- Email: [email protected]
Microangiographic and histologic analysis of the effects of hyperthermia on murine tumor vasculature
Inr. .I. Rudimon Onn~lo~.v Bd Phy.s.. Vol. Printed I” the U.S.A. All rights reserved.
15. pp. 41 I-420 Copyright
0360-3016/88 $3.00 + .OO 0 1988 Pergamon Press plc
??Original Contribution MICROANGIOGRAPHIC AND HISTOLOGIC ANALYSIS OF THE EFFECTS OF HYPERTHERMIA ON MURINE TUMOR VASCULATURE YASUMASA NISHIMURA, M.D., MASAHIRO HIRAOKA, M.D., SHIKEN Jo, M.D., KEIZO AKUTA, M.D., YUTAKA YUKAWA, R.T., YUTA SHIBAMOTO, M.D., MASAJI TAKAHASHI, M.D. AND MITSUYUIU ABE, M.D. Department of Radiology,Facultyof Medicine,Kyoto University,54 Shogoin-Kawahara-cho, Sakyo-ku,Kyoto 606,Japan The effects of hyperthermia on murine tumor vasculature were studied by microangiography and histological examination. The tumors used were SCC VII carcinoma and mammary adenocarcinoma of syngeneic 6H/He mice. For the quantitative analysis of microanglographic changes, the percent (%) vascular area, which was defined as the percentage of opacified tumor vessel area to the entire tumor area, was determined in each microanglogram. The % vascular area after heating in a water bath at 44°C for 30 min was minimized 24 hr after heating in both types of tumors. The histologic study revealed that the initial decrease of the 96 vascular area was due to congestion, thrombosis, and rupture of tumor vessels, and its subsequent increase was due to angiogenesis. SCC VII was more beat sensitive than mammary adenocarcinoma in terms of tumor growth delay, and tumor vessels of SCC VII were more vulnerable to heat than those of mammary adenocarcinoma. Histological examinations showed a marked difference in the architecture of vessels between the two types of tumors. Tumor vessels of mammary adenocarcinoma were supported by a connective tissue band, whereas those of SCC VII consisted of a single endothelial cell layer. Our findings suggest that the tumor vessels supported by a connective tissue band are less sensitive to heat than those without such support. The vascular damage of SCC VII was temperature dependent, and the critical temperature at which dramatic vascular damage appeared was between 42.7”C and 43.7”C. Hyperthermia, Tumor vasculature, Microangiography.
greater depth and length than can be accomplished by routine histologic methods. Using correlative histologic sections, microangiographic changes can be correlated to histopathologic lesions. By microangiography, circulating vessels can be distinguished from congested non-circulating vessels as contrast medium doses not enter the congested vessels. In addition, it can demonstrate the localization of vascular changes in a tumor, which cannot be detected by the functional studies. In the present study, the effect of hyperthermia on the microcirculation of two different types of transplantable mouse tumors was investigated by microangiography and on correlative histological sections.
INTRODUCTION studies have been performed to evaluate the changes in tumor blood flow following hyperthermia by functional methods, and significant changes of several circulatory parameters have been demonstrated in a variety of tumors as well as normal tissues.4’12*‘5*‘6724-26 The tumor vasculature can be markedly damaged at temperatures that may alter but do not damage the vasculature in normal tissues.4,5,‘6*‘4,24-26 In view of the morphological features of the tumor vessels, it is not surprising that the response of tumor vessels to heat differs from that in normal tissues. The tumor vascular bed, which is devoid of innervation, consists of single-layered endothelial cells without an external coat of elastic basement membrane. Tumor vessels are maximally dilated, irregularly constricted, tortuous, and bent abruptly.‘8,‘9*24 Microangiography is an excellent technique for observing vascular changes at the level of small arterioles and capillaries.22 The main advantage of microangiography is that it permits the study of the vasculature at a
Numerous
METHODS
AND
MATERIALS
Animals and tumors Eight-week-old C3H/He male mice were obtained from the Animal Center of Kyoto University. Mice were caged in groups of 8 or less at a constant temperature and given food and water ad libitum.
Reprint requeststo: Dr. Y. Nishimura. Acceptedfor publication2 March 1988.
Supportedin part by a Grant-in-Aidfor ScientificResearch (61010041,61015038)from the Ministry of Education, Science and Culture,Japan. 411
412
I. J. Radiation Oncology 0 Biology 0 Physics
The tumors used were spontaneous mammary carcinoma (MCa) of C3H/He mice and SCC VII carcinoma of C3H/He mice. Some of the biological characteristics of MCa and SCC VII were described previously.’ 1,23 MCa was minced, and preserved in several ampules in liquid nitrogen. When required, the content of an ampule was thawed and transplanted to new recipients, and the tumors from this second transplanted generation were used for the experiments. The tumors were minced with scissors in Hanks’s medium supplemented with 5% fetal calf serum. The suspension was kept in a test tube on ice for about 15 min to permit larger tissue pieces to settle. The supematant was then filtered through gauze to obtain a cell suspension. About 2 X 1O5viable cells in 10 ~1 were inoculated subcutaneously into the right hind thigh of the mice. The SCC VII carcinoma was thawed from original frozen stocks and maintained by alternate passage in syngeneic mice and in cell culture in Eagle’s minimum essential medium (MEM) supplemented with 12.5% fetal calf serum. Approximately 1 X lo5 cells collected from monolayer cultures were inoculated subcutaneously into the right hind thigh of mice. Hyperthermia The mice were held in a specially constructed jig with firm fixation of the tail and left leg by taping them to the jig. The right tumor-bearing leg was pulled down by a special sinker (approx. 45g), which was fixed to the skin of the toe with super glue* so as not to impair the blood flow of the leg. The fully awake mice were then placed on a circulating water bath? maintained at a desired temperature, and the extended right-llg was locally heated. The temperat& of the water bath was controlled to within 0.05”C. The tumor was immersed at least 10 mm below the water surface. The mice were air-cooled during the heat treatment. All temperatures mentioned in this paper refer to the water bath temperature. Intratumor temperature measurement Intratumor temperature was measured in 9 SCC VII tumors by inserting a 29 gauge needle type microthermocouple probe* into the tumors. The temperatures at the center and periphery of a tumor were obtained by moving the thermocouple probe within the tumor. Microangiography and histologic examination Microangiography was performed as previously described.22 The mice were anesthetized with sodium pentobarbital, and the thoracic aorta was cannulated with a 23 gauge needle and the vena cava was cut. After flushing
* Aronnarufa, Konishi Co. Ltd., Osaka, Japan. t Model ET-35P, Toyo Seisakusho Co. Ltd., Tokyo, Japan. $ MT-29/5, Sensortek, Clifton, N.J. 0 Ryubari Sol B, Maruishi Seiyaku Co. Ltd., Osaka, Japan.
August 1988, Volume 15, Number 2
the circulatory system with warmed heparinized saline, filtered barium sulfate solution$ (0.25g/ml) was injected with a pressure of 150 mmHg for 30 min. The animal’s heart was pulsated for about 5 min after the injection of the contrast medium. When the muscular vasculature of the left normal hind limb was not opacified sufficiently, the mouse was excluded from the study to avoid a poor filling artifact. After fixing the tumor with 10% buffered formalin, contact radiographs of 1 mm thick tumor slices were obtained** using fine-grain filmtt. The tumor slice was cut along a sagittal direction through its center with as much surrounding normal tissues as possible. Correlative histologic sections of 4 pm thickness were prepared for each tumor slice and stained with hematoxylin and eosin. The microangiographic findings were evaluated quantitatively by the following methods. On an enlarged (approx. 10X) microangiogram, avascular areas and the entire tumor were measured, using a digital planimeter$$. Thereafter, the percent (%) vascular area in each microangiogram was determined as follows; % vascular area =(l-
area of avascular portions (cm*) )X 100. area of the entire tumor (cm*)
The mean f SE of the % vascular area was obtained for 5-9 microangiograms. Growth delay study The response of the tumors to heat treatment was evaluated by the growth delay. Three perpendicular diameters of each tumor were measured every 2 days after treatment using a caliper. Experiments were performed when the tumors reached a volume-of approximately 600 mm3 (approx. 12 mm in the longest diameter). Each treatment group consisted of 1 l- 13 mice. RESULTS
Intratumor temperature measurements Tumor center temperatures were equilibrated within 3-4 min after immersion in the water bath, and remained 0.2-0.3”C below the water bath temperature (Fig. la). Temperature difference between the tumor center and the periphery was within 0.1°C (Fig. 1b). Sequential microangiographic and histologic changes following hyperthermia The time course of microangiographic and histologic changes following hyperthermia was investigated in SCC
** CSM-2, Softex Co. Ltd., Tokyo, Japan. tt Fuji Co. Ltd., Tokyo, Japan. $$ Planix 7, Tamaya Co. Ltd., Tokyo, Japan.
Microangiographic and histologic analysis 0 Y. NISHIMURA
I 0
I
I
I
I
2
4
6
8
Time
(a)
(mid
42.7 f O,l°C -
42,7 f 0,l “C 42,8 + O.l”c
Tumor
temperature
(b)
Water
bath
temperature
43,0 + 0,05”c
Fig. 1. (a) Tumor temperature as a function of time after immersion in a water bath kept at 43°C. Temperature profiles for the tumor center in 3 tumors. The tumor volume was approximately 500 mm3 (0), 600 mm3 (A) and 800 mm3 (e). The tumor center temperature stabilized to approximately 42.7”C within 3-4 min. (b) Intratumor temperature distribution when a tumor bearing leg was immersed for more than 5 min in the water bath maintained at 43’C.
413
et al.
24 hr after heating. Although tumor vessels showed nearly no filling in the microangiograms obtained 24 hr after heating, only small parts of tumor vessels adjacent to the underlying muscle were opacified (Fig. 3d). Histologically, dilated thrombotic tumor vessels with no apparent endothelial cells were observed in the entire tumor. Degeneration of tumor cells such as pyknosis and karyolysis was also noted. Three days after heating, the % vascular area increased to 29 + 17%. In the magnified microangiograms, saccular, and tapering capillary sprouts, which are typical morphological features of tumor angiogenesis,14 were observed at the inner edge of the opacified tumor vascular area (Fig. 3e). These angiogenesis inevitably occurred from the periphery of the tumors. Seven days after heating, the % vascular area further increased. Mouse mammary carcinoma (MCa). The unheated control tumors were totally opacified with irregular tumor vessels, and the control % vascular area was 99 ? 1%. Histologically, most of the large tumor vessels passed through the interlobular connective tissue (Fig. 4a), and the endothelial cells of small peripheral vessels were supported by collagen.’ In MCa, the % vascular area decreased slightly immediately and 3 hr after heating, and it decreased further 6 and 24 hr after heating (Fig. 2). The % vascular area differed significantly between the 2 types of tumors immediately and 3 hr after heating (p < 0.05, p < 0.001 respectively). Three and 7 days after heating, the % vascular area increased. Similar angiographic findings of angiogenesis observed in SCC VII were also noted in MCa. Histologically, some large vessels embedded in thick connective tissues were free from thrombosis 24 hr after heating (Fig. 4b). Tumor cells surrounding a bariumfilled circulating vessel were less damaged than those surrounding a congested or thrombosed vessel (Fig. 4~).
loo-
*.*
*.-. p”
:
mammary adenocarcinoma
2
VII and MCa. Figure 2 shows the changes in % vascular area for the tumors from 0- 168 hr after 30 min of heating at 44°C. XC VII carcinoma. The unheated control tumors were totally opacified with fine tumor vessels (Fig. 3a), and the % vascular area was 100 f 0% (mean + SE). Histologically, tumor vessels consisted of only a thin endothelial cell layer (Fig. 3b). The % vascular area decreased to 4 1 -+ 16% immediately after heating as focal avascular areas appeared on the microangiograms (Fig. 3~). Correlative histologic sections revealed the dilatation, congestion, and rupture of the tumor vessels. The % vascular area further decreased 3 hr after heating, and was lowest
f
50-
\
Time After
Heating
(hour)
Fig. 2. Changes in the % vascular area from 0 to 168 hr after heating at 44’C for 30 min. The % vascular area differed significantly between the tumors immediately and 3 hr after heating (p < 0.05, p < 0.00 1 respectively).
414
I. J. Radiation Oncology 0 Biology 0 Physics
August 1988, Volume 15,Number 2
Fig. 3. Microangiograms and histological sections of SCC VII tumors. (a) Fine tumor vessels were opacified in an unheated control tumor. (b) Histologically, tumor vessels consisted of only a sparse endothelial cell layer. (c) Immediately after heating at 44°C for 30 min, avascular areas appeared on the microangiogram. Opacified tumor vessels showed marked dilatation. (d) One day after heating, tumor vessels showed nearly no filling except for a small part of tumor vessels near the muscle (white arrows). (e) Three days after heating, angiogenesis was noted in the periphery of a tumor.
Microangiographic changes at dlyerent temperatures in see vzz As the maximum decrease of the % vascular area was observed 24 hr after heating (44°C 30 min) in both types of tumors, we investigated the changes in % vascular area 24 hr after heating at different temperatures. Figure 5 shows the changes in % vascular area 24 hr after heating at 40-44°C for 30 min in SCC VII. The % vascular area
of the tumors heated at 40°C was 98 f 1%. There were no substantial changes in the microangiograms and histologic sections at 40°C. The % vascular area decreased at temperatures of more than 4 1°C. In the microangiograms obtained 24 hr after heating at 4 1’C for 30 min, opacified vessels showed mild dilatation and small focal avascular areas appeared. The % vascular area at 41°C was 90 + 13%. % vascular areas
Microangiographic and histologic analysis 0 Y. NISHIMURAetal
(d)
415
Fig. 3. (Contd)
were further decreased at 42 and 43°C to 76 + 23% and 59 -t 19% respectively. Dilatation of tumor vessels and mosaic-like avascular areas were noted on the microangiograms in these temperature ranges. The % vascular area of tumors heated at 44°C decreased dramatically to 5 + 2%, and the difference in % vascular area between 43°C and 44°C was significant (p < 0.005). Growth delay study Figure 6 shows the growth curves for unheated control tumors and tumors heated at 44’C for 30 min. The expo-
nential part of the growth curves was nearly parallel for the unheated and the heated tumors. Table 1 shows the tumor growth time to reach twice its initial volume (volume-doubling time) for the tumors. Although these values for the unheated control tumors of SCC VII and MCa were nearly equal, those for the tumors heated at 44°C for 30 min differed significantly between the two types of tumors, indicating that SCC VII was more sensitive to heat than MCa. The growth delay in WC VII tumors heated at 4044°C for 30 min was also investigated. The volume-dou-
416
I. J. Radiation Oncology 0 Biology 0 Physics
(4
August 1988, Volume 15, Number 2
Fig. 3. (Contd)
bling times of the tumors heated at 40-42°C were equal to that of the control tumors (Table 1). The volume-doubling time of the tumors heated at 43°C was slightly prolonged, whereas that of the tumors heated at 44°C was markedly prolonged (p < 0.001). DISCUSSION Sequential changes in circulatory parameters of transplanted tumors following hyperthermia have been reported by several investigators using isotope-trapping methods.12,‘6*25,26Stewart et ~1.~~found that blood flow of the SAFA tumor was reduced 1 day after heating (42.5”C, 1 hr) and returned to the control range by 2-3 days. Kang et al. l2 observed that blood volume in SCK tumors was only one-tenth of the control value 5 to 12 hr after heating at 43.5”C for 30 min and started to recover slowly thereafter. In the present study, the change of % vascular area estimated by microangiography after 30 min of heating at 44°C showed similar tendency. The maximum decrease of the % vascular area was observed 24 hr after heating in both types of tumors, and it increased thereafter. From the functional analysis of the vascular changes, we could not determine whether the recovery of circulating parameters after heating were due to recanalization of tumor vessels or to angiogenesis. Our study revealed that the initial decrease of the % vascular area was due to congestion, thrombosis, and rupture of tumor vessels, and its subsequent increase was due to angiogenesis (Fig. 3). Recanalization of tumor vessels may not be ruled out completely only by observing the morphological features of tumor angiogenesis. However, only remnants of tu-
mor vasculature with no endothelial cells were observed in the histological sections obtained 24 hr after heating at 44°C. Similar histologic findings were made by Emami et al.’ and they have termed these vessels “ghosts of blood vessels”. Therefore, it is practically impossible that the re-opening of such degenerated tumor vessels occur. Vascular damage following hyperthermia at 44°C for 30 min was irreversible in small tumor vessels of murine tumors. As to localization of vascular changes in a tumor, tumor vessels adjacent to the underlying muscle (tumor peripheral zone) were to be protected from thermal damage (Fig. 3d). Consequently, angiogenesis after hyperthermia always occurred from the periphery of the tumor (Fig. 3e). Several experimental and clinical studies revealed similar vascular” and histological’0*“9’3 changes in the tumor after hyperthermia. One possible reason for this is the presence of a minute temperature difference between the tumor center and the peripheral zone. Although this difference was within O.l”C in our measurement (Fig. lb), the present invasive thermometry is too inaccurate to rule out this possibility. Another possibility is the difference in the architecture of the tumor vessels. The periphery of tumors is often well vascularized and well perfused while the blood supply to the tumor center is inadequate. This difference in vascular density and blood flow between the tumor center and the periphery might account for the differential vulnerability of tumor vessels by heat. Temperature dependency on vascular damage of murine tumor vasculature has been studied by several invesEmami et al.’found no significant changes tigators. 3*6.7~18 in the vasculature of the BA 1112 sarcoma after heating
Microangiographic and histologic analysis 0 Y. NISHIMURAet al.
for 40 min below 40.5”C, marked dilatation and congestion of the vessels after heating at 42.5’C, and massive hemorrhage and necrosis with rupture of vessel walls after heating at 44.5’C. Eddy,6 who studied a squamous cell carcinoma grown in transparent chambers, has observed reversible dilatation of the vessels at 41°C petechiation, stasis, thrombosis at 43°C and complete shutdown of the circulation after heating at 45°C for 30 min.
417
The present result on histological changes was quite similar to the observation by the investigators although the tumors used were different. In addition, our quantitative analysis of vascular changes by microangiography revealed a linear relationship between vascular damage and temperature ranging from 4 1‘C to 43°C and a “breaking point” temperature at which vascular damage occurred dramatically be-
(a)
04 Fig. 4. Histological sections of mouse mammary adenocarcinoma. (a) Most of the large tumor vessels passed through the interlobular connective tissue in a control tumor. (b) A large tumor vessel embedded in the thick connective tissue band was filled with contrast medium 24 hr after heating at 44°C for 30 min. (c) Tumor cells surrounding a barium-filled circulating vessel (arrow) were less damaged than those surrounding a congested vessel (open arrow).
418
I. J. Radiation Oncology 0 Biology 0 Physics
(c)
August 1988, Volume 15, Number 2
Fig. 4. (Co&d)
tween 43°C and 44°C (Fig. 5). Another quantitative analysis of heat-induced vascular damage was made by BergBlok et cd3 They calculated the time necessary to produce microcirculation stoppage in 50% of the treated tumors (ST50), using Rhabdomyosarcoma BA 1112 growing in observation chambers. The ST50 values for 42,42.5,43, and 43.5”C were 226, 152, 10 1, and 70 min respectively. Their data does not indicate the presence of the breaking point in the temperature range. The opposite conclusions on the breaking point may be due to a multitude of factors. Not only are different tumors used but also the assay system of vascular damage. Moreover, temperature ranges examined differ considerably, and
I
I
controls
40
,
,
41
42
Temperature
\* 43
44
(“Cl
Fig. 5. Changes of % vascular area in SCC VII 24 hr after heating (40-44’C, 30 min). % vascular area decreased dramatically
between 43’C and 44°C with a difference statistically significant (p < 0.005).
there may be an essential difference in timing of evaluation. As shown in Figure 2, vascular damage following hyperthermia continue to progress after the completion of heat treatments. Therefore, we evaluated the vascular damage 24 hr after heat treatments, whereas they investigated the microcirculation stoppage during heat treatments. The transition in Arrhenius plot is within the range of 42-43°C in almost all experiments ever done.’ As the intratumor temperature was 0.2-0.3”C below the water bath temperature in our heating method, the true transition temperature in vascular damage for SCC VII seems to be between 42.7’C and 43.7”C, which is the same temperature range as the transition temperature in Arrhenius plot. The results of the growth delay study in SCC VII following hyperthermia (40-44°C 30 min) well reflected the degree of vascular damage estimated 24 hr after the heat treatment. Thus, the transition temperature in heat responsive in vivo is suggested to correspond with that in vascular damage. Tumor vessels of SCC VII were more vulnerable to heat than those of MCa. Decrease in the % vascular area began more slowly in MCa than in SCC VII, and the maximum decrease of the % vascular area observed at 24 hr after heating was higher in MCa than in SCC VII (Fig. 2). Histological examinations showed a marked difference in the architecture of tumor vessels between the 2 tumors. Most of the larger vessels in MCa were embedded in the interlobular connective tissues and the smaller vessels were supported by collagen (Fig. 4a),8 whereas tumor vessels in SCC VII consisted of only sparse endothelial cells (Fig. 3b). Vascular occlusion following hyperthermia was not observed frequently in the large tumor vessels of MCa surrounded by thick connective tissues (Fig. 4b). Similar observations have been made by sev-
419
Microangiographic and histologic analysis 0 Y. NISHIMURAet al.
6r
I
5 Time
10 After
Treatment
15
20
(day)
Fig. 6. Growth curves for unheated control tumors (0: SCC VII, ??: mammary adenocarcinoma) and tumors heated at 44°C for 30 min (0: SCC VII, Cl:mammary adenocarcinoma).
era1 investigators. *,*I It is thus suggested that tumor vessels supported by collagenous tissues are less vulnerable to heat than those without such support. The mechanism of heat-induced vascular damage is still obscure. One possible explanation is as fol~ows.‘~~~‘~,*~ First, the heat-induced retardation of blood flow in the tumor occurs. This is due partly to the combined results of swelling or lysis of endothelial cells, sticking of leukocytes to vessel walls, and an increase in rigidity of red blood cells. Thereafter, a mechanical or passive dilatation of the vessels by the increased blood flow into the tumors from adjacent normal tissues damage the fragile tumor vessels irreversibly. We consider that tumor vessels surrounded by connective tissues are strong against the mechanical dilatation because of the perivascular supporting structures. Recently, Waterman et al.*’investigated the changes in blood flow in human tumors during local hyperthermia, and reported that the sharp reduction in blood flow Table 1. Volume-doubling times after heat treatment for mammary adenocarcinoma and SCC VII carcinoma
Treatment Mammary adenocarcinoma Control 44’C, 30 min SCC VII carcinoma Control 44”C, 30 min 43”C, 30 min 42”C, 30 min 41”C, 30 min 4o’C, 30 min
Volume-doubling time (day)* (Mean + SD) 4.1 * 0.9
8.0 k 1.7 4.6 k 0.4 14.2 k 2.6
6.7 5 3.9 4.9 * 1.1 4.4 + 1.1 4.3 + 0.9
* The growth time to double its initial tumor volume.
shown in most rodent tumors did not occur after 30-50 min of heating at approximately 44°C. Thus, the vascular structure of human tumors was not as sensitive to heat as that in most of the transplantable rodent tumors reported. With the exception of the rapidly growing sarcomas, most human neoplasms have fibrous stroma which provide structural support for the parenchyma and carries the nutritional blood supply.” In this respect, the vascular system of MCa is similar to that of human neoplasms. On the other hand, most of the transplantable rodent tumors including SCC VII are devoid of stroma1 connective tissues. This difference in vascular architecture may explain the above mentioned result. SCC VII carcinoma, the tumor vessels of which were more vulnerable to heat than those of MCa, was more heat sensitive than MCa in the tumor growth delay assay. There are a number of factors which could influence the heat sensitivity of the 2 tumors. The difference in inherent cellular sensitivity to heat or the cell cycle kinetics between the 2 tumors could contribute to variations in heat sensitivity. There are many in vitro data that the intrinsic cellular sensitivity varies considerably among various cell lines,4 and that heat sensitivity depend on the phase of a cell cycle.4 However, thermosensitivity of tumor cells in vivo is quite different from that in vitro mainly due to the acidic and nutritionally deprived intratumor environment, which is closely related to blood flow to a tumor. ‘8,21,24,25 The acidic and poor nutritional environment is known to enhance the thermal injury to mammalian cells in vitro.4,24Rofstad et ~1.,*~~*’ who studied human melanoma xenografts, have shown that the heat response in vivo was not positively correlated with the intrinsic heat sensitivity of the tumor cells in vitro, and that the heat response in vivo is partly governed by physiological and microvascular conditions. Another intriguing phenomenon in the response of tu-
420
I. J. Radiation Oncology 0 Biology 0 Physics
mor cells in vivo is the delayed or secondary cell death after heating, S2’V24 that is, cells are killed during a period lasting up to a few days after completion of heat treatments as a consequence of vascular damage in heated tumors. In the present study, a close relationship between the vascular damage and the degeneration of tumor cells was observed in the histological sections (Fig. 4~). The clusters of tumor cell degeneration around destructed tumor vessels may represent the delayed cell death. It follows from these facts that blood perfusion plays the cardinal role in the tumor response in vivo. ‘7S24
August 1988, Volume 15,Number 2
Therefore, it is very likely that the extent of vascular damage correlates with heat responsiveness in vivo. As already mentioned, vasculature surrounded by fibrous connective tissues seems less heat sensitive than vasculature that is composed only of endothelial cells. Our findings on vascular damage, histological examinations, and thermosensitivity of the 2 tumors suggest-that tumors with abundant connective tissues which support their tumor vessels are less sensitive to heat than those without such perivascular structures. If this is the general rule, heat sensitivity of a tumor might be predicted roughly by the histological sections before heat treatments.
REFERENCES 1. Von Ardenne, M., Bohme, G., Kell, E.: On the optimization of local hyperthermy in tumors based on a new radiofrequency procedure. J. Cancer Res. Clin. Oncol. 94: 163184,1979. 2. Badylak, S.F., Babbas, C.F., Skojac, T.M., Voorhees, W.D., Richardson, R.C.: Hyperthermia-induced vascular injury in normal and neoplastic tissue. Cancer 56: 9911000,1985. 3. Van den Berg-Blok, A.E., Reinhold, H.S.: Time-temperature relationship for hyperthermia induced stoppage of the microcirculation in tumors. Int. J. Radiat. Oncol. Biol. Phys. 10: 737-740, 1984. 4. Dewey, W.C., Hopwood, L.E., Sapareto, S.A., Gerweck, L.E.: Cellular responses to combinations of hyperthermia and radiation. Radiology 123: 463-474, 1977. 5. Dudar, T.E., Jain, R.K.: Differential response of normal and tumor microcirculation to hyperthermia. Cancer Res. 44: 605-612, 1984. 6. Eddy, H.A.: Alterations in tumor microvasculature during hyperthermia. Radiology 137: 5 15-52 1,198O. 7. Emami, B., Nussbaum, G.H., Hahn, N., Piro, A.J., Dritschilo, A., Quimby, F.: Histopathological study on the effects of hyperthermia on microvasculature. Znt. J. Radiat. Oncol. Biol. Phys. 7: 343-348, 198 1. 8. Falk, P.: The vascular pattern of the spontaneous C3H mouse mammary carcinoma and its significance in radiation response and in hyperthermia. Eur. J. Cancer 16: 203-2 17,198O. 9. Field, S.B., Morris, C.C.: The relationship between heating time and temperature: Its relevance to clinical hyperthermia. Radiother. Oncol. 1: 179-186, 1983. 10. Hiraoka, M., Jo, S., Akuta, K., Nishimura, Y., Takahashi, M., Abe, M.: Radiofrequency capacitive hyperthermia for deep-seated tumors. II. Effects of thermoradiotherapy. Cancer60: 128-135,1987. 11. Jo, S., Hiraoka, M., Akuta, K., Nishimura, Y., Furuta, M., Takahashi, M., Abe, M.: Histopathological changes in transplanted mouse mammary carcinoma following hyperthermia with or without radiation. Jpn. J. Cancer Res. (Gann)77: 1095-l 101, 1986. 12. Kang, M.S., Song, C.W., Levitt, S.H.: Role of vascular function in response of tumors in vivo to hyperthermia. CancerRes. 40: 1130-l 135,198O. 13. Overgaard J.: Simultaneous and sequential hyperthermia and radiation treatment of an experimental tumor and its surrounding normal tissue in vivo. Int. J. Radiat. Oncol. Biol. Phys. 6: 1507-l 5 17, 1980. 14. Peterson, HI.: Tumor blood circulation. In Angiogenesis,
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
26.
27.
Vascular Morphology and Blood Flow of Experimental and Human Tumors. Boca Raton, FL CRC Press Inc. 1979, pp. 49-76. Raaphorst, G.P., Romano, S.L., Mitchell, J.B., Bedford, J.S., Dewey, W.C.: Intrinsic differences in heat and/or X ray sensitivity of seven mammalian cell lines cultured and treated under identical conditions. Cancer Res. 39: 396401,1979. Rappaport D.S., Song, C.W.: Blood flow and intravascular volume of mammary adenocarcinoma 13726A and normal tissues of rat during and following hyperthermia. Int. J. Radiat. Oncol. Biol. Phys. 9: 539-547, 1983. Reinhold, H.S., Van den Berg-Blok, A.: Enhancement of thermal damage to the microcirculation of “sandwich” turnouts by additional treatment. Eur. J. Cancer 17: 781795,198l. Reinhold, H.S., Endrich, B.: Invited review. Tumor microcirculation as a target for hyperthermia. Znt.J. Hyperther. 2: 111-137, 1986. Robins, S.L., Angell, M.: Basic Rathology, 2nd edition. Philadelphia, Sunders Company. 1976, pp. 68- 106. Rofstad, E.K., Brustad, T.: Response to heat treatment (42.5’C) in vivo and in vitro of five human melanoma xenografts. Br. J. Radiol. 57: 1023-1032, 1984. Rofstad, E.K., Brustad, T.: Primary and secondary cell death in human melanoma xenografts following hyperthermic treatment. Cancer Res. 46: 355-36 1, 1986. Rubin, P., Casarett, G.W., Kurohara, S.S., Fujii, M.: Microangiography as a technique, radiation effect versus artifact. Am. J. Roentgenol. 92: 378-387, 1964. Shibamoto, Y., Yukawa, Y., Tsutsui, K., Takahashi, M., Abe, M.: Variation in the hypoxic fraction among mouse tumors of different types, sizes, and sites. Jpn. J. Cancer Res. (Gann) 77: 908-915,1986. Song, C.W.: Effect oflocal hyperthermia on blood flow and microenvironment: A review. Cancer Res. 44: (Suppl.) 472 ls-473Os, 1984. Song, C.W., Kang, M.S., Rhee, J.G., Levitt, S.H.: The effect of hyperthermia on vascular function, pH, and cell survival. Radiology 137: 795-803, 1980. Stewart, F., Begg, A.: Blood flow changes in transplanted mouse tumors and skin after mild hyperthermia. Br. J. Radiol. 56: 477-482, 1983. Waterman, F.M., Nerlinger, R.E., Moylan III, D.J., Leeper, D.B.: Response of human tumor blood flow to local hyperthermia. Int. J. Radiat. Oncol. Biol. Phys. 13:7582, 1987.
15. pp. 41 I-420 Copyright
0360-3016/88 $3.00 + .OO 0 1988 Pergamon Press plc
??Original Contribution MICROANGIOGRAPHIC AND HISTOLOGIC ANALYSIS OF THE EFFECTS OF HYPERTHERMIA ON MURINE TUMOR VASCULATURE YASUMASA NISHIMURA, M.D., MASAHIRO HIRAOKA, M.D., SHIKEN Jo, M.D., KEIZO AKUTA, M.D., YUTAKA YUKAWA, R.T., YUTA SHIBAMOTO, M.D., MASAJI TAKAHASHI, M.D. AND MITSUYUIU ABE, M.D. Department of Radiology,Facultyof Medicine,Kyoto University,54 Shogoin-Kawahara-cho, Sakyo-ku,Kyoto 606,Japan The effects of hyperthermia on murine tumor vasculature were studied by microangiography and histological examination. The tumors used were SCC VII carcinoma and mammary adenocarcinoma of syngeneic 6H/He mice. For the quantitative analysis of microanglographic changes, the percent (%) vascular area, which was defined as the percentage of opacified tumor vessel area to the entire tumor area, was determined in each microanglogram. The % vascular area after heating in a water bath at 44°C for 30 min was minimized 24 hr after heating in both types of tumors. The histologic study revealed that the initial decrease of the 96 vascular area was due to congestion, thrombosis, and rupture of tumor vessels, and its subsequent increase was due to angiogenesis. SCC VII was more beat sensitive than mammary adenocarcinoma in terms of tumor growth delay, and tumor vessels of SCC VII were more vulnerable to heat than those of mammary adenocarcinoma. Histological examinations showed a marked difference in the architecture of vessels between the two types of tumors. Tumor vessels of mammary adenocarcinoma were supported by a connective tissue band, whereas those of SCC VII consisted of a single endothelial cell layer. Our findings suggest that the tumor vessels supported by a connective tissue band are less sensitive to heat than those without such support. The vascular damage of SCC VII was temperature dependent, and the critical temperature at which dramatic vascular damage appeared was between 42.7”C and 43.7”C. Hyperthermia, Tumor vasculature, Microangiography.
greater depth and length than can be accomplished by routine histologic methods. Using correlative histologic sections, microangiographic changes can be correlated to histopathologic lesions. By microangiography, circulating vessels can be distinguished from congested non-circulating vessels as contrast medium doses not enter the congested vessels. In addition, it can demonstrate the localization of vascular changes in a tumor, which cannot be detected by the functional studies. In the present study, the effect of hyperthermia on the microcirculation of two different types of transplantable mouse tumors was investigated by microangiography and on correlative histological sections.
INTRODUCTION studies have been performed to evaluate the changes in tumor blood flow following hyperthermia by functional methods, and significant changes of several circulatory parameters have been demonstrated in a variety of tumors as well as normal tissues.4’12*‘5*‘6724-26 The tumor vasculature can be markedly damaged at temperatures that may alter but do not damage the vasculature in normal tissues.4,5,‘6*‘4,24-26 In view of the morphological features of the tumor vessels, it is not surprising that the response of tumor vessels to heat differs from that in normal tissues. The tumor vascular bed, which is devoid of innervation, consists of single-layered endothelial cells without an external coat of elastic basement membrane. Tumor vessels are maximally dilated, irregularly constricted, tortuous, and bent abruptly.‘8,‘9*24 Microangiography is an excellent technique for observing vascular changes at the level of small arterioles and capillaries.22 The main advantage of microangiography is that it permits the study of the vasculature at a
Numerous
METHODS
AND
MATERIALS
Animals and tumors Eight-week-old C3H/He male mice were obtained from the Animal Center of Kyoto University. Mice were caged in groups of 8 or less at a constant temperature and given food and water ad libitum.
Reprint requeststo: Dr. Y. Nishimura. Acceptedfor publication2 March 1988.
Supportedin part by a Grant-in-Aidfor ScientificResearch (61010041,61015038)from the Ministry of Education, Science and Culture,Japan. 411
412
I. J. Radiation Oncology 0 Biology 0 Physics
The tumors used were spontaneous mammary carcinoma (MCa) of C3H/He mice and SCC VII carcinoma of C3H/He mice. Some of the biological characteristics of MCa and SCC VII were described previously.’ 1,23 MCa was minced, and preserved in several ampules in liquid nitrogen. When required, the content of an ampule was thawed and transplanted to new recipients, and the tumors from this second transplanted generation were used for the experiments. The tumors were minced with scissors in Hanks’s medium supplemented with 5% fetal calf serum. The suspension was kept in a test tube on ice for about 15 min to permit larger tissue pieces to settle. The supematant was then filtered through gauze to obtain a cell suspension. About 2 X 1O5viable cells in 10 ~1 were inoculated subcutaneously into the right hind thigh of the mice. The SCC VII carcinoma was thawed from original frozen stocks and maintained by alternate passage in syngeneic mice and in cell culture in Eagle’s minimum essential medium (MEM) supplemented with 12.5% fetal calf serum. Approximately 1 X lo5 cells collected from monolayer cultures were inoculated subcutaneously into the right hind thigh of mice. Hyperthermia The mice were held in a specially constructed jig with firm fixation of the tail and left leg by taping them to the jig. The right tumor-bearing leg was pulled down by a special sinker (approx. 45g), which was fixed to the skin of the toe with super glue* so as not to impair the blood flow of the leg. The fully awake mice were then placed on a circulating water bath? maintained at a desired temperature, and the extended right-llg was locally heated. The temperat& of the water bath was controlled to within 0.05”C. The tumor was immersed at least 10 mm below the water surface. The mice were air-cooled during the heat treatment. All temperatures mentioned in this paper refer to the water bath temperature. Intratumor temperature measurement Intratumor temperature was measured in 9 SCC VII tumors by inserting a 29 gauge needle type microthermocouple probe* into the tumors. The temperatures at the center and periphery of a tumor were obtained by moving the thermocouple probe within the tumor. Microangiography and histologic examination Microangiography was performed as previously described.22 The mice were anesthetized with sodium pentobarbital, and the thoracic aorta was cannulated with a 23 gauge needle and the vena cava was cut. After flushing
* Aronnarufa, Konishi Co. Ltd., Osaka, Japan. t Model ET-35P, Toyo Seisakusho Co. Ltd., Tokyo, Japan. $ MT-29/5, Sensortek, Clifton, N.J. 0 Ryubari Sol B, Maruishi Seiyaku Co. Ltd., Osaka, Japan.
August 1988, Volume 15, Number 2
the circulatory system with warmed heparinized saline, filtered barium sulfate solution$ (0.25g/ml) was injected with a pressure of 150 mmHg for 30 min. The animal’s heart was pulsated for about 5 min after the injection of the contrast medium. When the muscular vasculature of the left normal hind limb was not opacified sufficiently, the mouse was excluded from the study to avoid a poor filling artifact. After fixing the tumor with 10% buffered formalin, contact radiographs of 1 mm thick tumor slices were obtained** using fine-grain filmtt. The tumor slice was cut along a sagittal direction through its center with as much surrounding normal tissues as possible. Correlative histologic sections of 4 pm thickness were prepared for each tumor slice and stained with hematoxylin and eosin. The microangiographic findings were evaluated quantitatively by the following methods. On an enlarged (approx. 10X) microangiogram, avascular areas and the entire tumor were measured, using a digital planimeter$$. Thereafter, the percent (%) vascular area in each microangiogram was determined as follows; % vascular area =(l-
area of avascular portions (cm*) )X 100. area of the entire tumor (cm*)
The mean f SE of the % vascular area was obtained for 5-9 microangiograms. Growth delay study The response of the tumors to heat treatment was evaluated by the growth delay. Three perpendicular diameters of each tumor were measured every 2 days after treatment using a caliper. Experiments were performed when the tumors reached a volume-of approximately 600 mm3 (approx. 12 mm in the longest diameter). Each treatment group consisted of 1 l- 13 mice. RESULTS
Intratumor temperature measurements Tumor center temperatures were equilibrated within 3-4 min after immersion in the water bath, and remained 0.2-0.3”C below the water bath temperature (Fig. la). Temperature difference between the tumor center and the periphery was within 0.1°C (Fig. 1b). Sequential microangiographic and histologic changes following hyperthermia The time course of microangiographic and histologic changes following hyperthermia was investigated in SCC
** CSM-2, Softex Co. Ltd., Tokyo, Japan. tt Fuji Co. Ltd., Tokyo, Japan. $$ Planix 7, Tamaya Co. Ltd., Tokyo, Japan.
Microangiographic and histologic analysis 0 Y. NISHIMURA
I 0
I
I
I
I
2
4
6
8
Time
(a)
(mid
42.7 f O,l°C -
42,7 f 0,l “C 42,8 + O.l”c
Tumor
temperature
(b)
Water
bath
temperature
43,0 + 0,05”c
Fig. 1. (a) Tumor temperature as a function of time after immersion in a water bath kept at 43°C. Temperature profiles for the tumor center in 3 tumors. The tumor volume was approximately 500 mm3 (0), 600 mm3 (A) and 800 mm3 (e). The tumor center temperature stabilized to approximately 42.7”C within 3-4 min. (b) Intratumor temperature distribution when a tumor bearing leg was immersed for more than 5 min in the water bath maintained at 43’C.
413
et al.
24 hr after heating. Although tumor vessels showed nearly no filling in the microangiograms obtained 24 hr after heating, only small parts of tumor vessels adjacent to the underlying muscle were opacified (Fig. 3d). Histologically, dilated thrombotic tumor vessels with no apparent endothelial cells were observed in the entire tumor. Degeneration of tumor cells such as pyknosis and karyolysis was also noted. Three days after heating, the % vascular area increased to 29 + 17%. In the magnified microangiograms, saccular, and tapering capillary sprouts, which are typical morphological features of tumor angiogenesis,14 were observed at the inner edge of the opacified tumor vascular area (Fig. 3e). These angiogenesis inevitably occurred from the periphery of the tumors. Seven days after heating, the % vascular area further increased. Mouse mammary carcinoma (MCa). The unheated control tumors were totally opacified with irregular tumor vessels, and the control % vascular area was 99 ? 1%. Histologically, most of the large tumor vessels passed through the interlobular connective tissue (Fig. 4a), and the endothelial cells of small peripheral vessels were supported by collagen.’ In MCa, the % vascular area decreased slightly immediately and 3 hr after heating, and it decreased further 6 and 24 hr after heating (Fig. 2). The % vascular area differed significantly between the 2 types of tumors immediately and 3 hr after heating (p < 0.05, p < 0.001 respectively). Three and 7 days after heating, the % vascular area increased. Similar angiographic findings of angiogenesis observed in SCC VII were also noted in MCa. Histologically, some large vessels embedded in thick connective tissues were free from thrombosis 24 hr after heating (Fig. 4b). Tumor cells surrounding a bariumfilled circulating vessel were less damaged than those surrounding a congested or thrombosed vessel (Fig. 4~).
loo-
*.*
*.-. p”
:
mammary adenocarcinoma
2
VII and MCa. Figure 2 shows the changes in % vascular area for the tumors from 0- 168 hr after 30 min of heating at 44°C. XC VII carcinoma. The unheated control tumors were totally opacified with fine tumor vessels (Fig. 3a), and the % vascular area was 100 f 0% (mean + SE). Histologically, tumor vessels consisted of only a thin endothelial cell layer (Fig. 3b). The % vascular area decreased to 4 1 -+ 16% immediately after heating as focal avascular areas appeared on the microangiograms (Fig. 3~). Correlative histologic sections revealed the dilatation, congestion, and rupture of the tumor vessels. The % vascular area further decreased 3 hr after heating, and was lowest
f
50-
\
Time After
Heating
(hour)
Fig. 2. Changes in the % vascular area from 0 to 168 hr after heating at 44’C for 30 min. The % vascular area differed significantly between the tumors immediately and 3 hr after heating (p < 0.05, p < 0.00 1 respectively).
414
I. J. Radiation Oncology 0 Biology 0 Physics
August 1988, Volume 15,Number 2
Fig. 3. Microangiograms and histological sections of SCC VII tumors. (a) Fine tumor vessels were opacified in an unheated control tumor. (b) Histologically, tumor vessels consisted of only a sparse endothelial cell layer. (c) Immediately after heating at 44°C for 30 min, avascular areas appeared on the microangiogram. Opacified tumor vessels showed marked dilatation. (d) One day after heating, tumor vessels showed nearly no filling except for a small part of tumor vessels near the muscle (white arrows). (e) Three days after heating, angiogenesis was noted in the periphery of a tumor.
Microangiographic changes at dlyerent temperatures in see vzz As the maximum decrease of the % vascular area was observed 24 hr after heating (44°C 30 min) in both types of tumors, we investigated the changes in % vascular area 24 hr after heating at different temperatures. Figure 5 shows the changes in % vascular area 24 hr after heating at 40-44°C for 30 min in SCC VII. The % vascular area
of the tumors heated at 40°C was 98 f 1%. There were no substantial changes in the microangiograms and histologic sections at 40°C. The % vascular area decreased at temperatures of more than 4 1°C. In the microangiograms obtained 24 hr after heating at 4 1’C for 30 min, opacified vessels showed mild dilatation and small focal avascular areas appeared. The % vascular area at 41°C was 90 + 13%. % vascular areas
Microangiographic and histologic analysis 0 Y. NISHIMURAetal
(d)
415
Fig. 3. (Contd)
were further decreased at 42 and 43°C to 76 + 23% and 59 -t 19% respectively. Dilatation of tumor vessels and mosaic-like avascular areas were noted on the microangiograms in these temperature ranges. The % vascular area of tumors heated at 44°C decreased dramatically to 5 + 2%, and the difference in % vascular area between 43°C and 44°C was significant (p < 0.005). Growth delay study Figure 6 shows the growth curves for unheated control tumors and tumors heated at 44’C for 30 min. The expo-
nential part of the growth curves was nearly parallel for the unheated and the heated tumors. Table 1 shows the tumor growth time to reach twice its initial volume (volume-doubling time) for the tumors. Although these values for the unheated control tumors of SCC VII and MCa were nearly equal, those for the tumors heated at 44°C for 30 min differed significantly between the two types of tumors, indicating that SCC VII was more sensitive to heat than MCa. The growth delay in WC VII tumors heated at 4044°C for 30 min was also investigated. The volume-dou-
416
I. J. Radiation Oncology 0 Biology 0 Physics
(4
August 1988, Volume 15, Number 2
Fig. 3. (Contd)
bling times of the tumors heated at 40-42°C were equal to that of the control tumors (Table 1). The volume-doubling time of the tumors heated at 43°C was slightly prolonged, whereas that of the tumors heated at 44°C was markedly prolonged (p < 0.001). DISCUSSION Sequential changes in circulatory parameters of transplanted tumors following hyperthermia have been reported by several investigators using isotope-trapping methods.12,‘6*25,26Stewart et ~1.~~found that blood flow of the SAFA tumor was reduced 1 day after heating (42.5”C, 1 hr) and returned to the control range by 2-3 days. Kang et al. l2 observed that blood volume in SCK tumors was only one-tenth of the control value 5 to 12 hr after heating at 43.5”C for 30 min and started to recover slowly thereafter. In the present study, the change of % vascular area estimated by microangiography after 30 min of heating at 44°C showed similar tendency. The maximum decrease of the % vascular area was observed 24 hr after heating in both types of tumors, and it increased thereafter. From the functional analysis of the vascular changes, we could not determine whether the recovery of circulating parameters after heating were due to recanalization of tumor vessels or to angiogenesis. Our study revealed that the initial decrease of the % vascular area was due to congestion, thrombosis, and rupture of tumor vessels, and its subsequent increase was due to angiogenesis (Fig. 3). Recanalization of tumor vessels may not be ruled out completely only by observing the morphological features of tumor angiogenesis. However, only remnants of tu-
mor vasculature with no endothelial cells were observed in the histological sections obtained 24 hr after heating at 44°C. Similar histologic findings were made by Emami et al.’ and they have termed these vessels “ghosts of blood vessels”. Therefore, it is practically impossible that the re-opening of such degenerated tumor vessels occur. Vascular damage following hyperthermia at 44°C for 30 min was irreversible in small tumor vessels of murine tumors. As to localization of vascular changes in a tumor, tumor vessels adjacent to the underlying muscle (tumor peripheral zone) were to be protected from thermal damage (Fig. 3d). Consequently, angiogenesis after hyperthermia always occurred from the periphery of the tumor (Fig. 3e). Several experimental and clinical studies revealed similar vascular” and histological’0*“9’3 changes in the tumor after hyperthermia. One possible reason for this is the presence of a minute temperature difference between the tumor center and the peripheral zone. Although this difference was within O.l”C in our measurement (Fig. lb), the present invasive thermometry is too inaccurate to rule out this possibility. Another possibility is the difference in the architecture of the tumor vessels. The periphery of tumors is often well vascularized and well perfused while the blood supply to the tumor center is inadequate. This difference in vascular density and blood flow between the tumor center and the periphery might account for the differential vulnerability of tumor vessels by heat. Temperature dependency on vascular damage of murine tumor vasculature has been studied by several invesEmami et al.’found no significant changes tigators. 3*6.7~18 in the vasculature of the BA 1112 sarcoma after heating
Microangiographic and histologic analysis 0 Y. NISHIMURAet al.
for 40 min below 40.5”C, marked dilatation and congestion of the vessels after heating at 42.5’C, and massive hemorrhage and necrosis with rupture of vessel walls after heating at 44.5’C. Eddy,6 who studied a squamous cell carcinoma grown in transparent chambers, has observed reversible dilatation of the vessels at 41°C petechiation, stasis, thrombosis at 43°C and complete shutdown of the circulation after heating at 45°C for 30 min.
417
The present result on histological changes was quite similar to the observation by the investigators although the tumors used were different. In addition, our quantitative analysis of vascular changes by microangiography revealed a linear relationship between vascular damage and temperature ranging from 4 1‘C to 43°C and a “breaking point” temperature at which vascular damage occurred dramatically be-
(a)
04 Fig. 4. Histological sections of mouse mammary adenocarcinoma. (a) Most of the large tumor vessels passed through the interlobular connective tissue in a control tumor. (b) A large tumor vessel embedded in the thick connective tissue band was filled with contrast medium 24 hr after heating at 44°C for 30 min. (c) Tumor cells surrounding a barium-filled circulating vessel (arrow) were less damaged than those surrounding a congested vessel (open arrow).
418
I. J. Radiation Oncology 0 Biology 0 Physics
(c)
August 1988, Volume 15, Number 2
Fig. 4. (Co&d)
tween 43°C and 44°C (Fig. 5). Another quantitative analysis of heat-induced vascular damage was made by BergBlok et cd3 They calculated the time necessary to produce microcirculation stoppage in 50% of the treated tumors (ST50), using Rhabdomyosarcoma BA 1112 growing in observation chambers. The ST50 values for 42,42.5,43, and 43.5”C were 226, 152, 10 1, and 70 min respectively. Their data does not indicate the presence of the breaking point in the temperature range. The opposite conclusions on the breaking point may be due to a multitude of factors. Not only are different tumors used but also the assay system of vascular damage. Moreover, temperature ranges examined differ considerably, and
I
I
controls
40
,
,
41
42
Temperature
\* 43
44
(“Cl
Fig. 5. Changes of % vascular area in SCC VII 24 hr after heating (40-44’C, 30 min). % vascular area decreased dramatically
between 43’C and 44°C with a difference statistically significant (p < 0.005).
there may be an essential difference in timing of evaluation. As shown in Figure 2, vascular damage following hyperthermia continue to progress after the completion of heat treatments. Therefore, we evaluated the vascular damage 24 hr after heat treatments, whereas they investigated the microcirculation stoppage during heat treatments. The transition in Arrhenius plot is within the range of 42-43°C in almost all experiments ever done.’ As the intratumor temperature was 0.2-0.3”C below the water bath temperature in our heating method, the true transition temperature in vascular damage for SCC VII seems to be between 42.7’C and 43.7”C, which is the same temperature range as the transition temperature in Arrhenius plot. The results of the growth delay study in SCC VII following hyperthermia (40-44°C 30 min) well reflected the degree of vascular damage estimated 24 hr after the heat treatment. Thus, the transition temperature in heat responsive in vivo is suggested to correspond with that in vascular damage. Tumor vessels of SCC VII were more vulnerable to heat than those of MCa. Decrease in the % vascular area began more slowly in MCa than in SCC VII, and the maximum decrease of the % vascular area observed at 24 hr after heating was higher in MCa than in SCC VII (Fig. 2). Histological examinations showed a marked difference in the architecture of tumor vessels between the 2 tumors. Most of the larger vessels in MCa were embedded in the interlobular connective tissues and the smaller vessels were supported by collagen (Fig. 4a),8 whereas tumor vessels in SCC VII consisted of only sparse endothelial cells (Fig. 3b). Vascular occlusion following hyperthermia was not observed frequently in the large tumor vessels of MCa surrounded by thick connective tissues (Fig. 4b). Similar observations have been made by sev-
419
Microangiographic and histologic analysis 0 Y. NISHIMURAet al.
6r
I
5 Time
10 After
Treatment
15
20
(day)
Fig. 6. Growth curves for unheated control tumors (0: SCC VII, ??: mammary adenocarcinoma) and tumors heated at 44°C for 30 min (0: SCC VII, Cl:mammary adenocarcinoma).
era1 investigators. *,*I It is thus suggested that tumor vessels supported by collagenous tissues are less vulnerable to heat than those without such support. The mechanism of heat-induced vascular damage is still obscure. One possible explanation is as fol~ows.‘~~~‘~,*~ First, the heat-induced retardation of blood flow in the tumor occurs. This is due partly to the combined results of swelling or lysis of endothelial cells, sticking of leukocytes to vessel walls, and an increase in rigidity of red blood cells. Thereafter, a mechanical or passive dilatation of the vessels by the increased blood flow into the tumors from adjacent normal tissues damage the fragile tumor vessels irreversibly. We consider that tumor vessels surrounded by connective tissues are strong against the mechanical dilatation because of the perivascular supporting structures. Recently, Waterman et al.*’investigated the changes in blood flow in human tumors during local hyperthermia, and reported that the sharp reduction in blood flow Table 1. Volume-doubling times after heat treatment for mammary adenocarcinoma and SCC VII carcinoma
Treatment Mammary adenocarcinoma Control 44’C, 30 min SCC VII carcinoma Control 44”C, 30 min 43”C, 30 min 42”C, 30 min 41”C, 30 min 4o’C, 30 min
Volume-doubling time (day)* (Mean + SD) 4.1 * 0.9
8.0 k 1.7 4.6 k 0.4 14.2 k 2.6
6.7 5 3.9 4.9 * 1.1 4.4 + 1.1 4.3 + 0.9
* The growth time to double its initial tumor volume.
shown in most rodent tumors did not occur after 30-50 min of heating at approximately 44°C. Thus, the vascular structure of human tumors was not as sensitive to heat as that in most of the transplantable rodent tumors reported. With the exception of the rapidly growing sarcomas, most human neoplasms have fibrous stroma which provide structural support for the parenchyma and carries the nutritional blood supply.” In this respect, the vascular system of MCa is similar to that of human neoplasms. On the other hand, most of the transplantable rodent tumors including SCC VII are devoid of stroma1 connective tissues. This difference in vascular architecture may explain the above mentioned result. SCC VII carcinoma, the tumor vessels of which were more vulnerable to heat than those of MCa, was more heat sensitive than MCa in the tumor growth delay assay. There are a number of factors which could influence the heat sensitivity of the 2 tumors. The difference in inherent cellular sensitivity to heat or the cell cycle kinetics between the 2 tumors could contribute to variations in heat sensitivity. There are many in vitro data that the intrinsic cellular sensitivity varies considerably among various cell lines,4 and that heat sensitivity depend on the phase of a cell cycle.4 However, thermosensitivity of tumor cells in vivo is quite different from that in vitro mainly due to the acidic and nutritionally deprived intratumor environment, which is closely related to blood flow to a tumor. ‘8,21,24,25 The acidic and poor nutritional environment is known to enhance the thermal injury to mammalian cells in vitro.4,24Rofstad et ~1.,*~~*’ who studied human melanoma xenografts, have shown that the heat response in vivo was not positively correlated with the intrinsic heat sensitivity of the tumor cells in vitro, and that the heat response in vivo is partly governed by physiological and microvascular conditions. Another intriguing phenomenon in the response of tu-
420
I. J. Radiation Oncology 0 Biology 0 Physics
mor cells in vivo is the delayed or secondary cell death after heating, S2’V24 that is, cells are killed during a period lasting up to a few days after completion of heat treatments as a consequence of vascular damage in heated tumors. In the present study, a close relationship between the vascular damage and the degeneration of tumor cells was observed in the histological sections (Fig. 4~). The clusters of tumor cell degeneration around destructed tumor vessels may represent the delayed cell death. It follows from these facts that blood perfusion plays the cardinal role in the tumor response in vivo. ‘7S24
August 1988, Volume 15,Number 2
Therefore, it is very likely that the extent of vascular damage correlates with heat responsiveness in vivo. As already mentioned, vasculature surrounded by fibrous connective tissues seems less heat sensitive than vasculature that is composed only of endothelial cells. Our findings on vascular damage, histological examinations, and thermosensitivity of the 2 tumors suggest-that tumors with abundant connective tissues which support their tumor vessels are less sensitive to heat than those without such perivascular structures. If this is the general rule, heat sensitivity of a tumor might be predicted roughly by the histological sections before heat treatments.
REFERENCES 1. Von Ardenne, M., Bohme, G., Kell, E.: On the optimization of local hyperthermy in tumors based on a new radiofrequency procedure. J. Cancer Res. Clin. Oncol. 94: 163184,1979. 2. Badylak, S.F., Babbas, C.F., Skojac, T.M., Voorhees, W.D., Richardson, R.C.: Hyperthermia-induced vascular injury in normal and neoplastic tissue. Cancer 56: 9911000,1985. 3. Van den Berg-Blok, A.E., Reinhold, H.S.: Time-temperature relationship for hyperthermia induced stoppage of the microcirculation in tumors. Int. J. Radiat. Oncol. Biol. Phys. 10: 737-740, 1984. 4. Dewey, W.C., Hopwood, L.E., Sapareto, S.A., Gerweck, L.E.: Cellular responses to combinations of hyperthermia and radiation. Radiology 123: 463-474, 1977. 5. Dudar, T.E., Jain, R.K.: Differential response of normal and tumor microcirculation to hyperthermia. Cancer Res. 44: 605-612, 1984. 6. Eddy, H.A.: Alterations in tumor microvasculature during hyperthermia. Radiology 137: 5 15-52 1,198O. 7. Emami, B., Nussbaum, G.H., Hahn, N., Piro, A.J., Dritschilo, A., Quimby, F.: Histopathological study on the effects of hyperthermia on microvasculature. Znt. J. Radiat. Oncol. Biol. Phys. 7: 343-348, 198 1. 8. Falk, P.: The vascular pattern of the spontaneous C3H mouse mammary carcinoma and its significance in radiation response and in hyperthermia. Eur. J. Cancer 16: 203-2 17,198O. 9. Field, S.B., Morris, C.C.: The relationship between heating time and temperature: Its relevance to clinical hyperthermia. Radiother. Oncol. 1: 179-186, 1983. 10. Hiraoka, M., Jo, S., Akuta, K., Nishimura, Y., Takahashi, M., Abe, M.: Radiofrequency capacitive hyperthermia for deep-seated tumors. II. Effects of thermoradiotherapy. Cancer60: 128-135,1987. 11. Jo, S., Hiraoka, M., Akuta, K., Nishimura, Y., Furuta, M., Takahashi, M., Abe, M.: Histopathological changes in transplanted mouse mammary carcinoma following hyperthermia with or without radiation. Jpn. J. Cancer Res. (Gann)77: 1095-l 101, 1986. 12. Kang, M.S., Song, C.W., Levitt, S.H.: Role of vascular function in response of tumors in vivo to hyperthermia. CancerRes. 40: 1130-l 135,198O. 13. Overgaard J.: Simultaneous and sequential hyperthermia and radiation treatment of an experimental tumor and its surrounding normal tissue in vivo. Int. J. Radiat. Oncol. Biol. Phys. 6: 1507-l 5 17, 1980. 14. Peterson, HI.: Tumor blood circulation. In Angiogenesis,
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
25.
26.
27.
Vascular Morphology and Blood Flow of Experimental and Human Tumors. Boca Raton, FL CRC Press Inc. 1979, pp. 49-76. Raaphorst, G.P., Romano, S.L., Mitchell, J.B., Bedford, J.S., Dewey, W.C.: Intrinsic differences in heat and/or X ray sensitivity of seven mammalian cell lines cultured and treated under identical conditions. Cancer Res. 39: 396401,1979. Rappaport D.S., Song, C.W.: Blood flow and intravascular volume of mammary adenocarcinoma 13726A and normal tissues of rat during and following hyperthermia. Int. J. Radiat. Oncol. Biol. Phys. 9: 539-547, 1983. Reinhold, H.S., Van den Berg-Blok, A.: Enhancement of thermal damage to the microcirculation of “sandwich” turnouts by additional treatment. Eur. J. Cancer 17: 781795,198l. Reinhold, H.S., Endrich, B.: Invited review. Tumor microcirculation as a target for hyperthermia. Znt.J. Hyperther. 2: 111-137, 1986. Robins, S.L., Angell, M.: Basic Rathology, 2nd edition. Philadelphia, Sunders Company. 1976, pp. 68- 106. Rofstad, E.K., Brustad, T.: Response to heat treatment (42.5’C) in vivo and in vitro of five human melanoma xenografts. Br. J. Radiol. 57: 1023-1032, 1984. Rofstad, E.K., Brustad, T.: Primary and secondary cell death in human melanoma xenografts following hyperthermic treatment. Cancer Res. 46: 355-36 1, 1986. Rubin, P., Casarett, G.W., Kurohara, S.S., Fujii, M.: Microangiography as a technique, radiation effect versus artifact. Am. J. Roentgenol. 92: 378-387, 1964. Shibamoto, Y., Yukawa, Y., Tsutsui, K., Takahashi, M., Abe, M.: Variation in the hypoxic fraction among mouse tumors of different types, sizes, and sites. Jpn. J. Cancer Res. (Gann) 77: 908-915,1986. Song, C.W.: Effect oflocal hyperthermia on blood flow and microenvironment: A review. Cancer Res. 44: (Suppl.) 472 ls-473Os, 1984. Song, C.W., Kang, M.S., Rhee, J.G., Levitt, S.H.: The effect of hyperthermia on vascular function, pH, and cell survival. Radiology 137: 795-803, 1980. Stewart, F., Begg, A.: Blood flow changes in transplanted mouse tumors and skin after mild hyperthermia. Br. J. Radiol. 56: 477-482, 1983. Waterman, F.M., Nerlinger, R.E., Moylan III, D.J., Leeper, D.B.: Response of human tumor blood flow to local hyperthermia. Int. J. Radiat. Oncol. Biol. Phys. 13:7582, 1987.