A direct comparison of biological response modulation and clinical side effects by interferon beta ser, interferon gamma, or the combination of interferons beta ser and gamma in humans

A direct comparison of biological response

modulation and clinical side effects by

interferon-beta ser, interferon-gamma, or the

combination of interferons beta ser and gamma

in humans.

J H Schiller, … , P L Witt, E C Borden

J Clin Invest.

1990;

86(4)

:1211-1221.

https://doi.org/10.1172/JCI114827

.

To directly compare clinical side effects and biological response modification, IFN-beta ser,

IFN-gamma, or the combination of IFN-beta ser plus IFN-gamma was administered to 21

cancer patients. Each IFN or the combination was given intravenously on days 1, 8, and 15

in varied order. Each IFN and the combination resulted in significant (P less than 0.05)

modulation of IFN-induced proteins. IFN-beta ser was more effective than IFN-gamma in

enhancing 2-5A synthetase activity (P = 0.001). gamma was more effective than

IFN-beta ser in enhancing serum IFN-beta 2 microglobulin expression (P = 0.05) and indoleamine

dioxygenase activity, as assessed by decreased serum tryptophan (P = 0.03). The

combination enhanced tryptophan catabolism more effectively than IFN-beta ser in a

dose-dependent manner (P less than 0.03). IFN-beta ser/IFN-gamma did not potentiate natural

killer cells or antibody-dependent cellular toxicity (ADCC). IFN-beta ser/IFN-gamma

enhanced monocyte guanylate cyclase activity, as assessed by serum neopterin, more

effectively than IFN-gamma alone (P = 0.005). Both IFNs and the combination resulted in

increases in HLA class II expression on monocytes. However, no significant difference in

the level of induction of HLA DQ and HLA DR expression between IFN-beta

ser/IFN-gamma and either IFN-beta ser or IFN-ser/IFN-gamma was noted. Although frequency and servity of

side effects of IFN-beta ser, IFN-gamma, or the combination were dose related, induction of

induced proteins (with […]

Research Article

Find the latest version:

A Direct

Comparison of Biological Response Modulation

and

Clinical Side

Effects by

Interferon-Betaser,

Interferon-Gamma,

or

the

Combination

of Interferons Betaser and Gamma

in Humans

Joan H. Schiller, Barry Storer, DonnaM.Paulnock, RaymondR.Brown, SurinderP.Datta,Patricia L.Witt,andErnest C.Borden

DepartmentsofHuman Oncology,Medical Microbiology, Medicine, and Statistics, University of Wisconsin Clinical CancerCenter,

Madison, Wisconsin53792;and William S. Middleton Veterans AdministrationHospital, Madison, Wisconsin 53705

Abstract

To directly compare clinical side effects and biological

re-sponsemodification,

IFN-#,,er,

IFN-'y,

orthecombination of

IFN-jfte,

plus

IFN-'y

wasadministered to 21 cancer patients. Each IFN or thecombinationwasgivenintravenouslyondays 1, 8, and 15 in varied order. Each IFN and the combination resulted in significant (P<0.05)modulation of IFN-induced proteins.

IFN-#,i.r

was moreeffectivethanIFN-yin enhancing 2-5A synthetaseactivity (P= 0.001). IFN--ywas more effec-tive than

IFN-jie

in enhancing serum 2 microglobulin

ex-pression (P=0.05)and indoleamine dioxygenase activity,as

assessedby decreasedserumtryptophan(P=_0.03).The

com-bination enhancedtryptophan catabolismmoreeffectivelythan

IFN-fser

in a dose-dependent manner (P < 0.03).

_IFN-#S,.r/

IFN-,y did notpotentiate natural killercells or antibody-de-pendent cellular toxicity (ADCC).

IFN-,8Ser/IFN-y

enhanced monocyte guanylate cyclase activity, as assessed by serum

neopterin,moreeffectivelythan

IFN-'y

alone(P=0.005).Both IFNs and thecombination resulted inincreasesin HLA class IIexpressionon monocytes.However,nosignificant difference in the level of induction of HLA DQ and HLADRexpression between

_{IFN-#,,/IFN-y}

and either

IFN-#,,,,

or IFN-7 was

noted. Although frequency and servity of side effects of

IFN-.t~er,

IFN-"Y,

orthecombinationweredose related, induc-tionof inducedproteins (with exception ofinfluenceson tryp-tophan catabolism) were not a function of dose administered

overthe10-fold range.Continuedtreatmentwith the combina-tionintravenouslythreetimesa weekfor 4 wksustained but did notfurther potentiate, mostof the changes in interferon-induced proteins. Thus,

_IFN-#ser

and IFN--y each resulted in effectiveandessentially equivalentpatterns ofinductionof in-ducedproteins. Whencombined,however, these IFNs did not result in potentiation of biological response modification in vivo.(J. Clin. Invest. 1990. 86:1211-1221.) Key words: clini-calcancertrial * interferon , biological responsemodification Introduction

IFNs are prototype

lymphokines

with a broad rangeof anti-viral,immunomodulatory, and antiproliferative effects. Three major classes (IFNs alpha, beta, and gamma) have been de-fined based on differences in amino acid structure. Biologic

Addressreprintrequests to Dr. Joan H.Schiller, University of

Wiscon-sin, Clinical Cancer Center, Rom K4/666, 600 Highland Ave.,

Madison, WI53792.

Receivedfor publication 15 December 1989 and in revisedform 25

April1990.

The Journalof ClinicalInvestigation, Inc. Volume86,October1990, 1211-1221

differences between the three groups also exist that may have therapeutic implications. IFN-,B has been shown to have more antiproliferative activity in vitro against many solid tumors

than IFN-a (1).

IFN-y

has uniqueimmunomodulatory prop-erties,including macrophage activation (2) and differentiating activity (3), and was more potent in augmenting Fc receptor expression (4) and HLA class II antigen expression on tumor cells(5), than IFN-aor

IFN-f3.

Clinically,IFN-a has well-defined activity against a variety ofneoplasms and is currently undergoing evaluation in phase IIItrials(6). Combinations of IFN-aor

_IFN-fler

and

IFN--y

are of clinical interest because ofsynergistic antiproliferative ef-fects in a variety of tumor types in both in vitro and murine model systems(7-9).Recentclinical trials of combinations of

IFN-#

orIFN-a with

IFN-y

(10, 11) have raised the question

as tohowacombination of IFNs will be tolerated when

com-pared with either IFN alone. Furthermore, results froma

phase I trial of IFN-a and

IFN-'y

suggest that this combination may not be tolerablewhenadministered overlongperiodsof time(1 1).

We conducteda phaseItrial of

IFN-3,B,,

IFN-'y,

andthe combination

(IFN-Bfer/IFN-y)

that was designed to directly compare in thesamepatientthese three alternatives. Specific clinicalobjectiveswere tocompare side effects and organ

tox-icities associated with three different IFN treatments, deter-mine whether the combined administration of

IFN43ser/IFN-y

would result inincreasedtoxicityascompared with eitherIFN alone,and determine whethera 10-foldincreasein IFN dose would resultin enhancedtoxicity.To comparemodulationof host responses, IFN-induced proteins(HLA class I and class II expression,2-5Asynthetase, andindoleamine 2-3 dioxygen-ase)and effector cell function natural killer

(NK)'

cell cytotox-icity, antibody-dependent cellular toxicity (ADCC), and monocyte activation) were assessed. These parameters were chosen for study because theyrepresent host responses that wereknown to be modulated by

_IFN-fl,,e

andIFN-y both in vitro and in vivo. The purpose ofthese assessments was to determine(a) theeffect ofbothsingle administrationand the combination on biological responses, (b) which oftwo dose levels of combined

_IFN-fe,

r/IFN-'y,

differingby 1O-fold,would produce the greatest augmentation ofthese biological re-sponses, and (c) whether the combined administration of

IFN-#,B/IFN-'y

would resultinsustainedincreases in biologi-cal responsesover 4wk.

Materials and Methods

IFNs.

_IFN-f,,,.

(specific antiviral activity 1-2X 108units/mg of

pro-tein; Triton Biosciences, Inc.) and IFN--y (specific antiviral activity

1.Abbreviations usedinthispaper:ADCC,

antibody-dependent

cellu-lartoxicity;MR,meanfluorescence intensity;NK,naturalkiller.

1.3-3.3 X107 units/mgof protein; Biogen, Cambridge, MA)(12)were provided through the Biological Response Modifiers Program,

Na-tional Institutes ofHealth (Frederick, MD).

IFN-3,,r

is a synthetic mutant protein with a substitution of serine for cysteine at amino acid

position 17(13).

Patientselection. This study was open to patients with

histologi-cally confirmed malignancies for which no standard therapy was known to beeffective.Eligibility criteria included an Eastern

Coopera-tive Oncology Group performancestatusof0 or 1 and adequate

hema-tologic,hepatic, and renal function. Patientscould havereceived no morethan twoprior cytotoxicchemotherapyregimens and could not havereceived cytotoxic chemotherapy foratleast 3 wkbefore initia-tion of IFN (6 wk formitomycinC ornitrosoureas);radiation therapy,

hormones,glucocorticosteroids, forat least 2 wk; or IFN-a or other

immunotherapy foratleast 4 wk. Patientswithclinically significant heartdisease,CNSinvolvement,orprior treatment withIFN43seror

IFN--ywereexcluded.While on study, patients were not allowed barbi-turates, aspirin, corticosteroids, or nonsteroidal antiinflammatory drugs.Allpatientsgaveinformedconsent.

21 patients with a variety oftumor types wereentered onstudy (Table I).One patientatthehigher dose level discontinued therapy

within thefirst 3 wk due to personal reasons and was not evaluatedfor

response.

Studydesign.Two dose levels were studied (TableII).The low-dose levelAconsisted of 3X 106IUof

_IFN-#,B3

and200Mgof IFN-y (1,000

_g 15 X 106IU).This dose levelrepresentedone-tenth thedose of thehigh-doselevel B(30X 106IUofIFN-Oserand2,000MAgofIFN-,y),

which was themaximum tolerated dosedetermined fromaprevious

phaseItrial(10). For allcombinedtreatments,IFN-,y wasgiven first intravenouslyover10minfollowed bya10-mininjection of

IFN-#ser.

Clinicalparameterswereevaluatedbeforeand 24 haftertheday1,8,

and 15 treatment. Each injection was followed by a 1-wk washout

period. Clinical parameters,biologicalresponse anddiseaseresponse were assessedagainafter4 wkof maintenancetherapy. The orderof administration ofIFN-Oser,IFN--y,andIFN-4ser/IFN-,y ondays 1, 8,

and 15 wasdistributed equally amongpatientsin each dose levelto

control forcumulative dose effects.

TableI.PatientCharacteristics

Low dose Highdose

Age range 39-70 26-69

Median 57 52

Sex

Male 4 5

Female 6 6

Priortreatment

None 4 1

Chemotherapy 5 4

Radiation therapy 1 7

Immunotherapy 4 1

Hyperthermiaorhormonaltherapy 0 2

Performancestatus:

0 5

1 5 10

Tumor

Lung 2 2

Melanoma 1 I

Renal 1 3

Other* 6 5

Totalnumber 10 11

IFNdosesweremodified for grade 3or4toxicities. Grade 3

symp-tomatology was definedas symptoms that required medication for relief, didnotresolve with repeatedIFN treatments,resultedin5-10% weightloss frombaseline,orresulted inadropinperformancestatus to

ECOG level 3. Grade 4 symptomatology includedsymptomsthatwere inadequately relieved by medication,causeda> 10% weight loss,or a dropinperformancestatus tolevel 4.

IFNtreatmentswerewithheld foranygrade3or4toxicityuntil the

parameterreturnedtoless thanagrade3toxicity.Therapywasthen resumed with the dose ofeach IFN reduced byonedose levelorto50% of dose level A.

IFNpharmacokinetics.Serumsampleswereobtainedimmediately before initiationof the IFN infusion andat15min, 1, 2, 4, 6,and24 h

thereafter.

Because an ELISAassay specific for_IFN-#ser wasnotavailable, totalserumantiviral activitywasassayed(byCetusCorp., Emeryville, CA) in acytopathic effect assay (CPE) thatuses human fibroblast GM2504E cells andVSVasthechallengevirus. An internallaboratory standard calibrated againstaninternational referencepreparation of human fibroblast IFN (G023-902-527) was included in eachassay.

Titerswerereportedininternational units/milliliterandarethemean of three titrationresults. BecauseIFN-#andIFN-ydonotshowequal activity on a per unit basis in the assay, total antiviral activity as measured here shouldnotbeinterpretedasthesumofthe actualIFN-fl and IFN--y titers. Serum IFN-y levelswere determined(byBiogen, Inc.) withamodified double-antibody sandwich ELISA. An internal laboratory standardwasusedtogenerateastandardcurve,which had been calibrated againstaninternational standard for antiviralpotency, G023-901-530,togive valuesexpressedin international units/milli-liter. Valuesrepresentthemeanof four titration results.

2-5Asynthetaseandserum[2microglobulin.The 2-SA

oligoade-nylate synthetaseassaywasperformedaspreviouslydescribedon

pe-ripheral blood lymphocytes (14).Acommerciallyavailable radioim-muneassaykit(Pharmacia Fine Chemicals, Piscataway, NJ)wasused for the quantitativedetermination ofserum[2microglobulinlevels.

Monocytesurface antigen expression.Determinationofmonocyte surface antigen expressionwasperformedaspreviouslydescribed(15). Both thepercentageofpositivecells and themeanchannel of fluores-cenceintensity(MFI)ofpositivecellsweredeterminedineach

analy-ses.Cells correspondingtothe monocytepopulationwereidentifiedon the basis of characteristic forwardangle and90° (right angle) light

scatter profile (16). Insome experiments, themonocyte population identified by light scatter propertieswasconfirmed by stainingwith

anti-Leu-M3 antibody. In all cases where these two methods were compared,identical resultswereobtained whether theidentification of monocyteswasdoneby lightscatterprofile analysisoronthe basisof

Leu-M3 stainingresults(datanotshown).

NKand ADCC cellcytotoxicity. Peripheralmononuclear cellswere separatedfrom whole blood by centrifugation over aFicollhypaque gradient.NKcell activitywasassessed ina4-h51Cr-releaseassaysusing

K562 target cells (17), and ADCC activity was assessed using 5'Cr labelled Changcells. ADCCwasassessedby addingeither mediumor rabbit anti-Changcell antiserum(finaldilution= _10-2)toappropriate cultures.NKand ADCC cell_activitywasdetermined inquadruplicate

at six effector-target ratios (3.125:1, 6.25:1, 12.5:5, 25:1, 50:1, and

100:1). Lytic units(LU) were calculated using amodification of the

exponential fitequation. One LUwasdefinedasthe number of lym-phocytes requiredtoresult in 30%specific 5'Crrelease.

Serum neopterin levels. Serum neopterin was assayed bya com-mercially available RIA kit (Neopterin-RIAcid; Henning Berlin

GmbH, distributed by DRG International, Inc., Mountainside, NJ). Samples were counted in a 10-well gamma counter (Isodata, Inc., Palatine, IL). Asimultaneously preparedstandard curve wasused to

calculateneopterin levels in the testsamples.

Urinary kynurenineandserum tryptophan levels. 24-h urine

col-lections, preserved byaddition ofthymol,were madebefore, and24

and 48 h afterIFN administration. Kynureninewas measured_by an

alkaline steam-volatile aminemethod previouslydescribed from this laboratory ( 18). Appropriatestandardcurvesand recoverieswere used

1212 Schiller, Storer, Paulnock, Brown, Datta, Witt, and Borden

*_{Breast, colorectal, pancreas, chordoma, adenocarcinoma of}

TableILSchema

Doselevels

A B DayI Day 8 Day 15 Maintenance therapy

IFN-fser 3X 106units 30x 106units

IFN-'y 200Ag 2,000/Ag

No.ofpatients

4 4 IFN-bser IFN-y IFN-flser/IFN-y IFN-flwr/IFN-y three_times/wk 3 3 IFN-y IFN-fser/IFN-y IFN-fser IFN-#ser/IFN-y

threetime/wk

3 4 _{IFN-lser/IFN-7 IFN-Oser} IFN-y

_{IFN-#ler/IFN-y}

threetimes/wk

Total 10 11

BiologicalResponse Assessment Day 1, 2 Day8,9 Day 15, 16 After4wkof maintenance therapy

forquantitationresults. The methodis sensitiveandspecificwithno

interference fromcommonly usedmedications.

Bloodserumsamples fortryptophanlevels wereobtained before

treatment,andat4, 8, and 24 hafter IFN administration. Serumor plasmatryptophan levelswere measured bya fluorometric method adaptedtotheTechnicon Autoanalyzer(TechniconInstruments, Tar-rytown, NY)(19).

Statisticalanalysis. Pairedt tests wereused to testthe null hypoth-esisthatmean treatmenteffectswerezero,bothwithineach dose level andforboth doselevelscombined,foreachinterferonused alone and

forthecombination. Necessarily, this analysiswas restrictedto pa-tients for whom both pre- and posttreatment determinationswere madeforaparticularIFN type. Thetrialwasbalancedwithrespect to the sequencingof IFNs. Approximately equal numbers ofpatients

receivedeachIFNasthefirst, second,orthirdinjections.

Pairedt tests were alsoused to test the nullhypothesisthat mean

differences between effects ofbeta and gammainterferon,and between

single-agentandcombinationtreatment,were zero.Thispaired analy-sisusedonlypatients forwhom pre- and posttreatmentmeasurements wereavailable for bothinterferons/combination inquestion.The

com-parisons were madewithin eachdoselevel andforbothdose levels combined.

Two-samplet tests were used to test the null hypothesis that the

meandifferencebetween treatment effects at the twodifferent dose

levels were zeroforeachinterferonused alone andforthe combina-tion.Wealsoassessedinfluence of injectionsequenceonclinicaland

biological endpoints;nosignificant differencesoccurred.Although ab-senceof suchanassociation doesnotprove lackofinteraction,wefeel

theseeffects,if any, wererelatively minor.

Analysis ofall biological response parameters was based on log

transformeddata. Thisisequivalent toanalysis ofalog stimulation index, where stimulation index is definedastheratio ofthe posttreat-ment topretreatmentvalues.For purposesof presentation intext and

figures, mean treatment effects have been reexpressed as percent

changesorfold increases.

The pre- and posttreatment meansfromthe raw data areindicated inthefigure legends. Because thefiguresarebaseduponreexpression of log transformed data,these may not correspond exactly.

Results

IFN-induced

proteins

HLA-DRexpressiononmonocytes. 24 hafterthe administra-tionof

IFN-fser,

IFN-,y,orthe

combination,

therewas a

signif-icant increase in the mean percentage of cells positive for

HLA-DR (Fig. 1A). Althoughin afewpatients the percentage of monocytespositivefor HLA-DRexpressiondecreased after

IFNs (Fig. 2 A), overall, themeanpercentage of cellspositive

for HLA-DR increased from 65.8 to 81.9% 24 h after

treat-mentwithIFN-flser,from 71.5%to78.7% 24 h aftertreatment

with IFN-,y, and from 69.9 to 81.8% 24 hours after IFN-1ser/ IFN-'ytreatment(P=0.011, 0.019,and <0.001, respectively). The percentage of cellspositiveforHLA-DRincreasedin15of

19patients receiving

IFN-,fser,

16 of 19receivingIFN-'y, and 18 of 21 receiving the combination. The level of antigen

ex-pression (MFI) for cells positive for HLA-DR in the three

treatmentgroups increased 37, 50, and 47%, respectively (P

=0.003, <0.001, and <0.001, respectively). The MFI for

cells positive for HLA-DR increased in 15 of 19patients

re-ceiving IFN-iser, 17 of 19 receiving IFN-'y, and 21 of21

re-ceiving the combination. Although these changes were not

correlated with monocyte counts, wehavepreviouslyshown that fluctuations in absolute numbers of monocytes did not

correlate withchangesinclassIIantigen expressionin patients treatedwith IFN-/3or

IFN-'y

(15).

The datawereanalyzedtodeterminewhether the combi-nationof

IFN-fser/IFN-i

significantly increasedHLA-DR

ex-pression more than IFN-fser or

IFN-'y.

Asimilar fraction of patientshadanincreasein positivemonocytesafterall

treat-ments (Fig. 2, A and B). The combination of

_IFN-fser

and

IFN-'y

was notsignificantly different fromeither IFN alone in itseffecton the MFIorpercentageofcellspositivefor

HLA-DR(P 2 0.25). Similarly, there was no significant difference between

_IFNis~er

and

IFN-'y

in their effect on the MFI or

percentage of cellspositive forHLA-DR (P. 0.08). No

signifi-cant(P>0.1)dose responserelationshipwasobservedfor any ofthethreetreatmentgroups(Fig. 2,AandB).

After4wkoftreatmentwith

IFN-,fser/IFN-'y,

the percent-ageofcellspositive forHLA-DR and the MFI ofthose cells remained significantly higher than pretreatment values (P

= 0.002 and0.04, respectively) (Fig. 1 C).

HLA-DQ expression on monocytes. Significant increases

werealso seen in the mean percentage ofmonocytespositive for HLA-DQ after

IFN-Aser, IFN--y,

and thecombination (Fig.

1 B) (P=0.02,<0.001, 0.001,respectively). The percentage of

A

B

C

Ui

₇₀

z*

) 50 *

Z 40

'U

z

oc 10 'U

2 0 -20 -30 -40

CELLS (%+) MR CELLS (%+) MR HLA-DR HLA-DO

Figure 1.Effects ofcombineddoselevelsof

_{IFN-fl., (fl),}

IFN-y(y),and

IFN-#f3,/IFN-y

(f/,y)on(A)HLA-DRand(B)HLA-DQexpressionon monocytes.(Percentage of cellspositive andMF.) Resultsaregivenasthemeanpercentchange (reexpressed fromanalysis of log transformed data) ofvaluesmeasured:24 h posttreatmentcompared withvalues measuredbeforetreatment; and(C) after4wkof maintenance therapy withIFN-Iser/IFN-y. Themeanpercentageofcellspositive forHLA-DRandDQ increased from65.8 to81.9%andfrom9.0 to 17.3% 24 h

after

_IFN-flser,

from 71.5 to 79.4% andfrom 10.1to 21.7%after IFN-y,andfrom69.9to81.8% andfrom9.5to 16.6%afterthecombination, respectively. TheMF1 on monocytespositive forHLA-DRandDQ increased from 43.0to56.6 andfrom 17.9to 18.524 hafter

IFN-f,B,,

from

44.5 to67.9andfrom 18.2to 19.3after IFN-,y, and 46.8to68.6andfrom 17.4to15.6after thecombination,respectively.After4wkof

main-tenancetherapy with

IFN-f3,,,

andIFN--y, themeanpercentageof cellspositiveforHLA-DRandDQ increased from 64.8to85.7% and 9.4to

15.9%,respectively. TheMFIforHLA-DRincreased from 47.8to70.7,andHLA-DQdecreasedfrom 19.4to12.0. Asterisks indicate

signifi-cantdifferences compared with baseline withP<0.05. Errorbars=SE.

monocytespositive for HLA-DQ increasedin 14of19patients receiving

_IFN-#fer,

15of19patients receiving IFN-y,and 16of 21 receivingthecombination.IncontrasttoHLA-DR expres-sion, nosignificant changes in the MFIofthosecellspositive for HLA-DQ were observed after treatment with

_IFN-Oser,

IFN-'y, and

_{IFN-fser/IFN-y}

(P=0.68, 0.86,and0.09, respec-tively) (Fig. 1 B). TheMFI of cellspositive for HLA-DQ in-creased in7of19patients receiving

IFN-3er,

7of19receiving

IFN-'y,

and 17of21 receivingthecombination. Similartothe results seen with HLA-DR, thecombination ofIFN-iser and IFN--ywas notsignificantly different from eitherIFNalone in its effecton the MFI orpercentageofcellspositive for HLA-DQ(P 2 0.25).

IFN-#,ser

didnotdiffersignificantly from IFN-y in itsabilitytoenhance the numberofmonocytesexpressing HLA-DQ(P=0.93).

The data was also analyzed to determine if there was a

significant dose response in any ofthe three IFN treatment groups.

IFN/3ser

at30X 106 unitswassignificantlymore effec-tiveinenhancingthe percentageof cells positive for HLA-DQ than 3 X 106 units (P = _0.04). The higherdose level of

IFN-,ser/IFN-'y

decreased the MFIofcells positive for HLA-DQ toagreater extentthan the lowerdose level (P=_{0.03). No} significant dose-response relationshipwasobservedfor

IFN-'y

(P.

0.18).

After4wkoftreatmentwith either ofthetwodose levels of

IFN-fl.,/IFN-y,

theMFR of DQ positivecells wassignificantly lower than pretreatment values (P= 0.007)(Fig. 1). The per-centageofcellspositive for HLA-DQ,however,didnotremain elevated(P=0.1 1).

(2

microglobulin expression

onmonocytes.The

expression

Of(2 microglobulinonmonocyteswasmeasured and used to

assessofHLAClassIexpressiononspecificcell types. There was nosignificantchange in the percentageof cells positive for

(2microglobulin24 h after

IFN-I3,,, IFN-y,

or

_{IFN-#,sr/IFN-'y}

(P < 0.43 for all threetreatment groups) (data not shown). Most monocytes

(98%,

datanot shown) however, were

posi-tive for this antigen before treatment making it difficult to

detectfurther increasesin the numberofpositivecells. How-ever, therewas a

significant

increasein the MFR (P= 0.011, 0.010,and0.001,

respectively)

of positivecells(Fig. 3A).The MFI of monocytespositive for

(2 microglobulin

increasedin 16of 20patients receiving

IFN-(ser,

16of20patients receiving IFN--y,and 16of21patientsreceivingthecombination.There were no significant differencesin the level of MR

enhance-mentbetween any ofthe IFN treatments (P. 0.32), or

be-tween dose levels(P > 0.22). After4 wk oftreatment with either ofthetwo

IFN-(s~er/IFN-,i

doselevels,theincreasein

(32

microglobulin on monocytes compared to pretreatment values, wasmaintained,althoughitwas nolongerstatistically significant (P. 0.08).

:1

.

PRE POST PRE POST PRE POST

IFNf

IFNy

_IFNf/y

PRE POST PRE POST PRE POST

IFNf

IFNy

_IFN#/y

Figure2.Percentage ofmonocytespositive forA) HLA-DR and B)

HLA-DQbefore(pre) and 24 hafter(post)IFN. Each linerepresents asample fromanindividual patient(-) lower dose;(----) lOX

higher dose.

Serum (2microglobulin. Significant elevations inserum(2

microglobulin were observed 24 h aftertreatment with all

threeIFN(s) (P<0.001) (Fig. 3 B). All 21 patients receiving

IFN-flser

and the combination hadanincrease intheirserum(2

microglobulin; 18 of19 patients receiving

IFN-y

had an in-crease.

IFN-fler

increased serum (2 microglobulin 1.29-fold, which wassignificantlyless thanthe 1.5-foldincreaseobserved after IFN--y, and the 1.63-fold increase observed after

IFN-#,ce/IFN-y

(P=0.05 and0.001, respectively).The differ-ence between

IFN-'y

and the combination was not statistically significant (P = _{0.08). Elevations} in serum 2

microglobulin

weremaintained after4wkoftreatment ateitherdose level(P

< 0.001) (Fig. 3 B). A

significant

dose response

relationship

was observed with either IFN-y or

IFN-,fser/IFN-y,

but not

IFN-fser

(P=0.04,0.01,and0.38,

respectively).

2-5A

synthetase

activity.

IFN-flser

increased serum 2-5A synthetase bya meanof 3.2-foldandin 17of 20

patients (P

<0.001) (Fig. 4). Similarly,

IFN-fl.,r/IFN-'y

increased 2-5A synthetase activity 2.3 times (P= _{0.004) and} in 15 of17 pa-tients assessed. IFN-y had no

significant

effect on 2-5A syn-thetaseactivity, increasing it in7of18patients (P=0.65).The differencein 2-5A synthetase enhancementbetween

IFN-Aser

and the combination was not statistically significant (P =

0.23), although both enhanced 2-5A

synthetase

more than IFN-y (P < 0.02). No significant dose-response

relationship

was observed foranyofthethree treatment arms(P. 0.38) (Fig.4A).The combination

_{IFN-i3,,r/IFN-y}

maintained2-SA synthetase activity when administered for 4 wk(P< 0.001) (Fig.4B).

Indoleamine 2,3 dioxygenase activity as measured by re-ductioninserum

tryptophan levels.

At the lowerdose,the com-bination of

IFN-fler/IFN-,

produced significant decreases (P ' .01) in serum tryptophanlevelsat 6 and 24 h

(Fig.

5 A). Except for

_IFN-#(3r

at 6 h, the individual interferonsat dose level A did not produce significant changes in tryptophan levels. At thehigher dose,

IFN-'y

andthecombination caused

significant

(P<0.04)decreases intryptophanatalltimes stud-ied(Fig. 5 B),decreasing itin 11 of11 patients,and 10of11 patients, respectively.

There was no significant difference between effects of

IFN-'y

andthecombinationat 24 h(P=0.5 1) although both ofthese treatmentsproducedgreater decreases in serum tryp-tophanthan

_IFN-fler

(P < 0.03). A

significant

dose-response effectwas seenforboth IFN-y andthecombinationat 24 h, but notfor

IFN-B3er

(P=_0.004,

0.005,

and0.94,respectively). Indoleamine 2,3 dioxygenase activity as

measured by

uri-nary

kynurenine

levels. Inparallel with the decreaseinserum tryptophan levels observed after IFN--y or

_{IFN-fle,/IFN-,y}

there was a

reciprocal

increase in urinary excretion of kyn-urenineat 24 hatbothdose levels (P<0.003)(Fig. 5, C and D). IFN-y increased urinary kynurenine levels in 17 of20 patients, and the combination increased levels in 18 of20 patients.Asignificant increasewasobservedfor

IFN-(ser

at the lowerdose levels(six of nine patients)(P= _{0.048) but not the} higher (five ofninepatients)(P=0.38). Aswithserum trypto-phan levels, the changes in urinary kynurenine induced by

IFN-'y

and

IFN-(3er/IFN-y

were not significantly different from one another (P = _0.24).

_{IFN-fler/IFN-'y,}

_however,

in-duced significantly higher urinary kynurenine levels at 24 h

than did

IFN-fler

(P = _0.004). A dose response relationship wasobserved for IFN-y andthecombination treatment, but notfor

IFN-(3,s.

(P=_{0.056,0.03,and0.67,respectively).}

Effector

cell

cytotoxicity

NKand ADCC activity. IFN-,y decreased NK cell activity by 32%when measured 24 haftertreatment (P=_{0.026) (Fig. 6} PhaseIBTrialInterferon-,ser, Interferon-y, and

Interferon-,ls,

Interferon-y 1215 10

U

0

o 5

F-40

hI

l)

U

hLVI

10

U)

hi

U)

0

a.

I--z hi

C)

I:

hi a.

10

0

'I

0

0

D.

-I.

-I.

-I -:1

I

I

9 ...

A

50

z

z

uj

a

LU

z

w

0

40 I

30

L

3

I

20 I

I

10

L

0

24 H

LU

(I)

LU

z

0

IL

2

1

4 WK

B

*

24

H

4

WK

Figure 3. Effects of combined dose levels of

IFN-#ser

(fl),IFN-y(y), and

_{IFN-(,,,/IFN-,y}

(#/y)onf2microglobulin expressionon(A)

mono-cytesand(B)serumlevels. Resultsaregivenas meanpercentchange (reexpressed fromanalysisoflog-transformed data)of values measured pretreatmentcomparedtothose measured24 hpostinjection,andafter4wkof maintenance treatment. TheMF1 onmonocytesincreased from 113.2to140.324 hafter

_IFN-#,er,

from 120.9to140.0after IFN-y, and from 129.6 to 155.5 after the combination. After4wkof mainte-nancewith thecombination, theMFRincreasedfrom 123.4to144.9.Mean serum(2 microglobulinincreased from 2.7to3.6mg/liter24h

after

_IFN-#,,r,

from 2.7to3.9mg/literafterIFN-'y,andfrom 2.7to4.4mg/literafter the combination. After4 wkof

_{IFN-#3,,/IFN--y,}

mean serum02 microglobulinincreased from 2.8to5.5mg/liter.Asterisksindicatesignificantdifferences compared with baseline (P<0.05).Error

bars=SE.

A). Although

IFN-Bser

and

_{IFN-flser/IFN-y}

alsodecreased NK activity compared with pretreatment values, these decreases were not statistically significant (P = 0.66 and 0.29, respec-tively). IFN-y did not depress NK cell activity significantly morethan

_IFN-Owr

orthecombination (P 2 0.21). ADCC was decreased by both

_IFN-flser

(13of20patients)(P=0.032) and IFN-y(14of19patients)(P=0.004), but not

IFN-Bser/IFN-i

(10 of 18 patients) (P = _{0.65) (Fig.} 6 B). Similar to results observed with NK cell activity,thecombination did not

de-crease ADCC significantly less than either IFN alone (P 20.36). Statistical analysis showed no significant dose re-sponserelationshipwasobserved for either NK cellactivityor

ADCCactivity by

IFN-0fler,

IFN-,y, orthecombination (P 2 0.09).

4wkof maintenance

IFN-fser/IFN-I

didnotenhance NK cellactivity (P=0.75)(Fig. 6 A). ADCC activity, however,was

enhanced after4wkof thehigherdose level (P=0.021)but

notthelower (P=0.54) (datanotshown).When datafromthe

A B Figure 4. Effects of low dose or higher dose 6 _{8* _}

(lOX)

of

IFN-#,..e

(al),

IFN--y

(y),

and

IFN-fle/IFN-'y (f/fy)

on2-5Asynthetase

7

*g

y wactivity (A) 24 h postinjection, and (B)

w after4wkof maintenancetreatment. U)

6

RX/Y

7 fl Mean 2-5A synthetase activity increased

UJ *m from11.3pmol/105cellsperhto20.1

5 T

mpmol/

I0O cellsperh 24 hafter the lower

z * doseofIFN-fl,,, from7.7to8.2pmol/105

4 4 * cellsperhafter the lower dose ofIFN-y,

O _{* and from 5.3 to14.1}

_pmol/105

_cells

_per

_h

0 3 - 21 - _ * + T after the lowerdoseof thecombination.24

Umm

hafterthe

high

doseof

_IFN-fl,0,,

IFN-y,

_

00_;:y

t/2y/

and thecombination,meanlevelschanged

1

______1_ from 10.0to30.1,from 16.3to 17.0,and

from 8.0to22.7pmol/105 cells per h, respectively.Mean2-5A synthetase

activ-DOSE (1x) DOSE

(l0x)

DOSE (lx) DOSE (10X) ityincreased from 4.7 to 27.9

_pmol/I05

cells perhafter4wkof thelowerdose of thecombination,and from 13.7to25.9pmol/105cells perhafterthehigherdose.Asterisks indicatesignificantdifferencescomparedwith baseline (P<0.05).Errorbars=SE.

1216 Schiller, Storer,Paulnock,Brown,Datta, Witt,and Borden

No

EM

y

SERUM

TRYPTOPHAN

z

4c C.)

I--z LU

LU

CL

z

44

LU

A B

4 H 6 H 24 H 4 H 6 H 24 H

KYNURENINE

LU 900

0

Z ₈₀₀ X ₇₀₀ 600

z LU 500 M 400

LU

aL 300

S 200

a 100

0

A

'U

C

z

z

U

I-z oc

0i

'U

c D

B

24H 4WK 24H 4WK

Figure6.Effects of combineddose levelsof

IFN-B(er

(,B), IFN-y (y), and

IFN-#

../IFN-y

(#/y)

on(A)NKcellactivityand(B) ADCC ac-tivity.Results aregivenas meanpercentchange (reexpressed from

analysisoflog-transformeddata) of valuesmeasuredpretreatment compared withthose measured 24 hpostinjection,andafter4wkof maintenancetreatment. MeanNKactivity changed from 194.5lytic units/107cellsto200.9lyticunits/107cells24 hafter

IFN-fler,

from 162.4to104.1 after

IFN-'y,

and from 183.0to193.3lyticunits/107

cellsafter thecombination.Mean ADCCactivity changed from 459.3lyticunits/107cellsto344.0lyticunits/107cells24 hafter ad-ministration of

IFN-#,r,

from 341.1 to225.6after IFN--y, and from 393.1 to396.7after thecombination. After4wkof maintenance

treatmentwiththecombination, mean NKactivity and ADCC activ-ity changed from 162.4 to 195.6lyticunits/107cells, and 345.1 to

655.8lytic

units/107

cells,respectively.Columns with asterisks show

significant differenceswithP <0.05.Errorbars=_SE.

Figure5. Effects of lower (1 X) or higher(lOX) dose of

IFN-fl.

(#),IFN-'y(y),and

IFN-,8fl,/IFN-y

(,/'y) onindoleamine 2,3 dioxygenase activity: (A and B) serum tryptophan levels and (C and D) urinary kyn-urenine levels.Results aregivenas meanpercent change (reexpressedfrom analysis of log-transformed data) of values measured pretreatment comparedto

thosemeasured 4, 6, and 24 hpostinjection (A and B), and 24 and 48 h posttreatment (C and D).Mean

serumtrypotophan levels changedfrom 1.2 to 1.1 mg/dl24 hafter the lower dose of

IFN-fle,,

from 1.1

to1.1 mg/dl afterthe lower doseofIFN-'y,andfrom 1.2 to 1.0mg/dl after the lower dose of the combina-tion. AfterthehigherdoseofIFN-#,B,,IFN--y,and thecombination, mean serum tryptophan levels

de-creasedfrom1.1 to 1.0, from 1.1 to0.6, andfrom 1.1

to0.5mg/dl, respectively.Meanurinary kynurenine concentrations increased from27.1 to 40.3gmol/d

24 hafter the lower concentration of

_IFN-#,.,,

from 26.7to36.9 after the lower concentration of IFN--y,

andfrom19.1to 37.1 umol/d after the lower concen-tration of the combination.24 hafterthe administra-tion of thehigher concentrations of

IFN-,6,e&,

IFN-'y,

and thecombination,meanurinary kynurenine

con-centrations increased from51.0 to48.9, from 29.5to 153.1,andfrom 57.0to246.4Mmol/d,respectively. Asterisksindicate significant differences compared with baseline with P<0.05. Error bars=SE.

two dose levelswerecombined(Fig.6B), ADCCactivitywas increased 63%(P=0.042).

Serum neopterin levels.

Significant

increases in serum

neopterinlevels were observed after all threeIFN treatments at

both dose levels (Fig. 7). Allthree IFN treatments increased

serum neopterinin allpatientsstudied. IFN-Iser/IFN--y

treat-mentincreasedserum neopterin levels bya mean of3.0-fold, anaugmentationthat wassignificantly higherthan the2.2-fold

increase induced by

IFN-'y,

but only marginally

significantly

higher than the 2.5-fold increase induced by IFN-flser (P =0.005,and0.06, respectively). Nodose response effectwas

noted inanyofthe treatment groups(P20.34). Serum neo-pterin levels were not obtained after 4 wk of maintenance treatmentwith

_{IFN-#,Be/IFN--y.}

Clinical tolerance

Single

injections.

Subjective

side effects were dose relatedand qualitatively similar to those typically associated with IFN

(Table III). The most common side effects of moderate or greaterseverityatthe lower dose werefatigueand headache. Symptoms ofmoderate or greaterseveritywere morefrequent atthe

higher

dose(Table III). Incidence of

anorexia,

nausea, and vomitingwas

slightly

greaterafterthecombination, and incidence offatigue was slightly higher after

IFN-'y.

Rigors requiringparenteralmeperidine(grade 2) were not observed at thelower dose but occurred in threepatientsatthehigherdose (onepatientonIFN-fserandtwo on

IF`N-fiser/IFN-y).

Vital signs including temperature, pulse, respiration and

4

w (I) w

C)

z

0

-j

10

U-3

2

2B/y

T

T

DOSE (I x)

1T

DOSE (1Ox)

Figure7.Effects of lower (I X) or higher (IOX) dose on serum neo-pterin. Results are given as mean percent change (reexpressed from analysis of log-transformed data) of values measured pretreatment compared with those measured 24 h postinjection, and after 4 wk of maintenance treatment.Meanserumneopterin levels increased from 9.6nmol/literto21.3 nmol/liter 24 h after the lower concentrations of

_IFN-#er,

from 8.2to 15.5nmol/liter after the lower concentration ofIFN-y,andfrom 9.0to25.3 nmol/liter after the lower concentra-tions of the combination. 24 h after administration of the higher concentrations of

_IFN-#,,r,

IFN-,y, and the combination, mean

serumneopterin levels increased from 13.3 to 31.9nmol/liter,from 14.4 to33.2 nmol/liter, and from 12.8 to 35.9 nmol/liter, respec-tively. Columns with asterisks show significant differences with P

<0.05.Errorbars=SE.

bloodpressure were recorded pretreatment and 1, 2, 4, and 6 h after administration ofIFN(s). There were no temperatures

>

40°C.

Mean maximum temperatures were equivalent for bothdoselevelsof

IFN-,l.3r

(37.9°C). For the higher dose level of

IFN-y,

mean temperature was higher (38.2°C, range 36.7-38.4)than at the lower(37.5°C,range37.0-39.7°C). The mean maximumtemperature at the higherdose was slightly higher for the combination (38.4°C) than either IFN alone. Acetaminophen (650mgp.o.) wasadministered30min before treatmentandevery 4 h foratotaldoseof 2,600mgwithall treatments atthe

higher

doselevel.

After4wkof maintenancetherapywith the combination at the low dose,

dose-limiting

toxicity was observed in two patients. One patient required dose reduction for grade 3 weightloss (9.4%of baseline)and anotherforgrade 3 fatigue

anda

performance

status decrease of2 2 levels. After 4 wk of therapy with thecombination, the incidence of moderate or

severefatigue, headache,

myalgias,

and anorexia decreased (TableIII). Fourofeight patientsatthehighdose level hada

performance

status decrease of2 2

levels,

whichwas attrib-utedto concomitantdiseaseprogression in three of these pa-tients. One

patient

withseverefatigue,nausea, andvomiting, andgrade4weightloss(10.3%ofbaseline) requireddose mod-ification.

Transaminase elevation (210

IU/liter

serum

glutamic

oxalo acetic acid transferase

(SGOT))

wasdoselimitinginone

patient who wasreceiving 1 X 106 IU of

IFN-l8ser

with 70

jg

IFN-'y

afteradosereduction forfatigue.

Laboratory

parameters

Single

injections. Hematologic parameters

were

measured

be-fore and 24 h after single weekly injections. Significant de-creasesinneutrophils, platelets, hemoglobin, and whiteblood countswereobserved (TableIV). Although these changes were statistically significant,none wereclinically significant.No sig-nificantalterationsinrenalorhepatic functionwere observed afteranyof the firstinjections.

After4 wkof maintenancetherapywiththecombination

IFN-,ser/IFN-I,

there were significantdecreasesin hemoglo-bin, hematocrit, neutrophils, total whitebloodcellcount, and significant increases in SGOT and lactate dehydrogenase (LDH) when compared with pretreatment values (Table V). The hepatic function changeswere not clinically

significant,

exceptinonepatient who developedanelevated SGOT of 210 IU/liter while receiving 1 X

106

IU

IFN-fscr

and 70

zg

IFN-y afteradosereduction forfatigue.The meandurationon main-tenancetherapywas 15.5 wkforthe doselevelA and 5.9wk forthe dose level B (range: 4-31 wk and 1-13

wk,

respec-tively).No

objective

responses were seenateitherdose level. IFN

pharmacokinetics

SerumIFN-yandtotal serum IFNlevelswereassessed follow-ingtheadministrationof

IFN-#ser,

IFN-'y,

or

_{IFN-#,,/IFN-,y.}

Because thefirst samplewascollected 15 minafterall IFNhad beeninjected,true"peak"values were notobtained. However, in general, serum IFNlevelsweredose related. Administration of lowdoses of IFN--y

(200 ,g) produced

maximum IFN-y titersof 35 U/ml, whereas administration of 2,000

tig

of IFN-,y producedmeanIFN-yantiviral titers of 150U/ml. Similarly,

TableIII.DifferencesinSideEffectsbetweenIFNTypes

Numberofpatientsexperiencingmoderateor severesymptoms

Low dose High dose

IFN-ft IFN-'y IFN-ft+ y Day 42 IFN-P IFN-'y IFN-ft+y Day 42*

Sideeffect n=10 n=10 n=10 n=₁₀ n=11 n=11 n=11 n=8

Anorexia 0 0 0 2 2 4 6 2

Fatigue ' 1 1 2 2 5 6 4 3

Chills 0 0 0 0 1 0 2 0

Nausea/vomiting 0 0 0 1 2 2 4 3

Headache 2 2 3 0 5 6 4 2

Myalgias 0 0 0 0 4 1 3 1

*_{Threepatientswithdrew beforeday 42.}

1218 Schiller, Storer, Paulnock,Brown,Datta, Witt,and Borden

TableIV.Hematological Changes24hafterthe AdministrationofIFN-,se,, IFN-iy, IFN-4ser/IFN-i

Lowerdose Higher dose

IFN-#..e IFN-y IFN-,8,e,/IFN--y IFN-#..r IFN-y IFN-#,e,,/IFN-,y

Hematocrit

(ml/dl) -0.1 0.5 -0.7 -1.5* -1.8* -0.5 Whiteblood count

(XJ03/mm3) -1 .8* -1.3 -2.6* -2.8* -3.1* -3.0*

Neutrophils

(XJ03/mm3) -1.6* -1.5* -3.0* -1.9* -2.1* -2.4*

Lymphocytes

(X103/mm3) -0.17 -0.15 -0.11 -0.44* -0.29* -0.57*

Platelets

(X103/mm3) -12.3 2.2 -24.8 -7.6 -48.1 -43.3*

*P<0.05bypairedttest, posttreatment valuescompared withpretreatment values.

3 X 106 U ofIFN-BfserorIFN-iser plus

IFN-'y

produced total groups was distributed equally among patients in each dose

IFNtiters of168-254 U/ml, whereas the 30X 106Udose level level to control for cumulative side effects. Based upon our

gavelevels of994-1,342 U/ml. These levels fell rapidly over prior studies (17, 20), a l-wkwashout period between

treat-thenext6 h to values that were notsignificantly different from ments allowedfor reversalofbiologicalandclinicaleffects.

baseline. Consideringthe large degree ofvariability associated Although there were slight differences in constitutional with serum IFN levels, we were not able to detect a significant symptomsafter the administration of a single dose of

_IFN-#,

alteration in the pharmacokinetics of eithertotal serum IFN IFN-,y, or thecombination, these were not clinically signifi-levels or

IFN-'y

levels after the administration of combined cant. Althoughconstitutional symptoms orclinicalside effects

IFN-flser/IFN-y

when compared to either

IFN-'y

or IFN-fser were more pronounced at the higher dose levels ofIFN-fl, alone. IFN-,y, orthecombination, these sideeffects were tolerable. Fourweeksof maintenancetherapy with the combination of Discussion IFN-fl and IFN-y wastolerated at both dose levels, although moreside effectsoccurred at thehigherdose.Withthe excep-This study, thefirsttodirectlycomparebiologicalandclinical tion of a mildincrease in the SGOT at the higher dose, none of effects ofatypeI(IFN-aor

IFN-fl)

and type II(IFN-y) within the changes in laboratory parametersafter 4 wkofIFN-4ser/ thesamepatient, providedtheuniqueopportunity todirectly

IFN-'y

wereofclinical importance. Thus, no notableclinical compare the side effects associated with two different IFN differences in side effects or toxicities between

IFN-0ser

and types, as well as to determine any qualitatively or quantita- IFN-,y were observed. Otherinvestigators have reported that tively differences associated withthecombination of

IFN-3ser

themaximallytolerated doseofan IFN-a andIFN-,y combi-and IFN--y. The orderof administration ofthe three treatment nation was well below the maximally tolerated doseof either

TableV.Hematologic,Hepatic,and RenalToxicitiesafter1 moofMaintenance

_{IFN-lser/IFN-y}

Lowerdose Higher dose

Mean values Meanvalues

After1 mo AfterI mo

On study IFN-~~,,,/IFN-y * On study IFN-m,,/IFN-y

Hemoglobin

(gm/dl) 13.0 12.0 0.036 12.1 11.5 0.05

Hematocrit

(ml/dl) 38.8 36.0 0.032 35.2 33.6 0.10

White bloodcount

(X103/mm3)

8.2 5.9 <0.001 7.1 6.6 0.74

Platelets

(X103/mm3) 375 370 0.90 385 375 0.88

Creatinine

(mg/dl) 1.1 1.1 0.50 1.0 1.1 0.17

SGOT

(lU/liter)

35.7 43.8 0.08 30.0 82.0 0.026

*

By

paired

t_test,Day42 valuescomparedwithbaselinevalues.

IFN alone;furthermore,patientshadpersistingfeversand pro-gressivefatigue with the daily intramuscular injections (11).In ourstudy, a higher dose level ofIFN-flrandIFN--y was tolera-blewhenadministered forprolonged periods of time, a finding thatconfirms ourpreviousobservations (21). The discrepan-cies between the combination ofIFN-a and

IFN-flser

with

IFN-'y

may result from the two different type I IFNs (IFN-alfa-2aand

IFN-fler),

bydifferencesin the two

IFN-'y

formu-lations whichwereused, or bydifferencesin route or schedule of administration.

Quantification of biologicalresponsemodificationby IFNs hasbeen usefultodefine differences in IFN types, the influ-enceof serumlevels,duration of response,andmodificationof effectorcellfunction invivo(2, 14, 15, 17, 20). Assessment of host responses was undertaken 24 haftertheadministration of IFNs, a timeidentifiedasyieldingmodulation in prior studies invitro and in vivo (2, 14, 15, 17, 20). While a kinetic analysis would also be ofinterest, sample volumes from patients are limited;wethuschosetofocusstudydesignon theinfluenceof dose. The biological parameters chosen for study were ones that are known tobe modulated byIFNs and may be media-torsof antitumoractivity.

Asynergisticenhancementin biologicalresponses was not observed. With the exception of tryptophan metabolism, the differences were not even additive. The combination of

IFN43ser/IFN-y

was more effective than

IFN-fl

in enhancing tryptophan metabolism and serum 2 microglobulin, and moreeffective than IFN-,y inenhancing the expression of 2-5A synthetase and serumneopterin levels.These results are consistent with previous reports that the

optimal

biological

dose was not related tothe total doseofIFN (15, 17), even when the IFNadministered consisted oftwoIFN types.

Onebiologicalresponse,whichwasenhanced by

IFN-B&,r/

IFN-,y inadose-dependentbut not

synergistic

fashionwasthe inductionof tryptophan metabolism. Indoleamine

2,3-dioxy-genaseisanenzyme present in the

lung

andintestinalmucosa

that degrades tryptophan and related indoles to

kynurenine

and othermetabolites (23),andhas beenfoundtobeinduced by viral infections, bacterial

lipopolysaccharides,

and IFN (23). We have observed tryptophan

degradation

induced by IFN in vitroand in vivo (24, 25). Addition ofexcess trypto-phantoculturemediahas been able to reverse the

antiprolif-erativeeffect of

IFN--y

in several celllines (26,

27).

Neopterin, excretedby human monocytes in response to IFN-y,has beenusedas anassayformonocyteactivation (28). In this study, the combination of

IFN-#,,r/IFN-y-enhanced

neopterin levelsoverIFN-y alone. Furthermore,theincrease inneopterinlevelsproduced bythe

combination,

when

com-paredto theincrease in

neopterin produced

by

IFN-fl.e

bor-deredonstatistical

significance

(P=

_0.057).

We havepreviouslyshown that

IFN--y

and

IFN-f

enhances HLA-DR andHLA-DQ expression on monocytes in cancer patients

(15,

29); the current study confirmsthose observa-tions, extends them to the combination of

_{IFN-#,Br/IFN-y,}

and determined that IFN-y was no more effective than

IFN-4ser,

Thislatter result differed from studies oftumorcell lines in vitro (30, 31). In contrast, IFN-,B hasbeen shownto

have anantagonistic effectonthe IFN-yinduced

expression

of

Ta

antigen inmurine macrophages(32). We foundno poten-tiation ofHLAclassI orII

expression by

thecombination of

IFN-flser/IFN-

overeither

_IFN-f3ser

orIFN--y.

We observed no enhancement over baseline in NK cell activity with

IFN-flBr/IFN-y.

These results differ fromthose observed in immune cells in vitro, in which lymphocyte NK cell activityby

IFN-y

waspotentiated by IFN-a or

_IFN-,ser

(33). The lack of NK cell enhancement by

IFN-flser

and the decrease in NK cell activity by

IFN-y

also differ from the potentiation of NKcell activity weand others have consis-tently observed with IFN-a and IFN-fl (17, 34, 35). Other investigators, however, have reported the absence of signifi-cant NK cell activity and ADCC activityin patients treated with 4-h infusions of

IFN-'y

(31), although thishas not been confirmed byother groupsusingadifferent

IFN-y

preparation givenintramuscularly (36). Inthis study, ADCCactivitywas enhancedafter 4 wk ofthecombination treatment. To deter-mine whetherthisvariability in responseresults from move-mentofnon-MHC-restricted effector cells from circulationto tumorsites, differences in disease stage or type, or active sup-pressionbythe IFNs administered, would require additional quantitative studies.

Insummary,whencomparedwith either

IFN-#l-r

or

IFN-'y

alone,thecombinationof

IFN-fler/IFN-'y

didresult ina

signif-icantlygreaterenhancement ofserum

12

microglobulin com-pared with that induced by

IFN-flser,

a

significantly

greater increase in urinary kynurenine levels compared to

IFN-flser,

and a significantly greater increase in neopterin levels and 2-5A synthetase activity compared with IFN-y. Since these changes were not synergistic,we conclude thatthe combina-tion of

_{IFN-#l.,/IFN-y}

atthisdose and scheduledidnot po-tentiatehost responses more thaneitherIFN alone.Consistent with thesefindings,dosesof

IFN-#1er

andIFN--y resulted in no significant addedhosttoxicity.

These results contrast withthesynergisticantiproliferative effects that have been observed invitro and in mousetumor

models (7-10). The role of

IFN-ler,

IFN-y,

and

_IFN-fl.r/

IFN-y in the treatment ofcancerhas yet to bedefined. As a basis forPhase ITtrials, chronic administrationof the combi-nation of

_IFN-#3,e,

and

IFN-'y

was tolerable. Whether direct effect on tumorcells orindirect effects onhost responses

ac-countsfor antitumor

activity

ofIFNsremainsanopen ques-tion.IfPhaseIT clinical trials using combinations ofIFNs show potentiated

therapeutic

activity,

resultsfrom this studywould support the

hypothesis

thatdirect

anti-proliferative

effects pre-dominate.

Acknowledgments

WewouldliketothankDiane Helgesen and YoshiFujimiyafor their

experttechnical assistance,KarenSielaff forhelpin thecoordination of thistrial,LizSchiferl forheraidindataanalysis,andMaryPankratz

forherassistance intypingof thismanuscript.

This workwassupported inpartby National Institutes of Health

and NationalCancer InstitutegrantNO1-CM47669 and Triton Bio-sciences. J. H. Schiller is arecipientofanAmerican CancerSociety

Clinical OncologyCareer DevelopmentAward. D. M.Paulnockisa

Shaw ScholaroftheMilwaukee Foundation.E.C. Bordenisan

Ameri-canCancerSociety Professor of Clinical Oncology.

References

1. Borden,E. C.,T. F.Hogan,and J.G. Voelkel. 1982. The

com-parative antiproliferativeactivityin vitroof naturalinterferonsaand13

fordiploidand transformed humancells. CancerRes.42:4948-4953. 2. Nathan, C. F., C.R. Horowitz,J.de laHarpe,S. Vadhan-Raj,

of recombinant interferon -y to cancer patientsenhances monocyte

secretion of hydrogen peroxide. Proc. Nati. Acad. Sci. USA. 82:8686-8690.

3. Sidman,C. L., J. D. Marshall, L.D.Shultz, P. W.Gray, and H. M.Johnson. 1984. Gamma-interferon isone of several directB

cell-maturing lymphokines.Nature(Lond.).309:801-804.

4.Guyre,P.M.,P. M.Morganelli, andR.Miller. 1983.

Recombi-nantimmuneinterferonincreases immunoglobulin GFc Receptors on

culturedhumanmononuclear phagocytes.J.Clin. Invest. 72:393-397.

5. Basham,T. Y., and T. C. Merigan. 1983. Recombinant

inter-feron--y increasesHLA-DR synthesis and expression. J. Immunol. 130:1492-1494.

6. Schiller, J.H., and E. C. Borden. 1988. Antiproliferative and clinicalantitumoreffects of interferons. InLeukolysins and Cancer.

J. H. Ransom and J.Ortaldo,J.,editors.Humana Press,Clifton,NJ.

273-301.

7.Schiller, J. H.,D.S.Groveman, S.M.Schmid,J. K. V.Willson,

K. B.Cummings, andE.C. Borden. 1986. Synergistic antiproliferative effects of human recombinanta54 or,l.e, with-y-interferononhuman celllines of varioushistogenesis. CancerRes.46:483-488.

8.Schiller,J.H.,J. K. V.Willson, G. Bittner,W. H.Wolberg,M. J.

Hawkins,and E. C. Borden. 1986. Antiproliferative effects of inter-feronsonhuman melanoma cells in the humantumorcolony-forming

assay.J.InterferonRes.6:615-625.

9.Czarniecki,C. W., C.W.Fennie,D. B.Powers, andD. A.Estell. 1984. Synergistic antiviral and antiproliferative activities of Esche-richia coli-derived human alpha, beta, andgamma interferons. J.

Virol. 49:490-496.

10.Schiller, J. H.,B. Storer,J. K. V.Willson, andE. C. Borden. 1987.Phase Itrial of combinationsof recombinant interferonsft. and y in patients withadvancedmalignancy. CancerTreat.Rep. 71:945-952.

11.Kurzrock, R.,M.G.Rosenblum,J. R.Quesada, S.A.Sherwin, L.M.Itri, andJ.U.Gutterman. 1986. PhaseIstudy ofacombination of recombinantinterferon-alpha and recombinant interferon-gamma incancerpatients. J. Clin. Oncol. 4:1677-1683.

12. van der Burg, M., M. Edelstein, L. Gerlis, C-M. Liang, M.

Hirschi, and A. Dawson. 1985. Recombinant interferon-y (immu-neron): results ofaphaseItrial inpatientswithcancer.J. Biol.

Re-sponseMod. 4:264-272.

13.Hawkins, M., S.Horning,M.Konrad, S.Anderson,K.Sielaff,

S. Rosno, J. Schiesel, T. Davis, D. DeMets, T. Merigan, and E. C. Borden. 1985.PhaseIevaluation ofasyntheticmutantofB-interferon.

CancerRes.45:5914-5920, 1985.

14.Merritt,J. A., E.C. Borden,and L. A. Ball.1985.Measurement

of2',5'-oligo-adenylate synthetase in patients receiving interferon-alpha.J.Interferon Res. 5:19 1-198.

15.Paulnock, D.M., K. A.Havlin,B.E.Storer,G.T.Spear,K. M.

Sielaff,andE.C. Borden. 1989. Inducedproteinsin human peripheral mononuclear cellsover arangeofclinicallytolerable doses of inter-ferongamma. J.Interferon Res.9:457-473, 1989.

16.Gonwa,T. A., J. P.Frost,and R. W. Karr. 1986.Allhuman monocyteshave thecapability ofexpressing HLA-DQ and HLA-DP

moleculesuponstimulation withinterferon-y.J.Immunol. 137:519-524.

17.Edwards,B.S.,M. J.Hawkins,and E.C. Borden. 1984.

Com-parativein vivoandin vitroactivation ofhuman naturalkillercells by tworecombinant a-interferons differingin antiviral activity. Cancer

Res.44:3135-3139.

18.Arend,R. A., J. E.Leklem,and R. R.Brown. 1970. Directand steamdistillation autoanalyzer methods forassayof diazotizable

aro-matic amine metabolites of tryptophan in urineandinserum.

Bio-chem. Med. 4:457-468.

19. Denckla,W. D., and H. K.Dewey. 1966. The determination of tryptophan inplasma, liver and urine. J. Lab. Clin. Med. 69:160-169.

20. Hawkins, M.J., E. C. Borden, J. A. Merritt, B. S. Edwards,

L. A.Ball, E.Grossbard, andK. J.Simon.Comparison of the biologic

effects of two recombinant human interferonsalpha (rAandrD)in humans.J.Clin. Oncol. 2(3):221-226.

21 and22.Deletedinpress.

23.Yasui,H., K.Takai,R.Yoshida, and 0. Hayaishi. 1986. Inter-feron enhances tryptophan metabolism by inducing pulmonary in-doleamine 2,3-dioxygenase: its possibleoccurrenceincancerpatients.

Proc.NaIL Acad. Sci. USA. 83:6622-6626.

24.Byrne,G. I.,L. K.Lehmann, J. G. Kirschbaum, E. C. Borden, C.M. Lee,and R. R. Brown. 1986. Induction of tryptophan degrada-tion in vitroandin vivo: a y-interferon-stimulatedactivity. J. Inter-feronRes.6:389-396.

25. Carlin,J. M., E. C. Borden,P. M.Sondel, and G. I.Byrne.

1987. Biological response modifier-induced indoleamine

2,3-dioxy-genaseactivity in humanperipheral blood mononuclear cell cultures.

J.Immunol. 139:2414-2418.

26. de laMaza,L.M., andE. M.Peterson. 1988.Dependence of

thein vitroanti-proliferativeactivity of recombinanthuman

'y-inter-feron on theconcentration of tryptophan inculture media. Cancer

Res.48:346-350.

27.Ozaki,Y., M. P.Edelstein,and D.S. Duch. 1988. Induction of indoleamine2,3-dioxygenase:amechanism of the antitumor activity of interferon-gamma. Proc.NatL.Acad. Sci. USA. 85:1242-1246.

28. Huber, C., J. R.Batchelor, D.Fuchs,A. Hansen, A.Lang, D.

Niederwieser, G. Reibnegger, P. Swetly, J. Troppmair, and H. Wachter. 1984.Immuneresponse-associated production of neopterin:

release from macrophages primarily under control of

interferon-gamma.J.Exp.Med. 160:310-316.

29. Spear, G. T., D. M. Paulnock, R. L. Jordan, D. M. Meltzer, J. A.Merritt,andE. C. Borden. 1987. Enhancementofmonocyte class I and IIhistocompatibility antigen expression inman by in vivo inter-feronfi.Clin.Exp.Immunol. 69:107-115.

30.Dolei,A., M. R.Capobianchi, andF.Ameglio. 1983.Human

interferon-gamma enhances the expression ofclass I andclassIImajor histocompatibility complex products in neoplastic cells more effec-tively than interferon-alpha and interferon-beta. Infect. Immun.

40:172-176.

31. Nickoloff, B. J., T. Y. Basham, T. C. Merigan, and V. B. Morhenn. 1985. Immunomodulatory andantiproliferative effect of recombinantalpha, beta, and gammainterferonsoncultured human

malignantsquamouscelllines, SCL-1, andSW-127 1.J. Invest.

Der-matol. 84:487-490.

32. Ling,P.D., M. K.Warren, andS.N. Vogel. 1985.Antagonist effect of

interferon-#

ontheinterferon-'y-inducedexpressionofIa

an-tigen in murine macrophages.J.Immunol. 135:1857-1863.

33. Weigent, D. A., M. P.Langford, W. R. Fleischmann, Jr., and

G. J. Stanton. 1983. Potentiation oflymphocyte natural killing by

mixtures ofalphaor betainterferon withrecombinantgamma

inter-feron.Infect.Immun.40:35-38.

34. Neefe,J.R., J. E. Sullivan, M.Ayoob, E. Phillips,and F. P.

Smith. 1985. Augmented immunity in cancerpatients treated with

a-interferon. CancerRes. 45:874-878.

35.Hawkins, M.J., S.E.Krown, E.C.Borden, M.Krimm,F. X. Real, B. S. Edwards, S. A. Anderson, S. Cunningham-Rundles, and H. F. Oettgen. 1984.American Cancer Society Phase I trial of naturally producedinterferonbeta. Cancer Res.44:5934-5938.

36. Kurzrock, R., J. Quesada, M. Talpaz, E. M. Hersh, J. M. Reuben, S. A.Sherwin, and J. U. Gutterman. 1986. Phase I study of

multipledoseintramuscularly administeredrecombinantgamma

in-terferon.J. Clin. Oncol. 4:1101-1109.