Gordon P. Buzby

Prognostic nutritional index in gastrointestinal surgery

Based on assessment of 161 nonemergency general surgical patients, a multiparameter index of nutritional status was defined relating the risk of postoperative complications to baseline nutritional status. When applied prospectively to 100 gastrointestinal surgical patients, this index provided an accurate, quantitative estimate of operative risk, permitting rational selection of patients to receive preoperative nutritional support.

The link between nutritional status and clinical outcome: can nutritional intervention modify it?

Most clinicians subjectively feel that malnutrition in surgical patients is associated with poor clinical outcome. This overview provides a chronologic review of studies relating poor nutritional status to increased surgical morbidity. Techniques for identifying surgical patients with clinically important nutritional deficits are discussed. Retrospective and/or non-randomized clinical studies evaluating the efficacy of perioperative forced feeding are reviewed. These data suggest a possible role for preoperative nutritional support of selected malnourished surgical candidates and provide the rationale for a large-scale nutrition-intervention clinical trial.

Study protocol: a randomized clinical trial of total parenteral nutrition in malnourished surgical patients.

CSP #221 is a randomized multiinstitutional clinical trial to assess the efficacy of 10 d of perioperative total parenteral nutrition (TPN) in reducing morbidity and mortality in malnourished patients undergoing intraperitoneal and/or intrathoracic operations. In this paper a detailed protocol for the clinical efficacy trial is presented primarily as a reference document for use in interpretation of the results of the clinical trial. It is also anticipated, however, that review of this protocol may be useful to other investigators planning future clinical nutrition intervention trials.

Energy expenditure in malnourished cancer patients.

It is widely believed that the presence of a malignancy causes increased energy expenditure in the cancer patient. To test this hypothesis, resting energy expenditure (REE) was measured by bedside indirect calorimetry in 200 heterogeneous hospitalized cancer patients. Measured resting energy expenditure (REE-M) was compared with expected energy expenditure (REE-P) as defined by the Harris-Benedict formula. The study population consisted of 77 males and 123 females with a variety of tumor types: 44% with gastrointestinal malignancy, 29% with gynecologic malignancy, and 19% with a malignancy of genitourinary origin. Patients were classified as hypometabolic (REE < 90% of predicted), normometabolic (90–110% of predicted) or hypermetabolic(>110% of predicted). Fifty-nine per cent of patients exhibited aberrant energy expenditure outside the normal range. Thirty-three per cent were hypometabolic (79.2% REE-P), 41% were normometabolic (99.5% REE-P), and 26% were hypermetabolic (121.9% REE-P) (p < 0.001). Aberrations in REE were not due to age, height, weight, sex, nutritional status (% weight loss, visceral protein status), tumor burden (no gross tumor, local, or disseminated disease), or presence of liver metastasis. Hypermetabolic patients had significantly longer duration of disease (p < 0.04) than normometabolic patients (32.8 vs. 12.8 months), indicating that the duration of a malignancy may have a major impact upon energy metabolism. Cancer patients exhibit major aberrations in energy metabolism, but are not uniformly hypermetabolic. Energy expenditure cannot be accurately predicted in cancer patients using standard predictive formulae.

A randomized clinical trial of total parenteral nutrition in malnourished surgical patients: the rationale and impact of previous clinical trials and pilot study on protocol design.

The rationale for a large-scale clinical trial of preoperative total parenteral nutrition (TPN) is described in the context of previous clinical trials that have attempted to demonstrate reduction of operative morbidity with preoperative TPN. Defects in study design or execution potentially compromising the validity of these studies are analyzed. Results of a single-institution pilot study performed during the planning phase of the multiinstitutional preoperative TPN trial are presented. This literature review and pilot study provided the data necessary to permit appropriate design of many critical elements in the protocol for the clinical trial including sample size, eligibility criteria, duration and intensity of treatment regimens, and end-point criteria. The rationale underlying critical decisions in protocol design are presented in detail to allow more meaningful interpretation of the results of the clinical trial.

Host‐tumor interaction and nutrient supply

Adequate parenteral nutritional support improves nutritional status in cancer patients, but its effect on tumor growth remains controversial. Using a transplantable mammary adenocarcinoma in a rat‐TPN model, the relative effect of different exogenous intravenous nutrients on tumor growth and host maintenance was studied. Relative to chow controls, starvation increased host depletion without reducing tumor growth. Adequate carbohydrate calories alone neither improved host maintenance nor stimulated tumor growth, yet adequate amino acids alone did improve host maintenance but also stimulated tumor growth. Adequate amino acids and carbohydrates given simultaneously maximized both host maintenance and tumor growth. In contrast, an isocaloric, isonitrogenous, intravenous diet providing non‐nitrogenous calories as fat promoted host maintenance equivalent to carbohydrate‐based TPN with no tumor stimulation. This apparent differential utilization of fat calories by normal and malignant cells may permit manipulation of the relative benefit of parenteral nutrition to host or to tumor, permitting host repletion without tumor stimulation or alternatively tumor stimulation at appropriate times to increase sensitivity to phase‐specific antineoplastic therapy.

Energy expenditure in malnourished gastrointestinal cancer patients.

Cancer cachexia, a common finding in patients with gastrointestinal (GI) malignancy, is frequently attributed to tumor‐induced aberrations in host energy expenditure. To characterize the frequency and severity of aberrations in energy expenditure in GI cancer patients, and to identify the potential influence of tumor characteristics in this group, the authors measured resting energy expenditure (REE) by indirect calorimetry in 173 patients and compared REE to predicted energy expenditure (PEE) from the Harris‐Benedict formulae based on current body weight. Fifty‐eight percent of patients had abnormal REE (normal REE = ±10% PEE); 36% (62 of 173) were hypometabolic (REE <90% PEE), and 22% (39 of 173) were hypermetabolic (REE >110% PEE). Host and tumor factors were compared between metabolic groups to identify potential determinants of abnormal energy expenditure. Differences between groups cannot be explained by differences in patient age, sex, body size, nutritional status, tumor burden, or duration of disease. Resting energy expenditure does not correlate with percent of weight loss, serum albumin, or duration of disease. Analysis by tumor site reveals patients with pancreatic or hepatobiliary tumors to be predominantly hypometabolic; gastric cancer patients tend to be hypermetabolic, whereas patients with colorectal or esophageal neoplasms are more evenly distributed across metabolic groups, the largest portion being normometabolic (X2 = 20.7, P <0.02). The majority of GI cancer patients have abnormal REE which is unpredictable and not uniformly hypermetabolic. The determinants of these abnormalities do not appear to be age, sex, body size, nutritional status or tumor burden. Primary tumor site is a major determinant of energy expenditure in GI cancer patients. Cancer 53:1265‐1273, 1984.

The efficacy of nutritional assessment and support in cancer surgery.

Malnutrition is common in cancer patients and may be an important determinant of operative morbidity and mortality. To determine whether preoperative nutritional assessment can be used to identify a group of high‐risk patients, and whether preoperative TPN decreases morbidity and mortality in this group, retrospective, nonrandomized review of 159 patients who were subjected to major cancer surgery was performed. All patients underwent preoperative multiparameter assessment. A previously developed and validated nutritional assessment model (Prognostic Nutritional Index) was used to evaluate the probability of operative complications. Based on predicted outcome (PNI), patients were assigned to either a high‐risk or low‐risk group for statistical comparison with actual outcome. The effect of preoperative TPN was then analyzed in both risk groups for determination of efficacy of preoperative nutritional support. Substantial malnutrition was found to exist among patients undergoing major cancer surgery and was closely correlated with subsequent morbidity and mortality. This predictive nutritional assessment model accurately identifies a subset of cancer surgery patients at increased risk of operative morbidity and mortality. In this high risk group (PNI ≥ 40%), preoperative nutritional support significantly reduces operative morbidity (P <0.001) and mortality (P <0.025).

Manipulation of TPN caloric substrate and fatty infiltration of the liver

Abstract Hepatic dysfunction frequently develops in severely malnourished patients receiving total parenteral nutrition (TPN) and may limit caloric delivery delaying nutritional repletion. Since hepatic dysfunction may reflect hepatic steatosis due to glucose delivery in excess of oxidative capacity, partial provision of calories as fat may permit increased caloric delivery without hepatic dysfunction. This hypothesis was evaluated using a protein-depleted rat-TPN model. After 4 weeks of protein depletion, rats were nutritionally repleted using one of four isonitrogenous, isovolemic intravenous diets of varying caloric content and/or composition. After 6 days of nutritional repletion massive hepatomegaly and fatty deposition were noted in animals repleted with regimens supplying all nonnitrogenous calories as fat or as carbohydrate (CHO). Hepatic steatosis was avoided when a “balanced” regimen (75% CHO; 25% fat) or a hypocaloric regimen (total calories decreased by 25%) were used, but hypocaloric TPN produced suboptimal nutritional repletion (serum albumin, transferrin, nitrogen balance, weight gain). All intravenous diets were similar to oral rat chow in their ability to restore liver nitrogen and maintain liver protein synthesis. The use of TPN regimens of “balanced” caloric composition similar to that of oral feedings (75% CHO; 25% fat) may provide a clinical alternative to hypocaloric TPN permitting optimal caloric provision and nutritional repletion without hepatic dysfunction.

Factors associated with a positive reexcision after excisional biopsy for invasive breast cancer

BACKGROUND Breast-conserving therapy followed by adjuvant radiotherapy represents an alternative to mastectomy as a treatment for invasive breast cancer. When excisional biopsy has been performed outside the parent institution, reexcision is often performed, with tumor being identified in 32% to 62% of the subsequent specimens. We analyzed not only the factors associated with a positive reexcision but also those factors associated with final surgical margins that are positive for tumor. METHODS Between 1978 and 1991, 956 female patients with American Joint Committee on Cancer clinical stage I or II breast cancer were treated with breast-conserving therapy where a total of 420 patients underwent reexcision after an initial excisional biopsy. Several factors were analyzed to determine their association with a positive reexcision, the status of the final surgical margin, and the nature of the disease present within the reexcision specimen. RESULTS Factors that correlated with a positive reexcision in both univariate and multivariate analysis were clinical tumor size, method of detection, the pathologic status of the axillary lymph nodes, and the histologic appearance. Those factors associated with finding invasive disease at the time of reexcision were clinical tumor size, clinical presentation, and nodal status. The single factor associated with finding residual in situ disease at the time of reexcision was histologic appearance of the primary tumor. A final positive margin was associated with method of tumor detection, age of the patient, and the presence of axillary lymph node metastases. CONCLUSIONS The most significant factors associated with a positive reexcision are clinical tumor size, method of tumor detection, pathologic nodal status, and histologic appearance. Patients with larger tumors or those that are detected by physical examination, as well as invasive lobular carcinomas, may require a more generous initial resection to achieve negative surgical margins and avoid the likelihood of reexcision.