ß 1998 Stockton Press.

Lupus (1998) 7, 654±659 All rights reserved 0961-2033/98 $12.00 http://www.stockton-press.co.uk/lup

REVIEW End-stage renal disease in lupus: Disease activity, dialysis, and the outcome of transplantation John H Stone, M.D., M.P.H.

Division of Rheumatology, Johns Hopkins University, Baltimore, Maryland, USA

Lupus nephritis remains a major cause of morbidity in SLE. Approximately 10% of patients with SLE develop end-stage renal disease (ESRD). In most SLE patients, disease activity diminishes as ESRD approaches. Consequently, the survival of SLE patients on dialysis (both hemo- and peritoneal) appears to be comparable to that of non-SLE patients. However, the role of antiphospholipid (aPL) antibodies in causing dialysis-related morbidity among patients with SLE requires further investigation. In contrast to the outcomes of dialysis, recent evidence suggests that renal transplantation outcomes among SLE patients are inferior to those of non-SLE patients, primarily because of the risk of recurrent lupus nephritis in the allograft and the effect of aPLrelated events on transplantation outcomes. Future avenues of investigation should be directed at developing better strategies to manage and prevent these complications. Keywords: dialysis; end-stage renal disease; lupus; transplantation

Introduction Despite current therapy for lupus nephritis (LN), a signi®cant proportion of SLE patients with renal involvement develop end-stage renal disease (ESRD). Among patients in the Johns Hopkins Lupus Cohort, approximately 15% had developed ESRD after 10 years of disease.1 Although the frequency of ESRD occurrence in SLE varies according to the population studied, 10% is a reasonable overall estimate.2 For some SLE patients with ESRD, dialysis (either peritoneal or hemodialysis) is the most prudent option for renal replacement therapy. For others, it is the only option, at least temporarily. However, because SLE patients with ESRD are strikingly younger than the general ESRD population (mean age 35 years, versus 473), many are excellent candidates for renal transplantation. If successful, this procedure permits a more normal daily routine and greatly enhances the quality of patients' lives.4 Recently, descriptions of renal transplantation outcomes in SLE and advances in the understanding of SLE patients' risk factors for

Correspondence: John H Stone, M.D., M.P.H., 1830 E. Monument Street, Suite 7500, Baltimore, MD 21205, USA.

vascular complications have indicated ways that transplantation outcomes might be improved for SLE patients. The effect of ESRD on SLE activity, the role of dialysis, and the outcome of renal transplantation in SLE are discussed in this review.

SLE activity in ESRD An old clinical pearl maintains that SLE becomes `burnt out' with the occurrence of ESRD.5 Indeed, several retrospective case series6 ± 9 support the concept that with the onset of azotemia, levels of antibodies to double-stranded DNA decline, serum complement levels rise, non-renal manifestations of SLE abate, and the disease becomes quiescent. Regarding the possible cause-and-effect relationship between azotemia and SLE remission, two points are worth noting. First, no prospective studies have compared the longitudinal disease activity of patients who progressed to ESRD to that of patients who did not develop ESRD. For a proportion of SLE patients, remission may be part of the disease's natural history, even without the occurrence of ESRD. In a study of 110 Dutch SLE patients, Swaak et al 10 noted that after the initial disease ¯are leading to the diagnosis of

ESRD in lupus JH Stone

SLE only half of the patients developed subsequent exacerbations. Moreover, if a patient had had no evidence of a ¯are during 5 years of follow-up, the risk of future exacerbations was very low. More recently, Drenkard et al 11 reported treatment-free remissions of at least 1 year in one-fourth of their SLE cohort (n ˆ 667), and found that more than half of those with a disease duration greater than 18 years had achieved remissions. Forty-one of 97 patients with renal involvement (42.3%) achieved disease remissions. Second, although some patients continue to have active disease up to the onset of ESRD and beyond, many SLE patients have only minimally active disease by the time they reach ESRD. That is, in many patients, quiescence precedes azotemia. In the study of 59 SLE patients by Cheigh et al,8 36% of the patients had clinically active SLE at the time dialysis was initiated. In 45%, disease remission had occurred before entry into ESRD and was maintained throughout the period of observation (mean 6.5 years). Thus, progression to ESRD may result not only from active LN but also from a variety of chronic renal injuries that are related to SLE only indirectly: hypertension, hyperlipidemia, steroid-induced diabetes mellitus, exposure to nephrotoxic drugs, hyper®ltration of remaining nephrons, and others. In summary, the precise relationship between ESRD and SLE activity remains incompletely understood, but most patients with advancing renal disease may anticipate signi®cant declines in SLE activity.

The initiation of dialysis SLE may cause rapidly progressive glomerulonephritis (de®ned as the loss of renal function occurring over an interval of less than 3 months12). To prevent ESRD in this setting, reversible renal dysfunction in SLE must be treated aggressively. However, because renal involvement in SLE may progress to ESRD over years rather than months,9 patients are at substantial risk for morbidity (and mortality) from the cumulative effects of treatment. At some point, the risk of treatmentrelated morbidities may outweigh the likelihood of good treatment outcomes. Thus, physicians treating LN must know when to quit. Avascular necrosis of the hips and other joints, life-threatening opportunistic infections, heightened risks of malignancy, obesity, and increased susceptibility to coronary artery disease are all possible sequelae of attempts to save kidneys, successful or not. If progressive azotemia fails to respond to transient increases in immunosuppression, a renal biopsy demonstrating sclerotic glomeruli and a

high chronicity index13 may signal the imminence of ESRD and justify the tapering of immunosuppression. The relatively long waiting times for cadaveric renal transplantation dictate that for SLE patients without living-related or -unrelated kidney donors, a signi®cant period of time will be spent on dialysis. The current waiting period for a cadaveric allograft in the USA is more than 2 years.14 For AfricanAmericans, a group disproportionately affected by SLE, the waiting period is typically even longer.15 Moreover, for a substantial number of patients with SLE, such as those with numerous co-morbidities or strong histories of the anti-phospholipid antibody (aPL) syndrome, hemodialysis or peritoneal dialysis may be the most appropriate choice of renal replacement therapy (see below). Few studies have evaluated the survival of SLE patients on dialysis, and most have failed to control for important demographic variables such as age and gender (because of the epidemiology of the disease, patients with ESRD secondary to SLE are predominantly young females). In general, however, the survival of SLE patients on dialysis is excellent,3,8,9 with 5-year survival rates approaching 90%. No substantial differences in survival between SLE patients on hemodialysis versus those on peritoneal dialysis have been reported. Early studies of SLE patient survival on hemodialysis noted enormous mortality within the ®rst 3 months of initiating dialysis treatment,16 frequently the result of infections rather thatn active SLE. More judicious use of immunosuppression may have reduced this early mortality. Beyond the ®rst 3 months of dialysis, infections and cardiovascular disease constitute the largest threats to SLE patients. Potential morbidity from aPLs in hemodialysis patients is controversial. Access problems (e.g. clotted ®stulas) are common in all hemodialysis patients, and it is not clear that SLE patients suffer disproportionately from these complications. No direct comparisons of the frequency of access problems between patients with SLE and those with other causes of ESRD exist. However, one study,17 which compared the prevalence of the relevant aPL antibodies in 51 SLE patients to their prevalence in 84 non-SLE patients (who had ESRD) and in 50 healthy controls, suggests that SLE patients might differ from other groups. The percentage of SLE patients with elevated anti-cardiolipin (aCL) antibody titers was signi®cantly higher than that of the other two groups (69% vs 14% and 4%; P < 0.005 for both comparisons). Furthermore, recurrent thrombotic events in the SLE patients were associated with very high aCL antibody titers or lupus anticoagulant activity. Unfortunately, the number of SLE patients with ESRD was not reported in this study

655

ESRD in lupus JH Stone

656

(only 9 had biopsy-proven lupus nephritis). Thus, although greater frequencies of aPLs and aPL-related events might be anticipated among SLE patients on dialysis, further study of this question is required.

When to transplant There are no ®rm rules regarding the optimal timing of renal transplantation in SLE. Roth et al,18 proposed a minimum dialysis period of 1 year, to permit patients' SLE to `burn out'. This recommendation was based only on the post-transplant serum creatinine levels of a small number of SLE patients (n ˆ 15) and has not been supported by subsequent studies. Indeed, some patients with living-related donors proceed successfully straight to transplantation, with no intervening dialysis period. In many cases, however, a 3-month dialysis period to ensure that spontaneous renal recovery will not occur is prudent. Kimberly et al 19 reported that 17 of 41 SLE patients reaching ESRD were able to discontinue dialysis (at least temporarily), and noted that rapid loss of renal function predicted those patients who were most likely to recover. At most transplant centers, demonstration of serological quiescence is an important (if not absolute) criterion for transplantion. However, just as serological parameters (double-stranded DNA antibody titers and complement levels) frequently fail to predict SLE ¯ares,20 serologies are imperfect guides to the appropriate timing of renal transplantation. In some SLE patients e.g. those with heritable de®ciencies of complement proteins, serum complement levels may never normalize. Among caucasian SLE patients, 37% have at least one C4A null allele.21 Thus, decisions about when to transplant patients with SLE must be made on an individual basis, considering all clinical and serological data. Ideally, the rheumatologist, transplant nephrologist, and transplant surgeon all participate in such discussions.

Renal transplantation in SLE In the early days of renal transplantation, the procedure was rarely offered to SLE patients. Transplant surgeons feared that recurrent LN would rapidly destroy the allograft. The initial published experience with renal transplantation in SLE allayed this fear: in 1975, a Renal Transplant Registry (RTR) study22 concluded that renal transplantation in SLE was associated with `parallel functional results' compared to use of the procedure in other ESRD

patients. During a mean follow-up period of only 2 years, the 56 SLE transplant patients in the RTR study had a modest allograft survival rate: 55%. This ®gure was believed to be comparable to that of non-SLE patients transplanted at the same centers during the same era, and led to the consensus that transplantation outcomes among patients with SLE are equivalent to those of patients with ESRD from other causes. It should be noted, however, that one-third of the SLE patients in the RTR study died during the relatively brief follow-up period. Since 1975, there has been surprisingly little critical analysis of renal transplantation outcomes in SLE. One recent study evaluated the renal transplantation outcomes of 97 SLE patients who underwent a total of 106 transplantation procedures at the University of California, San Francisco (UCSF) between 1984 and 1996.23 The SLE patients were matched to a group of non-SLE controls on six variables: age, gender, race, type of allograft (cadaveric versus living-related), number of previous renal transplants, and year of transplantation. All SLE patients and controls received post-transplantation regimens that included cyclosporine (or FK-506). Renal transplantation outcomes in the two groups were compared by a strati®ed Cox proportional hazards model. During the follow-up period (mean 6.3 years), 49.1% of the SLE patients and 34.9% of the controls lost their allografts. The 1-, 2-, 5-, and 10-year allograft survival probabilities for the two groups (SLE vs controls) were as follows: 81.7% vs 88.2% (1 year); 74.7% vs 84.4% (2-year); 45.9% vs 75.0% (5-year); and 18.5% vs 34.8% (10-year). In the multivariate Cox proportional hazards model, the risk of allograft loss associated with SLE as the cause of ESRD was 2.1 (95% CI ˆ 1.06 ± 4.06; P ˆ 0.0328). Thus, compared with matched controls, the SLE patients had more than a 2-fold risk of allograft loss. This increased risk occurred despite the fact that the control group included patients with diabetes as the cause of ESRD, which should have favored superior transplantation outcomes for the SLE group. Careful review of other renal transplantation outcome studies in SLE since 1975 supports this ®nding. The literature on renal transplantation in SLE was recently reviewed in detail.24 Among 10 studies that included comparison groups, the allograft survival rate for the comparison group was superior to that of the SLE group in six.25 ± 30 The other four studies found approximately equivalent allograft survival rates between the SLE and comparison groups.31 ± 34 Despite the SLE patients' younger average age at transplantation in all of the studies (except for the two in which agematched controls were used,29,32 none of the studies reported superior outcomes among the SLE patients.

ESRD in lupus JH Stone

A recent study from the University of Wisconsin,26 the second-largest single-center study to date, reported ®ndings similar to those of the UCSF study. The investigators included 69 lupus patients (80 renal transplant procedures) transplanted betwen 1971 and 1994. The comparison group was the entire cohort of 1966 patients transplanted during the same time period (excluding diabetes). When comparing the subset of 44 SLE patients who received cyclosporin to their non-SLE counterparts, the 5-year allograft survival for cadaveric renal transplants was sign®cantly worse among the SLE group (41% vs 71%, P ˆ 0.02). In contrast, results with living-related renal transplants were similar between the two groups. Postcyclosporin-era SLE patients who received livingrelated allografts demonstrated signi®cantly better outcomes at 5 years compared with SLE patients with cadaveric allografts (89% graft survival vs 41%, P ˆ 0.003). Thus, despite the common perception that SLE and non-SLE patients with ESRD fare equally well with renal transplantation, substantial evidence indicates that outcomes among SLE patients are inferior. Some possible explanations for this discrepancy are discussed below.

Reasons for inferior transplantation outcomes among SLE patients Results of the UCSF study23 suggest at least two reasons for inferior transplantation outcomes among SLE patients. First, recurrent LN may be more common than generally appreciated. Second, the role of aPL-related clinical events in causing adverse transplantation outcomes is probably under-recognized. Frequency of recurrent lupus nephritis

Among 14 studies of renal transplantation outcomes in SLE (involving a total of 823 patients) that commented on the presence or absence of recurrent LN, only 8 cases of recurrence were reported ( < 1%).24 This ®gure is undoubtedly low for several reasons: (1) under-reporting because the frequency of recurrence was not a principal question in most studies; (2) insuf®cient follow-up (recurrent LN has been reported more than 8 years after transplantation34,35); and (3) failure to diagnose recurrence, because distinguishing recurrent LN from other causes of impaired allograft function in the absence of biopsy (frequently not performed) is dif®cult.

In the UCSF study,35 all available part-transplant biopsies (135 of 143; 94%) were reviewed, as were the reports of biopsies not available for review (8 of 143; 6%). The post-transplant renal biopsy results were classi®ed as either recurrent LN or some other pathological diagnosis (e.g. acute rejection, chronic allograft nephropathy=rejection, cyclosporin toxicity, or thrombotic microangiopathy). Nine patients (8.4% of the 106 transplantation procedures) had pathological evidence of recurrent LN. Six of the patients with recurrence had cadaveric allografts, 2 had livingrelated allografts, and 1 had a living unrelated allograft. Recurrent LN occurred an average of 3.1 years after transplantation, with the longest interval being 9.3 years and the shortest 5 days. The histopathological diagnoses included 2 cases of diffuse proliferative glomerulonephritis (WHO class IV), 1 focal proliferative glomerulonephritis (class III), 2 membranous glomerulonephritis (one class Va and one class Vb), and 4 mesangial glomerulonephritis (class II). Recurrent LN contributed to allograft loss in 4 of the 9 cases (3.8% of all procedures). The results of this study suggest that LN recurs in renal allografts signi®cantly more frequently than reported in the literature. Nevertheless, recurrent LN still occurs in a minority of SLE transplant patients, and is not invariably associated with allograft loss. Although it appears to contribute to inferior transplantation outcomes in SLE, the risk of recurrence is not large enough to contraindicate renal transplantation in patients with this disorder.

The aPL syndrome and allograft loss

Understanding of the aPL syndrome, ®rst recognized in the mid-1980s, continues to evolve. Consequently, few studies of the impact of aPLs on transplantation exist. Reports of aPL-related morbidity among SLE transplant patients are limited by relatively small numbers of subjects available for study at any given center. In one study of SLE patients,36 8 patients with aCL antibodies were compared retrospectively to 6 patients without the antibodies (no tests for lupus anticoagulants were reported). Post-transplant thrombotic episodes occurred only in the aCL-positive group (3 vs 0), and included one pulmonary embolism and one case of biopsy-proven thrombotic microangiopathy in the transplanted kidney. The two groups did not differ signi®cantly in number of rejection episodes, rate of allograft loss, or renal function at follow-up. Obviously, the small number of patients in this study precludes ®rm conclusions about the impact of aPLs on renal transplantation in SLE.

657

ESRD in lupus JH Stone

658

In the UCSF study, aPLs were associated with 8 cases of allograft loss (including 3 deaths).23 Stated another way, 15.4% of all allograft failures suffered by the SLE patients were attributed to aPL-associated events. These events included thrombotic microangiopathy (2 cases), renal artery thrombosis (2 cases), renal vein thrombosis (1 case), and 3 deaths that occurred in patients whose allografts were functioning well when they died. In contrast, only 2 control patients suffered clinical events conceivably attributable to the aPL syndrome (P ˆ 0.012). The occurrence of aPL-associated complications early in the post-transplant period partly accounted for comparatively high number of early allograft losses in the SLE group. Because not all of the SLE patients and controls were tested for the presence of aPLs (the syndrome had barely been described when the ®rst patients in the study were transplanted), the results of the UCSF study are not de®nitive, but suggest important avenues for future study. The optimal strategy for performing renal transplantations in SLE patients with histories of aPL antibody complications remains unclear. Such patients are certainly at greater risk for thrombotic complications in the post-operative period. Case reports have described successful transplantation outcomes in these patients using intravenous heparin immediately following conclusion of the transplantation procedure, followed by the institution of coumadin therapy several days into the postoperative phase.37,38 Upon reviewing the causes of allograft loss in the UCSF study, if the allograft losses caused by recurrent LN (n ˆ 4) and aPL-related events (n ˆ 8) in the SLE group were removed, the number of allograft losses between the SLE and matched control groups would be nearly identical (40 vs 37, respectively).

Other possible causes of inferior transplantation outcomes

Other anticipated causes of differences between the groupsÐe.g. a greater risk of infection or allograft rejection among the SLE patientsÐwere not operative in this study. Although 98% of the SLE patients received pretransplantation immunosuppression for treatment of their underlying disorder (versus only 26% of the controls), the rates of post-transplantation infections were similar between the two groups. Although the SLE group had both more biopsyproven acute rejection reactions and more allograft losses caused by acute rejection, neither comparison was statistically signi®cant. In conclusion, the SLE patients' inferior transplantation outcomes were not attributable to a single factor, but rather to several

sources of allograft and patient morbidity, particularly recurrent LN and the aPL syndrome.

Conclusion In the past few decades, treatment of ESRD in SLE has improved remarkably. ESRD is now an unusual direct cause of death in this disease. In most SLE patients, disease activity diminishes as ESRD approaches. Although the role of aPLs in causing dialysis-related morbidity requires further de®nition, survival of SLE patients on dialysis (both hemodialysis and peritoneal dialysis) appears comparable to that of non-SLE patients. In contrast, renal transplantation outcomes among SLE patients appear to be inferior to those of non-SLE patients, primarily because of the risk of recurrent LN in the allograft and the effect of aPL-related events on transplantation outcomes. Transplantation outcomes in SLE may be improved by better understanding of ways to manage and prevent these complications.

References 1 Petri M. (personal communication) 1998. 2 Golbus J, McCune W. Lupus nephritis: classi®cation, prognosis, immunopathogenesis, and treatment. Rheum Dis Clin North Am 1994; 20: 213 ± 242. 3 Hellerstedt W et al. Survival rates of 2728 patients with end-stage renal disease. Mayo Clin Proc 1984; 59: 776 ± 783. 4 Johnson J, McCauley C, Copley J. The quality of life of hemodialysis and transplant patients. Kidney Int 1982; 22: 286 ± 291. 5 Fries J, Powers R, Kempson R. Late-stage lupus nephropathy. J Rheumatol 1974; 1: 166 ± 175. 6 Kimberly R et al. `End-stage' lupus nephritis: clinical course to and outcome on dialysis. Medicine 1981; 60: 277 ± 287. 7 Coplon NS, Diskin CJ, Petersen J, Swenson RS. The long-term clinical course of systemic lupus erythematosus in end-stage renal disease. N Engl J Med 1983; 308 (4): 186 ± 190. 8 Cheigh JS et al. Systemic lupus erythematosus in patients with endstage renal disease: long-term follow-up on the prognosis of patients and the evolution of lupus activity. Am J Kidney Dis 1990; XVI (3): 189 ± 195. 9 Nossent HC, Swaak TJ, Berden JH. Systemic lupus erythematosus after renal transplantation: patient and graft survival and disease activity. Ann Intern Med 1991; 114: 183 ± 188. 10 Swaak A et al. Systemic lupus erythematosus: II. Observations of the occurence of exacerbations in the disease course. Dutch experience with 110 patients studied prospectively. Ann Rheum Dis 1989; 48: 455 ± 460. 11 Drenkard C et al. Remission of systemic lupus erythematosus. Medicine 1996; 75 (2): 88 ± 98. 12 Kerr PG, Lan HV, Atkins RC. Schreier R, Gottschalk C (eds). Diseases of the Kidney. `Rapidly Progressive Glomerulonephritis.' 6th edn, Little, Brown: Boston, 1997 pp. 1619 ± 1644. 13 Balow J et al. NIH conference. Lupus nephritis. Ann Intern Med 1987; 106: 79 ± 94. 14 Suthanthiran M, Strom T. Renal transplantation. N Engl J Med 1994; 331 (6): 365 ± 376. 15 Kasiske B et al. The effect of race on access and outcome in transplantation. N Engl J Med 1991; 324 (5): 302 ± 307. 16 Coplon N, Siegel R, Fries J. Hemodialysis in end-stage lupus nephritis. Trans Am Soc Artif Intern Organs 1973; 19: 302 ± 304.

ESRD in lupus JH Stone 17 Sitter T, Spannagl M, Schif¯ H. Anticardiolipin antibodies and lupus anticoagulant in patients treated with different methods of renal replacement therapy in comparison to patients with systemic lupus erythematosus. Ann Hematol 1992; 65: 79 ± 82. 18 Roth D et al. Renal transplantation in systemic lupus erythematosus: one center's experience. Am J Nephrol 1987; 7: 367 ± 374. 19 Kimberly R et al. Reversible `end-stage' lupus nephritis: analysis of patients able to discontinue dialysis. Am J Med 1983; 74: 361 ± 368. 20 Esdaile J et al. Laboratory tests as predictors of disease exacerbations in systemic lupus erythematosus: why some tests fail. Arthritis Rheum 1996; 39: 370 ± 378. 21 Howard P, Bias W, Arnett F, McLean R. Relationship between C4 null genes, HLA-D region antigens, and genetic susceptibility to systemic lupus erythematosus in Caucasian and Black Americans. Am J Med 1986; 81: 187 ± 193. 22 Barnes BA et al. Renal transplantation in congenital and metabolic diseases: a report from the ASC=NIH Renal Transplant Registry. JAMA 1975; 232 (2): 148 ± 153. 23 Stone J, Amend W, Criswell L. Outcome of renal transplantation in SLE: 97 cyclosporine-era patients and matched controls. Arthritis Rheum 1998; 41 (8): 1438 ± 1445. 24 Stone JH, Amend WJ, Criswell LA. Outcome of renal transplantation in systemic lupus erythematosus. Semin Arthritis Rheum 1997; 27 (1): 17 ± 26. 25 Cattran DC, Aprile M. Renal transplantation in lupus erythematosus. Ann Intern Med 1991; 114 (11): 991. 26 Lochhead KM et al. Risk factors for renal allograft loss in patients with systemic lupus erythematosus. Kidney Int 1996; 49: 512 ± 517. 27 Krishnan G, Thacker L, Angstadt J, Capelli J. Multicenter analysis of renal allograft survival in lupus patients. Transplant Proc 1991; 23 (2): 1755 ± 1756.

28 Terashita GY, Cook DJ. Original disease of the recipient, in Clinical transplants. UCLA Tissue Typing Laboratory: Los Angeles. 1987; 372 ± 379. 29 Nyberg G et al. Renal transplantation in patients with systemic lupus erythematosus: increased risk of early graft loss. Scand J Urol Nephrol 1990; 24: 307 ± 313. 30 Hariharan S, Schroeder Y, Carey M, First M. Renal transplantation in patients with systemic lupus erythematosus. Clin Transplant 1992; 6: 345 ± 349. 31 Grimbert P et al. Renal transplantation in patients with systemic lupus erythematosus: a multicenter study. Transplant Proc 1997; 29: 2363 ± 2364. 32 Bumgardner GL et al. Single-center 1 ± 15-year results of renal transplantation in patients with systemic lupus erythematosus. Transplantation 1988; 46 (5): 703 ± 709. 33 Pollack CA, Ibels LS. Dialysis and transplantation in patients with renal failure due to systemic lupus erythematosus. The Australian and New Zealand Experience. Aust NZJ Med 1987; 17: 321 ± 325. 34 Sumrani N et al. Renal Transplantation in cyclosporine-treated patients with end-stage lupus nephropathy. Transplant Proc 1992; 24 (5): 1785 ± 1787. 35 Stone JH et al. Frequency of recurrent lupus nephritis among ninetyseven renal transplant patients during the cyclosporine era. Arthritis Rheum 1998; 41 (4): 678 ± 686. 36 Radhakrishnan J, Williams G, Appel G, Cohen D. Renal transplantation in anticardiolipin antibody-positive lupus erythematosus patients. Am J Kidney Dis 1994; 23 (2): 286 ± 289. 37 Knight R, Schanzer H, Rand J, Burrows L. Renal allograft thrombosis associated with the antiphospholipid antibody syndrome. Transplantation 1995; 60 (6): 614 ± 615. 38 Stone J, Amend W, Criswell L. (unpublished data).

659