When cross unions of lumbar transverse processes are found, the etiology, whether congenital, traumatic or neoplastic, should be differentiated, The relationship of such findings to DISH, Psoriasis, Idiopathic Myositis Ossificans and to malformations and variants must also be evaluated.

Our present purpose is to examine data on the clinical and radiological characteristics of such cross unions and their frequency, in order to differentiate their origin. We include all bony formations occurring between two or more transverse processes and between one transverse process and one rib or Os Ileum.

Shore (36 ) was probably the first to describe an intertransverse osseous bridge in 1930. He considered the lesion to be congenital. Other reports ( 28, 31, 23, 9, 29, 16) supported the congenital theory.

Louyot (25 ) and others thought that bridging was due to trauma, but mentioned the possibility of congenital origin. It was assumed that the lumbar spine could sometimes develop as the sacrum does. The well known (symmetric or asymmetric) "Sacralization of L5" and the "Lumbalization of S1" are considered developmental variations. Using this slogan the intertransverse osseous bridge of unclear origin is accepted by several authors as a possible variation of sacralization (Cataliotti 4 , Dunoyer 6, Suto 39).

Keats’s widely read book Variants of Bones and Joints Simulating Diseases (15) describes 3 cases of lumbar osseous bridges. No mention is made of either suspected or proven traumatic origin, and the reader is to assume these cases to be examples of congenital variants.

Summarizing the existing literature, we found 59 documented cases of osseous bridges.

Only 4 times was a bilateral bridge formation found, and never in exactly the same segments (4, 25, 32, 42).

Of the remaining 55 cases 24 osseous bridges are right-sided, 31 left-sided.

In 15 of 59 cases more than 2 segments are involved.

Vertebrae involved are:

Th12, 2 times;

L1, 11 times;

L2, 27 times;

L3, 50 times;

L4, 48 times;

L5, 18 times;

Os ileum, 5 times.

In 15 of 59 cases more than 2 segments are involved.

24 cases were considered by the authors to be of congenital origin, and 29 of traumatic origin ( no opinion in 6 reports). In only 10 cases was the traumatic origin proven by radiographs at the time of the accident and follow-up examination. Of those 10 cases 6 had scoliosis convex to the pathologic side, 4 were uncharacteristic.

The clinical significance of osseous bridges is controversial. Lumbar pain is found in approximately half of the patients. Some patients underwent operations, with varying results.

There is no doubt that the osseous bridge acts as an arthrodesis of one or more vertebral segments. The loss of function stops the degeneration but must be compensated by adjoining segments; this probably hastens the degeneration at the upper and lower ends of the bridging (7).

Like transverse process fractures in acute trauma, osseous bridging in the post-traumatic stage is only one aspect of a complex traumatisation, including the soft tissue of the retroperitoneum, thorax and pelvis.

Billet (2) described 3 forms of osseous bridging:

4. The base of the transverse process remains visible; an irregular osseous bridge links the middle section of 2 transverse processes. In a transverse position of the lumbar spine this resembles an "H" or "h" so it can be called the "H- or h-Form". It is found after both violent and mild traumata.

5. It is no longer possible to differentiate the original transverse processes, which are assumed to have been knocked off at their source and merged into a new bony formation. This "Knocked-off-shape" is present only in cases of violent traumata.

6. A regular bridge, which originates from the assumed base of a transverse process; known as the "O-form" or "kissing interspinaux" and assumed to be of congenital origin. But it is found only once in own material and 9 times in other literature.

Since the previous paper (2), more observations have been made. We observed the natural history of transverse process fractures and their remarkable polymorphy which improved our understanding. Moreover, we obtained CT examinations with 3D-reconstructions; osseous bridges hat not previously been demonstrated with this method.

We believe cross unions of lumbar transverse processes to be an unsolved problem. Our propose is to describe the "natural history" of transverse processes fractures, to compare the proven traumatic with accidental found osseous bridges and to document the role of CT and 3D reconstruction.