Long-term graft outcomes and patient survival are lower posttransplant in patients with a primary renal diagnosis of glomerulonephritis

clinical investigation

www.kidney-international.org

Long-term graft outcomes and patient survival are lower posttransplant in patients with a primary renal diagnosis of glomerulonephritis Rishi Pruthi1,6, Mark McClure2,6, Anna Casula1, Paul J. Roderick3, Damian Fogarty4, Mark Harber5 and Rommel Ravanan2 1

UK Renal Registry, Bristol, UK; 2Richard Bright Renal Unit, Southmead Hospital, Bristol, UK; 3Primary Care and Population Sciences, University of Southampton, Southampton, UK; 4Belfast HSC Trust, Belfast, Northern Ireland; and 5Royal Free Hospital, London, UK

Glomerulonephritis (GN) is the primary diagnosis in 20% to 40% of patients receiving a renal transplant. Here we studied patient survival and graft outcomes in patients with GN transplanted in the UK. UK Renal Registry data were used to analyze patient survival and graft failure in incident transplant patients between 1997 to 2009 who had a diagnosis of primary GN, in comparison to patients transplanted with adult polycystic kidney disease (APKD) or diabetes. Multivariable regression analysis adjusted for age, sex, donor type, ethnicity, donor age, time on dialysis, human leukocyte antigen mismatch, cold ischemic time, and graft failure (for patient survival). Patients were followed up through December 2012. Of 4750 patients analyzed, 2975 had GN and 1775 APKD. Graft failure was significantly higher in membranoproliferative glomerulonephritis (MPGN) type II (hazard ratio: 3.5, confidence interval: 1.9–6.6), focal segmental glomerulosclerosis (2.4, 1.8–3.2), MPGN type I (2.3, 1.6–3.3), membranous nephropathy (2.0, 1.4–2.9), and IgA nephropathy (1.6, 1.3–2.0) compared to APKD. Survival was significantly reduced in patients with MPGN type II (4.7, 2.0–10.8), and those with lupus nephritis (1.8, 1.1–2.9). Overall graft failure for patients with GN was similar to those with diabetes. Thus, in comparison to outcomes in APKD, graft survival is significantly lower in most GNs, with variation in outcomes between different GNs. This information should assist in pretransplant counseling of patients. Further study is required to understand the reduced survival seen in lupus nephritis and MPGN type II, and to improve overall graft outcomes. Kidney International (2016) 89, 918–926; http://dx.doi.org/10.1016/ j.kint.2015.11.022 KEYWORDS: APKD; diabetes; glomerulonephritis; IgA nephropathy; membranoproliferative glomerulonephritis (MPGN); systemic lupus erythematosus ª 2016 International Society of Nephrology

Correspondence: Rishi Pruthi, UK Renal Registry, Learning and Research Building, Southmead Hospital, Bristol BS10 5NB, UK. E-mail: Rishi.Pruthi@ nhs.net 6

Joint first authors.

Received 27 December 2014; revised 22 October 2015; accepted 12 November 2015; published online 21 January 2016 918

G

lomerulonephritis (GN) is a major cause of end-stage renal disease (ESRD), and is the primary diagnosis in 20% to 40% of patients receiving a renal transplant.1 The development of new immunosuppressive therapies has resulted in a significant increase in both short- and longterm graft survival from both living- and deceased-donor transplants over the last 3 decades.2 These therapies have largely been targeted toward controlling acute and chronic rejection, but have had limited impact on the incidence and outcome of recurrent and de novo GN after transplantation.3 Data from a number of studies suggest the cumulative incidence of recurrent GN varies from 2% to 18%.4,5 The impact of recurrence on graft survival is significant. The Renal Allograft Disease Registry in the United States reported that posttransplant allograft survival at 8 years was significantly different for patients with recurrent GN versus no recurrence (34% vs. 53%; P ¼ 0.003).3 A more recent report from the Canadian registry looked at 2026 sequential renal transplant recipients and showed the estimated graft survival at 15 years posttransplant was significantly reduced in patients with posttransplant GN (10.2% vs. 69.7%; P < 0.0001)6 as compared to those who did not develop posttransplant GN. Consequently, recurrence of GN is the third most frequent cause of allograft loss at 10 years, after chronic rejection and death with a functioning graft, and is therefore an important cause of allograft loss.7,8 While clinical recurrence has been reported for most types of GN, the frequency of recurrence may vary between the specific GNs. Focal segmental glomerulosclerosis (FSGS) is reported to recur after transplantation in approximately 20% to 30% of cases,9,10 which is similar to that reported for membranous nephropathy (10%–30%)11–13 and membranoproliferative glomerulonephritis (MPGN) type I (10%–30%).11,14,15 Even higher rates of recurrence are reported for IgA nephropathy (20%–60%)11,16,17 and MPGN type II (50%–100%).14,15,18 Despite the significant risk of disease recurrence and increasing prevalence of patients transplanted with GN, there is currently a paucity of published data on the long-term outcomes after first renal transplant for specific GNs. The primary aim of this study is to determine the risk of graft loss and patient survival for specific GNs, by comparing graft outcomes and survival of patients transplanted in the Kidney International (2016) 89, 918–926

clinical investigation

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

United Kingdom with a primary diagnosis of GN with those of a comparator group in whom the primary renal disease does not recur posttransplantation (adult polycystic kidney disease [APKD]). RESULTS Comparing GN outcomes with APKD patients

Of 5515 patients identified as being eligible for this study (based on the inclusion criteria), 14% (765 patients) were excluded from the analysis due to having either incomplete data for cold ischemic time or human leukocyte antigen (HLA) mismatch. Missing data were equally distributed across the GN group and the APKD group with no statistical differences. The distribution and categorization of primary renal diagnoses among the remaining 4750 patients based on European Renal Association–European Dialysis and Transplant Association coding are shown in Table 1, and the demographic characteristics in Table 2. Overall there were 2975 patients in the GN group and 1775 patients in the APKD control group. The median follow-up time for graft failure was 5.5 years (interquartile range: 3.8–8.3) and for patient survival was 5.9 years (interquartile range: 4.2–8.9). Patients in the GN group were significantly younger (median age 45 vs. 53; P < 0.0001), and although both groups had a male preponderance, this was greater in the GN group (66.7% vs. 54.5%; P < 0.0001). There were also more ethnic minorities in the GN group and more living kidney transplantation (33.9% vs. 28.5%; P < 0.0001), though patients with APKD had more preemptive transplantation (16.1% vs. 12.3%; P ¼ 0.0001). There was no significant difference in HLA mismatch between the 2 groups, and while cold ischemic times were significantly different (14.0 vs. 14.8; P < 0.0001), a separate sensitivity analysis showed that the difference in cold ischemic time disappeared once the higher rates of living donation seen in the GN group were accounted for. Unadjusted 10-year graft survival across the range of GNs analyzed varied with a range of 56.3% to 83.9% (Figure 1). Of

these, MPGN type II and FSGS had the worst outcomes at 56.3% and 65.7%, respectively, while those in the APKD comparison group had a 10-year graft survival of 84.8% (Supplementary Table S1 online). The 10-year risk of graft failure in the multivariable analysis was significantly higher in patients with MPGN type II (hazard ratio [HR]: 3.5, confidence interval [CI] 1.87–6.55), FSGS (HR: 2.39, CI 1.78–3.22), MPGN type I (HR: 2.33, CI 1.63–3.33), membranous nephropathy (HR: 1.99, CI 1.38–2.86), GN histologically proven (HR: 1.68, CI 1.31–2.17), lupus nephritis (HR 1.64, CI 1.13–2.4), and IgA nephropathy (HR: 1.59, CI 1.27–1.99). While MPGN type II had the worst outcomes, the confidence interval for MPGN type II is particularly wide reflecting the small cohort. Graft outcomes in patients with granulomatosis with polyangiitis or those labeled as having crescentic GN did not show any significant difference. Analysis of graft failure using a competing risk model did not show any differences, with subhazard ratios produced similar to those shown in Table 3. Patient survival

Unadjusted 10-year patient survival derived from Kaplan– Meier analysis for the APKD control group was 80.7%. This was similar to that of the different GNs analyzed ranging from 75.6% for MPGN type II to 85.6% for IgA nephropathy (Figure 2 and Supplementary Table S2). From the adjusted model, however (Table 4), 2 of the GNs analyzed were seen to have significantly reduced 10-year survival compared to APKD, with MPGN type II (HR: 4.7, CI 2.0–10.8) and lupus nephritis (HR: 1.8, CI 1.1–2.9) groups having reduced survival. These findings were seen even after adjusting for graft failure. Comparison with patients with diabetes mellitus

Separate analyses were performed to compare outcomes of the GN group (n ¼ 2975) with those of patients with diabetes mellitus (DM) (n ¼ 1873). Baseline patient characteristics of

Table 1 | Number and percentage of patients analyzed by primary renal diagnosis in multivariable model, with the corresponding number of deaths, failed grafts, and median follow-up time

Primary renal diagnosis APKD Crescentic GN FSGS GN histologically not examined GN histologically proven IgA nephropathy Lupus nephritis Membranous nephropathy MPGN type I MPGN type II GPA Total

Total number of patients with complete data, multivariable model

% of total sample

Number of graft failuresa

Number of deathsb

Median follow-up (to death) in years

1775 100 297 258 557 1069 190 183 171 30 120 4750

37.4 2.1 6.3 5.4 11.7 22.5 4 3.9 3.6 0.6 2.5 100

179 15 65 26 98 163 39 36 39 11 15 686

221 15 32 34 75 103 22 29 22 6 17 576

5.9 5.3 5.4 6.2 6.4 6.0 5.8 6.2 6.3 6.1 6.7 5.8

APKD, adult polycystic kidney disease; FSGS, focal segmental glomerulosclerosis; GN, glomerulonephritis; GPA, granulomatosis with polyangiitis; MPGN, membranoproliferative glomerulonephritis. a Over a maximum follow-up time of 10 years. b With functioning graft or after return to dialysis. Kidney International (2016) 89, 918–926

919

clinical investigation

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

Table 2 | Baseline patient characteristics of study cohort analyzed in multivariable model (n [ 4750) APKD (N [ 1775)

Glomerular disease (2975) N

%

Median (IQR)

N

%

Age at transplantation 16–44 45–54 55þ Overall

1461 707 807 2975

49.1 23.8 27.1

22.2 36.9 40.9

45 (35–56)

394 655 726 1775

Recipient gender Male Female

1985 990

66.7 33.3

967 808

54.5 45.5

<0.0001

Recipient ethnicity Asian Black Other White

202 148 74 2551

6.8 5 2.5 85.7

39 43 16 1677

2.2 2.4 0.9 94.5

<0.0001

Time on dialysis Preemptive transplantation <1 year 1–3 years >3 years Overall

366 584 1045 980 2975

12.3 19.6 35.1 32.9

286 367 581 541 1775

16.1 20.7 32.7 30.5

0.001

Donor type Donor after brainstem death Donor after circulatory death Living kidney transplant

1584 382 1009

53.2 12.8 33.9

976 294 505

55 16.6 28.5

<0.0001

Donor age <45 45–54 $55 Overall

1216 859 900 2975

40.9 28.9 30.3

605 534 636 1775

34.1 30.1 35.8

<0.0001

HLA mismatch 0 0DR & 0/1B 0DR & 2B or 1DR & 0/1B 1DR & 2B or 2DR

436 949 1179 411

14.7 31.9 39.6 13.8

234 553 707 281

13.2 31.2 39.8 15.8

Cold ischemic time Overall (hours)

2975

Median (IQR)

P value <0.0001

1.9 (0.7–3.6)

48 (37–56)

14.0 (3.1–18.4)

53 (46–59)

1.6 (0.5–3.5)

50 (41–58)

1775

<0.0001

0.0007

<0.0001 0.2

14.8 (3.9–19.0)

0.0002

APKD, adult polycystic kidney disease; HLA, human leukocyte antigen; IQR, interquartile range.

the DM group, GN group comparison, and data completeness are shown in Supplementary Table S3. Overall there was no significant difference in the unadjusted 10-year graft survival between the GN group (75.6%) and the DM group (71.0%), P ¼ 0.2; or in the results from the multivariable analysis of 10-year risk of graft failure in the DM group (adjusted HR: 1.09, CI 0.93–1.27, P ¼ 0.3) compared to the GN group (data not shown). In contrast, unadjusted 10-year patient survival (derived from Kaplan–Meier analysis) for the DM group (64.6%) was significantly reduced compared to that for the GN group (81.9%), P < 0.001, with results from the adjusted model also showing a significantly reduced survival in the DM group (HR: 2.47, CI 1.76–3.47, P < 0.001) compared to those with GN (data not shown). Comparison of subtypes of GN with IgA nephropathy

To further assess differences among individual GN subtypes, additional analyses comparing patients with different subtypes of GN to those with IgA nephropathy (the largest GN 920

subtype) were also performed. The 10-year risk of graft failure from these analyses showed an increased risk of graft failure in patients with MPGN type II (HR: 2.22, CI 1.19–4.14, P ¼ 0.012), FSGS (HR: 1.52, CI 1.13–2.05, P ¼ 0.006), and MPGN type I (HR: 1.49, CI 1.04–2.14, P ¼ 0.03) compared to patients with IgA nephropathy (Supplementary Table S4). In contrast, patient survival was lower only in patients with MPGN type II group (HR: 3.23, CI 1.38–7.58, P ¼ 0.007) with no other difference between patients with the other subtypes of GN and patients with IgA nephropathy (Supplementary Table S5). Overall group comparisons

Overall the unadjusted 10-year graft survival for patients with GN (75.6%) and those with diabetes (71%) was each significantly lower than that seen in patients transplanted with APKD (84.8%), P < 0.0001 (Figure 3), with the adjusted 10-year risk of graft failure also being lower (HR: 0.59, CI 0.49–0.70, P < 0.0001) in patients with APKD. In Kidney International (2016) 89, 918–926

clinical investigation

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

1

P < 0.0001 0.9

APKD

GraŌ Survival

0.8

CrescenƟc GN FSGS GN histologically not examined GN histologically proven

0.7

IgA nephropathy Lupus nephriƟs Membranous nephropathy

0.6

MPGN type I MPGN type II GPA

0.5 0

1

2

3

4

5

6

7

8

9

10

Years from TransplantaƟon

Figure 1 | Ten-year Kaplan–Meier unadjusted graft survival by primary renal diagnosis. Ten-year Kaplan–Meier unadjusted graft survival for patients transplanted in the United Kingdom between 1997 and 2009 with a primary renal diagnosis of adult polycystic kidney disease (APKD) and a range of different glomerulonephritis (GN): crescentic GN, focal segmental glomerulosclerosis (FSGS), GN histologically not examined, GN histologically proven, IgA nephropathy, lupus nephritis, membranous nephropathy, membranoproliferative glomerulonephritis (MPGN) type I, MPGN type II, and granulomatosis with polyangiitis (GPA).

comparison, the adjusted 10-year risk of graft failure was not significantly different between patients with GN and diabetes (P ¼ 0.2). As for patient survival, there was no significant difference between APKD and GN patients, though this was significantly reduced in the DM group (HR: 3.03, CI 2.24–4.09, P < 0.0001) (Table 5). DISCUSSION

To our knowledge, this is the first study undertaken in the United Kingdom to provide information on long-term graft and patient survival in patients transplanted with a range of different GNs. The relative rarity of such conditions necessitates registry-based evaluation of outcomes, as it is unlikely that any single institution may be able to accrue sufficient experience to describe comprehensive long-term outcomes. In comparison to outcomes in patients with APKD, this study showed that graft survival is significantly lower in patients with MPGN type I, MPGN type II, FSGS, membranous nephropathy, and IgA nephropathy. Poorer graft outcomes were also seen in MPGN I, MPGN II, and FSGS when using IgA nephropathy as the comparator group. This is not unexpected, as one would expect outcomes in patients with GN to be poorer due to recurring disease in their allograft. Primary disease recurrence can occur in every type of GN, and despite the overall progress in renal transplant care with improved short- and long-term outcomes as well as more effective immunosuppressive regimens, there has been limited advance in the area of recurrent GN. While data on disease recurrence were not available for analysis, by using graft failure as a surrogate marker and comparing outcomes with APKD patients (in whom disease does not recur posttransplant), it is highly likely that the differences in outcomes noted in this study are due to disease recurrence. Kidney International (2016) 89, 918–926

Additionally, the observation that differences in graft outcomes among GN subtypes persist (apart from those with membranous nephropathy) when using IgA as a comparator group highlights that “true” differences exist between different GN subtypes. Overall graft survival in patients transplanted with GN in this study appeared better than that reported in other older studies,3,6 which may reflect better use of renin–angiotensin– aldosterone system–blocking agents in recent times and also more effective approaches in managing recurrent disease. The survival analysis suggests that there is a significant patient survival disadvantage for transplant recipients who have a primary diagnosis of MPGN type II (irrespective of the comparator group being APKD or IgA nephropathy) or lupus nephritis when compared to APKD patients. One would assume that graft failure and return to dialysis would have a significant impact on patient survival. However, even after adjusting for graft failure, patients with MPGN II or lupus nephritis had significantly worse survival. This finding has not been previously documented in published literature. Systemic lupus erythematosus is a multisystem disease, and although the mortality rates and disease outcomes have greatly improved over the last 2 decades, the disease is still associated with a significantly reduced life expectancy.19,20 This may be a result of specific organ involvement (renal, central nervous system, lung), infection due to immunosuppression, and premature coronary artery disease, which may be independent of the cardiovascular risk attributed to established renal failure.20–22 MPGN type II (dense-deposit disease) results from abnormal regulation of the alternative complement pathway caused by the presence of C3 nephritic factor or mutations in factor H gene.15 Other than an association with monoclonal 921

clinical investigation

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

Table 3 | Adjusted hazard ratios for graft failure for each primary renal diagnosis from the multivariable analysis, and the hazard ratios for each factor included in the model Adjusted hazard ratio for graft failurea HR

95% confidence interval

N still at risk after 10 years

P value

Primary renal diagnosis APKD Crescentic GN FSGS GN histologically not examined GN histologically proven IgA nephropathy Lupus nephritis Membranous nephropathy MPGN type I MPGN type II GPA

1 1.53 2.39 0.93 1.68 1.59 1.64 1.99 2.33 3.5 1.16

(0.9–2.61) (1.78–3.22) (0.61–1.41) (1.31–2.17) (1.27–1.99) (1.13–2.4) (1.38–2.86) (1.63–3.33) (1.87–6.55) (0.68–1.98)

265 12 20 40 91 130 21 32 24 6 20

0.12 <0.0001 0.7 <0.0001 <0.0001 0.01 0.0002 <0.0001 <0.0001 0.6

Recipient age 1 year increase

0.984

0.977–0.991

-

<0.0001

Recipient gender Female Male

1.12 1

(0.95–1.32) -

-

0.17

Donor type Donor after brainstem death Donor after circulatory death Living kidney transplant

1.15 1.36 1

(0.83–1.60) (0.94–1.97) -

-

0.4 0.1

Recipient ethnicity Asian Black Other White

0.99 1.63 0.61 1

(0.69–1.40) (1.16–2.29) (0.32–1.16) -

-

0.9 0.005 0.13

Time on dialysis Preemptive transplantation <1 year 1–3 years >3 years

0.72 1.02 1 1.41

(0.53–0.97) (0.82–1.26) (1.17–1.70)

-

0.03 0.9 0.0003

Year of transplantation 1997–2009 (annual categories)

0.98

(0.95–1.01)

-

0.13

Donor age 1 year increase

1.024

1.018

-

<0.0001

HLA mismatch 0 0DR & 0/1B 0DR & 2B or 1DR & 0/1B 1DR & 2B or 2DR

0.82 1.01 1 1.08

(0.64–1.07) (0.84–1.22) (0.84–1.39)

-

0.14 0.9

Cold ischemic time 1 hour increase

1.01

(0.99–1.02)

-

0.3

0.5

APKD, adult polycystic kidney disease; FSGS, focal segmental glomerulosclerosis; GN, glomerulonephritis; GPA, granulomatosis with polyangiitis; HR, hazard ratio; MPGN, membranoproliferative glomerulonephritis. a Censoring at death with functioning transplant.

gammopathy23,24 (and hence higher malignancy-related mortality risk), it is not clear whether the poorer survival identified is a true effect (as compared to sample size) and, if so, what reasons may contribute to the same. Conversely, a recent review from the United Network for Organ Sharing database reported lower mortality rates for patients with MPGN types I and II compared with other forms of GN.18 It is important to emphasize that these cohorts reflect not just those reaching ESRD and needing dialysis but also those deemed fit to undergo renal transplantation. Ideally analyses 922

should consider the total number at risk of ESRD in systemic lupus erythematosus for instance, which makes the case for the merger of chronic kidney disease and ESRD registers. There are some limitations associated with this registrybased analysis, most notably the low patient numbers seen in several GN subtypes, which must be acknowledged when interpreting these results. Additionally, the lack of comorbidity and immunosuppression data and the assumption that the corrected variables affect each subgroup in a similar manner are further limitations. In this study, we cannot say Kidney International (2016) 89, 918–926

clinical investigation

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

1

P = 0.34 0.95

PaƟent Survival

0.9

APKD CrescenƟc GN FSGS

0.85

GN histologically not examined GN histologically proven IgA nephropathy

0.8

Lupus nephriƟs

Membranous nephropathy MPGN type I

0.75

MPGN type II GPA

0.7 0

1

2

3

4 5 6 7 Years from TransplantaƟon

8

9

10

Figure 2 | Ten-year Kaplan–Meier unadjusted patient survival by primary renal diagnosis. Ten-year Kaplan–Meier unadjusted patient survival for patients transplanted in the United Kingdom between 1997 and 2009 with a primary renal diagnosis of adult polycystic kidney disease (APKD) and a range of different glomerulonephritis (GN): crescentic GN, focal segmental glomerulosclerosis (FSGS), GN histologically not examined, GN histologically proven, IgA nephropathy, lupus nephritis, membranous nephropathy, membranoproliferative glomerulonephritis (MPGN) type I, MPGN type II, and granulomatosis with polyangiitis (GPA).

whether the increased risk of graft loss is definitely related to recurrence of the primary disease, or whether other factors were involved. The multivariable analysis tries to mitigate this by adjusting for conventional risk factors that may otherwise influence patient and graft survival in both GN and APKD cohorts, though despite this there may be some residual confounding that has not been accounted for. APKD constitutes a useful comparator group, as unlike GN, the disease does not recur in the transplant. We acknowledge, however, that the epidemiology, clinical course, and treatment of APKD differ significantly from those of GN (and even between subtypes of GN), which unavoidably introduces residual confounding that cannot be controlled. Among patients with APKD it is reassuring to see that the unadjusted 10-year graft survival (84.8%) is comparable to that noted by Jacquet et al. in their nationwide longitudinal study of patients with autosomal dominant polycystic kidney disease (ADPKD) in France.25 Though Jacquet et al. reported better graft survival in ADPKD patients compared to nonADPKD patients in their study, other studies have not always found superior outcomes in APKD patients.26 There is also evidence to suggest that specific complications may be more common or unique in patients with APKD who receive a renal transplant. These include posttransplant diabetes mellitus, polycythemia, symptomatic aneurysms, urinary tract infections, and diverticulitis.26–30 The additional analyses comparing outcomes of the GN group with patients with DM provide a balance of the effects of GN with the most common cause of ESRD. The findings of similar 10-year graft survival but significant reduction in 10-year patient survival in the DM group is unsurprising given the competing risk of higher mortality (despite transplantation) in patients with DM.31 Kidney International (2016) 89, 918–926

While lack of data on disease recurrence is clearly a limitation of this study, it also highlights the need for registries to collect robust data on causes of graft failure, and form data linkages with histopathology databases to enhance long-term outcome epidemiologic studies. Furthermore, histopathologyenhanced analysis of recurrent GN after transplantation would provide a unique opportunity to study individual GNs in the presence of various immunosuppressive regimens and could potentially offer further understanding of the pathogenesis of native GN. To conclude, this study provides unique long-term outcome data on transplant patients in the UK for a range of glomerular diseases. In comparison to outcomes in APKD, graft survival is significantly lower in most GNs, with variation in outcomes seen between different GNs. It also suggests reduced overall survival in patients with MPGN type II and those with lupus nephritis—the reasons for which are unclear. Further research is required to understand the reduced survival seen in lupus nephritis and MPGN type II, and to improve overall graft outcomes. The results of this study should assist the pretransplant counseling of patients with GN and enable more informed decision making. MATERIALS AND METHODS Study population and measurements Data for this observational cohort study were obtained from the UK Renal Registry (UKRR), a registered charity established for the development of care of patients with renal disease, which has been collecting data quarterly from UK renal units for over 15 years and is recognized as having one of the few high-quality clinical databases open to requests from researchers. The UKRR currently collects, analyzes, and reports on data from all 71 adult and 13 pediatric renal centers in the UK. It is funded by National Health Service (NHS) 923

clinical investigation

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

Table 4 | Adjusted hazard ratios for patient survival for each primary renal diagnosis from the multivariable analysis, and the hazard ratios for each factor included in the model Adjusted hazard ratio for patient survival HR

95% confidence interval

N still at risk after 10 years

P value

Primary renal diagnosis APKD Crescentic GN FSGS GN histologically not examined GN histologically proven IgA nephropathy Lupus nephritis Membranous nephropathy MPGN type I MPGN type II GPA

1 1.11 1.12 1.13 1.13 1.18 1.81 0.91 1.03 4.68 0.78

(0.65–1.9) (0.75–1.66) (0.78–1.63) (0.86–1.49) (0.92–1.52) (1.13–2.9) (0.61–1.36) (0.65–1.62) (2.03–10.81) (0.47–1.29)

300 16 35 43 114 184 32 38 31 7 24

Recipient age 1 year increase

1.07

(1.06–1.08)

-

<0.0001

Recipient gender Female Male

0.99 0.99

(0.83–1.19) (0.83–1.19)

-

0.94 0.94

Donor type Donor after brainstem death Donor after circulatory death Living kidney transplant

1.35 1.19 1

(0.92–1.97) (0.77–1.83) -

-

0.12 0.43

Recipient ethnicity Asian Black Other White

0.79 0.93 0.97 1

(0.49–1.27) (0.58–1.50) (0.47–2.01) -

-

0.33 0.78 0.93

Time on dialysis Preemptive transplantation <1 year 1–3 years >3 years

0.72 0.68 1 1.57

(0.49–1.06) (0.51–0.90) (1.29–1.92)

-

0.10 0.01

-

<0.0001

Year of transplantation 1997–2009 (annual categories)

0.97

(0.94–1.01)

-

0.15

Treatment modality Dialysis Transplantation

5.01 1

(4.12–6.10) -

-

<0.0001

Donor age 1 year increase

1.005

(0.998–1.011)

-

0.1506

HLA mismatch 0 0DR & 0/1B 0DR & 2B or 1DR & 0/1B 1DR & 2B or 2DR

1.24 1.15 1 1.06

(0.95–1.62) (0.94–1.42) (0.80–1.42)

-

0.1207 0.1835

-

0.6703

Cold ischemic time 1 hour increase

0.999

(0.983–1.015)

-

0.8994

0.7 0.6 0.5 0.4 0.2 0.013 0.7 0.9 0.0003 0.3

APKD, adult polycystic kidney disease; FSGS, focal segmental glomerulosclerosis; GN, glomerulonephritis; GPA, granulomatosis with polyangiitis; HR, hazard ratio; MPGN, membranoproliferative glomerulonephritis.

renal service commissioners, and holds demographic and clinical data on patients having renal replacement therapy in the UK. Renal centers return data electronically to the UKRR, and all data are validated. Full UKRR data set definitions are available at www. renalreg.com. All incident renal transplant patients in the UK (receiving their first renal transplant) between 1 January 1997 and 31 December 2009, aged >16 years with a primary renal diagnosis of GN or APKD, were considered eligible for this study. Patients registered at 924

renal centers that were not reporting to the UKRR at the time of transplantation were not considered eligible (so as to prevent introducing a survivor bias into the analyses), leaving a final cohort of 5515 patients. Patients with GN were coded according to the European Renal Association–European Dialysis and Transplant Association coding system.3 This included crescentic GN (code 16), FSGS (code 17), GN histologically not examined (code 10), GN histologically proven (code 19), IgA nephropathy (code 12), lupus nephritis (code 84), Kidney International (2016) 89, 918–926

clinical investigation

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

1

P < 0.0001 0.9

Gra Survival

0.8 APKD Diabetes 0.7

Glomerulonephris

0.6

0.5

0

1

2

3

4

5

6

7

8

9

10

Years from Transplantaon

Figure 3 | Ten-year Kaplan–Meier unadjusted graft survival for each overall group (glomerulonephritis [GN], adult polycystic kidney disease [APKD], and diabetes). Ten-year Kaplan–Meier unadjusted graft survival for patients transplanted in the United Kingdom between 1997 and 2009 with GN, APKD, and diabetes.

membranous nephropathy (code 14), MPGN type I (code 15), MPGN type II (code 13), and granulomatosis with polyangiitis (code 74). Patients with APKD were used as the main comparator group as this condition does not recur in the transplant. By using graft failure as a surrogate marker and comparing outcomes of patients with GN with APKD, it was our assumption that any differences were related to disease recurrence. While patients with APKD were used as the main comparator group, additional analyses were also performed (i) using an incident cohort of patients aged >16 years, transplanted with a primary renal diagnosis of DM, as a comparator group; and (ii) comparing outcomes among different GN subtypes against those with IgA nephropathy (the largest GN subtype). To enhance the analysis, additional data items (not collected by the UKRR) were obtained for each patient from the UK Transplant Registry held by the Organ Donation and Transplantation Directorate of NHS Blood and Transplant (NHSBT), a special health authority. Data items were obtained using patient identifiers, and included donor type, donor age, recipient ethnicity, HLA mismatch, and duration of cold ischemic time. The Organ Donation and Transplantation Directorate is a national body that administers organ donation and transplantation across the UK and records relevant information on patients on the waiting list as well as details of the episode of transplantation. Patients were followed up until 31 December 2012.

Statistical analyses Statistical analysis included the comparison of patient group characteristics using tests appropriate to their data and distribution. Kruskal–Wallis test was used to assess differences in medians for continuous variables, and categorical data were compared using c2 tests. Kaplan–Meier analysis was used to describe unadjusted graft survival by GN subtype, and a multivariable Cox regression model was developed to calculate the risk of graft loss over 10 years for types of GN versus APKD. The obtained hazard ratios (HRs) were adjusted for age, gender, type of transplant, ethnicity, donor age, time on dialysis pretransplantation, year of transplantation, HLA mismatch, and cold ischemic time. HLA mismatch was categorized (as per NHSBT) into 4 groups ranging from low to high levels of mismatch: group 1, 0 mismatch; group 2, 0DR and 0/1B mismatches; group 3, 0DR and 2B or 1DR and 0/1B mismatches; and group 4, 1DR and 2B or 2DR mismatches. Graft survival time was defined as time from date of transplantation to graft failure (defined as return to dialysis or preemptive retransplantation), death, loss to follow-up, or end of study period, whichever came first. Analysis of graft failure was conducted both censoring at death and using a competing risk model (considering death as a competing event rather than a censoring point). Patient survival analysis was performed using Kaplan–Meier unadjusted survival and Cox regression proportional hazards

Table 5 | Adjusted hazard ratios for graft failure and patient survival from multivariable analyses, for each overall group (GN, APKD, and diabetes) Adjusted hazard ratio for graft failurea

Adjusted hazard ratio for patient survivalb

Primary renal diagnosis

HR

95% confidence interval

P value

HR

95% confidence interval

P value

GN APKD Diabetes

1 0.59 1.10

(0.49–0.70) (0.94–1.28)

<0.0001 0.2

1 0.90 3.03

(0.76–1.07) (2.24–4.09)

0.24 <0.0001

APKD, adult polycystic kidney disease; GN, glomerulonephritis; HR, hazard ratio. a Adjusted for age, gender, type of transplant, ethnicity, donor age, primary renal diagnosis, time on dialysis pretransplantation, year of transplantation, HLA mismatch, and cold ischemic time. b Also adjusted for graft failure. Kidney International (2016) 89, 918–926

925

clinical investigation

models. This model adjusted for age, gender, type of transplant, ethnicity, donor age, time on dialysis pretransplantation, HLA mismatch, cold ischemic time, and graft failure. Change in modality was fitted as a time-dependent variable, which allowed adjustment for return to dialysis and retransplantation. The assumption of proportionality was assessed by graphical methods (Nelson–Aalen plots) and by Schoenfeld residuals in the final models. Robust errors were used to account for clustering in renal centers. All statistical analyses were performed using SAS software version 9.3 (SAS Institute, Cary, NC), and P values of <0.05 were considered as statistically significant. DISCLOSURE

All the authors declared no competing interests. ACKNOWLEDGMENTS

We are grateful to all UK renal centers for providing data on patients starting renal replacement therapy to the UK Renal Registry. We would also like to acknowledge NHS Blood and Transplant for providing additional data items for this analysis. This study was funded by the UK Renal Registry. SUPPLEMENTARY MATERIAL Table S1. Unadjusted KM-derived graft survival outcomes at 1 year and 10 years by primary renal diagnosis Table S2. Unadjusted KM-derived patient survival outcomes at 10 years by primary renal diagnosis Table S3. Baseline characteristics of patients transplanted in the UK between 1997 and 2009 with a primary renal diagnosis of diabetes and glomerulonephritis, analyzed in multivariable model (n ¼ 4848) Table S4. Unadjusted and adjusted hazard ratios for graft failure at 10 years for different glomerulonephritides (with IgA nephropathy as comparator group) Table S5. Unadjusted and adjusted hazard ratios for patient survival at 10 years for different glomerulonephritides (with IgA nephropathy as comparator group) Supplementary material is linked to the online version of the paper at www.kidney-international.org. REFERENCES 1. Chadban SJ. Glomerulonephritis recurrence in the renal transplant graft. J Am Soc Nephrol. 2001;12:394–402. 2. Hariharan S, Johnson CP, Bresnahan BA, et al. Improved graft survival after renal transplantation in the United States. N Engl J Med. 2000;342: 605–612. 3. Hariharan S, Peddi VR, Savin VJ, et al. Recurrent and de novo renal diseases after renal transplantation: a report from the renal allograft disease registry. Am J Kidney Dis. 1998;31:928–931. 4. Schwartz A, Krause PH, Offermann G, et al. Recurrent and de novo renal disease after kidney transplantation with or without cyclosporine A. Am J Kidney Dis. 1991;17:524–531. 5. O’Meara Y, Green A, Carmody M, et al. Recurrent glomerulonephritis in renal transplants: fourteen years experience. Nephrol Dial Transplant. 1989;4:730–734. 6. Chailimpamontree W, Dmitrienko S, Guiyan L, et al. Probability, predictors, and prognosis of post-transplant glomerulonephritis. J Am Soc Nephrol. 2009;20:843–851. 7. Briganti EM, Russ GR, McNeil JJ, et al. Risk of renal allograft loss from recurrent glomerulonephritis. N Engl J Med. 2002;347:103–109.

926

R Pruthi et al.: Long-term transplant outcomes in glomerulonephritis

8. Golgert W, Appel G, Hariharan S. Recurrent glomerulonephritis and renal transplantation: an unsolved problem. Clin J Am Soc Nephrol. 2008;3:800–807. 9. Atero M, Biava C, Amend W, et al. Recurrent focal segmental glomerulosclerosis: natural history and response to therapy. Am J Med. 1992;92:375–383. 10. Savin VJ, Sharma M, McCarthy ET, et al. Circulating factor associated with increased glomerular permeability to albumin in recurrent focal segmental glomerular sclerosis. N Engl J Med. 1996;334: 878–883. 11. Floege J. Recurrent glomerulonephritis following transplantation: an update. Nephrol Dial Transplant. 2003;18:1260–1265. 12. Denton MD, Singh AK. Recurrent and de novo glomerulonephritis in the renal allograft. Semin Nephrol. 2000;20:164–175. 13. Couchoud C, Pouteil-Noble C, Colon S, et al. Recurrence of membranous nephropathy after renal transplantation. Incidence and risk factors in 1614 patients. Transplantation. 1995;59:1275–1279. 14. Appel GB, Cook HT, Hageman G, et al. Membranoproliferative glomerulonephritis type II (dense deposit disease): an update. J Am Soc Nephrol. 2005;16:1392–1403. 15. Little MA, Dupont P, Dorman A, et al. Severity of primary MPGN, rather than MPGN type, determines renal survival and post-transplantation recurrence risk. Kidney Int. 2006;69:504–511. 16. Kim YS, Moon JI, Jeong HJ, et al. Live donor renal allograft in end-stage renal failure patients from immunoglobulin A nephropathy. Transplantation. 2001;71:233–238. 17. Ponticelli C, Traversi L, Feliciani A, et al. Kidney transplantation in patients with IgA mesangial glomerulonephritis. Kidney Int. 2001;60: 1948–1954. 18. Angelo J, Bell C, Braun M, et al. Allograft failure in kidney transplant recipients with membranoproliferative glomerulonephritis. Am J Kidney Dis. 2011;57:291–299. 19. Urowitz MB, Gladman DD, Tom BD, et al. Changing patterns in mortality and disease outcomes for patients with systemic lupus erythematosus. J Rheumatol. 2008;35:2152–2158. 20. Bernatsky S, Boivin JF, Joseph L, et al. Mortality in systemic lupus erythematosus. Arthritis Rheum. 2006;54:2550–2557. 21. Asanuma Y, Oeser A, Shintani AK, et al. Premature coronary-artery atherosclerosis in systemic lupus erythematosus. N Engl J Med. 2003;349: 2407–2415. 22. Gustafsson J, Gunnarsson I, Borjesson O, et al. Predictors of the first cardiovascular event in patients with systemic lupus erythematosus—a prospective cohort study. Arthritis Res Ther. 2009;11:R186. 23. Hill PA, Desmond M. Membranoproliferative glomerulonephritis type II (dense deposit disease) in association with monoclonal gammopathy. Nephrology (Carlton). 2007;12:419–420. 24. Sepandj F, Trillo A. Dense deposit disease in association with monoclonal gammopathy of unknown significance. Nephrol Dial Transplant. 1996;11: 2309–2312. 25. Jacquet A, Pallet N, Kessler M, et al. Outcomes of renal transplantation in patients with autosomal dominant polycystic kidney disease: a nationwide longitudinal study. Transpl Int. 2011;24:582–587. 26. Hadimeri H, Nordén G, Friman S, Nyberg G. Autosomal dominant polycystic kidney disease in a kidney transplant population. Nephrol Dial Transplant. 1997;12:1431–1436. 27. Hamer RA, Chow CL, Ong AC, McKane WS. Polycystic kidney disease is a risk factor for new-onset diabetes after transplantation. Transplantation. 2007;83:36–40. 28. Andreoni KA, Pelletier RP, Elkhammas EA, et al. Increased incidence of gastrointestinal surgical complications in renal transplant recipients with polycystic kidney disease. Transplantation. 1999;67:262–266. 29. Wijdicks EF, Torres VE, Schievink WI, et al. Cerebral haemorrhage in recipients of renal transplantation. Mayo Clin Proc. 1999;74:1111–1112. 30. Stiasny B, Ziebell D, Graf S, et al. Clinical aspects of renal transplantation in polycystic kidney disease. Clin Nephrol. 2002;58:16–24. 31. Rangel EB, de Sa JR, Melaragno CS, et al. Kidney transplant in diabetic patients: modalities, indications and results. Diabetol Metab Syndr. 2009;1:1–2.

Kidney International (2016) 89, 918–926