Armin K. Thron Vascular Anatomy Spinal Cord of the N euroradiological Investigations and Clinical Syndromes With collaboration of Ch. Rossberg and A. Mironov Springer-Verlag Wien New York Armin K. Thron, M.D. Professor of Neuroradiology Department of Neuroradiology, University of Tiibingen, Federal Republic of Germany Present address: Department of Neurology, Klinikum RWTH Aachen, Federal Republic of Germany This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocopying machine or similar means, and storage in data banks. © 1988 by Springer-Verlag Wien Softcover reprint ofthe hardcover 1st edition 1988 With 74 partly colored Figures ISBN-13: 978-3-7091-7446-3 e-ISBN-13: 978-3-7091-6947-6 DOl: 10.1007/978-3-7091-6947-6 Preface The idea for this treatise on the radiological anatomy of superficial and deep spinal cord vasculature evolved from daily routine neuroradiological work. This was also the reason for subdividing the monograph into a postmortem anatomical and a clinical part. The actual importance of a clear conception of radio anatomic fundamentals was made clear by many clinical conferences with neurologists, neurosurgeons and orthopedists, where a lack of knowledge about medullary syndromes of suspected vascular origin became evident. Also among neuroradiologists there is still widespread uncertainty in the interpretation of myelograms and angiographies in such cases. A study of the spinal cord's angioarchitecture is all the more justified and necessary considering the vast number of descriptions of cerebro vascular anatomy and pathology. The clinical challenge posed by patients suffering from partial or complete transverse spinal lesions has grown due to new diagnostic and therapeutic approaches. Myelography using water-soluble contrast media, X-ray computed tomography, magnetic resonance imaging and spinal angiography today allow and require both earlier and topographically and pathogenetically more exact classification of diseases of the spinal cord and its surrounding structures. Due to progress in microneurosurgery and interventional neuroradiology, even intramedullary lesions have become more and more accessible and treatable. Therefore this monograph mainly addresses those concerned with invasive therapeutic techniques and who are familiar with the interpretation of radio anatomic findings. A comprehensive description of medullary vascular syndromes would be beyond the scope of this treatise. Much more physiological and pathophysiological data will have to be collected for this to be possibly accomplished in the future. Our present diagnostic standard is illustrated by the neuroradiological studies in patients with spinal vascular malformations. The primary intent of this book, however, is a contribution to the widely unknown anatomy of the medullary venous system. Many angiographical findings indicate an important pathogenetic role of the spinal drainage system. Further clarification of these hemodynamic problems will require even more interdisciplinary cooperation among physiologists, neurologists, neuroradiologists, neurosurgeons and neuropathologists. Aachen, January 1988 Armin K. Thron Acknowledgement These studies were performed at the Department of Neuroradiology, University of Tubingen, and it is to the Department's head, Prof. Dr. K. Voigt, that I express my gratitude for constant support. Comprehensive postmortem and clinical investigations are impossible without the assistance of colleagues. In the first place I would like to thank Dr. Ch. RoJ3berg from the Neuropathological Department at the University of Marburg for removing and preparing the postmortem specimens. Without her assistance, the study could not have been realized. Furthermore, I owe thanks to my colleague Dr. A. Mironov for his help in postmortem preparations. Valuable support was also lent by Prof. Dr. J. Peiffer, Institute for Brain Research at the University of Tubingen and by his collaborators who kindly provided their laboratory facilities as well as by Prof. Dr. Dauber, who made the initial anatomical studies possible. Dr. Poremba provided the instructive schematic drawings of figures 2-5. Special thanks are further expressed to Mr. Wiehr for wonderful photographic work and his keen interest as well as to Mrs. Virginia Muller and Dr. H. Steinmetz for the translation. Furthermore, I would like to thank Prof. Dr. J. J. Merla,hd (Hopital Lariboisiere, Paris) and Prof. Dr. B. Kendall (National Hospital, Queens Square, London) for the i~spiration gained through their work and personal contact. Contents I. Introduction . . . . . . . . . . . . . . . . . . . . . . . 1 II. Previous Studies on the Spinal Vascular System . . . . . . 2 III. Experimental Methods and Clinical Examination Techniques 3 1. Anatomical Methods and Materials 3 2. Injection Techniques . . . . . . . . . . . . . . 3 3. Radiographic Techniques . . . . . . . . . . . . 4 3.1. Postmortem Macro- and Microangiography 4 3.2. Spinal Angiography in Animals 4 3.3. Clinical Spinal Angiography 5 3.4. Myelography . . . . . . . . . 5 4. Own Investigations . . . . . . . . . 5 IV. Anatomy of the Spinal Cord's Blood Supply 8 1. Embryology . . . . . . . . . . . . . 8 2. Sources of Arterial Blood Supply and Radicular Feeders 8 3. Extramedullary Venous Drainage .......... . 11 V. Postmortem Angiography and Microangiography of Spinal Cord Vessels. 13 A. Arteries . . . . . . . . . . . 13 1. Superficial Arterial System 13 1.1. Cervical Region . . . 13 1.2. Thoracic Region. . . 17 1.3. Thoracolumbar Region 18 2. Intrinsic Arterial System 26 2.1. Central System (Sulcal or Central Arteries) 26 2.2. Peripheral System (Vasocorona) 32 3. Arterial Territories of Supply . . . . . . . . . 34 4. Arterio-arterial Anastomoses . . . . . . . . . 35 5. Discussion of Arteriographical Findings-Physiological Aspects 37 B. Veins ........... . 39 1. Superficial Venous System 39 2. Intrinsic Venous Systems . 51 3. Transmedullary Venous Anastomoses 55 4. Discussion of Phlebographical Findings-Physiological Aspects 56 C. Capillaries of the Spinal Cord . . . . . . . . . 58 D. Spinal Cord Arteries in Pathological Conditions 58 VI. Clinical Applications . . . . . . . . . . . . . . 65 1. Normal Anatomy in Selective Spinal Angiography 65 2. Spinal Vascular Malformations . . . . . . . . . 65 2.1. Myelography . . . . . . . . . . . . . . . 65 2.2. Angiographical Findings and Their Classification 66 2.3. Discussion ............. .. 95 3. Vertebrospinal Space-occupying Lesions . . . . 105 VII. Pathomorphological and Pathophysiological Aspects 106 References 109 Subject Index . . . . . . . . . . . . . . . . . . . . . 113 L Introduction The relative inaccessibility of the medulla within the ficiencies in the venous system have occasionally re spinal column constitutes a major obstacle for both in ceived pathoanatomical and surgical attention (Stoch vivo and postmortem studies. This fact as well as the dorph, 1961, 1969; Jellinger, 1966; Neumayer, 1966; complex and highly variable blood supply of the spinal Wiillenweber, 1969). cord explain why our knowledge about the physiology The anatomy of the spinal vessels forms the basis and pathophysiology of medullary circulation is still for our understanding of primary or secondary spinal limited. As late as 1958, Gillilan mentions that knowl cord circulatory disorders. Standard handbooks usu edge of the spinal vascular supply was apparently so ally display this anatomy only schematically. The lack deficient among neurologists and neurosurgeons that of direct reproductions, controversial findings due to clinical studies almost always began with a synopsis of interindividual variations of spinal vessels, and the the normal medullary blood supply. Clinical syndromes present inability of in vivo angiography to describe the such as the anterior spinal artery syndrome, intermit entire spinal circulation, all these inadequacies led to tent spinal claudication (Verbiest, 1954, 1976, Jellinger an encompassing radiological study of the spinal vas and Neumayer, 1972) and the vascular myelopathies cular system in postmortem specimens. (Neumayer, 1967) are largely unexplained with regard This radioanatomical work, supplemented by com to their pathogenesis. From our present morphological parative studies of medullary lesions forms the first knowledge we cannot understand the course taken by part of this book: the repeatedly reported fibrocartilaginous emboli of It provides a systematical, three-dimensional repro intervertebral disk tissue into the anterior spinal artery duction of the superficial and intrinsic spinal arteries (Peiffer et al. 1976). Until recently, the subacute nec and veins using microradiographical techniques. rotizing myelitis (Foix-Alajouanine disease), later The clinical problems outlined above have gained termed "angiodysgenetic necrotizing myelopathy" actuality through progress in neurophysiological di (Scholz and Manuelidis, 1950), was considered a dis agnostics, microneurosurgical techniques and espe tinct clinical entity, and only lately could its underlying cially through the new neuroradiological methods of vessel anomaly be satisfactorily demonstrated and in examination. Selective spinal angiography, developed terpreted. Disorders whose pathogenesis also involves since the sixties, myelography using watersoluble con vascular mechanisms, such as the syndromes of the trast media and magnetic resonance imaging (MRI) narrow spinal canal or of slowly progressive tumor have all expanded the diagnostic capabilities in spinal compression, were usually explained in the past by processes. decreasing arterial supply. Despite the well-known The second, clinical section of this book is devoted pathoanatomical fact that arteriosclerosis of the med to these diagnostic improvements with emphasis on ullary vessels is extremely rare, literature on spinal cir spinal arteriovenous malformations (A VM). This clin culatory disorders abounds in explanations analogous ical work as well as pathomorphological and patho to the conditions in ischemic brain disease, where ob physiological considerations must be regarded as only structions in the arterial vasculature are often found. the present state of knowledge in a rapidly developing This becomes understandable in light of the fact that field of diagnostic potentials and therapeutic possibil even today the possible impact of venous disorders ities of interventional neuroradiology. cannot be clinically estimated, although circulatory de- II. Previous Studies on the Spinal Vascular System Our limited knowledge about the physiology and den, 1956; Hogan and Romanul, 1966; Adams, 1984), pathology of the spinal blood supply contrasts with neurological syndromes of assumed vascular origin anatomical descriptions of the spinal vascular system (Kalm, 1953; Ziilch, 1954, 1974; Gruner and Lapresle, from as early as the end of the 19th century by 1962), and the spinal vascular malformations that had Adamkiewicz (1881, 1882) and, with a precision that become diagnosable with the introduction of intravital has hardly been achieved since then, by Kadyi (1889). spinal angiography (Djindjian et aI., 1963, 1970; Di Earlier studies, such as those by Duret (1873) and Ross Chiro etaI., 1967; Doppmann etaI., 1969). These clin (1880) were less defined, incomplete and unsystematic. ical challenges have since then led to an expansion and Until about the middle of this century, the spinal vas deepening of our knowledge. Noteworthy are the stud cular system was seldom the object of thorough ana ies of Gillilan (1958, 1970), of Lazorthes and co-work tomical research. However, one must mention the com ers (1958 to 1973), Noeske (1958), Roll (1958), Perese prehensive works of Tureen (1938), Suh and Alexander and Fracasso (1959), Clemens and v. Quast (1960), (1939) as well as Herren and Alexander (1939), who Corbin (1961), Houdart etai. (1965), Romanes (1965), called back to mind the results of Kadyi (1889). Hassler (1966), Jellinger (1966, 1972), Mannen (1966), According to these previous investigations the spinal Turnbull etai. (1966), Fazio and Agnoli (1970), Ma cord is supplied by one ventral and two dorsal arterial nelfe et ai. (1972), Piscol (1972), Domisse (1975), Tveten anastomotic truncs. They are supplied at the cranial end (1976), Crock and Yoshizawa (1977), Crock etai. by feeders from the distal segment of the vertebral ar (1986). Except for a few authors such as Kadyi (1889), teries, as well as by a varying number of lateral feeders Suh and Alexander (1939), Clemens and v. Quast of different caliber entering with the nerve roots at dif (1961), or Crock and Yoshizawa (1977), the arterial ferent levels. These radicular feeding arteries originate system was more extensively, if not exclusively, ex from segmental arteries or homologous vessels. On the amined. Venous anatomy was focussed upon in the basis of this anatomical knowledge, Bolton (1939) car studies of v. Quast (1961) as well as Gillilan (1970). ried out injection studies to examine blood flow direc The vertebral venous system, connected to the veins of tion and vascularization territories of the medulla. As the medulla, was demonstrated quite early by Breschet of 1950, the interest in problems of spinal vascular (1828 to 1832) and later by Batson (1957) and Clemens supply was stimulated by progress and questions newly (1961). raised by clinical work. Some examples of these are Comparative animal experiments were conducted neurological complications following surgery of the mainly by Woollam and Millem (1955), Gouaze etai. aorta and spinal column (Adams and Van Geertruy- (1964), Jellinger (1966 a and b) and Tveten (1976). IlL Experimental Methods and Clinical Examination Techniques 1. Anatomical Methods and Materials ular feeders and the course and caliber of the ventral and dorsal spinal arteries (Jellinger, 1966a; Mannen, Anatomical investigative techniques for the spinal vas 1966). cular system are based upon preparations using mac roscopic-microscopic inspection, histological serial sec tions and numerous methods of vascular injection. The 2. Injection Techniques latter are especially valuable for demonstrating con tinuous vessel courses. In addition to non-hardening As feeding and draining vessels of the medulla originate from various vascular territories, particular problems staining fluids such as ink, usually mixed with gelatin are encountered in injection techniques. No method (Bolton, 1939; Herren and Alexander, 1939; Suh and exists as of yet which is qualitatively suitable for a Alexander, 1939; Gillilan, 1958; Corbin, 1961; Vuia complete demonstration of both the extra- and intra and Alexaniu, 1969), many hardening substances were spinal vessels. used as well. Adamkiewicz (1881) and Kaydi (1889) In principle, the most appropriate technique is con injected stained glutin and a Teichmann cement of sidered to be a global filling of all feeders via the aorta, chalk and linseed oil. Other notable substances are imitating natural conditions. This was employed by PANSCH-mass (Noeske, 1958; Roll, 1958), celloidin Gillilan (1958), Domisse (1975) and Tveten (1976) for (Bolton, 1939), latex solution (Gillilan, 1958; Clemens, all of the examined specimens, and by Corbin (1961), 1961; J ellinger, 1966 a; Domisse, 1975) or neoprene Hassler (1966) and Jellinger (1966a) only for a part. latex solution (Perese and Fracasso, 1959), synthetic Aside from the unfeasibility for routine examinations, resin (Lazorthes, 1958; Corbin, 1961) or polymerized the large injection volume of several liters necessary plastics such as plastoid (Gillilan, 1958; Piscol, 1972). even after ligation of the peripheral major arterial These and other substances which harden out well fa trunks poses an additional problem. This was reduced cilitate corrosive preparations after maceration in po by Gillilan (1958), Domisse (1975) and Tveten (1976) tassium hydroxide solution. by studying perinatal autopsy cases. This injection tech Not all substances are equally good for filling both nique is well-suited for determining the extraspinal large and small vessels and may easily cause extrava sources of arterial supply and their collatera1s. Corbin sation. Therefore, different methods were often applied (1961) and Hassler (1966) additionally carried out se separately, especially if visualization of the outer and lective filling of radicular arteries on the removed spinal inner medullary vessels was intended. For the study of cord to demonstrate the inner spinal vessels. vessels within the parenchyma, Adamkiewicz (1881) This selective injection technique had already been and Kaydi (1889) used a clearing method with clove applied by Adamkiewicz (1881) and Kadyi (1889). De oil, mentioned by Virchow (1857). This technique, fur spite numerous collaterals on the medullary surface, ther developed by Spalteholz (1914), renders tissue sec complete filling is usually impossible, unless several tions transparent, so that the stained vessels can be anterior radicular feeders are injected simultaneously pursued over a certain length under the microscope by or successively. This deficiency, on the other hand, is turning the knurled screw. advantageous for studying vascular territories on trans Even without vascular injection, the superficial verse and longitudinal sections if enough fractioned medullary vessels of the exposed spinal cord can be injections can be applied. Nevertheless, only very re studied both macroscopically and with a microscope. stricted conclusions concerning intramedullary in vivo This is easy with veins oflarge caliber (v. Quast, 1961), hemodynamics can be deduced from postmortem in but can also be applied to the determination of radic- jection experiments (Turnbull et aI., 1966; Piscol, 1972). 4 III. Experimental Methods and Clinical Examination Techniques The difficulties of an adequate filling technique are (Treichel et aI., 1977) it is likewise applicable to cap discussed at length by Noeske (1958), Roll (1958) and illary preparations. Piscol (1972) carried out flow tests Corbin (1961), yet the authors rarely quote exact data with iodine contrast medium, partially using serial an on the rate of success. giography. The quality and completeness of filling depend on The possibility of also demonstrating histological many factors: structures by the use of extremely fine-grained film and 1. type and quantity of the injected medium; soft X-rays led to the development of microangiog 2. the specimen's state of preservation (autolysis, raphy, thus named by Bellmann (1953). This technique thrombi); of radiological microscopy and microradiography is 3. vascular perforation in the vicinity of the injection based on the physical investigations of Cosslet et aI. site or intraparenchymal extravasations; (1957) and Engstrom et aI. (1960). Microangiograph 4. leakage of contrast medium from small disrupted ical techniques were conducted on the central nervous vessels in removed spinal cords; system with focus on the cerebral vessels (Saunders, 5. impossibility of filling very small radicular arteries. 1960; Lazorthes, 1961; Dor and Salomon, 1970; Sal For that reason the material for evaluating the outer omon 1973; Duvernoy, 1978). and inner microvascular architecture must be assem Neuroradiological studies of embryonic cerebral bled like a mosaic from a large number of cases in development by Voigt and Stoeter (1980) and of ex order to obtain a complete picture of all medullary perimental cerebral tumors by Schumacher (1981) em segments. One of the main reasons for divergent results ployed microangiographical techniques. These works can certainly be seen in these methodological problems include extensive surveys of the literature and meth which can only be solved at great expense. odological discussions on microangiography. The injection techniques applicable for the spinal Microangiography of spinal vascular structures has drainage system differ because global filling is usually received only sporadic attention. It was demonstrated impossible from extradural. This experience from post exemplarily by Saunders (1960) on a fetal spinal cord mortem investigations (Breschet, 1828 to 1832; Batson, in the context of an extensive study on the cerebral 1957; Clemens, 1961, Crock et aI. 1986) corresponds vessels. The first systematic, strictly microradiograph to results of spinal in vivo phlebography (Theron and ical investigations are the studies by Turnbull, Breig Moret, 1978; Vogelsang, 1980). Consequently, the few and Hassler (1966) on the cervical spinal cord and by existing studies on intradural and intramedullary veins Hassler (1966) on the thoraco-lumbar region; yet these are based on macroscopic-microscopic inspection studies only refer to the arterial component. Further (Quast. 1961; Jellinger, 1966 a; Domisse, 1975) or se more, one must mention the microangiographical in lective intradural filling of radicular veins (Adamkiew vestigations of the spinal dura mater by Manelfe et aI. icz, 1881, 1882; Kadyi, 1889; Suh and Alexander, 1939; (1972). Lazorthes et aI., 1962; Gillilan, 1970) Satisfactory post The studies of Tveten (1976) on the spinal vascular mortem filling through arteries and capillaries into the system in men and rats are partially based on micro veins is seldom achieved (Domisse, 1975). radiological methods. Extended application of this method can be found in animal experiments (Dopp man, 1975; Doppman and Girton, 1976; Doppman et aI., 1977, 1979). 3. Radiographic Techniques 3.1. Postmortem Macro- and Microangiography 3.2. Spinal Angiography in Animals Lazorthes etaI. (1958) and later Corbin (1961) com Animal experiments using intravital or postmortem bined classical methods and injection techniques with spinal angiography have gained importance not only postmortem X-ray arteriograms after injecting barium for comparative neuroradiological anatomy (Woollam sulphate, thereby contributing essentially to the ra and Millem, 1955; Tveten, 1976), but especially for the dioanatomy of the spinal vascular system. Several fur investigation of physiological and pathopyhsiological ther studies e.g. by Houdart et aI. (1965), Di Chiro and questions. (Gouaze et aI., 1964; Fried et aI., 1970; Fried Doppman (1975), Tveten (1976), Doppman et al. (1977, and Aparicio, 1973; Ramsay and Doppman, 1973; 1979) made use of the excellent contrast characteristics Doppman and Girton, 1976, 1977; Doppman etaI., of barium sulphate. With its average grain size of 0,5 ~ 1979)