Apr00i.doc

Pan American Society
For Clinical Virology

Volume 27Number 1April 2000

CONTENTS:

1.VOTE for PASCV Officers! Please use paper ballot sent to you. Note that
you should vote for 3 Councilors

2.Asessment of CMV Cellular Load in Aids Patients Using Viremia, pp65,
Antigenemia, and bDNA CMV 1.0, by I. Garrigue, I. Pellegrin, JL.
Pellegrin, M. Dupon, JM. Ragnaud, HJA. Fleury.

3.Renew your membership for year 2000.

ASSESSMENT OF CMV CELLULAR LOAD IN AIDS
PATIENTS USING VIREMIA, PP65 ANTIGENEMIA
AND bDNA CMV 1.0.

I. Garrigue, I. Pellegrin, JL. Pellegrin, M. Dupon, JM. Ragnaud, HJA. Fleury. Departments of
Virology and Infectious Diseases, Pellegrin and Haut-Lévêque Hospitals, University Victor
Segalen Bordeaux 2, France.

Cytomegalovirus (CMV) disease is a major cause of morbidity and mortality in
immunocompromised subjects, especially human immunodeficiency virus (HIV) infected
patients and transplant recipients. There is an increased interest in developing CMV markers
of viral load, to predict the risk of CMV disease, to diagnose CMV infection and to monitore
the efficacy of anti-CMV therapy.
The aim of this study was to compare the results of CMV shell vial culture (SV), CMV pp65
antigenemia (pp65 Ag) and branched DNA (bDNA) CMV 1.0 in AIDS patients with low CD4+
counts.

PATIENTS AND METHODS

AIDS patients with less than 100 CD4+ cells/μl were prospectively enrolled between
November 1995 and April 1997. When a CMV disease occured, patient received an infusion
of foscarnet (Astra France) 90 mg/kg twice daily over at least 10 days, the subsequent
treatment depended on the disease.
Blood samples were serially collected every two months, and more frequently when patient
was treated for a CMV infection (before and during treatment).
•SV culture : peripheral blood leukocytes were inoculated onto duplicate human fibroblast

(MRC-5) monolayers. An immunohistochemical staining was performed after a 48 hours
culture (anti-HCMV ab E13, Argène Biosoft (Varilhes, France), rabbit anti-mouse ab
RAM/IgG/PO, Nordic Immunology (Tilburg, The Netherlands)). This test was considered
to be positive when at least one viral inclusion was observed.

•

pp65 Ag were performed following the manufacturer’s recommandations (HCMV
antigenemia immunofluorescent kit, Argene Biosoft (Varilhes, France)). Results were
expressed as the number of CMV antigen positive cells per 200,000 neutrophils and were
considered to be positive when at least one fluorescent cell was observed.
bDNA CMV assay 1.0 (Chiron, Emeryville, CA, USA) was performed according to
manufacturer’s recommandations. The lower limit of detection of this quantitative assay
is 4400 equivalents of CMV DNA in 106leukocytes.

•

RESULTS

We included 78 AIDS patients (mean CD4+ counts : 25 cells/μl (1-98)). The mean follow-up
was 9.3 months (2-16) ; 19 out of the 78 patients met criteria for development of CMV
disease : retinitis in 11, retinitis and encephalitis in 1, colitis in 4, pneumonia in 3 ; 504
samples (mean of 6.5 / patient (2-22)) were evaluable by SV and pp65 Ag of which 70 were
also tested by bDNA CMV 1.0.
In the 70 samples (patients with or without CMV disease), the proportional Kappa
concordance coefficient calculated from qualitative results between pp65 Ag and bDNA was
Kp= 0.69 (good statistical concordance when Kp>0.6).
In the 30 samples pp65 Ag + / bDNA +, the quantitative results (number of positive
neutrophils / 200000 cells in relation to the number of Keq / 106cells) were significantly
correlated : r = 0.68 (p <0.02).

TABLES

Table 1: Results of SV and pp65 Ag during the follow-up of the 78 patients with or without

clinical CMV disease (504 samples).

CMV shell vial culture

pp65 Antigenemia

Patients without CMV
disease
(n = 59)
patients with CMV disease
(n = 19)

A patient was considered positive when at least 1 sample was (+) during his follow-up after
the appearance of clinical CMV disease, and (-) when all the samples were negative.
Shell vial culture :sensitivity = 58% (11/19) and specificity = 93% (55/59)
pp65 Antigenemia :sensitivity = 100% (19/19) and specificity = 58% (34/59)

Table 2: Results of SV, pp65 Ag and bDNA CMV 1.0 during the follow-up of 25 patients (70

samples) with or without clinical CMV disease.

CMV shell vial culture

pp65 Antigenemia

bDNA CMV 1.0

Patients without
CMV disease
(n = 10)

Patients with CMV
disease
(n = 15)

A patient was considered positive when at least 1 sample was (+) during his follow-up after

the appearance of clinical CMV disease, and (-) when all the samples were negative.

Shell vial culture :

sensitivity = 60% (9/15) and specificity = 100% (10/10)

pp65 Antigenemia :

sensitivity = 100% (15/15) and specificity = 30% (3/10)

bDNA CMV 1.0 :

sensitivity = 93% (14/15) and specificity = 40% (4/10)

Table 3: Results of pp65 Ag and bDNA CMV 1.0 in 70 samples.

bDNA CMV 1.0

pp65 Ag

13 samples (from 11 patients) were pp65 Ag + / bDNA -:

11 samples were belonging to 9 patients (6 with and 3 without CMV disease), who

did have a positive bDNA assay before or after this negative sample

1 sample, from a clinically infected patient had a bDNA result near the cut-off

value.

1 sample was belonging to a patient without CMV disease (pp65 Ag : 3 positive

cells).

Most of these 13 specimens had low grade antigenemia (7 were ≤5 positive cells)

and 6 presented between 12 and 58 positive cells.

7 samples (from 2 patients with and 4 without CMV disease) were pp65

Ag - / bDNA +: the bDNA results were ranging from 5 to 68 Keq / 106cells. All

of these 6 patients had a positive pp65 Ag before or after this negative result.

COMMENTS

In our study as in others (1-3), the pp65 Ag assay was more sensitive than SV culture
(100% vs 58%) to detect a CMV disease (Table 1). The pp65 Ag assay is rapid and
quantitative. However, the test is subjective, quite labor intensive and will have to compete
with semi automated and objective DNA based assay.
The bDNA CMV 1.0 assay, performed on 70 samples, had a sensitivity of 93% (Table 2).
Compared to PCR, a great advantage of the bDNA test is that anticontamination procedures
are not required. The assay can also be delayed when cells pellets are stored. Like Chernoff
et al. (4), we observed a good reproducibility, when testing identical pellets during separate
handling.

Since patients might develop CMV infection without clinical disease, pp65 Ag and bDNA
assays’ specificities are low ; SV is more specific (100%) of a CMV disease but less
sensitive.
When a test is sensitive, it detects patients with asymptomatic infection and its specificity
for determining CMV disease decreases. However, sensitive tests are required to detect
asymptomatic CMV infection and to predict CMV disease, so that patients may
receivepreemptive or early treatment. On the other hand, an asymptomatic infected patient

will not necessarily develop a CMV disease. The improvement of test sensitivity must not go
against the well recognition of patients who are at risk for developing a CMV disease.
However, this could be solved by the determination of accurate predictive threshold.

We have previously shown the similar and significant drop of CMV load evaluated by pp65
Ag and bDNA CMV 1.0 assays over 10 days of foscarnet therapy in HIV and CMV co-infected
patients (5). However, either one test or the other showed the CMV load decrease in a
shorter time : these 2 tests detect different viral components that could exhibit different
kinetics.
Therefore, a part of the discrepant results could be explained by the different kinetics of
DNA-emia and antigenemia. Among the 20 discordant results (Table 3), 7 specimens have
been collected during foscarnet treatment (from day 3 to day 10 of therapy) : 4 of these 7
had concordant negative results for the next withdrawal (the 3 others have not been tested
by bDNA assay in the next sample), so that we cannot consider that all the discordant
results are real discrepancies. These different kinetics require further investigation.

CONCLUSIONS

The pp65 Ag assay was confirmed to be more sensitive than shell vial culture. Although
bDNA CMV 1.0 was well correlated with pp65 Ag, the Kappa coefficient concordance pointed
out a less sensitivity of bDNA 1.0. The future bDNA CMV 2.0 will likely exhibit an enhanced
sensitivity for quantitation of CMV cellular load. It will help in the identification of patients at
highest risk for developing CMV disease, in the monitoring of anti-CMV therapies and in the
prompt recognition of drug-resistant CMV in HIV infected like in transplanted patients.

AKNOWLEDGMENTS
We wish to thank : B. Dumon, C. Leduc, C. Melon, all the technicians for excellent assitance
; Chiron France for providing bDNA CMV 1.0 kits.

REFERENCES

1. A. Erice, MA. Holm, PC. Gill et al. Cytomegalovirus (CMV) antigenemia assay is more
sensitive than shell vial cultures for rapid detection of CMV in polymorphonuclear blood
leukocytes. J. Clin. Microbiol. 1992 ; 30 : 2822-25.

2. ML. Landry and D. Ferguson. Comparison of quantitative cytomegalovirus antigenemia
assay with culture methods and correlation with clinical disease. J. Clin. Microbiol. 1993 ;
31 : 2851-56.

3. D. Francisci, A. Tosti, R. Preziosi, F. Baldelli, G. Stagni, S. Pauluzzi. Role of antigenemia
assay in the early diagnosis and prediction of human cytomegalovirus organ involvement
in AIDS patients. Eur. J. Clin. Microbiol. Infect. Dis. 1995 ; 14 : 498-503.

4.DN. Chernoff, RC. Miner, BS. Hoo and al. Quantification of cytomegalovirus DNA in
peripheral blood leukocytes by a branched-DNA signal amplification assay. J. Clin.
Microbiol. 1997 ; 35 : 2740-2744.

5. I. Devianne-Garrigue, JL. Pellegrin, I. Pellegrin, M. Dupon, JM. Ragnaud, HJA. Fleury.
Follow-up of CMV cellular load using bDNA CMV in AIDS patients treated with foscarnet.
1997. 4th Conference on Retroviruses and Opportunistic Infections. P 320.