Long-term antitumor necrosis factor antibody therapy in rheumatoid arthritis patients sensitizes the pituitary gland and favors adrenal androgen secretion.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 48, No. 6, June 2003, pp 1504–1512 DOI 10.1002/art.11036 © 2003, American College of Rheumatology Long-Term Anti–Tumor Necrosis Factor Antibody Therapy in Rheumatoid Arthritis Patients Sensitizes the Pituitary Gland and Favors Adrenal Androgen Secretion Rainer H. Straub,1 Georg Pongratz,1 Jürgen Schölmerich,1 Frieder Kees,2 Thomas F. Schaible,3 Christian Antoni,4 Joachim R. Kalden,4 and Hanns-Martin Lorenz4 tion of anti-TNF, which indicates a sensitization of the pituitary gland (not observed for the adrenal gland). During treatment, the ratio of serum cortisol to serum ACTH decreased, which also indicates a sensitization of the pituitary gland (P < 0.001), and which was paralleled by constant cortisol secretion. The adrenal androgen ASD significantly increased relative to its precursor 17(OH)progesterone (P ⴝ 0.013) and relative to cortisol (P ⴝ 0.009), which indicates a normalization of adrenal androgen production. The comparison of patients previously treated with prednisolone and those without previous prednisolone revealed marked differences in the central and adrenal level of this endocrine axis during long-term anti-TNF therapy. Conclusion. Long-term therapy with anti-TNF sensitizes the pituitary gland and improves adrenal androgen secretion in patients who have not previously received prednisolone treatment. These changes are indicative of normalization of the HPA axis and must therefore be considered as evidence of an additional antiinflammatory influence of anti-TNF treatment in patients with RA. Objective. New insights into the role of tumor necrosis factor (TNF) in the pathogenesis of rheumatoid arthritis (RA) have expanded our understanding about the possible mechanisms by which anti-TNF antibody therapy reduces local synovial inflammation. Beyond local effects, anti-TNF treatment may modulate systemic antiinflammatory pathways such as the hypothalamic–pituitary–adrenal (HPA) axis. This longitudinal anti-TNF therapy study was designed to assess these effects in RA patients. Methods. RA patients were given 5 infusions of anti-TNF at weeks 0, 2, 6, 10, and 14, with followup observation until week 16. We measured serum levels of interleukin-6 (IL-6), adrenocorticotropic hormone (ACTH), 17-hydroxyprogesterone (17[OH]progesterone), cortisol, cortisone, androstenedione (ASD), dehydroepiandrosterone (DHEA), and DHEA sulfate in 19 RA patients. Results. Upon treatment with anti-TNF, we observed a fast decrease in the levels of serum IL-6, particularly in RA patients who did not receive parallel prednisolone treatment (P ⴝ 0.043). In these RA patients who had not received prednisolone, the mean serum ACTH levels sharply increased after every injec- After its initial discovery in the early 1990s (1), anti–tumor necrosis factor (anti-TNF) antibody is now widely used as an antiinflammatory drug in patients with rheumatoid arthritis (RA) and in several other inflammatory diseases (2–4). The primary effect of anti-TNF therapy is most probably direct neutralization of proinflammatory TNF, but other antiinflammatory factors may also contribute to the favorable role of anti-TNF therapy. We speculated that the function of the hypothalamic–pituitary–adrenal (HPA) axis may recover under long-term anti-TNF therapy. During acute inflammation, a normal-functioning HPA axis is regulated by several factors (Figure 1). 1) Supported by the Deutsche Forschungsgemeinschaft (Str 511/11-1), and by Centocor, Inc. 1 Rainer H. Straub, MD, Georg Pongratz, MD, Jürgen Schölmerich, MD: University Hospital Regensburg, Regensburg, Germany; 2Frieder Kees, PhD: Institute of Pharmacy, University Regensburg, Regensburg, Germany; 3Thomas F. Schaible, PhD: Centocor Inc., Malvern, Pennsylvania; 4Christian Antoni, MD, Joachim R. Kalden, MD, Hanns-Martin Lorenz, MD: University of Erlangen– Nürnberg, Erlangen, Germany. Address correspondence and reprint requests to Rainer H. Straub, MD, Laboratory of Neuroendocrinoimmunology, Department of Internal Medicine I, University Hospital Regensburg, 93042 Regensburg, Germany. E-mail: email@example.com. Submitted for publication December 9, 2002; accepted in revised form February 27, 2003. 1504 INFLUENCE OF ANTI-TNF TREATMENT IN PATIENTS WITH RA Figure 1. The hypothalamic–pituitary–adrenal axis and the influence of cytokines on adrenal steroidogenesis. A line with an arrow at the end indicates that the respective mediator stimulates the enzyme step (interleukin-6 [IL-6]). A line with a bar at the end demonstrates that the respective mediator inhibits the enzyme step (tumor necrosis factor [TNF]). 3␤HSD ⫽ 3␤-hydroxysteroid dehydrogenase; 11␤HSD I and II ⫽ 11␤-hydroxysteroid dehydrogenase type I and type II; ACTH ⫽ adrenocorticotropic hormone; CRH ⫽ corticotropinreleasing hormone; DHEA ⫽ dehydroepiandrosterone; DHEAS ⫽ DHEA sulfate; DST ⫽ DHEA sulfotransferase; P450c11 ⫽ 11␤hydroxylase; P450c17 ⫽ 17␣-hydroxylase and 17/20-lyase (doubleenzyme step); P450c21 ⫽ 21␣-hydroxylase; P450ssc ⫽ side-chain cleavage enzyme; ST ⫽ DHEA sulfatase; StAR ⫽ steroidogenic acute regulatory protein. Corticotropin-releasing hormone (CRH) stimulates secretion of adrenocorticotropic hormone (ACTH), which stimulates cortisol secretion, and cortisol inhibits the hypothalamus and the pituitary gland by feedback inhibition. 2) A short-term administration of interleukin-6 (IL-6) stimulates the human hypothalamus, the pituitary gland, and the adrenals (5,6). 3) A short-term administration of TNF stimulates the hypothalamus and pitu- 1505 itary gland (7,8), but probably leads to inhibition of the adrenal gland (9) and other endocrine glands (10–13). Thus, on the peripheral level of the adrenal gland, IL-6 may act in a different way as compared with TNF. In a chronic inflammatory disease such as RA, the HPA axis demonstrates marked alterations. 1) There is inadequate secretion of ACTH relative to the extent of inflammation (14). 2) It has been described that patients with RA have inappropriately low levels of spontaneous and stimulated cortisol secretion, particularly in relation to inflammation (14–23). 3) During a long-term inflammatory disease such as RA, adrenal androgens dramatically decrease (24–32). The reasons for these changes are only partly understood, but striking changes on all levels of the HPA axis seem to play a role. During repetitive administration of IL-6 over 3 weeks, the stimulatory capacity of IL-6 on the central level is normally lost, but stimulation of the adrenal glands remains stable (5). In human subjects, this has never been tested with TNF, but one may expect similar adaptational changes on the level of the hypothalamus and pituitary gland. Thus, during chronic cytokine elevation, the hypothalamus and pituitary gland would not be adequately stimulated by IL-6 (or possibly TNF). However, IL-6–induced stimulation of the adrenal glands most likely remains unchanged, which was also suggested to be a mechanism during inflammatory cholestasis in rats and humans (33,34). Since TNF is the cytokine located upstream from IL-6, any increase or decrease in serum TNF is followed by an increase or decrease in serum IL-6. Thus, the effects of TNF may be mediated by the more stable and long-lived IL-6. Furthermore, the local and systemic concentrations of these immune mediators determine their influence on the levels of the hypothalamus, the pituitary gland, and the adrenals. Interactions of these different factors in chronic inflammatory diseases are complex. However, we believe that treatment with antiTNF opens a small window for understanding the complexity of the interwoven participants in the chronic inflammatory process of RA. Thus, it was the aim of the present study to investigate the effect of long-term anti-TNF therapy on the function of the HPA axis, including adrenal androgen secretion as well as cortisol and androgen inactivation (shuttle from cortisol to cortisone and dehydroepiandrosterone [DHEA] to DHEA sulfate, respectively) (Figure 1). Due to the fact that inflammation-induced changes of the HPA axis are long-lived, the patients were observed for 16 weeks during anti-TNF therapy. Furthermore, we compared patients with and without 1506 STRAUB ET AL Table 1. Characteristics of patients under investigation* Patients without parallel prednisolone Patients with parallel prednisolone 7 43.2 ⫾ 3.7 [27–55] 5/2 (female 71) 45.0 ⫾ 8.7 12 43.6 ⫾ 3.3 [26–57] 11/1 (female 92) 30.9 ⫾ 6.0 No. of patients Age, years Sex, no. female/no. male Initial erythrocyte sedimentation rate, mm/first hour Initial C-reactive protein, mg/liter Initial tender/swollen joint score Initial morning stiffness, minutes Additional therapy Prednisolone, no. of patients Mean daily dose prednisolone, mg Infliximab, mg/kg per injection Methotrexate, no. of patients NSAID, no. of patients 34.7 ⫾ 9.4 67.0 ⫾ 13.6 48.6 ⫾ 14.8 0 0⫾0 6.1 ⫾ 1.8 [1–10] 6 6 (86) 30.0 ⫾ 10.2 73.8 ⫾ 8.6 105.4 ⫾ 25.3 12 4.7 ⫾ 0.5 [2.5–7.5] 4.1 ⫾ 1.1 [1–10] 3 9 (75) * Except where indicated otherwise, values are the mean ⫾ SEM, with ranges in brackets and percentages in parentheses. Except in the frequency of prednisolone and methotrexate use (P ⫽ 0.028), the study groups were not different in the mentioned parameters. NSAID ⫽ nonsteroidal antiinflammatory drug. parallel prednisolone therapy, because this is the strongest influencing factor on hormone secretion. PATIENTS AND METHODS Patients and blood samples. In this study of anti-TNF therapy with infliximab (1–10 mg/kg per infusion), we included 19 white patients (16 women, 3 men) with long-standing RA that fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) criteria for RA (35). Patients of this study have been included in 2 other studies that dealt with other aspects of anti-TNF antibody therapy (36,37). A group of 7 patients did not receive parallel prednisolone therapy, whereas the other 12 patients received 2.5–7.5 mg prednisolone concurrently. Some of the patients were also administered methotrexate (Table 1). Since the frequency of patients receiving methotrexate was higher in the group without prednisolone versus those with prednisolone, we determined the influence of methotrexate on the clinical and hormonal parameters (no significant influence was detected by Kruskal-Wallis test; see statistical analysis). The patients were on a stable regimen with respect to prednisolone and methotrexate. Table 1 gives the characteristics of the 2 study groups. Patients were clinically investigated and blood was drawn between 8:00 and 9:00 in the morning when the patients visited the outpatient clinic on day 0, as well as on weeks 1, 2, 4, 6, 8, 10, 12, 14, and 16 of anti-TNF therapy. Anti-TNF antibodies were infused on day 0 and weeks 2, 6, 10, and 14. The blood was immediately centrifuged and serum was stored at ⫺80°C. The study was approved by the Ethics Committee of the University of Erlangen–Nürnberg in Germany. Laboratory parameters. Several adrenal hormones were considered in order to detect major adrenal pathways of steroidogenesis (Figure 1). We used radioimmunometric assays for the quantitative determination of serum levels of cortisol (Coulter Immunotech, Marseilles, France; detection limit 10 nmoles/liter, crossreactivity with prednisolone or prednisone ⬍6%). Serum levels of 17-hydroxyprogesterone (17[OH]progesterone) (IBL, Hamburg, Germany; detection limit 0.3 nmoles/liter), DHEA (Diagnostic Systems, Webster, TX; detection limit 0.13 nmoles/liter), androstenedione (ASD) (IBL; detection limit 0.3 nmoles/liter), and DHEA sulfate (IBL; detection limit 130 nmoles/liter) were measured by means of immunometric enzyme immunoassays. Serum levels of IL-6 (high-sensitivity Quantikine; R&D Systems, Minneapolis, MN; detection limit 0.2 pg/ml) were measured using the same technique. Intraassay and interassay coefficients of variation for all of the above-mentioned tests were below 10%. Routine parameters, mentioned in Table 1, were measured by standardized assays in the Department of Clinical Chemistry of the University of Erlangen–Nürnberg. Using a sensitive enzyme immunoassay for ACTH (Sangui BioTech, Santa Ana, CA, via IBL; detection limit 0.1 pmoles/liter), we were able to demonstrate a highly significant interrelation between ACTH measured in serum and ACTH assayed in plasma, with the following regression equation: ACTH (plasma) ⫽ 7.2508 ⫹ [1.8707 ⫻ ACTH (serum)] (R ⫽ 0.646, P ⬍ 0.000001, n ⫽ 112 healthy subjects) (38). In the present study, because no plasma samples were available, we measured ACTH in serum samples of patients with RA, using this enzyme immunoassay. HPLC assay for cortisone. Cortisone was determined by high-performance liquid chromatography (HPLC) with photometric detection at 245 nm, as adapted from a published method (39). Briefly, 200 l serum was buffered with 200 l of 0.2M sodium hydrogen carbonate, pH 9.6, and extracted with 2 ml dichloromethane. The organic layer was evaporated and the residue was reconstituted with 100 l mobile phase, of which 50 l was injected into the HPLC system. The chromatographic apparatus consisted of the solvent delivery system LC10AS, autosampler SIL-10A, ultraviolet detector LC10A, and Class LC10 controller and integrator (Shimadzu, Duisburg, Germany). Cortisol and cortisone were separated using 2 analytic columns (length ⫻ inner diameter 150 ⫻ 4.6 mm), a Synergi Polar-RP followed by a Synergi Max-RP (Phenomenex, INFLUENCE OF ANTI-TNF TREATMENT IN PATIENTS WITH RA 1507 Table 2. Reduction of inflammation in rheumatoid arthritis patients under long-term anti–tumor necrosis factor therapy* Erythrocyte sedimentation rate, mm/first hour Tender/swollen joint score Morning stiffness, minutes Serum interleukin-6, pg/ml Patients without prednisolone, Day 03 Week 1 Patients with prednisolone, Day 03 Week 1 45.0 ⫾ 8.7327.9 ⫾ 6.1 30.9 ⫾ 6.0316.9 ⫾ 4.6† 67.0 ⫾ 13.6326.1 ⫾ 4.9‡ 48.6 ⫾ 14.8339.3 ⫾ 12.3§ 41.0 ⫾ 11.3323.9 ⫾ 12.0§ 73.8 ⫾ 8.6329.3 ⫾ 4.1† 105.4 ⫾ 25.3342.9 ⫾ 10.0‡ 65.3 ⫾ 8.8314.1 ⫾ 7.1† * Values are the mean ⫾ SEM. Week 1 results are reported because the most dramatic changes occurred in this time period. † P ⬍ 0.002. ‡ P ⬍ 0.005. § P ⬍ 0.10. Aschaffenburg, Germany), with water–acetonitrile (70:30, volume/volume) as mobile phase. Cortisol eluted after 13.6 minutes and cortisone after 15.5 minutes (flow rate 1.0 ml/ minute, column temperature 40°C). The recovery of cortisone from serum was 98%. Intraassay and interassay variation coefficients were below 9%. Calibration curves for peak heights versus quantity were proved to be linear from 5 to 125 ng/ml cortisone, with a coefficient of correlation of ⬎0.996. The minimal detectable amount injected (signal:noise 3:1) was ⬃200 pg. The limit of quantitation was estimated to be 4–5 ng/ml. The method also allows the separation and simultaneous determination of prednisolone (retention time 13.0 minutes) and prednisone (retention time 14.5 minutes) with similar sensitivity and precision. The latter 2 factors were used to test whether or not patients were really treated with prednisolone. Statistical analysis. Changes in group medians between 2 different time points were compared by the nonparametric Wilcoxon signed rank test for paired data (SPSS/PC, Advanced Statistics, version 10.0.1; SPSS, Chicago, IL). The correlation analyses were performed with the Spearman’s rank test (SPSS/PC, Advanced Statistics, version 10.0.1; SPSS). The regression lines were derived from linear regression. A P value less than 0.05 was the significance level. Since our study group was relatively small (albeit typically large enough to study hormonal changes in a longitudinal way with 10 time points over 16 weeks), we first tested whether methotrexate or different infliximab doses (1, 3, and 10 mg/kg per infusion) influenced the clinical and laboratory parameters (using the Kruskal-Wallis test). With respect to all assessed clinical and hormonal parameters (including hormone ratios) during the entire observation period, we did not find any influence of methotrexate or different doses of infliximab (data not shown). Thus, patient groups were not further stratified. RESULTS Decrease of inflammation. Table 2 demonstrates the reduction in inflammatory indices during the longterm treatment with anti-TNF. All mentioned variables (Table 2) decreased, more or less, between day 0 and week 1 of therapy (Table 2). In order to test the continuous antiinflammatory effect of anti-TNF treatment, the decline in serum IL-6 throughout the entire study period is demonstrated in Figure 2. The decrease in IL-6 was similar in both groups, although RA patients without prednisolone demonstrated a more steady decline of serum IL-6 (Figure 2A), whereas mean serum IL-6 levels in patients with prednisolone particularly dropped during the first week (day 0 versus week 1 P ⫽ 0.018) (Figure 2B). Changes of serum ACTH and serum cortisol. In patients without prednisolone therapy, the mean levels of serum ACTH increased after every infusion of antiTNF (Figure 3A, thin, upward-directed arrows). This indicates that reduction of the inflammatory load (serum IL-6 and serum TNF) sensitizes the central parts of the hypothalamus and the pituitary, which induces an Figure 2. Course of serum IL-6 during 16 weeks of anti-TNF antibody therapy in patients with rheumatoid arthritis. The graph depicts patients without (A) and with (B) parallel treatment with prednisolone. The data are given as the mean and SEM. The Spearman’s rank correlation coefficient (RRank), its P value, and the linear regression line are shown. Arrows indicate the time point of anti-TNF antibody infusion. n.s. ⫽ not significant (see Figure 1 for other definitions). 1508 STRAUB ET AL both patient groups, but the changes did not reach the significance level (data not shown). Changes of serum cortisol relative to serum ACTH. In patients without parallel prednisolone treatment, the ratio of serum cortisol to serum ACTH decreased over the 16 weeks of anti-TNF therapy (Figure 4A). Under stable serum cortisol conditions, this again indicates a sensitization of ACTH secretion, with a relative increase of serum ACTH over serum cortisol (Figure 4A). This was corroborated by an initial increase in serum ACTH relative to serum IL-6 from day 0 to week 1 of therapy (700% increase of this ratio; P ⫽ 0.018) (Figure 4C). This particular ratio remained elevated up to 16 weeks of anti-TNF therapy (Figure 4C). The stable behavior of serum cortisol was confirmed by the relative increase in the ratio of serum cortisol to serum IL-6 (Figure 4E) because serum IL-6 had declined (Figure 2A). Figure 3. Course of serum ACTH and serum cortisol during 16 weeks of anti-TNF antibody therapy in patients with rheumatoid arthritis. The graph depicts patients without (A and C) and with (B and D) parallel treatment with prednisolone. The data are given as the mean and SEM. The Spearman’s rank correlation coefficient (RRank), its P value, and the linear regression line are shown. Thick upward arrows indicate the time point of anti-TNF antibody infusion. In A and B, the thin upward or downward arrows demonstrate the behavior of serum ACTH after anti-TNF antibody infusion. The plasma concentrations of ACTH would be 7.2508 ⫹ (1.8707 ⫻ serum ACTH) (see Patients and Methods). See Figures 1 and 2 for definitions. increase of serum ACTH, while serum cortisol did not markedly change (Figure 3C). This pattern was obviously different in the patients receiving prednisolone treatment, most probably due to the glucocorticoidinduced lack of CRH from the hypothalamus (Figure 3B, thin, downward-directed arrows). In these latter patients, the mean serum ACTH decreased after every infusion of anti-TNF, which indicates that ACTH secretion depends on the inflammatory load (serum IL-6 and serum TNF). Similar to the group without prednisolone, serum levels of cortisol did not change in these RA patients who were receiving prednisolone therapy (Figure 3D). The consideration of the ratio of serum cortisol to serum cortisone gives an idea about the extent of cortisol degradation. This ratio slightly decreased in Figure 4. Course of the ratios of serum cortisol to serum ACTH (A and B), serum ACTH to serum IL-6 (C and D), and serum cortisol to serum IL-6 (E and F) during 16 weeks of anti-TNF antibody therapy in patients with rheumatoid arthritis. The graph depicts patients without (A, C, and E) and with (B, D, and F) parallel treatment with prednisolone. The data are given as the means and SEM. The Spearman’s rank correlation coefficient (RRank), its P value, and the linear regression line are shown. Arrows indicate the time point of anti-TNF antibody infusion. See Figures 1 and 2 for definitions. INFLUENCE OF ANTI-TNF TREATMENT IN PATIENTS WITH RA In patients receiving prednisolone treatment, the ratio of serum cortisol to serum ACTH remained quite stable over the 16 weeks of anti-TNF therapy (Figure 4B). Under stable serum cortisol conditions, this does not indicate a sensitization of ACTH secretion. Since, in this group, serum ACTH and serum cortisol did not markedly change during the 16 weeks of anti-TNF treatment, the decrease in serum IL-6 led to elevated ratios of serum ACTH to serum IL-6 (Figure 4D) and serum cortisol to serum IL-6 (Figure 4F). Changes of adrenal androgens. In patients without and with parallel prednisolone, serum 17(OH)progesterone and serum ASD did not markedly change over the 16 weeks of anti-TNF treatment (Figures 5A–D). However, the ratio of these 2 hormones significantly increased in patients without prednisolone (Figure 5E), Figure 5. Course of serum levels of 17-hydroxyprogesterone (17OHP) (A and B), serum androstenedione (ASD) (C and D), and the ratio of serum ASD to serum 17OHP (E and F) during 16 weeks of anti-TNF antibody therapy in patients with rheumatoid arthritis. The graph depicts patients without (A, C, and E) and with (B, D, and F) parallel treatment with prednisolone. The data are given as the mean and SEM. The Spearman’s rank correlation coefficient (RRank), its P value, and the linear regression line are given. Arrows indicate the time point of anti-TNF antibody infusion. See Figures 1 and 2 for other definitions. 1509 Figure 6. Course of the ratio of serum ASD to serum cortisol during 16 weeks of anti-TNF antibody therapy in patients with rheumatoid arthritis. The graph depicts patients without (A) and with (B) parallel treatment with prednisolone. The data are given as the mean and SEM. The Spearman’s rank correlation coefficient (RRank), its P value, and the linear regression line are given. Arrows indicate the time point of anti-TNF antibody infusion. See Figures 1 and 5 for definitions. which indicates an activation of this pathway toward the direction of adrenal androgens (Figure 1). In contrast, patients with prednisolone treatment demonstrated a steady decrease in the ratio of serum ASD to serum 17(OH)progesterone (Figure 5F). Interestingly, serum ASD was more influenced by parallel prednisolone treatment as compared with 17(OH)progesterone (compare the levels of the hormones in Figures 5A and B, and Figures 5C and D). Patients with prednisolone had a generally lower mean serum ASD as compared with patients without parallel prednisolone (Figures 5C and D), which was corroborated by the ratio of serum ASD to serum 17(OH)progesterone (Figures 5E and F). This indicates that this particular enzyme step is subject to exogenous prednisolone therapy. Furthermore, the ratio of serum ASD to serum cortisol is another indicator of a change in secretion of androgens in relation to the glucocorticoid cortisol (Figure 1). This ratio increased in patients without parallel prednisolone (Figure 6A), whereas this ratio tended to decrease in patients with parallel prednisolone (Figure 6B). These data corroborate the results with respect to ASD and 17(OH)progesterone (Figures 5E and F). In both groups, there were no marked changes in the levels of serum DHEA, DHEA sulfate, and the ratio of serum DHEA to serum ASD or the ratio of serum DHEA sulfate to serum DHEA throughout the 16 weeks of anti-TNF therapy (data not shown). DISCUSSION Long-term therapy with anti-TNF in patients with RA leads to an overall reduction in the inflamma- 1510 tory load (serum IL-6, serum amyloid A, haptoglobin, fibrinogen), which has also been described in an earlier investigation of these same patients (37). In patients without prednisolone, this led to sensitization of the pituitary gland, which was demonstrated as a rapid increase in the average ACTH serum concentration after every infusion of anti-TNF. Thus, we propose that chronically elevated serum TNF inhibits hypothalamic CRH secretion or CRH-stimulated pituitary ACTH secretion, but certainly does not stimulate secretion of these 2 hormones (Figure 1). The pituitary sensitization was also demonstrated by the decrease in the ratio of serum cortisol to serum ACTH; since serum cortisol remained relatively stable during the course of the therapy and serum ACTH increased (particularly after every infusion of anti-TNF), the numeric value of the ratio decreased. The ratio of serum cortisol to serum ACTH was calculated for every patient, which better reflects the individual relationship of serum cortisol and serum ACTH. Consideration of the group mean ratios of one hormone to another hormone would not give equal insight, because we would lose the individual variation of both hormones in one patient. In this particular case (cortisol:ACTH), the relatively stable serum cortisol levels in relation to the increasing serum ACTH was only visible after building the ratio. Thus, in summary, ACTH serum levels increase in relation to serum cortisol, which is indicative of a sensitization of the hypothalamic–pituitary axis (the ACTH producer), but not of the adrenal glands (the cortisol producer). A marked increase in serum ACTH was also demonstrated in relation to serum IL-6 between day 0 and week 1 of therapy, which remained elevated throughout the observation period. Furthermore, the increase in the ratio of serum cortisol to serum IL-6 indicates the relatively stable behavior of serum cortisol, since serum IL-6 was obviously decreasing. Furthermore, the increase in the ratio of serum ASD to serum 17(OH)progesterone and serum ASD to serum cortisol indicates a step toward a normalization of adrenal androgen production, because ASD is the main precursor of adrenal androgens. This is a very interesting finding, because it shows that normalization of adrenal androgen production can occur even after long-term inflammation, when the inhibitory break (in this case, TNF) is removed. As expected, parallel stable prednisolone therapy completely changed the behavior of the HPA axis (this is not dependent on parallel methotrexate treatment, since we did not see any differences between patients with and STRAUB ET AL without methotrexate). First, after every infusion of anti-TNF, serum ACTH rapidly decreased, which indicates that the inflammatory load (serum IL-6 and serum TNF) stimulates the hypothalamic–pituitary axis under prednisolone-induced conditions. This is completely opposite to the above-mentioned situation without prior prednisolone therapy. Second, the hypothalamic– pituitary axis was not sensitized because the ratio of serum cortisol to serum ACTH did not change. Third, the ratio of serum ASD to serum 17(OH)progesterone and serum ASD to serum cortisol decreased, which indicates that under these conditions, TNF or IL-6 downstream may even stimulate these particular enzyme steps toward adrenal androgens (Figure 1). Under conditions with parallel prednisolone, stimulation of the remaining HPA axis depends more on the inflammatory load, as compared with the situation without parallel prednisolone. One may speculate that long-term prednisolone therapy inhibits hypothalamic CRH secretion, which removes the influence of the hypothalamus on the pituitary gland. Thus, ACTH secretion completely depends on the inflammatory load, but not on hypothalamic CRH. Interestingly, these patients who received parallel prednisolone had a clinical outcome similar to that of the patients without parallel prednisolone (36). This may indicate that during anti-TNF therapy, both the restored adrenal corticosteroids (in patients without parallel prednisolone) and the administered corticosteroids (patients with parallel prednisolone) favorably influenced the inflammatory process to a similar extent. Furthermore, parallel prednisolone therapy might have also decreased TNF secretion, which is an additional antiinflammatory factor acting in conjunction with anti-TNF antibodies. Such a situation would lead to a stronger reduction in the inflammatory load as compared with therapy with anti-TNF alone. These facts may partially explain the different HPA axis behavior in patients with and without parallel prednisolone therapy. In conclusion, this study with long-term anti-TNF therapy demonstrates a sensitization of the hypothalamus and pituitary gland in patients who have not received parallel prednisolone therapy. In addition, the adrenal androgen ASD increases relative to its precursor 17(OH)progesterone and cortisol, which indicates a step toward normalization of adrenal androgen production. As the systemic inflammation decreases, the function of the HPA axis begins to normalize over 16 weeks. This obviously demonstrates that during chronic inflammation, the HPA axis seems to support the systemic inflammation (desensitization of the hypothalamus, low INFLUENCE OF ANTI-TNF TREATMENT IN PATIENTS WITH RA cortisol in relation to inflammation, low adrenal androgens) rather than to counterbalance the chronic inflammatory process. 16. ACKNOWLEDGMENT We thank Alfonse Masi for helpful discussions at our poster at the 2002 American College of Rheumatology annual meeting in New Orleans. 17. 18. REFERENCES 19. 1. Elliott MJ, Maini RN, Feldmann M, Long-Fox A, Charles P, Katsikis P, et al. Treatment of rheumatoid arthritis with chimeric monoclonal antibodies to tumor necrosis factor ␣. Arthritis Rheum 1993;36:1681–90. 2. Van Dullemen HM, van Deventer SJ, Hommes DW, Bijl HA, Jansen J, Tytgat GN, et al. Treatment of Crohn’s disease with anti-tumor necrosis factor chimeric monoclonal antibody (cA2). Gastroenterology 1995;109:129–35. 3. Braun J, Brandt J, Listing J, Zink A, Alten R, Golder W, et al. Treatment of active ankylosing spondylitis with infliximab: a randomised controlled multicentre trial. Lancet 2002;359:1187–93. 4. Ogilvie AL, Antoni C, Dechant C, Manger B, Kalden JR, Schuler G, et al. Treatment of psoriatic arthritis with antitumour necrosis factor-alpha antibody clears skin lesions of psoriasis resistant to treatment with methotrexate. Br J Dermatol 2001;144:587–9. 5. Mastorakos G, Chrousos GP, Weber JS. Recombinant interleukin-6 activates the hypothalamic-pituitary-adrenal axis in humans. J Clin Endocrinol Metab 1993;77:1690–4. 6. Späth-Schwalbe E, Born J, Schrezenmeier H, Bornstein SR, Stromeyer P, Drechsler S, et al. Interleukin-6 stimulates the hypothalamus-pituitary-adrenocortical axis in man. J Clin Endocrinol Metab 1994;79:1212–4. 7. Michie HR, Spriggs DR, Manogue KR, Sherman ML, Revhaug A, O’Dwyer ST, et al. Tumor necrosis factor and endotoxin induce similar metabolic responses in human beings. Surgery 1988;104: 280–6. 8. Besedovsky HO, del Rey AE, Sorkin E. Immune-neuroendocrine interactions. J Immunol 1985;135:750–4. 9. Jäättelä M, Ilvesmaki V, Voutilainen R, Stenman UH, Saksela E. Tumor necrosis factor as a potent inhibitor of adrenocorticotropin-induced cortisol production and steroidogenic P450 enzyme gene expression in cultured human fetal adrenal cells. Endocrinology 1991;128:623–9. 10. Xiong Y, Hales DB. The role of tumor necrosis factor-alpha in the regulation of mouse Leydig cell steroidogenesis. Endocrinology 1993;132:2438–44. 11. Lin D, Sugawara T, Strauss JF III, Clark BJ, Stocco DM, Saenger P, et al. Role of steroidogenic acute regulatory protein in adrenal and gonadal steroidogenesis. Science 1995;267:1828–31. 12. Mauduit C, Gasnier F, Rey C, Chauvin MA, Stocco DM, Louisot P, et al. Tumor necrosis factor-alpha inhibits Leydig cell steroidogenesis through a decrease in steroidogenic acute regulatory protein expression. Endocrinology 1998;139:2863–8. 13. Budnik LT, Jahner D, Mukhopadhyay AK. Inhibitory effects of TNF alpha on mouse tumor Leydig cells: possible role of ceramide in the mechanism of action. Mol Cell Endocrinol 1999;150:39–46. 14. Straub RH, Paimela L, Peltomaa R, Schölmerich J, LeirisaloRepo M. Inadequately low serum levels of steroid hormones in relation to interleukin-6 and tumor necrosis factor in untreated patients with early rheumatoid arthritis and reactive arthritis. Arthritis Rheum 2002;46:654–62. 15. Van den Brink HR, Blankenstein MA, Koppeschaar HP, Bijlsma 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 1511 JW. Influence of disease activity on steroid hormone levels in peripheral blood of patients with rheumatoid arthritis. Clin Exp Rheumatol 1993;11:649–52. Hall J, Morand EF, Medbak S, Zaman M, Perry L, Goulding NJ, et al. Abnormal hypothalamic-pituitary-adrenal axis function in rheumatoid arthritis: effects of nonsteroidal antiinflammatory drugs and water immersion. Arthritis Rheum 1994;37:1132–7. Gudbjornsson B, Skogseid B, Oberg K, Wide L, Hallgren R. Intact adrenocorticotropic hormone secretion but impaired cortisol response in patients with active rheumatoid arthritis: effect of glucocorticoids. J Rheumatol 1996;23:596–602. Templ E, Koeller M, Riedl M, Wagner O, Graninger W, Luger A. Anterior pituitary function in patients with newly diagnosed rheumatoid arthritis. Br J Rheumatol 1996;35:350–6. Crofford LJ, Kalogeras KT, Mastorakos G, Magiakou MA, Wells J, Kanik KS, et al. Circadian relationships between interleukin (IL)-6 and hypothalamic-pituitary-adrenal axis hormones: failure of IL-6 to cause sustained hypercortisolism in patients with early untreated rheumatoid arthritis. J Clin Endocrinol Metab 1997;82: 1279–83. Cutolo M, Foppiani L, Prete C, Ballarino P, Sulli A, Villaggio B, et al. Hypothalamic-pituitary-adrenocortical axis function in premenopausal women with rheumatoid arthritis not treated with glucocorticoids. J Rheumatol 1999;26:282–8. Gutierrez MA, Garcia ME, Rodriguez JA, Mardonez G, Jacobelli S, Rivero S. Hypothalamic-pituitary-adrenal axis function in patients with active rheumatoid arthritis: a controlled study using insulin hypoglycemia stress test and prolactin stimulation. J Rheumatol 1999;26:277–81. Demir H, Kelestimur F, Tunc M, Kirnap M, Ozugul Y. Hypothalamo-pituitary-adrenal axis and growth hormone axis in patients with rheumatoid arthritis. Scand J Rheumatol 1999;28:41–6. Kanik KS, Chrousos GP, Schumacher HR, Crane ML, Yarboro CH, Wilder RL. Adrenocorticotropin, glucocorticoid, and androgen secretion in patients with new onset synovitis/rheumatoid arthritis: relations with indices of inflammation. J Clin Endocrinol Metab 2000;85:1461–6. Feher GK, Feher T, Zahumenszky Z. Study on the inactivation mechanism of androgens in rheumatoid arthritis: excretory rate of free and conjugated 17-ketosteroids. Endokrinologie 1979;73: 167–72. Masi AT, Josipovic DB, Jefferson WE. Low adrenal androgenicanabolic steroids in women with rheumatoid arthritis (RA): gasliquid chromatographic studies of RA patients and matched normal control women indicating decreased 11-deoxy-17-ketosteroid excretion. Semin Arthritis Rheum 1984;14:1–23. Cutolo M, Balleari E, Giusti M, Monachesi M, Accardo S. Sex hormone status of male patients with rheumatoid arthritis: evidence of low serum concentrations of testosterone at baseline and after human chorionic gonadotropin stimulation. Arthritis Rheum 1988;31:1314–7. Sambrook PN, Eisman JA, Champion GD, Pocock NA. Sex hormone status and osteoporosis in postmenopausal women with rheumatoid arthritis. Arthritis Rheum 1988;31:973–8. Deighton CM, Watson MJ, Walker DJ. Sex hormones in postmenopausal HLA-identical rheumatoid arthritis discordant sibling pairs. J Rheumatol 1992;19:1663–7. Hedman M, Nilsson E, de la Torre B. Low blood and synovial fluid levels of sulpho-conjugated steroids in rheumatoid arthritis. Clin Exp Rheumatol 1992;10:25–30. Valentino R, Savastano S, Tommaselli AP, Riccio A, Mariniello P, Pronesti G, et al. Hormonal pattern in women affected by rheumatoid arthritis. J Endocrinol Invest 1993;16:619–24. Mateo L, Nolla JM, Bonnin MR, Navarro MA, Roig-Escofet D. Sex hormone status and bone mineral density in men with rheumatoid arthritis. J Rheumatol 1995;22:1455–60. 1512 32. Mirone L, Altomonte L, D’Agostino P, Zoli A, Barini A, Magaro M. A study of serum androgen and cortisol levels in female patients with rheumatoid arthritis: correlation with disease activity. Clin Rheumatol 1996;15:15–9. 33. Swain MG, Maric M, Carter L. Defective interleukin-1-induced ACTH release in cholestatic rats: impaired hypothalamic PGE2 release. Am J Physiol 1995;268:G404–9. 34. Zietz B, Wengler I, Messmann H, Lock G, Schölmerich J, Straub RH. Early shifts of adrenal steroid synthesis before and after relief of short-term cholestasis. J Hepatol 2001;35:329–37. 35. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. 36. Maini RN, Breedveld FC, Kalden JR, Smolen JS, Davis D, Macfarlane JD, et al. Therapeutic efficacy of multiple intravenous infusions of anti–tumor necrosis factor ␣ monoclonal antibody STRAUB ET AL combined with low-dose weekly methotrexate in rheumatoid arthritis. Arthritis Rheum 1998;41:1552–63. 37. Charles P, Elliott MJ, Davis D, Potter A, Kalden JR, Antoni C, et al. Regulation of cytokines, cytokine inhibitors, and acute-phase proteins following anti-TNF-alpha therapy in rheumatoid arthritis. J Immunol 1999;163:1521–8. 38. Straub RH, Konecna L, Hrach S, Rothe G, Kreutz M, Schölmerich J, et al. Serum dehydroepiandrosterone (DHEA) and DHEA sulfate are negatively correlated with serum interleukin-6 (IL-6), and DHEA inhibits IL-6 secretion from mononuclear cells in man in vitro: possible link between endocrinosenescence and immunosenescence. J Clin Endocrinol Metab 1998;83:2012–7. 39. Santos-Montes A, Gonzalo-Lumbreras R, Izquierdo-Hornillos R. Simultaneous determination of cortisol and cortisone in urine by reversed-phase high-performance liquid chromatography: clinical and doping control applications. J Chromatogr B Biomed Appl 1995;673:27–33.