Morphogenesis of joint b2-microglobulin amyloid deposits

Nephrol Dial Transplant (2001) 16 wSuppl 4x: 3±7

Morphogenesis of joint b2-microglobulin amyloid deposits Charles van Ypersele de Strihou1, M. Jadoul1 and C. Garbar2 1 Department of Nephrology, Cliniques Universitaires St-Luc, Universite Catholique de Louvain, av. Hippocrate 10, 1200 Brussels and 2Department of Pathology, AZ VUB, 1090 Brussels, Belgium

Introduction The detection of amyloidosis in several patients on long-term dialysis came as a surprise w1x. The subsequent identi®cation of beta-2-microglobulin (b2m) as the main component of this amyloid w2,3x demonstrated that we were faced with a new, speci®c complication of long-term renal replacement therapy. The clinical picture of b2m amyloidosis (Ab2m) emerged progressively. The carpal tunnel syndrome (CTS), reported initially by Assenat w1x, was observed within 3 to 5 years after the onset of dialysis, its prevalence rising progressively with the duration of treatment, reaching 100% after 20 years on haemodialysis w4±6x. CTS was, however, not completely speci®c to Ab2m: small b2m deposits were detected in only 70% of the cases w6x. The preferential involvement of the joints subsequently became apparent: it extended to bones w7x and synovia w8x, and included arthralgias and destructive spondylarthropathy, whose prevalence also rose with dialysis duration w5,9x. Here again, the symptoms were not completely speci®c to Ab2m. Of interest, joint ultrasonography and bone X-rays allowed a more accurate diagnosis of b2m amyloid deposition w10±13x. Finally, b2m amyloid was also identi®ed in various organs, usually after )12 years of dialysis w14,15x, with the prevalence rising progressively thereafter. As yet, there is no fully adequate explanation to account for b2m precipitation into amyloid ®brils. A number of risk factors have been identi®ed. As already pointed out, duration of dialysis is a critical determinant of several clinical manifestations of b2m amyloidosis w5,12,16x. Age at onset of dialysis has also emerged as an independent risk factor w12,17x. Finally, the role of membrane characteristics has been strongly suggested w12,18±21x, but its mechanism remains disputed w22x. The actual determinants of b2m precipitation have been the object of a number of interesting studies; Correspondence and offprint requests to: C. van Ypersele de Strihou, Department of Nephrology, Cliniques Universitaires St-Luc, Universite Catholique de Louvain, av. Hippocrate 10, 1200 Brussels, Belgium. #

however, a ®nal conclusion remains elusive. b2m retention is clearly a prerequisite, but the respective contributions of enhanced b2m production and limited removal by dialysis remain speculative w22x. Modi®cation of b2m including partial proteolysis has been suggested but not con®rmed w23,24x. Advanced glycation of b2m (AGE b2m) has been demonstrated by Miyata et al. w25,26x, not only in circulating but also in precipitated b2m. The ability of several AGEmodi®ed proteins to crosslink has led to the hypothesis that advanced glycation may enhance b2m amyloid ®bril formation. A number of studies have further elaborated on this hypothesis, some suggesting that AGE modi®cation of synovial proteins enhances b2m precipitation w27x or that the presence of macrophages modi®es b2m characteristics and increases ®bril formation wreviewed in 28x. Histological evaluation has indeed demonstrated advanced glycation of the b2m amyloid ®brils w29±32x, as well as the presence of macrophages w33x within and around the b2m amyloid deposits.

Morphogenesis of Ab2m deposits It is of note that histological and immunohistochemical observations were made on advanced, tumour-like deposits. Evidence obtained in early deposits remained scarce w34x. A multi-centre prospective collection of joint samples obtained at autopsy was therefore undertaken w35x. Ab2m deposition in haemodialysis patients In the ®rst study w35x, 54 patients who had been on haemodialysis (HD) for an average of 47 months were included. Only 4% of them had clinical evidence of b2m amyloidosis, carpal tunnel syndrome surgery anduor bone cysts. The age at HD onset ranged from 20 to 80 years (median 63 years), and at the time of death from 28 to 82 years (median 69 years). A total of 153 joint samples were obtained (range: 1±3; median 2 per patient) from sternoclavicular joints (ns77), the shoulders (ns35), knees (ns28), hips (ns8), wrists (ns4) and the acromioclavicular joint (ns1). Thirtyfour control patients, of similar age but without

2001 European Renal Association±European Dialysis and Transplant Association

4

history of end-stage renal disease (ESRD), who died during the same time interval were also submitted for post-mortem joint sampling. This study led to three main conclusions. The ®rst relates to the distribution of Ab2m in the different joints. At least two different joints were available in 24 of the 26 HD patients with evidence of Ab2m. The sensitivity of each joint type for the detection of Ab2m was thus calculated. It was highest (97%) for sternoclavicular joints, followed by knees (91%) and shoulders (57%). In later studies, we therefore included only sternoclavicular joints. Interestingly, Zingraff et al. w36x also relied on synovial biopsies of the sternoclavicular joint, readily accessible during parathyroid surgery, to assess Ab2m prevalence. The second conclusion is derived from the control group. Whereas Ab2m was detected in almost half (26u54) of the HD patients, minute non-b2m amyloid deposits were found in a third of the elderly control patients. This observation suggests that serum amyloid P component (SAP) scintigraphy used to detect Ab2m has a lower speci®city than anticipated as it may bind to amyloid deposits of non-b2m type, present in a third of patients without renal failure w17x. The third and most striking observation was the presence of Ab2m in 48% of the HD patients despite a rather short mean follow-up. Ab2m was detected as early as within 23 months of HD. The prevalence increased with duration of HD, from 21% within the ®rst 2 years to 100% after )13 years (Figure 1). Histological evidence of Ab2m is thus present much earlier than anticipated on the basis of arthralgias or CTS w5x. Ab2m deposition in patients on chronic peritoneal dialysis Although symptomatic Ab2m had been reported occasionally in patients treated solely by peritoneal dialysis w37,38x, the actual prevalence of this complication remained unknown. On the basis of lower serum b2m levels, it had been suggested that prevalence should be lower in CAPD than in HD patients w22x. Still, convincing data remained unavailable as the clinical manifestations used to assess Ab2m (CTS

C. van Ypersele de Strihou et al.

and bone cysts) develop late after the onset of renal replacement therapy w39x. We therefore extended our prospective post-mortem study to 26 patients given solely peritoneal dialysis w40x for a median of 27 months (range: 4.5±126 months). Only one of them had been operated on for CTS prior to the onset of CAPD, but Ab2m was not found in the tissue removed during surgery. Age at CAPD onset ranged from 36 to 91 years (median 69 years), and at the time of death from 44 to 93 years (median 73 years). Ab2m was diagnosed in 31% of the patients, here again as early as within 21 months of CAPD. The prevalence ranged from 20% after 1±24 months to 50% after 25±48 months. In order to compare the prevalence in CAPD and HD patients, each of the 26 CAPD patients was paired for age and duration of dialysis with a patient treated solely by HD. Prevalence did not differ signi®cantly between the two groups, at 31 and 50% in CAPD and HD patients, respectively. Here again, the discovery that Ab2m affected 31% of the patients after a median of only 27 months of CAPD revealed a much earlier onset of Ab2m than previously appreciated.

The evolution of Ab2m deposits Both in CAPD and in HD patients, histological evidence antedates markedly the clinical symptoms of Ab2m: amyloid lesions observed after 2 years of dialysis are certainly not equivalent to those discovered after )10 years. It therefore appeared useful to re-analyse several controversies with respect to the onset of amyloidosis in light of the age of the deposit. The evolution of early into late amyloid deposits was therefore evaluated, with the hope of ®nding some clues on amyloidogenesis. We reviewed 54 dialysed and 24 control patients included in our prospective autopsy studies w27x. A large ()2 cm diameter) sternoclavicular joint was available in each patient. Among the dialysed patients, 34 had received solely HD, 16 solely CAPD, and four both methods. Dialysis had lasted from 3 to 244 months (median 46 months). Age at dialysis onset

Fig. 1. Prevalence of Ab2m as a function of haemodialysis duration. Data from Jadoul et al. w35x.

Joint b2-microglobin amyloid deposits

5

Fig. 2. Schematic representation of the three stages of Ab2m deposition. Ab2m is represented in black (a, bone; b, cartilage; c, synovium; d, synovium). Macrophages accumulate around synovial Ab2m in stage III (reprinted with permission from Garbar et al. w28x).

ranged from 42 to 91 years (median 64 years) and at time of death from 43 to 93 years (median 68 years). Ab2m was identi®ed by Congo Red (birefringence under polarized light) and an anti-b2m antibody (ab) staining, whereas Anonb2m was de®ned by Congo Red positivity but with negative anti-b2m ab staining. Among the 54 dialysed patients, 32 had Ab2m, eight had Anonb2m, and 14 had no amyloid deposits at all. Interestingly, among the 24 control patients, 12 had Anonb2m and 12 had no amyloid at all. The Anonb2m deposits were always small. These data con®rm those previously reported in HD patients w35x and stress again that SAP scintigraphy may yield false positive results, both in uraemic and non-uraemic subjects. An important further observation was that Ab2m precipitates ®rst on the surface of the cartilage as a thin layer, which will eventually thicken. No macrophages were ever present around cartilaginous deposits. The thin cartilaginous layer probably represents the earliest stage of Ab2m deposition since it was often observed in the absence of synovial or capsular deposits, whereas capsular deposits were never observed in the absence of cartilage deposits. The fact that Ab2m deposits ®rst in the cartilage has to be taken into consideration in all attempts to understand the initial steps of amyloid deposition. For instance, studies claiming that AGE modi®cation of synovial collagen w26x accounts for the initial formation of b2m amyloid ®brils should be re-interpreted, since Ab2m appears ®rst on the cartilage rather than in the synovium. Of equal importance is the observation that macrophages are absent at this initial stage. Their presence is therefore not required for b2m amyloid formation, in contrast to previous suggestions w33x: modi®cations of the b2m molecule by macrophages are not implicated in the initial stage of Ab2m formation. Ab2m deposits were present in the capsule anduor in the synovia, in 60 and 58% of the subjects with cartilaginous deposits, respectively. Clusters of macrophages were not uniformly associated with these deposits: they were identi®ed only ca. 75 and 25% of the synovial and capsular deposits, respectively.

On the basis of these observations we concluded that, after its initial deposition in the cartilage, Ab2m expanded to synovia and capsules. Macrophages were eventually attracted to these deposits in the ®nal stage. This sequence of events was supported by a quantitative analysis of Ab2m deposits in the cartilage. Their thickness proved to be highly correlated to dialysis duration and was thus utilized for the timing of Ab2m deposits. The early, cartilage-restricted Ab2m deposition was called stage 1. It was associated with small deposits (median 0.03 mm2) and occurred after a mean duration of 39 months. The further development of synovial anduor capsular deposits in the absence of macrophages was called stage 2. Cartilaginous deposits were larger (median 0.19 mm2) and dialysis duration averaged 56 months. In the last stage, called stage 3, macrophages appeared around capsular or synovial deposits. The size of the cartilaginous deposits was even larger (median 0.27 mm2) and dialysis duration averaged 111 months. Figure 2 illustrates schematically the three stages. Of interest, the prevalence of subchondral cysts was similar (ca. 40%) in patients with Ab2m, in patients with Anonb2m, and in control subjects with or without Anonb2m. Thus, the cysts appear to develop independently from Ab2m. By contrast, marginal bone erosions were observed only in stage 3 Ab2m. They were characterized by a cystic dilation of the synovial re¯ection, including variable amounts of ®brous tissue with small vessels and macrophages (Figure 3). We believe that these cysts are the forerunners of the bone cysts identi®ed by X-ray in long-term dialysis patients w12,13x. It is of note that the pattern of Anonb2m deposition in dialysed patients is similar to that found in control patients. Early deposits were identi®ed in the cartilage of all Anonb2m subjects, and in the synovia or capsule of only a small minority of them (in the absence of macrophages). The various stages identi®ed in this study reconcile various patterns of Ab2m deposition. Athanasou et al. w41x previously reported the absence of macrophages

6

C. van Ypersele de Strihou et al.

N N

C N N

M M M

M

Fig. 3. Marginal bone erosion delineated by the arrowheads. Large Ab2m deposits are present in the cartilage (arrows) and small Ab2m deposits within the area of erosion (Congo Red; magni®cation 2 3 20). C, capsule (reprinted with permission from Garbar et al. w28x).

around small Ab2m deposits in the cartilage in a heterogeneous series of joint samples. The size of the deposits, however, was not quantitated and macrophages were not detected by speci®c immunostaining. By contrast, Argiles et al. reported large clusters of macrophages around massive amyloid deposits w33x.

Conclusion Several implications for future studies accrue from the current observations. First, the presence of Ab2m in a sternoclavicular joint sample should not be ruled out unless cartilage is available. Secondly, as early Ab2m deposition does not require the participation of macrophages, further studies should concentrate on the cartilage as a nidus for Ab2m ®bril deposition. Our study has not addressed the last issue on Ab2m precipitation, i.e the role of advanced glycation of b2m and other joint proteins. The possibility that AGE modi®cation leads to protein cross-linking and ®bril formation, and the potential effect of AGE modi®cation on the recruitment of macrophages in the late stage of amyloidogenesis remain to be explored.

References 1. Assenat H, Calemard E, Charra B, Laurent G, Terrat JC, Vanel T. HeÂmodialyse, syndrome du canal carpien et substance amyloõÈde. Nouv Presse MeÂd 1980; 24: 1715 2. Gejyo F, Yamada T, Odani S et al. A new form of amyloid protein associated with chronic hemodialysis was identi®ed as b2-microglobulin. Biochem Biophys Res Commun 1985; 129: 701±706 3. Gorevic PD, Munoz PC, Casey TT et al. Polymerization of intact b2-microglobulin in tissue causes amyloidosis in patients on chronic hemodialysis. Proc Natl Acad Sci USA 1986; 83: 7908±7912 4. Bardin T. Arthropathy and carpal tunnel syndrome of b2microglobulin amyloidosis. In: van Ypersele C, DruÈeke TB, eds. Dialysis Amyloid. Oxford Clinical Nephrology Series, Oxford University Press, Oxford, 1996: 71±97 5. Charra B, Calemard E, Laurent G. Chronic renal failure treatment duration and mode: their relevance to the late dialysis periarticular syndrome. Blood Purif 1988; 6: 117±124 6. van Ypersele de Strihou C, Honhon B, Vandenbroucke JM, Huaux JP, Noel H, Maldague B. Dialysis amyloidosis. In: GruÈnfeld JP, Bach JF, Funck-Brentano JL, eds. Advances in Nephrology. Year Book Medical Publishers, Inc., Chicago, 1988: 117: 401±422 7. Huaux JP, NoeÈl H, Malghem J, Maldague B, Devogelaer JP, Nagant de Deuxchaisnes C. Erosive azotemic osteoarthropathy: possible role of amyloidosis. Arthritis Rheum 1985; 28: 1075±1076

Joint b2-microglobin amyloid deposits 8. MunÄoz-GoÂmez J, GoÂmez-PeÂrez R, SoleÂ-Arques M, LlopartBuisaÂn E. Synovial ¯uid examination for the diagnosis of synovial amyloidosis in patients with chronic renal failure undergoing haemodialysis. Ann Rheum Dis 1987; 46: 324±326 9. Kuntz D, Naveau B, Bardin T, DruÈeke T, Treves R, Dryll A. Destructive spondyloarthropathy in hemodialysed patients: a new syndrome. Arthritis Rheum 1984; 27: 369±375 10. Kay J, Benson CB, Lester S et al. Utility of high-resolution ultrasound for the diagnosis of dialysis-related amyloidosis. Arthritis Rheum 1992; 35: 926±932 11. Jadoul M, Malghem J, Vande Berg B, van Ypersele de Strihou C. Ultrasonographic detection of thickened joint capsules and tendons as marker of dialysis-related amyloidosis: a cross-sectional and longitudinal study. Nephrol Dial Transplant 1993; 8: 1104±1109 12. van Ypersele de Strihou C, Jadoul M, Malghem J, Maldague B, Jamart J, the Working Party on Dialysis Amyloidosis. Effect of dialysis membrane and patient's age on signs of dialysis related amyloidosis. Kidney Int 1991; 39: 1012±1019 13. Maldague B, Malghem J, Vande Berg B. Radiology of dialysis amyloidosis. In: van Ypersele C, DruÈeke TB, eds. Dialysis Amyloid. Oxford Clinical Nephrology Series, Oxford University Press, Oxford, 1996: 98±135 14. Gal R, Korzets A, Schwartz A, Rath-Wolfson L, Gafter U. Systemic distribution of b2-microglobulin-derived amyloidosis in patients who undergo long-term hemodialysis. Arch Pathol Lab Med 1994; 118: 718±721 15. Jadoul M, van Ypersele de Strihou C. b2-microglobulin amyloidosis. In: Brown E, Parfrey P, eds. Complications of Long Term Dialysis. Oxford Clinical Nephrology Series, Oxford University Press, Oxford, 1999: 121±144 16. Ohashi K, Hara M, Kawai R et al. Cervical discs are most susceptible to beta2-microglobulin amyloid deposition in the vertebral column. Kidney Int 1992; 41: 1646±1652 17. Schaeffer J, Floege J, Koch KM. Whole-body scintigraphy. In: van Ypersele C, DruÈeke TB. Dialysis Amyloid. Oxford Clinical Nephrology Series, Oxford University Press, Oxford, 1996: 145±155 18. Vandenbroucke JM, Jadoul M, Maldague B, Huaux JP, NoeÈl H, van Ypersele de Strihou C. Possible role of dialysis membrane characteristics in amyloid osteoarthropathy. Lancet 1986; 1: 1210±1211 19. Chanard J, Bindi P, Lauvaud S, Toupance O, Maheut H, Lacour F. Carpal tunnel syndrome and type of dialysis membrane. Br Med J 1989; 298: 867±868 20. KuÈchle D, Fricke H, Held E, Schif¯ H. High-¯ux hemodialysis postpones clinical manifestation of dialysis-related amyloidosis. Am J Nephrol 1996; 16: 484±488 21. Miura Y, Ishiyama T, Inomata A et al. Radiolucent bone cysts and the type of dialysis membrane used in patients undergoing long-term hemodialysis. Nephron 1992; 60: 268±273 22. van Ypersele de Strihou C, Jadoul M. Prevention and treatment of b2-microglobulin amyloidosis. In: van Ypersele C, DruÈeke TB. Dialysis Amyloid. Oxford Clinical Nephrology Series, Oxford University Press, Oxford: 1996; 261±276 23. Campistol JM. Determinants of ®bril precipitation and aggregation in amyloid deposits. In: van Ypersele C, DruÈeke TB. Dialysis Amyloid. Oxford Clinical Nephrology Series, Oxford University Press, Oxford, 1996: 246±258 24. ArgileÂs A, Derancourt J, Jauregui-Adell J, Mion C, Demaille JG. Biochemical characterization of serum and urinary beta 2

7

25. 26.

27.

28.

29.

30. 31. 32.

33.

34.

35. 36. 37. 38. 39. 40. 41.

microglobulin in end-stage renal disease patients. Nephrol Dial Transplant 1992; 7: 1106±1110 Miyata T, Jadoul M, Kurokawa K, van Ypersele de Strihou C. b-2 microglobulin in renal disease. J Am Soc Nephrol 1998; 9: 1723±1735 Miyata T, Oda O, Inagi R et al. b2-microglobulin modi®ed with advanced glycation end products is a major component of hemodialysis-associated amyloidosis. J Clin Invest 1993; 92: 1243±1252 Hou FF, Chertow G, Kay J et al. Interaction between b2-microglobulin and advanced glycation end products in the development of dialysis-related amyloidosis. Kidney Int 1997; 51: 1514±1519 Garbar C, Jadoul M, NoeÈl H, van Ypersele de Strihou C. Histological characteristics of sternoclavicular b2-microglobulin amyloidosis and clues for its histogenesis. Kidney Int 1999; 55: 1983±1990 Miyata T, Taneda S, Kawai R et al. Identi®cation of pentosidine as a native structure for advanced glycation end products in b2-microglobulin-containing amyloid ®brils in patients with dialysis-related amyloidosis. Proc Natl Acad Sci USA 1996; 93: 2353±2358 Niwa T, Miyazaki S, Katsuzaki T et al. Immunohistochemical detection of advanced glycation end products in dialysis-related amyloidosis. Kidney Int 1995; 48: 771±779 Niwa T, Sato M, Katsuzaki T et al. Amyloid b2-microglobulin is modi®ed with Ne-(carboxymethyl)lysine in dialysis-related amyloidosis. Kidney Int 1996; 50: 1303±1309 Niwa T, Katzsuzaki T, Miyazaki S et al. Amyloid b2microglobulin is modi®ed with imidazolone, a novel advanced glycation end product, in dialysis-related amyloidosis. Kidney Int 1997; 51: 187±194 ArgileÂs A, Mourad G, Kerr PG, GarcõÂa M, Collins B, Demaille JG. Cells surrounding haemodialysis-associated amyloid deposits are mainly macrophages. Nephrol Dial Transplant 1994; 9: 662±667 Athanasou NA. Description, localization and progression of b2-microglobulin dialysis-related amyloid deposition. In: van Ypersele C, DruÈeke TB, eds. Dialysis Amyloid. Oxford Clinical Nephrology Series, Oxford University Press, Oxford, 1996: 173±191 Jadoul M, Garbar C, NoeÈl H et al. Histological prevalence of b2-microglobulin amyloidosis in hemodialysis: a prospective post-mortem study. Kidney Int 1997; 51: 1928±1932 Zingraff J, NoeÈl LH, Bardin T, Kuntz D, Dubost C, DruÈeke T. A sternoclavicular joint biopsy study in hemodialysis patients. Clin Nephrol 1990; 33: 94±97 Jadoul M, NoeÈl H, van Ypersele de Strihou C. b2-microglobulin amyloidosis in a patient treated exclusively by continuous ambulatory peritoneal dialysis. Am J Kidney Dis 1990; 15: 86±88 Gagnon RF, Lough JO, Bourgouin PA. Carpal tunnel syndrome and amyloidosis associated with continuous ambulatory peritoneal dialysis. Can Med Assoc J 1988; 139: 753±755 Koch KM. Dialysis-related amyloidosis. Nephrology Forum. Kidney Int 1992; 41: 1416±1429 Jadoul M, Garbar C, Vanholder R et al. Prevalence of histological b2-microglobulin amyloidosis in CAPD patients compared with hemodialysis patients. Kidney Int 1998; 54: 956±959 Athanasou NA, Ayers D, Rainey AJ, Oliver DO, Duthie RB. Joint and systemic distribution of dialysis amyloid. Q J Med 1991; 287: 205±214