SEARCH

SEARCH BY CITATION

Keywords:

  • Acute rejection;
  • acute tubular injury;
  • BK virus;
  • chemokine;
  • kidney;
  • urine

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References

A noninvasive urinary test that diagnoses acute renal allograft dysfunction would benefit renal transplant patients. We aimed to develop a rapid urinary diagnostic test by detecting chemokines. Seventy-three patients with renal allograft dysfunction prompting biopsy and 26 patients with stable graft function were recruited. Urinary levels of CXCR3-binding chemokines, monokine induced by IFN-γ (Mig/CXCL9), IFN-γ-induced protein of 10 kDa (IP-10/CXCL10), and IFN-inducible T-cell chemoattractant (I-TAC/CXCL11), were determined by a particle-based triplex assay. IP-10, Mig and I-TAC were significantly elevated in renal graft recipients with acute rejection, acute tubular injury and BK virus nephritis. Using 100 pg/mL as the threshold level, both IP-10 and Mig had diagnostic value (sensitivity 86.4%; specificity 91.3%) in differentiating acute graft dysfunction from other clinical conditions. In terms of monitoring the response to antirejection therapy, this urinary test is more sensitive and predictive than serum creatinine. These results indicate that this rapid test is clinically useful.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References

Complications following renal transplantation such as acute rejection, acute tubular injury, BK virus nephritis, and chronic rejection are the major hurdles for successful long-term renal transplant function despite the improvement of organ preservation and the use of newer immunosuppressants (1). Diagnosis of these complications is best made on the histology of percutaneous needle biopsies. However, the biopsy procedure is expensive and has potential complications. Sampling errors of biopsy pose an additional challenge, and multiple samples may increase diagnostic accuracy (2,3). Accurate interpretation of the renal transplant biopsy demands the expertise of a pathologist with extensive experience in analyzing biopsy samples for evidence of renal transplant injury. Therefore, a noninvasive test that has high sensitivity and specificity for the diagnosis of renal graft dysfunction, including rejection, and that could be used for monitoring the effect of antirejection therapies would be of considerable value in supplementing the needle biopsy procedure.

Urine is a good sample source for monitoring kidney abnormalities for noninvasive tests. Li et al. (4) reported a reverse transcription-polymerase chain reaction (RT-PCR)-based method for the diagnosis of acute rejection of renal graft by measuring messenger RNA (mRNA) encoding perforin and granzyme B in urinary cells. The sensitivity and specificity of this test are both greater than 80%. Nevertheless, detection of soluble proteins in the urine would be technically simpler and require less stringent conditions for sample collection. We evaluated urinary chemokines to determine their usefulness for the diagnosis of renal allograft dysfunction. Existing data shows that chemokines are involved in at least three aspects of allograft biology. First, restoration of blood flow during the transplant surgery leads to ischemia-reperfusion injury in which chemokines are secreted and recruit leukocytes (5). Second, host responses to infection during immune suppression involve chemokines (5). Third, the inflammatory components of acute and chronic rejection are mediated in part by chemokines (5). Among the chemokines, CXCR3-binding chemokines, including monokine induced by IFN-γ (Mig/CXCL9), IFN-γ-induced protein of 10 kDa (IP-10/CXCL10), and IFN-inducible T-cell chemoattractant (I-TAC/CXCL11), are of particular interest. Activated immune cells that are present in the graft as infiltrating cells secrete these chemokines (6,7). Furthermore, human kidney tubular cells and mesangial cells stimulated in vitro by IFN-γ secrete IP-10 into the culture supernatant (8–10). Elevation of IP-10 mRNA has been observed in mouse kidneys after ischemia and reperfusion (11). Akalin et al. found by using high-density oligoarray technology that Mig mRNA was elevated 17-fold in human renal allograft biopsy samples with acute rejection (12). Segerer et al. studied human renal biopsies by RNase protection assay, and detected increased mRNA expression of IP-10 in biopsy samples with acute rejection (13). Furthermore, expression of the receptor CXCR3 on infiltrating immune cells is also strikingly up-regulated in biopsies with acute graft rejection (14).

In the present study, we took advantage of a recently developed technology, particle-based fluorescent immunoassay, which can simultaneously quantify multiple protein molecules in one reaction (15), to detect CXCR3-binding chemokines in urinary samples collected from recipients of renal allografts. Using this particle based rapid test to simultaneously quantify IP-10, MIG and I-TAC in urinary samples, we found that these chemokines were significantly elevated in urine of renal transplant recipients suffering from acute rejection, acute tubular injury and BK virus nephritis, but not in other causes of graft dysfunction or in the setting of normal graft function.

Participants and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References

Participants

Between July 2001 and June 2003, renal transplant patients were recruited from the Transplant Clinic of the University of Wisconsin Hospital and Clinics. The research protocol was approved by the University of Wisconsin Institutional Review Board, and all patients provided informed consent. On the day of biopsy, urinary samples (50 mL) were collected before biopsy by clean catch from patients who had an elevated serum creatinine (Cr) of 20% or greater above baseline and who were to undergo renal transplant biopsy as a diagnostic procedure. As controls, patients with stable renal transplant function and normal serum Cr were also enlisted. Patients with biopsy-proven acute rejection were hospitalized, treated with antirejection therapy, and donated daily urinary samples until the rejection resolved. Patients with tubular injury, chronic rejection, or borderline rejection were not admitted but rather treated as outpatients. Urinary samples were also collected from 16 healthy nontransplanted individuals as additional controls. The collected samples were centrifuged at 1500 r.p.m. for 10 min. Supernatant of each sample was aliquoted and stored at – 80 °C until use. At the time of the experiments, samples were thawed and evaluated for the levels of IP-10, Mig and I-TAC.

Patients with symptoms and elevated serum Cr underwent renal transplant biopsy, which was used as the diagnostic standard. Acute and chronic rejection was scored according to Banff criteria (16) by an experienced renal transplant pathologist. BK virus infection was diagnosed light microscopically by identifying pleomorphic, enlarged tubular epithelial cell nuclei containing characteristic ‘smudgy’ inclusions. The suspicion of BK virus infection was confirmed in all cases by immunohistochemistry using a polyclonal polyoma virus-reactive antibody.

Quantification of urinary IP-10, Mig and I-TAC

Luminex (Austin, TX) Multi-Analyte Profiling (xMAP) Technology and a Renovar human CXCR3-binding chemokines triplex assay kit (Madison, WI) were used for simultaneous quantification of urinary IP-10, Mig and I-TAC. Luminex xMAP is based on polystyrene particles (microspheres) that are internally labeled with two different fluorophores. When excited by a 635-nm laser, the fluorophores emit light at different wavelengths, 658 and 712 nm. By varying the 658-nm/712-nm emission ratios, these beads can be individually classified by the unique Luminex 100 IS analyzer (Luminex, Austin, USA). A third fluorophore coupled to a reporter molecule allows for quantification of the interaction that has occurred on the microsphere surface (15). The capture antibodies directed, respectively, at IP-10, Mig and I-TAC were separately preconjugated to their corresponding particles following the Luminex coupling protocol. The quantification of CXCR3-binding chemokines was conducted in 96-well flat-bottom plates. Twenty-five µL of mixed IP-10, Mig and I-TAC standards or urinary samples were added to wells containing 25 µL of assay buffer and 25 µL of precoated particles, and incubated on a 3-D rotator (Labline Instrument Inc., Melrose Park, IL) at 60 rounds/min at room temperature (RT) for 60 min. Mixed biotin-labeled detection antibodies directed at IP-10 (BD PharMingen, San Jose, CA), Mig and I-TAC (R & D Systems, Minneapolis, MN) were then added and incubated on the rotator at 60 rounds/min at RT for 60 min before the addition of streptavidin-PE (BD PharMingen). After an additional 30 min of incubation on the rotator at 60 rounds/min at RT, data acquisition and analysis were performed on a Luminex 100 IS analyzer.

Statistical analysis

The levels of urinary IP-10, Mig and I-TAC are expressed as mean value ± standard error (SE). The statistical significance of the findings was assessed by anova using computer software Prism 4 from GraphPad Software (San Diego, CA), and a p-value less than 0.05 was considered significant. The urinary chemokine threshold that gave the maximal sensitivity and specificity for the diagnosis of acute dysfunction of renal allograft was 100 pg/mL. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of the urinary chemokine test was calculated as follows:

sensitivity = number of true positive specimens (TP)/[TP + number of false-negative specimens (FN)]; specificity = number of true negative specimens (TN)/[TN + number of false-positive specimens (FP)]; PPV = TP/(TP + FP), and NPV = TN/(TN + FN)

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References

Ninety-nine renal allograft recipients were recruited and donated 350 urinary samples. Among the patients, 28 were diagnosed as acute rejection, nine as borderline rejection, six as BK virus nephritis, 10 as acute tubular injury, 20 as chronic rejection, and 26 as stable graft function without graft injury. Urinary samples were also collected from 16 healthy nontransplanted individuals.

Levels of urinary chemokines in patients and control individuals

Urinary IP-10, Mig and I-TAC were quantified simultaneously in each urine sample by the Luminex xMAP method. Urinary IP-10, Mig and I-TAC were significantly elevated (p < 0.01) in samples collected from recipients with acute rejection, BK virus nephritis, and acute tubular injury, but not in samples collected from recipients with borderline rejection, chronic rejection, and stable graft function (Figure 1). Furthermore, urinary samples collected from healthy individuals contained very low levels of the measured chemokines (Figure 1).

image

Figure 1. Urinary levels of CXCR3-binding chemokines IP-10, Mig and I-TAC in recipients of renal allografts. Urinary samples were obtained and the chemokine levels were determined by the triplex immunoassay. The levels of all three chemokines in the recipients with acute rejection (AR), acute tubular injury (ATI) or BK virus nephritis (BK VN) were significantly (*p < 0.01) higher than the levels in the recipients with borderline rejection (BR), chronic rejection (CR), stable graft function (SGF), and also significantly (p < 0.01) higher than levels in healthy controls (HCs).

Download figure to PowerPoint

Urinary chemokines as a sensitive and specific indicator of acute renal allograft dysfunction

Following renal transplantation, graft dysfunction may occur owing to treatable etiologies such as acute rejection, acute tubular injury, or BK virus nephritis. Other types of

graft injury occur more insidiously and usually do not require acute intervention, such as chronic rejection and recurrence of the original disease (1). Acute dysfunction refers herein to acute rejection, acute tubular injury and BK virus nephritis.

Levels of urinary IP-10, Mig and I-TAC varied greatly among the individuals investigated, ranging from 0 pg/mL to 2000 pg/mL, which was the highest concentration measurable by the present method. Using 100 pg/mL as the cutoff level for the urinary chemokines that gave the maximal sensitivity and specificity, as presented in Table 1, most of the renal graft recipients with acute rejection, BK virus nephritis and acute tubular injury had higher levels of urinary IP-10 and Mig, while most of the recipients with chronic rejection and stable graft function had lower levels of urinary IP-10 and Mig. For recipients with borderline rejection, four out of nine cases showed higher levels. None of the healthy controls had urinary IP-10, Mig or I-TAC greater than 100 pg/mL. The elevation of urinary IP-10 and Mig was far more prevalent than I-TAC in recipients with acute dysfunction caused by acute rejection, acute tubular injury and BK virus nephritis. As shown in Figure 1 and Table 1, elevation of urinary IP-10 and Mig indicated acute renal injury by one of the three etiologies. To evaluate the value of urinary IP-10 and/or Mig to differentiate the acute dysfunction from chronic rejection and stable graft function, we calculated the sensitivity, specificity, positive predictive value and negative predictive value as presented in Table 2. Both IP-10 and Mig are highly sensitive and specific.

Table 1. Kidney graft recipients or control individuals with urinary chemokine level greater than 100 pg/mL
 Acute rejection (n = 28)Borderline rejection (n = 9)BK virus nephritis (n = 6)Acute tubular injury (n = 10)Chronic rejection (n = 20)Stable graft function (n = 26)Healthy control (n = 16)
IP-1025 (89.3%)4 (44.4%)6 (100%)7 (70.0%)2 (10.0%)2 (7.6%)0 (0%)
Mig21 (75%)3 (33.3%)6 (100%)8 (80.0%)2 (10.0%)1 (3.8%)0 (0%)
I-TAC10 (35.7%)0 (0%)2 (33.3%)3 (30.0%)0 (0%)0 (0%)0 (0%)
IP-10 + Mig25 (89.3%)4 (44.4%)6 (100%)8 (80.0%)4 (20.0%)2 (7.6%)0 (0%)
Table 2. Value of urinary IP-10/Mig in differentiation of recipients with acute dysfunction* from recipients with chronic rejection and stable graft function
 Sensitivity (%)Specificity (%)Positive predictive value (%)Negative predictive value (%)
  1. *Acute dysfunction refers to acute rejection, acute tubular injury and BK virus nephritis.

IP-1086.491.390.587.5
Mig79.593.592.182.7
IP-10 and Mig88.687.086.688.9

Urinary chemokine levels were compared with the classic renal function indicator, serum Cr (17). When urinary IP-10 was used in the analysis, most of the recipients with acute dysfunction had increased urinary IP-10 and serum Cr. Recipients with chronic rejection had increased Cr, but not IP-10. Recipients with stable graft function had low IP-10 and Cr levels.

Decline of urinary chemokines after antirejection therapy

Among the 28 patients with acute rejection, daily urinary samples were collected from 24 of them during hospitalization for antirejection therapy. As presented in Figure 2, the urinary IP-10 and Mig tended to decline with the initiation of antirejection therapy on day 1, and most of the recipients reached a level less than 100 pg/mL in their last collected urinary sample.

image

Figure 2. Decline of urinary IP-10 and Mig in recipients with acute rejection after antirejection therapy. Urine samples were collected once daily from hospitalized patients receiving antirejection therapy. Day 1 is the time that acute rejection was diagnosed and antirejection therapy was initiated. Each line represents data for one patient.

Download figure to PowerPoint

Serum Cr is an important parameter used to judge the effectiveness of antirejection therapy (17). Therefore, we compared the levels of serum Cr and urinary chemokines in the recipients who received antirejection therapy. As represented in Figure 3, the serum Cr in the recipients who were hospitalized for 3–14 days could be classified into two patterns: started high and remained high for several days before declining (represented by UCR2, UCR5, UCR23, and UCR12), or started at the lower range of the abnormal level and maintained that level (represented by UCR20 and UCR22) during the hospitalization. In contrast to the Cr kinetics, urinary IP-10 declined in all patients, and this decrease started 2–5 days earlier than the serum Cr (Figure 3).

image

Figure 3. Urinary IP-10 (▴) declined several days earlier than serum creatinine (▪) in acute rejection patients receiving antirejection therapy. Urinary and serum samples were collected once daily from each of the patients, and urinary IP-10 and serum Cr were separately determined. UCR2, UCR5, UCR23, and UCR 12 are representative patients that had elevated serum Cr initially, which declined with antirejection therapy. UCR20 and UCR22 are representative patients that had elevated serum Cr initially, but Cr did not decline much with antirejection therapy during the hospitalization period.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References

It has been shown in previous studies that the mRNA expression of CXCR3-binding chemokines increases drastically in kidney biopsies during acute rejection (12,13). Elevation of IP-10 mRNA has also been observed in mouse kidneys after ischemia and reperfusion (11). Using a highly sensitive and specific triplex urinary assay, we found elevation of urinary IP-10, Mig and I-TAC in recipients of renal grafts with acute rejection, acute tubular injury and BK virus nephritis. Urinary chemokines were either negative or at very low levels in recipients with chronic rejection, stable graft function, and healthy individuals. Combined with previous evidence showing that tubular epithelial cells are able to secrete a large amount of IP-10 (8–10), our study suggests that the elevation of these urinary chemokines indicates acute damage or an acute inflammatory response of the tubular cells. As tubular injury is a common feature of acute rejection, acute tubular injury, and BK virus nephritis, we hypothesize that acute tubular injury is responsible for elevating these chemokines, although infiltrating immune cells may contribute to the elevation of the urinary chemokines (6,7).

The present urinary chemokine test has the advantage of being a completely noninvasive method for the diagnosis of acute renal allograft dysfunction. It could be used to complement allograft biopsy and serum Cr and to monitor response to therapy. Specifically, it could be used as a reference parameter in deciding whether and when a biopsy should be taken. Combining serum Cr with the urinary chemokine test distinguishes acute dysfunction of the renal allograft, which is a medical emergency and needs urgent treatment, from nonacute elevations of creatinine and dysfunction. For patients with elevated Cr and urinary IP-10/Mig, a biopsy should be immediately taken and an accurate diagnosis should be made before the initiation of therapy. However, in patients who have concurrent elevation of serum Cr and urinary chemokines, if a BK virus test on urinary cells either by microscopic observation or polymerase chain reaction is also positive (18), the chance of BK virus nephritis is very high, and a biopsy may be avoided. In the case of increased serum Cr but a normal level of urinary chemokines, this elevation of serum Cr is likely owing to chronic insidious damage to the renal graft. A biopsy may be delayed or even avoided. Urine chemokine levels may be useful in patients whose biopsy reveals borderline rejection. In our present study, four out of nine patients diagnosed as borderline rejection had an increased urinary level of IP-10/Mig. As documented previously, elevation of urinary IP-10/Mig reflects active damage or inflammation. Therefore, these four patients who had a diagnosis not warranting extra immunosuppression may prove to need antirejection therapy. The urine test may also help distinguish active low-grade damage in patients having dormant infiltrating immune cells. Indeed, many biopsies from renal grafts reveal numerous infiltrating immune cells. These patients usually have a normal serum Cr. This insidious infiltration of immune cells is currently regarded either as harmless (19) or as subclinical rejection (20). Although we do not have data to support that urinary chemokines differentiate these two clinical conditions, given the high sensitivity of the test in indicating active renal damage owing to physical and/or immunological causes, it would certainly be of interest to study this question further. Finally, the urinary chemokine test is an excellent parameter indicating the response to antirejection therapy. As indicated in Figures 2 and 3, urinary levels of IP-10 and Mig declined after the initiation of antirejection therapy. The decrease occurs days earlier than the serum Cr. This earlier decline reflects the fact that suppression of active alloimmune reaction is required for the resolution of acute rejection, and renal function recovers after the immune response subsides. Urinary IP-10 test would be especially useful in recipients with acute rejection that superimposed on chronic injury that causes an elevated baseline of serum Cr. In these patients serum Cr may not return to the baseline level, but the urinary IP-10 will decline as acute injury resolves.

It would be desirable to use noninvasive techniques to reduce biopsy procedures for the diagnosis of renal allograft dysfunction. Although our present test is highly specific and sensitive in identifying acute dysfunction in renal grafts when combined with serum Cr, the inability to differentiate acute rejection from acute tubular injury or BK virus nephritis means that biopsy is needed in conjunction with the urine test if both Cr and urine chemokine levels are elevated. Nevertheless, this urinary test is simple to conduct and it is possible to shorten the assay time to less than 2 h, making it suitable for clinical use. Furthermore, the platform of the present test, particle-based multiplex immunoassay, allows the addition of new parameters that are specific for each of these three causes of graft dysfunction. Considering the rapid development of proteomics, we aim to identify additional markers, and further develop a noninvasive diagnostic test using principles of the present study for recipients of renal allografts.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References

We thank Terry Sivesind for the administrative assistance at Renovar, Inc., and Dr Thomas Chin, Linda Jacobs, Nancy Radke, Jan Yakey, and Theresa Bergholz for patient consent and urinary sample collection at the UW Transplant Clinic. We are also grateful to Dr Bryan N. Becker and Dr Andreas Friedl for their critical reading of this manuscript.

Conflict of Interest Disclosure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References

H. Z. Hu is a part-time employee at Renovar, Inc. and a part-time employee at the University of Wisconsin, and holds stock options in Renovar, Inc. S J Knechtle is the President and a shareholder of Renovar, Inc. and M. M. Hamawy is a consultant and shareholder of Renovar, Inc. Both S. J. Knechtle and M. M. Hamawy are full-time employees of the University of Wisconsin.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Participants and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of Interest Disclosure
  9. References