Choice of Antibody Immunotherapy Influences Cytomegalovirus Viremia in Simultaneous Pancreas-Kidney Transplant Recipients Free

Thirty-nine consecutive patients received SPK transplants at the Leiden University Medical Center between October 1999 and May 2002. In all patients duodenocystostomy was used for exocrine drainage of the pancreatic graft. Patients were randomly assigned to receive either a single dose of ATG_F (9 mg/kg) intraoperatively or five consecutive doses of daclizumab (1 mg/kg) administered in 2-week intervals, starting before transplantation. Relevant patient characteristics were comparable between groups. No differences in clinical outcome were observed between either induction protocols or occurrence of CMV viremia with regard to transplant survival, insulin independence, and cumulative numbers of rejection episodes (Table 1). From 36 patients, sufficient plasma samples could be collected for the CMV DNA quantification used in this study. Two patients lost their pancreas graft at an early stage (3 and 4 days after transplantation, respectively) due to technical complications (venous graft thrombosis), and one patient died with functioning grafts 70 days after transplantation.

CMV serostatus of both donor and recipient was determined before transplantation. Patients at risk for CMV infection (based on donor [D]/receptor [R] serostatus: D+/R−, D+/R+, or D−/R+) received antiviral prophylaxis (1,000 mg ganciclovir orally three times daily for 3–4 months) starting within 14 days after transplantation. Maintenance immunosuppression in all patients consisted of cyclosporin A microemulsion (Neoral) with dose adjustments based on trough level monitoring, mycophenolate mofetil 1,000 mg twice per day, and prednisolone, which was gradually tapered to 10 mg/day by 3 months. Clinical rejection episodes were treated with high-dose intravenous steroids (Solu-Medrol 1,000 mg/day for 3 consecutive days). Recurrent or steroid-resistant rejection episodes were treated with a 10-day course of ATG_M (starting at 4 mg/kg), with subsequent dosing guided by absolute lymphocyte counts in peripheral blood.

EDTA plasma samples were collected at a frequency of about once a week for at least 180 days after transplantation and stored at −80°C until further processing. Nucleic acids were extracted from 0.2-ml plasma samples with the automated purification procedure of the MagNA Pure LC system (Roche Molecular Systems, Almere, the Netherlands) using the total nucleic acid isolation kit. Subsequently, CMV DNA quantification was performed using an internally controlled real-time quantitative CMV PCR. Sensitivity, specificity, and reproducibility of this assay were described in more detail previously (27). The course of CMV DNA load in plasma was documented longitudinally for each patient within 180 days of follow-up. Individual areas under the CMV viremia curves between 0 and 180 days after transplantation were calculated using the trapezoidal rule as described previously (25,28).

HLA-A2–restricted, CMV-specific phytoerythrin-labeled tetramers have been shown to be a valuable tool both for the detection of cytotoxic lymphocytes directed against CMV and potentially for diagnostic use (29). Blood from 16 HLA-A2–positive SPK transplant recipients was drawn and heparinized cross-sectionally 1–2 years after transplantation. Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll density gradient centrifugation and washed in 0.9% phosphate-buffered saline. One million cells were incubated in PBS containing 0.1% FCS at room temperature for 30 min with a CMV-specific tetramer developed in our lab. Cells were washed and stained with fluorescein isothiocyanate–labeled anti-CD3 monoclonal antibody (BD Biosciences, Oxford, U.K.) and allophyocyanin-labeled anti-CD8 monoclonal antibody (BD Biosciences) for 20 min at 4°C. After washing, fluorescence was measured immediately using a FACScan (BD Biosciences). Cells were analyzed using CellQuest software (BD Biosciences), measuring the percentage of CMV-specific cells in the CD3⁺/CD8⁺ living cell population.

Two-tailed Fisher’s exact test was used to determine differences between serologic groups. Disease-free survival data were presented as Kaplan-Meier survival curves with log-rank analysis and Cox proportional hazard regression to determine differences in survival. Differences in total viral load and T-cell counts were measured using nonparametric Mann-Whitney U tests, assuming non-Gaussian distribution.

With regard to the pretransplantation CMV serostatus of donor and recipient among the 36 SPK transplant recipients, 9 were D+/R−, 7 were D+/R+, and 6 were D−/R+. CMV viremia was detected in 13 of 16 patients (81%) receiving seropositive donor organs, compared with 2 of 20 patients (10%) receiving seronegative donor organs (P < 0.0001) (Table 2). In contrast, no significant difference was seen for the incidence of CMV viremia in seropositive recipients versus seronegative recipients (7 of 13 and 8 of 23, respectively). Regarding the serologic groups at risk for CMV, D+/R− patients tended to develop more CMV viremia, whereas D−/R+ patients showed a trend toward a reduced risk of CMV viremia compared with the other at-risk groups.

The two different antibody induction therapies were compared with regard to the moment CMV viremia occurred. CMV viremia was defined as detection of two consecutive CMV DNA loads of more than ₁₀log 2.7 (= 500) copies/ml plasma. In the total population, a trend was noted toward shorter CMV-free survival in the ATG_F-treated than in the daclizumab-treated patients (P = 0.10). Considering the population at risk for CMV infection (n = 22, D−/R− excluded), CMV-free survival was significantly shorter in the ATG_F group (P = 0.04) (Fig. 1A). Both patient groups were comparable regarding age, sex, incidence of rejection, and CMV serostatus. The median area under the viremia curve tended to be higher in the ATG_F group (Fig. 1B), indicating more severe CMV viremia.

In both groups, a number of patients received a 10-day course of ATG_M rejection treatment, influencing CMV load (see rejection treatment results below). Excluding these patients from the induction group analysis did not influence patient group characteristics, and shorter CMV-free survival (P = 0.01) and more severe infection (P = 0.05) were seen in the ATG_F compared with the daclizumab group (Fig. 1C and D, respectively).

Next, the correlation between rejection episodes treated with ATG_M and CMV viremia in the patient group at risk for CMV was assessed. One patient was excluded from this analysis because he received only Solu-Medrol as rejection treatment. Figure 1E shows the disease-free survival curves for patients receiving ATG_M rejection therapy versus patients without rejection episodes. A significantly shorter disease-free survival was seen in the ATG_M rejection therapy group (P = 0.02). In these patients, CMV viremia occurred after administration of rejection treatment, except for one patient in whom detection of CMV coincided with rejection treatment. Total viral load as measured by the area under the curve from 0 to 180 days was higher (P = 0.01) than in patients without rejection episodes (Fig. 1F).

Cox proportional hazard regression identified both ATG_M rejection therapy and ATG_F induction therapy as independent risk factors for shorter CMV-free survival (ATG_M hazard ratio 6.191 [95% CI 1.792–21.393], P = 0.004; ATG_F 5.447 [1.598–18.564], P = 0.007).

To further investigate the mechanism underlying the pattern of CMV viremia in this patient group, HLA-A2–restricted CMV-specific tetramer fluorescence-activated cell sorter staining was performed on PBMCs of 16 HLA-A2⁺ patients. Several patients showed distinct populations of CMV-specific cells in the CD3⁺/CD8⁺ T-cell population. In the patients at risk, a trend was noted toward a higher percentage of CMV-specific CD8⁺ T-cells in the daclizumab-treated group compared with the ATG_F-treated group (Fig. 2A). When we stratified for CMV viremia, a significantly lower percentage of CMV-specific CD8⁺ T-cells was seen in patients who developed CMV viremia (P = 0.01) (Fig. 2B). To test the possibility of an ongoing infection at the time of blood withdrawal for isolation of PBMCs, the serum samples were analyzed for CMV viremia. No CMV DNA was detected in any of the samples (not shown). As a further control, PBMCs from HLA-A2⁺ patients not at risk for CMV infection (D−/R−) were stained, showing no CMV specificity at all (Fig. 2B).

In this study, it is shown that CMV viremia not only occurred earlier but was also more severe in SPK transplant recipients receiving single-shot ATG_F induction therapy compared with five-dose daclizumab and after rejection episodes treated with a 10-day course of ATG_M. Despite the limited number of patients included in the study, several potentially clinically relevant differences were found to be significant. In our study, we aimed to compare two different, but well-established, induction protocols. Although variations in timing and dosage conceivably affect the clinical outcome, this was not the subject of our studies because these variables are inherent to the protocols of choice.

The impact of donor pretransplant CMV serology clearly shows from these data. Patients receiving an organ from a seropositive donor had a much higher chance of developing CMV viremia than those receiving an organ from a seronegative donor. Remarkably, no direct influence of the patient’s own pretransplant serology was noted. In the past, several studies have shown a higher risk for the development of CMV infection for patients who were de novo infected as a result of the transplantation (D+/R−) (17). In our patient group, only a trend in that direction was noted, conceivably due to the limited number of patients. Knowledge of pretransplant serology and subsequent adequate action could significantly decrease the risk of CMV infections. This is already being achieved by serological matching (positive organs to positive recipients and negative organs to negative recipients) (30). Unfortunately, donor shortage and limited ischemia times are restricting factors for the matching strategy. Another possible strategy would be to determine the immunosuppressive protocol individually for each patient based on CMV serology status.

Furthermore, this study stresses the need for careful monitoring of infections in patients treated with polyclonal ATG therapy. Antibody induction therapy for transplantation has become regular practice in recent years and in particular with SPK transplants (6). We conclude that antibody induction therapy with daclizumab (anti-CD25) is safer than antibody induction therapy with ATG_F regarding (re)activation of CMV in SPK transplant recipients because CMV viremia occurs later and the total viral load is lower. When patients receive ATG as rejection treatment, the effect on CMV viremia is even more pronounced. These findings are in accordance with findings in kidney transplant recipients (14) and can be explained by the proposed mechanisms through which both agents affect the immune system. Daclizumab treatment is said to affect activated T-cells only, thus leaving memory T-cell function relatively intact, whereas ATG profoundly depletes all T-cells, conceivably leading to a longer-lasting influence than with daclizumab (10,15). Nonetheless, in recent reports on nondepleting humanized anti-CD3 therapy in type 1 diabetes, it was suggested that modulation of T-cells can preserve β-cell function (31,32). The latter, however, was not the subject of our present studies.

Our findings are of importance because it is known that the consequences of CMV disease for morbidity and transplant survival are strongest in the first months after transplant. Furthermore, CMV disease indirectly affects transplant survival (33). In this study, however, none of the patients developed clinical CMV disease.

Tetramer staining for CMV-specific CD8⁺ T-cells gives additional insight into the mechanisms underlying the noted differences. The occasional high amounts of CMV-specific cells corresponded with absence of CMV viremia both in the first 6 months and at the time of staining rather than reflecting an ongoing infection. All three patients not developing CMV viremia (and with high CMV-specific T-cell counts) were treated with daclizumab, and, interestingly, the one patient developing CMV viremia in the daclizumab group had a low CMV-specific T-cell count. These findings suggest that having high CMV-specific tetramer counts is actually beneficial, rather than a surrogate for viremia, because they are correlated with low viremic activity after transplantation. In this respect, tetramer staining might become an important tool to prospectively identify patients at high risk for CMV infection in the future (29).

Although the number of patients limits definite conclusions, this study emphasizes the important role for cellular immunity in the prevention of CMV viremia after SPK transplantation and subsequently the impact antibody therapy has on the protective cytotoxic capacity of the immune system. With daclizumab induction therapy, this impact seems to be less vigorous than with ATG_F. Moreover, these results argue in favor of the use of daclizumab as induction therapy for pancreas and islet transplantation because of the reported potentiating effect of CMV on recurrent autoimmunity (23).

CMV disease in islet transplantation has not yet been studied extensively, but because recurrent autoimmunity may be an important reason for the long-term loss of islet allografts (34), such studies are warranted. This recommendation also applies to trials in which immunosuppressive agents are used to try to halt type 1 diabetes early in the course of the disease. For pancreas-kidney transplantation, it can be concluded that the differences between daclizumab and ATG_F induction on CMV infection are relevant when choosing a certain induction or rejection therapy, considering that no difference in immunosuppressive potential has been noted.

This study was supported by grants from the Juvenile Diabetes Research Foundation (42001-434) and the Dutch Diabetes Research Foundation (2001.06.001).

We thank Odette Tysma for expert technical assistance and Dr. Aan Kharagjitsingh for statistical advice.