You are on page 1of 5

Clinical Anatomy 20:956960 (2007)

REVIEW

The Evidence for the Spinal Segmental Innervation of Bone


JASON J. IVANUSIC*
Department of Anatomy and Cell Biology, University of Melbourne, Victoria, Australia

Dermatomes and myotomes are areas of skin and muscle, respectively, that are innervated by single spinal segmental nerves, and reect a principle of organization that appears in just about every clinical textbook available today. The evidence for the existence of dermatomes and myotomes has a long and substantial history. A lesser known, but similar principle exists for the skeletal system. The term sclerotome was rst used in the non-embryological sense by Inman and Saunders ([1944] J. Nerv. Ment. Dis. 99:660667) to dene a region of bone and periosteum that is innervated by a single spinal segment. It is used by clinicians in many healthcare settings to aid in the diagnosis and description of a variety of deep and/or skeletal tissue pathologies and pain syndromes. In this article, the evidence for the existence of the sclerotomes is described in detail. Early clinical studies that dene the sclerotomes, evidence from studies of the development of skeletal innervation, and the contributions of anatomical and physiological investigations are explored. It is suggested that there is in fact little direct evidence for the existence of discrete spinal segmental innervation patterns for the skeleton. Clin. Anat. 20:956960, 2007. V 2007 Wiley-Liss, Inc.
C

Key words: sclerotomes; segmental innervation of bone; skeletal innervation

INTRODUCTION
Dermatomes and myotomes are areas of skin and muscle, respectively, that are innervated by single spinal segmental nerves, and reect a principle of organization that appears in just about every clinical textbook available today. Whilst textbook descriptions are often simplistic (for example, they fail to account for considerable overlap in territories innervated by adjacent spinal nerves), they are used widely by clinicians in many healthcare settings to aid in the diagnosis and description of a variety of conditions. The evidence for the existence of dermatomes and myotomes has a long and substantial history (Warwick et al., 1989; Greenberg, 2003). A lesser known, but similar principle exists for the skeletal system. The term sclerotome was rst used in the nonembryological sense by Inman and Saunders (1944) to dene a region of bone and periosteum that is innervated by a single spinal segment. This denition appears as a secondary meaning in (some) medical dictionaries. The primary use of the term by most anatomists and clinicians is in descriptions of the ventromedial part of the embryological somite, which ultimately forms the axial skeleton (see

Evidence From Developmental Studies). This embryological denition cannot extend to the adult because the group of cells that constitute the somite do not exist beyond the embryonic stage, and the denition of Inman and Saunders cannot be applied to the early stages of development because the innervation of skeletal tissue is yet to fully develop. To avoid ambiguity in the present review, the term sclerotome is used to dene a region of bone and periosteum innervated by a single spinal segment. When referring to the embryological meaning, the term embryological sclerotome is used. The same convention is applied to the terms dermatome and myotome. A number of clinicians have used the sclerotomes, as dened by Inman and Saunders (1944), in early discus*Correspondence to: Dr. Jason Ivanusic, Department of Anatomy and Cell Biology, University of Melbourne, Melbourne, VIC 3010, Australia. E-mail: j.ivanusic@unimelb.edu.au Received 25 March 2007; Revised 15 August 2007; Accepted 4 September 2007 Published online 22 October 2007 in Wiley InterScience (www. interscience.wiley.com). DOI 10.1002/ca.20555

C 2007 V

Wiley-Liss, Inc.

The Sclerotomes
sions of backache (Rose, 1954; Wilson, 1967), cancer (Klingon, 1961, 1964), brositis (Yunus et al., 1981), herniated discs (Inman and Saunders, 1947), referred and deep pain (Southworth and Krohn, 1950), nerve blocks to treat pain (Dobney and Belam, 1963), and more recently, in discussions of bone disorders such as melorheostosis (Murray and McCredie, 1979; North and McCredie, 1987) and thalidomide induced malformations (McCredie, 1977; McCredie et al., 1984; McCredie and Willert, 1999). Chiropractors today use published maps of sclerotomes in describing deep, scleratogenous pain (Banner-Therapy-ProductsInc., 2007). However, upon reviewing the literature, few studies of the spinal segmental innervation of bone could be found. In this article, the evidence for the existence of the sclerotome is reviewed. It is suggested that there is in fact little direct evidence for the existence of the sclerotome.

957

CLINICAL OBSERVATIONS
Dejerine (1914) was the rst to construct a map of the spinal segmental innervation of the bony skeleton. His work was based on two principles. The rst was the observation that cutaneous (skin) and deep (muscle, tendon, bone, and periosteum) sensitivity to pressure could, in cases of certain central and peripheral nervous lesions, be dissociated because an anesthetized area of skin may overly deeper tissue that was not anesthetized, or vice versa. For example, a C5 lesion causing cutaneous anesthesia over a bone, without coincident deep anesthesia beneath the area of anesthetized skin, indicated to Dejerine that the deep tissue in that region was not innervated by the C5 spinal segment. The second was meticulous cadaver dissection of spinal nerves from their radicular origin. It is important to note however, that much of the pressure Dejerine applied to deep structures must have stimulated afferent bers in muscle, tendon and ligaments, rather than in periosteum and bone alone. There are only a few sites in which bone and periosteum are in close apposition to skin without other structures in between. Furthermore, it is difcult to conceive that such a dened representation of the innervation territory of single spinal nerves could be achieved by following the bers within a single spinal nerve through a peripheral plexus (e.g., lumbosacral plexus), multiple peripheral nerves, and the very ne branches that innervate bone and periosteum. Kellgren (1939) reported a similar map by examining the distribution of referred pain following injections of hypotonic saline into deep tissues such as ligaments, tendons, joint capsules, and muscles. When these deep tissues were injected, the author noted that the pain experienced by subjects was referred to cutaneous and muscular areas associated with dened dermatomes or myotomes, and proposed that the innervation of the deep tissue must have the same spinal segmental origin as the dermatome or myotome that the pain was referred to. Using this technique, Kellgren carefully identied the spinal segmental innervation patterns for deep tissue throughout the whole body. The similarity between Kellgrens and Dejerines maps conrms that what Dejerine mapped was most likely the innervation of muscle, tendon and ligaments, not bone and periosteum alone. Furthermore, at no stage was the term sclerotome used in the study of Kellgren, and no claim was made that this map was representing the segmental

innervation of bone and periosteum, yet some recent reports quote this work when dening the sclerotomes. A notable example is the published chiropractic charts of the sclerotomes (Banner Therapy Products Inc., 2007). The term sclerotome was not actually dened until later. This highlights the notion that the use of the term sclerotome has become somewhat ambiguous. In many cases, it has been used to imply the spinal segmental innervation of deep tissues, including muscle, tendon, joint capsule, and ligament as well as bone and periosteum, but this is an incorrect use of the term. Inman and Saunders (1944) were in fact the rst to use the term sclerotome in the non-embryological sense. They studied the distribution of referred pain to non-bony tissues in patients with focal bone lesions and a group of volunteers that submitted to direct periosteal stimulation (scratching with a ne needle), and reported a map of the spinal segments that were known to innervate the areas of referred pain for each stimulus site, in a manner similar to that of Kellgren. However, the distribution of referred pain is difcult to document accurately because it is generally described as diffuse and dull, and often perceived over a number of dermatomes or myotomes. In fact, Inman and Saunders acknowledged that: Our knowledge of the precise segmental innervation of the sclerotomal areas cannot at the moment be given with any great precision. They also commented that their aim was to map the extent of scleratogenous pain, not to produce a segmental map, contrary to the opinions of future authors. There exist a number of confounding issues regarding the sclerotomes as derived from these early clinical observations. The rst issue is that it is unclear as to how many points on the bony skeleton were actually sampled in these studies. It is difcult to imagine how these stimuli could be applied to the whole skeleton when deep tissues such as ligaments, tendons, and muscles overly bone at most points in the skeleton. The second issue is the notion that when stimuli were applied specically to skeletal tissues, they were applied to the periosteum (by pressure or scratching) and not to the inside of bone. We are now aware that primary afferent bers exist within bone (Bjurholm et al., 1988; Hill and Elde, 1991; Mach et al., 2002), and that they may be activated with many different stimulus types (Furusawa, 1970; Seike, 1976). The segmental innervation of the bone as opposed to the periosteum is yet to be dened at all. A third point of contention is that most of these studies were based on mapping the distribution of referred or deep pain, but because pain of these types is so dull and diffuse, it is difcult to ascribe the origin of the pain to a single dermatome or myotome and thus to a single spinal segment. Finally, referred pain is most likely an indication of common segmental spinal cord mechanisms. It is probable that spinal and supraspinal circuits inuenced the extent of referred pain perceived by subjects in these studies, confounding the results reported. The accuracy of the sclerotomes presented in these early studies was therefore questionable, and in line with the perceived lack of clinical importance at the time, led to their exclusion in the literature for many years. It should be emphasized that the sclerotome maps marketed at present (aimed mainly at chiropractic use e.g., Banner Therapy Products Inc., 2007) are based on these early clinical observations. As a result, they show only vague areas of deep, radiating or referred pain, and not specic skeletal territories supplied by single

958

Ivanusic
presents in the adult, and thus the pathological distribution of the sclerotic tissue can be compared with the sclerotome maps of Inman and Saunders (1944) with more condence than thalidomide induced limb reductions. Nonetheless, a clearer understanding of how segmental innervation patterns of skeletal tissues in both the embryo and the adult are organized is required to support the hypotheses that these authors present, and are indeed necessary to reinforce the idea that sclerotomes do exist in a form useful to such discussion.

spinal nerves. A more concise method for mapping the sclerotomes would result in a more effective use of the principle in the clinical setting. A number of more recent reports of patients with bone disorders such as melorheostosis (Murray and McCredie, 1979; North and McCredie, 1987) and thalidomide induced malformations (McCredie, 1977; McCredie et al., 1984; McCredie and Willert, 1999) have cited the sclerotomes to explain the pathology they observed. Thalidomide is wellknown for its anti-angiogenic and immuno-modulatory properties, but was also used as an antiemetic agent to control nausea in pregnant females in the 1960s (Perri and Hsu, 2003; Franks et al., 2004). However, babies born to mothers that had taken thalidomide developed gross deformities involving skin, muscle, and skeletal structures. Of the skeletal deformities, reductions of the limbs involving malformation of the long bones were most obvious (McCredie, 1977; McCredie et al., 1984; McCredie and Willert, 1999). McCredie (1974) hypothesized that this was a result of toxic damage to neural crest cells (the segmentally organized precursors of peripheral sensory neurons). She predicted that if a segment of neural crest fails to develop, then one would expect the skeletal structures supplied by the sensory nerves derived from that segment to be missing (or at least malformed), possibly because of impaired neurotrophic inuences. McCredie retrospectively examined the radiographs of limb reductions in thalidomide babies, and observed that most limb reductions could be approximated by theoretically subtracting a single sclerotome, as mapped by Inman and Saunders (1944), from a complete appendicular skeleton (McCredie, 1977; McCredie et al., 1984; McCredie and Willert, 1999). She argued that the sclerotomes provided a mechanism by which one could explain the effects of thalidomide poisoning, and therefore that they must indeed exist. However, there is some debate as to whether thalidomide affects neural crest cells directly, or causes secondary neuropathy related to an effect on target tissues (de Iongh, 1990). In addition, Strecker and Stephens (1983) showed that physically blocking the early development of peripheral nerves did not result in any limb reductions at all. DAmato et al. (1994) further showed that thalidomide is a potent inhibitor of angiogenesis, and Strecker and coworkers (Stephens et al., 2000; Stephens and Fillmore, 2000) subsequently suggested that thalidomide inhibits the transcription of genes that drive angiogenesis. Therefore, it appears that the observed malformations might result from a lack of blood supply to the developing limb, not from a direct toxic effect on developing segmental sensory neurons. This suggests that McCredies hypothesis regarding the importance of the segmentally organized neural crest cells in thalidomide teratogenicity may have been overstated, and raises the possibility that the implied existence of the sclerotomes may not be real. Melorheostosis is a disease that is characterized by abnormal longitudinal growth of sclerotic tissue in long bones. Murray and McCredie (1979) and North and McCredie (1987) loosely correlated the distribution of sclerotic tissue in patients with the distribution of single sclerotomes, as dened by Inman and Saunders (1944). They noted that in some cases, cutaneous disturbances associated with a single dermatome of the same segmental origin were also present, reinforcing the notion that single spinal segments could be associated with dened skeletal tissue territories. An important feature of these studies is that melorheostosis

EVIDENCE FROM DEVELOPMENTAL STUDIES


The relationship between developing, segmentally organized sensory and motor neurons in the embryo and spinal segmental innervation patterns of skin and muscle in the adult is well documented (Krull, 2001; Wang and Scott, 2002). Primary afferent neurons that innervate skin are derived from neural crest cells that migrate away from the developing neural tube and undergo signicant proliferation adjacent to primitive somites. Motorneurons originate in the basal plate of the developing neural tube and their developing axons project into the primitive somites. Somites are organized in a segmental manner, so when the peripheral processes of neural crest cells, or axons of developing motorneurons, extend into the adjacent somites they provide a template for the segmentation of these neurons. The somites subsequently differentiate into three regions: an embryological dermatome, embryological myotome, and embryological sclerotome. The embryological dermatome develops into axial skin and the embryological myotome develops into both axial and appendicular muscle tissue, all of which retain the segmental innervation patterns imposed by the developing neurons. Skin of the limbs develops from tissues associated with the outgrowth of the limb bud, so the peripheral processes of developing neural crest cells have to extend beyond the somite, and out to the developing limb bud to innervate their nal target tissue. There is signicant evidence that this outgrowth is mediated by either neurotrophic factors and/or contact guidance mechanisms, and that the segmentation is maintained when the processes travel over such long distances (Davies and Lumsden, 1990; Wang and Scott, 2002). Thus, it appears that the spinal segmental innervation patterns of skin and muscle of the adult develop in part from patterns imposed by somites during embryonic development. If the embryonic development of the innervation of skeletal structures follows the same principles that exist for skin and muscle, then it would seem reasonable to presume that developing neurons in the embryo might also impose segmental innervation patterns on the adult skeleton (and manifest as sclerotomes). In the development of the axial skeleton, neural crest cells extend peripheral processes into the surrounding embryological sclerotome, forming the basis for segmental innervation of the axial skeleton, including the vertebra and ribs. However, this is somewhat complicated because each embryological sclerotome is split: its cranial part forms the caudal part of the vertebra above, and its caudal part forms the cranial part of the vertebra below (Balling et al., 1992). This implies that in the adult, a single spinal segmental nerve should innervate two adjacent vertebrae. This is not acknowledged in the maps of

The Sclerotomes
Dejerine (1914). Furthermore, Kellgren (1939) and Inman and Saunders (1944) did not document segmental innervation patterns for axial tissues at all, and therefore offer no clarication on the matter. The case for the appendicular skeleton is also unclear. Peripheral processes of neural crest cells extend into the mesenchyme of the limb bud (Cameron and McCredie, 1982) which subsequently forms the bones of the developing limb (Olsen et al., 2000). However, the extent to which the developing primary afferent bers maintain segmental innervation patterns in developing limb bones is yet to be determined. Until these issues are resolved, some caution should be practiced before presuming that the development of segmentally organized nerves in the embryo imposes a segmental innervation pattern on adult skeletal tissues, as it appears to for cutaneous and muscular tissue. The evidence to imply that this might form the basis of the sclerotomes does not as yet exist.

959

that anatomical and physiological techniques need to be exploited further to study the sclerotomes. In particular, empirical mapping of connectivity between single spinal segmental nerves and bony tissues, as has been done for skin and muscle, is required to conrm the existence of the sclerotomes.

CONCLUDING REMARKS
The sclerotomes are used by clinicians in many healthcare settings to aid in the diagnosis and description of a variety of deep and/or skeletal tissue pathologies and pain syndromes. However, we are currently relying on early clinical or experimental observations of referred mechanisms of deep pain to approximate the sclerotomes. The maps reported in the early literature are not complete, and they should be interpreted with consideration of limits in the techniques used. They show only vague areas of deep, radiating or referred pain, and not specic skeletal territories supplied by single spinal nerves. Further to this, a denitive description of the sclerotomes in the adult is complicated by the lack of knowledge regarding the development of skeletal innervation in the embryo. We do not have the evidence to imply that the development of segmentally organized nerves imposes a segmental innervation pattern on skeletal tissues, as it appears to for cutaneous and muscular tissue. Finally, there is little evidence based on anatomical and physiological studies of the sort have been instrumental in conrming the existence of segmental innervation patterns in skin and muscle. Thus there appears to be little direct evidence for the existence of the sclerotomes. It is proposed that the nature of the segmental innervation of bony tissue requires further attention. In particular, more detail of the anatomical and physiological connections between single spinal nerves and bone is required. Like the dermatomes and myotomes presented in many of the textbooks today, this will reect the true patterns of segmental innervation of skeletal tissue in the adult, and provide a denitive basis upon which discussions of the sclerotomes can be based in the future.

ANATOMICAL AND PHYSIOLOGICAL INVESTIGATIONS IN ANIMAL STUDIES


It is relatively easy to study the connections between a single spinal nerve and either skin or muscle in animal studies. Connectivity has been established with electrophysiological recordings of spinal nerve activity evoked by stimulation of skin or muscle (Dykes and Terzis, 1981), muscle activity evoked by electrical stimulation of single spinal nerves (Browne, 1950; Thage, 1965), application of neuroanatomical tracers to skin and muscle (Takahashi et al., 2003), and by dye extravasation following electrical stimulation of single spinal nerves (Takahashi and Nakajima, 1996). Using these techniques, it has been possible to dene territories innervated by single spinal nerves and then reconstruct a concise map representing the innervation patterns of many spinal nerves in skin and muscle. Maps of dermatomes and myotomes produced in this way are very similar to maps produced from clinical (and experimental) observations in human studies (Warwick et al., 1989; Greenberg, 2003), and have provided empirical evidence that spinal segmental innervation patterns do indeed exist for skin and muscle. However, examining the connectivity between the bony skeleton and spinal nerves is difcult, because bone is a very hard, mineralized tissue that is difcult to work with. Most of the techniques applied to establishing dermatomes and myotomes are not appropriate for the determination of sclerotomes. A single study has attempted to address the issue of the segmental innervation of bony tissue using anatomical techniques. Gajda et al. (2004) placed a retrograde tracer (DiI) under the periosteum of the rat tibia, which is easy to expose because of its close proximity to the skin, and reported labeled neuronal cell bodies in the dorsal root ganglia conned predominantly to spinal cord levels L3 and L4. The author of the present paper has observed similar results following injections of another retrograde tracer (Fast Blue) into either the medullary cavity or epiphyses, in addition to the periosteum, of the rat tibia (Ivanusic, unpublished observations). However, using this approach, it would be very difcult and time consuming to place the tracer into enough points in the skeleton to be able to reconstruct the whole sclerotomal map. No other anatomical or physiological studies of connectivity between single spinal nerves (or spinal cord segments) and bone could be found in the literature. It is clear

ACKNOWLEDGMENTS
The author thanks Dr. Pascal Carrive and Dr. Marius Fahrer for their assistance in the translation of Dejerines monograph (Dejerine, 1914).

REFERENCES
Balling R, Lau CF, Dietrich S, Wallin J, Gruss P. 1992. Development of the skeletal system. Ciba Found Symp 165:132143. Banner-Therapy-Products-Inc. 2007. Sclerotome Pain Chart. URL: http://www.bannertherapy.com/ProductInfo.aspx?sclerotome-painchart&number35-201. [accessed August 2007]. Bjurholm A, Kreicbergs A, Brodin E, Schultzberg M. 1988. Substance Pand CGRP-immunoreactive nerves in bone. Peptides 9:165171. Browne KM. 1950. The spatial distribution of segmental nerves to striate musculature of the hindlimb of the rat. J Comp Neurol 93: 441455. Cameron J, McCredie J. 1982. Innervation of the undifferentiated limb bud in rabbit embryo. J Anat 134:795808. DAmato RJ, Loughnan MS, Flynn E, Folkman J. 1994. Thalidomide is an inhibitor of angiogenesis. Proc Natl Acad Sci USA 91:4082 4085.

960

Ivanusic
McCredie J, Willert HG. 1999. Longitudinal limb deciencies and the sclerotomes. An analysis of 378 dysmelic malformations induced by thalidomide. J Bone Joint Surg Br 81:923. McCredie J, North K, de Iongh R. 1984. Thalidomide deformities and their nerve supply. J Anat 139:397410. Murray RO, McCredie J. 1979. Melorheostosis and the sclerotomes: A radiological correlation. Skeletal Radiol 4:5771. North K, McCredie J. 1987. Neurotomes and birth defects: A neuroanatomic method of interpretation of multiple congenital malformations. Am J Med Genet Suppl 3:2942. Olsen BR, Reginato AM, Wang W. 2000. Bone development. Annu Rev Cell Dev Biol 16:191220. Perri AJ III, Hsu S. 2003. A review of thalidomides history and current dermatological applications. Dermatol Online J 9:5. Rose GK. 1954. Backache and the disc. Lancet 266:11431149. Seike W. 1976. Electrophysiological and histological studies on the sensibility of the bone marrow nerve terminal. Yonago Acta Med 20:192211. Southworth JL, Krohn S. 1950. Myotome and scleratome pain. Ann Surg 132:965971. Stephens TD, Fillmore BJ. 2000. Hypothesis: Thalidomide embryopathy-proposed mechanism of action. Teratology 61:189195. Stephens TD, Bunde CJ, Fillmore BJ. 2000. Mechanism of action in thalidomide teratogenesis. Biochem Pharmacol 59:1489 1499. Strecker TR, Stephens TD. 1983. Peripheral nerves do not play a trophic role in limb skeletal morphogenesis. Teratology 27:159 167. Takahashi Y, Nakajima Y. 1996. Dermatomes in the rat limbs as determined by antidromic stimulation of sensory C-bers in spinal nerves. Pain 67:197202. Takahashi Y, Chiba T, Kurokawa M, Aoki Y. 2003. Dermatomes and the central organization of dermatomes and body surface regions in the spinal cord dorsal horn in rats. J Comp Neurol 462:2941. Thage O. 1965. The myotomes L2S2 in man. Acta Neurol Scand Suppl 13:241243. Wang G, Scott SA. 2002. Development of normal dermatomes and somatotopic maps by abnormal populations of cutaneous neurons. Dev Biol 251:424433. Warwick R, Williams P, Dyson M, Bannister L. (eds.) 1989. Grays Anatomy, 37th Ed. Edinburgh, New York: Churchill Livingston. Wilson JC Jr. 1967. Low back pain and sciatica. A plea for better care of the patient. JAMA 200:705712. Yunus M, Masi AT, Calabro JJ, Miller KA, Feigenbaum SL. 1981. Primary bromyalgia (brositis): Clinical study of 50 patients with matched normal controls. Semin Arthritis Rheum 11:151 171.

Davies AM, Lumsden A. 1990. Ontogeny of the somatosensory system: Origins and early development of primary sensory neurons. Annu Rev Neurosci 13:6173. de Iongh RU. 1990. A quantitative ultrastructural study of motor and sensory lumbosacral nerve roots in the thalidomide-treated rabbit fetus. J Neuropathol Exp Neurol 49:564581. Dejerine J. 1914. Semiologie des Affections du Systeme Nerveaux. Paris: Masson. Reprinted in 1977 edition. Dobney GH, Belam OH. 1963. The scope of nerve root block in physical medicine. Proc R Soc Med 56:444446. Dykes RW, Terzis JK. 1981. Spinal nerve distributions in the upper limb: The organization of the dermatome and afferent myotome. Philos Trans R Soc Lond B Biol Sci 293:509554. Franks ME, Macpherson GR, Figg WD. 2004. Thalidomide. Lancet 363: 18021811. Furusawa S. 1970. A neurophysiological study on the sensibility of the bone marrow. Nippon Seikeigeka Gakkai Zasshi 44:365370. Gajda M, Litwin JA, Adriaensen D, Timmermans JP, Cichocki T. 2004. Segmental distribution and morphometric features of primary sensory neurons projecting to the tibial periosteum in the rat. Folia Histochem Cytobiol 42:9599. Greenberg SA. 2003. The history of dermatome mapping. Arch Neurol 60:126131. Hill EL, Elde R. 1991. Distribution of CGRP-, VIP-, D beta H-, SP-, and NPY-immunoreactive nerves in the periosteum of the rat. Cell Tissue Res 264:469480. Inman V, Saunders J. 1944. Referred pain from skeletal structures. J Nerv Ment Dis 99:660667. Inman V, Saunders J. 1947. Anatomicophysiological aspects of injuries to the intervertebral disc. J Bone Jt Surg 29:461467. Kellgren JH. 1939. On the distribution of pain arising from deep somatic structures with charts of segmental pain areas. Clin Sci 4: 3547. Klingon GH. 1961. Neurologic problems in cancer. Med Clin North Am 45:585599. Klingon GH. 1964. Some thoughts on pain localization. J Int Coll Surg 42:396403. Krull CE. 2001. Segmental organization of neural crest migration. Mech Dev 105:3745. Mach DB, Rogers SD, Sabino MC, Luger NM, Schwei MJ, Pomonis JD, Keyser CP, Clohisy DR, Adams DJ, OLeary P, Mantyh PW. 2002. Origins of skeletal pain: Sensory and sympathetic innervation of the mouse femur. Neuroscience 113:155166. McCredie J. 1974. Embryonic neuropathy. A hypothesis of neural chest injury as the pathogenesis of congenital malformations. Med J Aust 1:159163. McCredie J. 1977. Sclerotome subtraction: A radiologic interpretation of reduction deformities of the limbs. Birth Defects Orig Artic Ser 13:6577.