John Libbey Eurotext

Tumor necrosis factor-alpha (TNF-alpha), a protein cytokine, is a pluripotent agent playing an important role in the pathophysiology of various diseases. Its biological properties contribute significantly to vasodilatation, thrombosis, bone resorption, matrix degradation in cartilage, changes in liver metabolism, cachexia.

TNF-alpha is also believed to play an important role in body defences against infection and malignancy. It is now clear, however, that TNF-alpha not only inhibits the growth of certain tumor cells, but alters the growth, differentiation and metabolism of a variety of cells [1, 2]. As indicated earlier, TNF-alpha is produced by a large number of different cell types and organs under appropriate conditions (bacterial, viral, parasitic infection, trauma, drugs, autoimmune or degenerative diseases, tumors) [1, 2]. TNF-alpha levels are usually low or undetectable in the serum of healthy individuals, and the increase in it's activity has been correlated with a variety of physiological situations. Elevated levels have been detected in the urine of humans after long distance running [1, 3]. An increase in expression of TNF-alpha in adipose tissue of obese humans has been found. The level of the cytoxine can reflect various immune effector mechanisms. Studies demonstrate that the neuroendocrine system interferes with the immune system and they both build a complex network of cross-modulatory signals [1-5].

Little is known about the circadian variability of cytokine concentration in contrast to the endocrine hypothalamic-pituitary-adrenal axis where the circadian rhythms have been clearly described [6, 7]. Diurnal fluctuations of cytokine plasma levels have been shown only in a few studies. In patients with neoplasms, some studies have shown a spontaneous increase in TNF-alpha levels but did not present evidence of a regular pattern [8-11].

The aim of this study was to examine the spontaneous circadian secretion of endogenous
TNF-alpha in patients with colorectal cancer metastases.

From August 1995 to May 1996, 27 patients with metastatic colorectal cancer were registered. There were 15 male and 12 female with a mean age 54.5 years (range 34-70.5) (Table 1). All patients had measurable and histologically proven malignant disease.

Exclusion criteria were as follow: cerebral metastases, age greater than 75 years, WHO performance status greater than 2, history of acute and chronic infection, chronic metabolic disease, angina pectoris or myocardial infarction. Histological proof of metastasis was needed in patients with only a single metastatic lesion and normal tumor markers.

Patients with prior adjuvant or neoadjuvant chemotherapy were eligible for the study if they had had a disease-free period of 8 weeks or longer after completion of treatment.

None of the patients had been treated with prostaglandin inhibitors during the preceding 6 weeks before being investigated. Neither pentoxifylline nor systemic or local corticosteroids had been administered. Eighty percent of the patients were continued to be treated with diuretics, spasmolytics, opioid analgesics (morphine at 4 hours intervals) and parenteral nutrition.

We determined the weight and height of patients, complete blood cell count and levels of serum bilirubin, creatinine, urea, sodium, potassium, chloride, total proteins, alkaline phosphatases, CRP and carcinoembryonic antigen (CEA) and CA 19.9.

The control group included a comparable number of age-matched, healthy subjects. They were non-smokers, and had taken no medication in the preceding two months. They maintained their usual lifestyles throughout. None of the women was pregnant or using oral contraceptives.

All patients were informed about the aim of the study and gave their written informed consent. The trial was accepted by Ethical Committee of the Silesian Medical University.

Preparation of the subjects before blood collection was carefully controlled, and blood collection was performed under standarized conditions to minimize sources of preanalytical variations.

The levels of human TNF-alpha in serum were measured using commercial ELISA-type kits Quantikine HS (R & D Systems Austria). The assay employed the quantitative sandwich enzyme immunoassay technique. The minimum detectable serum dose for Quantikine HS is typically 180 fentograms/ml. The first blood sample was taken at 8 a.m. after overnight fasting. Thereafter, blood samples were collected at 2 p.m., 6 p.m., 10 p.m., 2 a.m. and 8 a.m. The lights were turned off at 10 p.m. after the blood sample collection, and turned on again at 7 a.m. After centrifugation, the samples were stored frozen for later analysis of the TNF-alpha levels.

Individual time series data were first displayed as chronograms.

Values are given as means, SEM (range). Each variable was first analysed for a time effect by a one-way analysis of variance (ANOVA), using data in original units as well as the percentage of individual means to minimize variations between individuals. A p < 0.05 was considered significant. In addition, data were analysed individually and as a group for rhythms by the fit of cosine curves of various periods using the method of least squares [12]. Rhythm detection (rejection of the 0 amplitude assumption) was considered to be statistically significant with p < 0.05.

In healthy controls, TNF-alpha levels were low and varied from 0.0 to 2.7 pg/ml. The mean concentration for these subjects was 0.6 pg/ml. Only 11 subjects showed detectable levels of
TNF-alpha in all collected blood samples. The cosine or least square analysis did not show evidence of a circadian rhythm in the control group.

Figure 1 presents serum levels of endogenous TNF-alpha mesured from 8 a.m to 8 a.m. (next day) in patients with advanced colorectal neoplasm. The mean serum concentrations of TNF-alpha increased significantly during the night hours in every patient.

TNF-alpha reached peak serum level at 2 a.m. (X = 372.8 ± 54.7 pg/ml) and the lowest value at 2 p.m (X = 9.3 ± 2.7 pg/ml). The lowest value for the TNF-alpha concentration noticed at 2 p.m. was 3.26% of the peak value.

After achieving the night peak level, the TNF-alpha concentration declined sharply to 94.3 ± 20.1 pg/ml at 8 a.m. From 2 p.m., the cytokine level gradually increased to X = 149 ± 19.2 pg/ml at 6 p.m. and X = 271 ± 29.8 pg/ml at 10 p.m.

The mean detectable level in patients was 167.9 pg/ml. The cosine or least square analysis revealed the presence of a circadian rhythm for TNF-alpha with the acrophase close to midnight: 0 hour 40 min (p < 0.05) (Figure 2). The amplitude of the circadian rhythm was 175.2 pg/ml.

The regulation of TNF-alpha expression requires considerable attention especially with reference to mechanism of immune cell activation and inhibition. As the effector cells produce the cytokines in response to stimulation, their serum levels can reflect the immune functional status. The circadian rhythmicity of the neuroendocrine system influenced by environmental stimuli such as light and dark cycle is closely associated with the immune system [13, 14]. Serum endogenous TNF-alpha levels in colorectal cancer patients varied widely in different studies [15, 16]. It is not clear if the measurements are entirely repetitive. The correlation between TNF-alpha, the immune status of patients and the clinical stage of neoplastic disease is still under consideration. Generally, the TNF-alpha concentration was measured only once a day. The circadian variation of TNF-alpha secretion has rarely been investigated [17, 18]. The results of our study have shown that the measured TNF-alpha levels exhibit a significant 24 hours flux range. Levels of the cytokine in colorectal cancer patients were lowest during the daytime and highest during the night. There could be several possible independent factors having a significant impact on endogenous TNF-alpha secretion. There is no doubt that the neuroendocrine system, particularly the hypothalamus-pituitary-adrenal axis, is the major regulator of endogenous TNF-alpha secretion. Interaction between circulating corticosteroids and TNF-alpha in blood may be one of the most important processes in the regulation of TNF-alpha production. In the present study, the TNF-alpha serum concentration achieved its peak value at 2 a.m., about 4 hours before the peak cortisol level, and thereafter steadily declined. The fact that gucocorticoids inhibit TNF-alpha production by decreasing the efficiency of TNF-alpha mRNA translation [19] may serve as an explanation of this event.

Another important factor that may help to explain the fluctuation of the endogenous TNF-alpha level in the patients is the possible influence of catecholamines during the day time. It is possible that endogenous adrenaline increased during the morning as a result of increased muscular, joint and mental activity, may contribute to the regulation of TNF-alpha production. Adrenaline inhibits
TNF-alpha production by elevating intracellular levels of cAMP [20].

Diurnal fluctuations of TNF-alpha may also contribute to the circadian fluctuation in tumor blood flow. In recent years, time-dependent fluctuations of tumor blood flow have been reported in animal studies. Results obtained by Hori et al. [21] revealed increased tumor blood flow in rats during the night. These data give rise to speculation that the enhanced blood flux may wash out the excess amount of the cytokine from tumor tissue infiltrated by macrophages.

Moreover, in the previous experience of the authors, circadian fluctuations of endogenous
TNF-alpha were followed by significant, time-dependent variations of soluble p-55 receptors serum concentrations in advanced gastrointestinal cancer patients [22]. Although statistical correlation between those factors was not found, the shapes of both curves exhibited regular patterns, with a delay of the TNF-alpha peak with respect to p-55 receptor peak. These facts may reflect the balance between ligand and its soluble receptors and may modify biological effects of TNF-alpha.

The present study has shown the presence of a circadian rhythm for endogenous TNF-alpha levels in patients with colorectal cancer metastases. The biological consequences of such fluxes have not yet been defined. Diurnal rhythmicity in blood levels of different cytokines was also found in other pathological states [22]. Awareness of the 24 hours rhythmicity of the TNF-alpha serum concentration may suggest the possibility of an optimal timing of medical interventions e.g. immunotherapy of cancer or therapeutic usage of TNF-alpha inhibitors in cachetic cancer patients. Our data may help to achieve a fuller understanding of TNF-alpha biomodulatory effects in malignant diseases and toxicity.

1. Aggarwal B B, Natarajan K. 1996. Tumor necrosis factor. Developments during the last decade. Eur. Cytokine Netw. 7: 93.

2. Beutler B D, Greenwald J D, Hulmes M. 1985. Identity of tumor necrosis factor and the macrophage secreted factor cachectin. Nature 316: 552.

3. Springer H C, Jacobs M, Nain A M. 1992. Enhanced release of cytokines, interleukin-2 receptors, and neopterin after long-distance running. Clin. Immunol. Immunopathol. 63: 188.

4. Basedovsky H O, Sorkin M, Keller M. 1975. Changes in blood levels during the immune response. Proc. Soc. Exp. Biol. Med. 150: 466.

5. Basedovsky H O, Rey A, Sorkin E. 1986. In: Oppenhaim J, Jackobs D, Liss A R eds. Leukocytes and defences. Proceedings of the 17th meeting of the International Leukocytes Culture Conference, Held Jointly White T. 22nd National Meeting reticuloendothelial Society Ithace 1985, New York.

6. Bernardini R, Kamilaris TC, Calogero A E, Johnson EO, Gomez MT, Gold PW, Chrousos G P. 1990. Interactions between tumor necrosis factor-alpha, hypothalamic-corticotropin-releasing hormone and adrenocorticotropic secretion in rats. Endocrinology 126: 2876.

7. Jaattela M, Ilvesmaki V, Voutilainen R, Stenman U H, Saksela E. 1991. Tumor necrosis factor-alpha as a potent inhibitor of adrenocorticotropin induced cortisol production and steroidogenic P 450 enzyme expression in cultured human fetal adrenal cells. Endocrinology 128: 623.

8. Orosz P, Echtenacher B, Falk W, Ruschoff J, Weber D, Mannel DN. 1993. Enhancement of experimental metastasis by tumor necrosis factor. J. Exp. Med. 177: 1391.

9. Hasday J D, Shah E, Lieberman A P. 1990. Macrophage tumor necrosis factor-alpha release in induced by contact with some tumors. J. Immunol. 145: 371.

10. Muc-Wierzgo´n M, Madej K, Baranowski M, Kokot T. 1996. Fluctuation of endogenous TNF-alpha concentration in plasma of advanced cancer patients. J. Biol. Regul. Homeostat. Agents 10: 25.

11. Muc-Wierzgo´n M, Madej K, Baranowski M. 1997. Diurnal rhythmicity of endogenous TNF-alpha concentration in plasma of advanced cancer. Oncol. Repp. 4 (3), 637.

12. Touitou Y, Haus E. 1994. Biologic rhythms in clinical and laboratory medicine. Springer Verlag, Berlin Heidelberg New York, 687.

13. Butler L D, Layman N K, Riedl P E, Cain R L, Shellhaas J, Evans G F, Zuckerman S H. 1989. Neuroendocrine regulation of in vivo cytokine production and effects. In vivo regulatory networks involving the neuroendocrine systems, interleukin-1 and TNF-alpha. J. Neuroimmunol. 24: 143.

14. Maestroni G J. 1993. The immunoneuroendocrine role of melatonin. (Review) J.Pineal Res. 14: 1.

15. Mackensen A, Galanos C, Engelhard R. 1991. Treatment of cancer patients with endotoxin induced release of endogenous cytokines. Pathobiology 59: 264.

16. Buchi K N, Moore J G, Hrushesky W J, Sothern R B, Rubin N H. 1991. Circadian rhythm of cellular proliferation in the human rectal mucosa. Gastroenterology 101: 410.

17. Young M R, Matthews J P, Kanabrocki E L, Sothern R B, Roitman-Johnson B, Scheving L E. 1995. Circadian rhythmometry of serum IL-2, IL-10, TNF-alpha, GM-CSF in man. Chronobiol. Int. 12: 19.

18. Smolensky M H, Halberg F, Sargents F S. 1991. Chronobiology of the life sequence. In: Ogata S L K, Yoshimura H. eds. Advances in Climatic Physiology, Tokyo, 281.

19. Parant M, Le Contel C, Parant F, Chedid L. 1991. Influence of endogenous glucocorticoid on endotoxin-induced production of circulating TNF-alpha. Lymphokine Cytokine Res. 10: 265.

20. Vassali P. 1992. The pathophysiology of tumor necrosis factors. Annu. Rev. Immunol. 10: 411.

21. Hori K, Suzuki M, Tanda S, Saito S, Shinozaki M, Zhang Q H. 1992. Circadian variation of tumor blood flow in rat subcutaneous tumors and its alteration by angiotensin II-induced hypertension Cancer Res. 52: 912.

22. Muc-Wierzgo´n M, Madej K, Baranowski M, Wierzgo´n J. 1996. Is there a correlation between tumor necrosis factor-alpha and soluble p-55 receptor plasma levels in advanced gastrointestinal cancer patients? J. Biol. Homeostat Agents 10: 4.

Circadian rhythmometry of serum endogenous tumor necrosis factor-alpha in patients with colorectal cancer metastases