INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

In the last few years, highly active antiretroviral therapy (HAART) has greatly reduced the incidence of human cytomegalovirus (HCMV) disease in human immunodeficiency virus (HIV)-infected patients in Western countries (5, 19, 20). However, at least two reasons suggest that careful monitoring of HCMV disease in HIV-infected patients is still important: first, some cases of HCMV disease in patients with relatively high CD4⁺ T-cell counts have been reported (7, 10, 11, 15); second, the guidelines for the discontinuation of maintenance anti-HCMV therapy are mainly based on the CD4⁺ T-cell count (12, 13, 16, 18, 24, 25, 27), regardless of any possible information on the reconstitution of the HCMV-specific immune response. Evaluation of specific anti-HCMV immunity could influence the clinical decision to discontinue or maintain anti-HCMV therapy in patients with previous HCMV disease and intermediate CD4⁺ T-cell counts. In this respect, evaluation of the HCMV-specific T-helper response is of particular interest (23). For several years, T-helper immune response has been evaluated by testing the antigen-specific lymphoproliferative response (LPR). However, the impact of the LPR assay in guiding clinical decisions is still limited, since the assay is time-consuming and poorly standardized and, being based on the use of tritiated thymidine, requires specific containment measures and facilities.

The most promising new alternative technique for the determination of HCMV-specific T-helper response is the evaluation of HCMV-specific CD4⁺ T-cell frequency by flow cytometry detection of intracellular cytokines after short-term antigen-specific activation of peripheral blood mononuclear cells (PBMC), as recently reported (1, 9, 14, 21, 26). This technique provides results in a few hours, does not require the use of radioactive compounds, is easy to standardize, and is applicable to frozen PBMC samples. Previous reports have shown that HCMV-specific CD4⁺ T cells are almost totally polarized toward Th1 phenotype and therefore that the frequency of CD4⁺ T cells producing tumor necrosis factor alpha (TNF-) or gamma interferon (IFN-) after exposure to HCMV antigens can be considered the overall frequency of HCMV specific CD4⁺ T cells (21, 26). Some authors have described the analysis of HCMV-specific CD4⁺ T-cell frequency in different stages of HIV disease (14, 21), but LPR remains the most commonly used test to evaluate the HCMV-specific T-helper response. At present, a single report has been published comparing the HCMV-specific CD4⁺ T-cell frequency and LPR in two groups of HIV-seropositive patients (9).

In this study, we compared the two techniques by testing a series of samples from HIV-infected patients. Samples from patients showing a wide range of CD4⁺ T-cell counts, either with or without HCMV-specific T-helper response, were examined. Our data suggest that HCMV-specific CD4⁺ T-cell frequency correlates with LPR and is a reliable alternative to the HCMV-specific LPR test.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Patients. A total of 78 blood samples from 65 adult patients (20 females, 45 males) that were both HIV and HCMV seropositive were analyzed. The number of CD4⁺ T cells per microliter of blood was assessed by the Ortho ImmunoCount Flow Cytometer System (Raritan, N.J.), while the HIV type 1 (HIV-1) viral load was determined by the bDNA technique (Bayer/Chiron Corp., Emeryville, Calif.). At first evaluation, of the 65 patients, 6 (3 with and 3 without HCMV disease) were HAART naive with a median HIV load of 272,551 RNA copies/ml (range, <50 to 1,800,000 copies/ml) and 51 CD4⁺ T cells/µl (range, 1 to 221 cells), and 6 (3 with and 3 without HCMV disease) had been treated with short-term HAART for 12 months (range, 8 to 20 months), reaching the level of <50 HIV RNA copies/ml and 271 CD4⁺ T cells/µl (range, 65 to 410 cells/µl). In addition, of the remaining 53 patients treated with long-term HAART for 46 months (range, 29 to 57 months), 11 with HCMV disease were tested either during anti-HCMV treatment (n = 6) or after its discontinuation (n = 5), while 42 were examined in the absence of HCMV disease. At initial observation, this patient group showed a median HIV load of 59 RNA copies/ml (range, <50 to 150,070 copies/ml) and a CD4⁺ T-cell count of 421/µl (range, 50 to 1,256 cells/µl).

PBMC preparation and storage. PBMC were isolated by standard Ficoll gradient centrifugation of heparinized blood samples and frozen by resuspending 5 × 10⁶ cell aliquots in freezing medium (RPMI 1640 supplemented with 10% dimethyl sulfoxide and 5% human albumin). When necessary, cells were thawed rapidly at 37°C and washed twice with RPMI 1640 containing 5% pooled human serum (AB type; Euroclone, Wetherby, West York, United Kingdom). Viability was evaluated by trypan blue exclusion staining. The rare samples with a viability of <95% were discarded.

HCMV-specific lymphoproliferative assay. Cells resuspended in RPMI 1640 containing 5% pooled human serum (AB type) were plated in triplicate at a 10⁵ concentration (200 µl of medium per well) in U-bottom 96-well microtiter plates. Viral and control antigens were prepared by glycine buffer extraction (0.1 M, pH 9.6) of AD169 strain-infected and uninfected human embryonic lung fibroblasts, followed by sonication and low-speed centrifugation. Supernatants were stored in aliquots at 80°C. Heat-inactivated HCMV antigen (or control uninfected fibroblast antigen) was added at a 1:1,000 dilution (optimal stimulating concentration). After 6 days of culture at 37°C in 5% CO₂ atmosphere, 1 µCi of [³H]thymidine per well was added, and the cultures incubated for another 18 h. DNA-incorporated thymidine was evaluated by harvesting cells and counting radioactivity in a TopCount 100 system (Packard Instrument Co., Meriden, Conn.). The net count-per-minute (cpm) response was calculated from the mean cpm of wells with HCMV antigen minus the mean cpm of wells without HCMV antigen. The stimulation index (SI) was the ratio of the mean cpm of the wells with HCMV antigen to the mean cpm of the wells without HCMV antigen.

HCMV-specific CD4⁺ T-cell frequency evaluation. HCMV-specific CD4⁺ T-cell frequency was determined as previously described (21) with minor modifications. Briefly, PBMC were resuspended in RPMI 1640 containing 10% fetal calf serum (Gibco, Grand Island, N.Y.) at 4 × 10⁶ cells/ml. Half-milliliter aliquots were dispensed in 15-ml polypropylene conical tubes (Becton Dickinson Labware, Franklin Lane, N.J.). Then, 0.5 µg of anti-CD28 (Becton Dickinson, San Jose, Calif.) and anti-CD49d (Southern Biotechnology Associates, Birmingham, Ala.) monoclonal antibodies were added to each aliquot. Next, 40 µl of a commercially available HCMV complement fixation antigen preparation (AD169 strain grown in human embryonic fibroblasts; BioWhittaker, Walkersville, Md.) at its optimal dilution or its negative control was added as a specific stimulus; in some experiments, staphylococcal enterotoxin B (SEB; Sigma, St. Louis, Mo.) at 2 µg/ml was used as an alternative stimulus. Tubes were incubated 1 h at 37°C in a slant position and then brefeldin A (Sigma) resuspended in ethanol was added at a 10-µg/ml final concentration. Cells were incubated for an additional 14 h under the same conditions, washed once with cold phosphate-buffered saline containing 2 mM EDTA, fixed and permeabilized by using a Fix & Perm kit (Caltag, Burlingame, Calif.) according to the manufacturer's instructions, and then labeled with the following monoclonal antibodies (Becton Dickinson): anti-CD4-fluorescein isothiocyanate, anti-TNF--phycoerythrin (PE), or anti-IFN--PE plus anti-CD69-PerCP. Cell suspensions were analyzed with a FACSCalibur flow cytometer (Becton Dickinson) equipped with CellQuest software. Viable lymphocytes were identified by scattering parameters. For each sample, 50,000 viable CD4⁺ T lymphocytes were evaluated. CD4⁺ cells expressing TNF- or IFN- and CD69 were considered activated cells. The HCMV-specific CD4⁺ frequency was calculated by subtracting the value of the control sample incubated with control antigen (consistently 0.05%). The value of HCMV specific CD4⁺ T cells per ml of blood was calculated by multiplying the HCMV-specific CD4⁺ T-cell frequency by the number of CD4⁺ T cells per milliliter of blood.

Statistical analysis. The correlation between absolute CD4⁺ T-cell counts and LPR to HCMV or HCMV-specific CD4⁺ T-cell frequency as well as that between LPR to HCMV and HCMV-specific CD4⁺ T-cell frequency was determined by calculating the coefficient of regression R in a univariate regression analysis.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Comparative analysis of LPR and HCMV-specific CD4⁺ T-cell frequency in fresh and frozen PBMC samples. HCMV-specific CD4⁺ T-cell frequency was first evaluated in both fresh and frozen PBMC samples from seven HCMV-seropositive subjects (six HIV-seropositive and one HIV-seronegative). TNF- was used as an activation marker because it gave a sharper peak than IFN- in flow cytometry analysis. The HCMV-specific CD4⁺ T-cell frequency found in frozen samples was 80.9% ± 67.7% of fresh samples. Overall, all of the five fresh samples that were considered positive for the presence of HCMV-specific CD4⁺ T cells (according to criteria described below) also remained positive when evaluated frozen. Individual data are reported in Table 1.

Reproducibility of HCMV-specific CD4⁺ T-cell analysis. PBMC samples from three normal subjects were frozen, and the frequency of HCMV-specific CD4⁺ T cells was evaluated in three independent test runs. The coefficients of variation were 31.9, 33.4, and 28.6% for the three subjects.

Definition of criteria for determination of cutoffs for HCMV-specific CD4⁺ T-cell frequency and LPR. Nine healthy HCMV-seropositive subjects and six healthy HCMV-seronegative subjects were evaluated for their HCMV-specific CD4⁺ T-cell frequency (Table 2). Based on these results and on the consideration that a flow cytometry determination of cell frequency of <0.1% was considered unreliable, we established the following criteria to define the presence of HCMV-reactive CD4⁺ T cells by flow cytometry analysis: the simultaneous presence of an HCMV-specific CD4⁺ T-cell frequency of 0.1% and 400/ml of blood. In parallel, the LPR response was evaluated in 23 samples from 10 immunocompetent HCMV-seropositive subjects and in 8 samples from 4 immunocompetent HCMV-seronegative subjects (Table 3). Patients were considered responders (R) for LPR to HCMV when the SI was 3 and the net cpm was 3,000 and defined as nonresponders (NR) when either one or both of these conditions were not satisfied. All of the nine HCMV-seropositive healthy subjects tested by both assays were found to be positive for both LPR to HCMV and a measurable frequency of HCMV-specific CD4⁺ T cells.

Comparison of HCMV-specific CD4⁺ T-cell frequency and LPR in HIV-infected patients Seventy-eight PBMC samples from 65 HCMV-seropositive HIV-infected patients (13 patients were evaluated twice) were tested by both techniques. In summary, according to the criteria reported above, 43 samples (55.1%) were found to be positive and 20 samples (25.6%) were found to be negative by both assays. In addition, seven samples (9.0%) were found to be positive for HCMV-specific CD4⁺ T-cell presence but negative for LPR, whereas eight samples (10.4%) were negative for HCMV-specific CD4⁺ T cells but positive for LPR (Table 4). Therefore, 80.7% of assayed samples gave concordant results by both techniques. As for discordant results, the evaluation of HCMV-specific CD4⁺ T-cell frequency was repeated for most samples, confirming previous results.

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Correlation of HCMV-specific CD4+ T-cell frequency and LPR as net cpm in 78 PBMC samples from 65 HIV-infected patients. Dotted lines indicate the cutoffs for the HCMV-specific CD4+ T-cell presence (0.1% of total CD4+ T cells) and for HCMV-specific LPR (3,000 net cpm).

Correlation of HCMV-specific CD4⁺ T-cell count and LPR. The correlation of HCMV-specific CD4⁺ T-cell count and LPR was significant (R = 0.327, P = 0.003) (Fig. 1), whereas no significant correlation was found between CD4⁺ T-cell count and either the net cpm of LPR (Fig. 2A) or the HCMV-specific CD4⁺ T-cell frequency (Fig. 2B). It was evident that several patients with CD4⁺ T-cell count of <150/µl showed a high anti-HCMV T-helper activity, as documented by both assays. On the other hand, an even greater number of patients with high CD4⁺ T-cell counts did not show any anti-HCMV T-helper activity by either assay. These data suggest that the CD4⁺ T-cell count alone is an insufficient criterion for deciding upon strategies of anti-HCMV therapy.

View larger version (20K): [in this window] [in a new window]   FIG. 2.   Correlation of CD4+ T-cell count per microliter of whole blood and LPR (A) or HCMV-specific CD4+ T-cell frequency (B) in 78 PBMC samples from 65 HIV-infected patients. Horizontal dotted lines indicate the cutoffs for the HCMV-LPR (3,000 net cpm, panel A) and for the HCMV-specific CD4+ T cells (0.1% of total CD4+ T cells, panel B). In both panels, the vertical dotted lines indicate the value of 150 CD4+ T cells/µl of blood. Open symbols indicate samples with an SI of <3.0 (A) or HCMV-specific CD4+ T cells <400/ml of blood (B).

Sequential evaluation of HCMV-specific CD4⁺ T-cell frequency and LPR in HIV-infected patients. To investigate whether the individual variability in the anti-HCMV T-helper response was due to fluctuations of the immune response, 13 HIV-infected patients were evaluated for both HCMV-specific CD4⁺ T-cell frequency and LPR during a median period of 8 months (range, 2 to 10 months). At the time of the first evaluation, two of these patients were HAART naive (patients 2 and 3), whereas the remaining patients had been HAART treated for 3 years (except patient 1, who was treated only for 9 months). Figure 3 shows comparative data of LPR (panel A) and HCMV-specific CD4⁺ T-cell frequency (panel B). Of the 13 patients, 2 (patients 3 and 4) did not respond by either assay over time, and 5 (patients 2, 5, 6, 7, and 13) showed a substantially stable response by both assays. On the other hand, 5 patients (patients 1, 8, 10, 11, and 12) showed declining levels of HCMV-specific T-helper response by both or either assay, and only one patient (patient 9) displayed a positive response by flow cytometry at the time of the second evaluation only (Fig. 3B).

View larger version (28K): [in this window] [in a new window]   FIG. 3.   HCMV-specific LPR (A) and CD4+ T-cell frequency (B) in 13 HIV-infected patients tested during a period of 2 to 13 months (median, 8 months). All patients (except patients 1, 2, and 3) had been HAART treated for at least 3 years prior to first evaluation. Only patients 3, 4, and 9 had a CD4+ T-cell count of <150/µl and an HIV load of >10,000 HIV RNA copies/ml. , first evaluation; , second evaluation.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

HAART has been found to be effective in restoring CD4⁺ T-cell counts and reconstituting the immune system function in the majority of HIV patients (2, 5, 17, 19, 20, 22). However, some cases of HCMV retinitis in HIV patients with high CD4⁺ T-cell counts have been reported (7, 10, 15), suggesting that the HCMV-specific immune response could not be restored in those patients, possibly due to failure of reconstitution of the complete T-cell-receptor repertoire (4, 6, 8). In addition, current guidelines suggest maintenance or interruption of anti-HCMV therapy only based upon CD4⁺ T-cell count (12, 13, 16, 18, 24, 25, 27). Furthermore, as shown in this study, some patients with low CD4⁺ T-cell counts possess an apparently adequate anti-HCMV immune response and so would not require anti-HCMV therapy.

As a matter of fact, one of the major aims of research in this field is to develop an immunoassay that can evaluate the HCMV-specific immune response (in particular, the T-helper response) and predict which patients will relapse or reactivate HCMV disease. Lack of this information is partially due to the poor standardization and the large intrinsic variability of the most common method, the LPR assay. In this respect, one of the most promising alternative techniques is the evaluation of HCMV-specific CD4⁺ T-cell frequency by cytokine flow cytometry. Despite the fact that this technique has been already used in several clinical studies (1, 14, 21, 26), its correlation with the well-known LPR assay has, to our knowledge, only preliminarily been reported (9).

The most interesting finding of the present study is that, in a population of HIV-seropositive patients with a wide range of CD4⁺ T-cell counts, analysis of presence of HCMV-specific T-helper response by LPR and HCMV-specific CD4⁺ T-cell frequency gave concordant results in more than 80% of cases. Agreement between the two assays is only slightly higher than that reported in a recent study, in which the HCMV-specific CD4⁺ T-cell frequency was determined by HCMV-specific stimulation of whole blood instead of PBMC (9). The lack of an even higher concordance between the two techniques could be explained by considering the two groups of discordant results separately. In the group of patients positive for LPR but negative for HCMV-specific CD4⁺ T-cell frequency, the positivity for LPR was consistently weak. As a matter of fact, it is possible that in a few samples very rare HCMV-specific CD4⁺ T cells are present with a high proliferative potential that can be detected by LPR but not by flow cytometry analysis. On the other hand, in a few samples HCMV-specific CD4⁺ T-cell presence was detected (in two cases, at high frequency), but the same PBMC samples were not able to proliferate in vitro in response to HCMV antigen. This discordance could be explained by considering that LPR measures the expansion of several antigen-specific T-cell clones, displaying different effector functions, whereas the HCMV-specific CD4⁺ T-cell frequency was calculated by taking into account only TNF-- and/or IFN--producing cells. It is conceivable that HIV-positive individuals showing a negative LPR, in the presence of a measurable frequency of HCMV-specific CD4⁺ T-cell frequency, lack some other antigen-specific T-cell subsets, such as IL-2-producing cells, that are necessary for optimal T-cell expansion. Alternative explanations could include (i) a suboptimal antigen-presenting-cell function, which might be inadequate for sustaining T-cell expansion but sufficient to activate TNF-- and/or IFN--producing cells, and (ii) the presence of a high proportion of HCMV-specific T cells with suppressor function that can inhibit lymphocyte proliferation (3).

The evidence that the HCMV-specific LPR assay and the evaluation of HCMV-specific CD4⁺ T-cell frequency in a proportion of HIV-positive individuals give rise to discrepant results suggests that the evaluation of HCMV-specific immune recovery after HAART therapy should include both assays at best. Nevertheless, the results of the present study suggest that, whenever the availability of PBMC is not sufficient to perform both assays, the evaluation of HCMV-specific CD4⁺ T-cell frequency is as reliable as LPR.

The concordance of results obtained by the two assays on the great majority of samples tested was confirmed by the similarity of results found in the short-term follow-up of the 13 HIV-infected patients tested some months apart. While a fair proportion of patients showed stable results, another substantial aliquot displayed results showing declining levels of the HCMV-specific CD4⁺ T-cell-mediated immune response by either one or both assays after 3 to 4 years of HAART. These findings are in agreement with results recently acquired in our laboratory showing a decrease or loss of the rescued LPR to HCMV in AIDS patients after 3 to 4 years of HAART (unpublished data).

Present data do not provide an explanation for this phenomenon. However, it is conceivable that immune recovery achieved after HAART therapy might bring about a decrease in the frequency of HCMV reactivation and, as a consequence, a reduction in the number of circulating HCMV-specific T cells, which are no longer stimulated by chronic antigen exposure.

The significant correlation of CD4⁺ T-cell frequency and LPR supports the concept that the two assays may be interchangeable. On the other hand, the lack of correlation between CD4⁺ T-cell count and either the CD4⁺ T-cell frequency or LPR documents that the absolute CD4⁺ T-cell count and the other two assays do not express a comparable immunological condition, thus indicating that the CD4⁺ T-cell count alone is insufficient for deciding upon strategies of discontinuation of anti-HCMV therapy in AIDS patients with HCMV retinitis.

In conclusion, our data suggest that the determination of HCMV-specific CD4⁺ T-cell frequency can be considered a valid alternative to the LPR test for the detection of HCMV-specific T-helper response in HIV-infected patients. This technique offers evident advantages over LPR. It does not require the use of radioactive compounds, and it is faster and easier to standardize. Furthermore, it could facilitate wider screening of anti-HCMV T-helper activity in HIV-infected patients, with potential benefits for clinicians in deciding strategies of discontinuation or maintenance of anti-HCMV therapy.