Reduction of Side Effects of Intravesical Therapy with Bacille Calmette-Guérin by Pentoxifylline? An In Vitro Approach

Transcription

1 S101 Reduction of Side Effects of Intravesical Therapy with Bacille Calmette-Guérin by Pentoxifylline? An In Vitro Approach A. Böhle, 1 A. Thanhäuser, 2 M. Ernst, 2 H.-D. Flad, 2 S. Rüsch-Gerdes, 3 D. Jocham, 1 and A. J. Ulmer 2 1 Department of Urology, Medical University of Lübeck, Lübeck, Germany; and 2 Department of Immunology and Cell Biology and 3 National Reference Center for Mycobacteria, Department of Microbiology, Forschungszentrum Borstel, Germany Immunotherapy with intravesical bacille Calmette-Guérin (BCG) is the treatment of choice against superficial bladder cancer recurrences. However, this therapy is associated with side effects that are considered to be the result of inflammatory cytokines. Since pentoxifylline is known to interfere with the production of cytokines, this drug was tested in vitro with regard to a later clinical application in BCG-treated patients. The cytokine release and the cytotoxicity of interleukin-2 or BCG-stimulated mononuclear cells were analyzed, and the growth of BCG under the influence of pentoxifylline was assayed. The results showed an inhibition of cytokine release of stimulated mononuclear cells. The cytotoxicity of BCG-stimulated mononuclear cells but not of lymphokine-stimulated mononuclear cells against bladder carcinoma cells was significantly inhibited. Restimulation with fresh BCG restored cytotoxicity. Direct coincubation of BCG and pentoxifylline resulted in a reduction of mycobacterial metabolism. From these data, we conclude that the use of pentoxifylline to reduce BCG-related side effects should be tested further in a clinical study. One of the most effective immunotherapies to date and one of the most successful therapies during the last decade for the treatment of carcinoma is intravesical administration of BCG [1, 2] against superficial bladder carcinoma recurrences or against carcinoma in situ. Local BCG instillation not only results in reduction of the recurrence rate [3, 4] in patients with superficial bladder carcinoma but also appears to improve the prognosis for a proportion of these patients [5 7]. On the other hand, this local therapy is associated with significant side effects. Sixty to eighty percent of patients suffer from local symptoms of cystitis, such as dysuria, pollakisuria, low-grade fever, and malaise [8, 9]. These local symptoms, however, result in a low acceptance of BCG by the patient and also by the treating urologist who, with regard to the therapeutic alternatives presented, does not accept the high clinical effectivity at the price of a high rate of local symptoms. From research on the mechanism of action of intravesical BCG therapy, it is known that a local granulomatous reaction of the bladder is induced [10, 11]. An accumulation of lymphocytes and macrophages is demonstrable in the bladder wall after administration of BCG, which persists for a prolonged period [12]. As a sign of the acute local immune reaction in the bladder after BCG instillation, significant increases in urinary Grant support: Deutsche Forschungsgemeinschaft (SFB 367, project C7) and Connaught Laboratories, Canada. Reprints or correspondence: Dr. A. Böhle, Dept. of Urology, Medical University of Lübeck, Ratzeburger Allee 160, D Lübeck, Germany (boehle@medinf.mu-luebeck.de). Clinical Infectious Diseases 2000;31(Suppl3):S by the Infectious Diseases Society of America. All rights reserved /2000/3103S3-0011$03.00 cytokines, such as IL-1, IL-2, IFN-g, and tumor necrosis factor a (TNF-a), have been demonstrated [13 15]. Clinically, a correlation has been observed between a pronounced cytokine secretion and a febrile reaction, together with the occurrence of influenza-like symptoms. Thus, an excessive overflow cytokine secretion could be responsible for the increased rate of side effects after intravesical BCG treatment, whereas a clinically less apparent degree of local inflammatory reaction might be necessary for the induction of the desired granulomatous delayed-type hypersensitivity reaction. Analysis of other groups with controlled endotoxemia has shown that within the complex immune reaction, the monokine TNF-a is responsible for the occurrence of fever and malaise [16]. Preliminary clinical data revealed that pentoxifylline (3.7- dimethyl-1-(5-oxo-hexyl)-xanthin) could reduce these symptoms effectively [16, 17]. Recently, it was shown that pentoxifylline is an effective drug for reducing the TNF-a related immune response in endotoxemia [16] or in OKT3-induced cytokine release associated side effects during renal allograft rejection therapy [18, 19]. Because of these specific effects on the immune system, pentoxifylline appeared to be a suitable drug for the reduction of side effects that occur during intravesical BCG therapy. Using several in vitro approaches, we tried to answer preclinical questions with regard to the feasibility of such adjuvant therapy. Methods Cytokine release of stimulated mononuclear cells. The release of cytokines after stimulation of peripheral blood mononuclear cells (PBMC) with phytohemagglutinin (PHA), as an unspecific T-

2 S102 Böhle et al. CID 2000;31 (Suppl 3) cell stimulus, and with BCG (7.5 mg/ml; Connaught strain, Cytochemia, Ihringen, Germany) was analyzed under the influence of different concentrations of pentoxifylline. PBMC (10 6 /ml) were incubated with BCG and PHA as stimuli. Pentoxifylline was added at a concentration of mg/ml at the beginning of coculture. Measurement of IL-2, IL-6, and TNF-a in the supernatant was performed on day 1, and IFN-g was measured on day 5, when maximal release had been demonstrated after coincubation with BCG [20]. The methods used to detect cytokines in the supernatant have been detailed previously [13]. IL-2 activity was measured with a CTL-6 assay (murine T-lymphoma cells). TNF-a activity was determined by the cytopathic effect of TNF-a on L929 cells (murine fibrous sarcoma cells) [21]. IL-6 release was measured by IL-6 dependent proliferation of murine B9.9-3A4 cells [21]. IFN-g was measured by means of a sandwich ELISA (a generous gift of Dr. Gallati, Hoffman La Roche, Basel, Switzerland) [22]. Analysis of cytotoxicity. Experiments were performed to measure the influence of pentoxifylline on BCG-induced or lymphokine-induced cellular cytotoxicity. Cellular toxicity was measured by means of a recently detailed 3 H-L-methionine cytotoxicity assay [23], in which target cells are labeled with tritium and radioactivity in the supernatant is measured after the killing of these target cells. PBMC were incubated for 7 days with BCG (7.5 mg/ml, Connaught strain) or human IL-2 (800 U/mL; a generous gift from Dr. Mohr, Blutspendedienst Niedersachsen, Springe, Germany) in 96- well round-bottom microtiter plates (Greiner, Nürtingen, Germany). Pentoxifylline was added at different concentrations (0 200 g/ml) during coincubation of BCG and mononuclear cells. Two natural-killer (NK) cell resistant human bladder carcinoma cell lines, BT-A and SBC-7 [20], were used as target cells. The resultant cytotoxicity was compared with the cytotoxicity of BCG-activated killer cells (BAK cells) [23], cytotoxicity of lymphokine-activated killer cells (LAK cells), or spontaneous cytotoxicity of NK cells. Analysis of growth of BCG. In order to examine the possible influence of pentoxifylline on growth and viability of BCG, different concentrations of pentoxifylline were coincubated with BCG and were analyzed in the BACTEC 460 assay (Becton-Dickinson, Heidelberg, Germany) [24, 25]. The 14 C release from metabolizing mycobacteria in this assay compared with that from a control is a good reflection of their growth and viability. Results were read daily, and the final comparison between the control and pentoxifylline-treated mycobacteria was made on day 16. In parallel experiments, BCG was coincubated with pentoxifylline in Lockemann liquid medium (Merck, Darmstadt, Germany). The resulting turbidity of this medium was assessed by a standardized photometric method and correlated to the growth of BCG by means of a standard curve. In order to prepare a graph of these in vitro assays, the mean of triplicates was used, and the standard deviation was!15%. Results Cytokine release of mononuclear cells. Coincubation with PHA, as well as with BCG, lead to a stimulation of IL-2 production. The addition of pentoxifylline (10 mg/ml) resulted in a reduction in production of IL-2, and 50 mg/ml completely blocked the production of IL-2 (figure 1). IFN-g was released after stimulation of PBMC with BCG and PHA and could also be inhibited in a dose-dependent manner by pentoxifylline, at concentrations!20 mg/ml. Concentrations 1100 mg/ml completely blocked IFN-g production. Dose-dependent reduction of TNF-a secretion was also detectable, whereas release of IL- 6 was not inhibited by pentoxifylline concentrations of up to 200 mg/ml. Examinations of cytotoxicity. To assess the effects of a possible systemic application of pentoxifylline, the influence of pentoxifylline on cellular cytotoxicity against bladder carcinoma cells was analyzed. BAK and LAK cells showed significant cytotoxicity against the NK cell resistant bladder carcinoma cell lines BT-A and SBC-7 (figure 2). After addition of pentoxifylline, a dose-dependent reduction of BAK cell cytotoxicity Figure 1. Cytokine release of mononuclear cells. Peripheral blood mononuclear cells (PBMC; 10 6 cells/ml) were incubated with BCG and phytohemagglutinin (PHA) as stimuli. Pentoxifylline was added at concentrations of mg/ml at the beginning of coculture. IL- 2, IL-6, and TNF-a were detected by biological assays of the supernatant on day 1, and IFN-g was measured by ELISA on day 5. Unstimulated PBMC (control) did not show any cytokine release. A significant dose-dependent reduction of PHA- and BCG-induced release of IL-2, IFN-g, and TNF-a was seen after coincubation with pentoxifylline, but no inhibition of the release of IL-6 was noted.

3 CID 2000;31 (Suppl 3) Fewer Side Effects of BCG with Pentoxifylline? S103 Figure 2. Influence of pentoxifylline on cellular cytotoxicity. Cellular toxicity against human bladder carcinoma cell lines (BT-A and SBC-7) [20] was measured by means of a 3 H-L-methionine cytotoxicity assay [23]. Peripheral blood mononuclear cells (PBMC) were incubated for 7 days with BCG (7.5 mg/ml) or human IL-2 (800 U/mL). Pentoxifylline was added at different concentrations (0 200 mg/ml) during coincubation. Incubation of PBMC with BCG resulted in significant BCG-activated killer cell cytotoxicity ( ) which can be inhibited in a dose-dependent manner by pentoxifylline. No influence of pentoxifylline on IL-2 induced lymphokine-activated killer cell cytotoxicity ( ) could be observed. For comparison, spontaneous cytotoxicity of unstimulated PBMC ( ) is shown. could be observed. It is interesting that IL-2 induced LAK-cell cytotoxicity was not influenced, even when pentoxifylline was added in concentrations up to 200 mg/ml. The inhibitory effect of pentoxifylline on BAK cell cytotoxicity was reversible. By washing the cells after 24 h and by restimulating with BCG in a pentoxifylline-free medium, complete cytotoxicity could be reinduced (figure 3). Examinations of the growth of BCG. BCG and pentoxifylline were coincubated, and the resultant growth of BCG was assessed. A pentoxifylline concentration dependent inhibition of growth of BCG was observed (figure 4). Inhibition of the growth was confirmed by a turbidity assay in Lockemann medium (data not shown). Discussion The most common side effects after intravesical BCG therapy are dysuria, pollakisuria, fever (temperature,!38.5 C), chills, and malaise [26 28]. These side effects may be seen as symptoms of local cystitis (dysuria or pollakisuria) and, even more important, as symptoms of a systemic cytokine reaction (fever, chills, or malaise). Similar symptoms of a systemic cytokine release have been noted in association with iv or sc therapy with IL-2 or IFN [29], as well as the cytokine-release syndrome during OKT3 therapy for renal allograft rejection [18, 19]. Because of the well-known intense local cytokine secretion following intravesical BCG instillation [14, 15, 30], an association between these side effects and a cytokine release seemed possible. Since cytokines do not have a direct cytotoxic effect on bladder carcinoma [23, 31] but exert a locally activating function on mononuclear cells of the bladder wall, inhibition of only the systemic overflow cytokine secretion may be a promising therapeutic option against BCG-related side effects. Our examinations were performed to pretherapeutically evaluate in vitro a strategy for reducing the side effects of intravesical BCG immunotherapy with use of pentoxifylline. However, in evaluation of the reduction of side effects of such immunotherapy, a possible influence on the immunotherapeutic effectivity itself must be taken into account [32, 33]. We therefore tested such effects of pentoxifylline in several in vitro experiments. First, it could be shown that pentoxifylline inhibits the release of IL-2, IFN-g, and TNF-a from activated mononuclear cells in a dose-dependent way. Secretion of IL-6 into the supernatant was not inhibited. From these experiments, it could be concluded that pentoxifylline indeed is a potentially useful drug against BCG-induced cytokine-related side effects. Next, we tested the influence of pentoxifylline on cytotoxic effector mechanisms. The cytotoxicity of NK cells [34 36], LAK cells [37, 38], and BAK cells (as recently described in [23] and in the article by Brandau et al. in this supplement [S94 100]) is recognized as a cellular effector mechanism of intravesical BCG therapy. Our experiments demonstrated a Figure 3. Reversibility of pentoxifylline-induced inhibition of cytotoxicity toward bladder carcinoma cell lines BT-A and SBC7 at effector-to-target ratio of 10:1. After exchange of medium and restimulation with fresh BCG ( ), inhibition of BCG-activated killer (BAK) cell cytotoxicity by pentoxifylline was discontinued. Without exchange of medium under the influence of pentoxifylline ( ), BAK cell cytotoxicity was reduced to control levels. Dashed line indicates the control level.

4 S104 Böhle et al. CID 2000;31 (Suppl 3) References Figure 4. Influence of pentoxifylline on the growth of BCG was analyzed with use of the BACTEC 460 assay [24, 25]. Release of 14 C from metabolizing mycobacteria was compared with that from a standard control. Final comparisons between control and pentoxifyllinetreated mycobacteria were made on day 16, and findings are indicated as a growth index (0, no growth; 10, background growth; 999, maximal growth). Coincubation of BCG and pentoxifylline revealed dose-dependent inhibition of BCG. dose-dependent, reversible impairment of BAK cell cytotoxicity by pentoxifylline, whereas no effects on LAK cell cytotoxicity were seen. Together with previously published results [20, 23, 31], these experiments revealed further evidence of the functional difference between LAK and BAK effector cells. Therefore, it could be concluded that continuous therapy with pentoxifylline for patients receiving BCG should be administered with caution, since it might impair the desired antitumorous immune response. The importance of viability and reproducibility of BCG for therapeutic effectiveness is well known from clinical [39], in vitro [23, 40], and animal studies [41]. The results of our coincubation experiments demonstrated inhibition of the growth of BCG by pentoxifylline. Therefore, the simultaneous intravesical application of both drugs might impair the efficacy of BCG. On the other hand, resumption of growth of BCG after removal of pentoxifylline was not tested. Before intravesical pentoxifylline is excluded as a possible therapeutic option, however, such experiments, together with animal studies involving local and systemic application of pentoxifylline, will be necessary. In conclusion, oral application of pentoxifylline could reduce BCG-related side effects effectively. Continuous high-dose therapy, however, would probably not be feasible because of possible impairment of cellular cytotoxicity, which can be predicted from our experiments. Rather, the presented data would suggest the short-term application of pentoxifylline. Finally, only clinical studies are suited to clarify this question. Therefore, our results argue in favor of clinical application of pentoxifylline for reducing side effects of intravesical BCG therapy in a comparative phase III study. 1. Chapman PB, Houghton AN. Non-antibody immunotherapy of cancer. Curr Opin Immunol 1993;5: Hudson ML, Ratliff TL. BCG therapy for superficial bladder cancer: fundamental aspects. In Vivo1991;5: Lamm DL. Long-term results of intravesical therapy for superficial bladder cancer. Urol Clin N Am 1992;19: Lamm DL. Carcinoma in situ. Urol Clin N Am 1992;19: Herr HW. Progression of stage T1 bladder tumors after intravesical bacillus Calmette-Guérin. J Urol 1991;145: Herr HW, Klein EA, Rogatko A. Local BCG failures in superficial bladder cancer: a multivariate analysis of risk factors influencing survival. Eur Urol 1991;19: Herr HW, Laudone VP, Badalament RA, et al. Calmette-Guérin therapy alters the progression of superficial bladder cancer. J Clin Oncol 1988;6: Brosman SA. Bacillus Calmette-Guérin immunotherapy: techniques and results. Urol Clin N Am 1992;19: Lamm DL. Complications of bacillus Calmette-Guérin immunotherapy. Urol Clin N Am 1992;19: Böhle A, Gerdes J, Ulmer AJ, et al. Effects of local bacillus Calmette-Guérin therapy in patients with bladder carcinoma on immunocompetent cells of the bladder wall. J Urol 1990;144: Prescott S, James K, Hargreave TB, et al. Intravesical Evans strain BCG therapy: quantitative immunohistochemical analysis of the immune response within the bladder wall. J Urol 1992;147: Böhle A, Busemann E, Gerdes J, et al. Long-term immunobiological effects of intravesical bacillus Calmette-Guérin against bladder carcinoma recurrences. In: Brown F, Revillard RP, eds. Developments in biological standardization: standardization of the immunopharmacology of natural and synthetic immunomodulators. Basel: Karger Verlag, 1992: Böhle A, Nowc C, Ulmer AJ, et al. Detection of urinary TNF, IL 1, and IL 2 after local BCG immunotherapy for bladder carcinoma. Cytokine 1990; 2: de Boer EC, de Jong WH, Steerenberg PA, et al. Induction of urinary interleukin-1 (IL-1), IL-2, IL-6, and tumor necrosis factor duringintravesical immunotherapy with bacillus Calmette-Guérin in superficial bladder cancer. Cancer Immunol Immunother 1992;34: Haaff EO, Catalona WJ, Ratliff TL. Detection of interleukin-2 in the urine of patients with superficial bladder tumors after treatment withintravesical bacillus Calmette-Guérin. J Urol 1986;136: Zabel P, Wolter DT, Schönharting MM, Schade UF. Oxpentifylline in endotoxaemia. Lancet 1989;30: Zabel P, Schade FU, Schlaak M. Pentoxifyllin ein Synthesehemmer für Tumor-Nekrose-Faktor-a. Immun Infekt 1992;20: Alegre M-L, Gastaldello K, Abramowicz D, et al. Evidence that pentoxifylline reduces anti-cd3 monoclonal antibody induced cytokine release syndrome. Transplant 1991;52: Zabel P, Schönharting MM, Schade UF, Schlaak M. Effects of pentoxifylline in endotoxinemia in human volunteers. Prog Clin Biol Res1991;367: Thanhäuser A, Böhle A, Flad H-D, et al. Induction of bacillus Calmette- Guérin activated killer cells from human peripheral blood mononuclear cells against human bladder carcinoma cell lines in vitro. Cancer Immunol Immunother 1993;37: Feist W, Ulmer AJ, Musehold, et al. Induction of tumor necrosis factor a release by lipopolysaccharide and defined lipopolysaccharide partialstructures. Immunobiol 1989;179: Ennen J, Ernst M, Flad H-D. The effect of interleukin-2 on FC-g receptor function of human monocytes requires specific intercellular interaction. Immunobiol 1988;179: Böhle A, Thanhäuser A, Ulmer AJ, et al. Dissecting the immunobiological

5 CID 2000;31 (Suppl 3) Fewer Side Effects of BCG with Pentoxifylline? S105 effects of bacillus Calmette-Guérin (BCG) in vitro: evidence of a distinct BCG-activated killer (BAK) cell phenomenon. J Urol 1993;150: Rüsch-Gerdes S, Schröder K-H, Fetting C. Untersuchungen mit dem System Bactec 460. Empfindlichkeitsprüfung bei Mycobacterium tuberculosis. Vergleich der randiometrischen mit der konventionellen Methode. Prax Klin Pneumol 1985;39: Rüsch-Gerdes S, Schröder K-H, Finnern J. Untersuchungen mit dem system Bactec Isolierung von Mycobacterium tuberculosis aus sputum: vergleich der radiometrischen mit der konventionellen methode. Prax Klin Pneumol 1987;41: Lamm DL, Blumenstein BA, Crawford ED, et al. A randomized trial of intravesical doxorubicin and immunotherapy with bacillus Calmette- Guérin for transitional-cell carcinoma of the bladder. N Engl J Med 1991; 325: Lamm DL, Blumenstein BA, Crissman JD, et al. Maintenance BCG immunotherapy for recurrent TA, T1 and carcinoma in situ transitional cell carcinoma of the bladder: a randomized prospective Southwest Oncology Group study [abstract]. J Urol 2000;163: Lamm DL, van der Meijden APM, Morales A, et al. Incidence and treatment of complications of bacillus Calmette-Guérin intravesical therapy in superficial bladder cancer. J Urol 1992;147: Atzpodien J, Poliwoda H, Kirchner H. Alpha-interferon and interleukin-2 in renal cell carcinoma. studies in nonhospitalized patients. Semin Oncol 1991;5: Böhle A, Nowc C, Ulmer AJ, et al. Elevations of cytokines, interleukin-1, interleukin-2, and tumor necrosis factor in the urine of patients after intravesical bacillus Calmette-Guérin immunotherapy. J Urol 1990;144: Böhle A, Thanhäuser A, Ulmer AJ, et al. On the mode of action of intravesical bacillus Calmette-Guérin: in-vitro characterization of BCG-activated killer cells. Urol Res 1994;22: de Boer EC, Steerenberg PA, van der Meijden APM, et al. Impaired immune response by isoniazid treatment during intravesical BCG administration in the guinea pig. J Urol 1992;148: Stassar MJJG, Vegt PDJ, Steerenberg PA, et al. Effects of isoniazid (INH) on the BCG-induced local immune response after intravesical BCG therapy for superficial bladder cancer. Urol Res 1994;22: Böhle A, Wang M-H, Flad H-D, Ulmer AJ. In vitro cellular cytotoxicity against human bladder carcinoma cell lines. In: Jocham D, Thüroff JW, Rübben H, eds. Investigative urology 4. Berlin: Springer Verlag, 1991: Herr HW, Bean MA, Whitmore WF Jr. Decreased ability of blood leukocytes from patients with tumors of the urinary bladder to act as stimulator cells in mixed leukocyte culture. Cancer Res 1976;36: Morales A, Ottenhof PC. Clinical application of a whole blood assay for human natural killer (NK) cell activity. Cancer 1983;52: Wang M-H, Chen Y-Q, Gercken J, et al. Specific activation of human peripheral blood g/d T lymphocytes by sonicated antigens of Mycobacterium tuberculosis: role in vitro in killing human bladder carcinoma cell lines. Scand J Immunol 1993;38: Wang M-H, Flad H-D, Böhle A, et al. Cellular cytotoxicity of human natural killer cells and lymphokine-activated killer cells against bladder carcinoma cell lines. Immunol Letters 1991;27: Kelley DR, Ratliff TL, Catalona WJ, et al. Intravesical bacillus Calmette- Guérin therapy for superficial bladder cancer: effect of bacillus Calmette- Guérin viability on treatment results. J Urol 1985;134: Brown CA, Brown IN, Sljivic VS. Suppressed or enhanced antibody responses in vitro after BCG treatment of mice: importance of BCG viability. Immunol 1979;38: Shapiro A, Ratliff TL, Oakley DM, Catalona WJ. Reduction of bladder tumor growth in mice treated with intravesical bacillus Calmette-Guérin and its correlation with bacillus Calmette-Guérin viability and natural killer cell activity. Cancer Res 1983;43: