Effect of Blood Flow Occlusion on Laser Hyperthermia for Liver Metastases

Effect of Blood Flow Occlusion on Laser Hyperthermiafor Liver Metastases

V. Muralidharan, FRACS, C. Malcontenti-Wilson, B.Sc., and Chris Christophi, M.D., FRACS

Department of Surgery, Monash University, Alfred Hospital, Commercial Road, Prahran 3181, Australia

Journal of Surgical Research 103, 165–174 (2002)doi:10.1006/jsre.2002.6365, available online at http://www.idealibrary.com on

Submitted for publication July 16,

Background. Interstitial laser hyperthermia (ILH) isan in situ ablative technique for the treatment of colo-rectal liver metastases. A significant factor limitingtumor destruction is hepatic blood flow. Modulation ofhepatic blood flow may increase the size of tumor ne-crosis achieved. Our aim was to investigate the effectof blood flow occlusion on ILH-induced necrosis inboth tumor and normal liver tissue.

Materials and methods. A model of colorectal livermetastases in male inbred CBA mice was used. ILHwas applied to normal liver and tumor tissue using abare optical quartz fiber from an SYL500 Nd:YAGsurgical laser generator, with and without hepaticblood flow occlusion, and the extent of necrosis wasstudied. Tumor blood flow was assessed by laserDoppler flowmetry and scanning electron micros-copy.

Results. Hepatic blood flow occlusion resulted in asignificant reduction in blood flow in normal liver tis-sue (37.9% � 5.8, P < 0.001) and in the periphery of thetumor (17.5% � 7.8, P < 0.001). It did not affect theblood flow in the center of the tumor (13.4% � 4.3, P �

0.07). ILH of normal liver tissue, at low power (2 W),with hepatic blood flow occlusion, resulted in a signif-icant increase in the diameter of necrosis. This effectwas not seen when higher power (5 W) was used innormal liver. No significant effect was noted withintumor tissue at either power setting.

Conclusion. The overall effect of hepatic blood flowocclusion in ILH-induced tissue necrosis appears to benegligible in tumor tissue. Its main applicability ap-pears to be at the tumor–host interface, where a de-crease in blood flow may lead to higher temperaturesand therefore to a greater degree of tumor celldestruction. © 2002 Elsevier Science (USA)

165

1; published online March 6, 2002

INTRODUCTION

Several local ablative techniques are now used totreat a specific subgroup of patients with colorectalliver metastases [1, 2]. The most commonly used mo-dalities are radiofrequency ablation and interstitiallaser hyperthermia. The major advantages of thesetechniques include a minimally invasive approach, lowmorbidity, and cost effectiveness. The main disadvan-tage is the limitation of the size of tumor necrosisachieved. At present, using a Nd:YAG source and asingle bare optical fiber, the maximum diameter oftissue necrosis approaches 10–15 mm [3, 4]. A numberof methods have been postulated to increase the size oftissue necrosis, including multiple fiber application [5,6], diffuser tipped fibers [7–9] and hepatic blood flowmodulation [10–13]. Hepatic blood flow in particularappears to be a significant factor limiting the size oftissue necrosis by rapid dissipation of heat, preventingthe temperature rising to cytocidal levels [10–12, 14].This study investigated the effects of hepatic blood flowocclusion on interstitial laser hyperthermia (ILH)-induced tumor necrosis in a murine model of colorectalliver metastases.

MATERIALS AND METHODS

Liver Metastases Model

A mouse model of colorectal liver metastasis, created by intrasplenicinjection of DMH-induced colon cancer cells, was used for the study.This model has been previously characterized in detail in our laboratoryand utilizes 4- to 6-week-old inbred male CBA mice [15, 16].

Preparation of Cell Suspension

Mice with flank tumors were killed by anesthetic overdose, and theflank tumor was dissected out, minced, and dissociated by incubating

Key Words: colorectal carcinoma; liver metastases;interstitial; laser; hyperthermia.

with 0.1% trypsin in 0.1% glucose for 10 min. Trypsin was inactivatedby an equal volume of Dulbecco’s modified eagle’s medium (DMEM) andcentrifuged for 5 min at 1000 rpm, producing 150g centrifugal force

200

0022-4804/02 $35.00© 2002 Elsevier Science (USA)

All rights reserved.

(Beckman Model TJ-6 centrifuge, Palo Alto, CA). The pellet was resus-pended in 5 ml of 0.02% EDTA in phosphate-buffered saline (PBS) andcentrifuged again in a similar manner. The second pellet was resus-

pended in Ringer’s solution with 0.1% glucose, a viable cell count per-formed by trypan blue exclusion (0.05% trypan blue in distilled water),and the cell concentration adjusted to 1 � 106 cells/ml.

FIG. 1. (a) Relationship of the energy applied to the size of necrosis at a 2-W power setting in liver tissue in the presence of normal bloodflow and under vascular occlusion. Vascular occlusion resulted in significantly larger areas of necrosis (*P � 0.05). (b) Relationship of theenergy applied to the size of necrosis at a 5-W power setting in liver tissue in the presence of normal blood flow and under vascular occlusion.Vascular occlusion made no difference to the area of necrosis attained (*P � 0.05).

166 JOURNAL OF SURGICAL RESEARCH: VOL. 103, NO. 2, APRIL 2002

Induction of Metastases

Mice were anesthetized with an anesthetic mixture consisting of2% ketamine (Parke Davis, Australia) and 0.1% xylazine (Bayer,

Australia) in 0.9% saline at a dose of 0.05–0.1 ml/10 g of body weight,given intraperitoneally. The spleen was exteriorized through a leftsubcostal incision and 0.1 ml of tumor cell suspension (100,000 cells)was slowly injected into it using a 30-gauge needle. The spleen was

FIG. 2. (a and b) Relationship of the energy applied to the size of necrosis at 2- and 5-W power settings in tumor tissue in the presence ofnormal blood flow and under vascular occlusion. Vascular occlusion made no difference to the area of necrosis attained in tumor tissue (*P � 0.05).

167MURALIDHARAN ET AL.: LASER HYPERTHERMIA FOR LIVER METASTASES

compressed for 2 min to prevent spillage and allow the tumor cells toenter the portal circulation. Splenectomy was then performed withcautery, hemostasis achieved, and the laparotomy closed in layers.Animals were observed until recovery and then monitored daily for21 days. This resulted in the development of multiple but discreetmetastases in the liver by day 21.

Interstitial Laser Hyperthermia

Animals with established liver metastases at 21 days after induc-tion of tumors underwent anesthesia by intraperitoneal injection of2% ketamine (Parke Davis) and 0.1% xylazine (Bayer). A singletumor, appropriate (�10 mm) for application of laser hyperthermia,was chosen from each experimental animal and a 600-�m-diameterbare optical quartz fiber was inserted into its center. In the controlgroups, the fiber was similarly inserted into normal liver tissue. Theoptical fibers were connected to an SYL5100 Nd:YAG (Laserex Int.,Australia) laser source, producing laser energy with a wavelength of1064 nm. Three separate groups and controls were used in thisstudy.

Study group A. In this group the effects of ILH on tumor tissuewere investigated in the presence of normal hepatic circulation.Following cessation of treatment the livers were removed and fixedin Formalin for measurement of tissue necrosis. Histological evalu-ation of the immediate effects of laser hyperthermia was performedusing H&E staining. Two power settings of 2 and 5 W were used forthis study. At each power setting the laser was activated for the

specific time exposures of 5, 10, 15, 20, 50, 70, 90, and 120 s. A totalof 10 measurements were obtained for each specific setting and themean value was calculated.

Study group B. In this group the effects of ILH on normal livertissue were investigated in the presence of normal hepatic circula-tion. Power and exposure times were matched to those in group Aand assessed similarly.

Study group C. The effect of ILH on tumor tissue was investi-gated in the presence of hepatic blood flow occlusion in this group.Blood flow occlusion was achieved by placing a vascular clamp acrossboth the hepatic artery and the portal vein prior to the commence-ment of treatment. Power and exposure times were matched to thosein group A and assessed similarly.

Study group D. The effect of ILH on normal liver tissue wasinvestigated in the presence of hepatic blood flow occlusion in thisgroup. Blood flow occlusion, power, and exposure times were simi-larly matched to those in group C.

Several measurements of the diameter of each thermal injury weremeasured and the mean was calculated. The relationship betweenthe total energy applied and the extent of tissue necrosis in normalliver and tumor tissue was determined (Figs. 1 and 2).

Laser Doppler Flowmetry

Following anesthesia and laparotomy, blood flow in the liver andtumor was measured using a Laserflo blood perfusion monitor (TSIModel 403A, MN). After initial stabilization for 10 min blood flow

FIG. 3. Tumor blood flow measured in relation to normal liver blood flow by laser Doppler flowmetry. Blood flow at the center decreasessignificantly with increasing tumor size.


measurements were obtained from normal liver and tumor tissue at1-min intervals. In tumor tissue readings were taken from the centerand the periphery. Eight reading were taken from each point and themean was calculated. The relative blood flow of the tumor wasexpressed as a percentage of normal liver blood flow in each animal.

After a period of stabilization, readings were taken from the centerof the tumor, the periphery of the tumor (at the tumor–host inter-face), and the surrounding normal liver tissue. A vascular clamp wasthen applied to the hepatic artery and the portal vein and thereadings were repeated. The mean of eight readings at each specificpoint was obtained. The relative tumor blood flow and the fall intumor and normal liver blood flow due to vascular occlusion wereexpressed as percentages of normal liver perfusion.

Electron Microscopy

The tumor microvasculature was assessed by corrosion castingand scanning electron microscopy. Following anesthesia a thoracot-omy was performed and the thoracic aorta canulated. The vascularsystem was flushed with warm saline containing heparin (10 IU/ml)(Weddell Pharmaceuticals, NSW, Australia), papaverine (12 mg/ml)(David Bull Laboratories, VIC, Australia), and 6% polyvinyl pyrolli-done (PVP40) (Sigma, Australia), the effluent being dischargedthrough a right atrial puncture. This was followed by the injection of5 ml of an acrylic resin of Mercox CL-2B (Okenshoji Co, Tokyo,Japan), methyl methacrylate (Sigma, St. Louis, MO), and catalystMA (Vilene Med Co., Japan) infused at a pressure of 160 mmHg toallow microvascular filling. The resin was allowed to polymerizeovernight at room temperature, and the liver excised and digested in20% potassium hydroxide (KOH) at 37°C. The casts were preparedand mounted on aluminium stubs with electrodag 415 (AchesonColloids, MI), gold coated with a Baltec SCD005 sputtercoater, andviewed on a scanning electron microscope (Hitachi SEM). Scanning

electron micrograph digital images for all tumors visible in eachspecimen were captured using Spectrum Imaging software at �40,�80, and �150 magnifications. SEM was performed on untreatedand ILH-treated tumors.

Histopathology

Histopathological assessments of both tumor and normal livertissue were carried out following ILH as previously described. Thecharacteristics of necrosis were assessed by H&E staining on paraf-fin sections.

Statistical Analysis

Results were expressed as means � SD and were normally dis-tributed. They were compared using a multiple analysis of variance(ANOVA) and by post hoc comparison using a least significant dif-ference (LSD) test. (Statsplus, Statsoft, OK). Probability values lessthan 0.05 were considered significant.

RESULTS

Power, Exposure Time, and Extent of Necrosis

In normal liver the initial rapid increase in the sizeof necrosis was followed by a more gradual increaseafter application of 50 J of energy at a power setting of2 W. ILH at a 2-W power setting in the presence ofhepatic blood flow occlusion produced a significantlylarger diameter of necrosis than that with normal he-patic blood flow (Fig. 1a). The maximum diameters of

FIG. 4. The effect of blood flow occlusion on laser Doppler flowmetry in normal liver and in tumors 10 mm in diameter. Relative tumorblood flow is measured at the tumor periphery and it’s center. Blood flow occlusion produces the greatest reduction in liver tissue followedby tumor periphery. Minimal change is seen in the tumor center.


necrosis achieved after 240 J of energy applied at 2 Wwith and without blood flow occlusion were 7.3 � 0.4and 6.8 � 0.7 mm, respectively (P � 0.05). Hepaticblood flow occlusion did not produce this effect whenILH was performed at a higher power setting of 5 W(Fig. 1b). The maximum diameters of necrosis achievedafter 240 J of applied energy at 5 W with and withoutblood flow occlusion were 7.9 � 0.4 and 7.8 � 1.3 mm,respectively (P � 0.05).

No significant increase in the diameter of necrosiswas achieved by hepatic blood flow occlusion when ILHwas applied to tumor tissue, at both 2- and 5-W powersettings (Figs. 2a and 2b). At a 2-W power setting themaximum diameter of necrosis achieved was 7.7 � 0.8mm with normal hepatic blood flow and 7.8 � 0.7 mmwith hepatic blood flow occlusion (P � 0.05). Simi-larly, at a 5-W power setting the maximum values were8.5 � 0.6 and 7.8 � 0.9 mm, respectively.

Laser Doppler Flowmetry

Blood flow in the tumor tissue was lower than that ofnormal liver. Relative blood flow at the center of thetumor decreased as the tumor size increased (Fig. 3). Intumors 10 mm or larger in diameter relative blood flowwas found to differ between the center and the periph-ery of the tumor (Fig. 4). Central tumor blood flow was29.8 � 16.9% of normal liver blood flow. It was 41.0 �10.6% of normal liver blood flow at the periphery of thetumor, nearing statistical significance (P � 0.07).

Hepatic blood flow occlusion resulted in a differential

reduction in relative blood flow rates in normal liverand tumor tissue. In normal liver tissue hepatic bloodflow occlusion reduced blood flow to 37.9 � 5.8% of itsnormal value. In tumor tissue this reduced blood flowfrom 29.8 � 17.1 to 17.6 � 7.8% (P � 0.07) at thecenter and from 41.0 � 10.6 to 14.4 � 5.1% (P �0.001) at the periphery (Fig. 4).

Scanning Electron Microscopy

Vascular corrosion casting revealed the structure ofnormal liver sinusoids to closely resemble human liverarchitecture (Fig. 5). Tumor vasculature consisted oflarge distorted sinusoidal vessels, which had multiplefeeding vessels communicating directly from the he-patic sinusoids. These feeding vessels were foundthroughout the tumor–host interface. In tumors largerthan 10 mm in diameter a thin rim of compressed liversinusoids formed the peripheral pseudocapsule of thetumor. In these the peripheral vessels connecting thecompressed liver sinusoids with the central tumorlakes were noted to be thin (Fig. 6). The central vascu-lar lakes were large and tortuous. In contrast, smalltumors had large and dilated peripheral vessels andshowed the absence of a compressed sinusoidal pseudo-capsule. Tumors as small as 1.0 mm in diameter werefound to have well-developed microvasculature (Fig. 7).Following ILH, casts of tumor vasculature revealed

FIG. 5. SEM of a vascular corrosion cast of normal liver tissue.The architecture is similar to human structure, with central veins(CV) and hexagonal distribution of portal triads.

FIG. 6. SEM of a cross section through a vascular corrosioncast of a liver metastasis. A thin rim of compressed liver sinusoidsform the periphery of the tumor (P), separating disordered andlarge tumor vessels (TV) from normal liver sinusoids (LS). Theincreasing size and complexity of tumor vessels toward the centercan be noted (C).


uniform tissue destruction in the center, as evidencedby the nonfilling of the central tumor vasculature. Theconsistent finding was the presence of intact tumorsinusoids at the tumor–host interface (Fig. 8).

Histopathology

The macroscopic and histopathological changes pro-duced by ILH in tumor and normal liver tissue havebeen described in our previous study. Central cavita-tion surrounded by concentric zones of charring, acel-lular coagulum, and thermally injured tissue were seenin both liver and tumor tissue. In liver a zone of hyper-emia at the periphery of the thermal injury demar-cated the zone of injury from uninjured tissue. Thiswas absent in tumor tissue. The demarcation betweenareas of thermal injury and surviving tumor tissuewere clearly evident (Figs. 9a and 9b). There were nomacroscopic or histopathological differences seen whenILH was performed in the presence of hepatic bloodflow occlusion in both tumor and normal liver tissue.

DISCUSSION

Surgical resection provides the only means of poten-tial cure for patients with colorectal liver metastases[17]. Liver resection, however, is associated with sig-nificant morbidity and is only applicable to a smallspecific group of patients (10%) [18]. Locally ablative

techniques, such as ILH and radiofrequency ablation(RFA), offer a further potential method for the treat-ment of colorectal liver metastases [19]. Advantages ofthese techniques over conventional liver resection in-clude wider applicability, minimal morbidity, and costeffectiveness. Apart from the lack of evidence regard-ing long-term survival the major limitation of thesetechniques is the size and completeness of tumor ne-crosis achieved.

In ILH several factors influence the size of thermalinjury. These include the type of fiber used, the char-acteristics of the tissue, and the range of temperaturesachieved. A cytocidal effect occurs at temperaturesabove 40°C. The role of hepatic blood flow in the dissi-pation of heat has been shown to be of importance inlimiting tissue necrosis and has been alluded to in alimited number of previous studies [10–12, 14].

Using hepatic blood flow occlusion and a single fiberILH technique Moller et al. produced a twofold increasein tissue necrosis in normal pig livers from 10.8 to 21.9mm [10]. Similarly, Heisterkemp et al. reported a five-fold increase in coagulated volume in the normal pigliver using a combination of hepatic blood flow occlu-sion and multiple diffuser tipped fibers. Of interest, inthis study, portal vein occlusion with a patent hepaticartery increased the volume of tissue necrosis from 6.4to 30.6 cm3. The additional occlusion of hepatic arteryflow did not result in any further significant increase[12, 13]. Other studies using a combination of hepaticblood flow occlusion and multiple diffuser-tipped fibers

FIG. 7. SEM of a vascular corrosion cast of a small liver metas-tasis, less than 1.0 mm in diameter. Feeding vessels originatingdirectly from the surrounding sinusoids leading into the large tumorvascular lakes can be distinctly seen (arrows). Note that the smallertumor has relatively large peripheral vasculature (P).

FIG. 8. SEM of a vascular corrosion cast of a liver metastasisfollowing application of ILH. Complete and uniform central tumordestruction is surrounded by segments of intact tumor vasculatureat the tumor–host interface (arrow).


in normal liver tissue have confirmed these findings[20, 21].

Few data exist on the effect of hepatic blood flowocclusion on ILH-induced necrosis in tumor tissue.Sturesson et al. reported the effects of ILH in thepresence of hepatic blood flow occlusion in an im-planted model of liver tumor in rats, the diameter oftissue necrosis increasing from 13.1–14.2 mm with nor-

mal hepatic blood flow to 19.2–21.0 mm when hepaticblood flow was occluded. However, the mean diameterof tumor prior to treatment (11.6 � 3.0 mm) was sig-nificantly smaller than both results; thus, the increasein tissue necrosis with hepatic blood flow occlusion maybe attributable to its effect on the normal liver tissuesurrounding the tumor [11].

In attempting to quantify the effects of blood flow

FIG. 9. (a and b) H & E-stained histological section (3 �m) of a colorectal liver metastasis after application of ILH. Note the uniform areaof ILH-induced tumor necrosis (N) denoted by cells with shrunken nuclei and poorly stained cytoplasm which contrast with the deeply stainedviable tumor cells with large intact nuclei (T). The demarcation between the two zones is easily determined (arrows).


occlusion on ILH-induced tumor destruction we consid-ered a number of end points, namely, temperaturemeasurements and tissue necrosis. Measurement oftissue temperature at the various zones, using ther-mistor probes, presented some technical difficulty inthe relatively small tumor model we utilized, althoughthe temperature changes in these zones in relation toblood flow modulation would have added more infor-mation on the physiological changes that take placeduring ILH. However, the ultimate aim of this studywas to assess the ability of blood flow occlusion toenhance the ability of ILH to produce tumor necrosis,and it is this effect we chose as the end point for thestudy.

We found a significant increase in tissue necrosiswithin normal liver following ILH at 2 W in the pres-ence of hepatic blood flow occlusion (Fig. 1a). WhenILH was performed at the higher power of 5 W, thisdifference in normal liver tissue was negated (Fig. 1b).It is likely that the higher rate of energy deposition bythe bare quartz fiber at 5 W is able to overcome theheat sink effect of hepatic blood flow. In contrast, he-patic blood flow occlusion produced no significant dif-ference in the size of necrosis achieved within tumortissue when compared to that produced in the presenceof normal hepatic blood flow (Figs. 2a and 2b).

The conflicting results may be due to the consider-able heterogeneity of hepatic tumor vasculature [22,23]. However, the role of the hepatic artery as thepredominant source of blood flow to liver metastasesremains inconclusive, with widely varying experienceby researchers [23]. Although blood flow studies re-vealed a reduction in tumor blood flow resulting fromhepatic arterial ligation in experimental settings [24–26], its therapeutic value has been shown to be mini-mal. Tumor vasculature also has minimal response tophysiological and pharmacological stimuli, unlike nor-mal liver microvasculature [27–29].

The study of the microvasculature of the liver me-tastases in our model [16] has confirmed the presenceof multiple microvessels feeding the tumor vasculaturedirectly from hepatic sinusoids. This may well be re-sponsible for the minimal therapeutic response to he-patic arterial ligation seen in previous studies. Rela-tive blood flow within the center of the tumor wassignificantly less than normal liver flow and reducedfurther with increasing tumor diameter (Fig. 3). Intumors larger than 10 mm, the relative blood flow atthe periphery of the tumor was significantly higherthan that at the center, while still being lower thannormal liver blood flow (Fig. 4).

The greatest reduction in blood flow with hepaticblood flow occlusion occurred within normal liver tis-sue, the flow falling to 37.9 � 5.8% (P � 0.001) ofnormal. In tumor tissue this effect was less pro-nounced, but significant at the periphery, at the

tumor–host interface (41.0 � 10.6 to 14.4 � 5.1%, P �0.001). There was no significant reduction in bloodflow at the center of the tumor (29.8 � 17.1 to 17.6 �7.8% P � 0.007). The overall effect of hepatic bloodflow occlusion was to reduce the differential blood flowgradient between normal liver and tumor tissue. Theexisting low blood flow within central tumor tissue andthe insignificant reduction brought about by blood flowocclusion strongly support the fact that the heat sinkeffect is produced by the higher blood flow in the nor-mal liver tissue.

In conclusion hepatic blood flow occlusion appears toproduce a significant increase in tissue necrosis in nor-mal liver tissue but has minimal effect within tumortissue. This may be attributable to the existing lowblood flow rates evident within the center of tumortissue and the minimal effects of blood flow occlusion.The greatest reduction in tumor blood flow occurred atthe periphery of the tumor (tumor–host interface).Therefore, one possible application of blood flow occlu-sion in achieving greater tumor cell destruction is atthe tumor–host interface, where tumor blood flow ap-proaches that of normal liver blood flow. Viable tumorcells have been evident at the periphery of ILH-treatedtumors in some clinical studies. The reduction in pe-ripheral tumor blood flow may achieve increased tem-peratures at the tumor periphery, resulting in morecomplete tumor cell destruction at the tumor–host in-terface.

REFERENCES

1. Vogl, T. J., Mack, M. G., Muller, P. K., et al. Interventional MR:Interstitial therapy. Eur. Radiol. 9: 1479, 1999.

2. De Sanctis, J. T., Goldberg, S. N., and Mueller, P. R. Percuta-neous treatment of hepatic neoplasms: A review of currenttechniques. Cardiovasc. Intervent. Radiol. 21: 273, 1998.

3. Matthewson, K., Coleridge-Smith, P., O’Sullivan, J. P., et al. Bio-logical effects of intrahepatic Neodymium:Yttrium-Aluminium-Garnet laser photocoagulation in rats. Gastroenterology 93: 550,1987.

4. Matsumoto, R., Selig, A. M., Colucci, V. M., and Jolesz, F. A.Interstitial Nd:YAG laser ablation in normal rabbit liver: Trialto maximize the size of laser induced lesions. Lasers Surg. Med.12: 650, 1992.

5. Steger, A. C., Lees, W. R., Hovron, P. S., et al. Multiple fiberlow-power interstitial laser hyperthermia: Studies in the nor-mal liver. Br. J. Surg. 79: 139, 1992.

6. Steger, A. C., Bown, S. G., and Clark, C. G. Interstitial laserhyperthermia—Studies in the normal liver. Br. J. Surg. 75:598, 1988.

7. Moller, P. H., Lindberg, L., Henriksson, P. H., et al. Interstitiallaser thermotherapy: Comparison between bare fiber and sap-phire probe. Lasers Med. Sci. 10: 193, 1995.

8. Sweetland, H. M., Wyman, A., and Rogers, K. Evaluation of theeffect on the normal liver of interstitial laser hyperethermiausing artificial sapphire probes. Lasers Med. Sci. 8: 99, 1993.

9. Heisterkamp, J., van Hillegersberg, R., Sinofsky, E., and Ijzer-mans, J. N. M. Heat resistant cylindical diffuser for interstitial


laser coagulation: Comparison with the bare tip fiber in a por-cine liver model. Lasers Surg. Med. 20: 304, 1997.

10. Moller, P. H., Hannesson, P. H., Ivarsson, K., et al. Interstitiallaser thermotherapy in pig liver: Effect of inflow occlusion onextent of necrosis and ultrasound image. Hepatogastroenterol-ogy 44: 1302, 1997.

11. Sturesson, C., Liu, D. L., Stenram, U., and Andersson-Engels,S. Hepatic inflow obstruction increases the efficacy of intersti-tial laser-induced thermotherapy in rats. J. Surg. Res. 71(1):67, 1997.

12. Heisterkamp, J., van Hillegersberg, R., Mulder, P. G. H., et al.Importance of eliminating portal blood flow to produce largeintrahepatic lesions with interstitial laser coagulation. Br. J.Surg. 84: 1245, 1997.

13. van Hillegersberg, R., Heisterkamp, J., Sinofsky, E. L., et al.Increasing lesion size in interstitial laser coagulation: The im-portance of eliminating hepatic perfusion. In Anderson, R. R.(Ed.), Lasers in Surgery: Advanced Characterization Therapeu-tics and Systems VII, Vol. 2970. San Jose, California: SPIE—Int. Soc. Opt. Eng., 1997. Pp. 538–542.

14. Albrecht, D., Germer, C. T., Bock, A., et al. Laparoscopic ap-proach for laser induced thermotherapy (LITT) of liver tumors.Br. J. Surg. 81: 84, 1994.

15. Kuruppu, D., Christophi, C., Bertram, J. F., and O’Brien, P. E.Characterization of an animal model of hepatic metastasis. J.Gastroenterol. Hepatol. 11: 26, 1996.

16. Kuruppu, D., Christophi, C., and O’Brien, P. E. Microvasculararchitecture of hepatic metastases in a mouse model. HPBSurg. 10: 149, 1997.

17. Steele, G., and Ravikumar, T. S. Resection of hepatic metasta-ses from colorectal cancer. Ann. Surg. 210(2): 127, 1989.

18. Nakamura, S., Suzuki, S., and Baba, S. Resection of liver me-tastases of colorectal carcinoma. World J. Surg. 21: 741, 1997.

19. Vogl, T. J., Mack, M. G., Straub, R., et al. Percutaneous MRI-guided laser induced thermotherapy for hepatic metastases forcolorectal cancer. Lancet 350: 29, 1997.

20. Albrecht, D., Germer, C. T., Isbert, C., et al. Interstitial lasercoagulation: Evaluation of the effect of normal liver blood per-fusion and the application mode on lesion size. Lasers Surg.Med. 23: 40, 1998.

21. Germer, C. T., Albrecht, D., Roggan, A., et al. Experimentalstudy of laparoscopic laser-induced thermotherapy for liver tu-mors. Br. J. Surg. 84: 317, 1997.

22. Peterson, H. I. Modification of tumor blood flow—A review. Int.Radiat. Biol. 60(1/2): 201, 1991.

23. Suzuki, T., Sarumaru, S., Katsumi, K., and Honjo, I. Study ofvascularity of tumors of the liver. Surg. Gynaecol. Obstet. 134:27, 1972.

24. Ackerman, N. B. Experimental studies on the circulatory dy-namics of intrahepatic tumor blood supply. Cancer 29(2): 435,1972.

25. Mooney, B., and Taylor, I. The effect of hepatic arterial ligationon spontaneously developing colorectal liver metastases in therat. Clin. Oncol. 8: 231, 1982.

26. Ackerman, N. B., and Hodgson, W. B. Vascular patterns of livertumors and their consequences for different therapeutic ap-proaches. Recent Results Cancer Res. 100: 248, 1986.

27. Inoue, Y., Machida, K., Honda, N., et al. Effect of angiotensin IIon arteriovenous shunting assessed by hepatic arterial perfu-sion scintigraphy. Am. J. Clin. Oncol. 20(3): 237, 1997.

28. Forrest, E. H., Jalan, R., Redhead, D. N., and Hayes, P. C.Differing actions of the acute administration of propranolol andisosorbide-5-mononitrate on the portal circulation. AlimentPharmacol. Ther. 10(5): 795, 1996.

29. Burton, M. A., and Gray, B. N. Redistribution of blood flow inexperimental hepatic tumors with noradrenaline and propran-olol. Br. J. Cancer. 56: 585, 1987.


Documents

Effect of Blood Flow Occlusion on Laser Hyperthermia for Liver Metastases