vascular reactivity and perfusion characteristics in 7,12...
TRANSCRIPT
[CANCER RESEARCH 43, 363-366, January 1983]0008-5472/83/0043-OOOOS02.00
Vascular Reactivity and Perfusion Characteristics in 7,12-Dimethylbenz-(a)anthracene-induced Rat Mammary Neoplasia1
Ragnar Hultborn, Egil Tveit, and Lilian Weiss2
Departments of Histology and Oncology ¡R. H.¡, Surgery [E. T.], and Physiology and Surgery [L. W.I, University of Göteborg, Box 33031, S-40033 Göteborg, Sweden
ABSTRACT
Blood flow during "resting conditions" and during noradren-
aline infusion were studied by the labeled microsphere technique in dimethylbenz(a)anthracene-induced mammary tu
mors, skin, muscle, and lung in the rat. Intratumoral distributionof flow was studied by autoradiography of spheres trapped inthe vascular beds of the tumors. Histological examination wasperformed and correlated to the blood flow data. Mean bloodflow to the tumors during "resting conditions" was relatively
high (49 ml/min/100 g tissue) but was substantially decreased(5 ml/min/100 g tissue) during noradrenaline infusion whichproduced a pressure elevation of 35 mm Hg. Thus, vascularresistance of the tumor tissue increased dramatically. Cardiacoutput increased, but total systemic resistance was unchanged. Vascular resistance in muscle was unchanged incontrast to an increase seen in skin. Shunted systemic bloodflow to the lungs and bronchial arterial flow decreased indicating reactivity of abnormally large arteriovenous passages in thetumors. Poorly differentiated tumors had a higher vascularresistance than did well-differentiated tumors. Autoradiography
revealed a nodular flow distribution with a slight tendency ofhigher perfusion in the periphery of these tumors.
INTRODUCTION
Knowledge of regulatory mechanisms for tumor blood supplymay potentially improve results of irradiation and cytotoxicdrug therapy and contribute to the understanding of neoplasticprogression (3, 13). Most studies of tumor vascularity havebeen performed on fast-growing transplanted tumors using
morphological methods. The tumors seem to have an irregularvascular network comprised of large vessels with poorly developed smooth muscle which are not differentiated into arte-
rioles and venules (14). Whether or not tumor vessels react tohumoral vasoactive stimulants is a matter of debate. Studiesusing pharmacoangiographic methods indicate that tumor vessels are not able to constrict (2), while other studies usingisotope techniques have shown that blood vessels of varioustransplanted tumors have this ability (1, 4, 7, 9, 11). Thisdiscrepancy of results obtained for tumor blood flow and vascular reactivity may be ascribed to a number of factors, suchas different tumor models, different methods of analysis, andvariable physiological status of the vascular beds at the actualmoment of investigation. Studies of tumor angiogenesis suggest that tumor vessels are derived from the surrounding normal tissue (3). This would indicate that the vascular reactivityof tumor vessels is similar to that of the surrounding host tissue.
' This investigation was supported by the Swedish Cancer Society (1193-B82-
04X), Konung V:s Research Foundation, and Syskonen Svenssons Foundation.2 To whom requests for reprints should be addressed.
Received December 21, 1981; accepted December 12, 1982.
Thus, with this as a background, this study of the vascularhemodynamics of a slowly growing nontransplanted rat mammary adenomatous neoplasm induced by DMBA3 (Sigma
Chemical Co., St. Louis, Mo.) was undertaken. The histopath-
ological type of tumor was correlated to resting blood flow andblood flow during noradrenaline infusion, which was measuredby means of a double-isotope microsphere technique. The
distribution of blood perfusion within the tumors was alsostudied by autoradiography.
MATERIALS AND METHODS
Tumor Model. Female Sprague-Dawley rats (Anticimex, Stockholm,
Sweden), 50 to 55 days old, were fed DMBA (16 mg dissolved in 1 mlolive oil) by gavage while under brief ether anesthesia (5, 12). Six to 8weeks later, nodules could be palpated along the mammary ridge.Experiments were performed 10 to 12 weeks after induction.
Cardiac Output and Regional Blood Flow Measurement. Cardiacoutput and regional blood flow were measured by means of a double-isotope microsphere technique (10) during "resting conditions" and
during constant noradrenaline infusion. The radioactivity of the tumorswas not only studied in a well-type auto-y-spectrometer but also by
autoradiography to illustrate the intratumoral distribution of blood flow(6).
The procedure was as follows. The rats were anesthetized withNembutal (50 mg/kg body weight) i.p. The right carotid artery wascannulated to allow injection of microspheres into the left ventricle.One femoral artery and the caudal artery were cannulated for obtainingreference blood samples and for the respective monitoring of bloodpressure and heart rate. The right brachial artery was cannulated forthe infusion of NaCI solution (9 g/liter) simulating "resting conditions"
or of noradrenaline (0.005 mg/ml) added to this solution. The rate ofnoradrenaline infusion was adjusted individually to produce a substantial increase of arterial blood pressure (30 to 40 mm Hg). Attemptswere made to maintain the body temperature around 37°and to keep
the depth of anesthesia, the preparation time, the blood loss, and therespiratory status as constant as possible. Polystyrene spheres (3MCo., St. Paul, Minn.), diameter 15 ± 3 (S.D.) ftm, were labeled with14'Ce and 85Sr, respectively. The number of spheres injected per animal
was approximately 600,000 per isotope. The microspheres, whichwere suspended in 1 ml NaCI solution (9 g/liter) including dextran andTween, were injected within 45 sec. Reference blood samples werewithdrawn at 0.6 ml/min during 90 sec. Two measurements were madein each animal. Thus, microspheres were injected first during "resting"
steady state conditions and then while infusing noradrenaline solution.The rats were killed, and the tumors and samples from abdominal
skin, muscle, and lung were dissected out, wet weighed, fixed informaldehyde solution, and processed further for activity measurements in a well-type -/-spectrometer.
Whole-body y detection, including dissected tissues and reference
samples, was performed by use of a Packard Armac scintillationcounter. Cardiac output and regional perfusion could then be calcu-
3 The abbreviation used is: DMBA, 7,12-dimethylbenz(a)anthracene.
JANUARY 1983 363
on July 9, 2018. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
R. Hultborn et al.
lated according to the equation:
Organ flow Reference sample flow
Organ activity Reference sample activity
Autoradiography. Intratumoral blood flow distribution was furtherinvestigated by means of autoradiography since tumor tissue is nothomogenous and blood flow measurements obtained per unit of totaltumor weight are not considered to be adequate. Slices (2 mm thick)of the formaldehyde-fixed tumors with weights over 3 g were cut andplaced on X-ray film (Kodak Industrex C; Eastman Kodak Co., Roch
ester, N. Y.) for 2 weeks. A sagittal section of the kidneys and acoronar section of the brain from each animal were also included. Boththe injected isotopes contributed to the exposure of the film, and noattempt was made to quantify blood flow or vascular reactivity in thispart of the study.
Histopathology of the Tumors. Tumors induced along the mammaryridge by DMBA are mainly of adenomatous origin, but the histologicalappearance varies widely (12). They rarely metastasize, and the criteriafor malignancy are harder to define than for human mammary tumors.Therefore, the tumors were subgrouped into adenomatous and mes-
enchymal neoplasms. The adenomatous tumors were then furtherclassified according to (a) the epithelial architecture of the ducts, i.e.,well-differentiated epithelial lining versus disorganized and multilay-
ered lining, and to (b) the epithelial cell nuclear appearance, i.e.,condensed small nuclei without prominent nucleoli versus large vesicular nuclei with distinct nucleoli. The tumors were further subdividedby application of the above criteria into a well-differentiated group, a
poorly differentiated group, and an intermediate group. Histologicalsections were taken adjacent to the thick section that was processedfor autoradiography in order to allow for a correlation between bloodflow pattern and morphology.
Statistical Analysis. Blood flow estimations obtained by means ofmicrosphere tracer technique were normalized by logarithmic transformation in recent work by Jirtle ef al. (8). All statistical tests performedin this study followed the same procedure. The linear correlationcoefficient was used to determine the extent of correlation betweentumor weight and organ vascular resistance. The same test was usedto correlate total tumor burden to cardiac output and radioactivitytrapped in lung. Student's f test group comparison was used to com
pare the ratio of vascular resistance at rest and during noradrenalineinfusion between tumor groups of different differentiation. This test wasalso used to compare blood flow and vascular resistance in tumors ofdifferent differentiation in matched-weight groups. All other statisticalanalyses were performed with Student's i test pairing design. Values
for p less than 0.05 were considered significant. Mean values andstandard errors are presented.
RESULTS
Tumor Yield. Experiments were performed 10 to 12 weeksafter tumor induction. The mean weight of the 10 rats at thisage (1 7 to 19 weeks) was about 300 g. The total number oftumors was 62, i.e., an average of 6.2 per animal. The tumorsweighed between 0.02 and 15 g, with a mean of 1.95 g and amedian of 0.78 g. The majority of tumors were located in thecranial halves of the mammary ridges. Skin ulcérations werepresent above 4 of the larger tumors.
Histopathology. All tumors were subjected to histologicalanalysis using van Gieson staining. Two of them were fibromas.The others were adenomatous neoplasms, which were furthersubdivided into well-differentiated (n = 18), poorly differentiated (n = 16), and intermediate (n = 27) groups. Tumors withdifferent degrees of differentiation were found to be present inthe same animal. The mean and median weights of the well-
differentiated tumors were 2.12 and 0.75 g as compared to2.75 and 1.47 g for the poorly differentiated ones. The histo-pathological pattern of the large tumors showed a markedheterogeneity within each tumor. Necrotic areas were seenespecially in large and poorly differentiated tumors. Skin ulcer-tation was not related to poor differentiation.
Cardiac Output, Regional Blood Flow, and Vascular Resistance. Blood pressure and cardiac output were increasedto the same extent during the noradrenaline infusion. Bloodflow to muscle did not change while blood flow to abdominalskin showed a moderate decrease. Tumor blood flow sharplydropped during the infusion of noradrenaline. The flow to thelungs (bronchial arterial and shunted flow) also decreasedduring noradrenaline infusion (Chart 1). Changes in total systemic resistance per 100 g body weight and organ vascularresistance per 100 g tissue (Chart 2) were related to the-changes in resistance vessel radius or total cross-sectionalarea of the resistance vessels and were calculated by dividingblood pressure with blood flow. Thus, as seen in Chart 2, thetotal systemic resistance did not change during noradrenalineinfusion while that of abdominal skin increased. The tumorsshowed a pronounced increase in vascular resistance. Bloodflow and vascular resistance were compared between thedifferent tumor groups at various stages of differentiation. Nodifference was found between matched-weight groups of tu
mors with wet weights below 2.0 g at resting conditions.However, larger poorly differentiated tumors had a significantlylower blood flow (30.8 ± 5.0 ml/min/100 g) and highervascular resistance (4.03 ±0.43 mm Hg/min/100 g/ml) thandid well-differentiated tumors (59.8 ±7.9 and 1.90 ±0.21,
respectively). There was no significant difference in the ratio ofvascular resistance at restrvascular resistance during noradrenaline infusion between tumors with different degrees ofdifferentiation. There was a positive correlation between weightand vascular resistance at resting conditions for poorly differentiated tumors but not for intermediately and well-differen
tiated tumors. Vascular resistance during noradrenaline infusion was not related to tumor weight. The 2 fibromas had ahigher vascular resistance at rest (6.5 and 8.0 mm Hg/min/
4 bloodpressuremmHg150no50rfrf3O20Or
(Tito-outpulml/minlOOgiÕIO5'Hood
flowskrml/millOOgIO
ITÕ5>bkx)d
flowmuscleml/min
lOOgrU1i'est
NA resi NA rest NA rest NA
P «O.OO5 «0.05 «=0.1 ns60
SO40302010'blood
flowtumorml
/min 100g500*4OO3OO20000•*!•
arterialMshunt
edftowtothébnçsml/min
•lOOqTHiiiresi
NA resi NA
«DOCH «O.O05
Chart 1. Histogram illustrating the blood pressure, cardiac output, and bloodflow of abdominal skin, muscle, tumors, and arterial and shunted flow to the lungsduring NaCI solution and noradrenaline (NA) infusion. Bars, S.E. The degree ofsignificance of the difference between resting and noradrenaline values is indicated as the p value, ns, not significant.
364 CANCER RESEARCH VOL. 43
on July 9, 2018. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Hemodynamic Characteristics in DMBA Tumors
40
30
»Un Timor
10
10
ai
dJ0.001
k
p< 0.05
Chart 2. Histogram illustrating the mean of the vascular resistance units (mmHg per min per 100g tissue per ml) in abdominal skin, muscle, and tumors andthe total systemic resistance (mm Hg per min per 100 g body weight per ml)during rest and noradrenaline (A/A), ns, not significant; bars, S.E.
100 g/ml) than did adenomatous neoplasms but reacted similarly to noradrenaline.
The amount of radioactivity trapped in lung tissue increasedwith increasing total tumor mass (range, 2 to 40 g; mean, 12g) during resting conditions and noradrenaline infusion. Anominal lung flow comprised of bronchial perfusion andshunted flow from normal tissues was calculated by extrapolation of the tumor weight to zero and found to be 165 ml/min/100 g at rest. This result corresponded well with values obtained in normal non-tumor-bearing rats." There was found to
be a significant decrease in the amount of trapped radioactivityin the lungs by one-third during noradrenaline infusion as
compared to resting conditions. Blood flow through the tumorswas decreased by one-tenth, which indicated that blood flow
through arteriovenous passages larger than 15 firn was decreased.
Analysis of "shunting" in relation to tumor differentiation is
impossible to make in this study due to the presence of tumorsof several types within the same animal. No correlation betweentotal tumor weight and cardiac output could be detected in thislimited material. A central and a peripheral part from the tumorsweighing more than 3 g were dissected out in a standardizedfashion. The mean weight of the central part was 0.42 g and ofthe peripheral anulus was 1.21 g. No statistical difference inblood flow could be obtained between the central and peripheral parts of the tumors. This could probably be explained bythe fact that well-perfused regions were not strictly peripheralbut were distributed across the section of the tumors, whichwas also illustrated by the autoradiograms.
Autoradiography. The autoradiograms of coronar brain sections showed a homogenous distribution of spheres and aresolution capable of visualizing the well-perfused cortical and
central gray substances which were in contrast to the poorerperfusion of the white matter. Sections from kidneys produced
dense and homogenous blackening in the cortical areas. Sections of 15 tumors with weights over 3 g produced a greatvariety in the patterns of blackening (Fig. 1). In general, therewas a heterogenous "nodular" blackening with a tendency to
appear at the periphery. Regions of low blackening corresponded well to areas of necrosis as judged from histologicalanalysis. So far, no direct correlation between intratumoralperfusion pattern and histology can be found.
DISCUSSION
Several studies have demonstrated the presence of vascularreactivity to noradrenaline and other pressor drugs in varioustransplanted tumors (1, 4, 7, 9, 11). Very few hemodynamicstudies have been performed on induced tumors. DMBA-in-
duced mammary neoplasia was chosen as the model of thepresent study since these tumors seem to be closely related tomammary tumors in man from both morphological and functional points of view (5, 12). The tumors are derived frommammary epithelium, i.e., they are of epidermal origin. Thehistological appearance of the tumor is known to be both of abenign adenoma-adenofibroma type as well as of a malignantadenocarcinoma type. Thus, the possibility of relating functional characteristics to morphology is offered. The micro-
sphere tracer technique was chosen for blood flow analysisbecause flow at resting conditions and during noradrenalineinfusion could be studied in various tissues of the same animal.This reduction of error due to biological variation is of specialimportance because of the heterogeneity of tumor morphology.Furthermore, autoradiographic analysis of intratumoral flowdistribution is made possible.
The microsphere tracer technique measures only flowthrough vessels with dimensions less than the sphere diameter,which in this case was 15 /im. Thus, blood flow data is anatomically determined in contrast to results obtained from blood
1cm
' R. Hultborn, E. Tveit. and L. Weiss, unpublished observations. Fig. 1. Autoradiogram from a tumor section.
JANUARY 1983 365
on July 9, 2018. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
R. Hultborn et al.
flow analysis with diffusible tracer substances. Arteriovenouspassages larger than 15 /urn are known to exist in the skin aspart of the thermoregulatory system and probably in tumortissue. The blood flow through these passages in this studymay be only indirectly estimated by the amount of activitytrapped in the lungs. Results are presented both as blood flowand vascular resistance. The former gives direct information asto the access of nutrients to the tissues, but the latter parameteris more informative concerning reactivity in the vascular bedand provides an indirect measure of the cross-sectional area
of the vascular bed under study. Blood flow in tumors at restingconditions is 49 ±2 ml/min/100 g. Similar values were foundin S.C., i.m., and intrahepatically autotransplanted tumors ofthis type.4 However, autoradiography reveals a considerable
heterogeneity of flow within the tumors. The background of thisheterogeneity is unknown but may represent areas with differences in growth pattern and cellularity. The vast heterogeneityof flow in the adenomatous neoplasms may involve severalfactors such as cellularity, follicles, connective tissue, andnecrosis. Further studies to elucidate the importance of thesefactors for blood flow are under way. The finding of a decreasedperfusion in larger tumors is confirmed here but only for poorlydifferentiated tumors weighing more than 1 g. This fits well withthe concept of angiogenesis and tumor progression (3) inwhich hypoperfused areas occur in tumors weighing more than1 g. The lack of this relation for well- and intermediately
differentiated tumors indicates the importance of growth characteristics. The significantly lower blood flow (higher vascularresistance) of poorly differentiated as compared to well-differ
entiated tumors might be even more marked if blood flow wererelated to cell number instead of to wet weight. An increasedinterstitial pressure in tumors has been demonstrated to increase with tumor weight in other studies. This finding mayimply that vascular compression compromises perfusion oflarger tumors (17).
It is evident that blood flow through tumor tissue is drasticallyreduced and vascular resistance is increased during a moderate infusion of noradrenaline. The vascular bed of tumorsreacts more than those of skin and muscle, which might indicate a higher sensitivity to noradrenaline. Such a conclusionmay be erroneous since the degree of smooth muscle contraction at "resting conditions" may vary from tissue to tissue.
Smooth muscle of the tumor vascular bed might be morerelaxed than that of skin and muscle at rest for metabolicreasons and therefore may react more to a particular vasoac-tive drug. Dose-response curves determined for various organsin preliminary in vitro experiments have confirmed the hyper-sensitivity of the tumor vascular bed to noradrenaline (16). Anabnormally high number of spheres were trapped in the lungsof tumor-bearing rats as compared to controls. A positivecorrelation between the amount of radioactivity and the totaltumor mass was found. The amount of spheres in the lungswas reduced by approximately one-third during noradrenaline
infusion irrespective of the tumor burden. Values obtained forbronchial flow and for shunted flow from normal tissues byextrapolation of the tumor weight to zero in a tumor mass-lung
activity graph were well in accordance with those results foundfor normal rat." Thus, the fraction of lung activity derived fromtumor "shunting" can be determined both at rest and during
noradrenaline infusion. It is clear from these calculations thatarteriovenous passages larger than 15 ¿imreact but to a lesserdegree than do the smaller ones. At least 40% of tumor bloodwas demonstrated to pass through arteriovenous connectionslarger than 15 /¿min experiments using larger spheres (25 firn)(15).
In conclusion, resting blood flow of DMBA-induced mammary
tumors is relatively high. Vascular resistance is higher in poorlydifferentiated tumors than in well-differentiated ones and in
creases with increasing tumor weight in poorly differentiatedtumors. There is also an abnormal passage of blood througharteriovenous passages larger than 15 ¡im. Noradrenalinedrastically increases vascular resistance of tumors with themost profound reduction of flow through passages less than15 ftm in diameter. Perfusion of blood in tumors is heterogeneous, with a tendency to be greater at the periphery of thetumor.
ACKNOWLEDGMENTS
We are indebted to Ake Cederblad, Department of Radiopnysics, and to CarlWidmark for technical assistance.
REFERENCES
1. Edlich, R. F., Rogers, W., OeShazo, C. V., Jr., and Aust, J. B. Effects ofvasoactive drugs on tissue blood flow in the hamster melanoma. CancerRes., 26: 1420-1424, 1966.
2. Ekelund, L., Göthlin, J., Jonsson, N., and Sjögren, H. O. Pharmaco-angiog-raphy in experimental tumors. Evaluation of vasoactive drugs. Acta Radiol.Diagn., 17: 329-324, 1976.
3. Folkman, J., and Cotran, R. Relation of vascular proliferation to tumourgrowth. Int. Rev. Exp. Pathol., Õ6.208-245, 1976.
4. Gullino, P., and Grantham, F. Studies on the exchange of fluids betweenhost and tumor. Ill Regulation of blood flow in hepatoma and other rattumors. J. Nati. Cancer Inst., 28. 211-229, 1962.
5. Muggins, C. B. Experimental Leukemia and Mammary Cancer. Chicago:University of Chicago Press, 1979.
6. Hultborn, R., and Weiss, L. Gamma-radiation autoradiography of micro-spheres in the study of regional blood flow in experimental tumours. Micro-vase. Res., 17: S1 20, 1979.
7. Jirtle, R.. Clifton, K., and Rankin. J. Effects of several vasoactive drugs onthe vascular resistance of MT-W9B tumors in W/Fu rats. Cancer Res., 38.2385-2390, 1978.
8. Jirtle, R., Clifton, K., and Rankin, J. Measurement of mammary tumor bloodflow in unanesthetized rats. J. Nati. Cancer Inst.. 60. 881-886. 1978.
9. Mattson, J., Appelgren, L., Karlsson, L., and Peterson, H.-l. Influence ofvasoactive drugs and ischaemia on intratumor blood flow distribution. Eur.J. Cancer, 14: 761-764, 1978.
10. McDevitt, D. G., and Nias, A. Simultaneous measurement of cardiac outputand its distribution with microspheres in the rat. Cardiovasc. Res., 10: 494-498, 1976.
11. Rankin, J., Jirtle, R., and Phernetton, T. Anomalous response of tumorvasculature to norepinephrin and prostaglandin E2 in the rabbit. Circ. Res.,41: 496-502, 1977.
12. Young, S. and Hallowes, R. Tumours of the mammary gland. In: V. S.Turusov (ed.), Pathology of Tumours in Laboratory Animals, Vol. 1, Part 1,pp. 31-72. Lyon, France: International Agency for Research on Cancer.1973.
13. Tannock, I. Oxygen distribution in tumours: influence of cell proliferation andimplications for tumour therapy. Adv. Exp. Med. Biol., 75: 597-603, 1976.
14. Warren, B. A. The vascular morphology of tumors. In: H.-l. Peterson (ed.).Tumor Blood Circulation; Angiogenesis, Vascular Morphology and BloodFlow of Experimental and Human Tumors, pp. 1-47. Boca Raton, Fla.: CRCPress, Inc., 1979.
15. Weiss, L., Hultborn, R., and Tveit, E. Blood flow characteristics in inducedrat mammary neoplasia. Microvasc. Res.. 17: S119, 1979.
16. Weiss, L., Tveit, E. and Hultborn, R. Vascular reactivity in experimentalmammary tumours and normal vascular beds as studied in vitro with artificialperfusion. Acta Physiol. Scand., 708. 47A, 1980.
17. Wiig, H., Tveit, E., Hultborn, R., and Weiss, L. Interstitial fluid and vascularpressures in rat mammary tumours. Bibl. Anat., 20: 620-623, 1981.
366 CANCER RESEARCH VOL. 43
on July 9, 2018. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
1983;43:363-366. Cancer Res Ragnar Hultborn, Egil Tveit and Lilian Weiss Neoplasia7,12-Dimethylbenz-(a)anthracene-induced Rat Mammary Vascular Reactivity and Perfusion Characteristics in
Updated version
http://cancerres.aacrjournals.org/content/43/1/363
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
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/43/1/363To request permission to re-use all or part of this article, use this link
on July 9, 2018. © 1983 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from