Relation of pro- and anti-inflammatory cytokines and the production of nitric oxide in patients receiving high-dose immunotherapy with interleukin-2

ARTICLE

INTRODUCTION

Within the past 15 years immunotherapy of malignancies, involving by administration of various cytokines, has become an important treatment option [1]. In particular, interleukin-2 (IL-2), a 15-kDa glycoprotein released by activated T lymphocytes, has proven its efficacy in the treatment of different tumor entities [2]. Although several side effects of IL-2 treatment have been described [2], the exact mechanism underlying the most common adverse effect, the capillary (vascular) leak syndrome, remains unclear. Capillary leak is defined as an increase in capillary permeability, thus leading to fluid and colloidal loss into the interstitium of visceral organs and soft tissues and ­ in severe cases ­ to fulminant respiratory and renal insufficiency [3-5].

Recently, we demonstrated that the endothelium may play an active role in the development of vascular leak [6]: IL-2 and consecutively released cytokines such as tumor necrosis factor alpha (TNF-alpha) or interferon gamma (IFN-gamma) were shown to activate the endothelium as indicated e.g. by significant increases in endothelin-1 (ET-1) and the circulating endothelial leukocyte adhesion molecule-1 (cELAM-1).

The systemic effects of IL-2 include tachycardia, decrease in blood pressure, fever, transient deterioration of renal and hepatic function, and respiratory failure [3-5, 7], similar to the manifestations of sepsis or the so called systemic inflammatory response syndrome [8, 9].

It has been demonstrated that nitric oxide (NO), expressed in large amounts in the sepsis syndrome, is one of the most potent vasodilators known [10, 11]. NO itself has a short half-life in vivo and is rapidly converted into nitrite and nitrate, its stable metabolites. In the systemic circulation, nearly all nitrite is converted into nitrate by oxyhaemoglobin; therefore, serum or plasma levels of nitrate are frequently used as an indirect parameter for the production of NO [12]. Once established as being an important mediator in a variety of diseases [13], the role of NO in immunotherapy-associated hypotension had been investigated. Not unexpectedly, a significant increase in plasma nitrate and nitrite levels accompanying treatment with IL-2 could be demonstrated in animals [14] as well as in man [15]. This increase in NO was through to be responsible for the haemodynamic compromising effects, as well as for the development of the capillary leak syndrome [16]. In support of this view, inhibition of NO synthase (NOS) was shown to result in reduction of vascular leakage [16].

NO production, or more precisely the induction of the inducible form of NOS, is regulated in particular by pro-inflammatory cytokines such as IL-1, TNF-alpha, or IFN-gamma [11]. Anti-inflammatory cytokines such as IL-10 were through to reduce the production of NO [17]. The aim of the present investigation was to establish a relationship between the patterns of pro- and anti-inflammatory cytokines as well as the production of NO in cancer patients receiving continuous intravenous infusion of high-dose IL-2.

METHODS

Patients

Eight patients with advanced malignancies failing to respond to conventional cytostatic chemotherapy (patients' characteristic see Table 1), were enrolled in a prospective pilot study after written informed consent [6]. The study protocol was approved by the local ethics committee of the university hospital. All data were collected during each patient's first cycle of immunotherapy. Data from patient # 3 were determined over 5 consecutive treatment cycles.

Study design

The study protocol has been described previously [6]. Briefly, recombinant human (rh) IL-2 (Eurocetus BV, Amsterdam, The Netherlands) was administered at a dose of 18 MU of per m² body surface area per day for 120 hours as a continuous infusion. rh IL-2 was diluted in distilled water containing 250 mmol/L glucose and 0.3 mmol/l human albumin. The specific activity of rh IL-2 was 18 x 10⁶ IU/mg of protein equivalent to 3 x 10⁶ Cetus units. For laboratory assays, patients' blood samples were collected before the start of the IL-2 infusion, and 24, 48, 72, 96, and 120 hours thereafter. Plasma and serum were obtained by centrifugation of blood for 15 min at 1,500 g at 4° C. All samples were stored at ­ 80° C until analyzed.

Cytokines

The pro-inflammatory cytokines IL-6 and IL-8, as well as the anti-inflammatory cytokines IL-10, and the soluble TNF-alpha receptors type I and II (sTNFr-I, sTNFr-II) were determined in serum with commercially available enzyme immuno assays (R&D; Research and Diagnostic Systems, Minneapolis, MN, USA) according to instructions. Reference values for all cytokines (i.e. normal range) were obtained from the manufacturer.

Nitrate

The quantitation of nitrate in patient samples was performed using the Griess assay as followed: plasma samples were transfered into Centrisart I ultrafiltration tubes (Sartorius, Goettingen, Germany) with a molecular weight cut-off at 20,000 Daltons and centrifuged at 2,500 g, at 4° C for 2 hours. The filtrate was immediately removed from the inner tube and prepared for measurement. Stock solutions were prepared freshly: L1 consisted of 0.1 mmol/l flavine adenine dinucleotide (FAD; Sigma, USA), and 2 mmol/l nicotinamide adenine dinucleotide phosphate (NADPH; Sigma) in H₂O; L2 consisted of 2 U/mL nitrate reductase (Boehringer Mannheim, Mannheim, Germany) in H₂O; L3 consisted of 200 mmol/l sodium pyruvate (Sigma) in H₂O; and L4 consisted of 275 U/ml lactate dehydrogenase (LDH; Boehringer Mannheim) in H₂O. One hundred mul of the filtrate were transfered into 400 mul H₂O. Then, 25 mul of L1 and 25 mul of L2 were added and the samples were incubated at 37° C. After 60 minutes of incubation, 25 mul of L3 and 25 mul of L4 were added and incubated for a further 15 min at the same temperature. The samples were transfered onto ice and 25 mul of 25 mmol/l sulfanilamide (Sigma), 25 mul 2.5 M HCl and 25 mul of 0.1% N-(1-naphtyl)ethylenediamine (Sigma) were added. Prior to measurements, the samples were stored at room temperature for 30 min. Absorption was measured at 546 nm with a dual beam Hitachi U-2000 spectrophotometer (Tokyo, Japan). The absorbance data were confirmed by at least three readings of the same sample at different time intervals.

Reference values for nitrate plasma levels were obtained from 25 healthy controls.

Statistical analysis

Results are presented as arithmetic means and standard deviation (± SD). Serum level changes during the whole observation period were calculated for each parameter with the non-parametric Friedman-test. In addition, we calculated the mean values of IL-6, IL-8, IL-10, sTNFR-I and -II, and nitrate daily for all patients. To assess a possible association between the variables over time, we used the rank Spearman correlation. P-levels less than 0.05 were considered statistically significant.

RESULTS

Pro-inflammatory cytokines

IL-6 was slightly elevated before therapy (19.1 ± 9.4 pg/ml; normal range: < 5 pg/ml). However, the treatment with IL-2 led to a continuous and significant increase of IL-6 up to 282.7 ± 90.1 pg/ml on day 5 of therapy (p < 0.001; Figure 1). This increase was reproducible in patient # 3 over 5 consecutive treatment cycles (15.3 ± 11.0 versus 156.1 ± 120.7 pg/ml; p < 0.002). IL-8 serum levels prior to IL-2 infusion were clearly elevated (190 ± 238 pg/ml; normal range: < 31.2 pg/ml), and only moderately increased up to 356 ± 553 pg/ml on day 4 of therapy, followed by a subsequent decrease within the last 24 hours (p < 0.001). These changes were also observed in patient # 3 (p < 0.001; Figure 1).

Anti-inflammatory cytokines

During the observation period, IL-10 (normal range: < 7.8 pg/ml) increased from 1.65 ± 4.0 pg/ml (pt. # 3: 0.0 pg/ml) up to 94.7 ± 61.0 pg/ml on day 5 of therapy (pt. # 3: highest level 68.2 ± 101.9 pg/ml, reached on day 4), however, the increases were not statistically significant (Figure 2). Pretherapeutic serum levels of sTNFR-I (2,035 ± 1,642 pg/ml; pt. # 3: 1,707 ± 636 pg/ml) were found to be slightly elevated compared with the normal range (749 ­ 1,966 pg/ml). IL-2 treatment led to a significant increase, with the highest levels observed within 96 (12,999 ± 4,292 pg/ml) to 120 hours (11,484 ± 4,013 pg/ml in pt. # 3) after the initiation of therapy (p < 0.02; Figure 2). Serum levels of sTNFR-II (5,057 ± 2,817 pg/ml; pt. # 3: 3,820 ± 688 pg/ml) were also above the normal range (1,003 ­ 3,170 pg/ml) and again increased up to 42,435 ± 11,909 pg/ml (pt. # 3: 41,882 ± 13,218 pg/ml) on day five of therapy (p < 0.03; Figure 2).

Nitrate

Plasma nitrate levels, before the initiation of IL-2 therapy were 36.4 ± 23.7 mumol/l, comparable to those of a healthy population (34.9 ± 12.8 mumol/l). During immunotherapy, a continuous increase up to 215.1 ± 84.0 mumol/l was observed (p < 0.05; Figure 3). Plasma nitrate levels over 5 consecutive cycles of IL-2 treatment revealed similar kinetics (28.1 ± 11.3 versus 307.6 ± 64.7 mumol/l; p < 0.05).

Correlation analysis

The correlation between nitrate and the pro-inflammatory cytokine IL-6 was highly significant (r_s = 0.94, p < 0.02), in contrast to the correlation between nitrate and IL-8 (r_s = 0.14, p = 0.80). The correlation between nitrate and the anti-inflammatory cytokine IL-10 (rs = 0.83, p = 0.058) just failed to reach significance, but the correlations between nitrate and sTNFR-I (r_s = 0.94, p < 0.02), and nitrate and sTNFR-II (r_s > 0.99, p < 0.01) were highly significant. Positive correlations were also observed between IL-6 and IL-10 (r_s = 0.94, p < 0.02), IL-6 and sTNFr-I (rs = 0.89, p < 0.04), and IL-6 and sTNFR-II (r^_s = 0.94, p < 0.02). There were no significant correlations between Il-6 and IL-8 (r_s = 0.09, p = 0.92), between IL-8 and sTNFr-I (r_s = 0.43, p = 0.42), or between IL-8 and sTNFr-II (r_s = 0.14, p = 0.80), respectively.

Discussion

In 1992, nitric oxide was named as the "molecule of the year" [18, 19]. Since then, NO has been identified as a biological mediator, modulator, and effector [13] in e.g. sepsis [11], surgical infections [20], peripheral arterial occlusion disease [21], or following liver transplantation [22]. Nevertheless, NO is not only an important factor involved in cellular damage, but is also a physiological regulator and exerts cytotoxic as well as cytoprotective properties.

Nitric oxide is synthesized by three different isoformes of nitric oxide synthase. Two of these NOS are expressed constitutively in vascular endothelial cells (eNOS or type III NOS) and in neurones (nNOS or type I NOS), whereas the third one is inducible (iNOS or type II NOS) in a variety of cells. This induction is mediated by cytokines [11]. Constitutive isoforms of NOS (types I and III) are calcium-dependent. Nitric oxide reacts with soluble adenylate cyclase, which leads to a decrease in intracellular calcium in smooth muscle cells and consequent muscle relaxation [11]. This latter mechanism is believed to be responsible for the observed decrease in blood pressure in patients receiving intravenous IL-2 [23].

The cytokines mainly involved in the induction of type II NOS are TNF-alpha, IFN-gamma, and IL-1 [11, 17, 20]. In addition, IL-6 as well as IL-8 have been demonstrated to correlate significantly with serum nitrate levels [17]. Recently, we demonstrated increases in TNF-alpha and IFN-gamma in patients receiving continuous high-dose infusions of IL-2 [6]. In the present investigation, we were able to show significant changes in levels of IL-6 and IL-8 as a result of the administration of IL-2, as documented by others [24]. In accordance with the literature and consistent with our findings, plasma levels of nitrate, the stable metabolite of NO, significantly increased during the observation period. Whether the observed increases in pro- and anti-inlammatory cytokines and NO really reflect the definite peak levels remains unknown, since the observation period was stopped 120 hours after the initiation of therapy.

However, correlation analyses revealed high significance between the levels of nitrate and IL-6, but not between nitrate and IL-8, probably due to the only moderate increase within the first 96 hours and the decrease in IL-8 levels within the last 24 hours of therapy. Retrospective analyses of the correlation between serum levels of TNF-alpha [6] and plasma levels of nitrate also confirmed a high significance (r_s = 0.94, p < 0.02). These observations, i.e. the alterations in pro- and anti-inflammatory mediators, are in excellent agreement with those demonstrated in patients suffering from severe sepsis [17], although it cannot be ruled out that some of the positive correlations might have been due to chance due to the number of tests performed.

However, these data, taken together, suggest that IL-2 leads to a release of IL-6, IL-8, TNF-alpha and IFN-gamma. The latter pro-inflammatory cytokines (TNF-alpha and IFN-gamma) seem to trigger both the generation of NO as well as (perhaps in combination with IL-6 and IL-8) the release of anti-inflammatory cytokines. It is worth noting that, in contrast to the publication of Groeneveld et al. [17], we were unable to confirm the observation that an increase in anti-inflammatory cytokines leads to decreasing nitrate levels.

As mentioned above, increased NO production was attributed as the cause of transient hypotension [23] and the development of capillary leak syndrome in patients receiving intravenous IL-2 [16]. Therefore, attempts to reduce the induction of iNOS and the consecutive NO production by administration of NOS inhibitors are currently under investigation. Mostly analogs of the natural substrate for NOS, namely the amino acid L-arginine, were administered [14, 16, 24, 26]. It has to be borne in mind, however, that all of the above studies used non-selective inhibitors of the L-arginine NO pathway, which prevent the formation of NO by both cNOS and iNOS [11].

One NOS inhibitor, N^G-monomethyl-L-arginine (L-NMMA), has recently been introduced as treatment for catecholamine refractory septic shock. A dose-dependent increase in mean arterial blood pressure, systemic vascular resistance, pulmonary vascular resistance, central venous pressure, and pulmonary arterial occlusion pressure, and a decrease in cardiac output and heart rate was reported [27, 28]. However, until now the clinical use of L-NMMA in sepsis has failed to reduce mortality. Moreover, severe side effects have been described [29], and an interim analysis of an international sepsis trial showed a significant increase in the mortality rates of patients receiving the NOS inhibitor, thus leading to premature termination of this study [30]. Therefore, and in the light of a most limited understanding of this enormously complex system, the broad use of non-selective NOS inhibitors cannot be generally recommended yet [31].

Previously, we demonstrated a significant increase in ET-1, one of the most potent vasoconstrictors, during IL-2 administration and discussed the possibility of it serving as an endogenous counterregulation against the haemodynamic effects of IL-2 [6]. Reviewing the literature, there are several hints of an interaction between NO and ET-1, more complex than yet assumed. ET-1 was demonstrated to inhibit the expression of iNOS [32, 33]. On the other hand, NO itself has been reported to play an inhibitory role in the regulation of ET-1 [34-37]. Although these interactions warrant further investigation, it is worth noting that ET-1 has also been reported to enhance NO-induced cytotoxicity [38]. The latter finding may be relevant in the present context. As mentioned above, NO exerts direct cytotoxic properties. We propose that immunotherapy using IL-2 leads to high serum levels of IL-6, one pro-inflammatory cytokine closely associated with increasing serum nitrate levels. Increasing levels of circulating IL-6 during IL-2 immunotherapy in tumor-bearing patients have been shown to be associated with clinical response [39]. Moreover, it has been demonstrated that nitric oxide synthesis contributes to IL-2-induced antitumor response in an experimental model [40]. Taking these findings in context, one could assume that the patients responding in the study of Deehan and co-workers were those with the highest serum or plasma nitrate levels, and that ­ at least in part ­ the cytotoxic effects of NO may have contributed to tumor regression. We further hypothesize that the previously observed increase in ET-1 might also be involved.

There are also reports that NO might exert quite different effects on the generation of the capillary leak syndrome than generally discussed, namely an attenuation of IL-2-induced lung injury [41]. This effect was attributed to an inhibition of neutrophil superoxide anion synthesis and adherence to endothelial cells. Moreover, reduction of IL-2 inducible NO to normal levels did not affect the incidence of pulmonary edema in another study [42]. This seems to be of special interest in the context of the hypotheses discussed above. However, it has to be admitted that there are conflicting reports showing that IL-2 therapy-induced NO may well compromise the antitumor effects of IL-2: addition of a NOS inhibitor to IL-2 therapy was shown to augment lymphokine-acitivated killer cell development in vivo [43]. Summarizing these considerations, the interaction between IL-2, NO, and ET-1 warrants definite elucidation.

From today's view, and in the light of the present data it seems to be clear that immunotherapy with (intravenous) IL-2 leads to a pro-inflammatory state associated with an increase of plasma nitrate levels. Compensatory anti-inflammatory responses accompany but do not abrogate this process. The currently available literature does not definitely prove that NO is responsible for the side effects of IL-2 therapy. Whether an inhibition of NO formation by non-selective NOS inhibitors will reduce the incidence of severe side effects without reducing the antitumor effects of IL-2 must be the goal of extended randomized trials. Meanwhile, the broad use of NOS inhibitors to overcome transient hypotension or the development of capillary leak syndrome cannot be recommended.

CONCLUSION

REFERENCES

1. Rosenberg S A, Longo D L, Lotze M T. 1989. Principles and application of biologic therapy. In: DeVita V T, Hellmann S, Rosenberg S A, Eds. Cancer, Principles and Practices of Oncology, ed 3. Philadelphia: Lippincott, 301.

2. Rosenberg S A. (Karnofsky Memorial Lecture). 1992. The immunotherapy and gene therapy of cancer. (Review) J. Clin. Oncol. 10: 180.

3. Lotze M T, Chang A E, Seipp C A, Simpson C, Vetto J T, Rosenberg S A. 1986. High dose recombinant interleukin-2 in the treatment of patients with disseminated cancer. Responses, treatment related morbidity and histologic findings. JAMA, 256: 3117.

4. Rosenstein M, Ettinghausen S E, Rosenberg S A. 1986. Extravasation of intravascular fluid mediated by the systemic administration of recombinant interleukin-2. J. Immunol. 137: 1735.

5. Vial T, Descotes J. 1992. Clinical toxicity of interleukin-2. (Review) Drug. Saf. 7: 417.

6. Locker G J, Kapiotis S, Veitl M, Mader R M, Stoiser B, Kofler J, Sieder A E, Rainer H, Steger G G, Mannhalter C, Wagner O F. 1999. Activation of endothelium by immunotherapy with interleukin-2 in patients with malignant disorders. Br. J. Haematol. 105: 912.

7. Nakagawa K, Miller F N, Sims D E, Lentsch A B, Miyazaki M, Edwards MJ. 1996. Mechanisms of interleukin-2-induced hepatic toxicity. Cancer. Res. 56: 507.

8. Gaynor E R, Vitek L, Sticklin L, Creekmore S P, Ferraro M E, Thomas J X Jr, Fisher S G, Fisher SI. 1988. The hemodynamic effects of treatment with interleukin-2 and lymphokine-activated killer cells. Ann. Intern. Med. 109: 953.

9. Ogibene F P, Rosenberg S A, Lotze M, Skibber J, Parker M M, Shelhamer J H, Parillo J E. 1988. Interleukin-2 administration causes reversible hemodynamic changes and left ventricular dysfunction similar to those seen in septic shock. Chest. 94: 750.

10. Evans T, Carpenter A, Kinderman H, Cohen J. 1993. Evidence of increased nitric oxide production in patients with the sepsis syndrome. Circ. Shock. 41: 77.

11. Parrat J R. 1998. Nitric oxide in sepsis and endotoxaemia. J. Antimicrob. Chemoth. 41: SA 31.

12. Moncada S, Higgs A. 1993. The L-arginine-nitric oxide pathway. (Review) N. Engl. J. Med. 329: 2002.

13. Radomski M W. 1995. Nitric oxide: biological mediator, modulator and effector. (Review) Ann. Med. 27: 321.

14. Orucevic A, Lala PK. 1996. NG-nitro-L-arginine methyl ester, an inhibitor of nitric oxide synthesis, ameliorates interleukin 2-induced capillary leakage and reduces tumour growth in adenocarcinoma-bearing mice. Br. J. Cancer. 73: 189.

15. Citterio G, Pellegatta F, Lucca G D, Fragasso G, Scaglietti U, Pini D, Fortis C, Tresoldi M, Rugarli C. 1996. Plasma nitrate plus nitrite changes during continuous intravenous infusion interleukin-2. Br. J. Cancer. 74: 1297.

16. Orucevic A, Hearn S, Lala P K. 1997. The role of active inducible nitric oxide synthase expression in the pathogenesis of capillary leak syndrome resulting from interleukin-2 therapy in mice. Lab. Invest. 76: 53.

17. Groeneveld P H, Kwappenberg K M, Langermans J A, Nibbering P H, Curtis L. 1997. Relation between pro- and anti-inflammatory cytokines and the production of nitric oxide (NO) in severe sepsis. Cytokine 9: 138.

18. Koshland D E Jr. 1992. The molecule of the year (editorial). Science 258: 1861.

19. Culotta E, Koshland D E Jr. 1992. NO news is good news. (Review) Science 258: 1862.

20. Johnson M L, Billiar T R. 1998. Roles of nitric oxide in surgical infection and sepsis. World. J. Surg. 22: 187.

21. Boger R H, Bode-Boger S M, Thiele W, Junker W, Alexander K, Frolich J C. 1997. Biochemical evidence for impaired nitric oxide synthesis in patients with peripheral arterial occlusive disease. Circulation 95: 2068.

22. Ioannidis I, Hellinger A, Dehmlow C, Rauen U, Erhard J, Eigler FW, De Groot H. 1995. Evidence for increased nitric oxide production after liver transplantation in humans. Transplantation 59: 1293.

23. Shahidi H, Kilbourn R G. 1998. The role of nitric oxide in interleukin-2 therapy induced hypotension. Cancer Metastasis Rev. 17: 119.

24. Weidmann E, Bergmann L, Stock J, Kirsten R, Mitrou P S. 1992. Rapid cytokine release in cancer patients treated with interleukin-2. J. Immunother. 12: 123.

25. Kilbourn R G, Fonseca G A, Griffith O W, Ewer M, Price K, Striegel A, Jones E, Logothetis C J. 1995. NG-methyl-L-arginine, an inhibitor of nitric oxide synthase, reverses interleukin-2 induced hypotension. Crit. Care. Med. 23: 1018.

26. Samlowski W E, Yim C Y, McGregor J R, Kwon O D, Gonzales S, Hibbs J B Jr. 1995. Effectiveness and toxicity of protracted nitric oxide synthesis inhibition during IL-2 treatment of mice. J. Immunother. Emphasis Tumor Immunol. 18: 166.

27. Petros A, Bennett D, Vallance P. 1992. Effect on nitric synthase inhibitors on hypotension in patients with septic shock. Lancet 328: 1557.

28. Petros A, Lamb G, Leone A, Moncada S, Bennett D, Vallance P. 1994. Effects of a nitric oxide synthase inhibitor in humans with septic shock. Cardiovasc. Res. 28: 34.

29. Locker G J, Burgmann H, Staudinger T, Knapp S, Laczika K F, Frass M. 1996. Fatal cardiac side effects of L-NMMA? (letter). Crit. Care. Med. 24: 1930.

30. Nasraway S A Jr. 1999. Sepsis research: we must change course. Crit. Care. Med. 27: 427.

31. Fink MP. 1996. Just say no to NO, yes? (editorial). Crit. Care. Med. 24: 1772.

32. Markewitz BA, Michael JR, Kohan DE. 1997. Endothelin-1 inhibits the expression of inducible nitric oxide synthase. Am. J. Physiol. 272: 1078.

33. Ikeda U, Yamamoto K, Maeda Y, Shimpo M, Kanbe T, Shimada K. 1997. Endothelin-1 inhibits nitric oxide synthesis in vascular smooth muscle cells. Hypertension 29: 65.

34. Kourembanas S, McQuillan L P, Leung G K, Faller D V. 1993. Nitric oxide regulates the expression of vasoconstrictors and growth factors by vascular endothelium under both normoxia and hypoxia. J. Clin. Invest. 92: 99.

35. Richard V, Hogie M, Clozel M, Loffler B M, Thuillez C. 1995. In vivo evidence of an endothelin-induced vasopressor tone after inhibition of nitric oxide synthesis in rats. Circulation 91: 771.

36. Hinson J P, Kapas S, Cameron L A. 1996. Differential effects of endogenous and exogenous nitric oxide on the release of endothelin-1 from the intact perfused rat adrenal gland in situ. FEBS Lett. 379: 7.

37. Ahlborg G, Lundberg JM. 1997. Nitric oxide-endothelin-1 interaction in humans. J. Appl. Physiol. 82: 1593.

38. Nakahashi T, Fukuo K, Inoue T, Morimoto S, Hata S, Yano M, Ogihara T. 1995. Endothelin-1 enhances nitric oxide-induced cytotoxicity in vascular smooth muscle. Hypertension. 25: 744.

39. Deehan D J, Heys S D, Simpson W G, Broom J, Franks C, Eremin O. 1994. In vivo cytokine production and recombinant interleukin-2 immunotherapy: an insight into the possible mechanisms underlying clinical responses. Br. J. Cancer. 69: 1130.

40. Yim C Y, McGregor J R, Kwon O D, Bastian N R, Rees M, Mori M, Hibbs J B Jr, Samlowski W E. 1995. Nitric oxide synthesis contributes to IL-2 induced antitumor responses against intraperitoneal Meth A tumor. J. Immunol. 155: 4382.

41. Bouchier-Hayes D, Abdih H, Kelly C J, Barry M, Redmond H P, Burker P, Tanner A, Bouchier-Hayes D J. 1997. Nitric oxide attenuates interleukin-2 induced lung injury. Br. J. Surg. 84: 540.

42. Leder G H, Oppenheim M, Rosenstein M, Shah N, Hoffmann R, Lotze M T, Beger H G. 1996. Inhibition of nitric oxide synthesis does not improve interleukin-2 mediated antitumor effects in vivo. Eur. Surg. Res. 28: 167.

43. Orucevic A, Lala P K. 1996. Effects of N(G)-Nitro-L-arginine methyl ester, an inhibitor of nitric oxide synthesis, on IL-2-induced LAK cell generation in vivo and in vitro in healthy and tumor-bearing mice. Cell. Immunol. 169: 125.