Cytomegalovirus (CMV) infection is still a major complication after kidney transplantation. Although cytotoxic CMV-specific T cells play a crucial role controlling CMV survival and replication, current pretransplant risk assessment for CMV infection is only based on donor/recipient (IgG)-serostatus. Here, we evaluated the usefulness of monitoring pre- and 6-month CMV-specific T cell responses against two dominant CMV antigens (IE-1 and pp65) and a CMV lysate, using an IFN-γ Elispot, for predicting the advent of CMV infection in two cohorts of 137 kidney transplant recipients either receiving routine prophylaxis (n = 39) or preemptive treatment (n = 98). Incidence of CMV antigenemia/disease within the prophylaxis and preemptive group was 28%/20% and 22%/12%, respectively. Patients developing CMV infection showed significantly lower anti-IE-1-specific T cell responses than those that did not in both groups (p < 0.05). In a ROC curve analysis, low pretransplant anti-IE-1-specific T cell responses predicted the risk of both primary and late-onset CMV infection with high sensitivity and specificity (AUC > 0.70). Furthermore, when using most sensitive and specific Elispot cut-off values, a higher than 80% and 90% sensitivity and negative predictive value was obtained, respectively. Monitoring IE-1-specific T cell responses before transplantation may be useful for predicting posttransplant risk of CMV infection, thus potentially guiding decision-making regarding CMV preventive treatment.
biopsy-proven acute rejection
enzyme-linked immunosorbent spot assay
peripheral blood mononuclear cells
rabbit antithymocyte globulin
receiver operating characteristic
Human cytomegalovirus (CMV) infection is still a major complication after kidney transplantation. Because of T cell immunosuppression, transplant recipients are at increased risk to develop CMV infection a short time after transplantation, critically challenging both graft and also patient survival [1, 2]. On the one hand, CMV infection may directly lead to persistent viremia and tissue-invasive injury such as pneumonitis, enteritis or retinitis, and on the other, indirect-related CMV effects have also been associated to acute and chronic allograft rejection, diabetes and accelerated atherosclerosis [3, 4].
Noteworthy, the advent of preventive strategies using either universal prophylaxis or preemptive treatment initiated on the basis of viral detection in blood has significantly helped to reduce morbidity and mortality. Indeed, while recent reports have shown that routine prophylaxis with valganciclovir may reduce the incidence of posttransplant CMV infection and improve long-term kidney graft survival [5-8], other groups have also shown that preemptive therapy is also able to decrease the incidence of CMV disease, avoiding development of anti-viral resistance, drug toxicity [7, 9, 10] and the advent of late-onset CMV infection [11, 12]. Furthermore, some CMV seronegative patients receiving a kidney allograft from a CMV seropositive donor never develop CMV infection despite not receiving any prophylaxis treatment [13-15]. Altogether, it suggests that current immune assessment of the CMV risk before kidney transplantation exclusively evaluating detectable circulating CMV IgG titers is not accurate and informative enough to predict the risk of CMV infection in all transplant recipients.
Cytotoxic CMV-specific T cells play a crucial role controlling CMV survival and replication [16, 17]. While host CD8+ T cells may target a wide range of CMV immunogenic proteins, particular dominant T cell responses against immediately early-1 (IE-1) antigens and to phosphoprotein 65 (pp65) seem to be essential for CMV control [18-20]. Recent relevant reports using different T cell immune-monitoring tools have shown the importance of such CMV-specific T cell responses for controlling CMV infection after transplantation. However, most of them have mainly focused on the posttransplant period, gathering different solid organ transplants and assessing rather low numbers of kidney transplant recipients [21-25].
Since all kidney transplant patients display an intrinsic baseline functionality of CMV-specific T cell responses, thus predisposing to CMV replication after transplantation, we aimed to evaluate the clinical usefulness of monitoring prior to transplantation CMV-specific T cell responses against dominant CMV antigens (IE-1 and pp65) and a CMV lysate, using an IFN-γ Elispot assay, for predicting the advent of posttransplant CMV infection in two cohorts of kidney transplant recipients either receiving routine prophylaxis (n = 39) or preemptive treatment (n = 98). Furthermore, changes in 6-month posttransplant CMV-specific T cell responses were also analyzed in both groups of patients.
Patients and study groups
This is a single-center retrospective study performed at our Renal Transplant Unit at Bellvitge University Hospital in Barcelona, Spain. Between June 2009 and June 2011, consecutive kidney adult renal transplant recipients were enrolled to the study if pretransplant peripheral blood mononuclear cells (PBMCs) were available. The study was approved by the Ethics Committee of our center.
Patients were divided in two groups, depending on the CMV preventive strategy performed; either prophylaxis or preemptive therapy was done following the clinical protocol established in our Transplant Unit during the study time period. Until June 2010, prophylaxis treatment posttransplantation was restricted to CMV seronegative transplant recipients receiving a seropositive donor (R−/D+), and preemptive therapy was carried out in all CMV positive recipients either receiving a positive or a negative donor allograft (R+/D+ and R+/D−, respectively), including those receiving T cell depleting antibodies. Subsequently, from July 2010 on, a prophylaxis policy was also extended to all CMV positive transplant recipients (R+) receiving T cell depleting antibodies. In addition, six R− patients because of either hypersensitivity history to acyclovir or showing posttransplant absolute leukocyte count <2000 cells/µL, platelet count <100 000 cells/µL or hemoglobin levels lower than 8.0 g/dL, preemptive therapy was assigned.
CMV preventive strategies
In the prophylaxis group, including those transplants recipients receiving T cell depleting agents such as rATG, patients received 900 mg (2 × 450 mg) per day oral valgancyclovir tablets starting within 14 days after transplantation until Day 100 posttransplantation, and in the preemptive group, quantitative CMV monitoring by means of antigenemia was performed once weekly at weeks 1–4; every 2 weeks at weeks 6–12; every 4 weeks at months 4–6; and every 3 months at months 9 and 12, or additionally as clinically indicated.
Patients in either group who tested positive (detectable CMV antigenemia higher than 20 positive cell/2 × 105 PBMC) at any time after transplantation received 1800 mg (2 × 900 mg) per day oral valganciclovir for at least 14 days, until CMV antigenemia became negative on two consecutive assessments within 1 week. Thereafter, secondary prophylaxis was given using 900 mg (2 × 450 mg) per day oral valganciclovir for 1 month. In case of CMV disease or if the patient was unable to take oral medication, intravenous ganciclovir at 2 × 5 mg/kg body weight per day was permitted. In all cases, doses of all antiviral regimens were adjusted by kidney allograft function.
Clinical data and definitions
CMV antigenemia was defined as a positive antigenemia for CMV with no symptoms. CMV disease included both viral syndrome and tissue invasive disease. Identification of the viral syndrome caused by CMV required the following: (1) positive antigenemia for CMV; (2) temperature of >38°C with no other source to account for it and (3) one of the following findings: leukocyte count of <4000/mm3, atypical lymphocytes of >3%, elevation of transaminases and platelet count of <100 000/mm. Tissue invasive disease required histopathological evidence of CMV, with or without virus culture of the tissue. This included identification of inclusion bodies or viral antigens in biopsy material or in bronchoalveolar lavage specimen cells by immunocytochemistry [3, 22-25].
Surveillance by means of CMV antigenemia was routinely performed (approximately every 1–2 weeks) during the first 3 months after transplantation in both preemptive and prophylaxis strategies. CMV antigenemia was determined in polymorphonuclear Leukocytes, obtained by dextran sedimentation, formaldehyde fixed, stained and read under a fluorescence microscope (rapid antigenemia anti-human CMV ppUL83, Argene, Varilhes, France; Ref 14-002). The maximum sensitivity of the method in our laboratory was 1 positive cell/2 × 105 PBMCs.
ELISA for CMV-IgG
CMV serostatus was determined using a commercial CMV IgG ELISA Kit (BioCheck, Inc., Burlingame, CA) according to the manufacturer's instructions.
Pools of peptides derived from a peptide scan (15 mers with 11aa overlap), covering the whole antigen length through the immediate-early protein 1 (IE-1) and through the 65 kDa phosphoprotein (pp65; Jerini Peptide Technologies, Swiss-Prot ID: P13202 and Swiss-Prot ID: P06725, respectively) of Human CMV (HHV-5), as well as a CMV lysate (Autoimmune Diagnostik®, Strasberg, Germany), were used as stimuli for the IFN-γ Elispot assay, allowing us to avoid HLA restrictions.
Anti-CMV T cell immune response assessment
IFN-γ Elispot assay
A multiscreen, 96-well filtration plate (AID®, Strasberg, Germany) coated with antihuman IFN-γ antibody (AID®, Autoimmune Diagnostika) was used. Cryopreserved PBMCs from either pretransplantation and/or 6 months after transplantation were thawed and incubated for at least 3 h at 37°C before peptide stimulation. Thereafter, 3 × 105 of lymphocytes (in a 100 μL volume) were added to each well together with each different peptide, medium alone as a negative control and with PHA (Sigma–Aldrich®, Madrid, Spain) as a positive control. All Elispot assays were carried out in triplicate. After 18 h incubation at 37°C/5% CO2, cells were removed by washing the plates four times with PBS containing 5% Tween 20 and twice with PBS. Fifty microliters of biotinylated anti-IFN-γ antibody was added (1:1000 dilution, 7-B6-1-biotin; Mabtech) and incubated for 3 h at room temperature. The Elispot plate was washed a further six times with PBS/Tween 20 and incubated for 2 h with streptavidin-ALP substrate (AID®) followed by the addition of an alkaline phosphatase conjugate substrate (50 lL; AID®, Autoimmune Diagnostika). The resulting spots were counted semi-automatically with an Elispot reader (AID® Elispot Reader HR, 4th generation). Results were expressed as percentage of cells secreting IFN-γ after subtracting the number of spots due to spontaneous IFN-γ release (measured in the control wells) from the number of spots obtained in the wells incubated with each peptide.
IFN-γ flow cytometry
Following incubation with respective peptides, PBMC were tested for intracellular IFN-γ production by the cytokine flow cytometry assay in five transplant recipients showing relevant anti-viral T cell responses against all three evaluated stimuli in the Elispot assay. PBMC were washed and stained for 30 min in ice with APC-conjugated mAb anti-CD3 (clone HIT3a; BD®, Madrid, Spain), PE-conjugated mAb anti-CD4 (clone RPA-T4; BD®) and PERP-CY.5-conjugated mAb anti-CD8 (clone RPA-T8; eBioscience®, Barcelona, Spain) in PBS + 5% FBS, containing 5% human immunoglobulin and 0.01% sodiumazide. Cells were then washed with PBS + 5% FBS, fixed and permeabilized using the FIX and PERMJ kit (BD®), according to the manufacturer's instructions, and stained for 45 min with FITC-conjugated mAb anti-IFN-γ (clone 4s.B3; BD®).
All data are presented as mean ± SD. Groups were compared using the χ2 test for categorical variables, the one-way ANOVA or t-test for normally distributed data and the nonparametric Kruskal–Wallis or Mann–Whitney U test for non normally distributed variables. Both CMV antigenemia and disease were considered the outcome variables of the study. Bivariate correlation analyses were done using Pearson or Spearman test for non-parametric variables. A sensitivity/specificity ROC curve test was done to investigate the value of the Elispot test for predicting the advent of posttransplant CMV infection. The statistical significance level was defined as 2-tailed p < 0.05.
As shown in Figure 1, 137 consecutive kidney transplant recipients were assessed for their anti-CMV T cell response before transplantation. Of these, 39 patients received posttransplant CMV prophylaxis and 98 followed a preemptive protocol. Six-month CMV-specific T cell responses could be evaluated in 58 patients, 21 receiving prophylaxis and 37 preemptive therapy. Mean follow-up of the study was 25 months (range 37–15 months).
Main demographic and baseline characteristics of all patients are depicted in Table 1. Incidence of posttransplant antigenemia or CMV disease was not different between patients receiving prophylaxis or preemptive therapy. No CMV infection events were observed beyond 6 months after transplantation. Among prophylactic-treated patients, the advent of CMV antigenemia appeared in all but one patient after completing valganciclovir treatment with a median of 45 days after stopping treatment. Within preemptive-followed patients, most CMV infection episodes occurred during the first 3 months after transplantation with a median of 38 days after transplantation. Incidence of CMV recurrence after treatment was equally distributed between both groups (three within preemptive and two among prophylaxis). Among patients receiving preemptive therapy, the advent of CMV antigenemia was significantly more common in older recipients (54.8 ± 9 vs. 48.3 ± 13, p < 0.005) and in those experiencing delayed graft function (DGF; 45.5% vs. 14.4%, p < 0.005). To note, T cell depletion induction treatment was associated to a significantly increased risk of both posttransplant antigenemia and CMV disease (63.6% vs. 38% and 80% vs. 40% for antigenemia and disease, respectively, p < 0.005). Conversely, type of maintenance immunosuppression was not associated with CMV infection. At 6 months, allograft function was significantly worse among those patients experiencing either CMV antigenemia or disease as compared to those that did not.
|All patients (N = 137)||Prophylaxis (N = 39)||Preemptive (N = 98)|
|Age (years, mean ± SD)||48.9 ± 13.2||46.4 ± 14.7||49.8 ± 12.7|
|Type of kidney TX (living/deceased)||63/74||20/19||43/55|
|Pre-TX CMV donor (D)/recipient (R) serostatus|
|R−/D+ (%)||28 (20.4)||22 (56.4)*||6 (6)|
|R+/D+ (%)||83 (60.6)||12 (30.8)||71 (72.5)|
|R+/D− (%)||26 (19)||5 (1.2)||21 (21.5)|
|CNI-based (CsA/TAC; %)||8 (6)/112 (82)||3 (8)/3 (8)||5 (5)/81 (83)|
|CNI-free (mTor-i; %)||17 (12)||5 (13)||12 (12)|
|Mycophenolate mofetil (%)||137 (100)||39 (100)||98 (100)|
|No Induction therapy (%)|
|Induction immunosuppression||10 (7.3)||1 (2.5)||9 (9.2)|
|rATG (%)||68 (49.7)||25 (6)||43 (43.8)|
|Anti-CD25 monoclonal Ab (%)||59 (43)||13 (33.4)||46 (47)|
|BPAR (%)||18 (13)||6 (15)||12 (12)|
|Allograft function (eGFR; mL/min)|
|Month 6||40.6 ± 25||45.8 ± 26||38.4 ± 24|
|Month 12||52.6 ± 15||53.4 ± 17||52.2 ± 14|
|Month 18||52.5 ± 16||53.8 ± 14||52 ± 16|
|Pre-TX anti-CMV IgG titers (UA/mL)||166.7 ± 99||99.9 ± 116*||190.7 ± 81|
|Pre-TX anti-CMV T cell response (spots/3 × 105 PBMC)|
|CMV lysate||128.9 ± 183||61.9 ± 112*||155.5 ± 198|
|Pp65 antigen||101.7 ± 168||39.4 ± 65*||126.5 ± 189|
|IE-1 antigen||39.8 ± 86.1||21.5 ± 29*||47 ± 99|
|CMV infection (antigenemia/disease)||33 (24)/18 (13)||11 (28)/8 (20)||22 (22)/10 (10)|
|Exitus (%)||8 (5.8)||2 (5)||6 (6)|
Main clinical data of patients with and without CMV antigenemia/disease within preemptive and prophylactic-treated patients are displayed in Tables 2 and 3, respectively. Most CMV infections in both cohorts of patients were asymptomatic CMV-detected antigenemia (28% and 22% in prophylactic and preemptive, respectively) and clinical disease was observed in 20% and 12% of prophylactic and preemptive groups, respectively. To note, the majority of clinical diseases were diagnosed as viral syndromes (11/18) whereas tissue invasive diseases were observed in seven patients, located in the gastro-intestinal tract and two in the pulmonary tract.
|Preemptive strategy (N = 98)||CMV antigenemia||CMV disease|
|Yes (N = 22)||No (N = 76)||Yes (N = 10)||No (N = 88)|
|Age (years, mean ± SD)||54.8 ± 9*||48.3 ± 13||53.3 ± 10||49.5 ± 13|
|Type of kidney TX (deceased/living)||6/16||37/39||1/9||42/46|
|BPAR (%)||3 (14)||9 (12)||2 (20)||10 (11)|
|Pre-TX CMV donor (D)/recipient (R) serostatus|
|R−/D+ (%)||1 (4.5)||5 (6.5)||1 (10)||5 (5.5)|
|R+/D+ (%)||17 (77)||60 (79)||7 (70)||72 (82)|
|R+/D− (%)||4 (18)||11 (14.5)||2 (20)||11 (12.5)|
|Pre-TX anti-CMV IgG titers||216.6 ± 59||183.4 ± 85||177.9 ± 85||192 ± 81|
|CNI-based (CsA/TAC; %)||0/19||5/62||0/9||5/72|
|CNI-free (mTor-i; %)||3||9||1||11|
|Mycophenolate mofetil (%)||22||76||10||88|
|Mycophenolate acid tough levels (mean ± SD)|
|Month 1||2.9 ± 1.9||2.9 ± 1.9||2.9 ± 2.1||2.9 ± 1.9|
|Month 3||2.9 ± 2||3.3 ± 2||2.7 ± 1.8||3.2 ± 1|
|Month 6||3.5 ± 2||2.7 ± 1.7||3.2 ± 1||2.8 ± 1.7|
|No induction therapy (%)||1 (4.4)||8 (10.5)||0 (0)||9 (10)|
|rATG (%)||14 (63.6)*||29 (38)||8 (80)*||35 (40)|
|Anti-CD25 monoclonal Ab (%)||7 (32)||39 (51.5)||2 (20)||44 (50)|
|Allograft function (eGFR; mL/min)|
|Month 6||28.3 ± 23*||41.2 ± 23||12.3 ± 14*||41.2 ± 22|
|Month 12||47.1 ± 17||53.8 ± 12||48.1 ± 19||53 ± 13|
|Month 18||50.1 ± 19||52.7 ± 15||53 ± 22||52 ± 15|
|Exitus (%)||3 (13)||3 (4)||1 (10)||5 (5.6)|
|Prophylaxis treatment (N = 39)||CMV antigenemia||CMV disease|
|Yes (N = 11)||No (N = 28)||Yes (N = 8)||No (N = 31)|
|Age (years, mean ± SD)||44.7 ± 15||47.1 ± 14||38.8 ± 13||48.5 ± 13|
|Type of kidney TX (deceased/living)||3/8||17/11||3/5||17/14|
|Pre-TX CMV donor (D)/recipient (R) serostatus|
|R−/D+ (%)||5 (45.5)||1 (3.5)||4 (50)||18 (58)|
|R+/D+ (%)||4 (36.4)||8 (28.5)||3 (37.5)||9 (29)|
|R+/D− (%)||2 (18.2)||3 (11)||1 (12.5)||4 (13)|
|Pre-TX anti-CMV IgG titers||90 ± 117||104 ± 118||50.1 ± 92||115 ± 120|
|Mycophenolate acid tough levels (mean ± SD)|
|Month 1||3.1 ± 1.6||3.2 ± 1.8||3.1 ± 1.4||3.2 ± 1.8|
|Month 3||3.7 ± 2||3.6 ± 2||3.4 ± 2||3.7 ± 2|
|Month 6||3.1 ± 2||3.6 ± 2.4||3.2 ±||3.1 ± 2|
|No induction therapy (%)||0 (0)||1 (3.5)||0 (0)||1 (3.2)|
|-rATG (%)||8 (73)||17 (61)||7 (87.5)||18 (58)|
|-Anti-CD25 monoclonal Ab (%)||3 (27)||10 (35.5)||1 (12.5)||12 (38.8)|
|Allograft function (eGFR; mL/min)|
|Month 6||41.7 ± 23||47.5 ± 28||48.2 ± 21||45.1 ± 28|
|Month 12||50.1 ± 11||55.3 ± 20||53.4 ± 8||53.4 ± 20|
|Month 18||48.6 ± 12||57 ± 16||53.8 ± 13||54.6 ± 15|
|Exitus (%)||1 (9)||1 (3.5)||0 (0)||2 (6)|
Pp65- and IE-1-specific T cell responses are predominantly provided by the CD8+ T cell compartment
While T cell responses against both pp65 and IE-1 CMV peptides were predominantly CD8+, CD4+ T cell responses could also be detected against the CMV lysate (Figure 2).
Low pretransplant IE-1-specific T cell responses is associated with posttransplant CMV infection
All anti-CMV T cell responses within prophylactic patients were significantly lower as compared to patients with preemptive therapy (Table 1). Pretransplant pp65 and CMV lysate but not anti-IE-1-specific T cell responses positively correlated with pretransplant CMV IgG titers (r = 0.298, p = 0.001 and r = 0.325, p < 0.001, respectively). Although pretransplant CMV-specific T cell responses could be detected among some seronegative transplants recipients (12/28), they were significantly lower than within seropositive recipients (Figure 3).
Patients receiving either preemptive or prophylaxis therapy developing CMV infection showed significantly lower anti-IE-1 T cell responses as compared to patients that did not. No association was observed between pretransplant anti-pp65 and CMV lysate T cell responses and incidence of CMV infection (Figure 4). Similar findings were observed among those patients receiving rATG (Figure 5). Furthermore, prophylaxis-treated transplant recipients developing CMV disease, did also show lower pretransplant pp65-specific T cell responses as compared to those that did not. When all patients of the study were assessed together, those with posttransplant CMV infection showed significantly lower pretransplant anti-IE-1 T cell responses than patients not experiencing CMV infection (data not shown). Patients under mTor-i did not show a different CMV-specific T cell immunity as compared to those receiving CNI-based regimens.
Frequencies of pretransplant anti-IE-1 T cell responses independently predict the risk of posttransplant CMV infection
Receiver operating characteristic curve (ROC) analysis for predicting either posttransplant antigenemia or disease in patients receiving prophylaxis and preemptive therapy is depicted in Figure 6. As shown, considerably high AUC, ranging from 0.635 up to 0.760, were obtained for pretransplant anti-IE-1 T cell responses for prediction of both CMV antigenemia and disease in the different treatment groups, respectively. Sensitivity and specificity of anti-IE-1 T cell IFN-γ Elispot is summarized in Table 4. No additive effect for predicting posttransplant CMV infection was observed when using pp65 and IE-1 T cell responses together (data not shown).
|Treatment group||Variables||Cut-off values (IFN-γ spots)||Predictive value|
|Specificity (%)||Sensitivity (%)||NPV (%)|
|Prophylaxis||Pre-TX anti-IE-1 (CMV infection)||8 spots 3 × 105 PBMCs||65||82.5||89.5|
|Preemptive||Pre-TX anti-IE-1 (CMV infection)||7 spots 3 × 105 PBMCs||55||80||95.7|
When risk of CMV infection was categorized as a binary variable, taking into account most sensitive and specific cut-off values of pretransplant IE-1-specific T cell IFN-γ Elispots for each group of transplant recipients (7 and 8 IFN-γ spots per 3 × 105 stimulated PBMCs, for preemptive and prophylaxis-treated patients, respectively) a higher than 80% and 90% sensitivity and negative predictive value were obtained, respectively (Table 4).
Anti-CMV T cell responses at 6 months after transplantation between CMV infected and noninfected transplant recipients
As all CMV infection events appeared before the first 6 months after transplantation, changes in 6-month CMV-specific T cell responses were evaluated. In general, anti-CMV T cell responses significantly increased after transplantation despite that patients were receiving immunosuppression (128.9 ± 183 vs. 278 ± 433, p = 0.012; 101.7 ± 168 vs. 127 ± 183, p = 0.006 and 39.8 ± 86.1 vs. 126 ± 454, p < 0.001, for CMV-lysate, pp65 and IE-1 for 3 × 105 stimulated PBMCs, respectively). There were no differences between 6-month anti-CMV T cell responses among patients having received prophylaxis or preemptive therapy, CNI or non-CNI-based immunosuppression or different type of induction therapy (data not shown). However, when patients with or without posttransplant CMV infection were compared regarding their change in the CMV-specific T cell response at 6-month, patients having experienced CMV infection showed a significantly increase in pp65 and IE-1-specific T cell responses as compared to those that did not (Figure 7).
While current clinical immune assessment of the CMV risk of infection before transplantation exclusively relies on donor and recipient CMV IgG-serostatus, our study shows that CMV-specific T cell response, particularly against the IE-1 dominant CMV antigen, may improve the identification of those kidney allograft recipients at high-risk for CMV infection. Importantly, our approach is capable to discriminate such patients already before transplantation, with high sensitivity and specificity, regardless the type of preventive strategy used. Furthermore, the high negative predictive value of the test highlights the usefulness of such approach.
Noteworthy, our study shows that monitoring IE-1 CMV-specific T cell frequencies before transplantation would help transplant physicians on the one hand to better discriminate those patients with no need of CMV prophylactic treatment from those in whom prophylaxis should preferentially be indicated and on the other to better predict those patients in whom prophylaxis treatment could safely be discontinued. Interestingly, intrinsic impairment of the IE-1-specific T cell response was not only associated with the advent of posttransplant CMV infection but also with the development of CMV disease, thus reinforcing the importance of such functionally active CMV-specific T cell precursors for achieving CMV control under immunosuppression.
The observation that patients receiving T cell depleting antibodies experiencing CMV infection were those with significantly lower pretransplant IE-1-specific T cell frequencies, suggests that the increased susceptibility for CMV infection after T cell depletion is particularly facilitated by the impairment of IE-1-specific T cell precursors already before transplantation rather than to a generalized T cell subset depletion after rATG therapy. Differently from what has been shown among normal individuals , within our seronegative group of chronic kidney disease patients, CMV-specific T cell responses were also detectable in a group of them, though at significantly lower frequencies than among seropositive patients. Nevertheless, only patients with adequate pretransplant anti-IE-1-speciifc T cell frequencies were at significant low-risk for CMV infection. This finding supports the notion that although CMV triggers both humoral and cellular responses, only the latter and particularly that directed to IE-1 CMV antigens seem to be crucial for posttransplant viral replication control, therefore being the former of limited utility in the clinical practice [27, 28]. Nonetheless, whether the detection of CMV-specific peptide T cell responses among CMV-seronegative patients could result from cross-reactive recognition of CMV epitopes by memory T cells originated from distinct (e.g. non-HCMV) antigenic exposures or if a more accurate assessment of CMV-specific memory B-cell IgG frequencies would increase the sensitivity to detect patients already sensitized to CMV antigens deserves further evaluation.
To date, studies in transplant recipients evaluating the impact of CMV-specific cellular responses have mainly focused at the posttransplant period and used different cellular immune assays with distinct CMV stimuli [21-26, 29-32]. Our study is in consonance with these previous reports, but also shows that the increased risk to develop posttransplant CMV infection (even after a course of prophylactic treatment) seems to rely in an individual immune susceptibility already manifested prior to transplantation. Likewise, but in lung and heart transplant patients, Bunde et al.  showed that frequencies of IE-1 but not pp65-specific CD8+ T cells already at Day 0, discriminated patients who did not develop CMV disease from patients at risk. Although focusing on the association between allogeneic and CMV-specific effector T cell responses, Nickel et al.  reported similar findings in a group of 36 kidney transplant patients.
Although different studies have suggested a preponderant role of CD8+ T cells for CMV control [21, 24, 26, 29, 30], others have also shown the concomitant key function of CD4+ T cells, which seem to even have a preferential role conferring long-lasting protection [29, 32]. In our study, we found that pp65 and IE-1-specific T cell responses are predominantly but not exclusively restricted to CD8+ thus, CD4+ T cells responses could similarly be required to confer long-term protection against CMV infection.
Even though T cell responses may target multiple CMV-specific proteins [18, 34, 35], it appears that protective cellular immunity is mainly directed against the tegument protein ppUL83 and to the immediately early protein ones [19, 21, 33, 36]. To note, IE-1 is the first protein expressed upon CMV reactivation , thus IE-1-specific T cells would be the first to be activated and directed to sites of replication [38, 39]. Hence, this mechanism could explain why high levels of IE-1 but not other CMV-specific T cells would be associated with protection from CMV disease. Some other groups have shown lack of correlation with exclusive IE-1-specific T cell responses and risk of CMV disease [40, 41]. To note, most of them focused at the posttransplant setting and evaluated a rather low number of transplant recipients. In our study, at 6 months while there was a general increase of all CMV-specific T cell responses (both against IE-1 and also pp65), this feature was specifically observed within those having recovered from CMV infection, suggesting that broader CMV-antigen specific T cell responses might be also necessary for controlling CMV replication. This data reinforces the potential value of preventive strategies using recombinant CMV proteins as vaccines, preferentially containing immunogenic IE-1 antigens already before transplantation.
There are some limitations in this study. First, although we used a non-standardized immune assay, the IFN-γ Elispot has already been shown to be highly reproducible for measuring antigen-specific cellular responses in other relevant fields of medicine [42-44], allowing a comprehensive quantitative-dynamic idea of the antigen-specific cellular strength at a single cell level. Another limitation is the lack of PCR-CMV viremia monitoring in our study that could have induced misleading diagnosis. Nonetheless, although PCR-CMV viremia has shown higher sensitivity as compared to CMV antigenemia , the incidence of CMV antigenemia or disease among our two cohorts of kidney transplant recipients fitted with that reported in the literature using PCR-based assays [46, 47].
In conclusion, we have shown that monitoring frequencies of IE-1-specific T cell responses before transplantation may be useful for predicting posttransplant risk of CMV infection, thus being potentially valuable for guiding decision-making regarding CMV preventive treatment. To further support this result and validate its potential clinical utility, large-scale prospective randomized trials are highly warranted and should be preferentially performed in the context of multicenter cooperative networks.
The authors gratefully acknowledge all our Nephrology staff for their clinical assistance in this study.
Funding sources: This work was supported in part by a grant from Fundación Mutua Madrileña (2008–2011), two national public grants from Instituto de Investigación Carlos III (PI10/01786) and the Red de Investigación Renal (REDinREN, ISCIII 06/0016) and by a European Commission grant within the RISET consortium (EU 512090).
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.