INT. J. RADIAT. BIOL ., 1991, VOL . 59, NO . 2, 433-445

Influence of the 100% w/v perfluorooctyl bromide (PFOB)emulsion dose on tumour radiosensitivity

C. THOMASt$, J. RIESS§ and M. GUICHARDt

t Laboratoire de Radiobiologie cellulaire (Unite INSERM 247), InstitutGustave Roussy, 94805 Villejuif Cedex, France§Unite de Chimie moleculaire associee au CNRS, Universite de Nice, ParcValrose, 06034 Nice Cedex, France

(Received 6 April 1990; revision received 2 July 1990 ; accepted 7 August 1990)

The radiosensitizing effect of a 100% w/v emulsion of a fluorocarbon, PFOB,which carries 4 times more oxygen than does Fluosol-DA 20% emulsion, wasstudied on two human tumour xenografts (HRT18 and HT29) and the murinetumour EMT6 . This effect was compared with that obtained with carbogenalone . The fluorocrit (amount of fluorocarbon in the blood) and haematocritremained unchanged from 7 to 65 min post-injection of the emulsion (8 ml/kg) .Tumour-bearing mice were pretreated with 100% w/v PFOB emulsion dosesranging from 2 to 15 ml/kg in the presence of carbogen for 30 min prior to andduring irradiation . The fluorocrit increased from 1 .5% to 9 . 5% as the dose of100% w/v PFOB emulsion increased from 2 to 15 ml/kg . The haematocritremained the same for all the fluorocarbon emulsion doses used. Tumourradiosensitization varied with the fluorocarbon emulsion dose . Clinically rele-vant doses (2-4ml/kg) of the 100% w/v PFOB emulsion plus carbogenproduced significantly more radiosensitization than carbogen alone, with sensi-tizing enhancement ratios of 1 . 4 for EMT6 and 1 .7 for HRT18 . The radiosensi-tivity of HRT18 cells was thus very close to that obtained with normallyoxygenated cells . For higher doses (8-15 ml/kg) the radiosensitizing effect of100% w/v PFOB emulsion plus carbogen becomes comparable to that ofcarbogen alone. These experiments show that clinically useful doses of 100%w/v PFOB plus carbogen produced tumour radiosensitization only at relativelylow fluorocrits. Thus the fluorocrit, and hence the fluorocarbon's oxygen-carrying capacity, is not the only factor involved in radiosensitizing tumour cellsby oxygen-carrying fluorocarbon emulsions .

1 . IntroductionFluorocarbon emulsions are promising new tools for overcoming the radioresis-

tance of hypoxic tumour cells . Many studies have shown that injection of theFluosol-DA (FDA) 20% emulsion in animals breathing normobaric carbogen,normobaric or hyperbaric oxygen, significantly increases the radiosensitivity ofrodent solid tumours to both single and fractionated doses of radiation (Teicher andRose 1984a,b, 1986, Rockwell 1985, Rockwell et al. 1986, 1988, Song et al . 1985,1987, Lee et al . 1987, Martin et al . 1987a,b, Moulder and Fish 1988). Thisemulsion plus carbogen or 100% oxygen produces little (Rockwell et al . 1986, Leeet al . 1987), or undetectable (Mason et al . 1985, Rockwell et al . 1986) radio-sensitization of healthy tissues .

$Address correspondence to : Dr Charles Thomas, c/o Professor Adams, MRC Radiobiology Unit,Chilton, Didcot, Oxford OXI I ORD, UK .

0020-7616/91 $3 .00 © 1991 Taylor & Francis Ltd

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

434

C. Thomas et al .

One of the fluorocarbon emulsions recently introduced in an effort to producemore stable, more effective and less toxic emulsions than FDA 20% emulsion, isthe perfluorooctylbromide (C 8 F17Br) (PFOB) 100% w/v emulsion (Long et al .1988b, Arlen et al. 1988). Several studies have shown that the toxicity of PFOBemulsion is lower than that of FDA 20% emulsion (high LD 0 (Burgan et al. 1988))and that its retention time is short (Riess and Le Blanc 1988, Long et al. 1988a) . Fora given injected volume the concentration of fluorocarbon in the emulsion is timesgreater for 100% w/v PFOB emulsion than for FDA 20% emulsion (Riess 1990) .Thus the oxygen-carrying capacity of the 100% w/v PFOB emulsion is 4 timesgreater than that of FDA 20% emulsion . Radiosensitization of normal tissues issmall or nil (Thomas et al . 1989b, Lartigau et al. 1989). The 100% w/v PFOBemulsion should be preferable to FDA 20% emulsion for therapeutic tests ofhypoxic cell radiosensitization . However, tests in our laboratory (Thomas et al .1989a) indicate that 1 ml/kg 100% w/v PFOB emulsion plus carbogen producesonly a slightly higher radiosensitization of a human tumour xenograft than doescarbogen alone, although this effect of carbogen was not as large as that of the fulloxygen effect in vitro .

Relatively low doses of FDA 20% emulsion (7-1 ml/kg) have been found toproduce considerable radiosensitization, but higher doses (20-60 ml/kg) causedeither no increase or a reduction in radiosensitization (Teicher and Rose 1984a,1986, Martin et al . 1987a, Rockwell et al . 1988). Sasai et al . (1989) also recentlyshowed that the radiosensitivity of RIF-1 tumour was unchanged by injection of20 ml/kg FDA 20% emulsion plus carbogen, although this tumour was radiosensit-ized by injection of 1 ml/kg (Song et al . 198 , 1987) .

The present study was designed to examine the influence of the dose of 100%w/v PFOB emulsion plus carbogen on the radiosensitivity of two human tumourxenografts and the murine tumour EMT6 . The dose of 100% w/v PFOB emulsionproducing optimal tumour radiosensitivity was determined, and the haematocritand fluorocrit (the amount of fluorocarbon in the blood) were measured as afunction of 100% w/v PFOB emulsion injected, and also as a function of the timepost-injection .

2 . Materials and methods2.1 . Tumour cell studies

Human rectal adenocarcinoma (HRT18) cells, human colon adenocarcinoma(HT29) cells and EMT6 mouse mammary tumour cells were maintained in vitrobefore transplantation into 3- -month-old athymic nude mice . The characteristicsand maintenance of the cell lines, the tumour transplantation technique and thein vitro colony assay used to determine survival level have already been published(HRT18, HT29: Guichard et al . 1983 and EMT6: Guichard et al . 1977). Briefly,for the in vitro colony assays, tumours were removed immediately after irradiation,dispersed to single cells using 0 .2% trypsin solution ; after 20 min the cell suspen-sion was filtered through a stainless-steel filter, cells were collected by centrifuga-tion and seeded in plastic flasks at the appropriate concentration . The culturemedium (MEM, Earle's salts) was supplemented with 20% foetal calf serum andantibiotics ( 0 mg kanamycin and 1 mg gentamycin per litre) . All cultures werethen placed in an incubator (air + % C02 ) at 37 °C. The tumours had a meandiameter of 7mm (volume of about 180mm 3 ) on the day of irradiation, i .e . 4-weeks (HRT 18), 3-4 weeks (HT29) and about 2 weeks (EMT6) after cell injection .

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

100% w/v PFOB + carbogen and radiosensitization

435

Unanaesthetized unrestrained mice were placed in a 21 cm x 21 cm x 3 cmPlexiglas box and were whole-body irradiated with y-rays ( 137Cs-dose rate about1 .7 Gy/min) . The homogeneity of the dose over the irradiation field was checkedwith Agfa Structurix D2 radiography film (§3%) . As this irradiation techniquediffers from that previously used in the laboratory, the radiobiological hypoxictumour cell fractions were determined . Tumours were irradiated under fourexperimental conditions : (1) air-breathing mice ; (2) mice breathing carbogen (95%02-5 % CO2 at atmospheric pressure, flow rate 101/min) 30 min prior to and duringirradiation ; (3) mice sacrificed by cervical dislocation 10 min before the irradiation ;(4) mice pretreated with the 100% w/v PFOB emulsion (injection i .v. into theretro-orbital sinus) and breathing carbogen 30 min prior to and duringirradiation . Experimentation and care of animals were in accordance with theAnimal Act No. 0167 given from the Ministere de ]'Agriculture et de la Foretto Dr Marcelle Guichard .

2.2 . Perfluorocarbon emulsion and fluorocritPerfluorooctylbromide (PFOB) is a linear fluorocarbon obtained from oligomers

of tetrafluoroethylene (Long et al. 1988b). The 100% w/v PFOB emulsion (Batch806103) was kindly supplied by Alliance Pharm . Corp., San Diego, California,USA . The fluorocarbon was emulsified with egg yolk phospholipid or lecithin (4 . 5w/v %). The 100% w/v PFOB emulsion contains 100% by weight of fluorocarbon,which represents 52% by volume . Experiments were performed with 2-15 ml/kg100% w/v PFOB emulsion doses (2-15 g/kg fluorocarbon doses) injected in theretro-orbital sinus . The retro-orbital sinus route of injection has, in our hands,proven equal to or more reliable than tail-vein injection . (When tail-vein and sinusinjections were compared in nude mice, mean fluorocrit values (95% confidenceintervals) were 4 .5% (3 . 7-5 .2) and 44% (3-9-4-8), respectively) .

In a first step the time-course responses of the amount of fluorocarbon in the blood(fluorocrit) and haematocrit were measured after an injection of 100% w/v PFOB ata dose of 8 ml/kg into 2-month-old female BALB/c mice and into 2-5-month-oldfemale nude mice. In a second step the dose responses of fluorocrit and haematocritwere measured in nude mice after an injection of the 100% w/v PFOB emulsion atdoses ranging from 2 to 15 ml/kg . The fluorocrit and haematocrit were determinedby a modification of the method of Tremper et al . (1980) . After the injection of theemulsion, mice were lightly anaesthetized with ether (<30s) and blood sampleswere taken from the retro-orbital sinus using a heparinized microhaematocrit tube .The tube was centrifuged for 10min at 10000 rpm and the fluorocarbon contentwas measured by the length of the white band lying under the sedimented cells andthe plasma . The measured fluorocrit was probably an overestimation as the whitecolour indicates the presence of lecithin vesicles . Haematocrit and fluorocrit weredetermined on the same tube using the same reference scale .

3 . ResultsAll confidence intervals were calculated for a 5 % error.

3 .1 . Fluorocrits and haematocrits3 .1 .1 . Time-course response . In a first step the fluorocrit and haematocrit wereextensively studied into 2-month-old female Balb/c mice : they remainedunchanged from 7 to 65 min post-injection of the 100% w/v PFOB emulsion(8 ml/kg) (Figure 1) . We have verified that fluorocrits obtained from blood samples

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

436

i000LL

7-

6-

-

C. Thomas et al .

I 1

0

1

30

4

60

7

Time After PFOB Emulsion Administration (min)

Figure 1 . Time-course response of fluorocrit and haematocrit after PFOB 100% w/vemulsion injection (8 ml/kg) in Balb/c mice . Fluorocrit (∎) and haematocrit (A) afteri .v . injection via the retro-orbital sinus in lightly anaesthetized mice ; blood sampletaken from the same sinus . Fluorocrit( •) and haematocrit ( •) after i .v . injection viathe retro-orbital sinus in lightly anaesthetized mice ; blood sample taken from thecontralateral retro-orbital sinus . Control haematocrit of non-injected mice (hatchedarea). Mean and 9 % confidence intervals from two or three experiments are shown .

taken from the sinus which had been used to administer the emulsion did not differsignificantly when emulsion was introduced and blood was sampled from thecontralateral sinus (Figure 1) . In a second step the fluorocrit and haematocrit weremeasured 10 and 60 min after i .v. injection in 2- -month-old nude mice; they werenot dependent on the age of the mice . Mean fluorocrits (9 % confidence intervals)measured 10 and 60 min after the injection were . 0% (4 .7- .2) and -1% (4.9- . 3),respectively .

3 .1 .2 . Dose response . The fluorocrit was measured in nude mice immediately afterirradiation, i .e. about 30 min after i .v. injection of the PFOB emulsion . It increasedfrom 1- % to 9- % as the emulsion dose increased from 2 to 1 ml/kg (Figure 2,panel 1). The haematocrit remained unchanged 30 min after the emulsion injectionregardless of the dose injected (Figure 2, panel 1) .

3 .2 . Radiosensitivity3 .2 .1 . Radiosensitivity as a function of the dose of 100% w/v PFOB emulsion . Therelationship between radiosensitivity and dose of PFOB emulsion injected(2-1 ml/kg) was determined for the HRT18, HT29 and EMT6 tumours . Radio-sensitivity was studied under three experimental conditions : air-breathing mice,carbogen-breathing mice and carbogen-breathing mice injected with the 100% w/vPFOB emulsion (Figure 2, panels 2 to 4) . A single dose of irradiation was used inthe initial experiments (10 Gy for HRT18 and HT29, 20 Gy for EMT6) . Themajority of surviving cells were severely hypoxic at these radiation doses . Theplating efficiencies remained unchanged in all the tumour cell lines for all doses ofPFOB emulsion tested. The mean plating efficiency for doses of PFOB emulsion

0

Bd- 40 0

n

ag

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

84

00

U-

10-1 HRT18 O

100% w/v PFOB + carbogen and radiosensitization

437

2

4

6

8

10

12

1 '4

1'6

0-'

•

0

•

10 -2 - qLL

LL

a

OIC

C

~, 10

HT29 310" 1 \\\\\\\\\\\\

10-2

,

0

2

4

EMT6

10

12

14

16

10 -1

10-4

0246810121416

0246810121416'PFOB Emulsion Dose (ml/kg)

PFOB Emulsion Dose (ml/kg)

Figure 2 . Panel 1 : effect of the 100% w/v PFOB emulsion dose on the fluorocrit andhaematocrit, measured 30 min after injection . Fluorocrit ( •) ; haematocrit (A) ;control haematocrit (>W). Panels 2 to 4: effect of 100% w/v PFOB emulsion dose ontumour cell survival . The pretreatment time with emulsion and/or carbogen wasalways 30 min prior to and during irradiation . 9 % confidence intervals for tumoursirradiated in air-breathing mice (hatched area : \\\\) and in carbogen-breathing mice(hatched area : ////) are shown . Tumours irradiated in mice pretreated with theemulsion and carbogen (∎) . Each experiment was performed two or three times .Number of tumours per point : HT29 (five to fifteen), HRT 18 (five to fifteen), EMT6(eleven to fifteen). Means and 9 % confidence intervals are shown .

from 2 to 1 ml/kg is shown in Table 1 . Neither carbogen-breathing alone or PFOBemulsion plus carbogen changed the plating efficiencies of any of the tumour celllines (t-test non-significant, Table 1) .

The radiosensitizing effect of carbogen alone was compared to that of 100% w/vPFOB emulsion plus carbogen : carbogen-breathing produced an 18-fold reductionin the survival fraction of HRT18 cells and a 20-fold reduction in that of HT29cells, with respect to those of the same cell lines in air-breathing mice (t-test,p=10 -2 for HRT18 and p<10 -3 for HT29) (panels 2 and 3) . However, thesurvival of EMT6 cells in carbogen-breathing mice was not significantly modified(Figure 2, panel 4) .

The radiosensitivity of tumour cells irradiated in carbogen-breathing miceinjected with 100% w/v PFOB emulsion varied with the dose of PFOB emulsion .For HRT18 tumours the cell survival levels at a PFOB emulsion dose of 2 ml/kgand 3 ml/kg were significantly lower (7- and -fold respectively) than that of the

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

438

C. Thomas et al .

Table 1 . Plating efficiencies for the cell lines HRT18, HT29 and EMT6 derived from solidtumour in mice breathing air, carbogen or treated with carbogen and 100% w/vPFOB emulsion (2-1 ml/kg)

Group

Carbogen-breathingmice pretreated with

Carbogen-breathing

PFOB emulsionTumours

Air-breathing mice

mice

(2-1 ml/kg)

HRT18

0. 4

0. 8

0. 1(0.4-0 •68)

(0.43-0 •73)

(0.43-0• 8)HT29

0.24

0.29

0.21(0 . 18-0 •3 )

(0.21-0 .37)

(0. 1 -0•27)EMT6

0.30

0.2

0.26(0 .21-0 •3 9)

(0.07-0 •44)

(0.22-0•29)

Means and 9 % confidence interval are shown; six to thirty tumours per point for thedetermination of the mean plating efficiency .

same cells in carbogen-breathing mice (p=1 . 3 x 10 -2 and 3 . 8 x 10 -2 , respectively) .Cell survival increased for PFOB emulsion doses from 3 to 1 ml/kg and theradiosensitivity obtained for 8 and 1 ml/kg was comparable to that obtained withcarbogen alone (Figure 2, panel 2) .

For HT29 tumours the radiosensitization of tumour cells in carbogen-breathingmice given 2 ml/kg PFOB emulsion was the same as that of tumour cells incarbogen-breathing mice . At a dose of 4 ml/kg, PFOB emulsion significantlyincreased carbogen-induced radiosensitization ; cell survival was 2-fold lower thanwith carbogen alone (p=0 .02). Cell survival increased as the dose of PFOBincreased from 4 to 1 ml/kg; the radiosensitizing effect of 8 ml/kg PFOB emulsionplus carbogen was not significantly different from that of carbogen alone . Finally,1 ml/kg PFOB emulsion plus carbogen showed a radioprotective effect comparedto carbogen alone (p<10-3) (Figure 2, panel 3). For EMT6 tumours, carbogenplus low doses of PFOB emulsion produced 3-fold (2 ml/kg) and -fold (4 ml/kg)reductions in cell survival compared to the cell survival in animals breathingcarbogen alone (t-test, p < 10 -2 for 2 ml/kg and p < 10 -3 for 4 ml/kg)). The cellsurvival fractions increased for PFOB emulsion doses of 4-1 ml/kg, and wascomparable to that in carbogen-breathing mice at 8 and 1 ml/kg (Figure 2,panel 4) .

3.2 .2. Survival curves . Cell survival experiments were performed for HRT18 andEMT6 tumours irradiated under three experimental conditions : in air-breathingmice, in mice sacrificed 10 min before irradiation and in carbogen-breathing micepretreated with PFOB emulsion (Figures 3 and 4) . The dose of PFOB emulsionused was that giving maximum radiosensitization in carbogen-breathing mice . Thesurvival curves parameters are reported in Table 2 .

The average percentage of severely hypoxic cells in HRT18 tumours irradiatedin air-breathing mice was 7% (Table 2) . Cell radiosensitivity in carbogen-breathingmice pretreated with 2 ml/kg PFOB emulsion was very close to that of normallyoxygenated cells (irradiated in vitro h after excision of the solid tumour) (Figure

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

100% w/v PFOB + carbogen and radiosensitization

1=

10-2

10-4 ;

10-4

10-8

,

,

,

i0

10

1

20

2

30DOSE (Gy)

Figure 3 . Effect of 100% w/v PFOB emulsion (2 ml/kg) on HRT18 cell survival . Allanimals were pretreated with emulsion and/or carbogen for 30 min prior to and duringirradiation . Tumours irradiated in mice breathing air (A), carbogen (*), sacrificed bycervical dislocation 10 min before irradiation (∎) and pretreated with emulsion pluscarbogen (•). Cells dissociated from the tumour and irradiated in vitro immediately(Guichard et al. 1983) or

h later (Guichard et al

. 1983, Thomas et al . 1989a) ( •) .Each experiment was performed two or three times, five to fifteen tumours per point .Means and 9 % confidence intervals are shown . Curves were fitted by linearregression .

439

0,10

1

20DOSE (Gy)

2

30

•

10-' -

•

10-1

Figure 4. Effect of 100% w/v PFOB emulsion (4 ml/kg) on EMT6 cell survival . All animalswere pretreated with emulsion and/or carbogen for 30 min prior to and duringirradiation. Tumours irradiated in mice breathing air (A), carbogen (*), sacrificed bycervical dislocation 10 min before irradiation (∎) and pretreated with emulsion pluscarbogen ( •) . Cells dissociated from the tumour and irradiated in vitro immediatelyafter ( •). Each experiment was performed two or three times, five to fourteentumours per point . Means and 9 % confidence intervals are shown . Curves werefitted by linear regression .

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

440

C. Thomas et al .

NV.0

V0

4 4 O

wV y

A

T

q

CO

n0000

1-

•

~p N

pp M

N

M p

M

N ;nO

M -

nM

171M

I

•

N

> a-0 CQ Va+

U

Nb O• •~

u bV

4) cN

V Cq

yC . ccV cy

ate.+ .~ Ly

C > ~

Gy w,CC•

O O

O p

o a• •~i. O

V

3 " >

cd•

00 U ~~O VQ N .O O

3 ~a

Vv. 0

COV

MOOOO

r-

Ln

10

Ln

> 00U 0 CW IVo 0 oL cd y

d

VOO tn tn

m .b UVc(d G:a Uv, C. >V .17

cd

O Cy Vcr

3

7 0 Or-16

O

C7

r .

00

N

O ' COV .y V

on O .O

O 0 0 NA 0

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

100% w/v PFOB + carbogen and radiosensitization

441

3); the ratio: Do obtained with mice sacrificed 10 min before the irradiation/Doobtained with mice treated with carbogen plus PFOB (D0(N2)D0 (C + PFOB)) was3 . 1 . The sensitizing enhancement ratio (SER), defined as the ratio of the radiationdose giving the same survival fraction in the absence and presence of emulsion pluscarbogen, increased with the radiation dose (Table 3) . The mean SER was 1 .7; thehypoxic cell fraction in HRT18 tumours (calculated at a survival level of 2 x 10 -4 inmice treated with PFOB emulsion plus carbogen) was reduced to about 0 . 2%(Figure 3) .

For EMT6 tumours the mean percentage of severely hypoxic cells in air-breathing mice was 6% (Table 3) . The radiosensitivity of EMT6 cells in PFOBemulsion (4 ml/kg) injected carbogen-breathing mice was higher than in air-breathing mice (Figure 4) . The ratio D0(N 2)D0 (C + PFOB) was 1 .9, which is morethan half the full oxygen effect . The SER was 1 . 4, and was independent of cellsurvival (Table 3), indicating that PFOB emulsion plus carbogen acted as a dose-modifying factor . The hypoxic cell fraction in the EMT6 tumour was reduced toapproximatively 0 .4% (calculated for a survival fraction of 3 x 10 -4 in mice treatedwith PFOB emulsion plus carbogen) .

4 . DiscussionSeveral studies have shown that experimental results depending directly on the

degree of tissue oxygenation may be significantly influenced by the physicalrestraint of the animal during irradiation (Zanelli and Lucas 1976, Shibamoto et al.1987, Tozer 1987, Stevens and Michael 1988, Thomas and Guichard unpublishedresults) . In particular, for certain tumours the percentage of severely hypoxic cellsis dependent on the irradiation modality. It is lower when animals are unrestrainedduring irradiation . The hypoxic cell fractions in mice restrained during irradiationare 13% for the SCCVII tumour (Shibamoto et al. 1987), 14% for the HRT18tumour (Guichard et al . 1983) and 44% for the EMT6 tumour (Guichard et al .1977); in mice unrestrained during irradiation these figures are 4% for SCCVII(Shibamoto et al . 1987), and 7% and 6% for HRT18 and EMT6 respectively (thisstudy) . All these studies have been performed with tumours of a comparable size .However, the hypoxic cell fraction is not always influenced by the way the animal isrestrained; indeed in the HT29 tumour this fraction remained unchanged at about10% in mice restrained or unrestrained during irradiation (unpublished results) .

The duration of tumour oxygenation by oxygen carrier is an importantparameter in the radiosensitization of tumours by fluorocarbon emulsions . This

Table 3. Sensitizing enhancement ratio calculated for different cell survival levels in micepretreated with 100% w/v PFOB emulsion plus carbogen

Mean 1 . 7

Mean 1 .4

Sensitizing enhancement ratioSurviving

level HRT18 tumour EMT6 tumour

0. 1 1 . 33 1 . 390 .01 1 . 68 1 .420.001 1 . 82 1 . 430.0003 1 . 9 1 . 44

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

442

C. Thomas et al .

effect was indirectly evaluated by determining the fluorocrit as a function of timeafter the 100% w/v PFOB emulsion injection (8 g/kg of fluorocarbon) . Thefluorocrit and haematocrit remained the same from 7 to 6 min post-injection(Figure 1) . This stability can be compared to results obtained in humans : after anincrease of the fluorocrit during the slow infusion of FDA 20% emulsion thefluorocrit reached a plateau over the following 2 h (Tremper et al . 1982). Incontrast, in rats the fluorocarbon concentration in the blood measured by gaschromatography decreased by 12% of the control value during the first hour afterinjection of FDA 20% emulsion (4 g/kg of fluorocarbon) (Watanabe et al. 1988) .This difference might be due to the host (mouse versus rat) or to the techniqueemployed to measure the amount of fluorocarbon in the blood (fluorocrit versus gaschromatography) .

The fluorocarbon content of the emulsion was used as the basis for comparingthe radiosensitizing capacities of 100% w/v PFOB and FDA 20% emulsions(always in carbogen-breathing animals) . In our experiments, although the fluoro-crit increases with the dose of fluorocarbon administered, tumour radiosensitivitydoes not always increase as a function of the fluorocarbon dose . The radiosensitiz-ing capacity of 100% w/v PFOB emulsion increased with the fluorocrit for low dosesof fluorocarbon (2-4 g/kg), but did not increase further at higher doses (4-1 g/kg) .It decreased in some cases, and radioprotection even occurred, compared tocarbogen alone. This result is comparable to that obtained with FDA 20%emulsion, for which a fluorocarbon dose of 3 g/kg (1 ml emulsion/kg) alsoproduced maximal radiosensitization (Rockwell et al . 1988, Martin et al . 1987a,Teicher and Rose 1984a, 1986) . With Therox, another perfluorochemical emulsion,the dose of 4 g/kg produced the largest radiosensitizing effect (Teicher et al . 1989) .

It has been suggested that the increase in oxygen transport with the fluorocritmay be offset by the reduction of haematocrit by hemodilution : Martin et al .(1987a) observed a reduction in radiosensitization with FDA 20% emulsion and areduction in haematocrit when the emulsion dose increased from 20 to 60 ml/kg(4-12g fluorocarbon/kg); however, these experiments were performed with rats,thus weakening the comparison with mice as blood volume/body weight ratios arenot readily comparable between those two species . Rockwell et al . (1988) observedno modification in EMT6 tumour radiosensitization for doses of FDA 20%emulsion from 20 to 4 ml/kg (4-9 g fluorocarbon/kg) which did not change thehaematocrit . In our experiments the hemodilution effect is times lower than withFluosol DA 20%, and similar to what was observed by Rockwell et al . (1988), nomodification of the haematocrit was observed for doses of fluorocarbons whichvaried from 2 to 1 g/kg .

We could speculate on other factors which could influence the radiosensitizationinduced by PFOB. The oxygen carried by the fluorocarbon is taken up by thetissues before removal of haemoglobin-bound oxygen begins (Tremper et al . 1982) .Also, oxygen is more easily extracted from fluorocarbon than from haemoglobinbecause of the lack of chemical bonding between fluorocarbon and oxygen, the3-fold faster rate of oxygenation/deoxygenation and the much greater (several100-fold) exchange surface (Riess and Le Blanc 1982) . Finally, the lower viscosity offluorocarbon emulsions than that of blood should allow a faster flow of oxygenatingmedium (Naito and Yokohama 1978) . These phenomena together should result inan increase in tumour radiosensitization with increasing dose of fluorocarbon, and

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

100% w/v PFOB+carbogen and radiosensitization

443

this is not what was observed whatever the fluorocarbon used . An off-loading ofoxygen from red blood cells in the presence of high concentrations of oxygenatedPFC emulsions had been suggested (Teicher et al. 1989), but it has also beensuggested that fluorocarbon emulsions could help oxygen delivery by red bloodcells by providing a type of relay (S . Faithfull private communication) .

Part of our results could be explained by the variations in the volume injected inthe mice: it varied in our experiments from 0 .06 ml (2 g/kg of fluorocarbon) to0 .4 ml (1 g/kg of fluorocarbon) . Using Therox, Teicher et al . (1989), found that avery important parameter to be considered was the volume injected : an injectedvolume higher than 0 .2 ml led to a decrease of the radiosensitization of the Lewislung carcinoma for a given amount of fluorocarbon injected .

Finally, the reduction in the radiosensitizing effect of oxygenated emulsionsat fluorocarbon doses above 4 g/kg does not seem to be linked to a reductionin intratumoral blood flow. Blood flow (BF) through a tissue is given by :BF=(Pa -P,,)/rl •Z where (Pa -P,,) is the arterial-venous pressure difference, tl isblood viscosity and Z is the vessel geometric resistance (see Jain 1988) . Administra-tion of oxygenated fluorocarbons with carbogen should produce an increase in BF,as (Pa -P,,) increases (Tremper et al. 1982), the viscosity of the blood-emulsion mixis lower (Naito and Yokohama 1978) and the vascular resistance Z should bereduced as carbogen is known to cause peripheral vasodilation (Olive and Inch1973). Fluorocarbon emulsion has no effect, to our knowledge, on vessel geometry .A recent study (Watanabe et al . 1988) confirmed this analysis by showing thatintratumoral BF increased by 0% after administration of 20 ml/kg FDA 20%emulsion (4g fluorocarbon/kg) to carbogen-breathing animals . Moreover, theincreasing blood volume with emulsion dose should therefore increase the blood-emulsion pressure and hence tumour blood flow . Preliminary studies have shownthat tumour blood perfusion, as measured by the 86Rb extraction technique, is notreduced in the SCCVII tumour, after an injection of 1 ml/kg of 100% w/v PFOBemulsion (Thomas et al . unpublished results). Studies are in progress to see if thetumour perfusion is modified after administration of 100% w/v PFOB emulsion innude mice bearing tumours .

. ConclusionsSignificant tumour radiosensitization was obtained with relatively low amounts

of 100% w/v concentrated emulsion of PFOB plus carbogen . The maximumradiosensitization occurs at a low fluorocarbon dose of about 3 g/kg . These resultsare comparable to those obtained by others with Fluosol-DA 20% or Theroxemulsion . Since this radiosensitization occurs only at relatively low fluorocritwithout modification of the haematocrit, the oxygen-carrying capacity of thefluorocarbon is not the only factor involved in radiosensitization of tumour cells,regardless of the effect of carbogen on radiosensitivity .

AcknowledgementsThe authors wish to thank Professor G . Adams, Dr E. P. Malaise and Dr D .

Long for many helpful discussions. They also thank Mr Patrice Ardouin forproduction of animals, Mrs Valerie Frascogna for animal care and technicalassistance, Mr H . Bounick for carrying out the dosimetry and Mrs Joelle Encinas

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

444

C. Thomas et al .

for typing the manuscript. This work was supported in part by the' Ligue NationaleFrangaise Contre le Cancer' (Comite des Hauts de Seine) .

ReferencesARLEN, C., FOLLANA, R., LE BLANC, M., LONG, C., LONG, D., RIESS, J. G. and VALLA, A.,

1988, Formulation of highly concentrated fluorocarbon emulsions and assessment bynear-total exchange perfusion of the conscious rat . Biomaterials, Artificial Cells andArtificial Organs, 16, 44 -4 7 .

BURGAN, A. R ., HERRICK, W. C., LONG, D. M. and LONG, D. C., 1988, Acute and subacutetoxicity of 100% PFOB emulsion . Biomaterials, Artificial Cells and Artificial Organs,16,681-682 .

GUICHARD, M., DERTINGER, H . and MALAISE, E. P., 1983, Radiosensitivity of four humantumour xenografts . Influence of hypoxia and cell-cell contact . Radiation Research, 9 ,602-609 .

GUICHARD, M ., LACHET, B. and MALAISE, E. P., 1977, Measurement of RBE, OER andrecovery of potentially lethal damage of a 64 MeV helium ion beam using EMT6cellls . Radiation Research, 71, 413-419 .

JAIN, R. K., 1988, Determinants of tumor blood flow : a review. Cancer Research, 48,2641-26 8 .

LARTIGAU, E., THOMAS, C ., LE BLANC, M., RIESS, J. G ., LONG, D., LONG, C., MALAISE, E. P .and GUICHARD, M ., 1989, New high 02 carrying perfluorochemical emulsions :toxicity, radiosensitivity of GM-CFC and development of metastases in mice .International Journal of Radiation Oncology, Biology, Physics, 16, 11 3-11 6 .

LEE, I ., LEVITT, S . H. and SONG, C .W., 1987, Effects of Fluosol-DA 20% and carbogen on theradioresponse of SCK tumors and skin of A/J mice . Radiation Research, 112,173-182 .

LONG, D. M ., LONG, D. C ., MATTREY, R. F ., LONG, R. A ., BURGAN, A. R ., HERRICK, W. C . andSHELLHAMER, D. F., 1988a, An overview of perfluoroctylbromide-Application as asynthetic oxygen carrier and imaging agent for X-ray, ultrasound and nuclear magneticresonance . Biomaterials, Artificial Cells and Artificial Organs, 16, 411-420 .

LONG, D. C ., RIESS, J. G ., FOLLANA, R., BURGAN, A. R ., MATTREY, R. F. and LONG, D. M .,1988b, Preparation and applications of highly concentrated PFOB emulsions. Bio-materials, Artificial Cells and Artificial Organs, 16, 441-442 .

MARTIN, D. F ., PORTER, E . A., FISHER, J. J . and ROCKWELL, S ., 1987a, Effect of aperfluorochemical emulsion on the radiation response of BAI112 rhabdomyosarco-mas . Radiation Research, 112, 4 - 3 .

MARTIN, D. F ., PORTER, E. A ., ROCKWELL, S . and FISHER, J . J ., 1987b, Enhancement of tumorradiation response by the combination of a perfluorochernical emulsion and hyperbaricoxygen . International Journal of Radiation Oncology, Biology, Physics, 13, 747-7 1 .

MASON, S . A., WITHERS, H. R ., and STECKEL, R. J ., 198 , Acute affects of a perfluorochemicaloxygen carrier on the normal tissues of the mouse . Radiaion Research, 104, 387-394 .

MOULDER, J. E., and FISH, B . L., 1988, Intermittent use of a perfluorochemical emulsion(Fluosol-DA 20%) and carbogen breathing with fractionated irradiation . InternationalJournal of Radiaton Oncology Biology Physics, 1 , 1193-1196 .

NAITO, Y . and YOKOHAMA, K., 1978, Perfluorochemical blood substitutes . Technical Informa-tion Series no . S and no . 7 (The Green Cross Corp ., Osaka, Japan) .

OLIVE, P . L. and INCH, W. R., 1973, Effect of inhaling oxygen/carbon dioxide mixtures onoxygenation and cure by X rays of the C3H BA mouse mammary carcinoma .Radiology, 106, 376-678 .

RIESS, J. G., 1990, Assessment of advances in fluorocarbon-based injectable oxygen carriers .Artificial Organs (In press) .

RIESS, J. G. and LE BLANC, M., 1982, Solubility and transport phenomena in perfluorochemi-cals relevant to blood substitution and other biochemical applications . Pure AppliedChemistry, 4, 2383-2406 .

RIESS, J . G. and LE BLANC, M., 1988, Preparation of perfluorochemical emulsions forbiomedical use . Materials and Methods . Blood Substitutes, Preparation, Physiology andMedical Applications, edited by K . C . Lowe (Ellis Horwood, Chichester), pp . 94-124 .

ROCKWELL, S ., 198 , Use of a perfluorochemical emulsion to improve oxygenation in a solidtumor . International Journal of Radiation Oncology, Biology, Physics, 11, 97-103 .

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.

100% w/v PFOB + carbogen and radiosensitization

44

ROCKWELL, S., MATE, T . P ., IRVIN, C . G . and NIERENBURG, M ., 1986, Reactions of tumors andnormal tissues in mice to irradiation in the presence or absence of a perfluorochemicalemulsion . International Journal of Radiation Oncology, Biology, Physics, 12,131 -1318 .

ROCKWELL, S ., IRVIN, C . G . and KELLEY, M ., 1988, Preclinical studies of a perfluorochemicalemulsion as an adjunct to radiotherapy . International Journal of Radiation Oncology,Biology, Physics, 1 , 913-920 .

SASAI, K., ONO, K., NISHIDAI, T., SHIBAMOTO, Y., TAKAHASHI, M . and ABE, M ., 1989,Variation in tumor response to Fluosol-DA . International Journal of RadiationOncology, Biology, Physics, 16, 1149-11 2 .

SHIBAMOTO, Y., SASAI, K . and ABE, M ., 1987, The radiation response of SCCVII tumor cellsin C3H/He mice varies with irradiation conditions . Radiation Research, 109, 3 2-3 4 .

SONG, C. W., ZHANG, W. I ., PENCE, D. M., LEE, I. and LEVITT, S. H., 198 , Increasedradiosensitivity of tumors by perfluorochemicals and carbogen . International Journalof Radiation Oncology, Biology, Physics, 11, 1833-1836 .

SONG, C. W., LEE, I ., HASEGAWA, T ., RHEE, J . G ., and LEVITT, S. H., 1987, Increase in PO,and radiosensitivity of tumors by Fluosol-DA 20% and carbogen . Cancer Research, 47,442-446 .

STEVENS, G .N. and MICHAEL, B . D., 1988, Breathing conditions for animals in radiobiologicalexperiments . British Journal of Radiology, 61, 863-864 .

TEICHER, B., HERMAN,T. and JONES, S., 1989, Optimization of perfluorochemical levels withradiation therapy in mice. Cancer Research, 49, 2693-2697 .

TEICHER, B. A. and ROSE, C. M., 1984a, Oxygen-carrying perfluorochemical emulsion as anadjuvant to radiation therapy in mice . Cancer Research, 44, 428 -4288 .

TEICHER, B . A. and ROSE, C. M ., 1984b, Perfluorochemical emulsions can increase tumorradiosensitivity. Science, 223, 934-936 .

TEICHER, B .A. and ROSE, C. M., 1986, Effects of dose and scheduling on growth delay of thelewis lung carcinoma produced by the perfluorochemical emulsion, Fluosol-DA .International Journal of Radiation Oncology, Biology, Physics, 12, 1311-1313 .

THOMAS, C., LARTIGAU, E., MALAISE, E. P . and GUICHARD, M ., 1989a, New high 0 2 carryingperfluorochemical emulsions and/or carbogen: reactions of a human tumor xenograftto irradiation . International Journal of Radiation Oncology, Biology, Physics, 16,11 7-1160 .

THOMAS, C., DE VATHAIRE, F., LARTIGAU, E., MALAISE, E. P. and GUICHARD, M., 1989b,Radiosensitivity of mouse lip mucosa : influence of anesthesia, carbogen and a new high02 carrying perfluorochemical emulsion . Radiation Research, 118, 476-487 .

TOZER, G . M., 1987, Some artefacts involved in the radiation response of mouse tumorsarising from anesthesia and physical restraint . Rodent Tumor Models in ExperimentalCancer Therapy, edited by R . F. Kallman (Pergamon, Oxford), pp . 47-49 .

TREMPER, K. K., LAPIN, R., LEVINE, E., FRIEDMAN, A. and SCHOEMAKER, W . C ., 1980,Hemodynamic and oxygen transport effects of a perfluorochemical blood substitute,Fluosol-DA 20% . Critical Care Medicine, 8, 738-741 .

TREMPER, K. K., FRIEDMAN, A. E ., LEVINE, E. M., LAPIN, R . and CAMARILLO, D., 1982, Thepreoperative treatment of severely anemic patients with a perfluorochemical oxygen-transport fluid, Fluosol-DA . New England Journal of Medicine, 307, 277 .

WATANABE, M., OKAMOTO, H., OHYANAQI, H., SAITOH, Y ., YANG, K., TSUDA, Y., UEDA, Y.and YOKOHAMA, K., 1988, Oxygenation effect of Fluosol-DA on hypoxic tumor tissueand retention of PFCs in the tumor-bearing rats . Biomaterials, Artificial Cells andArtificial Organs, 16, 47- 6 .

ZANELLI, G . D. and LUCAS, P . B., 1976, Effect of stress on blood perfusion and vascular spacein transplanted mouse tumours . British Journal of Radiology, 49, 382-383 .

Int J

Rad

iat B

iol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y IN

SER

M o

n 04

/19/

12Fo

r pe

rson

al u

se o

nly.