ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 11, No. 5 Copyright © 1981, Institute for Clinical Science, Inc.

Gold Nephropathy TATIANA T. ANTONOVYCH, M.D.

Registry of Nephropathology, American Registry of Pathology, Armed Forces Institute of Pathology, and the Veterans Administration Special Reference Laboratory for Pathology, Washington, DC 20306

ABSTRACT The early use of in medicine and dentistry dates back to the ancient Chinese and Egyptians. The discovery in 1890 that were toxic in vitro to tubercle bacilli led to the extensive treatment of tuberculosis with gold salts in the first three decades of this century. Eventually, gold therapy was extended to arthritis and lupus erythematosus, because of the belief that these diseases were forms of tuberculosis. Because of its beneficial effect particularly on active , chrysotherapy has remained one of the most widely used treatments of rheumatoid arthritis for the past half century. Toxicity of gold salts includes hypersensitivity reaction of skin and mucous membranes, bone marrow depression, and nephrotoxicity. The nephrotoxic clinical manifestations are renal insufficiency, proteinuria and hematuria, and the nephrotic syndrome. The pathologic changes are tubular degeneration, acute tubular necrosis or immune complex glomerulone­ phritis. The justification that any of these possible changes are the result of gold therapy rests clinically upon the time relationship of gold therapy and the renal symptoms, and pathologically upon the presence of gold inclusions (aurosomes) in proximal tubular epithelial cells. Aurosomes can at times be visualized by light microscopy, are usually seen by electron microscopy, and can be identified by microprobe analysis. Their pathology will be illustrated and pathogenic mechanisms discussed.

is most probably related to the total body Introduction concentration of gold, since only 40 per­ Chrysotherapy has been one of the most cent of gold is excreted each week on a widely used medications in the manage­ standard weekly dose, leading to a pro­ ment of patients with rheumatoid arthritis gressive increase in its body tissue stores. for the past 50 years, despite a rather high The most common toxic effects are those incidence of side effects. Untoward reac­ involving skin, mucous membranes, bone tions may occur early or late in the course marrow, and the kidney. The nephrotoxic of therapy. The incidence of toxic reaction clinical and pathological manifestations 386 GOLD NEPHROPATHY 387 have been well documented. Proteinuria, peruvic dehydrogenase suggest that the hematuria, the nephrotic syndrome, and therapeutic effects of gold salts are due to renal insufficiency are well recognized inhibition of sulfhydryl-related enzymes. complications of gold therapy. The renal The most commonly used gold com­ lesions associated with chrysotherapy in­ pounds in the treatment of rheumatoid ar­ clude acute tubular necrosis, tubular de­ thritis are (Solganal) and generation, and immune complex gold sodium thiomalate (Myochrysin). glomerulonephritis. Similar lesions can Both compounds contain approximately be produced in experimental animals 50 percent gold. Water soluble com­ with gold salts. The pathogenesis of the pounds are rapidly absorbed after in­ tubular lesions seem to be a direct drug tramuscular injection. Serum gold con­ “toxicity” of gold ions, while the immune centration peaks within two to four hours, complex glom erulonephritis is an im- after which it declines gradually. Ninety- munologically mediated drug toxicity. two percent of gold in the blood is protein The earliest use of gold in medicine and bound, of which 95 percent is present in dentistry dates back to the ancient Egyp­ the albumin fraction.10 Gold is widely dis­ tians and Chinese.21 The therapeutic use tributed throughout the body.20 The high­ of gold in modem medicine was stimu­ est concentration of soluble compounds is lated by the observation of Robert Koch of found in the kidneys, while insoluble the adverse reaction of gold salts on the compounds are recovered in larger tubercle bacillus in 1890.8 Despite disap­ amounts from the reticulo-endothelial pointing clinical results, gold was used in system. Gold compounds are selectively the treatment of tuberculosis, particularly concentrated within the inflamed syno­ in western Europe, until the late 1930’s. vial tissue of active rheumatoid arthritis. Eventually gold therapy was extended to Soluble gold compounds are excreted the treatment of arthritis and lupus ery- mainly by the kidney, while insoluble thematosis because of the belief of some compounds are excreted in the bile. How­ that these diseases were forms of tuber­ ever, only 40 percent of administered gold culosis. The treatment of rheumatoid ar­ is excreted each week during standard thritis with gold, however, was first treatment; therefore, with weekly in­ popularized by Jaques Forestier in the jections, the body stores increase progres­ early 1930’s.9 He reasoned that gold prob­ sively. Gold was found to persist in ably stimulated the defense mechanism of various tissues up to 23 years after chry­ the body in tuberculosis and might be­ sotherapy was stopped. In the past, unduly have similarly in rheumatoid arthritis. large doses of gold compounds used in the Fifty years of use of gold in the man­ treatment of tuberculosis and later of agement of rheumatoid arthritis has had a rheumatoid arthritis resulted in a high in­ mixed acceptance. Despite a rather high cidence of severe toxic reactions.3 Unto­ incidence of side effects, some prognosti- ward reactions to gold may occur early or cally serious, chrysotherapy remains one later in the course of therapy. The inci­ of the most widely prescribed medica­ dence of toxic reactions to gold is unre­ tions in the management of patients with lated to its plasma levels but is probably rheumatoid arthritis. related to its total body concentration. The Gold exists in monovalent and trivalent most common toxic effects of chryso­ forms. All significant preparations are au- therapy are those involving the skin and reous (monovalent) salts in which gold is the mucous membranes. Less common attached to sulfur. The strong affinity of but more severe in their consequences are monovalent gold for sulfur and the in­ the hematologic and renal manifestations hibitory affects of certain gold salts on of gold toxicity.6'13 388 ANTONOVYCH

Renal symptoms of gold toxicity range Gold deposits (aurosomes) of soluble from renal proteinuria and hematuria and gold compounds consist of electron dense the nephrotic syndrome13,14 to renal insuf­ granules, their projecting rod like and ficiency and even fatal anuric failure.3,11 linear structure giving them a feathery The renal lesions associated with chryso- appearance, whereas aurosomes pro­ therapy have been well documented. duced by colloidal gold are made up of Large doses of gold salts produce acute aggregates of spherical electron dense tubular necrosis.3 The main lesion is con­ structures (figure 2).J2,17,18,20 fined to the proximal convoluted tubules. The primary site of gold deposition in The evolution of the tubular lesion was the kidney is the m itochondria of the prox­ extensively studied in experimental ani­ imal convoluted tubules. The damage to m als.12,17,18 the mitochondria is proportional to the By light microscopy, in hematoxylin amount of gold injected. Accumulation of and eosin stained sections, the proximal gold results in mitochondrial disruption tubular cells appear swollen and show and consequent degeneration and necro­ areas of rarifaction or vacuolization in sis of the lining cells. The tubular base­ which at times small yellowish-brown, re- ment membrane remains intact. The fractile, poorly defined granules can be distal tubules become filled with cast of recognized. These may be shown in light cellular debris and clumps of gold microscopic sections to be gold by a deposits.12,17,18 photochemical method described by The percentage of patients reported Gilg,7 or recognized by their distinct mor­ who develop proteinuria or the nephrotic phologic structure by electron microscopy syndrome during chrysotherapy varies or identified by electron probe micro­ from 0.2 to 10 percent. Although most pa­ analysis (figure 1). tients presenting with proteinuria have

F ig u r e 1. Light micro­ scopic section of proximal convoluted tubule. The cytoplasm is vacuolated and shows poorly defined refractile granules. Hem­ atoxylin and eosin x 630 (AFIP 80-10684). GOLD NEPHROPATHY 389

F ig u r e 2. Electro­ micrograph of a portion of proximal convoluted tu­ bules, showing de­ generating mitochondria and gold inclusions, x 10,800 (AFIP 80-10685).

received a substantial amount of gold, pro­ in endocapillary cells, or show focal or teinuria has been reported after as little as diffuse thickening of capillary walls (fig­ 10 mg of gold sodium thiomalate. The ure 3). The thickened capillary walls in glomeruli in renal biopsies of these pa­ methenamine silver stained sections tients viewed by light microscopy may show basement membrane spiking and, appear normal, or show minimal increase in trichrome stained sections, granular

F ig u r e 3. Light micro­ scopic section of glomeru­ lus showing focal thicken­ ing of capillary walls. 5?t ir v * Hematoxylin and eosin x • > * :• 280 (AFIP 80-10683). ' * r 390 ANTONOVYCH

F ig u r e 4. E lectro- micrograph of a portion of glomerulus. Widely spaced subepithelial de­ posits with early base­ ment membrane spiking is present. There is diffuse effacement of epithelial cell foot processes, x 3700 (AFIP 80-10686).

magenta red deposits. By electron micros­ segmental, the prognosis is good. First, copy, electron dense deposits are seen on there is progressive and, finally, complete the epithelial surface of the glomerular disappearance of proteinuria, followed by capillary basement membranes (figure loss of deposits and restoration of 4).2-4 Depending on the time interval be­ glomerular capillary basement membrane tween the first gold injection, the appear­ as revealed by repeated biopsies.4 ance of proteinuria, and the day of the The tubular damage in the course of kidney biopsy, the deposits may be few chrysotherapy is caused by the deposition and the underlying basement membrane of gold inside mitochondria, which appear intact, or there may be basement mem­ to be the target organelles of injury.17,23 brane spiking or finely incorporation of The exact pathogenesis of mitochondrial deposits into the thickened remodeled injury is not known, but gold probably basement membrane. Thus, all stages of enters the mitochondria in ionic form, membranous glomerulonephritis may be since it can be seen in mitochondria seen in gold glomerulonephropathy ex­ which have an intact outer membrane. It cept that in the latter, irrespective of the is possible that they become overloaded stage, the number of deposits varies and a by large numbers of gold ions, which number of loops may be totally free of interfere with their normal functions of deposits.16 If properly searched for, auro- selective binding and unbinding of somes will be found in proximal tubular cations. epithelium and rarely, visceral glomeru­ Several hypothesis have been proposed lar epithelial cells.18,19 Fluorescence concerning the pathogenesis of the im­ microscopy shows the deposits to contain mune complex glomerulonephritis re­ IgG and complement in a granular pattern lated to gold therapy. It was originally along the glomerular capillary basement suggested that gold acts as a hapten when membrane. If gold therapy is discon­ combined with tissue proteins; however, tinued soon after the proteinuria appears, up to date it has never been detected and the glomerular lesions are early and within the immune deposits. It has been GOLD NEPHROPATHY 391 suggested that the deposits may be com­ 6. G ih b o n s , R. B.: Complications of chryso- therapy. Arch. Intern. Med. 139:343-346,1979. plexes of rheumatoid autoantibodies and 7. GlGL, E.: A photochemical method for micro­ that gold may facilitate the glomerular detection of gold in tissue sections. Acta deposition.15 Another hypothesis, recently Psychiat. Scand. 2 7 :43-46, 1952. 8. G o o d m a n , L. S. and G ilm a n , A.: The Phar­ widely supported, is that the gold- macological Basis of Therapeutics, 4th ed. damaged tubules liberate autoantigens, Philadelphia, The MacMillan Company, 1970, which are then deposited with antibody in pp. 969-974. 9. HERSPERGER, W. G.: Gold therapy for the glomeruli.11 Finally, there is growing rheumatoid arthritis: A current evaluation. Ann. evidence that immunologically mediated Int. Med. 36:571-582, 1952. drug toxicity may be controlled by genes 10. International W orkshop and Symposium on G old Therapy in Rheumatic Disease. J. of the histocompatibility system. It is Rheumat. Suppl. 5, 1979. known that patients with rheumatoid ar­ 11. K a t z , A. and L i t t l e , A. H.: Gold nephropathy. thritis may have HLA-88 or HLA-DRw3 Arch. Path. 96:133-136, 1973. 12. N a g i, A. H., Alexander, F., and Barabas, A. antigens. It was recently reported that Z.: Gold nephropathy in rats—Light and elec­ DRw3 positive patients with rheumatoid tron microscopic studies. Exp. Mol. Path. arthritis have a high titer of rheumatoid J5:354-362, 1971. 13. S a m u e ls , B., L e e , J. C., Engleman, E. P., and factor and treatment with aurothiomal- H o p p e r , J.: Membranous nephropathy in pa­ ate1 results in a high incidence of tients with rheumatoid arthritis. Relation to proteinuria.22 gold therapy. Medicine 57:319-327, 1977. 14. Silverberg, D. S., K id d , E. G., S h n i t k a , T. K., The diagnosis of gold induced mem­ and U l a n , R. A.: Gold nephropathy. Arthritis branous glomerulonephritis is justified Rheum. 13:812-825, 1970. only when there is a relationship between 15. Skrifvars, B. V., T ornroth, T. S., and Tallqvist, G. N.: Gold induced immune com­ gold therapy and renal symptoms, and au- plex nephritis in seronegative rheumatoid ar­ rosomes are found in the proximal convo­ thritis. Ann. Rheum Dis. 36:549-556, 1977. luted tubules. At times they are difficult to 16. Tornroth, T. and Skrifvars, B.: Gold ne­ phropathy prototype of membranous glomer­ detect in conventionally stained electron ulonephritis. Amer. J. Path. 75:573-584, 1974. microscopic sections, but they are rela­ 17. S t u v e , J. and G a l l e , P.: Role of mitochrondria tively easily visualised in unosmicated in the handling of gold by the kidney. J. Cell. Biol. 44:667-676, 1970. unstained sections.5 18. Strunk, S. W. and Z i f f , M.: Ultrastructural studies of the passage of gold thiomalate across References the renal glomerular capillary wall. Arthritis Rheum. 13:39-52, 1970. 1. Bu r g e r , R., B r in e r , J., G ü r t l e r , B., and 19. V a a m o n d e , C. A. and H u n t , F. R.: The ne­ BlNSWANGER, U.: Unusual incidence of mem­ phrotic syndrome as a complication of gold branous glomerulonephritis after gold therapy therapy. Arthritis Rheum. 13:826-833, 1970. for rheumatoid arthritis. Kidney Internat. 15 - 20. Vernon-Roberts, B., D o r e , J. L., J e s s o p , J. 5:584, 1979. D., and Henderson, W. J.: Selective concen­ 2. D a v ie s , D . J., D o w l in g , J., and Xip e l l , J. M.: tration and localization of gold in macrophages Gold nephropathy. Pathology 9:281-288,1977. of synovial and other tissues during and after 3. D e r o t , M., Ka h n , J., Ma z a l t o n , A., and chrysotherapy in rheumatoid patients. Ann. PEYRAFORT, J.: Nephrite anurique aigue mor- Rheum. Dis. 35:477-486, 1976. telle apres traitement aurique chrysocyanose 21. W ilson, W. J.: Alchemy in China. Ciba Sym- associee. Bull. Mem. Soc. Med. Hop. 70:244- pos. 2:594-623, 1940. 239, 1959. 22. W ooley, P. H., G r i f f i n , J., Panayi, G. S., 4. D itscherlein , G., Sc h n e id e r , W., Mu l l e r , B atchelor, J. R., W e l s h , K. I., and G ib s o n , T. V., Kr u m h a a r , I., and STORCH, W.: Long term J.: HLA-DR antigens and toxic reaction to observation on membranous glomerulo­ and D- in nephritis due to gold and penicillamine patients with rheumatoid arthritis. New Eng. J. therapy. Internat. Acad. Path. 195, 1980. Med. 303:300-302, 1980. 5. G h a d ia l l y , F. N., T h o m a s, I., and L a l o n d e , J. 23. Y a ro m , R., Stein, H., P eters, P. D., S l a v i n , M. A.: Comparative ultrastructural morphology S., and H a l l , T. A.: Nephrotoxic effect of par­ of aurosomes produced by collodial gold and enteral and intraarticular gold. Arch. Path. soluble gold salts. J. Path. 123:181-185, 1977. 99:36-43, 1975.