thoracolumbar mortice joint: radiological and histological observations

Clinical Biomechanics 1989; 4: 137-l 43

Thoracolumbar mortice joint: radiological and histological observations K P Singer DipPT, MSc

Department of Anatomy & Human Biology, The University of Western Australia, Australia

Summary

The superior articular processes and mammillary processes of T,, and T12 and L, often form a ‘mortice-like’ enclosure for the inferior articular processes of the vertebra above. From 630 thoracolumbar junction CT investigations, 297 individuals demonstrated one or more mortice joints. The frequency and variety of mortice joints from Tlo-,, to T 12-L1 were classified into three main types, according to segmental level and orientation of their zygapophysial joints. The T,l_12 level accounted for 77% of all mortice joints, with 2Q% at T12-L, and 3% at T,,,_,,. In 27% of cases a unilateral mortice joint was demonstrated, often in association with marked zygapophysial joint tropism. Radiographic features of 43 cadaveric mortice joints were compared with their histological characteristics.

Relevance

Apposition of the thoracolumbar mortice joint in orthograde or extended postures may constrain vertebral motion at the transition between the lumbar and thoracic regions of the human vertebral column. This articulation putatively limits axial rotation during normal activities but may also contribute to localization of injuries at the thoracolumbar junction.

Key words: Thoracolumbar junction, zygapophysial joints, mortice joint, spinal segmental motion

Introduction

The thoracolumbar (T-L) transitional region of the human vertebral column is a common site of serious injury due to the marked structural and functional change occurring between the thoracic and lumbar reg- ions14. Experimental studies on T-L autopsy spines have indicated that the transitional and upper lumbar segments exhibit high torsional stiffness’-‘. This has been attributed largely to the zygapophysial joints which. at these levels. are orientated predominantly in the sagittal plane’. In 1877, Topinard’ described a ‘mortice’ like zygapophysial joint congruency at the T-L junction formed between the superior articular processes (SAP) and mammillary processes (MP). with the inferior ar- titular processes (IAP) of the vertebra above. Obser- vations regarding the specific anatomy of these

Received: 23 November I988 Accepted: 24 April 19x9 Correspondence and reprint requests to: Kevin P. Singer, Centre for Human Biology, Department of Anatomy & Human Biology, The University of Western Australia, Nedlands. WA 6009 Australia

0 1989 Butterworth & Co (Publishers) Ltd 0268~0033/89#30137-07 $03.00

zygapophyses were subsequently reported by Le Dou- ble’“, Kaplan”, Davis” and Malmivaara et al.“. Davis” speculated that when the cranial IAP was approximated into the mortice. as in normal orthograde or postures incorporating extension, greater stability would be af- forded to the T-L transition. Progressive unlocking of this joint mechanism would occur during flexion.

The objective of this study was to record the frequency and variety of thoracolumbar mortice joint (TLMJ) types demonstrated from a radiological survey of T-L junction CT scans of T,,,_, ,, Tll-r2 and Tr2-L, zygapophysial joints. Radiographic features of cadaveric mortice joints were compared with their histological characteristics.

Methods

A retrospective review of 630 T-L junction CT studies was undertaken to select scans orientated close to the superior end-plates of TN,, T12 and L,, respectively. Scans were excluded due to (i) poor image quality, (ii) structural, surgical or pathological changes or (iii), asymmetrical scan planes. Vertebral levels were deter- mined with the aid of digital scanograms of the pelvis and thoracolumbar spine, in addition to plain films if

138 C/in. Biomech. 1989; 4: No 3

available. The lowest thoracic vertebra was recognized as having ribs originating symmetrically from the root of the pedicle14. Equivocal cases were deleted from the series. The final sample comprised patient scans of the upper abdomen (87%), T-L junction vertebrae (6%) and cadaveric T-L junctions (7%).

Where evident, bilateral mortice joints were classified according to an interpretation of the three groups used by Davis’*. These appeared to be related to three characteristic patterns of zygapophysial joint orientations found at the T-L junctiona, namely: those joints directed towards either the coronal and sagittal planes, and those situated in an intermediate orientation (Figure 1). These distinctions between bilateral mortice types, although arbitrary, could be recognized as follows: (i) a Type I TLMJ specifies a zygapophysial joint orientated toward the sagittal plane and is produced when the SAPS and MPs envelope the IAPs, (ii) a Type II mortice joint is formed by a zygapophysial joint orientated in between the sagittal and coronal planes; its IAPs enclosed by prominent MPs situated posteriorly and (iii), a Type III mortice joint classification denoted a coronally orientated zygapophysial joint in which the IAPs were bounded posteriorly by MPs (Figure 2b). In general, a type I TLMJ appeared to afford the greatest depth of IAP enclosure; however, this aspect depended upon the development and disposition of the MP in relation to the IAP; that is, very prominent MPs associated with types II and III mortice joints would also encircle the lateral aspect of the cranial IAPs.

In addition to these three principal categories, unilateral mortice joints were also demonstrated (Figure 2a). A fourth group was identified in association with zygapophysial joint asymmetry (tropism), in which a unilateral MP was often identified on the same side as the more coronally facing zygapophysis and occasionally representing a combination of two mortice types at the same level (Figure 2~).

The frequency of each mortice joint type was recorded with regard to whether there was an abrupt or gradual change from coronal to sagittal plane orientation of the T-L junction zygapophysial jointsx.

To inspect the condition of the TLJM articular surfaces, T-L junction segments from TI~,I to Ll_2, removed from 58 cadavers, were sectioned on a bandshaw and prepared for histological examination. A modification of the method described by Giles and Taylor” was employed, involving decalcification, dehydration and embedding the vertebral joints in a 1O:l mixture of low-viscosity nitrocellulose and Necoleidine. Serial transverse sections were cut in the plane of the superior end-plate at a thickness of - 200 pm using a Reichert- Jung Polycut-S motorized microtome, then stored in 70% ethanol between numbered sheets of paper. Each joint was divided into tenths and these sections were stained in Ehrlich’s haematoxylin with Light Green counterstain and mounted in DePeX on glass slides for light microscopy. The mid-joint section, corresponding to the level of the superior end-plate, was used to record the frequency and type of mortice joints. Radiographic

Figure 1. Schematic representation of three types of bilateral thoracolumbar mortice joints. Main differences relate to zygapophysial joint orientation and the extent to which the inferior articular processes are enclosed by the superior articular and mammillary processes (shaded). The lower axial sections designated Types I, II and Ill, represent the plane of the CT scans parallel with the superior vertebral end-plate.

Figure 2. CT scans demonstrating variations in thoracolumbar mortice joints. a depicts a Type I mortice at T,2-L, showing greater development of the right mammillary process (arrow) compared to the left. b demonstrates a bilateral Type II mortice joint at Tll_12 showing the intermediate orientation of the zygapophysial joints characteristic of a gradual transition pattern. The mammillary processes (arrows) project medially to enclose the inferior articular processes. c represents a combined form of mortice joint with a Type Ill joint on the right and aType II on the left of this T,,-12 zygapophysial joint.

features of these mortice joints were compared with their histological characteristics.

Results

Mortice joints were demonstrated in 297 cases out of the 630 CT studies (47%); this subset comprised 190 males and 107 females. Twenty-nine of these individuals (3%) exhibited two mortice joints at adjacent levels, to produce a total of 326 TLMJs. Bilateral mortice joints made up 73% of the total. Of these, Type II TLMJs accounted for 44%) followed by Type I (14%) and Type III (15%) (Table 1). The remainder represented unilateral forms of mortice joints which were associated with Type I (7%), Type III (6%) or Type II (3%) TLMJs

Singer: Thoracolumbar mortice joint 139

(Figure 3a-c). The remaining 11% comprised unilateral or combined mortice joints that were commonly ob- served in cases of marked zygapophysial joint tropism ( > 30”) (Figure 3d) or those that exhibited two mortice types, for example Types II and III (Figure 2c). In the CT study, Types I and III mortice joints were most commonly associated with an abrupt pattern of transition (Table 2) (Figure 4), whereas the Type II mortice was frequently related to a gradual transition pattern (Table 2). The greater proportion of mortice joints were encountered at Tl,_12 (77%), with 20% at T,l-L, and 3% at Tlo-,, (Table 3).

Of the 58 histological cases, there were 35 individuals who exhibited mortice joints; 8 of these possessing two TLMJs (Table 1). These were commonly Type I TLMJs (56%) associated with an abrupt transition. The Type I mortice joints showed variable enclosure of the IAPs dependent upon the medial development of the MP (Figure 5). In some cases there was histological evidence that the MPs formed an extension to the articular surface of the SAPS (Figures 3a, b and Figure 5). Examples of mortice joints are demonstrated in the photomicrographs (Figures 3a-d, 4 and 5).

Ffgure 3. Photomicrographs of 200 pm thick horizontal sections through four Tl,-,2 zygapophysial joints, demonstrating the variability of joint orientations frequently seen at this level and their associated mortice joint types. a represents a well developed Type I mortice at T, ,_Q in a 75- year-old female. The mammillary processes (MP) afford almost complete enclosure of the inferior articular processes (IAP). b is an example of a Type I mortice at T, 1_12 in a 82-year- old female. Histological evidence of articular cartliage is seen on the right mammillary process (MP) and corresponding surface of the adjacent inferior articular process (IAP). Partial ossification of the ligamentum flavum (LF) is noted. c is an example of a type Ill mortice joint from a 54-year-old male at T, 1_12 showing the paired mammillary processes behind the coronally orientated zygapophyses. d shows articular tropism at T, 1-12 in a 31 -year-old male. The right zygapophysial joint shows an isolated mammillary process (MP) (Type II mortice joint) which forms a posterior reinforcement to the joint (+ = posterior joint capsule). SAP = superior articular process, IAP = inferior articular process, MP = mammillary process, Lf = ligamentum flavum, AC = articular cartilage, R = rib head, Rt = right side, All scale markers = 5 mm.

Discussion

The French anthropologist Paul Topinard’ recognized that the 1APs of the 11th and 12th thoracic, and first lumbar vertebrae, were frequently recessed into a mortice-like enclosure, thereby increasing the congruency of the T-L junction zygapophyses. This observation was extended by Le Double”‘, Davis”.‘h and recently by Malmivaara et al.‘” (Table 3). The early descrip- ~~~~sY~~l~l~l~~lh suggested an almost universal occurrence of this anatomical feature. exclusive to the T-L junction. Although Davis” reported that a high proportion of vertebral columns examined by him possesed a mortice- type joint, pre-selection of his skeletal material may have resulted in over-estimation of the incidence (Table 3). In a recent patho-anatomical study of the T-L junction involving 24 autopsy spines, Malmivaara et al. I3 noted 17 cases of mortice joints and drew attention to the frequency of unilateral mortice joints (65%) in this small series. Consistent with other reports (Table 3). they noted a predominance of mortice joints at Tll_17.

In the present study, 297 (47%) of the 630 CT cases demonstrated mortice joints. This relatively low incidence may reflect slight variation in the scan planes as TLMJs would be optimally identified from scans through the caudal aspect of the zygapophysial joints. In particular, Type III mortice joints may have been overlooked if the scan plane was slightly cranial to the end-plate, as the T,, MPs are short compared to T,,“. Differences in scan factors between the various CT investigations, notably slice thickness and field of view, may have also contributed to these difficulties due to volume averaging effects”. Image quality might also have led to differences in grading mortice joints from CT compared with the histological cases. As no sagittal reconstructions were available with the CT series or the

140 C/in. Biomech. 1989; 4: No 3

Figure 4. An example of an abrupt transition between Tll_,2 and T12-L, from a 63-year-old male. A mammillary process (MP) atT11_,2 (a) corresponds with a unilateral Type III mortice joint. AType I mortice joint is evident atthe level of the superior end-plate at TIP-L, (b) demonstrating the congruent articular surfaces. Photomicrographs of the same joints, orientated as for their CT scans, are depicted to detail the condition of the articular surfaces. LF = ligamentum flavum, IAP = inferior articular process, SAP = superior articular process, Scale markers = 5 mm.

Table 1. Frequency of mortice joint types seen in 630 CT investigations of the thoracolumbar junction and in a series of 58 cadavers

Sample

CT cases* (n = 297 from 630 cases)

Cadaveric casest (n = 35 from 58 cases)

Bilateral Unilateral

Bilateral Unilateral

I

47 22

11 13

Mortice type II

142 IO

6 5

111

49 19

2 4

Combined/tropism

37 Grand total =

2 Grand total =

Total

238 88

326

19 24 43

*Twenty-nine CT cases showed two mottice joints giving a total of 326 TLMJs tEight individuals in the cadaveric series showed two mortice joints at adjacent levels to derive a total of 43 TLMJs

Table 2. Relationship of mortice joint types to abrupt and cadaveric sample, the relative height of the MP to SAP gradual patterns of zygapophysial joint transition at the could not be routinely ascertained. However, this fea- thoracolumbar junction recorded from 297 CT investigations* turc was assessed indirectly from the serial histological

Transition type Mortice type Combined/

sections. and with para-sagittal scans of cadavers.

I II 111 tropism Total A similar incidence to the present histological study was reported by Malmivaara ct al.‘?, from their investi-

Abrupt transitiont 54 17 40 Gradual transition* 15 135 28

Totals 69 152 66

37

37

111 215

326

*Twenty-nine individuals showed two mortice joints at adjacent levels to derive a total of 326 TLMJs tAbrupt transition = change from coronal to sagittal orientation over two joint levels *Gradual transition = change from coronal to sagittal orientation over three joint levels

gation of 24 cadavcric T-L junctions (Table 3). In part, this may reflect greater recognition of mortice joint characteristics from dry vertebrae or thin histological sections, compared to single scans variously located about the superior vertebral end-plate.

The three mortice types, designated groups I, II and III by Davis”, were similar in description to three characteristic T-L junction zygapophysial joint orientations previously reported’ (see Figure 1). tn general,

Singer: Thoracolumbar mortiCe joint 141

Figure 5. Multiple 200 pm thick sections cut in the plane of the superior end-plate in a T,l_12 joint from a 75-year-old female, to illustrate the progressive development of the postero-medially directed mammillaty processes (MP). The middle section (C) corresponds to the superior end-plate level. Arrows depict the cranial appearance of MPs (6).

Types I and III were identified with an abrupt transition, whereas Type II mortice joints corresponded with a gradual transition form (Table 2). The 29 individuals possessing two mortice joints at adjacent levels usually demonstrated a Type I mortice joint at T,I--LI and a Type III at Tl,_iz.

The extent to which the IAPs were enclosed by the MPs varied considerably (Figures 2, 4 and 5). Histo- logically, the MPs often formed an extension to the articular surfaces of the SAPS (Figure 3b), particularly

in Type I mortice joints where the MPs projected medially to enclose the IAPs (Figures 3a, 5) and occasionally in Type II TLMJs. Several combined forms of mortice joint were demonstrated in which zygapophysial joint tropism was a common associated characteristic (Figures 2c and 3d). In the majority of these cases, the more coronally displaced zygapophysis had a prominent MP situated strategically behind it, which provided a secure anchor for the posterior and lateral aspects of the joint capsule (Figure 3d). In this way, the joint pair may behave functionally as a symmetrical unit. Further, the multifidus muscle arising from the MP would reinforce the posterior joint capsule and the MP itself may act as a bony check against dorsal displacement of the coronally orientated joint.

In an anatomical study of the lower thoracic vertebrae in 21 skeletons, Kaplan” documented the height of the MPs in relation to the adjoining SAPS. He noted that the MPs were most prominent at Tr?, accounting for almost 50% of the height of the SAPS, but were less developed at T,, where they formed only 14% of SAP height. In contrast, the MPs at L, were continuous with the lateral aspect of the SAP. These trends were sup- ported with present observations from para-sagittal scans of the T-L zygapophysial joints (Figure 6). Kap- lan” regarded the large Tlz MPs as a useful distinguish- ing landmark to indicate this level during spinal surgery and he recommended excision of the MPs to achieve a successful exposure of the zygapophysial joints prior to fusion of the T-L junction posterior elements.

The biomechanical influence of zygapophysial joints in guiding and limiting normal ranges of spinal segmental motion has been postulated in the classical anatomical literature’.2*‘X and confirmed in contempor- ary experimental studies’-‘.“. At the T-L transitional junction, increased stiffness under axial compression, coupled with zygapophysial joints orientated toward the sagittal plane, tends to result in limited rotation capabil- ity of these jointss-7. The degree to which the various types of mortice joint might further restrict segmental motion, particularly rotation, would depend upon the depth of the recess, disposition of the MPs, spinal post- ure, and the extent of axial compression loading. On theoretical grounds, a Type I mortice would tend to restrict most movements by affording an optimal ‘close packed’ position, produced by the vertical incline of the SAPS and the enveloping MPs (Figures 3a, b, 4, 5 and 7). Even in T-L transitional joints without demonstrable mortices, the medial taper of the SAPS (Figure 8) and predominantly sagittal orientation of the zygapophysial joints may be sufficient to resist torsional shear. This speculation would hold mainly for orthograde postures or positions incorporating spinal extension. Flexed postures would tend to produce distraction of the zygapophyses with a resultant ‘unlocking’ of the joint12. These arguments may also explain, at least in part, the mechanism of torsional injuries at the T-L junction, in that an abrupt transition, with a mortice joint configuration, may act to localize the level of injury. Specific biomechanical studies of axial rotation and careful

142 C/in. Biomech. 1989; 4: No 3

Figure 6. Para-sagittal CT section through the thoracolumbar junction zygapophysial joints in a 68-year-old male cadaver. Arrows illustrate the relative heights of the mammillary processes (MP) at T,,, T,2 and L, compared to their adjacent superior articular processes (S). During spinal extension at T,,_,2, the inferior articular processes (IAP) would abut onto the lamina (L) between the SAP and MP (see Figure 8). Dotted lines indicates the typical scan plane selected for imaging these posterior elements.

Figure 7. Schematic illustration of a Type I mottice to emphasize the medial taper effect of the superior articular processes (SAP) and the posterior enclosure formed by the mammillary processes (MP).

examination of radiological records of spinal injured patients, noting the type and level of transition, would be necessary to elucidate the influence of the TLMJ on vertebral motion and level of trauma.

In conclusion, a mortice-like configuration of the zygapophysial joint is frequently demonstrated on mid- joint axial scans at the T11_12 transitional level. The three principal mortice types reflect typical coronal, sagittal and intermediate zygapophysial joint orientations recorded at the thoracolumbar transition. Uni- lateral or combined forms may also be encountered. In cases of tropism, a unilateral mortice frequently associated with the coronally orientated zygapophysis may serve to protect the joint from torsional stress. The mortice articulation putatively limits axial rotation during normal activities but may also contribute to localization of injuries at the thoracolumbar junction.

Acknowledgements

These studies stem from collaboration with Dr Peter D Breidahl, MD, FRCR, FRACR, FRACP and Mr Robert E Day, BE. I record my thanks to Dr Lynton GF Giles, DC.PhD, for advice on the histological methods used in this study.

Table 3. Frequency of mortice joints according to spinal level recorded by various investigators

Segmental level Author Sample size (n = ) TVJ-7, T,,-,z Tw-L,

Topinardg 66 51 21 Le Double” 400 z 297 21 Davis”’ 67 5 46 16 DavisI 75 2 42 18 Malmivaaraet al.13 24 0 13 4

Present study: CT seriest 630 10 250 66 Cadavers 58 0 26 17

‘Davis selected 69 skeletons out of 149 for this survey tTotal cases = 297 from sample of 630 CTs. From the 297 cases, 326 TLMJs were demonstrated

Sample incidence

100% 100% 97% 92% 71%

47% 74%

Singer: Thoracolumbar mortice joint 143

References

Macalister A. A textbook of human anatomy. London: Griffin, 1889, 129 Levine A, Edwards CC. Lumbar spine trauma. In: Camins M, O’Leary P, eds. The lumbar spine. New York: Raven Press, 1987; 183-212 Holdsworth FW. Review article: Fractures, dislocations and fracture-dislocations of the spine. J Bone Jt Surg 1970; 52A: 1534-51 Rehn J. Die knochernen verlezungenderwibelsaule. Bedeutung des erstbefundes fur die spatere begutachtung. Die Wirbelslule in Forschung und Praxis 1968; 40: 131-8 White AA. Analysis of the mechanics of the thoracic spine in man. Acta Orthop Stand 1969; Suppl 127: 7-105

6 Kazarian L. Dynamic response characteristics of the human vertebral column. An experimental study on human autopsy specimens. Acta Orthop Stand 1972; Suppl 146: 7-165

7 Markolf KL. Deformation of the thoracolumbar intervertebral joints in response to external loads. J Bone Jt Surg 1972; S4A: 511-33

8 Singer KP, Breidahl PD, Day RE. Posterior element variation at the thoracolumbar transition: a morphometric study using computed tomography. Clin Biomech 1989; 4: 80-6

9 Tooinard P. Des anomalies de nombre de la colonne

10

11

12

13

I4

15

16

17

I8

19

veitebrale chez I’homme. Revue D’Anthropologie 1877; 6: 577-649 Le Double A-F. Traite des variations de la colonne vertebrale de l’homme. Paris: Vigot Freres, 1912:243 Kaplan EB. The surgical and anatomic significance of the mammillary tubercle of the last thoracic vertebrae. Surgery 1945; 17: 78-92 Davis PR. The thoraco-lumbar mortice joint. J Anat 1955; 89: 370-7 Malmivaara A. Videman T, Kuosma E. Troup JDG. Facet joint orientation, facet and costovertebral joint osteoarthrosis, disc degeneration, vertebral body osteophytosis. and Schmorl’s nodes in the thoracolumbar junctional region of cadaveric spines. Spine 1987; 13: 458-63 Terry RJ, Trotter M. Section III: Osteology. In: Parsons Schaeffer J, ed. Morris’ human anatomy. I Ith ed. New York: McGraw-Hill, 1953: 102. 196 Giles LGF, Taylor JR. Histological processing of large vertebral specimens. Stain Technol 1983; 58: 4.5-9 Davis PR. The thoraco-lumbar mortice joint in West Africans. J Anat 1961; 95: 589-93 Wegener OH. Whole body computerized tomography. Basel: Karger, 1983: l-50 Winslow JB. Exposition anatomique de la structure du corps humain. Paris: Deprez, 1732 [cited by White?] Stokes IAF. Mechanical function of the facet joints in the lumbar spine. Clin Biomech 1988; 3: 101-S

Figure 8. A frontal plane CT scan of the thoracolumbar junction demonstrating the tapered nature of the zygapophysial joints below T,, . A Type II mortice joint is formed at Tll-12 where the inferior articular processes (iap) abut down onto the base of the mammillary processes (mp) (compare with Figure 6).

thoracolumbar mortice joint: radiological and histological observations

Documents