Osteoporosis

Osteoporosis degrades the ability of instrumentation to engage the bones of the spine. The more osteoporotic the spine, the more likely hardware will fail to provide internal stabilization until the goal of bony fusion is achieved. The diagnosis of osteoporosis can be made by dual-energy X-ray absorptiometry (DXA) or quantitative computed tomography (CT). The most common method to test bone density is by utilization of the DXA scan, which provides a two-dimensional method to quantify bone density. It is accurate, inexpensive, safe, and normalizes the results to a distribution curve based on age and sex. A T score and a Z score is generated. The former normalizes one’s score to a distribution curve of normal individuals of the same sex who are 30 years of age. A value between -1.0 to -2.5 indicates the presence of osteopenia, and scores below -2.5 are indicative of osteoporosis. The Z score normalizes an individual’s score against the distribution curve of a group of individuals of the same age and sex. The Z score is not as useful in the elderly osteoporotic population. Though there is no absolute contraindication to instrumentation of the thoracolumbar spine with pedicle screws, a DXA score less than -2.5 is a relative contraindication to utilization of pedicle screws, especially for short segment fixation in the absence of anterior support.

Quantitative CT imaging can also be used to determine osteoporosis. Weiser et al designed a study to test fatigue strength in the setting of osteoporosis based on the quantitative CT diagnosis of osteoporosis. Twenty-one donors aged 12 to 96 had their T12 vertebrae evaluated with a quantitative CT protocol. The osteoporotic group had bone densities below 80 mg/cm3, and a normal group consisted of those with bone densities above 120 mg/cm3. Their vertebrae were then tested in a mechanical hydraulic testing machine with a single pedicle screw within the T12 vertebral body. The osteoporotic group achieved only 45% of the cycles to failure with only 60% of the fatigue load compared to the normal bone density group. ¹

Though this study potentially predicts a worse outcome with utilizing instrumentation in the setting of osteoporosis, this is not an absolute contraindication. Having the information enables the surgeon to alter his or her intraoperative strategy to decrease the risk of hardware failure.

Kyphosis

Kyphosis occurs with aging. Mean thoracic kyphosis is less than 30°in those under 30 years of age, and increases to 66°in the 75-year-old age group. ² The pathogenesis of degenerative kyphosis is multifactorial and caused by either asymmetric disc collapse or vertebral body collapse. Hyperkyphosis is associated with overall impairment in physical function, worsening pulmonary function, worsening of gait, increasing falls, chronic pain, and fractures of all kinds. Pulmonary dysfunction associated with thoracic kyphosis greater than 40°results in increased mortality in the older population. In the elderly female population, the combination of hyperkyphosis and vertebral body fracture predicts a greater mortality compared to either condition alone. ³ In the aging population, worsening kyphosis results in weakening of the extensor musculature, further imbalance, and greater degeneration of the anterior discs. The sequela of worsening kyphosis results in an even greater degree of kyphotic deformity over time.

Thoracic kyphosis greater than 40°(T5 to T12) is associated with increased risks of proximal junctional kyphosis following spinal fusions ⁴ and is thus a relative contraindication to spinal instrumentation. The biomechanical forces on a screw–rod construct in the kyphotic spine are different than the nonkyphotic spine. Force vectors are often aligned with the pedicle screws when kyphotic individuals stand, thus creating less resistance to screw pull-out.

Cervical Spine

In the cervical spine there are advantages to using longer screws in those individuals with osteoporosis, and in cases spanning three or more motion segments. In the past when performing anterior cervical reconstructions with the early, non-locking plates, bicortical fixation was necessary. Fluoroscopy was utilized to verify bicortical purchase. Additionally, measurements on a sagittal reconstructed CT scan can determine an ideal screw length so as to be able to engage the posterior cortex (Fig. 21‑2). This becomes an important part of one’s armamentarium when considering long constructs (three or more motion segments), and when performing multilevel corpectomies, especially in the osteoporotic individual. Biomechanical studies have shown that C1 screw fixation is optimized by engaging both cortices (Fig. 21‑3). Care must be taken to avoid extending the screw tip beyond 1 to 2 mm from the distal cortex to avoid injury to retropharyngeal structures. The anterior portion of the C2 vertebral body contains weaker cancellous bone. Therefore bicortical fixation should be considered at the cephalad end of the construct when performing an anterior cervical reconstruction that will include the C2 vertebral body (Fig. 21‑4).

Thoracic Spine

In the thoracic spine, safe havens for bicortical purchase exist. Ventral to the vertebral bodies of T1 and T4, no critical structures exist that create a significant danger for bicortical fixation. Structures anterior to the upper thoracic spine include the esophagus, segmental vessels, the mediastinal structures, the sympathetic chain, the phrenic and vagus nerves. The trachea lies much more anteriorly within the mediastinum, and the aortic arch is positioned closer to the T4 level. The aforementioned structures are difficult to injure with bicortical fixation utilizing the classically-described entry points and trajectories for thoracic pedicle screw placement (Fig. 21‑5). When considering bicortical fixation lower in the thoracic spine, the position of the aorta on the left side of the vertebra must be considered. A left-sided vertebral body screw with an entry point near the costovertebral joint and a trajectory which crosses the vertebral body provides a method for bicortical fixation at any segment of the thoracic spine (Fig. 21‑6).

The thoracic extrapedicular technique is an excellent example of multicortical screw fixation in the thoracic spine. By choosing an entry point within the pedicle, at the level of the costotransverse joint, one can pierce the pedicle laterally, and then regain entry through the costovertebral joint, and gain access into the vertebral body. This technique, originally described by Dvorak et al, provides sound biomechanical fixation of the thoracic spine (Fig. 21‑7). ⁵ Pullout testing between traditional pedicle screws and the extrapedicular technique reveals no significant differences, though longer screws are able to be utilized with the extrapedicular technique. ⁶

Lumbar Spine

The lower lumbar spine contains wider pedicles, allowing more variation in trajectory. The screw trajectory can either be maintained in the sagittal plane or begin more laterally, thus applying triangulation. The cortical bone can be engaged in the pedicle, along the superior endplate and through bicortical purchase of the anterior vertebral body. Determining the optimal size and trajectory of the screw is accomplished by preoperative planning to select the optimal diameter and length for pedicle screws. Measuring the pedicle diameter and selecting the appropriate screw diameter can provide optimal engagement of the pedicle screw threads with cortical bone. Care must be taken not to oversize the screw, as this risks pedicle fracture. The long axis of the screw should be aimed in close approximation to the rostral endplate of the vertebral body (Fig. 21‑8). This will engage the long axis of the screw to the cortex of the superior endplate to optimize cortical bone engagement.

Design changes to pedicle screws that aid to prevent pullout include changes to the thread shape (v-shaped as opposed to rectangular shape), as well as the thread depth. Dual lead-threaded screws have been reported to help in settings of osteoporosis. However, biomechanical studies reveal that the optimal way to make a significant difference on screw pullout is to use a screw that abuts the internal cortex of the pedicle. ⁷ The thread pattern does not seem to matter.

A variation of lumbar pedicle screw placement, called the cortical trajectory, provides another option for lumbar fixation. This technique involves beginning the screw at the inferomedial portion of the pedicle, and applying a trajectory towards the superolateral portion of the pedicle (Fig. 21‑9). This screw trajectory engages more of the cortical bone as the screw passes obliquely through the pedicle. At the pedicle level, a safe haven exists just behind the transverse process for bicortical purchase. Further posteriorly along L5, one has to be careful to avoid the femoral nerve and the traversing L4 nerve root. Similarly, anterolateral to the sacrum, one has to be careful to avoid the L5 nerve root.

A tricortical fixation technique has been described for fixating the sacrum, which proves useful when performing multilevel fixation. Aiming the sacral screw through the sacral promontory, provides better fixation, especially in severely degenerated cases. ⁸ Fixation of the sacrum through the more densely calcified superior articular process (as opposed to an entry point more lateral) tends to provide better fixation of the sacrum.

Finally, an underutilized region to engage cortical bone is the anterior portion of the inferior vertebral body in the setting of spondylolisthesis. For instance, in the setting of an L4–5 degenerative spondylolisthesis, the L5 vertebral body will often have a separation with the iliac vessels of 5 mm or more (Fig. 21‑10). Engaging the anterior cortex in this setting is safe.

The preoperative CT scans can be used to plan appropriate trajectories and screw lengths. The axial view provides the best method to measure for an appropriate screw length. By measuring the distance between the transverse process/superior articular process border and the midline/ventral portion of the vertebral body, by adding 5 mm to that distance, one is able to obtain a very accurate screw length that can provide bicortical purchase. The potential for error or complications varies by the vertebral body level. For instance, when placing bicortical ventral vertebral body screws (ventral fixation) from the left side of the spine at the L2 level there are rarely critical structures to injure, even if one is 5 mm longer than intended. On the other hand, a mid-cervical vertebral body screw should not pierce the posterior cortex by more than 1 or 2 mm. Practice can be obtained either by the use of cadavers, or by attempting to predict with unicortical screws, the resulting correspondence between the intraoperative radiograph and the predicted depth of the screw. It is very important to consider that the manufacturer of the instrumentation is an important variable when one intends to utilize the preoperative CT scan measuring technique. Some manufacturers only utilize the screw shaft length, where others utilize the head/shaft length to determine screw length.

Technology is playing a larger role in the ability to precisely and accurately place pedicle screws. Three-dimensional (3-D) computerized navigation technology can be utilized to obtain exact trajectories for screw placement. Other advantages include the ability to precisely determine screw length so as to be able to apply a bicortical technique, or the ability to place a screw in a specific region of the pedicle for optimal screw purchase. Importantly, soft tissue structures can be visualized and avoided with 3-D navigation. Disadvantages include cost, patient radiation exposure, as well as additional time requirements.

An ineffective technique, hubbing, refers to the placement of the entire screw, including the head, into the bone (Fig. 21‑11). Hubbing results in weakening of the screw-bone interface by damaging the bone as it engages the screw threads. This results in a loss of up to 40% of pullout strength. ⁹