- Email: [email protected]
Tumor necrosis factor-α and anorexia—Cause or effect?
Tumor
Necrosis
Factor-cx and Anorexia-Cause Nachum
Vaisman
or Effect?
and Talia Hahn
Tumor necrosis factor-a (TNF-a) is a principal cytokine that may induce weight loss in malignancies and certain chronic infections. Short-term caloric deprivation has been found to facilitate in vitro TNF-c~ production, while increased spontaneous production of TNF-a has been found in patients with anorexia nervosa (AN). In the present work, we studied in vitro TNF-cy production in other types of chronic undernutrition and the changes in TNF-a production during the refeeding of patients with AN. Undernutrition was evaluated by calculating fat body mass (FBM) from skinfold measurements and lean body mass (LBM) by total body potassium (TBK) counting. Spontaneous and induced TNF-a production by peripheral blood mononuclear cells (PBMC) was studied in six chronically malnourished patients with no intercurrent infections, seven patients with AN, and 16 age-matched normal healthy subjects. Spontaneous TNF-a production was in the normal range in the chronically undernourished subjects (4.3 ? 1.5 v 5.0 2 1.9 U/mL), but significantly increased in the seven patients with AN (221 5 327 Y 5.0 t 1.9 U/mL, P < .OOOS).During refeeding of patients with AN, TNF-cu production decreased to the normal range concomitantly with weight gain. We concluded that chronic undernutrition, in general, is not always associated with increased TNF-a production and that it still remains to be determined whether TNF-cy plays a primary role in the pathogenesis of AN. Copyright 0 1991 by W.B. Saunders Company
T
HE COMPLEXITY of the interactions between nutrition and infections is well recognized. These interactions are partially mediated through cytokines produced by certain tissues.’ Tumor necrosis factor-a (TNF-a), or cachectin, is a principal cytokine with numerous antitumoral, antiviral, metabolic, and immunologic activities.’ Several lines of evidence suggest that TNF-a may induce weight loss in malignancies and in certain chronic infections.’ It has also been shown that repeated administration of sublethal doses of TNF-(w to rats rapidly induces anorexia, resulting in cachexia and depletion of lipid and protein stores.4 These effects may be attributed to suppression of lipoprotein lipase activity, catabolic effects, and central suppression of food intake.‘.6 Based on the above, it was natural to question whether TNF-a may also play a role in weight loss and cachexia of otherwise normally healthy subjects during energy deprivation. We have studied, therefore, a group of healthy subjects on 6 days of very low caloric diet (100 kcal/d), and found that acute starvation may be associated with facilitated TNF-a production.’ In addition, we found increased spontaneous TNF-a synthesis in patients with anorexia nervosa (AN).’ The present study was designed to investigate whether TNF-(U production is altered in other types of chronic malnutrition and to examine the possibility that TNF-a plays a role in the anorexia and resistance to treatment observed in AN and non-AN patients. MATERIALS
AND METHODS
Patients Seven patients with AN, age 13 to 18 years old, were studied. All patients
met the American
Psychiatric
Association’s
diagnostic
From the Clinical Nutrition Clinic and the Paediattic Research Institute, Kaplan Hospital, Rehovot; and The Jerusalem Medical School. Hebrew Universily,Jerusalem, Israel. Supported by a grantfrom the Milton Rosenblum Foundation, Address reprint requests to Nachum Vaisman, MD, Paediattic Research Institute, Kaplan Hospital, Rehovot 76100, Israel. Copyright 0 1991 by W B. Saunders Company 0026-049519114007-0011$03.OOlO 720
criteria for AN.” Patient characteristics on admission to the hospital are given in Table 1. Dietary management involved behavior modification. According to the behavioral modification protocol, each patient was expected to gain 1.0 to 1.5 kgiwk. An eventual discharge weight was established by the physician. Each patient was seen by a psychiatrist at least once a week. Six chronically malnourished patients with no intercurrent infections were studied. These patients were recruited from our clinical nutrition clinic. All of them were referred for long-standing poor weight gain and were monitored for a few months with no significant weight gain. All subjects had a thorough investigation to exclude reasons such as occult cancer for their underweight. No organic reason was found for the below normal weight other than low calorie intake compared with normal children of their age. During this period, 7-day food intake records were kept, which confirmed poor caloric intake for their age and weight. Nutritional status was assessed by four skinfold measurements and total body potassium determination, as described below. Patient characteristics, including nutritional status, are shown in Table 2. Sixteen age-matched normal healthy subjects served as controls. Informed consent was obtained from all subjects. Blood was drawn from all subjects in the postabsorptive state between 8:00 to 1O:OOAM for routine hematological and biochemical testing, as well as for examination of TNF+ production.
Preparation of Effector Cells Ten milliliters of peripheral blood was drawn into heparinized syringes and subsequently separated by density centrifugation on Ficoll-Hypaque gradients (Pharmacia Fine Chemical, Uppsala. Sweden). Separated peripheral blood mononuclear cells (PBMC) were washed with phosphate-buffered saline (PBS), pH 7.1. and suspended in RPM1 1640 medium, supplemented with antibiotics and 10% heat-inactivated fetal calf serum (RPMI-FCS).
Target Cells The target cells used were the human carcinoma cell line, HeLa (ATCC CCL 2.1). The cells were maintained in RPMI-FCS. This cell line is sensitive to the cytotoxic activity of both the adherent mononuclear phagocytic cells and the nonadherent natural killer and natural cytotoxic lymphocytes, as measured by 4- and 12-hour “Cr release assays (unpublished data).
Spontaneous and Induced Production of TNF-a PBMC were suspended RPMI-FCS. and 200-FL
at a concentration of 5 x 10” cells/mL in aliquots were incubated at 37°C for 16
Metabolism, Vol40, No 7 (July), 1991: pp 720.723
TNF AND ANOREXIA
721
Table 1. Individual Values of Nutritional Status and TNF Production in Patients With AN Studied on Admission Patient NO.
Age lvl
W/H VU)
TEIK (9)
TNF Production(UlmL)Stimulation
Fat (%I
N0lle
LPS
PHA 165.0
1
15.0
74.5
74.6
21.4
115.0
122.0
2
13.5
76.0
69.3
20.8
26.4
113.8
129.3
3
15.0
78.0
21.0
313.0
340.0
314.0
4
15.5
84.0
68.6
20.4
23.6
30.8
22.0
5
14.3
80.0
69.0
21.3
26.4
23.0
146.0
6
18.0
80.3
66.8
21.0
926.0
982.0
859.0
7
14.4
76.3
69.9
23.9
119.8
423.0
30.7
Mean ? SD
15.1 f 1.4
78.4 + 3.2
69.7 r 2.6
21.4 2 1.2
221 + 327
291 r 340
238 ? 290
hours in the absence of inducer or in the presence following inducers: bacterial lipopolysaccharide extract i&a coli (LPS; Makor Chemicals, Jerusalem, Israel) phytohemagglutinin (PHA: Wellcome. Dartford,
of one of the from Escher10 PLgimL, or England) 20
PLglmL.
Quantification
of TNF-CY Activity
Cytotoxicity was quantified as described previously.“’ Twentyfour hours after seeding HeLa cells in 9-mm microwells at 5 x 10’ cells/well, cell-free PBMC supernatants were applied to the cells in serial dilutions in the presence of 40 ug/mL cyclohexamide (CHI; Sigma, St Louis, MO). Sixteen hours later. cell death was quantified by measuring the uptake of neutral red. One TNF-a unit was defined as the concentration at which 50% of the target cells were killed. An internal standard was included in each TNF-(Y bioassay. TNF-(Y assays (patients and controls) were performed in two sessions, and representative cases were repeated in the different bioassays to detect any interassay differences.
Nutritional Assessment Weight and height were recorded and plotted on the growth and development charts of Tanner et al.” Weight as a percentage of ideal weight for height (W/H%) was calculated for each subject. Four skinfold thickness measurements were used to calculate fat body mass (FBM) based on the method of Durnin and Rahaman.” Total body potassium (TBK) was studied as an indicator for fat-free body mass (FFBM).”
difference in hematological and biochemical profiles of AN patients and chronic malnourished subjects (not shown), nor were there differences in daily energy intakes: 800 to 1,300 kcal/d as compared with 1,600 to 2,100 kcal/d in the normal controls. A significant statistical difference in spontaneous production of TNF-o was found between the AN and control patients (P < .0006), as well as between the AN and the undernourished patients (P < .Ol). There were no significant differences between all three groups with respect to stimulated TNF-cx production (Table 4). Seven AN patients were repeatedly studied during the refeeding process. Both spontaneous and induced TNF-(U release decreased gradually during refeeding, approaching normal values. Figure 1 illustrates the decrease in TNF-rx release as a function of time calculated by dividing the TNF-cx level at a certain time during refeeding by the TNF-IX level on admission (before refeeding). The levels of spontaneous TNF-cx production at 3 weeks’ posttreatment were already different from pretreatment levels at P = .06 using the Wilcoxon signed-rank test. Later. these continued to decrease.
DISCUSSION
Statistical Analysis The Kruskal Wallis test was used for nonparametric ANOVA. Simultaneous comparison of pairs of groups was performed by the Wilcoxon test with a Bonferroni correction of the Pvalues. RESULTS
Despite state,
as
their
marked
weight
in Table
indicated
loss
3, there
and
malnourished
was no significant
The facilitated release of TNF-a by stimulated PBMC from short-term, acutely starved subjects has previously been reported.’ These results, together with the observation of increased TNF-o release by nonstimulated PBMC from AN patients,” led us to postulate the possible involvement of TNF-cx in the metabolic changes occurring in malnutritional states. This study investigated whether increased TNF production is, in general, associated with
Table 2. individual Values of Nutritional Status and TNF Production in Chronic Undernutrition
Age Diagnosis
(yr)
W/H (%I
TBK
TNF Production(U/mL)Stimulation
(!3)
Fat (%I
N0fle
LPS
PHA
Intestinal pseudo-obstruction
21
73.3
53.5
26.5
4.1
62.5
139.3
Chronic poor intake
16
74.0
74.0
20.9
6.2
33.1
17.6
7
80.0
23.5
20.0
5.1
3.7
5.5
18
65.5
52.7
24.8
5.5
23.1
248.0
16.5
76.0
71.0
21.0
2.3
121.7
5.1
15.5
72.0
74.0
22.2
2.8
28.3
3.0
15.7 f 4.7
73.5 + 4.8
58.1 * 19.6
22.6 ? 2.5
4.3 f 1.5
Gastroesophageal
reflux
Fanconi anemia Chronic poor intake S/P tracheoeosophageal Mean -t SD
fistula
41.9 -c 44.9
JO + 102
VAISMAN AND HAHN
722
Table 3. Nutritional Status of Subjects W/H 1%)
FBW (%)
TBK (9)
O--O
o--o
Normal controls
A-A
(n = 16)
100.3” + 3.7
27.5” 2 4.3
86.4” t 5.4
AN (n = 7)
78.4b -t 3.2
21.4b 2 1.2
69.7b -+ 2.6
73.5b + 4.8
22.6b 2 2.5
58.1b + 19.6
Non-stimulated LPS PHA
~ I I
Chronic undernutrition (n = 6)
NOTE. Values are mean i SD. Means with the same superscript are not significantly different at the .05 significance level.
undernutrition. Therefore, we compared AN patients with chronically undernourished non-AN patients with the same nutritional parameters. We also examined changes occurring in TNF-(w production in AN patients during therapeutic refeeding. Interestingly, TNF-a production by PBMC from all non-AN, chronically undernourished subjects was normal, whereas TNF-(Y production in AN patients was consistently increased. Furthermore, during the therapeutic refeeding of AN patients, both spontaneous and induced TNF-CYproduction were reduced and approached normal values. This occurred concomitantly with weight gain and improvement in their nutritional status, despite the absence of significant changes in body image and psychological attitude toward eating. Both groups, AN and undernourished patients, were similar in some aspects of their body composition. Weight as a percentage of ideal weight for height, percentage of FBM and lean body mass (LBM) expressed by TBK were similar. Energy intake also did not differ between these groups of patients. Although the nutritional parameters in the AN and non-AN subjects studied were similar at the time of the study, the chronically non-AN undernourished subjects maintained almost constant weight during the period preceding the study, whereas weight tended to fluctuate in the AN patients, probably due to sporadic changes in eating habits. Fluctuations in carolic intake may not have occurred in non-AN chronic undernutrition. We were unable to investigate other differences in metabolic responses of these two groups, such as resting energy expenditure, which was found to be reduced in AN patientsI These results may indicate that a principal metabolic defect in AN lies in an excess production of TNF-a. Refeeding and behavioral modification somehow alter this syndrome with a concomitant suppression of TNF-(-uproduction by an unknown mechanism. On the other hand, these results tempt us to speculate that short-term, acute starvation may sensitize the TNF-o-producing cell to stimulation, while prolonged chronic undernutrition may lead also to spontaneous TNF-a production in certain situations. A
Table 4. Spontaneous and Stimulated TNF Production (U/mL) Stimulation NOM
LPS (10
kg/mL)
Control (n = 16)
X5.0” * 1.9
60” t 35
AN (n = 7)
221b + 327
291” + 340
4.3” a 1.5
42” + 45
PHA (20 wg/mL) 75” + 102 238” 2 290
Chronic undernutrition (n = 6)
70” f 102
NOTE. Values are mean 2 SD. *Means with the same superscript are not significantly different at the .05 significance level.
Weeks of refeeding Fig 1. Changes in spontaneous and stimulated (by LPS or PHA) TNF production during refeeding of patients with AN, expressed as a percentage of TNF levels on admission [mean 2 SD).
maintenance of a constant state of energy deprivation (ie, constantly poor energy intake), may eventually lead to a new metabolic balance or adjustment, resulting in the return to normal TNF-a production. However. sudden changes in energy intake that occur in AN patients may maintain the spontaneous production of TNF-a and the increased responsiveness to stimulation. In this case, the observed decrease of TNF-(Y with therapeutic refeeding may indicate that TNF-(w probably does not play a primary role in AN and is most probably induced by poor and inconsistent energy intake. The relation between TNF-a and nutritional states has been suggested in the past. The cytokine TNF-cr (cachectin) has been implicated as a mediator of cachexia.li The induction of AN by repeated administration of large sublethal doses of human cachectin to rats has been demonstrated.3 Furthermore, the inoculation of TNF-a-secreting tumor cells into nude mice caused more advanced wasting and more rapid death than control tumor cells.” Michie et al found that the metabolic changes induced by TNF-CYare dose-dependent.” In their study, an increase in intravenous infusion of recombinant human TNF-(x from 45 l.&m’/24 h to 108 kg/m’/24 h resulted in a significant decrease in nitrogen intake, thus causing a negative nitrogen balance. Some, although not all. studies report increased serum TNF-a in cachectic cancer patients,‘*.” as well as in other cachectic patients.“’ The metabolic effects of TNF-a can be attributed to different mechanisms. One mechanism may involve an alteration in firing rates of glucose-sensitive neurons in the feeding centers of the lateral hypothalamus.’ Another mechanism may involve “cellular cachexia,” due to reduced lipogenic enzyme transcription, leading subsequently to a depletion of intracellular lipid stores.” The documented effects of TNF-a on lipid metabolism lead us to consider the possible role of that cytokine during prolonged energy deprivation. Low levels of TNF-(w may mediate beneficial effects such as normal tissue remodeling and the destruction of undesirable body cells. However, prolonged overproduction of the cytokine may have deleterious effects resulting in the damage of normal tissue. Indeed, increased TNF-cx production in AN may additively contribute to weight loss, cachexia, and possibly to other metabolic impairments associated with AN.
TNF AND ANOREXIA
723
REFERENCES
1. Grimble RF: Cytokines: Their relevance to nutrition. Eur J Clin Nutr 43:217-230, 1989 2. Beutler B, Cerami A: Cachectin (tumor necrosis factor): A macrophage hormone governing cellular metabolism and inflammatory response. Endocr Rev 9:57-r%, 1988 3. Tracey KJ, Lowry SF, Cerami A: Cachectin: A hormone that triggers acute shock and chronic cachexia. J Infect Dis 157:413-420, 1988 4. Tracey KJ, Wei H, Manogue KR, et al: Cachectin/tumor necrosis factor induces cachexia. anemia and inflammation. J Exp Med 167:1211-1227,1988 5. Plata-Salaman CR, Oomura Y, Kai Y: Tumor necrosis factor and interleukia 18: Suppression of food intake by direct action in the central nervous system. Brain Res 448:106-114,1988 6. Tracey KJ. Vlasarra H, Cerami A: Cachectinitumor necrosis factor. Lancet 1:1122-1126. 1989 7. Vaisman N, Schattner A, Hahn T: Tumor necrosis factor production during starvation. Am J Med 87:115,1989 8. Schattner A, Steinbock M, Tepper R, et al: Tumour necrosis factor production and cell-mediated immunity in anorexia nervosa. Clin Exp Immunol79:62-66.1990 9. American Psychiatric Association Task Force on Nomenclature and Statistics: Diagnostic and Statistical Manual of Mental Disorders (ed 3). Washington, DC, APA. 1980 10. Hahn T, Schattner A. Handzel ZT, et al: Possible role of natural cytotoxic activity in the pathogenesis of AIDS. Clin Immunol Immunopathol50:53-56,1989 11. Tanner JH, Whitehouse RH, Takaishi M: Standards from birth to maturity for height, weight, height velocity and weight velocity. British Children, 1965. Part II. Arch Dis Child 41:613-635, 1966
12. Durnin JV, Rahaman MM: The assessment of the amount of fat in the human body from measurements of skinfold thickness. Br J Nutr 21:681-689. 1967 13. Reba RC, Cheek DB, Leitnaker FC: Body potassium and lean body mass, in Cheek DB (ed): Human Growth. Philadelphia, PA, Saunders, 1963, pp 19-28 14. Vaisman N, Rossi MF, Goldberg E, et al: Energy expenditure and body composition in patients with anorexia nervosa. J Pediatr 113:919-924, 1988 15. Cerami A, Ikeda Y, Le Trang N, et al: Weight loss associated with an endotoxin-induced mediator from peritoneal macrophages: The role of cachectin (tumor necrosis factor). Immunol Lett 11:173-177,1985 16. Oliff A, Defeo-Jones D, Boyer M, et al: Tumor secreting human TNFicachectin induce cachexia in mice. Cell 50:555-563. 1987 17. Michie HR, Sherman ML, Spriggs DR. et al: Chronic TNF infusion causes anorexia but not accelerated nitrogen loss. Ann Surg 209:19-24,1989 18. Balkwill F, Osborne R, Burke F, et al: Evidence for tumor necrosis factoricachectin production in cancer. Lancet 2:12291232,1987 19. Socher SH, Martinez DM, Craig JB, et al: Tumor necrosis factor not detectable in patients with clinical cancer cachexia. JNCI 80:595-598, 1988 20. Aderka D. Fischer S, Levo Y, et al: Cachectihn/tumor necrosis factor production by cancer patients. Lancet 2:1190, 1985 21. Pekala PH, Kawakami M, Angus CW, et al: Selective inhibition of synthesis of enzymes for de novo fatty acid biosynthesis by an endotoxin-induced mediator from exudate cells. Proc Nat1 Acad Sci USA 80:2743-2747,1983
Necrosis
Factor-cx and Anorexia-Cause Nachum
Vaisman
or Effect?
and Talia Hahn
Tumor necrosis factor-a (TNF-a) is a principal cytokine that may induce weight loss in malignancies and certain chronic infections. Short-term caloric deprivation has been found to facilitate in vitro TNF-c~ production, while increased spontaneous production of TNF-a has been found in patients with anorexia nervosa (AN). In the present work, we studied in vitro TNF-cy production in other types of chronic undernutrition and the changes in TNF-a production during the refeeding of patients with AN. Undernutrition was evaluated by calculating fat body mass (FBM) from skinfold measurements and lean body mass (LBM) by total body potassium (TBK) counting. Spontaneous and induced TNF-a production by peripheral blood mononuclear cells (PBMC) was studied in six chronically malnourished patients with no intercurrent infections, seven patients with AN, and 16 age-matched normal healthy subjects. Spontaneous TNF-a production was in the normal range in the chronically undernourished subjects (4.3 ? 1.5 v 5.0 2 1.9 U/mL), but significantly increased in the seven patients with AN (221 5 327 Y 5.0 t 1.9 U/mL, P < .OOOS).During refeeding of patients with AN, TNF-cu production decreased to the normal range concomitantly with weight gain. We concluded that chronic undernutrition, in general, is not always associated with increased TNF-a production and that it still remains to be determined whether TNF-cy plays a primary role in the pathogenesis of AN. Copyright 0 1991 by W.B. Saunders Company
T
HE COMPLEXITY of the interactions between nutrition and infections is well recognized. These interactions are partially mediated through cytokines produced by certain tissues.’ Tumor necrosis factor-a (TNF-a), or cachectin, is a principal cytokine with numerous antitumoral, antiviral, metabolic, and immunologic activities.’ Several lines of evidence suggest that TNF-a may induce weight loss in malignancies and in certain chronic infections.’ It has also been shown that repeated administration of sublethal doses of TNF-(w to rats rapidly induces anorexia, resulting in cachexia and depletion of lipid and protein stores.4 These effects may be attributed to suppression of lipoprotein lipase activity, catabolic effects, and central suppression of food intake.‘.6 Based on the above, it was natural to question whether TNF-a may also play a role in weight loss and cachexia of otherwise normally healthy subjects during energy deprivation. We have studied, therefore, a group of healthy subjects on 6 days of very low caloric diet (100 kcal/d), and found that acute starvation may be associated with facilitated TNF-a production.’ In addition, we found increased spontaneous TNF-a synthesis in patients with anorexia nervosa (AN).’ The present study was designed to investigate whether TNF-(U production is altered in other types of chronic malnutrition and to examine the possibility that TNF-a plays a role in the anorexia and resistance to treatment observed in AN and non-AN patients. MATERIALS
AND METHODS
Patients Seven patients with AN, age 13 to 18 years old, were studied. All patients
met the American
Psychiatric
Association’s
diagnostic
From the Clinical Nutrition Clinic and the Paediattic Research Institute, Kaplan Hospital, Rehovot; and The Jerusalem Medical School. Hebrew Universily,Jerusalem, Israel. Supported by a grantfrom the Milton Rosenblum Foundation, Address reprint requests to Nachum Vaisman, MD, Paediattic Research Institute, Kaplan Hospital, Rehovot 76100, Israel. Copyright 0 1991 by W B. Saunders Company 0026-049519114007-0011$03.OOlO 720
criteria for AN.” Patient characteristics on admission to the hospital are given in Table 1. Dietary management involved behavior modification. According to the behavioral modification protocol, each patient was expected to gain 1.0 to 1.5 kgiwk. An eventual discharge weight was established by the physician. Each patient was seen by a psychiatrist at least once a week. Six chronically malnourished patients with no intercurrent infections were studied. These patients were recruited from our clinical nutrition clinic. All of them were referred for long-standing poor weight gain and were monitored for a few months with no significant weight gain. All subjects had a thorough investigation to exclude reasons such as occult cancer for their underweight. No organic reason was found for the below normal weight other than low calorie intake compared with normal children of their age. During this period, 7-day food intake records were kept, which confirmed poor caloric intake for their age and weight. Nutritional status was assessed by four skinfold measurements and total body potassium determination, as described below. Patient characteristics, including nutritional status, are shown in Table 2. Sixteen age-matched normal healthy subjects served as controls. Informed consent was obtained from all subjects. Blood was drawn from all subjects in the postabsorptive state between 8:00 to 1O:OOAM for routine hematological and biochemical testing, as well as for examination of TNF+ production.
Preparation of Effector Cells Ten milliliters of peripheral blood was drawn into heparinized syringes and subsequently separated by density centrifugation on Ficoll-Hypaque gradients (Pharmacia Fine Chemical, Uppsala. Sweden). Separated peripheral blood mononuclear cells (PBMC) were washed with phosphate-buffered saline (PBS), pH 7.1. and suspended in RPM1 1640 medium, supplemented with antibiotics and 10% heat-inactivated fetal calf serum (RPMI-FCS).
Target Cells The target cells used were the human carcinoma cell line, HeLa (ATCC CCL 2.1). The cells were maintained in RPMI-FCS. This cell line is sensitive to the cytotoxic activity of both the adherent mononuclear phagocytic cells and the nonadherent natural killer and natural cytotoxic lymphocytes, as measured by 4- and 12-hour “Cr release assays (unpublished data).
Spontaneous and Induced Production of TNF-a PBMC were suspended RPMI-FCS. and 200-FL
at a concentration of 5 x 10” cells/mL in aliquots were incubated at 37°C for 16
Metabolism, Vol40, No 7 (July), 1991: pp 720.723
TNF AND ANOREXIA
721
Table 1. Individual Values of Nutritional Status and TNF Production in Patients With AN Studied on Admission Patient NO.
Age lvl
W/H VU)
TEIK (9)
TNF Production(UlmL)Stimulation
Fat (%I
N0lle
LPS
PHA 165.0
1
15.0
74.5
74.6
21.4
115.0
122.0
2
13.5
76.0
69.3
20.8
26.4
113.8
129.3
3
15.0
78.0
21.0
313.0
340.0
314.0
4
15.5
84.0
68.6
20.4
23.6
30.8
22.0
5
14.3
80.0
69.0
21.3
26.4
23.0
146.0
6
18.0
80.3
66.8
21.0
926.0
982.0
859.0
7
14.4
76.3
69.9
23.9
119.8
423.0
30.7
Mean ? SD
15.1 f 1.4
78.4 + 3.2
69.7 r 2.6
21.4 2 1.2
221 + 327
291 r 340
238 ? 290
hours in the absence of inducer or in the presence following inducers: bacterial lipopolysaccharide extract i&a coli (LPS; Makor Chemicals, Jerusalem, Israel) phytohemagglutinin (PHA: Wellcome. Dartford,
of one of the from Escher10 PLgimL, or England) 20
PLglmL.
Quantification
of TNF-CY Activity
Cytotoxicity was quantified as described previously.“’ Twentyfour hours after seeding HeLa cells in 9-mm microwells at 5 x 10’ cells/well, cell-free PBMC supernatants were applied to the cells in serial dilutions in the presence of 40 ug/mL cyclohexamide (CHI; Sigma, St Louis, MO). Sixteen hours later. cell death was quantified by measuring the uptake of neutral red. One TNF-a unit was defined as the concentration at which 50% of the target cells were killed. An internal standard was included in each TNF-(Y bioassay. TNF-(Y assays (patients and controls) were performed in two sessions, and representative cases were repeated in the different bioassays to detect any interassay differences.
Nutritional Assessment Weight and height were recorded and plotted on the growth and development charts of Tanner et al.” Weight as a percentage of ideal weight for height (W/H%) was calculated for each subject. Four skinfold thickness measurements were used to calculate fat body mass (FBM) based on the method of Durnin and Rahaman.” Total body potassium (TBK) was studied as an indicator for fat-free body mass (FFBM).”
difference in hematological and biochemical profiles of AN patients and chronic malnourished subjects (not shown), nor were there differences in daily energy intakes: 800 to 1,300 kcal/d as compared with 1,600 to 2,100 kcal/d in the normal controls. A significant statistical difference in spontaneous production of TNF-o was found between the AN and control patients (P < .0006), as well as between the AN and the undernourished patients (P < .Ol). There were no significant differences between all three groups with respect to stimulated TNF-cx production (Table 4). Seven AN patients were repeatedly studied during the refeeding process. Both spontaneous and induced TNF-(U release decreased gradually during refeeding, approaching normal values. Figure 1 illustrates the decrease in TNF-rx release as a function of time calculated by dividing the TNF-cx level at a certain time during refeeding by the TNF-IX level on admission (before refeeding). The levels of spontaneous TNF-cx production at 3 weeks’ posttreatment were already different from pretreatment levels at P = .06 using the Wilcoxon signed-rank test. Later. these continued to decrease.
DISCUSSION
Statistical Analysis The Kruskal Wallis test was used for nonparametric ANOVA. Simultaneous comparison of pairs of groups was performed by the Wilcoxon test with a Bonferroni correction of the Pvalues. RESULTS
Despite state,
as
their
marked
weight
in Table
indicated
loss
3, there
and
malnourished
was no significant
The facilitated release of TNF-a by stimulated PBMC from short-term, acutely starved subjects has previously been reported.’ These results, together with the observation of increased TNF-o release by nonstimulated PBMC from AN patients,” led us to postulate the possible involvement of TNF-cx in the metabolic changes occurring in malnutritional states. This study investigated whether increased TNF production is, in general, associated with
Table 2. individual Values of Nutritional Status and TNF Production in Chronic Undernutrition
Age Diagnosis
(yr)
W/H (%I
TBK
TNF Production(U/mL)Stimulation
(!3)
Fat (%I
N0fle
LPS
PHA
Intestinal pseudo-obstruction
21
73.3
53.5
26.5
4.1
62.5
139.3
Chronic poor intake
16
74.0
74.0
20.9
6.2
33.1
17.6
7
80.0
23.5
20.0
5.1
3.7
5.5
18
65.5
52.7
24.8
5.5
23.1
248.0
16.5
76.0
71.0
21.0
2.3
121.7
5.1
15.5
72.0
74.0
22.2
2.8
28.3
3.0
15.7 f 4.7
73.5 + 4.8
58.1 * 19.6
22.6 ? 2.5
4.3 f 1.5
Gastroesophageal
reflux
Fanconi anemia Chronic poor intake S/P tracheoeosophageal Mean -t SD
fistula
41.9 -c 44.9
JO + 102
VAISMAN AND HAHN
722
Table 3. Nutritional Status of Subjects W/H 1%)
FBW (%)
TBK (9)
O--O
o--o
Normal controls
A-A
(n = 16)
100.3” + 3.7
27.5” 2 4.3
86.4” t 5.4
AN (n = 7)
78.4b -t 3.2
21.4b 2 1.2
69.7b -+ 2.6
73.5b + 4.8
22.6b 2 2.5
58.1b + 19.6
Non-stimulated LPS PHA
~ I I
Chronic undernutrition (n = 6)
NOTE. Values are mean i SD. Means with the same superscript are not significantly different at the .05 significance level.
undernutrition. Therefore, we compared AN patients with chronically undernourished non-AN patients with the same nutritional parameters. We also examined changes occurring in TNF-(w production in AN patients during therapeutic refeeding. Interestingly, TNF-a production by PBMC from all non-AN, chronically undernourished subjects was normal, whereas TNF-(Y production in AN patients was consistently increased. Furthermore, during the therapeutic refeeding of AN patients, both spontaneous and induced TNF-CYproduction were reduced and approached normal values. This occurred concomitantly with weight gain and improvement in their nutritional status, despite the absence of significant changes in body image and psychological attitude toward eating. Both groups, AN and undernourished patients, were similar in some aspects of their body composition. Weight as a percentage of ideal weight for height, percentage of FBM and lean body mass (LBM) expressed by TBK were similar. Energy intake also did not differ between these groups of patients. Although the nutritional parameters in the AN and non-AN subjects studied were similar at the time of the study, the chronically non-AN undernourished subjects maintained almost constant weight during the period preceding the study, whereas weight tended to fluctuate in the AN patients, probably due to sporadic changes in eating habits. Fluctuations in carolic intake may not have occurred in non-AN chronic undernutrition. We were unable to investigate other differences in metabolic responses of these two groups, such as resting energy expenditure, which was found to be reduced in AN patientsI These results may indicate that a principal metabolic defect in AN lies in an excess production of TNF-a. Refeeding and behavioral modification somehow alter this syndrome with a concomitant suppression of TNF-(-uproduction by an unknown mechanism. On the other hand, these results tempt us to speculate that short-term, acute starvation may sensitize the TNF-o-producing cell to stimulation, while prolonged chronic undernutrition may lead also to spontaneous TNF-a production in certain situations. A
Table 4. Spontaneous and Stimulated TNF Production (U/mL) Stimulation NOM
LPS (10
kg/mL)
Control (n = 16)
X5.0” * 1.9
60” t 35
AN (n = 7)
221b + 327
291” + 340
4.3” a 1.5
42” + 45
PHA (20 wg/mL) 75” + 102 238” 2 290
Chronic undernutrition (n = 6)
70” f 102
NOTE. Values are mean 2 SD. *Means with the same superscript are not significantly different at the .05 significance level.
Weeks of refeeding Fig 1. Changes in spontaneous and stimulated (by LPS or PHA) TNF production during refeeding of patients with AN, expressed as a percentage of TNF levels on admission [mean 2 SD).
maintenance of a constant state of energy deprivation (ie, constantly poor energy intake), may eventually lead to a new metabolic balance or adjustment, resulting in the return to normal TNF-a production. However. sudden changes in energy intake that occur in AN patients may maintain the spontaneous production of TNF-a and the increased responsiveness to stimulation. In this case, the observed decrease of TNF-(Y with therapeutic refeeding may indicate that TNF-(w probably does not play a primary role in AN and is most probably induced by poor and inconsistent energy intake. The relation between TNF-a and nutritional states has been suggested in the past. The cytokine TNF-cr (cachectin) has been implicated as a mediator of cachexia.li The induction of AN by repeated administration of large sublethal doses of human cachectin to rats has been demonstrated.3 Furthermore, the inoculation of TNF-a-secreting tumor cells into nude mice caused more advanced wasting and more rapid death than control tumor cells.” Michie et al found that the metabolic changes induced by TNF-CYare dose-dependent.” In their study, an increase in intravenous infusion of recombinant human TNF-(x from 45 l.&m’/24 h to 108 kg/m’/24 h resulted in a significant decrease in nitrogen intake, thus causing a negative nitrogen balance. Some, although not all. studies report increased serum TNF-a in cachectic cancer patients,‘*.” as well as in other cachectic patients.“’ The metabolic effects of TNF-a can be attributed to different mechanisms. One mechanism may involve an alteration in firing rates of glucose-sensitive neurons in the feeding centers of the lateral hypothalamus.’ Another mechanism may involve “cellular cachexia,” due to reduced lipogenic enzyme transcription, leading subsequently to a depletion of intracellular lipid stores.” The documented effects of TNF-a on lipid metabolism lead us to consider the possible role of that cytokine during prolonged energy deprivation. Low levels of TNF-(w may mediate beneficial effects such as normal tissue remodeling and the destruction of undesirable body cells. However, prolonged overproduction of the cytokine may have deleterious effects resulting in the damage of normal tissue. Indeed, increased TNF-cx production in AN may additively contribute to weight loss, cachexia, and possibly to other metabolic impairments associated with AN.
TNF AND ANOREXIA
723
REFERENCES
1. Grimble RF: Cytokines: Their relevance to nutrition. Eur J Clin Nutr 43:217-230, 1989 2. Beutler B, Cerami A: Cachectin (tumor necrosis factor): A macrophage hormone governing cellular metabolism and inflammatory response. Endocr Rev 9:57-r%, 1988 3. Tracey KJ, Lowry SF, Cerami A: Cachectin: A hormone that triggers acute shock and chronic cachexia. J Infect Dis 157:413-420, 1988 4. Tracey KJ, Wei H, Manogue KR, et al: Cachectin/tumor necrosis factor induces cachexia. anemia and inflammation. J Exp Med 167:1211-1227,1988 5. Plata-Salaman CR, Oomura Y, Kai Y: Tumor necrosis factor and interleukia 18: Suppression of food intake by direct action in the central nervous system. Brain Res 448:106-114,1988 6. Tracey KJ. Vlasarra H, Cerami A: Cachectinitumor necrosis factor. Lancet 1:1122-1126. 1989 7. Vaisman N, Schattner A, Hahn T: Tumor necrosis factor production during starvation. Am J Med 87:115,1989 8. Schattner A, Steinbock M, Tepper R, et al: Tumour necrosis factor production and cell-mediated immunity in anorexia nervosa. Clin Exp Immunol79:62-66.1990 9. American Psychiatric Association Task Force on Nomenclature and Statistics: Diagnostic and Statistical Manual of Mental Disorders (ed 3). Washington, DC, APA. 1980 10. Hahn T, Schattner A. Handzel ZT, et al: Possible role of natural cytotoxic activity in the pathogenesis of AIDS. Clin Immunol Immunopathol50:53-56,1989 11. Tanner JH, Whitehouse RH, Takaishi M: Standards from birth to maturity for height, weight, height velocity and weight velocity. British Children, 1965. Part II. Arch Dis Child 41:613-635, 1966
12. Durnin JV, Rahaman MM: The assessment of the amount of fat in the human body from measurements of skinfold thickness. Br J Nutr 21:681-689. 1967 13. Reba RC, Cheek DB, Leitnaker FC: Body potassium and lean body mass, in Cheek DB (ed): Human Growth. Philadelphia, PA, Saunders, 1963, pp 19-28 14. Vaisman N, Rossi MF, Goldberg E, et al: Energy expenditure and body composition in patients with anorexia nervosa. J Pediatr 113:919-924, 1988 15. Cerami A, Ikeda Y, Le Trang N, et al: Weight loss associated with an endotoxin-induced mediator from peritoneal macrophages: The role of cachectin (tumor necrosis factor). Immunol Lett 11:173-177,1985 16. Oliff A, Defeo-Jones D, Boyer M, et al: Tumor secreting human TNFicachectin induce cachexia in mice. Cell 50:555-563. 1987 17. Michie HR, Sherman ML, Spriggs DR. et al: Chronic TNF infusion causes anorexia but not accelerated nitrogen loss. Ann Surg 209:19-24,1989 18. Balkwill F, Osborne R, Burke F, et al: Evidence for tumor necrosis factoricachectin production in cancer. Lancet 2:12291232,1987 19. Socher SH, Martinez DM, Craig JB, et al: Tumor necrosis factor not detectable in patients with clinical cancer cachexia. JNCI 80:595-598, 1988 20. Aderka D. Fischer S, Levo Y, et al: Cachectihn/tumor necrosis factor production by cancer patients. Lancet 2:1190, 1985 21. Pekala PH, Kawakami M, Angus CW, et al: Selective inhibition of synthesis of enzymes for de novo fatty acid biosynthesis by an endotoxin-induced mediator from exudate cells. Proc Nat1 Acad Sci USA 80:2743-2747,1983