Microvascular Architecture of Experimental Colon Tumors in ... · Tumor Induction. Colon tumors were induced by s.c. injections of dimethylhydrazine dihydrochloride (Sigma Chemical

[CANCER RESEARCH 50. 2411-2417, April 15. 1990]

Microvascular Architecture of Experimental Colon Tumors in the Rat1

Stewart A. Skinner, Peter J. M. Tutton,2 and Paul E. O'Brien1

Monash University, Department of Surger)', Alfred Hospital, Prahran, Victoria, Australia 3181

ABSTRACT

Tumor cell proliferation is dependent upon concurrent growth of asupporting vasculature. This study aims to characterize the structuralfeatures of the microvasculature within a primary tumor model. Therewere 22 colon tumors induced in 16 rats by repeated administration ofdimethylhydrazine. A cast of the microvessels was prepared by intraar-terial administration of acrylic resin (Mercox). After corrosion of thetissue, the cast was examined by scanning electron microscopy.

Tumors 2.6 to 12.0 mm in diameter were examined. Within polypoidcarcinomas up to 5.7 mm in diameter, there were two distinct vascularzones, a luminal vascular zone continuous with the vasculature of normalmucosa and a central zone continuous with the normal submucosa andmuscularis propria vessels. Within both vascular zones, the organizationof microvessels had the same general pattern as in normal mucosa.However, in tumors with diameters >5.7 mm, the vasculature was seento be disorganized and of a greater density than normal.

In the smallest tumors, few morphological changes were seen in theindividual microvessels when compared to normal. However, with tumorgrowth, there was elongation and increased diameters of the microvesselswithin the tumor. Microvessels within the luminal zone of the tumorswhich could definitely be traced to veins had diameters of 50 to 100 urn(compared to 12 to 30 urn for normal venules). Individual microvesselsvaried in diameter along their course forming saccular dilations in places.Networks of frequently anastomosing microvessels were formed. Extravasation of resin occurred from some microvessels. Elongated vessels ofuniform diameter which travel distances up to 2 mm without branchingwere seen and were probably arterioles.

These appearances indicate that there are two distinct stages ofdevelopment of the vasculature within primary tumors, an early phasewhere the tumor is supplied by the preexisting host microvessels, followedby a phase of proliferation of new vessels with abnormal morphologicalcharacteristics.

INTRODUCTION

Tumor cell proliferation is dependent on the concurrentgrowth of a supporting vasculature (1). In their studies oftransplantable tumors, Gimbrone et al. (2) found that, after aninitial prevascular phase during which tumor cells will proliferate, growth of an avascular tumor will stop when the tumordiameter reaches 1-2 mm. Further growth in the tumor cellpopulation must be preceded by an increase in the microvesselsthat supply the tumor. This neovascularization has a markedeffect on the rate of growth of the tumor. The development ofthe vasculature is also associated with the capacity of the tumorto metastasize (3). The stimulus for the development of themicrovasculature appears to arise from the tumor itself (4) anda number of soluble factors, derived from both neoplastic andnonneoplastic tissues, are capable of inducing angiogenesis inexperimental systems (5, 6).

Studies of the microvasculature within normal tissues haveshown that the microcirculation has an important role in thespecialized functions of that tissue and that the morphological

Received 12/1/88; revised 11/27/89.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 inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1This study was supported in part by a grant from the Anti-Cancer Council ofVictoria.

3 Present address: Department of Anatomy. Monash University, Clayton,Victoria, Australia.

3To whom requests for reprints should be addressed.

241

features and spatial arrangement of the microvessels haveadapted to facilitate those functions (7-9). Furthermore, thevessels are organized in a manner that results in efficientdelivery of oxygen and nutrients to the tissue.

The morphology and spatial arrangement of the tumor vasculature have been studied almost exclusively in transplantabletumors. These studies describe a chaotic or disorderly array ofcapillaries within the substance of the tumor implant that issupplied by dilated arterioles and drained by large veins in thesurrounding host tissues (10, 11). Capillaries within the tumorare tortuous and sinusoidal with diameters up to 200 ^m. Theyform a dense vascular network which is profusely anastomotic(11). Thin and tapered capillaries were also noted at the edgeof the necrotic center of the tumor (12). The sequential development of the microvasculature in tumor implant models hasbeen described (11, 13) and indicates that while the tumor isnourished by the host vasculature initially, there is progressivedilation and tortuosity of the host venules with bulbous formations along their length. This is followed by sprouting ofendothelial buds from these venules and from capillaries toeventually form a dense network of capillaries. With growth ofthe tumor implant, the relative vascular volume decreases andthe center of the tumor becomes avascular and necrotic (10,12).

Little is known of the characteristics of the individual micro-vessels and their spatial organization within primary in situtumors. This has been noted by other authors (14). Furthermore, primary tumors arise within tissues which already havetheir own vasculature. Therefore, there are two possible sourcesof vascular supply within primary tumors. They may receivetheir vascular supply (a) by modification of the preexisting hostvascular bed or (b) by the development of new vessels (angiogenesis). The relative role of each of these possible sources ofvascular supply during tumor growth is not established. Therefore, the aims of this study are to determine the origin of thevascular supply within an experimental primary tumor and tocharacterize the spatial arrangement and morphological features of the microvessels within a primary tumor model.

MATERIALS AND METHODS

Animals. Male Sprague-Dawley rats (400-500 g) were housed in atemperature-controlled room with 12 h of light daily. They were fedClark King GR2+ pellets and water ad libitum.

Tumor Induction. Colon tumors were induced by s.c. injections ofdimethylhydrazine dihydrochloride (Sigma Chemical Co., St. Louis,MO), at a dose of 20 mg/kg. Rats received weekly injections from theage of 4 weeks for a total of 24 weeks.

Tumor Detection. Following completion of dimethylhydrazine injections, colonoscopy was performed weekly to detect the presence oftumors. Under light ether anesthesia, the colon was examined using aflexible nasopharyngoscope (Fujinon NAP-F, Fuji Photo Optical Co.,Ltd., Japan). The site, size, and number of colon tumors were recorded.

Vascular Casting. Vascular casts were prepared at least 24 h aftercolonoscopy. Each rat was anesthetized with methohexitone sodium(Eli Lilly & Co., New South Wales, Australia), 50 mg/kg i.p. The renalpedicles on both sides were then clamped through loin incisions. Thethoracic aorta was exposed and a cannula (internal diameter, 1.57 mm)was inserted caudally and secured with ties. Blood was then washed outfrom the vascular system by infusing a filtered 0.154 M NacÃsolution

1

Research. on September 26, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

MICROVASCULATURE OF COLON TUMORS

containing heparin, 10.000 units/liter (Weddel Pharmaceuticals, NewSouth Wales, Australia), l Â¿IMpapaverine (David Bull Laboratories,Victoria, Australia), and polyvinylpyrrolidone 40, 60 mg/liter (Sigma)at 37-40Â°Cthrough the aortic cannula at a pressure of 150 Â±5 (SD)

mm Hg. An incision of the right atrium allowed free efflux of thewashout. When the effluent from the right atrium was clear (approximately 3-4 min after commencing the infusion) an acrylic resin solution, prepared by mixing 15 ml of Mercox CL-2B, 0.1 g of Catalyst-MA (Vilene Med. Co., Japan), and 5 ml of methylmethacrylate monomer (BDH Chemicals, Poole, United Kingdom) in a 20-ml syringe,was infused via the aortic cannula. The mean intraarterial pressure wasmeasured in the first 5 rats via a cannula in the femoral artery. It wasfound that when a pressure of 200 mm Hg was applied to the castingmedium in the syringe, this resulted in a mean arterial pressure in therat of 110 Â±5 mm Hg. This reflects normal mean arterial pressure inthe rat. When the resin was flowing from the right atrium, the aorticcannula and the inferior vena cava were clamped. The resin was allowedto fully polymerize at room temperature overnight. On the followingday, the colon was excised and tumors were identified. Each tumor wasthen divided in half through its pedicle and in continuity with thesurrounding colon.

One half of the tumor was fixed in 10% formalin overnight (12-18h) for subsequent histological examination. The other half of the tumorwas placed in a solution of 3.5 M KOH to digest the tissue away fromthe vascular cast. The resultant cast was dried at room temperature,mounted on an aluminum stub, gold coated, and examined in a scanningelectron microscope (Cambridge; Stereoscan 250). The cast was photographed and examined using stereopairs. Measurements were madeusing the automatic micrometer scale at the bottom of the photographs.

Histology. The tumor half fixed in formalin was processed for histology by dehydrating in a graded series of ethanol solutions (50%absolute). It was then embedded in plastic (JB4 embedding medium)

Fig. 1. Normal rat colon in cross-section. Scanning electron micrograph ofvascular cast. Bar, 40 urn. The vasculature within the mucosa and submucosa isdisplayed. Within the submucosa, arteries (A) and veins (K) are seen runningtogether in parallel. Arteriolcs (a) divide at submucosal level into capillarybranches. Capillaries drain into venulcs (>â€¢)at the luminal aspect of the mucosaonly. Venules then pass to submucosal veins without receiving any furthercapillary branches through the mucosa.

Fig. 2. Normal rat colon viewing the luminal aspect. Scanning electron micrograph of vascular cast. Bar. 100 Â¡im.Mucosa! capillaries form a regular honeycomb-like plexus around openings of mucosa! glands. Capillaries are quite uniform in their diameters (5-8 Mm)-

using a Polaron LM embedding kit (Polaron Equipment, Ltd., Watford,United Kingdom). Sections (5 firn) were cut and stained with hematox-ylin and eosin.

Colon Tumors. Vascular casts of 22 tumors in 16 rats were prepared.There were 17 polypoid tumors the histologies of which revealedmoderate or well-differentiated adenocarcinomas, 4 benign polyps, andone sessile adenocarcinoma.

RESULTS

The organization of the microvasculature within the normalrat colon has previously been characterized in detail (8). In thisstudy, casts of the colon were prepared in 4 normal rats andour findings support those of Browning and Gannon (8).Briefly, it was found that in the mucosa of normal rat colon thespatial arrangement of the microvessels had a consistent patternthroughout the colon. Arterioles do not pass beyond the submucosa (Fig. 1). They branch at the border between mucosaand submucosa to supply the mucosal capillary network whichdrains into venules at the luminal surface only. These venulespass to submucosal veins without receiving further branchesthrough the mucosa (Fig. 1). Capillaries within the mucosa arearranged in a regular honeycomb-like pattern around the mucosal glands (Fig. 2).

The polypoid adenocarcinomas ranged in diameter from 2.6to 12.0 mm. The spatial organization of the microvessels in allof these tumors retained many features in common with thenormal colon. In all tumors, two distinct vascular regions couldbe defined, but this feature was most obvious in the smallertumors. There was a central region of larger vessels that werecontinuous with the vessels of the submucosa and muscularispropria layers of the adjacent normal colon. On the luminalaspect of this, there was an outer region of nutrient microvesselsthat were continuous with the mucosal layer of adjoining normal colon. This pattern is best seen in Figs. 3 and 5.

In the outer vascular zone of polypoid tumors up to 3.5 mmin diameter, the organization of microvessels had the same

2412




general pattern as seen in normal rat colonie mucosa (Fig. 4).As in normal mucosa, capillaries in the outer zone of the tumorswere supplied by arterioles that pass along the submucosa anddivide into their branches at the border between the two vascularzones (Fig. 4). Capillaries then drain into venules mainly at theluminal surface. Within the central vascular zone of tumors upto 3.5 mm in diameter, tortuous, large diameter veins were themost obvious feature (Fig. 4). However, the spatial arrangementof the microvessels in this zone was also similar to that ofnormal colon.

Few changes in the morphological features of the individualmicrovessels were seen in tumors less than 3.5 mm in diameter.All microvessels were elongated but there was no apparentchange in the diameter of arterioles which ranged from 15 to25 urn. Capillaries had slightly greater diameters than normal(5-20 firn compared to normal, 5-8 urn), while venules wereobviously of greater diameters than normal with a range of 15-70 /Â¿m(normal, 12-30 urn) (Fig. 4).

In the tumors >5.7 mm in diameter, the vasculature lackedfeatures of the orderly structure seen in normal colon. Thisdisorderly structure was characterized principally by variationsin size of vessels, variations in direction, and increased numbersof interconnections without a perceptible pathway of flow (Fig.8). The pattern was seen most prominently on the luminalsurface and within the luminal zone of the tumor. Althoughquantitative studies have not been done, it gives the appearanceof greater vascular density than the vasculature of normal colonand of smaller tumors. In the central vascular zone, the micro-vessels also showed a disorderly structure with marked variationin vessel diameters, irregularity in contour, and variations indirection.

In tumors >3.5 mm in diameter, a number of characteristicmorphological features of the microvessels were seen. Thediameters of the microvessels within the luminal vascular zone

Fig. 3. Well-differentiated adenocarcinoma of rat colon in cross-section. Scanning electron micrograph of vascular cast. Bar. 400 ^m. Low magnification viewof tumor 3.1 mm in diameter, demonstrating the gradual change in the vasculaturcof normal colon as it becomes tumor vasculature. Within the tumor, two vascularzones can be distinguished: an outer zone continuous with the normal mucosalvasculature; and a central zone continuous with the vasculature of the submucosaand muscularis propria of the normal colon.

Fig. 4. Well-differentiated adenocarcinoma of the rat colon in cross-section.Scanning electron micrograph of vascular cast of 2.6-mm-diameter tumor. Bar.200 ^m. This micrograph shows detail of the junction between tumor and normalcolon vasculature. Within the outer vascular zone of the tumor, when comparedto normal, all microvessels are elongated, but capillaries have diameter similar tonormal ones (5-20 Mm). However, venules have greater diameters, up to 75 ^m.Capillaries are supplied by arterioles (arrows) running along the border betweenouter and central zones. These tumor arterioles are of a caliber similar to that ofnormal (15-25 firn). Marked dilation and tortuosity of veins (I") in the central

zone are also seen.

of the tumors varied over a wide range (5-100 urn) and ingeneral were much greater than in the mucosa of normal colon(Fig. 5). The majority of microvessels ranged in diameter from5 to 50 urn, while microvessels 50 to 100 urn in diameter wereless frequent and often could be traced to veins. In all tumorsthat were approximately >3.5 mm in diameter, extravasationof the casting medium was seen to occur from the microvesselsin the luminal zone. The greater proportion of extravasationoccurred from microvessels near the surface of the tumor, sothat resin was extruded into the lumen of the colon (Fig. 6).Smaller amounts of extravasation of resin could also be seenwithin the depth of the tumor. Many vessels were irregularalong their length, varying frequently in diameter and givingrise to saccular dilations (Fig. 7). In places, microvessels anastomosed repeatedly with each other and formed a network offrequently anastomosing vessels (Fig. 9). Elongated microvessels of uniform diameters (40-50 Â¿Â¿m)that travel distances upto 2 mm before giving any branches were seen and were probablyarteriolar (Fig. 10).

Histopathological examination showed 17 of the polypoidtumors to be well or moderately well differentiated adenocar-cinomas. There was no correlation between the degree of differentiation of the tumor and the nature of the vascular changes.Because of the presence of the casting resin within the lumen,the blood vessels were easy to identify in the histolÃ³gica! sections. An abrupt transition between normal colon vasculatureand abnormal tumor vasculature could be seen within thevascular casts of the larger tumors (Fig. 5). Examination of thehistolÃ³gica! sections of the corresponding half of the tumorsrevealed that the abrupt transition between the normal mucosalepithelial cells and the tumor cells correlated directly with theabrupt changes seen in the vasculature (Fig. 11). Within the

2413




Fig. 5. Moderately differentiated adenocarcinoma of rat colon. Cross-sectionof polypoid tumor 5.7 mm in diameter. Bar, I mm. In this tumor there is obviousextravasation of resin from surface microvessels into the lumen of the colon(arrows). Two distinct vascular zones within the tumor, continuous with thevasculature of normal colon, can still be distinguished. In contrast to Fig. 3, allvessels within the outer /.one of the tumor have greater diameters than normal.Within the central /one in this tumor, the microvessels are arranged in basket-like plexuses. Arrow, point of abrupt transition from normal colon vasculature totumor vasculature at one edge of the tumor. This corresponds to the point ofclear demarcation between tumor cells and normal colon mucosa seen on histolÃ³gica!sections.

Fig. 6. Same tumor as Fig. 5. Bar, 50 Â¿im.This view of the surface of thetumor shows that the regular honeycomb pattern seen on the surface of normalcolonie mucosa (see Fig. 2) is no longer present. Instead vessels pass to thesurface, loop over, and pass back into the tumor, after giving an irregular numberof branches to adjacent microvessels. Extravasation of resin can be seen in thebackground (arrow).

Fig. 7. Same tumor as Fig. 5. Bar, 40 >im. Scanning electron microscopy ofvascular cast demonstrating microvessels on the surface of the tumor. Vesseldiameters range from 5 to 50 Â»mand thus are wider and more varied in diameterthan normal colonie microvessels. Variations in calibre and frequent sacculardilations can be seen along their course.

Fig. 8. Well-differentiated adenocarcinoma of rat colon. Bar, 200 /<m. Thisview of the vasculature on the surface of this I0-mm-diameter tumor demonstratesthe disorderly structure of microvessels seen in these larger tumors. There isvariation in size and direction of microvessels and the appearance of greatervascular density.

smaller tumors (up to 3.5 mm in diameter) there was obviousdilation of veins in the region of the submucosa invaded bytumor cells, but vascular changes within the tumor itself werenot obvious on the histological sections, nor were there anyvascular changes seen within the surrounding normal colon.

Four tumors were classified histologically as benign polyps.2414




Fig. 9. Same tumor as Fig. 8. Bar, 200 firn. At a higher magnification thismicrograph shows that the diameters of the microvessels vary over a wide range(5-100 urn). Each vessel branches frequently and in places forms a network offrequently anastomosing vessels (arrow). The disorganization of the vasculatureis seen.

Fig. 10. \\ ell-differentiated adenocarcinoma of rat colon. Cross-sectional \ ie\\of 10-mm-diameter tumor. Bar, 400 ><m.A characteristic specific to larger tumorsis the presence of elongated vessels of uniform diameters (40-50 /Â¿m)which canbe traced for I or 2 mm before giving any branches (arrows). They are probablyarteriolar.

Since there may have been malignant invasion in the half of thetumor that was corroded to examine the vascular cast, thisclassification is not conclusive. Nevertheless, there appeared tobe no difference in the vascular structure of these tumors whencompared to the carcinomas.

The one remaining tumor was a large well-differentiatedadenocarcinoma 14 mm in diameter (Fig. 12). It differed mor-

Fig. 11. Cross-section of well-differentiated adenocarcinoma of the rat colon.4.5 mm in diameter. This light micrograph shows the clear point of transitionfrom normal colon mucosa to tumor epithelium (arrow). Numerous blood vesselscan be identified within the tumor as clear spaces (filled by the casting resin) linedby cndothelium (arrowheads). H & E, x 40.

Fig. 12. Well-differentiated adenocarcinoma of rat colon. Bar. I mm. This14-mm-diameter tumor is different morphologically from all others examined, asthe bulk of the tumor is invading and expanding the muscularis propria. A largeavascular zone (x) in the center of the tumor corresponds to an area of nccrotictissue. On the luminal aspect of the tumor there is a vascular zone continuouswith the normal mucosa! layer (arrows). Its vasculature has an arrangementsimilar to that of normal mucosa. This layer then overlies a zone of disorganizedvasculalure which surrounds the avascular center.

phologically from the other tumors in that the bulk of the tumorhad invaded and expanded the muscularis propria. A centralpedicle of large diameter vessels had not developed as in othertumors, and it was the only tumor in which an avascular areahad occurred. On the luminal surface of the tumor there was alayer of microvessels continuous with the normal mucosa! layer

2415




which had an organizational structure similar to that seen innormal mucosa.

DISCUSSION

The majority of the studies of the tumor microcirculationhave been of transplantable tumors observed in transparentchambers in rodents (11, 13, 15-17). This technique allowsdirect visualization of the microcirculation in vivo over a periodof time. However, the tumor can only grow in a lateral directionand is limited in size by being compressed by the chamber.Furthermore, the tumor is studied in an implantation site oftenremote from its tissue of origin and the tumor cell line has beenpassaged many times so that its morphological and proliferativecharacteristics may be quite different from those of the originaltumor (18). In contrast, we have studied primary tumors thatcan be related to their tissue of origin.

The microvasculature of dimethylhydrazine-induced colontumors in rats (19) and human colorectal tumors (20) havepreviously been studied using histolÃ³gica! techniques to obtainmorphometric data of the degree of vascularization within thesetumors. These techniques, however, do not demonstrate thespatial organization of the vasculature within these tumors;neither do they reliably demonstrate all vessels within the tissue.Furthermore, it was not the aim of these studies to demonstratethe development of the vasculature within these tumors. In thisstudy microvascular casting was used since it will demonstratethe spatial organization and structural features of the microves-sels within the tumor and will reliably cast the smallest micro-vessels due to the low viscosity of the casting resin.

In small tumors, the spatial arrangement of the microvesselshad the same general pattern as the vasculature of normal colonand individual microvessels had similar morphological featuresbut differed from normal colonie microvessels in being elongated and dilated. With tumors of increasing size, changes areseen in the spatial arrangement of the tumor vasculature andin the morphological features of the microvessels. Eventuallyin the largest tumors the vasculature is completely disorderlyand the individual microvessels are characterized by a numberof abnormal morphological features.

These appearances indicate that there are two distinct stagesof development of the vasculature within primary tumors. During the early phase of growth, the tumor is supplied by thepreexisting host microvessels. These microvessels are then modified by the tumor but retain a similar organization to that inthe normal host vascular bed. Then with further growth thetumor induces proliferation of new microvessels. These newvessels are clearly different in morphology and spatial arrangement from the host microvessels and would appear to be characteristic of tumor-induced microvessels. This Finding in primary tumors correlates with the findings of many studies oftumor implants which have also shown that the tumor implantis supplied both by vessels recruited from the preexisting hostvasculature and by new vessels resulting from tumor-inducedangiogenesis (11, 13, 16, 21, 22).

Folkman et al. (1) were able to demonstrate that tumor cellproliferation is dependent on the concurrent growth of a supporting vasculature and that tumors induce angiogenesis fromhost vessels by release of soluble angiogenic factors. A numberof angiogenic factors that are secreted by tumor cells, endothe-lial cells, and macrophages or that are stored in the interstitialmatrix have recently been identified. The role of each individualfactor and how they are interrelated is still to be determined(5). Some of these factors stimulate endothelial motility, others

induce endothelial mitosis, and others cause both (23). Ourresults indicate that at least two types of angiogenic responsesby the vasculature occur with the presence of tumor cells. Inthe early stages there is elongation and dilation of existing hostmicrovessels, and at a later stage there is proliferation of newmicrovessels. It is possible that different angiogenic factors areacting separately in each of these two stages of development ofthe tumor vasculature.

Previous studies of the development of the vasculature oftransplantable tumors showed that, after a phase of vascularproliferation and ingrowth of vessels into the expanding tumormass, there is a further stage of blood vessel rarefaction andreduced vascular volume within the tumor (15, 24). Interestingly, in this study in only one tumor was an avascular zoneseen.

Areas of hemorrhage within experimental tumors have beennoted in several previous studies (10, 25), and carcinoma of thecolon is well known to cause blood loss into the lumen of thecolon (26). In this study, extravasation of the casting mediumwas seen to occur from microvessels on the luminal surface andwithin the depth of tumors >3.5 mm in diameter. This extravasation of resin may be the result of increased permeability ofthe tumor microvessels or may be due to hemorrhage. It isunlikely, however, that the resin is able to pass through endothelial cell pores or interendothelial cell junctions because ofits physical properties. A large break in the integrity of themicrovessel wall must precede resin extravasation. Therefore,this resin extravasation probably represents hemorrhage fromthe tumor microvessels.

The mechanisms responsible for hemorrhage in tumors areunknown although a number have been postulated. Endrich etal. (27) observed that hemorrhage occurred from enlarged veinsand venules at the periphery of tumor implants. These vesselswere thought to be mechanically fragile and damaged by theadvancing tumor. Others have noted hemorrhage to occur nearthe necrotic center of tumor implants which is the furthestpoint from the arterial supply (10, 25). This may imply thatnecrotic tissue leads to damage to microvessels in this area insome way. In a further study, Endrich et al. (28) observedintermittent blood flow rates in different areas of tumor implants and noted that petechial hemorrhages developed in areaswhere flow was restored after slow or zero blood flow, thusimplying hypoxic damage to microvessels. In this study therewas no evidence of necrosis in any of the polypoid tumors >3.5mm in diameter, despite the presence of resin extravasation.Van Den Brenk et al. (29) found that extravasation of erythro-cytes occurs through the tips of capillary sprouts which developin the process of tumor-induced angiogenesis. Ausprunk andFolkman (13) further demonstrated that during migration ofendothelial cells to form capillary sprouts, there was looseningof the cell junctions and dissolution of the basal lamina forminggaps in the blood vessel wall through which erythrocytes hadextravasated. The association of hemorrhage with areas ofvascular proliferation within tumors has also been noted byother workers (21) and is thought to be due to the presence ofnewly formed, immature blood vessels that are fragile and proneto rupture. Our results would support this, since resin extravasation was seen only from tumors in which there was clearevidence of the development of new blood vessels with abnormalmorphological features.

Increasing tumor cell hypoxia and cell death with tumorgrowth has been a consistent finding in many studies (15, 28).Deficiencies of the tumor vasculature are believed to be responsible and a number of mechanisms have been postulated (30).

2416




Although an avascular zone with corresponding tumor cellnecrosis was present in only one tumor in this study, a numberof possible inadequacies of the vasculature can be identified: (a)large diameter vessels are less efficient as exchange vesselsbecause they have a lower surface area:volume ratio (31, 32);(b) although the outer vascular zone of the tumor is muchthicker than normal mucosa, arterioles supplying this zone inthe smallest tumors (<3.5 mm) appear not to increase theircaliber in response to the presence of the tumor cell mass inthe same way that capillaries and venules do. In larger tumors,elongated vessels which traverse relatively long distances beforebranching appear to be arteriolar and may be expected to havea greater resistance compared to normal arterioles; (c) increasedvascular resistance within tumor microvessels may also resultfrom the frequent irregularities seen along the course of thevessels and from their irregular branching pattern.

This study has confirmed that the developing vascular systemwithin primary tumors has two components (22). This hasimplications in two main areas: (a) current models for examining the tumor microcirculation using tumor implants in thecornea or in transparent chambers will demonstrate the development of abnormal tumor-induced microvessels but do notdemonstrate the early stage where the tumor is supplied principally by the preexisting host microvessels; (b) any study of atherapeutic approach which aims to control primary tumorgrowth through its direct action on the tumor microvesselsmust consider its effect on each of the two components of thevascular supply of the tumor.

REFERENCES

1. Folkman, J. Tumor angiogenesis. Adv. Cancer Res., 43: 175-203, 1985.2. Gimbrone, M. A., Leapman, S. B.. Colran. R. S., and Folkman, J. Tumor

dormancy in vivo by prevention of neovascularization. J. Exp. Med., 136:261-276. 1972.

3. Liotta. L. A., Kleinerman. J.. and Saidel, G. M. Quantitative relationshipsof intravascular tumor cells, tumor vessels, and pulmonary mÃ©tastasesfollowing tumor implantation. Cancer Res., 34: 997-1004, 1974.

4. Folkman. J., and Cotran, R. Relation of vascular proliferation to tumorgrowth. Int. Rev. Exp. Pathol., 16: 207-243, 1976.

5. Folkman. J.. and Klagsbrun. M. A family of angiogenic peptides. Nature(LonA.).329: 671-672, 1987.

6. Schor, A. M., and Schor, S. L. Tumor angiogenesis. J. Pathol.. 141: 385-413, 1985.

7. Gannon. B., Browning. J.. O'Brien. P.. and Rogers, P. Mucosal microvascular

architecture of the fundus and body of human stomach. Gastroenterology.9Â«:866-875. 1984.

8. Browning. J.. and Gannon. B. Mucosa] microvascular organisation of the ratcolon. Acta Anat.. 126: li-li, 1986.

9. Gannon. B., Browning. J.. and O'Brien. P. The microvascular architectureof the glandular mucosa of rat stomach. J. Anal., 135: 667-683, 1982.

10. Warren, B. A. The vascular morphology of tumours. In: H-I. Peterson (ed.).Tumour Blood Circulation; Angiogenesis. Vascular Morphology and BloodFlow of Experimental and Human Tumors, pp. 1-47. Boca Raton. FL: CRCPress. Inc.. 1979.

11. Eddy, H. A., and Casarett, G. W. Development of the vascular system in thehamster malignant neurilemmoma. Microvasc. Res., 6:63-82, 1973.

12. Egawa. J., Ishioka. K.. and Ogata. T. Vascular structure of experimentaltumors. Acta Radiol. Oncol.. IS: 367-375. 1979.

13. Ausprunk, D. H., and Folkman, J. Migration and proliferation of endothelialcells in preformed and newly formed blood vessels during tumor angiogenesis.Microvasc. Res.. 14: 53-65, 1977.

14. Konerding, M. A.. Steinberg. F.. and Budach. V. The vascular system ofxenotransplanted tumor-scanning electron and light microscopic studies.Scanning Microsc., 3: 327-336, 1989.

15. Endrich. B.. and Messmer, K. Microcirculation of transplanted tumours.Arzneim-Forsch., 31: 2007-2011, 1981.

16. Warren. B. A., and Shubik, P. The growth of the blood supply to melanomatransplants in the hamster cheek pouch. Lab. Invest., 15: 464-478. 1966.

17. Cullino. P. M.. and Grantham. F. H. The vascular space of growing tumors.Cancer Res., 24: 1727-1732. 1964.

18. Fidler, I. J., and Balch. C. M. The biology of cancer metastasis and implications for therapy. Curr. Prob. Surg., 24: 166-177. 1987.

19. Gabbert, H., Wagner, R., and HÃ¶hn, P. The relation between tumor cellproliferation and vascularization in differentiated and undiffercntiated coloncarcinomas in the rat. Virchows Arch. Cell Pathol.. 41: 119-131, 1982.

20. Mlynek, M-L., van Beunigen. D.. Leder. L-D.. and Streffer. C. Measurementof the grade of vascularization in histolÃ³gica! tumour tissue sections. Br. J.Cancer. 52: 945-948, 1985.

21. Witte. S.. and Goldenbcrg. D. M. The early vascular reactions to heterotran-splanted human tumours in the hamster cheek pouch. Bibl. Anat., 9: 396-402. 1966.

22. Jain, R. K. Determinants of tumor blood flow. A review-. Cancer Res.. 48:2641-2658. 1988.

23. Folkman, J., and Klagsbrun, M. Angiogenic factors. Science (Wash. DC),255:442-447, 1987.

24. Yamaura, H., and Sato, H. Quantitative studies on the developing vascularsystem of rat hepatoma. J. Nati. Cancer Inst., 53: 1229-1240, 1974.

25. Peters. W'.. Teixeira, M., (maglietta. M., and Gross, J. F. Microcirculatory

studies in rat mammary carcinoma. I. Transparent chamber method, development of microvasculature. and pressures in tumor vessels. J. Nati. CancerInst.. 65:631-642, 1980.

26. Winawer, S. J. Preventive screening and early diagnosis. In: J. J. DeCosse(ed.). Clinical Surgery International. Large bowel cancer. Vol. 1, pp. 46-62.New York: Churchill Livingstone. 1981.

27. Endrich, B., Reinhold. H. S., Gross. J. F.. and Intaglietta. M. Tissue perfusioninhomogeneity during early tumor growth in rats. J. Nati. Cancer Inst.. 62:387-395. 1979.

28. Endrich. B.. Hammersen. F., GÃ¶tz,A., and Messmer, K. Microcirculatoryblood flow, capillary morphology, and local oxygen pressure of the hamsteramelanotic melanoma A-Mel-3. J. Nati. Cancer Inst., 68: 475-485, 1982.

29. Van Den Brenk, H. A. S., Crowe, M.. Kelly. H., and Stone, M. G. Thesignificance of free blood in liquid and solid tumours. Br. J. Exp. Pathol.,58: 147-159, 1977.

30. Vaupel. P. Hypoxia in neoplastic tissue. Microvasc. Res., 13: 399-408. 1977.31. Vogel. A. W. Intnuiinioial vascular changes with increased size of a mam

mary adenocarcinoma: new method and results. J. Nati. Cancer Inst., 34:571-578, 1965.

32. Hilmas. D. E.. and Gillette, E. L. Morphometric analyses of the microvasculature of tumors during growth and after X-irradiation. Cancer (Phila.),33: 103-110. 1974.

2417



1990;50:2411-2417. Cancer Res Stewart A. Skinner, Peter J. M. Tutton and Paul E. O'Brien the RatMicrovascular Architecture of Experimental Colon Tumors in

Updated version

http://cancerres.aacrjournals.org/content/50/8/2411

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

[email protected] 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/50/8/2411To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Documents

Microvascular Architecture of Experimental Colon Tumors in ... · Tumor Induction. Colon tumors were induced by s.c. injections of dimethylhydrazine dihydrochloride (Sigma Chemical