Z Rheumatol 59: Suppl 2II/65–II/69 © Steinkopff Verlag 2000

M. CutoloM. GiustiL. FoppianiB. SerioloM. BriataA. BissoF. FaelliL. FelliC. PreteC. PizzorniA. Sulli

The hypothalamic-pituitary-adrenocorticaland gonadal axis function in rheumatoid arthritis

eindeutig als relative Nebennieren-insuffizienz zu bezeichnen sind. Andro-gene dürften in der Pathogenese vonAutoimmunerkrankungen, einschließ-lich der RA, als physiologische Immu-nosuppressoren eine Rolle spielen. Beimännlichen RA-Patienten wurden nied-rige Testosteronspiegel sowohl imPlasma als auch in der Synovia gefun-den, bei RA-Patientinnen waren dieentsprechenden Spiegel für DHEASerniedrigt.

Die Häufung der RA um die Meno-pause legt weiter eine pathophysiolo-gische Bedeutung sinkender Spiegelvon Östrogenen und/oder Progesteronnahe. Zahlreiche Befunde deuten aufsuppressive Effekte von Östrogenen aufdie zelluläre Immunität, während siestimulierend auf die humorale Immu-nität wirken (bzw. ein Östrogendefizitsteigert die Th 1-Typ-vermittelteImmunität). Genetische Polymorphis-men von Enzymen der Steroidbiosyn-these scheinen zusätzlich die Rolle vonSexualsteroiden in der Entwicklungvon autoimmunen Reaktionen zubegünstigen. Darüberhinaus beeinflus-sen erworbene Formen eines veränder-ten Steroidmetabolismus die peripherenSpiegel der Sexualsteroide. Zusammen-fassend ist die pathogenetische Rele-vanz eines komplexen Zusammen-wirkens der Achsen Hypothalamus-Hypophyse-Nebennierenrinde undHypothalamus-Hypophyse-Gonadenbei RA evident.

Schlüsselwörter Achse Hypothala-mus-Hypophyse-Nebennierenrinde –Androgene – Östrogene – rheumatoideArthritis – Entzündung – Sexualhor-mone – Zytokine

Summary The altered cortisol andadrenal androgen (i.e., dehydro-epiandrosterone sulfate = DHEAS)secretion, observed during testing inrheumatoid arthritis (RA) patients nottreated with corticosteroids, should beclearly regarded as a “relative adrenalinsufficiency” in the setting of a sus-tained inflammatory process, as shownby high serum IL-6 levels. Androgensseem implicated in the pathophysiol-ogy of autoimmune disorders, includ-ing RA, as natural immunosuppressors.Low plasma and synovial fluid testos-terone concentrations are observed inmale RA patients; low plasma DHEASlevels are mainly observed in femaleRA patients.

The menopausal peak of RA sug-gests that estrogens and/or proges-terone deficiency also play a role in thedisease, and many data indicate thatestrogens suppress cellular immunity,but stimulate humoral immunity (i.e.,deficiency promotes cellular Th1-typeimmunity). Gene polymorphisms forenzymes involved in the steroidogene-sis seem to further complicate the roleof sex hormones in the susceptibility toautoimmunity. Acquired changes of sex

M. Cutolo (Y) · B. Seriolo · M. BriataA. Bisso · F. Faelli · C. Prete · C. PizzorniA. SulliDivision of RheumatologyDepartment of Internal MedicineUniversity of Genova-ItalyViale Benedetto XV, 6I-16132 Genova, Italy

L. FelliDepartment of OrthopedicsUniversity of GenovaGenova, Italy

M. Giusti · L. FoppianiDivision of EndocrinologyDISEMUniversity of GenovaGenova, Italy

Die Funktionen der Achsen Hypothalamus-Hypophyse-Neben-nierenrinde und Hypothalamus-Hypophyse-Gonaden bei rheumatoider Arthritis

Zusammenfassung Untersuchungenvon bis dahin mit Glukokortikoidenunbehandelten Patienten mit rheuma-toider Arthritis (RA) zeigen Abwei-chungen der Spiegel für Kortisol undadrenalen Androgenen (vor allemDehydroepiandrosteronsulfat,DHEAS), die in der Relation zu gleich-zeitig hohen Serumspiegeln von IL-6

II/66 Zeitschrift für Rheumatologie, Band 59, Suppl 2 (2000)© Steinkopff Verlag 2000

Introduction

Rheumatoid arthritis (RA) results from the combination ofseveral predisposing factors that include the relationshipsbetween epitopes of the trigger agent (i.e., virus) and histo-compatibility epitopes (i.e., HLA), the status of the stressresponse system (hypothalamic-pituitary-adrenocortical axis= HPA) and the gonadal hormones (hypothalamic-pituitary-gonadal axis = HPG), with estrogens implicated as enhancersof the immune response and androgens and progesterone asnatural suppressors (1–4).

The hypothalamic-pituitary-adrenocortical axisin rheumatoid arthritis

The inflammatory cytokines (i.e., IL-6, IL-1, TNFa) as solu-ble products of the activated immune system stimulate the pro-duction of corticotropin-releasing hormone (CRH) in thehypothalamus: CRH release leads to pituitary production ofadrenocorticotropic hormone (ACTH), followed by glucocor-ticoid secretion by the adrenal cortex and indirect perturba-tions of gonadal function (see Fig. 1 – left side) (5, 6).

It is now recognized that young females, affected by recentintensive stressful conditions (interpersonal stressors, surgicalor infectious events) activating the HPA, with associated lowplasma adrenal androgens (i.e., dehydroepiandrosterone sul-fate = DHEAS) and recent use of contraceptive pills, are thebest candidates for the onset of autoimmune disorders, includ-ing RA (7–10). The stress system has multiple levels and iscomprised of neuroendocrinological (i.e., HPA axis), psycho-logical and environmental components (11, 12). The physio-logic balance between the stress and the immune system maybe disrupted as a consequence of various pathological insults,including sustained exposure to stress, abnormal immune reac-tions to infections or both (13).

Recently, intact ACTH secretion but impaired cortisolresponse in patients with active RA has been described and thisobservation was consistent with a relative adrenal glucocorti-coid insufficiency, the latter already suggested forty yearsearlier (14, 15). Increased HPA axis function is a normalresponse to the stress of inflammation and might be mediatedby central and peripheral actions of circulating cytokines.

Besides, IL-1 and tumor necrosis factor-a (TNFa), IL-6appears to be a major factor mediating interactions between

the activated immune system and both the anterior pituitarycells and the adrenal steroidogenesis (see Fig. 1 – right side)(16). However, recent studies in RA patients have shown thatoverall activity of the HPA axis remains inappropriatelynormal and apparently is insufficient to inhibit ongoing inflam-mation at least in early untreated artritic patients (17).

More recently, another study showed a significantly alteredsecretion of adrenal androgens in non-glucocorticoid-treatedpremenopausal RA patients (18). Baseline concentrations ofDHEA and its sulfate metabolite DHEAS were found to besignificantly lower in RA than in normal subjects. In addition,during the low-dose ACTH testing, the DHEA production wasfound to be significantly lower in RA patients than in controls(18). In the same study, low levels of plasma DHEA andDHEAS were found to be significantly correlated with earlymorning low cortisol concentrations and high basal levels ofIL-6 in RA patients (18). Early morning IL-6 peak values wererecently found to be higher in RA patients than in controls, andsignificantly correlated to morning CRP levels and Ritchie’sindex (19). The observation of reduced DHEA production,combined with normal cortisol production during oCRH and

steroid metabolism seem to also play a role in the peripheral sex hormonelevels. In conclusion, a complex inter-action between the hypothalamus-pitu-

itary-adrenocortical and gonadal axisfunctions is evident in RA.

Key words Hypothalamic-pituitary-

adrenocortical axis – androgens –estrogens – rheumatoid arthritis –inflammation – sex hormones –cytokines

Fig. 1 Stress and inflammatory cytokines stimulate the production of cor-ticotropin-releasing hormone (CRH) in the hypothalamus: CRH releaseleads to pituitary production of adrenocorticotropic hormone (ACTH),followed by glucocorticoid secretion by adrenal cortex, influences onimmune cells and indirect perturbations of gonadal function (left side).Inflammatory cytokines (i.e., IL-6, IL-1, TNFa), as soluble products ofthe activated immune system, stimulate both the anterior pituitary cellsand the adrenal steroidogenesis (right side). Th1/Th2 and Ts cells aremodulated by adrenal and gonadal steroids (bottom). (Mf macrophages,DHEAS dehydroepiandrosterone sulfate, Th1-Th2 T helper, Ts Tsuppressor)

M. Cutolo et al. II/67The hypothalamic-pituitary-adrenocortical and gonadal axis function

ACTH testing, further support the concept of the presence ofan adrenal hypofunction in active RA patients (20).

IL-6 had a strong effect on steroid release and may be oneof the factors controlling the long-term adrenal response tostress, because this cytokine is able to act synergistically withACTH on the adrenal cells to stimulate the release of corti-costerone (21, 22). Therefore, the altered cortisol and adrenalandrogen secretion, observed during testing in RA patients nottreated with corticosteroids, should be clearly regarded as a“relative adrenal insufficiency” in the setting of a sustainedinflammatory process, as shown by high IL-6 levels (18).

In a very recent investigation on salivary cortisol levels inpatients with recent-onset RA, afternoon concentrations inpatients with high disease activity did not drop, as did thecortisol levels in healthy controls and RA patients with lowdisease activity (23).

The hypothalamus-pituitary-adrenocortical and gonadalaxis product interactions in rheumatoid arthritis

Androgens seem implicated as natural immunosuppressors inthe pathophysiology of autoimmune disorders, including RA.The disease is uncommon in men under age 45, but its inci-dence increases rapidly in older men and approaches the inci-dence in women. Low plasma and synovial fluid testosteronelevels are observed in male RA patients; low plasma DHEASlevels are mainly observed in female RA patients (4).

Direct exposure of T cells, T cell clones or T cell hybrido-mas to low concentrations of DHEA was found to increase thesecretion of type-1 cytokines (Th1), at least IL-2. Similarresults have been observed in cultured T-cells from systemiclupus erythematosus (SLE) patients who are characterized bylow DHEA serum levels (24).

Therefore, a causative role for reduced levels of gonadal(testosterone) and adrenal androgens (DHEA and DHEAS)has been suggested in the pathogenesis of RA, as well as inother immune/inflammatory arthritis (i.e., SLE) (see Fig. 1 –bottom) (25). The reduced basal concentration of adrenalandrogens in RA might be due both to a decreased pooling anda reduced sensitivity of the adrenal gland to exogenous corti-cotropin, or alternatively, to a partial enzymatic defect (i.e., 17,20 – lyase). Recently, a significantly altered steroidogenesis ofadrenal androgens (DHEA and DHEAS) in non-glucocorti-coid-treated premenopausal RA patients was confirmed (26,27). The observation of reduced DHEA production, combinedwith unexpected normal cortisol levels during oCRH andACTH testing, further support the concept of the presence ofadrenal hypofunction in active RA patients in the setting of a sustained inflammatory process as shown by high IL-6 andIL-12 concentrations.

Testosterone replacement therapy in RA patients is charac-terized by concomitant changes of DHEAS concentrations(28). In addition, adrenal and gonadal androgens, in particularDHEAS and both testosterone and dihydrotestosterone(DHT), have been found to repress the expression and activityof the human IL-6 gene promoter thus supporting the conceptof antiinflammatory/immunosuppressive effects of androgens(29, 30). Therefore, the well assessed deficiency of gonadaland adrenal androgens in RA patients seems to represent a rel-evant factor involved in the pathophysiology of the disease(31, 32).

The menopausal peak of RA suggests that estrogens and/orprogesterone deficiency also play a role in the disease, andmany data indicate that estrogens suppress cellular immunity,but stimulate humoral immunity (i.e., deficiency promotescellular Th1-type immunity) (33). On the other hand, recentobservations suggest that progesterone, as well as pregnancy(elevated estrogens, progesterone, corticosteroids and testos-terone), stimulates a switch from Th1- to Th2-type immuneresponses.

These data taken together indicate that gonadal and adrenalsteroid deficiency (altered HPA/HPG axis), plus prolactinincrease (i.e., breastfeeding), probably facilitates the expres-sion of Th1-type immunity, which is now considered to becritical in the pathogenesis of RA (33). On the contrary, normallevels of adrenal and gonadal steroids seem to suppress thedevelopment of RA.

In addition, recent studies suggest that testosterone maydirectly suppress anti-DNA antibody production in peripheralblood mononuclear cells from SLE patients by inhibiting Bcell hyperactivity and, indirectly, once again by downregulat-ing IL-6 production in monocytes (34). Furthermore, in an invivo investigation testosterone therapy ameliorated experi-mental autoimmune encephalomyelitis and induced a T helper2 bias in the autoantigen-specific T lymphocyte response (35).

Further factors involved in the androgen and estrogeneffects in the susceptibility to rheumatoid arthritis

Gene polymorphisms for enzymes involved in the steroido-genesis seem to further complicate the role of sex hormones inthe susceptibility to autoimmunity. Recently a relationshipbetween CYP17 genotypes and the age at onset of rheumatoidarthritis (RA) in female patients has been observed (36). TheCYP17 gene, coding for the cytochrome P450c17a, mediatesboth steroid 17a-hydroxylase and 17,20-lyase activities thatrepresent the key points in human steroidogenesis. A singlebase change in the 51 promoter region of the CYP17 creates an additional Sp-1-type promoter site which might causeincreased expression (37). A new recognition site was foundas two alleles (A1 and A2) in this RA study. Interestingly, the

II/68 Zeitschrift für Rheumatologie, Band 59, Suppl 2 (2000)© Steinkopff Verlag 2000

authors showed that female RA patients with the A2 alleletended to develop the disease at a younger age than thosewithout, and having the A2 allele was a protective factor fromolder age onset female RA. The results of the study suggestthat the A2 allele is related to early onset, and the A1 allele tothe late onset. As a matter of fact, the A2 allele, being expres-sion of increased CYP17 activities, is tought to be linked toelevated production of both estrogens and androgens throughincreased transcription. The authors suggest that the effects ofthe androgen increase induced by the A2 presence might notbe biologically influential in the fertile age (younger RAfemale patients), which is characterized by high estrogens(immunostimulant). However, the same induction of increasedproduction of androgens (immunosuppressive) might becomean influential resisting factor in older women, who are charac-terized by physiologically reduced estrogens and higherandrogens. Parallel to these results, with a large sample usingsibpair analysis methods, another study showed that the estro-gen synthase (CYP19) locus was linked to RA and that this

linkage was strongest in patients with onset after 50 years ofage, when serum hormone levels decline (41). The functionalactivities of the CYP19 gene variants are unknow at thepresent.

Acquired changes of sex steroid metabolism seem to alsoplay a role in the peripheral sex hormone levels. For example,an increase of the estrogen/androgen ratio in relation withenhanced peripheral enzymatic activities (i.e., aromatase =conversion of testosterone to 17b-estradiol; 17b-hydroxys-teroid dehydrogenase = conversion of estrone to more active17b-estradiol) induced by local inflammatory cytokines (IL-1, IL-6, TNFa), might explain low androgen levels found inRA synovial fluids and also in lacrimal/salivary fluids in RA-associated conditions such as Sjögren’s syndrome (SS) (42,43). As a consequence, local high levels of estrogen metabo-lites are observed in synovial fluids of RA patients (44).

In conclusion, a complex interaction between hypothala-mus-pituitary-adrenocortical and gonadal axis functions isevident in RA (45).

1. Cutolo M,Accardo S (1991) Sex hormones,HLA and rheumatoid arthritis. Clin ExpRheumatol 9: 641–646

2. Wilder RL (1995) Neuroendocrine-immune system interactions and autoim-munity. Annu Rev Immunol 13: 307–338

3. Cutolo M, Sulli A, Seriolo B, Accardo S,Masi AT (1995) Estrogens, the immuneresponse and autoimmunity. Clin ExpRheumatol 13: 217–226

4. Cutolo M, Castagnetta L (1996) Immuno-modulatory mechanisms mediated by sexhormones in rheumatoid arthritis. Ann N YAcad Sci 784: 534–541

5. Masi AT, Feigenbaum SL, Chatteron RT,Cutolo M (1995) Integrated Hormonal-Immunological-Vascular (“H-I-V” Triad)system interactions in rheumatic diseases.Clin Exp Rheumatol 13: 203–216

6. Chrousos GP (1995) The hypothalamic-pituitary-adrenal axis and immune-medi-ated inflammation. N Engl J Med 332:1351–1362

7. Man KS, Burleson M, Castro Z, Zautra AJ,Roth S (1992) Effects of stress on theneuroendocrine axis in rheumatoid arthritisand osteoarthritis. Arthritis Rheum 35:24–27

8. Haller C, Holzner B, Mur E, Gunther(1997) The impact of live events on patientswith rheumatoid arthritis: a psychologicalmyth? Clin Exp Rheumatol 15: 175–179

9. Masi AT, da Silva JAP, Cutolo M (1996)Perturbations of the hypothalamic pituitarygonadal axis in rheumatoid arthritis. In:Chikanza IC (ed) Neuroendocrine ImmuneMechanisms of Rheumatic Diseases. Bal-lière’s Clinical Rheumatology 10: 295–331

10. Cutolo M. Foppiani L, Prete C et al. (1999)Hypothalamic-pituitary-adrenocorticalaxis in premenopausal rheumatoid arthritispatients: not treated with glucocorticoids. JRheumatol 26: 282–288

11. Walker JG, Littlejohn G, McMurray NE,Cutolo M (1999) Stress system activation inrheumatoid arthritis: a multilevel approach.Rheumatol 38: 1050–1057

12. Cutolo M, Prete C, Walker J (1999) Is stressa factor in the pathogenesis in autoimmunerheumatic diseases? Clin Exp Rheumatol17: 515–518

13. Pietrini P, Guazzelli M (1997) Life eventsin the course of chronic diseases: a psycho-logical or a psycho-neuro-biochemicalloop? Clin Exp Rheumatol 15: 125–128

14. Gudbjörnsson B, Skogseid B, Öberg B,Wide L, Hällgren R (1996) Intact adreno-corticotropic hormone secretion butimpaired cortisol response in patients withactive rheumatoid arthritis. Effect of gluco-corticoids. J Rheumatol 23: 596–602

15. West HF (1993) Corticosteroid metabolismand rheumatoid arthritis. Ann Rheum Dis16: 173–181

16. Crofford LJ, Kalogeras KT, Mastorakos Get al. (1997) Circadian relationship betweeninterleukin (IL)-6 and hypothalamus-pitu-itary-adrenocortical axis hormones: failureof IL-6 to cause hypercortisolism inpatients with early untreated rheumatoidarthritis. J Clin Endocr Metab 82: 1279–1283

17. Masi AT, Chrousos GP (1996) Hypothala-mic-pituitary-adrenal-glucocorticoid axisfunction in rheumatoid arthritis. J Rheuma-tol 23: 577–581

18. Cutolo M. Foppiani L, Prete C et al. (1999)Hypothalamic-pituitary-adrenocorticalaxis in premenopausal rheumatoid arthritispatients: not treated with glucocorticoids. J Rheumatol 26: 282–288

19. Arvidson GN, Gudbjørsson G, Elfman L,Rydén AC, Titterman TH, Hällgren R(1994) Circadian rhythm of serum inter-leukin-6 in rheumatoid arthritis. AnnRheum Dis 53: 521–524

20. Templ E, Koeller M, Riedl M, Wagner O,Graninger W, Luger A (1996) Anteriorpituitary function in patients with newlydiagnosed rheumatoid arthritis. Br JRheumatol 35: 350–356

21. Mastorakos G, Chrousos GP, Weber JS(1993) Recombinant interleukin-6 activatesthe hypothalamic-pituitary-adrenal axis inhuman. J Clin Endocr Metab 77: 1690–1694

References

M. Cutolo et al. II/69The hypothalamic-pituitary-adrenocortical and gonadal axis function

22. Salas MA, Evans SW, Levell MJ, WhicherJT (1990) Interleukin-6 and ACTH actsynergistically to stimulate the release ofcorticosterone from adrenal gland cells.Clin Exp Immunol 79: 470–473

23. Dekkers JK, Greenen R, Godaert GLR, vanDoornen LJP, Bijlsma LWJ (2000) Diurnalrhythm of salivary cortisol levels in patientswith recent-onset rheumatoid arthritis.Arthitis Rheum 43: 465–467

24. Sukuki N, Suzuki T, Sakane T (1996) Hor-mones and lupus: defective DHEA activityinduces impaired interleukin-2 activity of Tlymphocytes in patients with systemiclupus erythematosus. Ann Med Interne147: 248–252

25. Hennebold JD, Poynter ME, Daynes RA(1995) DHEA and immune function: activ-ities and mechanism of action. Sem ReprodEndocrinol 13: 257–269

26. Cutolo M, Foppiani L, Giusti M et al.(1997) The adrenal response to oCRH andlow dose ACTH in premenopausal rheuma-toid arthritis patients. Arthritis Rheum 40:S253, 1331

27. Foppiani L, Sulli A, Prete C, Sessarego P,Seriolo B, Cutolo M (1999) Desmopressin,ovine CRH, and low-dose ACTH tests:tools for the study of the hypothalamic-pituitary-adrenal axis in premenopausalrheumatoid arthritis patients. Ann N Y AcadSci 876: 83–87

28. Cutolo M (1996) Effects of gonadal andro-gens on adrenal androgens. Clin Endocrinol44: 490–491

29. Bellido T, Jilka R, Boyce B et al. (1995)Regulation of interleukin-6, osteoclasto-genesis and bone mass by androgens. J ClinInvest 95: 2886–2895

30. Keller T, Chang C, Ershler WB (1996)Inhibition of NFkB activity through main-tenence of IkBa levels contributes todihydrotestosterone-mediated repression ofIL-6 promoter. J Biol Chem 271: 26267–26275

31. James WH (1997) Further evidence thatlow androgen values are a cause of rheuma-toid arthritis: the response of rheumatoidarthritis to seriously stressful life events.Ann Rheum Dis 56: 566

32. Cutolo M (1997) Do sex hormones modu-late synovial macrophages in rheumatoidarthritis? Ann Rheum Dis 56: 281–284

33. Wilder RL (1996) Adrenal and gonadalsteroid hormone deficiency in the patho-genesis of rheumatoid arthritis. J Rheuma-tol 44: 10–12

34. Kanda N, Tsichida T, Tamaki K (1997)Testosterone suppress anti-DNA antibodyproduction in peripheral blood mononu-clear cells from patients with systemiclupus erythematosus. Arthritis Rheum 40:1703–1711

35. Mira D, Sookhyun K, Rhonda RR (1997)Testosterone therapy ameliorates experi-mental autoimmune encephalomyelitis andinduces a T helper 2 bias in the autoantigen-specific T lymphocyte response. J Immunol159: 3–6

36. Huang J, Ushiyama K, Mori K, Hukuda S(1999) Possible association of CYP17 genepolymorphism with the onset of rheuma-toid arthritis. Clin Exp Rheumatol 17: 721–724

37. Cutolo M, Villaggio B, Sulli A, Seriolo B,Giusti M (2000) CYP17 polymorphism andandrogen levels in post-menopausalpatients with rheumatoid arthritis. Clin ExpRheumatol 18: 420–421

38. Carey AH, Waterworth D, Patel K et al.(1994) Polycystis ovaries and prematuremale pattern baldness with one allele of thesteroid metabolism gene CYP17. HumanMol Genet 3: 1873–1876

39. Cutolo M, Balleari E, Giusti M et al. (1986)Sex hormone status in women sufferingfrom rheumatoid arthritis. J Rheumatol 13:1019–1023

40. Cutolo M (1998) The role of the hypothal-amus-pituitary-adrenocortical and gonadalaxis in rheumatoid arthritis. Clin ExpRheumatol 16: 3–6

41. John S, Myecough A, Eyre S et al. (1999)Linkage of a marker in intron D of the estro-gen synthase locus to rheumatoid arthritis.Arthritis Rheum 42: 1617–1620

42. Macdiarmid F, Wang D, Duncan LJ et al.(1994) Stimulation of aromatase activity inbreast fibroblasts by tumor necrosis factoralpha. Mol Cell Endocr 106: 17–25

42. Nestler JE (1993) Interleukin-1 stimulatesthe aromatase activity of human placentalcytotrophoblasts. Endocrinol 132: 566–572

43. Castagnetta L, Cutolo M, Granata OM et al.(1999) Endocrine end-points in rheumatoidarthritis. Ann N Y Acad Sci 876: 180–187

44. Masi AT, Chrousos GP, Bornstein SR(1999) Enigmas of adrenal androgen andglucocorticoid dissociation in premeno-pausal onset rheumatoid arthritis. JRheumatol 26: 247–250