Abstract

1 |. INTRODUCTION

Focal segmental glomerulosclerosis (FSGS) is a common cause of kidney failure in children; specifically, it was found in a recent report by NAPRTCS (North American Pediatric Renal Trials and Collaborative Studies) registry to be the cause of kidney failure in 15.1% of children transplanted between 2002 and 2016.¹ Recurrent nephrotic syndrome occurs in up to 60% of transplanted pediatric patients with idiopathic FSGS^2,3 with inferior graft survival compared to transplanted patients with other glomerular disease and non-glomerular causes of kidney failure.^4–6 Limited data suggest that kidney transplant recipients with graft failure secondary to FSGS recurrence are at an increased risk of recurrence and graft failure after a subsequent transplant.⁷

Recurrent FSGS leading to graft failure often leads to consideration for re-transplantation. However, FSGS recurrence and graft failure after re-transplantation are a significant concern.^1,7 Jungraithmyar et al. reported nephrotic syndrome recurrence rates of 36%, 48%, and 100% in first, second, and third transplants, respectively, in those with an idiopathic form of FSGS, compared to 0% recurrence in patients with genetic forms of FSGS. In their series of 38 patients with idiopathic FSGS, 40% of those with recurrent nephrotic syndrome in the first graft lost their grafts due to recurrence, and all of them had recurrent nephrotic syndrome after the second transplant.⁷ The recent NAPRTCS study also reports an increased risk of graft failure in FSGS patients who had a prior transplant that failed due to recurrent nephrotic syndrome OR 1.58 (1.10–2.27), p = .013.¹ Given this unfavorable data, many physicians and pediatric kidney transplant centers may be reluctant to re-transplant patients due to concerns of recurrence of FSGS and graft loss, increased morbidity, poor outcomes, and lack of sufficient data regarding mitigation strategies and therapies to prevent or minimize the impact of recurrence of FSGS after re-transplantation.

The purpose of this study was to evaluate outcomes in pediatric kidney transplant recipients with FSGS who had first graft failure due to recurrent nephrotic syndrome and received a second transplant and to evaluate potential factors affecting outcome. In addition, this study surveyed factors influencing physicians’ decisions to re-transplant or not and institution practices throughout North America.

2 |. MATERIALS AND METHODS

3 |. RESULTS

3.1 |. Patient outcomes after FSGS recurrence

Twenty patients from 10 centers were enrolled, with one center enrolling 6 subjects, 2 centers enrolling 3 subjects each, 1 center enrolling 2 subjects, and 6 centers enrolling 1 subject each. Fifty percent of patients were male, 37% were white, 12.5% were black, and 50% were of other races. The mean age was 5.9 ± 4.8 years at initial FSGS diagnosis, 8.8 ± 5.3 years at diagnosis of kidney failure, 9.8 ± 4.8 years at first transplant, and 15.9 ± 4.9 years at re-transplant. Initial transplants took place between 1998 and 2013, and re-transplants took place between 2006 and 2018. Fifty percent of primary transplant patients and 50% of re-transplanted patients received a kidney from a deceased donor. Among the 20 pediatric kidney transplant recipients whose primary transplant failed due to recurrent FSGS, more patients underwent a pre-transplant nephrectomy prior to the initial transplant as compared to allograft nephrectomy prior to the second transplant (47.4% vs. 9.1%, p = 0.05) (Table 1).

TABLE 1

Patient demographics

	First transplant (N = 20)	Second transplant (N = 20)	p-Value
Sex (male)		50%
Race
White		37%
Black		12.50%
Other		50%
Mean age at FSGS diagnosis (years)	5.9 ± 4.8
Mean age at diagnosis of kidney failure (years)	8.8 ± 5.3
Mean age at transplant (years)	9.8 ± 4.8	15.9 ± 4.9
Transplant period	1998–2003	2006–2018
Deceased donor transplant (%)	50%	50%	1
Pre-transplant nephrectomy (%)	47.4%a (9/19)	9.1%b (1/11)	.05

Open in a separate window

^aNative nephrectomy.

^bAllograft nephrectomy.

Thymoglobulin was more commonly used for induction therapy in re-transplants compared to primary transplants (47.4% vs. 10.5%, p = .02) (Table 2). Tacrolimus was the primary calcineurin inhibitor (CNI) used in 63.2% and 74.8% of primary and re-transplants, respectively (p= NS). CNI switch between tacrolimus and cyclosporine occurred in 29.4% of first transplants and 15.8% of re-transplants (p = .43). Pre-transplant and post-transplant use of anti-CD20 therapy was not statistically significant between primary and re-transplants. Pre-transplant plasmapheresis was used in 7 re-transplanted patients (38.9%) vs. only 1 primary transplant patient (5.3%) (p = .03). Post-transplant plasmapheresis was similar in both transplants (57.9% vs. 63.2%, p = 1.00) (Table 2).

TABLE 2

Immunosuppression and plasmapheresis

	First transplant (N = 20)	Second transplant (N = 20)	p-Value
Induction immunosuppression a
Basiliximab	(8/19) 42.1%	(6/19) 31.6%	1
Thymoglobulin	(2/19) 10.5%	(9/19) 47.4%	0.02
Alemtuzumab	(3/19) 15.8%	(4/19) 21.1%	1
None	(2/19) 10.5%	(0/19) 0%
Maintenance immunosuppression
Initial CNI—tacrolimus	(12/17) 70.6%	(15/19) 78.9%	.29
Initial CNI—cyclosporine	(5/17) 29.4%	(4/19) 21.1%	1
Switch in CNI	(5/17) 29.4%	(3/19) 15.8%	.43
Plasmapheresis and anti-CD20 therapy
Pre-transplant plasmapheresis	(1/19) 5.3%	(7/18) 38.9%	.03
Mean (range) # of sessions per patient	6	3.9 (1–10)b
Post-transplant plasmapheresis	(11/19) 57.9%	(12/19) 63.2%	1
Mean # of sessions per patient	19.4 (2–60)	10.2 (8–24)
Pre-transplant anti-CD20 therapy	(7/19) 36.8%	(6/19) 31.6%	1
Post-transplant anti-CD20 therapyc	(5/19) 26.3%	(6/19) 31.6%	1

Open in a separate window

^aUse of more than one induction therapy was reported for some patients.

^b10 sessions in 1 patient, 7 sessions in 2 patients, and 1 session in 3 patients.

^cRituximab in all except one patient who received ofatumumab after the second transplant.

Nephrotic syndrome recurred in 70% of children after the second transplant and was less severe in the second transplant (p = .03) (odds ratio for severe recurrence in second transplant 0.19, 95CI 0.047, 0.81, p = .04). Recurrence after the second transplant was diagnosed a median of 6 days post-transplant compared to 12 days after the primary transplant, p = .01. The nadir protein/creatinine ratio (mg/mg) was statistically significantly lower after the second transplant compared to the first (0.73 vs. 3.75, p < .001), and peak values were higher in first transplants but not significant (37.4 and 12.0, for first and second transplants, respectively, p = .06) (Table 3). One month after re-transplantation, 68.4% were no longer nephrotic, and 42% were in complete remission. At 1 year, 68.4% were in complete remission, 10.5% had proteinuria in the sub-nephrotic range, and only 20.1% continued to have nephrotic-range proteinuria (2 had moderate recurrence, and 2 had severe recurrence), with an overall remission (complete and partial) rate of 78.9%. At last follow-up, 66.7% were still in complete remission, 11.1% had sub-nephrotic proteinuria, and only 16.7% were still with nephrotic-range proteinuria, with an overall remission rate of 77.8% (Table 3). There were no statistically significant differences in delayed graft function, serum creatinine values, or eGFR at one month or one year post-transplant (Table 4), and eGFR was significantly higher at last follow-up in the second transplant (p < .001). Recurrent nephrotic syndrome resulted in graft loss in 30% of those who underwent a second transplant compared to 100% after the first transplant (p < .001), with no difference in the median follow-up time or median time to graft failure. Of the 6 patients who had graft failure, 2 patients had severe recurrence and 4 patients had moderate recurrence. The biopsy findings did not show any statistically significant differences between the first and second transplants. Rates of acute tubular necrosis, acute cell-mediated rejection, and FSGS lesion seen on biopsy were numerically higher in the primary transplants (Table 4). Kaplan-Meier graft survival showed significantly better graft survival in re-transplanted patients (p = .009) (Figure 1).

Open in a separate window

FIGURE 1

Kaplan Meier survival analysis. Graft survival in first and second transplants, Long-rank test p = .009

TABLE 3

Recurrence of FSGS and proteinuria post-transplant

	First transplant (N = 20)	Second transplant (N = 20)	p-Value
Recurrence rate	20 (100%)	14 (70%)	.01
Median time to recurrence (days)	12.0 (IQR 1, 467) (n = 16)	6.0 (IQR 1, 51) (n = 11)	.01
Severity of post-transplant recurrence
Mild	1/17 (5.9%)	5/19 (26.3%)	.03
Moderate*	5/17 (29.4%)	4/19 (21.1%)
Severe*	11/17 (64.7%)	5/19 (26.3%)
Odds ratio for severe recurrence	1	0.19 (0.047, 0.81)	.04
Urine protein/creatinine ratio (mg/mg)
Peak	37.4 IQR (11.6, 30.6)	12.0 IQR (2.0, 12.0)	.06
Nadir	3.75 IQR (0.2, 5.6)	0.73 IQR (0.2, 0.8)	<.001
1 month	17.6 IQR (0.1, 30.3)	2.03 IQR (0.26, 2.29)	.04
1 year	10.8 IQR (1.88, 12.7)	1.40 IQR (0.14, 1.54)	.04
Last follow-up	11.8 IQR (6.3, 18.0)	0.55 IQR (0.10, 0.65)	.16
Remission rates after second transplant a	NA	1 year (n = 19)	Last follow-up (n = 18)
Complete remission		68.40%	66.70%
Sub-nephrotic proteinuria		10.50%	11.10%
Nephrotic-range proteinuria		21.10%	16.70%
Overall remission		78.90%	77.80%

Open in a separate window

^aData were not available for first transplant for comparison.

^*p = .04 when only severe recurrence is compared, and p = .003 when combining and comparing both moderate and severe recurrence.

TABLE 4

Graft function and biopsy data post-transplant

	First transplant (N = 20)	Second transplant (N = 20)	p-Value
Delayed graft function	21% (4/19)	15.8% (3/19)	1
Serum creatinine post-transplant (mg/dl)
1 month	1.22 (IQR 0.7, 1.46)	1.24 (IQR 0.7, 1.42)	.28
1 year	1.31 (IQR 0.77, 1.70)	1.07 (IQR 0.7, 1.44)	.28
Last follow-up	9.89 (IQR 6.15, 10.60)	1.56 (IQR 0.79, 1.90)	.25
Estimated glomerular filtration rate (ml/min/1.73 m2)
1 month	107 (IQR 68, 121) (n = 9)	99 (IQR 59, 123) (n = 16)	.79
1 year	106 (IQR 64, 117) (n = 8)	112 (IQR 71, 139) (n = 14)	.83
Last follow-up	10 (IQR 6, 11) (n = 9)	87 (IQR 53, 124) (n = 16)	<.001
CKD stage
1 month	n = 8	n = 15
Stage 1	50%	53.30%
Stage 2	25%	20%
Stage 3	12.50%	13.30%
Stage 4	12.50%	6.70%
Stage 5	0	6.70%
1 year	n = 7	n = 13
Stage 1	57.10%	61.50%
Stage 2	14.30%	23.10%
Stage 3	14.30%	15.40%
Stage 4	14.30%	0
Stage 5	0	0
Last follow-up	n = 8	n = 15
Stage 1	0	46.70%
Stage 2	0	20%
Stage 3	0	13.30%
Stage 4	25%	20%
Stage 5	75%	0
Graft failure
Graft failure rate	100% (n=20)	30% (n=6)	<.001
Median time to failure (months)	36.0 (IQR 7, 54) (n = 19)	25.2 (IQR 2.5, 37.4) (n = 6)	.23
Median follow-up (months)	66.0 (IQR 36.5, 76.2) (n = 15)	40.0 (IQR 19.0, 68.7) (n = 17)	.23
Graft biopsy (at any point)
Cell-mediated acute rejection	10 (50%)	4 (20%)	.11
Acute tubular necrosis	11 (55%)	6 (30%)	.3
Light microscopy changes of FSGS	12 (60%)	9 (45%)	.51

Open in a separate window

3.1.1 |. Physician practice patterns survey

Thirty-one pediatric nephrologists among 174 surveyed from 21 PNRC centers responded to the survey (17.8% response rate). Of the respondents, 94% reported they would re-transplant patients with graft failure due to FSGS recurrence. Within those who would re-transplant patients, 44.4% indicated they would require a minimum waiting period before re-transplanting patients, 25% of those chose 6 months as a minimum waiting period, 66.7% chose 1 year, 8.3% chose >1–3 years, and 55.6% responded they have no requirement for a waiting period.

Some physicians preferred a living donor (36.4%), and 22.2% of physicians indicated having a protocol for re-transplanting patients after FSGS recurrence at their centers. Most physicians reported deciding on re-transplantation on an individual patient basis (92.6%). For respondents who answered they would not re-transplant such patients, the most important factors influencing the decision not to re-transplant, on a 10-point scale with 1 being most important, were risk of recurrence (2.83) and risk of graft failure (3.16), followed by family refusal (4.12), risk of poor center outcomes and regulatory audit (4.33), refusal by other team members (4.7), refusal by surgeon (5.17), high sensitization (6.29), cost (6.75), and lack of resources (6.75) (breakdown of responses is shown in Figure 2). For the most effective therapy when considering re-transplantation in order of priority on a 5-point scale, nephrologists reported plasmapheresis (1.41) as the most effective therapy, followed by anti-CD20 therapy (2.52), CNI (3.04), and steroids (3.52) (breakdown is shown in Figure 3). The Likert graphs show the various limiting factors influencing decisions regarding re-transplantation, and the therapies viewed as most effective when considering re-transplantation.

Open in a separate window

FIGURE 2

Nephrologist survey responses – factors affecting decision to re-transplant. Numbers reflect priority of response, with the most significant factor assigned the number 1, and least important factor assigned the number 10. Percentages on the horizontal bars reflect the proportion of responses to the specific question

Open in a separate window

FIGURE 3

Nephrologist survey responses – most effective therapies. Numbers reflect priority of response, with the therapy viewed as most effective assigned the number 1, and the least effective assigned the number 4 (the number 5 was for other – there were no responses). Percentages on the horizontal bars reflect the proportion of responses to the specific question

4 |. DISCUSSION

We are encouraged by the unique results of this study which show that failure of a first transplant due to recurrent nephrotic syndrome in pediatric kidney transplant recipients with FSGS is not associated with worse outcome after re-transplantation, contrary to many previous reports. In fact, our study found that second transplants are associated with better graft survival compared to the primary transplants, in spite of high recurrence rates, likely because recurrence after the second transplant in our patients was less severe compared to that in the first transplant. The earlier diagnosis of recurrence in the second transplants does not seem to be associated with worse outcome and is likely due to increased vigilance in evaluating for recurrence of nephrotic syndrome in the second transplant. Although the numbers are small and the design is retrospective, the study is unique in that the populations in the first and second transplants were identical, which allows for excellent case-control matching for sex, race, ethnicity, and socioeconomic status. Most studies focus on the incidence of recurrence after re-transplantation, without stratifying for severity of recurrence. In our study, we examined the severity of recurrence. Even though the definition of severity is somewhat arbitrary, it was defined a priori based mostly on the practical experience of the authors. Our study confirms the previously reported high rates of recurrence after re-transplantation⁷; however, the recurrence was less severe after the second transplant, with significantly lower proteinuria, high rate of complete remission, and improved graft survival. Although our sample size did not allow us to examine the relationship between severity of recurrence and graft failure, proteinuria is a known risk for progression of renal disease, and lower proteinuria is likely associated with improved graft survival. Additionally, many of the re-transplanted children with recurrence achieved remission and were no longer nephrotic and no longer had proteinuria. The small number of patients also limits our ability to analyze the effect of therapies such as plasmapheresis, choice of CNI, and anti-CD20 therapy on outcomes.

Plasmapheresis in our study was used more frequently in re-transplants, with several sessions performed pre-transplant in a few patients. Plasmapheresis has become an essential component of therapy for recurrent nephrosis in transplanted patients with FSGS.^10–13 However, the timing and length of therapy vary greatly and are not well established. The role of prophylactic or pre-emptive plasmapheresis is still in question. Verghese et al. reported no difference in the incidence or timing of recurrent nephrosis or graft survival in 57 patients who received pre-emptive plasmapheresis including pre- and post-transplant.¹⁴ However, post-transplant plasmapheresis therapy was limited per protocol to only 5 sessions of plasmapheresis, including in those with recurrence, and only 2 patients with recurrence received rituximab therapy. Several studies including a meta-analysis showed plasmapheresis to be associated with high rates of complete or partial remission post-transplantation,^15–17 with early initiation of therapy associated with higher likelihood of achieving remission. A recent consensus statement from the CERTAIN (Cooperative European Pediatric Renal Transplant Initiative) study group recommends the early use of plasmapheresis or immunoadsorption in the management of kidney transplant patients with FSGS and a history of previous graft failure due to recurrent nephrotic syndrome.¹⁸ In our study, there were more patients receiving pre-emptive plasmapheresis therapy before the second transplant compared to the first transplant (38.9% vs. 5.3%, p = .03) with the number of plasmapheresis sessions ranging from 1 to 10 sessions. The hypothesis behind pre-transplant plasmapheresis is that it may prevent or decrease the severity of recurrence and improve proteinuria and graft survival by removing circulating protein permeability factors associated with recurrence.^19–22 The recent NAPRTCS report shows improved graft survival in a recent cohort of transplanted patients with FSGS; however, the causes for improvement were not determined, but were likely multifactorial.¹

In our study, the frequency of anti-CD20 therapy utilization was not different between primary and re-transplants. Several studies examining the efficacy of anti-CD20 therapy suggest a positive effect in terms of inducing remission and preventing recurrence although it was most frequently used in conjunction with plasmapheresis.^23–25 The mechanisms of action of anti-CD20 therapy in treating glomerular disease including FSGS and recurrent post-transplant nephrosis are not fully understood. Modulating B- and T-cell interaction has been proposed, and more recently, rituximab has been shown to act in a manner that affects the podocyte cytoskeleton and function by direct modulation of sphingomyelinase and sphingomyelin-phosphodiesterase-acid-like-3b (SMPDL-3b).^24,26

The effect of donor type (living vs deceased) could not be analyzed in this study as 50% of donors were living donors in both primary and second transplants. Although a survival benefit from living donation could not be assessed in this study,^1,27 living donation is associated with less delayed graft function. Delayed graft function often delays the introduction and optimization of CNI use, which is felt to be critical in managing nephrosis post-transplant. In addition, delayed graft function has been shown to be independently associated with greater risk of graft failure in transplanted FSGS patients.¹ Therefore, living donation may be beneficial by lowering the risk of delayed graft function and by allowing patients the ability to receive pre-transplant therapies such as plasmapheresis and anti-CD20 therapies before recurrent disease is well established and potentially modifying the risk and severity of recurrence post-transplant. Conversely, graft failure after living donation can be an even more traumatic and emotional experience, both for the donor and for the recipient. Although living donation as an independent factor has not been shown to be associated with lower recurrence or survival benefit,¹ it would be interesting to evaluate living donor effect in the context of pre-transplant plasmapheresis.

Graft failure due to recurrent FSGS in transplanted patients with FSGS is often a traumatic experience for the patient, family, and the medical team and comes at a high cost, both financially and emotionally. Delayed graft function, edema, infectious complications, frequent procedures, prolonged or frequent hospitalizations, and frequent clinic visits are the hallmarks of this disease, which also has significant psychosocial implications.^28–30 The decision to re-transplant after graft failure due to recurrent nephrotic syndrome is often a difficult one, both for the physicians and for the patients and their families. Reluctance to re-transplant is often driven by fear of repeating the traumatic and emotionally charged experience and concern about another recurrence of nephrotic syndrome and subsequent graft loss. This may be justifiable on the grounds of having such past experience resulting in post-traumatic stress and even depression³¹ and also on the grounds of having data that suggest higher rates of recurrence and graft loss with re-transplantation.⁷

In our survey of nephrologist members of the Pediatric Nephrology Research Consortium, the majority indicated they would re-transplant their patients, with 44% of surveyed nephrologists requiring a waiting period of 1 year before re-transplantation. Waiting 1 year between graft failure and re-transplantation may be reasonable considering the increased risk of early acute rejection, humoral rejection, graft failure, and overall mortality as well as death with a functioning graft in patients re-transplanted within 6 months of graft failure.³² However, the time lag between graft failure and re-transplantation in our cohort is much longer (median of 37 months) than what surveyed nephrologists reported as a preference and longer than previously reported median waiting time of 1.6 years (IQR 0.5, 3.6) for re-transplanted children in the United States.³³ In their report, Van Arendonk et. al. did not find a difference in rates of re-transplantation or waiting times between FSGS patients and other patients. Reasons for the longer waiting times to re-transplantation in our cohort were not examined, but are likely multifactorial. These factors may include morbidities related to nephrotic syndrome, high sensitization status resulting in longer waiting times, changes in waiting times due to changes in the allocation system, psychosocial and socioeconomic factors,³³ hesitancy to re-transplant older patients due to reported higher graft failure rates,^34,35 and to a lesser extent, center-related factors such as lack of consensus among transplant center members, and center outcome and cost considerations, as reported by surveyed nephrologists in this study. The fact that only 22% of centers have an established protocol for re-transplanting such patients, and that the majority of nephrologists would consider re-transplantation on an individual basis, reflects the difficulty in the decision-making process for this challenging disease and unique patient population.

The small sample size and the retrospective study design limit the generalizability of the results of our study. Moreover, there were some missing data elements, especially with the first transplant as many of these transplants were performed at a different center, and before the wide utilization of electronic medical records, which limits the ability to extract data.

5 |. CONCLUSIONS

Our study provides additional information and outcome data, which should help better inform the pediatric nephrologists and transplant centers when considering re-transplantation after recurrence of nephrotic syndrome in kidney transplant recipients with FSGS. Knowing that re-transplants are not destined to fail and that severe recurrence is less likely in a second transplant may encourage more frequent and more timely re-transplantation in this unique population and improve patient overall survival and quality of life. We believe that a prospective study with carefully designed treatment protocols based on the findings of this study and existing data is warranted to guide management and subsequently improve outcomes in this special and very challenging patient population.

Abbreviations:

REFERENCES