Preoperative Identification of the Surgical Cancer Patient in Need of Postoperative Supportive Total Parenteral Nutrition

MICHAEL M. MEGUID, MD, PHD, FACS, AND VICTORIA MEGUID, MD

In the absence of specific therapy, nutrition was the mainstay of medicine in ancient times. Because of the current emphasis on modern treatment modalities in the fight against cancer, the provision of adequate nutrition is frequently overlooked. Because of the inconsistent results obtained from random- ized trials of total parenteral nutrition (TPN) in cancer patients undergoing chemotherapy and radiation therapy, ambivalence about the usefulness of TPN as an adjunct to cancer therapy (particularly as it pertains to surgical patients) is further confused by the lack of appropriate criteria for the use of TPN postoperatively. The incidence of malnutrition in relation to certain cancer types is high. Malnutrition is associated with a higher incidence of both postoperative complications and mortality when compared to the well nourished patient. Consequently, preoperative criteria were developed to identify that group of cancer patients requiring abdominal operation who are at high risk and in whom planned nutritional support should be initiated postoperatively. Use of these criteria provides a rational basis for the use of TPN postoperatively.

Cancer 55258-262, 1985.

N A HILL at Kom Ombo by the banks of the Nile 0 stands the well preserved temple built about 200 BC by the Pharoah Ptolomy 11. Dedicated to two gods, the crocodile-headed Sobek and the falcon hawk- headed Horus, this temple was a center focusing on the craft of medicine. Its murals depict Horus offering fruit to Ptolomy I1 after a successful eye operation (Figs. 1 A and 1B). It is probably the first record of postoperative feeding, a common practice of ancient times' in the ab- sence of specific treatment.

The current established indications for total paren- teral nutrition (TPN) are summarized in Table 1. De- spite considerable controversy regarding the use and clinical efficacy of TPN in the cancer patient,2 data cur- rently exist which strongly suggest that 6 to 10 days of preoperative TPN improves postoperative outcome after elective oncologic operation^,^-^ as summarized by Silbermad in the preceding paper. Mullen et aL7 and Rombeau et al.* have obtained similar results in patient populations with a variety of diseases.

More recently, Warnold and Lundholm9 reported the clinical significance of the correlation between preopera- tive nutritional status and postoperative outcome in 2 15

From the Department of Clinical Nutrition, Division of Surgery, City of Hope National Medical Center, Duane, California.

Supported in part by a Cancer Core Grant #CA 16434, to the City of Hope National Medical Center, Duane, California.

Address for reprints: Michael M. Meguid, MD, PhD, FACS, Depan- ment of Surgery, University Hospital, Upstate Medical Center, 750 E. Adams Street, Syracuse, NY 13210.

patients with a variety of benign diseases. Patients un- derwent major vascular, minor vascular or abdominal operation. Overall complication rates were higher in malnourished patients than in well-nourished patients (48% and 2346, respectively; P < 0.0 l), confirming the observations of others who reported their results pre- v i o u ~ l y . ~ , ~ Malnourished patients spent 29 k 5 days (M f SE) and nourished patients 14 k 1 days in hospi- tal ( P < 0.01).

In the present climate of cost containment, these data seem not to have made a sufficiently strong impact on the practice of surgery in the United States. Because of the inconsistent results obtained from randomized trials of TPN in cancer patients undergoing chemotherapy and radiation therapy, the ongoing controversy has helped to confuse the issue of nutritional support in sur- gical patients as related to benefit and outcome. These randomized trials are elegantly summarized by Burt and Brennan.lo The main criticisms of these studies are: (1) patient populations were heterogeneous and frequently not stratified according to disease stage; (2) the popula- tion samples were small and not stratified according to degree of protein-calorie malnutrition; and (3) many patients had advanced disease in which the primary ther- apy was relatively ineffective, with low, i.e.. <25%, re- sponse rates.

As a consequence of this confusion, TPN is used in an inconsistent fashion in the management of the surgical cancer patient. Usually TPN is given when the patient appears grossly malnourished, when a protracted post-

258

No. 1 CRITERIA FOR POSTOPERATIVE TPN Meguid and Meguid 259

FIGS. I A A N D IB. (A, left) The Pharoah Ptolomy I 1 eating fruit immediately after a successful eye operation under the God Horus' supervision. (B, right) Details of the operating room table depicting surgical instruments (Photos: M. M. Meguid).

operative course associated with prolonged starvation is anticipated, following alarming postoperative weight loss, or when a complication has occurred. More often, it is not used at all because of the lack ofclear criteria which specify the appropriate indications for its use.

We developed criteria" for preoperatively identifying a high-risk group of cancer patients, requiring abdomi- nal operation, in whom planned nutritional support would be initiated postoperatively to prevent further metabolic losses and to support recovery. These criteria were based on the patient's nutritional status as deter- mined at the time of admission and on the length of the inadequate oral nutrient intake period (IONIP) after ad-

mission. A patient was considered malnourished if there was a serum albumin less than 3.5 g/dl plus any two of the following four factors: a recent weight loss greater than 10%; a weight for height less than 10th percentile; mid-arm muscle circumference less than 10th percen- tile; triceps skinfold thickness less than 10th percentile. ** IONIP was defined as the number of days elapsing after admission until the patient was eating 60% of the pre- dicted caloric requirements. Patient data were analyzed according to age, site ofthe primary tumor, disease stage, and whether they underwent a curative or palliative pro- cedure. IONIP was divided into three cohorts: < 7, 7 - 10, and > 10 days.

TABLE 1. Summary of Current Usage and Efficacy of TPN

Therapy Efficacy Disease Outcome

Primary Proven GI fistula

Short bowel syndrome Supportive Proven Regional enteritis

Burns Anorexia nervosa

Radiation enteritis Acute hepatic failure

Supportive Questionable Acute renal failure

Acute pancreatitis

Supportive Suggestive Cancer cachexia

Reduces morbidity and mortality. Increases incidence of

Reduces morbidity and mortality. Increases temporary remission rates Life saving Life saving Reduced postoperative complications Augments survival Improves hepatic function and possibly survival Reverses rise in BUN, K & PO4. Possibly decreases length of

anuric phase and possibly improves survival Improves survival, but does not ameliorate respiratory or renal

failure

spontaneous closure rate.

260 CANCER January 1 Supplement 1985

O' lb ;o 3b 40 50 6b 7b ;o So loo

Vol. 55

70

v) 5 6 0 - F q 2 50-

8 40- 0

8 30-

aJ

COMP = 099 MAL + 6 3 R 2 = 09, p<OOOl

80

-

MALNOURISHED~44%

ABD CARCINOMATOSIS (46) 0 PANCREAS (15) 0

OTHER (14) 0 ( ~ B O O U I N A L S b R C O M b S l

GU-GYN ( 9 4 ) 0

0 COLORECTUM (101) LYMPHO- 0 LIVER-GALLBLADDER ( 6 ) PROLIFERATIVE

(35)

NOURISHED= 56%

FIG. 2. Frequency of mz~..utrition related to disease site and linear correlation between the incidence of malnutrition and complications.

2o t

During a 2-year period, 365 patients with gastrointes- tinal cancer were observed. Forty-four percent suffered predominantly from a marasmic type of malnutrition on admission and prior to operation.

The incidence of malnutrition was related to disease site (Fig. 2), and was greatest in patients with Stage 111 disease (Fig. 3). The incidence of complications rose linearly with the incidence of malnutrition (Fig. 2). The median length of the IONIP within each diagnostic group was directly related to the nutritional state and to

whether a complication occurred (Table 2). Thus, in each diagnostic group, well nourished patients without a complication had the shortest IONIP, while those who were malnourished had a significantly longer IONIP ( P < 0.01). If a complication occurred, the IONIP in both well-nourished and malnourished patients was fur- ther prolonged ( P < 0.01), irrespective of the age, the diagnosis, the operative procedure or the sample size. Furthermore, only a few malnourished patients had an IONIP < 10 days (Fig. 4).

I STAGE (n = 99)

(n = 123)

go/ 17% 9 12% 21% a

(n= 148) b"T", FIG. 3. Incidence of malnutrition related to

disease stage in 365 patients with cancer tabulated according to nutritional status.

No. I CRITERIA FOR POSTOPERATIVE TPN . Meguid and Meguid 26 1

TABLE 2. Length of Inadequate Postoperative Oral Nutrient Intake (Days)

Well nourished Malnourished

Organ system involved N o complication Complication No complication Complication

Oesophagus Gastroduodenum Liver/gallbladder Pancreas Colorectum Lymphatic system AM. carcinomatosis Genitourinary Other

Figures in italic type show number of patients.

Stepwise linear regression was used to predict the ef- fect of nutritional status, diagnosis and age on IONIP. From this analysis, three mutually exclusive groups of patients were identified: Group 1 included the high-risk patients in whom TPN should be started in the immedi- ate postoperative period, and were those of any age with preoperative criteria of ( 1 ) being malnourished or (2) having a diagnosis of carcinoma of the esophagus, the gastroduodenum, the liver -gallbladder, the pancreas, intra-abdominal retroperitoneal sarcoma, or abdominal carcinomatosis. Group 2 were well-nourished patients younger than age 40 years who did not require TPN regardless of their primary carcinoma site or stage. Group 3 included malnourished patients of all ages with malignancies of lymphoproliferative, genitourinary - gynecologic (GU-GYN), and colorectal origin. Patients in this group need to be observed carefully, with the anticipation that IONIP would last from 7 to 10 days. Ifa complication occurs in patients of either Group 2 or 3, supportive TPN should be started without delay.

In summary, using the concept of IONIP, we delin- eated preoperative characteristics of the high-risk pa- tients who had prolonged postoperative periods of func- tional starvation. These preoperative criteria identified those who are candidates for immediate postoperative TPN, thereby providing a rational basis for the use of postoperative parenteral nutrition.

Yamada ef a l l3 randomized 96 high-risk Stage 111 and IV gastric carcinoma patients to either a TPN group or a control group, and examined the relationships between nutritional status, immune reactivity, morbidity, and mortality during the first 3 weeks after operation. The effects of a TPN-5-fluorouracil (FU) combination on nutritional and immune status, on morbidity and mor- tality for 3 weeks after gastrectomy, and on cumulative 3-year survival rate and disease-free interval, were also evaluated.

Cell-mediated immunocompetence was restored in all patients who received postoperative TPN and who were immunoimpaired preoperatively. Improvement of im-

Reprinted with permission from Meguid MM. Debonis D, Meguid V el ul. Nutritional support in cancer. Lancet 1983: 2:230-23 I .

paired cell-mediated immunity was obtained in patients treated with a TPN- 5-FU combination as an adjunct to surgery. TPN increased tolerance for 5-FU, allowing achievement of the chemotherapy target dose; a satisfac- tory 3-year survival rate was obtained. In patients who underwent non-curative gastrectomy there was a signifi- cant difference in the 3-year survival rates (54% for TPN - 5-FU versus 0% for non-TPN- 5-FU; P < 0.05). The authors concluded that TPN during surgical adju- vant chemotherapy led to diminished morbidity in pa- tients undergoing gastrectomy for gastric cancer, and possibly to prolonged survival time. They speculated that the mechanism responsible for the late results asso- ciated with postoperative TPN treatment was an in- creased tolerance for 5-FU which resulted from im- proved nutrition and increased cell-mediated immunity.

90

70

40 lL 10 0

0 NOURISHED = MALNOURIMD * P < 0001

-

7-10 >10

INADEQUATE ORAL NUTRIENT INTAKE PERIOD (DAYS)

FIG. 4. Relation of nutritional status to IONIP.

262 CANCER January I Supplement 1985 VOl. 55

TABLE 3. Relationship Between Doubling Time and Tumor Volume in a Cancer Model+

No. of Doublings 0 10 20 30 40 Volume fim’ 0.001 mm’ I mm’ I cm’ lo00 cm’ No. of Cells I lo00 I Ob 109 10l2 Diameter 10 pmm 0. I24 mm 1.24 mm 1.24 cm 12.4 cm

* The assumption is made that the average solid tumor doubling time

Although this is an encouraging report, the number of patients involved in each treatment arm were small.

Considerable disagreement exists regarding the use and clinical efficacy of TPN in the cancer patient. The proponents argue that TPN given to patients as an ad- junct to aggressive surgical, chemotherapeutic and radi- ation treatment of cancer is associated with (1) few com- plications; (2) no evidence for inappropriate increased tumor growth; (3) diminution and reversal of weight loss; (4) prevention of further malnutrition; ( 5 ) correc- tion of existing specific nutritional deficits; ( 6 ) decreased anergy; and (7) increased feeling of well being. These effects are brought about by meeting the patient’s nu- trient requirements during the altered metabolic state induced by the cancer and by the sequelae of therapy. In contrast, opponents argue that the adjunctive use of TPN in cancer patients has not affected the patient’s course by either prolonging survival, reducing complica- tions, or improving tolerance to therapy.

In the light of this controversy, the impact of postoper- ative nutritional support on outcome for high-risk surgi- cal cancer patients is currently being tested. It is antici- pated that the incidence of morbidity, and maybe even mortality, may be lessened. The obvious remaining question is whether postoperative nutritional support in the high-risk cancer group improves long-term survival. This may not be a pertinent question. One must view cancer as a long-term process; the induction phase may be as long as 25 to 30 years, the cancer in situ phase about 5 to 10 years, the invasive phase between 1 and 5 years, and the cancer dissemination phase about 1 to 5 years- all depending on the tumor type, of course. If one as- sumes that the average solid tumor doubling time is about 100 days, then the average tumor is diagnosed and treated at between 30 and 40 doubling times (Table 3). Thus, although it is frequently and justifiably questioned whether the provision of TPN for a limited time during tumor reduction therapy would influence an inexorable pathologic process initiated months to years before, the

is about 100 days. Hence, the average tumor is diagnosed and treated at about 30 to 40 doubling times.

truly appropriate question is whether TPN can permit effective cancer treatment (relative to survival) in a chronically debilitated patient.

Given that malnutrition is associated with a higher incidence of postoperative complications and higher mortality, then supportive therapy directed at (1 ) meet- ing the needs for nutrients during the stress ofthepostop- erative period; (2) preventing further nutritional deple- tion due to functional starvation or semi-starvation; and (3) treating existing nutritional deficits, would seem a p propriate. To this end, the development of preoperative criteria provides a rational basis for identifying those patients in need of postoperative supportive TPN.

REFERENCES

1. Darby WJ, Ghalioungui P, Grivetti L. Food: The Gift of Osiris.

2. Elliot J. Quaring off over total parenteral nutrition. JAMA 1980; New York: Academic Press, 1977.

243:1610-1616. - 3. Monhissi K. Hornshaw J. Teasdale PR et al Parented nutrition

in carcinima of the esophagus treated by surgery: Nitrogen balance and clinical studies. Br J Slrrg 1977; 64: 125 - 128.

4. Heatley RV, Williams RHP, Lewis MH. Preoperative intrave- nous feeding: A controlled trial. PostgradMed J 1979; 55:541-545.

5. Mueller JM, Brenner U, Dieust et a/ . Preoperative parenteral feeding in patients with gastrointestinal carcinoma. Lance? 1982; i:68- 7 I .

6. Silberman H. The role of preoperative parenteral nutrition in cancer patients. Cancer 1985; 55(Suppl):254-257.

7 . Mullen JL, Buzby GP, Matthews DC et al. Reduction of opera- tive morbidity and mortality by combined preoperative and postoper- ative nutritional support. Ann Surg 1980; 192:604-713.

8. Rombeau JL, Barot LR, Williamson CE ef al. Preoperative total parenteral nutrition and surgical outcome in patients with inflamma- torybowel disease. Am JSurg 1982; 143:139-143.

9. Warnold I, Lundholm K. Clinical significance of preoperative nutrition in 215 cancer patients. Ann Surg 1984; 199:299-305.

10. Burt ME, Brennan MF. Nutritional support of the cancer pa- tient. In: Yarborough MF, Curreri PW, eds. In Surgical Nutrition, vol. 3. New York: Churchill, Livingston, 1981; 77-97.

I I . Meguid MM, Debonis D, Meguid V et al. Nutritional support in cancer. Lancef 1983; ii:230-23 I .

12. McLaren DS, Meguid MM. Nutritional assessment at the cross- roads. J Parent Ent Nutr 1983; 7:(6)575-579.

13. Yamada N, Koyama H, Hioki K et a/. Effect of postoperative total parenteral nutrition (TPN) as an adjunct to gastrectomy for ad- vanced gastric cancer. Br J Surg 1983; 70~267-274.