A number of previous studies have reported perioperative and postoperative complication rates in adult spinal deformity surgery of up to or more than 40%
. One major complication is a compression fracture of the last instrumented and/or the supra-adjacent vertebral body
, this requires reintervention and extension of spinal fusion.
Augmentation of pedicle screws with PMMA or calcium phosphate cement has been shown to improve the initial fixation and fatigue strength of instrumentation in osteoporotic vertebrae, and it also decreases the likelihood of compression fracture at the same level
. Additional vertebroplasty of the neighboring 1 to 2 uninstrumented levels has been applied attempting to avoid the junctional kyphosis and compression fractures of cranial vertebrae.
As mentioned before, Watanabe et al.
 described two patterns of proximal vertebral fracture following spinal deformity surgery in adults receiving segmental pedicle screw instrumentation: upper instrumented vertebral collapse + adjacent vertebral subluxation and supra-adjacent vertebral fracture alone. In an attempt to avoid these complications in our patient, we augmented the upper instrumented vertebra and its adjacent level, but this created excessive distractive stress in the posterior element, resulting in a compression fracture of the upper instrumented vertebral body (even after being augmented with cement) with distraction of posterior elements and pedicles through the upper edge of the screw and subluxation of the adjacent level.
We hypothesize that vertebral augmentation could increase the risk of this complicated fracture. There are several factors that might lead to this complication:
Incomplete cement filling of the vertebra can create an area of weakness between two cemented areas (from the edge of the upper screws and the cranially-adjacent cemented vertebra) that behaves like an osteoporotic vertebra between two cemented vertebras.
The insertion of the pedicular screw leads to pedicle weakness.
Ending the instrumentation construct just below the apex of the postoperative thoracic kyphosis is too-frequently associated with junctional kyphosis/fractures
The case presented here is concerning. The juxtaposition of these factors, partial vertebral cement filling, a positive sagittal balance and structurally-weakened pedicles may set the scene for a very early posterior element distraction failure combined with a compression fracture in the uncemented area.
Ending the construct too low, just below the apex of the kyphosis, can predispose to junctional kyphosis and compression fracture at the cranial level. According to the preoperative radiographs, the upper instrumented vertebra (T10) was located several levels below the apex (T8). In an attempt to improve the preoperative sagittal imbalance we surgically increased the lumbar lordosis (from 45ª preoperatively to 62ª in the postoperative x-ray), this may have led to a new sagittal profile and a greater kyphosis, moving the level of the apex. Attention should be paid to this point, because excessive bending of the rod in the lumbar spine can facilitate the migration of the thoracic kyphosis apex.
We think that the augmentation of the upper instrumented vertebra and the supra-adjacent level, in an attempt to avoid the complications described by Watanabe et al.
, does not solve the problem and may lead to further and more severe fractures, like the one described in this paper, although this opinion requires biomechanical and/or clinical investigation for confirmation.
Our service has used this technique successfully in more than 15 patients in the last two years. To date, this is the only occurrence of this complication. When successful, vertebral augmentation of the cranial level in a long thoracolumbar fusion solves the topping off syndrome and avoids major disabilities.
This is a single case report and does not completely explain high incidence of complications in the junctional level after surgery for adult spinal deformity. Our purpose is to alert the orthopedic community to this phenomenon.
We would advise that any patient with a long thoracolumbar fusion with cranial level augmentation be closely monitored.