Abstract
A total of 1,305 blood samples from 85 solid organ transplant (SOT) recipients and 25 stem cell transplant (SCT) recipients at risk for cytomegalovirus (CMV) infection were prospectively collected and tested using the shell vial assay (SVA) and a leukocytic qualitative PCR (q-PCR). Of these, 462 specimens were further tested by direct quantification of CMV antigenemia by flow cytometry (FC-Ag), 125 were tested with a quantitative competitive PCR, and 200 were tested for pp65 antigenemia using the slide method (S-Ag). Laboratory data were statistically analyzed according to the presence of CMV-related symptoms. In SOT and SCT recipients, active CMV infection occurred in 63.5 and 36%, respectively, and CMV disease occurred in 53 and 24%, respectively. FC-Ag results correlated better with q-PCR and S-Ag than with SVA. The first test found to be positive during follow-up was FC-Ag in 73% of cases. In SOT recipients, FC-Ag showed the highest sensitivity and negative predictive value for the diagnosis of any grade of CMV disease. For FC-Ag, the threshold beyond which CMV disease was highly probable seemed to lie at 0.20% positive polymorphonuclear leukocytes. FC-Ag appears to be a useful test for the early detection of CMV infection and the prediction of CMV disease.
Human cytomegalovirus (CMV) infection remains a frequent and sometimes life-threatening complication of immunosuppressed states, notably during immunosuppression after solid organ (SOT) or allogeneic stem cell (SCT) transplantation (28). The efficacies of several different strategies for anti-CMV treatment (delayed, prophylactic, and preemptive) have been demonstrated (10, 14, 17, 24). The use of such strategies hinges on the singling out of a virological tool capable of predicting the development of CMV disease as accurately as possible, while also allowing active infection to be detected sufficiently early (for the preemptive therapeutic approach). The choice of technique has to take into account not only the qualities of the selected method for each different context in which immunosuppression occurs but also its ease of use in routine laboratory testing. Recent studies have shown that the intensity of viral load is correlated with the occurrence of CMV-related symptoms (reviewed by Boeckh and Boivin [3]). pp65 antigenemia is one of the most widespread methods for quantifying CMV viral load, but the conventional procedures used, indirect immunofluorescence and quantification by fluorescence microscopy, are relatively time-consuming and the interpretation of slides involves a subjective element (2, 3, 5). In an effort to overcome these disadvantages, we and others have developed methods for the direct quantification of CMV antigenemia by flow cytometry (FC-Ag). Initial results have shown such methods to be feasible and rapid, and they have the great advantage that quantification is objective (7, 19, 22). The present large-scale prospective study was designed to evaluate the usefulness of FC-Ag as a test for the early detection of CMV infection and for quantifying viral load with a view to predicting CMV disease in immunocompromised posttransplant patients.
MATERIALS AND METHODS
Study population.
One hundred and ten consecutive patients undergoing allograft transplantation at Nantes University Hospital and at risk of CMV active infection were prospectively enrolled in this study during the period May 1996 to June 1998 (Table 1). For the 85 patients who underwent SOT, immunosuppressive treatment consisted of variable triple-drug therapy regimens according to the different ongoing protocols in use in the unit. Sixteen patients received a combination of prednisolone, azathioprine, and cyclosporine; 63 received prednisolone, mycophenolate mofetil (MMF) (2 g/day), and cyclosporine (including 20 patients who also received antithymocyte globulin [ATG] as induction therapy); and 6 received MMF combined with other treatments. The attribution of grafts was independent of the CMV serological status of both donor and recipient. Twenty-five patients received allogeneic SCT. Patients were prepared for transplantation with total body irradiation and cyclophosphamide. Prophylaxis of graft-versus-host disease consisted of a standard association of cyclosporine and methotrexate, and active graft-versus-host disease was treated with steroids (2 mg/kg of body weight per day). All CMV-seronegative recipients received blood transfusions from seronegative donors.
TABLE 1.
Characteristics of study population
| Characteristic | SOT recipient (n = 85) | SCT recipient (n = 25) |
|---|---|---|
| Type(s) of graft (n) | Renal (79)/renal-pancreas (6) | BM (19), PBSCa (6) |
| Mean age (yr) (range) | 46.2 (11–70) | 39 (25–52) |
| No. of men/no. of women | 53/32 | 12/13 |
| Graft no. | ||
| 1 | 72 | 24 |
| 2 | 12 | 1c |
| 3 | 1 | |
| CMV serological statusb | ||
| D+, R+ | 19 | 6 |
| D−, R+ | 17 | 2 |
| D+, R− | 42 | 8 |
| D?, R+ | 7 | 9 |
| Median (range) days of posttransplant follow-up | 117 (42–217) | 152 (42–304) |
| Median (range) no. of samples analyzed/patient | 11 (5–18) | 15 (4–28) |
BM, bone marrow cells; PBSC, peripheral blood stem cells.
Donor (D) and recipient (R) CMV status: +, positive; −, negative; ?, unknown.
First one was autologous.
Virological follow-up.
Follow-up for CMV infection was conducted using shell vial assay (SVA) and leukocytic qualitative PCR (q-PCR) weekly during the first 2 postoperative months in all patients and then twice a month, until the fourth month in SOT recipients and until the fifth in SCT recipients. When either SVA or q-PCR was positive, FC-Ag was carried out on at least one sample preceding the first instance of positivity and subsequently until all CMV tests were negative. A total of 1,305 blood samples (923 and 382 for SOT and SCT recipients, respectively) were collected, and FC-Ag was carried out on 462 samples: 378 of these were from 63 infected patients (54 SOT recipients and 9 SCT recipients), and the remaining samples (n = 84) were obtained between day 20 and day 60 following transplantation from 24 uninfected patients (16 SOT recipients and 8 SCT recipients). A quantitative competitive PCR (QC-PCR) was also carried out on 125 samples (110 from 32 SOT recipients and 15 from 9 SCT recipients) from those found to be positive by q-PCR. The corresponding FC-Ag data were available in all cases. Finally, with a view to comparing FC-Ag with the slide method (S-Ag), which is currently the technique most widely used for the detection of antigenemia, 200 samples (186 from 50 SOT recipients and 14 from 7 SCT recipients) were also evaluated using the latter methodology. Most of these samples were from patients with active CMV infection.
Antiviral therapy.
No CMV prophylaxis was given to SOT recipients, whereas 16 of 25 CMV-seropositive SCT recipients received high doses of acyclovir (500 mg/m2, three times daily) during the first month after transplantation (31).
The medical teams in charge of the patients were not aware of the viral load results. In SOT recipients, ganciclovir (GCV) therapy (5 mg/kg twice daily for 15 days, adjusted according to renal function) was thus initiated in accordance with clinical data and current laboratory tests, i.e., positive viremia or DNAemia. Two patients with very severe disease received foscavir (FCV) as a third course of antiviral therapy (2 g twice daily for 15 days). In SCT patients, GCV (5 mg/kg twice daily for 15 days) or FCV (3 g twice daily for 5 days and then 3 g daily for 10 days) was started following the second instance of positive CMV DNAemia (21).
Definitions.
Active infection was defined as the finding of positive viremia or DNAemia at least once during follow-up.
CMV disease was defined as the observation of the following symptoms during active infection, and was scored according to different grades of severity (20). In all patients, absence of CMV-related symptoms was scored as grade 0. In SOT recipients, grade I was defined as the occurrence of unexplained fever (>38°C) for 2 or more days or of leukopenia (white blood cell count < 3.5 × 109/liter); grade II was defined as the occurrence of both fever and leukopenia; and grade III was defined as the occurrence of fever, leukopenia, and one or more of the following: thrombocytopenia (platelet count < 100 × 109/liter), raised liver enzymes (alanine transaminase concentration >50 U/liter), raised amylase with or without abdominal pain, unexplained asthenia, pneumonitis associated with isolation of CMV from bronchoalveolar lavage (BAL) fluid when no other cause was apparent, retinitis with a positive PCR using aqueous humor, and encephalitis with a positive PCR using cerebrospinal fluid. In SCT recipients, grade I was defined as the occurrence of leukopenia, and grades II and III were defined as the occurrence of pulmonary signs, whether mild (arterial oxygen percent saturation > 0.90) or severe (requiring O2 therapy), together with the isolation of CMV from BAL fluid or the occurrence of digestive tract involvement with at least one episode (grade II) or recurrent episodes (grade III) of diarrhea, associated with the isolation of CMV from biopsy specimens.
CMV tests.
Viremia was determined from buffy coat cells isolated from 10 ml of heparinized blood by shell vial culture, using the anti-CMV monoclonal antibody (MAb) E13 (Argène-Biosoft, Varihles, France) (22). BAL or biopsy specimens were inoculated on MRC-5 monolayers and observed over 4 weeks for any cytopathic effect.
Qualitative and quantitative PCRs were performed on DNA extracted from a 200-μl suspension adjusted to 2.5 × 106 polymorphonuclear leukocytes (PMNLs)/ml obtained from EDTA blood samples as previously described (30). Results were expressed as the logarithm of the number of copies of viral genome per 106 PMNLs. The threshold for the PCR was defined as 2.7 log genome copies/106 PMNLs. In q-PCR, inhibitors were detected by the addition of a low concentration of the internal standard used in QC-PCR (10 copies).
FC-Ag was carried out using a procedure developed in the laboratory (22) involving sequential incubation with 1% paraformaldehyde (Merck, Nogent sur Marne, France) and cold 80% methanol for the fixation and permeabilization of white blood cells, followed by 1C3 anti-CMV MAb (Argène-Biosoft) or related-isotype immunoglobulin G1 control (Caltag/Tebu, Marnes la Coquette, France). After the fixation-permeabilization step, cells were frozen at −20°C (cells can be kept at −20°C at least 1 month without any influence on the test results [34; personal observation]). Results were expressed as the percentage of PMNLs carrying pp65 antigen. The threshold for positivity was determined in our previous study as being 0.05% of positive PMNLs (22).
Conventional S-Ag was performed using the CMV CINAKIT (Argène-Biosoft) according to the manufacturer's instructions. Cells positive for CMV pp65 antigen were recognized and counted by fluorescence microscopy after incubation of 2 × 105 PMNLs with a mixture of 1C3 and AYM-1 MAbs (recognizing two different epitopes of the lower matrix phosphoprotein pp65).
Statistical analysis.
Data were analyzed using SAS (Cary, N.C.) software. Univariate analyses were carried out using either the chi-squared or Fisher exact test for proportion and Student's t test or the Duncan test for means. Multivariate analyses were performed using logistic regression.
CMV laboratory tests were compared using the kappa test (qualitative results) or the linear correlation coefficient (quantitative results).
In all cases, the confidence interval was 95%; values of P of <0.05 were considered to be significant.
RESULTS
Population characteristics (i) SOT recipients.
Active CMV infection (as defined in Materials and Methods) occurred in 54 of 85 patients (63.5%), and CMV disease occurred in 45 of 85 patients (53%). Grades I, II, and III occurred in 12, 7, and 26 patients respectively. Eighteen patients experienced one relapse of active infection, and one patient experienced two relapses. One patient remained actively infected until death.
Thirty-eight of the 54 actively infected patients were treated with GCV for CMV disease (all patients with severe disease and more than half of patients with moderate or mild disease). Of these, 30 received only one course of antiviral therapy and 8 received two courses. FCV was administered as a third-line antiviral treatment in two patients. Of the 16 other infected patients, active infection resolved spontaneously in 9 and after tapering-off of immunosuppressive drugs in 7. Only administration of ATG was identified as a risk factor for active CMV (P = 0.0003; RR = 1.82 [range, 1.42 to 2.35]).
(ii) SCT recipients.
Of the 25 SCT recipients, 9 developed at least one episode of active CMV infection. CMV disease was observed in six patients (three grade I and three grade III). No difference in these incidences was noted according to whether or not prophylactic treatment had been given.
Six patients whose PCRs were positive on two consecutive occasions were treated with either GCV (n = 5) or FCV (n = 1). There was a relapse of active infection in three patients, associated in two cases with severe organ involvement and requiring a second course of GCV (n = 2) or FCV (n = 1) and a third course of GCV in one patient.
Virological data.
Of the 1,305 blood samples analyzed by both SVA and q-PCR, 108 (8.3%) and 198 (15.2%), respectively, yielded positive results. FC-Ag was positive for 217 of 462 samples, with the viral load ranging from 0.05 to 4.91% positive PMNLs (mean, 0.23%). S-Ag was positive in 52 of 200 specimens (range, 1 to 900 positive cells/2 × 105 PMNLs; mean 61 cells). DNA viral load as estimated in 125 specimens by QC-PCR ranged from 2.7 to 5.7 log/106 PMNLs (mean, 3.5 log). Comparison of the intensities of viral load as estimated by FC-Ag with QC-PCR (n = 125 samples) and with S-Ag (n = 200 samples) showed linear coefficients of correlation of R = 0.38 (P < 0.0001) and R = 0.76 (P < 0.0001), respectively (Fig. 1).
FIG. 1.
Comparison of viral load intensity estimated by FC-Ag and by S-Ag (a) QC-PCR (b).
Concordance and agreement between FC-Ag, as evaluated in terms of positive (>0.05% positive PMNLs) or negative result, and the three other CMV tests are presented in Table 2. FC-Ag results correlated better with PCR and S-Ag than with SVA. In concordant positive cases (n = 159), i.e., when a positive FC-Ag result corresponded to positive SVA and/or q-PCR results, mean viral load intensities were 0.24% positive PMNLs, 3.6 log/106 PMNLs, and 84 positive cells/2 × 105 PMNLs, as determined by FC-Ag, QC-PCR, and S-Ag, respectively. In all cases of discrepancy between FC-Ag and the other tests, low levels of CMV were present, i.e., less than 0.3% positive PMNLs as evaluated with FC-Ag, equal to or less than 3 log/106 PMNLs as evaluated with QC-PCR, and less than six positive cells/2 × 105 PMNLs as evaluated with S-Ag. These discrepancies were mainly observed for patients with active CMV infection during very early or late phases of the episode of infection. Among the 24 of 47 patients negative by both q-PCR and SVA, six (nine samples) were positive by FC-Ag. Four of these patients had CMV-related symptoms or a concomitant positive result by S-Ag (see Table 5) [patient 7]).
TABLE 2.
Concordance and kappa coefficient agreement between the results of FC-Ag and other CMV assays
| Assay and results | No. of samples | Results of FC-Ag
|
Total concordancea | Agreementb | ||
|---|---|---|---|---|---|---|
| Positive | Negative | NDc | ||||
| Viremia | 0.63 | 0.276 | ||||
| Positive | 108 | 72 | 18 | 18 | ||
| Negative | 311 | 124 | 168 | 19 | ||
| Toxic effect | 83 | 21 | 59 | 3 | ||
| DNAemia | 0.795 | 0.586 | ||||
| Positive | 198 | 139 | 26 | 33 | ||
| Negative | 291 | 69 | 209 | 13 | ||
| PCR inhibitor | 22 | 9 | 10 | 3 | ||
| S-Ag | 0.795 | 0.496 | ||||
| Positive | 52 | 36 | 16 | |||
| Negative | 148 | 25 | 123 | |||
Number of concordant results/(number of concordant results + number of discordant results). Calculations were made after elimination of untested samples or those with toxic effect or PCR inhibitor.
Kappa coefficient.
ND, not done.
TABLE 5.
Virological and clinical follow-up of seven patients with CMV disease
| Patient no. | Sexc | Age (yr) | Type of graft | Immunosuppressive regimen | CMV statusa
|
Day(s) posttransplant | Biological diagnosisa
|
Grade of CMV disease | Antiviral therapy (days of treatment) | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Donor | Recipient | Viremia | DNAemia | FC-Ag (% positive PMNLs) | ||||||||
| 1 | M | 35 | Renal | MMF | + | − | 20 | − | − | 0.07 | ||
| 28 | − | + | 0.15 | |||||||||
| 35 | + | + | 0.42 | III | GCV (35–49) | |||||||
| 39 | − | + | 0.44 | |||||||||
| 42 | − | + | 1.67 | |||||||||
| 49 | − | + | 0.27 | |||||||||
| 54 | + | + | NDb | |||||||||
| 63 | + | + | ND | |||||||||
| 90 | + | + | 0.39 | III | GCV (90–104) | |||||||
| 112 | − | − | <0.05 | |||||||||
| 2 | M | 36 | Renal | MMF | + | − | 0–14 | − | − | <0.05 | ||
| 21 | + | + | 0.35 | |||||||||
| 30 | + | + | 0.13 | |||||||||
| 35 | + | + | 0.19 | GCV (35–49) | ||||||||
| 42 | − | + | 0.66 | II | ||||||||
| 50 | − | + | 0.08 | |||||||||
| 54 | + | + | 0.18 | |||||||||
| 61 | + | + | 1.81 | II | ||||||||
| 72 | + | + | 0.30 | |||||||||
| 83 | − | + | <0.05 | |||||||||
| 99 | − | − | <0.05 | |||||||||
| 3 | F | 36 | Renal | MMF | + | − | 20 | − | − | <0.05 | ||
| 27 | − | − | 0.05 | |||||||||
| 40 | + | + | 0.54 | III | GCV (40–54) | |||||||
| 48 | − | + | 0.05 | |||||||||
| 50 | − | − | 0.06 | |||||||||
| 69 | − | − | <0.05 | |||||||||
| 4 | M | 53 | Renal | MMF + ATG | + | − | 0–13 | Toxic effect | − | <0.05 | ||
| 20 | − | − | 0.07 | |||||||||
| 29 | + | + | 0.05 | III | GCV (23–37) | |||||||
| 42 | − | + | 0.24 | |||||||||
| 46 | − | + | 0.09 | |||||||||
| 54 | + | + | ND | II | GCV (52–77) | |||||||
| 70 | Toxic effect | − | 0.20 | |||||||||
| 84 | − | − | 0.23 | |||||||||
| 98 | + | + | 0.19 | III | GCV (100–115) | |||||||
| 112 | Toxic effect | − | <0.05 | |||||||||
| 124 | − | − | <0.05 | |||||||||
| 5 | M | 39 | Second renal | MMF | + | + | 13 | − | − | 0.13 | ||
| 31 | − | + | 0.05 | |||||||||
| 38 | + | + | Technical problem | |||||||||
| 48 | − | + | 0.11 | III | GCV (48–62) | |||||||
| 52 | − | + | 0.18 | |||||||||
| 55–78 | − | − | <0.05 | |||||||||
| 6 | F | 29 | Bone marrow | ? | + | 149 | − | − | <0.05 | |||
| 158 | − | − | <0.05 | |||||||||
| 163 | + | + | 0.06 | I | FCV (167–177) | |||||||
| 167 | + | Inhibitor | 0.07 | |||||||||
| 174 | − | − | <0.05 | |||||||||
| 177 | − | − | 0.05 | |||||||||
| 192 | − | − | 0.05 | GCV (192–206) | ||||||||
| 212 | − | − | 0.30 | III | ||||||||
| 220 | − | − | ND | |||||||||
| 230 | − | + | 0.09 | III | GCV (230–244) | |||||||
| 237 | − | + | 0.28 | |||||||||
| 7 | F | 30 | Renal-pancreas | MMF | + | − | 0–28 | − | − | <0.05 | ||
| 42 | − | − | 0.14 | |||||||||
| 52 | − | − | 0.3 | I | GCV (52–66) | |||||||
| 62 | − | − | <0.05 | |||||||||
| 70 | − | − | <0.05 | |||||||||
+, positive; −, negative; ?, unknown.
ND, not done.
M, male; F, female.
Clinical significance of CMV tests.
In SCT recipients, because of the small number of episodes of disease and the use of preemptive therapy, specificity was not calculated. Means of maximum viral load found during episodes of CMV disease as determined by FC-Ag were 0.12% (range, 0.07 to 0.17%) for grade I and 0.26% (range, 0.24 to 0.28%) for grade III. FC-Ag was always negative in patients with grade 0 CMV disease.
In SOT recipients, the results of FC-Ag, whether positive (>0.05% positive PMNLs) or negative, showed the highest sensitivity and negative predictive value (NPV) for the diagnosis of all grades of CMV disease, whereas SVA was the least sensitive but most specific method (Table 3). When only grades higher than II were considered, sensitivity and NPV reached 100 and 97% for q-PCR and FC-Ag, respectively (data not shown).
TABLE 3.
Comparison of FC-Ag with other assays in diagnosis of the first episode of CMV disease in SOT recipients (all stages of severity)a
| Parameter | Viremia | DNAemia | FC-Ag threshold viral load
|
||
|---|---|---|---|---|---|
| 0.05% | 0.20% | 1.12% | |||
| Sensitivity | 0.75 | 0.85 | 0.96 | 0.56 | 0.15 |
| Specificity | 0.82 | 0.73 | 0.59 | 0.95 | 1 |
| PPV | 0.90 | 0.87 | 0.84 | 0.96 | 1 |
| NPV | 0.60 | 0.70 | 0.87 | 0.50 | 0.35 |
Analysis was done on 85 SOT patients for viremia and DNAemia assay and on 70 SOT patients (54 infected and 16 uninfected) for FC-Ag.
Means of maximum viral load as observed by FC-Ag among patients with grade 0, I, II, and III CMV disease as determined by FC-Ag were 0.12% (range, 0.02 to 0.19%), 0.16% (range, 0.01 to 0.53%), 0.47% (range, 0.08 to 1.81%), and 0.61% (range, 0.04 to 4.91%), respectively. Maximum virus levels among patients with CMV disease compared to those without CMV disease, and among patients with grade 0 disease compared to those with grade II or III disease, were thus significantly different (P < 0.05). In addition, statistical analysis showed that the risk of developing CMV disease of any grade and of grades II and III only was maximal when FC-Ag results were over 0.20 and 1.12%, respectively (for P < 0.05). In view of these two thresholds, it is not surprising that a higher specificity and positive predictive value (PPV) were observed (Table 3).
Kinetics of CMV markers. (i) Kinetics in the whole population of infected patients.
Active infection, as determined by the first positive q-PCR or SVA result obtained during follow-up, occurred 14 to 175 days after transplantation (Table 4). For the 63 patients with active infection, FC-Ag, q-PCR, or SVA was the first to turn positive (either by itself or simultaneously with one or both of the other assays) in 46, 30, and 30 patients, respectively. FC-Ag became positive before the other tests in 33% of cases (difference range, 3 to 30 days; median, 10 days) and before the first CMV-related symptoms appeared in 57% of cases (range, 3 to 48 days; median, 9 days). The duration of active infection was longer when estimated by FC-Ag. In 27 of the 36 patients with grade II or III disease, the viral level as determined by FC-Ag was either initially higher than 0.20% or showed a progressive increase (Table 5 [patients 1 to 6]). The nine remaining patients had a maximum viral load as determined by FC-Ag of less than 0.15%: eight of them received GCV rapidly, before a new sample could be analyzed by FC-Ag. Finally, the last test to turn negative (either by itself or simultaneously with one or both of the other assays) was FC-Ag in 76% of cases, q-PCR in 60% of cases, and SVA in 22% of cases.
TABLE 4.
Virological characteristics in the 63 infected patients (54 SOT and 9 SCT)
| Characteristic and group | CMV test
|
||
|---|---|---|---|
| SVA | q-PCR | FC-Ag | |
| No. of patients with at least one positive result | |||
| SOT recipients | 44 | 49 | 51 |
| SCT recipients | 7 | 8 | 8 |
| Median (range) day of onset of active infection | |||
| SOT recipients | 35.5 (16–89) | 34 (14–85) | 28 (12–90) |
| SCT recipients | 62 (51–175) | 51 (42–165) | 51 (35–90) |
| Median (range) duration of active infection (days) | |||
| SOT recipients | 9 (0–77) | 15 (0–98) | 20.5 (0–98) |
| SCT recipients | 9 (0–28) | 14.5 (0–28) | 17 (5–32) |
(ii) Effect of antiviral therapy.
It was possible to analyze the effect of antiviral therapy on the different tests in 41 episodes of active infection occurring in 39 patients who received GCV or FCV. In 15 patients there was an increase in viral load during the first week after starting antiviral treatment (mean increase in positive PMNLs 0.37%; range, 0.09 to 1.25%) (patients 1 and 2 in Table 5). All patients became negative for viremia less than a week after GCV or FCV had been started in 80% of cases, and by the end of antiviral treatment all patients were negative for viremia except for three patients who received a second course of antiviral treatment. FC-Ag and q-PCR results remained positive at the end of treatment in 54% (22 of 41) and 44% (18 of 41) of cases, respectively (Fig. 2). The finding of maximum viral load intensity by FC-Ag before the beginning of antiviral treatment was identified as a risk factor for testing to remain positive upon discontinuation of treatment (mean, 0.74%, compared to 0.19% in patients testing negative upon discontinuation; P = 0.001). The delay between the onset of active infection and the start of GCV treatment had no influence on either maximum viral load as determined by FC-Ag or positive testing on discontinuation of treatment. Of the 17 patients with negative FC-Ag results at the end of antiviral therapy, 4 developed a relapse of active infection 12 to 30 days after stopping treatment. Of the 22 patients with positive FC-Ag results at the end of treatment, further analysis was possible in 19. On discontinuing treatment, the 9 patients whose FC-Ag results were either higher than 0.20% positive PMNLs or higher than 0.10% and showing progressive increase subsequently developed grade II or III disease. The other 10 patients (including 6 with positive PCR) remained asymptomatic.
FIG. 2.
Percentage of leukocyte samples tested that remained positive for CMV by SVA, q-PCR, and FC-Ag during the period of antiviral therapy.
DISCUSSION
This study was undertaken to test the usefulness of FC-Ag for detecting CMV infection and for quantifying viral load with a view to predicting CMV disease in immunocompromised patients. With these goals, we designed a prospective longitudinal study of SOT and SCT recipients with a theoretical risk of developing active CMV infection (CMV seropositive recipient and/or donor). Screening for active infection was carried out using two CMV tests in widespread current use, viremia assay (as determined by SVA with inoculation of a large number of vials to provide a level of information similar to that of conventional culture), and qualitative leukocyte DNAemia assay, chosen for its greater sensitivity.
Both CMV infection and disease were found to occur with incidences similar to those reported elsewhere (8, 18, 20). Multivariate analysis only identified the use of ATG as a risk factor for the development of CMV infection or CMV disease (16) and confirmed the absence of any increased risk when MMF is used at the standard dose of 2 g/day (26, 33).
The biological value of FC-Ag was first evaluated by comparing it with the other currently used CMV diagnostic tests. In qualitative terms, we found that FC-Ag agreed closely both with pp65 antigenemia (S-Ag) and with qualitative leukocyte DNAemia assay, thus demonstrating its high degree of sensitivity. This was confirmed by longitudinal evaluation, which showed that FC-Ag results allowed active infection to be detected earlier and also remained positive for a longer time than PCR or viremia assay results. Isolated instances of positive FC-Ag results could be explained as being due either to nonspecific binding of MAbs, despite the use of AB sera to limit this possibility (25) or to the detection of very low levels of CMV beyond the range of other techniques. We thus demonstrated that FC-Ag may be a useful tool in screening for active infection. When intensities of viral load were compared, we found that FC-Ag correlated more closely with S-Ag than with QC-PCR. From a comparison of the antigenemia levels obtained, we realized that the two methods do not calculate the absolute number of PMNLs expressing antigen in the same way (so that a result of 0.05% positive PMNLs in FC-Ag does not correspond to 100 positive cells on a slide). This is probably due to the fact that flow cytometry analyzes each cell individually and is a more sensitive detector of low antigen densities than is visual microscopy (23). An additional reason may be the significant cell loss that occurs during preparation of the slide (up to 40%). The weakness of agreement between FC-Ag results and DNA load that we observed may be related to the small number of quantifications carried out.
The second part of this study was the evaluation of FC-Ag in the clinical situation, as an indicator of the occurrence of CMV disease. The notable differences in the natural history of CMV infection and disease according to immunosuppressive context (32) made it preferable to study SOT and SCT recipients separately.
In SCT recipients we did not observe any association between FC-Ag results or DNA viral load and severity of the disease (6, 32). The likely explanation is primarily that preemptive treatment was used in our study, starting with the second instance of positive CMV DNAemia (8, 21). In addition, in most cases the severe disease in such patients is pneumonitis, which is mediated by immunopathologic mechanisms that occur even at low levels of viral load (4, 9, 15).
In SOT recipients, as previously reported using the slide method, we showed that there was a clear relationship between the level of antigenemia and the presence of clinically significant CMV infection (1, 12, 27). Moreover, FC-Ag level was related to the severity of disease, allowing the test to discriminate between CMV infection or benign disease and more severe CMV disease, as we have previously found using a method for quantifying viral DNA load (20). In view of the viral load as determined by FC-Ag for patients with grade II or III disease and the results at the time of discontinuing antiviral treatment, we postulate that the finding of 0.20% positive PMNLs may represent a threshold of positivity, beyond which tissue involvement is highly probable. The finding of lower levels of antigenemia, however, does not allow the occurrence of CMV disease to be definitively eliminated, as has been previously observed in studies using S-Ag or DNA viral load (20, 35).
As a laboratory tool for assessing the efficacy of antiviral treatment, the usefulness of FC-Ag appears less clear than in predicting CMV disease, notably because antigenemia rebounds during the first week of treatment, as previously reported in a study using the slide assay (13). One hypothesis that might explain this phenomenon is that while antiviral treatment quickly blocks virus replication, pp65 already synthesized may continue to be phagocytosed by PMNLs for several days after treatment has begun (13). PCR has previously been identified as a useful test for assessing the definitive clearance of virus from the blood and for adjusting the duration of antiviral treatment (8, 21). In our study, we observed that, of patients with positive PCR at the time of discontinuing antiviral therapy, only those patients with FC-Ag-determined viral loads higher than 0.20% or with progressively increasing antigenemia levels subsequently developed symptomatic relapse. Again, FC-Ag results appear to be a marker predictive of the occurrence of CMV disease.
This prospective study demonstrates that direct quantification of pp65 antigenemia using flow cytometry can match the sensitivity of PCR and can thus be used for the screening of CMV infection in both SCT and SOT recipients. We have also defined a threshold viral load, namely, 0.20% positive PMNLs, beyond which the occurrence of tissue involvement becomes highly probable in SOT recipients. In these patients, this tool may thus represent a useful marker for the initiation of preemptive therapy. For virology laboratories with easy access to a flow cytometer, we conclude that FC-Ag is a valuable alternative to the slide assay, with the main advantage of rapid, objective quantification of viral load.
ACKNOWLEDGMENTS
This work was supported by a clinical research grant from Nantes University Hospital (Délégation à la Recherche Clinique).
We thank Bernard Besse for excellent technical assistance.
REFERENCES
- 1.Baldanti F, Revello M G, Percivalle E, Gerna G. Use of the cytomegalovirus (HCMV) antigenemia assay for the diagnosis and monitoring of HCMV infections and detection of antiviral drug resistance in immunocompromised patients. J Clin Virol. 1998;11:51–60. doi: 10.1016/s0928-0197(98)00040-3. [DOI] [PubMed] [Google Scholar]
- 2.Bek B, Boeckh M, Lepenies J, Bieniek B, Arasteh K, Heise W, Deppermann K M, Bornhoft G, Stoffler-Meilicke M, Schuller I, Hoffken G. High-level sensitivity of quantitative pp65 cytomegalovirus (CMV) antigenemia assay for diagnosis of CMV disease in AIDS patients and follow-up. J Clin Microbiol. 1996;34:457–459. doi: 10.1128/jcm.34.2.457-459.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Boeckh M, Boivin G. Quantitation of cytomegalovirus: methodologic aspects and clinical applications. Clin Microbiol Rev. 1998;11:533–554. doi: 10.1128/cmr.11.3.533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Boeckh M, Bowden R A, Goodrich J M, Pettinger M, Meyers J D. Cytomegalovirus antigen detection in peripheral blood leukocytes after allogeneic marrow transplantation. Blood. 1992;80:1358–1364. [PubMed] [Google Scholar]
- 5.Boeckh M, Woogerd P M, Stevens-Ayers T, Ray C G, Bowden R A. Factors influencing detection of quantitative cytomegalovirus antigenemia. J Clin Microbiol. 1994;32:832–834. doi: 10.1128/jcm.32.3.832-834.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Boivin G, Quirk M R, Kringstad B A, Germain M, Jordan M C. Early effects of ganciclovir therapy on the quantity of cytomegalovirus DNA in leukocytes of immunocompromised patients. Antimicrob Agents Chemother. 1997;41:860–862. doi: 10.1128/aac.41.4.860. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Detrick B, Hooks J J, Keiser J, Tabbara I. Detection of cytomegalovirus proteins by flow cytometry in the blood of patients undergoing hematopoietic stem cell transplantation. Exp Hematol. 1999;27:569–575. doi: 10.1016/s0301-472x(98)00076-9. [DOI] [PubMed] [Google Scholar]
- 8.Einsele H, Ehninger G, Hebart H, Wittkowski K M, Schuler U, Jahn G, Mackes P, Herter M, Klingebiel T, Loffler J, Wagner S, Muller C A. Polymerase chain reaction monitoring reduces the incidence of cytomegalovirus disease and the duration and side effects of antiviral therapy after bone marrow transplantation. Blood. 1995;86:2815–2820. [PubMed] [Google Scholar]
- 9.Enright H, Haake R, Weisdorf D, Ramsay N, McGlave P, Kersey J, Thomas W, McKenzie D, Miller W. Cytomegalovirus pneumonia after bone marrow transplantation. Risk factors and response to therapy. Transplantation. 1996;55:1339–1346. doi: 10.1097/00007890-199306000-00024. [DOI] [PubMed] [Google Scholar]
- 10.Gane E, Saliba F, Valdecasas G J, O'Grady J, Pescovitz M D, Lyman S, Robinson C A. Randomised trial of efficacy and safety of oral ganciclovir in the prevention of cytomegalovirus disease in liver-transplant recipients. Lancet. 1997;350:1729–1733. doi: 10.1016/s0140-6736(97)05535-9. [DOI] [PubMed] [Google Scholar]
- 11.Gerna G, d'Armininio Monforte A, Zavattoni M, Sarasini A, Testa L, Revello M G, Moroni M. Sharp drop in the prevalence of human cytomegalovirus leuko-DNAemia in HIV-infected patients following highly active antiretroviral therapy. AIDS. 1998;12:118–120. [PubMed] [Google Scholar]
- 12.Gerna G, Percivalle E, Baldanti F, Sarasini A, Zavattoni M, Furione M, Torsellini M, Revello M G. Diagnostic significance and clinical impact of quantitative assays for diagnosis of human cytomegalovirus infection/disease in immunocompromised patients. New Microbiol. 1998;21:293–308. [PubMed] [Google Scholar]
- 13.Gerna G, Zavattoni M, Percivalle E, Grossi P, Torsellini M, Revello M G. Rising levels of human cytomegalovirus (HCMV) antigenemia during initial antiviral treatment of solid-organ transplant recipients with primary HCMV infection. J Clin Microbiol. 1998;36:1113–1116. doi: 10.1128/jcm.36.4.1113-1116.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Goodrich J M, Mori M, Gleaves C A, Du Mond C, Cays M, Ebeling D F, Buhles W C, Dearmond B, Meyers J D. Early treatment with ganciclovir to prevent cytomegalovirus disease after allogeneic bone marrow transplantation. N Engl J Med. 1991;325:1601–1607. doi: 10.1056/NEJM199112053252303. [DOI] [PubMed] [Google Scholar]
- 15.Gor D, Sabin C, Prentice H G, Vyas N, Man S, Griffiths P D, Emery V C. Longitudinal fluctuations in cytomegalovirus load in bone marrow transplant patients: relationship between peak virus load, donor/recipient serostatus, acute GVHD and CMV disease. Bone Marrow Transplant. 1998;21:597–605. doi: 10.1038/sj.bmt.1701139. [DOI] [PubMed] [Google Scholar]
- 16.Hassan-Walker A F, Kidd I M, Sabin C, Sweny P, Griffiths P D, Emery V C. Quantity if human cytomegalovirus (CMV) DNAemia as a risk factor for CMV disease in renal allograft recipients: relationship with donor/recipient CMV serostatus, receipt of augmented methylprednisolone and antithymocyte globulin (ATG) J Med Virol. 1999;58:182–187. [PubMed] [Google Scholar]
- 17.Hebart H, Kanz L, Jahn G, Einsele H. Management of cytomegalovirus infection after solid-organ or stem-cell transplantation. Current guidelines and future prospects. Drugs. 1998;55:59–72. doi: 10.2165/00003495-199855010-00005. [DOI] [PubMed] [Google Scholar]
- 18.Ho M. Cytomegalovirus infection and indirect sequelae in the immunocompromised transplant patient. Transplant Proc. 1991;23:2–7. [PubMed] [Google Scholar]
- 19.Honda J, Okubo Y, Imamura Y, Kusaba M, Saruwatari N, Oizumi K. Flow cytometric detection of cytomegalovirus antigen in peripheral blood cells after bone marrow transplantation. Br J Haematol. 1997;99:415–417. doi: 10.1046/j.1365-2141.1997.3863200.x. [DOI] [PubMed] [Google Scholar]
- 20.Imbert-Marcille B M, Cantarovich D, Ferre-Aubineau V, Richet B, Soulillou J P, Billaudel S. Usefulness of DNA viral load quantification for cytomegalovirus disease monitoring in renal and pancreas/renal transplant recipients. Transplantation. 1997;63:1476–1481. doi: 10.1097/00007890-199705270-00018. [DOI] [PubMed] [Google Scholar]
- 21.Imbert-Marcille B M, Milpied N, Coste-Burel M, Richet B, Moreau P, Harousseau J L, Billaudel S. Clinical and practical value of human cytomegalovirus DNAemia detection by semi-nested PCR for follow-up of BMT recipients. Bone Marrow Transplant. 1995;15:611–617. [PubMed] [Google Scholar]
- 22.Imbert-Marcille B M, Robillard N, Poirier A S, Coste-Burel M, Cantarovich D, Milpied N, Billaudel S. Development of a method for direct quantification of cytomegalovirus antigenemia by flow cytometry. J Clin Microbiol. 1997;35:2665–2669. doi: 10.1128/jcm.35.10.2665-2669.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Jenson H B, Grant G M, Ench Y, Heard P, Thomas C A, Hilsenbeck S G, Moyer M P. Immunofluorescence microscopy and flow cytometry characterization of chemical induction of latent Epstein-Barr virus. Clin Diagn Lab Immunol. 1998;5:91–97. doi: 10.1128/cdli.5.1.91-97.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lowance, D., H. H. Neumayer, C. M. Legendre, J. P. Squifflet, J. Kovarik, P. J. Brennan, D. Norman, R. Mendez, M. R. Keating, G. L. Coggon, A. Crisp, and I. C. Lee. Valacyclovir for the prevention of cytomegalovirus disease after renal transplantation. International valacyclovir cytomegalovirus prophylaxis transplantation study group. N. Engl. J. Med. 340:1462–1470. [DOI] [PubMed]
- 25.McSharry J J. Uses of flow cytometry in virology. Clin Microbiol Rev. 1994;7:576–604. doi: 10.1128/cmr.7.4.576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Moreso F, Seron D, Morales J M, Cruzado J M, Gil-Vernet S, Perez J L, Fulladosa X, Andres A, Grinyo J M. Incidence of leukopenia and cytomegalovirus disease in kidney transplants treated with mycophenolate mofetil combined with low cyclosporine and steroid doses. Clin Transplant. 1998;12:198–205. [PubMed] [Google Scholar]
- 27.Niubo J, Perez J L, Martinez-Lacasa J T, Garcia A, Roca J, Fabregat J, Gil-Vernet S, Martin R. Association of quantitative cytomegalovirus antigenemia with symptomatic infection in solid organ transplant patients. Diagn Microbiol Infect Dis. 1996;24:19–24. doi: 10.1016/0732-8893(95)00248-0. [DOI] [PubMed] [Google Scholar]
- 28.Noble S, Faulds D. Ganciclovir. An update of its use in the prevention of cytomegalovirus infection and disease in transplant recipients. Drugs. 1998;56:115–146. doi: 10.2165/00003495-199856010-00012. [DOI] [PubMed] [Google Scholar]
- 29.Palella F J, Delaney K M, Moorman A C, Loveless M O, Fuhrer J, Satten G A, Aschman D J, Holmberg S D. Declining morbidity among patients with advanced human immunodeficiency virus infection. HIV outpatient study investigators. N Engl J Med. 1998;338:853–860. doi: 10.1056/NEJM199803263381301. [DOI] [PubMed] [Google Scholar]
- 30.Poirier-Toulemonde A S, Imbert-Marcille B M, Ferre-Aubineau V, Besse B, Le Roux M G, Cantarovich D, Billaudel S. Successful quantification of cytomegalovirus DNA by competitive PCR and detection with capillary electrophoresis. Mol Cell Probes. 1997;11:11–23. doi: 10.1006/mcpr.1996.0071. [DOI] [PubMed] [Google Scholar]
- 31.Prentice H G, Kho P. Clinical strategies for the management of cytomegalovirus infection and disease in allogeneic bone marrow transplant. Bone Marrow Transplant. 1997;19:135–142. doi: 10.1038/sj.bmt.1700630. [DOI] [PubMed] [Google Scholar]
- 32.Saltzman R L, Quirk M R, Jordan M C. High levels of circulating cytomegalovirus DNA reflect visceral organ disease in viremic immunosuppressed patients other than marrow recipients. J Clin Investig. 1992;90:1832–1838. doi: 10.1172/JCI116059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Sarmiento J M, Munn S R, Paya C V, Velosa J A, Nguyen J H. Is cytomegalovirus infection related to mycophenolate mofetil after kidney transplantation? A case-control study. Clin Transplant. 1998;12:371–374. [PubMed] [Google Scholar]
- 34.Schimenti K J, Jacobberger J W. Fixation of mammalian cells for flow cytometric evaluation of DNA content and nuclear immunofluorescence. Cytometry. 1992;13:48–59. doi: 10.1002/cyto.990130109. [DOI] [PubMed] [Google Scholar]
- 35.Toyoda M, Carlos J B, Galera O A, Galfayan K, Zhang X, Sun Z, Czer L S, Jordan S C. Correlation of cytomegalovirus DNA levels with response to antiviral therapy in cardiac and renal allograft recipients. Transplantation. 1997;63:957–963. doi: 10.1097/00007890-199704150-00009. [DOI] [PubMed] [Google Scholar]