Anti-glomerular basement membrane antibody disease is an uncommon cause of end-stage renal disease

Transcription

1 & 2012 International Society of Nephrology clinical investigation Anti-glomerular basement membrane antibody is an uncommon cause of end-stage renal Wen Tang 1,2,3, Stephen P. McDonald 1,4, Carmel M. Hawley 1,3, Sunil V. Badve 1,3, Neil C. Boudville 1,5, Fiona G. Brown 1,6, Philip A. Clayton 1,7, Scott B. Campbell 1,3, Janak R. de Zoysa 1,8 and David W. Johnson 1,3 1 ANZDATA Registry, Adelaide, South Australia, Australia; 2 Division of Nephrology, Peking University Third Hospital, Beijing, China; 3 Department of Nephrology, University of Queensland at Princess Alexandra Hospital, Brisbane, Queensland, Australia; 4 Department of Nephrology and Transplantation Services, University of Adelaide at Central Northern Adelaide Renal and Transplantation Services, Adelaide, South Australia, Australia; 5 School of Medicine and Pharmacology, Sir Charles Gairdner Hospital Unit, The University of Western Australia, Perth, Western Australia, Australia; 6 Department of Nephrology, Monash Medical Centre, Melbourne, Victoria, Australia; 7 Department of Renal Medicine, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia and 8 Department of Renal Medicine, North Shore Hospital, Auckland, New Zealand There are few reports regarding outcomes of anti-glomerular basement membrane (GBM) in patients who underwent renal replacement therapy. To help define this we studied all patients with anti-gbm who started renal replacement therapy for end-stage renal (ESRD) in Australia and New Zealand (ANZDATA Registry) between 1963 and 2010 encompassing 449 individuals (0.8 percent of all ESRD patients). The median survival on dialysis was 5.93 years with death predicted by older age and a history of pulmonary hemorrhage. Thirteen patients recovered renal function, although 10 subsequently experienced renal death after a median period of 1.05 years. Of the 224 patients who received their first renal allograft, the 10-year median patient and renal allograft survival rates were 86% and 63%, respectively. Six patients experienced anti-gbm recurrence in their allograft, which led to graft failure in two. Using multivariable Cox regression analysis, patients with anti-gbm had comparable survival on dialysis or following renal transplantation (hazard ratios of 0.86 and 1.03, respectively) compared to those with ESRD due to other causes. Also, renal allograft survival (hazard ratio of 1.03) was not altered compared to other s requiring a renal transplant. Thus, anti-gbm was an uncommon cause of ESRD, and not associated with altered risks of dialysis, transplant or first renal allograft survival. Death on dialysis was predicted by older age and a history of pulmonary hemorrhage. Kidney International (2013) 83, ; doi: /ki ; published online 19 December 2012 Correspondence: David W. Johnson, Department of Renal Medicine, Level 2, Ambulatory Renal and Transplant Services Building, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Brisbane, Queensland 4102, Australia. david_johnson@health.qld.gov.au Received 1 June 2012; revised 16 August 2012; accepted 31 August 2012; published online 19 December 2012 KEYWORDS: anti-gbm ; anti-glomerular basement membrane antibody ; end-stage renal ; kidney failure; renal function recovery; renal transplantation Anti-glomerular basement membrane (anti-gbm) is generally a fulminant and rapidly progressive that is caused by autoantibodies to the noncollagenous domain of the a3 chain of type IV collagen. It can present as an isolated glomerulonephritis, or as a pulmonary renal syndrome with severe lung hemorrhage (Goodpasture s ). 1 3 The incidence of anti-gbm is estimated to be cases per million per year in both European Caucasoid 4 6 and Asian populations. 7 9 It is responsible for 1 5% of all types of glomerulonephritis and is the cause in 10 20% of patients with crescentic glomerulonephritis. 4 Even with intensive treatment with immunosuppression and/or plasma exchange, many patients will develop end-stage renal (ESRD) necessitating life-long renal replacement therapy (RRT). 10,11 One-year patient and renal survival rates following a diagnosis of anti-gbm have been reported to be 73% and 25%, respectively. 7 Although the course, treatment, and outcomes from the onset of anti-gbm have been well described, 4,12 15 there are few reports regarding the predictors and outcomes of anti-gbm once long-term RRT (dialysis or kidney transplantation) has been commenced. The aim of this study was to investigate the characteristics, treatments, and outcomes of all cases of ESRD due to anti- GBM in the Australian and New Zealand dialysis populations, using data from the Australia and New Zealand Dialysis and Transplant (ANZDATA) Registry. RESULTS Patient characteristics Between 15 May 1963 and 31 December 2010, 58,422 individuals started RRT for ESRD. Of these, 449 individuals Kidney International (2013) 83,

2 clinical investigation W Tang et al.: ESRD Table 1 of all patients with ESRD secondary to anti-gbm or other causes in Australia and New Zealand ( ) ESRD (n ¼ 449) Other ESRD (n ¼ 57,973) P-value Age (years) 45.8± ±17.4 Age category (years) (0%) 572 (1%) (9%) 1596(3%) (26%) 3807 (7%) (7%) 5772(10%) (11%) 9105 (16%) (15%) 12,299 (21%) (15%) 12,810 (22%) (13%) 9698 (17%) (3%) 2314(4%) Gender: male 257 (57%) 33,746 (58%) 0.68 Racial origin European 405 (90%) 45,180 (78%) ATSI 4 (1%) 3484 (6%) MPI 31(7%) 5083(9%) Asian 7 (2%) 2477 (4%) Other 2 (0%) 1749 (3%) RRT era (8%) 2800 (5%) (20%) 6452 (11%) (24%) 12,054 (21%) (12%) 9807 (17%) (16%) 12,331 (21%) (21%) 14,529 (25%) Ever smoked Current 61 (14%) 6105 (11%) Former 122 (27%) 17,633 (30%) Never 120 (27%) 22,408 (39%) Missing 146 (33%) 11,827 (20%) Hypertension Yes 115 (26%) 24,438 (42%) No 80 (18%) 4476 (8%) Missing 254 (57%) 29,059 (50%) Diabetes mellitus Yes 16 (4%) 17,886 (31%) No 336 (75%) 31,616 (55%) Missing 97 (22%) 8446 (15%) Chronic lung Yes 43 (10%) 7112 (12%) No 298 (66%) 41,017 (71%) Missing 108 (24%) 9844 (17%) Coronary artery Yes 41 (9%) 17,594 (30%) No 295 (66%) 30,657 (53%) Missing 113 (25%) 9722 (17%) Peripheral vascular Yes 19 (4%) 11,436 (20%) No 320 (71%) 36,694 (63%) Missing 110 (25%) 9843 (17%) Table 1 Continued ESRD (n ¼ 449) Other ESRD (n ¼ 57,973) P-value Cerebrovascular Yes 16 (4%) 6538 (11%) No 328 (73%) 41,606 (72%) Missing 105 (23%) 9829 (17%) BMI (kg/m 2 ) 26.4± ± Late referral Yes 173 (39%) 9219 (16%) No 99 (22%) 33,760 (58%) Missing 177 (39%) 14,995 (26%) First RRT Hemodialysis 388 (86.4%) 38,910 (67%) Peritoneal dialysis 59 (13.1%) 17,580 (30%) Renal transplant 2 (0.45%) 1483 (3%) Abbreviations: ATSI, Aboriginal and Torres Strait Islander; BMI, body mass index; ESRD, end-stage renal ; GBM, glomerular basement membrane; MPI, Maori and Pacific Islander; RRT, renal replacement therapy. (0.8%) had ESRD secondary to anti-gbm. The baseline characteristics of ESRD patients with and without anti-gbm are displayed in Table 1. Using multivariable logistic regression analysis, anti-gbm was significantly associated with younger age and Caucasian racial origin compared with other forms of ESRD (Table 2). In a supplementary analysis using a more contemporary cohort ( ) in which complete data were available on comorbidities (n ¼ 36,884 including 217 patients with anti- GBM ), anti-gbm was independently associated with smoking, late referral to a renal unit, and lower probabilities of hypertension, diabetes mellitus, cerebral vascular, coronary artery, and Aboriginal and Torres Strait Islander racial origin (Table 2). Patients with pulmonary hemorrhage in the setting of ESRD secondary to anti-gbm (n ¼ 300) were independently predicted by younger age (odds ratio per decade 0.75, 95% confidence interval (CI) , P). No associations were observed with cigarette smoking or the presence of chronic lung. Patient survival on dialysis Death occurred in 147 (33%) anti-gbm ESRD patients and 26,166 (46%) patients with ESRD due to other causes (P). The causes of death were cardiac (29% vs. 41%, respectively), dialysis withdrawal (29% vs. 24%), infections (16% vs. 13%), vascular (13% vs. 10%), malignancy (6% vs. 6%), and other (8% vs. 6%). Median survival of patients with anti-gbm on dialysis (5.93 years, 95% CI years) was superior to that of patients with other causes of ESRD (4.41 years, 95% CI years) (log-rank score 13.5, P) (Figure 1). Respective survival rates in the two groups were 88.4% vs. 88.1% at 1 year, 78.3% vs. 76.5% at 2 years, and 56.7% vs. 44.9% at 5 years. 504 Kidney International (2013) 83,

3 W Tang et al.: ESRD clinical investigation Table 2 Multivariable binary logistic regression analysis of predictors of ESRD due to anti-gbm, as opposed to an alternative cause of ESRD Entire cohort ( ) (n ¼ 58,442) b Odds ratio (95% CI) b Contemporary cohort ( ) (n ¼ 36,884) Odds ratio (95% CI) Age (per decade, ( ) NS NS years) Gender: male NS NS NS NS Racial origin European Reference Reference ATSI ( ) ( ) MPI ( ) NS NS Asian ( ) NS NS Other ( ) NS NS Cumulative survival ESRD cause Others Others-censored -censored RRT era NS NS NS NS Ever smoked NA NA Current ( ) Former ( ) Never Reference Hypertension NA NA ( ) Diabetes mellitus NA NA ( ) Chronic lung NA NA NS NS Coronary artery NA NA ( ) Peripheral vascular NA NA NS NS Cerebrovascular NA NA ( ) BMI (kg/m 2 ) NA NA NS NS Late referral NA NA ( ) Abbreviations: ATSI, Aboriginal and Torres Strait Islander; BMI, body mass index; CI, confidence interval; ESRD, end-stage renal ; GBM, glomerular basement membrane; MPI, Maori and Pacific Islander; NA, not analyzed due to incomplete data prior to 1996; NS, not statistically significant; RRT, renal replacement therapy. Dependent factor, 1¼ ESRD patients due to anti-gbm, 0¼ ESRD patients with other causes. was independently associated with lower mortality on dialysis in the entire cohort after adjustment for age, gender, racial origin, and dialysis era (adjusted hazard ratio (HR) 0.85, 95% CI , P ¼ 0.049), and was associated with a lower mortality hazard in the supplementary analysis of the contemporary ( ) cohort after adjustment for age, gender, racial origin, dialysis era, body mass index, smoking status, history of hypertension, chronic lung, cerebrovascular, ischemic heart, diabetes mellitus, peripheral vascular, and late referral (HR 0.65, 95% CI , P ¼ 0.005). In light of the possibility of informative censoring due to a higher rate of renal transplantation in the anti-gbm group, a competing risks approach was used in the entire cohort, which showed that anti-gbm was not significantly associated with mortality on dialysis (HR 0.86, 95% CI , P ¼ 0.08). When only patients with anti-gbm were considered, death on dialysis was predicted by older age (HR per decade 1.54, 95% CI , P), dialysis era (HR per period 0.73, 95% CI , P), and a history of pulmonary hemorrhage (HR 1.49, 95% CI , P ¼ 0.024). In the contemporary cohort ( ), death was predicted by age (HR 1.79 per decade, 95% CI , P), diabetes mellitus (HR 5.59, 95% CI , P ¼ 0.017), and a history Time (years) At risk Other 56,490 29,129 14, Figure 1 Kaplan Meier survival curves for anti-glomerular basement membrane (anti-gbm) end-stage renal (ESRD) and other causes of ESRD in Australian and New Zealand dialysis patients between 1963 and The difference between the groups was statistically significant (log-rank score 13.5, P). Outcomes were censored for renal function recovery, loss to followup, renal transplantation, and end of study. of pulmonary hemorrhage (HR 2.52, 95% CI , P ¼ 0.022). None of the other parameters included in the multivariable models was associated with death on dialysis in patients with anti-gbm. Recovery of renal function Recovery of dialysis-independent renal function occurred in 13 (3%) anti-gbm patients and 879 (1.5%) ESRD patients with other forms of renal (P ¼ 0.02). Anti- GBM was associated with a significantly shorter time to renal recovery (log-rank score 6.49, P ¼ 0.011) and independently predicted renal function recovery in the entire cohort (HR 1.93, 95% CI , P ¼ 0.02). However, of the 13 anti-gbm patients who recovered renal function, 10 (77%) experienced renal death (1 died and 9 returned to dialysis) within a median period of 1.05 years (minimum 0.1 years, maximum 7.92 years). Renal transplant graft survival A total of 224 (50%) patients with anti-gbm received 286 renal allografts during the study period (Table 3). Compared with other renal transplant recipients, patients with anti-gbm were significantly more likely to be younger, Caucasian, current smoker, referred late to a renal unit, and to have re-transplants. They were less likely to have diabetes mellitus, hypertension, and coronary artery. The median time from dialysis commencement to first renal transplant was longer for anti-gbm patients (1.63 years, interquartile range ) than non-anti-gbm Kidney International (2013) 83,

4 clinical investigation W Tang et al.: ESRD Table 3 of all patients with ESRD secondary to anti-gbm or other causes in Australia and New Zealand who underwent renal transplantation during the period ESRD (n ¼ 224) Other ESRD (n ¼ 19,455) P-value Age (years) 35.6± ±14.9 Gender: male 134 (60%) 11,622 (60%) 0.98 Racial origin European 203 (91%) 16,653 (86%) ATSI 1 (0.4%) 489 (3%) MPI 14 (6%) 802 (4%) Asian 5 (2.2%) 921 (5%) Other 1 (0.4%) 590 (3%) Transplant era (8%) 1925 (10%) (25%) 3416 (18%) (25%) 4562 (23%) (13%) 2663 (14%) (13%) 3146 (16%) (17%) 2743 (19%) Ever smoked Current 38 (17%) 1532 (8%) Former 45 (20%) 3637 (19%) Never 50 (22%) 7809 (40%) Missing 91 (41%) 6477 (33%) Hypertension Yes 66 (29%) 8550 (44%) No 46 (21%) 1875 (10%) Missing 112 (50%) 9030 (46%) Diabetes mellitus Yes 1 (0%) 2098 (11%) No 178 (80%) 13,367 (69%) Missing 45 (20%) 3988 (21%) Chronic lung Yes 5 (2%) 577 (3%) No 164 (73%) 14,278 (73%) Missing 55 (25%) 4600 (24%) Coronary artery Yes 1 (0%) 1203 (6.2%) No 163 (73%) 13,428 (69%) Missing 60 (27%) 4824 (25%) Peripheral vascular Yes 2 (1%) 702 (4%) No 165 (74%) 14,082 (72%) Missing 57 (25%) 4671 (24%) Cerebrovascular Yes 1 (0%) 393 (7%) No 169 (75%) 14,464 (58%) Missing 54 (24%) 4598 (23%) BMI (kg/m 2 ) 25.7± ± Late referral Yes 64 (29%) 1713 (9%) No 58 (26%) 10,292 (53%) Table 3 Continued ESRD Other ESRD (n ¼ 224) (n ¼ 19,455) P-value Missing 102 (46%) 7450 (38%) Donor type 0.72 Deceased 164 (73%) 14,448 (74%) Living 60 (27%) 5007 (26%) Subsequent grafts nd 47 (16%) 2703 (12%) 3rd 13 (5%) 430 (2%) 4th 2 (1%) 60 (0%) 5th 0 (0%) 5 (0%) Abbreviations: ATSI, Aboriginal and Torres Strait Islander; BMI, body mass index; ESRD, end-stage renal ; GBM, glomerular basement membrane; MPI, Maori and Pacific Islander. patients (1.43 years, interquartile range , P ¼ 0.02). Death-censored renal allograft survival rates were generally comparable between patients with anti-gbm ESRD and those with ESRD due to other causes, regardless of donor type (living or deceased) and transplant era. Median death-censored, first graft survival rates were 21.2 years (95% CI years) in the anti-gbm group and 18.9 years (95% CI years) in the other group (log-rank score 1.23, P ¼ 0.27) (Figure 2). The respective 10-year graft survival rates were 63% vs. 67%. Using multivariable Cox proportional hazards model analysis of first grafts, anti-gbm was not associated with renal allograft survival in the entire cohort (HR 1.03, 95% CI ). For subsequent grafts, median graft survival in patients with anti-gbm (8.1 years, 95% CI years) was nonsignificantly lower than that in patients with other causes of ESRD (11.5 years, 95% CI ; log-rank score 3.01, P ¼ 0.083). The respective 10-year graft survival rates were 45 and 53%. Using multivariable Cox proportional hazards model analysis, it was found that anti-gbm was not associated with subsequent renal allograft survival (HR 1.29, 95% CI , P ¼ 0.13). When only anti-gbm patients with first renal allografts were considered, mean death-censored first renal allograft survival in anti-gbm patients with a history of pulmonary hemorrhage (16.3 years, 95% CI years) tended to be inferior to that in non-anti-gbm ESRD patients (20.9 years, 95% CI years) (log-rank score 3.83; P ¼ 0.050). The respective graft survival rates in the two groups were 72% vs. 80% at 5 years and 59% vs. 70% at 10 years. Using multivariable Cox proportional hazards model analysis of first grafts in anti-gbm patients, older age at transplant (HR 0.98, 95% CI , P ¼ 0.034) was an independent predictor of renal allograft survival after adjusting for gender, a history of pulmonary hemorrhage, race, donor type, and time from first dialysis to transplant. For subsequent transplants, no independent predictors of renal allograft failure were identified. When anti-gbm patients were analyzed according to whether they 506 Kidney International (2013) 83,

5 W Tang et al.: ESRD clinical investigation ESRD cause Others Others-censored -censored Time to transplant <1 Year 1 Year <1 Year-censored 1 Year-censored Cumulative survival Cumulative survival At risk Other , ,333 Time (years) Figure 2 Kaplan Meier death-censored first graft survival curves for anti-glomerular basement membrane (anti-gbm) end-stage renal (ESRD) and other causes of ESRD undergoing renal transplantation in Australia and New Zealand between 1963 and The difference between the groups was not statistically significant (log-rank score 1.23, P ¼ 0.27). 0.0 At risk <1 Year 1 Year Time (years) Figure 3 Kaplan Meier death-censored first graft survival curves for patients with anti-glomerular basement membrane (anti-gbm) end-stage renal (ESRD) who received a transplant o1 year or X1 year after starting dialysis. The difference between the groups was not statistically significant (log-rank score 0.73, P ¼ 0.39). received a transplant o1 year or X1 year after starting dialysis, the median renal allograft survival rates were 19.0 years (95% CI ) and 21.2 years (95% CI ), respectively (log-rank score 0.73, P ¼ 0.39; Figure 3). recurrence in renal transplants Six (2.7%) patients with anti-gbm experienced renal biopsy confirmed recurrence in their allografts, which led to graft failure in 2 (0.9%) cases after 0.08 and 0.48 years. In contrast, 1 (0.013%) patient with non-anti-gbm ESRD (Alport s ) experienced de novo anti-gbm causing renal allograft failure. The remaining causes of renal allograft failure in the anti-gbm and nonanti-gbm ESRD groups were chronic allograft nephropathy (54% vs. 48%, respectively), acute rejection (15% vs.19%), hyperacute rejection (2% vs. 3%), renal artery thrombosis (2% vs. 3%), renal vein thrombosis (5% vs. 3%), glomerulonephritis other than anti-gbm (3% vs. 4%), non-compliance (5% vs. 2%), and others (14% vs. 18%). Renal transplant patient survival When first renal allografts were considered, the median survival of anti-gbm patients was comparable to that of patients with other causes of ESRD (log-rank score 3.09, P ¼ 0.08) (Figure 4). The 10-year patient survival rates were 86 and 78%, respectively. Ten-year survival rates were also comparable between the two groups for deceased donor transplants (83% vs. 75%, P ¼ 0.21) and living donor transplants (94% vs. 90%, P ¼ 0.21). The causes of death in Cumulative survival At risk Other Time (years) ESRD cause Others Others-censored censored ,455 10, Figure 4 Kaplan Meier patient survival curves for antiglomerular basement membrane (anti-gbm) end-stage renal (ESRD) and other causes of ESRD undergoing first renal transplantation in Australia and New Zealand between 1963 and The difference between the groups was not statistically significant (log-rank score 3.09, P ¼ 0.08). the anti-gbm (n ¼ 41) and non-anti-gbm groups (n ¼ 4137) were cardiac (24% vs. 30%), vascular (20% vs. 13%), malignancy (24% vs. 25%), infection (20% Kidney International (2013) 83,

6 clinical investigation W Tang et al.: ESRD vs. 19%), therapy withdrawal (2% vs. 4% ), and others (10% vs. 9%). Using multivariable Cox proportional hazards model analysis in the entire cohort, anti-gbm was not associated with patient mortality following renal transplantation (HR 1.03, 95% CI , P ¼ 0.87). When only anti-gbm patients were analyzed, the mean survival of anti-gbm patients with a history of pulmonary hemorrhage (n ¼ 160) (18.4 years, 95% CI years) was lower than those without pulmonary hemorrhage (n ¼ 64) (26.2 years, 95% CI years), although it did not reach statistical significance (log-rank score 3.54, P ¼ 0.06). The respective transplant patients survival rates in the two groups were 93% vs. 92% at 5 years and 86% vs. 86% at 10 years. Using multivariable Cox proportional hazards model analysis of first transplant survival, age at transplant (HR 1.07, 95% CI , P) was a significant predictor of mortality after adjusting other factors, including a prior history of pulmonary hemorrhage. DISCUSSION This retrospective, multicenter, multicountry registry analysis examined the outcomes of 449 ESRD patients with anti-gbm compared with 57,973 patients with ESRD due to other causes. Our study found that anti-gbm was not associated with altered dialysis survival, renal transplant patient survival, or renal allograft survival, although it was associated with an increased probability of renal recovery (albeit short-lived). In addition, in patients with anti-gbm, a history of pulmonary hemorrhage was associated with an increased risk of mortality on dialysis. Until now, the rarity of anti-gbm as a cause of ESRD has impeded the study of outcomes of this condition once RRT has commenced. Limited outcome data from observational cohort studies of anti-gbm populations that included patients without ESRD reported 1-year patient survival rates in the range of 65 93%. 3,9,12,16 A longterm outcome study over a 25-year period ( ) of 39 patients with anti-gbm and dialysis-dependent endstage renal failure reported 1- and 5-year dialysis survival rates of 65 and 44%, respectively. 12 The results of this study, which covered a somewhat longer time period ( ), demonstrated slightly higher 1- and 5-year dialysis survival rates of 88 and 57%, respectively. A novel finding of the current investigation was that the survival rates of patients with anti-gbm on dialysis or following renal transplantation were comparable to those of patients with ESRD due to other causes. In contrast to previous studies, which found that age was not associated with anti-gbm patient survival, 2,12 the present study observed that age was a significant, independent predictor of mortality on dialysis, such that each additional decade was associated with a 54% increase in mortality in the entire cohort and a 79% increase in mortality in the contemporary cohort. The apparent disparity in findings may relate to the fact that previous smaller studies lacked statistical power to identify an association of age with survival. In addition to age, our study also observed that the presence of pulmonary hemorrhage in anti-gbm patients predicted a greatly increased risk of death on dialysis in both the entire cohort (49% increase) and the contemporary cohort (152% increase) despite these patients being younger. Although there have been no prior reports examining the outcomes of Goodpasture s in ESRD populations per se, Levy et al. 12 found that the median time to death was 2 months in patients with pulmonary hemorrhage compared with 18 months in those without. These findings may relate to pulmonary hemorrhage connoting the presence of more serious or necessitating more intense immunosuppression. The proportion of patients experiencing pulmonary hemorrhage in our study (67%) was higher than that reported from centers in New Zealand (40%), 6 Hong Kong (40%), 9 France (48%), 17 and China (26 to 46%), 2 but similar to that reported in the United Kingdom (62%) 12 and the United States (61%). 18 A possible reason for the more frequent occurrence of pulmonary hemorrhage in the present population is that, unlike the aforementioned studies, our study only included ESRD patients, thereby representing the more severe end of the spectrum. Moreover, the ethnic composition of patients in the present study differed from those in Asian centers. In keeping with the findings of other series, 2,12 pulmonary hemorrhage was associated with younger age. Similar to the report by Taylor et al, 6 smoking was independently associated with anti-gbm, raising the possibility that smoking may induce an autoimmune response either directly or indirectly by uncovering epitopes. In our series, recovery of dialysis-independent renal function in anti-gbm ESRD patients was uncommon (3%) and toward the lower end of the range reported by other studies (3 18%). 12,19 21 This finding may be related to the fact that patients were registered with ANZDATA, if they were considered to have ESRD requiring long-term RRT. Thus, patients not requiring dialysis or those requiring shortterm dialysis for reversible acute kidney injury may not have been registered and therefore not included in this analysis. Importantly, in those patients in the present study who did experience renal recovery, the duration of this recovery was found to be short-lived, such that approximately threequarters of patients returned to dialysis around a median period of 1 year. This suggests that patients experiencing renal recovery on dialysis should have their renal function monitored closely. Another important finding of this study was that transplant patient survival and renal allograft survival rates in individuals with ESRD secondary to anti-gbm were comparable to those of patients with ESRD due to other causes following multivariable adjustment for potentially confounding factors. These results differ from those previously reported by the European Renal Association- European Dialysis and Transplant Association (ERA-EDTA) Registry for 190 patients with Goodpasture s receiving first transplants between 1982 and In that study, the 508 Kidney International (2013) 83,

7 W Tang et al.: ESRD clinical investigation overall survival of Goodpasture s patients following renal transplantation was significantly higher at 5 years than in patients with other causes of ESRD (92% vs. 86%, Po0.05), whereas renal allograft survival rates were lower (44% vs. 58%, p not significant). Multivariable analysis was not performed in the latter paper. Although 5-year patient survival rates in the ERA-EDTA registry study were similar to those of the present investigation (92% vs. 93%, respectively), 5-year renal allograft survival rates were considerably lower (44% vs. 72%). It should be noted that the definition of graft failure in the present study (censored for patient death) and the ERA-EDTA registry study (death with a functioning graft counted as a graft failure) may account for the apparent difference in survival outcomes. In addition, an appreciably higher rate of Goodpasture s recurrence in the ERA-EDTA study, which led to 14% of renal allograft failures (compared with only two allograft failures in the present study), may have accounted for the differences seen. This may reflect different immunosuppressive treatment strategies between Europe and Oceania and/or different approaches to the timing of renal transplantation following the onset of anti-gbm. Although early case series showed frequent biopsy-demonstrated recurrence in the allograft (up to 50%), 23 reports of recurrent leading to renal allograft failure with modern therapeutic approaches are rare. 11,24,25 In 1982, Cameron suggested that recurrence could be reduced to less than 5% if the patient was dialysed for 6 12 months before renal transplantation. 26 However, in the present study, renal allograft survival was comparable between patients with anti-gbm ESRD transplanted o1 year vs. X1 year after commencing dialysis. This analysis was limited by the absence of information regarding anti-gbm antibody titers and duration of remission before renal transplantation. It is also possible that a number of patients transplanted within 1 year of dialysis commencement had experienced anti-gbm some time before they started on dialysis. In this respect, it should be noted that at least 22% of patients with anti-gbm had been referred to a renal unit at least 3 months before the commencement of RRT. The main weakness of this study was the limited depth of data collection. ANZDATA does not collect important information, such as severity of comorbidities, concomitant medications, patient compliance, individual unit management protocols (including plasma exchange and immunosuppression), laboratory values (such as anti-gbm antibody measurements), and detailed renal histopathology. Even though we adjusted for a large number of patient characteristics, the possibility of residual confounding could not be excluded. In common with other Registries, ANZDATA is a voluntary Registry and there is no external audit of data accuracy, including the diagnosis of anti-gbm. In conclusion, anti-gbm is an uncommon cause of ESRD, which was not associated with altered risks of dialysis survival or renal transplant patient or allograft survival, although it was associated with an increased probability of renal recovery (albeit short-lived). Death on dialysis in anti-gbm patients was predicted by older age and a history of pulmonary hemorrhage. MATERIALS AND METHODS Patient population All patients with ESRD enrolled in the ANZDATA registry, who commenced RRT between 15 May 1963 and 31 December 2010, were included in the study. Demographic and clinical data were collected throughout the calendar year by medical and nursing staff in each renal unit and submitted every 6 months to the ANZDATA Registry until 2005 and then annually thereafter. Patients with a primary renal diagnosis of anti-gbm (defined as proliferative glomerulonephritis with linear immunoglobulin G with or without lung hemorrhage) were compared with the remainder of the cohort with an alternative primary renal diagnosis. Patients with anti-gbm were subcategorized according to whether they had a history of pulmonary hemorrhage. Dialysis and transplant eras were determined by the dialysis and transplant commencement dates, respectively: 15 May 1963 to 31 December 1975, 1 January 1976 to 31 December 1985, 1 January 1986 to 31 December 1995, 1 January 1996 to 31 December 2000, 1 January 2001 to 31 December 2005, and 1 January 2006 to 31 December The primary outcomes were patient survival on dialysis (censored for renal function recovery, loss to follow-up, renal transplantation and end of study), time from dialysis commencement to recovery of dialysis-independent renal function (censored for death, loss to follow-up, renal transplantation, and end of study), renal transplant patient survival (censored for allograft failure, loss to follow-up, and end of study), and renal allograft survival (censored for death, loss to follow-up, and end of study). Recovery of dialysisindependent renal function was considered to have occurred if the treating renal unit had recorded that the patient had recovered renal function and completed dialysis therapy. The onset of recovery was defined as the date of the last dialysis treatment. Statistical analysis Results were expressed as frequencies and percentages for categorical variables, mean±s.d. for continuous normally distributed variables, and median (interquartile range) for continuous variables that were not normally distributed. ESRD patients with or without anti-gbm were compared using w 2 tests, two-tailed unpaired t-tests, or Mann Whitney tests, depending on data distribution. The independent predictors of anti-gbm ESRD were assessed by multivariable logistic regression analysis. Time to event analyses were evaluated by Kaplan Meier and multivariable Cox proportional hazards survival analyses. The covariates included in the model for the entire cohort were age, gender, racial origin, ESRD cause, and dialysis era or transplant era (as well as donor type, renal allograft number, and time from dialysis commencement to renal transplantation for renal transplant analyses). In light of the possibility of informative censoring due to differential transplantation rates between patients with and without anti-gbm ESRD, multivariable Cox proportional hazards survival analysis using a competing risks approach was also performed for dialysis patient survival analyses. 27 A supplementary, fully adjusted analysis was also conducted using a contemporary cohort ( ), in which data were available on body mass index, smoking status, history of hypertension, chronic lung, cerebrovascular, ischemic heart, diabetes mellitus, peripheral vascular Kidney International (2013) 83,

8 clinical investigation W Tang et al.: ESRD, and late referral. Statistical analysis was performed using the SPSS software, version 13.0 (SPSS., Chicago, IL) and Stata/SE version 11.2 (StataCorp. CollegeStation, TX). P-values o0.05 were considered statistically significant. DISCLOSURE DWJ is a consultant for Baxter Healthcare Pty and has previously received research funds from this company. He has also received speakers honoraria and research grants from Fresenius Medical Care and is a current recipient of a Queensland Government Health Research Fellowship. FGB is a consultant for Baxter and Fresenius and has received travel grants from Amgen and Roche. SPMcD has received speaking honoraria from AMGEN Australia, Fresenius Australia, and Solvay Pharmaceuticals, and travel grants from AMGEN Australia, Genzyme Australia, and Jansen-Cilag. CMH has received research funds from Amgen, Roche, Shire, and Abbott, travel grants from Amgen, speaking honoraria from Amgen, Roche, Shire, Genzyme, and Fresenius. NCB has previously received research funds from Roche, travel grants from Roche, Amgen, and Jansen Cilag, and speaking honoraria from Roche. All the other authors declared no competing interests. ACKNOWLEDGMENTS We acknowledge the substantial contributions of the entire Australian and New Zealand nephrology community (physicians, surgeons, database managers, nurses, renal operators, and patients) in providing information for and maintaining the ANZDATA Registry database. WT was supported by grants from National Natural Science Foundation of China (Project ), Beijing Municipal Science & Technology Commission (D ) and Fund of Peking University Third Hospital ( ). REFERENCES 1. Salama AD, Levy JB, Lightstone L et al. Goodpasture s. Lancet 2001; 358: Cui Z, Zhao J, Jia XY et al. Clinical features and outcomes of antiglomerular basement membrane in older patients. Am J Kidney Dis 2011; 57: KDIGO Clinical Practice Guideline for Glomerulonephritis. Anti-glomerular basement membrane antibody glomerulonephritis. Kidney Int Suppl 2012; 2: Kluth DC, Rees AJ. Anti-glomerular basement membrane. JAm Soc Nephrol 1999; 10: Peto P, Salama AD. Update on antiglomerular basement membrane. Curr Opin Rheumatol 2011; 23: Taylor DM, Yehia M, Simpson IJ et al. in Auckland. Intern Med J 2012; 42: Chen M, Cui Z, Zhao MH. ANCA-associated vasculitis and anti-gbm : the experience in China. Nephrol Dial Transplant 2010; 25: Cui Z, Zhao MH, Xin G et al. and prognosis of Chinese patients with anti-glomerular basement membrane. Nephron Clin Pract 2005; 99: c49 c Li FK, Tse KC, Lam MF et al. Incidence and outcome of antiglomerular basement membrane in Chinese. Nephrology (Carlton) 2004; 9: Salant DJ. Goodpasture s new secrets revealed. N Engl J Med 2010; 363: Khandelwal M, McCormick BB, Lajoie G et al. Recurrence of anti-gbm 8 years after renal transplantation. Nephrol Dial Transplant 2004; 19: Levy JB, Turner AN, Rees AJ et al. Long-term outcome of anti-glomerular basement membrane antibody treated with plasma exchange and immunosuppression. Ann Intern Med 2001; 134: Johnson JP, Moore J Jr., Austin HA 3rd et al. Therapy of anti-glomerular basement membrane antibody : analysis of prognostic significance of clinical, pathologic and treatment factors. Medicine (Baltimore) 1985; 64: Jindal KK. Management of idiopathic crescentic and diffuse proliferative glomerulonephritis: evidence-based recommendations. Kidney Int Suppl 1999; 70: S33 S Cui Z, Zhao MH. Advances in human antiglomerular basement membrane. Nat Rev Nephrol 2011; 7: Cui Z, Zhao J, Jia XY et al. Anti-glomerular basement membrane : outcomes of different therapeutic regimens in a large single-center Chinese cohort study. Medicine (Baltimore) 2011; 90: Herody M, Bobrie G, Gouarin C et al. : predictive value of clinical, histological and serological data. Clin Nephrol 1993; 40: Briggs WA, Johnson JP, Teichman S et al. Antiglomerular basement membrane antibody-mediated glomerulonephritis and Goodpasture s syndrome. Medicine (Baltimore) 1979; 58: Andrews PA, Sheerin NS, Hicks JA et al. Unusual presentations of anti-glomerular basement membrane antibody mediated are associated with delayed diagnosis and poor outcome. Clin Nephrol 1995; 44: Merkel F, Pullig O, Marx M et al. Course and prognosis of anti-basement membrane antibody (anti-bm-ab)-mediated : report of 35 cases. Nephrol Dial Transplant 1994; 9: Walker RG, Scheinkestel C, Becker GJ et al. Clinical and morphological aspects of the management of crescentic anti-glomerular basement membrane antibody (anti-gbm) nephritis/goodpasture s syndrome. QJMed1985; 54: Briggs JD, Jones E. Renal transplantation for uncommon s. Scientific Advisory Board of the ERA-EDTA Registry. European Renal Association-European Dialysis and Transplant Association. Nephrol Dial Transplant 1999; 14: Choy BY, Chan TM, Lai KN. Recurrent glomerulonephritis after kidney transplantation. Am J Transplant 2006; 6: Deegens JK, Artz MA, Hoitsma AJ et al. Outcome of renal transplantation in patients with pauci-immune small vessel vasculitis or anti-gbm. Clin Nephrol 2003; 59: Sauter M, Schmid H, Anders HJ et al. Loss of a renal graft due to recurrence of anti-gbm despite rituximab therapy. Clin Transplant 2009; 23: Cameron JS. Glomerulonephritis in renal transplants. Transplantation 1982; 34: Lunn M, McNeil D. Applying Cox regression to competing risks. Biometrics 1995; 51: Kidney International (2013) 83,