Immunological and Virological Changes in Antiretroviral NaõÈve

Human Immunode®ciency Virus Infected Patients Randomized to

G-CSF or Placebo Simultaneously with Initiation of HAART

H. ALADDIN, H. ULLUM, T. KATZENSTEIN, J. GERSTOFT, P. SKINHéJ &

B. KLARLUND PEDERSEN

Departments of Infectious Diseases, Copenhagen University Hospital (Rigshopitalet), M7641, Tagensvej 20, 2200 Copenhagen N, Denmark

(Received 20 July 1999; Accepted in revised form 6 December 1999)

Aladdin H, Ullum H, Katzenstein T, Gerstoft J, Skinhùj P, Klarlund Pedersen B. Immunological and

Virological Changes in Antiretroviral NaõÈve Human Immunode®ciency Virus Infected Patients Randomized

to G-CSF or Placebo Simultaneously with Initiation of HAART. Scand J Immunol 2000;51:520±525

To determine the ef®cacy of combined G-CSF and highly active antiretroviral treatment (HAART), a

randomized, double blind, placebo controlled study was conducted. Treatment naive human immuno-

de®ciency virus (HIV) infected patients were randomized to receive either placebo or G-CSF (0.3 mg/ml, 3

times a week) for 12 weeks and HAART simultaneously. The trial was terminated prematurely after interim

analysis performed because of a case of severe encephalopathia in the G-CSF group. At that point 11 HIV

infected patients with a CD4� T cell count < 350/mm3 had been randomized to the G-CSF group (n� 6) or

placebo group (n� 5). In both groups plasma HIV RNA decreased signi®cantly in response to HAART.

However, plasma HIV RNA changed signi®cantly different between the two groups with the decrease being

less pronounced in the G-CSF group (P� 0.02). The concentrations of CD4� memory T cells and CD8� naive

and memory T cells increased in response to HAART, and there was a trend towards more pronounced

increases in several T-cell subpopulations in the G-CSF group. The CD56� NK cells increased signi®cantly

more in the G-CSF group compared with placebo (P� 0.000). All patients in the G-CSF group reported bone

pain. The present data do not support simultaneous administration of G-CSF with initiation of HAART in

treatment naive HIV infected patients.

Hassan Aladdin, Department of Infectious Diseases, M7641, Rigshospitalet, Tagensvej 20, 2200 Copenhagen N,

Denmark

INTRODUCTION

The introduction of highly active antiretroviral therapy

(HAART) has had a tremendous effect on both the morbidity

and the mortality in human immunode®ciency virus type 1 (HIV)

infected patients [1, 2]. Patients infected with HIV who receive

HAART achieve substantial reductions in plasma HIV RNA and

increases in CD4� cell count associated with clinical improve-

ment [1, 3, 4]. HAART also induces improvement in several

additional immune functions; however, immune reconstitution is

incomplete and may be insuf®cient in some cases [4]. This has

renewed the interest in possible adjunct immunostimulating

therapy to increase the degree of immunological reconstitution.

The recombinant form of G-CSF, a hematopoietic growth factor,

acts primarily on neutrophil precursors and mature neutrophils.

Studies of short-term treatment with G-CSF in patients infected

with HIV have demonstrated an increases in the CD4� cell count

[5] and a partial restoration of in vitro interleukin (IL)-2

production [6]. Use of recombinant G-CSF also increases neutro-

phil counts and decreases bacterial infections in neutropenic

HIV-infected patients [7, 8]. In a previous double blind, ran-

domized, placebo controlled study, 30 HIV infected patients on

stable HAART (24 weeks) were randomized to G-CSF or

placebo. G-CSF enhanced the CD4� cell count without inducing

changes in HIV RNA [9].

In this study we investigated the effect of G-CSF given

simultaneously with initiating HAART on viral load and

lymphocyte subsets in treatment naõÈve HIV infected patients.

Scand. J. Immunol. 51, 520±525, 2000

q 2000 Blackwell Science Ltd

METHODS

Patients and study design. The study was approved by the Danish

Board of Health and the local ethical committee. All patients gave

written informed consent before enrolling in the study. The study was

intended as a pilot study evaluating the safety and immunological

changes in a small group of HIV-1 infected patients. Based on a case

of severe encephalopathy in one G-CSF treated patient (to be published

in detail elsewhere), an interim analysis was performed. The results of

this analysis made us terminate the trial after the inclusion of 11 patients

(9 men, 2 women). The median age of the patients was 44 years (range,

35±46 years) in the G-CSF-treated group (n� 6) and 35 years (range,

25±55 years) in the placebo treated group (n� 5). One patient in the

placebo group and 2 in the G-CSF group had been diagnozed with AIDS

prior to enrolment. The inclusion criteria were a CD4� count < 350 cells/

mm3 and no prior antiretroviral treatment. All patients received at least

2 of the following reverse transcriptase inhibitors-zidovudine (AZT),

stavudine (d4T), lamivudine (3TC), nevirapine and 1 or 2 of the

following protease inhibitors-ritonavir, indinavir, or nel®navir (Table 1).

Patients were on a double blind basis randomized to G-CSF administered

at a dose of 0.3 mg/ml subcutaneously 3 times weekly for 12 weeks or

placebo medication with saline given at similar time schedule. All other

medications, other than HAART (e.g. prophylaxis for Pneumocystis

carinii pneumonia [PCP]) remained unchanged throughout the study.

Blood samples were collected for viral load quanti®cation at enrolment,

2, 4, 6, 8, and 12 weeks and for immune monitoring at enrolment, weeks

4, and 12. Based on severe side effects in one patient (to be published

elsewhere), an interim analysis was performed. The results of this

analysis made us terminate the trial after the inclusion of 11 patients.

Plasma HIV RNA. Plasma HIV RNA was quanti®ed in plasma using a

standardized reverse transcriptase polymerase chain reaction (RT-PCR)

assay (Amplicor HIV-1 Monitor; Roche Diagnostic System, Inc;

Branchburg, NJ, USA). Detection level was 20 copies/ml.

Flow cytometry. Whole blood was collected from patients and incu-

bated for 15 min with antibodies: phycoerythrin-Texas Red conjugated

CD4-ECD (clone SFCI2T4D11, Immunotech [T4]) and carbocyanine-5

conjugated CD8-Cy5 (clone DK25, Dako, Glostrup, Denmark) in

combination with [1]; phycoerythrin (PE)-conjugated CD45RA (clone

4KB5, Dako) or PE-conjugated CD45RO (clone UCHL1, Dako) and

¯uorescein isothiocyanate (FITC)-conjugated CD62L (clone DREG56,

Immunotech, Marseille, France); or [2] CD38 FITC (clone AT13/5,

Dako), CD28 PE (clone L293, Becton Dickinson, Oxnard, CA, USA).

Natural killer (NK) cells were distinguished with the following anti-

bodies and Cy5-conjugated CD3-Cy5 (clone UCHT1, Dako) in combi-

nation with PE-Cy5-conjugated CD56 (clone MY31, Becton Dickinson)

and FITC-conjugated CD16 (Leu-11a) (clone NKP15, Becton Dick-

inson). Labelled cells were then lysed for 10 min using FACS Lysing

solution (Becton Dickinson) and centrifuged for 5 min at 2000 r.p.m. The

pellet was resuspended in 0.5 ml 3% phosphate-buffered saline (PBS)

and analysed by ¯ow cytometry (Epics XL-MCL, Coulter, Florida,

USA). Computer analyses were carried out using PC Lysis software

(Becton Dickinson). To determine background staining, cells were

incubated with ¯uorescein/PE/carbocyanine-5-conjugated IgG1 (clone

679.1Mc7, Immunotech); IgG1 ECD (clone 679.1Mc7, Immunotech)

was the negative control. Lymphocytes were distinguished from mono-

cytes on the basis of their forward-versus-right-angle light scatters and

controlled by a FITC-PE-conjugated CD4 5 FITC-CD14 PE (clone 2D1,

MwP9)-stained sample. A lymphocyte gate was used for all analyses.

The CD4� lymphocyte count was independently measured in whole

blood by ¯ow cytometry (Epics XL-MCL, Coulter, FL, USA), applying

the following combination of Three-Color reagent antibody: CD4-

FITC/CD8-PE/CD3-Per-cp (Becton Dickinson).

Statistical analysis. The values measured were represented as mean

and standard deviation (SD), and were analysed in a two way ANOVA with

time, group and time ´ group as factors. If the ANOVA showed time ´group effects, values at different time points were compared between

groups applying Tukey adjusted post hoc tests.

RESULTS

Baseline characteristics, clinical characteristics, and patient

disposition

There were no signi®cant differences between the groups receiv-

ing G-CSF and placebo with respect to baseline values of

lymphocyte subpopulations, HIV RNA, or neutrophil counts

(Table 2).

This study was not designed to gather data about adverse

events or side effects associated with the use of G-CSF; however,

all patients in the group treated with G-CSF reported bone pain

(a labelled side effect) and ¯u like symptoms. One patient

developed severe encephalopathy (to be published elsewhere).

Plasma HIV RNA concentrations

The concentration of HIV RNA did not differ signi®cantly

between the 2 groups at enrolment. Both groups experienced

an immediate decrease in viral load as expressed in a signi®cant

effect of time in the ANOVA (Table 2, Fig. 1). However, in the

placebo group, this decrease was both more profound and more

sustained compared with the G-CSF group as indicated by an

signi®cant group effect in the ANOVA despite similar levels at

inclusion (Table 2, Fig. 1). A 3-log decrease was experienced in

both groups at week 12. However, the 3 month of follow up

revealed that the viral load remained low in the placebo group,

while 3 patients (patient 7, 8, and 10, Table 1) showed a 1-log

increase from week 12±24 in the G-CSF group (Table 2, Fig. 1).

In patient number 7 the reported increase in viral load was caused

by lack of compliance to HAART. This patient number 7 stopped

his protease inhibitor treatment at week 8 and continued with 2

nucleosides only thereafter (Table 1). Patients number 8 and

10 also showed increases in viral load, but reported both

full compliance when interviewed about their compliance to

treatment.

Concentrations of neutrophils and lymphocytes

Administration of G-CSF induced increases in neutrophil con-

centration compared to placebo (P� 0.06) (Table 2). Both

groups showed signi®cant increases in number of lymphocytes

following treatment with a tendency to a larger increase in the

G-CSF group (Table 2).

Concentrations of T-cell subpopulations and NK cells

The CD4� cell count during follow up increased in both

Immunological and Virological Changes 521

q 2000 Blackwell Science Ltd, Scandinavian Journal of Immunology, 51, 520±525

522

H.

Ala

ddin

etal.

q2000

Black

well

Scien

ceL

td,

Sca

ndin

avia

nJo

urn

al

of

Imm

unolo

gy,

51,

520

±525

Table 1. Viral load (HIV RNA), CD4� levels at enrolment and HAART treatment during the study

CD4� cell count Plasma HIV RNA

Patient (cells/mm3) (log10 (copies/ml)) AIDS Week-0 Week-4 Week-12 Week-24

1 201 4,8 No AZT, 3TC, ritonavir AZT, 3TC, ritonavir AZT, 3TC, ritonavir AZT, 3TC, ritonavir

2 223 4,4 No AZT, 3TC, ritonavir AZT, 3TC, ritonavir AZT, 3TC, ritonavir AZT, 3TC,

indinavir ritonavir, indinavir

3 284 4,9 No Indinavir 3TC, 4dT indinavir 3TC, 4dT indinavir 3TC, 4dT indinavir 3TC, 4dT

4 330 3,2 No 3TC, 4dT,nel®navir 3TC, 4dT, nel®navir 3TC, 4dT, nel®navir 3TC, 4dT, nel®navir

5 100 4,6 Yes 3TC, 4dT, 3TC, 4Dt 3TC, 4dT, 3TC, 4dT,

nel®navir, nevirapine nel®navir, nevirapine nel®navir, nevirapine nel®navir, nevirapine

6 164 4,2 No indinavir 3TC, 4dT, indinavir 3TC, 4dT indinavir 3TC, 4dT indinavir 3TC,

G-CSF G-CSF G-CSF nevirapine

7 568 4,9 No AZT, 3TC, 3TC, 4dT, 3TC, 4dT, G-CSF 3TC, 4dT

ritonavir, G-CSF ritonavir G-CSF

8 11 5,5 No AZT, 3TC, AZT, 3TC, AZT, 3TC, AZT, ritonavir, 3TC

ritonavir, G-CSF ritonavir, G-CSF ritonavir, G-CSF

9 150 5,4 Yes 4dT, nel®navir, 4dT, nel®navir, 4dT, nel®navir, 4dT, nel®navir,

nevirapine, G-CSF nevirapine, G-CSF nevirapine, G-CSF nevirapine

10 183 5,2 No AZT, nel®navir, AZT, nel®navir,3TC, AZT, nel®navir, AZT, 3TC

3TC, G-CSF G-CSF 3TC, G-CSF nel®navir

11 43 4,8 Yes Viracept, AZT,3TC, Viracept, AZT, 3TC, Viracept, AZT, 3TC, Viracept, AZT, 3TC,

G-CSF G-CSF G-CSF

AZT (Retrovir), 3TC (Epivir), 4dT (Stavudine).

Imm

unolo

gica

land

Viro

logica

lC

hanges

523

q2

00

0B

lackw

ellS

cience

Ltd

,S

can

din

avia

nJo

urn

al

of

Imm

un

olo

gy,

51,

520

±525

Table 2. Changes in concentration of lymphocyte subpopulation (cells/mm3) and viral load (copies/ml) in the two groups (GCSF/Placebo). The values are

given as mean 6 SD

Placebo (n� 5) G-CSF (n� 6)

WEEK WEEK P-value in two-way ANOVA

Variable 0 4 12 24 0 4 12 24 Time Group Time*group

Log10(HIVRNA) 4.40 6 0.17 2.58 6 0.03 1.48 6 0.21 1.42 6 0.26 5.01 6 0.45 3.13 6 1.05 1.88 6 0.76 2.95 6 0.19 0.000 0.02 0.2

Leukocytes 5.5 6 1.5 5.2 6 1.44 5.1 6 1.8 5.3 6 1.03 5.1 6 4.6 7.5 6 14.1 4.6 6 7.4 3.1 6 5.21 0.04 0.07 0.07

Lymphocytes 1.7 6 0.61 2.16 6 1.05 2.08 6 0.94 2.06 6 0.87 1.43 6 0.47 3.05 6 1.61 2.61 6 2.15 2.46 6 1.47 0.006 0.4 0.3

Neutrophils 3.12 6 1.05 2.44 6 0.65 2.4 6 0.87 3.24 6 1.42 2.63 6 1.13 9.64 6 10.38 3.96 6 3.63 2.38 6 0.52 0.08 0.06 0.09

CD4� 454 6 170 565 6 267 545 6 233 541 6 215 365 6 128 794 6 539 710 6 686 611 6 440 0.03 0.4 0.2

CD4� CD45RO� 62Lÿ 143 6 99 223 6 223 198 6 134 164 6 102 94 6 60 368 6 415 369 6 512 260 6 354 0.04 0.4 0.4

CD4� CD45RO� 62L� 176 6 45 222 6 62 202 6 59 213 6 64 211 6 138 323 6 211 251 6 191 150 6 148 0.09 0.4 0.2

CD4� CD45RA� 62Lÿ 13 6 22 27 6 35 11 6 13 9 6 8 36 6 44 77 6 85 72 6 97 68 6 131 0.6 0.2 0.3

CD4� CD45RA� 62L� 111 6 64 127 6 70 148 6 68 168 6 69 149 6 172 98 6 142 106 6 94 122 6 133 0.7 0.9 0.4

CD4� CD28� 212 6 111 227 6 138 268 6 100 241 6 111 110 6 79 275 6 241 311 6 270 248 6 241 0.03 0.6 0.2

CD4� CD38� 44 6 66 108 6 207 51 6 55 52 6 60 56 6 49 139 6 171 136 6 257 66 6 121 0.3 0.6 0.7

CD8� 671 6 261 885 6 497 863 6 444 854 6 446 570 6 257 1257 6 822 1085 6 957 938 6 646 0.02 0.4 0.3

CD8� CD45RO� 62Lÿ 379 6 115 504 6 324 452 6 210 431 6 237 358 6 234 759 6 511 607 6 609 415 6 411 0.08 0.3 0.4

CD8� CD45RO� 62L� 68 6 47 157 6 62 137 6 79 117 6 52 95 6 64 164 6 82 117 6 31 81 6 54 0.003 0.8 0.4

CD8� CD45RA� 62Lÿ 132 6 58 180 6 122 199 6 120 206 6 134 240 6 132 487 6 344 471 6 508 344 6 317 0.09 0.07 0.3

CD8� CD45RA� 62L� 61 6 42 104 6 51 100 6 46 127 6 64 63 6 48 98 6 73 91 6 38 88 6 63 0.000 0.9 0.3

CD8� CD28� 121 6 103 167 6 125 200 6 66 192 6 58 69 6 82 120 6 90 173 6 157 185 6 186 0.007 0.8 0.6

CD8� CD38� 297 6 215 383 6 406 306 6 244 307 6 259 295 6 105 583 6 436 509 6 432 321 6 240 0.02 0.4 0.2

NK/CD56� 35 6 15 30 6 22 30 6 28 17 6 12 48 6 33 118 6 76 73 6 61 60 6 34 0.0001 0.003 0.0001

groups with a tendency for larger increases in the G-CSF group.

The G-CSF group especially tended to recruit more

CD4�CD45RO�62 lÿ memory cells. Noteworthy was also a

total absence of increased levels of naõÈve CD4�CD45RA�62 l�

in the G-CSF group. However, the G-CSF group showed the

largest increase in CD4� cell counts with a normal expression of

the CD28� T-cell co-receptor (Table 2).

Both groups increased their total CD8� cell count with a trend

for a larger increase in the G-CSF treated patients. The increase

in the CD8� cell number resulted from a combination of

increases in numbers of memory CD8�CD45RO�62 l� and

naõÈve CD8�CD45RA�62 l� T cells. Also other subsets of

CD8� T cells expressing CD8�CD38� and CD8�CD28�

marker showed larger increases without any striking differences

between the groups (Table 2).

Natural killer cells showed a signi®cant increase following

G-CSF treatment (P< 0.0001). The concentration of CD3ÿ/

CD56� NK cells increased in the G-CSF group, but decreased

in the group on placebo. A follow up of 12 weeks revealed that

the concentrations of NK/CD56� cells remained higher than

pretreatment levels (Table 2).

DISCUSSION

The present randomized placebo controlled study aimed to

investigate the effect of G-CSF given simultaneously with

initiation of HAART. The trial was terminated prematurely

after interim analysis showed that G-CSF had an unwanted

effect on plasma HIV RNA. The viral load decreased signi®-

cantly in both groups as a consequence of HAART. However, 12

weeks of follow up showed a rebound in the viral load in 3

patients who received G-CSF. The rebound in the viral load in

one patient was caused by an initial lack of compliance and

subsequent complete cessation of HAART. The increased viral

load in the other 2 patients could be owing to insuf®cient viral

control by the regimen HAART caused by resistance develop-

ment. In a previous double blind, placebo controlled study,

including 30 patients on stable HAART for 6 months, 30 patients

were randomized to either G-CSF for 12 weeks or placebo. In the

latter study, G-CSF did not induce any changes in HIV RNA.

Many patients experience side effects when they start on anti-

retroviral treatment, but after some time the side effects usually

disappears, they either learn to tolerate the side effects or have to

be changed to other treatment combinations with less side

effects. It is therefore possible that the lack of compliance in

one patient has been caused by accumulated side effects to

both HAART and G-CSF. However, since an increase in the

viral load was observed also in 2 patients with reported complete

compliance to HAART a direct biological effect of G-CSF on

control of HIV replication can be suspected. This could possibly

be explained by an increase in the memory CD4 cells as potential

targets for HIV. The size of the present study and the biological

data do unfortunately not allow a de®nite conclusion to be drawn

on a biological effect of G-CSF leading to loss of viral control in

HAART treated patients.

The trend was that the G-CSF administration enhanced the

number of neutrophils and the number of several CD4� and

CD8� T-cell subpopulations in agreement with previous ®ndings

on G-CSF administration to patients on stable HAART [9] and in

agreement with ®ndings in patients with neutropenia [10] and

malignancy [11, 12]. The lymphocyte changes did not reach

statistical signi®cance. However, G-CSF induced a signi®cant

increase in the number of CD56� NK cells in accordance with

previous ®ndings [9].

In conclusion, although G-CSF administration may enhance

the number of CD4� T cells, the ®nding of a signi®cantly less

pronounced decrease in viral load in the G-CSF group make us

recommend that the G-CSF is at present not given simulta-

neously with initiation of HAART. The results in this study on

viral load may partly be ascribed to the G-CSF induced side

effects and a following lack of compliance with HAART.

524 H. Aladdin et al.

q 2000 Blackwell Science Ltd, Scandinavian Journal of Immunology, 51, 520±525

Fig. 1(A, B). HIV RNA levels in HIV infected patients randomized to

G-CSF or placebo for 12 weeks simultaneously with initiation of

HAART and were followed for 24 weeks. Detection levels were log

(20).

However, since an increase in viral load was also observed in 2

patients compliant to HAART, an additional direct biological

effect of G-CSF on viral replication can not be excluded.

Furthermore, administration of G-CSF as a double blind placebo

controlled study is impossible in this patient group owing to the

severe side effects they experience. Given the results on G-CSF

administration on stable HAART which was not associated with

a viral rebound [9], we suggest that possible future studies

evaluate the effect of intermittent G-CSF dose regimens

mainly in patients who have received HAART for a long

period without suf®cient immune reconstitution, and particularly

in patients with a decreased number of neutrophils during

HAART therapy.

REFERENCES

1 Autran B, Carcelain G, Li TS et al. Positive effects of combined

antiretroviral therapy on CD4� T-cell homeostasis and function in

advanced HIV disease. Science 1997;277:112±6.

2 Palella FJJ, Delaney KM, Moorman AC et al. Declining morbidity

and mortality among patients with advanced human immunode®-

ciency virus infection. HIV Outpatient Study Invest N Engl J Med

1998;338:853±60.

3 Kelleher AD, Carr A, Zaunders J, Cooper DA. Alterations in the

immune response of human immunode®ciency virus (HIV)-infected

subjects treated with an HIV-speci®c protease inhibitor, ritonavir. J

Infect Dis 1996;173:321±9.

4 Powderly WG, Landay A, Lederman MM. Recovery of the immune

system with antiretroviral therapy: the end of opportunism? JAMA

1998;280:72±7.

5 Nielsen SD, Afzelius P, Dam-Larsen S et al. Effect of granulocyte

colony-stimulating factor (G-CSF) in human immunode®ciency

virus-infected patients: increase in numbers of naive CD4 cells and

CD34 cells makes G-CSF a candidate for use in gene therapy or to

support antiretroviral therapy. J Infect Dis 1998:177:1733±6.

6 Hartung T, Pitrak DL, Foote M, Shatzen EM, Verral SC, Wendel A.

Filgrastim restores interleukin-2 production in blood from patients

with advanced human immunode®ciency virus infection. J Infect Dis

1998;178:686±92.

7 Hermans P, Rozenbaum W, Jou A et al. Filgrastim to treat neutro-

penia and support myelosuppressive medication dosing in HIV

infection. G-CSF 92105 Study Group AIDS 1996;10:1627±33.

8 Kuritzkes DR, Parenti D, Ward DJ et al. Filgrastim prevents severe

neutropenia and reduces infective morbidity in patients with

advanced HIV infection: results of a randomized, multicenter,

controlled trial. G-CSF 930101 Study Group AIDS 1998;12:65±74.

9 Aladdin H, Ullum H, Neilsen S et al. G-CSF increases CD4� T-cell

counts in HIV-infected patients on stable highly active antiretroviral

therapy (HAART): results from a randomized, placebo-controlled

study. J Inf Dis, in press.

10 Stricker RB, Goldberg B. Increase in lymphocyte subsets following

treatment of HIV-associated neutropenia with granulocyte colony-

stimulating factor. Clin Immunol Immunopathol 1996;79:194±6.

11 Gabrilove JL, Jakubowski A, Fain K et al. Phase I study of

granulocyte colony-stimulating factor in patients with transitional

cell carcinoma of the urothelium. J Clin Invest 1988;82:1454±61.

12 Kerrigan DP, Castillo A, Foucar K, Townsend K, Neidhart J.

Peripheral blood morphologic changes after high-dose antineo-

plastic chemotherapy and recombinant human granulocyte colony-

stimulating factor administration. Am J Clin Pathol 1989;92:

280±5.

Immunological and Virological Changes 525

q 2000 Blackwell Science Ltd, Scandinavian Journal of Immunology, 51, 520±525