- Email: [email protected]
Intraoperative liver radiation after partial hepatectomy in a rat model
JOURNAL
OF SURGICAL
RESEARCH
lntraoperative
53,287-292
(19%)
Liver Radiation after Partial Hepatectomy
in a Rat Model
BURTON L. EISENBERG, M.D., FACS,* RACHELLE M. LANCIANO, M.D.,? MICHAEL L. NUSSBAUM, M.D.,* ANDRES KLEIN-SZANTO, M.D.$ AND DOUGLAS D. TAYLOR, PH.D.* Fox Chase Cancer $Pathology,7701
Center, Department Burholme Avenue, Submitted
for publication
Academic
Press,
March
tRadiation Therapy, and Pennsylvania 19111 14, 1991
been used to enhance local control in the post-hepatectomy liver when resection margins are inadequate. Gastrointestinal toxicity in IORT trials is well documented [6]. However, because of the lack of clinical and laboratory experience, there is little information regarding the effect of a large fraction of irradiation to the post-resection liver. In this study, a rat model was employed to evaluate histological changes and DNA turnover as an indication of hepatic regenerative capacity following partial hepatectomy and IORT. The purpose of the study was to evaluate response of the normal liver parenchyma to varying doses of electron beam irradiation directed to the post-resection liver edge.
Hepatic resection of metastatic tumor is a treatment option in selected patients. Resection margin is a prognostic factor of hepatic recurrence and survival. Although intraoperative radiation therapy (IORT) has been clinically useful in some gastrointestinal cancers, there is little information regarding its use following hepatic metastasectomy. In this study, a rat model was employed to evaluate histological changes and DNA synthesis as an indication of hepatic regenerative capacity following hepatectomy and liver IORT. All rats (N = 40) had a partial hepatectomy and were divided into four random groups: a nonradiated group and three groups of 1000, 2000, and 3000 cGy given by IORT. The only deaths occurred in the 3000 cGy group. Routine H and E staining of liver sections after 3, 6, and 10 days suggested progressive hepatocyte damage notably in the 3000 cGy group. Comparison of the final average liver weights at 10 days confirmed a diminished liver mass in the 2000 and 3000 cGy animals. DNA synthesis in hepatocytes measured by [3H]thymidine label incorporation 3, 6, and 10 days after hepatectomy and IORT indicated a comparative and overall decrease in Day 6 peak activity between the three IORT groups. This study demonstrated delayed but substantial hepatic regeneration in the post-resected liver within the clinically useful IORT dose range (1000-2000 cGy) needed to control minimal residual tumor. This model has importance concerning the feasibility of IORT to the hepatic resection bed for patients where resection margins are inadequate. 0 1992
of *Surgery, Philadelphia,
MATERIALS
AND METHODS
Fisher 344 (12-week-old, 350 g) male rats (N = 40) were divided into four equal groups. All groups were subjected to a partial hepatectomy and then randomly assigned to receive no radiation (control) or 1000, 2000, and 3000 cGy delivered by IORT. Additionally, three rats served as nonhepatectomy controls for [3H]thymidine labeling. All rats were anesthetized with intraperitoneal brevitol (8 mg/rat). A right upper quadrant incision was used and the liver was delivered into the operative field. The anterior three lobes of the liver were resected with mass ligature with the remaining liver consisting of the two posterior lobes [7]. A lead shield was used to protect surrounding organs during IORT. The IORT was given as follows: 6 MeV electrons were delivered through a 10 X lo-cm cone with a 3-cm lead cutout for treatment; dose was prescribed to the 90% isodose line (depth 1.53 cm at 100 cm source to surface distance). The entire liver remnant was encompassed by the electron cutout. Following IORT the animals were observed daily and fed standard rat chow and water ad lib. Three rats from each radiation group and from the nonradiated hepatectomized control group were sacrificed by CO, asphyxiation at 3, 6, and 10 days after hepatectomy and IORT. After sacrifice, the remaining liver was removed and a portion was submitted for histological study. The histol-
Inc.
INTRODUCTION Surgical excision of primary or metastatic liver cancer has become standard therapy [l, 21. However, the margin of the resection is often important in long-term disease control within the liver and ultimately in patient survival [31. Intraoperative radiation therapy (IORT) has been utilized to improve local control rates for a variety of gastrointestinal neoplasms [4, 51. It has not, however, 287
Copyright 0 1992 All rights of reproduction
0022-4804/92 $4.00 by Academic Press, Inc. in any form reserved.
288
JOURNAL
FIG.
1.
Normal
hepatic
tissue
OF
SURGICAL
and wound
healing
RESEARCH:
VOL.
of the hepatectomy
ogy interpretation was made blinded to the IORT dose involved. We examined the regenerative capacity of the liver by quantitating nuclear DNA synthesis (as measured by [3H]thymidine incorporation). Briefly, 1 hr prior to sacrifice, rats were injected intraperitoneally with 100 &i [3H]thymidine. After sacrifice, livers were harvested, weighed, and washed with ice-cold phosphate-buffered saline. A portion was then sectioned, weighed, and homogenized in a sucrose containing phosphate buffered solution with 0.1% Triton X 100 and then centrifuged at 20,OOOg for 30 min to isolate nuclei. The nuclear pellet was solubilized with protosol (NEN DuPont) and 100 ml of this solution was applied to a glass fiber filter. Each filter was washed three times with 10% trichloroacetic acid and twice with 100% ethanol. The filters were dried and counted by liquid scintillation. The studies were repeated in triplicate and the results expressed as counts per minute incorporated per gram of liver tissue. RESULTS
Three deaths occurred (all in the 3000 cGy group) within the study interval. There was no evidence of intra-
53, NO.
border
could
3, SEPTEMBER
be observed
1992
after
6 days
1000 cGy.
X190
H&E
peritoneal bleeding, ascites, or gross hepatic necrosis in these animals although the livers were small and fibrotic. Liver sections from the control and the three treated groups were examined by H&E staining. There was evidence within 6 days in the 1000 cGy group for normal hepatic tissue and wound healing at the hepatectomy border (Fig. 1). In addition in this group at 6 days, there were areas of both normal and swollen hepatocytes with occasional mitotic figures indicative of regeneration (Fig. 2). This was quite similar to the histological appearance of the control hepatectomy group at 3 days. In contrast in the 2000 cGy animals, groups of necrotic cells could be seen as late as 10 days after radiation exposure (Fig. 3). Although normal hepatocytes were seen, findings suggestive of normal regeneration were somewhat limited. This was even more evident in the 3000 cGy group at 10 days, where residual pyknosis and steatosis were noted and no evidence of regeneration was observed in these sections suggesting significant and permanent liver injury (Fig. 4). Figure 5 summarizes the [3H]thymidine labeling data for the study groups (IORT) with a nonhepatectomized and nonradiated rat liver serving as a baseline label for comparison. In the 1000 cGy group, [3H]thymidine incor-
FIG. H&E
2.
1000 cGy, 6 days. Areas
of normal
FIG. 3. 2000 cGy, 10 days. Necrotic (1000 cGy, 6 days) in which no or only
and swollen
hepatocytes
with
occasional
cells could be seen as late as 10 days after exposure minor alterations were detected. X390 H&E
mitotic
figures
to the higher
were seen throughout
doses. This contrasts
the sections.
with
X380
Figs. 1 and 2
290
JOURNAL
FIG.
4.
Residual
pyknosis
OF
SURGICAL
and steatosis
RESEARCH:
is seen 10 days
poration peaks at 6 days and is statistically higher at all time intervals in comparison to the 2000 or 3000 cGy groups (P < 0.05). Although activity appears to peak at 6 days in all radiated groups, there is a progressive decrease in counts as radiation dose increases. The rat nonradiated hepatectomized controls peak with 10,000 cpm/g liver at 3 days. Additional evidence of quantitative differences in the capacity for liver regeneration can be found in the final weight of the regenerated livers expressed as a percentage of the original liver weight (Table 1). As noted, the average final liver weight for the nonradiated hepatectomized rats is the same as the 1000 cGy group. There is progressive decrease in liver mass in the 2000 and 3000 cGy groups. DISCUSSION
Hepatic resections are an increasingly utilized treatment modality for those patients whose metastatic disease is limited to the liver [8, 91. Studies have demonstrated that disease-free and overall survival are inversely related to close or positive tumor margins [lo]. Unfortunately due to anatomic considerations or lesion size, it is not always feasible to obtain wide resection
VOL.
after
53, NO.
3000 rads.
3, SEPTEMBER
No regeneration
1992
was observed.
X190
H&E
margins. Thus far with the exception of one study, adjuvant chemotherapy has not shown survival benefit in those patients with close margins [ll]. Although IORT has been studied to attempt an improvement in local regional control in other GI cancers, this modality has not been employed in the post-resection liver. A successful trial of IORT in the form of brachytherapy, however, has been reported in patients with liver metastases that were either unresectable or had close resection margins [12]. Since there is some indication that radiation therapy can provide local control within the liver substance, our study was designed as a model of the acute changes that might occur following hepatic resection and IORT. The rat model was chosen because of the ease of partial hepatectomy and the rapid regenerative sequence of normal nonradiated rat liver. Although liver tolerance to external radiation has been investigated both in animals and in humans, the effects of a directed cone-down radiated field has been the subject of only a few animal studies [13, 141. These studies, however, were primarily directed toward the deleterious effects on hepatic function following resection and IORT and did not relate these findings to hepatocyte DNA synthesis. Prior clinical experience with external fractionated
EISENBERG
ET
AL.:
INTRAOPERATIVE
LIVER
291
RADIATION
A
Bdays
0
Gdays
0
lOdays
c1
mean values
A
0
w
1’
7f A I
HE;$Tk’;;y
1000CG~
NOI’+;E$‘A’A;~TOM
FIG. 5. Results of hepatic [3H]thymidine control; (3) hepatectomy and 1000 cGy IORT; was statistically greater than the corresponding
labeling in five groups: (1) hepatectomy nonradiated (4) hepatectomy and 2000 cGy IORT, (5) hepatectomy 2000 cGy or 3000 cGy groups, P < 0.05, two-tailed
irradiation to the entire liver suggeststhat hepatic tissue tolerance is acceptable to 3000 cGy in 200-cGy fractions [15] Post-radiation, hepatic toxicity may occur in both acute and chronic forms and consists mainly of radiation hepatitis to progressive fibrosis leading to cirrhosis and liver failure. This is dependent on radiation dose and the presence of intrinsic liver disease. Our study indicates that unfractionated IORT dosesof 3000 cGy, a dramatically higher overall biological equivalent dose compared to the same dose in a conventional fractionated treatTABLE Liver
I Weights
at 10 Days final
Group
Liver
weight
=
initial
Control (hepatectomy, no radiation) 1000 cGy
96% 96%
2000cGy 3000cGy
76% 68%
Note. Initial liver weight was calculated the resected liver by a factor of 5/3.
2ooocGy
3ooocGy
(non-Radiited)
(non-Radiated)
Regenerative
q e
weight liver
by multiplying
weight
x 100
the weight
of
control; (2) nonhepatectomy and 3000 cGy IORT. *The Student’s t test.
nonradiated 1000 cGy group
ment schedule, have profound acute effects on regenerative liver capacity. Because of the size of the remaining liver in this rat model, it was not possible to limit the IORT treatments to the cut edge of the liver as it would correspond to a treatment scheme in a patient trial. However, the model does provide valuable information relative to hepatic parenchymal response to radiation injury in a surgically impaired liver. The only treatment-related deaths in the radiated animals occurred in the 3000 cGy group, which experienced a 20% mortality from hepatic failure. There was, however, no evidence of bile leak or liver substance suture breakdown contributing to acute hemorrhage in any of the IORT-treated groups. This suggests that the common postoperative complications of major hepatic resections should not be exaggerated by IORT, although this conclusion can only be substantiated by a clinical trial. The histology represents progressive evidence of hepatocyte damage characterized by fibrosis and necrosis as the IORT dose increased. This interpretation is corroborated by the [3H]thymidine labeling data. The mean value of peak activity (6 days) in all treated groups indicates a radiation effect on hepatocyte function most evident in the higher dose range. This suggests that re-
292
JOURNAL
OF
SURGICAL
RESEARCH:
generative capacity as represented by DNA synthesis diminishes in the 2000 cGy group and substantially diminishes in the 3000 cGy group. When comparison is made to the peak activity (3 days) of the nonradiated hepatectomized controls it is noted that the radiated livers all have a delay in cellular repair. The whole liver weights at 10 days in the higher dose range indicate a similar finding with reference to regenerative capacity. Clinically relevant IORT dose ranges of 1000 to 2000 cGy would be the planned dose for control of microscopic residual disease in patient trials. This study indicates that within this relevant dose range, there was good evidence for hepatic regeneration (albeit delayed), substantiated by histological review and DNA turnover data. Although this animal study did not address longterm toxicity questions such as bile duct stricture, IORT clinical trials that employed dose levels to 2000 cGy in the management of bile duct carcinoma suggested only a small incidence of late onset biliary sepsis or stricture [16]. Additionally, if chronic hepatic fibrosis would occur in the human liver, it would be restricted to a small volume with the use of electron irradiation. Adjuvant treatment strategies after partial hepatectomy for primary liver cancer or metastatic disease to the liver must address the patterns of failure (i.e., both local and distant). IORT can be integrated into these trials in patients with close or positive surgical margins. The information provided by this model indicates the feasibility of a clinical trial and may assist in future protocol planning to optimize results. APPENDIX:
TWO-TAILED
Day 3 looovs looovs 2ooovs Day 6 looovs looovs 2ooovs Day 10 looovs looovs 2ooovs
STUDENT’S
53, NO.
3, SEPTEMBER
1992
REFERENCES 1.
2.
Fortner, J. G., Silva, J. S., Golbey, R. B., Cox, E. B., and Maclean, B. J. Multivariate analysis of a personal series of 247 consecutive patients with liver metastases from colorectal cancer. I. Treatment by hepatic resection. Ann. Surg. 199: 306, 1984. Adson, M. A., and Van Heerden, J. A. Major hepatic resections for metastatic colorectal cancer. Ann. Surg. 191: 576, 1980.
3
Hughes, K. S., Simon, R., Songhorabodi, S., Adson, M. A., et al. Resection of the liver for colorectal carcinoma metastases: A multi-institutional study of pattern of recurrence. Surgery 100: 278,1986.
4.
Sindelar, W. F., Hoekstra, H., and Kinsella, T. J. Surgical approaches and techniques in intraoperative radiotherapy for intra-abdominal, retroperitoneal, and pelvic neoplasms. Surgery 103: 247,1988.
5
Gunderson, f external
6.
Cromack, D. T., Maher, tions in intraoperative conventional treatment?
7.
Higgins, G. M., and Anderson, R. M. Experimental pathology of the liver. I Restoration of the liver of the white rat following partial surgical removal. Arch. Pathol. 12: 186, 1931.
8.
Steele, G., and Ravikumar, T. S. Resection of hepatic from colorectal cancer: Biologic perspectives. Ann. 127,1989.
9.
Hoeffer, R. A., Bowers, G. J., and Eisenberg, resection in a residency training program. 1025, 1990.
L. L., Tepper, beam radiation.
J. E., Briggs, P. J., et al. Intraoperative Curr. Probl. Cancer 7: 1, 1983.
M. M., Hoekstra, H., et al. Are complicaradiation therapy more frequent than in Arch. Surg. 124: 229, 1989.
metastases Surg. 2 10:
B. L. Major hepatic South. Med. J. 83:
10.
Holm, A., Bradley, E., and Aldrete, J. S. Hepatic resection of metastasis from colorectal carcinoma: Morbidity, mortality, and pattern of recurrence. Ann. Surg. 209: 428, 1989.
11.
Patt, ative with liver.
t TEST
Y. Z., McBride, C. M., Ames, F. C., et al. Adjuvant perioperhepatic arterial Mitomycin C and Floxuridine combined surgical resection of metastatic colorectal cancer in the Cancer 59: 867, 1987.
2000 3000 3000
P = 0.0057 P = 0.0038 P = 0.7771
12.
Dritschilo, A., Harter, K. W., Thomas, D., et al. Intraoperative radiation therapy of hepatic metastases: Technical aspects and report of a pilot study. Int. J. Radiut. Oncol. Biol. Phys. 14: 1007, 1988.
2000 3000 3000
P = 0.0008 P = 0.0002 P = 0.0563
13.
Geraci, J. P., Jackson, K. L., Mariano, M. S., et al. Hepatic injury after whole-liver irradiation in the rat. Rudiut. Res. 101: 508, 1985.
14.
2000 3000 3000
P = 0.0055 P = 0.0116 P = 0.6103
Bossola, M., Merrick, H. W., Eltaki, A., et al. Rat liver tolerance for partial resection and intraoperative radiation therapy: Regeneration is radiation dose dependent. J. Surg. Oncol. 45: 196, 1990.
15.
Borgelt, B. B., Gelber, R., and Brady, L. W. The palliation of hepatic metastases. Results of Radiation Therapy Oncology Group Pilot Study. Int. J. R&at. Oncol. Biol. Phys. 7: 587,1981.
16.
Busse, P. M., Stone, M. D., Sheldon, T. A., et al. Intraoperative radiation therapy for biliary tract carcinoma: Results of a 5-year experience. Surgery 105: 724, 1989.
ACKNOWLEDGMENTS Special thanks Patrick Stafford.
VOL.
for technical Ph.D.
assistance
to James
Chu,
Ph.D.,
and
OF SURGICAL
RESEARCH
lntraoperative
53,287-292
(19%)
Liver Radiation after Partial Hepatectomy
in a Rat Model
BURTON L. EISENBERG, M.D., FACS,* RACHELLE M. LANCIANO, M.D.,? MICHAEL L. NUSSBAUM, M.D.,* ANDRES KLEIN-SZANTO, M.D.$ AND DOUGLAS D. TAYLOR, PH.D.* Fox Chase Cancer $Pathology,7701
Center, Department Burholme Avenue, Submitted
for publication
Academic
Press,
March
tRadiation Therapy, and Pennsylvania 19111 14, 1991
been used to enhance local control in the post-hepatectomy liver when resection margins are inadequate. Gastrointestinal toxicity in IORT trials is well documented [6]. However, because of the lack of clinical and laboratory experience, there is little information regarding the effect of a large fraction of irradiation to the post-resection liver. In this study, a rat model was employed to evaluate histological changes and DNA turnover as an indication of hepatic regenerative capacity following partial hepatectomy and IORT. The purpose of the study was to evaluate response of the normal liver parenchyma to varying doses of electron beam irradiation directed to the post-resection liver edge.
Hepatic resection of metastatic tumor is a treatment option in selected patients. Resection margin is a prognostic factor of hepatic recurrence and survival. Although intraoperative radiation therapy (IORT) has been clinically useful in some gastrointestinal cancers, there is little information regarding its use following hepatic metastasectomy. In this study, a rat model was employed to evaluate histological changes and DNA synthesis as an indication of hepatic regenerative capacity following hepatectomy and liver IORT. All rats (N = 40) had a partial hepatectomy and were divided into four random groups: a nonradiated group and three groups of 1000, 2000, and 3000 cGy given by IORT. The only deaths occurred in the 3000 cGy group. Routine H and E staining of liver sections after 3, 6, and 10 days suggested progressive hepatocyte damage notably in the 3000 cGy group. Comparison of the final average liver weights at 10 days confirmed a diminished liver mass in the 2000 and 3000 cGy animals. DNA synthesis in hepatocytes measured by [3H]thymidine label incorporation 3, 6, and 10 days after hepatectomy and IORT indicated a comparative and overall decrease in Day 6 peak activity between the three IORT groups. This study demonstrated delayed but substantial hepatic regeneration in the post-resected liver within the clinically useful IORT dose range (1000-2000 cGy) needed to control minimal residual tumor. This model has importance concerning the feasibility of IORT to the hepatic resection bed for patients where resection margins are inadequate. 0 1992
of *Surgery, Philadelphia,
MATERIALS
AND METHODS
Fisher 344 (12-week-old, 350 g) male rats (N = 40) were divided into four equal groups. All groups were subjected to a partial hepatectomy and then randomly assigned to receive no radiation (control) or 1000, 2000, and 3000 cGy delivered by IORT. Additionally, three rats served as nonhepatectomy controls for [3H]thymidine labeling. All rats were anesthetized with intraperitoneal brevitol (8 mg/rat). A right upper quadrant incision was used and the liver was delivered into the operative field. The anterior three lobes of the liver were resected with mass ligature with the remaining liver consisting of the two posterior lobes [7]. A lead shield was used to protect surrounding organs during IORT. The IORT was given as follows: 6 MeV electrons were delivered through a 10 X lo-cm cone with a 3-cm lead cutout for treatment; dose was prescribed to the 90% isodose line (depth 1.53 cm at 100 cm source to surface distance). The entire liver remnant was encompassed by the electron cutout. Following IORT the animals were observed daily and fed standard rat chow and water ad lib. Three rats from each radiation group and from the nonradiated hepatectomized control group were sacrificed by CO, asphyxiation at 3, 6, and 10 days after hepatectomy and IORT. After sacrifice, the remaining liver was removed and a portion was submitted for histological study. The histol-
Inc.
INTRODUCTION Surgical excision of primary or metastatic liver cancer has become standard therapy [l, 21. However, the margin of the resection is often important in long-term disease control within the liver and ultimately in patient survival [31. Intraoperative radiation therapy (IORT) has been utilized to improve local control rates for a variety of gastrointestinal neoplasms [4, 51. It has not, however, 287
Copyright 0 1992 All rights of reproduction
0022-4804/92 $4.00 by Academic Press, Inc. in any form reserved.
288
JOURNAL
FIG.
1.
Normal
hepatic
tissue
OF
SURGICAL
and wound
healing
RESEARCH:
VOL.
of the hepatectomy
ogy interpretation was made blinded to the IORT dose involved. We examined the regenerative capacity of the liver by quantitating nuclear DNA synthesis (as measured by [3H]thymidine incorporation). Briefly, 1 hr prior to sacrifice, rats were injected intraperitoneally with 100 &i [3H]thymidine. After sacrifice, livers were harvested, weighed, and washed with ice-cold phosphate-buffered saline. A portion was then sectioned, weighed, and homogenized in a sucrose containing phosphate buffered solution with 0.1% Triton X 100 and then centrifuged at 20,OOOg for 30 min to isolate nuclei. The nuclear pellet was solubilized with protosol (NEN DuPont) and 100 ml of this solution was applied to a glass fiber filter. Each filter was washed three times with 10% trichloroacetic acid and twice with 100% ethanol. The filters were dried and counted by liquid scintillation. The studies were repeated in triplicate and the results expressed as counts per minute incorporated per gram of liver tissue. RESULTS
Three deaths occurred (all in the 3000 cGy group) within the study interval. There was no evidence of intra-
53, NO.
border
could
3, SEPTEMBER
be observed
1992
after
6 days
1000 cGy.
X190
H&E
peritoneal bleeding, ascites, or gross hepatic necrosis in these animals although the livers were small and fibrotic. Liver sections from the control and the three treated groups were examined by H&E staining. There was evidence within 6 days in the 1000 cGy group for normal hepatic tissue and wound healing at the hepatectomy border (Fig. 1). In addition in this group at 6 days, there were areas of both normal and swollen hepatocytes with occasional mitotic figures indicative of regeneration (Fig. 2). This was quite similar to the histological appearance of the control hepatectomy group at 3 days. In contrast in the 2000 cGy animals, groups of necrotic cells could be seen as late as 10 days after radiation exposure (Fig. 3). Although normal hepatocytes were seen, findings suggestive of normal regeneration were somewhat limited. This was even more evident in the 3000 cGy group at 10 days, where residual pyknosis and steatosis were noted and no evidence of regeneration was observed in these sections suggesting significant and permanent liver injury (Fig. 4). Figure 5 summarizes the [3H]thymidine labeling data for the study groups (IORT) with a nonhepatectomized and nonradiated rat liver serving as a baseline label for comparison. In the 1000 cGy group, [3H]thymidine incor-
FIG. H&E
2.
1000 cGy, 6 days. Areas
of normal
FIG. 3. 2000 cGy, 10 days. Necrotic (1000 cGy, 6 days) in which no or only
and swollen
hepatocytes
with
occasional
cells could be seen as late as 10 days after exposure minor alterations were detected. X390 H&E
mitotic
figures
to the higher
were seen throughout
doses. This contrasts
the sections.
with
X380
Figs. 1 and 2
290
JOURNAL
FIG.
4.
Residual
pyknosis
OF
SURGICAL
and steatosis
RESEARCH:
is seen 10 days
poration peaks at 6 days and is statistically higher at all time intervals in comparison to the 2000 or 3000 cGy groups (P < 0.05). Although activity appears to peak at 6 days in all radiated groups, there is a progressive decrease in counts as radiation dose increases. The rat nonradiated hepatectomized controls peak with 10,000 cpm/g liver at 3 days. Additional evidence of quantitative differences in the capacity for liver regeneration can be found in the final weight of the regenerated livers expressed as a percentage of the original liver weight (Table 1). As noted, the average final liver weight for the nonradiated hepatectomized rats is the same as the 1000 cGy group. There is progressive decrease in liver mass in the 2000 and 3000 cGy groups. DISCUSSION
Hepatic resections are an increasingly utilized treatment modality for those patients whose metastatic disease is limited to the liver [8, 91. Studies have demonstrated that disease-free and overall survival are inversely related to close or positive tumor margins [lo]. Unfortunately due to anatomic considerations or lesion size, it is not always feasible to obtain wide resection
VOL.
after
53, NO.
3000 rads.
3, SEPTEMBER
No regeneration
1992
was observed.
X190
H&E
margins. Thus far with the exception of one study, adjuvant chemotherapy has not shown survival benefit in those patients with close margins [ll]. Although IORT has been studied to attempt an improvement in local regional control in other GI cancers, this modality has not been employed in the post-resection liver. A successful trial of IORT in the form of brachytherapy, however, has been reported in patients with liver metastases that were either unresectable or had close resection margins [12]. Since there is some indication that radiation therapy can provide local control within the liver substance, our study was designed as a model of the acute changes that might occur following hepatic resection and IORT. The rat model was chosen because of the ease of partial hepatectomy and the rapid regenerative sequence of normal nonradiated rat liver. Although liver tolerance to external radiation has been investigated both in animals and in humans, the effects of a directed cone-down radiated field has been the subject of only a few animal studies [13, 141. These studies, however, were primarily directed toward the deleterious effects on hepatic function following resection and IORT and did not relate these findings to hepatocyte DNA synthesis. Prior clinical experience with external fractionated
EISENBERG
ET
AL.:
INTRAOPERATIVE
LIVER
291
RADIATION
A
Bdays
0
Gdays
0
lOdays
c1
mean values
A
0
w
1’
7f A I
HE;$Tk’;;y
1000CG~
NOI’+;E$‘A’A;~TOM
FIG. 5. Results of hepatic [3H]thymidine control; (3) hepatectomy and 1000 cGy IORT; was statistically greater than the corresponding
labeling in five groups: (1) hepatectomy nonradiated (4) hepatectomy and 2000 cGy IORT, (5) hepatectomy 2000 cGy or 3000 cGy groups, P < 0.05, two-tailed
irradiation to the entire liver suggeststhat hepatic tissue tolerance is acceptable to 3000 cGy in 200-cGy fractions [15] Post-radiation, hepatic toxicity may occur in both acute and chronic forms and consists mainly of radiation hepatitis to progressive fibrosis leading to cirrhosis and liver failure. This is dependent on radiation dose and the presence of intrinsic liver disease. Our study indicates that unfractionated IORT dosesof 3000 cGy, a dramatically higher overall biological equivalent dose compared to the same dose in a conventional fractionated treatTABLE Liver
I Weights
at 10 Days final
Group
Liver
weight
=
initial
Control (hepatectomy, no radiation) 1000 cGy
96% 96%
2000cGy 3000cGy
76% 68%
Note. Initial liver weight was calculated the resected liver by a factor of 5/3.
2ooocGy
3ooocGy
(non-Radiited)
(non-Radiated)
Regenerative
q e
weight liver
by multiplying
weight
x 100
the weight
of
control; (2) nonhepatectomy and 3000 cGy IORT. *The Student’s t test.
nonradiated 1000 cGy group
ment schedule, have profound acute effects on regenerative liver capacity. Because of the size of the remaining liver in this rat model, it was not possible to limit the IORT treatments to the cut edge of the liver as it would correspond to a treatment scheme in a patient trial. However, the model does provide valuable information relative to hepatic parenchymal response to radiation injury in a surgically impaired liver. The only treatment-related deaths in the radiated animals occurred in the 3000 cGy group, which experienced a 20% mortality from hepatic failure. There was, however, no evidence of bile leak or liver substance suture breakdown contributing to acute hemorrhage in any of the IORT-treated groups. This suggests that the common postoperative complications of major hepatic resections should not be exaggerated by IORT, although this conclusion can only be substantiated by a clinical trial. The histology represents progressive evidence of hepatocyte damage characterized by fibrosis and necrosis as the IORT dose increased. This interpretation is corroborated by the [3H]thymidine labeling data. The mean value of peak activity (6 days) in all treated groups indicates a radiation effect on hepatocyte function most evident in the higher dose range. This suggests that re-
292
JOURNAL
OF
SURGICAL
RESEARCH:
generative capacity as represented by DNA synthesis diminishes in the 2000 cGy group and substantially diminishes in the 3000 cGy group. When comparison is made to the peak activity (3 days) of the nonradiated hepatectomized controls it is noted that the radiated livers all have a delay in cellular repair. The whole liver weights at 10 days in the higher dose range indicate a similar finding with reference to regenerative capacity. Clinically relevant IORT dose ranges of 1000 to 2000 cGy would be the planned dose for control of microscopic residual disease in patient trials. This study indicates that within this relevant dose range, there was good evidence for hepatic regeneration (albeit delayed), substantiated by histological review and DNA turnover data. Although this animal study did not address longterm toxicity questions such as bile duct stricture, IORT clinical trials that employed dose levels to 2000 cGy in the management of bile duct carcinoma suggested only a small incidence of late onset biliary sepsis or stricture [16]. Additionally, if chronic hepatic fibrosis would occur in the human liver, it would be restricted to a small volume with the use of electron irradiation. Adjuvant treatment strategies after partial hepatectomy for primary liver cancer or metastatic disease to the liver must address the patterns of failure (i.e., both local and distant). IORT can be integrated into these trials in patients with close or positive surgical margins. The information provided by this model indicates the feasibility of a clinical trial and may assist in future protocol planning to optimize results. APPENDIX:
TWO-TAILED
Day 3 looovs looovs 2ooovs Day 6 looovs looovs 2ooovs Day 10 looovs looovs 2ooovs
STUDENT’S
53, NO.
3, SEPTEMBER
1992
REFERENCES 1.
2.
Fortner, J. G., Silva, J. S., Golbey, R. B., Cox, E. B., and Maclean, B. J. Multivariate analysis of a personal series of 247 consecutive patients with liver metastases from colorectal cancer. I. Treatment by hepatic resection. Ann. Surg. 199: 306, 1984. Adson, M. A., and Van Heerden, J. A. Major hepatic resections for metastatic colorectal cancer. Ann. Surg. 191: 576, 1980.
3
Hughes, K. S., Simon, R., Songhorabodi, S., Adson, M. A., et al. Resection of the liver for colorectal carcinoma metastases: A multi-institutional study of pattern of recurrence. Surgery 100: 278,1986.
4.
Sindelar, W. F., Hoekstra, H., and Kinsella, T. J. Surgical approaches and techniques in intraoperative radiotherapy for intra-abdominal, retroperitoneal, and pelvic neoplasms. Surgery 103: 247,1988.
5
Gunderson, f external
6.
Cromack, D. T., Maher, tions in intraoperative conventional treatment?
7.
Higgins, G. M., and Anderson, R. M. Experimental pathology of the liver. I Restoration of the liver of the white rat following partial surgical removal. Arch. Pathol. 12: 186, 1931.
8.
Steele, G., and Ravikumar, T. S. Resection of hepatic from colorectal cancer: Biologic perspectives. Ann. 127,1989.
9.
Hoeffer, R. A., Bowers, G. J., and Eisenberg, resection in a residency training program. 1025, 1990.
L. L., Tepper, beam radiation.
J. E., Briggs, P. J., et al. Intraoperative Curr. Probl. Cancer 7: 1, 1983.
M. M., Hoekstra, H., et al. Are complicaradiation therapy more frequent than in Arch. Surg. 124: 229, 1989.
metastases Surg. 2 10:
B. L. Major hepatic South. Med. J. 83:
10.
Holm, A., Bradley, E., and Aldrete, J. S. Hepatic resection of metastasis from colorectal carcinoma: Morbidity, mortality, and pattern of recurrence. Ann. Surg. 209: 428, 1989.
11.
Patt, ative with liver.
t TEST
Y. Z., McBride, C. M., Ames, F. C., et al. Adjuvant perioperhepatic arterial Mitomycin C and Floxuridine combined surgical resection of metastatic colorectal cancer in the Cancer 59: 867, 1987.
2000 3000 3000
P = 0.0057 P = 0.0038 P = 0.7771
12.
Dritschilo, A., Harter, K. W., Thomas, D., et al. Intraoperative radiation therapy of hepatic metastases: Technical aspects and report of a pilot study. Int. J. Radiut. Oncol. Biol. Phys. 14: 1007, 1988.
2000 3000 3000
P = 0.0008 P = 0.0002 P = 0.0563
13.
Geraci, J. P., Jackson, K. L., Mariano, M. S., et al. Hepatic injury after whole-liver irradiation in the rat. Rudiut. Res. 101: 508, 1985.
14.
2000 3000 3000
P = 0.0055 P = 0.0116 P = 0.6103
Bossola, M., Merrick, H. W., Eltaki, A., et al. Rat liver tolerance for partial resection and intraoperative radiation therapy: Regeneration is radiation dose dependent. J. Surg. Oncol. 45: 196, 1990.
15.
Borgelt, B. B., Gelber, R., and Brady, L. W. The palliation of hepatic metastases. Results of Radiation Therapy Oncology Group Pilot Study. Int. J. R&at. Oncol. Biol. Phys. 7: 587,1981.
16.
Busse, P. M., Stone, M. D., Sheldon, T. A., et al. Intraoperative radiation therapy for biliary tract carcinoma: Results of a 5-year experience. Surgery 105: 724, 1989.
ACKNOWLEDGMENTS Special thanks Patrick Stafford.
VOL.
for technical Ph.D.
assistance
to James
Chu,
Ph.D.,
and