and chronic purulent rhinosinusitis. A double-blind cross-over trial

Clin. exp. Immunol. (1990) 80, 304-313

Beneficial effects of the thymic hormone preparationthymoxtimulin in patients with defects in cell-mediated immunityand chronic purulent rhinosinusitis. A double-blind cross-over trial

on improvements in monocyte polarization and clinical effects

M. TAS, J. A. LEEZENBERG* & H. A. DREXHAGE Laboratory for Clinical Immunology, Pathological Institute, and

*Department of Otorhinolaryngology, Head and Neck Surgery, Free University Hospital, Amsterdam, The Netherlands

(Acceptedfor publication 8 January 1990)

SUMMARY

Twenty patients with chronic purulent rhinosinusitis were treated with TP- 1 (Serono; 1 mg/kg bodyweight), in a double-blind cross-over trial. TP-1 was administered by daily i.m. injections for the first14 days followed by two injections/week for 6 further weeks. The patients were immunologicallyspecial in that they had defects in their cell-mediated immune system. Fourteen showed a decreasedchemotactic responsiveness of their peripheral blood monocytes as measured in the polarizationassay. This defective function can probably be ascribed to the presence in serum of low molecularweight factors (LMWFs; < 25 kD). As reported earlier, this factor shows a structural homology to

the envelope protein of murine and feline leukaemia virus (P1SE). Thirteen patients showed a

defective delayed-type hypersensitivity (DTH) skin test reactivity towards candidin and/orstreptokinase-streptodornase (Sk/Sd) antigen, 14 had a defective MIF production from theirperipheral blood lymphocytes towards candidin, Sk/Sd and/or Haemophilus influenzae antigen.Eighteen patients completed the TP-1 trial and showed clinical improvements: 12 out of 15 were

feeling better during TP- 1 therapy and the nasal mucosa showed on inspection absent mucopurulentsecretion in 13 patients. Positive bacterial culture rates for the nose decreased from 14 out of 16 to fiveout of 15. Placebo treatment had no significant effects. The clinical improvements were accompaniedby a better performance of the cell-mediated immune system; the most significant effects were

recorded in the monocyte polarization assay. The suppressive P1 SE-like LMWFs in serum clearlydecreased during TP-1 treatment. In vitro TP-1 neutralized the immunosuppressive effect of theLMWFs. The restoring effects of TP-1 on monocyte polarization and its neutralizing activity ofP1SE-like LMWFs could explain the beneficial effects of thymic hormone treatment reported inadults with clinical signs of immunodeficiency in the presence of a full T cell repertoire.

Keywords rhinosinusitis immunosuppression monocyte chemotoxis TP-I

INTRODUCTION

A disorder frequently encountered in ENT and paediatricpractice is a relapsing of chronic purulent rhinosinusitis. Thepathogenic factors and treatment modalities of this disorder areat present not entirely clear. Surgical interventions to improvethe drainage of the sinuses and frequent courses with antibioticsare normally given, but it is our experience that in about 15% ofthe patients complaints of general malaise, nasal blockage,purulent nasal discharge and headache remain.

Correspondence: M. Tas, Department of Immunology, ErasmusUniversity, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands.

As reported earlier in detail, 60% of patients with rhinosi-nusitis refractory to the current therapies show a defective cell-mediated immunity (CMI) towards the commensal micro-organisms (Drexhage et al., 1983).

At first we detected T cell impairment in the patients usingdelayed-type hypersensitivity (DTH) skin testing and a macro-

phage migration inhibition factor (MIF) assay on peripherallymphocytes; both tests showed faulty reactions towards a panelof microbial antigens (Drexhage et al., 1983, 1985; van derPlassche-Boers et al., 1985, 1986). Other T cell functions, viz.,the blastogenic responsiveness towards these antigens and thenumber of peripheral T cells, were normal (van der Plassche-Boers et al., 1986).

304

Thymostimulin and chronic purulent rhinosinusitis

Later we found that an impaired polarization of monocytesaccompanied the T cell defects; polarization of monocytestowards chemo-attractants is an early event that precedes theirchemotactic responsiveness (Cianciolo & Snyderman, 1981).The defective monocyte polarization could be explained by afactor detectable in the patient serum: addition of serumfractions of < 25 kD to healthy donor monocytes resulted in aninhibition of polarization. The suppressive factor appeared tobe P1SE-related, since monoclonal antibodies directed againstP1 SE neutralized the inhibitory effect (van der Plassche-Boers etal., 1988). P1SE is the transmembranous part of the envelopeprotein GP85 of murine and feline leukaemia virus and isresponsible for the immunosuppressive effects of these retrovir-uses in their hosts.

P1SE-related low molecular weight factors capable ofinhibiting the function ofmonocytes have been reported earlierin detail in other patient groups characterized by defects in CMI.Increased levels have been documented in patients sufferingfrom several types of tumours (Pike & Synderman, 1976;Maderazo, Anton& Ward, 1978; Snyderman, Pike & Cianciolo,1980; Cianciolo et al., 1981; Cianciolo, Phipps & Snyderman,1984; Snyderman & Cianciolo, 1984; Balm et al., 1984; Tan etal., 1986a,b) and have also been detected in sera of HIV-seronegative homosexuals (Tas, Drexhage & Goudsmit, 1988).

The origin of the P15E-related factors in the serum ofpatients is unknown. The speculative possibilities range from anexogenous infection with an unknown retrovirus producingproteins that share structural homology with P1 SE, to anendogenous production of the immunosuppressive factor;P15E-related factors were detectable in squamous cell carcino-mas, in epithelial cells overlaying areas of inflammation and innormal thymic epithelial cells (Tan et al., 1987).

Physiological substances known to have strong effects on thefunctioning of the cell-mediated imune system are the so-calledthymic hormones. In neonates they induce a terminal differen-tiation in T cells, and are able to stimulate T lymphocytes toproduce lymphokines (Aiuti et al., 1975; Falchetti et al., 1977;Skotnicki et al., 1984). In adults, the administration of thymichormones has also been described as being effective and lowlevels of endogenous thymic hormones can be found in severalsecondary immunodeficiency states, as well as in viral infectionsand other states of impaired adult host defence (Lewis et al.,1977; Iwata et al., 1981; Iwata et al., 1983). This report describesthe effects of a double-blind cross-over trial with the thymichormone preparation thymostimulin (TP- 1, Serono) in 20patients characterized by the above-described impairments ofcell-mediated immunity and chronic purulent rhinosinusitis.The in vitro effects of TP-1 on the in vitro activity of P1SE werealso tested.

MATERIALS AND METHODS

PatientsTwenty patients, 17 women and three men, aged 23 to 71 years,median 44 years, with a relapsing ofpurulent rhinosinusitis wereincluded in a double-blind cross-over trial with TP-1. Criteriafor inclusion in the study were (Drexhage et al., 1983, 1985): (1) adefective CMI as indicated by (a) a defective monocytepolarization at previous testing, and/or (b) a defective skin testtowards candidin, Streptokinase-Streptodornase (Sk/Sd) and/or H. influenzae antigen at previous testing, and/or (c) a

defective MIF-production towards candidin, Sk/Sd and/or H.infiuenzae at previous testing; (2) relapsing of chronic purulentrhinosinusitis as indicated by (a) duration of the disease of atleast 18 months, (b) a positive culture for H. influenzae, S.pneumoniae, other Streptococci, or Staphylococci at one ormore occasions, (c) no response to or only temporary relief aftertreatment with several courses of antibiotics, (d) failure ofsurgery to give a permanent cure by improving the drainage ofthe ethmoidal and maxillary sinus, (e) no gross disturbances inmucociliary transport (exlusion of Kartagener and relatedsyndromes, van der Baan et al., 1987).

All patients had normal levels of total serum IgG, IgM andIgA, normal total numbers of peripheral blood leucocytes and anormal differenital lymphocyte count.

A treatment with antibiotics and/or other drugs known toinfluence the immune system was not given during the TP-1trial.

Bovine thymic extract (thymostimulin, TP-J)Thymostimulin is a bovine thymic extract prepared by Serono,according to Bergesi & Falchetti (1977) and Falchetti et al.(1977), using the following procedures. Calf thymus glands wereminced and extracted with ammonium acetate. The extract washeated to 70°C, filtered, and precipitated with ammoniumsulphate. The precipitate was dissolved in water and subjectedto ultrafiltration on an Amicon PM-10 membrane. The filtratewas desalted on Sephadex G-25 and gel-filtered on Sephadex G-50. The fractions which show two characteristic bands onpolyacrylamide gel electrophoresis with Rf 0 22 and 0 24 arecombined and termed thymostimulin (TP-1). The extract waslyophilized and its activity expressed as units of T cell rosetteformation per milligram of protein. It did not contain endo-toxin. In toxicological studies, the extract did not cause anytoxic or other side-effects in doses up to 100 mg/kg whenadministered to mice for 21 days, to rats for 31 days or to cats ordogs for 180 days in doses up to 50 mg/kg (Falchetti et al., 1977).The extract did not alter neuromuscular transmission either invitro or in vivo (Bergesi & Falchetti, 1977). The preparation hasshown beneficial effects in patients with primary cellularimmunodeficiency and combined immunodeficiency (Lin et al.,1985, 1986, 1987, 1988). TP-1 is commercially available in Italy,Spain and Germany.

Double-blind cross-over treatment with thymic factor TP-J andplaceboPatients gave informed consent and the ethical committee of thehospital approved the double-blind trial. The treatment con-sisted of an intramuscular injection of either TP- 1 (1 mg/kgbody weight) or its placebo [solvent (pyrogen-free sodiumchloride solution) + carrier (mannitol); specifically prepared bySerono]. Patients were randomly allocated to two groups; onegroup started with TP-l injections, the other with placeboinjections. Treatment was given for 8 weeks. For the first 2weeks injections were given daily, followed by twice a week for 6weeks. On week 9 no treatment was given to test patients' CMI.Thereafter the schedule was crossed over (TP- 1 to placebo, andplacebo to TP-1) and treatment was given for 8 more weeks.

The week prior to the trial the following tests were carriedout in each patient (for techniques see below): (1) an inspectionof the nose, (2) a bacterial culture from the nose, (3) a BSR, (4)DTH skin tests with bacterial antigens, (5) lymphocyte subset

305

M. Tas, J. A. Leezenberg & H. A. Drexhage

determinations in the peripheral blood, (6) an MIF assay, (7) apolarization assay with peripheral blood monocytes, (8) adetermination of the PiSE-like factor in serum.

Tests 1, 2, 3, 7 and 8 were repeated at week 4, 8, 13 and 17(end of trial). Tests 4, 5 and 6 were only repeated at week 8(before cross-over) and at week 17 (end of trial).

In all patients, parameters relevant to liver function (biliru-bin, ALAT, ASAT) and kidney functions (creatinine, protein-urea) were checked during treatment.

DTH skin testsThe following skin test antigens were used (see also Drexhage etal., 1985): (1) 250 pg/ml of a somatic H. influenzae antigen,prepared as described elsewhere (Drexhage et al., 1983; van derPlassche-Boers et al., 1985); (2) two commercially availablepreparations; viz., 1% Candidal antigen (HAL allergens, Haar-lem, The Netherlands) (= Candidin); and 100 U/ml Sk and 50U/ml Sd (Varidase, Lederle, Wayne, MI).

Delayed responsiveness was tested by intradermal injectionof0 1 ml suspension ofeach antigen preparation in the forearm.The skin reactions were read at 30 min, 6, 24 and 48 h and thediameter of the induration, expressed as the average of twomeasurements at right angles, was recorded.

Enumeration of total peripheral blood lymphocytes and lympho-cyte subsetsThe percentage of lymphocytes and lymphocyte subsets wasdetermined by reacting peripheral blood lymphocytes isolatedby Ficoll-Isopaque density gradient centrifugation (Pharmacia,Uppsala, Sweden) with CD3 + antibodies against T cells (Leu 4;Becton Dickinson, Mountain View, CA), CD2+ antibodiesagainst active T cells (OKT I 1; Orthoclone Ortho, Raritan, NJ),CD4+ helper/inducer T cells (Leu 3a), CD8+ suppressor/cyto-toxic T cells (Leu 2a), and CD24+ B cells (BA-1; Hybritech, SanDiego, CA) as indicated by the manufacturer. Two-hundredcells were counted in a fluorescent microscope; the tests weredone in duplicate. Absolute numbers ofT cells and T cell subsetswere calculated by multiplication from the total peripherallymphocyte counts.

Macrophage migration inhibition factor testMIF production was estimated with an indirect microdropletagarose assay as described in detail by van der Plassche-Boers etal. (1986). In brief, peripheral blood mononuclear cells(2-5 x 106) were cultured with the antigens of H. influenzae,Candidin and Sk/Sd. Supernatants were also prepared using themitogen Concanavalin A (Con A, Sigma, St Louis, MO).Supernatants were collected after 3 days of culture (370C, 5%CO2 in air) and stored at - 20°C until testing for MIF activity.

The agarose microdroplet assay was performed according toThurman et al. (1983) using the human monocytoid U937 asindicator cells (Singh & Khan, 1982). From the cells (2 x 107cells/ml) in 0-2% agarose (Marine Colloids, Rockland, USA) 1Mu droplets were centrally placed in the wells of flat-bottomedmicrotitre plates (Nunc, Denmark) using a Hamilton RepeatingDispenser with a 0 05 ml gas-tight syringe (Hamilton, Reno,AR). The droplets were left to solidify at 4°C for 10-20 min, andcarefully overlaid with 0-1 ml of thawed supernatant diluted 1: 1with fresh medium. Each supernatant was tested five times.After incubation of the covered plates for 21 h at 37°C, and 5%CO2 in air, migration areas (cell migration area minus area ofthe

agarose droplet) were computed using a projection microscopeand a graphic tablet, connected to a computer. MIF productionwas expressed as per cent migration inhibition:

MI= 100-Mean migration area in antigen-stimulated cultures

Mean migration area in medium x 100%

The isolation ofperipheral blood monocytes and the polarizationassayPeripheral blood mononuclear cells (20 x 106) isolated byFicoll-Isopaque density gradient centrifugation were washedtwice in phosphate-buffered saline (PBS), pH 7-4 containing0 5% bovine serum albumin (BSA), and counted in suspensionemploying positive staining with non-specific esterase (Mullinket al., 1979). The percentage of NSE-positive cells varied, at 5-25%. An enrichment for the monocytes in the Ficoll-Paque-isolated fraction was obtained by Percoll gradient centrifuga-tion (Pertoft et al., 1980). After washing, the pellet containingthe monocytes was resuspended in the above-mentionedmedium and carefully underlayed with an equal volume ofPercoll 1.063 (Pharmacia, Uppsala, Sweden). After centrifuga-tion (40 min, 450 g) the cells were collected from the interface,washed twice in medium (10 min, 500 g) and counted: thesuspension now contained 70-95% NSE-positive cells.

The Cianciolo & Snyderman (1981) assay for monocytepolarization was performed with slight modifications (Tan etal., 1986a); 0-2-ml aliquots of the Percoll- or electicator-purifiedcell suspension containing 0-2 x 106 monocytes were added to12-75 mm polypropylene tubes (Falcon Labware Division ofBecton Dickinson Co., Oxnard, CA) containing 0 05 ml ofeither medium or N-formyl-methionyl-leucyl-phenylalanine(FMLP) in medium, to reach a final concentration of 10 nm. Allexperiments were carried out in duplicate. The tubes wereincubated at 37°C in a waterbath for 15 min. The incubation wasstopped by addition of 0-25 ml ice-cold 10% formaldehyde in0 05% PBS (pH 7 2). The cell suspensions were kept at 4°C untilcounting in a haemocytometer using an ordinary light micro-scope (magnification 250 x ). The test was read 'blindly' by twopersons; 200 cells were counted from each tube. A cell was'polarized' if any of the following occurred: (1) elongated ortriangular shape, (2) broadened lamellopodia, (3) membraneruffling.

The percentage of polarized monocytes was calculated asfollows:

% total cells polarized x 100%% NSE-positive cells

Lymphocytes do not exhibit any polarization activity in thisassay (Cianiolo & Snyderman, 1981).

The chemotactic responsiveness of a monocyte population isexpressed as the percentage of polarized monocytes in thepresence ofFMLP minus the percentage ofpolarized monocytesin the absence of FLMP. The assay has proven to be a rapidmethod to test monocyte chemotaxis and outcomes of the assaycorrelate well with outcomes of the conventional Boydenchamber assay to measure chemotaxis to casein (Tan et al.,1986a). Of 77 healthy control individuals, a chemotacticresponsiveness was found of 33 (mean, s.d. 1 1; ranges from 18 to70), there were no differences between female and male indi-

306


Monocyte polarization

( a)

Beforetreatment

P aceboprior to

TP-1 therapy2

TP-1 therapy

L2

Placebo ofterTP-1 therapy

L2

[% polarization]

Presence of LMWFs

[°h inhibition of polarization]

n-5 2]

i

60 50 40 30 20 10

Fig. 1. Monocyte polarization (a) and inhibition of polarization (b). (a) FMLP-induced polarization of patients' monocytes. Fourteenout of 18 patients had a polarization lower than normal before any treatment. (b) The activity of serum fractions < 25 kD (LMWFs)from these patients expressed as the inhibition of FMLP-induced polarization of healthy donor monocytes. Monocytes were obtainedvia Ficoll/Percoll gradient centrifugation. The hatched areas represent the values found in healthy controls. Statistically significantdifferences are given (* P< 0-01, t P< 0-05; Wilcoxon).

viduals (mean 33, s.d. 9 (n= 36); or 32, s.d. 13; n= 41) or betweenindividuals less than 50 years and over 50 years of age (mean 34,s.d. 11 (n=66); or 31, s.d. 8; n= I1). When one monocytepopulation was tested several times (inter-assay variation)outcomes were 24 (mean, s.d. 3; n = 13). The intra-assayvariation also never exceeded 15% (n = 77). On the basis of theseoutcomes, FMLP-induced polarization values of less than 20%are considered to be abnormal.

Determination in patient serum of low molecular weight factors(LMWFs) inhibiting monocyte polarizationSera were collected from the patients by venepuncture anddiluted 1:1 in saline. These dilutions were subjected to ultrafilt-ration through Amicon CF25 Centriflo cones (Amicon,Danvers) for 15 min at 700g (molecular weight 'cut offpoint' 25kD). The residues were resuspended and stored at - 70'C untilfurther use.

The capability of the serum fractions to inhibit FMLP-induced polarization of healthy donor monocytes was deter-mined by incubating the monocytes (1 x 106/ml) during 15 minat 370C either with FMLP alone or with FMLP in combinationwith a serum fraction (final dilution 1: 60).

Addition of serum fractions alone to donor monocytes didnot affect the polarization. The percentage of inhibition ofFMLP-induced minus spontaneous polarization caused byaddition of the serum fractions was calculated.

The in vitro effects ofanti-PiSEmonoclonal antibodies and TP-Ion the activity ofserum LMWFsIn this series of experiments, elutriator-purified monocytes (deBoer & Roos, 1986) of one healthy donor were used as an

indicator system. In brief, mononuclear cells were separatedfrom 450 ml whole blood via percoll centrifugation (20 min,1000 g, room temperature). Thereafter the mononuclear cellswere injected into an elutriation centrifugation system (Beck-man J21 centrifuge with a JE-6 elutriation rotor). The elutria-tion medium was PBS with 13 mM trisodium citrate and 5 mg ofhuman albumin per ml. To separate the different cell popula-tions, the flow rate was kept constant at 20 ml/min while therotor speed was diminished from 4000 to 0 rev/min. The fractionat 2500 rev/min was collected. After Percoll gradient centrifuga-tion this fraction contains 93-97% monocytes as judged bypositivity for non-specific esterase activity. Monocytes were

stored in liquid nitrogen until use.

With this indicator system the P1 5E-like character of theLMWFs was validated. Adsorption experiments were carriedout by incubating the serum fractions with a combination oftwoP15E-specific monoclonal antibodies (see below) in a finaldilution of 1:200 at 40C for 16 h, followed by Amiconultrafiltration to remove formed complexes; this adsorptionprocedure was carried out twice (Tan et al., 1986b). Themonoclonal antibodies used were a combination of 4F5 and19F8 (anti-PI SE isotypes IgG2a and IgG2b; kindly provided byDr G. J. Cianciolo, Genentech Inc., Pharmacological Sciences,

307

v b )

Beforetreatment

Placeboprior toTP-I therapy

TP-1 therapy

Placebo afterTP-I therapy

2]


50

ena)

0

0

E

a1)

0

0

40

30

20

10

o,-- - - -

- 0

-I- '=- 7= = ~~- - -

0 5 10 15 20 24Months

Fig. 2. The fluctuations in time of FMLP-induced monocyte polarization. * healthy controls; 0 patients without TP- 1.

South San Francisco, CA). As a control antibody we used an

anti-human IgG (Tago, Burlingame, CA).Patient serum fractions of < 25 kDa diluted in culture fluid

or incubated with TP-I (final dilution 67 jug/ml) and serum

fractions after adsorption with the monoclonal antibodies were

tested in the monocyte polarization assay.

RESULTS

Ofthe 20 patients enrolled in the double-blind cross-over trial 18completed the 17 weeks course of treatment. One patient (afemale, 23 years old) was removed from the trial because shereacted with fever, nausea and malaise to the injections, an

allergy to calf thymic peptides was established by a positive skintest to the preparation. The other patient (a female, 53 years old)was wrongly enrolled in the trial; she had no CMI defectsdetectable before treatment. The 18 patients who completed thetrial did not have clinically adverse side effects of the treatment,biochemical blood changes such as abnormalities in the level ofbilirubin, ALAT, ASAT or creatinin were also not detectable.

When we decoded the patients at the end of the double-blindcross-over trial it appeared that seven patients had beenallocated to the group receiving placebo prior to TP- I therapy,the remaining 11 patients had received TP-1 first. This left us

with four situations to be evaluated: (1) performance before anytreatment (all 18 patients); (2) performance during and afterplacebo treatment prior to TP-1 therapy (seven patients); (3)performance during and after TP- 1 therapy (all 18 patients); (4)performance during and after placebo treatment after TP-1therapy (11 patients).

Monocyte polarizationWith regard to the polarization of the patients monocytes, 14 ofthe 18 patients included in the trial had a polarization lower thannormal before any treatment. The mean value found in these 14patients was 13% (s.d. 4%; Fig. la). Normal values in healthycontrols are over 20% (see Materials and Methods). Figure 2

shows the fluctuations in time of the monocyte chemotaxismeasured by polarization: when healthy donors are followed upfor 2 years, fluctuations range from 20% to 45%. In a series ofexperiments prior to this trial, 10 patients were left untreated;only one showed values of 19% to 22% in 3 months time, theothers did not reach values exceeding the 20% level. In contrast,there was a significant rise in the percentage of polarized

monocytes in this patient group after 2 months of TP- I therapyand values of 21% (mean)±8% (s.d.) were reached (Fig. Ia);besides seven out of 13 patients were now in the range ofnormalhealthy individuals. This restoring effect continued during theplacebo period after TP-I therapy and values of 25% (mean,s.d. + 7%, 1 month) and 22% (mean, s.d. + 4%, 2 months) were

reached, while six out of nine of the patients were in the normalrange. Placebo treatment prior to TP-I therapy had no signifi-cant effects on the defective monocyte polarization.

Serum PFSE-like factorsBefore any treatment, the 14 patients with a defective monocytepolarization had immunosuppressive LMWFs detectable intheir serum (see Fig. Ib and 3): serum fractions of <25 kDinduced a 51% (mean, s.d. + 9%) inhibition of FMLP-inducedpolarization of healthy donor monocytes obtained via Percollgradient centrifugation (Fig. Ib) and a 45% (mean, s.d. + 7%)inhibition of FMLP-induced polarization of healthy donormonocytes obtained via centrifugation with an elutriator (Fig.3). Figure 3 shows that the inhibitory effect of patient LMWFson the polarization of healthy donor monocytes could beneutralized by treating the LMWFs with a combination of twomonoclonal antibodies to P1 SE. A control antibody had no

such effect. This shows the P15E-like character of the LMWFsof these patients, as has been established before (van derPlassche-Boers et al., 1988).

After TP-I therapy patients serum LMWFs were no longerdetectable and values of inhibition of polarization decreased to

30% (s.d.+ 11%, I month therapy) and 29% (s.d.+8%, 2months therapy) respectively (Fig. lb). LMWFs isolated fromthe serum of healthy individuals also have such low inhibitoryeffects on healthy donor monocytes (<35%) inhibition ofpolarization of monocytes obtained via Percoll isolation (vander Plassche-Boers et al., 1988); < 30% inhibition of elutriator

purified monocytes (mean 14, s.d. 6; n = 34) (Fig. 3).Placebo treatment prior to TP- I therapy had no statistically

significant effects on the disappearance of the LMWFs fromserum, and the LMWFs soon reappeared in the placebo periodafter TP-I therapy (Fig. Ib).

In vitro TP-1 also appeared to be able to neutralize thesuppressive activity of patient serum LMWFs when added tothe system of elutriator-purified monocytes (Fig. 3). TP-1 had

no direct effect on spontaneous or FMLP-induced polarizationof the healthy donor monocytes.

308

I


30

lu

E

0c 200

c

0L._

0 100

FMLPTP-I

Anti-pIl5EAnti-IgGLMWFs

Fig. 3. Per cent polarization of elutriator-purified healthy donor monocytes with or without FMLP and the activity of serum fractions< 25 kD (LMWFs) from healthy donors and patients diluted in medium or incubated with TP- I (final dilution 67 jug/ml) and serumfractions after adsorption with the monoclonal antibodies anti-PI SE and anti-human IgG. Statistically significant differences are given(Wilcoxon).

DTH skin testing and MIFproductionThere were some improvements in DTH skin test reactivityduring TP-1 therapy (Table la). Regarding the Candidin skintest only 8 out of 18 patients showed a positive DTH test beforeany treatment, after 2 months of TP-1 therapy 14 out of 15patients showed a positive recording; values stayed high duringthe placebo period after TP-1 therapy (8 out of 10). Placebotreatment prior to TP-1 therapy had no such effects.

Regarding the Sk/Sd skin test, 5 out of 18 patients had a

positive recording before treatment, after TP-1 therapy 10 out of17 and 6 out of 10 (post TP- I placebo period) showed a positiveDTH, respectively. However, placebo treatment prior to TP- Itherapy had a curious effect; 6 out of 7 patients had a positiverecording.

The MIF production towards Con A, H. influenzae, Candi-din and Sk/Sd showed defective responses before treatment(Table I b); only nine out of 18, nine out of 18, 10 out of 18 andfour out of 18 patients were positive, respectively. Placebotreatment prior to TP-l therapy had no effect on MIFproduction. After 2 months of TP-1 therapy there was an

increase in responder rates in the case of Sk/Sd (seven out of 10)and H. influenzae (13 out of 17), the values decreasing againduring the placebo period after TP- I therapy (to one out of 1 Iand to six out of 11, respectively). However, increases inresponder rates did not reach statistical significance.

Surface markersWith regard to the number of CD3+ (total T), CD4+ (Th) andCD8+ (Ts/Tc) peripheral blood lymphocytes, near normalnumbers were found before any treatment, and numbers stayedwithin these normal limits throughout the period of treatmentirrespective of whether this was TP-1 or placebo. However, as

noticed earlier (Drexhage et al., 1985), there was a lowered CD2expression on peripheral lymphocytes in 40% of the patients atentrance to the trial ( < 63% of peripheral blood lymphocytes,normal values in our laboratory range from 63-79% oflymphocytes with a CD2-receptor). At the end of placebotreatment prior to TP- 1 therapy, 33% of the patients still had a

lowered CD2 expression on peripheral lymphocytes, but at theend of the 2 months ofTP- I therapy this value was decreased to13% of the patients. After TP-1 therapy the phenomenon of thelower expression of the CD2 marker reappeared and now 64%of the patients had diminished percentages of CD2+ lympho-cytes.

With regard to the peripheral B cells (as measured as CD24+cells) there was a significant increase in percentage during TP- 1

therapy.

Clinical improvementsThe clinical condition of the 18 patients showed a considerableimprovement during TP-1 therapy: subjectively 12 out of 15

309

Donor monocytes Effects of donor serum Effects of patient serumfractions <25 kD fractions <25 kD

0000~~

00 00 0

000~~~~~000 0

0 0~~~~~~~~0

0 0~~~~~~0

O 0 0~~~~~~~~~~~~~0~0

000 _

00~~~~~

_ * ..

0

0 0

0


Table 1. (a) Positive DTH skin test reactivities to the antigens H.influenzae (25 p1) and Candidin (1% solution). The diameter ofinduration (in mm) was recorded 24 h after antigen injection. Percent-ages of positive skin tests >20 mm induration) and P-values forstatistical differences (x2) are given. (b) Positive MIF production to themitogen ConA and the antigens H. influenzae, Candidin and Sk/Sd asmeasured as the macrophage migration inhibition. A percentage

inhibition of > 25% was considered as positive

(a) DTH skin tests

Candidin SK/SD

Before treatment 8/18 5/18

Placebo prior ¶to TP-1 therapy 4/6 6/7

TP-l therapy 14/15 10/17

Placebo afterTP-l therapy 8/10 6/10

(b) MIF production

ConA H.I. Candidin Sk/Sd

Before treatment 9/18 9/18 10/18 4/18Placebo prior

to TP-1 therapy 3/6 3/6 2/6 1/6TP-l therapy 9/17 13/17 9/17 7/10Placebo after 6/11 6/11 7/11 1/11

* P=0-35; t P=0-003; tP=0 07; §P=0-009; P=0 06; ** P=0 09.

were clearly feeling better during specific therapy (Fig. 4a) andshowed an absent mucopurulent secretion on inspection of thenasal mucosa (Fig. 4b). Placebo treatment prior to TP- I therapyhad no such effects and rapid relapses were recorded during theplacebo period after TP-I therapy.

The prevalence of a positive nasal bacterial culture rate (H.influenzae, Branhamella, Citrobacter, Klebsiella, Moraxella, S.

pneumoniae, other Steptococci and/or Staphylococci) decreasedfrom 14 out of 16 before any treatment to five out of 15 duringTP-I treatment (Fig. 4c), followed by a rise to again 8 out of 9

positivity 2 months after TP-l therapy. TP-I treatment had no

effects on the BSR. Only one of the patients had besides the CMI

defects signs of an atopic allergy (raised serum IgE, positiveRAST, asthmatic attacks). TP- 1 treatment had no effect on the

atopic complaints and IgE levels stayed high.

DISCUSSION

This report shows the beneficial clinical effects of TP-I treat-

ment in a double blind cross-over trial on the clinical condition,the purulent nasal discharge and the bacterial invasion of the

upper airways in patients with chronic purulent rhinosinusitis

resistant to the current therapies of surgery and antibiotics.The clinical improvements during TP- 1 therapy were accom-

panied by a raised performance of the cell-mediated immune

system of the patients, which probably forms the basis for their

improved clinical condition. The most predominant and signifi-cant effects were recorded in monocyte chemotaxis as measuredin the polarization assay. MIF production by peripheral bloodlymphocytes and DTH skin test reactivity to microbial antigenswere also enhanced during TP-1 treatment, but in generalstatistical significance was not reached, except for the CandidinDTH skin test.

The number of patients in this double-blind cross-over trialis limited and the period of active treatment (8 weeks) is rathershort. However, similar effects were obtained in an open non-controlled trial with TP-l in 10 patients with CMI defects and arelapsing of CPR (carried out before this double-blind, cross-over trial). All treated patients showed clinical improvements:they were feeling better and their nasal mucosa showed less toabsent mucopurulent secretion on inspection. The positivebacterial rate decreased from eight out of 10 to three out of 10.The clinical improvements were accompanied by a betterperformance of their CMI; the clearest effects were againrecorded in the monocyte polarization assay (before treatment10 (mean, s.d. 5; n=8); after 8 weeks of TP-l treatment 27(mean, s.d. 16; n= 8).

At an earlier occasion we presented evidence that theimpairment in monocyte polarization found in our patientgroup is most probably due to the presence in serum of a lowmolecular weight factor of <25 kD showing a structuralhomology with P1 SE of murine and feline leukaemia virus (vander Plassche-Boers et al., 1988). The data presented in thisreport further strengthen this view. P1SE-like factors weredetected in the serum of 14 out of 18 patients before anytreatment; these patients also showed an impaired monocytepolarization. There was a functional disappearance or loweringof the activity of the P1 SE-like factors during TP- 1 treatment inthe double-blind cross-over trial. In the above-mentioned opentrial five patients were tested for the presence of the P1 SE-likefactor in serum; in these five patients the factor also disappearedafter TP-1 treatment, expressed as the inhibition of monocytepolarization: before treatment 52% (mean, s.d. 13), after 8weeks of TP-l treatment 19% (mean, s.d. 16). Whether thisshows a real molecular disappearance of the factor from theserum ofTP- I treated patients (non-production) or a functionalneutralization of the factor by thymic hormones in the serum,needs further study. Also, in vitro TP- I neutralized the suppres-sive activity of patient serum LMWFs. TP-1 had no effect on

spontaneous or FMLP-induced polarization of healthy donormonocytes.

The most well-known function of thymic hormones is theireffect on the maturation and differentiation of T cells in thethymic micro-environment. Many studies have focused on thestimulating effects of TP-l on T cell maturation and reactivityboth in primary as well as in secundary immune deficiency states

(Aiuti et al., 1980; Fiorelli et al., 1985; Liberati et al., 1988). Thebeneficial outcomes of TP- I treatment in primary immunodefi-ciencies in neonates can be attributed to this effect.

From our data it is, however, tempting to speculate thatsome of the compounds of the thymic peptide hormonepreparation may exert their effect via cells of the mononuclearphagocyte system, thereby indirectly influencing T cell mediatedimmunity. Such function could explain the beneficial effects ofthymic hormone preparations seen in adults with secundaryimmunodeficiencies and a full T cell repertoire. Reports on theeffects of thymic hormones on the functioning of cells of the

310


( a )

No complaints

Complaints

( b)

Subjective improvements*

.

I

IF

II I I Il

1 2 2 2

Mucopurulent secretion*

#~~ .. 1fPI ItXtt"

.. * e

1 2 2 2

Present

Absent F

( c)

Positive forcom ensal

bacteria

Negative I

Before Placebo priortreatment to TP-1 therapy

Nasal bacterial culture rate*~~~~~~~~~~~

I t I

:"PW

2 2 2 -Months

TP-1 therapy Placebo afterTP-1 therapy

Fig. 4. Clinical improvements. (a) Subjective improvements after TP-1 therapy. (b) absence of mucopurulent secretion after TP-1

therapy; (c) the prevalence ofpositive bacterial culture rate (H. influenze, Branhamella, Klebsiella, Moraxella, Citrobacter, Streptococciand/or Staphylococci) decreased after TP-1 therapy. *Statistically significant difference; x2; P <0001, t P< 0 002, 1 P< 0 05).

mononuclear phagocyte system are, however, extremely scarce.

Davis et al. (1984) reported an effect ofTP- 1 treatment on serum

lysozyme levels in patients with Hodgkin's disease, whichsuggested an increase in monocyte function and/or mass. Thereis also a report on a disturbed monocyte natural cytotoxicity inviral hepatitis, the function of which became normal after invitro incubation of the monocytes with thymic preparations(Kiczka et al., 1985). Zatz et al. (1985) stated that theaugmentation ofIL-2 production ofperipheral blood leucocytesinduced by thymosin fraction 5 was prevented by monocytedepletion. They interpreted their data as demonstrating that thethymic hormone effect in lymphocyte IL-2 production was

mediated directly or indirectly via a monocyte population. Thisview that TP- I may also act on accessory cells is furthersupported by preliminary experiments carried out in our

laboratory, which show that P- I SE inhibited clustering ofdendritic cells (a monocyte-like cell; Kabel et al., 1989) is alsorestored to normal by the in vitro addition of TP- 1.

Current treatment protocols of chronic purulent rhino-sinusitis form an agony for both patient and physician (frequentconsultations, surgical interventions and several courses ofantibiotics). Our data support a further critical evaluation oftheuse of thymic hormone preparations and justify larger con-

trolled trials. Purification procedures are necessary to determine

311

I

I I IIl-a

312 M. Tas, J. A. Leezenberg & H. A. Drexhage

which thymic hormone fraction is responsible for the neutraliz-ing effects on the activity of P1 5E-like factors in monocytepolarization.

ACKNOWLEDGMENTS

The authors wish to express their thanks to A. de Boer and A. Heenemanfor their help in preparing the manuscript and Meeny de Haan-Meulman and Diny Knol for technical assistance.

REFERENCES

AIUTI, F., FIORELLI, M. & SCHIRMACHER, V. (1975) Effect of thymus

factor on human precursor T lymphocytes. Clin. exp. Immunol.20,499.

AIUTI, F., D'AMELIO, R. & GIUNCHI, G. (1980) Thymic hormones andtheir clinical use in immunodeficiencies. In Thymus, Thymic Hor-mones and Tlymphocytes, Serono Symposia No. 38 (ed. by F. Aiuti &H. Wigzell) p. 375. Academic Press, London and New York.

BALM, A.J.M., VON BLOMBERG-VAN DER FLIER, B.M.E., DREXHAGE,H.A., DE HAAN-MEULMAN, M. & SNOW, G.B. (1984) Mononuclearphagocyte function in head and neck cancer: depression of murinemacrophage accumulation by low molecular weight factors derivedfrom head and neck carcinomas. Laryngoscope, 94, 223.

CIANCIOLO, G.J., MATTHEWS, T.J., BOLOGNESI, D.P. & SYNDERMAN, R.

(1980) Macrophage accumulation in mice is inhibited by lowmolecular weight products from murine leukemia viruses. J. Immu-nol. 124, 2900.

CIANCIOLO, G.J. & SNYDERMAN, R. (1981) Monocyte responsiveness to

chemotactic stimuli is a property of a subpopulation of cells that can

respond to multiple chemoattractants. J. clin. Invest. 67, 60.CIANCIOLO, G.J., HUNTER, J., SILVA, J., HASKILL, J.S. & SNYDERMAN, R.

(1981) Inhibitors of monocyte responses to chemotaxins are presentin human cancerous effusions and react with monoclonal antibodiesto the P15(E) structural protein of retroviruses. J. clin. Invest. 68, 831.

CIANCIOLO, G.J., PHIPPS, D. & SNYDERMAN, R. (1984) Human malig-

nant & mitogen-transformed cells contain retroviral p1SE relatedantigen. J. exp. Med. 159, 964.

DAVIS, S., CERNETTI, C., SPINOZZI, F. & GRIGNANI, F. (1984) Hodgkin'sDisease: role of thymostimulin (TP-1) as an immunomodulator. InThymic Factor Therapy (ed. by N. A. Byrom & J. R. Hobbs) p. 365.Serono Symposia Publications, Raven Press, New York.

DE BOER, M. & Roos, D. (1986) Metabolic comparison betweenbasophils and other leukocytes from human blood. J. Immunol.136, 3447.

DREXHAGE, H.A., VAN DER PLASSCHE-BOERS, E.M., KOKJE-KLEINGELD,M. & LEEZENBERG, J.A. (1983) Abnormalities in cell-mediatedimmune functions to Haemophilus influenzae in chronic purulentinfections of the upper respiratory tract. Clin. Immunol. immuno-pathol. 28, 218.

DREXHAGE, H.A., VAN DER PLASSCHE-BOERS, E.M., KOKJE-KLEINGELD,M. & LEEZENBERG, J.A. (1985) Subtle antigen-selective T-cell defectsas a cause of relapsing purulent infections of the upper respiratorytract. In Recent Advances in Primary and Aquired Immunodeficiencies,vol. 28 (ed. by F. Aiuti, F. Rosen & M. D. Cooper) p. 395. SeronoSymposia Publications, Raven Press, New York.

FALCHETTI, R., BERGESI, G., ESHKOL, A., CAFIERO, C., ADORINI, L. &

CAPRINO, L. (1977) Pharmacological and biological properties of a

calf thymus extract (TP-1) Drugs Exp. Clin. Res. 3, 39.FIORELLI, M., CRESCENZI, M. & AIUTI, F. (1985) Thymic hormone

therapy of viral infections. Eos. Riv. immunol. Immunofarm. 5, 72.IWATA, T., INCEFY, G.S., CUNNINGHAM-RUNDLES, S., SMITHWICK, E.,

GELLER, N., O'REILLY, R. & GOOD, R.A. (1981) Circulating thymichormone activity in patients with primary and secondary immunode-ficiency diseases. Am. J. Med. 71, 385.

IWATA, T., INCEFY, G.S., CUNNINGHAM-RUNDLES, S., DARDENNE, M.,

KAPOOR, N., KIRKPATRICK, D. & O'REILLY, R.J. (1983) Circulatingthymic hormone levels in severe combined immunodeficiency. Clin.Exp. Immunol. 53, 1.

KABEL, P.J., DE HAAN-MEULMAN, M., VOORBIJ, H.A.M., KLEINGELD,M., KNOL, E.F. & DREXHAGE, H.A. (1989) Accessory cells with a

morphology and marker pattern of dendritic cells can be obtainedfrom elutriator purified blood monocyte fractions. An enhancingeffect ofmetrizamide in this differentiation. Immunobiology, 179, 395.

KICZKA, W., SZKARADKIEWICZ, A. & FLIEGER, J. (1985) Studies on

natural cytotoxicity of human monocytes in viral hepatitis. Immunol.Lett. 9, 87.

LEWIS, V., TWOMEY, J.J., GOLDSTEIN, G., O'REILLY, R., SMITHWICK, E.,PAHWA, R., PAHWA, S., GOOD, R.A., SCHULTE-WISSERMAN, H.,HOROWITZ, S., HONG, R., JONES, J., SIEBER, O., KIRKPATRICK, C.,POLMAR, S. & BEALMAR, P. (1977) Circulating thymic hormoneactivity in congenital immunodeficiency. Lancet ii, 471.

LIBERATI, A.M., BALLATORI, E., FiZZOTTi, M., SCHIPPA, M., CINI, L.,CINIERI, S., PROIETTI, M.G., DI MARZIO, R., SENATORE, M. &GREGNANI, F. (1988) A randomized trial to evaluate the immunores-torative properties of thymostimulin in patients with Hodgkin'sdisease in complete remission. Cancer Immunol. Immunother. 26, 87.

LIN, C.Y., Hsu, H.C. & HSIEH, H.C. (1985) Treatment of progressiveBacillus Calmette-Guerin infection in an immunodeficient infant witha specific bovine extract (Thymostimulin). Ped. Infect. Dis. 4, 402.

LIN, C.Y., Hsu, C.H., Liu, K.C., CHEN, C.L. & Hsu, H.C. (1986) Serialimmunologic and histologic studies in the treatment of necrotizingfasciitis with combined immunodeficiency by a bovine thymic extract

(Thymostimulin). J. Ped. Surg. 21, 1000.LIN, C.Y., Hsu, H.C., CHEN, C.L. & SHEN, E.Y. (1987) Treatment ofcombined immunodeficiency with thymic extract (Thymostimulin).Ann. Allergy 58, 379.

LIN, C.Y., Kuo, Y.C., LIN, C.C. & Ou, B.R. (1988) Enhancement ofinterleukin-2 and gamma-interferon production in vitro on cordblood lymphocytes and in vivo on primary cellular immunodeficiencypatients with thymic extract (Thymostimulin). J. clin. Immunol.8, 103.

MADERAZO, E.G., ANTON, T.F. & WARD, P.A. (1978) Defectivemonocyte and polymorphonuclear leukocyte chemotaxis in atopicdisease. J. Allergy clin. Immunol. 61, 288.

MULLINK, H., VON BLOMBERG-VAN DER FLIER, M., WILDERS, M.,DREXHAGE H.A. & ALONS, C.L. (1979) A simple cytochemical methodfor distinguishing EAC rosettes formed by lymphocytes and mono-

cytes. J. Immunol. Methods, 29,133.PERTOFT, H. JOHNSSON, A., WXRMEGARD, B. & SELJELID, R. (1980)

Separation of human monocytes on density gradients of Percoll. J.Immunol. Methods, 33, 221.

PIKE, M.C. & SNYDERMAN, R. (1976) Depression of macrophagefunction by a factor produced by neoplasms: a mechanism forabrogation of immune surveillance. J. Immunol. 117, 1243.

SINGH, K. & KHAN, A. (1982) The usefulness of a monocytoid cell line(U-937) in the macrophage migration inhibition assay. J. Clin.Hemat. Oncol. 12, 29.

SNYDERMAN, R., PIKE, M.C. & CIANCIOLO, G.J. (1980) An inhibitor ofmacrophage accumulation produced by neoplasms: its role inabrogating host resistance to cancer. In Mononuclear Phagocytes.Functional Aspects, part I. (ed. by R. van Furth) p. 569. MartinusNijhoff Publishers, The Hague, Boston, London.

SNYDERMAN, R. & CIANCIOLO, G.J. (1984) Immunosuppressive activityofthe retroviral envelope protein p1 SE and its possible relationship toneoplasia. Immunol. Today, 5, 240.

SKOTNICKI, A.B., DABROWSKA-BERNSTEIN, B.K., DABROSKI, M.P.,GORSKY, A.J., CZARNECKI, J. & ALEKSANDROWICZ, J. (1984) Biologi-cal properties and clinical use of calf thymus extract TFX-Polfa. InThymic Hormones and Lymphokines (ed. by A. L. Goldstein) p. 545.Plenum Press, New York.

TAN, I.B., DREXHAGE, H.A., SCHEPER, R.J., VON BLOMBERG-VAN DER

FLIER B.M.E., DE HAAN-MEULMAN M., SNOW, G.B. & BALM, A.J.M.

Thymostimulin and chronic purulent rhinosinusitis 313

(1986a) Defective monocyte chemotaxis in patients with head andneck cancer. Restoration after treatment. Arch. Otolaryngol. HeadNeck Surg. 112, 541.

TAN, I.B., DREXHAGE, H.A., SCHEPER, R.J., VON BLOMBERG-VAN DERFLIER, B.M.E., DE HAAN-MEULMAN, M., SNOW, G.B. & BALM, A.J.M.(1986b) immunosuppressive retroviral P1SE-related factors in headand neck carcinomas. Arch. Otolaryngol. Head Neck Surg. 112, 942.

TAN, I.B., DREXHAGE, H.A., MULLINK, H., HENZEN-LOGMANS, S.C., DEHAAN-MEULMAN, M., SNOW, G.B. & BALM, A.J.M. (1987) Immuno-histochemical detection of retroviral p1SE-related material in carci-nomas of the head and neck. Otolaryngol. Head Neck Surg. 96, 251.

TAS, M., DREXHAGE, H.A. & GOUDSMIT, J. (1988) A monocytechemotaxis inhibiting factor in serum of HIV infected men sharesepitopes with the HIV transmembrane protein gp4l. Clin. Exp.Immunol. 71, 13.

THURMAN, G.B., STULL, H.B., MILLER, P.J., STEVENSON, H.C. &OLDMAN, R.K. (1983) Utilization of purified human monocytes in theagarose droplet assay for measuring migration inhibitory factors. J.Immunol. Methods, 65, 41.

VAN DER BAAN, S., VEERMAN, A.J.P., HEIDENDAHL, G.A.K., DEN

HOLLANDER, W. & FEENSTRA, L. (1987) Primary ciliary dyskinesiaand nasal mucociliary clearance. Respiration, 52, 69.

VAN DER PLASSCHE-BOERS, E.M., DREXHAGE, H.A. & KOKJt-KLEIN-GELD, M. (1985) The use of somatic antigen of Heamophilus influenzaefor the monitoring of T cell mediated skin test reactivity inman. J.immunol. Methods, 83, 353.

VAN DER PLASSCHE-BOERs, E.M., DREXHAGE, H.A., KOKIJ-KLEINGELD,M. & LEEZENBERG, J.A. (1986) Parameters of T-cell mediatedimmunity to commensal micro-organisms in patients with chronicpurulent rhinosinusitis: a comparison between delayed type hyper-sensitivity skin test, lymphocyte transformation test and macrophagemigration inhibition factor assay. Clin. exp. Immunol. 66, 516.

VAN DER PLASSCHE-BOERS, E.M., TAS, M., DE HAAN-MEULMAN, M.,KLEINGELD, M. & DREXHAGE, H.A. (1988) Abnormal monocytechemotaxis in patients with chronic purulent rhinosinusitis; an effectof retroviral plSE-related factors in serum. Clin. exp. Immunol.73, 348.

ZATZ, M.M., SKOTNICKI, A., BAILEY, J.M., OLIVER, J.H. & GOLDSTEIN,A.L. (1985) Mechanism of action of thymosin. II. Effects of aspirinand thymosin on enhancement of 11-2 production. Immunopharmaco-logy, 9, 189.

Documents

and chronic purulent rhinosinusitis. A double-blind cross-over trial