Prediction of survival prognosis after surgery in patients with symptomatic metastatic spinal cord compression from non-small cell lung cancer

Background The aim of this study was to develop a scoring system for prediction of survival prognosis after surgery in patients with symptomatic metastatic spinal cord compression (MSCC) from non-small cell lung cancer (NSCLC). Methods We retrospectively analyzed nine preoperative characteristics for survival in a series of 64 patients with NSCLC who were operated with posterior decompression and spine stabilization for MSCC. Characteristics significantly associated with survival on multivariate analysis were included in the scoring system. The scoring point for each significant characteristic was derived from the hazard ratios on Cox proportional hazards model. The total score for each patient was obtained by adding the scoring points of all significant characteristics. Results Eastern Cooperative Oncology Group (ECOG) performance status, number of involved vertebrae, visceral metastases, and time developing motor deficits had significant impact on survival on multivariate analysis and were included in the scoring system. According to the prognostic scores, which ranged from 4 to 10 points, three prognostic groups were designed: 4–5 points (n = 22), 6–7 points (n = 23), and 8–10 points (n = 19). The corresponding 6-month survival rates were 95, 47 and 11 %, respectively (P < 0.0001). In addition, the functional outcome was worse in the group of patients with 8–10 points compared with other two prognostic groups. Conclusions The new scoring system will enable physicians to identify patient with MSCC from NSCLC who may be a candidate for decompression and spine stabilization, more radical surgery, or supportive care alone. Patients with scores of 4–5, who have the most favorable survival prognosis and functional outcome, can be treated with more radical surgery in order to realize better local control of disease and prevent the occurrence of local disease. Patients with scores of 6–7 points should be surgical candidates, because survival prognosis and functional outcome are acceptable after surgery, while patients with scores of 8–10 points, who have the shortest survival time and poorest functional outcome after surgery, appear to be best treated with radiotherapy or best supportive care.


Background
Metastatic spinal cord compression (MSCC) is a severe complication of cancer that occurs in 28 % of patients with lung cancer and can become symptomatic, which involves intractable pain, disability, and incontinence [1][2][3], negatively impacting the patient's quality of remaining life. The optimal treatments for patients with MSCC are analgesics, corticosteroids, chemotherapy, radiotherapy and surgery, and most often these treatments are combined to give the maximum palliative effect with a minimum of operative morbidity and mortality [1,4,5], positively improving the patient's quality of remaining life. Recently, an increasing number of studies supported the use of decompressive surgery as an effective treatment for MSCC due to the evolvement of surgical techniques [1,2,6], while only a few studies specifically addressed surgical treatment of MSCC in lung cancer [7,8], which was often associated with high morbidity and mortality [8]. A major problem in selection patients for surgery is to avoid operating on those who are likely to die very soon after surgery, so life expectancy is the most important selection criteria for surgery. While for patients with very short survival time radiotherapy or best supportive care alone are recommended, for patients with more favorable prognosis can be treated with decompressive surgery, or even more radical surgery such as excisional procedures [4,9,10].
Some scoring systems were designed to estimate the survival time of each patient and select the optimal treatment strategy among supportive care, palliative radiotherapy, palliative surgery, and excisional surgery [9][10][11][12][13][14][15]. However, some old and commonly-used scoring systems have underestimated the life expectancy of lung cancer patients with spinal metastases because of the increased survival time for this patient group in recent years [16][17][18][19]. Notably, it is critical to regard patients with MSCC from a particular primary tumor type as a separate group of patients for individual treatment, because primary tumors vary with respect to their biological behavior. Therefore, our present study is designed to develop a new survival score particularly for patients with MSCC from non-small cell lung cancer (NSCLC) after surgery.

Patients
Sixty-four patients with NSCLC operated with decompression and spine stabilization for MSCC were retrospectively analyzed in the study at the Affiliated Hospital of Academy of Military Medical Sciences, Beijing, between May 2005 and May 2015. The diagnosis of bone metastasis in NSCLC patients was confirmed histologically, adequate diagnostic imaging including spinal CT or MRI, as well as bone scan.
Patients with an estimated survival less than 3 months or health too poor to undergo surgery were excluded. Of the total series of 64 patients, six patients were treated with radical resection of primary lung cancer, while others weren't. The data were collected from patients, their family members, treating surgeons, and patients' files. The Medical Research Ethics Board of the Affiliated Hospital of Academy of Military Medical Sciences approved this retrospective study and required neither patient approval nor informed consent for review of patients' images and medical records. The data were retrospective in nature and anonymized by the Medical Research Ethics Board. The postoperative survival was defined as the time between the date of surgery and death or the latest followup. For the present study, we included all 64 patients with NSCLC who had decompressive surgery and spine stabilization due to spinal cord compression. None of the patients were excluded for any reason. 5 patients were still alive by the end of the study period, with a mean follow-up of 9.7 months in those patients. In patients who had surgery for more than one metastasis, all

Surgery and functional evaluation
The indication for surgery was neurological deficit due to spinal cord compression. All patients were operated with posterior decompression and stabilization in our department. Local radiotherapy, systemic chemotherapy, and targeted therapy with gefitinib were performed after the wound healed, about 3-4 weeks after the surgery. Postoperative functional outcome was analyzed according to the scoring system. Neurological function was graded based on Frankel et al. [20] preoperatively and 4 weeks postoperatively (Patients with Frankel D and E have the ability to walk). Time developing motor deficits was defined as the time between deterioration of motor function to disability or surgery. Deterioration of motor function was defined as a change of at least one Frankel grade.

Statistical analysis
The univariate analysis of survival was performed using the Kaplan-Meier method and the log-rank test. The significant prognostic factors (P < 0.05) were additionally evaluated in a multivariate analysis performed with the Cox proportion hazards model (multiple Cox regression, selection = stepwise). The prognostic factors that were significant in the multivariate analysis were included in the scoring system. The prognostic factors that were excluded by Cox proportion hazards model (multiple Cox regression, selection = stepwise) were not included in the scoring system. The scoring point for each significant factors was derived from the hazard ratios on Cox proportional hazards model (simple Cox regression). The total prognostic score for each patient was determined by adding the scoring points of every significant factor. Neurological outcome in risk groups was compared with Chi-square test and Fisher exact test. A P value of 0.05 or less was considered statistically significant. Statistical analysis was performed using SAS 9.2 software.

Patient characteristics and survival
A total of 64 patients were included in the study, 34 % (22/ 64) of patients were female, and 66 % (42/64) were male. The overall median age was 57 years old. The median time interval from diagnosis to surgery was 80 days, and the median time developing motor deficits was 14 days.
For all patients, the overall median survival time was 6.3 months (95 % confidence interval, 4.5-7.4 months), 6-month and 12-month survival rates were 52.6 and 23 %, respectively. At the latest follow-up, 5 patients were still alive, with a mean follow-up of 9.7 months.

Scoring system
On the univariate analysis, survival was significantly associated with preoperative ambulatory status (P = 0.003), ECOG performance status (P < 0.001), number of involved vertebrae (P = 0.001), visceral metastases (P = 0.002), and time developing motor deficits (P < 0.001,    Table 1). On Cox proportional hazards model (multiple Cox regression, selection = stepwise), four of above five factors, ECOG performance status (P = 0.017), number of involved vertebrae (P = 0.021), visceral metastases (P = 0.022), and time developing motor deficits (P = 0.002), maintained significant impact on survival and were included in the scoring system ( Table 2). The scoring points for each of the four significant factors obtained from the hazard ratios on Cox proportional hazards model (simple Cox regression) were seen in Table 3. The prognostic score for each patient was calculated by adding the scoring points of the four significant characteristics. The addition resulted in prognostic scores of 4, 5, 6, 7, 8, 9, 10 points. The 6-month survival rates of the prognostic scores were shown in Fig. 2. Taking into account the 6month survival rates of the prognostic scores, the following three survival groups were formed: 4-5 points (group A, n = 22), 6-7 points (group B, n = 23), and 8-10   ), respectively, and 6-month survival rates were 95, 47 and 11 %, respectively (P < 0.001, Fig. 3).

Functional outcome
The functional outcome was worse in the group of patient with 8-10 points (group C) compared with the other two prognostic groups (

Discussion
Individually treatment needs to be planned for each patient with MSCC to give the maximum palliative effect: reduction in pain, recovery of function, and improvement in the patient's quality of remaining life. Selection of the optimal treatment for the individual patient with MSCC should take into account patient's estimated survival time, as well as functional outcome after therapies. Only those who survive long enough, more than 3 months, can benefit from surgery [21,22]. In contrast, patients with very short survival time and poor functional outcome appear to be best treated with radiotherapy or even best supportive care alone, which means less discomfort for these debilitated and enervated patient [4,10]. Remarkably, it is also critical to regard patients with MSCC from a particular primary tumor type as a separate group of patients for optimal treatment, because primary tumors vary with respect to their biological behavior. Crnalic et al. [23] presented a score specifically for predicting survival of patients with prostate cancer after surgery for MSCC. However, who may benefit from surgery, and what kind of patients are appropriate for supportive care, remains nuclear in NSCLC patients with MSCC.
Several scoring systems have been proposed for predicting survival in patient with spinal metastasis on the basis of retrospective data from various primary tumors treated with surgery or radiotherapy alone. However, these scores comprised relatively small number of patient with lung cancer (Tokuhashi 6 [9], revised Tokuhashi 26 [10], Tomita 10 [11], Van der Linden 68 [12], Sioutos 45 [13], Bauer 6 [14], Bartels 28 [15], more details were seen in Table 5), making it difficult to draw conclusions on this specific tumor type.
Although the revised Tokuhashi was found to be useful to predict survival for patients with spinal metastases from breast cancer alone [4] or solid cancers [24,25], which seems to be a suboptimal tool for the prediction of an individual prognosis in the group of patients with lung cancer (Hessler et al. [16]). In their study, 67 patients with spinal metastasis from lung cancer, all of them underwent surgical treatment. Hessler et al. [16] concluded that the Tokuhashi scoring system underestimated the life expectancy of lung cancer patients due to the increased survival time for this patient group. In 2013, Morgen et al. [17] also found a statistically significant increase in survival over the years for lung cancer patients with MSCC (n = 2321, 499 patients with lung cancer, 103 lung cancer patients received surgical treatment). For patients with lung cancer who underwent surgery for MSCC, survival increased from 9 % in year 2005 up to 30 % in year 2010 (P = 0.047). More recent studies have reported improvements among patients with advanced lung cancer because of the new treatment options [18,19]. Therefore, with the increasing survival time of patients with lung cancer during recent years, the Tokuhashi scoring system and other scores may no longer be suitable for patients with lung cancer.
Furthermore, these scores were designed for patients with spinal metastasis in general, not particularly for patients with motor impairment due to MSCC. Rades et al. [26] developed and validated a scoring system for survival of patients (n = 356, all patients with lung cancer) with MSCC from NSCLC who had been treated with radiotherapy alone. Aside from the Rades score, the above mentioned scoring systems included relatively small number of patients with spinal metastasis from various primary tumors. In fact, participants in Rades score received radiotherapy alone, and the functional outcome was not considered either. Moreover, patients who had prior surgery to the involved parts of the spinal cord were excluded in their study. In our study, a score was developed based on the data derived from 64 patients with NSCLC who underwent decompressive surgery and spine stabilization for MSCC. The indication for surgery was neurological deficits. Functional outcome was also considered according to the scoring system. The patient's individual situation, therefore, is taken more into account in the present scoring system. Patients with scores of 4-5 survived more than 1 year in median time, and 86 % patients were ambulatory 4 weeks after surgery. More radical surgery, such as widely excision of vertebra metastasis, can be considered in order to realize better local control of disease and prevent the occurrence of local disease in those patients. Patients with scores of 6-7 points should be surgical candidates, because survival prognosis and functional outcome were favorable after surgery. Patients with scores of 8-10 points, who survived 2.7 months in median time and had the worst functional outcome after surgery compared with other two prognostic groups, appeared to be best treated with radiotherapy or best supportive care alone. Functional outcome was acceptable in the entire cohort of 64 patients, 68.8 % (44 of 64) patients were able to walk 4 weeks after decompression; 51.6 % (16/31) of nonambulatory patients before operation regained the ability to walk. 74-84 % patients were able to walk after surgery [6,7,27] and 22-68 % of nonambulatory patients became ambulatory again in other studies [7 28].
However, patients with asymptomatic MSCC were not included in our study, so this scoring system doesn't pertain to those patients. Besides, our score was based on retrospective data, and the statistical analysis didn't include a relatively larger number of patients, and data on systemic treatment following treatment was not available in most patients. Despite good predictive value in our scoring system, the score still warrants a prospective study to be confirmed.

Conclusion
We present a new score for predicting survival of patients with NSCLC operated with posterior decompression and spine stabilization for MSCC. Functional outcome after surgery was also considered in our study. The scoring system can help select the individual treatment for patients with MSCC from NSCLC. Patients with scores of 4-5, who have the most favorable survival prognosis and functional outcome, can be treated with more radical surgery in order to realize better local control of disease and prevent the occurrence of local disease. Patients with scores of 6-7 points should be surgical candidates, because survival prognosis and functional outcome are acceptable after surgery, while patients with scores of 8-10 points, who have the shortest survival time and poorest functional outcome after surgery, appear to be best treated with radiotherapy or best supportive care. Still, a prospective study is needed.