Lumbar spine was the most frequently affected region, followed by thoracic spine. Saffudin reviewed 78 English reports on spinal OO and osteoblastoma, and found 163 OOs at lumbar spine and 80 at thoracic spine [5]. We got similar results, and the ratio of lumbar spine to thoracic spine approximated 2:1. Most OO presented as isolated lesion below T8, and only one case of OO occurred at T6 in our cohort. OO distributed evenly at thoracic region, while been more likely seen at L3 and L4 for lumbar region. However, no matter where they occur, their curve patterns are similar. It demonstrated that pain-evoked muscle spasm had an effect on the whole spine no matter where it generates. Most OOs occur in the posterior element of the spine, including lamina, facet and pedicle. OOs occurring at the lamina tend to be unilateral and surround the facet. Similarly, no matter where OOs occur, curve pattern of scoliosis is similar.
Scoliosis secondary to OO was always misdiagnosed as other spinal diseases such as idiopathic scoliosis and scoliosis associated with lumbar disc herniation [11]. Scoliosis secondary to OO and scoliosis associated with lumbar scoliosis were both pain-related. The pain of OO is localized and may be aggravated with motion. It is more severe at night and relieved by non-steroidal anti-inflammatory drugs (NSAIDs). Night pain was the characteristic manifestation of scoliosis secondary to OO, which played a great role in the diagnosis [12]. The etiopathogenesis of scoliosis associated with lumbar disc herniation was different. Nerve irritation and decrease of the weight-bearing capacity of healthy lower limbs were thought to contribute to scoliosis associated with disc herniation [13]. Therefore, the majority of the pain improved due to lying in bed, and there was no nocturnal pain. Another important differential diagnosis point was the relationship between location of pain and direction of curve. Pain of OO was always located at the concave side of scoliosis, while for lumbar disc herniation, pain could be located at both sides of scoliosis [14]. Sagittal alignment could also help differentiate scoliosis secondary to OO from scoliosis associated with lumbar disc herniation. Patients with scoliosis associated with lumbar disc herniation frequently display decreased lumbar lordosis and thoracic kyphosis, which are believed a pain-relief posture [15]. However, for scoliosis secondary to OO, we did not note this phenomenon. Patients presented similar sagittal alignment to normal population, and there were no significant difference of sagittal parameters before and after surgery.
Nonsteroid anti-inflammatory drugs (NSAIDs) were recommended for pain control. However, it was not effective for scoliosis. Curettage was thought to be the standard surgical treatment for OO, which could obtain radical resection of nidus, thus halting the muscle spasm [16]. In our cohort, more than 70% patients got significant correction of scoliosis within 3 months. At last follow-up, all the patients obtained satisfactory outcomes with regards to both pain relief and scoliosis correction. The use of internal fixation mainly depends on the stability of the spine after curettage. In our cohort, only one patient received curettage without fixation, as lesion was located at midline of laminae without involving facet. The majority of the cases of out study had lesions at facet or pedicle, and curettage would inevitably damage the facet generating iatrogenic instability of the spine. Additional posterior fixation could permit long-term stability of the spine, especially for adolescents with great growth potential. One case had a lesion at vertebral body, and we performed curettage and fixation by anterior approach. Minimally invasive methods, such as CT-guided thermocoagulation and percutaneous radiofrequency ablation have also been adopted for the treatment of OO [17]. However, its effect on scoliosis remains to be observed. As a benign spinal tumor, the recurrence of OO is very low. In this study, no tumor recurrence was found at the last follow-up. Complete resection of the nidus can effectively avoid tumor recurrence.
In conclusion, patients with spinal OO had a significantly high incidence of scoliosis. Patients could get rapid relief of pain and scoliosis with low occurrence. Night pain, pain at the concave side of curve, normal sagittal alignment could help differentiate it from scoliosis associated with lumbar disc herniation.