Arthritis and Rheumatism O$cial Journal of- the gmeuican Rheumatismassociation FEBRUARY, 1962 VOL. V, NO. 1 The Kidney in Rheumatoid Arthritis : Studies by Renal Biopsy By VICTORE. POLLAK, CONRAD L. PIRANI, IRVING E. STECKAND ROBERT M. KARK Renal biospsies and renal function tests were done on 41 rheumatoid arthritis patients. Histologically, vascular disease was found in 14, amyloidosis in 4, and lupus nephritis in 2. The kidney was‘ normal in 21. “Rheumatoid glomerulitis was not found in any biopsy, but occurred in 6 of 16 autopsies. The significance of these results is discussed with particular reference to the effect of gold salts on the kidney and the interrelationships of cortisone treatment and amyloidosis. Biopsias renal e tests de function renal esseva effectuate in 41 patientes de arthritis rheumatoide. Histologicamente, morbo vascular esseva trovate in 14, amyloidosis in 4, e nephritis a lupus in 2. Le ren esseva normal in 21. Glomerulitis “rheumatoide” non esseva trovate in ulle biopsia sed Occurreva in 6 inter 16 necropsias. Le signification de iste resultatos es discutite con referentias particular a1 effect0 de sales de auro in le ren e a1 relation inter therapia a cortisona e amyloidosis. G LOMERULITIS has been observed relatively frequently in postmortem studies on the kidneys of patients dying with rheumatoid arthritis.1*9 During life, however, proteinuria, abnormalities of the urinary sediment, and decompensation of renal function have been reported to be relatively uncomWhen clinical evidence of reiial mon in patients with rheumatoid abnormalities has been found, unrelated intrinsic renal disease or amyloidosis secondary to rheumatoid arthritis have often been implicated to sxplain the proteinuria or a z ~ t e m i a . ~Clearly, v ~ * ~ further studies on the kidney in patients with rheumatoid arthritis are needed, as emphasized by Short, Bauer, and Reynold~.~ They considered ( 1) that more “histologic observations on the kidney in rheumatoid arthritis derived from autopsy and biopsy material” were important, and ( 2 ) that “quantitative measurements of the excretion of of albumin, cellular elements, and casts in this disease along with tests of renal function by modern techniques,” were necessary. From the Departments of Medicine, Presbyterian-St. Luke’s, Cook County, and Research and Educational Hospitals, and from the Departments of Medicine and Pathology, University of Illinois College of Medicine, Chicago, 111. Presented in part at the Annual Meeting of the American Rheumatism Association, June 90-21, 1958, San Francisco, Calif. Supported by grants from tlxd Illinois Chapter of the Arthritis and Rheumatism Foundution, and from the U . S . Public Health Seruice, Bethesda, Md. (H-2253 and A-968). 1 ARTHRITIS& RHEUMATISM,VOL. 5, No. 1 (FEBRUARY), 1962 2 POLLAK, PIRANI, STECK AND KARK In this communication we will report on clinical findings and rend biopsy studies in 41 patients with rheumatoid arthritis. In studying the renal biopsies, particular attention was paid to the glomerulitis reported by others in necropsy studies,lS2J’ to the diagnosis of lupus nephritis,1° to renal amyloidosis, and to possible histologic evidence of toxic effects in the nephron due to treatment with gold salts. MATERIAL AND METHODS Clinical. Most patients were admitted to hospital for study from the Arthritis Clinic of the University of Illinois Hospitals; a few were studied in other hospitals. In all, 41 patients were investigated. The diagnosis of rheumatoid arthritis was made using the criteria of the American Rheumatism Association.11 None had ankylosing spondylitis. The patients were not selected a t random. Many were selected because of long-standing severe rheumatoid arthritis with marked deformities. Some patients were chosen for study because proteinuria liad been found on routine urinalysis in the outpatient clinic, and others because they had been treated with gold salts. In fact, the patients were selected for renal biopsy because it was thought that they were the patients most likely to have renal abnormalities. Tha average duration of rheumatoid arthritis was over 8 years (range: 2 months to 25 years). Eighteen patients had been treated with gold salts for periods of 12 weeks to more than 4 years; thirteen had received at least one Gm. of Myochrisine. Twenty-one had received ACTH or cortisol and its analogs. Most patients had been treated with 37.5 to 75 mg. of cortisol daily or its equivalent. Eight had been treated with both gold salts and ACTH or cortisol. Eleven had received neither drug. Measurements of renu2 functions. The methods used for the measurement of renal functions and for urinalysis have been described elsewhere.10 Renal biopsies. The methods used to obtain specimens of renal tissue by percutaneous needle biopsy have been described, as have the methods of handling the renal tissue.10 Where indicated, sections stained with Congo red and Crystal violet were examined for ainyloid deposits.12 RESULTS Histologic (Table 1).No histologic abnormality was detected in the kidneys of 21 patients. Arterio- and arteriolosclerosis or arteriolar nephrosclerosis were the most common histologic abnormalities, and were observed in 14 biopsies. In most patients these lesions were relatively mild. The average age of the patients with vascular changes was 50 years; that of those with normal kidneys was 37 years. Amyloidosis was found in the kidneys of four patients, three of whom were being treated with cortisol. Lupus nephritis was diagnosed in two. Other evidence of systemic lupus erythematosus was subsequently demonstrated in both patients with lupus nephritis.* Glomerulitis of the type described in rhcumatoid arthritis by Baggenstoss and Rosenbergl and by Fingerman and Andrus2 was not observed. *Both of these patients were admitted to hospital for study from the Arthritis Clinic where the diagnosis of rheumatoid arthritis had been made according to the. criteria of the American Rheumatism Association. The finding of lesions consistent with lupus nephritis and the subsequent clinical evolution and laboratory findings led us to conclude that thc correct diagnosis was systemic lupus erythematosus. 3 THE KIDNEY IN RHEUMATOID ARTHRITIS Table 1.-Histologic Diagnoses in Patieiits with Rheumatoid Arthritis on Whom Renal Riovsies Were Made Histologic diagnosis Normal kidney Arterio- & arteriolosclerosis Arteriolar nephrosclerosis Amyloidosis Lupus nephritis “Rheumatoid” glomerulitis Total Patients with proteinuria Patients with no proteinuria Total 3 2 2 4 2 0 18 10 0 0 0 0 21 12 2 4 2 0 13 28 41 CZinical. Thirteen patients had persistent proteinuria ( table I). I: was associated with amyloidosis in four patients, lupus nephritis in two, and arteriolar nephrosclerosis in two. Proteinuria also occurred in two patients with rrnal arterial and arteriolar sclerosis but no interstitial or glomerular fibrosis. It is uncertain whether or not the vascular disease was the cause of the proteinuria. In three patients with normal kidneys, no underlying histologic abnormality was found which might explain the proteinuria. Tests of renal function were within normal limits in almost all the patients. After dehydration overnight for 14 hours, the average urinary specific gravity was 1.026; the average serum creatinine and urea nitrogen levels were 0.8 mg./100 ml. and 12.0 mg./100 ml. respectively. The average creatinine clearance was 100 ml./min., and the average urea clearance was 85 per cent of average normal renal function. Only five patients had a slight elevation of the serum urea nitrogen level. Two had arteriolar nephrosclerosis and two had renal amyloidosis. No renal lesion was found in the fifth patient. The &ect of gold salts o n the kidnay. One patient first became ill with rheumatoid arthritis in the summer of 1955. When she first attended the Arthritis Clinic one year later, her urine contained no protein. Shcrtly after treatment with Myochrisine was started, protein was noted in her urine, and proteinuria persisted thereafter. A renal biopsy was done in February 1958. Arteriosclerosis was diagnosed but, in addition, there was some degeneration and fatty metamorphosis of the convoluted tubules. Whether or not these nonspecific degenerative changes in the tubules were due to gold treatment must remain an open question at present. We found no histologic evidence for a toxic effect of gold salts on the kidney in the other 17 patients who had received this drug. Proteinuria was noted in only 4 of the 18 patients treated with gold salts and in 3 of 17 patients who had never received gold salts (excluding from consideration patients with lupus nephritis and renal amyloidosis ) . DISCUSSION The incidence and cause of proteinicrin in rheunzatoid arthritis. In the second edition of Osler’s Modern Medicine, McCrae reported that proteinuria occurred in 20 per cent of 500 patients with arthritis deformans, including those 4 POLLAK, PJRANI, STECK AND KARK with spondylitis. Cylinduria occurred in over one-half of those with protein~ r i a Eaton . ~ and Cocheu found albumin in the urine in 48 of 111 cases with “hypertrophic arthritis” and in 48 of 170 cases of “atrophic arthritis,” but the criteria used by these authors for the diagnosis of these disorders were not clearly delineated in their paper.13 In 1943 Fingerman and Andrus reported on a retrospective study of 61 ca5es of rheumatoid arthritis from the autopsy files of the University of Minnesota Hospitals. Protein had been observed in the urine during life in IS of these patients, at least 9 of whom were shown to have renal amyloidosis.2 In 1955 Fearnley and Lackner5 attempted to assess the incidence and significance of proteinuria in rheumatoid arthritis. They studied all 183 patients attending an arthritis clinic during a one year period, and tested the urine for protein at each visit to the clinic. Persistent proteinuria was found in 24. The proteinuria appeared to be associated with some obvious renal disease such as glomerulonephritis or pyelcnephritis or with congestive cardiac failure in 15 patients. Amyloidosis was demonstrated by retention of Congo red or by liver bic;psy in 8 cf the remaiaing 12 patients; the other 4 could not be investigated for the underlying cause of proteinuria. Short, Bauer and Reynolds havc reported detailed cbjervations cn 293 patients with rheumatoid arthritis. They attempted to exclude frcm consideration those patients with primary urinary tract disorders, and noted that seven per cent of their patients had proteinuria not associated with primary urinary tract disease.3 Our observations on a selected group of patients suggest that when proteinuria occurs in patients with rheumatoid arthritis, it is usually not attributable directly to rheumatoid involvement of the kidney; it may result from complicating amyloidosis, cr from other diseases such as lupus nephritis, or frcm renal vascula; disease. The incidence of amyloidosis in rheumntoid arthritis. Secondary arriyloidosis in patients with rheumatoid arthritis may cause the nephrotic syndrome or death in renal failure.7.R~1~*’~ We found amybid deposits in the renal biopsies of four patients, but this does not reflect the incidence of amyloidosis in rheumatoid arthritis as the patients were selected. The best available estimate of the incidence of amyloidosis appears to be that of Fearnley and Lackner. They studied 183 unselected patients with rheumatoid arthritis attending an arthritis clinic in a one year period, and found amyloidosis in eight.5 All had proteinuria, and the diagnosis of amyloidosis was made by liver biopsy, by necropsy, or by the Congo red test. Unger and his colleagues examined 56 patients with rheumatoid arthritis and found evidence of amyloidosis in six.l6 These clinical observations indicate that amyloidosis is the most common cause of serious renal disease in patients with rheumatoid arthritis. In postmortem studies, amyloid disease has also been found frequently. In a careful review of the literature, Missen and Taylor found 42 acceptable cases of amyloidosis complicating rheumatoid arthritis in reports by others of small series or isolated cases1? They also reviewed the incidence of amyloidosis in several larger postmortem series of patients with rheumatoid arthritis. A total of 66 instances of amyloidosis was found among 353 cases studied at autopsy 5 THE KIDNEY IN RHEUMATOID ARTHRITIS Table 2.--Incidence of Amyloidosis in Rheumatoid Arthritis (Autopsy Studies! Authors Teilum & Lindahl ( 19-54)? Gedda (1955)s Fingerman & Andrus ( 1943)2 Missen & Taylor (1956)17 Solomon (1943)lS Bayles ( 1943) 19 Sinclair & Cruickshank (1956)e Young & Schwedel (1944)zO Unger et al. (1948)lO Baggenstoss & Rosenberg (1943)l Total Amyloidosia Number of autopsies Number Per cent 28 45 61 47 7 23 16 38 58 30 17 11 13 8 1 3 2 5 4 2 61 24 21 17 14 13 13 13 7 7 353 66 18.7 (table 2) .' No other cause of amyloidosis was present in 57, whereas in 9 there was another possible cause such as tuberculosis in addition to rheumatoid arthritis. In most of the series the reported incidence of amyloidosis was less than 20 per cent (table 2),2l but Teilum and Lindahl found a much higher incidence of amyloidosis.7 They restudied the autopsies on 28 patients and found severe amyloid disease in 10 and lesser deposits of amyloid h i 7. All 10 patients with severe disease had had proteinuria, and 7 had died in uremia. They had been ill with rheumatoid arthritis for an average of 28 years. Four of those with less severe disease had deposits of amyloid in the kidney, but none died in uremia. They had had rheumatoid arthritis for an average of 12 years. These observations re-emphasize the importance of searching liistologic sections diligently for evidence of amyloidosis. Of the 17 cases of ainyloiciosis diagnosed by these authors, only 2 had been diagnosed when the init;al routine postmortem examination was made. The other 15 were brought to light by the systematic study of the renal and other tissues after staining for amyloid with both Congo red and methyl violet. Possible effects of treatment with cortbol and its analogs on the imclmce of amyloidosis. Teilum has demonstrated that when mice were trtated for several weeks with injections of casein in amounts insufficient to produce amyloidosis, a few injections of cortisone resulted in the prompt appearance of amyloid in the spleenJ2 These observations were confirmed by Christenset~~3 On the other hand, cortisone administration started prior to and continued during casein treatment was found either to reduce the incidence or to pre'In their paper, Missen and Taylor include one series of patients with rheumatoid arthritis which has been excluded from table 2. This was the series of Bennett.31 Missen and Taylor quote Bennett as having found no case of amyloidosis among 48 autopsies on patients with rheumatoid arthritis. However, it is clear from the discussion which follows Bennett's paper that Bennett found 11 cases of amyloid infiltration. Although it seems likely that these were among the patients with rheumatoid arthritis studied by him, it is not clear from this discussion whether the 11 cases of amyloid infiltration were found only among patients with rheumatoid arthritis or among patients with both rheumatoid arthritis and rheumatic fever. For this reason Bennett's series have been omitted from table 2. 6 POLLAK, PIRANI, STECK AND KARK vent the formation of a m y l ~ i d .An ~ ~analogous *~~ situation might conceivably apply to cortisol treatment of rheumatoid arthritis in man, and a few observations have been reported which suggest a possible precipitating effect of ACTH or cortisol on amyloid formation.25vz6The possibility that amyloidosis results from cortisol treatment in our patients cannot be excluded, but it is pertinent to recall that amyloidosis was found frequently in patients with rheumatoid arthritis before the introduction of treatment with ACTH and cortisol (table 2 ) . Certainly there is as yet no proof of an amyloid-promoting effect of cortisol therapy in man. Well controlled serial biopsy studies will be necessary to prove this point. Glomerulitis in rheumatoid arthritis. The incidence of glornerulitis in three series of postmortem studies is given in table 3. Glomerulitis was fouiid in 28 of 107 autopsies on patients with rheumatoid arthritis; we h w e found glomerulitis in 8 of lG autopsies."7 The lesions were characterized by mild endothelial hyperplasia without significant epithelial cell changes and without exudative features. There was little or no evidence of titbular degeneration nor of interstitial fibrosis and inflammation. In striking contrast with these findings was the complete absence of glomerulitis in the tissues studied by renal biopsy. It is unlikely that the absence of glomerulitis in biopsy specimens was due to sampling error, as the endothelial hyperplasia seen in postmortem specimens was diffuse and involved all or almost all glomeruli. The highest incidence (63 per cent) of glomerulitis was found by Baggenstoss and R0senberg.l Nine of their 19 patients with glomerulitis had acute or chronic pulmonary infections when they died. Bell has demonstrated a proliferative glomerulitis in patients dying of pneumonia and other bacterial but Baggenstoss and Rosenberg pointed out that the pdmonary infection could not account per se for the high incidence of glomerulitis. T!iey concluded that these lesions were probably due to or associated with the rheumatoid disease. If they are indeed a manifestation of rheumatoid disease, the failure to find such lesions in renal biopsies suggests that glomerulitis may be a late or preterminal manifestation, and that it has little clinical significance. This view is supported by the observations of Duihie and his colleagues, who restudied 283 patients with rheumatoid arthritis an average of six ycars after their discharge from hospital for active rheumatoid diseaseGAt this time the average duration of the disease was about 12 years. Of the 36 patients who died, 4 died of renal disease, 3 of pyelonephritis, and 1of amyloidosis. There is e~perirnental~~ and clinical evidence30 to suggest that either the Table 3.--Incidence of Glomerulitis in Rheumatoid Arthritis (Autopsy Studies) Authors Baggenstoss & Rosenberg (1943)l Fingerman & Andrus ( 1943)z Sinclair & Cruickshank (1956)9 Total Number of Glomerulitia autopsies Number Percentage 30 61 16 19 8 1 107 28 63 13 6 26 7 THE KIDNEY IN RHEUMATOID ARTHRITIS pathcgenic stimulus to the glomerular endothelial cells or their reactivity to such a stimulus may be controlled by adequate treatment with cortisone. Thus, it is conceivable but unlikely that the absence of glomerulitis might be explained in part by the fact that cne-half of the patients studied had been treated with ACTH or cortisol and its analogs. SUMMARY 1. Renal biopsies and tests of renal function were made on 41 selected patients with rheumatoid arthritis. The patients were chosen for studv because it was thought that these patients were most likely to have renal abnormalities. 2. No histologic abnormality was found in 21 cases. Arterio- and arteriolosclerosis or arteriolar nephrosclerosis was found in 14 biopsies. Amyloidosis occurred in 4 biopsies, and lupus nephritis was diagnosed in 2. Glomerulitis of the type described by others in postmcrtem studies was not observed. 3. Proteinuria was observed persistently in 13 patients. It was associated with amyloidosis in 4, with lupus nephritis in 2, and with renal vascular disease in 4.In 3 patients no underlying histolcgic abnormality was found. Tests of renal function were within normal limits in almost all the patients. Slight elevation of the serum urea nitrogen level was found in only 5, 2 of whom had renal amyloidosis and 2 had arteriolar nephrosclerosis. 4. Eighteen patients had been treated with gold salts. Although proteinuria first appeared in one patient after treatment with gold salts was started, in the series as a whole proteinuria occurred as frequently in those who had never received gold therapy as in thcse who were being treated with gold. We found no histologic evidence for a toxic effect of gold salts on the kidney except for the presence of slight fatty metamorphosis in the convoluted tubules of one patient. REFERENCES l. Bsggenstoss, A. H., and Rosmnberg, E. 2. 3. 4. 5. F.: Visceral lesions associated with chronic infectious (rheumatoid) arthritis. Arch. Path. 35:503, 1943. Fiqgerman, D. L., and Andrus, F. C.: Visceral lesions associated with rheumatoid arthritis. Ann. Rhemat. Dis. 3:168, 1943. Short, C. L., Bauer, W., and Reynolds, W. E.: Rheumatoid Arthritis. Cambridge, Mass., Harvard University Press, 1957. McCrae, T.: Arthritis Deformans. In Modern Medicine, 2nd ed., Osler, W. and McCrae, T., eds.: Philadelphia, Lea & Febiger, 1913-1915. Fearnley, G. R., and Lackner, R.: Amyloidosis in rheumatoid arthritis, 6. 7. 8. 9. and significance of “unexplainsd” albuminuria. A report of eight cases. Brit. Med. J. 1:1129, 1955. Duthie, J. J. R., Brown, P. E., Knox, J. D. E., and Thompson, M.: Course and prognosis in rheumatoid arthritis. Ann Rheumat. Dis. 16:411, 1957. Teilum, G., and Lindahl, A.: Frequency and significance of amyIoid changes in rheumatoid arthritis. Acta. med. scandinav. 149:449, 1954. Gedda, P. 0.: On amyloidosis and other causes of death in rheumatoid arthritis. Acta med scandinav. 150:443, 1955. . Sinclair, R. J. G., and Cruickshank, B. A.: Clinical and pathological study of sixteen cases of rheumatoid arthri- 8 POLLAK, PIRANI, STECK AND KARK tis with extensive visceral involvement 20. Young, D., and Schwedel, J. B.: The (“rheumatoid disease”). Quart. J. heart in rheumatoid arthritis: A study Med. 25:313, 1956. of thirty-eight autopsy cases. Am. 10. Muehrcke, R. C., Kark, R. M., Pirani, Heart J. 28:1, 1944. C. L., and Pollak, V. E.: Lupus 21. Ropes, M. W.: Massachusetts General nephritis: a clinical and pathologic Hospital Case No. 31212. New England J. Med. 232:632, 1945. study based on renal biopsies. Medi- . 22. Teilum, G. : Cortisone-ascorbic acid cine 36:1, 1957. 11. Ropes, M. W., Bennett, G. A,, Cobb, interaction and the pathogenesis of S., Jacox, R., and Jessar, R. A.: amyloidosis. Mechanism of action of Diagnostic criteria for rheumatoid cortisone on mesenchymal tissue. arthritis. Bull. Rheumat. Dis. 7 :121, Ann. Rheumat. Dis. 11:119, 1952. 1956, and Ann. Rheumat. Dis. 16: 23. Christensen, H. E.: Effects of cortisone in experimental mouse amyloidosis. 118, 1957. 12. Lillie, R. D.: Histopathologic Terhnic Proc. Fourth Internat. Congr. Rheuand Practical Histochemistry. New matology, Rome, 1961. 24. Bestetti, A., Pirani, C. L., and CatchYork, Blakiston, 1954. 13. Eaton, E. R., and Cocheu, L. F.: pole, H. R.: Studies on experimental Chronic arthritis. Original biochemiamyloidosis. Arth. & Rheumat. I: 274, 1958. cal studies with a review of the literature. J. Am. Inst. Homeop. 25: 25. West, H. F., and Newns, C. R.: Some effects of long-continued cortisone 485, 1932. therapy in rheumatoid arthritis. Lan14. Trasoff, A., Schneeberg, N., and Scarf, M.: Recovery from multiple rheumncet 2:515, 1952. toid arthritis complicated by amy- 26. Frenkel, M., and Groen, J.: Amyloidosis na behandeling met ACTH. Nederl. loidosis in a child. Report of a case tijdschr. geneesk. 98:2352, 1954. and review of the literature. Arch. 27. Pirani, C. L., and Pollak, V. E.: UnInt. Med. 74:4, 1944. published observations. 15. Reece, J. M., and Reynolds, T. B.: Amyloidosis complicating rheumatoid 28. Bell, E. T.: Renal Diseases, ed. 2. Philadelphia, Lea & Febiger, 1950. arthritis. Am. J. Med. Sc. 228:554, 29. Teilum, G., Engbaek, H. C., Harboe, 1954. 16. Unger, P. N., Zuckerbrod, M., Beck, G. N., and Simonsen, M.: Effects of J., and Steele, J. M.: Amyloidosis in cortisone on experimental glomerulonephritis. J. Clin. Path. 4:301, 1951. rheumatoid arthritis. A report of ten cases. Am. J. Med. Sc. 216:51, 1948. 30. Pollak, V. E., Pirani, C. L., and Kark, R. M.: The effect of large doses of 17. Missen, G. A. K., and Taylor, J. D.: Amyloidosis in rheumatoid arthritis. prednisone on the renal lesions and life span of patients with lupus J. Path. & Bact. 71:179, 1956. glomerulonephritis. J. Lab. & Clin. 18. Solomon, W. M.: Amyloidosis in chronic atrophic arthritis. Ann. Int. Med. 18: Med. 57:495, 1961. 846, 1943. 31. Bennett, G. A.: Comparison of the 19. Bayles, T. B.: Rheumatoid arthritis and pathology of rheumatic fever and rheumatic heart disease in autopsied rheumatoid arthritis. Ann. Int. Med. 19:111, 1943. cases. Am. J. Med. Sc. 205:42, 1943. 9 THE KIDNEY IN RHEUMATOID ARTHRITIS Vidor E. Pollak, M.B., M.R.C.P.E., Research Assistant Professor of Medicine, University of Illinois College of Medicine; Attending Physiciiin, Research and Edlicaiional Hospiials; Assistant Attending Physician, Presbyterian-St. Luke‘s Eioupital, Chicago, 111. Established Investigator of the Americari Heart Association, supported by the Illinois Heart Association. Conrad L. Pirnni, M.D., Professor of Putthology, Unioersity of Illinois College of Medicine, Chicago, 111. Irving E . Steck, M.D., Clinical Associate Professor of Medicine, University of Illinois College of Medicine, and Director, Arthritis Clinic, Research and Educaiional Hospitals, Chicago, 111. Robert M . Kark, F.R.C.P., F.A.C.P., Profexor of Medicine, University of Illinois College of Medicine; Attending Physician, Presbytcrian-St. Luke’s Hospita!, Cook Counly Hospital, and Researoh and Educational Hospitals, C h i a g o , Ill.