Pretransplant Interferon-γ Secretion by CMV-Specific CD8+ T Cells Informs the Risk of CMV Replication After Transplantation


Corresponding author: Sara Cantisán


In this prospective study we analyzed pretransplant interferon-γ secretion by cytomegalovirus (CMV)-specific CD8+ T cells to assess its possible utility in determining the risk of CMV replication after solid organ transplantation. A total of 113 lung and kidney transplant patients were enrolled in the study but only 55 were evaluable. All CMV-seronegative recipients were pretransplant “nonreactive” (IFNγ <0.2 IU/mL) (11/11), whereas 30/44 (68.2%) CMV-seropositive (R+) recipients were “reactive” (IFNγ ≥0.2 IU/mL) and 14/44 (31.8%) were “nonreactive”. In the R(+) “nonreactive” group, 7/14 (50%) developed posttransplant CMV replication, whereas the virus replicated only in 4/30 (13.3%) of the R(+) “reactive” patients (p = 0.021). According to the best multivariate model, pretransplant “nonreactive” recipients receiving an organ from a CMV-seropositive donor had a 10-fold increased risk of CMV replication compared to pretransplant “reactive” recipients (adjusted OR 10.49, 95% CI 1.88–58.46). This model displayed good discrimination ability (AUC 0.80) and calibration (Hosmer–Lemeshow test, p = 0.92). Negative and positive predictive values were 83.7% and 75%, respectively. The accuracy of the model was 82%. Therefore, assessment of interferon-γ secretion by cytomegalovirus (CMV)-specific CD8+ T cells prior to transplantation is useful in informing the risk of posttransplant CMV replication in solid organ transplant patients.


area under the curve






CMV-seropositive recipient


CMV-seronegative recipient


CMV-seropositive donor


CMV-seronegative donor




mammalian target of rapamycin.


Traditionally, the risk of cytomegalovirus (CMV) infection in solid organ transplant recipients has been defined by donor and recipient CMV serostatus, the transplanted organ and additional immunosuppressive therapy. Therefore, these parameters have been used to design the prevention strategy [1-3]. CMV serology has been considered a surrogate marker of the presence of CMV-specific cellular immunity. Consequently, it is considered that because CMV-seronegative recipients (R−) lack specific immunity they are at highest risk of CMV replication when they receive an organ from a CMV-seropositive donor (D+/R−) and antiviral prophylaxis protocol is recommended. On the contrary, CMV-seropositive recipients (R+) are considered to have CMV-specific immunity and have an intermediate risk of CMV replication after transplantation. In this case, antiviral prophylaxis is replaced by viral monitoring after transplantation applying a therapy when symptomatic replication is detected (preemptive therapy).

However, this traditional serological classification of patients has some limitations. Firstly, although R(+) recipients have an intermediate risk of CMV replication, CMV reactivates in some recipients after transplantation producing CMV-related complications [4]. Secondly, although only a few intermediate risk R(+) recipients will develop CMV disease if they are appropriately managed, all of them follow a rigorous and expensive viral monitoring protocol [5, 6]. On the other hand, although most patients receiving antiviral prophylaxis will never develop CMV replication after discontinuation of the prophylaxis, the extension of the prophylaxis period or continuation with preemptive therapy has been proposed.

Our hypothesis is that pretransplant CMV-serostatus determination, by itself, may not be a good marker for identifying the risk of CMV primary infection/reactivation and that the analysis of CMV-specific T cell response may allow for a better prediction. CD8+ T cell response is a fundamental effector process controlling CMV replication [7, 8]. Recently, many studies have reported the relationship between posttransplant functional impairment of CD8+ T cells and failure to suppress CMV replication after solid-organ transplantation [9-11]. However, there are no studies addressing whether pretransplant functional impairment of CMV-specific CD8+ T cells might be associated with the risk of posttransplant CMV replication. In this prospective study we have analyzed pretransplant CMV-specific CD8+ T cells secreting interferon-γ (IFNγ) using the QuantiFERON®-CMV assay to determine whether this test provides prognostic value of CMV replication after transplantation.


Study population and design

This longitudinal study was carried out in two centers of the REIPI network (Reina Sofia University Hospital, Cordoba, and Cruces University Hospital, Barakaldo, Spain). Adult patients were eligible for the study if they were awaiting lung or kidney transplantation. Patients at high risk of CMV replication, such as CMV-seronegative recipients of organs from CMV-seropositive donors (D+R−), recipients receiving thymoglobulin induction therapy or lung recipients, received antiviral prophylaxis. Standard prophylaxis consisted of intravenous ganciclovir (5 mg/kg/day) or valganciclovir (900 mg/day) adjusted for renal function. In lung recipients, the recommended duration of prophylaxis was 6 months, while in high-risk kidney recipients it was 3 months. Patients at intermediate risk of CMV replication received preemptive therapy, which was administered when the viral load was higher than 1500 copies/mL. Viral replication episodes were treated for at least 2 weeks or until a negative PCR was obtained. The study was approved by the ethics committee of the aforementioned institution and informed consent was obtained from all patients included in the study.

Patients were recruited from March 2009 to June 2010. Patients who received a graft were monitored for CMV replication for 24 months following transplantation. Specifically, CMV viral load was determined at least weekly during hospitalization, every 2 weeks until the third month, monthly until the first year and when clinically indicated. IFNγ production by CMV-specific CD8+ T cells was assessed pretransplant when infectologic evaluation was performed and at 3 or 6 months posttransplant (coinciding with the discontinuation of antiviral prophylaxis in high-risk kidney or lung transplant recipients, respectively).

Patients with CMV replication were classified as asymptomatic or as having CMV disease according to standard definitions [1].

Immunosuppressive and rejection therapy

Immunosuppression was indicated according to the protocols of each center. All solid organ transplant recipients received a standard triple-therapy immunosuppression regimen consisting of cyclosporine/tacrolimus, mycophenolate mofetil and steroids. Induction therapy with thymoglobulin or basiliximab was used according to each center's practice. Steroids were used to treat acute rejection episodes.

Determination of anti-CMV IgG antibodies and CMV viral load

Pretransplant anti-CMV IgG antibodies were determined in donors and recipients by chemoluminiscence (Diasorin SA, Spain). CMV load was determined by real-time PCR (COBAS® AmpliPrep/COBAS® TaqMan®, Roche Diagnostics, Branchburg, NJ, USA). Viral DNA was isolated from citrate-anticoagulated whole blood samples according to the manufacturer's instructions. The lower limit of the assay was 150 copies/mL. Peak viral load was defined as the maximum viral load detected at any point within the posttransplant period. Episodes of viral replication were treated for at least 2 weeks or until the PCR result was negative. The duration of CMV replication was calculated as the number of days from the first positive PCR to the first negative PCR. In patients with two or more replication episodes, the total number of days of the different episodes was considered.

QuantiFERON-CMV Assay

The QuantiFERON-CMV (QF-CMV) test was performed according to the manufacturer's instructions (Cellestis, a QIAGEN company, Melbourne, Australia). In brief, 1 mL of heparinized whole blood was collected in three QF-CMV blood collection tubes. Tubes contained either (i) a mix of 22 CMV peptides from a variety of proteins; (ii) no antigens (negative control) or (iii) phytohemagglutinin (positive mitogen control). After collection, the tubes were shaken vigorously and incubated for 16–24 h at 37°C. Subsequently, supernatants were recovered and analyzed for IFNγ (IU/mL) by standard ELISA. A result for the CMV antigen was considered “reactive” when the CMV antigen response minus the negative control response was greater than 0.2 IU/mL of IFNγ. According to the manufacturer's instructions, a result was considered “indeterminate” when the IFNγ level in the CMV antigen tube minus the negative control was less than 0.2 IU/mL and the IFNγ level in the mitogen tube (once the negative control is subtracted) was less than 0.5 IU/mL. For the purpose of statistical analysis “indeterminate” results was considered “nonreactive”.

Statistical analysis

Statistical analysis was performed using the SPSS 15.0 program. All categorical variables were compared with the chi-square test or with Fisher's exact test. The probability of change in QF-CMV response after transplantation was evaluated using the McNemar Test, which compares paired binary data before and after a specific event. Logistic regression was used to evaluate the association between various pretransplant factors and the onset of posttransplant CMV replication during the follow-up period. We included some parameters of clinical or biological significance in the multivariate analysis in spite of not being statistically significant in the univariate analysis. In order to prevent collinearity between independent variables, only noninterrelated parameters were included in the same multivariate model. Goodness-of-fit was performed using the Hosmer–Lemeshow test. The area under the curve (AUC), sensitivity, specificity, positive and negative predictive values and the accuracy of the model were also assessed. Values were considered statistically significant when the p value was < 0.05.


Demographic characteristics of patients

A total of 113 patients awaiting lung or kidney transplantation were enrolled in the study. Of these, six patients awaiting a lung transplant died while they were on the waiting list. Of the remaining 107 patients, 61 received a solid organ transplant during the study period and satisfied the “inclusion criteria”. Of these, six patients died in the 2 months following the lung transplantation. Therefore, we have analyzed a total of 55 evaluable patients. The demographic characteristics of the patients are shown in Table 1.

Table 1. Patient characteristics
CharacteristicsN = 55
  1. CNI = calcineurin inhibitor; MMF = mycophenolate mofetil; mTOR = mammalian target of rapamycin.

Age, median (range)50 (21–75)
Sex, n (%) 
 Female 17 (30.9)
 Male38 (69.1)
Donor (D)/recipient (R) CMV-Serostatus, n (%) 
 D−R−3 (5.4)
 D+R−8 (14.5)
 D−R+11 (20.0)
 D+R+33 (60.0)
Type of transplant, n (%) 
 Lung23 (41.8)
 Kidney32 (58.2)
Antiviral prevention strategy, n (%) 
 Preemptive therapy23 (43.6)
 Universal prophylaxis31 (56.4)
Rejection, n (%)14 (25.4)
Induction therapy, n (%) 
 Thymoglobulin1 (1.8)
 Basiliximab15 (27.3)
Maintenance immunosuppressiona, n (%) 
 Corticosteroids/CNI/MMF55 (100)
 mTOR inhibitor10 (18.2)

There were 11 R(−) and 44 R(+) recipients. Of the R(−) patients, 3 were D−R− and 8 D+R−. Of the R(+) patients, 11 were D−R+ and 33 D+R+. All lung transplanted patients (n = 23) and D+R− kidney recipients (n = 5) received antiviral prophylaxis. The majority of R(+) kidney recipients (n = 24) received preemptive therapy, with the exception of three patients who received induction therapy and were therefore administered antiviral prophylaxis instead of preemptive therapy. The total incidence of CMV replication and CMV disease was 16/55 (29.1%) and 5/55 (9.1%), respectively. Of the D+R− patients, 5/8 (62.5%) developed CMV replication after transplantation, while only 11/44 (25%) R(+) patients experienced a posttransplant CMV replication episode. Of these 11 R+ patients with CMV replication, 6 received preemptive therapy and 5 were administered universal prophylaxis. The median duration of CMV replication was 31 days (range 9–78). Only one patient experienced more than one replication episode and in this case the total number of days of the different episodes was considered. The median peak CMV load was 13 650 copies/mL (range 991–1.4×106). The median posttransplant day of CMV replication was 153 (range 31–734) in the whole group and 106 (range 31–274) in R+ patients.

QuantiFERON-CMV assay results

The pretransplant QF-CMV test was “reactive” (QF-CMVR) in 30/55 recipients (54.5%). No patient had an “indeterminate” pretransplant QF-CMV test. The median IFNγ level in these patients was 5.81 IU/mL (range 0.37–147.18 IU/mL). The posttransplant QF-CMV test was QF-CMVR in 28/55 recipients (50.9%) and the median IFNγ level was 3.90 IU/mL (range 0.24–60.21 IU/mL). At posttransplant timepoint, eight patients (14.5%) showed an “indeterminate” result. Six of these patients were “nonreactive” (QF-CMVNR) and two were “reactive” at the pretransplant timepoint. For the analysis purposes, these eight patients were considered as QF-CMVNR.

We separately analyzed the frequency of QF-CMVR recipients in the group receiving prophylaxis versus preemptive therapy at both timepoints to study changes in relation to the pretransplant results. Pretransplant QF-CMV was “reactive” in 19 of 24 (79.2%) patients receiving preemptive therapy (intermediate risk group). One pretransplant QF-CMV “nonreactive” patient changed to “reactive” at the posttransplant timepoint. This patient had CMV replication in this period. Three pretransplant “reactive” recipients changed to “nonreactive” (one of them “indeterminate”) (McNemar test, p = 0.625).

Pretransplant QF-CMV was “reactive” in 11 of 31 (35.5%) patients receiving prophylaxis (high risk group). Three recipients who were pretransplant QF-CMV “nonreactive” changed to “reactive” at the second point. Two of them were D+R− and required the temporary suspension of prophylaxis due to the toxicity of valganciclovir. Both patients had CMV replication. Three pretransplant QF-CMV “reactive” recipients changed to “nonreactive” (one of them “indeterminate”) in the same period (McNemar Test, p = 1.00).

Factors associated with CMV replication after solid organ transplantation

Logistic regression was used to determine the factors associated with the risk of CMV replication after transplantation. In the univariate analysis (Table 2), pretransplant QF-CMV was statistically significant and QF-CMVNR patients had a higher risk of posttransplant CMV replication than QF-CMVR patients (OR 6.0, 95% CI 1.66–22.30). However, in high-risk recipients (N = 31), QF-CMVNR at the end of the prophylaxis period was not associated with an increased risk of CMV replication after transplantation compared with QF-CMVR recipients (OR 8.18, 95% CI 0.87–76.60). When donor CMV-serology was considered, D+R- serostatus was associated with an increased risk of developing CMV replication after transplantation compared with R(+) recipients (OR 5.0, 95% CI 1.02–24.41). We included a new variable combining donor CMV-serostatus with pretransplant recipient QF-CMV and observed that D(+)QF-CMVNR recipients were strongly and independently associated with an increased risk of developing posttransplant CMV replication compared with QF-CMVR recipients (OR 8.12, 95% CI 1.99–33.09). Other factors, such as age, the transplanted organ, type of immunosuppression, antiviral prevention strategy or graft rejection, were not associated with an increased risk of CMV replication.

Table 2. Univariate logistic regression model relating some parameters with posttransplant CMV replication (n = 55)
VariableCMV replication (n = 16)No CMV replication (n = 39)OR (95% CI)p-Value
  1. 1QF-CMVNR, QuantiFERON-CMV “nonreactive” recipients (<0.2 IU/mL IFNγ); QF-CMVR, QuantiFERON-CMV “reactive” recipients (≥0.2 IU/mL IFNγ).

  2. 2

    Only patients receiving antiviral prophylaxis (N = 31) were analyzed.

  3. 3

    N/A: not applicable. Unable to calculate OR due to the absence of CMV replication in D−R− patients.

  4. CI = confidence interval.

Age (per 1 year increase)43.5 (26–69)50 (21–75)0.97 (0.93–1.01)0.193
 Female3 (17.6)14 (82.4)0.41 (0.1–1.70)0.220
 Male13 (34.2)25 (65.8)1 
Recipient (R) CMV-serostatus    
 R−5 (45.4)6 (54.5)2.50 (0.64–9.83)0.190
 R+11 (25.0)33 (75.0)1 
Pretransplant recipient QF-CMV1    
 QF-CMVNR12 (48.0)13 (52.0)6.00 (1.66–22.30)0.007
 QF-CMVR4 (13.3)26 (86.7)1 
Posttransplant recipient QF-CMV2    
 QF-CMVNR9 (45.0)11 (55.0)8.18 (0.87–76.60)0.065
 QF-CMVR1 (9.1)10 (90.9)1 
Donor (D)/recipient (R) CMV-serostatus    
 D−/R−03 (100)N/A3 
 D+/R−5 (62.5)3 (37.5)5.00 (1.02–24.41)0.047
 R+11 (25.0)33 (75.0)1 
Donor CMV-serology/pretransplant recipient QF-CMV    
 D(−)/QF-CMVNR2 (28.6)5 (71.4)2.60 (0.37–18.25)0.337
 D(+)/QF-CMVNR10 (55.6)8 (44.4)8.12 (1.99–33.09)0.003
 QF-CMVR4 (13.3)26 (86.7)1 
Type of transplant    
 Lung6 (26.1)17 (73.9)0.77 (0.23–2.56)0.678
 Kidney10 (31.3)22 (68.7)1 
Antiviral prevention strategy    
 Prophylaxis10 (32.2)21 (67.7)1.43 (0.43–4.70)0.558
 Preemptive therapy6 (25.0)18 (75.0)1 
 Yes3 (21.4)11 (78.6)0.59 (0.14–2.47)0.468
 No13 (31.7)28 (68.3)1 
 Yes6 (40.0)9 (60.0)2.00 (0.57–7.03)0.280
 No10 (25.0)30 (75.0)1 
mTOR inhibitor    
 Yes5 (50.0)5 (50.0)3.09 (0.75–12.71)0.118
 No11 (24.5)34 (75.5)1 

In the multivariate analysis (Table 3), we developed three separate multivariate models since D/CMV-serostatus and D/QF-CMV are related and they cannot be included in the same model. In addition, some parameters such as “antiviral prevention strategy” could not be included in the same multivariate model with “type of transplant” or “D/R CMV-serostatus” due to collinearity. Thus, model 1 included D/R CMV-serostatus and the variables we thought might affect posttransplant CMV replication even though they were not statistically significant in the univariate analysis. These included recipients” age and type of transplant. In model 2 we substituted D/R CMV-serostatus for D/QF-CMV and also included the use of basiliximab. Model 3 was identical to model 2 with the exception of “type of transplant”, which was substituted for “antiviral prevention strategy”. As observed in model 1, none of the parameters were statistically significant, including D/R CMV-serostatus. In model 2, however, D(+)/QF-CMVNR versus QF-CMVR recipients was the only statistically significant variable, and was independently associated with a 10-fold increased risk of CMV replication after transplantation compared with QF-CMVR recipients (adjusted OR 10.18 95% CI 2.07–50.13). In model 3, D(+)/QF-CMVNR versus QF-CMVR recipients was also the only statistically significant variable, and was associated with a 10-fold increased risk of CMV replication after transplantation compared with QF-CMVR recipients (adjusted OR 10.49 95% CI 1.88–58.46). The Hosmer–Lemeshow test result (p = 0.61 for model 2 and p = 0.92 for model 3) indicated that multivariate model 3 was the best model. Model 3 displayed good calibration and discrimination ability (AUC 0.80, 95% CI 0.67-0.94). Sensitivity and specificity were 56.2% and 92.3%, respectively. Negative and positive predict values were 83.7% and 75%, respectively. The accuracy of the model was 82%.

Table 3. Multivariate logistic regression models relating some parameters with posttransplant CMV replication (n = 55)
Multivariate models Adjusted OR95% CIp
  1. 1In order to calculate the odds ratio for D/R CMV-serostatus, D−R− recipients (n = 3) were removed as there were 0 individuals in this category.

  2. 2QF-CMVNR, QuantiFERON-CMV “nonreactive” recipients (<0.2 IU/mL IFNγ); QF-CMVR, “reactive” recipients (≥0.2 IU/mL IFNγ)

  3. 3Goodness-of-fit: Hosmer–Lemeshow test result (model 2, p = 0.61; model 3, p = 0.92)

  4. 4Area under the curve (AUC): 0.80; sensitivity: 56.2%; specificity: 92.3%; negative predictive value: 83.7%; positive predictive value: 75%.

  5. CI = confidence interval.

Model 1Donor (D)/recipient (R) CMV-serostatus1   
  D+/R− vs. R+3.710.70–19.690.123
 Age (per 1 year increase)0.970.92–1.020.242
 Type of transplant   
  Lung vs. kidney1.060.29–3.880.934
Model 2*Donor CMV-serostatus/pretransplant recipient QF-CMV2   
  D+/QF-CMVNR vs. QF-CMVR10.182.07–50.130.004
  D−/QF-CMVNR vs. QF-CMVR3.340.32–34.510.311
 Age (per 1 year increase)0.950.89–1.010.084
 Type of transplant   
  Lung vs. kidney0.890.12–6.410.910
Model 3* §Donor CMV-serostatus/pretransplant recipient QF-CMV   
  D+/QF-CMVNR vs. QF-CMVR10.491.88–58.460.007
  D−/QF-CMVNR vs. QF-CMVR3.340.35–31.740.294
 Age (per 1 year increase)0.950.89–1.000.066
 Antiviral prevention strategy   
  Prophylaxis vs. preemptive therapy0.880.15–5.190.886

Five patients developed CMV disease (Table 4). However, this scarce number prevented us from performing a multivariate analysis.

Table 4. Characteristics of patients who developed CMV disease (n = 5)
Patient codeAgeType of transplantPretransplant recipient QF-CMV1D/R CMV- serostatus2Antiviral prevention strategy3Posttransplant day of CMV replicationTotal duration of CMV replication (days)4Peak viral load (copies/mL)Type of CMV disease
  1. 1QuantiFERON-CMV “nonreactive” (<0.2 IU/mL IFNγ).

  2. 2Donor (D) and recipient (R) CMV-serostatus.

  3. 3The duration of prophylaxis was 3 months for Tx147 and Tx124 and 6 months for Tx96.

  4. 4Tx111 had three CMV replication episodes and total duration was considered.

Tx11154KidneyNonreactiveD−R+Preemptive317873 579Syndrome
Tx12955KidneyNonreactiveD+R+Preemptive373328 753Syndrome
Tx14730KidneyNonreactiveD+R−Prophylaxis7342329 400Hepatitis
Tx12434KidneyNonreactiveD+R−Prophylaxis461441 387 387Gastrointestinal
Tx9633LungNonreactiveD+R+Prophylaxis2301813 800Syndrome

QuantiFERON-CMV and recipient CMV-serostatus agreement

In order to determine whether “CMV-serostatus” and “pretransplant QF-CMV” were independent variables or, on the contrary, were associated, we performed a test of independence using a chi-squared test. The test result (p < 0.001) indicated that CMV-serostatus and pretransplant QF-CMV were associated. In fact, 100% of the R(−) recipients (11/11) were QF-CMVNR and 68.2% of the R(+) recipients (30/44) were QF-CMVR. However, 14 of the 44 R(+) recipients (31.8%) were QF-CMVNR. Since these differences could be clinically relevant, we analyzed the R(+) group separately.

CMV-seropositive recipients

Incidence of posttransplant CMV replication in R(+)QF-CMVNR versus R(+)QF-CMVR

In order to determine how the QF-CMV results before transplantation could affect future CMV replication after transplantation, we performed a separate analysis of R(+) recipients (n = 44). Strikingly, we observed a higher incidence of posttransplant CMV replication in R(+) recipients with pretransplant QF-CMVNR. Specifically, 7/14 (50%) R(+)QF-CMVNR recipients developed CMV replication after transplantation, whereas the virus replicated only in 4/30 (13.3%) of R(+)QF-CMVR patients (p = 0.021; corrected chi-square). The 26 R(+)QF-CMVR recipients that did not experience a posttransplant CMV replication episode showed a median IFNγ production of 7.0 IU/mL (interquartile range 1.5–32.0) and the 4 recipients with CMV replication showed a median of 5.8 IU/mL (interquartile range 3.8–13.4).

Kinetics of CMV replication and incidence of CMV disease

To determine whether R(+)QF-CMVNR and R(+)QF-CMVR also behaved differently at the onset of posttransplant CMV replication and the kinetics of CMV replication, we analyzed parameters such as posttransplant day of CMV replication, peak CMV load and duration of CMV replication episodes. Since in lung recipients CMV replicates after discontinuing antiviral prophylaxis, Table 5 only shows the values for kidney transplant recipients. In R(+)QF-CMVNR recipients, the virus replicated earlier than in R(+)QF-CMVR patients. In addition, the peak CMV load was higher and the duration of CMV replication was longer than in R(+)QF-CMVR patients. Two of the R(+)QF-CMVNR recipients developed CMV disease, while the three R(+)QF-CMVR recipients were asymptomatic. Therefore, although we could not carry out a statistical analysis due to the low number of patients in each group (n = 3 + n = 3), we believe that the differences observed between both groups are suggestive of a causal association of R(+)QF-CMVNR with the severity of CMV infection.

Table 5. Kinetic of CMV replication in CMV-seropositive kidney transplant recipients
R(+)QF-CMV1Posttransplant day of CMV replicationPeak CMV load (copies/mL)Total duration of CMV replication (days)2CMV prevention strategy3A/D4
  1. 1CMV-seropositive recipients and pretransplant QuantiFERON-CMV test result. R(+)QF-CMVNR, “nonreactive” recipients (<0.2 IU/mL IFNγ); R(+)QF-CMVR, “Reactive” recipients (≥0.2 IU/mL IFNγ).

  2. 2Tx111 had three CMV replication episodes and total duration was considered.

  3. 3All patients but one (Tx121) received antiviral treatment. Tx121 did not reach the threshold of 1500 copies/mL.

  4. 4(A) Asymptomatic or (D) CMV disease. Tx111 and Tx129 had viral syndrome.

 Tx1113173 57978PreemptiveD
 Tx1293728 75333PreemptiveD
 Tx1223957 36424PreemptiveA
 Tx1211061 14729PreemptiveA
 Tx1337017 8379PreemptiveA
 Tx112598 2349PreemptiveA


This study shows that pretransplant QuantiFERON®-CMV assay predicts the risk of CMV replication after transplantation in solid organ transplant recipients. A complete agreement of the QF-CMV assay with CMV-serology is observed in R(−) recipients since all of them (11/11) are pretransplant QF-CMVNR. However, a subset of R(+) recipients were QF-CMVNR before the transplantation, indicating that in spite of having been in contact with the virus they did not develop CMV-specific CD8+ T immunity. Alternatively, they might have developed CMV-specific CD8+ T cells, but these cells are not able to produce IFNγ. This could be associated with an overall impairment of cellular immunity (possibly due to hemodialysis) [12, 13] or with the impaired functionality of CMV-specific CD8+ T cells related to high maturation stage and phenotype [14-19]. Thus, R(+)QF-CMVNR recipients are at an increased risk of posttransplant CMV replication compared with R(+)QF-CMVR recipients. Changes in QF-CMV response from pretransplant to the second timepoint were observed in our study. Thus, one intermediate risk QF-CMVNR patient became QF-CMVR because the patient developed CMV replication in this period. Two high-risk D+R− patients also became QF-CMVR because prophylaxis was temporally suspended and CMV was able to trigger the development of functional CMV-specific CD8+ T cell immunity. Three intermediate risk and three high-risk QF-CMVR patients became QF-CMVNR (one “indeterminate” in each group). Overall, eight patients converted to “indeterminate” after transplantation. The considerable number of patients that converted to “nonreactive” or “indeterminate” may reflect overimmunosuppression. When donor CMV-serostatus is also considered, QF-CMVNR recipients receiving an organ from a D(+) have a higher risk of CMV replication after transplantation than QF-CMVR recipients. The use of the QuantiFERON®-CMV test before transplantation could therefore be useful in more accurately predicting the risk of posttransplant CMV replication in solid organ transplant patients.

Several studies have reported the utility of determining IFNγ production to predict the risk of CMV replication, but all of them measured IFNγ production after transplantation [4, 11, 20]. In another study, a reduced production of posttransplant IFNγ by CMV-specific CD8+ T cells was coincident with symptomatic CMV recrudescence in transplant recipients [10]. Kumar et al. [21] reported that the measurement of CMV-specific immunity close to the end of prophylaxis was predictive of CMV disease. Although they performed pretransplant QF-CMV assay in a group of recipients, they did not report any results regarding this parameter.

In our study, we did not find any relationship between recipients” age or use of mTOR inhibitors and the risk of CMV replication, contrary to what has been observed by other authors previously [3, 11, 22]. Although basiliximab did not increase the risk of CMV replication both in our study and others [23], it contributed to improving the predictive accuracy of the final models.

According to our results, pretransplant QuantiFERON®-CMV assay might aid in deciding how to manage the recipients once they receive the allograft. Thus, if intermediate-risk R(+) recipients are QF-CMVR before transplantation, they will have a low risk of posttransplant CMV replication or, if the virus replicates, the episode will be of low severity. In such a situation it might therefore be possible to avoid CMV load monitoring of the patient. On the contrary, if intermediate-risk R(+) recipients are QF-CMVNR before transplantation, they should be managed as high-risk recipients and antiviral prophylaxis might be considered.

Pretransplant QF-CMV assay in intermediate-risk patients might also be useful in predicting the timing and the kinetics of CMV replication although the low frequency of patients impedes carrying out any statistical analyses. When we compared R(+)QF-CMVNR versus R(+)QF-CMVR kidney transplant recipients, we observed that in R(+)QF-CMVNR recipients: (i) CMV replicated earlier after transplantation; (ii) CMV reached a higher peak viral load; (iii) CMV replication was longer in duration and (iv) there was a higher frequency of CMV disease. In fact, this is consistent with what has been recently published by Lisboa et al. [24] in solid organ transplant patients with low-level asymptomatic viremia. High IFNγ production by CMV-specific CD8+ T cells shortly after the onset of CMV viremia was associated with spontaneous CMV clearance. It has also been reported that, even in high-risk transplant recipients, the decline in the incidence of CMV replication episodes is inversely correlated with the acquisition of the CMV-specific T cell response, and that after acquisition of the immune response, some replication episodes were cleared without treatment [9].

The main limitation of our study is that given that CMV replication occurred in a small number of R(+) kidney transplant recipients, we were unable to perform any statistical analyses to compare the posttransplant day to CMV replication and the kinetics of the episodes between R(+)QF-CMVNR and R(+)QF-CMVR recipients, although we did observe some relevant differences. We believe that if we analyzed the pretransplant QuantiFERON®-CMV assay in a higher number of patients, we would obtain statistical significance in these parameters.

In summary, we have reported that one third of R(+) transplant candidates are QF-CMV “nonreactive” and they are at higher risk for CMV replication after transplantation. Thus, determining pretransplant IFNγ production by CMV-specific CD8+ T cells in lung or kidney transplant recipients using the QuantiFERON®-CMV assay may be useful in predicting the risk for CMV replication after transplantation. This determination might aid in reclassifying patients awaiting solid-organ transplantation, and identifying which patients will be at higher risk of CMV replication once they receive the graft. Hence, it may allow the individualization of CMV infection management after solid-organ transplantation [25]. R(+) kidney transplant patients with pretransplant QF-CMVNR should therefore be managed as high-risk patients. Consequently, this pretransplant assay could be used in the clinical setting to guide physicians in administering the adequate anti-CMV therapy after solid-organ transplantation.


This work was supported by the Spanish Ministry of Science and Innovation and the Carlos III Health Institute (grant numbers FIS09/0723 to R.S., FIS08/0336 to J.T.C.), and cofinanced by the European Regional Development Fund and the Spanish Network for Research in Infectious Diseases (REIPI RD06/0008). We are grateful to Mr. George Dragovic and Cellestis, a QIAGEN company, for providing the QuantiFERON-CMV kits.


The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. J.T.C. and S.C. have received an unrestricted research grant from Roche Pharma.