BK-Virus and the Impact of Pre-Emptive Immunosuppression Reduction: 5-Year Results


* Corresponding author: Daniel C. Brennan, brennan@wudosis.wustl.edu


A 1-year, single-center, randomized trial demonstrated that the calcineurin inhibitor or adjuvant immunosuppression, independently, does not affect BK-viruria or viremia and that monitoring and pre-emptive withdrawal of immunosuppression was associated with resolution of BK-viremia and absence of clinical BK-nephropathy without acute rejection or graft loss. A retrospective 5-year review of this trial was conducted. In cases of BK viremia, the antimetabolite was withdrawn and for sustained viremia, the calcineurin inhibitor was minimized. Five-year follow-up was available on 97% of patients. Overall 5-year patient survival was 91% and graft survival was 84%. There were no differences in patient-survival by immunosuppressive regimen or presence of BK-viremia. Immunosuppression and viremia did not influence graft survival. Acute rejection occurred in 12% by 5-years after transplant, was less common with tacrolimus versus cyclosporine (9% vs. 18%; p = 0.082), and was lowest with the tacrolimus-azathioprine regimen (5%, p = 0.127). Tacrolimus was associated with better renal function at 5-years (eGFR 63 FK vs. 52 CsA mL/min, p = 0.001). Minimization of immunosuppression upon detection of BK-viremia was associated with excellent graft survival at 5-years, low rejection rates and excellent renal function. It is a safe, short and long-term strategy that resulted in freedom from clinically evident BK-virus nephropathy.


The BK-virus has emerged as one of the most important infectious agents in renal transplantation. Many questions remain regarding appropriate prevention and treatment. The reactivation of BK-virus has been associated with tacrolimus (FK) and mycophenolic acid (1–7). A trial performed at our center demonstrated that the choice of calcineurin inhibitor or antimetabolite did not independently influence BK-viruria or viremia. Furthermore, suspension of the antimetabolite in patients positive for BK-viremia and minimization of the calcineurin inhibitor in cases of sustained viremia effectively prevented clinical BK-nephropathy at 1-year of follow-up.

Although, prevention of BK-nephropathy was successful in the 1-year follow-up, long-term follow-up is warranted to determine the safety and efficacy of immunosuppressive suspension and minimization on long-term outcomes. The purpose of this paper is to report 5-year outcomes of this trial.


Study design

This study was designed as a retrospective 5-year review of an open-label, prospective trial of renal transplant recipients, randomized to receive either FK or cyclosporine (CsA). Other questions that were studied included the effect of immunosuppression on the incidence of BK-virus infection and the safety and efficacy of pre-emptive reduction of immunosuppression after detection of BK-viremia to prevent progression to BK-virus nephropathy. Inclusion and exclusion criteria have been described previously (8). The Human Research Protection Office at Washington University approved the original and current study.


All patients except 6-antigen HLA-matched living related allograft recipients were induced with four doses (1.5 mg/kg) of rabbit-antithymocyte globulin (Thymoglobulin, Genzyme, Cambridge, MA). Calcineurin inhibitors were instituted when patients achieved a brisk diuresis (but no later than 3 days postoperatively). Patients were randomized to FK with levels targeted between 5 and 10 ng/mL or CsA with targeted 12-hour whole blood trough levels of 250–300 ng/mL during the first three months and 100–250 ng/mL thereafter. Patients received azathioprine (AZA) or mycophenolate mofetil (MMF) if they had a history of glomerulonephritis or other immunologic basis for endstage renal disease (ESRD), a second transplant, high sensitization with PRA > 20% or a history of gout. Steroids were tapered by 3 months to 5–7.5 mg per day.


Plasma and urine samples for BK-virus polymerase chain reaction (PCR) were collected pre-transplant, weekly for 16 weeks, and at months 5, 6, 9 and 12. BK-viruria was defined by the presence of BKV-DNA in the urine, and was taken as evidence of active viral infection. BK-viremia was defined by detection of BKV-DNA in either whole blood or plasma. Intensity of infection was defined as follows; transient viruria–presence of BK-virus in the urine for less than 1 month, sustained viruria—presence of BK-virus in the urine for longer than 1 month, transient viremia—presence of BK-virus in the blood for less than 1 month or sustained viremia-presence of BK-virus in the blood for longer than 1 month. All patients with a rise in creatinine who received biopsies were simultaneously assessed for BK viremia.

Management of BKV

Active viral infection, as detected by a positive urine BK-PCR (viruria), triggered review of the patient's immunosuppression regimen and confirmation or adjustment as appropriate in accordance with our center's practices. Progression to BK-viremia (defined by positive plasma PCR at any level) triggered suspension of the antimetabolite component of the immunosuppressive regimen (AZA or MMF). If viremia failed to clear within 3–4 weeks despite withdrawal of the antimetabolite (sustained viremia), the calcineurin inhibitor was tapered to minimum acceptable levels (typically targeting CsA 12-hour trough levels of 100–200 ng/mL and FK levels of 3–5 ng/mL).

Allograft biopsy

All patients who developed an unexplained elevation of the serum creatinine underwent a kidney biopsy. Acute rejection was defined by the Banff criteria.

Endpoints and outcomes

The endpoints were the rates of acute rejection, patient and graft survival, renal function, new onset diabetes after transplantation (NODAT), and malignancy at 5 years. A triple composite endpoint of freedom from acute rejection, graft loss or death and a quadruple endpoint of freedom from acute rejection, graft loss, death or BK-viremia were established. Renal function was determined by the estimated glomerular filtration rate (eGFR) calculated by the modification of diet in renal disease (MDRD) equation (9).

Database and patient follow-up

Patients were followed with a minimum of monthly labs and yearly clinic visits. Patients were tracked at least yearly and their clinical status was reported to UNOS in accordance with standard procedures at our center. Our center uses an electronic medical record (OTTR–HKS Medical Information Systems, Omaha, NE). In addition, research assistants monitor and record all pertinent data from the medical records of all patients in a SAS database maintained for research purposes (SAS version 8, Carey, NC).


Differences in the characteristics of patients were tested with Student's t-test for continuous variables, Fisher's exact test for binary categorical variables, chi-square for multi-categorical variables. Incidence of patient survival, graft survival and acute rejection were calculated using survival analysis techniques. All statistical tests were two-tailed. An intent to treat analysis was performed.


Between December 2000 and October 2002, 200 patients were enrolled and randomized in a 2:1 fashion to receive either FK or CsA modified as part of a quadruple immunosuppressive regimen. By the twelfth month of the trial, 70 patients (35%) developed viruria and 23 patients (11.5%) developed viremia. Viremia resolved in 95% of patients, with reduction of immunosuppression without complications, including acute rejection, renal impairment or graft loss within the first year. Five-year follow-up was available on 97% (n = 194) of patients enrolled in the trial.

Patient survival

Overall 5-year patient survival was 91%. There were a total of 19 deaths (9%) during the 5-year follow-up. Five patients expired from malignancy: 2 from posttransplant lymphoproliferative disorder, 2 from lung cancer and 1 from pancreatic cancer. Four patients died from cardiovascular causes, 3 from infection (pneumonia–systemic inflammatory response syndrome, acute respiratory distress syndrome and endocarditis, respectively), 2 from suicide, 1 from anoxic encephalopathy, and 4 from undetermined causes. Patients with sustained viremia had lower 5-year survival rates than patients without sustained viremia, (sustained viremia 73% vs. no viremia 92%; p = 0.044; Figure 1). There were no differences in survival between type of calcineurin inhibitor (Table 1) and immunosuppressive regimen (Table 2). Patient survival tended to be best in the CsA-MMF group (96%, p = 0.134, Table 2).

Figure 1.

Patient survival at 5 years: Sustained viremia compared to no sustained viremia was associated with worse patient survival (72.7 vs. 91.5%, p = 0.044).

Table 1.  Five-year outcomes by calcineurin inhibitor
 Total (n = 200)CsA (n = 66)FK (n = 134)p-Value
  1. Triple composite endpoint–freedom from death, graft loss and acute rejection.

  2. Quadruple composite endpoint—freedom from death, graft loss, acute rejection and viremia.

  3. *One patient had an early and late acute rejection:

  4. eGFR = estimated glomerular filtration rate; NODAT = new onset diabetes after transplantation; PTLD = posttransplant lymphoproliferative disorder.

Patient survival91%94%89%0.227
Cause of death, #19415 
Graft Survival84%86%83%0.483
Death censored graft survival93%93%93%0.879
Cause of graft loss, #32923 
 Death with function18414 
 Chronic allograft nephropathy422 
 Recurrent disease220 
Rejection, #(%)24 (12%)12 (18%)12 (9%)0.082
 Grade I16106 
 Grade II312 
 Grade III211 
 Early (<1 year)11560.510
 Late (>1 year)14860.073
Viremia23 (12%)7 (11%)16 (12%)1.00
Serum creatinine (mg/dL)1.4 ± 0.71.7 ± 1.01.3 ± 0.50.01
eGFR (mL/min)60 ± 2052 ± 1963 ± 190.001
NODAT18 (9%)3 (5%)15 (11%)0.188
Malignancy19 (10%)6 (9%)13 (10%)0.968
 Solid or hematological725 
Triple composite endpoint48 (76%)19 (71%)29 (78%)0.327
Quadruple composite endpoint65 (68%)24 (64%)41 (69%)0.495
Table 2.  Five-year outcomes by immunosuppressive regimens
Immunosuppressive regimenFK-MMF (n = 66)CsA-MMF (n = 24)FK-AZA (n = 67)CsA-AZA (n = 42)p-Value
  1. Triple composite endpoint—freedom from death, graft loss and acute rejection.

  2. Quadruple composite endpoint—freedom from death, graft loss, acute rejection and viremia.

  3. *One patient had an early and late acute rejection.

  4. eGFR = estimated glomerular filtration rate; NODAT = new onset diabetes.

Patient survival62 (94%)23 (96%)56 (84%)39 (93%)0.134
Graft survival58 (88%)20 (83%)53 (79%)37 (88%)0.483
Death censored graft survival61 (92%)21 (88%)64 (96%)40 (95%)0.544
Rejection 9 (14%) 4 (17%)3 (5%) 8 (19%)0.127
 Early rejection42230.697
 Late rejection52160.085
BK Viremia 7 (13%)1 (4%) 9 (12%) 6 (15%)0.598
Serum creatinine (mg/dL)1.3 ± 0.51.7 ± 0.61.3 ± 0.41.6 ± 1.10.013
eGFR (mL/min)63 ± 2253 ± 2364 ± 1752 ± 170.013
NODAT 7 (11%) 3 (13%) 8 (12%)00.145
Malignancy    0.276
Triple composite endpoint53 (80%)17 (71%)52 (78%)30 (71%)0.767
Quadruple composite endpoint47 (71%)24 (67%)46 (69%)26 (62%)0.855

Graft survival

Overall 5-year graft survival was 84%. There were a total of 32 (16%) grafts lost during the 5-year period. Overall, 18 grafts (9%) were lost due to death with a functioning graft, 4 due to chronic allograft nephropathy, 2 due to rejection (antibody-mediated rejection that occurred 10 days after transplantation and acute rejection that occurred 18 months after transplantation in a noncompliant patient), 4 due to thrombosis, 2 due to recurrent focal segmental glomerulosclerosis, 1 due to an immunological cause with a gouty attack, and 1 due to acute tubular necrosis with cardiomyopathy. No autopsies were performed. Graft survival and censored graft survival were not influenced by the calcineurin inhibitor, immunosuppressive regimen, or intensity of the BK infection.


Rejection occurred in 24 patients (12%) by 5 years after transplant. Eleven patients (6%) suffered acute rejection in the first year of transplant and 14 (7%) suffered rejection during years 2–5 after transplant. One patient had both an early and late rejection. Acute rejection was less common in the FK arm (CsA 18% vs. FK 9%; p = 0.082, Figure 2). Five-year acute rejection differed by immunosuppressive regimen: FK-AZA (5%), FK-MMF (14%), CsA-MMF (17%), CsA-AZA (19%); p = 0.127 (Figure 3). Overall rejection rates were not influenced by the intensity of the BK infection (Table 3 and 4).

Figure 2.

Acute rejection at 5 years: Tacrolimus was associated with less acute rejection than cyclosporine (9 vs. 18%, p = 0.082).

Figure 3.

Late Acute Rejection by Immunosuppressive Regimen at 5 years: The lowest rejection rates were with the combination of tacrolimus and azathioprine: FK-AZA (5%), CsA-MMF (17%), FK-MMF (14%), CsA-AZA (19%); p = 0.127.

Table 3.  Five-year outcomes by viruria and viremia
 No viruria (n = 130)Viruria (n = 70)p-ValueNo viremia (n = 177)Viremia (n = 23)p-Value
  1. Triple composite endpoint—freedom from death, graft loss and acute rejection.

  2. Quadruple composite endpoint—freedom from death, graft loss and acute rejection and viremia.

  3. *One patient had an early and late acute rejection eGFR-estimated glomerular filtration rate.

Patient survival120 (92%)61 (87%)0.235161 (91%)20 (87%)0.464
Graft survival108 (83%)60 (86%)0.627149 (84%)19 (83%)0.769
Death censored graft survival117 (90%)69 (99%)0.022164 (93%)22 (96%)0.999
Rejection11 (8%)13 (19%)0.036 19 (11%) 5 (22%)0.164
 Early rejection740.922830.119
 Late rejection590.0381220.666
Serum creatinine (mg/dL)1.4 ± 0.71.4 ± 0.50.8951.4 ± 0.71.4 ± 0.40.774
eGFR (mL/min)59 ± 2061 ± 200.88360 ± 2057 ± 180.678
Viremia 0 (0%)23 (33%)<0.0001  
Sustained Viremia 0 (0%)11 (18%)<0.0001  0 (0%)11 (47%)<0.0001 
Triple composite endpoint100 (77%)52 (74%)0.677135 (76%)17 (74%)0.803
Quadruple composite endpoint100 (77%)35 (50%)<0.0001  
Table 4.  Five-year outcomes by the intensity of the BK infection
 No viruria or viremia (n = 130)Transient viruria (n = 11)Sustained viruria (n = 36)Transient viremia (n = 12)Sustained viremia (n = 11)p-Value
  1. Triple composite endpoint—freedom from death, graft loss and acute rejection.

  2. Quadruple composite endpoint—freedom from death, graft loss and acute rejection and viremia.

  3. *One patient had an early and late acute rejection eGFR-estimated glomerular filtration rate.

Patient survival120 (92%)10 (91%) 31 (86%) 12 (100%)8 (73%)0.158
Graft survival108 (83%)10 (91%) 31 (86%) 11 (92%) 8 (73%)0.707
Death censored graft survival117 (90%)11 (100%)36 (100%)11 (92%) 11 (100%)0.185
Rejection11 (8%)3 (27%)5 (14%)3 (25%)2 (18%)0.170
 Early rejection701120.314
 Late rejection534200.015
Serum creatinine (mg/dL)1.4 ± 0.81.8 ± 0.81.3 ± 0.41.5 ± 0.51.2 ± 0.30.274
eGFR (mL/min)60 ± 2050 ± 1862 ± 2152 ± 1466 ± 210.274
Triple composite endpoint100 (77%)8 (73%)27 (75%)9 (75%)8 (73%)0.995
Quadruple composite endpoint100 (77%)8 (73%)27 (75%)0 (0%) 0 (0%) <0.0001 

Sixteen (67%) Banff-grade I rejections occurred, of which all were steroid responsive. Three Banff-grade II rejections were diagnosed. One noncompliant patient developed a Banff IIA rejection approximately 8 months after transplantation, was treated with thymoglobulin and steroids, and had a serum creatinine of 2.7 mg/dL at 5 years after transplant. One patient was diagnosed with a Banff-grade II rejection by the pathologist (felt to be a Banff-grade I rejection by the nephrologists), responded to treatment with steroids, and had a serum creatinine of 1.1 mg/dL at 5 years after transplant. The third patient with an intravenous immunoglobulin-responsive Banff-grade IIB rejection, had persistent BK viremia, rejection, and died after a stroke with a functioning graft (serum creatinine 4.5 mg/dL) just short of his 5-year transplant anniversary.

Two patients suffered Banff-grade III rejections. One non-compliant patient presented with a serum creatinine of 10.5 mg/dL at 16 months after transplantation and after unsuccessful treatment with steroids returned to dialysis 3 months later. The other extremely non-compliant patient presented 19 months after transplantation with a serum creatinine of 19.6 mg/dL. His Banff-grade III rejection was treated unsuccessfully with steroids and he returned to dialysis within 1 week.

Additionally there were two plasma cell and one antibody mediated rejections. One patient with a plasma cell rejection treated with rituximab, remains off dialysis with a serum creatinine of 2.5 mg/dL, at 5 years after transplantation. The other patient with a plasma cell rejection responded to thymoglobulin, intravenous immunoglobulin and steroids and has a serum creatinine of 1.4 mg/dL 5 years after transplant. In one sensitized patient, antibody mediated rejection developed 10-days after a second transplant. Treatment with plasmapheresis and IVIG was unsuccessful and he returned to dialysis.

Composite endpoint-(freedom from death, graft loss or acute rejection)

Nineteen patients (29%) in CsA arm and 29 patients (22%) in the FK arm suffered from acute rejection, graft loss or death. Thus, the 5-year triple composite endpoint was similar (71% CsA vs. 78% FK, respectively, p = 0.327). The quadruple composite endpoint was also similar.

Renal function

For those patients with a functioning graft, the overall mean 5-year serum creatinine was 1.4 ± 0.7 mg/dL and the mean 5-year eGFR was 60 ± 20 mL/min. The CsA group had a higher serum creatinine compared to FK (1.7 ± 1.0 mg/dL vs. 1.3 ± 0.5 mg/dL, p < 0.01) and lower eGFR (52 ± 19 mL/min vs. 63 ± 19 mL/min, p < 0.001), respectively. Viremia and sustained viremia did not influence renal function. However, by immunosuppressive regimen, the FK-AZA and FK-MMF had the best eGFR compared to the other regimens (P = 0.013, Table 2).

In a per protocol analysis of calcineurin inhibitor use, 5 year renal function was better in the FK group for surviving patients that did not switch calcineurin inhibitors. The FK group had a mean serum creatinine of 1.3 ± 0.4 versus 1.6 ± 1.0 mg/dL in the CsA group (P = 0.005) and eGFR of 63 ± 20 versus 53 ± 18 mL/min (P = 0.008), respectively.


By 5 years, 18 patients (9%) developed NODAT. New onset diabetes after transplantation occurred more frequently in those receiving FK although the difference was not statistically different (11% FK vs. 5% CsA, p = 0.188). The incidence of NODAT was 11–13% in each of the immunosuppressive regimens with the exception of the CsA–AZA group in which no cases occurred (P = 0.145, Table 2). Nineteen patients (9.5%) developed malignancy; 12 were cutaneous malignancies and 7 were solid or hematological malignancy. There were no significant differences among groups by calcineurin inhibitor or immunosuppressive regimen (Tables 1 and 2).

Long-term BK virus outcomes

No clinically driven biopsies revealed BKV nephropathy through a mean follow up 6.3 ± 1.2 years. Additionally, although routine monitoring of BK-virus in urine and blood was performed only in the first 12 months after transplantation, clinically driven investigation, performed on all patients at the time of renal biopsy, did not reveal episodes of viremia.

Effect of serology

Outcomes including patient survival, graft survival, acute rejection and renal function were not influenced by donor/recipient BK serology, the baseline or peak BK antibody response (data not shown; 10).

Effect of intensity of infection

The 5-year outcomes by the intensity of the infection are shown in Tables 3 and 4. Death-censored graft rejection, overall rejection and late rejection occurred more commonly among those with viruria compared to no viruria (Table 3).

Influence of reducing immunosuppression in patients with BK viremia

Of the patients with sustained BK viremia (n = 11), all patients remained on a calcineurin inhibitor and most patients 91% (10/11) still had remained off treatment with an antimetabolite at 5 years after transplant. One patient restarted the antimetabolite at 3.5 years after transplant in an attempt to minimize FK concentrations and improve renal function.

Three patients (27%) with sustained viremia died during the 5-year follow-up. These three patients had their antimetabolite discontinued upon detection of viremia and their calcineurin inhibitor reduced after one month of sustained viremia. All patients had cleared the virus in blood at one year of follow-up, while virus remained in the urine. Their antimetabolite was not restarted. One patient developed BK-viremia 3-months after transplantation, suffered two Banff-grade I rejections at 5 and 6-months after transplantation, and died with a functioning graft (terminal serum creatinine 1.3 mg/dL) due to acute respiratory distress syndrome at 4.5 years after transplantation. The second patient developed BK-viremia 2 months after transplantation, suffered a Banff-grade II rejection at 5-months after transplantation, and died from a stroke with a functioning graft (terminal serum creatinine 4.5 mg/dL) 12-days before his 5-year transplant anniversary. The third patient developed BK-viremia 1-month after transplantation and died from lung cancer with a functioning graft (terminal serum creatinine 0.8 mg/dL) 13-months after transplantation.

Of the patients with any BK-viremia 6 out of 23 (26%) either died, had graft loss, or acute rejection after transplantation. These complications were comparable in patients with and without BK-viremia (26% with viremia vs. 28% without viremia, p = 0.787). In addition to the three patients with sustained BK-viremia mentioned earlier, three patients with transient viremia developed complications. One patient developed BK-viremia at 2-months after transplantation, suffered a Banff grade I rejection 7-months after transplantation, and lost her allograft from chronic allograft nephropathy 3-years after transplantation. Two additional patients developed Banff-grade I, late rejections.

Immunosuppressive therapy

At the time of transplant 134 patients (67%) were randomized to FK and 66 (33%) to CsA. Of those patients with functioning grafts at 5 years after transplant, 17 patients (11%) had changed calcineurin inhibitors (12 from CsA to FK and 5 from FK to CsA). Seventeen patients (11%) had their calcineurin inhibitor discontinued, 7 in the CsA arm and 10 in the FK arm. At 5 years the on-therapy eGFR was better in those on FK compared to CsA (52 ± 19 mL/min CsA vs 63 ± 19 mL/min FK, p = 0.001).

At the time of transplant 109 patients (55%) were prescribed AZA. Of the patients with functioning graft, 40 (44%) had discontinued AZA and 15 (16%) switched to MMF by 5 years after transplant. The most common reason for discontinuing AZA was leukopenia and infection and the most common reason for switching to MMF was rejection.

At the time of transplant 89 (45%) of the patients were prescribed MMF. Of the patients with functioning grafts, 22 (29%) had discontinued MMF by 5 years after transplantation. No patients were switched to AZA and the most common reason for discontinuing MMF was infection. One patient continued on hydroxychloroquine, a pretransplant medication used for rheumatoid arthritis after transplant, and one patient was switched to leflunomide for chronic allograft nephropathy (11). All patients with surviving grafts remained on prednisone at 5 years.


Several studies have shown that reducing immunosuppression can prevent or resolve BK-nephropathy early after transplantation (12–17), but few studies have investigated the influence of individual agents on BK-viremia (8). Many studies support our claim that the overall immunosuppressive load, rather than individual agents, is associated with development of BK-viremia and BK-nephropathy (8,12–16). In the present study the antimetabolite was discontinued upon development of BK-viremia and the calcineurin inhibitor was reduced in cases of sustained viremia. The 1-year follow-up of the trial demonstrated that the choice of calcineurin inhibitor or antimetabolite, did not independently influence BK-viruria or viremia. After 1-year we assessed BK-viremia for cause only. At 5 years we observed no new cases of viremia or nephropathy; thus the influence of individual agents on development of viremia did not change. There continued to be differences in outcomes between modified-CsA-and FK treated patients. Patient survival and graft survival were similar, while FK treated patients had less acute rejection and improved renal function at 5 years of follow-up.

Treatment of immunosuppressed patients with BK-viremia is largely supportive and directed toward minimizing immunosuppression. Currently, there is no known effective antiviral agent for BK, and viral infection in a transplant is commonly recognized as a consequence of excessive immunosuppression. We hypothesized that BK-viremia would be a consequence of excessive immunosuppression and could be managed safely and effectively by immediate withdrawal of the antimetabolite in the immunosuppressive regimen. This hypothesis was proven in the 1-year trial that effectively prevented progression of BK viremia to nephropathy. Logically the next question was whether or not this strategy would prevent BK-nephropathy without the consequences of under-immunosuppression, clinically manifested as acute rejection, renal dysfunction or graft loss. At 5-years of follow-up, patients with viremia had similar death-censored allograft outcomes as those patients without viremia. Patients with sustained viremia, however, were more likely to die compared to those without viremia. They died with functioning grafts so it seems unlikely that this was from under-immunosuppression related to withdrawal of the antimetabolite. It may be from an indirect effect of BK or simply a Type 1 error.

In this trial, the development of sustained viruria predicted viremia and sustained viremia. A recent manuscript written by Babel and colleagues supports our prior observations that viremia does not develop without viruria and that sustained viruria predicts viremia (17).

This is the first long-term report of the effects of a pre-emptive immunosuppression reduction strategy to prevent BK-nephropathy. Others have reported successful immunosuppressive minimization at 1–2 years of follow-up (12,13,15,16). The current study is a comprehensive study because 5-year follow-up was available on 97% of patients enrolled in the trial.

Study limitations exist. Protocol biopsies were not performed during the study period, but biopsies were obtained for cause, thus only clinically significant acute rejection, BK-nephropathy and graft dysfunction were recognized. The incidence of histological chronic allograft nephropathy could not be quantified. Second, the number of patients with sustained BK-viremia was small and a larger number of patients may be needed to determine the full impact of reducing immunosuppression on long-term outcomes.

Another limitation to the study may be that the histologic diagnosis of BK-virus can be complicated and often requires confirmation by immunostaining. In this trial, BK viremia was assessed in all patients who underwent a biopsy. BK staining was only performed on those with histological concern for BK or when BK viremia was detected. Although the numbers are small there does not appear to be a difference in chronic allograft nephropathy among those with (n = 1, 4.3%) or without BK viremia (n = 3, 1.6%).

Overall 5-year patient survival, graft survival and acute rejection rates were excellent in this trial. Furthermore this pre-emptive technique resulted in absence of clinical BK-nephropathy at 5 years of follow-up with no differences in overall graft loss between patients with and without viremia or sustained viremia despite withdrawal of the antimetabolite. Although there were few patients with sustained viremia, these patients were likely to die with a functioning graft and suffer acute rejection. Minimization of immunosuppression upon detection of BK viremia is a safe, short- and long-term management strategy that resulted in freedom from BK-virus nephropathy. Further study of patient survival among patient with sustained BK viremia is warranted.


This work was supported by a grant from Astellas Pharmaceuticals to Washington University, the NIDDK K24DK002886 (DCB) and NIDDK P30 DK079333 (DCB).