[CANCER RESEARCH 55. 3752-3756, September 1, 1995|

Advances in Brief

Vascular Permeability in a Human Tumor Xenograft: Molecular SizeDependence and Cutoff Size1

Fan Yuan, Marc Dellian, Dai Fukumura, Michael Leunig, David A. Berk, Vladimir P. Torchilin, and Rakesh K. Jain

Edwin L Steele Laboratory, Department of Radiation Oncology ¡F.Y.. M. D., D. F., M. L, D. A. B., R. K. ].], and Center for Imaging and Pharmaceutical Research ¡V.P. TJ.Massachusetts General Hnspital and Harvard Medical School. Boston. Massachusetts 02114

Abstract

Molecular size is one of the key determinants of transvascular transportof therapeutic agents in tumors. However, there are no data in theliterature on the molecular size dependence of microvascular permeabilityin tumors. Therefore, we measured microvascular permeability to variousmacromolecules in the human colon adenocarcinoma I SI74T transplanted in dorsal skin chambers in severe combined immunodeficientmice. These molecules were fluorescently labeled and injected i.v. intomice. The microvascular permeability was calculated from the fluorescence intensity measured by the intravital fluorescence microscopy technique. The value of permeability varied approximately 2-fold in the range

of molecular weight from 25,000 to 160,000. These data indicate thattumor vessels are less permselective than normal vessels, presumably dueto large pores in the vessel wall. The transport of macromolecules appearsto be limited by diffusion through these pores. The cutoff size of the poreswas estimated by observations of transvascular transport of stericallystabilized liposomes of 100-600 inn in diameter. We found that tumor

vessels in our model were permeable to liposomes of up to 400 mn indiameter, suggesting that the cutoff size of the pores is between 400 and600 mil in diameter.

Introduction

Previous quantitative studies have demonstrated that the permeability of tumor vessels (Pv)2 is in general higher than that of normal

vessels (1-6). The increase in permeability is hypothesized to be anecessary condition for angiogenesis in tumors or wound-healing

process (7). Vascular permeability may depend on tumor type (5, 8)and increase with tumor size (9, 10) and growth rate (11). It may behigher in the periphery than in the central region of tumors (12-14).

However, the regulation of tumor vascular permeability is not wellunderstood. Certain brain tumors possess a tight blood-tumor barrier

to the transport of molecules (5, 10, 15); angiogenesis in these tumorsdoes not require the vessels to be hyperpermeable (5).

The present study was designed to address two critical questionsregarding the transvascular transport in tumors: (a) how does molecular size influence transport? and (b) what is the maximum size ofparticles that can cross the tumor vessel wall? The size of moleculesor particles is one of the key determinants of the microvascularpermeability because it can vary by several orders of magnitude. Ourhypothesis is that tumor vessels may lack permselectivity to macro-

Received 7/5/95; accepted 7/21/95.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1This work was supported by National Cancer Institute Outstanding Investigator Grant

R35-CA56591 (to R. K. J.); M. D. and M. L. are recipients of Feodor Lynen Fellowships

of the Humboldt Foundation; D. A. B. is a NRSA fellow (CA59255); and D. F. is aDuPont Merck fellow. This work was presented at the 42nd Annual Meeting of Micro-circulation Society, 1995. and Experimental Biology, 1995.

2 The abbreviations used are: Pv, microvascular permeability; D. diffusion coefficientin water; PEG-DSPE. polyethylene glycol conjugated with distearoylphosphatidyleth-anolamine; Rho-PE, Rhodamine-phosphatidylethanolamine; SCID. severe combined immunodeficient; VPF/VEGF, vascular permeability factor/vascular endothelial growthfactor; PS, permeability-surface area product.

molecules or even liposomes due to the existence of large porestructures (1). Experiments were performed in the human colon adenocarcinoma LS174T implanted in the dorsal skin fold chamber inSCID mice (16). The microvascular permeability to proteins withdifferent molecular weight (25,000-160,000) was measured using anintravital fluorescence microscopy technique (3-5). The maximum

size of pores in vessels of LS174T tumors was estimated by monitoring extravasation of i.v. injected liposomes of different sizes (100-

600 nm in diameter).

Materials and Methods

Tracer Molecules and Liposomes. The microvascular permeability toseven macromolecules was measured (Table 1). The mouse IgG fragments [Fc,Fab, and F(ab')2] were produced by enzymatic digestion and labeled with

indocarbocyanine Cy3 (Jackson ImmunoResearch Laboratories, Inc., WestGrove, PA). The other three tracer molecules were labeled with tetramethyl-rhodamine (Molecular Probes, Eugene, OR). The SDS-PAGE analysis showed

that the molecular weight of IgG fragments (Fc and Fab) were different fromthose reported in the literature (17). This discrepancy may be related to themethod of preparation of these molecules using enzymes. There were twofluorescence bands of F(ab')2 in SDS-PAGE analysis (25,000 and 110,000),

and they were separated by the size exclusion column (Econo-Pac 10DG;Bio-Rad Laboratories, Hercules, CA). The Stokes-Einstein radius of each

molecule was calculated assuming that (a) the radius of albumin is 3.5 nm and(b) the diffusion coefficient is inversely proportional to the cube root of themolecular weight (18). The free fluorescent dye in the solution was removedby passing the solution through the size exclusion columns (Econo-Pac 10DG;Bio-Rad). The final solute concentrations were 1.3 mg/ml for IgG and its

fragments and 6.5 mg/ml for the other three tracers.Sterically protected long-circulating PEG liposomes were prepared by the

detergent (octyl glycoside) dialysis method or by ultrasonication method andfluorescently labeled with membrane-incorporated Rho-PE (MolecularProbes). The molar ratio of egg phosphatidylcholine:cholesterol:PEG-DSPE;Rho-PE was 10:5:0.8:0.1. Lipid mixture was argon-dried from chloroform,

vacuumed, solubilized with octyl glycoside in HBSS (pH 7.4) with a final totallipid concentration of 20 mg/ml, and dialyzed overnight against HBSS at 4°C.

Alternatively, dried lipid mixture was supplemented with HBSS, hydrated for30 min, and briefly sonicated in a bath-type ultrasonicator. Liposomes obtained

were sized by multiple passing through the polycarbonate filters (Poretics,Livermore, CA) with pore diameters of 0.6, 0.4, 0.2, and 0.1 /xm depending onthe target size. The liposome size in the final preparation was determined witha Coulter N4 MD Submicron Particle Size Analyzer (Coulter Electronics,Hialeah, FL). All liposome preparations had a narrow size distribution (95% ofliposomes within a ±15-nm interval for 100- and 200-nm liposomes and a±25-nm interval for 400- and 600-nm liposomes). In addition, surface-modified long-circulating polystyrene latex beads (Sigma Chemical Co., St. Louis,

MO) with similar diameters as liposomes were prepared. To modify the beads,PEG-DSPE and Rho-PE were mixed in chloroform, argon-dried, vacuumed,

and supplemented with latex suspension in HBSS. The weight ratio of latex:PEG-DSPE:Rho-PE was 1:10:0.1. The mixture was briefly sonicated and thenstirred overnight at the room temperature. Surface-modified beads were separated from micelles of free PEG-DSPE and Rho-PE by centrifugation and

washing with HBSS.Measurement of Molecular Weight and Charge. The molecular weight

and charge of proteins used were analyzed by SDS-PAGE (Mini-Protean II

3752

Research. on January 16, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

VASCULAR PERMEABILITY IN A HUMAN TUMOR XENOORAFT

Table 1 Characteristics of tracer molecules and the corresponding microvascular permeability in tumors

NameMouse

FcfragmentMouse

FabfragmentOvalbuminBSAConcanavalin

AMouse

F(ab')2fragmentMouse

IgGM,25,00025,00045,00066.0IX)104,000110,000160,000r

(nm)"2.542.543.083.504.074.154.70P"4.84.84.2-5.24.57.4-8.44.8-5.86.0s/v(mm~/mnv)269(148-376)198

(147-262)255

(220-303)285

(203-392)235

(202-278)218

(171-275)170

(122-254)Median

K(range) (100s)15.8

(13.2-16.0)19.0

(18.1-19.3)6.4

(6.2-8.7)80.5(77.4-130)23.3

(22.6-23.4)60.7

(46.5-105)50.0(33.7-82.2)Median

Pv (range)(10~7cm/s)3.74

(2.65-6.80)4.61

(3.6-6.13)5.77

(4.36-7.07)1.61

(0.65-1.93)1.53

(1.34-3.04)1.51

(0.99-1.57)2.82

(1.47-4.07)Median

PJD (range)(cm'1)0.33

(0.23-0.59)0.40(0.23-0.53)0.61

(0.46-0.75)0.19

(0.08-0.23)0.22

(0.19-0.43)0.22

(0.14-0.23)0.46

(0.24-0.67)

" r, Stokes-Einstein radius; pi, isoelectric point; S/V, vascular surface area per unit vascular volume; K, time constant of concentration decay in the plasma; Pv, microvascular

permeability; D, diffusion coefficient in water.

Electrophoresis Cell; Bio-Rad) and isoelectric focusing electrophoresis (Model111 Mini IEF Cell; Bio-Rad), respectively. The sample buffer for SDS-PAGEwas prepared with or without reducing agent (3-mercaptoethanol becauseß-mercaptoethanol may break the disulfide bonds in IgG and its fragments

(17).Experimental Procedure. The LS174T tumor was grown in the dorsal

skin chamber in SCID mice following the procedure described in Ref. 16.After injection of tracer molecules (0. l ml/25 g body weight), the fluorescenceintensity of the tumor tissue was measured and used to calculate the tumorvascular permeability, as described previously (3—5).The time constant of

concentration decay in the plasma was determined within 30 min after tracerinjections (4). Extravasation of liposomes was observed 24 h after injection ofliposomes.

Mann-Whitney U test was used to compare the differences in permeabilitybetween two groups. Kruskal-Wallis test was used when more than two groups

needed to be compared. The correlation between permeability and molecularweight was checked using the Spearman Rank test.

Results and Discussion

Molecular Weight Dependence. Fig. 1A shows the molecularweight dependence of tumor microvascular permeability, where themolecular size is given as the Stokes-Einstein radius (Table 1). Thevalue of permeability varied approximately 2-fold in the molecular

weight range from 25,000 to 160,000. The value of Pv was comparedamong all molecules, and the P values are given in Table 2. SpearmanRank test indicated that there was no correlation between permeabilityand molecular weight (P = 0.11).

To study the influence of diffusion on the transvascular transport,we normalized the permeability by the diffusion coefficient (D) of themacromolecules in water (Fig. IB), where D was estimated as3.6 X 10~5 (Mr)~'°4 (cm2/s) (18). The value of PV/D was compared

among all molecules, and the P values are given in Table 2. TheSpearman Rank test indicated that there was no correlation betweenPV/D and the molecular weight (P = 0.78). No statistical difference

between PV/D values of Fc, Fab, concanavalin A, and IgG wasobserved by Kruskal-Wallis test (P = 0.18). Even for large particles

[i.e., sterically stabilized liposomes (90 nm in diameter)], the PJDvalue (Fig. IB) obtained from Yuan et al. (4) was comparable withthat of albumin (P = 0.47), where the diffusion coefficient of theliposome was calculated from the Stokes-Einstein equation (18).

Convective transport across the vessel wall of tumors used in ourstudies was likely to be small for the following reasons. These tumorswere sandwiched between the host skin tissue (bottom) and a glass

i

•t.!

E

£

o.u-6.0-Fab1A1

14.0-2.0-0.0-1

n1.U0.8-0.6-0.4

-02-nn1Oval

]ConA¡1-

r ll1~LF(ab')2J

11•

11"*"»**1 'LLipo/

/T1ifB..

45

Radius (nm)

Fig. 1. A, tumor Pv versus the Stokes-Einstein radius of various macromolecules andsterically stabilized liposome (Lipo). Note that the value of permeability varied approximately 2-fold in the molecular weight range from 25,000 to 160,000. Spearman Rank test

indicated that there was no correlation between permeability and molecular weight(P = 0.11). B, ratio of permeability to diffusion coefficient (PJD) versus the Stokes-

Einstein radius of the same macromolecules and liposome. No statistical differencebetween PJD values of Fc, Fab, concanavalin A (Con A), and IgG was observed byKruskal-Wallis test (P = 0.18). The PJD value of liposome was comparable with that ofBSA (P = 0.47). Points, medians; bars, ranges. Oval, ovalbumin.

coverslip (top) (16). Because no fluid could leak out at the top surfaceof tumors, we would not expect the interstitial fluid pressure andoncotic pressure there to be significantly different from that in thecentral region of tumors, where these two pressures in interstitium aresimilar to those in microvessels (19, 20).3 Thus, the convective

transport may be negligible in the top and central regions. In the

3 M. Stohrer, Y. Boucher, and R. K. Jain, unpublished data.

3753

Research. on January 16, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

VASri'LAK f'l RMI ABILITY IN A Hl'MAN Tl'MOR XCNfXiRArT

Table 2 Statistical comparison between Pv or between P^/D (in ¡xirenihtdifferent molecules

'sis) of

FabOvalhuminFc0.465" 0.076

(0.465)(0.028)Fab

0.117(0.016)OvalhuminBSAF(ab'),C'oncanavalin

ABSA0.006

(0.011)0.006

(0.006)0.006

(0.1X16)F(ab')2

Concanavalin AIgG0.009

(0.00'))(UKW

(O.(KW)O.IKW

(O.(KN)0.648

(0.784)0.016

(0.117)0.<K«

(0.076)(UK«

(»UK«)0.465

(0.144)0.347

(11.465)0.117(0.602)0.02X

(0.917)(UHM

(0.076)0.144

(01)06)0.060

(O.IK«)0.2%

(0.076)"P values of Mann-Whitney U lest.

bottom region of tumors, convection is important because the interstitial pressure in the host skin tissue is close to zero (21). However,the fluorescence from the bottom region is significantly attenuated bythe tumor tissue and limited by the depth of light collection of thefluorescence microscope before being detected by the photomultiplier(3). These results, in combination with those shown in Fig. Iß,suggest that the convettive extravasation is minimal, and thus, trans-

vascular transport of therapeutic agents in our measurements is likelyto be limited by diffusion.

The normalized permeability (PV/D) shown in Fig. Ißvaried between 0.19 and 0.61 (cm"1), a 3-fold difference. In particular, the

tumor microvascular permeability to ovalbumin was significantlyhigher than that to other tracer molecules (Fig. Iß).The highervascular leakiness to ovalbumin also occurred in normal tissues, asreflected by the rapid decrease in its plasma concentration. We foundthat the median time constant, K, was 640 s (Table 1), which was 2and 12 times shorter than that of Fc fragment and BSA, respectively.The enhancement of permeability to ovalbumin both in tumor andnormal tissues cannot be explained by the charge effect because thenet charge can only account for approximately 2-fold difference in

permeability if the charge effect on permeability in both normal andtumor tissues is similar to that in the frog mesentery (22). In fact, it islikely that the charge effect becomes less important when vessels areleaky. This is presumably because interactions between molecules andchannels in vascular endothelium are reduced in these vessels. Immune reaction to ovalbumin may cause vascular leakiness, but ovalbumin cannot induce microvascular leakiness in the nonimmunizedanimals (23).

Cutoff Size of Pores in Tumor Vessels. The permeability datashown in Table 1 and Fig. 1 suggest that the permselectivity of tumorvessels to different molecules is less than that of normal vessels(24-26). Our hypothesis for explaining this difference is that large

pore structures exist in the tumor vessel wall (1), through whichmacromoleculcs or even liposomes can pass. These large pores couldbe either open clefts between endothelial cells or channels formed byvesicles/vacuoles (7). Permeability measurements and extravasationobservations in our study do not provide sufficient information todistinguish between these two structures. However, the maximumdiameter of vesicle/vacuole channels in tumors, shown in the literature, is less than 200 nm (27), which is smaller than the maximum sizeof liposomes that can extravasate in LS174T tumors, as will be shownlater. This evidence is still insufficient to exclude the possibility ofvesicle/vacuole channels as a pathway for transendothelial transportof liposomes because the maximum pore size may be tumor dependent. In limited cases, liposomes might be phagocytized by endothelial

cells in tumors (28). However, there is no evidence showing thatliposomes are transported across endothelial cells by large vacuoles.

To estimate the maximum pore size in the LS174T tumor, weinjected sterically stabilized liposomes of different sizes. We foundthat the maximum diameter of liposomes that can escape from thelumen of tumor vessels was between 400 and 600 nm because theliposomes of 400 nm in diameter could penetrate into tumor intersti-

tium, whereas liposomes of 600 nm in diameter were excluded fromextravascular space (Fig. 2). The maximum pore size of the tumorvessel wall was assumed to be the maximum size of liposomes, whichwas around 500 nm as determined above. The validity of this estimation was based on the study of normal liver tissue, where Scherphofet al. (29) demonstrate that the liver sinusoids are permeable toliposomes of 100 nm in diameter, the size that is almost the same asthe median diameter of pores (106 nm) in the sinusoids revealed bythe electron microscopy. Larger liposomes (e.g.. 500 nm in diameter)cannot cross these vessels (29). The cutoff size of the tumor vesselwall was also confirmed by latex beads coated with phospholipids(data not shown). Latex beads are more rigid; they are unlikely todeform when crossing the vessel wall.

Comparison with Other Studies. Despite its clinical importance,there are few data quantifying the relationship between the microvascular permeability and the molecular size of drugs in solid tumors (1).Only the PS has been estimated for selected molecules with differentsizes (30, 31). However, if the surface area of tumor vessels isassumed to be the same for different molecules, PS is equivalent to Pvwhen the relative values are compared. Peterson and Appelgren (30)demonstrate that there is no significant difference in PS for albuminand IgG in chemical-induced sarcomas. Imoto et (il. (31) have measured the extravascular concentration of various radiolabeled macro-

molecules [dextrans (M, 10,000 and 70,000), inulin (M, 5,200), andBSA] in isolated Walker 256 carcinoma implanted in rat ovariantissue. The concentration of these tracer molecules in vessels wasmaintained at a constant level by continuous perfusion. They find thatthe ratios of the interstitial concentrations between small and largedcxtran molecules and between inulin and BSA are 1.6 and 2.0,respectively. Therefore, our results arc qualitatively in agreement withthese studies. In contrast to tumor vessels, the permeability of normalvessels depends significantly on the molecular size (24-26). Simi

larly, the ratio of lymph to plasma concentrations of various moleculesin skeletal muscle drops significantly (5-fold) when the molecular

radius increases from 2 to 4 nm, whereas the same ratio decreases only2-fold in the liver (32). Sinusoids in the liver are known to be

hyperpermeable to macromolecules due to the discontinuity in endothelial junctions and the lack of continuous basement membrane. Thisstructure may be similar to that of tumor vessel wall ( 1). In light of theliver data, our data support the notion that tumor vascular endotheliumis less selective for different macromolecules compared with mostnormal host vascular endothelium.

As stated in "Introduction," tumor microvascular permeability may

depend on tumor size and growth rate, and is in general higher thanpermeability of normal vessels. However, exceptions have also beenreported in the literature. The discrepancy among different studiesmay be tumor related (i.e., each tumor line may have unique mechanisms for regulating its microvascular permeability). Alternatively,common mechanisms of regulation may exist only at cellular ormolecular levels. For instance, VPF/VEGF has been identified, purified, and sequenced (7). It is expressed by several tumor and normalcells, and can function as a mitogen to endothelial cells to induceangiogenesis or as a factor to increase vascular permeability both intumors and in normal tissue (7). In addition to VPF/VEGF, othertissue environmental factors may change the structure of the vesselwall because the vascular permeability of some tumors depends on the

3754

Research. on January 16, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

VASCULAR PERMEABILITY IN A HUMAN TUMOR XENOGRAFT

Fig. 2. Extravasation of sterieally stabilized liposomes in tumor tissue 24 h after injection. A,mean diameter of liposomes *= 400 nm. B, meandiameter of liposomes •=600 nm. Both photos have

the same magnification. Bar, 100 firn. Note that theliposomes of 400 nm in diameter could penetrateinto tumor interstitium as indicated by the brightfluorescence spots, whereas liposomes of 600 nmin diameter were excluded from extravascularspace.

location of tumor transplantation (13). In our previous study, wefound that microvessels of a human glioblastoma HGL21 transplantedinto the cranial window in SCID mice are almost impermeable toLissamin green (M, 577) (5), but they become highly leaky to this dyewhen the tumors grow s.c., although the angiogenesis/vascular densityis similar at two sites.4 Our current working hypothesis is that VPF/

VEGF secretion by HGL21 cells may be reduced in the brain compared to that in s.c. tissue. This hypothesis is supported by ourpreliminary observation that the permeability of HGL21 in the cranialwindow increases significantly after superfusion with VPF/VEGF.

In summary, tumor vessels in general are more leaky and lesspermselective than normal vessels. The vascular permeability in ourstudy is likely governed by diffusion across the vessel wall. Largepores may exist in the tumor vessel wall that allow the penetration ofliposomes up to the size of 400 nm in diameter. Other physical and

4 Unpublished observations.

physicochemical properties such as charge, hydrophobicity, and configuration of therapeutic agents may also influence the permeability(1). The role of these molecular properties needs to be studied in thefuture with an approach similar to that used here.

Acknowledgments

We thank Julia Kahn and Yi Chen for chamber preparations, and Drs.Laurence T. Baxter and Robert L. Dedrick for their helpful comments.

References

1. Jain, R. K. Transport of molecules across tumor vasculature. Cancer Metastasis Rev.,6: 559-593, 1987.

2. Gerlowski, L. E., and Jain, R. K. Microvascular permeability of normal and neoplastictissues. Microvasc. Res., 31: 288-305, 1986.

3. Yuan, F., Leunig, M., Berk, D. A., and Jain, R. K. Microvascular permeability ofalbumin, vascular surface area, and vascular volume measured in human adenocar-cinoma LS174T using dorsal chamber in SCID mice. Microvasc. Res., 45: 269-289,1993.

4. Yuan, F. Leunig, M., Huang. S. K., Berk. D. A., Papahadjopoulos, D., and Jain, R. K.

3755

Research. on January 16, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

VASCULAR PERMEABILITY IN A HUMAN TUMOR XENCXÃŒRAFT

Microvascular permeability and interstitial pcnclratkm of sterically stabilized(Stealth) liposomes in a human tumor xenograft. Cancer Res.. 54: 3352-33.%. 1994.

5. Yuan, F.. Salehi. H. A.. Boucher. Y.. Vasihare. U. S.. Tuma. R. F.. and Jain. R. K.Vascular permeability and microcirculation of gliomas and mammary carcinomastransplanted in rat and mouse cranial windows. Cancer Res.. 5V: 4564-4568, 1994.

h. Wu. N. 7... Kin/in.ni. B.. Rosner, G., Nccdhain. D., and Dewhirst. M. W. Measurement of material extravasation in microvascular networks using fluorescence video-microscopy. Microvasc. Res., 46: 231-253, 1993.

7. Dvorak, H. F., Brown, L. F., Detmar, M., and Dvorak, A. M. Vascular permeabilityfactor/vascular endothelial growth factor, microvascular hypcrpermeability. and an-giogenesis. Am. J. Pathol.. I4f>: 1029-1039, 1995.

8. Sands, H., Jones. P. L. Shah. S. A., Palme, D.. Vessella. R. L...and Gallagher. B. M.Correlation of vascular permeability and blood How with monoclonal antibody uptakeby human Clouser and renal cell xenografts. Cancer Res., 4X: 188-193, 1988.

9. Blasherg. R. G., Kobayashi, T., Patlak. C. S., Shinohara, M.. Miyoaka. M., Rice,J. M., and Shapiro, W. R. Regional blood How, capillary permeability, and glucoseutili/ation in two brain tumor models: preliminary observations and pharmacokineticimplications. Cancer Treat Rep.. 6.5 (Suppl. 2): 3-12, 1981.

10. Zhuang. R-D., Price. J. E., Fujimaki, T.. Bucana. C. D., and Fidler, I. J. Differentialpermeability of the blood-brain barrier in experimental brain métastasesproduced byhuman neoplasms implanted into nude mice. Am. J. Pathol.. 141: \ 115-1124. 1992.

11. Heuser. L. S.. and Miller. Y. N. Differential macromolecular leakage from thevasculature of tumors. Cancer (Phila.). 57: 461-464. 1986.

12. Vriesendorp, F. J., Peagram, C.. Bigner. D. D., and Groothuis. D. R. Concurrentmeasurementsof blood How and transcapillary transport in xenotransplanted humangliomas in immunosuppressed rats. J. Nati. Cancer Inst.. 79: 123-130, 1987.

13. Hasegawa.H.. Ushio, Y., Hayakawa, T.. Yamada. K., and Mogami. H. Changesof theblood-brain harrier in experimental metastatic brain tumors. J. Neurosurg.. 5^: 304-310, 1983.

14. Dvorak. H. F., Nagy. J. A.. Dvorak. J. T., and Dvorak. A. M. Identification andcharacteri/alion of the blood vessels of solid tumors that are leaky to circulatingmacromolcculcs. Am. J. Pathol.. 133: 95-109. 1988.

15. Schackert. G., Simmons. R. D., Buzbcc, T. M.. Hume, D. A., and Fidler. I. J.Macrophage infiltration into experimental brain métastases:occurrence through anintact blood-brain barrier. J. Nati. Cancer Inst., HO: 1027-1034, 1988.

16. Leunig, M.. Yuan. F.. Mcnger, M. D.. Boucher. Y.. Goetz. A. E.. Messmer. K.. andJain. R. K. Angiogenesis, microvascular architecture, microhcmodynamics. and interstitial fluid pressure during early growth of human adenocarcinoma LS174T inSCID mice. Cancer Res.. 52: 6553-656(1. 1992.

17. Colcher. D.. Bird, R., Roselli. M.. Hardman, K. D., Johnson, S., Pope, S.. Dodd.S. W.. Pantoliano. M. W.. Milenic. D. E.. and Schlom. J. In vnv>tumor targeting ofa recombinant single-chain antigen-binding protein. J. Nail. Cancer Inst.. #2: 1191-1197, 1990.

18. Berk, D. A., Yuan, F., Leunig. M.. and Jain. R. K. Fluorescencephotobleaching with

spatial Fourier analysis: measurementof diffusion in light-scattering media. Biophys.J., 65: 2428-2436, 1993.

19. Jain. R. K.. and Baxter. L. T. Mechanisms of heterogeneous distribution of monoclonal antibodies and other macromolecules in tumors: significance of elevatedinterstitial pressure. Cancer Res.. 4K: 7022-7032. 1988.

20. Boucher. Y.. and Jain. R. K. Microvascular pressure is the principle driving force forinterstitial hypertension in solid tumors: implication tor vascular collapse. CancerRes., 52: 5110-5114, 1992.

21. Boucher, Y, Baxter, L. T., and Jain. R. K. Interstitial pressure gradients in tissuc-isolated and subcutaneous tumors: implications for therapy. Cancer Res.. 50: 4478-4484, 1990.

22. Adamson. R. H.. Huxley. V. H.. and Curry. F. E. Single capillary permeability toproteins having similar size hut different charge. Am. J. Physiol., 25V: H304-H312,1988.

23. Adamski, S. W.. Langone, J. J.. and Grega. G. J. Modulation of macromolecularpermeability by immune complexes and a ß-adrenoceptorstimulant. Am. J. Physiol..253: H1586-H1595. 1987.

24. Rippe, B., and Haraldsson. B. Transport of macromolecules across microvascularwalls: the two-pore theory. Physiol. Rev., 7V: 163-219. 1994.

25. Dedrick, R. L.. and Flessncr. M. F. Pharmacokinelic considerations on monoclonalantibodies. In: M. S. Mitchell (ed.). Immunity to Cancer, Vol. 2, pp. 429-438. NewYork: Alan R. Liss, 1989.

26. C'urry. F-R. E. Regulation of water and solute exchange in microvascular endothe-

lium: studies in single perfused capillaries. Microcirculation, /: 11-26, 1994.27. Ou-Hong, Nagy. J. A.. Scngcr. D. R., Dvorak, H. F., and Dvorak. A. M. Ultrastruc

tural localization of vascular permeability factor/vascular endothelial growth factor(VPF/VEGF) to the abluminal plasma membrane and vesiculovacuolar organdíesoftumor microvascular endothelium. J. Histochem. Cytochem.. 43: 381-389. 1995.

28. Huang. S. K.. Martin. F. J., Jay, G., Vogel, J., Pupahadjopoulos, I)., and Friend. D. S.Extravasation and transcytosis of liposomes in Kaposi's sarcoma-like dermal lesions

of transgcnic mice hearing the HIV tal gene. Am. J. Pathol., 143: 10-14, 1993.29. Scherphof. G.. Roerdink. F.. Dijkslra, J., Ellens, H., de Zanger. R., and Wisse, E.

Uptake of liposomes by rat and mouse hepatocytes and Kupffer cells. Biol. Cell. 47:47-57, 1983.

30. Peterson, H-I.. and Appelgren. L. Tumour vessel permeability and transcapillaryexchange of large molecules of different size. Bihl. Anal.. IS: 262-265, 1977.

31. Imoto, H., Sakamura, Y., Ohkouchi, K.. Atsumi, R.. Takakura, Y., Sezaki, H., andHashida, M. Disposition characteristics of macromolecules in the perfused tissue-isolated tumor preparation. Cancer Res., 52: 4396-4401. 1992.

32. Taylor, A. E., and Granger, D. N. Exchange of macromolecules across the microcirculation. In: E. M. Renkin and C'. C. Michel (eds.). Handbook of Physiology: The

cardiovascular system. Volume 4. Section 2. pp. 467-520. Bethcsda, MD: AmericanPhysiology Society, 1984.

3756

Research. on January 16, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

1995;55:3752-3756. Cancer Res   Fan Yuan, Marc Dellian, Dai Fukumura, et al.   Size Dependence and Cutoff SizeVascular Permeability in a Human Tumor Xenograft: Molecular

  Updated version

  http://cancerres.aacrjournals.org/content/55/17/3752

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/55/17/3752To request permission to re-use all or part of this article, use this link

Research. on January 16, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from