~) INSTITUTPASTEUR/ELsEVIER Paris 1992
Res. Virol. 1992, 143, 17-22
HIV1 quantitatiol in infected patients: a comparison of cell viraemia, plasma viraemia and R-HEV B. Masquelier, R. Faivre, M.-C. Paty and H.J.A. Fleury Laboratoire d f Virologic, Universit~ de Bordeaux II, 33076 Bordeaux Cedex (France)
SUMMARY Anti-HIV1 trials require sensitive, reproducible and reliable parameters which measure virus load and virus production. The aim of the present study of 14 patients at different CDC stages was to compare results provided by three techniques which evaluate, respectively, cell-associated HIV1 infectivity (cell viraemia), plasma-associated HIV1 infectivity (plasma viraemia) and radiation-resistant expression ex vivo (R-HEV). For each patient, TCD4 lymphocytes were numbered and an HIV1 antigenaemia determination was carried out. We found no correlation between cellular viraemia and R-HEV; on the contrary, data concerning plasma viraemia and R-HEV were significantly correlated. While cellular viraemia and R-HEV were not correlated with TCD4 levels, plasma viraemia exhibited a significant negative correlation with TCD4 counts. Despite theoretical and practical differences, plasma viraemia and R-HEV may measure events in the viral replication cycle which bear a close resemblance; however, as designed, R-HEV seems to be far less discrimina.tive than plasma viraemia.
Key-words: HIV, AIDS, Viraemia, Lymphocyte; CD4, Replication, R-HEV, Plasma and cell viraemia, Quantitation, Chemotherapy. INTRODUCTION Research and development of anti-HIV1 drugs is now one of the main challenges in the fight against AIDS. Phase 2 efficacy (antiviral) studies must thus be carried out on numerous putative anti-HIV1 compounds. For such trials, several parameters need to be studied: p24 antigenaemia has this far been the most frequently used, but it cannot measure HIVI replication with good accuracy (Paul et al., 1987; Masquelier et el., 1991). Viral quantitation, that is, ceil-associated or plasma virus infectivity, would be a better parameter to evaluate the in vivo efficacy of a
Submitted August 13, 1991, acceptedDecember2, 1991.
drug. Several methods of titration have recently been proposed, and the present study will attempt to compare two of them. 1) The first involves ANRS consensus protocoles: such techniques are currently being studied by a French medical virology group in a collaborative work supported by the ANRS (Agence nationale de Recherche sur le SIDA) based on previously described techniques (Coombs et el., 1989; Ho et al., 1989) for separately evaluating cell-associated HIVI and plasma-associated HIV1 infectivity. 2) The second method (Mathez et al., 1989) measures radiation-resistant HIV1 expression ex
B. MASQUELIER E T AL.
18
vivo (R-HEV) using gamma-irradiation o f infected ceils. Because g a m m a - i r r a d i a t i o n arrests cell division, this technique should evaluate patient cells engaged in active HIV1 expression in vivo; radioresistant cells can synthesize in vitro HIVI genomic R N A a n d propagate infected material to target cells during a 24-28-h post-irradiation period. In contrast, radiosensitive cells carrying latent HIV1 provirus can no longer produce infectious particles. Both methods were tested and compared using samples f r o m 14 H I V l - i n f e c t e d patients.
1 ml at a concentration of 1 x 106/ml for the 1st dilution and 2 × 106/ml for the next 5 dilutions. Each dilution was tested in quadruplicate. B medium was changed twice a week (0.75 ml); 3× 105 fresh cells per well were added at days 7 and 14 and cultures were maintained for 28 days. p24 (Core protein) production was checked in culture supernatants at days 14 and 28 using DuPont p24 EIA. Cell-associated HIV quantitation was calculated in TCID50/106 cells according to Reed and Muench (1938).
2) Plasma-associated HIV1 quantitation (plasma
viraemia)
MATERIALS AND METHODS Patients
Fourteen HIVl-antibody carriers, medically followed at the "CHRU de Bordeaux", provided samples for this study; 7 of them were asymptomatic and 7 had AIDS or AIDS-related illnesses. T C D 4 lymphocytes and HIV1 antigenaemia
TCD4 lymphocytes were numbered by flow cytometry; HIVI antigenaemia (HIVI-Ag) was determ[ned by an enzyme immunoassay (Abbott-EIA).
Plasma was separated from samples of 30 ml heparinized blood by centrifugation (15 min, 1,200 g). Plasmas were submitted to a second centrifugation (3,000 g 15 min) and then distributed in 24-well plates in a volume of 0.5 ml at 6 dilutions in B medium. All dilutions were tested in quadruplicate; dilutions were l / l , 1/5, 1/25, 1/125, 1/625 and 1/3125. Healthy donor PBMC were added to a volume of I ml per well, 2 × 106 cells/well. Twice a week, the medium was changed and fresh cells were added (3 × 105 per well) at days 7, 14 and 21 ; p24 production was checked as above at days 14 and 28 and plasma viraemia was calculated in TCIDs0/ml.
A N R S consensus quantitation protocoles
1) Cell-associated HIV1 quantitation viraemia)
(cellular R-HEV assay
Healthy donor PHA-stimulated PBMC (kindly provided by the "Centre R~gional de Transfusion Sanguine", Bordeaux) were added to a volume of
The R-HEV score was determined as previously described (Mathez et al., 1989). Briefly, 2.5 × 10s fresh PBMC from patients were suspended in I-2 ml of B medium and exposed to a converged f'vcedsource of gamma-rays within a biological laboratory irradiator (INSERM-U.306, Bordeaux, courtesy of Dr. Brothier) for a total dose of 45 grays. Irradiated cells were distributed in 24-well plates. For each sample, 4 wells with 5 × 105 cells/well and 4 wells with l × 105 cells/well were cultivated with a total volume of l ml B-medium/well; 5×I05 PHAstimulated PBMC from healthy donors were added to each replicate; l ml of B medium was added at day 5 and cultures were maintained for 10 days. p24
ANRS "Agence nationale de Recherche sur le SIDA". CDC = Centers for Disease Control. HIVi-Ag = HIVI antigenaemia.
PBMC = peripheral blood mononuclear cell. PHA -- phytohaemagglutin. R-HEV = radiation-resislantexpression ex vivo.
Peripheral blood mononuclear cells (PBMC) from patients' blood were isolated by Ficoll gradient and seeded in 24-well plates (Costar) in RPMI-1640 culture medium supplemented with faetal calf serum, antibiotics, interleukin-2 and anti-c×interferon serum (B medium). Patient PBMC were distributed in a volume of 0.5 ml at 6 dilutions from a starting concentration of 2 × 106 cells/m[. Dilutions were I / l , I/5, 1/25, 1/125, 1/625 and 1/3125.
19
HIV1 Q U A N T I T A TION I N INFECTED P A T I E N T S
Table I. Results of cellular and plasma viraemia and R-HEV scores.
Patient
Cellular viraemia TCIDso/106 cells
Plasma viraemia TCIDs0/ml
1 2 3 4
5,000 500 5,000 500
100 0 1,000 100
5 6
500 0
0 0
R-HEV score
0 50 87 100
TCD4 /~l
% % % °70
12 429 271 189
I00 070 0 070
354 770
0 070
HIV1-Ag (-) (- ) (+) (-) (- )
Clinical status (CDC) IV A II II IV C l
(- )
II II
7
> 7,000
0
228
(- )
II
8 9 I0 II
> 7,000 125 3 > 7,000
I0,000 250 250 > 14,000
I00 100 0 I00
°70 070 070 %
224 35 235 210
(- ) (+) (- ) (-)
II IVCI, D II IVB, Ct
12 13
625 3,125
> 14,000 6,250
100 070 100 07o
003 000
(-) (+)
IVB, C~ IVA, C 2, E
14
280
250
62 070
185
(-)
Ill
TCD4 counts weremadeusinga cytofluorimeter(FACscan. Becton-Dickinson).HIV-I antigenaemia(HIVI-Ag) was determinedusing a sandwichimmunoassaytest (Abbott, HIVAG-I).
production was checked in supernatants at day 10. The R-HEV score was expressed as percent p24-positive wells at day 10.
Statistical analysis Data from the different techniques were compared using the Spearmann test according to Siegel et ai., 1989. The statistical analysis was done at INSERM-U.330, Bordeaux. RESULTS All results concerning cellular viraemia, plasm a viraemia and R - H E V are summarized in table I. All patients except patient 6 had positive cellular viraemia. Patient 6 also had negative plasm a viraemia and an R - H E V score o f 0 °70, with a high CD4 count and negative p24 antigenaemia. Three patients (8, 11 and 12) had very high values o f plasma viraemia and R-HEV scores at 100 070, with surprisingly negative p24 antigenaemia.
All patients with negative plasma viraemia were asymptomatic (CDCII). A m o n g 4 patients (1, 6, 7 and 10) with R - H E V scores o f 0 070, 3 were asymptomatic, but one (1) showed a low CD4 count and developed AIDS-related disease. F r o m these results, we attempted to compare and find some correlation between results using A N R S techniques and the R-HEV scores.
Comparison of cellular viraemia and R-HEV C o m p a r a t i v e data are presented in figure 1. We found no significant correlation between cellular viraemia (TCID50) and R - H E V scores (r = 0.212, S p e a r m a n n ' s test). Results on 2 patients (1 and 7) showed a total dissociation between the two.
Comparison of plasma viraemia and R-HEV Comparative data are shown in figure 2. We f o u n d significant correlation between these two parameters ( r = 0 . 5 5 8 , p < 0.05, S p e a r m a n n ' s
20
B. MASQUELIER E T AL. CELLULAR lOS VIREMIA (TCID50/106 cells)
1i
10000
1000
•
t
100
I0.
R-HEY SCORE([)
Fig. 1. Comparison of cellular viraemia versus R-HEV score. r = 0.212, Spearmann's test (no correlation).
test). Two patients showed total dissociation (2 and 5), while 8 (3, 6, 7, 8, 11, 12, 13 and 14) had concordant results. No conclusions could be reached concerning the other 4.
Comparison of cellular viraemia, plasma viraemia, R-HEV and TCD4 count Plasma viraemia showed a significant negative correlation with TCD4 count (r = - 0.571, p < 0.05, Spearmann's test); cellular viraemia (r = - 0.229) and R-HEV (r = - 0.406) were not significantly correlated with TCD4 levels.
peripheral viral load, taking into account productively infected as well as latently infected cells due to possible proviral activation during the culture test. R-HEV measures only HIV1 cell load, corresponding to cells with active viral transcription at the time when samples are collected.
Plasma viraemia/R-HEV
Cell viraemia/R-HEV
Correlation between these two parameters was expected, since both techniques are considered to measure HIV replication activity; our data confirm that they corresponded to closely related events. However, some of the results (on patients 1, 2, 5, 9 and 10) appeared to be dissociated. This could be to different features.
Here, no correlation was anticipated, since cellular viraemia and R-HEV do not measure the same biological events. Cellular viraemia implies
1) Plasma viraemia and R-HEV measure HIV1 replication activity in PBMC and plasma. Dissociation between certain results could be due to the in vivo difference between HIV-replicating
DISCUSSION
HIV1 Q U A N T I T A T I O N I N I N F E C T E D P A T I E N T S PLASNA VIREHIA
21
10s'
(TC I Oso/ml )
I
10000
1000'
100
10.
d.s
z'~
,b
-'5 A
,;.i
,~0
R-HEV SCORE (g)
Fig. 2. Comparison of plasma viraemia versus R-HEV score. r=0.558, p < 0.05, Spearmann's test (significant correlation).
cells and viral particles still circulating in the plasma. In terms of kinetics, the R-HEV score could measure replicating events over a very short time, during which plasma viraemia corresponds to more cumulative replicating events. 2) Plasma viraemia would appear to be more accurate than R/HEV. Indeed, for 4 patients with a negative R-HEV score, plasma viraemia was between 0 and 250 TCIDs0. For 7 patients with an R-HEV score of 100 %, plasma viraemia had values of 0, 100, 250, 6,250, 10,000 and > 14,000 (2 times). From a technical point of view, this lack of discrimination in the R-HEV score might be explained as follows: - - p24 monitoring at day 10 (versus day 28 for plasma viraemia) may decrease the sensitivity of the method; - - t h e number of dilutions tested (2 for R-HEV score and 6 for plasma viraemia) would favour the latter technique; it is worth noting that the R-HEV score was recently modified with a greater number of dilutions (Leibowitch et aL, personal communication).
Comparison of results of quantitation assays with TCD4 counts
Plasma viraemia (Rouzioux et al., 1991), R-HEV (Mathez et al., 1990) and cellular viral load (Venet et al., 1991) have been shown to be correlated with CD4 counts; in our study, we found a negative correlation between plasma viraemia and TCD4, but we failed to find one between R-HEV and TCD4, perhaps because of the low number of patients enrolled. In the same way, an absence of significant correlation between cellular viraemia and TCD4 was observed: cellular viraemia is a positive marker regardless of the stage of HIV infection, and is thus useful in the monitoring of anti-HIV trials; however, its correlation with HIV disease evolution is less clearly established than for plasma viraemia. In conclusion, plasma viraemia and R-HEV seem to be correlated, thus indicating that they involve closely related events in the replication cycle; however, in our series, R-HEV yielded less
22
B. M A S Q U E L I E R E T A L .
discriminating data than plasma viraemia. A longitudinal follow-up would be useful in comparing the value o f the two groups o f techniques tested in terms o f H I V I infection monitoring. Longitudinal studies using ANRS techniques on AZT-treated patients are currently in progress, and should provide biologists and clinicians with rapid i n f o r m a t i o n on the use o f quantitation m e t h o d s for clinical trials.
Acknowledgements We wish to thank the group "Vir~mie Quantitative" of "Action Coordonn/'e n* 11" (AC 11) (C. Rouzioux, J. Puel, H. Agut, F. Brun-V~zinet,F. Ferchal, H. Fleury, C. Tamalet) of the "Agence Nationale de Recherche sur le SIDA" (ANRS), J. Leibowitch and D. Mathez for having received and taught the R-HEV technique to one of us (BlVl)and J.M. Huraux (AC 11/ANRS coordinator) and D. Dormont (CEA Fontenay aux Roses) for critical reading of the manuscript. We thank G. Chene (INSERM-U.330) for assistance with statistical analysis, and F. Latapie and J. Bar,re for preparation of the manuscript and figures.
Quantification du HIV1 chez des patients infect~s: comparaison de la vir~mie cellulaire, de la vir~mie plasmatique et du R-HEV Les essais antiviraux de l'infection HIVI n~cessitent l'obtention de param~tres sensibles, reproductibles et fiables, permettant d'~valuer la charge et la production virales. Le but de la pr~sente ~tude ~tait de comparer chez 14 patients ~ diffc~rents stades du CDC les r~sultats de trois techniques qui ~valuent respectivement la charge virale (vir~mie cellulaire), la production virale (vir~mie plasmatiq'ue) et l'expression radior~sistante ex vivo (R-HEV). Pour chaque patient, les lymphocytes TCD4 ont ~t~ num&~s et l'antig6n~mie HIV1 d~termin~e. Nous n'avons pas trouv~ de correlation entre la vir~mie cellulaire et le R-HEV; par contre, les donn~w.s de la vir~mie plasmatique et du R-HEV ont ~t~ significativement corrc~16es. Alors que la vir~mie cellulaire et le R-HEV n'ont pas ~t6 corral,s avec le nombre de lymphocytes TCD4, la vir~mie plasmatique a pr~sent~ une correlation n~gative significative avec ces m~mes lyrnphocytes. En d~pit de diff&ences th~oriques et pratiques, la vir6mie plasmatique et le R-HEV pourraient me.surer des ~v6nements du cycle de r~plication qui ont une signification proche, mais, tel qu'il
est coneu, le R-HEV semble &re nettement moins discriminant que la vir6mie plasmatique. Mots-clds: VIH, SIDA, Vir~mie, Lymphocyte; CD4, R6plication, R-HEV, Vir~mies plasmatique et cellulaire, Mesures, Chimioth6rapie.
References Coombs, R.W., Collier, A.C., Allain, J.P., Nikora, B., Leuther, M., Gjerset, G.F. & Corey, L. (1989), Plasma viremia in human immunodeficiency virus infection. New Engl. J. Med., 321, 1626-31. Ho, D.D., Mougdil, T. & Alam, M. (1989), Quantitation of human immunodeficiencyvirus type I in the blood of infected persons. NewEngl. J. Med., 321, 1621-25. Masquelier, B., Combeau, T., Poveda, J.P., Delord, B., Pellegrin, J.L., Sallafranque-Andreola, M.L., Tarrago-Litvak, L. & Fleury, H.J.A. (1991), In vitro assays show a dissociation of reverse transcriptase activity and core antigen (p24) production in two HIV-I isolates from a patient receiving long-term treatment with Zidovudine (ZDV). J. AIDS, 4, 499-505. Mathez, D., Paul, D., De Belilovsky,C., Sultan, Y., Gorin, I., Deleuze, J., Saurin, W., Decker, R. & Lelbowitch, J. (1989), HIV expression levels ex vivo: a surrogate marker for clinical end points in the evaluation of AIDS treatments? Colloque des Cents-Gardes, Pasteur-M&ieux, Marnes-La-Coquette,France, October 22. Mathez, D., Paul, D., De Belilovsky, C., Sultan, Y., Deleuze, J., Gorin, I., Saurin, W., Decker, R. & Leibowitch, J. (1990), Productive HIV infection levels correlates with AIDS-related manifestations in ,the patient. Proc. nat. Acad. Sci. (Wash.), 87, 7438-42. Paul, D.A., Falck, L.A., Kessler, H.A., Chase, R.M., Blaauw, B., Chudwin, D.S. & Landay, A.L. (1987), Correlation of serum HIV antigen and antibody with clinical status in HIV-infectedpatients. J. med. Virol., 22, 357-63. Reed, L.J. & Muench, H.A. (1938), A simple method of estimating fifty per cent end points. Amer. J. Hyg., 27, 493-97. Rouzioux, C., Puel, J., Agut, H., Brun-V6zinet, F., Ferchal, F., Fleury, H. & Tamalet, C. (1991), Groupe tt Virc~miequantitative ~>,Action Coordonn~e n ° 11, Agence Nationale de Recherches sur le SIDA. Quantitative plasma viremia: comparison of two techniques. Abstract MA 1107 VIIth International Conference on AIDS, Florence. Siegel, S. & Castellan, N.J. (1989), Non parametric statistics for the behavioral sciences, second edition. McGraw Hill Book Co. Gmbk, Hamburg. Venet, A., Lu, W., Bedjord, K. & Andrieu, J.M. (1991), Correlation between CD4 cell counts and cellular and plasma viral load in HIV-I seropositive individuals. AIDS, 5, 283-288.
Res. Virol. 1992, 143, 17-22
HIV1 quantitatiol in infected patients: a comparison of cell viraemia, plasma viraemia and R-HEV B. Masquelier, R. Faivre, M.-C. Paty and H.J.A. Fleury Laboratoire d f Virologic, Universit~ de Bordeaux II, 33076 Bordeaux Cedex (France)
SUMMARY Anti-HIV1 trials require sensitive, reproducible and reliable parameters which measure virus load and virus production. The aim of the present study of 14 patients at different CDC stages was to compare results provided by three techniques which evaluate, respectively, cell-associated HIV1 infectivity (cell viraemia), plasma-associated HIV1 infectivity (plasma viraemia) and radiation-resistant expression ex vivo (R-HEV). For each patient, TCD4 lymphocytes were numbered and an HIV1 antigenaemia determination was carried out. We found no correlation between cellular viraemia and R-HEV; on the contrary, data concerning plasma viraemia and R-HEV were significantly correlated. While cellular viraemia and R-HEV were not correlated with TCD4 levels, plasma viraemia exhibited a significant negative correlation with TCD4 counts. Despite theoretical and practical differences, plasma viraemia and R-HEV may measure events in the viral replication cycle which bear a close resemblance; however, as designed, R-HEV seems to be far less discrimina.tive than plasma viraemia.
Key-words: HIV, AIDS, Viraemia, Lymphocyte; CD4, Replication, R-HEV, Plasma and cell viraemia, Quantitation, Chemotherapy. INTRODUCTION Research and development of anti-HIV1 drugs is now one of the main challenges in the fight against AIDS. Phase 2 efficacy (antiviral) studies must thus be carried out on numerous putative anti-HIV1 compounds. For such trials, several parameters need to be studied: p24 antigenaemia has this far been the most frequently used, but it cannot measure HIVI replication with good accuracy (Paul et al., 1987; Masquelier et el., 1991). Viral quantitation, that is, ceil-associated or plasma virus infectivity, would be a better parameter to evaluate the in vivo efficacy of a
Submitted August 13, 1991, acceptedDecember2, 1991.
drug. Several methods of titration have recently been proposed, and the present study will attempt to compare two of them. 1) The first involves ANRS consensus protocoles: such techniques are currently being studied by a French medical virology group in a collaborative work supported by the ANRS (Agence nationale de Recherche sur le SIDA) based on previously described techniques (Coombs et el., 1989; Ho et al., 1989) for separately evaluating cell-associated HIVI and plasma-associated HIV1 infectivity. 2) The second method (Mathez et al., 1989) measures radiation-resistant HIV1 expression ex
B. MASQUELIER E T AL.
18
vivo (R-HEV) using gamma-irradiation o f infected ceils. Because g a m m a - i r r a d i a t i o n arrests cell division, this technique should evaluate patient cells engaged in active HIV1 expression in vivo; radioresistant cells can synthesize in vitro HIVI genomic R N A a n d propagate infected material to target cells during a 24-28-h post-irradiation period. In contrast, radiosensitive cells carrying latent HIV1 provirus can no longer produce infectious particles. Both methods were tested and compared using samples f r o m 14 H I V l - i n f e c t e d patients.
1 ml at a concentration of 1 x 106/ml for the 1st dilution and 2 × 106/ml for the next 5 dilutions. Each dilution was tested in quadruplicate. B medium was changed twice a week (0.75 ml); 3× 105 fresh cells per well were added at days 7 and 14 and cultures were maintained for 28 days. p24 (Core protein) production was checked in culture supernatants at days 14 and 28 using DuPont p24 EIA. Cell-associated HIV quantitation was calculated in TCID50/106 cells according to Reed and Muench (1938).
2) Plasma-associated HIV1 quantitation (plasma
viraemia)
MATERIALS AND METHODS Patients
Fourteen HIVl-antibody carriers, medically followed at the "CHRU de Bordeaux", provided samples for this study; 7 of them were asymptomatic and 7 had AIDS or AIDS-related illnesses. T C D 4 lymphocytes and HIV1 antigenaemia
TCD4 lymphocytes were numbered by flow cytometry; HIVI antigenaemia (HIVI-Ag) was determ[ned by an enzyme immunoassay (Abbott-EIA).
Plasma was separated from samples of 30 ml heparinized blood by centrifugation (15 min, 1,200 g). Plasmas were submitted to a second centrifugation (3,000 g 15 min) and then distributed in 24-well plates in a volume of 0.5 ml at 6 dilutions in B medium. All dilutions were tested in quadruplicate; dilutions were l / l , 1/5, 1/25, 1/125, 1/625 and 1/3125. Healthy donor PBMC were added to a volume of I ml per well, 2 × 106 cells/well. Twice a week, the medium was changed and fresh cells were added (3 × 105 per well) at days 7, 14 and 21 ; p24 production was checked as above at days 14 and 28 and plasma viraemia was calculated in TCIDs0/ml.
A N R S consensus quantitation protocoles
1) Cell-associated HIV1 quantitation viraemia)
(cellular R-HEV assay
Healthy donor PHA-stimulated PBMC (kindly provided by the "Centre R~gional de Transfusion Sanguine", Bordeaux) were added to a volume of
The R-HEV score was determined as previously described (Mathez et al., 1989). Briefly, 2.5 × 10s fresh PBMC from patients were suspended in I-2 ml of B medium and exposed to a converged f'vcedsource of gamma-rays within a biological laboratory irradiator (INSERM-U.306, Bordeaux, courtesy of Dr. Brothier) for a total dose of 45 grays. Irradiated cells were distributed in 24-well plates. For each sample, 4 wells with 5 × 105 cells/well and 4 wells with l × 105 cells/well were cultivated with a total volume of l ml B-medium/well; 5×I05 PHAstimulated PBMC from healthy donors were added to each replicate; l ml of B medium was added at day 5 and cultures were maintained for 10 days. p24
ANRS "Agence nationale de Recherche sur le SIDA". CDC = Centers for Disease Control. HIVi-Ag = HIVI antigenaemia.
PBMC = peripheral blood mononuclear cell. PHA -- phytohaemagglutin. R-HEV = radiation-resislantexpression ex vivo.
Peripheral blood mononuclear cells (PBMC) from patients' blood were isolated by Ficoll gradient and seeded in 24-well plates (Costar) in RPMI-1640 culture medium supplemented with faetal calf serum, antibiotics, interleukin-2 and anti-c×interferon serum (B medium). Patient PBMC were distributed in a volume of 0.5 ml at 6 dilutions from a starting concentration of 2 × 106 cells/m[. Dilutions were I / l , I/5, 1/25, 1/125, 1/625 and 1/3125.
19
HIV1 Q U A N T I T A TION I N INFECTED P A T I E N T S
Table I. Results of cellular and plasma viraemia and R-HEV scores.
Patient
Cellular viraemia TCIDso/106 cells
Plasma viraemia TCIDs0/ml
1 2 3 4
5,000 500 5,000 500
100 0 1,000 100
5 6
500 0
0 0
R-HEV score
0 50 87 100
TCD4 /~l
% % % °70
12 429 271 189
I00 070 0 070
354 770
0 070
HIV1-Ag (-) (- ) (+) (-) (- )
Clinical status (CDC) IV A II II IV C l
(- )
II II
7
> 7,000
0
228
(- )
II
8 9 I0 II
> 7,000 125 3 > 7,000
I0,000 250 250 > 14,000
I00 100 0 I00
°70 070 070 %
224 35 235 210
(- ) (+) (- ) (-)
II IVCI, D II IVB, Ct
12 13
625 3,125
> 14,000 6,250
100 070 100 07o
003 000
(-) (+)
IVB, C~ IVA, C 2, E
14
280
250
62 070
185
(-)
Ill
TCD4 counts weremadeusinga cytofluorimeter(FACscan. Becton-Dickinson).HIV-I antigenaemia(HIVI-Ag) was determinedusing a sandwichimmunoassaytest (Abbott, HIVAG-I).
production was checked in supernatants at day 10. The R-HEV score was expressed as percent p24-positive wells at day 10.
Statistical analysis Data from the different techniques were compared using the Spearmann test according to Siegel et ai., 1989. The statistical analysis was done at INSERM-U.330, Bordeaux. RESULTS All results concerning cellular viraemia, plasm a viraemia and R - H E V are summarized in table I. All patients except patient 6 had positive cellular viraemia. Patient 6 also had negative plasm a viraemia and an R - H E V score o f 0 °70, with a high CD4 count and negative p24 antigenaemia. Three patients (8, 11 and 12) had very high values o f plasma viraemia and R-HEV scores at 100 070, with surprisingly negative p24 antigenaemia.
All patients with negative plasma viraemia were asymptomatic (CDCII). A m o n g 4 patients (1, 6, 7 and 10) with R - H E V scores o f 0 070, 3 were asymptomatic, but one (1) showed a low CD4 count and developed AIDS-related disease. F r o m these results, we attempted to compare and find some correlation between results using A N R S techniques and the R-HEV scores.
Comparison of cellular viraemia and R-HEV C o m p a r a t i v e data are presented in figure 1. We found no significant correlation between cellular viraemia (TCID50) and R - H E V scores (r = 0.212, S p e a r m a n n ' s test). Results on 2 patients (1 and 7) showed a total dissociation between the two.
Comparison of plasma viraemia and R-HEV Comparative data are shown in figure 2. We f o u n d significant correlation between these two parameters ( r = 0 . 5 5 8 , p < 0.05, S p e a r m a n n ' s
20
B. MASQUELIER E T AL. CELLULAR lOS VIREMIA (TCID50/106 cells)
1i
10000
1000
•
t
100
I0.
R-HEY SCORE([)
Fig. 1. Comparison of cellular viraemia versus R-HEV score. r = 0.212, Spearmann's test (no correlation).
test). Two patients showed total dissociation (2 and 5), while 8 (3, 6, 7, 8, 11, 12, 13 and 14) had concordant results. No conclusions could be reached concerning the other 4.
Comparison of cellular viraemia, plasma viraemia, R-HEV and TCD4 count Plasma viraemia showed a significant negative correlation with TCD4 count (r = - 0.571, p < 0.05, Spearmann's test); cellular viraemia (r = - 0.229) and R-HEV (r = - 0.406) were not significantly correlated with TCD4 levels.
peripheral viral load, taking into account productively infected as well as latently infected cells due to possible proviral activation during the culture test. R-HEV measures only HIV1 cell load, corresponding to cells with active viral transcription at the time when samples are collected.
Plasma viraemia/R-HEV
Cell viraemia/R-HEV
Correlation between these two parameters was expected, since both techniques are considered to measure HIV replication activity; our data confirm that they corresponded to closely related events. However, some of the results (on patients 1, 2, 5, 9 and 10) appeared to be dissociated. This could be to different features.
Here, no correlation was anticipated, since cellular viraemia and R-HEV do not measure the same biological events. Cellular viraemia implies
1) Plasma viraemia and R-HEV measure HIV1 replication activity in PBMC and plasma. Dissociation between certain results could be due to the in vivo difference between HIV-replicating
DISCUSSION
HIV1 Q U A N T I T A T I O N I N I N F E C T E D P A T I E N T S PLASNA VIREHIA
21
10s'
(TC I Oso/ml )
I
10000
1000'
100
10.
d.s
z'~
,b
-'5 A
,;.i
,~0
R-HEV SCORE (g)
Fig. 2. Comparison of plasma viraemia versus R-HEV score. r=0.558, p < 0.05, Spearmann's test (significant correlation).
cells and viral particles still circulating in the plasma. In terms of kinetics, the R-HEV score could measure replicating events over a very short time, during which plasma viraemia corresponds to more cumulative replicating events. 2) Plasma viraemia would appear to be more accurate than R/HEV. Indeed, for 4 patients with a negative R-HEV score, plasma viraemia was between 0 and 250 TCIDs0. For 7 patients with an R-HEV score of 100 %, plasma viraemia had values of 0, 100, 250, 6,250, 10,000 and > 14,000 (2 times). From a technical point of view, this lack of discrimination in the R-HEV score might be explained as follows: - - p24 monitoring at day 10 (versus day 28 for plasma viraemia) may decrease the sensitivity of the method; - - t h e number of dilutions tested (2 for R-HEV score and 6 for plasma viraemia) would favour the latter technique; it is worth noting that the R-HEV score was recently modified with a greater number of dilutions (Leibowitch et aL, personal communication).
Comparison of results of quantitation assays with TCD4 counts
Plasma viraemia (Rouzioux et al., 1991), R-HEV (Mathez et al., 1990) and cellular viral load (Venet et al., 1991) have been shown to be correlated with CD4 counts; in our study, we found a negative correlation between plasma viraemia and TCD4, but we failed to find one between R-HEV and TCD4, perhaps because of the low number of patients enrolled. In the same way, an absence of significant correlation between cellular viraemia and TCD4 was observed: cellular viraemia is a positive marker regardless of the stage of HIV infection, and is thus useful in the monitoring of anti-HIV trials; however, its correlation with HIV disease evolution is less clearly established than for plasma viraemia. In conclusion, plasma viraemia and R-HEV seem to be correlated, thus indicating that they involve closely related events in the replication cycle; however, in our series, R-HEV yielded less
22
B. M A S Q U E L I E R E T A L .
discriminating data than plasma viraemia. A longitudinal follow-up would be useful in comparing the value o f the two groups o f techniques tested in terms o f H I V I infection monitoring. Longitudinal studies using ANRS techniques on AZT-treated patients are currently in progress, and should provide biologists and clinicians with rapid i n f o r m a t i o n on the use o f quantitation m e t h o d s for clinical trials.
Acknowledgements We wish to thank the group "Vir~mie Quantitative" of "Action Coordonn/'e n* 11" (AC 11) (C. Rouzioux, J. Puel, H. Agut, F. Brun-V~zinet,F. Ferchal, H. Fleury, C. Tamalet) of the "Agence Nationale de Recherche sur le SIDA" (ANRS), J. Leibowitch and D. Mathez for having received and taught the R-HEV technique to one of us (BlVl)and J.M. Huraux (AC 11/ANRS coordinator) and D. Dormont (CEA Fontenay aux Roses) for critical reading of the manuscript. We thank G. Chene (INSERM-U.330) for assistance with statistical analysis, and F. Latapie and J. Bar,re for preparation of the manuscript and figures.
Quantification du HIV1 chez des patients infect~s: comparaison de la vir~mie cellulaire, de la vir~mie plasmatique et du R-HEV Les essais antiviraux de l'infection HIVI n~cessitent l'obtention de param~tres sensibles, reproductibles et fiables, permettant d'~valuer la charge et la production virales. Le but de la pr~sente ~tude ~tait de comparer chez 14 patients ~ diffc~rents stades du CDC les r~sultats de trois techniques qui ~valuent respectivement la charge virale (vir~mie cellulaire), la production virale (vir~mie plasmatiq'ue) et l'expression radior~sistante ex vivo (R-HEV). Pour chaque patient, les lymphocytes TCD4 ont ~t~ num&~s et l'antig6n~mie HIV1 d~termin~e. Nous n'avons pas trouv~ de correlation entre la vir~mie cellulaire et le R-HEV; par contre, les donn~w.s de la vir~mie plasmatique et du R-HEV ont ~t~ significativement corrc~16es. Alors que la vir~mie cellulaire et le R-HEV n'ont pas ~t6 corral,s avec le nombre de lymphocytes TCD4, la vir~mie plasmatique a pr~sent~ une correlation n~gative significative avec ces m~mes lyrnphocytes. En d~pit de diff&ences th~oriques et pratiques, la vir6mie plasmatique et le R-HEV pourraient me.surer des ~v6nements du cycle de r~plication qui ont une signification proche, mais, tel qu'il
est coneu, le R-HEV semble &re nettement moins discriminant que la vir6mie plasmatique. Mots-clds: VIH, SIDA, Vir~mie, Lymphocyte; CD4, R6plication, R-HEV, Vir~mies plasmatique et cellulaire, Mesures, Chimioth6rapie.
References Coombs, R.W., Collier, A.C., Allain, J.P., Nikora, B., Leuther, M., Gjerset, G.F. & Corey, L. (1989), Plasma viremia in human immunodeficiency virus infection. New Engl. J. Med., 321, 1626-31. Ho, D.D., Mougdil, T. & Alam, M. (1989), Quantitation of human immunodeficiencyvirus type I in the blood of infected persons. NewEngl. J. Med., 321, 1621-25. Masquelier, B., Combeau, T., Poveda, J.P., Delord, B., Pellegrin, J.L., Sallafranque-Andreola, M.L., Tarrago-Litvak, L. & Fleury, H.J.A. (1991), In vitro assays show a dissociation of reverse transcriptase activity and core antigen (p24) production in two HIV-I isolates from a patient receiving long-term treatment with Zidovudine (ZDV). J. AIDS, 4, 499-505. Mathez, D., Paul, D., De Belilovsky,C., Sultan, Y., Gorin, I., Deleuze, J., Saurin, W., Decker, R. & Lelbowitch, J. (1989), HIV expression levels ex vivo: a surrogate marker for clinical end points in the evaluation of AIDS treatments? Colloque des Cents-Gardes, Pasteur-M&ieux, Marnes-La-Coquette,France, October 22. Mathez, D., Paul, D., De Belilovsky, C., Sultan, Y., Deleuze, J., Gorin, I., Saurin, W., Decker, R. & Leibowitch, J. (1990), Productive HIV infection levels correlates with AIDS-related manifestations in ,the patient. Proc. nat. Acad. Sci. (Wash.), 87, 7438-42. Paul, D.A., Falck, L.A., Kessler, H.A., Chase, R.M., Blaauw, B., Chudwin, D.S. & Landay, A.L. (1987), Correlation of serum HIV antigen and antibody with clinical status in HIV-infectedpatients. J. med. Virol., 22, 357-63. Reed, L.J. & Muench, H.A. (1938), A simple method of estimating fifty per cent end points. Amer. J. Hyg., 27, 493-97. Rouzioux, C., Puel, J., Agut, H., Brun-V6zinet, F., Ferchal, F., Fleury, H. & Tamalet, C. (1991), Groupe tt Virc~miequantitative ~>,Action Coordonn~e n ° 11, Agence Nationale de Recherches sur le SIDA. Quantitative plasma viremia: comparison of two techniques. Abstract MA 1107 VIIth International Conference on AIDS, Florence. Siegel, S. & Castellan, N.J. (1989), Non parametric statistics for the behavioral sciences, second edition. McGraw Hill Book Co. Gmbk, Hamburg. Venet, A., Lu, W., Bedjord, K. & Andrieu, J.M. (1991), Correlation between CD4 cell counts and cellular and plasma viral load in HIV-I seropositive individuals. AIDS, 5, 283-288.
