Spinal metastases are frequently observed complications in the malignant disease at terminal stage . At least 30 to 90% of patients who died of cancer had spinal metastasis in cadaveric studies [3, 8, 38]. The average age of our series was 63.6 years. Indeed, spinal metastasis has a high incidence in people aged 40 to 65 years, that match with the period where the risk of cancer is at the highest . Men are most susceptible to develop spinal metastasis, thus probably reflecting the men’s slight higher prevalence of developing spinal metastasis of lung cancer and of prostate cancer against symptomatic breast cancer . This man predominance is considerably observed in our series (39 men/13 women).
The NSCLC of adenocarcinoma type was the most common in our series (69.2%). Histological type varies according to the studies. Sun & Al , in 2011, reported an incidence of 23% of osseous metastasis in a retrospective series of 1166 NSCLC.
Predominance of dorsal localization (30 %) is also reported by Patricia & Al in their series (49 %) . In 36.5% of cases, it was 1 or 2 metastasis and in 63.5% of cases, more than 2 metastases. Metastasis was synchronous in 82.7% of cases and metachronous in 17.3 % of cases. Sioutos and Col  noticed that ambulatory patients in preoperative, and whose metastasis only concerned one vertebra, had statistically survived more than non-ambulatory patients whose metastasis concerned several vertebral levels.
Depending on the extent and the localization, they cause variable symptoms from pain to neurological deficit (complete or partial) . Those osseous spinal metastases occur most often at the posterior part of vertebral body . When they extend to the epidural space and are compressing the spinal cord, they are presented as a medullar compression of neoplastic origin . In our series, pain was the main symptom associated in 59.6% to a medullar compression presentation. The preoperative SINS (spinal instability neoplastic score) was evaluated in all patients as previously described . A score under 7 means stability, between 7 and 12 corresponds to an indeterminate or upcoming instability and a score from 13 to 18 suggests instability. This score can be used as a guide for making the surgical decision, because unstable lesions are most susceptible to require stabilization for a long-term management of metastatic diseases of the vertebral column .
In our series 41 patients (78.9%) had a score equal or above 7 and 11 patients (21.1%) under 7.
Extra spinal metastasis was present in 57.6 % of cases in our series. Tumorous extension was evaluated by TNM classification, the stage 4 was the most common with 88.6% of cases. Functional consequences were appreciated by Karnofsky Index, which has an important part in our decision making. In our series, it was good in 28 % of cases (80–100), moderate in 67.4% (50–70) and poor in 3.8% (10–40%). It was equal or above 70% in 33 patients and under 70% in 9 patients for Patricia and Al’s series .
Survival was evaluated by Tokuhashi Index, some authors covered the topics of survival after a surgery for spinal metastasis [35, 36]. Indeed, Tokuhashi and Col  have studied the results of 64 patients who underwent surgery for spinal metastasis. They included all primary diagnosis and surgery performed for diverse indications including pain and paralysis. Based on this, they formulate specific parameters including general condition, number of osseous extra spinal metastasis, vertebral metastasis number, presence of lesions in other internal organs, primary lesion localization, spinal cord lesion severity.
A score of 9 or above was predictive of at least 12 months of survival, a score of 8 or weaker indicate survival of less than 12 months and a score of 5 and under predicted 3 month or less of survival. Based on those data, it appears that the Tokuhashi system can be a precious tool in preoperative discussions and the decision making; its values were confirmed by other authors. Enkaoua, & al.  found it predictive of survival after surgery for most of patients presenting spinal metastasis. Although Tomita's scale  should not be overlooked, it still can be used to predict survival (Table 4). In our series, 48 patients had survival prediction of less than 6 months (Tokuhashi 0–8) and 4 patients a prediction of 6 to 12 months (9–11). There was no patient with a survival prediction above 12 months; as a matter of fact, the nature of the vertebral ailment which is secondary to a lung cancer is a bad prognosis element [13, 34, 35]. Thus, the Van der Linden team  found a median survival of 2.9 months for the patients suffering from spinal metastasis of lung cancer unlike our series where we found a median survival of 16 months. In our series, Tokuhashi's group C (12–15) and Tomita's group A (2–3), which correspond to good prognosis groups, were not found because of the high severity and mortality of bronchopulmonary cancers at the time of diagnosis. This partly proves the inadequacy and relativity of the classic prognosis scores (Tokuhashi and Tomita) in PBCs. The non-representation of good prognosis groups in PBC should lead us to set up a specific prognostic classification of PBC spinal metastases considering all groups of patients.
Surgery was essentially palliative in our series; the main objective was to relieve pain and restore the stability. The palliative surgery allowed pain control, vertebral column stabilization, dural sheath release in case of medullar or root compression . There is a lot of technics and surgical approaches which can be chosen depending on the area to treat, the medullar compression severity, the risk of instability and the patient general condition . Vertebroplasty is indicated for spinal mechanical pain before or after surgery and provides a rapid relief in less than 12 hours . The other consolidation technics are kyphoplasty (Fig. 2) and stentoplasty or percutaneous osteosynthesis when we doubt of the sufficiency of cementoplasty alone or with stent to stabilize the spine . Other surgical options are corpectomy associated to a reconstruction by vertebral prosthesis (Fig. 3), laminectomy and stabilization by osteosynthesis (Fig. 4) and vertebrectomy [10, 18]. The surgical treatment has allowed the regression of the painful symptomatology in 84.6% of patients of our series, and allowed to stabilize the spine for all patients.
The after-surgery pain relief rate for metastatic tumors was generally high. Weigel and al.  noticed a moderate relief of pain in 89 % of patients after they underwent decompressive anterior surgery and stabilization. Likewise, Rompe & al  and Hussein & al.  have documented a pain relief in almost 90 % of their cases, while Shimizu & al  Cahill and Kumar  noticed a pain relief of 100% after posterior approach surgery. The motor deficit was improved in 48.3% of our patients with a gain of at least one grade of Frankel. Similar rates of neurological improvement were reported by King and al.  with a rate of 50% of improvement of at least one grade of Frankel in 20 patients after an anterior decompression combined with posterior stabilization without laminectomy. Also, a meta-analysis published in 2005  resuming the results of 28 studies including 1542 patients with medullar metastases, found 1.3 times more important rates of ambulatory patients in case of surgery compared to radiotherapy alone (capacity to walk for 85% of operated patients versus 64% of irradiated patients)
In our series, postoperative complications were present in 28.8% of cases. They were essentially pulmonary complications. Panayiotis & al.  report that 17 % of postoperative complications were dominated by pulmonary complications in their series, covering 109 cases of solid cancer thoracic spine metastasis. Survival prediction above 12 months was null in our study and there were only 4 patients between 6 and 12 months. However, after surgery, 8 patients had a survival above 12 months and 8 others between 6 and 12 months. Factors which influenced this survival in our series were complementary treatments: RT/CT (p = 0.072), the histological type essentially the epidermoid carcinoma (p = 0.068) and preoperative Karnofsky 80–100% (p = 0.097). A possible explanation of this prediction can be the enhanced availability of adjuvant therapy. Indeed, new therapies have been introduced to improve patients affected by lung cancer survival [12, 29]. The median survival of patients with advanced lung cancer improved from 6 to 12 months with introduction of new treatment diagram combined with monoclonal antibodies, which proceed as angiogenesis inhibitor [28, 29]. Immunotherapy, which efficiency in improving the survival of bone metastases was first demonstrated by the retrospective analysis of Tamiya et al .