Archivum lmmunologlac Cl Thcrapiac Expcrimcnralis. 2002. :;0. 111-120 PL ISSN 0004-06lJX
Review
Potential for Immune Reconstitution through G -CSF Treatment of HIV Patients
SONJA YON AULOCK and THO MAS HA RT UNG*
Biochemical Pharmacology. University of Konsranz. Konstanz.Germany
Abstract. New treatment strategies for HIY/ AIDS are very successful in reducing viral loa d. However. recon- stitution of the immune system takes about one year ami may be insu fficient or remain incomplete. During this time the patient remains prone to opportunistic infections as a result of the complex immunedysfunctioncaused by the virus. Rccombinant granulocytecolo ny-stimulating factor(G-CSF) has diverse imrnunomodulatory proper- ties whichmay be beneficial in aiding immune reconstitution.
Key words: HIY; AlDS~ G-CS F~ immune reconst itut ion.
Introduction
Acquired immunodeficiency syndrome (AIDS), a disease that was firstdescribed in the early 1980 's , is the consequence of an infection with human immune- deficiency virus(H!Y) . This pandemic is still sprea ding very fast, especially in Africa and Asia: worldwide, 36.1 million people were living with HIV/ AID S mthe year 2000, 5.3 million new infections and 3 million deaths were recorded in that year (hup:llwww.unaids. orglwac/2000/wadOO/filcs/WAD3pidemic_repon.htrn).
HIY infection results in complex immune dysfunc- tion. Although T helper (Th) cells, which are part of the specific immune system, arc the primary targets of HIV infection, cells of the non-specific immune sys- tem,such as macrophages and dendritic cells, are also infected. The numbers and functions of ncutrophils.
monocytes and lymphocytcs arc pathologically altered
in the course of HIY disease . Furthermore, HIV-in- fccted patients develop major defects in the immune signaling system, in particular the cyrokine-rnediator system". Imbalances in the network of cytokincs may ultimately trigger tile immune system collapse caused by HIyH. Abnormal cytokine production may explain not only the immune defects in HIV infection. butalso the developmentof clinical abnormalities ,such as ca- chexia, encephalitis. Kaposi sarcoma and lymphoma".
For a long time, zidovudinc (AZT) and other nucle- oside analogs were the only type of medication avail- able to slow the replication of HIY. However, they were also associated with severe side effects, including myelosuppression. Recent therapeutic protocols for HIY infection have introduced cocktails of drugs that counteract de 11OFO infection of cells (reverse transcrip- tase inhibitors) and prevent virus assembly in infected cells (protease inhibitors).This highly activeanti-rctro-
Abbreviations used: AIDS - acquired immunodeficiency syndrome. ARC - AIDS·relaled complex. CTL- cytotoxic T lympho cytc . G-CSF- granulocyrccolony-stimulatingIactor, HAART- highly active anti-rctroviralthcrupy.my - huma n immunodeficiency virus.
IFN- interferon. IL- intcrleukin.LPS - lipopolysacch aride, OKT-3- :lI1ti-CD3 antibody.['HA - phytohcrnagghninin. PBMC- periph- eral blood mononuclear cells, SEB - staphylococcalenterotoxin B. Th - T helper. TKF- turner necrosisfactor.
"Correspondence 10: Thomas llanung. M.D.PI1.D.. BiochemicalPharmacology. POB 655. University of Konstanz. 78457 Konstanz, Germany. tcl.: +49 75 31 8841 t6, fax: +49 75318841 17.c.mail:Thomas.Hanung@uni -ko nqanl..dc
First publ. in: Archivum Immunologiae et Therapiae Experimentalis 50 (2002), 2, pp. 111-120
Konstanzer Online-Publikations-System (KOPS) URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-79422
URL: http://kops.ub.uni-konstanz.de/volltexte/2009/7942/
112
Fig. I. Model of immune reconstitution under highly active anti- -rctro vira l therapy (HAART)
viral therapy (HAART) not only reduces viral burden drastically, but also affects T cell dynamics?'. The clinical benefits of HAART arc demonstrated by the resolving of opportunistic infections and malignancies, as well as declining hospital ization and mortality rates.
Howe ver, while substantial increases in CD4+ cell count and functional improvement are observed in pa- tients receiving HAART, normal values are generally still not attained after 2 years of therapy, despite sus- tained decreases in plasma viremiu to below detectable levels" (Fig. 1).
Consequently, the aims of an approach to supportive therapy forthe immunesystem would include fortifica- tion ofthe non-specific immune system and an acceler- ation of the increase in counts of functional CD4+ and CD8+ cells by stimulation of lymphopoiesis without concomitant activation and without unfuvorablc inter- actions with HAART drugs.
Granulocyte colony-stimulating factor (G-CSF) is a hematopoietic growth factor and pivotal cytokine in the immune mediator network. Recombinam human G-CSF is available in different forms: filgrastim (trade names:
Neupogen. Gran) is expressed in Escherichia coli and, therefore, unglycosylated; lenograstim (trade names:
Neutrogin, Granocyre) is expressed in CHO cells and therefore, glycosylated. G-CSF has been approved for the mobilization of stem cells for bone marrow trans- plantation or the treatment of neutropenia, e.g. as a consequence of chemotherapy. G-CSF is also ap- provedfor thetreatment of neutropenia inHlY-infectcd individuals in some countries. Howeve r, neutropenia has become a less frequent complication in HIY pa- tients since HAART was imple mented.In recent years, a number of immunomodularory effects of G-C SF treatment have been recognized which might aid im- mune reconstitution in HIY patients. These are sum- marized and discussed in this review.
The predominant neutrophil-stimulatory activity of G-CSF has been studied most intensively so far. In a number of settings, including various infectious dis- eases, both the induced neutrophilia and the augmenta- tion of neutrophil function have been found safe and, in some cases, beneficial". The role ofG-CSr in infec- tion and inflammation andthe potential clinical uses of G-CSF compared with those of granulocytc-macrophage colony-stimulating factor (GM-CSF) are discussed in detail in38. 39.
Data from animal models suggest that G-CSF may also support the immune system in its defense against diseases caused by opportunistic pathogens which are relevant in AIDS patients. Prophylactic administration Effects of G·CS F on Neutrophil Counts
and Functions and Innate Dcfcnsc Neutr openia, Neutrophil Dysfunction and Opportunistic Infections
Neutrophils pose the first line of immune defcnsc against invading pathogens. Studies in cancer patients receiving chemotherapy, which attacks all replicating tissue including leukocytes. have shown that the dura- tion and depth of neutropenia is correlated with the risk of infections. This also holds true for patients with ad- vanced HIY infcction'", Until the advent of HAA RT , neutropenia was common in the advanced stages of AIDS (75-90%), often caused directlybythe retroviral infection, by antiretrovirul and other drug therapy, sys- temic infections, and auroimmune mechanisms!" :5. 68.
Furthermore, a number of functional defects of neu- trophi!s have been reported for Hlv-infected patients, including impaired chemotaxis, phagocytosis, oxidative metabolism. intracellular enzyme activity and killing of bacterial and fungal pathogens and Hlv-infected cells, as well as accelerated apoptosis'" 26.55. 61. 70. 71.74.
Apoptosis and functional impairment of neutrophils may further exacerbate the characteristic underlying immune defects and, as with neutropenia, increase the risk of secondary and/or certain opportunistic infec-
rtons":
7J. 74. These are, in turn, associated withincreased levels of prointlammatory cytokines and concomitant viremia. a vicious cycle of virus replica- tion and cellular activation in response to this replica-
uon":
49. 86. Infections associated with neutropenia and neutrophil dysfunction in AIDS patients include bac- teremia and fungemia, pulmonary aspergillosis, pyomyositis, malignant external otitis, neutropenic en- terocolitis and pseudomonaskcratitis'" 45. 46. 63, M naivememory cells
1-2years _----,~Tnormal
- .
C04' cells
\ virus
\ copies
\
-"--
"-'"
8 weeks
tstart ofHAART
of G~CSF before infection with Streptococcus pneu- moniaesignificantly increase d surv ivalcompared with placebo treatment in rats and also in splenect omized rnice'". This latter study was associated withdecreased numbers of viable pneumococci recovered from tra- cheobronchial lymph nodes from G-CSF~treated mice compared with saline-treated control mice. When
G~CSFtreatment was initiated afterexperimentalinfec- tion with S, pneumoniaein rats, survival was signifi- cantly improved over controlrats". Even immunosup- pressed mice which received G-CSF after experimental pulmonaryinfection with Pneumocystis carinii also had increased survival compared with control mice that re- ceived a placebo".
Also, the combinationof G-CSF with antibiotics or antifungal treatmentseems to be more beneficialthan either treatment alone. LAzARD et al.:.4 showed that G-CSF in combination with clarithromycin was more effectivethanclarithromycinalone againstMycobacte- rium avium complex-related lung and spleen infection in mice. In a mouse model of disseminated candidiasis, prophylactic or postinfec tion treatmentwith G-CSFand fluconazole resulted in increased survivaland reduced kidney tissue counts of Candida compared with mice receiving fluconazole or G~CSF alone.". The mechan- isms underlying the survival benefits seen withG~CS F
administration in animal models of infect ion are most likelymultifacrori af" .
Functionaldefectsofneutrophilsfrom patients with HIV infection have been improved or corrected by in vitro or ill vivo administ ration of G-CSF, Defectivebac- teria-killing capacity of neutrophils frompatien ts with AIDSorAIDS·related complex (ARC) was restored in vitro as wellas ex vivo28.n. G-CSF also corrected the impairmentof respiratory burst ofP. carillii-stimula ted ncutrophils'".Exvivo, antifung alactivityof neut rophils from HIV-infected patients was significantlyenhanced after administration of G-CSF. This correlated with augmented superoxide anion and leukotriene produc- tionin response to pathogens'S 8R.Acceleratedsponta- neous ex vivo apoptosis of neutrophils from AIDS pa- tients was reduced by incubation with G_CSF14, Also, cytotoxic function of neutrophils from AIDS patients againstHtV-infectcdcells was markedlyaugmentedby G-CSFillvitro6,
Filgrastimtreatment hasbeenshownto reverseneu- tropenia inHIV-infected patientsand to decreaseinfec- tion and increase survival without increasing viral load27•41-..43. 48. 64--67. &-I. A retrospective cohort analys is found that G-CSFtreatmentfor neutropen ia was asso- ciated with prolongedsurvival inHlV-infected patients with disseminated M. avium complex infection". In-
113
creased survival was also seenin thegroup of patients who received clarithromycin and G-CSF versus clari- thromycin alone". Furthermore, a recentstudy showed that regularadministration of G-CSF (daily orintermit- tent) can actuallyprevent severe neutropeniain patients with HIV infection andthat itreduces infection-related morbidity".
Defects in Monocyte Function in HIVlA I DS
Both HIV-infected bone marrow stroma cells and monocytcs have shown a reduced ability to produce G-CSF and other hcmatopcieric growth factors in re- sponse to intcrlcukin la (IL-la)or lipopolysaccharide (LPS). However, secretion of proinflammarory cyto- kines (IL-l[3, lL-6, IL-8 and tumor necrosis factor a-TNF-a ) was upregulated incomparison with the re- sponse of uninfecredcontrol cells": 22. eo. 69. As there was no difference in the frequency or intensity of cell- -associated immunocytochemical cytokinc staining be- tween HIV and mock-infected monocyte-derived mac- rophages. the altered cytokine release seems to result not from dysregulated productive. but rather from de- funct secretory activity". TNF-aandinterferon y(IFN-y) have also beenshown to suppress proliferationof bone marrowcens" and. like IL-6, enhance HIV replication
ill vitro12,».75, This shift in the cytokine pattern may contribute to the pathogenesis of HIV-mediated dis- ease, e.g.by exacerbatingtissue damage associated with opportunistic infcctions-" ia,25, 69,
Therelease of high levels of these proinflammatory cytokines by isolated peripheralblood cells and tissue macrophages from HIV-infected patients cultured in vitro was observed in many, though not all, studies, reviewed in22• However, ex vivo LPS-inducible mono- kinerelease in wholeblood from Hl v -infected patients did not differ from that of normal or high-risk volun- reers" .
Prolonged periods of afebrile neutropenia in lIIV- -seropositive individuals did not induce an increase in G-CSF serum levcls'", which is the normal. adequate response seen in I-IIV-negative neutropenic patients". However, elevated serum levels of proinflammatory cy- tokines weredetected in Hlv-infectedpatients in some studies, thoughone study reported normal concentra- tions of cyrokines in serum, discussed in22. 53. Chronic production of monoki nes over the time span of many yearsmay be responsible for clinical manifestations such as cachexia, as demonstratedby the aruicachetic effect re- suiting from administrationof anti-lL--6 monoclonal anti- bodies toHIV-infected patients with lymphoma-" 86.
114
Anti-Inflammatory Effe cts of G-CSF Might Limit HIV Prog ression
Monocytes express G-CSFreceptors':' and respond
[0 G-CSFinan anti-inflammatory manner.G-CSF ad- ministered to healthy volunteers reduced the release of proinflammatory TNF-a , IL- l ~ and IL-12 per mono- cyte in responseto LPS stimulation of whole blood ex
vivo.Furthermore, the release of the anti-inflammatory soluble TNF receptors I and If as wellas IL-l receptor antagonist (Il.clra) wasaugmented'?-36, The reduction in TNF-u and IL-12 release by monocytes by G-CSF directly resultsinless activationoflymphocyresto pro- duce the proinflammatory cytokine IFN-yD,
Both HIV itself and opportunistic infections over- balance cyrokine regulation towards proinflammatory cytokine production together with diminished growth factor response. as describedabove.This suggests that supportivetherapy witha safe and well-toleratedhem a- topoieric growth factor such as G-CSF, whichhas anti- -inflammato ry properties, might help to rebalance cy- tokine relations andprevent infectionsby strengthening the nonspecific defense system, as wellas ameliorating morbidity associated with proinflamrnatory cytokines and breaking the vicious cycle of immune activation andvirus replication(Fig. 2). Although the suppressive effectof G-CSFaddedin vitro on LPS-inducible TNF-a and IFN-yformation in whole bloodof normal volun- teers wasnot found in bloodofHlv -infectcd patients" , this may be different when G-CSF is administered in vivoto HIV-infected patients.
infection7
T
Fig. 2. Putativeeffects of G-CSF ontheinnate immunesystemand on HIVreplication
Lymphope nia and Lymphocyte Dysfunc tion in lIlVIAIDS
In contrast to CD4+T lymphocyte depletion, the hallmark of HIV infection, a pronounced expansion of
CD8+T lymphocytes occurs early on during HIV in- fection' ", Anti-viral cytotoxic T lyrnphccytes (CTL) cons titute the strongest mechanism by which the im- mune system can partially control HIV spread during the latency period by two mechanisms: killing of in- fected cells, e.g. via perforin and granzymes, and re- leasing soluble mediators which interfere with viral replication, e.g. RANTES, macrophage inflammatory protein la and IW7. 20.Paradoxically,however,theCD8+
population becomes unresponsiveto T cellsignalingill vitro and displays decreased ill vitro cloning potential in thecourse of HIV infection.These findings suggest an increase in anergic or apoptotic CD8+T cells in HIV-infectedpcrsons-" 56.
The numbers of both naive and memory CD4+ and CD8+T cells, including the HIV-spcciflc CTL precur- sors,declinein the blood duri ngHlV disease progress- ion, reflecting alterations in T cell regenerationcapac- ity andvirus-mediated cell death.The loss ofmemory (specific)CD4+Tcell reactivityagainst recall antigens.
which is an early event in HIY disease progression, results inincreased susceptibilityto opportunisticinfec- tions and tumors. Primary responses of naive CD4+
cells towards new pathogens are suppressed evenear- lier in the course of HIVdisease- 23.19. 59. 79. 81.82.
Furthermore, the progression of HIV disease is characterized by a major T cell activat ion,as assessed by cell surface expression of various T cell activation markers such as CD2S (thea chain of IL-2 rcccptor), CD38or HLA-DR, on both CD4+and CD8+T cells5. Abnormalities inthe activationand functionalstates of B cells, macrophagesand dendritic cells have also been observed?' 16. 20,n. 53.
Based on theirrelease of different combinations of lymphokines, theTh cells havebeen grouped into ThI and Th2 cells.The Th l cytokines,IL-2 and IFN~y, are involved in stimulating the "cell-mediated responses"
that help the host eliminate cells infected with patho- gens. In contrast, Th2 cytckines, such as IL-4, IL~IO
and others, activate antibody production. Attenuated production ofIL-2 inadvanced HlY infection appears to play a major role in the loss of Th cells. IL-2 is considered to be the pri mary T cell growth factor. In addition to stimulating the proliferation of CD4+and CD8+lymphocytes in response to antigen activat ion, IL-2 also enhances cytolytic and oncolytic activity of suppressor cells, augments natural killer cell activity, and increases lFN-y production.Moreover. IL-2 forma- tion by CD4+T cellsappears to be mandatory in main- taining CD8'" T cell control of virus replication".A re- duction of Thl type lymphokinc (i.e. primarily IL-2) and concomitant increase of Th2 type lympbokine
(such asIL-4 and IL- lO)productionin response to mi- togens. such as phytohemagg lutinin (PHA), alloanti- gens, tetanus toxoid, etc., was suggested to characteriz e and predictthe change from asymptomatic HIV infec- tion to AIDS.However, there is also contradictory evi- deuce!': 15. 16. 32. 58, 78.
The addition of IL-2 in vitroto cultures of periph- eral blood mononuclear cells (PBMC) from HIV-in- fected individuals initiated a polyclonal expansion of CD4+and CD8+ cells and partially or completely re- stored certain defective cellular immune responses, including the proliferative response to antigen stimu- lation, and reduced the spontaneous and activation-in- duced programmed cell death, but was also associated with increases in HIYreplication and spread":49.Treat- ment of CD8+cells from long-term survivors with the Th l cytokine IL-2 enhanced their anti-HlV activity, whereas exposure of these cells to the Th2 cytokines IL-4 or IL-lO reduced their ability to suppress HIY replication and to produce IL_29•
In a number of clinical trials, the administration of antiretroviral agents plus IL-2 produced either transient (less than I year)S5.90 or sustained (1-3 year)so. 83 in- creases inCD4+ T cellnumbers in asymptomatic HIY- -infected individuals . Among the functional responses reported to be increasedduring and, in some cases, after cessation of IL-2 treatment werenatural kille(>2·S3. 85. 9U and lymphokine-activated killer activity'" ss, delayed typehypersensitivity", major histocompatibility com- plex (MHC)-restricte d HIY-specific cytotoxicity'" 90
and in vitro proliferative responses to IL-2 and mi- togcn".Increases innumbers of eosinophils, CD8+and CDI6+cells have also been reported'S62.Transientin- creases in the expressionof activation markers (CD38 andHLA-DR) were reported for bothCD4+ and CD8+
cells, and an increase of CD25 (a chain of the IL-2 receptor) expression was sustained for monthsafterdis- continuationof IL-2 therapy". Sustained increases in viral burden were observed only in HIv-infecred indi- viduals with <200 CD4+ T cclls/ml'", also reviewed in49• Although IL-2 treatment seems to hold some benefit for HlY-infected patients, one must consider that IL-2ind uces clonal proliferation of lymphocytcs in the periphery and not production ofnew lymphocytcs in the bone marrow and thymus. In thissense, it risks inducing the replicat ion of infected cells without promising to patch up the holes in the specific im- mune system by increasing the diversity of healthy lymphocytes.
liS
Effects ofG-CSF on Lymphocyte Counts and Functions
We conducted a study" where 24 healthy volun- teers wererandomizedto receive G-CSF(filgrastim75, 150, 300 ug/d) or a placebo for 12 days. Daily blood sampleswere assayed forJL-2 production andlympho- eyre proliferation in response to anti-CD3 antibodies (OKT-3),PHA, or staphylococcalenterotoxin B(SEB).
Volunteers receiving filgrastim showed a transient dose-dependent lymphocytosis with a doubling of CD4+cells by day 8. Twenty four hours after the first injection of filgrastim,ex vivoIL-2 release by OKT-3- -stimulated PBMC was increased by 110% inthe group receiving 300 ug filgrastim. Lyrnphocyte proliferation was augmented in response to PHA or OKT-3, though not to SEB, by 50% 24 h after the first injection of filgrastim. At later time points, however,proliferation was suppressed in comparison to placebo values.
A suppressionof lyrnphocyre proliferation was also de- scribed when serum from healthy volunteers receiving G-CSF was incubated with allogeneic lymphocytes stimulatedwith PHA~oand when PBMC from G-CSF- -treared breast cancer patients were stimulated withthe mitogens PHA, anti-CD3and concanavalin A76. In the latter study, a tendency towards a decreased prolif- erative response was already noted 24 h after the first injection of G-CSr which was more pronounced (p ~
0.05) after 4 days of treatment. No changes in IL·2 productionwere observedin eitherof these two studies.
The observation that both the numbers and the pro- liferative response of lymphocytes were initially augmented butthen decreased under prolonged G-CSF therapy suggested that intermittent administration of G-CSF might be more effective in boosting this activity to stable, higher levels. Therefore, wetreated 4 groups of 6 healthy volunteers each with different regimens of filgrastim over the time period of 8 weeks. One group received saline only, onc group received300)lg filgras- tim once per week,another group was given fllgrustim 3 times per week on alternate days, and the last group received filgrastim for 5 consecutive days in week I and week 5, otherwise saline. An increase in lympho- eyre counts could only be attained by daily injectionof filgrastim. not by intermittent treatment" . We found thatthedoubling of lymphocyte counts in thegroup of healthy volunteers that received fi!grastim for 5 conse- cutive days resulted from proportionateincreasesinTh, cytotoxicT, natural killer (NK)and B cells, suggesting that G-CSF promotes theproduction of new cells in the bone marrow, not lymphocyte proliferation in the pe- riphery. The expression of activation or proliferation
116
markers onT cells was unch anged or slightly decreased and the proliferative response and IL-2 production of isolated mononuclear cells ex .,jvo were unchanged.
thus supporting this hypothesis. Lym phoc yte cou nts had returned to baseline by the next measureme nt
1 week later" . Further srudies are required for the adaptation of treatmentregimens and doses originally established for neutropenic cancer patients after che- motherapyto the requirements of non-neutrope nic HIV patients. Perhaps cyclic treatme nt schedules, where G-CSFisadmin istered fore.g. 5 consecu tive daysfoll- owed by e.g. 2 days brea k, or treatment with the new longer- lastingformof G-CSFcurrentlybeing tested in clinical trials, may offeran effecti ve strateg y for maxi- mizing the increase in Iymphocyte counts by G-C SF treatment,
In a study conducted by ourgroup. blood samples from 31 HlV-infected patients as well as norma l vol- unteers and 8 subjects at high risk for HIV infection due tointravenous drug use were collected" andstimu- lated with SEBin thepresenceor absenceof G-CSFin vitro. Whole blood from patients with advanced HIV infection showed reduced IL-2 release inthe presence of SEB. whichwas part ially restored in the presenceof G_CSF31.
The phenotype and functionalcapacityof progenitor cells mobilized by G-C SF were compared with those of unprimed progenitor cells collected from the same patients prior to treat ment. Ten HIV patient s received 300 ug filgrastim per day for 5 consecutive days.The absolute numbe rs of immature and T cell prog enitors did notincrease and the meannumber of Iymphoc ytes generated per C034+cellina thym ic orga nculture de- creased. Also, the number ofCD4+ cells generated per CD34+cell was significantly reducedafter G-C SFtreat -
mem" .
Althoug h studies of G-CSF effects in patientswith HIV ha ve pre viou sly focused on cha nges in neutroph il levels. an approximate dou bling of total Iymphoc yte count. C04+ and C08+ cell num be rs. augmented lyrn- phocy teproliferative response. and inc reased bone mar- row cellularity have beenreported following the daily use of escalating concentrations of G-CSF alone or in combinatio n with erythropoietin in neutropenic and anemic Hl v -infected patients. Although opportunistic infections occurred in 14of the 22patients. these were treated successfully with myelosupp ressive anti micro- bial agents without develop ment of neutrope nia'". Sev- eral other studies ha ve also demonstra ted increases in lymphocytc, NKcell. C04+ and C08+ counts in neu- tropenic HIV pati ents under daily43. 1(0 or intermittent (thrice weekly5 ug/k g subcutancousj " G-CSFther apy.
Howe ver.inanotherstudy withtwelve neutrope nic pa- tients withAIDS orARC who received dailysubc uta- neous injections ofG-CSFina weeklyincreasing dose (0.4-10 Jlglk g) with 4 weeks of subseque nt mainren- ance thera py. neutrophil and monocyte counts in- creased. but no effects on Iymphoc yte counts were ob- served".
Ina recent randomized, double blind. placebo- con- trolled study . filgrastim was given to 30 Hlv -infected patients whohadbeen onHAARTfor atleast6 months but still had a CD4+ Iymphocyte count smalle r than 350/Jll. The patients rece ived eithe r300 ug filgrastim thrice weeklyfor 12 weeks or a placebo. A significan t increase in absolute numbersof circulatingC034+cells as well as their co lony-forming unit capacity was de- tec ted in the treatme nt grou p. G-CSF treatme nt en- hanced the total IyrnpbocytccountandincreasedC03+.
C04+and C08+Tcell counts as wellas natural killcell numbers. The increases in C04+ and C08+ cells re- sultedfrom increases in C045RO+mem oryT cells and cellsexpressingtheactivationmarke rC038.There was no correlat ion between the numbers of progenitors and the CD4count. There were also no signi fica ntcha nges in the exp ressionof activa tio n markers C025 or C069.
Lyrnphoc yte proliferat ive responses to PHA and Call- dida antige n decreased. whereas NK cell activity and plasma HIV RNA did not change. Also. there was no changeinthe meantelomere len gth in PBMC . All cell countsand immunological parameters had returnedtoba- selinevalues 12 weeks aftercessationoftreatment':3.7J.
In a follow-u p study, treatment-naive Hlv -infected patients were randomizedtorecei ve either a placeboor filgrastim (300 pg, 3 times a wee k) for 12 weeks and HA ARTsimultaneously.The study wasterminatedpre- maturely because one patient de veloped severe cnce- phalopathia .Until then,6patientshad beentreated with G· CSFand 5 with place bo.CD4+memory and C08+
naive and memory T cells inc reased in response to HAART and there was a trend towards more pro- nounced inc reases in the G-C SF group. NK cells in- creased significantly more in the G-CSF group. How- ever. plasma HIV RNA decrease was less pronounced in the group that recei ved G-CSF and rebounded in 2 patients ofthe G-CS Fgroup despitecompliance with HAART drugs. although it cannot be determined whether thiswas dueto the G·CSF treatment or resist- ance devetop menr'. Further investigatio ns ofthese pa- tients also showed that. although the telo mere leng ths of PBMC from patient s on HAART alone increased.
thiswasnot observedinthe groupthat receiv edG-CSF inadditionto HAAR r .
Theanalysisof thetypesandcapa cityof progenitors
recruited by G-CSF and of the telomere lengths in PBMC have indicated that G·CSF may induce produc- tionof new Iymphocytes,red istribution oflymphocytes from the Iymphoreticular system. and proliferation of peripheral lymphocytes. However. the mechan isms by whichthese diverse actions might beachieved are un- known.
Sum mary and Ou tloo k
The advent of HAART has revolutionized HIV/AIDStherapyandconcomitantly the prospects of HlV-infected patients. Still, our knowl ed ge of the changes in theimmu ne statusunder HAART islimited.
The mainfactorsthatare monitored arcHlVRNA plas- ma levels and CD4+ cellcounts, some times also some lymphocyt efunctions such asthereactiontorecall anti- gens. The innate immune system is the first line of defen se against infections. However. weare not aware of studies whichadditiona lly monitored the change of stat us of the innate immune system under HAART, i.e. neutroph il functi ons and cytokine release. Such data are necessary for a clear definitionof the require- ments to be fu lfilled by adjuvent thera pies. Patients who do not respondto HAARTor whodevelop resist- anceto the drugs shou ldbe evaluatedsepa rately in their requirements forimmunesupport.
Building on the information available, G-CSF seems a good candidate to aid immune reconstitution . Benefit s to patie nts might include :
• prev ent ion of neutrope nia due to HIV infection or myelosuppressi vedrugs; substitutionofimpaired G- -CS F producti on,
• improved defense against opportunistic bacterial and fungal infections by induction of neutrophilia, improvement of neutrophil functions, balancingand strengthe ning ofthenon-specifi cimmune system,
• increases inlymphoc yrecounts,incl uding reco very of CD4+cellcounts. and protect ion against oppor- tunistic diseasesbyrec onstitu tion of thespec ific im- mune system.
• restorationof IL-2 formation and Iym phoc yte pro- liferation in response to antigens and mitogens.
• attenuationof virus replicationby reduction of pro- -inflammatory cyto kine release.
Toinvestigate the potentialof G·CS Fas an adjuvant therapy for immune reconstitutio n in HIV patien ts, tai- loringofthe dosage, applicationregimen and timing in relation to the onset ofHAARTdrugs is stilloutstand- ing. The longer-lasting form of G-CS F mayoffernew possibilities inthisrespect. So far, treatment regimens
117 of G·CSF were designed 10 overcome neut ropeni a,but despite this focus, promising effects as to a broader imm une reco nstitution were observed. There findings call for clinical studies exploiting the imm unom od ula- rory propertiesofG-CSF on variousleukocytesubsets.
In suchtrials a comprehens ive immune status analysis should beperfo rmed to evaluate the effects of G-CS F on both the acquired and the innate immune syste m.
e.g.response of monocytesand neutrophils to bacterial stimu li. Only with all this information can a clear evaluatio nof G-CSF administra tion inthesesettings be made.G-CSFtreatment inHIVpatien ts might proveto have impactfarbeyond fight ing neutropeni a.
References
I. ALADDlN11.,UI.LUMH.,OAMNIELSENS.,ESPERSENC,MATH· IESEN L.. KAl7.ENSTEIN T. L., GEI!.Sru~T I.,$KINHOJ P. and PEl>ERSENB.K.(2000); Granulocyrecolony-stimulating factor increases CD4+T cell counts of human immunodeficiency virus-infected patients receivingstable, highlyactiveantirerro- viral therapy: results from a randomiz.ed. placebo-controlled mal. J.lnfccl.Ois.,181.1148-1152.
2. ALADOIN H..Uu.uMH.•KATZENSTEINT..GERSlUHJ.•SKIN·
HOJ P. and KLARLUND PEOERSEN B. (2000): Immunological andvirological changes inamirct rovua lnaive humanimmuno- deficiency virus infected patients random ized to G-CSF or placebo simultaneously with initiation of HAART. Scand, J. Immunol.,51.520-525.
3. ALADOIN H.•Uu.uaH.•5cHJERUNGP..SKOV JE.""SENM.•OAM
Nlf.lSE.'1S.,MAllUESENL..GERSTOFTJ..SKINHOJP.and KLAR·
LUNO PEOERSEN B. (2000): Effects of G·CSF on telomere lengths in PBMCs from human immunodeficiency virus-in- fected patients: results from a randcmized,placebo-controlled trial.Scand .J.lmmunol.,52,212-216.
4. ALADDlNH.,VON EsSE:-IM., SCHJERLlNGP.,KATZENSTEl NT., GERSTOFTJ.,SKINHOJ P.,KLA RI.UND PEDERSENB.andULLUM I!. (2001) :Ticell mean telomere lengthchanges in treatment naive Hlv-infecred patientsrand omiz cd to G-CS For placebo simultaneouslywithinitiation ofHAA RT.Scand.J.lmrnunol., 5.\.30 1-305.
5. AtrrRAN B.,CARCELAIl\'TG.,1.JT.S.,GOROCHO VG., BLANC C.,RENAUDM.. DURAU M.,MAlll EZD.. CALVEZV.•UlBQ.
WtTCHJ.,KAl1Jt MAC. andDEBREP.(1999):Restorationof the immunesystem withanti-reuoviraltherapy.Immunol.Len.,66, 207-21J.
6. BALDWING.C..FuLLER N.D.,ROBERTS R.L..HoD.O. and GoLoEO. W. (1989); Granulocytc- and granuloc yte-mac rc- phage colony-stimulatingfactors enhance neutrophilcytotox- icilytowardHlv-intectedcells. Blood ,74,1 673-1677.
7. BALDWIN G.C.,GASSONJ.C..QUANS.G.,FlLISCI-IMAN:'IlJ., WEISBART R.,OETIE D..MlTSUYASUR.T.andGoLDeD.W.
(1988); Granulocytc-macrophage colony-sumu laungfactoren- hancesneutrophilfunctioninacquiredimmunodeficiency syn- drome paticnts.Proc. Nail.Acad.Sci. USA,8S. 2763-2766.
8. BALTERM.(1995):AIDS research. Cymkincsmove fromthe marginsinto the spotlight. Science,2611.205-206.
9. BARKERE., MACKEWICZ C.E. and LEVYJ. A.(1995): Effects
118
of THI and TH2 cytokines on COS· cell response against human immunodeficiencyvirus: implications forlon g -te rm sur- vival. Proc. Natl. Acad. Sci. USA.92, IJ135-11139.
10.BERN STEL" Z.P., PoRTERM.M.. GoULOM.. I.JPMAN B.•Bt.u.
MAN E.M., STEWART C.C•• HewrrrR.G., FYFE G..PorESZ B.andC"UGIURI M. A. (1995): Prolonged administration of low-dose interlcukin -2in human immunodeficiency virus-asso- ciated malignan cy results inselective expansionofinnate im- mune effectorswithout significantclinicalroxicuy. Blood ,86.
3287-3294.
11. B1NOEl..S P. 1., KROL A., RODS M., VEE."S'TRAJ.•VANoex
HO£KJ.A., MIEDE.\ IA F. and counsuoR.A. (1995): The pre- dictive value of T cell functionin vitro and pre.AIDS zido- vudine use for survival after AIDS diagnosis in a cohort of homosexual menin Amsterdam.J.Infect.Oi5.. 172. 97-104.
12. B1SWASP.,Pou G.,ORE~STEL'IJ.M.and FAua A. S. (1994):
Cytokine-med imed induction of human immunode ficiency virus (Hlv)expressionand celldeath in chronically infected UJ ce lls: do tumor necrosis factor alpha and gamma interferon selectivelykill Hlv-infectcd cells'?J. Yim!.,6H.2598-2~ .
13. BONEBERG E. M..HARENG L,GANTNER F.,WEN DElA. and HARl1JSG T. (2000):Human mo nocyr es express func tion alre- ceprors for granulocytc colony-stimulating factor that mediate suppression of monoldnesand interferon-gam ma. Blood. 95, 270--276.
14. BORNEMAS.'1M. A .. VERHOEFJ.and PETERsos P.K. (1997):
Mac roph agcs . cytokincs. and HIY. J. Lab. Cnn. Med., 129, 10-16.
15. CLERICI M., BALOTI A C.,MERONI L..FERRARIOE., RIVA C ..
TRABATIONID..RIDOLFO A.,VII.LAM.,SHEARER G.M., Mo - RONtM.andGALUM.(1996): Type Icyto kine prod uc tionand lo w prevalence of viralisolation correla tewith long-term non - progression in m y infection. AIDS Res.Hum . Re rrc viruses, 12. 1053-1061.
16 . CURiaM. and SHEARBl. G.M. (1994): The ThI-Th 2 hypo - thesis ofHIV infection: new insights. Immunol . Today, 15.
575-581.
17. COCCHI F., DEVICOA. L, GARz INo -Demo A., ARYA S. K.•
GALLaR.C.and Lusso P. (199 5):Identification of RANTES, MIP-I alpha, and MIP-I beta as the major Hlv-supprc ssive fac to rs producedby CD8+Tcells . Science,270, 1811-1815.
18. COFFEYM.J.,Pll ARES. M.,GEORGES., PETERS-GOLDEN M.
andKAzA.'lJlA~ P. H.(1998): Granulocyte colony-st imulating factor administration to HlV-infected subjects augments re- duced leukotriene synthesis and anticryprococcal activity in neutrophils. J. Clin. In vest.•102. 663-670.
19. COYl.ET. E. (1997): Hema tolo gic complicat ionsofhu ma n im- munode ficiencyvirus in fectionand the acquiredimmunodefi- ciencysyndrome. Med. Clin .Nort h Am., 81, 449-470.
20. EMILlED.,ROilR.,]ARRO USSEB.•MARFAING-KOKAA.,MEII·
111101'1 D.. DEVEIIGNE0 .. CREVONM.C., MAILW1' M. C.and GALANAUDP. (199 4) :Cyto kinc sin Hlv infection .lnt. J. Im- muno pharmacol ..16.39 1-396.
21. EssERR., GU8'O KE W., A.'lDREESEN R., UNGERR. E..KREt.rrz M., RUBSAMEN-WAIGMASS H. and vox BRlESE."IH. (1998):In- dividual cell analysis of the cyrckinerepe rtoire in hu man immu- nodeficiency virus- I-in fected monocyteslmacrophages by a combination ofimm unoc ytoc he mis tryand in siluhybndiza- tion. Blood.91, 4752-4760.
22. EsSERR.•GUENKEW., VON BlttESENH., RUBSAMEN-WAIG·
MAN NH.andANDItEESEN R. (1996):Differential regulatio nof
proinflammatory and hematopoieticcytckinesinhuman macro- phages after
m recuoo
with hum an immunodeficiency virus.Blood.88_3474-3481.
23. EYlJ\R E. H.•BAEZL YAZQUEZA..COLO~ ·M ARTlSEZS.L and YA!'.IAMURAY.(199 5): Suppre ssed prolife rative respon se andinterleukin-Zproduction inbispanicH1Y+andAIDS'r-eel!
subsets. Cell.Mo l.BioI. (Noisy-le-grand), 41, 525- 33 . 24. FAUCIA.S. (I9%):Ho st factors andthe patbogencsis of HIV-
-indu ceddisease. Nature ,384,529-534.
25. FRUMKINL R. (1997): Ro le of granulocytc colony-stimulating factor andgranulocyte-macrophagecolony-stimulating factorin lhe treatment of patients with HIVinfec tion. CUlT. Opin. Hema- tol., -t 200-206.
26. GABRJLOVICH D ..lv As o v AL., SEREBIIOVSKAYA L, SHEPEI£VA G. and PoKROVSKY Y.(1994): Clinical significanceof neutro-- philfunctio nalactiv ity in HIVinfec tion .Scand. J.Infe ct.Dis..
26,4 1-47.
27. GAII.A VELLlP. LandBERllP. (1993) :Efficacyofreco mbina nt granu loc yte colo ny- s timu lat ing fac tor in the long- te rm treat- ment of AIDS-related ne utropen ia. AIDS,7, 589-590.
28. GEORGES.• COI-HY M.•CrsttS., COLLi:'lSJ" BItOWNM. and
KWAZA~Jl A~P. (1998):Ne utro p hilsfrom AIDS patie nts treated with granulocyte colony-stimulating factor demonstrate en- hanc ed kill ing ofMyc obocu rium avium. J. In fect. Dis.. 178. 1530-1533.
29. GouGEON M. L. LECOEURH. and SASAKI Y.(19 99 ): Apo p- tos is andthe COO5syste m in HIYdisease:impact of highl y active ami-rcrrcviral therap y (HA ART). Immu nol. Le u., 66. 97-103.
30 . GRAYBILLJ.R.,BOCANEGRA R.and LUTHER M.(1995):Anti- fungal co mbin ationtherapy withgranu locy rccotony-stirnulat- ing facto rand flucon azo le in experime ntaldisseminated candi- diasis. Eu r.J.Ch n. ~iicrobiol . Infect .Dis., 1-1. 700-703.
31. GRl.ITI.RS R. A.• TElI.P$TRA F.G., DE JO....G R.. VANNOE.sa C.J.~1.• VA.... LIF.R R. A. W. and MIEDE\IA F. (1990): Selec- tive loss ofT cell functions in differentstages ofmvinfection.
Earlyloss of anti-CD3 -indueed Tcellprol ife ratio n followedby decreased ant i-C D3 -ind uced cytotoxic T lymphocyte gener - ation inAIDS·relatcdcomplexand AIDS. Eur.J. 1mmunol.,20. 1039-1044.
32. HAGIWARA E.,SACKS T., LEITM AN-KuNMAN S,F. and KLIN· MAND. M. (1996):Effect ofHIVinfe ctiononthcfrequencyof cytokine-secrctingcells inhuma nperipher al blood. AIDS Res.
Hum.Re rro viru ses .12. 127-133.
33. HAS X., BECKER K., DEGEN H. J.•JABl.IDlOWSKI H. and STROHMEYERG. (1996): Synergisticstimulatory effects of tu- mou r nec rosi s factoralphaand interferongamma on replication ofhuma nimm uood efici ency virus type I and on apcptosisof HIV-I-in fected hos t cell s. Eur.J.Clin.Inve st.,26, 286-292.
34. HARTUNGT. (1999 ): Im muno rnod ulation by colony- stim ula t- ing factors .Re v.Physiol. Biochcm.Pharmacol., 136. 1-164.
35. HARTUNG T.. DOECKEW.D.,flUNl)SCHUlI 0.. roore M. A .•
GANTNER F., lIERMANN
c..
LENZA., MILW EE 5., RICH B..StMO.'l B.,VOLKH.D., vox AULOCK S. andWE~DElA.(1999):
Effect of filgrastim trea tme nt on inflammatory cytokines and lym p hoc yre functions. Clin. Pharmacol.The r.,66. 415-424.
36. HARTUI';G T., OoECKEW. D., GA.'l'ThERF..KRIEGER G.. SAtJERA., STEVE.-';S p.. YOlKH.D. andwe-on,A.(1995): Effect of gra- nuloc yte colony-stimulating factor treatmentonex \'ivoblood cytckin eresponsein human volunteers . Blood,8S,248 2-2489.
37. HARl1JNGT..PITRAK D.L.,FOOTEM..SIlATZIiNE.M" Vu-
RAL S. C. and WENOEL A. {I998): Filgrasurn restores inter- leukin-2 production in blood from patients with advanced human immunodeficiency virus infection.J. Infect. D!s.. 17H.
686-692.
38. HARTUNGT.,VONAULOCK S. and WENDELA.(1998): Role of grenulocyrc colony-stimulating factor in infection and inflaru- mation. Med. Microbiol.fmmunol.,I1l7,61-69.
39. HAl(l\JNG T..VO:-l AULOCK S.and WE."IDEL A. (2000): Growth factors G-CSF and GM-CSF: clinical options. In BAUEA. E..
FAtSTE.and FRY D.E.:Multiple organ failure- pathophysio- logy, prevention andtherapy.Springer.NewYork-Bcrlin-Hci- delberg, 621-629.
40. HEBERT J. C., O'REtLLYM.and GAMELLt R.l.(1990): Protec- tive effectof rccombinnnt human granulocyte colony-stimulat- ing factor against pneumococcal infections in splenectomized mice. Arch. Surg.,125. 1075--1078.
41. HENGGE U. R..B!(OCKMEYERN. H.andnoosM.(1992): Ora- nulocytecolony-stimulatingfactortreatment in AIDSpatients.
Cfin.lnvest.. 70.922-926.
42. HERMANS P.(1995): Hacrnuropoietic growth factorsassupper- rivctherapy in Hlv-infecred patients. AIDS.9(suppl. 2). 59- 514.
43. HERMANS P.. ROZENBAUM W., Jou A.. CASTELLtF.,BORLEfFS J.,GRAY S.. WARDN..GORIA..DE BONA A..FERRE
c..
loNeI'.M.,LANG J. M., AMMASSARt A. and CLUMECKN.(1996): Fil- grasum to treat neutropenia and support myclosupprcssivc medication dosing in IIIV infection. G-CSF 92105 Study Group AIDS.10, 1627-1633.
44. IEKl R.. FURUTA T.,ASA:-IO S.. MORI S.,KUDOH S., KIMURA H. and TAKAKU F.(1989): Effect ofrccombtnnnthuman granu- locyrc colony-stimulating factor onPneumocystiscarinilinfec- tion in nude mice. Jpn. 1.Exp.Med .,59.51-58.
45. JACOHSON M. A.. LIUR. C.. DAVIES D. and COHEN P. T.
(1997): Human immunodeficiency virus disease-related ncu- rropcnia and the risk of hospitalization for bacterial Infection.
Arch.Intern. Med..157,1825~1831.
46. KEISER P.,HICGSE.and SMITHJ. (1996): Neutropenia is asso- ciatedwithbacrercmiain patients infected with the human im- munodeficiency virus.Am. J.Med. Sci.,312. 118-122.
47. KEISER P.,RAllEMACHER S.. SMITH J. and SKI ESTD. (1998):
G-CSFassociation with prolonged survivalinHIV-infcctcd pa- tients with disseminatedMycobacterium aviumcomplex infec- tion.lm.J.STDAIDS,9, 394-399.
48, KElSER P.. RADEMAClIER S., SMITH J. W., SKIEST D. and VADDE V. (1998): Granulncyte colony-stimulating factor use is associated with decreased bacteremia and increased survival in neutropenic Hfv-infccted patients. Am. 1. Med.,IO·t4&-55. 49, KINTER A. and FAUCI A. S. (1996): lmcrlcukin-Z and human
immunodeficiency virus infection: pathogenic mechanisms and potential for immunologic enhancement. Immunol. Rcs., 15,
1-15.
50. KOVACS 1.A.,BASELER M.. OEWAR R 1., YOGEL S., DAVEYR.T.
JR., FALt.OON J., POLlS M. A.. W."'LKER R. E.. STEVE:-JS R., S,\LZMAN N. P., METCALfJ. A.. MASUR H. and LANEH. C.
(1995): Increases in CD4 T lymphocytes with intermittent courses of interleukin-2 in patients with human immunodefi- ciency virus infection. A preliminary study. N. Engl. J. Med., 332. 567-575.
51. KOVACSJ. A.,VOGEL S.. METeAL!' J. A., BASELER M.. STE-
\lENSR.,ADELSBERGER 1., LEMPICKI R.. HEI'GELR.L.,seesnI., LAMBERTL.. DEWAR R.L., DAVEY R. T. JR.. WALKER R. E..
119
FALLOO:-JJ..POllS M. A., MASURH. and LA:-IEH.C.(2001):
Interleukin-2 induced immune effects in human immunodefi- ciency virus-infectedpatients receiving intermittent interleukin-2 immunotherapy.Eur. J. lmmunol., 31.1351~1360.
52. KtJRrrI.KF.SD.R., PARENTI D., WARD D. J., RACHusA" WO~G
R. J..MALLDNK.P.,R1CIl W.J. andJACOBSONM.A. (1998):
Pilgrastim prevents severe neutropenia and reduce, infective morbidityin patients with advanced IIIV infectio n:results of a randomized. muhicenter. controlled trial. G-CSP 930101 Study Group AIOS. 12. 65-74.
53. LATHEY J. L.. KANAl'iGAT S.. ROUSE B. T., AGOSTI J.M. and SPECTOR S. A. (1994): Dysregulation of cytokine expressionin monocytcsfrom Htv -positivcindividuals. 1. Leukoc. BioI.,56. 347-352.
54. LAZARDT..PERRONNEC.,COllENY.,GROSSET J .. VILDE J.L.
and PoclDALOJ.J.(1993): Efficacy
or
granulocyrccolony-stimu- lating factor andRU-40555 in combination with clarithromycin againstMycobaeteriUIIl al'illlll complex infection inC57BU 6 mice. Anumicrub. Agents. Chemother..37,692---695.55. LAZlARIN A., UOE\l.T1 FOPPA C.• GALLt M.. MANTOVANI A..
POll G.• FRANZElTl P. and NOVATI R. (1986): Impairment of polymorphonuclear leucocyte function in patients with acquired immunodeficiency syndrome and with lymphadenopathy syn- dromc. Clin. Exp. lmmunol.,65, 105-111.
56. LEWIS D. E.•TA:-IG D. S.. ADU-OPPO:-IG A., SCHOBER W. and RODGERS J. R. (1994): Anergy and apoptosisin CD8+Tcells fromHtv -infected persons.J.lmmullol.. 153, 4 12-4 20, 57. LlSTER P.D.• GI'NT\l.YM.J. and PREHEIM L.C. (1993): Granu-
locytc colony-stimulating factor protects control rats but not ethanol-fedruts from fatal pneumococcal pneumonia. J.Infect.
Dis.,168, 922-926.
58. MAGGI E.and ROMAGNANI S. (1995): Role ofT cell derived cytokines in HIV infection. J.BioI. Regul. Homeost.Agents.
9. 37-41.
59. MANCA F.. FENOGLlO D.,LrPJI{AG.•GENNARIF..DELPUPL.
and PALU G. (1999): Rational reconstitution of the immune repertoire in AIDSwith autologous, antigen-specific,illvitro- -expandedCD4 lymphncytcx.Immunol. Len.,66. 117-120.
60, MAUSS S.. STEINMETZ H. T., JABLONOWSKI H. and HAUSS- INGERD.(19%): Lack ofinduction of granulocyte cclony-stimu- Iating factor in human immunodeficiency virus-seropositive in- dividuals. Blood,88. 1897- 1898.
61. MAZZONE A.. RICEVLrt'l G.• FlORA vANTI A.. PSAOlTtD.,vrrro A. R., GAMBA G.. DEGIULlO R., SACCHI S.and NOTARIO A.
(1989): Defective phagocyte functions in HIV patients. Med.
Bio!. Environ.•17, 245-251.
62. McMAIlO:-l D. K..AR~ISTRONG1. A.,HUANG X.L.,RI:-lALDO C.R.,JR., GUPTA P.,WIIITESIDE T. L..PAZtN G.L,TRIPOLIC.
and Ho M. (1994): A phase Istudy of subcutaneous rccombi- nant interlcukin-2in patients withadvanced HIV disease while on zidovudinc. AIDS.8.59---66.
63. MEYER C.N.,SKINllOJ P. and PRAG J. (1994): Bactcremia in Hlv-positivc and AIDS patients: incidence, species distribu- tion. risk-factors. outcome,and influence of long-termprophy- lactic antibiotic treatment. Scene. 1. Infect. Dis.,26.635-642.
64. MILES S. (1995):The use of hematopoietic growth factors in treating HIV infection. Curr. Opin. Hematol..2. 227-233.
65. MILES S. A.(1992): Hernatopoietic growth factors as adjuncts to antirctroviral therapy. AIDS Res. Hum. Rcrroviruscs, 8.
1073-1080.
66. MILES S. A., MITSUYASU R. T .. MORENO J., BALDWIN G.,
