment with peritonitis, aspiration pneumonia, gastrointestinal intolerance, and metabolic abnormalities. In those patients unsuitable for enteral support, parenteral nutrition remains an important option. These solutions are hypertonic and therefore require catheterization of the central venous system to reduce the incidence of phlebitis and venous thrombosis. Acute therapy is usually provided by percutaneous catheterization of the subclavian or internal jugular vein. Chronic support is administered through a subcutaneous catheter (Hickman), which lies in the superior vena cava. Complications of parenteral therapy include pneumothorax, venous thrombosis, arterial trauma, sepsis, electrolyte disorders, and metabolic abnormalities such as glucose intolerance and nonketotic hyperosmolar coma. Caloric and nitrogen requirements must be calculated for each patient. The nonprotein caloric requirement is based on measurement of REE by indirect calorimetry or the Harris-Benedict equation: Male: Basal energy expenditure (kcal) = 66 + 13.7 + 5 X Height (cm) - 6.S X Age (yrs) Female: Basal energy expenditure (kcal) = 65 + 9.6 + 1. 7 X Height (cm) - 4.7 X Age (yrs)
Weight (kg)
X
X
Weight (kg)
In general, 130% to 150% of the REE is provided as nonprotein calories.
532
JOHN
M.
DALY ET AL.
These nonprotein calories include carbohydrate and lipid in an approximate ratio of 2:1 to 1:1, respectively. The administration of protein is based on a determination of nitrogen balance. Typically, 1.5 to 2.5 gm/kg of protein per day is provided, with the higher value reserved for severely depleted or septic patients. Close monitoring of the patient's progress is crucial in adjusting dietary formulation and requirements.
EFFECTS OF NUTRITIONAL SUPPORT ON HOST, TUMOR, AND ONCOLOGIC THERAPY The merit of nutritional support in the cancer patient must be evaluated in terms of changes in host nutritional status, tumor biology, and efficacy of oncologic therapy. The results of detailed studies support the idea that the success of nutritional therapy in terms of repleting the host and reversing metabolic abnormalities is dependent on tumor type as well as the cause and degree of malnutrition. When the malnutrition is attributed to diminished food intake, then nutritional intervention is quite effective. However, malnutrition associated with metabolic abnormalities or advanced malignancy is much more difficult to treat. In a study of patients with small-cell carcinoma, REE and metabolic aberrations did not respond to total parenteral nutrition. However, patients responding to chemotherapy experienced normalization of REE and metabolic status. 48 In other trials, parenteral and enteral nutrition have proved efficacious in reversing metabolic abnormalities induced by malignancy. For example, Burt et al8 noted normalization of glucose turnover, suppression of gluconeogenesis, increased protein synthesis, and decreased metabolism in esophageal cancer patients treated with total parenteral nutrition or enteral support. Bennegard and associates 4 similarly noted improvement in glucose metabolism in enterally fed cancer patients. The variability in patient response to nutritional therapy is probably related to the severity of metabolic derangement. Although nutritional goals are more difficult to accomplish in malnourished cancer patients than in patients without cancer, it is safe to conclude that nutritional therapy will most likely improve host nutritional indices. The potential for stimulating tumor growth is an obvious concern when patients with malignant disease are given nutritional support. Although numerous animal studies demonstrate significant acceleration of tumor growth during nutritional repletion, gross tumor stimulation by exogenous nutrient provision has not been clearly documented in humans. Mullen et al40 demonstrated no difference in protein synthesis rates of upper gastrointestinal cancers in 25 patients receiving total parenteral nutrition compared with patients nourished orally. Tumor fractional protein synthesis rates were identical in the two groups of patients despite a substantial increase in protein-calorie intake in parenterally nourished patients. Nixon et al41 noted no direct clinical or biochemical evidence of tumor stimulation with total parenteral nutrition for patients with metastatic colon cancer compared with patients fed orally.
SUPPORT IN GASTROINTESTINAL CANCER
533
Ota et al44 measured erythrocyte polyamine levels in cancer and noncancer patients before and after total parenteral nutrition. They noted significant increases in polyamine levels in cancer patients compared with noncancer patients during nutritional support. However, most patients also undergo oncologic treatment during nutritional support, which confounds the issue. Selective administration of certain nutrients may also affect tumor growth. In our laboratory, recent studies have focused on the dibasic semiessential amino acid arginine and its ability to augment host immune function and modulate tumor growth. Using a subcutaneously transplanted murine neuroblastoma, we found decreased tumor growth and prolonged survival in AlJ mice supplemented enterally with 1% arginine versus isonitrogenous glycine. This finding was associated with enhanced immune response to mitogens and specifically to the tumor in mixed lymphocytetumor culture. In a prospective trial, patients with upper gastrointestinal cancer were randomized to receive postoperative jejunostomy feedings supplemented with either arginine or glycine. Along with modest improvement in nitrogen balance, arginine supplementation significantly augmented lymphocyte response to mitogen and increased the percentage of T lymphocytes expressing the helper phenotype (CD4).1l While it appears that nutritional support in humans does not feed the tumor preferentially, a definitive study on tumor growth in humans is lacking. Nutritional repletion usually is given concurrently with surgery, chemotherapy, and radiation. Thus, a rationale for feeding cancer patients comes from an association of improved nutritional measures with improved outcome after various forms of oncologic therapy. Retrospective studies have correlated malnutrition with adverse clinical outcomes after surgery. These studies also reported decreased perioperative morbidity and mortality rates in patients treated with total parenteral nutrition. Several prospective randomized trials examined the role of total parenteral nutrition in decreasing perioperative morbidity and mortality rates. Heatley and associates 19 reported a significant reduction in wound complications in a group of patients with upper gastrointestinal cancer receiving preoperative parenteral nutrition. Mueller et al39 studied the efficacy of 10 days of total parenteral nutrition in surgery patients with gastrointestinal tumors. Postoperative morbidity was significantly lower in patients given parenteral nutrition (17%) than in the control group (32%). The mortality rate was 4% in the former group and 16% in the latter. However, other randomized trials noted no significant advantage in terms of operative morbidity and mortality with the provision of total parenteral nutrition. The accumulated data suggest that severely malnourished patients with upper gastrointestinal cancer benefit from preoperative parenteral nutrition. Patients with less severe nutritional deficits will probably not benefit from preoperative nutritional support in terms of decreasing operative morbidity and mortality rates and may be harmed by increased infectious complications. Prospective clinical trials have evaluated the effects of nutritional support as an adjunct to chemotherapy. 30 Tumor response to treatment and toxicity and survival were analyzed in these trials. The major toxicities
534
JOHN
M.
DALY ET AL.
evaluated were gastrointestinal, infectious, and hematologic. No difference between treated and control groups was noted in three studies, improvement in the treated group was noted in one study, and worse stomatitis in the treated group was noted in two studies. Two of eleven evaluable studies demonstrated less hematologic toxicity. Finally, two of five trials showed that infectious complications were more common in the group given total parenteral nutrition than in the control group. Similarly, these prospective trials have failed to demonstrate treatment response or survival advantage in patients receiving total parenteral nutrition. Radiation therapy increases the potential for the development of severe nutritional deficits. 14 The onset and degree of malnutrition relate to the tumor's location and the dosage of radiation. Radiotherapy may have the greatest detrimental effect on host nutrition when used to treat head and neck cancer. Such radiotherapy affects the patient's ability to eat secondary to xerostomia and mucositis and induces changes in taste perception. A randomized trial of patients with head and neck cancer treated with radiation therapy and enteral support showed that these patients were able to attain nutritional goals in contrast to the control patients. Although nutritional intervention may improve nutritional status, several clinical trials have not shown improved treatment response or tolerance, better survival, or reduced complications of therapy. 30 SUMMARY Malnutrition is extremely common in patients with malignant disease. Whereas the causes are multifactorial, the predominant factor is the imbalance between nutrient intake and host nutrient requirements. Furthermore, the evidence suggests that cachexia is related to abnormal changes in host intermediary metabolism induced by host-tumor interactions, and endogenous peptides such as TNF may be important mediators. The role of nutritional therapy in cancer patients remains to be defined. Clearly, patients with severe malnutrition benefit from nutritional intervention. However, the benefit of nutritional therapy in less severe cases of malnutrition as an adjuvant to oncologic therapy has yet to be established. REFERENCES 1. Aderka D, Fisher S, Levo Y, et al: Cachectin/tumour necrosis factor production by cancer patients. Lancet 2:1190, 1985 2. Asher A, Mule JJ, Reichert CM, et al: Studies on the anti-tumor efficacy of systematically administered recombinant tumor necrosis factor against several murine tumors in vivo. J Immunol 138:963-974, 1987 3. Balkwill F, Burke F, Talbot D, et al: Evidence for tumour necrosis factor/cachectin production in cancer. Lancet 1:1229-1232, 1987 4. Bennegard K, Eden E, Eisman L, et al: Metabolic response of whole body and peripheral tissues to enteral nutrition in weight losing cancer and noncancer patients. Gastroenterology 85:92-99, 1983 5. Beutler B, Cerami A: Cachectin and tumor necrosis factor as two sides of the same biological coin. Nature 320:584-588, 1985 6. Beutler B, Milsark IW, Cerami AC: Passive immunization against cachectin/tumor necrosis factor protects mice from lethal effect of endotoxin. Science 229:1925-1937, 1985
SUPPORT IN GASTROINTESTINAL CANCER
535
7. Burt ME, Brennan MF: Nutritional support of the patient with esophageal cancer. Semin Oncol11:127-135, 1984 8. Burt ME, Stein TP, Schwade JG, et al: Whole body protein metabolism in cancer-bearing patients. Cancer 53:1246-1254, 1984 8a. Buzby GP, Mullen JL, Matthews DL, et al: Prognostic nutritional index in gastrointestinal surgery. Am J Surg 139:160-167, 1980 9. Carswell EA, Old LJ, Casse RL, et al: An endotoxin-induced serum factor that causes necrosis of tumors. Proc Natl Acad Sci USA 72:3666-3670, 1975 10. Chlebowski RT: Abnormal glucose metabolism in patients with advanced cancer. In Heber D (moderator): Malnutrition and cancer: Mechanisms and therapy. Nutrition International, 1986 11. Daly JM, Reynolds J, Thom A, et al: Immune and metabolic effects of arginine in the surgical patient. Ann Surg 208:512-523, 1988 12. Dempsey DT, Feurer ID, Know LS, et al: Energy expenditure in malnourished gastrointestinal cancer patients. Cancer 53:1265-1273, 1984 13. DeWys WD, Begg C, Lavin PT, et al: Prognostic effect of weight loss prior to chemotherapy in cancer patients. Am J Med 69:491-497, 1980 14. Donaldson SS: Nutritional support as an adjuvant to radiation therapy. JPEN 8:302-309, 1984 15. Eden E, Edstrom S, Bennegard K, et al: Glucose flux in relation to energy expenditure in malnourished patients with and without cancer during periods of fasting and feeding. Cancer Res 44:1717-1724, 1984 16. Emery PW, Lowvell L, Rennie MJ: Protein synthesis measured in vivo in muscle and liver of cachectic tumor-bearing mice. Cancer Res 44:2779-2784, 1984 17. Gutman A, Thilo E, Biren S: Enzymes of gluconeogenesis in tumor-bearing rats. Isr J Med Sci 5:998-1001, 1969 18. Hammond KD, Balinsky D: Activities of key gluconeogenic enzymes and glycogen synthase in rat and human livers and hepatoma cell cultures. Cancer Res 38:13171322, 1978 19. Heatley RV, Williams R, Lewis M: Preoperative intravenous feeding: A controlled trial. Postgrad Med J 55:541-545, 1979 20. Heber D, Byerley LO, Chlebowski RT: Metabolic abnormalities in the cancer patient. Cancer 55:225-233, 1985 21. Heber D, Chlebowski RT, Ishibashi DE, et al: Abnormalities in glucose and protein metabolism in noncachectic lung cancer patients. Cancer Res 42:4815-4819, 1982 22. Holroyde CP, Reichard GA: Carbohydrate metabolism in cancer cachexia. Cancer Treat Rep 65:55-59, 1981 23. Holroyde CP, Gabuzda G, Putman RC, et al: Altered glucose metabolism in metastatic cancer. Cancer Res 35:3710-3714, 1975 24. Holroyde CP, Skutches CL, Boden G, et al: Glucose metabolism in cancer cachectic patients with colorectal cancer. Cancer Res 44:5910-5913, 1984 25. Jasani B, Donaldson LJ, Ratcliffe JG, et al: Mechanism of impaired glucose tolerance in patients with neoplasia. Br J Cancer 38:286-292, 1978 26. Jeevanandam M, Lowry SF, Horowitz GD, et al: Cancer cachexia and protein metabolism. Lancet 1:1423-1426, 1984 27. Karlberg HI, Kern KA, Fischer JE: Albumin turnover in sarcoma-bearing rats in relation to cancer anorexia. Am J Surg 145:95-101, 1983 28. Kawamura I, Moldawer LL, Keenan RA, et al: Altered amino acid kinetics in rats with progressive tumor growth. Cancer Res 42:824-829, 1982 29. Kern KA, Norton JA: Cancer cachexia. JPEN 12:286-298, 1988 30. Koretz RL: Parenteral nutrition: Is it oncologically logical? I Clin Oncol 2:534-548, 1984 31. Krause R, Humphrey C, Von Meyetfeldt M, et al: A central mechanisms for anorexia in cancer: A hypothesis. Cancer Treat Rep 65(suppl 5):15-23, 1981 32. Lundholm K, Holm G, Schersten T: Insulin resistance in patients with progressive malignant disease. Cancer Res 39:1968-1972, 1979 33. Lundholm K, Edstrom S, Karlberg I, et al: Glucose turnover, gluconeogenesis from glycerol, and estimation of net glucose cycling in cancer patients. Cancer 50:1142-1150, 1982 34. Marks PA, Bishop IS: The glucose metabolism of patients with malignant disease and of normal subjects as studied by means of an intravenous glucose tolerance test. J Clin Invest 36:254-264, 1957
536
JOHN
M.
DALY ET AL.
35. Mathison JC. Wolfson E, Ulevitch RJ: Participation of tumor necrosis factor in the mediation of gram negative bacterial lipopolysaccharide induced injury in rabbit. J Clin Invest 81:1925-1937, 1988 36. Michie HR, Manogue KR, Spriggs DR, et al: Detection of circulating tumor necrosis factor after endotoxin administration. N Engl J Med 318:1481-1486, 1988 37. Moley JF, Morrison SD, Gorschboth CM, et al: Body composition changes in rats with experimental cancer cachexia: Improvement with exogenous insulin. Cancer Res 48:2784-2787, 1988 38. Moody TW, Pert CB, Gazdar AR, et al: High levels of intracellular bombesin characterize human small-cell lung carcinoma. Science 214:1246-1248, 1981 39. Mueller J, Brenne V, Dienst J: Peroperative parenteral feeding in patients with gastrointestinal carcinoma. Lancet 1:68-71, 1982 40. Mullen JL, Buzby GP, Gertner MH, et al: Protein synthesis dynamics in human gastrointestinal malignancies. Surgery 87:331-338, 1980 41. Nixon DW, Moffitt S, Lawson DH, et al: Total parenteral nutrition as an adjunct to chemotherapy of metastatic colorectal cancer. Cancer Treat Rep 65(suppl):121-128, 1981 42. Norton JA, Stein TP, Brennan MF: Whole body protein synthesis and turnover in normal man and malnourished patients with and without known cancer. Ann Surg 194:123128, 1981 43. Norton JA, Shamberger MD, Stein TP, et al: The influence of tumor-bearing on protein metabolism in the rat. J Surg Res 30:456-462, 1981 44. Ota D, Nishioka K, Foulkes M, et al: Nutritional parameters affecting erythrocyte polyamine levels in cancer patients. J Clin Oncol 2: 1157-1164, 1984 45. Peacock JL, Norton JA: Impact of insulin on survival of cachetic tumor-bearing rats. JPEN 12:260-264, 1988 45a. Pettigrew RA, Hill GL: Indicators of surgical risk and clinical judgement. Br J Surg 73:47-51, 1986 46. Roh MS, Ekman L, Jeevanandam M, et al: Gluconeogenesis in tumor-influenced hepatocytes. Surgery 96:427-433, 1984 47. Rohdenberg GS, Bernhard A, Krehbiel 0: Sugar tolerance in cancer. JAMA 72:15281529, 1919 48. Russell D, Shike M, Marliss E, et al: Effects ofTPN and chemotherapy on the metabolic derangements in small cell lung cancer. Cancer Res 44:1706-1711, 1984 49. Schein PS, Kisner D, Haller D, et al: Cachexia of malignancy: Potential role of insulin in institutional management. Cancer 49:2070-2076, 1979 50. Shaw JHF, Wolfe RR: Glucose and urea kinetics in patients with early and advanced gastrointestinal cancer: The response to glucose infusion, parenteral feeding, and surgical resection. Surgery 101:181-191, 1986 51. Shaw JHF, Humberstone DM, Wolfe RR: Energy and protein metabolism in sarcoma patients. Ann Surg 207:283-289, 1988 52. Theologides A: Cancer cachexia. Cancer 43:2004-2012, 1979 53. Tracey KJ, Fong Y, Hesse DG, et al: Anti-cachectinffNF monoclonal antibodies prevent septic shock during lethal bacteremia. Nature 330:662-664, 1987 54. Warren RS, Jeevanandam M, Brennan MF: Protein synthesis in the tumor-influenced hepatocyte. Surgery 98:275-282, 1985 55. Warren RS, Starnes F, Gabrilove JL, et al: The acute metabolic effects of tumor necrosis factor administration in humans. Arch Surg 122:1396-1400, 1987 56. Waterhouse C: Lactate metabolism in patients with cancer. Cancer 33:66-71, 1974 57. Waterhouse C, Jeanpetre N, Keilson J: Gluconeogenesis from alanine in patients with progressive malignant disease. Cancer Res 39:1968-1972, 1979 58. Waterlow JC, Goldend M, Picou J: The measurement of rates of protein turnover, synthesis, and breakdown in man and the effect of nutritional status and surgical injury. Am J Clin Nutr 30:1333-1337, 1977
Address reprint requests to John M. Daly, MD Department of Surgery The University of Pennsylvania School of Medicine 3400 Spruce Street Philadelphia, PA 19104
0039--6109/91 $0.00
+ .20
Nutritional Support of Patients with Cancer of the Gastrointestinal Tract
John M. Daly, MD, * H.P. Redmond, BSc, FRCSI, t Michael D. Lieberman, MD,+ and Lori Jardines, MD§
Among hospitalized patients, those with cancer have the highest incidence of protein-calorie malnutrition. Significant malnutrition occurs in more than 50% of cancer patients undergoing major operative procedures. Clinically significant malnutrition is a result of diminished nutrient intake, increased nutrient losses, and tumor-induced derangements in host metabolism. In the absence of adequate exogenous nutrients, the body utilizes endogenous substrates to satisfY the ongoing requirements of both host and tumor for energy and protein. In patients with malignant obstruction of the gastrointestinal tract, the tumor itself may induce diminished nutrient intake. Present-day treatment modalities, including gastrointestinal resection, chemotherapy, and radiotherapy, compound these metabolic derangements, which further increase the risk of Significant postoperative morbidity and death. The presence of malnutrition in cancer patients has major prognostic significance. In a review of more than 3000 cancer patients, DeWys and colleagues found better survival rates in those patients without weight loss compared with patients who had lost just 6% of their body weight. 13 Numerous other studies have noted increased postoperative morbidity and mortality rates associated with malnutrition. The development and refinement of enteral and parenteral nutrition support techniques have provided the opportunity for studying the relation between nutritional supplementation and postoperative prognosis. Therefore, because cancer patients are often malnourished and because malnutrition is associated with increased postoperative morbidity, nutritional support is often instituted to improve From the Division of Surgical Oncology, Department of Surgery, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania *Jonathan E. Rhoads Professor of Surgery, and Chief, Division of Surgical Oncology tResearch Fellow :j:American Cancer Society Fellow in Surgical Oncology §Assistant Professor of Surgery
Surgical Clinics of North America-Vol. 71, No.3, June 1991
523
524
JOHN
M.
DALY ET AL.
nutritional status and potentially reduce the risks of postoperative complications. This article addresses the role of nutritional therapy in patients with cancer of the gastrointestinal tract.
NUTRITIONAL STATUS AND CANCER The term cancer cachexia encompasses a varied group of symptoms and signs including inanition, anorexia, weakness, organ dysfunction, and eventually death. Although this cachexia is clearly related to the presence of malignancy, there is no consistent relationship between the development of cachexia and the stage of the disease, tumor histology, primary site, or duration of illness. The cause of cancer cachexia is thought to be multifactorial (Table 1). Anorexia Anorexia associated with cancer cachexia develops for a number of reasons. First, local effects of tumor that diminish nutrient intake are often involved, particularly when the tumor arises from or impinges on the gastrointestinal tract. Tumors arising in the pharynx and the esophagus can cause various degrees of dysphagia secondary to partial or complete obstruction. Gastric neoplasms may cause early satiety or outlet obstruction, and enteric or colonic obstruction can result in intestinal obstruction, blind loop syndrome, and fistulas. Second, alterations in taste perception and appetite contribute to diminished nutrient intake. Patients often have an increased threshold for sweet taste and a concomitant decrease in sour and salty taste thresholds. Deficiencies in trace elements such as zinc have been implicated in these changes in some instances. Third, patients with extensive hepatic involvement by tumor may develop nausea and anorexia secondary to high lactic acid levels produced by anaerobic tumor metabolism. Fourth, neuroendocrine control of appetite is centered in the lateral hypothalamus, and a number of factors may alter afferent and efferent stimuli to and from this area. Tryptophan, a precursor of the neurotransmitter serotonin, has been shown to be elevated in brain tissue of tumor-bearing animals, and alterations in serotonin levels are associated with anorexia. Finally, chemotherapeutic agents can act centrally, in the chemoreceptor trigger zone, or locally on the gastrointestinal tract, inducing nausea and vomiting or stomatitis, mucositis, glossitis, and pharyngitis. Table 1. Factors Related to Cancer Cachexia Anorexia Local tumor effects Altered taste perception Elevated lactic acid levels Neuroendocrine factors Chemotherapeutic agents Abnormal carbohydrate metabolism Abnormal lipid metabolism Abnormal protein metabolism
SUPPORT IN GASTROINTESTINAL CANCER
525
The extent of the malnourished state associated with malignant disease is often greater than expected from the decrease in dietary intake. Furthermore, the degree and type of malnutrition vary in the cancer patient. Several studies have demonstrated that the presence of tumor appears to affect the resting energy expenditure (REE). Although it appears that some cancer patients have altered resting and total energy expenditure, those factors that determine elevation or depression of energy expenditure remain unclear. Dempsey et al studied 173 patients with gastrointestinal malignancies, noting that gastric cancer patients were more likely to be hypermetabolic than were patients with pancreatic and hepatobiliary tumors. 12 Curative resection results in a return to normal levels of REE, but patients who undergo palliative resection fail to achieve this normalcy. Carbohydrates Abnormalities in host carbohydrate metabolism are common in cancer patients. Peripheral utilization of glucose, hepatic gluconeogenesis, and consequently glucose turnover have been demonstrated to be abnormal. 10, 22, 23, 34 Altered glucose tolerance was recognized many years ago and was suggested as early as 1919 to be a possible diagnostic aid in patients with gastrointestinal symptoms. 47 Glucose intolerance has been correlated with insulin resistance 32, 49 and may involve an attenuated insulin secretory response. 20 Recent studies in sarcoma-bearing rats suggest that supplemental insulin administration may preserve host lean body mass and even influence survival. 37, 45 Cancer patients appear to be glucose intolerant and also have been found to synthesize glucose at an elevated rate. Recently, Shaw and Wolfe 50 noted that patients with gastrointestinal tumors had elevated rates of basal hepatic glucose production, They noted a direct relation between the extent of the tumor burden and the increase in gluconeogenesis. Furthermore, patients with the largest tumor burdens failed to suppress their own endogenous glucose production during glucose infusions. In studies on patients with sarcomas, those workers also demonstrated that administration of glucose did not suppress endogenous protein catabolism. 51 The reason for the elevated rates of hepatic gluconeogenesis is unclear. Increased levels of appropriate substrates from tissue breakdown in the periphery may fuel a substrate-driven process. Increased gluconeogenesis occurs in cachectic patients who are challenged with the substrates alanine, glycerol, or lactate. 33,56,57 Plasma concentrations oflactate are often elevated in cancer patients. At least part of the explanation resides with the host hepatocytes themselves, as hepatocytes isolated from glycogen-depleted tumor-bearing rats not only produce more glucose from available substrate than do hepatocytes from similar nontumor-bearing controls but also respond by producing more glucose when challenged with alanine and lactate. 46 Such increases in the hepatic capacity to synthesize glucose have also been demonstrated by Gutman and associates 17 as well as by Hammond and Balinsky.18 These findings are consistent with radioisotope tracer studies demonstrating increases in glucose recycling. 16, 33 The Cori cycle is a metabolically futile cycle in which glycolytic production of lactate serves as substrate for
526
JOHN
M.
DALY ET AL.
hepatocyte glucose production. Because the energy required to synthesize glucose outweighs the energy created during glycolysis, increased use of this pathway by an organism can lead to significant waste of energy. Holroyde et al found levels of Cori cycle activity to be increased in cancer patients compared with starved controls.23, 24 Burt and Brennan also demonstrated increased rates of glucose turnover in 12 patients with localized esophageal carcinoma. 7 Gluconeogenic activity is often inappropriately high in cancer patients compared with non tumor-bearing controls. Eden et al studied eight cancer patients and seven control patients after a 14-day period of enteral nutrition. 15 In both the basal and the re-fed states, cancer patients had increased levels of glucose turnover and recycling. The implications for energy loss in such a situation are substantial. The increased glucose flux observed amounted to 42% of the daily glucose intake in the cancer patient population, or between 250 and 350 kcal per day.
Lipids Alterations in lipid metabolism in cancer patients include changes in body fat composition and increased lipid mobilization. Decreases in total body fat are common in these patients and may be related to insulin deficiency. Decreased levels oflipoprotein lipase correlate with the degree of weight loss, Increased lipolysis allows glycerol and fatty acids to be utilized for gluconeogenesis and ~mergy production. Fatty acids have been implicated as the main substrates in some patients with neoplastic disease, and these patients fail to suppress lipolysis after glucose administration.
Protein Although abnormal carbohydrate cycling has been implicated as the central mechanism for host nutritional depletion, abnormalities in protein metabolism occur and are interrelated. Intuitively, severe wasting of host muscle mass along with depletion of visceral and circulating proteins (especially albumin) occurs secondary to a combination of increased protein breakdown and decreased protein synthesis. Yet studies in both animals and humans suggest that a more complex derangement of protein metabolism is induced in the tumor-bearing host. Both tumor-bearing animals and persons with cancer seem to be unable to respond appropriately to diminished nutrient supplies. 58 However, Kawamura et al found that sarcoma-bearing rats incorporated radiolabeled tyrosine into total body protein at a rate 33% faster than controls. 28 At the same time, incorporation of tyrosine into skeletal muscle (gastrocnemius and rectus abdominis) was decreased from 6% to 10%. The reconciliation of increased total protein flux in the face of decreased skeletal muscle synthesis may come from increases in hepatic fractional protein synthesis. In studies comparing sarcoma-bearing rats with nontumor-bearing pair-fed controls, Norton et al43 demonstrated that despite equal nitrogen balance between the groups, the tumor-bearing animals had increased total protein flux and diminished muscle protein synthesis but increased hepatic protein synthesis. Protein synthesis within the tumor did not fluctuate with nutritional state.
SUPPORT IN GASTROINTESTINAL CANCER
527
In humans, a similar pattern of deranged protein metabolism has been documented. Norton et al42 noted that three of seven cancer patients had elevated levels of labeled glycine incorporated into protein. In the face of decreased nutrient intake, even the normal levels observed in the other four cancer patients become suspect. Heber and associates found increased levels of whole-body protein turnover and increased excretion of 3-methylhistidine (evidence of muscle breakdown) in a group of 12 patients with nonsmall-cell lung cancer.21 Jeevanandam et al,26 in a study comparing cancer patients with non tumor-bearing malnourished patients and healthy controls, found markedly increased levels of protein turnover in the cancer group. Evidence of decreased skeletal muscle protein synthesis and increased hepatic protein synthesis is consistent in both animals and humans. Warren et al 54 used isolated hepatocytes from sarcoma-bearing rats to demonstrate enhanced synthesis of both secretory and fixed cellular proteins compared with pair-fed controls. Furthermore, these increases were proportional'to tumor load. This group of researchers postulated an increase in hepatic protein degradation to explain their finding of equivalent protein levels in tumor-bearing and non tumor-bearing livers. Karlberg et aP7 have examined albumin production in a similar model and found it to be elevated. However, increases in the rate of albumin degradation and subsequent expansion of the intravascular volume result in hypoalbuminemia. Concomitant increases in muscle protein degradation may playa role in muscle wasting (and overall protein flux), as evidenced by increased levels of urinary 3-methylhistidine in studies of both animals and humans with tumor. Emery et aP6 documented an 80% increase in the intramuscular level of free 3-methylhistidine in mice bearing tumors. Thus, in the tumorbearing organisms, levels of protein flux depend on the relative rates of synthesis and degradation in both skeletal muscle and liver. Theologides52 has proposed that a tumor-secreted peptide is responsible for the changes in intermediary metabolism that occur in the tumor-bearing host. However, it appears unlikely that tumors themselves are universally capable of inducing the array of metabolic abnormalities described above. Although in isolated instances tumors elaborate products with anorexigenic effects (such as carcinoids [which elaborate serotonin] and small-cell lung carcinomas [which produce bombesin]), no tumor-derived product with a systemic effect has yet been identified. 31, 38 Recently, studies have evaluated the host-produced factors associated with the endogenous immune response to a tumor. These factors, or cytokines, are produced by immunocytes as a paracrine communication and are integral to a coordinated immune response. Interleukins 1, 2, 4, and 6; interferons alpha, beta, and gamma; and lymphotoxin all have powerful effects on cells of the immune system. One such macrophage-elaborated cytokine, tumor necrosis factor-alpha (TNF) or cachectin, appears to exert its effects beyond the immune system and produces many of the systemic effects seen in cachexia. Tumor necrosis factor was first defined as the primary antitumor agent found in the serum of BCG-primed, lipopolysaccharide-treated mice and has since been found to be cytotoxic or cytostatic to a wide variety of
528
JOHN
M.
DALY ET AL.
tumors in vivo and in vitro. 2 , 9 Interest in TNF as a possible mediator of cancer cachexia grew from observations that trypanosomiasis in rabbits was accompanied by severe tissue wasting and hypertriglyceridemia. Beutler and Cerami attributed these effects to TNF after demonstrating that TNF was capable of inducing hypertriglyceridemia and carcass fat loss in mice by inhibiting transcription of the gene for lipoprotein lipase. 5 As reduced levels of lipoprotein lipase activity have been identified in the plasma of cancer patients, and as increased levels of TNF have been found both in the serum of cancer patients and as a product of monocytes isolated from cancer patients, investigation of TNF as a mediator of cancer cachexia has intensified. 1, 3 Receptors for TNF have been found on metabolically important liver, adipose, and muscle cells. In vitro studies of cultured cells from these tissues have demonstrated mixed results depending on isolation techniques, length of exposure to TNF, and other factors. In both animals and humans, TNF has been administered to normal and noncachetic tumor-bearing controls. In animal studies, rats given sublethal doses of TNF have demonstrated increased liver amino acid uptake, glucose turnover, and levels of both acute-phase reactants and stress hormones. Concomitantly, recipient animals demonstrate diminished appetite with little change in anabolic hormone levels. When antibody to TNF has been administered to animals given endotoxin or live Escherichia coli, a protective effect has been noted, ranging from abrogation of the stress response (smaller doses) to survival after previously defined lethal challenge. 6, 35, 53 In humans, Warren et al55 have measured the acute response to TNF in healthy volunteers. The changes included a rise in C-reactive protein, a decrease in serum zinc levels, and a doubling of the forearm amino acid flux. This last effect was primarily attributable to increases in the gluconeogenic amino acids alanine and glutamine. Michie et al36 examined the relation of endotoxin to TNF in healthy volunteers. They found a brief pulse of TNF in the subjects' plasma after bolus endotoxin administration, followed by the characteristic increases in temperature, heart rate, ACTH, and epinephrine. Significantly, administration of ibuprofen blunted these responses while having no effect on TNF levels. Emerging parallels between the metabolic changes found in cancer cachexia and after the administration of cytokines led Kern and Norton to speculate about the mechanisms for host cachexia. 29 They proposed that cytokines elaborated by activated immunocytes in response to tumor have secondary effects on host organs. In the short term, these effects would promote an acute-phase response, rerouting nutrients from the periphery to the liver. However, over the long term, such teleologic strategies backfire, resulting in anorexia and widespread abnormalities in carbohydrate, protein, and lipid metabolism. CONSEQUENCES OF MALNUTRITION IN THE CANCER PATIENT Significant perioperative risk factors in cancer patients include age, obesity, chronic illness, and malnutrition. The consequences of malnutrition for the tumor-bearing host have been well documented.
SUPPORT IN GASTROINTESTINAL CANCER
529
Table 2. The Nutritionally High-Risk Patient Gross underweight: weight-for-height below 80% of standard Gross overweight: weight-for-height above 120% of standard Recent loss of 12% or more of usual body weight Alcoholism No oral intake for more than 10 days on simple intravenous solutions Protracted nutrient losses Malabsorption syndromes Short-gut syndromes/fistulas Renal dialysis Draining abscesses, wounds Increased metabolic needs Protracted fever and infection Trauma
In human studies, evidence of increased morbidity and mortality rates with depressed immunocompetence exists. Poor wound healing, increased w0und infection rates, and prolonged postoperative ileus and hospital stay have all been linked to poor nutritional status in cancer patients. Preoperative assessment of high-risk patients, in particular evaluating reversible factors such as malnutrition, is essential for optimal outcome (Table 2). Nutritional assessment techniques include a detailed history, noting the degree of weight change as well as dietary and alcohol intake, and a thorough physical examination (Table 3). A global clinical assessment stratifies patients into one of three nutritional categories: normal status, mild to moderate malnutrition, or severe malnutrition. Other measures studied are serum albumin and transferrin levels. Total circulating lymphocyte counts, delayed cutaneous hypersensitivity, grip strength, creatinineheight index, and anthropometry are used to a much lesser extent. A number of authors have combined indices of nutritional status to predict postoperative sequelae. The prognostic nutritional index, described by Buzby et al,8a combines serum albumin and transferrin levels, body fat (skinfold thickness), and delayed cutaneous hypersensitivity skin testing. Pettigrew and Hill45a compared body composition measurements, grip strength, prognostic indices, and clinical judgment. A careful assessment Table 3. Assessment of Nutritional Status History and physical examination Extent of disease Weight loss Diet history Anthropometric analysis Laboratory evaluation Hepatic protein synthesis Liver function tests BUN/creatinine Cholesterol/triglyceride Amylase Electrolytes Glucose Immune function Estimation of energy expenditure
530
JOHN
M.
DALY ET AL.
of medical risk, noting in particular cardiorespiratory disease and preexisting sepsis, as well as nutritional state, was as effective as other currently used indications of risk.
NUTRITIONAL SUPPORT The challenge of nutritional support is to prevent or reverse the cachexia of malignancy. The consequences of malnutrition are severe and include impairment of immune function with increased susceptibility to infection and perhaps tumor growth, increased perioperative morbidity, and ultimately death. When one is attempting to treat malnutrition, the patient with cancer must have a complete nutritional assessment, determination of the optimal route of nutrient administration, and meticulous monitoring of nutritional therapy. Standard methods for assessing nutritional status include a thorough history and physical examination, anthropometric measurements, laboratory evaluation of specific visceral organ function, measurement of immune function, and estimation of energy expenditure. Completion of a thorough history and physical examination is important in terms of documenting the extent of disease, identifYing coexisting illness, and quantifying the rate and degree of weight loss and associated protein-calorie deficits. A loss of more than 10% of the usual body weight correlates with increased morbidity and mortality rates. Anthropometry is a noninvasive method of determining body composition based on measurements of skinfold thickness (total body fat) and midarm muscle circumference (lean body mass). Anthropometrics alone has limited value in diagnosing malnutrition, predicting morbidity, or monitoring the effectiveness of nutritional therapy. Laboratory determination of hepatic protein synthesis (i. e., albumin, thyroxine-binding prealbumin, transferrin), liver function tests, blood urea nitrogen, creatinine, triglycerides, cholesterol, amylase, electrolytes, and glucose levels should be performed on all cancer patients being considered for nutritional support. Immune competence, as determined by delayed-type hypersensitivity, may be measured. Finally, patients with severe nutritional deficits should have periodic measurements of REE and 24-hour nitrogen balance. No single clinical or laboratory test accurately predicts the severity of malnutrition. In fact, clinical judgment is superior or equal to single objective measures in predicting nutritionally associated complications. Multiparameter nutritional indices have been developed to identifY patients at increased risk of nutritionally related morbidity and death. One useful predictive index, the prognostic nutritional index, relates four variables (serum albumin, serum transferrin, triceps skinfold, and delayed-type hypersensitivity) to postoperative morbidity and mortality rates. 8a Once it has been shown that volitional oral intake is inadequate and the patient's need for nutritional maintenance or repletion has been evaluated, the route for nutrient delivery must be determined. The alternatives are enteral or intravenous delivery or both. Enteral and parenteral techniques are equally efficacious in meeting nutritional requirements. Enteral feeding is the preferred method of nutritional support in
531
SUPPORT IN GASTROINTESTINAL CANCER
Table 4. Contraindications to Enteral Nutritional Support Absolute Complete mechanical intestinal obstruction Severe diarrhea High-output external fistula (>500 ml/day) Severe acute pancreatitis Shock Relative Immediate postoperative or poststress period «12 hour) Acute severe enteritis Complete absence of small intestine
terms of preserving gastrointestinal architecture and preventing bacterial translocation from the gut. The specific contraindications for enteral feeding are listed in Table 4. Enteral feedings may be delivered by a variety of techniques. It is preferable to administer nutrients distal to the ligament of Treitz to avoid tHe complication of aspiration pneumonia and gastric ileus. This may be accomplished by placement of a small (S-F) silicone rubber nasoduodenal tube, whose position is always confirmed radiographically, or by surgical construction of a feeding jejunostomy. If it is anticipated that the duration of enteral support will be longer than 4 weeks, a jejunostomy is performed. Gastric feeding is acceptable via nasogastric tube or gastrostomy in an alert patient with an intact gag mechanism who has a patent gastrointestinal tract. The gastrostomy may be placed percutaneously with the aid of endoscopy or by an operative Stamm procedure. Complications related to enteral therapy include abdominal wound infection, bowel obstruction, tube dislodgment with peritonitis, aspiration pneumonia, gastrointestinal intolerance, and metabolic abnormalities. In those patients unsuitable for enteral support, parenteral nutrition remains an important option. These solutions are hypertonic and therefore require catheterization of the central venous system to reduce the incidence of phlebitis and venous thrombosis. Acute therapy is usually provided by percutaneous catheterization of the subclavian or internal jugular vein. Chronic support is administered through a subcutaneous catheter (Hickman), which lies in the superior vena cava. Complications of parenteral therapy include pneumothorax, venous thrombosis, arterial trauma, sepsis, electrolyte disorders, and metabolic abnormalities such as glucose intolerance and nonketotic hyperosmolar coma. Caloric and nitrogen requirements must be calculated for each patient. The nonprotein caloric requirement is based on measurement of REE by indirect calorimetry or the Harris-Benedict equation: Male: Basal energy expenditure (kcal) = 66 + 13.7 + 5 X Height (cm) - 6.S X Age (yrs) Female: Basal energy expenditure (kcal) = 65 + 9.6 + 1. 7 X Height (cm) - 4.7 X Age (yrs)
Weight (kg)
X
X
Weight (kg)
In general, 130% to 150% of the REE is provided as nonprotein calories.
532
JOHN
M.
DALY ET AL.
These nonprotein calories include carbohydrate and lipid in an approximate ratio of 2:1 to 1:1, respectively. The administration of protein is based on a determination of nitrogen balance. Typically, 1.5 to 2.5 gm/kg of protein per day is provided, with the higher value reserved for severely depleted or septic patients. Close monitoring of the patient's progress is crucial in adjusting dietary formulation and requirements.
EFFECTS OF NUTRITIONAL SUPPORT ON HOST, TUMOR, AND ONCOLOGIC THERAPY The merit of nutritional support in the cancer patient must be evaluated in terms of changes in host nutritional status, tumor biology, and efficacy of oncologic therapy. The results of detailed studies support the idea that the success of nutritional therapy in terms of repleting the host and reversing metabolic abnormalities is dependent on tumor type as well as the cause and degree of malnutrition. When the malnutrition is attributed to diminished food intake, then nutritional intervention is quite effective. However, malnutrition associated with metabolic abnormalities or advanced malignancy is much more difficult to treat. In a study of patients with small-cell carcinoma, REE and metabolic aberrations did not respond to total parenteral nutrition. However, patients responding to chemotherapy experienced normalization of REE and metabolic status. 48 In other trials, parenteral and enteral nutrition have proved efficacious in reversing metabolic abnormalities induced by malignancy. For example, Burt et al8 noted normalization of glucose turnover, suppression of gluconeogenesis, increased protein synthesis, and decreased metabolism in esophageal cancer patients treated with total parenteral nutrition or enteral support. Bennegard and associates 4 similarly noted improvement in glucose metabolism in enterally fed cancer patients. The variability in patient response to nutritional therapy is probably related to the severity of metabolic derangement. Although nutritional goals are more difficult to accomplish in malnourished cancer patients than in patients without cancer, it is safe to conclude that nutritional therapy will most likely improve host nutritional indices. The potential for stimulating tumor growth is an obvious concern when patients with malignant disease are given nutritional support. Although numerous animal studies demonstrate significant acceleration of tumor growth during nutritional repletion, gross tumor stimulation by exogenous nutrient provision has not been clearly documented in humans. Mullen et al40 demonstrated no difference in protein synthesis rates of upper gastrointestinal cancers in 25 patients receiving total parenteral nutrition compared with patients nourished orally. Tumor fractional protein synthesis rates were identical in the two groups of patients despite a substantial increase in protein-calorie intake in parenterally nourished patients. Nixon et al41 noted no direct clinical or biochemical evidence of tumor stimulation with total parenteral nutrition for patients with metastatic colon cancer compared with patients fed orally.
SUPPORT IN GASTROINTESTINAL CANCER
533
Ota et al44 measured erythrocyte polyamine levels in cancer and noncancer patients before and after total parenteral nutrition. They noted significant increases in polyamine levels in cancer patients compared with noncancer patients during nutritional support. However, most patients also undergo oncologic treatment during nutritional support, which confounds the issue. Selective administration of certain nutrients may also affect tumor growth. In our laboratory, recent studies have focused on the dibasic semiessential amino acid arginine and its ability to augment host immune function and modulate tumor growth. Using a subcutaneously transplanted murine neuroblastoma, we found decreased tumor growth and prolonged survival in AlJ mice supplemented enterally with 1% arginine versus isonitrogenous glycine. This finding was associated with enhanced immune response to mitogens and specifically to the tumor in mixed lymphocytetumor culture. In a prospective trial, patients with upper gastrointestinal cancer were randomized to receive postoperative jejunostomy feedings supplemented with either arginine or glycine. Along with modest improvement in nitrogen balance, arginine supplementation significantly augmented lymphocyte response to mitogen and increased the percentage of T lymphocytes expressing the helper phenotype (CD4).1l While it appears that nutritional support in humans does not feed the tumor preferentially, a definitive study on tumor growth in humans is lacking. Nutritional repletion usually is given concurrently with surgery, chemotherapy, and radiation. Thus, a rationale for feeding cancer patients comes from an association of improved nutritional measures with improved outcome after various forms of oncologic therapy. Retrospective studies have correlated malnutrition with adverse clinical outcomes after surgery. These studies also reported decreased perioperative morbidity and mortality rates in patients treated with total parenteral nutrition. Several prospective randomized trials examined the role of total parenteral nutrition in decreasing perioperative morbidity and mortality rates. Heatley and associates 19 reported a significant reduction in wound complications in a group of patients with upper gastrointestinal cancer receiving preoperative parenteral nutrition. Mueller et al39 studied the efficacy of 10 days of total parenteral nutrition in surgery patients with gastrointestinal tumors. Postoperative morbidity was significantly lower in patients given parenteral nutrition (17%) than in the control group (32%). The mortality rate was 4% in the former group and 16% in the latter. However, other randomized trials noted no significant advantage in terms of operative morbidity and mortality with the provision of total parenteral nutrition. The accumulated data suggest that severely malnourished patients with upper gastrointestinal cancer benefit from preoperative parenteral nutrition. Patients with less severe nutritional deficits will probably not benefit from preoperative nutritional support in terms of decreasing operative morbidity and mortality rates and may be harmed by increased infectious complications. Prospective clinical trials have evaluated the effects of nutritional support as an adjunct to chemotherapy. 30 Tumor response to treatment and toxicity and survival were analyzed in these trials. The major toxicities
534
JOHN
M.
DALY ET AL.
evaluated were gastrointestinal, infectious, and hematologic. No difference between treated and control groups was noted in three studies, improvement in the treated group was noted in one study, and worse stomatitis in the treated group was noted in two studies. Two of eleven evaluable studies demonstrated less hematologic toxicity. Finally, two of five trials showed that infectious complications were more common in the group given total parenteral nutrition than in the control group. Similarly, these prospective trials have failed to demonstrate treatment response or survival advantage in patients receiving total parenteral nutrition. Radiation therapy increases the potential for the development of severe nutritional deficits. 14 The onset and degree of malnutrition relate to the tumor's location and the dosage of radiation. Radiotherapy may have the greatest detrimental effect on host nutrition when used to treat head and neck cancer. Such radiotherapy affects the patient's ability to eat secondary to xerostomia and mucositis and induces changes in taste perception. A randomized trial of patients with head and neck cancer treated with radiation therapy and enteral support showed that these patients were able to attain nutritional goals in contrast to the control patients. Although nutritional intervention may improve nutritional status, several clinical trials have not shown improved treatment response or tolerance, better survival, or reduced complications of therapy. 30 SUMMARY Malnutrition is extremely common in patients with malignant disease. Whereas the causes are multifactorial, the predominant factor is the imbalance between nutrient intake and host nutrient requirements. Furthermore, the evidence suggests that cachexia is related to abnormal changes in host intermediary metabolism induced by host-tumor interactions, and endogenous peptides such as TNF may be important mediators. The role of nutritional therapy in cancer patients remains to be defined. Clearly, patients with severe malnutrition benefit from nutritional intervention. However, the benefit of nutritional therapy in less severe cases of malnutrition as an adjuvant to oncologic therapy has yet to be established. REFERENCES 1. Aderka D, Fisher S, Levo Y, et al: Cachectin/tumour necrosis factor production by cancer patients. Lancet 2:1190, 1985 2. Asher A, Mule JJ, Reichert CM, et al: Studies on the anti-tumor efficacy of systematically administered recombinant tumor necrosis factor against several murine tumors in vivo. J Immunol 138:963-974, 1987 3. Balkwill F, Burke F, Talbot D, et al: Evidence for tumour necrosis factor/cachectin production in cancer. Lancet 1:1229-1232, 1987 4. Bennegard K, Eden E, Eisman L, et al: Metabolic response of whole body and peripheral tissues to enteral nutrition in weight losing cancer and noncancer patients. Gastroenterology 85:92-99, 1983 5. Beutler B, Cerami A: Cachectin and tumor necrosis factor as two sides of the same biological coin. Nature 320:584-588, 1985 6. Beutler B, Milsark IW, Cerami AC: Passive immunization against cachectin/tumor necrosis factor protects mice from lethal effect of endotoxin. Science 229:1925-1937, 1985
SUPPORT IN GASTROINTESTINAL CANCER
535
7. Burt ME, Brennan MF: Nutritional support of the patient with esophageal cancer. Semin Oncol11:127-135, 1984 8. Burt ME, Stein TP, Schwade JG, et al: Whole body protein metabolism in cancer-bearing patients. Cancer 53:1246-1254, 1984 8a. Buzby GP, Mullen JL, Matthews DL, et al: Prognostic nutritional index in gastrointestinal surgery. Am J Surg 139:160-167, 1980 9. Carswell EA, Old LJ, Casse RL, et al: An endotoxin-induced serum factor that causes necrosis of tumors. Proc Natl Acad Sci USA 72:3666-3670, 1975 10. Chlebowski RT: Abnormal glucose metabolism in patients with advanced cancer. In Heber D (moderator): Malnutrition and cancer: Mechanisms and therapy. Nutrition International, 1986 11. Daly JM, Reynolds J, Thom A, et al: Immune and metabolic effects of arginine in the surgical patient. Ann Surg 208:512-523, 1988 12. Dempsey DT, Feurer ID, Know LS, et al: Energy expenditure in malnourished gastrointestinal cancer patients. Cancer 53:1265-1273, 1984 13. DeWys WD, Begg C, Lavin PT, et al: Prognostic effect of weight loss prior to chemotherapy in cancer patients. Am J Med 69:491-497, 1980 14. Donaldson SS: Nutritional support as an adjuvant to radiation therapy. JPEN 8:302-309, 1984 15. Eden E, Edstrom S, Bennegard K, et al: Glucose flux in relation to energy expenditure in malnourished patients with and without cancer during periods of fasting and feeding. Cancer Res 44:1717-1724, 1984 16. Emery PW, Lowvell L, Rennie MJ: Protein synthesis measured in vivo in muscle and liver of cachectic tumor-bearing mice. Cancer Res 44:2779-2784, 1984 17. Gutman A, Thilo E, Biren S: Enzymes of gluconeogenesis in tumor-bearing rats. Isr J Med Sci 5:998-1001, 1969 18. Hammond KD, Balinsky D: Activities of key gluconeogenic enzymes and glycogen synthase in rat and human livers and hepatoma cell cultures. Cancer Res 38:13171322, 1978 19. Heatley RV, Williams R, Lewis M: Preoperative intravenous feeding: A controlled trial. Postgrad Med J 55:541-545, 1979 20. Heber D, Byerley LO, Chlebowski RT: Metabolic abnormalities in the cancer patient. Cancer 55:225-233, 1985 21. Heber D, Chlebowski RT, Ishibashi DE, et al: Abnormalities in glucose and protein metabolism in noncachectic lung cancer patients. Cancer Res 42:4815-4819, 1982 22. Holroyde CP, Reichard GA: Carbohydrate metabolism in cancer cachexia. Cancer Treat Rep 65:55-59, 1981 23. Holroyde CP, Gabuzda G, Putman RC, et al: Altered glucose metabolism in metastatic cancer. Cancer Res 35:3710-3714, 1975 24. Holroyde CP, Skutches CL, Boden G, et al: Glucose metabolism in cancer cachectic patients with colorectal cancer. Cancer Res 44:5910-5913, 1984 25. Jasani B, Donaldson LJ, Ratcliffe JG, et al: Mechanism of impaired glucose tolerance in patients with neoplasia. Br J Cancer 38:286-292, 1978 26. Jeevanandam M, Lowry SF, Horowitz GD, et al: Cancer cachexia and protein metabolism. Lancet 1:1423-1426, 1984 27. Karlberg HI, Kern KA, Fischer JE: Albumin turnover in sarcoma-bearing rats in relation to cancer anorexia. Am J Surg 145:95-101, 1983 28. Kawamura I, Moldawer LL, Keenan RA, et al: Altered amino acid kinetics in rats with progressive tumor growth. Cancer Res 42:824-829, 1982 29. Kern KA, Norton JA: Cancer cachexia. JPEN 12:286-298, 1988 30. Koretz RL: Parenteral nutrition: Is it oncologically logical? I Clin Oncol 2:534-548, 1984 31. Krause R, Humphrey C, Von Meyetfeldt M, et al: A central mechanisms for anorexia in cancer: A hypothesis. Cancer Treat Rep 65(suppl 5):15-23, 1981 32. Lundholm K, Holm G, Schersten T: Insulin resistance in patients with progressive malignant disease. Cancer Res 39:1968-1972, 1979 33. Lundholm K, Edstrom S, Karlberg I, et al: Glucose turnover, gluconeogenesis from glycerol, and estimation of net glucose cycling in cancer patients. Cancer 50:1142-1150, 1982 34. Marks PA, Bishop IS: The glucose metabolism of patients with malignant disease and of normal subjects as studied by means of an intravenous glucose tolerance test. J Clin Invest 36:254-264, 1957
536
JOHN
M.
DALY ET AL.
35. Mathison JC. Wolfson E, Ulevitch RJ: Participation of tumor necrosis factor in the mediation of gram negative bacterial lipopolysaccharide induced injury in rabbit. J Clin Invest 81:1925-1937, 1988 36. Michie HR, Manogue KR, Spriggs DR, et al: Detection of circulating tumor necrosis factor after endotoxin administration. N Engl J Med 318:1481-1486, 1988 37. Moley JF, Morrison SD, Gorschboth CM, et al: Body composition changes in rats with experimental cancer cachexia: Improvement with exogenous insulin. Cancer Res 48:2784-2787, 1988 38. Moody TW, Pert CB, Gazdar AR, et al: High levels of intracellular bombesin characterize human small-cell lung carcinoma. Science 214:1246-1248, 1981 39. Mueller J, Brenne V, Dienst J: Peroperative parenteral feeding in patients with gastrointestinal carcinoma. Lancet 1:68-71, 1982 40. Mullen JL, Buzby GP, Gertner MH, et al: Protein synthesis dynamics in human gastrointestinal malignancies. Surgery 87:331-338, 1980 41. Nixon DW, Moffitt S, Lawson DH, et al: Total parenteral nutrition as an adjunct to chemotherapy of metastatic colorectal cancer. Cancer Treat Rep 65(suppl):121-128, 1981 42. Norton JA, Stein TP, Brennan MF: Whole body protein synthesis and turnover in normal man and malnourished patients with and without known cancer. Ann Surg 194:123128, 1981 43. Norton JA, Shamberger MD, Stein TP, et al: The influence of tumor-bearing on protein metabolism in the rat. J Surg Res 30:456-462, 1981 44. Ota D, Nishioka K, Foulkes M, et al: Nutritional parameters affecting erythrocyte polyamine levels in cancer patients. J Clin Oncol 2: 1157-1164, 1984 45. Peacock JL, Norton JA: Impact of insulin on survival of cachetic tumor-bearing rats. JPEN 12:260-264, 1988 45a. Pettigrew RA, Hill GL: Indicators of surgical risk and clinical judgement. Br J Surg 73:47-51, 1986 46. Roh MS, Ekman L, Jeevanandam M, et al: Gluconeogenesis in tumor-influenced hepatocytes. Surgery 96:427-433, 1984 47. Rohdenberg GS, Bernhard A, Krehbiel 0: Sugar tolerance in cancer. JAMA 72:15281529, 1919 48. Russell D, Shike M, Marliss E, et al: Effects ofTPN and chemotherapy on the metabolic derangements in small cell lung cancer. Cancer Res 44:1706-1711, 1984 49. Schein PS, Kisner D, Haller D, et al: Cachexia of malignancy: Potential role of insulin in institutional management. Cancer 49:2070-2076, 1979 50. Shaw JHF, Wolfe RR: Glucose and urea kinetics in patients with early and advanced gastrointestinal cancer: The response to glucose infusion, parenteral feeding, and surgical resection. Surgery 101:181-191, 1986 51. Shaw JHF, Humberstone DM, Wolfe RR: Energy and protein metabolism in sarcoma patients. Ann Surg 207:283-289, 1988 52. Theologides A: Cancer cachexia. Cancer 43:2004-2012, 1979 53. Tracey KJ, Fong Y, Hesse DG, et al: Anti-cachectinffNF monoclonal antibodies prevent septic shock during lethal bacteremia. Nature 330:662-664, 1987 54. Warren RS, Jeevanandam M, Brennan MF: Protein synthesis in the tumor-influenced hepatocyte. Surgery 98:275-282, 1985 55. Warren RS, Starnes F, Gabrilove JL, et al: The acute metabolic effects of tumor necrosis factor administration in humans. Arch Surg 122:1396-1400, 1987 56. Waterhouse C: Lactate metabolism in patients with cancer. Cancer 33:66-71, 1974 57. Waterhouse C, Jeanpetre N, Keilson J: Gluconeogenesis from alanine in patients with progressive malignant disease. Cancer Res 39:1968-1972, 1979 58. Waterlow JC, Goldend M, Picou J: The measurement of rates of protein turnover, synthesis, and breakdown in man and the effect of nutritional status and surgical injury. Am J Clin Nutr 30:1333-1337, 1977
Address reprint requests to John M. Daly, MD Department of Surgery The University of Pennsylvania School of Medicine 3400 Spruce Street Philadelphia, PA 19104
