American Journal of Hematology 1: 105-114 (1976)
ARTERIOVENOUS SHUNTS IN THE HUMAN SPLEEN Marion 1. Barnhart, Charles A. Baechler, and Jeanne M. Lusher Wayne State University School of Medicine and Children’s Hospital of Michigan, Detroit
The mission of this study was to determine whether or not arteriovenous connections, indicative of a “closed” type of circulation, existed in the human spleen. Spleens from four patients requiring therapeutic splenectomy were the basis for this report. Scanning electron microscopy of plastic corrosion casts, prepared from these four spleens, revealed direct vascular conduits between splenic pulp arteries or arterial capillaries and the venous sinuses in the red pulp. Also demonstrated were a few arteriovenous shunts between pulp arteries or arterial capillaries and pulp or trabecular veins. Inclusion of sized microspheres in low-viscosity perfusion plastic illustrated that some diameters of the connecting shunts were 7- 10 pm, with other shunts even smaller. Not only do arteriovenous connections exist in human spleens, but their frequency, as revealed by methods accentuating threedimensional aspects of the splenic microcirculation, justify future reconsiderations of the functional significance of this closed type of circulation. Examination of samples of the same intact spleens, prepared by freezefracture and conventional critical-point drying, also revealed an “open” type circulation structure, namely, pore-patterned sinus walls that could facilitate blood cell movement from pulp cords into venous sinuses. Scanning electron microscopy thus has provided direct evidence that human spleens have both “open” and “closed” circulatory pathways in their microvasculature. Key words: spleen, A-V shunts, circulation, hereditary spherocytosis, Banti’s syndrome, idiopathic thrombocytopenia
INTRODUCTION
The exact manner of blood circulation through the microvasculature of the spleen has remained controversial since Billroth accurately described the splenic sinuses over 100 yr ago (1-3). Transillumination studies by Knisley (4) were interpreted in favor of a “closed)’ type of circulation, while those by MacKenzie ( 5 ) were supportive of an “open” type. Transmission electron microscopy in several species has not demonstrated a closed circulation (6-lo), but thin-section work is necessarily very selective and provides only two-dimensional information. Even scanning electron microscopy (SEM) of intact spleens, which has offered striking three-dimensional views of the splenic architecture (1 1-14), thus far has not been instructive concerning the existence of a closed circulation. It has been tacitly assumed in the absence of contrary evidence that the splenic circulation is anatomically open but possibly physiologically closed. Address reprint requests to Dr. M. I. Barnhart, Department of Physiology, Wayne State University School of Medicine, 540 East Canfield, Detroit, Michigan 48201.
105
0 1976 Alan R. Liss, Inc.,
150 Fifth Avenue, New York, N.Y. 10011
106
Barnhart, Baechler, and Lusher
Recent observations by SEM of plastic casts of the splenic vasculature provided three-dimensional visual evidence of arteriovenous connections to Barnhart and Baechler (1 5) for the dog, and to Murakami et al. (16) for the rat. Although there are species similarities, differences also exist so that extrapolations to other species are unwise. Our observations (1 S), contrary to those of Murakami et al. (16), led us to propose that there is anatomical evidence for both closed and open types of circulation through the canine splenic red pulp. A unique opportunity was afforded us t o examine the structural physiology of the blood circulation in several human spleens immediately after therapeutic splenectomy. The experimental design utilized scanning electron microscopy of plastic corrosion casts of the splenic vasculature (1 0, 12). In some of these, calibrated plastic microspheres were incorporated to permit judgments on the relative dimensions of various small vessels. Our primary mission was to discover if there was any evidence for a closed type of circulation in the human spleen.
MATERIALS AND METHODS Human Material
Four spleens supplied information for this report. Written statements of informed parental consent permitted us to study these spleens immediately following therapeutic splenectomy for hereditary spherocytosis (one), chronic ITP (one), Banti’s syndrome (one), and hypersplenism in a child with acute lymphoblastic leukemia who had been in continuous remission for several years. Organ Perfusion
Spleens were perfused via splenic artery beginning within 3-5 min from time of excision. Various perfusion fluids at room temperature were delivered with a ColeParmer Masterflex Pump (Chicago) operating at 26 ml/min and at physiologic pressure (1 15 t 10 mm Hg). Following placement of the arterial cannula (an Argyl umbilical artery catheter, size 3% French), about 1 liter of oxygenated, heparinized (2 U/ml) Tyrode’s solution was perfused. The effluent from splenic veins, cannulated after the artery, was collected in calibrated reservoirs. Clearing of blood from the spleen was evident by the color change of the spleen. With careful attention to maintaining the perfusion circuit, small specimens were generally cut at this time for the pathologist. The second perfusion solution, about 1 liter of 1.0%wt/vol glutaraldehyde in 0.06 M cacodylate buffer at pH 7.3, had an osmolarity of 300 mOs and chased the Tyrode’s solution to achieve fixation of the spleen. Specimens cut at this time were suitable for conventional SEM processing (1 5). Plastic Casts
Plastic casts of the vasculature (1 7) were prepared by continuing the perfusion with the substitution of about 150 ml of low-viscosity plastic solution (Batson No. 17 plastic from Polysciences Inc., Warrington, Pa.). When desired, 0.1-mg plastic microspheres (3M Corp., St. Paul, Min.), with diameters of 7-10 pm, were included in the Batson plastic solution which hardens within 30 min of mixing its components. The hardened spleen was suspended and immersed in hot (60°C) KOH (10-20%) overnight. Generally, the
107
Splenic A-V Shunts
processing was improved by a repeat exposure to fresh, hot KOH, which completed digestion of the splenic tissues leaving behind a clean, intact plastic cast of the vasculature. Following washing in cold tap water, the cast was air dried and stored in a dust-free container for later study by SEM. Selection of suitable regions for SEM study was aided by observation and dissection under a Zeiss operation dissection microscope. Conventional Spleen Processing for SEM
Specimens of intact spleen taken prior to plastic perfusion were fixed further by immersion overnight in fresh, cold (4°C) buffered glutaraldehyde. Subsequently, samples were placed in chilled (4°C) sucrose cacodylate buffer for 24-48 hr and then dehydrated in ascending concentrations of ethanol (30, 50,70, 80,90,95, and 100%). Critical-point drying was accomplished immediately in a Parr bomb containing Freon 13 or followed use of Humphrey's freeze-fracture procedure (18) on fixed tissue fractured in liquid nitrogen with reimmersion in 100% ethanol. Electron Microscopy
Segments of the plasticized vasculature or conventional and freeze-fractured tissues were gold sputtered under vacuum. Casts were observed and photographed using an ETEC Autoscan microscope operating at 2% kV to minimize beam damage. Observation and photography of SEM specimens of intact spleen were at accelerating voltages of 10-20 kV. Stage tilting from 0" to 45" and rotations through 360" improved judgments on the continuity of vascular connections. RESULTS
Vasculature of the spleen was strikingly displayed by the plastic corrosion casts which presented replicas of vessel wall interiors (Fig. 1). Observations at low magnifications were helpful in orienting the specimen and for recognition of various regions of the splenic vasculature. Trabecular veins and arteries, the large sausage-like or ropey structures, could be distinguished from one another by their surface appearances and by their continuities with other vessels. Red pulp, chiefly because of its generally extensive and communicating sinus system, was not difficult to identify. The microvasculature of this region was carefully examined for evidence of direct connections between arterial inflow and venous outflow. We elected to examine first spleens from those pathologic states (hereditary spherocytosis or Banti's syndrome) characterized by venous congestion and expected to have a preponderance of red pulp. In vascular casts from both spleens arteriovenous (A-V) connections were found. A-V Shunts Between Pulp Arteries and Sinuses
The splenic cast from the child with hereditary spherocytosis had a very extensive red pulp vasculature. Arteries (Fig. 2, arrow) within the spleen stood out as tapering corrugated conduits. Small arteries were revealed as ropey strands that, in many cases, led directly, or via a capillary, into one or more splenic sinuses (Fig. 2, asterisk). It was not unusual to find several A-V shunts in a field viewed at 200 diam, but not all sinuses had demonstrable connections with arterial inflow. The voIuminous splenic sinuses of the red pulp were frequently enlarged, blunt-ended pouches interconfiecting extensively with one
108
Barnhart, Baechler, and Lusher
another or pulp veins ultimately to converge on trabecular veins. The vacant spaces between the sinus casts represent the splenic cord. The occasional surface depressions in the sinus and pulp vein casts (Fig. 2) have an irregular distribution and denote sites where nuclei of endothelial cells bulged into the lumen (Fig. 3).
Fig. 1. Plastic corrosion cast of splenic vasculature from a case of chronic idiopathic thrombocytopenic purpura. The largest white tubes are trabecular veins. X 16.
Fig. 2. Plastic corrosion cast of vasculature of splenic red pulp. Arterial cast (arrow) branches to empty directly (*) into several venous sinuses. Numerous craters in sinuses and trabecular veins mark sites where endothelial cell nuclei protruded into lumen. X 144. Bar = 50 um.
109
Splenic A-V Shunts
Fig. 3. Freeze-fracture specimen of sinus (S) organization in the splenic red pulp from a case of hereditary spherocytosis. Sinuses have been fractured in both a longitudinal plane and also in crosssectional aspect. The nuclear (N or n) regions of longitudinally oriented endothelial cells bulge into the sinus lumen. Erythrocytes (e), probably neutrophilic leukocytes (L), platelets (p), probably lymphocytes of (t) and (b) varieties can be seen. A macrophage (m) shows its surface ruffles as it lies against the endothelial cells of a sinus, while another is shown “on the move.” A reticular (R) cell has been fractured so that some of the nuclear as well as cytoplasmic interior is evident in addition to the reticular processes that partition the splenic cord (C) region. X 1,600.
Inclusion of plastic microspheres (7-10 pm) in the perfusion plastic and subsequent location of the spheres in the casts permitted judgments on the relative size of the A-V connections. Microspheres accumulated at branch points in the pulp artery too small for them to enter (Fig. 4, arrow). A single bead occasionally blocked the A-V shunt at its point of entry into the splenic sinus (Fig. 4,*). Consequently, some of the A-V shunts were about 7-10 pm in diameter and of adequate size to permit blood cells to pass in single file from the arterial capillary into the venous sinus.
A-V Shunts Between Pulp Arteries and Trabecular Veins
The cast from the spleen of the child with Banti’s syndrome had a very expanded venous component consistent with the clinical problem of portal hypertension (Fig. 5). Some A-V shunts were found connecting arterial capillaries with sinuses. Most spectacular was the finding of A-V shunts permitting arterial inflow into pulp or trabecular veins (Fig. 5, central Y configuration) in the cast from the patient with Banti’s syndrome. To ensure that continuity actually existed between these vessels the specimen was tilted through 45” and rotated through 360” while observing and rephotographing the suspected A-V connections (Fig. 6). Two of the four suspected A-V shunts (Fig. 6) were clearly not visual artifacts of three-dimensional display, superimposition images, but qualified as real replicas of direct connections. This contributing pulp artery, diam 15-20 pm, gave off two branch arterial capillaries (10-15-pm diam) which, by all our criteria, appeared confluent with the trabecular vein (1 13-1 60 pm).
110
Barnhart, Baechler, and Lusher
Fig. 4. Arteriovenous connection of 7-10-pm diam contains a microsphere in the cast (*). Obstruction (arrow) formed where branches of a pulp artery were too narrow to permit passage of spheres. x 312. Bar = 25 pm.
Fig. 5. Plastic corrosion cast of vasculature of red pulp from a case of Banti’s syndrome. Arteries (A) appear as ropey and longitudinally striated tubes. Sinuses are odd-shaped expansions interconnecting with pulp and trabecular veins (V) which are the larger vessel replicas. Centrally, note the Y-shaped configuration made by two arterial capillaries branching from a pulp artery. Both of these capillaries terminate on the adjacent trabecular vein. X 160.
111
Splenic A-V Shunts
Fig. 6 . Enlargement of arteriovenous shunts connecting arterial capillaries of pulp artery (A) with a trabecular vein (V). The specimen was rotated 60” and tilted to show the termination of one of the capillaries on the vein replica. X 1,344.
Arteriovenous connections were identified in red pulp regions of splenic casts from two other patients (the child with acute lymphoblastic leukemia in remission and one with chronic ITP). The former cast exhibited both A-V shunts connecting with sinuses (Fig. 7, arrows), and also an occasional direct pathway from artery to trabecular vein (Fig. 8). DISCUSSION
The pathway(s) for splenic blood flow in that intermediate region between arterial input and venous outflow has continued to intrigue investigators since Billroth’s initial observations (1) and proposal of an unusual or “open” type of circulation (2). Although seldom cited, by 1862 his injection studies in human spleen led him to reverse h s position. He then suggested a “closed” type circulation with direct continuity between terminal arterial capillaries and sinuses in human spleen ( 3 ) . Subsequently, other students of the spleen were also equivocal, although generally agreeing that the open-type circulation was the more important, while direct vessel continuities, if they existed in the intermediate zone, had an insignificant functional role (7,8, 14, 19-21). It is not the purpose of this communication to consider fully the evidence for or against the opposing views. The intent is to recognize that methodology applied in the past has proved unsuitable and has not supplied a satisfactory structural basis for the known splenic hemodynamics. This is largely because direct evidence in three-dimensional form is required t o comprehend the intricacy of the splenic microvasculature. A new type of microscopy (SEM) with advances in microtechnique and plastics chemistry, permitting three-dimensional displays of an organ’s microvasculature, encouraged our re-examination of this provocative question of splenic microcirculation.
112
Barnhart, Baechler, and Lusher
Fig. 7. Plastic corrosion cast of vessels of splenic red pulp from a case of hypersplenism in a child with acute lymphoblastic leukemia in remission. Fine arteriovenous shunts (arrows) provide direct connections between arterial capillaries of pulp arteries (A) and venous sinuses (S). X 160.
Only one aspect of the question is addressed in the present paper: namely, d o direct vascular connections exist between arterial terminations and the venous outflow in human spleens? The answer was unequivocally affirmative for the four types of splenic pathology examined. While it is probable that similar A-V connections also exist in normal human spleen, these have not been demonstrated because we have yet to acquire such a vascular replica. Some comments are in order regarding the relative frequency of these A-V shunts. They are not as rare as suggested in the literature based on two-dimensional microscopy (7,8,14, 19-21). They are more numerous than we appreciated in our preliminary work when we were inadvertently breaking connections by our handling of the brittle casts (15). Now it is not unusual to find several, and sometimes 10 or more A-V shunts in a red pulp area. Moreover, specimens taken at random from other and remote regions of the spleen ordinarily contain some A-V shunts, providing sinuses are intact. This is not to imply that an A-V connection can be traced to every sinus. In fact, many apparent A-V connections, on critical examination involving stage tilting and specimen rotation, turn out to be just running along or over a pulp sinus or vein to other destinations, rather than terminating there. Thus, it becomes essential to verify each suspected A-V shunt for its exact vessel continuities, which not infrequently are at a distance. With conventional histology or electron microscopy on tissue sections, the probability of identifying such A-V connections would appear to be vanishingly small. A-V connections varied not only with respect to the terminal vessel, but also with respect to the diameter of the shunt. Sizing was aided by inclusion of solid plastic microspheres of known dimensions in a reasonable volume of plastic solution. In our study, diameters of A-V shunts varied from 2 p m to 13 pm. Most probably, these dimensions reflected minimum diameters of the shunts since volumes and flow rates were not allowed
113
Splenic A-V Shunts
Fig. 8. An arteriovenous shunt (arrow) connecting a pulp artery (A) with a trabecular vein (V). The specimen was taken from the preceding hypersplenism plastic corrosion cast. X 320.
to exceed physiologic values. The fact that some vessel terminations measured 2-3 pm attests to the adequacy of the plastic perfusion penetrating the smallest of vessels. Most A-V shunts were about 7-10 pm in diameter with dimensions similar to arterial capillaries. Certainly these were of sufficient size that normal blood cells could pass with relative ease directly into the venous circuit. Not infrequently, a plastic microsphere was visible within the small arterial vessels. Occasionally a plastic microsphere remained at the entry point of an arterial capillary discharging into a sinus or pulp vein. Our study thus far has been primarily concerned (1) with making practical and consistent the plastic casting technique on fresh human material, and (2) with establishing better habits for handling the spleen replicas t o preserve most of their vascular integrity. Naturally, we have developed impressions about the comparative aspects noted in this spleen series. These impressions need to be seasoned and adjusted or balanced against the more conventional SEM views of companion specimens cut prior t o preparation of the plastic casts. It does not seem justified at this time to express opinion regarding comparative differences in the closed pathways seen in the splenic pathologies examined. However, it n o longer seems to be an impossible task for the future to acquire, through careful exploration, enough data on frequency of A-V shunts and their characteristics to present the data in a comparative and quantitative manner.
CONCLUSIONS
A-V connections have been clearly identified in four human spleens of diverse pathology. These were demonstrated in sufficient numbers and size to warrant renewed attention to the “closed” type of splenic circulation and to its significance in splenic hemodynamics in health and disease.
114
Barnhart, Baechler, and Lusher
ACKNOWLEDGMENTS
The cooperation of Drs. A. J. Brough, Director of Laboratories, S. Austin, Director of Anesthesiology, J. Hertzler, Chief of Surgery, A. Philippart, and V. Von Berg, Department of Surgery of Children’s Hospital of Michigan was essential for these studies and is gratefully acknowledged. This work was supported by research grant HL 14142-04 from the National Heart and Lung Institute, National Institutes of Health, United States Public Health Service. REFERENCES I . Billroth, T. Beitraege zur vergleichender histologie der milz. Arch. Anat. Physiol. Wissench. Med. 88-108 (1857). 2. Billroth, T. Zur normalen und pathologischen anatomie der menschlichen milz. Arch. Path. Anat. 20:409425 (1861). 3. Billroth, T. Zur normalen und pathologischen anatomie der menschlichen milz. Virchow’s Arch. 231457-477 (1 862). 4. Knisley, M.H. Spleen studies: microscopic observations on the circulatory systems of living unstimulated mammalian spleens. Anat. Rec. 65:23-50 (1936). 5. MacKenzie, D. W., Whipple, A. O., and Winterstein, M. P. Studies on the microscopic anatomy and physiology of living transilluminated mammalian spleens. Am. J. Anat. 68:397-456 (1941). 6. Weiss, L. A study of the structure of splenic sinuses in man and the albino rat with light microscopy and the electron microscope. J. Biophys. Biochem. Cytol. 3:599-610 (1957). 7. Weiss, L. The structure of fine splenic arterial vessels in relation to hemoconcentration and red cell destruction. Am. J. Anat. 111:131-179 (1962). 8. Chen, L. T., and Weiss, L. Electron microscopy of the red pulp of human spleen. Am. J. Anat. 134:425-458 (1 972). 9. Burke, J. S., and Simon, G. T. Electron microscopy of the spleen: 1. Anatomy and microcirculation. Am. J. Pathol. 58:127-155 (1970). 10. Galindo, B., and Freeman, J. A. Fine structure of the splenic pulp. Anat. Rec. 147:25-41 (1973). 11. Fujita, T. A scanning electron microscopy study of the human spleen. Arch. Histol. Japan. 37: 187-2 16 (1 974). 12. Weiss, L. A scanning electron microscopic study of the spleen. Blood 43:665-691 (1974). 13. Leblond, P. F. Etude au microscope electronique a balayage, de la migration des cellules sanguines travers, les parois des sinusoids splenigues et medullaires chez la rat. Nouv. Rev. Fr. Hematol. 13:771-788 (1973). 14. Suzuki, T. Application of scanning electron microscopy in the study of the human spleen: Threedimensional fine structure of the normal red pulp and its changes as seen in splenomegalias associated with Banti’s syndrome and cirrhosis of the liver. Acta Hematol. Japan. 35:506-522 (1972). 15. Barnhart, M. I., and Baechler, C. A. Human and canine splenic vasculature: Structure and function. Scanning Electron Microscopy/l974 (Part 111). Illinois Institute of Technology Research Institute, Chicago, Ill. 705-712 (1974). 16. Murakami, T., Fukita, T., and Miyoshi, M. Closed circulation in the rat spleen as evidenced by scanning electron microscopy of vascular casts. Experientia 29: 1374-1375 (1974). 17. Nowell, J. A., and Lohse, C. L. Injection replication of the microvasculature for SEM. Scanning Electron Microscopy/l974, Illinois Institute of Technology Research Institute, Chicago, Ill. 267-274 (1974). 18. Humphreys, W. J., Spurlock, B. O., and Johnson, J. S. Critical point drying of ethanol-infiltrated, cryofractured biological specimens for scanning electron microscopy. Scanning Electron Microscopy/1974, Illinois Institute of Technology Research Institute, Chicago, 111. 275-282 (1974). 19. MacNeal, W. J., Otani, S., and Patterson, M. B. The finer vascular channels of the spleen. Am. J. Pathol. 3: 111-136 (1927). 20. Bjorkman, S. E. The splenic circulation. Acta Med. Scand. (Suppl. 191) 128:l-89 (1947). 21. Snook, T. The histology of vascular terminations in the rabbit spleen. Anat. Rec. 130: 71 1-723 (1958).
ARTERIOVENOUS SHUNTS IN THE HUMAN SPLEEN Marion 1. Barnhart, Charles A. Baechler, and Jeanne M. Lusher Wayne State University School of Medicine and Children’s Hospital of Michigan, Detroit
The mission of this study was to determine whether or not arteriovenous connections, indicative of a “closed” type of circulation, existed in the human spleen. Spleens from four patients requiring therapeutic splenectomy were the basis for this report. Scanning electron microscopy of plastic corrosion casts, prepared from these four spleens, revealed direct vascular conduits between splenic pulp arteries or arterial capillaries and the venous sinuses in the red pulp. Also demonstrated were a few arteriovenous shunts between pulp arteries or arterial capillaries and pulp or trabecular veins. Inclusion of sized microspheres in low-viscosity perfusion plastic illustrated that some diameters of the connecting shunts were 7- 10 pm, with other shunts even smaller. Not only do arteriovenous connections exist in human spleens, but their frequency, as revealed by methods accentuating threedimensional aspects of the splenic microcirculation, justify future reconsiderations of the functional significance of this closed type of circulation. Examination of samples of the same intact spleens, prepared by freezefracture and conventional critical-point drying, also revealed an “open” type circulation structure, namely, pore-patterned sinus walls that could facilitate blood cell movement from pulp cords into venous sinuses. Scanning electron microscopy thus has provided direct evidence that human spleens have both “open” and “closed” circulatory pathways in their microvasculature. Key words: spleen, A-V shunts, circulation, hereditary spherocytosis, Banti’s syndrome, idiopathic thrombocytopenia
INTRODUCTION
The exact manner of blood circulation through the microvasculature of the spleen has remained controversial since Billroth accurately described the splenic sinuses over 100 yr ago (1-3). Transillumination studies by Knisley (4) were interpreted in favor of a “closed)’ type of circulation, while those by MacKenzie ( 5 ) were supportive of an “open” type. Transmission electron microscopy in several species has not demonstrated a closed circulation (6-lo), but thin-section work is necessarily very selective and provides only two-dimensional information. Even scanning electron microscopy (SEM) of intact spleens, which has offered striking three-dimensional views of the splenic architecture (1 1-14), thus far has not been instructive concerning the existence of a closed circulation. It has been tacitly assumed in the absence of contrary evidence that the splenic circulation is anatomically open but possibly physiologically closed. Address reprint requests to Dr. M. I. Barnhart, Department of Physiology, Wayne State University School of Medicine, 540 East Canfield, Detroit, Michigan 48201.
105
0 1976 Alan R. Liss, Inc.,
150 Fifth Avenue, New York, N.Y. 10011
106
Barnhart, Baechler, and Lusher
Recent observations by SEM of plastic casts of the splenic vasculature provided three-dimensional visual evidence of arteriovenous connections to Barnhart and Baechler (1 5) for the dog, and to Murakami et al. (16) for the rat. Although there are species similarities, differences also exist so that extrapolations to other species are unwise. Our observations (1 S), contrary to those of Murakami et al. (16), led us to propose that there is anatomical evidence for both closed and open types of circulation through the canine splenic red pulp. A unique opportunity was afforded us t o examine the structural physiology of the blood circulation in several human spleens immediately after therapeutic splenectomy. The experimental design utilized scanning electron microscopy of plastic corrosion casts of the splenic vasculature (1 0, 12). In some of these, calibrated plastic microspheres were incorporated to permit judgments on the relative dimensions of various small vessels. Our primary mission was to discover if there was any evidence for a closed type of circulation in the human spleen.
MATERIALS AND METHODS Human Material
Four spleens supplied information for this report. Written statements of informed parental consent permitted us to study these spleens immediately following therapeutic splenectomy for hereditary spherocytosis (one), chronic ITP (one), Banti’s syndrome (one), and hypersplenism in a child with acute lymphoblastic leukemia who had been in continuous remission for several years. Organ Perfusion
Spleens were perfused via splenic artery beginning within 3-5 min from time of excision. Various perfusion fluids at room temperature were delivered with a ColeParmer Masterflex Pump (Chicago) operating at 26 ml/min and at physiologic pressure (1 15 t 10 mm Hg). Following placement of the arterial cannula (an Argyl umbilical artery catheter, size 3% French), about 1 liter of oxygenated, heparinized (2 U/ml) Tyrode’s solution was perfused. The effluent from splenic veins, cannulated after the artery, was collected in calibrated reservoirs. Clearing of blood from the spleen was evident by the color change of the spleen. With careful attention to maintaining the perfusion circuit, small specimens were generally cut at this time for the pathologist. The second perfusion solution, about 1 liter of 1.0%wt/vol glutaraldehyde in 0.06 M cacodylate buffer at pH 7.3, had an osmolarity of 300 mOs and chased the Tyrode’s solution to achieve fixation of the spleen. Specimens cut at this time were suitable for conventional SEM processing (1 5). Plastic Casts
Plastic casts of the vasculature (1 7) were prepared by continuing the perfusion with the substitution of about 150 ml of low-viscosity plastic solution (Batson No. 17 plastic from Polysciences Inc., Warrington, Pa.). When desired, 0.1-mg plastic microspheres (3M Corp., St. Paul, Min.), with diameters of 7-10 pm, were included in the Batson plastic solution which hardens within 30 min of mixing its components. The hardened spleen was suspended and immersed in hot (60°C) KOH (10-20%) overnight. Generally, the
107
Splenic A-V Shunts
processing was improved by a repeat exposure to fresh, hot KOH, which completed digestion of the splenic tissues leaving behind a clean, intact plastic cast of the vasculature. Following washing in cold tap water, the cast was air dried and stored in a dust-free container for later study by SEM. Selection of suitable regions for SEM study was aided by observation and dissection under a Zeiss operation dissection microscope. Conventional Spleen Processing for SEM
Specimens of intact spleen taken prior to plastic perfusion were fixed further by immersion overnight in fresh, cold (4°C) buffered glutaraldehyde. Subsequently, samples were placed in chilled (4°C) sucrose cacodylate buffer for 24-48 hr and then dehydrated in ascending concentrations of ethanol (30, 50,70, 80,90,95, and 100%). Critical-point drying was accomplished immediately in a Parr bomb containing Freon 13 or followed use of Humphrey's freeze-fracture procedure (18) on fixed tissue fractured in liquid nitrogen with reimmersion in 100% ethanol. Electron Microscopy
Segments of the plasticized vasculature or conventional and freeze-fractured tissues were gold sputtered under vacuum. Casts were observed and photographed using an ETEC Autoscan microscope operating at 2% kV to minimize beam damage. Observation and photography of SEM specimens of intact spleen were at accelerating voltages of 10-20 kV. Stage tilting from 0" to 45" and rotations through 360" improved judgments on the continuity of vascular connections. RESULTS
Vasculature of the spleen was strikingly displayed by the plastic corrosion casts which presented replicas of vessel wall interiors (Fig. 1). Observations at low magnifications were helpful in orienting the specimen and for recognition of various regions of the splenic vasculature. Trabecular veins and arteries, the large sausage-like or ropey structures, could be distinguished from one another by their surface appearances and by their continuities with other vessels. Red pulp, chiefly because of its generally extensive and communicating sinus system, was not difficult to identify. The microvasculature of this region was carefully examined for evidence of direct connections between arterial inflow and venous outflow. We elected to examine first spleens from those pathologic states (hereditary spherocytosis or Banti's syndrome) characterized by venous congestion and expected to have a preponderance of red pulp. In vascular casts from both spleens arteriovenous (A-V) connections were found. A-V Shunts Between Pulp Arteries and Sinuses
The splenic cast from the child with hereditary spherocytosis had a very extensive red pulp vasculature. Arteries (Fig. 2, arrow) within the spleen stood out as tapering corrugated conduits. Small arteries were revealed as ropey strands that, in many cases, led directly, or via a capillary, into one or more splenic sinuses (Fig. 2, asterisk). It was not unusual to find several A-V shunts in a field viewed at 200 diam, but not all sinuses had demonstrable connections with arterial inflow. The voIuminous splenic sinuses of the red pulp were frequently enlarged, blunt-ended pouches interconfiecting extensively with one
108
Barnhart, Baechler, and Lusher
another or pulp veins ultimately to converge on trabecular veins. The vacant spaces between the sinus casts represent the splenic cord. The occasional surface depressions in the sinus and pulp vein casts (Fig. 2) have an irregular distribution and denote sites where nuclei of endothelial cells bulged into the lumen (Fig. 3).
Fig. 1. Plastic corrosion cast of splenic vasculature from a case of chronic idiopathic thrombocytopenic purpura. The largest white tubes are trabecular veins. X 16.
Fig. 2. Plastic corrosion cast of vasculature of splenic red pulp. Arterial cast (arrow) branches to empty directly (*) into several venous sinuses. Numerous craters in sinuses and trabecular veins mark sites where endothelial cell nuclei protruded into lumen. X 144. Bar = 50 um.
109
Splenic A-V Shunts
Fig. 3. Freeze-fracture specimen of sinus (S) organization in the splenic red pulp from a case of hereditary spherocytosis. Sinuses have been fractured in both a longitudinal plane and also in crosssectional aspect. The nuclear (N or n) regions of longitudinally oriented endothelial cells bulge into the sinus lumen. Erythrocytes (e), probably neutrophilic leukocytes (L), platelets (p), probably lymphocytes of (t) and (b) varieties can be seen. A macrophage (m) shows its surface ruffles as it lies against the endothelial cells of a sinus, while another is shown “on the move.” A reticular (R) cell has been fractured so that some of the nuclear as well as cytoplasmic interior is evident in addition to the reticular processes that partition the splenic cord (C) region. X 1,600.
Inclusion of plastic microspheres (7-10 pm) in the perfusion plastic and subsequent location of the spheres in the casts permitted judgments on the relative size of the A-V connections. Microspheres accumulated at branch points in the pulp artery too small for them to enter (Fig. 4, arrow). A single bead occasionally blocked the A-V shunt at its point of entry into the splenic sinus (Fig. 4,*). Consequently, some of the A-V shunts were about 7-10 pm in diameter and of adequate size to permit blood cells to pass in single file from the arterial capillary into the venous sinus.
A-V Shunts Between Pulp Arteries and Trabecular Veins
The cast from the spleen of the child with Banti’s syndrome had a very expanded venous component consistent with the clinical problem of portal hypertension (Fig. 5). Some A-V shunts were found connecting arterial capillaries with sinuses. Most spectacular was the finding of A-V shunts permitting arterial inflow into pulp or trabecular veins (Fig. 5, central Y configuration) in the cast from the patient with Banti’s syndrome. To ensure that continuity actually existed between these vessels the specimen was tilted through 45” and rotated through 360” while observing and rephotographing the suspected A-V connections (Fig. 6). Two of the four suspected A-V shunts (Fig. 6) were clearly not visual artifacts of three-dimensional display, superimposition images, but qualified as real replicas of direct connections. This contributing pulp artery, diam 15-20 pm, gave off two branch arterial capillaries (10-15-pm diam) which, by all our criteria, appeared confluent with the trabecular vein (1 13-1 60 pm).
110
Barnhart, Baechler, and Lusher
Fig. 4. Arteriovenous connection of 7-10-pm diam contains a microsphere in the cast (*). Obstruction (arrow) formed where branches of a pulp artery were too narrow to permit passage of spheres. x 312. Bar = 25 pm.
Fig. 5. Plastic corrosion cast of vasculature of red pulp from a case of Banti’s syndrome. Arteries (A) appear as ropey and longitudinally striated tubes. Sinuses are odd-shaped expansions interconnecting with pulp and trabecular veins (V) which are the larger vessel replicas. Centrally, note the Y-shaped configuration made by two arterial capillaries branching from a pulp artery. Both of these capillaries terminate on the adjacent trabecular vein. X 160.
111
Splenic A-V Shunts
Fig. 6 . Enlargement of arteriovenous shunts connecting arterial capillaries of pulp artery (A) with a trabecular vein (V). The specimen was rotated 60” and tilted to show the termination of one of the capillaries on the vein replica. X 1,344.
Arteriovenous connections were identified in red pulp regions of splenic casts from two other patients (the child with acute lymphoblastic leukemia in remission and one with chronic ITP). The former cast exhibited both A-V shunts connecting with sinuses (Fig. 7, arrows), and also an occasional direct pathway from artery to trabecular vein (Fig. 8). DISCUSSION
The pathway(s) for splenic blood flow in that intermediate region between arterial input and venous outflow has continued to intrigue investigators since Billroth’s initial observations (1) and proposal of an unusual or “open” type of circulation (2). Although seldom cited, by 1862 his injection studies in human spleen led him to reverse h s position. He then suggested a “closed” type circulation with direct continuity between terminal arterial capillaries and sinuses in human spleen ( 3 ) . Subsequently, other students of the spleen were also equivocal, although generally agreeing that the open-type circulation was the more important, while direct vessel continuities, if they existed in the intermediate zone, had an insignificant functional role (7,8, 14, 19-21). It is not the purpose of this communication to consider fully the evidence for or against the opposing views. The intent is to recognize that methodology applied in the past has proved unsuitable and has not supplied a satisfactory structural basis for the known splenic hemodynamics. This is largely because direct evidence in three-dimensional form is required t o comprehend the intricacy of the splenic microvasculature. A new type of microscopy (SEM) with advances in microtechnique and plastics chemistry, permitting three-dimensional displays of an organ’s microvasculature, encouraged our re-examination of this provocative question of splenic microcirculation.
112
Barnhart, Baechler, and Lusher
Fig. 7. Plastic corrosion cast of vessels of splenic red pulp from a case of hypersplenism in a child with acute lymphoblastic leukemia in remission. Fine arteriovenous shunts (arrows) provide direct connections between arterial capillaries of pulp arteries (A) and venous sinuses (S). X 160.
Only one aspect of the question is addressed in the present paper: namely, d o direct vascular connections exist between arterial terminations and the venous outflow in human spleens? The answer was unequivocally affirmative for the four types of splenic pathology examined. While it is probable that similar A-V connections also exist in normal human spleen, these have not been demonstrated because we have yet to acquire such a vascular replica. Some comments are in order regarding the relative frequency of these A-V shunts. They are not as rare as suggested in the literature based on two-dimensional microscopy (7,8,14, 19-21). They are more numerous than we appreciated in our preliminary work when we were inadvertently breaking connections by our handling of the brittle casts (15). Now it is not unusual to find several, and sometimes 10 or more A-V shunts in a red pulp area. Moreover, specimens taken at random from other and remote regions of the spleen ordinarily contain some A-V shunts, providing sinuses are intact. This is not to imply that an A-V connection can be traced to every sinus. In fact, many apparent A-V connections, on critical examination involving stage tilting and specimen rotation, turn out to be just running along or over a pulp sinus or vein to other destinations, rather than terminating there. Thus, it becomes essential to verify each suspected A-V shunt for its exact vessel continuities, which not infrequently are at a distance. With conventional histology or electron microscopy on tissue sections, the probability of identifying such A-V connections would appear to be vanishingly small. A-V connections varied not only with respect to the terminal vessel, but also with respect to the diameter of the shunt. Sizing was aided by inclusion of solid plastic microspheres of known dimensions in a reasonable volume of plastic solution. In our study, diameters of A-V shunts varied from 2 p m to 13 pm. Most probably, these dimensions reflected minimum diameters of the shunts since volumes and flow rates were not allowed
113
Splenic A-V Shunts
Fig. 8. An arteriovenous shunt (arrow) connecting a pulp artery (A) with a trabecular vein (V). The specimen was taken from the preceding hypersplenism plastic corrosion cast. X 320.
to exceed physiologic values. The fact that some vessel terminations measured 2-3 pm attests to the adequacy of the plastic perfusion penetrating the smallest of vessels. Most A-V shunts were about 7-10 pm in diameter with dimensions similar to arterial capillaries. Certainly these were of sufficient size that normal blood cells could pass with relative ease directly into the venous circuit. Not infrequently, a plastic microsphere was visible within the small arterial vessels. Occasionally a plastic microsphere remained at the entry point of an arterial capillary discharging into a sinus or pulp vein. Our study thus far has been primarily concerned (1) with making practical and consistent the plastic casting technique on fresh human material, and (2) with establishing better habits for handling the spleen replicas t o preserve most of their vascular integrity. Naturally, we have developed impressions about the comparative aspects noted in this spleen series. These impressions need to be seasoned and adjusted or balanced against the more conventional SEM views of companion specimens cut prior t o preparation of the plastic casts. It does not seem justified at this time to express opinion regarding comparative differences in the closed pathways seen in the splenic pathologies examined. However, it n o longer seems to be an impossible task for the future to acquire, through careful exploration, enough data on frequency of A-V shunts and their characteristics to present the data in a comparative and quantitative manner.
CONCLUSIONS
A-V connections have been clearly identified in four human spleens of diverse pathology. These were demonstrated in sufficient numbers and size to warrant renewed attention to the “closed” type of splenic circulation and to its significance in splenic hemodynamics in health and disease.
114
Barnhart, Baechler, and Lusher
ACKNOWLEDGMENTS
The cooperation of Drs. A. J. Brough, Director of Laboratories, S. Austin, Director of Anesthesiology, J. Hertzler, Chief of Surgery, A. Philippart, and V. Von Berg, Department of Surgery of Children’s Hospital of Michigan was essential for these studies and is gratefully acknowledged. This work was supported by research grant HL 14142-04 from the National Heart and Lung Institute, National Institutes of Health, United States Public Health Service. REFERENCES I . Billroth, T. Beitraege zur vergleichender histologie der milz. Arch. Anat. Physiol. Wissench. Med. 88-108 (1857). 2. Billroth, T. Zur normalen und pathologischen anatomie der menschlichen milz. Arch. Path. Anat. 20:409425 (1861). 3. Billroth, T. Zur normalen und pathologischen anatomie der menschlichen milz. Virchow’s Arch. 231457-477 (1 862). 4. Knisley, M.H. Spleen studies: microscopic observations on the circulatory systems of living unstimulated mammalian spleens. Anat. Rec. 65:23-50 (1936). 5. MacKenzie, D. W., Whipple, A. O., and Winterstein, M. P. Studies on the microscopic anatomy and physiology of living transilluminated mammalian spleens. Am. J. Anat. 68:397-456 (1941). 6. Weiss, L. A study of the structure of splenic sinuses in man and the albino rat with light microscopy and the electron microscope. J. Biophys. Biochem. Cytol. 3:599-610 (1957). 7. Weiss, L. The structure of fine splenic arterial vessels in relation to hemoconcentration and red cell destruction. Am. J. Anat. 111:131-179 (1962). 8. Chen, L. T., and Weiss, L. Electron microscopy of the red pulp of human spleen. Am. J. Anat. 134:425-458 (1 972). 9. Burke, J. S., and Simon, G. T. Electron microscopy of the spleen: 1. Anatomy and microcirculation. Am. J. Pathol. 58:127-155 (1970). 10. Galindo, B., and Freeman, J. A. Fine structure of the splenic pulp. Anat. Rec. 147:25-41 (1973). 11. Fujita, T. A scanning electron microscopy study of the human spleen. Arch. Histol. Japan. 37: 187-2 16 (1 974). 12. Weiss, L. A scanning electron microscopic study of the spleen. Blood 43:665-691 (1974). 13. Leblond, P. F. Etude au microscope electronique a balayage, de la migration des cellules sanguines travers, les parois des sinusoids splenigues et medullaires chez la rat. Nouv. Rev. Fr. Hematol. 13:771-788 (1973). 14. Suzuki, T. Application of scanning electron microscopy in the study of the human spleen: Threedimensional fine structure of the normal red pulp and its changes as seen in splenomegalias associated with Banti’s syndrome and cirrhosis of the liver. Acta Hematol. Japan. 35:506-522 (1972). 15. Barnhart, M. I., and Baechler, C. A. Human and canine splenic vasculature: Structure and function. Scanning Electron Microscopy/l974 (Part 111). Illinois Institute of Technology Research Institute, Chicago, Ill. 705-712 (1974). 16. Murakami, T., Fukita, T., and Miyoshi, M. Closed circulation in the rat spleen as evidenced by scanning electron microscopy of vascular casts. Experientia 29: 1374-1375 (1974). 17. Nowell, J. A., and Lohse, C. L. Injection replication of the microvasculature for SEM. Scanning Electron Microscopy/l974, Illinois Institute of Technology Research Institute, Chicago, Ill. 267-274 (1974). 18. Humphreys, W. J., Spurlock, B. O., and Johnson, J. S. Critical point drying of ethanol-infiltrated, cryofractured biological specimens for scanning electron microscopy. Scanning Electron Microscopy/1974, Illinois Institute of Technology Research Institute, Chicago, 111. 275-282 (1974). 19. MacNeal, W. J., Otani, S., and Patterson, M. B. The finer vascular channels of the spleen. Am. J. Pathol. 3: 111-136 (1927). 20. Bjorkman, S. E. The splenic circulation. Acta Med. Scand. (Suppl. 191) 128:l-89 (1947). 21. Snook, T. The histology of vascular terminations in the rabbit spleen. Anat. Rec. 130: 71 1-723 (1958).
