Total parenteral nutrition (TPN) was not used initially in patients with malignant diseases because of the fear of sepsis from the indwelling superior vena cava catheter in immunosuppressed and myelosuppressed patients and the potential for the concentrated nutrient solutions to stimulate tumor growth.
Upon joining the staff at the University of Texas M.D. Anderson Hospital in 1972, I encountered a group of malnourished cancer patients who were being supported by almost every means available then in an attempt to deliver appropriate oncologic therapy.
Many of these patients were still not candidates for aggressive antineoplastic therapy because the toxic side effects of treatment superimposed on the malnourished state would have resulted in unacceptably high morbidity and mortality rates. Yet these patients' tumors were regionalized and, predictably, would have responded to appropriate therapy if it could have been delivered safely.
In this setting TPN was first used to replenish nutritionally cancer patients who had failed all attempts at enteral nutrition. The risks of stimulating tumor growth and septic complications were acceptable because adequate oncologic therapy was otherwise impossible.
In 1974, we reported 120 cancer patients treated with TPN.1 Because of careful attention to the catheter as a nutritional lifeline rather than as a means of measuring central venous pressure or withdrawing blood, the catheter sepsis rate was only 2.2%.2 No stimulation of tumor growth was noted in any patient.
Unfortunately, the indications for TPN in cancer patients have been liberalized since those early days when it was reserved only for patients with responsive tumors and severe malnutrition. Many randomized trials have failed to show any advantage for TPN in patients receiving chemotherapy, primarily because patients with true protein-calorie malnutrition are not included in these studies.3
In the surgical realm, the Veterans Administration trial identified only one subgroup that benefited from preoperative TPN—patients with severe malnourishment.4 In the absence of malnutrition, TPN does not result in an increase in lean body mass but rather in weight gain as fat mass and water volume, neither of which justifies the expense or possible complications of treatment. Also, catheter sepsis rates have risen dramatically with the advent of double- and triple-lumen catheters, which welcome frequent violations of the integrity of the nutritional delivery system.
TPN candidates today, 25 years later, remain individuals with responsive tumors who cannot tolerate the toxicity of the oncologic therapy because of antecedent malnutrition. Exceptions may be patients with bone marrow transplants or pediatric patients undergoing dose-escalation therapy in whom adequate antineoplastic therapy may be predicated upon maintenance of an initial optimal nutritional state.
John M. Daly, MD, and colleagues5 are largely responsible for showing the efficacy of defined formula diets for nutritional repletion of malnourished cancer patients who have a functional gastrointestinal tract. Their work with the amino acid arginine as an immunomodulator is particularly noteworthy. In this issue of CA, Dr. Daly's group reviews the state-of-the-art of nutritional support for cancer patients.6
The amino acid glutamine also is finding ever-increasing nutritional and metabolic uses. As a precursor to glutathione, it can reduce the toxic side effects of several chemotherapeutic agents on the gastrointestinal tract without decreasing the cytotoxic properties of the agents on the malignancy.7 Similarly, the acute effects of radiation therapy on the gastrointestinal mucosa can be prevented if glutamine is used before radiation or reversed if it is used afterward.8,9 Glutamine is abundant in most enteral formulas but is not yet available for parenteral use in the United States.
The fear that nutritional repletion will stimulate tumor growth comes from studies of animal tumor models. Such data are not directly applicable to human beings, however, because the time from tumor implantation to death of the animal may be only 7 to 8 weeks, and the tumor may constitute 30% to 40% of the animal's body weight. A week to 10 days of nutritional repletion in an animal would affect tumor growth dramatically, whereas tumor doubling times in human beings are much longer, and 10 days of nutritional repletion probably would be inconsequential.
Nevertheless, Daly and colleagues10 have shown that cell-cycle specific chemotherapeutic agents have a much better cytotoxic effect in tumors stimulated to grow by nutritional repletion than occurs in their protein-depleted counterparts. Evidence for tumor growth stimulation by TPN in human beings is circumstantial, such as an increase in tumor DNA synthesis11 and an increase in red blood cell poly amine levels.12 Many thousands of cancer patients worldwide have been nutritionally replenished by both enteral and parenteral nutrition, however, without the scientific community noting tumor growth stimulation to be a clinical problem. If it were, as Dr. Daly's team has shown in animals, a therapeutic advantage for human beings might be realized.