Transplant nephrectomy

Abstract

About 10% of all renal allografts fail during the first year of transplantation and thereafter approximately 3%-5% yearly. Given that approximately 69 400 renal transplants are performed worldwide annually, the number of patients returning to dialysis following allograft failure is increasing. A failed transplant kidney, whether maintained by low dose immunosuppression or not, elicits an inflammatory response and is associated with increased morbidity and mortality. The risk for transplant nephrectomy (TN) is increased in patients who experienced multiple acute rejections prior to graft failure, develop chronic graft intolerance, sepsis, vascular complications and early graft failure. TN for late graft failure is associated with greater morbidity and mortality, bleeding being the leading cause of morbidity and infection the main cause of mortality. TN appears to be beneficial for survival on dialysis but detrimental to the outcome of subsequent transplantation by virtue of increased level of antibodies to mismatched antigens, increased rate of primary non function and delayed graft function. Many of the studies are characterized by a retrospective and univariate analysis of small numbers of patients. The lack of randomization in many studies introduced a selection bias and conclusions drawn from such studies should be applied with caution. Pending a randomised controlled trial on the role of TN in the management of transplant failure patients, it is prudent to remove failed symptomatic allografts and all grafts failing within 3 mo of transplantation, monitor inflammatory markers in patients with retained failed allografts and remove the allograft in the event of a significant increase in levels.

Key Words: Allograft intolerance syndrome, Hemorrhage, Immunosuppression, Infection, Panel reactive antibody, Patient survival, Subsequent graft survival

INTRODUCTION

It is estimated that 7%-10% of all renal allografts fail during the first year of transplantation. Thereafter, approximately 3%-5% allografts fail yearly[1,2]. If the 69 400 renal transplants performed worldwide in 2008[3] is an indication of annual transplant activity, then the number of patients returning to dialysis following failure of their renal allograft is increasing in absolute numbers year after year. A failed transplant kidney, whether maintained by low dose immunosuppression or not, elicits an inflammatory process characterized by hypoalbuminemia, erythropoietin (EPO) resistance, anemia, high ferritin and elevated C-reactive protein (CRP); and is associated with an increased incidence of infectious and cardiac complications, failure to thrive and increased morbidity and mortality[4-6]. The United States Renal Data System data suggest that overall annual adjusted death rates were > 3-fold higher after graft loss as compared to before graft loss (9.42% vs 2.81%). Death after graft loss is strongly associated with infection, acute rejection or thrombosis related graft failure[7]. Low-dose immunosuppressive medication is often continued in patients returning to dialysis with a failed renal allograft in situ in order to reduce the risk of rejection but this is associated with increased mortality both from infectious and cardiovascular diseases[8]. However, continuation of immunosuppressive medication does not result in fewer rejections[8] and in another series that did not continue immunosuppressive therapy, increased rejection rates were not reported[9].

Following allograft failure, patients may be classified into the following categories: permanent dialysis/unsuitable for re-transplantation; bridge dialysis/waiting list for re-transplantation; and unsuitable for dialysis or re-transplantation. Current controversies relate to what to do with a failed allograft in patients on dialysis awaiting re-transplantation and the role of transplant nephrectomy (TN) in asymptomatic patients or dialysis patients unsuitable for renal transplantation[10]. In asymptomatic patients, the risk of surgical morbidity and mortality and a rising number of circulating antibodies associated with TN are among the arguments to support non intervention. On the other hand, chronic inflammation, the potential for malignancy, infection and the need for low-dose immunosuppression are concerns often addressed by performing a pre-emptive nephrectomy[11]. Other authors argue that TN should not be routinely performed but be reserved for those patients who develop symptoms related to the allograft or those who require space for re-transplantation[1,12]. About 20% of patients with established renal failure on the waiting list for renal transplantation in the US have had a TN[13]. The effect of TN on outcome of subsequent transplantation has been investigated by several authors[14-16] but it remains unclear whether removal of the failed allograft is beneficial or not. It is not well understood whether removal of the failed renal allograft affects patient survival while receiving long-term dialysis[13]. The aim of this review is to provide an update on current practice regarding TN with a view to proffering recommendations.

NEED FOR TRANSPLANT NEPHRECTOMY

Reasons for transplant nephrectomy

Urgent and non urgent reasons for TN may occur during the early or late phase of transplantation or transplant failure. Indications vary according to the time course after transplantation. The common indications for TN are shown in Table 1.

Table 1 Indications for transplant nephrectomy.

Type	Reason	Comments
Cause of failure	Renal vein thrombosis	Transplant pyonephrosis
	Renal artery occlusion
	Acute rejection refractory to treatment[17]
	Sepsis
Allograft associated complications	Primary non function	Apart from tumors, these occur in the early phase after transplantation
	Bleeding post biopsy[17]
	Transplant rupture[18]
	Rupture of pseudoaneurysm
	Malignant tumors[18]
Following failure	Pain due to uncontrolled rejection[5]
	Graft tenderness[14]
	Anemia resistant to treatment
	Recurrent urinary tract infection
	Persistent hematuria
	Sepsis
Miscellaneous	Preliminary measure prior to re-transplantation[19]	For example, failed kidney-pancreas transplant awaiting combined re-transplantation
	Recurrence of primary disease
	Polyoma virus nephropathy

With the increasing use of kidneys derived from elderly donors, it is likely that there is a higher risk of developing neoplasms[20,21]. Malignant degeneration of a chronically rejected kidney allograft has been reported several years after the transplant[22]. Treatment options for renal cell carcinoma (RCC) in a renal allograft include radical nephrectomy and nephron-sparing surgery (NSS). The risk of local recurrence of an RCC in kidney allograft raises the need for continued surveillance after NSS[23]. With respect to post transplant lymphoproliferative disease involving the allograft, TN alone has superior patient survival over other measures without allograft removal[24].

Renal allograft rupture is a rare but potentially serious complication and may be associated with acute rejection[25], renal vein thrombosis or severe acute tubular necrosis in the absence of acute rejection. Frequently, nephrectomy is necessary but conservative surgical treatment could be attempted to preserve the allograft in selected cases[26]. In cases of later graft failure, the main indication for TN is graft related complications associated with chronic rejection in 58.2% of cases[18].

Risk factors for transplant nephrectomy

Prior rejection: In a study of 41 patients, Madore et al[27] used univariate analysis to show that allograft nephrectomy was significantly more frequent in patients with a history of two or more episodes of acute rejection than in patients with no rejection episode (83% vs 30%, P = 0.03) or if the immunosuppressive regimen included cyclosporine (62% vs 27.3%, P = 0.04). Following multivariate analysis of the same data, the number of previous episodes of rejection was found to be the only significant predictor for allograft nephrectomy. Patients with a history of numerous rejection episodes may either suffer continuing subclinical rejection or have an increased propensity to develop acute rejection when immunosuppression is reduced or stopped. They should thus be considered more likely to require allograft nephrectomy once immunosuppression is withdrawn. The introduction of an effective immunosuppressive agent in the early 1980s led to a reduction in the need for TN. Of the 280 recipients undergoing transplantation before 1984 (pre-cyclosporin era), 70 (25%) underwent TN, whereas only 61 (12.5%) of the 486 recipients in the cyclosporine period had TN (P < 0.01)[28].

Chronic graft intolerance: It is currently standard practice to leave failed kidney transplants in place upon return to HD and to treat symptomatic graft intolerance syndrome with immunosuppression. While this approach may reduce clinical symptoms in the short term, treatment failure necessitating TN occurs in the majority of cases. Medical treatment of graft intolerance syndrome has not been shown to reduce chronic inflammation or decrease mortality. Similarly, embolization of failed kidney transplants is associated with a high rate of treatment failure and has not been shown to reduce chronic inflammation[29]. Therefore, biochemical evidence of chronic inflammation increases the risk of TN prior to the development of clinical symptoms[30].

Sepsis: Dialysis patients are at risk for sepsis and the risk may be even higher among transplant failure patients because of previous or ongoing immunosuppression. Analysis of the incidence of sepsis among 5117 patients initiating dialysis after transplant failure between 1995 and 2004 in the US showed the highest rates in the first 6 mo - 35.6 per 100 patient years (95% CI: 29.4-43.0) during 0-3 mo and 19.7 per 100 patient years (95% CI: 17.2-22.5) during 3-6 mo[31]. In comparison, the sepsis rate among incident dialysis patients between 3 and 6 mo after dialysis initiation was 7.8 per 100 patient years (95% CI: 7.3-8.3). Smak Gregoor et al[4] compared the morbidity and mortality due to infections between patients with retained failed allograft and those undergoing TN and reported more serious and life-threatening infections associated with those who have retained failed allografts on low-dose immunosuppression.

Vascular complications: Anastomotic pseudoaneurysm following renal transplantation is uncommon but may result in graft loss. TN was required in five of six patients who underwent surgical excision of a ruptured anastomotic pseudoaneurysm[32,33]. Dorffner at al[34] evaluated the outcome in seven patients in whom iatrogenic vascular complications of renal transplants were treated with catheter embolization using coils. Angiographic success with total occlusion of the vascular injury was achieved in five of the seven patients but in two cases nephrectomy was necessary because of renal artery occlusion or acute hemorrhage at the renal artery anastomosis, respectively.

Percutaneous embolization: Gonzalez-Satue et al[29] successfully managed 28 (85%) of 33 patients by percutaneous embolization without complications but the remaining five (15%) required TN. If it can be proved that percutaneous embolization does not make subsequent TN more hazardous, then it may become an attractive option in managing symptomatic failed allografts with the possibility of reducing the necessity for TN.

Probability of transplant nephrectomy

Allograft nephrectomy is not routinely performed at the time of graft failure when loss of graft function occurs more than 6 mo after transplantation. However, little is known about the characteristics that make patients more likely to require allograft nephrectomy. The probability of TN is highly dependent on the duration of allograft function prior to failure. There is no agreement on the definition of early allograft failure with various time points being used - < 2 mo[35], < 6 mo[36] and <

12 mo[10,17]. Early graft failures are far more likely to result in TN. The National Health Service Blood and Transplant (UK) statistics show that 41% of allografts failing during the first 3 mo after transplantation are removed compared to 23% during 3 to < 12 mo, 9% during 12 to < 24 mo and 4% after 2 years[37]. Zerouali et al[38] studied the outcome of 182 renal transplants who were managed according to a conservative treatment policy for allografts failing after 1 mo of transplantation. Of 63 failed allografts, 53 grafts (84%) were removed: 100% of the group failing during 0-1 mo, 86% 1-12 mo and 68% > 1 year. Further evidence is provided by a large observational study of the likelihood of TN among 19 107 transplant failure patients between 1995 and 2003 in the US. Johnston et al[10] showed that among 3707 patients with early transplant failure (graft survival < 12 mo), nephrectomy was performed in 56% compared to 27% among 15 400 patients with late transplant failure (graft survival ≥ 12 mo). Secin et al[18] found that the steepest rise in the Kaplan-Meier cumulative incidence curve was within the first 2 years after transplantation, reaching 11% (95% CI: 9-15). About half of all TN in their series were performed within 1 year of patients returning to dialysis.

The cumulative incidence of TN is 4.5%-84.4%[10,17,18,35,39-41]. Following a policy of removing early and late failed grafts only in symptomatic patients, 70 of 762 (9.2%) failed allografts were removed by Mazzucchi et al[35]. In a review of 631 renal transplants performed in 598 patients and 91 transplant nephrectomies in 85 patients in one institution from 1970 to 2000, Secin et al[18] reported the cumulative incidence of TN of 74% (95% CI: 49-90) at 10 years after return to dialysis[18]. The advent of cyclosporine significantly decreased the TN rate at the expense of fewer graft failures but not at the expense of a lower amount of graft related symptoms after patients returned to dialysis[18].

Hansen et al[42] retrospectively examined the courses of 34 graft failures leading to TN in 19 patients. Having diagnosed graft failure, the immunosuppressive treatment was continued for about 2-3 mo and then tapered slowly. They reported no deaths related to graft failure but in three cases, a delay in TN caused complications such as sepsis and coagulopathy. They concluded that continuing immunosuppression a few months after having diagnosed graft failure may postpone or avoid TN.

TECHNICAL CONSIDERATIONS

TN is performed using the extracapsular or intracapsular technique (Table 2)[35,43] via the same oblique incision in the iliac fossa as the original transplantation. In order to reduce the complication rates, Zomorrodi et al[44] described a debulking technique, leaving an intact ureter with intracapsular instillation of betadine. In their technique, the capsule of the allograft was exposed and incised from pole to pole for the renal parenchyma to be bluntly dissected free with an index finger. After TN, small doses of immunosuppressive drugs were continued for at least 2 mo. This surgical approach to allograft nephrectomy was applied in 25 patients between 1991 and 2006 with no significant complications. To avoid bleeding during and following TN, some have adopted a slight modification of the technique of renal extracapsular nephrectomy, performing three running sutures between the two sides of the renal capsule. Ghinolfi et al[45] used this technique in nine patients without any complications.

Table 2 Techniques of transplant nephrectomy.

	Extracapsular	Intracapsular
Description of technique[43]	Dissection outside kidney capsule to remove kidney including capsule, ureter and most of the transplant vessels	Dissection within kidney capsule which is densely ardent to surrounding tissues by the time of late graft removal
Mean procedure time[35]	125.3 min	109.4 min
Blood loss[35]	385 (60-1500) mL	638 (70-2200) mL
Blood transfusion[35]	6/17 (35.3%)	16/30 (53.3%)
Complication rate[35]	3/17 (17.6%)	6/30 (20%)
Comments	Suitable for early graft failures	Generally used for late graft failures

Among patients developing a renal tumor in the kidney allograft, TN reduced the quality of life due to the loss of renal function, prompting the search for alternate treatment options such as NSS, and local ablative procedures (like radiofrequency ablation or cryoablation) have been described[23]. An important issue is to find the balance between the preservation of transplant function on the one hand, which is dependent on the maintenance of an immunosuppressive regimen, and a sufficiently radical tumor therapy on the other hand[21,46].

In the case of TN due to anastomotic pseudoaneurysms, positioning a stent in the external iliac artery prior to the procedure avoids ligature of the iliac artery in the majority of cases[33]. In an attempt to decrease blood loss, Neschis et al[47] performed intraoperative pre-nephrectomy coil embolization of the transplant renal artery during 13 consecutive transplant nephrectomies and compared them with the 13 most recently performed consecutive transplant nephrectomies without coil embolization. They demonstrated a reduction in the estimated blood loss (465 mL vs 198 mL, P = 0.035) and shorter operating time in the embolization group.

COMPLICATIONS

TN is considered to be a procedure with high morbidity (17%-60%) and mortality (1.5%-14%) rates due to immunosuppression, co-morbid conditions of patients and the technical difficulty of the procedure[11,18,35,38-40,44,48-50]. In the pre-cyclosporin era, TN was considered a risky intervention with a mortality ranging from 7.3%-38.7%[11,49,50]. Death was mainly due to septic complications. Since the more widespread use of cyclosporin and the lower doses of steroids, TN has become a safer procedure with almost no mortality and a low incidence of major morbidity[42,51].

Infection is the most frequent complication of TN followed by hemorrhage (Table 3). Early reports of TN after failed transplantation revealed a high incidence of wound infection and sepsis due to the effects of immunosuppression[52]. TN site sepsis or generalised sepsis was encountered in 56% of cases in an earlier report, where it was observed that sepsis often remained occult until subsequent transplantation and immunosuppression[53]. Without prophylactic antibiotics, wound infections after TN were common (20% with 81% due to staphylococcal organisms) such that wounds containing a pre-existing focus of infection or at a high risk for infection were left open for secondary healing[54]. The use of modern techniques of wound closure/management and more powerful antibiotics have improved the outlook for these patients.

Table 3 Complications of transplant nephrectomy.

Complication	Comments
Hemorrhage	Commonest problem
Hematoma
Infection: superficial deep	Extremely high in earlier series[39]
Lymphocele
Pseudoaneurysm of iliac vessels	Treat by endovascular techniques
Urinary fistula
Bowel injury
Injury to obturator nerve or lateral cutaneous nerve of the thigh	Uncommon
Miscellaneous

Mazzucchi et al[35] reviewed the surgical complications of 70 consecutive patients who underwent TN between May 1994 and April 2002, noting that the mean blood loss, likelihood of blood transfusion and severity of complications were higher in patients undergoing TN for late allograft failure. The technique of TN also influenced the complication rates (Table 2).

The incidence of vascular complications following TN ranges from 0.9%-14%[17,55] and are associated with a significantly poor outcome. The presence of sepsis is a significant risk factor for vascular complications[56]. Nine patients (5.6%) sustained significant vascular complications with three deaths - two from overwhelming sepsis and one from an intra-cerebral hemorrhage. Immediate attempts to reconstruct the vascular supply to the lower limb are associated with a high morbidity rate. Where possible, vascular reconstruction should be deferred and the external iliac artery ligation can be performed safely with a low risk of limb ischemia[17]. Iliac pseudoaneurysm may also develop following TN[57-59]. Endovascular treatment with a covered stent is a safe and effective alternative to open surgery in the treatment of a symptomatic pseudoaneurysm arising from a TN site[60].

OUTCOME

One series reported the need for blood transfusion in 58.6% during TN but found a significant improvement in the hematological, biochemical and clinical parameters - EPO resistance index, serum albumin, prealbumin, ferritin, fibrinogen, CRP and erythrocyte sedimentation rate. Six months after allograft failure, those with TN had higher Hb and serum albumin levels, and lower CRP and EPO resistance index in comparison to incident hemodialysis patients. Parameters in those without TN showed no change during follow-up[5]. Another series demonstrated complete resolution of pain, fever and macroscopic hematuria in all patients and hypertension in 8 (36%) of 22 patients[38]. Wan et al[61] described an unusual case of refractory inflammatory ascites, along with cachexia, hypoalbuminemia and EPO resistance, associated with the chronic inflammatory state induced by a failed kidney transplant with no other identifiable cause of the ascites. The inflammatory ascites did not respond to antibiotic therapy but promptly resolved, along with the other manifestations of the chronic inflammatory state, after TN.

Whether the timing of TN after allograft failure has any significant effect on outcome is not fully known. Toledo-Pereyra et al[11] compared the morbidity, mortality and hospitalisation costs of 37 patients undergoing TN within 14 d after graft failure and return to dialysis, with 31 patients undergoing delayed TN (more than 14 d after graft failure and return to dialysis). Although there were no significant differences in patient morbidity and mortality between these groups, there was, however, a substantial increase (P < 0.05) in the cost of hospitalization in the delayed nephrectomy group[11]. It can be argued that a 14 d cut off is probably too short a period to make any meaningful difference.

TN reduces the risk of sepsis in patients returning to dialysis after transplant failure[31]. TN was not associated with septicemia but patients who were hospitalized for sepsis had an increased risk for death with a hazard ratio (HR) of 2.93 (95% CI: 2.64-3.24, P < 0.001). The risk of death following TN was dependent on whether it was performed for early (associated with an increased risk of death) or late (decreased risk of death) graft failure - HR were 1.13 (95% CI: 1.01-1.26) and 0.89 (95% CI: 0.83-0.95) respectively. Strategies to prevent sepsis during the transition from transplantation to dialysis may improve the survival of patients with allograft failure.

TN improves survival on dialysis. Ayus et al[40] identified all adults who received a kidney transplant and returned to long-term dialysis after renal allograft failure between January 1994 and December 2004 from the US Renal Data System. Among 10,951 transplant recipients who returned to long-term dialysis, 3451 (31.5%) received an allograft nephrectomy during follow-up. Overall, 34.6% of these patients died during follow-up. TN was associated with a 32% lower adjusted relative risk for all-cause mortality (adjusted HR 0.68, 95% CI: 0.63-0.74) after adjustment for sociodemographical characteristics, comorbidity burden, donor characteristics, interim clinical conditions associated with TN and propensity to receive TN. Study of this large, nationally representative sample of high-risk patients returning to long-term dialysis after failed kidney transplant revealed that TN was independently associated with improved survival[40]. However, the group that underwent TN was composed of younger individuals who were less likely to have diabetes and/or cardiovascular disease, more likely to be black, to have required the use of T cell-depleting antibodies and to have experienced anemia, sepsis and urinary tract infections. Despite these complications, the rate of death within 30 d of the TN was only 1.5% (53 deaths of 3451 patients). In their series, those undergoing TN were more likely (10% vs 4.1%, P < 0.001) to receive a second transplant when compared with those who did not undergo a nephrectomy[40]. However, Perl et al[13] were unable to demonstrate whether immunosuppression reduction or TN affected survival.

EFFECT OF TN ON SUBSEQUENT TRANSPLANTATION

TN increases the likelihood of developing antibodies to mismatched human leukocyte antigens (HLA), partly due to the absence of immunosuppression and the removal of the organ that adsorbs the antibodies. Rosenberg et al[62] found that 30% (10/34) of their patients had antibodies to all of the mismatched HLA, 43% had antibodies to some and 27% did not develop antibodies to any of the mismatched antigens. Sixty percent of the patients who developed antibodies to all of the mismatched HLA had had a TN. In an earlier study by Adeyi et al[63], sera from 27 patients with HLA-mismatched kidney transplants that had been removed following rejection were screened for HLA-specific antibodies by direct complement-dependent lymphocytotoxicity with HLA-typed cell panels. Circulating donor-specific antibodies were detected in 3 cases (11%) before and in 26 cases (97%) after allograft nephrectomy. These findings demonstrate the production of donor-specific antibodies in patients with rejected transplants but, in most cases, they were undetectable before nephrectomy because the graft had adsorbed them. This has important implications with respect to utilising “unacceptable antigens” in an allocation system for patients awaiting a second transplant who have had a TN[62].

Sumrani et al[14] studied 95 consecutive cyclosporin treated re-transplant patients - 52 without primary allograft nephrectomy, 35 with TN prior to re-transplantation due to symptoms and 8 had TN at the time of re-transplantation. Nephrectomy of the primary allograft prior to re-transplantation was associated with a significant subsequent rise in preformed cytotoxic antibody levels, a significantly higher incidence of delayed graft function among re-transplants and a trend toward decreased allograft survival in the subgroup who lost their primary allografts in the first year post transplant. The incidence of acute rejection and 3-year post transplant renal function in re-transplants were not, however, influenced by nephrectomy of the primary allograft. The effect of TN on subsequent renal transplant function is shown in Table 4[10,14-16,19].

Table 4 Effect of transplant nephrectomy on subsequent transplant function.

Study	TN	No	PRA (%) > 30%	HLA (mm)	Delayed graft function (%)	Acute rejection (%)	Graft survival (%)		Comments
							1 yr	5 yr
Sumrani et al[14], 1992	Yes	35	57		63				AR rates similar between groups
	No	52	33		30
Yagmurdur et al[16], 2005	Yes	21		1.9 ± 1.1		43	83	45	Retrospective
	No	32		1.0 ± 0.6		38	69	68	TN has no advantage
Johnston et al[10], 2007	Yes	6213	14.7	64.8%1	33.6				USRDS 1995-2003
	No	12 894	12.6	53.4%1	24.4				Early TN associated with lower risk of transplant failure
Ahmad et al[15], 2009	Yes	68	372			49.1	83.8	66.2	PRA level significantly influences graft survival independent of TN
	No	21	292			31.2	94.7	69.5
Schleicher et al[19], 2011	Yes	121	16	2.2 ± 1.5	29.3	29.7			Retrospective study
	No	45	2	2.0 ± 1.7	20	13.3			Graft survival worse in TN group

¹4-6 mismatches;

²Proportion of positive patients.

Abouljoud et al[36] studied 192 patients receiving a primary and a subsequent kidney transplant between 1980 and 1992, reporting that patients having primary TN had a worse second allograft outcome than patients who kept their failed grafts (P = 0.0003). They demonstrated a significant relationship between primary allograft survival and re-transplant outcome which persisted even after excluding patients whose first graft failed within 6 mo of transplantation. Multivariate analysis identified primary allograft nephrectomy, older donor age, longer interval from nephrectomy to re-transplant and lack of antibody induction as negative risk factors[36]. In another retrospective comparison of outcomes in 121 patients who underwent TN prior to re-transplantation with 45 who did not, Schleicher et al[19] showed TN led to increased panel reactive antibody (PRA) levels prior to re-transplantation, increased rates of primary non function (PNF, P = 0.05) and acute rejection (P = 0.04). Overall graft survival after re-transplantation was significantly worse in those who had preliminary TN compared with those who had not (P = 0.03). On the multivariate analysis, pre transplant graft nephrectomy and PRA > 70% were independent and significant risk factors associated with graft loss after kidney re-transplantation[19]. However, Ahmad et al[15] undertook a retrospective analytical study of 89 patients with kidney re-transplants to determine the effect of removal of a failed kidney allograft on the outcome of subsequent transplant and came to a different conclusion. They reported no significant difference in the two groups (68 had a TN while 21 had retained failed grafts) in the PRA level at the time of re-transplantation. Multivariate analysis showed that PRA level significantly influenced graft survival independent of nephrectomy (P = 0.04) and concluded that nephrectomy of a failed allograft does not seem to significantly influence the survival of a subsequent graft.

LIMITATIONS OF STUDIES

No randomised controlled trials of the role of TN were encountered during the literature search for this article. The series reported by Ayus et al[40], one of the largest nationally representative sample in recent times, is also subject to important limitations. As pointed out by Rubin et al[64], if the need for TN was triggered by an immune event (for example, humoral rejection), then the procedure may just be a marker of high immune responsiveness and an indication of an adverse outcome with repeated transplantation (selection bias). The study design (observational analysis of administrative data) did not permit a random selection of allograft nephrectomy and represented a potential source of bias, not completely removed by the adjustments such as sensitivity analyses and propensity scores applied by the authors.

Many of the studies were characterized by a retrospective and univariate analysis of small numbers of patients[36,65]. Given the patient population undergoing TN, it is arguable that the reason for the decreased mortality in the group of patients undergoing nephrectomy was that they were a younger, healthier population as compared with those with retained transplants, who may have been considered too high risk to undergo a surgical procedure[40]. The lack of randomization in many studies introduces a selection bias and conclusions drawn from such articles must be applied with caution.

In a retrospective review of 345 patients with failed kidney transplants, 79% of patients ultimately required nephrectomy primarily for clinical symptoms. The finding of significant inflammation on histological examination of the TN specimens in the vast majority of cases does not mean that asymptomatic patients with failed allografts would have similar histopathological features[12].

Studies of the effect of pre-transplant allograft nephrectomy may be comparing patients of different backgrounds. The interval between transplant failure, or TN where applicable, and re-transplantation was not stated in many of the reports. Most patients undergoing TN lost their grafts during the earlier phase post transplantation (< 2 years) and may be predisposed to higher immunogenicity and may also have spent a longer period on dialysis prior to re-transplantation[16,19].

CONCLUSION

A growing number of patients are returning to dialysis after a failed transplant and it is vital to optimize their treatment and to consider the potential role of TN in this subset that have a high risk of morbidity and mortality. TN appears to be beneficial for survival on dialysis but detrimental to the outcome of subsequent transplantation. In case of early graft failure, early nephrectomy and complete interruption of immunosuppressive therapy is recommended. When graft failure occurs later, the graft can be left in place, while immunosuppressive therapy is tapered and eventually stopped. Patients must be followed carefully as, in at least one third of the patients, side-effects will occur necessitating subsequent TN.

The role of nephrectomy in the management of dialysis treated transplant failure patients and the implications of nephrectomy for repeat transplantation should be further studied in prospective studies. Such a study is likely to clarify the indications for TN. In addition to testing various immunosuppression attrition rates, it may be possible to discern a protocol that minimizes drug exposure while leading to reduced nephrectomy rates after returning to dialysis. Pending such a randomised controlled trial, it is prudent to remove failed symptomatic allografts and all grafts failing within 3 mo of transplantation, monitor inflammatory markers in patients with retained failed allografts and remove in the event of a significant change in levels. It is important to select an appropriate technique of TN to minimise blood loss and other complications.