An Electron Microscopic Study of the Effects of Portacaval Shunts on the Ultrastructure of the Rat Liver after Partial Hepatectomy Charles Marks, MD, PhD, New Orleans, Louisiana C. Markey, MD, New Orleans, Louisiana I?. Dyer, PhD, New Orleans, Louisiana M. Ft. Vaupel, PhD, New Orleans, Louisiana
The clinical frequency with which partial hepatic resections have been and still are being carried out for trauma, cyst, or neoplasm has brought into sharp focus the ability of the liver to restore the total deficit after such drastic resection. (Figure 1.) This is readily demonstrable in man by postoperative radionuclide scintiscan technics. (Figure 2.) The latent capacity of the liver for regeneration is based on the unique ability of the naturally long-lived hepatocytes to undergo mitosis even though fully differentiated. A precise autoregulatory mechanism leads to cessation of such growth once the deficit has been restored, so that the original total quantity of hepatic mass is never exceeded. This obscure hepatotropic mechanism regulates the process of hepatic regeneration and provides a critical point of distinction between regenerative hyperplasia and autonomous unregulated neoplasia. The process of restitution takes four to six months in man, six to eight weeks in dogs, and ten to fourteen days in rats. This temporal difference makes the rat an expedient experimental model for study of the regenerative processes after partial hepatectomy and provides a basis for rapid study
From the Department of Surgery and Anatomy, Louisiana State University School of Medicine, and the Department of Anatomy, Tulane University, New Orleans, Louisiana. Reprint requests should be addressed to Dr Charles Marks, Department of Surgery, Louisiana State University School of Medicine, 1542 Tulane Avenue, New Orleans, Louisiana 70112. Presented at the Fifteenth Annual Meeting of the Society for Surgery of the Alimentary Tract, San Francisco, California, May 21 and 22, 1974.
156
of the effects of alterations in hepatic blood flow on the regenerative patterns. Experimental alteration of the hepatic circulation can be produced by the surgical construction of an end to side portacaval anastomosis. Since the first successful portacaval anastomosis by Eck [I] and the subsequent experimental application of the technic by Pavlov [2], this procedure has been utilized frequently in liver circulation research. Mann [3] and Grindlay and Bollman [4] reported that construction of an Eck shunt in the dog hindered hepatic regeneration. They suggested that portal blood contained factors that enhanced liver regeneration and that the shunt decreased portal pressure, thereby retarding regeneration. The concept of a regeneration-enhancing factor in the portal blood, however, has been invalidated by Child et al [5]. In a dog model the caudal end of the portal vein was anastomosed to the cephalic end of the inferior vena cava and the caudal end of the vena cava was anastomosed to the cephalic end of the portal vein. All portal blood was consequently shunted away from the liver while the systemic venous blood passed through that organ. Despite this venous transposition, satisfactory regeneration occurred after partial hepatectomy. Fisher et al [6] also noted diminished liver regeneration after portacaval shunt, although subsequent studies by Fisher et al [7] noted that an increase in hepatic blood flow by substitution of arterial blood for portal blood provided greater restitution of hepatic parenchyma after partial hepatectomy than that found in animals with normal hepatic blood flow. A reduced
The American Journal of Surgery
Effects of Portacaval Shunts on Liver
capacity of the liver for restoration after partial hepatectomy and partial ligation of the portal vein was demonstrated by Stephenson [8]. Weinbren [9], however, indicated that portal blood is unnecessary for hepatic regeneration and stressed that the liver itself is capable of generating conditions that regulate hepatocytic proliferation. Standard morphologic changes occurring under conditions of 70 per cent hepatectomy without alteration of hepatic blood flow have been studied by many investigators [Z&12]. The following study was designed to alter the hemodynamic patterns of hepatic blood flow and to investigate the effects on cellular ultrastructure during the first seventy-two hours of hepatic regeneration. A series of rats were subjected to standard 70 per cent hepatectomy and the regenerative changes were compared with those in a series of rats that had been subjected to standard 70 per cent hepatectomy after preliminary construction of a portacaval anastomosis. Material and Methods Male Sprague-Dawley rats weighing 350 to 560 gm were used. All animals were housed under identically controlled environmental conditions without variations in lighting. Constant maintenance of environmental temperature was provided during the surgical procedures as well as in the pre- and postoperative stages. The animals were housed four per cage and permitted water and standard laboratory chow ad libitum. Inasmuch as diurnal periodicity is a characteristic of many cellular functions, all operative procedures were carried out between 8 and 10 AM. Postoperative sacrifice of the animals and collection of regenerating hepatic tissue for appropriate study were carried out at specific intervals of time, that is, three, six, nine, twelve, twenty-four, forty-eight, and seventy-two hours post hepatectomy. The animals were sacrificed in groups of six at each appropriate interval. Sections of regenerating liver of all the animals were prepared for examination by light microscopy, and sections of liver of three animals per group were prepared for electron microscopic examination. Surgical Procedures. Sham operations: A series of animals were subjected to laparotomy and appropriate closure of the abdominal wound without any manipulation of the liver and were subsequently sacrificed to study the effects of anesthesia and laparotomy only at the same designated intervals of time. Partial hepatectomy: A series of animals were subjected to standard partial hepatectomy carried out by the technic described by Higgins and Anderson [13] whereby under ether anesthesia the median and the left lateral lobes were removed after ligating the base of these lobes near the hilum. Pack and Islami [14] have shown that removal of less than two thirds of the liver
volume 129, February 1975
Figure 1. Liver scan utillzng radioiodinatad rose bengal demonstrates filing defect in right lobe due to malignant hepatoma. Right hepatk lobe resected surgically. results in a reduced and more gradual regenerative response whereas Harkness [I51 demonstrated that removal of more than two thirds of the liver results in an increase in size of the organ rather than in the number of cells. For our purposes, partial hepatectomy represents the removal of the median and left lateral lobes of the rat liver comprising 70 per cent by weight of the whole liver. Portacaoal shunts: In the experimental models an end to side portacaval anastomosis was carried out through a midline abdominal incision after the abdomen was shaved and the area prepared with ether. In a series of fifteen rats the changes in the portal pressure induced by the portacaval anastomosis were measured. At the
Figure 2. Repeated liver scan six months after right hepatk lobectomy demonstrates normal uptake of radiokdinated rose bengal due to complete regeneration of lobe.
157
Marks
TABLE
et al
PortalPressureMeasurementsin
I
No. of Time after Rats Shunt (min) 3 3 3 3 3
9 10 12 15 18
Rats
Mean lntrasplenic Pressure (cm of water) Before 15.5(+0.7) 14.8(+0.5) 15.1(+0.5) 13.9(+0.3) 14.4(+0.3)
After 12.8(10.3) 12.1(+0.2) 11.8(*0.2) ll.O(ztO.2) 11.2(+0.2)
Change in Mean Pressures 2.7 2.7 3.3 2.9 3.2
time of the portacaval anastomosis the spleen was delivered through the midline laparotomy incision and a hollow needle (gauge 22) was introduced into its parenchyma. The needle was connected by polyethylene tubing to a pressure chamber containing a stainless steel diaphragm. The movement of the diaphragm activated a strain gauge producing an impulse that was amplified to register the pressure. The hollow system from needle to diaphragm contained heparinized normal saline (50 mg/lOO cc). Calibration was set before each reading in centimeters of water, taking the pubic symphysis of the rat as zero. The normal mean intrasplenic pressure was I5 cm of water. Pressure measurements were repeated ten minutes after completion of the end to side portacaval shunt with a mean reduction to 12 cm of water providing a 3 cm change in mean pressure. (Table I.) The sensitivity of the model was demonstrated by the almost immediate response of the intrasplenic pressure to changes in portal venous pressure produced by occlusion and subsequent release of the portal vein. At the time of re-exploration for subtotal hepatectomy, portal pressure measurements were taken in three shunted animals and confirmed the fact that the reduced portal pressure persisted. Liver Tissue Samples. In hepatic regeneration not all elements are restored in parallel, and it therefore be-
came important to assess the first cycle of cellular growth and division during regeneration, since in later stages it was noted that the patterns were modified by poor mitotic synchrony. Harkness [15] has indicated that parenchymal cells constitute about 90 per cent of the total hepatic cellular mass yet only 60 per cent of the cell population and that maximal cellular changes occur at the periphery of the lobule, spreading only gradually towards the center within the ensuing eight hours. For purposes of this study, only sections of liver tissue obtained from the periportal areas were utilized for appropriate studies. Thin slices of liver tissue fixed in phosphate buffered formalin (pH 7.25) for twenty-four hours were prepared for appropriate light microscopic examination. The tissue slices were embedded in paraffin, sectioned at 4 p, and stained with hematoxylin and eosin. The presence of mitoses was measured by using a square ocular and the incidence expressed per 1,000 nuclei counted in the sections stained with hematoxylin and eosin. Changes in cellular glycogen content were studied by periodic acid-Schiff (PAS) histochemical staining technics. Any doubt that PAS-positive substance in the cells was glycogen was solved before staining by pretreating a similar section with diastase, which breaks down glycogen to free soluble sugars. Electron Microscopy. Segments of regenerating liver tissue 3 mm in size were fixed at 2°C for forty-eight hours in 0.2 M phosphate buffered glutaraldehyde at pH 7.2 followed by treatment with buffered 1 per cent aqueous osmium tetroxide for two hours. The specimens were then dehydrated by passage through three series of 50 per cent alcohol, then 70 per cent alcohol for twentyfive minutes, and then brought to room temperature. The specimens were rinsed three times in 95 per cent alcohol and then in absolute alcohol followed by propyl-
Figure 3. An electron micrograph of parenchymal cell of rat liver twenty-four hours after sham operation demonstrates glycogen granules, normal rough endoplasmlc reticulum, and mitochondria. An occaslonal iysosome, fat droplets, and microbodles are evident. (Fixatkm in glutaraldelyde and osmium tetroxide; double stained with uranyi acetate and lead citrate; original magnification X 24,750.)
The American Journal of Surgery
Effects of Portacaval Shunts on Liver
ene oxide. Each vial of specimens was filled with a 50 per cent solution of propylene oxide and epoxy resin (Epon 812). Half of each specimen was infiltrated with a 50 per cent solution of propylene oxide and Maraglas and the rest with a 50 per cent solution of propylene oxide and Araldite. Each specimen was then placed in 100 per cent Epon. The specimens were left overnight, embedded in individual capsules, and placed in an oven at 60°C for three to seven days. Sections of 2 p were prepared and stained with methylene blue. Selected areas of tissue were then chosen and sectioned at 600 to 900 angstroms, double stained with uranyl acetate and lead citrate, and examined with the Philips 200 electron microscope.
Resufts Effects of Sham Operations. Handling of the rat liver without resection or interference with its vascular supply produced transient morphologic changes similar to those occurring during the first six hours of regeneration. Histologic examination demonstrated swelling, atrophy, and degeneration of hepatocytes associated with a brief and unsustained elevation in the mitotic index to 3.4 per 1,000 cells at six hours. There was a subsequent rapid return to normal. Reduction of liver glycogen was noted at three hours, persisting to six hours. By twelve hours after sham operations reglycogenation was complete and the hepatocytic morphologic characteristics were restored to normal. Electron microscopic examination confirmed the loss of glycogen, which is associated with widespread, albeit transient, lysosomal activity and autophagocytosis. Within six to nine hours no persistent abnormalities were noted apart from the presence of occasional fat droplets at forty-eight hours. (Figure 3.) Partial Hepatectomy. Within three hours after partial hepatectomy light microscopy demonstrated a dispersal of basophilic substance, spreading from the periportal to the central part of the lobule at six hours. Accumulation of cytoplasmic inclusion bodies was noted to be associated with this progressive vacuolization. A conspicuous increase in fat globules became apparent. A rapid and progressive depletion of glycogen stores was noted, first at three hours and reaching its maximal phase at nine hours. After twelve hours the reappearance of glycogen stores became manifest, the reglycogenation being complete by forty-eight hours. The cellular nuclei and nucleoli enlarged progressively, and at twenty-four hours each nucleus had doubled in size. The hepatocytes enlarged in size as the fat content increased, distortvolume 129, February 1975
ing the normal lobular pattern. The cells were noted to bulge into the sinusoidal spaces. The mitotic index increased slowly at first with a rapid rise occurring between twelve and t,wenty-four hours, rising from 7 to 34 during this time. It then fell progressively but remained elevated beyond seventy-two hours. Electron microscopic study demonstrated the disappearance of granular endoplasmic reticulum which was first noted at three hours with the presence of free polysomes in the cytoplasm. The normally flattened cisternae became distended and dilated, providing a prominent feature. At this time mitochondria became pale and swollen and alterations in their contours were noted. Depletion of glycogen was associated with an increase in the number and size of lysosomes that engulf and digest fragments of the endoplasmic reticulum, glycogen granules, and mitochondria. At six to nine hours the loss of granular endoplasmic reticulum and the mitochondrial changes were most prominent and autophagosomal activity was most notable. (Figure 4.) At nine to twelve hours the residual endoplasmic reticulum increased and clusters of smooth vesicles representative of smooth endopiasmic reticulum increased in number in close approximation to the mitochondria. At twelve hours the quantity and quality of the endoplasmic reticulum and mitochondria improved and were restored to normal at twenty-four hours. Disengagement of hepatocytes from their apposed surfaces was noted. The cell membranes appeared to separate except at the intercellular junction bordering the bile canaliculi. The most dramatic electron microscopic features were seen at six to twelve hours and were represented by the following (Figure 5): (1) total loss of glycogen, (2) accumulation of lipid, (3) increased quantity of smooth endoplasmic reticulum, (4) increased prominence of lysosomes, (5) autophagocytosis, (6) dispersal of granular endoplasmic reticulum, (7) dilatation of cisternae, (8) loss of microvilli, and (9) changes in shape and contour of mitochondria. Portacaval Anastomosis. Diversion of portal blood from the liver led to gross and recognizable changes in the liver. The organ immediately lost its normal turgidity and became discernibly flabby. Within two weeks of the operation the liver had become pale, mottled, and yellow. Within six weeks areas of nodular cirrhosis had developed. Examination by light microscopy demonstrated atrophy of the cells, accumulation of lipid, and hyalinization of hepatocytes within three to six 159
Marks et al
Figure 4. An electron mkrograph of parenchymal cell of rat liver six hours after partial hepatectomy. The presence of lysosomes, dense granules, and autophagocytk vacuotes is apparent. (Fixation in ghdaraktelyde and osmium tetroxtde; double stained with uranyl acetate and lead c&ate; ortgkai magnifkatton X 31,250.)
hours. These changes were most marked in the periportal areas extending subsequently to the central portion of the lobules. Depletion of glycogen began at three hours and was maximal at nine hours. The patterns of deglycogenation were recognizable and specifically attributable to the deviation of hepatic blood flow. Mitotic figures were conspicuous and resembled the patterns attributable to regenerative processes even though the liver was intact except for its reduced blood flow. An increased mitotic index of 7 occurred at three hours and reached a peak of 18 at twenty-four hours, so that the rise was more
rapid than that after partial hepatectomy but did not reach as great a peak at twenty-four hours as that found after partial hepatectomy. The rapid atrophy of liver tissue deprived of portal blood flow, however, was self-limiting, and equilibrium was established and maintained by a mechanism different from that initiating hepatic regeneration although the responsive mechanisms were the same. Electron microscopic examination confirmed the rapid and marked depletion of glycogen and early lipid infiltration. There was prominent dilatation of the granular endoplasmic reticulum
Figure 5. An electron mkrograph of parenchymal cell of rat liver nine hours after partial hepatectomy demonstrates separatton of cell membranes and disruption of vtlli wtthin space of Dtsse. (Flxatton in gktaraldelyde and osmium tetroxide; double stained wtth uranyi acetate and lead citrate; original magnification X 18,750.)
160
The Amerlcsn Journal of Surgery
Effects of Portacaval Shunts on Liver
Figure 6. An electron micqraph of fat liver twenty-four tnnrrs after portacaval anastomoeis. Wycogen deposits and dense in&&n bodks are present and irregular ctkternae are apparent kr the pertnuctear zone. (Fixation in glutarakWyde and osmium tetroxkie; double stained with uranyl acetate and lead citrate; original magnifkation X 18,750.)
which was most marked at twelve hours. By twenty-four hours inclusion bodies were greatly increased in number and reglycogenation became apparent while large, round, and irregular cisternae were noted in the perinuclear zones of the cytoplasm. (Figure 6.) Partial Hepatectomy after Preliminary Portaof a portacaval caval Shunts. The establishment anastomosis in the rat provided a stimulus to regenerative hepatocytic activity. Subsequent partial hepatectomy resulted in an enhancement of cellular and lobular responses with exaggeration of the mitotic index so that at six hours post hepatectomy the index was 15, increasing at twenty-four hours to 48, thereby attaining a level of mitotic activity that was greater than the response to partial hepatectomy alone. Deprivation or reduction of portal blood flow, therefore, does not inhibit regeneration after partial hepatectomy but actually enhances it. Thus, portal blood flow cannot be considered the primary controlling factor in its maintenance. Light microscopy demonstrated rapid depletion of glycogen as well as a devc.ilipment of cytoplasmic vacuolization at three tX .x hours. A marked increase in lipid was obsel . ,!e at twelve hours with a progressive increase 1forty-eight hours. These changes were simila / z:flected in the electron microscopic studies. Electron microscopy den,. .,trated a marked reduction in the granular endoplasmic reticulum at six hours whereas an increased number of microbodies were discernible at twelve hours. By
Volume 129, February 1975
twenty-four hours new granular endoplasmic reticulum was noted to develop in the hepatocytes and glycogen redeposition occurred. At forty-eight hours the hepatic cells appeared normal except for the persistence of the high lipid content. (Figure 7.) Comparison between electron microscopic effects of partial hepatectomy in animals with an intact portal circulation and animals with a previous portacaval shunt demonstrated the following: Glycogen depletion: The degree of glycogen depletion is much greater in animals with shunts. The reglycogenation commenced in both series at twenty-four hours, but the process was greater in animals without shunts over the ensuing fortyeight hours. Granular endoplasmic reticulum: A much greater degree and extent of dilatation of the granular endoplasmic reticular membranes occurred in animals with shunts with a correspondingly higher number of large cytoplasmic cisternae. Mitochondria: The edema, swelling, and alteration of mitochondrial contours were greater in animals with shunts than in animals with conventional hepatectomy. Inclusion bodies: Although an increased number of inclusion bodies was noted in animals without shunts, the number of inclusion bodies at twentyfour hours in animals with shunts was unquestionably greater. Lipid content: The degree of lipid infiltration in animals with shunts was of a greater order than that in animals without shunts. In animals without
161
Marks et al
Figure 7. A high magnifkation efectron mkqraph of parenchymal ceU of rat liver forty-eight hours after partial hepatectomy in rat wfth a shunt. Glycogen defnWtkn is minimal with per&Hence of mkrobodies and prominent Golgl compfex. The mitochondrial internal structure Is still abnormal. (Fixation In gktarahtelyde and osmium te&oxide; double staln8d with uranyl acetate and had c&ate; orfginal magnifkation X 57,500.)
shunts, lipid droplets were infrequently noted after twenty-four hours, but in animals with shunts, lipid deposition persisted beyond fortyeight hours, representing lower clearing. Lysosomes: An increased number of lysosomes was apparent in animals with shunts with a greater evidence of autophagocytosis. Conclusions The normally quiescent stable adult liver has a generous capacity for reparative hypertrophy and hyperplasia after loss of functional tissue. The large reserve of the liver’s functional capacity permits survival of the animal even if over 70 per cent of its liver is removed. It retains an inherent capacity for regenerative growth which subsides once the original organ deficit is restored. This study attempted to resolve the question of whether alteration in hepatic hemodynamics affects the regenerative stimulus of the liver after partial (70 per cent) hepatectomy. It has shown that the liver remnant regenerates after reduction of portal blood flow by construction of a portacaval anastomosis. The diversion of blood from the liver exerts its own histologic and electron microscopic effects on the liver. Reduction of portal blood flow affects the temporal patterns of regeneration after partial hepatectomy but does not prevent completion of the regenerative process. Correlation of this study with the biochemical data available in the literature indicates that the structural changes in the cellular organelles during the process of regeneration re-
162
events that are based fleet dynamic biochemical on a predetermined genetic code representing the key to life that is uniquely found in the liver. References 1. Eck NV: Concerning ligation of the vena porta. Voen Med Zh (St. Petersburg) 130: 1, 1877. 2. Pavlov IP: Modification of the Eck fist&. Arch Bkd /U&r 2: 580, 1893. 3. Mann FC: The portal circulation and restoration of the liver after partiil removal. Surgery 8: 225, 1940. 4. Grindfay JH, Bollman JL: Regeneration of the liver in the dog after partial hepatectomy. Surg Gynecol Obstet 94: 491, 1952. 5. Child CG Ill, Barr D, Holswade GR, Harrison SC: Liier regeneration following portacaval transposition in dogs. Ann Surg 138: 800, 1953. 8. Fisher B, Lee SH, Fisher ER, Saffer E: Liver regeneration following portacaval shunt. Surgery52: 88, 1962. 7. Fisher B, Russ C, Updegraff H, Fisher ER: Effect of increased hepatic blood flow upon liver regeneration. Arch Surg69: 263, 1954. 8. Stephenson’GN: Restoration of the liver after partial hepatectomy and partial ligation of the portal vein. Arch P&ho/ 14: 484, 1932. 9. Weinbren K: The portal blood suppfy and regeneration of the liver. Br J Exp Pathoi36: 583, 1955. 10. Gfinos AD, Grey Go: Humoral factors involved in the induction of liver regeneration in the rat. hoc Sot Exp Biol A&d 80: 421, 1952. 11. Grisham JW: Morphologic study of DNA synthesis and cell proliferation in regenerating rat liver: autoradiography with thymidine-H3. Cancer Res 22: 842, 1969. 12. Bucher NLR: Experimental aspects of hepatic regeneration. N Engl J Med277: 686, 1967. 13. Higgins M, Anderson RM: Experimental pathology of the liver. 1. Restoration of the liver of the white rat following partial surgical removal. Arch P&ho/ 12: 186, 1931. 14. Pack GT. lstami AH: Extent of hepatic resection and the regenerative response. Surgery 40: 6 11, 1956. 15. Harkness RD: Regeneration of the liver. Br A&d Bull 13: 1387, 1957.
The Amarkan
Journal
of Surgery
The clinical frequency with which partial hepatic resections have been and still are being carried out for trauma, cyst, or neoplasm has brought into sharp focus the ability of the liver to restore the total deficit after such drastic resection. (Figure 1.) This is readily demonstrable in man by postoperative radionuclide scintiscan technics. (Figure 2.) The latent capacity of the liver for regeneration is based on the unique ability of the naturally long-lived hepatocytes to undergo mitosis even though fully differentiated. A precise autoregulatory mechanism leads to cessation of such growth once the deficit has been restored, so that the original total quantity of hepatic mass is never exceeded. This obscure hepatotropic mechanism regulates the process of hepatic regeneration and provides a critical point of distinction between regenerative hyperplasia and autonomous unregulated neoplasia. The process of restitution takes four to six months in man, six to eight weeks in dogs, and ten to fourteen days in rats. This temporal difference makes the rat an expedient experimental model for study of the regenerative processes after partial hepatectomy and provides a basis for rapid study
From the Department of Surgery and Anatomy, Louisiana State University School of Medicine, and the Department of Anatomy, Tulane University, New Orleans, Louisiana. Reprint requests should be addressed to Dr Charles Marks, Department of Surgery, Louisiana State University School of Medicine, 1542 Tulane Avenue, New Orleans, Louisiana 70112. Presented at the Fifteenth Annual Meeting of the Society for Surgery of the Alimentary Tract, San Francisco, California, May 21 and 22, 1974.
156
of the effects of alterations in hepatic blood flow on the regenerative patterns. Experimental alteration of the hepatic circulation can be produced by the surgical construction of an end to side portacaval anastomosis. Since the first successful portacaval anastomosis by Eck [I] and the subsequent experimental application of the technic by Pavlov [2], this procedure has been utilized frequently in liver circulation research. Mann [3] and Grindlay and Bollman [4] reported that construction of an Eck shunt in the dog hindered hepatic regeneration. They suggested that portal blood contained factors that enhanced liver regeneration and that the shunt decreased portal pressure, thereby retarding regeneration. The concept of a regeneration-enhancing factor in the portal blood, however, has been invalidated by Child et al [5]. In a dog model the caudal end of the portal vein was anastomosed to the cephalic end of the inferior vena cava and the caudal end of the vena cava was anastomosed to the cephalic end of the portal vein. All portal blood was consequently shunted away from the liver while the systemic venous blood passed through that organ. Despite this venous transposition, satisfactory regeneration occurred after partial hepatectomy. Fisher et al [6] also noted diminished liver regeneration after portacaval shunt, although subsequent studies by Fisher et al [7] noted that an increase in hepatic blood flow by substitution of arterial blood for portal blood provided greater restitution of hepatic parenchyma after partial hepatectomy than that found in animals with normal hepatic blood flow. A reduced
The American Journal of Surgery
Effects of Portacaval Shunts on Liver
capacity of the liver for restoration after partial hepatectomy and partial ligation of the portal vein was demonstrated by Stephenson [8]. Weinbren [9], however, indicated that portal blood is unnecessary for hepatic regeneration and stressed that the liver itself is capable of generating conditions that regulate hepatocytic proliferation. Standard morphologic changes occurring under conditions of 70 per cent hepatectomy without alteration of hepatic blood flow have been studied by many investigators [Z&12]. The following study was designed to alter the hemodynamic patterns of hepatic blood flow and to investigate the effects on cellular ultrastructure during the first seventy-two hours of hepatic regeneration. A series of rats were subjected to standard 70 per cent hepatectomy and the regenerative changes were compared with those in a series of rats that had been subjected to standard 70 per cent hepatectomy after preliminary construction of a portacaval anastomosis. Material and Methods Male Sprague-Dawley rats weighing 350 to 560 gm were used. All animals were housed under identically controlled environmental conditions without variations in lighting. Constant maintenance of environmental temperature was provided during the surgical procedures as well as in the pre- and postoperative stages. The animals were housed four per cage and permitted water and standard laboratory chow ad libitum. Inasmuch as diurnal periodicity is a characteristic of many cellular functions, all operative procedures were carried out between 8 and 10 AM. Postoperative sacrifice of the animals and collection of regenerating hepatic tissue for appropriate study were carried out at specific intervals of time, that is, three, six, nine, twelve, twenty-four, forty-eight, and seventy-two hours post hepatectomy. The animals were sacrificed in groups of six at each appropriate interval. Sections of regenerating liver of all the animals were prepared for examination by light microscopy, and sections of liver of three animals per group were prepared for electron microscopic examination. Surgical Procedures. Sham operations: A series of animals were subjected to laparotomy and appropriate closure of the abdominal wound without any manipulation of the liver and were subsequently sacrificed to study the effects of anesthesia and laparotomy only at the same designated intervals of time. Partial hepatectomy: A series of animals were subjected to standard partial hepatectomy carried out by the technic described by Higgins and Anderson [13] whereby under ether anesthesia the median and the left lateral lobes were removed after ligating the base of these lobes near the hilum. Pack and Islami [14] have shown that removal of less than two thirds of the liver
volume 129, February 1975
Figure 1. Liver scan utillzng radioiodinatad rose bengal demonstrates filing defect in right lobe due to malignant hepatoma. Right hepatk lobe resected surgically. results in a reduced and more gradual regenerative response whereas Harkness [I51 demonstrated that removal of more than two thirds of the liver results in an increase in size of the organ rather than in the number of cells. For our purposes, partial hepatectomy represents the removal of the median and left lateral lobes of the rat liver comprising 70 per cent by weight of the whole liver. Portacaoal shunts: In the experimental models an end to side portacaval anastomosis was carried out through a midline abdominal incision after the abdomen was shaved and the area prepared with ether. In a series of fifteen rats the changes in the portal pressure induced by the portacaval anastomosis were measured. At the
Figure 2. Repeated liver scan six months after right hepatk lobectomy demonstrates normal uptake of radiokdinated rose bengal due to complete regeneration of lobe.
157
Marks
TABLE
et al
PortalPressureMeasurementsin
I
No. of Time after Rats Shunt (min) 3 3 3 3 3
9 10 12 15 18
Rats
Mean lntrasplenic Pressure (cm of water) Before 15.5(+0.7) 14.8(+0.5) 15.1(+0.5) 13.9(+0.3) 14.4(+0.3)
After 12.8(10.3) 12.1(+0.2) 11.8(*0.2) ll.O(ztO.2) 11.2(+0.2)
Change in Mean Pressures 2.7 2.7 3.3 2.9 3.2
time of the portacaval anastomosis the spleen was delivered through the midline laparotomy incision and a hollow needle (gauge 22) was introduced into its parenchyma. The needle was connected by polyethylene tubing to a pressure chamber containing a stainless steel diaphragm. The movement of the diaphragm activated a strain gauge producing an impulse that was amplified to register the pressure. The hollow system from needle to diaphragm contained heparinized normal saline (50 mg/lOO cc). Calibration was set before each reading in centimeters of water, taking the pubic symphysis of the rat as zero. The normal mean intrasplenic pressure was I5 cm of water. Pressure measurements were repeated ten minutes after completion of the end to side portacaval shunt with a mean reduction to 12 cm of water providing a 3 cm change in mean pressure. (Table I.) The sensitivity of the model was demonstrated by the almost immediate response of the intrasplenic pressure to changes in portal venous pressure produced by occlusion and subsequent release of the portal vein. At the time of re-exploration for subtotal hepatectomy, portal pressure measurements were taken in three shunted animals and confirmed the fact that the reduced portal pressure persisted. Liver Tissue Samples. In hepatic regeneration not all elements are restored in parallel, and it therefore be-
came important to assess the first cycle of cellular growth and division during regeneration, since in later stages it was noted that the patterns were modified by poor mitotic synchrony. Harkness [15] has indicated that parenchymal cells constitute about 90 per cent of the total hepatic cellular mass yet only 60 per cent of the cell population and that maximal cellular changes occur at the periphery of the lobule, spreading only gradually towards the center within the ensuing eight hours. For purposes of this study, only sections of liver tissue obtained from the periportal areas were utilized for appropriate studies. Thin slices of liver tissue fixed in phosphate buffered formalin (pH 7.25) for twenty-four hours were prepared for appropriate light microscopic examination. The tissue slices were embedded in paraffin, sectioned at 4 p, and stained with hematoxylin and eosin. The presence of mitoses was measured by using a square ocular and the incidence expressed per 1,000 nuclei counted in the sections stained with hematoxylin and eosin. Changes in cellular glycogen content were studied by periodic acid-Schiff (PAS) histochemical staining technics. Any doubt that PAS-positive substance in the cells was glycogen was solved before staining by pretreating a similar section with diastase, which breaks down glycogen to free soluble sugars. Electron Microscopy. Segments of regenerating liver tissue 3 mm in size were fixed at 2°C for forty-eight hours in 0.2 M phosphate buffered glutaraldehyde at pH 7.2 followed by treatment with buffered 1 per cent aqueous osmium tetroxide for two hours. The specimens were then dehydrated by passage through three series of 50 per cent alcohol, then 70 per cent alcohol for twentyfive minutes, and then brought to room temperature. The specimens were rinsed three times in 95 per cent alcohol and then in absolute alcohol followed by propyl-
Figure 3. An electron micrograph of parenchymal cell of rat liver twenty-four hours after sham operation demonstrates glycogen granules, normal rough endoplasmlc reticulum, and mitochondria. An occaslonal iysosome, fat droplets, and microbodles are evident. (Fixatkm in glutaraldelyde and osmium tetroxide; double stained with uranyi acetate and lead citrate; original magnification X 24,750.)
The American Journal of Surgery
Effects of Portacaval Shunts on Liver
ene oxide. Each vial of specimens was filled with a 50 per cent solution of propylene oxide and epoxy resin (Epon 812). Half of each specimen was infiltrated with a 50 per cent solution of propylene oxide and Maraglas and the rest with a 50 per cent solution of propylene oxide and Araldite. Each specimen was then placed in 100 per cent Epon. The specimens were left overnight, embedded in individual capsules, and placed in an oven at 60°C for three to seven days. Sections of 2 p were prepared and stained with methylene blue. Selected areas of tissue were then chosen and sectioned at 600 to 900 angstroms, double stained with uranyl acetate and lead citrate, and examined with the Philips 200 electron microscope.
Resufts Effects of Sham Operations. Handling of the rat liver without resection or interference with its vascular supply produced transient morphologic changes similar to those occurring during the first six hours of regeneration. Histologic examination demonstrated swelling, atrophy, and degeneration of hepatocytes associated with a brief and unsustained elevation in the mitotic index to 3.4 per 1,000 cells at six hours. There was a subsequent rapid return to normal. Reduction of liver glycogen was noted at three hours, persisting to six hours. By twelve hours after sham operations reglycogenation was complete and the hepatocytic morphologic characteristics were restored to normal. Electron microscopic examination confirmed the loss of glycogen, which is associated with widespread, albeit transient, lysosomal activity and autophagocytosis. Within six to nine hours no persistent abnormalities were noted apart from the presence of occasional fat droplets at forty-eight hours. (Figure 3.) Partial Hepatectomy. Within three hours after partial hepatectomy light microscopy demonstrated a dispersal of basophilic substance, spreading from the periportal to the central part of the lobule at six hours. Accumulation of cytoplasmic inclusion bodies was noted to be associated with this progressive vacuolization. A conspicuous increase in fat globules became apparent. A rapid and progressive depletion of glycogen stores was noted, first at three hours and reaching its maximal phase at nine hours. After twelve hours the reappearance of glycogen stores became manifest, the reglycogenation being complete by forty-eight hours. The cellular nuclei and nucleoli enlarged progressively, and at twenty-four hours each nucleus had doubled in size. The hepatocytes enlarged in size as the fat content increased, distortvolume 129, February 1975
ing the normal lobular pattern. The cells were noted to bulge into the sinusoidal spaces. The mitotic index increased slowly at first with a rapid rise occurring between twelve and t,wenty-four hours, rising from 7 to 34 during this time. It then fell progressively but remained elevated beyond seventy-two hours. Electron microscopic study demonstrated the disappearance of granular endoplasmic reticulum which was first noted at three hours with the presence of free polysomes in the cytoplasm. The normally flattened cisternae became distended and dilated, providing a prominent feature. At this time mitochondria became pale and swollen and alterations in their contours were noted. Depletion of glycogen was associated with an increase in the number and size of lysosomes that engulf and digest fragments of the endoplasmic reticulum, glycogen granules, and mitochondria. At six to nine hours the loss of granular endoplasmic reticulum and the mitochondrial changes were most prominent and autophagosomal activity was most notable. (Figure 4.) At nine to twelve hours the residual endoplasmic reticulum increased and clusters of smooth vesicles representative of smooth endopiasmic reticulum increased in number in close approximation to the mitochondria. At twelve hours the quantity and quality of the endoplasmic reticulum and mitochondria improved and were restored to normal at twenty-four hours. Disengagement of hepatocytes from their apposed surfaces was noted. The cell membranes appeared to separate except at the intercellular junction bordering the bile canaliculi. The most dramatic electron microscopic features were seen at six to twelve hours and were represented by the following (Figure 5): (1) total loss of glycogen, (2) accumulation of lipid, (3) increased quantity of smooth endoplasmic reticulum, (4) increased prominence of lysosomes, (5) autophagocytosis, (6) dispersal of granular endoplasmic reticulum, (7) dilatation of cisternae, (8) loss of microvilli, and (9) changes in shape and contour of mitochondria. Portacaval Anastomosis. Diversion of portal blood from the liver led to gross and recognizable changes in the liver. The organ immediately lost its normal turgidity and became discernibly flabby. Within two weeks of the operation the liver had become pale, mottled, and yellow. Within six weeks areas of nodular cirrhosis had developed. Examination by light microscopy demonstrated atrophy of the cells, accumulation of lipid, and hyalinization of hepatocytes within three to six 159
Marks et al
Figure 4. An electron mkrograph of parenchymal cell of rat liver six hours after partial hepatectomy. The presence of lysosomes, dense granules, and autophagocytk vacuotes is apparent. (Fixation in ghdaraktelyde and osmium tetroxtde; double stained with uranyl acetate and lead c&ate; ortgkai magnifkatton X 31,250.)
hours. These changes were most marked in the periportal areas extending subsequently to the central portion of the lobules. Depletion of glycogen began at three hours and was maximal at nine hours. The patterns of deglycogenation were recognizable and specifically attributable to the deviation of hepatic blood flow. Mitotic figures were conspicuous and resembled the patterns attributable to regenerative processes even though the liver was intact except for its reduced blood flow. An increased mitotic index of 7 occurred at three hours and reached a peak of 18 at twenty-four hours, so that the rise was more
rapid than that after partial hepatectomy but did not reach as great a peak at twenty-four hours as that found after partial hepatectomy. The rapid atrophy of liver tissue deprived of portal blood flow, however, was self-limiting, and equilibrium was established and maintained by a mechanism different from that initiating hepatic regeneration although the responsive mechanisms were the same. Electron microscopic examination confirmed the rapid and marked depletion of glycogen and early lipid infiltration. There was prominent dilatation of the granular endoplasmic reticulum
Figure 5. An electron mkrograph of parenchymal cell of rat liver nine hours after partial hepatectomy demonstrates separatton of cell membranes and disruption of vtlli wtthin space of Dtsse. (Flxatton in gktaraldelyde and osmium tetroxide; double stained wtth uranyi acetate and lead citrate; original magnification X 18,750.)
160
The Amerlcsn Journal of Surgery
Effects of Portacaval Shunts on Liver
Figure 6. An electron micqraph of fat liver twenty-four tnnrrs after portacaval anastomoeis. Wycogen deposits and dense in&&n bodks are present and irregular ctkternae are apparent kr the pertnuctear zone. (Fixation in glutarakWyde and osmium tetroxkie; double stained with uranyl acetate and lead citrate; original magnifkation X 18,750.)
which was most marked at twelve hours. By twenty-four hours inclusion bodies were greatly increased in number and reglycogenation became apparent while large, round, and irregular cisternae were noted in the perinuclear zones of the cytoplasm. (Figure 6.) Partial Hepatectomy after Preliminary Portaof a portacaval caval Shunts. The establishment anastomosis in the rat provided a stimulus to regenerative hepatocytic activity. Subsequent partial hepatectomy resulted in an enhancement of cellular and lobular responses with exaggeration of the mitotic index so that at six hours post hepatectomy the index was 15, increasing at twenty-four hours to 48, thereby attaining a level of mitotic activity that was greater than the response to partial hepatectomy alone. Deprivation or reduction of portal blood flow, therefore, does not inhibit regeneration after partial hepatectomy but actually enhances it. Thus, portal blood flow cannot be considered the primary controlling factor in its maintenance. Light microscopy demonstrated rapid depletion of glycogen as well as a devc.ilipment of cytoplasmic vacuolization at three tX .x hours. A marked increase in lipid was obsel . ,!e at twelve hours with a progressive increase 1forty-eight hours. These changes were simila / z:flected in the electron microscopic studies. Electron microscopy den,. .,trated a marked reduction in the granular endoplasmic reticulum at six hours whereas an increased number of microbodies were discernible at twelve hours. By
Volume 129, February 1975
twenty-four hours new granular endoplasmic reticulum was noted to develop in the hepatocytes and glycogen redeposition occurred. At forty-eight hours the hepatic cells appeared normal except for the persistence of the high lipid content. (Figure 7.) Comparison between electron microscopic effects of partial hepatectomy in animals with an intact portal circulation and animals with a previous portacaval shunt demonstrated the following: Glycogen depletion: The degree of glycogen depletion is much greater in animals with shunts. The reglycogenation commenced in both series at twenty-four hours, but the process was greater in animals without shunts over the ensuing fortyeight hours. Granular endoplasmic reticulum: A much greater degree and extent of dilatation of the granular endoplasmic reticular membranes occurred in animals with shunts with a correspondingly higher number of large cytoplasmic cisternae. Mitochondria: The edema, swelling, and alteration of mitochondrial contours were greater in animals with shunts than in animals with conventional hepatectomy. Inclusion bodies: Although an increased number of inclusion bodies was noted in animals without shunts, the number of inclusion bodies at twentyfour hours in animals with shunts was unquestionably greater. Lipid content: The degree of lipid infiltration in animals with shunts was of a greater order than that in animals without shunts. In animals without
161
Marks et al
Figure 7. A high magnifkation efectron mkqraph of parenchymal ceU of rat liver forty-eight hours after partial hepatectomy in rat wfth a shunt. Glycogen defnWtkn is minimal with per&Hence of mkrobodies and prominent Golgl compfex. The mitochondrial internal structure Is still abnormal. (Fixation In gktarahtelyde and osmium te&oxide; double staln8d with uranyl acetate and had c&ate; orfginal magnifkation X 57,500.)
shunts, lipid droplets were infrequently noted after twenty-four hours, but in animals with shunts, lipid deposition persisted beyond fortyeight hours, representing lower clearing. Lysosomes: An increased number of lysosomes was apparent in animals with shunts with a greater evidence of autophagocytosis. Conclusions The normally quiescent stable adult liver has a generous capacity for reparative hypertrophy and hyperplasia after loss of functional tissue. The large reserve of the liver’s functional capacity permits survival of the animal even if over 70 per cent of its liver is removed. It retains an inherent capacity for regenerative growth which subsides once the original organ deficit is restored. This study attempted to resolve the question of whether alteration in hepatic hemodynamics affects the regenerative stimulus of the liver after partial (70 per cent) hepatectomy. It has shown that the liver remnant regenerates after reduction of portal blood flow by construction of a portacaval anastomosis. The diversion of blood from the liver exerts its own histologic and electron microscopic effects on the liver. Reduction of portal blood flow affects the temporal patterns of regeneration after partial hepatectomy but does not prevent completion of the regenerative process. Correlation of this study with the biochemical data available in the literature indicates that the structural changes in the cellular organelles during the process of regeneration re-
162
events that are based fleet dynamic biochemical on a predetermined genetic code representing the key to life that is uniquely found in the liver. References 1. Eck NV: Concerning ligation of the vena porta. Voen Med Zh (St. Petersburg) 130: 1, 1877. 2. Pavlov IP: Modification of the Eck fist&. Arch Bkd /U&r 2: 580, 1893. 3. Mann FC: The portal circulation and restoration of the liver after partiil removal. Surgery 8: 225, 1940. 4. Grindfay JH, Bollman JL: Regeneration of the liver in the dog after partial hepatectomy. Surg Gynecol Obstet 94: 491, 1952. 5. Child CG Ill, Barr D, Holswade GR, Harrison SC: Liier regeneration following portacaval transposition in dogs. Ann Surg 138: 800, 1953. 8. Fisher B, Lee SH, Fisher ER, Saffer E: Liver regeneration following portacaval shunt. Surgery52: 88, 1962. 7. Fisher B, Russ C, Updegraff H, Fisher ER: Effect of increased hepatic blood flow upon liver regeneration. Arch Surg69: 263, 1954. 8. Stephenson’GN: Restoration of the liver after partial hepatectomy and partial ligation of the portal vein. Arch P&ho/ 14: 484, 1932. 9. Weinbren K: The portal blood suppfy and regeneration of the liver. Br J Exp Pathoi36: 583, 1955. 10. Gfinos AD, Grey Go: Humoral factors involved in the induction of liver regeneration in the rat. hoc Sot Exp Biol A&d 80: 421, 1952. 11. Grisham JW: Morphologic study of DNA synthesis and cell proliferation in regenerating rat liver: autoradiography with thymidine-H3. Cancer Res 22: 842, 1969. 12. Bucher NLR: Experimental aspects of hepatic regeneration. N Engl J Med277: 686, 1967. 13. Higgins M, Anderson RM: Experimental pathology of the liver. 1. Restoration of the liver of the white rat following partial surgical removal. Arch P&ho/ 12: 186, 1931. 14. Pack GT. lstami AH: Extent of hepatic resection and the regenerative response. Surgery 40: 6 11, 1956. 15. Harkness RD: Regeneration of the liver. Br A&d Bull 13: 1387, 1957.
The Amarkan
Journal
of Surgery
