Graft dysfunction is a problematic complication of kidney transplantation. The diagnosis and treatment varies by the time period after transplantation. This chapter reviews the major etiologies of graft dysfunction and diagnostic measures. Acute rejection, calcineurin inhibitor nephrotoxicity and BK virus are reviewed in detail.
KeywordsAcute rejection, BK nephropathy, Chronic allograft injury, Delayed graft function
Delayed Graft Function, 605
Differential Diagnosis, 606
Prediction and Prevention of Delayed Graft Function, 607
Management of Delayed Graft Function, 608
Diagnostic Studies in Persistent Oliguria or Anuria, 608
Other Causes of Graft Dysfunction During the First Week After Transplantation, 609
Nonimmunological Causes, 609
Long-Term Impact of Immediate Graft Dysfunction, 609
Graft Dysfunction in the Early Posttransplant Period, 610
Acute Rejection, 610
Types of Acute Rejection, 610
Acute T-Cell–Mediated Rejection, 611
Acute Antibody-Mediated Rejection, 611
Noninvasive Diagnostic Biomarkers, 611
Calcineurin Inhibitor-Mediated Nephrotoxicity, 612
Histological Features of Calcineurin Inhibitor-Mediated Nephrotoxicity, 612
Thrombotic Microangiopathy, 612
Vascular Complications, 613
Renal Artery Stenosis, 613
Allograft Thrombosis, 614
Ureteral Obstruction, 614
Perinephric Fluid Collections, 614
Late Graft Dysfunction, 614
Differential Diagnosis of Chronic Allograft Injury, 615
Antigen-Dependent Causes of Late Graft Loss, 615
Prior Sensitization, 617
Alloantigen-Independent Factors, 617
Donor Age, 617
Chronic CNI Nephrotoxicity, 617
Histopathological Features of Chronic Graft Dysfunction, 617
BK Nephropathy, 619
Recurrent Diseases, 620
The authors would like to thank Phuong-Thu T. Pham, MD, Cynthia C. Nast, MD, Phuong-Chi T. Pham, MD, and Gabriel Danovitch, MD, for their contributions to this chapter in the previous edition.
Despite being a well-established procedure, kidney transplantation is associated with risk for complications. Depending on the period since transplant, the etiology and management of graft dysfunction will vary. Hence, the differential diagnosis is best approached by considering the different posttransplant periods. In this chapter, graft dysfunction will be discussed in three phases:
Delayed graft function (DGF), occurring in the immediate posttransplant period
Early graft dysfunction, occurring within the first 3 posttransplant months
Late graft dysfunction, occurring beyond the first 3 months posttransplant
Delayed Graft Function
The term delayed graft function (DGF) has been used to describe marginally functioning grafts that take several days to weeks to recover function. DGF must be distinguished from primary nonfunction, where the kidney allografts never function and allograft nephrectomy is usually indicated. Various indices, including urine volume, dialysis requirement, and serum creatinine (SCr), have been used to define DGF.
From the standpoint of registry reporting, DGF is defined by the requirement for dialysis in the first posttransplant week. Using the need for dialysis alone to define DGF, however, may lead to underdiagnosis, particularly if there is some residual native kidney function. In studies evaluating the causes and management of DGF, a modified definition of “the need for more than one dialysis session” is sometimes applied to take into account the need for a single postoperative dialysis session for the management of hyperkalemia, or diuretic-resistant volume overload. An elevated SCr concentration of greater than 400 mmol/L (>4.5 mg/dL) 1 week after transplantation has been suggested to be a more sensitive and specific measure of DGF.
A urine output greater than 20 mL/kg/day in the immediate postoperative period is generally a good clinical indicator of adequate kidney function; however, it is limited in cases where large urine volume is still produced by the native kidneys. Alternatively, no increase in urine output postoperatively does not necessarily indicate DGF. The term slow graft function (SGF) is sometimes used to describe nonoliguric patients who usually do not require dialysis but who experience a delayed fall in SCr.
Although most kidneys with DGF when biopsied demonstrate findings consistent with acute tubular necrosis (ATN), it should be noted that these terms are not synonymous. Unless an allograft biopsy is performed, posttransplant ATN should be a diagnosis of exclusion. The differential diagnosis of DGF is shown in Box 39.1 . Similar to nontransplant acute kidney injury, a systematic approach to the evaluation of DGF may be divided into prerenal, intrinsic, and postrenal causes. Although uncommon, vascular causes of DGF must be excluded, particularly in the early postoperative period.
Nephrotoxic drugs and high CNI levels
Arterial or venous thrombosis
Renal artery stenosis
Acute tubular necrosis
Accelerated or acute rejection
Recurrence of primary glomerular disease (particularly FSGS)
Perinephric fluid collection (lymphocele, urine leak, hematoma)
Intrinsic (blood clots, technical issues, ureteral slough)
Extrinsic (ureteral kinking)
Benign prostatic hypertrophy
CNI, Calcineurin inhibitor; FSGS , focal segmental glomerular sclerosis.
Prerenal causes such as intravascular volume depletion or a significant hypotension are usually suggested by a careful review of patients’ preoperative history and intraoperative report. Knowing the dialysis dry weight and preoperative weight may aid in assessing recipient volume status in the immediate postoperative period.
Because calcineurin inhibitors (CNIs) cyclosporine and tacrolimus cause a dose-related reversible afferent arteriolar vasoconstriction that may temporarily worsen kidney function, it is possible, though unproven, that these agents could delay recovery of allograft function in the immediate posttransplant period as well. Target of rapamycin (TOR) inhibitors have been found to delay recovery from DGF and are seldom used in the early posttransplant setting.
Intrinsic renal causes of DGF typically include ATN, acute rejection, infection, thrombotic microangiopathy, and recurrence of glomerular diseases affecting the native kidneys. The incidence of ATN varies widely among centers and has been reported to occur in 20% to 25% of patients. The difference in reported incidence may in part be related to variation in use of marginal donor organs across centers. Both donor and recipient factors are important determinants of early allograft dysfunction.
ATN found in the posttransplant setting is an ischemic injury that may be synergistically enhanced by both immunological and nephrotoxic insults. All transplanted kidneys are subjected to injury stressors at various stages from donor death to organ procurement, surgical reanastomosis, and in the early postoperative period. The kidney encounters ischemia during organ procurement and storage and then reperfusion when transplanted; this process upregulates endothelial human leukocyte antigen (HLA) expression, causing the release of chemokines, cytokines, and adhesion molecules within the allograft, which leads to the activation and recruitment of T lymphocytes. In a multicenter cohort study of deceased-donor kidney transplants procured from March 2010 through April 2012 by participating organ procurement organizations (OPOs), it was found that preimplant biopsy-reported ATN was modestly associated with the development of DGF only in donation after cardiac death (DCD) kidneys. Their findings suggested that there was substantial variability in pathology reports and that acute structural injury was often underreported.
Although ischemic injury is a major risk factor for the development of posttransplant ATN and DGF, several lines of evidence suggest that immunological factors contribute as well. DGF is more prevalent in retransplant compared with primary transplant recipients, particularly in highly sensitized patients.
Some data suggest that ischemic injury may contribute to the upregulation of inflammatory cytokines and increase expression of class I and II major histocompatibility complex antigens, thus increasing the immunogenicity of the transplanted kidney. Nitric oxide produced by the inducible nitric oxide synthase enzymes in response to ischemic cell injury has been suggested as a link between ischemic reperfusion injury and graft rejection. Renal epithelial regeneration mediated by growth factors and cytokines, such as epidermal growth factor and transforming growth factor β, after ischemic damage may culminate in low-grade inflammation and fibrosis observed with chronic rejection.
The stress of donor brain death itself likely has a similar effect to ATN on the risk for allograft rejection. In one animal study, brain death was found to be associated with upregulation of macrophage (interleukin-1 [IL-1], IL-6, and tumor necrosis factor α [TNFα]) and T-cell–associated products (IL-2 and interferon-α [IF-α]) in peripheral organs, rendering them more susceptible to host inflammatory and immunological responses.
Immunohistochemical analysis of pretransplant living and deceased donor biopsy specimens demonstrated increased E-selectin expression and interstitial leukocyte accumulation in deceased compared with living donor kidneys, suggesting that brain death initiates an inflammatory reaction in the human kidney. Highly matched kidneys have been suggested to be less susceptible to the harmful effects of DGF, presumably because ATN exposes the mismatched kidney to a more aggressive immune attack.
A number of studies have found that developing posttransplant ATN does not have long-term consequences provided that rejection does not occur, lending support to the theory that it is the immunological consequences of ATN that are responsible for its prognostic significance.
Studies also suggest that there is a sex-based difference in DGF risk, with the female hormonal milieu being protective. Aufhauser et al. demonstrated in a rodent transplant model that female recipient sex was protective against DGF, regardless of donor sex, and supplemental estrogen given to female mice was further protective. Analyses of the United Network of Organ Sharing (UNOS) data set found greater risks of DGF in male transplant recipients, and this difference was not explained by variation in body weight between donor and recipient or nephron dosing. Trials examining supplemental estrogen as a means of reducing DGF rates are underway.
Histologically, allograft and native kidney ATN are similar. Tubular epithelial cells show necrosis, often with sloughed, degenerated, or apoptotic epithelial cells in the tubular lumina. Proximal cell brush border staining of proximal tubular epithelium is focally absent with flattening of tubular cells, and there may be regeneration in the form of mitotic figures ( Fig. 39.1 ). Calcium oxalate deposition in tubular lumina is associated with early graft dysfunction and ATN prolongation. The interstitium is variably edematous, with a minimal or patchy interstitial lymphocytic infiltrate; however, there is no associated tubulitis or glomerular or vascular changes.
Prediction and Prevention of Delayed Graft Function
The uncontrolled circumstances surrounding brain death, as well as the complex deceased donor organ procurement process, inevitably result in varying degrees of ischemic damage that adversely affect allograft function.
Preexisting donor factors are important predictors of early and late graft function. Kidneys from older donors have a higher incidence of ATN, attributed to the diminished capacity of the aging vasculature to vasodilate adequately to protect the kidney from anoxic damage. For this reason, these organs were historically rejected by most centers. However, the critical shortage of organs has resulted in increased use of kidneys from more marginal donors. Until recently, these were labeled “expanded criteria donor” kidneys and were associated with a 70% greater risk for allograft failure compared with kidneys from young, healthy donors. This terminology has since been supplanted by the Kidney Donor Profile Index (KDPI), which incorporates into its assessment of donor “quality” the characteristics of age, height, weight, race/ethnicity, history of hypertension, history of diabetes, hepatitis C status, cause of death, and brain versus cardiac death. KDPI helps understand how long a deceased donor kidney is expected to function relative to all the kidneys recovered in the United States during the last year. Lower KDPI scores are associated with longer estimated graft function, whereas higher KDPI scores are associated with shorter estimated graft function.
To optimize allograft outcomes using these “marginal” kidneys, UNOS has implemented a system to expedite high KDPI kidney placement. Despite the efforts to decrease cold ischemia time, the use of these kidneys inevitably increases the incidence of posttransplantation ATN. Similarly, the use of DCD kidneys, which are exposed to increased warm ischemia in addition to cold ischemia, has resulted in an increased incidence of DGF. In a single-center study of 456 renal transplants, DGF occurred in 17% of heart-beating deceased donor kidneys compared with 95% of DCD kidneys. Nonetheless, graft survival in the DCD recipients with DGF was significantly better compared with recipients of a heart-beating donor kidney transplant that developed DGF at 3 years (84% vs. 73%, respectively; P < 0.05) and 6 years (84% vs. 62%, respectively). Although this appears to be paradoxical, it has been proposed that brainstem death causes cytokine release and inflammatory reactions, events that do not occur in donation after cardiac death. In addition, surgeons may be more cautious in accepting organs that have experienced additional “insults” and therefore may select only the “best” DCD organs for transplantation. Analysis of data from the United States Renal Data System revealed that recipients of DCD kidneys experienced nearly twice the incidence of DGF compared with donation after brain death (DBD) kidneys (42.3% vs. 23.3%, respectively). Nonetheless, DCD kidneys had comparable allograft survival to kidneys from DBD donors at 6 years of follow-up (73.2% vs. 72.5%, respectively; P = NS). Significant risk factors for allograft loss for DCD kidney recipients included retransplant, DGF, donor age older than 35 years, and head trauma as a cause of death.
Because the potential for ischemic injury to kidney transplants begins with procurement, the goal of donor management during this period is to maintain adequate organ perfusion before rapid cooling and flushing of the kidneys to minimize warm ischemia. Warm ischemia time refers to the period between circulatory arrest and the commencement of cold storage. Ischemia at body temperature can be tolerated for only a few minutes, after which irreversible injury begins to occur and the organ becomes nonviable within 30 minutes. Cold ischemia time refers to the period from initiation of cold storage or machine perfusion (MP) until the transplanted kidney is reperfused in vivo. For the purpose of transplantation, anaerobic metabolism can maintain renal cellular energy requirements for up to 48 hours, provided the organ is cooled to about 4°C with an appropriate preservation solution. Increasing warm and cold ischemic time both can result in higher rates of DGF and a progressive decline in graft survival. Ideally, kidneys are transplanted without significant warm ischemia and with cold ischemia time less than 24 hours. In DCD donation, rapid institution of donor cooling to reduce warm ischemia is particularly vital because of the absence of blood circulation. Hypothermia-induced reduction in tissue metabolism (for every 10°C of organ cooling, metabolism is decreased by approximately 50%) alleviates ischemic injury.
Pulsatile hypothermic machine perfusion, first developed by Belzer in the late 1960s, was used by many centers to preserve kidneys until EuroCollins preservation solution was introduced in 1969. There has been renewed interest in using MP because of its reported beneficial effects in lowering the DGF incidence and improvement in allograft function. A metaanalysis on the effectiveness of MP versus cold storage found that MP led to a 20% reduction in the incidence of DGF in both DBD and DCD kidneys compared with cold storage, with no significant difference in 1-year graft survival between the two preservation methods. Nonetheless, MP may permit identification of kidneys at high risk for primary nonfunction, thus sparing potential recipients the morbidity associated with the transplant operation and the development of allosensitization. The perfusion index (PFI) is used to assess organ function while being perfused.
Because ATN may render allografts more susceptible to immunological injury, the use of lymphodepleting antibody induction may have a beneficial effect on DGF rates. Intraoperative thymoglobulin administration has been reported to be associated with significantly less DGF and better early allograft function compared with administration initiated postoperatively, attributed to modulation and attenuation of ischemia-reperfusion injury. A small molecule mimetic of hepatocyte growth factor/scatter factor (HGF/SF) whose activity is expected to preserve tissue viability and attenuate dysfunction in the setting of organ injury is being tested in kidney transplant recipients to decrease the rates of DGF. HGF/SF is a renotrophic factor that exerts significant cytoprotective, antifibrotic, and antiapoptotic effects in renal epithelial cells. The molecule being tested was found to be efficacious in animal models of ischemia-reperfusion injury as well as toxin-induced injury.
Management of Delayed Graft Function
Most patients with DGF are oliguric or anuric. As discussed earlier, information about both native kidney urine output and donor kidney characteristics is essential. For example, if postoperative oliguria occurs immediately after living donor kidney transplantation, surgical complications such as arterial or venous thrombosis must be immediately considered. In contrast, when a patient receives a deceased donor kidney from a high KDPI donor, DGF may be anticipated. The mate kidney from a deceased donor often behaves in a similar manner, and information on its function can be useful.
Oliguria in the peritransplant period typically refers to a urine output of less than 50 mL/h. Evaluation of oliguria should commence with assessment of volume status, fluid balance, and Foley catheter patency. If clots are present, the catheter should be removed while gentle suction is applied in an attempt to capture the clot. Thereafter, replacement with a larger catheter may be required. For patients with hypervolemia, high-dose furosemide may be given intravenously. If the patient is judged to be hypovolemic, or if a clinical assessment cannot be made, a trial of isotonic saline infusion may be given, with or without subsequent administration of furosemide as dictated by the patient’s response to saline infusion alone.
Indications for dialysis in the transplant recipient with DGF are essentially the same as in any patient with postoperative kidney dysfunction. Hyperkalemia must be monitored closely and treated aggressively. It is usually safest to dialyze patient if the potassium level is greater than 5.5 mg/dL. Temporizing measures, such as intravenous calcium and glucose with insulin, can be used but do not obviate the need for dialysis. Sodium polystyrene sulfonate (Kayexalate) should not be administered perioperatively because its use has been associated with gut ischemia.
Patients with DGF often become volume overloaded in the early posttransplant period. Ultrafiltration with or without dialysis may be required. When dialyzing patients who have DGF, care must be taken to avoid hypotension, which may plausibly perpetuate graft dysfunction. In patients with established DGF, dialysis requirements should be assessed daily until graft function improves.
Diagnostic Studies in Persistent Oliguria or Anuria
Failure to respond to volume challenge and furosemide administration warrants further evaluation to determine the cause of the early posttransplant oligo/anuric state.
Imaging studies are used to confirm the presence of blood flow to the graft and the absence of a urine leak or obstruction. Blood flow studies are performed scintigraphically or by Doppler ultrasound. The typical scintigraphic finding in ATN is relatively good flow to the graft but with poor excretion. If the Doppler reveals no demonstrable blood flow, a prompt surgical reexploration is necessary to evaluate the allograft vasculature. Kidneys without arterial circulation are seldom salvageable and are usually removed during reexploration. If imaging studies indicate adequate blood flow to the allograft, the possibility of ureteral obstruction or urinary leak needs to be considered and can be evaluated by the same modalities. In the first 24 hours after transplantation, as long as the Foley catheter has been providing good bladder drainage, the obstruction or leak is almost always at the ureterovesical junction and represents a technical problem that requires surgical correction.
Other Causes of Graft Dysfunction During the First Week After Transplantation
Hyperacute rejection occurs immediately after transplantation and is due to preformed antibodies, usually to blood group antigens or donor HLAs. The rejection occurs after an amnestic response where a critical level of antibodies is produced and results in an irreversible vascular damage. Hyperacute rejection may be evident at reperfusion or may be “hidden,” manifesting as primary nonfunction of the kidney allograft. Patients are usually anuric or oliguric and might have associated graft tenderness. The renal scan shows little or no uptake, a finding that differentiates this cause of graft dysfunction from ATN. There may be evidence of intravascular coagulation. Prompt surgical exploration of the allograft is often indicated, and often an intraoperative biopsy is performed to determine viability.
In hyperacute rejection, preformed antibodies activate the complement cascade. Nephrectomy specimens have arterial and glomerular thrombi, which often contain neutrophils. The interstitium is edematous and variable parenchymal necrosis or infarction is observed, depending on the length of time from thrombosis to nephrectomy. There is no significant vascular or tubulointerstitial inflammation. Immunofluorescence microscopy reveals fibrin within the intravascular thrombi, and immunoglobulin M (IgM), IgG, C3, and fibrin may be found in arterial and capillary lumina or lining or within the intimas. Most allografts are nonviable and need to be removed. With advances in crossmatching techniques incorporating flow cytometry and Luminex bead-based assays, hyperacute rejection is very rare.
Accelerated Acute Rejection
Accelerated acute rejection or delayed hyperacute rejection typically occurs within 24 hours to a few days after transplantation and may involve both antibody-mediated and cellular immune mechanisms. Accelerated acute rejection probably represents a delayed amnestic response to prior sensitization. It is most commonly observed after ABO- and HLA-incompatible transplants and may also be seen after donor-specific transfusions in recipients of living-donor transplant as a result of a primed T-cell response. HLA sensitization through repeat transplants, multiple pregnancies, or multiple transfusions are well-substantiated risk factors for hyperacute or accelerated acute rejection.
Early T-Cell Acute Rejection
Early T-cell rejection, with a typical interstitial mononuclear cell infiltrate and tubulitis, including CD4 and CD8 T cells, can rarely be detected in the latter part of the first week after transplantation, although it typically occurs somewhat later. It may be difficult to recognize clinically in patients with DGF. If suspected, there should be a low threshold to perform an allograft biopsy. The prognosis for long-term function of these grafts is poor, although adequate function may be achieved, if the ATN reverses and the rejection responds to intensification of immunosuppression.
Nonimmunological causes of DGF (other than ATN) may occur in the first posttransplant week or any time thereafter and are discussed under Graft Dysfunction in the Early Posttransplant Period.
Long-Term Impact of Immediate Graft Dysfunction
Studies on the impact of DGF on long-term graft function have yielded conflicting results. Transplant registry data revealed that DGF reduced 1-year graft survival from 91% to 75% ( P < 0.001) and graft half-life from 12.9 to 8 years, independent of early acute rejection episodes. This deleterious effect of DGF remained significant after adjusting for discharge SCr less than 2.5 mg/dL as well. DGF has also been found to abrogate the survival advantage of well-matched kidneys (0 to 1 mismatch) over those with lesser degrees of matching (5 to 6 mismatches). Some investigators have found that DGF, when combined with rejection, had an additive negative effect on allograft survival, whereas others suggested that DGF is deleterious to graft outcome only when associated with reduced renal mass and hyperfiltration injury. A metaanalysis of 33 DGF studies found that DGF was associated with a 41% increased risk for graft loss (risk ratio [RR] 1.41, 95% confidence interval [CI], 1.27 to 1.56) with no clear deleterious effect on patient survival (RR 1.14, 95% CI, 0.94 to 1.39). Patients who experienced DGF had higher posttransplant SCr levels and an increased risk for rejection (RR 1.38, 95% CI, 1.29 to 1.47). The harmful effect of DGF may also be more pronounced when marginal donor kidneys are used.
There are several explanations for the contradictory findings in these studies. First, most studies did not incorporate transplant biopsies and DGF was attributed to ATN; second, there is no universally defined criteria for DGF; third, donor and recipient characteristics were quite variable across studies.
In addition, the duration of DGF has also been suggested to be impactful on outcomes. For example, in one study, 5-year graft survival for patients with immediate function and DGF lasting 8 days or less was similar (89%, 85%) and higher than in patients with DGF lasting for more than 8 days (50%). Given the impact of DGF on morbidity and health resource use, trials are underway to prevent or attenuate this detrimental complication.
Graft Dysfunction in the Early Posttransplant Period
The early posttransplant period typically spans the period after discharge from the transplant hospitalization until the end of the third month, by when most patients have achieved stable graft function and immunosuppressive regimen. Although a somewhat arbitrary differentiation, most acute rejection episodes and surgery-related complications occur during this period as well.
By the second week, DGF caused by ATN typically begins to improve, although some patients remain dialysis dependent for several weeks. In dialysis-independent patients, SCr measurement is the main indicator of allograft function and outcome predictor and should be evaluated in the context of the donor type, size mismatch between donor and recipient, and function of the mate kidney if applicable. The SCr level reached by the second week may predict long-term graft function; a baseline level greater than 2 mg/dL warrants further investigation.
Elevations in SCr greater than 25% from baseline almost always indicate a clinically significant and potentially graft-endangering event, and it is advisable to repeat within 48 hours. The clinical algorithm in approaching SCr elevations (or failure to reach a low baseline value) is similar, in principle, to that used in the nontransplant setting: “prerenal,” “renal,” and “postrenal” causes need to be considered. In the early posttransplantation period, acute rejection is the most important potentially reversible threat to graft function. In addition, anatomical or surgical problems must also be considered before medical diagnoses are sought to explain deteriorating graft function.
Acute rejection conventionally describes the cellular and/or humoral immune response to the transplant that produces enough inflammation and destruction to cause detectable allograft dysfunction. The majority of patients with acute rejection are asymptomatic, and clinical suspicion arises on the basis of SCr elevation. In symptomatic patients a concomitant search for alternative causes of graft dysfunction is warranted. For example, a tender, swollen allograft with a rising SCr concentration and fever is a typical presentation of acute transplant pyelonephritis. For this reason, a thorough history, physical examination, standard laboratory tests, and evaluation for infection must be obtained before the diagnosis and treatment of rejection. Because of the interplay between rejection and viral infections, testing for cytomegalovirus (CMV) and BK virus via polymerase chain reaction (PCR) is also a reasonable part of the evaluation of allograft dysfunction.
Although SCr can define the occurrence of allograft dysfunction and is highly convenient, it is insensitive in detecting subclinical rejection. Approximately 15% to 80% of protocol biopsy specimens acquired in the first 3 to 6 months posttransplant in patients with well-functioning grafts have been reported to have histopathological lesions of acute rejection, although this varies markedly depending on the induction and maintenance immunosuppression regimens used. Although it may be a risk for chronic rejection and progressive allograft failure, treatment of subclinical rejection has not been consistently found to prevent late allograft deterioration or fibrosis in studies with serial biopsies ; controlled trials in this area are lacking.
Although nonspecific for demonstrating rejection, imaging studies are performed to exclude alternative causes of acute allograft dysfunction. In mild acute rejection episodes, imaging studies may be normal. The resistive index can be elevated, though this can be observed in other causes of graft dysfunction that result in increased intrarenal vascular resistance.
Given the current radiological limitations, allograft biopsy, although invasive, remains the gold standard for differentiating acute rejection from other causes of acute allograft function.
Biopsy evaluation for acute rejection should be performed in unscarred portions of the renal cortex. Rejection can present focally and may be missed on biopsy; therefore at least two cores with two arteries and seven slides of tissue are required for adequate assessment.
The timing and frequency of kidney biopsies vary among centers. We advocate for transplant biopsy to be performed in any patient with elevated SCr for whom an explanation is not apparent, to determine the presence and severity of rejection and its subsequent treatment.
Types of Acute Rejection
On the basis of the underlying immunopathogenic mechanisms (discussed in Chapter 35 ), acute rejection can be divided into T-cell–mediated and antibody-mediated rejection. The vast majority of acute rejection episodes are T-cell mediated, whereas 12% to 37% of all acute rejection episodes have a humoral component. It should be noted that the histopathological findings of acute T-cell rejection and acute antibody-mediated rejection (ABMR) may occur in a renal biopsy simultaneously. The diagnostic criteria for cellular and ABMR are set forth in the Banff classification schema, which are regularly updated.
Acute T-Cell–Mediated Rejection
The 2015 Banff update to the classification of acute T-cell–mediated rejection (TCMR) recognizes category 3, borderline changes suspicious for acute TCMR, and category 4, acute TCMR. “Borderline” changes consist of foci of tubulitis or interstitial inflammation present on biopsy but not meeting category 4 criteria. Depending on the clinical scenario and indication for biopsy, clinicians may choose not to treat borderline rejections, augment maintenance immunosuppression, or give intravenous steroids. In one study in which researchers performed follow-up biopsies for 24 patients with “borderline” TCMR findings on the initial specimen, nearly 80% had mild or moderate cellular rejection on repeat sampling, calling into question the supposedly benign nature of the finding.
Acute TCMR (category 4) can be further subdivided on the basis of the severity of inflammation and the involvement of blood vessels ( Fig. 39.2 ). Grades Ia and Ib acute TCMR both require an interstitial inflammatory infiltrate that involves more than 25% of the cortex with tubulitis (>4 lymphocytes/tubular cross section for Ia and >10 lymphocytes for Ib). Some practitioners treat with steroids alone, though some use lymphocyte-depleting therapy as well. Grade II rejections are characterized by the presence of intimal arteritis (>25% luminal area compromised for IIb) and do not necessarily require interstitial inflammation or tubulitis; these rejections are treated with lymphodepleting antibodies. Grade III rejections are less common and display fibrinoid necrosis changes along with transmural arteritis; these severe rejections require prompt treatment with lymphocyte depleting therapies but often portend irreparable damage to the allograft.
Acute Antibody-Mediated Rejection
The diagnosis of acute ABMR, according to the 2013 update to the Banff criteria, requires evidence of renal parenchymal injury (microvascular inflammation, arteritis, thrombotic microangiopathy, or acute tubular injury), antibody interaction with the vascular endothelium (C4d staining in the peritubular capillaries, microvascular inflammation, or increased expression of endothelial injury gene transcripts), and the presence of donor-specific antibodies (DSAs) ( Fig. 39.3 ). These may occur with or without features of acute TCMR. Furthermore, C4d is not always pathological and can be identified in allograft biopsy specimens lacking morphological evidence of rejection, such as immune accommodation in ABO incompatible transplants.
Noninvasive Diagnostic Biomarkers
Over the last decade, gene analysis using molecular biology techniques such as reverse transcription–PCR has been studied as a noninvasive tool in the diagnosis of acute rejection. Elevated peripheral blood or urine levels of perforin, granzyme B, Fas ligand, and serpin proteinase inhibitor 9 have variably been reported to indicate the presence of ongoing acute TCMR. It has also been suggested that measurement of urine perforin messenger RNA (mRNA) and granzyme B mRNA may offer a noninvasive means of diagnosing acute TCMR, with a sensitivity of 79% to 83% and a specificity of 77% to 83% in one study. In one single-center study, perforin and granzyme B had sensitivities of 50% and specificities of 95% in predicting rejection when a cutoff value of 140 was used.
There have been other attempts at developing noninvasive markers of rejection, including measurement of donor-derived cell-free DNA (dd-cfDNA). Cellular damage leads to the release into the recipient’s circulation of donor-derived DNA, termed cell-free DNA . This DNA can be amplified by digital droplet PCR or massive parallel shotgun sequencing. A study of 102 kidney transplant recipients correlated plasma dd-cfDNA results with allograft biopsy specimens; a dd-cfDNA greater than 1% was predictive of acute rejection on biopsy (area under the curve [AUC] 0.74, 95% CI, 0.61 to 0.86). Although there was a trend toward increasing dd-cfDNA levels with more severe rejections by Banff classification, the study was underpowered to demonstrate this.
CD4 + CD25 + FoxP3 regulatory T cells (Treg) are involved in the maintenance of immune homeostasis and tolerance to self and nonself antigens. A single-center study in kidney recipients with DGF suggested that measurement of FOXP3 gene (X-linked forkhead/winged helix transcription factor) in peripheral blood leukocytes and urinary cells provides an accurate marker of acute rejection with a reported sensitivity, specificity, and positive and negative predictive values between 94% and 100%. In a single-center case-control study of stable recipients undergoing tacrolimus dose reduction with ultimate withdrawal of the drug, it was demonstrated that before tacrolimus dose reduction, the CD8 + :Treg ratio was higher in rejectors compared with nonrejectors. Rejectors also had a higher memory CD4 + : Treg ratio. In rejectors, an increase over time was observed in the percentage of naïve T cells in the peripheral blood, with a reciprocal decrease in the percentage of effector T cells. Whether the combination of the memory T cells:Treg ratio and the changes in T-cell subsets over time might prove useful in the detection of acute rejection remains to be studied. The use of immune monitoring as a noninvasive tool in the prediction and diagnosis of acute rejection is a subject of intense, ongoing research. But despite these encouraging preliminary results, noninvasive methods have failed to supplant biopsy in the diagnosis of rejection.
Calcineurin Inhibitor-Mediated Nephrotoxicity
CNIs cyclosporine and tacrolimus are two potent immunosuppressive agents with similar mechanisms of action (see Chapter 38 on immunosuppression) and pathological patterns of nephrotoxicity. CNIs produce a dose-related, reversible, renal vasoconstriction that particularly affects the afferent arteriole, manifesting clinically as a blood-level–dependent elevation in SCr concentration that may be difficult to distinguish from other causes of graft dysfunction. High blood levels of CNIs do not preclude a diagnosis of rejection. In the acute phase, tubular function is intact.
CNI nephrotoxicity may develop at low drug levels, especially in malnourished patients with diminished protein binding where the free drug concentration is increased; some degree of toxicity may be intrinsic to their use. In practice, particularly when SCr elevation is modest, it is reasonable to initially presume that a patient with a very high CNI level probably has nephrotoxicity and that a patient with deteriorating graft function and a very low drug level is likely undergoing rejection. CNI toxicity usually resolves within 24 to 48 hours of dose reduction. Progressive elevation of the plasma creatinine level, even in the face of persistently high drug levels, warrants consideration of rejection.
CNI levels are most commonly measured using mass spectrometry and immunoassays. Although mass spectrometry–based measurements are more accurate, these techniques are not standardized across institutions, leading to interlaboratory variability. Immunoassays also have limitations because they are prone to measuring metabolites in addition to the parent drug and may overestimate CNI levels. It is important for clinicians to be aware of which testing methodology their laboratory uses and encourage patients to be consistent in the laboratory they use, especially in the early posttransplant period when CNI levels are highly variable and therapeutic dosing is most critical.
Histological Features of Calcineurin Inhibitor-Mediated Nephrotoxicity
Cyclosporine and tacrolimus nephrotoxicity have similar appearances in renal allografts. The most common form of acute toxicity is a variant of ATN, with scattered individual necrotic tubular cells, considerable dilatation of tubular lumina, and epithelial cell flattening without extensive loss of brush border staining. The characteristic feature, often not present, is isometric vacuolization of proximal tubular cell cytoplasm, which tends to involve all tubular cells in few tubular profiles. There is mild interstitial edema without significant inflammation or with focal aggregates of inactive lymphocytes often in perivenous locations; tubulitis is absent. There are no specific glomerular abnormalities. Early in the course of toxicity, arterioles have muscular hypertrophy and individual smooth muscle cells may undergo necrosis; in these locations there is accumulation of rounded plasma protein collections (insudates) in the outer aspect of the muscular walls. The juxtaglomerular apparatus are enlarged.
Chronic CNI toxicity is characterized by focal “striped” interstitial fibrosis, thought to be cortical medullary rays with associated tubular atrophy and little interstitial or tubular inflammation. Arteries are normal, but arterioles may have muscular hypertrophy and rounded insudates in the walls, particularly the outer portion ( Fig. 39.4 ). The glomeruli are unremarkable or have mild ischemic change.